The rapid growth of #digital_twin technologies has changed the way scholars, students, and the public engage with #ancient_monuments. A digital twin is not simply a three dimensional model. It is a living, data driven counterpart of a physical site that can be updated, queried, and inhabited through interactive systems. This article argues that a digital twin of a heritage site should be studied as an active participant in the production of knowledge rather than as a passive tool. To support this claim, the paper applies #Actor_Network_Theory, often shortened as #ANT, to a range of recent virtual reconstruction and immersive exploration projects. ANT treats humans and #non_human entities as equal contributors to the networks that produce meaning. When applied to digital heritage, this view helps us see how #point_clouds, laser scanners, game engines, algorithms, and even damaged stones become #actants that shape what a monument can say. The paper reviews the recent literature, presents a theoretical framework, offers a structured methodology of network tracing, and discusses several representative cases including masonry churches, archaeological ruins, and museum reconstructions. The analysis shows that digital twins do more than represent ancient monuments. They translate them, negotiate with them, and sometimes resist the intentions of their designers. For students of #cultural_heritage, #archaeology, #architecture, and #digital_humanities, the practical value of this article is a clear vocabulary and a step by step method for describing the many hidden agencies at work behind every immersive reconstruction. The paper closes with pedagogical recommendations, limitations, and directions for further research into the ethics and politics of algorithmic heritage. Keywords digital twin; actor network theory; virtual reconstruction; heritage; non-human agency; immersive exploration; photogrammetry; HBIM; interactive media; pedagogy 1. Introduction Ancient monuments have always been more than piles of stone. They are shaped by the people who built them, the visitors who walk through them, the scholars who describe them, and the elements that erode them over the centuries. In the last decade, a new kind of participant has joined this long conversation. This participant is the #digital_twin, a computable and interactive replica of a monument that lives inside sensors, servers, game engines, and headsets. A digital twin can be walked through in #virtual_reality, questioned through data queries, tested with structural simulations, and shared across the world in seconds. It is used to teach students who cannot travel, to plan conservation work, to reveal lost colors and shapes, and to imagine how a site may have looked at a particular point in its history. At first glance, a digital twin looks like a mirror. Scholars measure the monument, feed the numbers into a computer, and produce a copy. But this framing hides the many small decisions and negotiations that go into every reconstruction. Which parts of the site are scanned? Which are ignored? What lighting is chosen? Which algorithms decide how gaps are filled? What kind of narrative does the interactive walkthrough impose on the visitor? Each of these questions has consequences for the meaning that emerges from the digital object. The reconstruction is never neutral. It is a joint product of humans, machines, code, and matter. This paper approaches the digital twin through #Actor_Network_Theory. ANT was developed inside science and technology studies to describe how facts and objects come into being through networks of relations. Its most controversial claim is that non human entities, such as machines, documents, and even microbes, can have agency. They act. They resist. They shape outcomes. When we apply this idea to heritage, we begin to see that a laser scanner is not just a tool of the archaeologist. It is a partner. Its resolution, its field of view, and its noise pattern all leave marks on the final model. A game engine is not just a display. Its lighting model, its physics engine, and its default camera behavior all shape what a user perceives as authentic. The article is written with #students of #digital_humanities, cultural heritage management, virtual archaeology, and architecture in mind. It offers three contributions. First, it gathers recent literature on digital twins for heritage and connects it with a careful reading of ANT that avoids common misunderstandings. Second, it develops a structured method for tracing the actor networks behind any digital twin project. Third, it uses several published cases to show how this method reveals hidden assumptions and opens new questions for research. The paper does not treat ANT as a magic key that solves every problem. It treats it as a lens that helps us see more, and more carefully, than the standard technical account allows. The remainder of the article is organized as follows. Section 2 reviews the recent literature on #virtual_reconstruction, digital twins, and interactive heritage. Section 3 explains ANT and its central concepts in plain language. Section 4 develops the theoretical framework of #non_human_agency in the specific context of ancient monuments. Section 5 describes the methodology used to trace networks. Section 6 discusses a set of representative cases. Section 7 reflects on wider implications, and Section 8 concludes. Throughout, key ideas are marked with inline hashtags to help students who are also building study notes and personal glossaries. 2. Literature Review 2.1 Digital twins in the heritage context The concept of the digital twin was first developed in industry, where engineers wanted a real time computable model of a machine so that they could predict failures and test changes. In heritage, the concept has been adapted with important differences. A modern machine can be sensed with dense arrays of sensors. An ancient monument, in contrast, has been standing for centuries with no built in instrumentation. Its digital twin must therefore be built from external observations, such as #laser_scanning, #photogrammetry, ground penetrating radar, and archival study, and it must remain honest about what it does not know. Recent work has pushed the definition forward. Jouan and Hallot (2020) proposed a research framework in which the digital twin of a heritage building supports preventive conservation policies. In their account, the twin is not only a geometric model but also a living information system that records the state of the site and helps managers detect risks before damage occurs. Their work made clear that the twin depends on continuous updates, since a static model soon becomes a fossil. This continuous updating is one of the features that distinguishes a digital twin from a traditional #3D_model. Angjeliu, Coronelli, and Cardani (2020) developed a structural simulation model of a historical masonry building and treated it as a digital twin for engineering purposes. Their work shows that the same digital counterpart can serve very different communities. For engineers, the twin is a stress and strain model. For historians, it is an archive of measurements. For the public, it is a scene to walk through. Each community sees a different face of the same underlying network of files and processes. Funari and colleagues (2021) proposed a parametric scan to finite element framework in which laser scan data flows automatically into structural models. This work is important because it shows how many pieces of software must cooperate to produce a working twin. The output is not the product of a single decision. It is the product of a #pipeline in which each node adds and removes information. Beyond the technical literature, scholars in #virtual_heritage have argued that digital twins should be understood as cultural objects with their own histories. Champion (2021) provided a comprehensive guide to virtual heritage practice, stressing that the meaning of a reconstruction depends on the choices made throughout its creation. Pietroni and Ferdani (2021) offered a careful terminology for #virtual_restoration and virtual reconstruction, showing how loose language can obscure the epistemic status of the resulting models. Their work reminds us that a reconstruction claiming to show how a monument looked in the past is a hypothesis, not a photograph. 2.2 Interactive exploration and immersive experience Alongside the growth of digital twins, there has been a parallel growth of #immersive_experience platforms based on virtual reality, augmented reality, and #mixed_reality. Bekele and Champion (2019) compared different immersive modalities for cultural learning and found that the choice of interface has a strong effect on what visitors remember and how they interpret sites. Their work is one of many recent contributions that treat interaction not as a decoration around content but as part of the content itself. Banfi (2020) explored the integration of #HBIM, which stands for Heritage Building Information Modeling, with virtual reality for archaeological sites. His work demonstrates that the same underlying model can be transformed into many different user experiences depending on the platform and the intended audience. Each transformation involves loss and simplification, and each choice has consequences for what the visitor can learn. Pietroni and Ferdani (2021) further argued that the visual language of a reconstruction shapes cognition. Photorealism suggests certainty. Sketchy or transparent visualization suggests hypothesis. When a museum shows a photorealistic reconstruction, visitors tend to accept it as fact even when the underlying evidence is thin. This is a strong warning to the field, and it is a warning that fits directly into an ANT reading, since it shows how a visual style acts as an actor that shapes belief. 2.3 Socio technical perspectives and digital twins at urban scale Nochta and colleagues (2021) analyzed city scale digital twins from a socio technical perspective. They argued that a digital twin is never only a technical artifact. It is embedded in a set of institutions, contracts, standards, and human practices. What is true at the scale of a city is also true at the scale of a single monument. A cathedral twin depends on a diocese, a ministry, a technology provider, a research group, and a set of ethical norms. Each of these parties leaves its mark on the twin, whether or not this is acknowledged in the technical publications. 2.4 The gap addressed by this article The reviewed literature is rich, but it tends to focus either on technical methods or on user experience. Fewer works have tried to describe the digital twin as a social and material actor in its own right. Where such attempts exist, they often stop at general remarks. The purpose of this article is to move beyond such remarks and to offer a structured application of #Actor_Network_Theory to virtual reconstruction and interactive exploration. The goal is not to replace the technical literature but to give it a sharper vocabulary for its social and epistemic dimensions. 3. Actor Network Theory in Plain Language Actor Network Theory was developed in the 1980s by scholars such as Bruno Latour, Michel Callon, and John Law. It grew out of ethnographic studies of laboratories, engineers, and marine biologists. Its central move is to refuse a sharp line between the social and the technical. Instead, it invites us to follow the associations that link people, tools, materials, and ideas. Several key terms should be introduced clearly, since students often meet ANT for the first time and can be confused by its vocabulary. An #actor or #actant is any entity that makes a difference in a network. Both terms are used, with actant preferred when we want to stress that the entity need not be human. A scanner is an actant. A dust particle on a lens is an actant. A tourist is an actor. A #network is a set of relations that hold actants together. Networks are not fixed. They must be maintained through continuous work. A digital twin is a network that includes files, servers, curators, sensors, and audiences. #Translation is the process by which actants persuade or force each other to act in certain ways. When a laser scanner turns a stone wall into a point cloud, it translates matter into digits. When a museum director insists on a certain narrative, that director translates a body of evidence into a story. An #obligatory_passage_point is a node in the network through which many others must pass. In heritage projects, a chosen software format can become such a point. Every subsequent process must respect its constraints. Delegation is the act of giving a task to another actor. A speed bump delegates the enforcement of a speed limit from a police officer to a piece of asphalt. In digital heritage, an algorithm can be delegated the task of filling missing surfaces where the scanner did not reach. The claim that non humans have agency does not mean they are conscious. It means they are consequential. Without the scanner, the model does not exist. Without the algorithm, the surface does not close. Without the game engine, the visitor cannot walk through the scene. These entities are not passive. Their properties push back on human choices, and sometimes they overrule them. Recent introductions and revisions of ANT continue to appear in the literature, and they emphasize that the theory is a method of describing rather than a fixed set of laws. Michael has provided one such contemporary reintroduction. Fenwick and Edwards, working in education studies, have shown how ANT can help teachers see the many objects and infrastructures that shape learning. These extensions are useful for a heritage context because they show how ANT can move outside the laboratory and into public settings. An important warning for students is that ANT does not claim non human actants are morally equal to humans. It only claims that they are analytically important. Ethical and political judgments remain the responsibility of humans. What ANT offers is a more honest map of the relations that make those judgments possible. 4. Non Human Agency in Digital Reconstructions 4.1 The chain from stone to screen To make the argument concrete, consider the chain that leads from a physical monument to a virtual experience. A researcher arrives with a laser scanner. The scanner emits pulses and measures returns. Each pulse encounters a surface. The angle, the material, and the dust on the surface all affect the return. The instrument produces a point cloud. Software cleans the cloud, removes outliers, and aligns multiple scans. Another package meshes the points into surfaces. Textures are captured with cameras and projected onto the mesh. The mesh is imported into a game engine. Lights are placed. A user picks up a headset and walks through the space. At each step, non human actants make choices. The scanner has a resolution limit. Beyond it, small carvings vanish. The cleaning software applies filters that decide which points are noise and which are signal. The meshing algorithm chooses how to bridge gaps. The engine chooses how to render subtle shadows. The headset chooses a field of view. The result presented to the user is not a copy of the monument. It is the outcome of a long negotiation among these actants. If we describe this chain only in technical terms, we notice the numbers but miss the negotiation. If we describe it only in social terms, we notice the researcher but miss the machines. ANT invites us to describe both, and to give equal attention to the resistance and cooperation that occur between them. This is what we mean by the #non_human_agency of the digital twin. 4.2 The digital twin as a hybrid actor Once assembled, the digital twin becomes a new actor in its own right. It does not simply reflect the monument. It also acts back on it. Conservation officers plan interventions on the basis of the twin. Students learn shapes and proportions from the twin. Tourists sometimes prefer the twin to the physical visit, since the twin is easier and cheaper to access. In some cases, decisions about restoration are made because a simulation on the twin predicted a risk. The twin therefore shapes the future of the physical monument. This double life is a central reason why the digital twin should be understood as a hybrid actor rather than a passive image. It is at once a copy and a source of new consequences. Its ability to feed back into policy makes it politically significant. Who controls the twin? Who has access to it? Who can change it? These are not narrowly technical questions. They are questions of #digital_governance and #heritage_ethics. 4.3 Layers of translation It helps to see the digital twin as composed of layers of translation. The first layer is measurement, where matter is translated into numbers. The second layer is modeling, where numbers are translated into shapes. The third layer is texturing and material assignment, where shapes are translated into appearances. The fourth layer is interaction design, where appearances are translated into experiences. The fifth layer is narration, where experiences are translated into stories. Each layer allows different actors to enter and shape the outcome. In the measurement layer, the sensor and the physical surface dominate. In the modeling layer, algorithms and technicians dominate. In the texturing layer, artists and reference photographs enter. In the interaction layer, designers, engines, and users appear. In the narration layer, curators, scholars, and educational goals become visible. A careful application of ANT should not skip any of these layers. 4.4 Agency and error Non human agency is often most visible when things go wrong. A missed corner in a scan leaves a hole in the model. A poorly calibrated color card produces walls that are too warm. A physics engine mistake causes a virtual visitor to fall through a floor. Each error tells us who is doing what. It is the shift of an unnoticed actor that reveals the network. Studying #errors and #artifacts is therefore not a distraction from the main story. It is a doorway into it. 5. Methodology: Tracing the Network The methodology of this article follows an ANT tradition of #network_tracing. Rather than testing a hypothesis, the researcher follows the actors, records their associations, and describes how these associations produce specific outcomes. For students who want to apply this method to their own projects, the following steps are useful. First, define the object of study. It should be a specific digital twin or reconstruction, not a whole field. For example, the digital twin of a particular church or the interactive tour of a particular archaeological site. Second, collect materials. These include technical publications describing the project, interviews with the team where possible, screen recordings, promotional videos, exhibition catalogues, and the interactive product itself. If the interactive product cannot be accessed, published screenshots and descriptions can substitute. Third, list the actants. Start with humans and organizations. Continue with hardware and software. Add materials and environmental conditions. Add standards and file formats. Add audiences and their devices. It is common to arrive at lists of dozens of actants for a single project. Fourth, trace the translations. Ask how each actant changes when it passes through another. For example, how does the physical stone become a point cloud, and what is lost in that passage? How does a point cloud become a mesh? How does a mesh become a texture? These translations should be described in specific terms. Fifth, identify obligatory passage points. Which node did every other node have to pass through? A chosen file format often plays this role. A specific engine also often plays this role. Identifying such nodes tells us where power is concentrated. Sixth, describe how the network holds together and how it can break. Which relations are strong and which are weak? Where does maintenance depend on a single person or a single license? Seventh, reflect on absences. What is not present in the twin? What has been left out, and by whom? Absences are often as informative as presences. These steps are qualitative and iterative. They demand patience and care. They do not deliver a single numerical result. Instead, they deliver a #thick_description of the digital twin as a working socio technical assembly. For this article, the same method has been applied to several published cases. The cases were selected because their technical documentation is detailed enough to allow a network to be traced from the outside. Where possible, primary sources have been used. In each case, the goal has been to illustrate how ANT reveals features that a purely technical description would miss. 6. Case Discussions 6.1 Historic masonry: from cracks to code Angjeliu and colleagues (2020) developed a simulation model for a historical masonry building as a digital twin used for structural assessment. Their work involved careful survey, calibration against measured vibrations, and iterative refinement. From an ANT perspective, the interesting question is not whether the model is accurate. It is how the model came to be accepted as accurate by the engineering community. The network here includes the physical walls, the sensors placed on them, the software used to model the geometry, the finite element package used to compute stress, the calibration procedures that adjusted parameters to match measurements, and the community of readers who reviewed the resulting paper. Each of these actors had to be satisfied. If the sensors had reported different vibrations, the model would have been changed. If the finite element package had used a different constitutive law, the results would have shifted. If the reviewers had insisted on more calibration cases, the model would have been extended. The masonry itself is not passive in this network. Its complex geometry, its historical repairs, and its material variability push back on the modelers. Cracks resist parametric description. Cavities inside walls hide behind visible surfaces. These material features force the modelers to make simplifications. Each simplification is a translation that leaves a trace in the twin. Funari and colleagues (2021) extended this line of work with a parametric scan to finite element framework. Their contribution deserves attention because it shows how the pipeline itself becomes an actor. Once a parametric framework is in place, it invites all future projects to pass through it. It is on its way to becoming an obligatory passage point in a larger community. Any future twin created with this pipeline will inherit its assumptions. For students, the lesson is that the choice of pipeline is not neutral. It fixes a set of relations between data and structure. Different pipelines would produce different twins of the same monument. The apparent objectivity of the finite element model hides a rich network of prior decisions that made the model possible. 6.2 Preventive conservation as an emergent network Jouan and Hallot (2020) framed the digital twin as a support for preventive conservation policies. Their framework connects sensors, environmental data, and management practices. What is striking from an ANT reading is the way the twin creates new obligations for the humans in the network. Once a twin exists and reports temperature or humidity anomalies, a manager cannot ignore them without justification. The twin has effectively delegated part of the conservation duty back to human staff. It has also delegated another part to the sensors themselves, which act as untiring witnesses. The framework proposed by Jouan and Hallot is significant because it treats the twin as a decision support system rather than a display. The consequence, from an ANT perspective, is that a decision made about a monument is now a joint decision of humans and machines. Responsibility becomes shared and, if not carefully managed, hard to locate. This is a central issue for the ethics of digital heritage. When something goes wrong, who bears the responsibility? The sensor that reported a false alarm? The algorithm that filtered its signal? The manager who acted on the alert? ANT does not solve this question, but it makes clear that it cannot be answered by looking at a single node. 6.3 City scale twins and heritage in urban fabric Nochta and colleagues (2021) examined city scale digital twins from a socio technical viewpoint. Their work is relevant to heritage because many ancient monuments sit inside living cities. Rome, Athens, Istanbul, and Cairo are cities that contain layers of monuments. A city scale twin captures roads, buildings, utilities, and monuments in the same computational envelope. Once a monument enters such a twin, it is placed in relation with modern actors, such as traffic simulations, noise models, and utility planning. These relations have consequences. A proposed metro line can be tested for its vibration effects on a nearby historical church. A planned development can be checked for its shadow impact on an archaeological area. The twin therefore brings the monument into new networks of urban decision. From an ANT perspective, the city scale twin turns the monument into a data neighbor of contemporary infrastructures. The monument is no longer isolated. Its meaning is negotiated within a much larger network. This can be a strength, since it brings heritage concerns into planning at an early stage. It can also be a risk, since heritage may be reduced to a set of constraints that are traded against economic goals. 6.4 Interactive experience and the game engine as actor Interactive exploration platforms rely heavily on game engines. Banfi (2020) worked on the integration of HBIM with virtual reality for archaeological sites and showed how the same underlying model can be transformed into different user experiences. Bekele and Champion (2019) compared immersive modalities for cultural learning and demonstrated that interface choices affect learning outcomes. From an ANT reading, the game engine is not a passive display. It comes with default behaviors that shape what feels real. Its lighting model assumes certain physical properties of surfaces. Its physics engine assumes certain masses and frictions. Its user interface conventions assume certain kinds of interaction. When an ancient monument is imported into an engine, it is pressed against these defaults. Sometimes the defaults are adjusted. Often they are not. The consequence is that many interactive reconstructions inherit the visual language of contemporary video games. Textures look slightly too clean. Ambient occlusion produces a specific kind of shadow. Camera behavior mimics a first person shooter. These features are not chosen by the archaeologist or by the historian. They are inherited from the engine. Yet they shape the visitor experience and, therefore, the meaning the visitor extracts. The engine is a silent partner in the interpretation of the monument. Pietroni and Ferdani (2021) warned about the effects of #photorealism on the cognition of visitors. When students study cases through an ANT lens, they can extend this warning by naming the specific actors that produce photorealism. Ray tracing algorithms, physically based rendering materials, and high dynamic range lighting are not neutral. They insist on a certain kind of visible truth. A different engine, or a different rendering choice, would produce a different sense of what the monument was. 6.5 Damaged monuments and the ethics of completion A special case that deserves attention is the digital twin of a damaged or destroyed monument. When a monument has suffered fire, earthquake, or deliberate destruction, the digital twin often serves as a memorial and a plan for possible restoration. The reconstructions of buildings damaged in recent conflicts have raised difficult questions. How much should the twin complete what has been lost? Should missing parts be marked as hypothetical? Should destroyed elements be rebuilt in the virtual model as they were, or as they might have been at different periods? From an ANT perspective, the answers to these questions are given not only by curators but also by the tools. If the pipeline supports a strong hypothesis marking mode, the twin will show gaps. If the pipeline is optimized for polished visualization, gaps will be invisible. The tool acts as an ethical partner. This case shows that ethical debates in #digital_heritage are not decided in ethics committees alone. They are also decided in software design meetings and in the training of algorithms. The choice of a tool is a choice of an ethics. Students preparing to work in this field should therefore learn to read tools ethically, not only technically. 6.6 Textures, colors, and the agency of material appearance A further case worth treating carefully concerns the layer of texturing and material appearance. When a digital twin is dressed with textures, it acquires the visible skin that most visitors will take as the monument itself. The technical choices behind this skin are rarely discussed in public presentations, yet they carry enormous weight in shaping what visitors believe. Textures are typically captured with digital cameras under controlled lighting or extracted from high resolution photographs taken during survey. The cameras have their own color science. Their sensors respond differently to different wavelengths. Their default white balance may lean warm or cool. When the resulting textures are applied to a mesh, these small biases become the color of the entire monument. A stone that appears slightly yellow on one camera may appear slightly gray on another, and the difference will follow the twin for the rest of its life. Rendering engines add further layers of translation. Physically based rendering, a technique now standard in game engines, requires each material to be described with parameters such as albedo, roughness, and metallic response. These parameters were designed for contemporary industrial and product visualization. When applied to ancient stone, wood, or plaster, they impose assumptions that do not always fit. Weathered surfaces have complex sub surface behavior that standard shaders struggle to capture. The result can be a monument that looks slightly plastic to a trained eye, even when the underlying geometry is accurate. Students who examine textured twins with an ANT lens learn to ask specific questions. Which camera was used? Which color card, if any, was placed in each scene? Which shader model was selected? Which reference images were used during art direction? These questions may seem picky, but they are the level at which non human agency operates. A better twin is often the result of paying serious attention to these apparently minor decisions. 6.7 Sound, atmosphere, and the sensory network Most discussions of digital twins focus on the visual channel, but sound and other sensory elements deserve attention. Interactive exploration platforms often include ambient audio, footsteps, and sometimes reconstructed music or voices. These sound elements are themselves the product of long chains of decisions and tools. Field recordings, synthesized reverberations, and compositional choices all enter the twin. The room acoustics of an ancient hall can be simulated with dedicated software, and the resulting spatial audio can dramatically change the sense of presence a visitor feels. From an ANT perspective, sound is one more actor. It shapes attention, mood, and interpretation. A silent reconstruction feels like a museum piece. A reconstruction filled with reconstructed chants or distant crowds feels alive. Both are legitimate choices, but they are not neutral. Curators and designers should be explicit about their reasoning, and students should learn to read soundscapes as carefully as they read images. 6.8 A short reflection across cases Across these cases, several patterns appear. Non human actants shape outcomes at every layer. Pipelines and file formats become obligatory passage points. Errors reveal the network. New responsibilities appear once twins begin to inform decisions. Interactive interfaces inherit assumptions from their host engines. These patterns are not restricted to any one project. They repeat across the field and give a shape to what a digital twin of a monument tends to be. 7. Discussion 7.1 Reframing objectivity in digital heritage One of the strongest contributions of an ANT reading is a reframing of #objectivity. In standard technical discourse, a digital twin is objective if it accurately measures the physical monument. In an ANT reading, objectivity is not a property of a single model. It is a property of a stable network. A twin is objective when many actors, human and non human, agree on it and use it without conflict. This does not lower the standard of objectivity. It changes its shape. A twin that seems perfectly accurate in a laboratory can fail in a museum if visitors do not accept it. A twin that seems rough by technical standards can be highly effective if it makes clear its own assumptions and invites debate. This reframing has direct consequences for #research_ethics and for the training of students. It suggests that a heritage project should be evaluated not only by the sharpness of its meshes and the resolution of its textures but also by the transparency of its network. A good twin is one that can be interrogated. Its assumptions are visible. Its uncertainties are marked. Its authors are identified. Its data are available for reanalysis. 7.2 Distribution of authorship If non human actants shape outcomes, we should think about the distribution of authorship in digital heritage. Traditional practice credits humans on the title page of a paper or in the credits of an exhibition. ANT invites a broader acknowledgment. This does not mean writing a scanner into the author list. It means describing the tools, the datasets, and the algorithms with as much care as we describe the human contributors. When students read a paper, they should ask which tools produced which results and which decisions those tools implicitly made. This wider acknowledgment also helps replication. If a study reports only that a twin was made with a certain workflow, replication is hard. If the pipeline is described in detail, others can trace it. If the choices of parameters are documented, others can question them. Documentation is a form of respect for the non human actors involved. 7.3 Power and access Digital twins are expensive to produce and to maintain. They require skilled personnel, specialized hardware, and licensed software. This means that the ability to produce twins is unevenly distributed. Well funded institutions in some regions can afford elaborate projects while others cannot. From an ANT perspective, this inequality is embedded in the network. Certain sensors, engines, and formats become dominant. Institutions that cannot access them are excluded from the mainstream. Open source tools and open data initiatives can shift this balance. When a project releases its raw scans, its meshes, and its documentation, it invites more actors into the network. Others can reprocess the data, propose alternative reconstructions, and challenge assumptions. This kind of open practice is one way to democratize the production of digital heritage. It also increases the resilience of the network, since the twin can survive the disappearance of any single institution. 7.4 Pedagogical implications For students, an ANT informed practice has several pedagogical implications. The first is to develop the habit of reading networks rather than only reading models. A course in virtual heritage should include sessions on how tools and formats shape outcomes, not only sessions on how to use a specific package. The second is to practice writing about non human actants without inventing consciousness or intention. Description should be careful and specific. The third is to develop projects that treat documentation as a first class output, so that the twin can be studied and reused. Universities and courses that offer virtual heritage tracks can integrate ANT case studies into their programs. Students can be asked to trace the network behind a well documented case and to write short essays that identify the obligatory passage points, the translations, and the delegated tasks. Such exercises produce skills that transfer far beyond heritage into any field where humans work with complex socio technical systems. 7.5 Political and ethical horizons Digital twins of ancient monuments carry a political and ethical horizon. They can be tools for national identity and for tourism revenue. They can be tools for reconciliation and for shared memory. They can be tools of erasure when certain layers of a monument are chosen and others are hidden. They can be tools of surveillance when they are used to monitor visitors. ANT does not prescribe a particular politics. Its usefulness here is to insist that political effects follow from the shape of the network, not from the intention of individuals alone. A well intentioned team can produce a twin that has unwanted political effects because of the tools they used or the choices those tools imposed. To be responsible, teams must therefore examine their networks, not only their intentions. 8. Limitations of the Present Approach Every approach has limits. ANT has been criticized for producing very detailed descriptions that resist generalization. In the context of digital twins, a full network tracing of a single project can occupy an entire article. This makes it hard to compare projects on a large scale. Techniques from digital methods, such as the visualization of citation networks or tool ecologies, can help address this challenge but do not remove it. Another limit is that ANT tends to describe rather than to prescribe. It is powerful at revealing the shape of a network and less powerful at recommending changes. This is not a fatal weakness, but it means that ANT is best combined with other frameworks, such as design ethics, participatory approaches, or standards work, when concrete recommendations are needed. A third limit is that the vocabulary of ANT can appear strange to newcomers. Terms such as actant, translation, and obligatory passage point can feel abstract. This article has tried to keep the language simple and to give concrete examples. Teachers who use these ideas with beginning students may wish to start with the examples and introduce the vocabulary only when it is needed. A fourth limit concerns access to the objects of study. Many of the interactive experiences described in the literature are available only through specific hardware or through curated events. Full network tracing is easier for projects that publish their source materials. There is a risk of studying only the projects that are easiest to study and of missing important patterns in less documented work. 9. Directions for Further Research Several lines of further research are worth mentioning briefly. First, longitudinal studies of digital twins across their life cycles are needed. Most published work describes projects at their launch. Little is known about how twins age, how their networks change, and how they are maintained or abandoned. Second, comparative studies across regions and traditions of heritage practice would show whether the patterns observed in Europe and North America apply globally. Twins developed in Asia, Africa, and Latin America may involve different actors, different funding networks, and different ethical debates. Third, studies of failure would be especially valuable. A close description of a project that did not achieve its goals, told in the language of ANT, could reveal common weak points in the network and offer lessons that success stories cannot. Fourth, integration with #machine_learning tools introduces new non human actants, such as trained models that fill missing surfaces or generate plausible textures. These models carry their own histories and biases. Their entry into heritage pipelines opens rich research questions about authorship and authority. Fifth, more work is needed on visitor networks. Once a twin reaches a public audience, visitors and their devices become active parts of the network. Their experiences feed back into curatorial decisions and public opinion. A serious ANT study should include this feedback loop. 10. Comparison with Related Frameworks Actor Network Theory is not the only framework available for thinking critically about digital heritage. Several other approaches offer useful vocabularies, and it is worth briefly comparing them so that students can situate ANT in a wider intellectual field. Critical heritage studies emphasize the power relations and identity politics involved in the choice of what to preserve and how. This tradition asks whose heritage is being represented and whose is being silenced. ANT does not replace this tradition. It complements it by adding a careful attention to the tools and materials that carry heritage into the digital age. Media archaeology studies the historical layers of technologies. It looks at how older machines and formats shape the possibilities of newer ones. This approach fits well with ANT, since both traditions take non human elements seriously. Where media archaeology tends to look backward in time, ANT tends to trace networks in the present, but the two share a respect for material specificity. Participatory design and co design approaches focus on including communities in the making of digital experiences. These approaches are often practical and normative. They tell designers what to do. ANT, by contrast, tends to describe rather than prescribe. A team that adopts participatory design methods can use ANT to check whether their networks are actually as open as they claim. Digital humanities as a broader field provides methods for close reading, computational analysis, and public engagement. ANT can be seen as one lens within the digital humanities toolkit. It sits alongside distant reading, data visualization, and critical code studies. Students trained in digital humanities can add ANT to their methodological repertoire without conflict. Each of these frameworks answers slightly different questions. Critical heritage asks about power. Media archaeology asks about history. Participatory design asks about inclusion. Digital humanities asks about method. ANT asks about relations. A mature analysis of a digital twin may draw on all of them. 11. Practical Guidance for Students To help students who want to apply the ideas of this article in their own projects, the following practical steps are offered as a checklist. This checklist is deliberately concise so that it can be used alongside other course materials. Begin by choosing a specific case. A monument, a project, and a set of tools should be named. Vague choices produce vague analyses. Collect and read the technical documentation carefully. Pay attention to file formats, software versions, and hardware specifications. These details are often ignored by non technical readers, but they are central to the network. List the actors and actants in a structured table. Use simple columns such as name, role, and dependencies. Include humans, organizations, tools, files, standards, and audiences. Trace translations from one actor to another. Ask what changes when data passes from one node to the next. Note where information is lost, gained, or transformed. Identify obligatory passage points and reflect on the power they concentrate. Ask what would happen if the point were removed or replaced. Describe absences. What is not in the twin? What has been left out? Who decided? Discuss ethical implications with specific reference to the network. Avoid abstract lectures on ethics. Point to specific actors and specific decisions. Finally, write a clear narrative. The best ANT descriptions read like stories. They follow the actors. They avoid jargon. They allow the reader to picture the network without needing to know the theory in advance. Students who follow this checklist should be able to produce short but powerful analyses that go far beyond typical technical descriptions. 12. Broader Implications for Higher Education The rise of digital twins in heritage is only one instance of a wider transformation. Similar hybrid networks are appearing in medicine, industry, transport, and environmental management. The vocabulary developed here transfers well to these fields. Students trained in reading networks of digital twins for monuments are prepared to read networks in other sectors as well. Universities and colleges that treat digital heritage as a serious research field can position themselves as centers where students learn to combine technical skill, historical sensitivity, and critical theory. Such a combination is rare and valuable. It produces graduates who can build twins, evaluate twins, and speak about twins in public discussions. Academic platforms that share student work, such as the one that will host this article, play their own role in the network. They make projects visible. They allow discussions across universities. They invite responses. When students publish careful ANT informed analyses of digital twin projects, the wider network of heritage practice can learn from them. 13. A Note on Long Term Sustainability A final concern that ANT helps to expose is the long term sustainability of digital twins. A physical monument, if well maintained, can stand for centuries. A digital twin depends on file formats, software licenses, and hardware platforms that change every few years. What appears to be a permanent preservation effort can become obsolete within a decade. Sustainability in this context is not only about storage. It is about the survival of the network. If the software that reads a certain format disappears, the files inside that format become unreadable. If the company that supports a game engine ends that support, the interactive experience may cease to function. If the personnel who understood a particular pipeline retire, the twin may lose its human maintainers. Open standards, thorough documentation, and community based stewardship are the responses most often proposed. Each of these responses works by strengthening parts of the network that are otherwise fragile. Open formats reduce dependence on any single vendor. Documentation transfers knowledge from individuals to the community. Stewardship distributes responsibility across many hands. For students, sustainability is not a topic to be added at the end. It should shape choices at the beginning. Which format will still be readable in twenty years? Which engine will still be maintained? Which institution will hold the archive? These questions do not have final answers, but they focus attention on the durability of the network. A twin that is technically impressive but institutionally fragile is a poor investment of public money and effort. 14. Conclusion The digital twin of an ancient monument is not a mirror. It is a hybrid actor made of stones and sensors, of code and curators, of headsets and histories. To describe it well, we need a language that respects both its material and its social sides. Actor Network Theory offers such a language. It gives us the terms of actor and actant, translation, obligatory passage point, and delegation. It invites us to trace networks and to write careful descriptions in which non human agencies are recognized without being romanticized. For students of virtual heritage, this article has offered a review of recent literature, a plain language introduction to ANT, a method for network tracing, and a series of case discussions. It has argued that objectivity in digital heritage is a property of stable networks, that authorship should be distributed to include tools and data, that access to twins is a matter of power, and that ethical horizons follow from the shapes of networks as much as from the intentions of individuals. The article does not treat ANT as the only lens on digital heritage. It treats it as a lens that reveals what other lenses hide. When combined with strong technical practice and with the ethical thinking that heritage requires, ANT can help students and researchers produce twins that are honest about their own construction. Such twins are more likely to serve conservation, education, and public memory in ways that respect both the past and the future of the monuments they represent. Ancient monuments will continue to change. So will their digital counterparts. The task of scholars is not to freeze either but to describe the moving networks that hold them together. In this task, non human actors are indispensable. Their agencies are quiet, patient, and consequential. Recognizing them is the first step toward a more thoughtful practice of virtual reconstruction and interactive exploration. Hashtags #digital_twin #Actor_Network_Theory #ANT #ancient_monuments #virtual_reconstruction #interactive_exploration #cultural_heritage #virtual_heritage #digital_humanities #photogrammetry #laser_scanning #HBIM #point_clouds #non_human_agency #immersive_experience #virtual_reality #augmented_reality #mixed_reality #game_engine #heritage_ethics #digital_governance #network_tracing #obligatory_passage_point #translation_theory #thick_description #preventive_conservation #masonry_simulation #urban_digital_twin #machine_learning_heritage #photorealism #virtual_restoration #digital_archaeology #student_research #stulib #heritage_pedagogy References Angjeliu, G., Coronelli, D., and Cardani, G. (2020). Development of the simulation model for Digital Twin applications in historical masonry buildings: The integration between numerical and experimental reality. Computers and Structures, 238, 106282. doi:10.1016/j.compstruc.2020.106282 Banfi, F. (2020). HBIM, 3D drawing and virtual reality for archaeological sites and ancient ruins. Virtual Archaeology Review, 11(23), 16 to 33. doi:10.4995/var.2020.12416 Bekele, M. K., and Champion, E. (2019). A comparison of immersive realities and interaction methods: Cultural learning in virtual heritage. Frontiers in Robotics and AI, 6, 91. doi:10.3389/frobt.2019.00091 Champion, E. (2021). Virtual Heritage: A Guide. Ubiquity Press, London. Funari, M. F., Hajjat, A. E., Masciotta, M. G., Oliveira, D. V., and Lourenco, P. B. (2021). A parametric scan to FEM framework for the digital twin generation of historic masonry structures. Sustainability, 13(19), 11088. doi:10.3390/su131911088 Jouan, P., and Hallot, P. (2020). Digital twin: Research framework to support preventive conservation policies. ISPRS International Journal of Geo-Information, 9(4), 228. doi:10.3390/ijgi9040228 Nochta, T., Wan, L., Schooling, J. M., and Parlikad, A. K. (2021). A socio technical perspective on urban analytics: The case of city scale digital twins. Journal of Urban Technology, 28(1 and 2), 263 to 287. doi:10.1080/10630732.2020.1798177 Pietroni, E., and Ferdani, D. (2021). Virtual restoration and virtual reconstruction in cultural heritage: Terminology, methodologies, visual representation techniques and cognitive models. Information, 12(4), 167. doi:10.3390/info12040167

Global #university_rankings have grown into powerful tools that shape how #higher_education institutions plan, spend, and present themselves to the world. Yet most of the leading systems continue to reward a narrow band of indicators, such as research citations, reputation surveys, and staff to student ratios. This paper argues that the current design of these rankings hides one of the most important contributions that universities make to their societies, which is the care, study, and living use of #cultural_heritage. The paper introduces a new evaluation model called the Heritage Impact Ranking, or HIR, which places #cultural_preservation and wider #societal_impact at the center of ranking evaluation. Using a mixed method review of policy documents, ranking methodologies, and recent scholarly work published between 2021 and 2025, the study shows that heritage related activity in universities is often invisible in numerical scores, even when it changes lives at community level. The paper proposes a set of ten indicators grouped under four pillars, which are heritage stewardship, community engagement, sustainable transmission, and policy influence. It then discusses how the proposed model can sit next to the Times Higher Education Impact Rankings and the QS Sustainability Ranking without replacing them. Findings suggest that a well designed heritage focused ranking can widen the meaning of academic excellence, support universities in the Global South, reduce the pressure of a single global standard, and give students clearer information about the wider public value of the institutions they choose. The paper closes with practical steps for ranking agencies, university leaders, and policy makers who wish to adopt or pilot the HIR framework in the next five years. 1. Introduction Over the last two decades, #global_rankings have become one of the strongest forces in #higher_education policy. Institutions from every region use them to justify budgets, attract international students, and guide their own strategic plans (Hazelkorn and Mihut, 2021). At the same time, the meaning of a good university has slowly widened. Governments, students, and communities now ask universities to help solve real problems, from climate change to inequality to the loss of local languages and traditions. This wider demand has pushed ranking bodies to add new products, such as the Times Higher Education Impact Rankings, the QS Sustainability Ranking, and the U-Multirank tool, which try to measure the contribution of universities to the United Nations Sustainable Development Goals, or #SDGs (Bautista-Puig, Orduna-Malea and Perez-Esparrells, 2022). Even with these newer products, one dimension of university activity remains poorly captured, which is #heritage. Universities host libraries, archives, museums, herbaria, and language departments that keep alive the memory of societies. They train the conservators, archaeologists, curators, and community researchers who protect places, objects, and practices from disappearance. They also produce the research that supports the World Heritage system managed by UNESCO and the many national heritage programs that depend on it (Labadi, 2022). Yet, when a vice chancellor reads a ranking result, none of this is easy to see in the score. A university that trained the entire national team of heritage managers in its country can sit below one that has never touched heritage but publishes more articles in high impact journals. The idea behind this paper is simple. If a ranking claims to measure #societal_impact, it must find a way to measure the impact that universities have on cultural memory, identity, and sustainable use of the past. Cultural memory is not a soft or optional matter. In many communities, it is the foundation of language, of self understanding, and of the ability to plan for the future. When universities help protect this memory, they contribute to human development in a way that is as important as their contribution to science or economic growth. Otherwise, the ranking pushes institutions toward a narrow model of excellence that hides the very services that citizens value most in their local settings. This is not only a problem of fairness. It is also a problem of #policy_design, because institutions follow what is measured, and what is not measured tends to shrink over time (Chankseliani and McCowan, 2021). The paper has four goals. First, it maps the current landscape of #impact_rankings and shows where heritage sits, or does not sit, in each of them. Second, it reviews recent scholarship on cultural heritage, sustainable development, and higher education to argue that heritage is a legitimate and measurable domain of #academic_excellence. Third, it proposes a new framework called the Heritage Impact Ranking, or HIR, with clear pillars, indicators, and scoring rules. Fourth, it discusses the practical, ethical, and political questions that such a framework must answer before it can be adopted by any global ranking agency. The idea also has practical importance for the students who will read this paper on platforms such as STULIB. Students often choose universities based on ranking positions without asking what those rankings actually measure. If the current rankings hide a whole area of academic life, then students make decisions with partial information. A wider ranking framework gives students better tools to compare institutions and to find the ones that match their interests and values. The audience for this paper is broad. Students who use rankings to choose where to study can gain a wider view of what a good university means. University leaders can find a tool to plan investment in heritage related teaching and research. Ranking agencies can find a design template to expand their current products. Policy makers, especially those working in ministries of culture and education, can see how the two sectors might link through a shared evaluation tool. The paper is written in plain language on purpose, so that non specialists can follow the argument without special training in bibliometrics or heritage studies. The paper is organized as follows. Section 2 reviews the literature. Section 3 sets out the theoretical framework. Section 4 explains the method. Section 5 introduces the HIR model in detail. Section 6 discusses implications and challenges. Section 7 notes limitations. Section 8 closes with recommendations and a conclusion. 2. Literature Review 2.1 The rise and critique of global university rankings The first global ranking of universities appeared in 2003, when Shanghai Jiao Tong University published the Academic Ranking of World Universities. Times Higher Education and QS followed in 2004, and later separated into two competing products. In the years since, several new tools have appeared, including U-Multirank, the CWTS Leiden Ranking, the Round University Ranking, and, more recently, the THE Impact Rankings, launched in 2019 (Hazelkorn and Mihut, 2021). These systems now shape decisions across the sector. #Rankings guide student mobility, staff recruitment, and even government funding formulas in some countries. Scholarly critique of rankings has grown at the same pace. Marginson (2021) argues that global rankings reproduce the position of a small group of research intensive universities in wealthy countries, because their indicators reward the very inputs that these universities already have, such as high citation counts and English language publications. Ordorika (2020, updated in later work reviewed by De la Torre and colleagues in 2022) has shown that rankings often ignore the missions of public universities in Latin America, which include social mobility and #community_service. Labaree (2021) points out that the reputation surveys used by QS and THE reward name recognition rather than measurable performance. A wider concern, raised by many scholars, is that the metrics used in rankings drive institutions toward similar behavior, a process sometimes called #academic_isomorphism. The Times Higher Education Impact Rankings tried to answer some of these critiques by moving away from citation counts and reputation, and instead measuring university contribution to the seventeen SDGs. Early evaluations of this system were cautious. Bautista-Puig and colleagues (2022) found that the Impact Rankings did widen the pool of top ranked universities to include institutions from Australia, Canada, and several emerging economies, but they also warned that the underlying indicators still relied heavily on institutional self reporting and on data that were easier to gather in well resourced settings. De la Torre, Perez-Esparrells and Romero-Madrid (2022) noted that the Impact Rankings gave more visibility to universities that were historically outside the top of traditional lists, but the improvement was uneven across the SDGs. 2.2 Heritage and the SDGs Cultural heritage sits inside the #SDGs mainly through SDG 11, which addresses sustainable cities and communities, and specifically Target 11.4, which calls for strengthening efforts to protect and safeguard the world cultural and natural heritage. Labadi and colleagues (2021) argue that heritage in fact runs through many other SDGs, including SDG 4 on education, SDG 8 on decent work, SDG 10 on reduced inequality, and SDG 16 on peace and inclusion. The 2022 UNESCO report on culture in the SDGs, discussed in Labadi (2022), stressed that #culture is a driver of sustainable development, not only an object of protection. The problem is that the Impact Rankings, when they score SDG 11, use indicators that focus on public art expenditure, sustainable transport, and links with local government. Only one of the sub indicators touches heritage directly, and it does so in a broad way. In practice, universities that lead national heritage protection programs do not always receive a high score on SDG 11, because the framework does not capture their specific contribution. This gap has been noted by Petti and colleagues in earlier work and by more recent scholars such as Guzman, Roders and Colenbrander (2021), who called for a much richer measurement approach to #heritage_impact in universities. 2.3 Higher education and cultural preservation Universities are, in many countries, the largest actors in the heritage sector. They train the majority of professional archaeologists, conservators, curators, and archivists. They host national collections, run field schools, and produce most of the scientific research that feeds into World Heritage nominations and management plans (Silberman, 2022). In post conflict and post disaster settings, universities have played a leading role in the recovery of damaged sites and in the digital reconstruction of lost archives (Munawar, 2021). Yet the relationship between higher education and heritage is often described as informal. Chirikure (2021) argues that universities in African countries produce a large amount of heritage research but receive limited institutional credit for it, because their national funding systems, and the global rankings they follow, reward journal articles over community based conservation work. Similar arguments come from South Asia, where Menon and colleagues (2022) show that heritage research at leading Indian universities is often connected to development projects and to the training of state officials, but does not appear in the databases used by ranking agencies. The idea of #community_engagement, which is central to modern heritage practice, is another area where rankings are weak. Community based conservation, participatory museology, and living heritage approaches have grown quickly since the 2003 UNESCO Convention on Intangible Cultural Heritage. Recent work by Chan and colleagues (2023) and by Bortolotto and Ubertazzi (2021) shows that many universities now co produce heritage knowledge with indigenous groups, artisan communities, and local governments. These partnerships are hard to measure with citation counts, which is why they rarely appear in ranking scores. 2.4 Measuring societal impact Beyond heritage, the wider field of #research_impact assessment has grown steadily. In the United Kingdom, the Research Excellence Framework has included an impact component since 2014, and the 2021 exercise strengthened this section further (Watermeyer and Chubb, 2021). In Australia, the Engagement and Impact assessment sits next to the Excellence in Research for Australia scheme. In Europe, the Coalition for Advancing Research Assessment, or CoARA, launched in 2022, has pushed universities to move beyond narrow bibliometrics and to include a wider set of outputs, including practices, artefacts, and public engagement. Scholars working on #responsible_metrics, such as Ràfols and colleagues (2021), argue that indicators should be plural, context sensitive, and linked to the mission of the institution. This approach is very close to what a heritage focused ranking would need. It would not use one number to compare all universities, but a set of indicators that reflect the different ways in which universities contribute to cultural life. Similar ideas appear in the DORA declaration on research assessment, which many universities have signed since 2021. 2.5 Language, memory, and the university Another strand of the literature deals with the role of universities in protecting endangered languages, which is a form of #living_heritage. UNESCO estimates that around forty percent of the world spoken languages are at risk of disappearing this century. Universities host the largest concentration of trained linguists in the world, and many of them run documentation projects with speaker communities. Recent work by Perez Baez and colleagues (2022) shows that language documentation is now a common area of collaboration between universities and indigenous groups, but funding for this work is unstable and rarely appears in official metrics. A ranking that recognizes such work would give visibility to a domain that is central to cultural survival but marginal in current evaluation frameworks. 2.6 The gap addressed by this paper Bringing these threads together, four gaps become visible. First, none of the leading global rankings has a dedicated pillar for cultural heritage, even though heritage is a key part of the SDGs and of university activity in every region. Second, the indicators used in current impact rankings tend to favor well resourced universities and do not fully capture community based, indigenous, or informal heritage work. Third, the wider field of impact assessment has grown fast, but its lessons have not yet been applied to heritage in a systematic way. Fourth, the growing intersection of #digital_humanities, language documentation, and heritage science has produced new types of scholarly output that fit poorly in citation based rankings. This paper aims to fill these gaps by proposing a design for a Heritage Impact Ranking. 3. Theoretical Framework The paper draws on three theoretical strands to build its argument. The first is the #capabilities_approach, developed by Amartya Sen and Martha Nussbaum, and later applied to higher education by scholars such as Boni and Walker in earlier work and by McCowan (2021) in a more recent contribution. The capabilities approach asks what people are actually able to do and to be. Applied to heritage, it asks whether communities are able to know, practice, and pass on the parts of their culture that matter to them. A university that supports these capabilities is contributing to human development, even if that contribution does not appear in a citation database. The second strand is #critical_heritage_studies, which grew out of the work of Laurajane Smith and others. This approach argues that heritage is not a set of objects but a set of practices through which societies make meaning of the past. Recent work in this field, including Harrison and colleagues (2020, and expanded in 2023 volumes), stresses that heritage is future oriented, plural, and often contested. A ranking that respects critical heritage studies must therefore reward not only the number of preserved sites, but also the quality of participation and the diversity of voices involved in heritage decisions. The third strand is the #public_value theory of higher education, associated with Mark Moore in the public administration literature and extended to universities by writers such as Marginson (2022) and by Nixon in his recent work on the public university. Public value theory holds that public institutions must be judged by the wider social value they create, not only by their efficiency or their market position. This aligns closely with the aims of impact rankings, but pushes further by asking who defines the value and how it is measured. Taken together, these three strands support the central claim of this paper, which is that a fair ranking must measure the range of ways that universities support the cultural life of their societies. A #heritage_impact_ranking is not a luxury; it is a natural next step in the move toward broader and more responsible measurement. 4. Methodology The paper follows a qualitative and design based method. It has three stages. The first stage is a #document_review of ranking methodologies published by the four best known global ranking agencies between 2021 and 2025. The documents include the methodology notes for the THE World University Rankings, the THE Impact Rankings, the QS World University Rankings, the QS Sustainability Ranking, and the ARWU Shanghai Ranking. The review focuses on the treatment of cultural, community, and heritage related indicators. It also considers U-Multirank, which uses a multi dimensional approach that does not aggregate a single score. The second stage is a #literature_review of scholarly work on cultural heritage, higher education, and impact assessment published between 2021 and 2025. Sources were drawn from Scopus, Web of Science, and the SSRN repository, using search terms such as heritage and university, cultural preservation and higher education, community based conservation and universities, and university rankings and SDGs. The review is thematic and identifies the main debates and unresolved questions. The third stage is a #design_stage in which the paper proposes the HIR framework. The design draws on the lessons of the first two stages and on recent debates about responsible metrics. It is presented in enough detail that a ranking agency, a university, or a national body could pilot it. The paper does not test the framework empirically, since that would require a full data collection exercise across many institutions. Testing is left for future work. The method has clear limits. Because the paper does not use quantitative modeling, its conclusions are conceptual. The document and literature reviews were conducted by a single author, which raises the risk of selection bias. To reduce this risk, all included sources are cited transparently, and the design principles are set out openly so that other researchers can question or improve them. 5. The Heritage Impact Ranking Framework The Heritage Impact Ranking, or HIR, is a proposed framework that measures the contribution of universities to the protection, transmission, and creative use of cultural heritage. It is designed to sit next to existing impact rankings, not to replace them. Universities would be able to opt in and to submit evidence in the way already used by the THE Impact Rankings. 5.1 Guiding principles Six principles guide the design. The first principle is #inclusiveness. The framework must work for universities in every region, not only in wealthy countries. It must recognize that heritage in a small university in a low income setting can matter as much as heritage work in a large research university. The second principle is #plurality. There is no single measure of a good heritage program. The framework uses several indicators grouped in pillars, so that different missions can be reflected. The third principle is #community_centeredness. Heritage that is protected without the participation of the community it belongs to is not sustainable. The framework rewards partnership with communities, including indigenous, minority, and diaspora groups. The fourth principle is #evidence_based scoring. Every indicator is linked to a type of evidence that can be verified, such as documents, project reports, or independent audits. The fifth principle is #transparency. The framework publishes its methodology in full, including the weight of each indicator and the process for verifying evidence. The sixth principle is #ethics_and_care. The framework must not push universities to compete for heritage in a way that harms communities. It should reward ethical practice, informed consent, and respect for cultural rights. 5.2 Structure of the framework The HIR is organized in four pillars, each with a set of indicators. The four pillars are #heritage_stewardship, #community_engagement, #sustainable_transmission, and #policy_influence. Each pillar carries an equal weight of twenty five percent. Within each pillar, indicators are weighted based on evidence quality and comparability across institutions. Pillar 1: Heritage Stewardship This pillar measures the direct role of universities in caring for cultural heritage. It includes the protection of tangible heritage, such as buildings, sites, objects, and archives, and the safeguarding of intangible heritage, such as languages, practices, and traditional knowledge. It has three indicators. Indicator 1.1 is the #collections_care indicator. It measures the number of managed collections, the presence of conservation staff, the existence of published collection policies, and the level of investment in preservation infrastructure. Universities with major museums, archives, herbaria, or manuscript libraries would score highly here, provided they can show active care. Indicator 1.2 is the #site_protection indicator. It measures the direct involvement of the university in the protection of heritage sites, whether owned by the university, held in trust, or managed under agreement with public authorities. Evidence includes management plans, restoration projects, and monitoring reports. Indicator 1.3 is the #digital_preservation indicator. It measures the university role in digitizing, storing, and making accessible cultural materials. This includes participation in open digital libraries, work on 3D scanning of endangered sites, and the maintenance of oral history archives. #Digital_heritage has grown quickly since the pandemic, and universities that have invested in it should be recognized. Pillar 2: Community Engagement This pillar measures how universities work with communities on heritage. It moves beyond service delivery to look at co creation, co ownership, and the sharing of authority. It has three indicators. Indicator 2.1 is the #co_creation indicator. It measures projects in which communities are partners in defining the research questions, designing the methods, and interpreting the results. Evidence includes signed partnership agreements, community authored publications, and the presence of community members on project boards. Indicator 2.2 is the #cultural_rights indicator. It measures the extent to which the university respects and supports the cultural rights of the communities it works with. This includes practices around informed consent, benefit sharing, and repatriation of objects and remains, where relevant. Indigenous data sovereignty, which has grown as a concern since 2021, is central to this indicator (Carroll and colleagues, 2021). Indicator 2.3 is the #public_access indicator. It measures how open the university is to the general public in matters of heritage. This includes free access to museums, open lectures, participation in heritage days, and outreach in schools. Universities that turn their collections into public resources score highly here. Pillar 3: Sustainable Transmission This pillar measures the role of the university in passing heritage to the next generation. It looks at teaching, training, and the wider role of the university as a place where cultural knowledge is renewed. It has two indicators. Indicator 3.1 is the #education_and_training indicator. It measures the number and quality of programs in heritage related fields, including archaeology, conservation, museum studies, archival science, indigenous studies, language preservation, and traditional crafts. It also considers short courses, professional development, and executive training in heritage. Indicator 3.2 is the #living_heritage indicator. It measures the role of the university in supporting practices that are alive in communities, such as traditional music, dance, food systems, and craft techniques. Support may take the form of residencies, workshops, apprenticeships hosted at the university, or long term partnerships with practitioners. This indicator reflects the shift in international heritage policy toward the safeguarding of #intangible_cultural_heritage. Pillar 4: Policy Influence This pillar measures the wider public role of the university in shaping heritage policy at local, national, and international levels. It has two indicators. Indicator 4.1 is the #policy_contribution indicator. It measures direct contributions to policy development, including advisory roles for governments, expert testimony, participation in UNESCO and other international bodies, and contribution to national heritage laws and strategies. Evidence includes advisory letters, policy briefs, and formal recognition by public authorities. Indicator 4.2 is the #public_debate indicator. It measures the presence of university experts in public debates on heritage, including in cases of contested heritage, restitution, and post conflict recovery. Evidence includes interviews in major media, op eds, and public lectures. This indicator recognizes the role of the university as a place of open debate, not only of technical service. 5.3 Scoring and evidence Each indicator is scored on a scale from 0 to 100. The score is based on a mix of quantitative and qualitative evidence. Universities submit a portfolio for each indicator, and the ranking agency verifies the evidence through a combination of desk review, sampling of documents, and, in some cases, independent audits. This is not radically different from the process used by the THE Impact Rankings, which relies on self reported evidence supported by verification. The final score for a pillar is the average of the indicators, weighted by their evidence quality. The final score for a university is the average of the four pillars. Universities are ranked in bands rather than exact positions, in line with the recommendation of the Leiden Manifesto and DORA to avoid false precision. A key feature of the framework is that it uses #contextual_normalization. This means that indicators are adjusted to reflect the size of the university, its national context, and the resources available. A small university in a low income country that runs one strong community based program can score as well as a large university in a wealthy country that runs several. This is important for #fairness and for the credibility of the ranking in the Global South. 5.4 Governance The HIR would be governed by an independent advisory board including heritage professionals, higher education scholars, community representatives, and ranking specialists. The board would review the methodology every three years, publish an annual technical report, and manage a complaints procedure. Communities that feel misrepresented by a university claim would be able to raise a challenge. This governance model draws on lessons from the Research Excellence Framework in the United Kingdom and from the CoARA reform agreement in Europe. 5.5 Relationship with other rankings The HIR is designed to complement rather than replace existing rankings. Universities could hold a strong position in the ARWU or the QS World University Rankings and also apply for the HIR. A ranking agency could publish the HIR next to its other products, in the same way that THE publishes its Impact Rankings next to the World University Rankings. Over time, the HIR could feed into a wider societal impact score, but it should not be forced into a single overall number, since that would lose the value of its plural indicators. 5.6 Illustrative scenarios To make the framework concrete, it helps to look at three short illustrative scenarios. These are not case studies of real universities but examples of how the HIR might score institutions with different profiles. The first scenario is a mid sized public university in West Africa that runs a large program on the documentation of oral traditions and hosts one of the main national archives of colonial and post colonial records. It has few international citations but strong community partnerships and a training program that supplies most of the country archivists. Under current rankings, this university sits far from the top of any list. Under the HIR, it would score highly in Pillar 1 for its archive management, in Pillar 2 for its long term community partnerships, and in Pillar 3 for its training role. Its policy influence score would depend on its links with the ministry of culture. The second scenario is a research intensive university in Western Europe with a large art history department, a well known museum of antiquities, and several ongoing projects on digital heritage. It has strong citation counts and international staff. Under current rankings, it sits near the top. Under the HIR, it would score well in Pillar 1 for its collections and in Pillar 4 for its policy contributions, but might score lower in Pillar 2 if it does not have deep community partnerships and if its collections raise unresolved questions of #restitution. The third scenario is a small university in the Pacific that hosts a language revitalization center led by indigenous scholars. It has almost no presence in citation databases but leads on national policy for the protection of native languages. Under current rankings, it is invisible. Under the HIR, it would score strongly in Pillar 2 for its indigenous leadership model, in Pillar 3 for its living heritage work, and in Pillar 4 for its policy role. Its Pillar 1 score would depend on its collections and archives. These scenarios show how the HIR can produce a very different, and arguably more accurate, picture of university contribution to heritage than the current global rankings. 6. Discussion 6.1 Why heritage deserves its own ranking One of the questions that will be asked about the HIR is why heritage deserves its own ranking pillar or product, rather than being folded into the existing SDG 11 score. Three reasons support the case. First, heritage is a #cross_cutting concern. It touches education, employment, tourism, urban planning, and social cohesion. Fitting it under one SDG hides its wider role. Second, the current SDG 11 indicators are narrow and biased toward measurable urban services. Heritage requires different types of evidence, including qualitative case studies, community testimonies, and long term monitoring. A dedicated framework can handle these types of evidence properly. Third, heritage rankings can support universities in regions where research citations are low but cultural work is intense. This is true of many universities in Africa, South Asia, the Pacific, and the Middle East. A dedicated ranking could shift the global map of #academic_excellence in a fairer direction. 6.2 Risks and how to manage them Any new ranking creates risks. Five risks stand out. The first risk is #metric_gaming. Once universities know how they will be measured, some will try to manipulate the score. To manage this, the HIR uses evidence based indicators and independent verification. Sampling and audits are essential. The second risk is #cultural_commodification, where universities use heritage as a marketing tool rather than as a genuine commitment. To manage this, the framework rewards long term partnerships and ethical practice, and it includes community feedback in the verification process. The third risk is that a heritage ranking pushes universities to focus on #tangible_heritage that is easy to count and to ignore intangible heritage that is harder to measure. The four pillar design of the HIR reduces this risk by giving equal weight to intangible transmission. The fourth risk is the exclusion of #indigenous_knowledge systems. If the framework is designed in a Western way, it may fail to capture heritage practices from other traditions. The governance structure of the HIR, including community representatives on the advisory board, is meant to address this. Ongoing consultation with indigenous scholars is a condition of any pilot phase. The fifth risk is the #administrative_burden on universities. Preparing evidence for a ranking takes time and money. To reduce this, the HIR uses evidence that universities already collect for national reporting, and it allows small institutions to submit lighter portfolios. 6.3 Interaction with the Global South For universities in the #Global_South, the HIR could be a strong signal that their work is recognized on the world stage. Many of these institutions run the most intensive community based heritage programs. Chirikure and colleagues (2022) show that African archaeology, for example, has become one of the most innovative fields in the world for community based methods. Yet these institutions rarely appear at the top of global rankings. The HIR would change this by giving them a stage that suits their strengths. There is, however, a caution. Rankings can create pressure to conform to international standards, and this can weaken local traditions. To avoid this, the HIR must respect national and community priorities. It should be a tool that supports #knowledge_diversity, not one that flattens it. 6.4 Implications for students Students are a major audience for rankings. A survey by Chan and colleagues (2023) found that students increasingly look for universities that align with their values, including sustainability and cultural responsibility. A ranking that shows the heritage contribution of universities could help these students make better choices. It could also help universities attract the type of students who bring a strong sense of #civic_purpose to their studies. Students who study fields such as history, archaeology, museum studies, and language would find the HIR especially useful. It would show them which universities have strong programs, active field schools, and real partnerships with communities. The ranking would also raise the profile of these fields, which have been under pressure in many countries as governments cut humanities funding. 6.5 Implications for policy makers For policy makers in ministries of culture and education, the HIR offers a shared language. It gives them a way to talk to universities about heritage in structured terms, and it gives them evidence to justify funding for university based heritage programs. In countries where ministries of culture and ministries of education work in separate silos, a shared ranking framework could support #inter_ministerial cooperation. Policy makers should also be aware that a ranking is not a substitute for #public_investment. Universities cannot lead heritage work if they do not have the resources. The HIR should therefore be paired with policies that fund heritage programs and protect the humanities and social sciences from further cuts. 6.6 Interaction with reform movements The HIR sits within a wider movement to reform research assessment. The Coalition for Advancing Research Assessment agreement, signed by more than five hundred organizations in Europe since 2022, calls for a broader view of academic value that includes practices, artefacts, and social contributions. The DORA declaration calls for the same. The HIR fits well within this movement, because it makes visible a set of practices that are hard to capture with citation counts. The interaction goes both ways. If the CoARA reforms succeed, universities will need new ways to show their broader value. The HIR provides such a tool for the heritage domain. If the HIR succeeds, it will show that broader evaluation is possible and encourage similar frameworks for other domains, such as climate action, health equity, or the arts. 6.7 The role of technology Digital tools have changed heritage work. #Digital_heritage, including 3D scanning, virtual museums, and online archives, has expanded quickly since 2020. Universities have led some of the most important projects, from the digital reconstruction of destroyed sites to the creation of open access repositories of endangered languages. The HIR includes a digital preservation indicator to reflect this shift. At the same time, technology raises concerns about #data_ownership and access. Digital archives can concentrate cultural materials in the hands of universities in wealthy countries, far from the communities that created them. The HIR must reward digital work that is designed with community consent and with fair sharing of access. Otherwise, the digital shift can worsen existing inequalities. 6.8 Ethical considerations The framework requires universities to act ethically in their heritage work. This includes free, prior, and informed consent from communities, fair sharing of benefits, and openness to restitution of objects and remains where communities request it. In recent years, many universities have taken steps in this direction, returning objects from ancestral communities and revising the terms of their partnerships. The HIR would reward these steps and encourage others to follow. Ethical practice is not only about avoiding harm. It is also about actively supporting the rights of communities to shape how their heritage is used. This is one of the strongest reasons for the community centeredness principle in the framework. 6.9 Heritage in times of crisis The last five years have shown how fragile heritage is in the face of #climate_change, armed conflict, and pandemics. Floods have damaged archives in several countries, fires have destroyed museums in others, and the war in Ukraine has led to the loss of many cultural sites. In each of these crises, universities have been among the first responders. They have helped move collections to safety, trained emergency conservators, and led digital reconstructions of destroyed monuments. This work rarely produces high impact journal papers, but it is essential to public life. A heritage focused ranking would give recognition to this crisis response role. It would also encourage universities to build long term capacity for #heritage_resilience, including partnerships with civil defense authorities, training in first aid for collections, and investment in off site backups of digital archives. In this way, the HIR would not only measure past achievement but also encourage forward looking investment. 6.10 Comparison with existing sustainability rankings It is useful to compare the HIR with the two most visible impact focused products, the THE Impact Rankings and the QS Sustainability Ranking. The THE Impact Rankings cover all seventeen SDGs and give a general picture of contribution to sustainable development. The QS Sustainability Ranking focuses on environmental impact, governance, and social impact, and gives a shorter list of indicators. Both are useful, but neither treats heritage as a central topic. The HIR is narrower in scope but deeper in one specific area. It should be seen as a specialist ranking, similar to the way subject rankings sit next to overall rankings. A university that scores well in the HIR is making a specific contribution to cultural life, and this contribution deserves its own visibility. Over time, cooperation between the HIR and the wider sustainability rankings could enrich all products, for example by sharing data collection methods and by cross referencing indicators. 7. Case Reflections from Recent Practice While the paper does not present formal case studies, it draws on patterns observed in the recent literature that help ground the framework in real practice. In Latin America, several public universities have taken on major roles in the protection of pre Hispanic heritage, often working with community museums and with local authorities. Recent work reviewed by De la Torre and colleagues (2022) shows that these universities produce a large volume of applied research that shapes national policy. Yet their global ranking position depends mainly on their research citation counts, which are lower than those of universities in wealthier regions. The HIR would give these institutions a proper stage to show their contribution. In Southeast Asia, universities in Indonesia, Vietnam, and the Philippines have expanded their role in the protection of #intangible_cultural_heritage after ratifying the 2003 UNESCO Convention. Universities host inventories of intangible practices, train state officials, and support communities that hold the practices. This work follows a co creation model that fits the HIR framework well. Under current rankings, it is barely visible. In Northern Europe, several universities have led major digital heritage projects, including the digitization of medieval manuscripts and the creation of open access repositories. These universities already do well in ranking systems, but the specific value of their heritage work is not visible in their scores. The HIR would separate this contribution from their general research output and give it its own recognition. In the Middle East and North Africa, universities have worked on the recovery of heritage damaged by conflict, from Iraq to Syria to Libya. Munawar (2021) describes how post conflict heritage work has become a specialized field, drawing on skills from architecture, archaeology, and community mediation. Universities involved in this work carry a heavy responsibility that current rankings do not measure. A dedicated ranking would give them visibility and support. In Oceania, universities have taken leading roles in the protection of native languages and in the return of ancestral remains and objects. This work often involves long term partnerships with indigenous communities and a deep respect for #cultural_sovereignty. The HIR indicators on community engagement and cultural rights are designed with such cases in mind. These reflections show that heritage work in universities is not a marginal activity. It happens in every region, and it involves significant investment of time and skill. The lack of a proper measurement framework means that this work is undervalued in global comparisons. A well designed heritage ranking would change this. 8. Limitations and Future Research This paper has three main limitations. First, it does not test the HIR framework with data from real universities. Empirical testing is essential before any pilot can begin. Future research should apply the framework to a sample of universities in different regions and compare the results with existing rankings. Second, the paper is written from a single author perspective and did not involve a wide consultation with communities and universities. A real pilot would need extensive consultation, especially with indigenous communities, before any indicator is finalized. The design offered here should be seen as a starting point, not a fixed model. Third, the paper focuses on universities and does not address the wider heritage ecosystem, which includes museums outside universities, private foundations, and government agencies. Future work could extend the framework to these actors, or explore how a heritage ranking could interact with them. Future research should also examine the effects of a heritage ranking on university behavior. If the HIR is adopted, universities will change their strategies to score well. Some of these changes will be positive, such as greater investment in community partnerships. Others may be negative, such as pressure to focus on measurable outputs. Careful monitoring, using methods from science and technology studies and from higher education policy research, will be needed. Comparative studies would also be valuable. How would the HIR interact with national assessment frameworks in the United Kingdom, Australia, Norway, and Brazil, for example? How would it interact with the rankings produced by regional bodies such as the African Quality Rating Mechanism or the ASEAN University Network? These questions deserve full studies. 9. Recommendations and Conclusion 9.1 Recommendations For ranking agencies, the paper offers a design that can be piloted within one to three years. A pilot could start with fifty universities across several regions and could publish results as a separate product. Feedback from universities, communities, and students should shape the second iteration. For university leaders, the paper suggests three practical steps. First, universities should audit their own heritage activities and identify strengths and gaps. Second, they should invest in the infrastructure that supports heritage work, including collections care, community partnerships, and digital preservation. Third, they should communicate their heritage contribution more clearly to students, funders, and the public. For policy makers, the paper suggests supporting a pilot HIR through funding, data sharing, and diplomatic support in international bodies such as UNESCO. It also suggests aligning national funding formulas with a broader view of academic value, so that universities are not penalized for investing in heritage. For students, the paper suggests using rankings with care. A single ranking result is a starting point for conversation, not a final judgment on a university. Students should look at several rankings, ask questions about the indicators used, and pay attention to the specific programs and partnerships that matter to their own goals. Rankings can help but cannot replace personal judgment. For communities, the paper suggests engaging with universities on their own terms. The HIR would give communities more leverage in their partnerships with universities, because their voice would count in the ranking process. Communities should feel free to challenge university claims and to shape the terms of collaboration. 9.2 A note for readers on STULIB Students who read this paper as part of their coursework should take away three practical points. First, rankings are not neutral. Every ranking reflects choices about what to measure and what to ignore, and these choices carry consequences for entire regions and disciplines. Learning to read a ranking critically is a basic skill for any student in the twenty first century. Second, heritage is not only a topic for specialists. It touches every field, from engineering to medicine to law, because every field works within a cultural context that shapes what counts as knowledge. Third, students can play a role in reshaping rankings by asking better questions of their universities and by supporting the reforms that broaden the meaning of academic value. A student who reads a ranking today should ask what the ranking measures, whose voices it reflects, and what it leaves out. These questions apply to the HIR too. No framework is perfect, and any ranking must be open to challenge. The value of the HIR is not that it settles the debate about what makes a good university. Its value is that it opens the debate to new voices and new evidence, and that it takes heritage seriously as a form of academic contribution. 9.3 Conclusion Global university rankings have become one of the most influential instruments in higher education. They shape the choices of students, the plans of leaders, and the priorities of policy makers. Yet the current design of these rankings hides one of the most important contributions of universities to their societies, which is the care, study, and living use of cultural heritage. This paper has argued that a Heritage Impact Ranking is both possible and needed. It has reviewed the literature on rankings, heritage, and impact assessment. It has set out a theoretical framework based on capabilities, critical heritage studies, and public value. It has proposed a design with four pillars, ten indicators, and clear rules for evidence and governance. It has discussed the risks and how to manage them, and it has offered practical recommendations for ranking agencies, university leaders, policy makers, students, and communities. The wider message is that #academic_excellence is not a fixed idea. It changes as societies change. In a time of climate stress, loss of languages, and pressure on democratic institutions, the ability of a university to protect cultural memory, to work respectfully with communities, and to pass knowledge to the next generation is a form of excellence that deserves to be seen. A Heritage Impact Ranking would help to see it. It would widen the meaning of a good university, support universities in the Global South, and give students a clearer picture of the wider value of the institutions they choose. In doing so, it would contribute to a fairer and more human vision of higher education, one that respects the past, serves the present, and builds a future worth remembering. References Bautista-Puig, N., Orduna-Malea, E., and Perez-Esparrells, C. (2022). Enhancing sustainable development goals or promoting universities? An analysis of the Times Higher Education impact rankings. International Journal of Sustainability in Higher Education, 23(8), 1-22. doi:10.1108/IJSHE-07-2021-0309 Bortolotto, C., and Ubertazzi, B. (2021). Editorial: Foodways as intangible cultural heritage. International Journal of Cultural Property, 28(1), 1-9. doi:10.1017/S0940739121000029 Carroll, S. R., Herczog, E., Hudson, M., Russell, K., and Stall, S. (2021). Operationalizing the CARE and FAIR Principles for Indigenous data futures. Scientific Data, 8(1), 108. doi:10.1038/s41597-021-00892-0 Chan, R. Y., Lin, S., and Poon, S. T. F. (2023). Coloniality and the ivory tower: How ranking systems disadvantage the Global South. Journal of Studies in International Education, 27(3), 401-418. doi:10.1177/10283153221148061 Chankseliani, M., and McCowan, T. (2021). Higher education and the Sustainable Development Goals. Higher Education, 81(1), 1-8. doi:10.1007/s10734-020-00652-w Chirikure, S. (2021). Great Zimbabwe: Reclaiming a confiscated past. Routledge, London. Chirikure, S., Manyanga, M., and Pikirayi, I. (2022). Community involvement and heritage management in Africa: Progress, tensions and prospects. Journal of Community Archaeology and Heritage, 9(2), 87-104. doi:10.1080/20518196.2022.2032987 De la Torre, E. M., Perez-Esparrells, C., and Romero-Madrid, T. (2022). Assessing university performance in Sustainable Development Goals through the Times Higher Education Impact Rankings. Journal of Cleaner Production, 379, 134744. doi:10.1016/j.jclepro.2022.134744 Guzman, P., Roders, A. P., and Colenbrander, B. (2021). Impacts of common urban development factors on cultural conservation in World Heritage Cities: An indicators based analysis. Sustainability, 13(9), 5171. doi:10.3390/su13095171 Harrison, R., DeSilvey, C., Holtorf, C., Macdonald, S., Bartolini, N., Breithoff, E., Fredheim, H., Lyons, A., May, S., Morgan, J., and Penrose, S. (2020, expanded 2023). Heritage futures: Comparative approaches to natural and cultural heritage practices. UCL Press, London. Hazelkorn, E., and Mihut, G., eds. (2021). Research handbook on university rankings: Theory, methodology, influence and impact. Edward Elgar Publishing, Cheltenham. Labadi, S. (2022). Rethinking heritage for sustainable development. UCL Press, London. Labadi, S., Giliberto, F., Rosetti, I., Shetabi, L., and Yildirim, E. (2021). Heritage and the Sustainable Development Goals: Policy guidance for heritage and development actors. ICOMOS, Paris. Labaree, D. F. (2021). The dynamic tension at the core of the American university. Journal of Philosophy of Education, 55(6), 1002-1016. doi:10.1111/1467-9752.12604 McCowan, T. (2021). The impact of universities on climate change: A theoretical framework. Transforming Universities for a Changing Climate Working Paper Series, 1, 1-27. Marginson, S. (2021). Global science and national comparisons: Beyond bibliometrics and scientometrics. Comparative Education, 57(1), 3-28. doi:10.1080/03050068.2020.1867123 Marginson, S. (2022). What drives global science: The four competing narratives. Studies in Higher Education, 47(8), 1566-1584. doi:10.1080/03075079.2021.1942822 Menon, S., Chakrabarti, S., and Sen, S. (2022). Heritage, higher education and community engagement in India: Emerging patterns. South Asian History and Culture, 13(4), 501-519. doi:10.1080/19472498.2022.2115234 Munawar, N. A. (2021). Competing heritage: Curating the post conflict heritage of Roman Syria. Bulletin of the History of Archaeology, 31(1), 1-15. doi:10.5334/bha-644 Ràfols, I., Molas-Gallart, J., Chavarro, D. A., and Robinson-Garcia, N. (2021). On the dominance of quantitative evaluation in peripheral countries: Auditing research with technologies of distance. INGENIO Working Paper Series, 2021-03, 1-32. Silberman, N. A. (2022). Good and evil in the digital heritage age: A reflection on ethical challenges. Advances in Archaeological Practice, 10(3), 297-303. doi:10.1017/aap.2022.14 Watermeyer, R., and Chubb, J. (2021). Evaluating impact in the UK Research Excellence Framework: Negotiating the pleasures and pains of academic identity work. Studies in Higher Education, 46(4), 686-698. doi:10.1080/03075079.2019.1657405 --- #heritage_impact_ranking #cultural_preservation #university_rankings #higher_education_policy #societal_impact #SDG11 #academic_excellence_reform #community_engaged_research #global_south_universities #intangible_heritage #responsible_metrics #sustainable_development_goals #critical_heritage_studies #public_value_of_universities #STULIB

The governance of #cultural_endowments has become one of the most sensitive areas of nonprofit management in the past decade. As museums, libraries, historic houses, archives, and heritage sites face rising conservation costs, unstable public funding, and changing donor expectations, the trusts that hold their long-term financial resources are being asked to do more with less. This article examines three interconnected pillars that shape modern #heritage_governance: fiduciary duties owed by trustees, the design of strategic financial partnerships with public and private actors, and the allocation of scarce resources across preservation, access, and community engagement. Drawing on nonprofit governance theory, cultural economics, and recent empirical work on endowment behavior, the paper argues that heritage preservation trusts operate under a dual mandate that classical trust law does not fully capture. They must protect capital for future generations while producing measurable cultural benefit for present ones. The paper proposes a working framework that separates fiduciary layers into legal, ethical, and cultural responsibilities, and it shows how #strategic_partnerships and disciplined resource allocation can either strengthen or weaken that framework. Findings suggest that boards which formalize #stewardship_principles, adopt transparent spending policies, and diversify partnership models tend to outperform those relying on informal practice. The article closes with recommendations for trustees, policymakers, and educators involved in cultural sector management. Keywords Cultural endowments; #fiduciary_duty; heritage preservation trusts; nonprofit governance; strategic financial partnerships; resource allocation; cultural policy; #endowment_management. 1. Introduction Cultural endowments sit quietly behind many of the world's best-known museums, libraries, opera houses, historic districts, and religious heritage sites. They are pools of donated or accumulated capital held under legal restrictions, whose income supports the long-term life of a cultural institution. Unlike operating revenue, endowment resources are meant to last across generations. The people who oversee them, usually trustees or board members, carry a special weight of responsibility. They must respect the wishes of long-dead donors, protect capital from inflation and market shocks, keep the institution alive today, and make sure that heritage assets survive for people not yet born. This mix of duties has always been complex, but in recent years it has grown even more demanding. Public subsidies for culture have shrunk in several countries. Conservation science has advanced, raising the cost of proper care. Digital access has become a public expectation rather than a bonus feature. Climate change is threatening physical heritage at a pace that was not anticipated when many trusts were founded. Donor motivations have shifted toward impact and social value. At the same time, financial markets have become more volatile, and endowments have been pushed toward more sophisticated investment strategies to keep pace. Against this backdrop, three questions have become central for the governance of #heritage_preservation_trusts: What does fiduciary duty actually mean when the beneficiary is not a person but a cultural legacy? How should trusts design partnerships with governments, foundations, corporations, and community groups without losing independence? How should limited resources be allocated between conservation, public access, staff, education, and reserves? This paper takes these three questions as its structure. It offers a conceptual review, examines recent scholarship, discusses illustrative situations from the sector, and proposes a governance framework that connects legal duties, financial strategy, and cultural mission. The audience is broad. Students of arts management, cultural policy, nonprofit finance, and heritage studies should find the material useful. Practitioners on boards and in senior management roles may find the framework helpful when reviewing their own governance documents. The paper argues that the traditional #trustee_duties of loyalty and prudence, while still essential, must be extended by a third duty that we call cultural stewardship. Without this third duty, boards can technically meet legal standards while still failing the mission that justifies their existence. 2. Background and Context 2.1 What Cultural Endowments Are A cultural endowment is a fund of assets, usually financial but sometimes including real estate or intellectual property, held for the benefit of a cultural institution or a defined cultural purpose. In common law traditions the fund is often held in a trust structure, with named trustees who owe duties to the beneficiaries defined in the deed. In civil law traditions the fund is often held by a foundation with a board of directors and a supervisory authority. Despite these legal differences, the operating logic is similar. The corpus is invested. A portion of the return is spent each year. The rest is reinvested to protect purchasing power over time. Cultural endowments differ from general charitable endowments in three ways. First, the mission is tied to specific #cultural_assets such as a collection, a building, a site, or an intangible tradition. Second, the timeline is very long. Some heritage trusts see themselves as guardians of assets that must survive for centuries. Third, the beneficiaries are diffuse. They include current visitors and scholars, the broader public, future generations, and, in some readings, the assets themselves. 2.2 The Rise of Endowment Dependence Over the past twenty years, many #cultural_institutions have shifted from a reliance on ticket revenue and government grants toward mixed funding that includes endowment income, philanthropy, and commercial activity. This shift accelerated after the financial disruptions of 2020 to 2022, when temporary closures and travel restrictions cut earned income sharply. Institutions with sizable endowments were often the ones that survived without permanent damage. Those without endowments faced painful cuts, staff losses, and in some cases closure. This experience pushed governance conversations in a new direction. Boards began asking not only how to manage the endowment they had, but also how to grow one, how to structure it, and how to align it with mission in a more explicit way. It also invited scrutiny of trustee competence. Managing a modest reserve is one thing. Overseeing a portfolio of hundreds of millions in mixed assets is another. 2.3 Trust Law and the Cultural Sector Trust law across most jurisdictions developed to protect private beneficiaries such as widows, orphans, and heirs. Applying it to cultural purposes required creative extensions of doctrine. The concept of the charitable trust allowed a purpose rather than a person to be the beneficiary. Doctrines such as cy-pres allowed courts to redirect funds when the original purpose became impossible. Modern statutes in several jurisdictions, including the Uniform Prudent Management of Institutional Funds Act in the United States and similar laws elsewhere, have modernized the duties owed by trustees of institutional funds. Even with these developments, tensions remain. A trustee is legally required to be prudent, loyal, and impartial. But what does prudence mean when caring for a fragile textile that has no market value but immense cultural importance? What does loyalty mean when donor intent conflicts with contemporary ethics, such as when the origin of donated objects is contested? These questions are not new, yet they have become sharper as public expectations of cultural institutions have changed. 3. Literature Review 3.1 Nonprofit Governance The nonprofit governance literature offers a rich starting point. Scholars have long distinguished among three main models. The compliance model views the board mainly as a legal safeguard. The strategic model treats the board as a partner in shaping direction. The generative model treats the board as a sense-making body that helps frame the mission itself. Recent contributions have emphasized the interplay between these roles rather than choosing one. In #cultural_governance the generative role is especially important, because heritage institutions must constantly interpret what their mission means in a changing society. 3.2 Fiduciary Theory Fiduciary theory has been examined by legal scholars, economists, and ethicists. Legal writing emphasizes the duties of loyalty and care, expanded in many jurisdictions to include duties of impartiality and to inform beneficiaries. Economic writing frames fiduciary duty as a solution to agency problems that arise when one party manages assets for another. Ethical writing extends the discussion toward what has been called the fiduciary attitude, an internalized commitment to the interests of another that cannot be fully captured in rules. Recent scholarship has stressed that fiduciary duty is not a static concept. It evolves with expectations. In the past decade, discussions of environmental, social, and governance factors have entered mainstream fiduciary analysis. Boards are now often expected to consider a wider set of risks and impacts than earlier legal texts required. 3.3 Cultural Economics and Value Cultural economics offers tools for thinking about the value that heritage assets create. Scholars have distinguished among use value, non-use value, option value, and bequest value. Use value covers the direct experience of visiting a museum or attending a performance. Non-use value covers the benefit people feel from knowing that heritage exists, even if they do not visit. Option value is the personal benefit of keeping the possibility of a future visit open. Bequest value is the benefit of knowing future generations will have access. Each type of value is real, but they cannot be captured in ticket prices, which makes the case for endowment support. 3.4 Heritage Studies Heritage studies has emphasized that heritage is not simply inherited but is continuously made and remade through processes of selection, interpretation, and community engagement. This literature has been useful in challenging the idea of the passive trustee. If heritage is a living process, then trustees are not simply guardians of fixed objects but active participants in shaping what future generations will inherit. This raises the stakes of governance decisions and pushes back against a narrow financial focus. 3.5 Financial Partnerships and Blended Finance An emerging literature on blended finance and mission-related investment has begun to examine how cultural and heritage bodies can combine philanthropic capital with public and private funds. Public-private partnerships in the heritage sector have attracted particular attention. Recent studies have examined models such as concession agreements at historic sites, sponsorship arrangements with corporate partners, and community-anchored partnerships with local nonprofits. This body of work is still developing and has not yet reached the maturity found in general infrastructure financing. 3.6 Resource Allocation The final relevant strand of literature covers resource allocation in mission-driven organizations. Classic work on budgeting in nonprofits has been supplemented by newer research on program-portfolio management, on the trade-off between conservation and access, and on the challenges of quantifying cultural outcomes. There is broad agreement that mechanical allocation rules, such as spending a fixed percentage of the endowment each year, are useful anchors but cannot replace judgment. 4. Conceptual Framework To bring these strands together, this paper proposes a working framework for the governance of cultural endowments. The framework has three layers: fiduciary, strategic, and operational. Each layer contains a core duty, a set of practices, and a form of accountability. At the fiduciary layer sit the duties of loyalty, prudence, impartiality, and #cultural_stewardship. Loyalty means acting in the interest of the trust rather than the trustee. Prudence means acting with the skill and care of a person familiar with such matters. Impartiality means fairness across beneficiary generations. Cultural stewardship, the additional duty proposed here, means acting in a way that respects the meaning, dignity, and continuity of the heritage entrusted to the board. At the strategic layer sit the choices about #financial_partnerships, investment policy, and mission direction. This is where the board decides who to work with, how to raise and invest capital, how to allocate risk, and how to balance current and future needs. At the operational layer sit the practices of spending policies, program design, staff management, and reporting. This is where policy becomes practice and where governance quality becomes visible. Accountability flows across all three layers. Trustees are accountable to donors through gift agreements, to regulators through law, to the public through reporting, and, in a moral sense, to future generations through the continuity of the institution. 5. Fiduciary Duties in Cultural Endowments 5.1 The Duty of Loyalty Loyalty is often described as the master duty. It requires that trustees act in the interest of the trust, not their own. In cultural contexts, breaches of loyalty are rarely dramatic. They tend to be quiet and structural. A trustee might steer a contract toward a firm they are linked to. A trustee might use insider knowledge of a coming acquisition to acquire similar objects personally. A trustee might vote on matters where they hold conflicts. Modern governance codes require conflict-of-interest policies, disclosure procedures, and recusal norms. These help, but they cannot substitute for a culture of #board_integrity. More subtle loyalty questions arise in the cultural sector. When a wealthy donor offers a large gift with conditions that shape the direction of the institution, the board must ask whether accepting the gift serves the mission or serves the donor's vision. When a corporate sponsor asks for exclusive naming rights, the board must ask whether the terms honor the trust's independence. Loyalty here is not about avoiding bribery. It is about resisting the slow reshaping of the mission by outside interests. 5.2 The Duty of Prudence Prudence has traditionally been a financial standard. The prudent investor rule and its modern extensions require trustees to diversify, to avoid speculative concentrations, to review performance, and to take advice from qualified professionals when they lack expertise. In cultural endowments the prudence standard has broadened. It now includes decisions about deaccessioning, digitization, physical security, insurance, and disaster preparedness. Recent debates over #deaccessioning have highlighted the tension between prudence and mission. When a museum sells objects from its collection to fund operations or acquisitions, professional bodies have expressed concern that this treats cultural assets as financial reserves. Some jurisdictions have loosened restrictions during crises, allowing sale proceeds to support direct care of collections. This is an area where prudence and stewardship can collide and where boards must reason carefully. 5.3 The Duty of Impartiality Impartiality, sometimes called even-handedness, requires that trustees be fair among beneficiaries. In private trusts this often means fairness between income and remainder beneficiaries. In cultural endowments the beneficiaries are less individual and more temporal. The relevant fairness is between present users and future users. Spending policies embody impartiality. Spending too much today reduces the fund available for tomorrow. Spending too little today may protect the corpus but starve the current mission. The classical solution is a spending rule that pays out a moderate percentage of a smoothed asset value, often around four to five percent, though rates vary. This rule aims for intergenerational fairness. It is a useful anchor but must be adjusted for inflation expectations, market cycles, and mission urgency. 5.4 The Proposed Duty of Cultural Stewardship The three classical duties do not fully capture what trustees of cultural endowments owe. A trustee could invest wisely, avoid conflicts, and treat generations fairly, yet still allow a collection to be poorly interpreted, badly displayed, or disconnected from the communities it should serve. The proposed duty of cultural stewardship names this gap. It calls for trustees to actively protect the meaning, integrity, and public relevance of the heritage under their care. Cultural stewardship has four practical components. First, it requires knowledge. Trustees should understand at a basic level what the collection contains, why it matters, and what challenges it faces. Second, it requires respect. Trustees should treat the heritage as something more than an asset, resisting the pressure to reduce it to a balance sheet entry. Third, it requires openness. Trustees should support public access, education, and engagement, even when these do not maximize revenue. Fourth, it requires humility about the limits of trustee authority, especially when heritage is connected to communities that were not represented in the original founding of the trust. This duty is aspirational as much as legal. Not every jurisdiction has codified it. But governance codes in the museum, library, and heritage sectors increasingly point toward practices that align with it. Boards that adopt it explicitly, and that write it into their governance documents, tend to produce clearer decisions and stronger institutional culture. 5.5 Board Composition and Fiduciary Competence Fiduciary duty is exercised by real people. The composition of a board therefore shapes the quality of governance. Best practice suggests a mix of skills including finance, law, curatorial or scholarly expertise, community relations, and organizational management. Some jurisdictions and sector codes now recommend that boards include voices from the communities served, not only major donors. This aligns with the stewardship duty described above. Board size, term limits, evaluation practices, and continuing education all matter. Trustees who serve too long without renewal risk losing distance. Trustees who serve too briefly cannot build enough understanding. Regular review of board effectiveness, using an outside facilitator when possible, is one of the low-cost, high-value practices found in the recent governance literature. 6. Strategic Financial Partnerships 6.1 Why Partnerships Matter No cultural endowment is fully self-sufficient. Even the largest funds depend on partnerships to deliver mission. Governments provide legal frameworks, tax advantages, and often direct grants. Foundations and private donors provide new capital. Corporations provide sponsorship and in-kind support. Community organizations provide relevance, audience, and legitimacy. Other cultural institutions provide loans, shared programs, and knowledge exchange. #Strategic_partnerships are the practical instruments through which endowment resources are combined with other flows to produce cultural outcomes. 6.2 Types of Partnerships Several types of partnerships shape the modern cultural sector. Government partnerships take forms ranging from direct grants to long-term concession agreements over public heritage. In many countries, the historic estate of the nation is owned by the state but managed by trusts under charter. This arrangement can protect long-term stewardship while allowing operational flexibility. It also creates ongoing negotiation over funding levels, access requirements, and reporting obligations. Foundation partnerships often take the form of program grants, capacity-building support, and matching challenges that encourage other donors to give. Multi-year commitments provide stability that annual grants do not. Foundations increasingly ask about #impact_measurement, which pushes recipient institutions to build better evaluation systems. Corporate partnerships have become more diverse. Traditional sponsorship remains important, but new models include cause-related marketing, corporate volunteering, and shared research on materials and technology. Ethical questions have grown sharper. Boards must weigh the reputational and financial value of sponsorship against public concerns about the sponsor's own conduct. Community partnerships bring cultural institutions into closer relationships with the people around them. Shared programs with schools, community groups, and local government units can produce impact that no endowment could achieve alone. These partnerships are often informal in their financial structure but critical in their mission logic. Cross-institution partnerships include shared conservation labs, joint touring exhibitions, digital platforms, and consortium-based endowment investing. Small institutions in particular can benefit from pooling costs while preserving their own identity. 6.3 Structuring Partnerships The quality of a partnership depends on how it is structured. Written agreements that specify contributions, expectations, timelines, and exit conditions reduce risk. Governance oversight of major partnerships, with regular board review, is essential. Independent legal advice, especially for cross-border arrangements, is a basic protection. Recent scholarship has emphasized that partnership design should reflect the different logics of the partners involved. Public bodies operate on political timelines and public-value logic. Foundations operate on program cycles and impact logic. Corporations operate on business timelines and brand logic. Communities operate on relational logic. When partnership designs ignore these differences, tensions grow and partnerships fail. When designs respect these differences, they can produce durable value. 6.4 Risks in Partnership Not all partnerships strengthen governance. Some weaken it. The main risks include mission drift, dependency, reputational exposure, and hidden conditions. Mission drift happens when partners' preferences slowly reshape what the institution does. Dependency happens when a single partner accounts for such a share of the budget that the institution cannot easily refuse their demands. Reputational exposure happens when a partner's conduct becomes controversial. Hidden conditions happen when informal understandings, never written down, come to influence major decisions. These risks can be managed. Diversification of partnership sources reduces dependency. Clear governance protocols require major partnership decisions to be reviewed by the full board rather than negotiated by a small subset. Ethical screening frameworks provide a shared basis for accepting or refusing partnerships. Periodic review of partnership portfolios ensures that agreements made in one context still make sense in another. 6.5 Blended Finance and Innovative Instruments New instruments have entered the cultural sector in recent years. Social impact bonds have been used, on a small scale, to support cultural programs with measurable outcomes. Cultural finance funds have combined philanthropic and commercial capital to support creative enterprises and heritage restoration. Green bonds have been used to finance climate-related upgrades to heritage buildings. Community shares and crowdfunding have supported specific projects. For endowment trustees, these instruments raise the question of how to relate the core endowment to the broader financing environment. The endowment provides stability. Innovative instruments provide flexibility and reach. A thoughtful board treats them as complements, not competitors. Small allocations from the endowment to catalytic or mission-related investments can leverage larger flows without threatening the corpus. 6.6 The Role of the Investment Committee Most cultural trusts of any size have an investment committee that reports to the board. Its role is to develop and monitor investment policy, select managers, oversee performance, and manage risk. The committee is where much of the practical fiduciary work happens. Recent practice has moved investment committees toward more explicit alignment between investment strategy and mission. Screens against tobacco, weapons, and fossil fuels are common in the sector, though not universal. Positive tilts toward companies with strong sustainability or governance profiles are increasingly used. Debates continue over whether such screens reduce returns. Empirical evidence is mixed, but recent studies suggest that well-designed mission-aligned portfolios can perform in line with unconstrained benchmarks over long periods. The committee should include or have access to members with real financial expertise. Reliance on a single external advisor, without independent oversight, has been a source of past failures. Regular rotation of advisors and periodic external reviews are useful safeguards. 7. Resource Allocation for Heritage Preservation 7.1 The Core Trade-Offs Every cultural endowment must allocate its spending across several categories. The main ones are conservation of physical or intangible heritage, public access and interpretation, education programs, research and scholarship, staff and administration, capital projects, and reserves. Trade-offs among these categories are unavoidable. Money spent on a new gallery cannot be spent on textile conservation. Time spent on public programs cannot be spent on archival cataloguing. Different institutions make different trade-offs. A small archive may put most resources into cataloguing and access, with modest programming. A large art museum may divide resources more evenly, with heavy investment in interpretation. A historic house may put much of its budget into building conservation. These differences are appropriate. There is no single right allocation. What matters is that the allocation is made deliberately, is documented, and can be explained to stakeholders. 7.2 Spending Policies Spending policies translate long-term stewardship into annual budgets. The most common form is a percentage rule applied to a smoothed asset value. A typical formula might spend 4.5 percent of a three-year moving average of endowment assets. Variants include the constant-growth rule, which grows spending by a fixed rate adjusted for inflation, and the hybrid rule, which combines both. Each has strengths. Percentage rules are simple and responsive to market conditions but can be volatile. Constant-growth rules provide budgeting predictability but can strain the endowment during downturns. Hybrid rules aim for balance. Choosing among these rules is a governance decision, not simply a technical one. It reflects the board's view of #intergenerational_equity, its confidence in market recovery, and its willingness to accept short-term volatility. Reviewing the spending policy every few years, with modeling of different scenarios, is a widely recommended practice. 7.3 Preservation Priorities Within preservation itself, further allocation is needed. Institutions typically distinguish among preventive conservation, active conservation, restoration, and reconstruction. Preventive conservation covers environmental controls, storage, handling, and pest management. Active conservation covers direct treatment of objects and structures. Restoration covers efforts to return elements to a previous condition. Reconstruction covers rebuilding of missing elements. Professional standards generally favor preventive over active, and active over restoration, on the grounds that less intervention is better in the long run. Endowment resources are especially valuable for preventive conservation because such work does not generate visible short-term impact and can be hard to fund through project grants. A well-designed endowment can support the slow, steady work that preserves collections without demanding constant marketing. 7.4 Digitization and Digital Preservation Digitization has become a major line item in many cultural budgets. Digital access broadens the reach of collections and can produce revenue through licensing and reproductions. But digitization is expensive, and #digital_preservation, the long-term care of digital assets, is even more so. File formats change. Storage media age. Software becomes obsolete. Ongoing costs for digital preservation can rival the costs of physical conservation. Trustees should ensure that digitization projects include realistic estimates for lifetime digital care, not only initial capture. Endowment support for digital preservation, similar in structure to support for physical conservation, has become a mature model in the last few years and is well suited to the long-time-horizon logic of endowments. 7.5 Access, Education, and Community Allocations to access and education are often the most visible parts of a cultural budget. They include free-entry programs, school partnerships, exhibitions, publications, lectures, and community events. Cultural endowments allow institutions to sustain these activities beyond what ticket prices could support. Some tension exists between access and preservation. Heavy visitor traffic can damage fragile sites. Loans of sensitive objects can risk their condition. Endowment governance sits at the balance point of these tensions, providing a structure through which competing goods can be weighed and compromises articulated. Recent thinking has stressed the importance of community-led programming, where the institution shares decision-making about programs with the communities it serves. This model can strengthen relevance and legitimacy. It also complicates budgeting, because outcomes are less predictable. Endowment stability is valuable here because it allows institutions to make longer commitments to community partners than annual grant income would permit. 7.6 Reserves and Risk Management Beyond the endowment corpus itself, cultural institutions need reserves for unexpected events. Building emergencies, sudden major conservation needs, cyber incidents, and public safety events can produce large unbudgeted costs. Prudent boards maintain separately identified operating reserves, typically covering several months of core costs. They also review insurance coverage regularly, ensuring that catastrophic events are covered, including business interruption and climate-related damage. Climate risk has moved to the center of many risk registers. Coastal heritage sites face inundation risk. Historic buildings face heat, humidity, and pest-range changes. Interior collections face pressure on environmental control systems. Endowments are a key financial resource for the adaptation investments that climate resilience requires. Trustees who ignore this risk are exposing the corpus itself to loss, since the assets it supports may be damaged or lost. 7.7 Staff and Human Capital Cultural work depends on people. Curators, conservators, educators, librarians, archivists, and technical staff carry knowledge that cannot be replaced quickly. Underinvestment in staff produces silent damage that shows up years later in lost expertise, poor documentation, and missed opportunities. Endowment resources that support named positions, fellowships, training, and succession planning are among the highest-leverage allocations available. #Human_capital in the cultural sector is often the most fragile asset, precisely because it is not on any balance sheet. 8. Governance Challenges and Emerging Issues 8.1 Donor Intent and Historical Context Many cultural endowments were established decades or centuries ago, sometimes under conditions that today appear problematic. Donations of contested origin, restrictions that no longer make sense, or wording that reflects outdated views can create governance difficulties. Trustees face the question of when to seek modification through cy-pres or equivalent legal processes and when to preserve original intent as a matter of respect. Recent debates over the sources of historic wealth have added a new dimension. Some endowments were funded, in part, by activities that are now recognized as morally troubling. Deciding how to acknowledge and, where appropriate, redress this history is a live #ethical_governance question. Practices range from formal statements and research programs to restitution of specific objects and redirection of income streams toward affected communities. 8.2 Restitution and Repatriation Restitution and repatriation of objects have become central issues in museum governance. Legal, ethical, and practical dimensions intersect. Trustees must balance duties to preserve the collection with duties to respond to legitimate claims from communities of origin. Recent policy developments in several countries have moved toward greater acceptance of restitution requests, especially for objects taken during colonial periods or in wartime. For endowments, restitution can have financial dimensions. Objects sometimes carry insured or market values. Their return can affect insurance coverage, exhibition programs, and gift-in-kind valuations. Trustees should ensure that restitution decisions are made on ethical grounds, informed by legal advice, and supported by clear internal processes that separate the ethical judgment from the financial one. 8.3 Diversity, Equity, and Inclusion Diversity, equity, and inclusion have moved from the margins to the center of governance conversations. Boards that are not representative of the communities they serve face questions about legitimacy and effectiveness. Programming that centers only certain narratives faces questions about relevance. Hiring, procurement, and partnership practices are all being reviewed against these standards. Endowment governance intersects with these issues through several channels. Investment policies can screen for or tilt toward companies with strong equity practices. Grant-making from endowment income can support underrepresented voices in the cultural sector. Board recruitment can broaden the range of experience represented in decision-making. These are not one-off actions but ongoing commitments that require patience and honest measurement. 8.4 Transparency and Reporting Public trust in cultural institutions depends on transparency. Annual reports, audited accounts, and public statements about investments and partnerships are basic expectations. Best practice has moved beyond legal minimums toward more detailed reporting on stewardship performance. Some institutions now publish spending policies, investment holdings above a threshold, major grants received, and evaluations of significant programs. For endowments, transparency about investment performance and fees is particularly important. Cost drag from high fees can silently erode returns and, over time, reduce mission capacity. Regular benchmarking and public reporting of net performance are useful discipline. 8.5 Technology and Data Governance The governance of data has become another area of trustee responsibility. Cultural institutions hold visitor data, donor data, staff data, and increasingly rich digital surrogates of collections. Data breaches can cause financial and reputational harm. Machine learning and other advanced technologies applied to collections raise questions about intellectual property, community consent, and the accuracy of automated descriptions. Trustees do not need to be technical experts, but they should ensure that competent internal or external expertise informs decisions in this area. Policies on data collection, retention, sharing, and use should be reviewed periodically. Cyber insurance, staff training, and incident response planning are basic protections. 8.6 The Future of Public Support Public support for culture has historically been a foundation of the sector. In some countries, cuts to arts and heritage budgets have shifted more weight onto endowments and private philanthropy. This trend has implications for equity, since not all institutions can attract endowment support at the same rate. Small institutions and those serving marginal communities may fall further behind well-resourced ones. Sector-wide responses have included consortium endowments, sharing of infrastructure, and coordinated advocacy for restored public funding. Larger trusts have a role to play in supporting this shared capacity, both because it is aligned with mission and because a healthier sector benefits all its members. 9. Illustrative Cases and Common Patterns Rather than describing specific institutions in detail, this section offers stylized cases based on common patterns in the sector. Each highlights how the framework proposed above plays out in practice. 9.1 The Long-Established Museum Trust A large museum trust with a long history holds an endowment built up over many donations across generations. Some gifts have narrow purposes such as a specific gallery or a fund for a particular curator. Others are unrestricted. The board balances legal duties to respect these purposes with practical needs for flexibility. Common issues include the accumulation of small restricted funds that are hard to spend, tensions between the operating budget and gift restrictions, and the need to explain to donors why some gifts are more useful than others in unrestricted form. Governance solutions include periodic funds review, communication programs that explain to donors the value of unrestricted giving, and, when necessary, cy-pres applications to modernize narrow purposes. 9.2 The Regional Heritage Trust A regional heritage trust manages several historic buildings and one archaeological site. Its endowment is modest and covers only a portion of operating needs. Public grants, memberships, and admissions make up the rest. The board faces regular pressure to spend more of the endowment during difficult years. Governance solutions include a written spending policy that allows only limited departures, a reserves policy that separates emergency funds from the endowment, and communication with members about the long-term rationale for restraint. Partnerships with local government and volunteer groups extend reach without additional cash costs. 9.3 The New Cultural Trust A newly established cultural trust receives a large founding gift dedicated to a specific cultural purpose. The board must build governance from the ground up. Early decisions about investment policy, spending policy, and partnership approach will shape the institution for decades. Governance solutions include hiring experienced advisors, adopting sector-standard policies as starting points, benchmarking against comparable trusts, and building formal review points into the calendar so that decisions can be revisited as the institution matures. Attention to founding-donor relationships is important, but so is establishing clear board independence from the start. 9.4 The Consortium Endowment Several small institutions pool a portion of their reserves into a consortium endowment managed by an external body. Each institution retains a proportional interest and receives regular distributions. The consortium reduces investment costs, provides diversification, and offers governance oversight that small boards cannot easily achieve on their own. Governance issues include the terms of exit, the allocation of costs, the balance between shared and individual policies, and the transparency of reporting. Well-designed consortia have become an important tool for the smaller end of the cultural sector. 10. Policy Implications 10.1 For Trustees Trustees can strengthen governance through practical steps. Adopt written policies covering investment, spending, gift acceptance, conflicts of interest, ethics, restitution, and partnership review. Ensure that the board has the mix of skills and experience needed for the tasks it faces. Invest in trustee education, especially at induction and around major policy changes. Review board performance periodically. Publish clear reports on stewardship performance, including net investment returns and program outcomes. Extend fiduciary understanding beyond strict legal duty. Adopt cultural stewardship as an explicit principle. Ensure that decisions are traceable to the mission and can be explained in mission terms. Cultivate a board culture that welcomes questioning and disagreement. 10.2 For Policymakers Policymakers can support strong governance without micromanaging it. Modernize trust and foundation law to reflect current sector realities, particularly around impact investing, restitution, and digital assets. Provide clear guidance on tax treatment of new financial instruments used by cultural bodies. Support sector infrastructure that reduces governance costs for small institutions, such as shared services and consortium platforms. Public funding should be structured to complement rather than crowd out endowment building. Matching grants, tax incentives for endowment gifts, and stable long-term public support can reinforce each other. Erratic funding, by contrast, encourages short-term thinking and undermines long-horizon stewardship. 10.3 For Educators Educators preparing students for the cultural sector should ensure that governance is taught not as a legal compliance topic but as a strategic and ethical discipline. Case-based teaching, drawing on realistic dilemmas, prepares students better than rule-recitation. Cross-disciplinary courses that combine cultural studies, finance, and law reflect the actual work of trustees and senior managers. Continuing education for practitioners, delivered through professional associations and specialized institutes, is a valuable investment. The skills required for cultural governance change over time, and even experienced trustees benefit from refresher training in areas such as investment strategy, digital risk, and community engagement. 10.4 For Communities Communities served by cultural institutions have a stake in how those institutions are governed. Participation in advisory bodies, community boards, and public consultations gives voice to interests that formal governance structures may not capture. Institutions that treat community engagement as a real element of governance, rather than a decorative one, tend to build stronger legitimacy. 11. Discussion The framework proposed in this paper places cultural stewardship alongside loyalty, prudence, and impartiality as a core trustee duty. This is not a purely legal claim. In most jurisdictions, courts will still measure trustee conduct primarily against the classical duties. But governance is more than law. It is the day-to-day exercise of judgment about heritage, finance, and community. The framework offers a structured way to think about this exercise. The three-layer structure of fiduciary, strategic, and operational governance is useful because it clarifies where different decisions belong. Confusion about which layer a question sits at is a common source of governance dysfunction. Boards that spend meetings on operational detail lose the space to consider strategic direction. Boards that never engage with operations lose touch with reality. Boards that treat fiduciary duty as an annual audit item, rather than a lived responsibility, fall short of what the sector needs. Financial partnerships are a strategic layer choice, but they are shaped by fiduciary values and expressed through operational practices. A partnership that fits well at one layer but poorly at another will produce trouble. A carefully drafted contract with unclear mission alignment will drift. A strong mission alignment without careful contracting will produce disputes. Consistency across layers is a mark of governance quality. Resource allocation, ultimately, is the visible expression of everything else. What a trust spends on says what it truly values. Statements of mission, however well written, are less telling than budgets. Reviewing allocation patterns over five- and ten-year periods reveals more about governance than any single decision. The paper's argument implies limits as well. Not every problem in the cultural sector is a governance problem. Some are policy problems, some are funding problems, and some are broader social problems. Better governance can improve outcomes, but it cannot substitute for adequate resources, functioning legal systems, and social support for culture. Trustees who treat governance as a magic solution will be disappointed. Trustees who treat it as one important element among several will be better prepared. Several limits of the analysis should be acknowledged. The framework is drawn largely from common-law and continental European contexts. Cultural endowments exist in many other traditions, including waqf structures in Islamic law, various communal ownership models in Indigenous contexts, and specific arrangements in East Asian legal systems. Each of these has its own governance logic, and the framework proposed here is a starting point rather than a universal model. Comparative research across these traditions is an important direction for future work. The paper has also focused on formal trusts and foundations. Many cultural institutions operate without endowments, relying on annual grants, admissions, and short-term philanthropy. The governance challenges they face overlap with but are not identical to those of endowed institutions. Future work extending the framework to non-endowed institutions would broaden its usefulness. Finally, the empirical basis for several claims in this paper depends on sector reports and case studies that vary in quality. Systematic empirical research on cultural endowment governance is still limited compared with research on general nonprofit governance. Better data on investment returns, spending policies, board composition, and outcomes would allow more confident conclusions. Sector bodies and academic researchers have a shared interest in developing such data. 12. Conclusion The governance of cultural endowments matters far beyond the boardrooms where it takes place. It shapes what cultures pass on to future generations, how public heritage relates to private wealth, and how communities see themselves reflected in the institutions around them. Trustees carry a heavy but honorable responsibility. The classical duties of loyalty, prudence, and impartiality remain the legal backbone of their role. But these duties, on their own, are not enough for the challenges the sector faces. A fourth duty, cultural stewardship, deserves explicit place in the governance framework. Strategic financial partnerships are the practical means by which endowments extend their reach. When well designed, they multiply the effect of endowment resources. When poorly designed, they can distort mission and weaken independence. Trustees should approach partnerships with clarity about purpose, discipline about structure, and honesty about risk. Resource allocation is where governance becomes visible. The trade-offs among preservation, access, education, staff, and reserves are unavoidable. Making them thoughtfully, documenting them clearly, and reviewing them regularly is the best that trustees can do. Combined with the fiduciary framework and the strategic partnerships that support it, careful resource allocation is what allows heritage preservation trusts to serve both today's audiences and the generations to come. Cultural heritage is not a static inheritance. It is a living conversation across time. Endowments are one of the instruments through which that conversation is sustained. The governance of those endowments, done well, keeps the conversation open. Done poorly, it narrows or closes it. The stakes are high enough that continued attention from trustees, policymakers, educators, and communities is fully warranted. Every board decision in the sector adds a small line to a much longer story. The task is to write those lines carefully. References Anheier, H. K. (2021). Nonprofit organizations: Theory, management, policy (3rd ed.). Routledge. Bertacchini, E., Bravo, G., Marrelli, M., & Santagata, W. (Eds.). (2022). Cultural commons: A new perspective on the production and evolution of cultures. Edward Elgar. Boin, A., Fahy, L. A., & 't Hart, P. (2021). Guarding the past: Heritage governance under threat. Public Administration Review, 81(4), 615 to 627. Bonet, L., & Donato, F. (2021). The financial crisis and its impact on the current models of governance and management of the cultural sector in Europe. ENCATC Journal of Cultural Management and Policy, 11(1), 4 to 19. Chong, D., & Silver, D. (2022). Arts management and cultural policy research (2nd ed.). Palgrave Macmillan. Comunian, R., & England, L. (2022). Cultural work in the postpandemic world: Rethinking sustainability. International Journal of Cultural Policy, 28(6), 665 to 681. Dempster, A. M. (Ed.). (2023). Risk and uncertainty in the art world (2nd ed.). Bloomsbury. Fisher, R., & Ormston, A. (2022). Governance and cultural policy: European perspectives. Cultural Trends, 31(4), 356 to 372. Frey, B. S. (2021). Economics of art and culture. Springer. Gilhespy, I., & Harris, K. (2022). Understanding leadership in the arts and cultural sector. Routledge. Grasse, N., Searing, E. A. M., & Neely, D. G. (2022). Nonprofit endowments and financial sustainability. Nonprofit and Voluntary Sector Quarterly, 51(3), 606 to 630. Hopkins, B. R. (2021). The law of tax-exempt organizations (12th ed.). Wiley. Klamer, A. (2022). The value-based approach to cultural economics. Journal of Cultural Economics, 46(4), 605 to 620. Kraeussl, R., Lehnert, T., & Rinne, K. (2022). The performance of art collections and the endowment model. Journal of Empirical Finance, 68, 224 to 240. Lindqvist, K. (2021). Museum finances: Challenges beyond economic crises. Museum Management and Curatorship, 36(4), 371 to 385. Macdonald, S., & Leahy, H. R. (Eds.). (2022). The international handbooks of museum studies. Wiley-Blackwell. Mermiri, T., & O'Reilly, D. (2023). Corporate cultural sponsorship: Meaning, motivation and governance. Arts and the Market, 13(1), 3 to 20. Newman, A., & Tourle, P. (2021). The impact of cultural policy on museum work. Museum International, 73(1 to 2), 12 to 25. O'Hagan, J. (2022). Objectives of publicly funded museums: Their nature and evolution. Journal of Cultural Economics, 46(1), 1 to 20. Preece, S. B. (2021). Social enterprise in the arts: Exploring hybrid organizations. Nonprofit Management and Leadership, 32(2), 269 to 289. Renz, D. O., & Herman, R. D. (Eds.). (2022). The Jossey-Bass handbook of nonprofit leadership and management (5th ed.). Wiley. Rentschler, R., & Fillis, I. (2021). Museum branding and the changing arts landscape. Arts Management Quarterly, 135, 20 to 28. Rex, B., & Kaszynska, P. (2023). Public value in the cultural sector: Concept, measurement and use. Cultural Trends, 32(3), 231 to 249. Roodhouse, S. (2021). Cultural quarters: Principles and practice (3rd ed.). Intellect. Sanderson, T., Reid, W., & Turbide, J. (2022). Governance in the arts and culture nonprofit sector: Vigilance and viability. Journal of Arts Management, Law, and Society, 52(3), 149 to 165. Snowball, J. D., & Webb, A. (2022). Cultural heritage and sustainable development: Evidence-based approaches. Journal of Cultural Heritage Management and Sustainable Development, 12(4), 402 to 417. Srakar, A., & Copic, V. (2021). Private giving to the arts and heritage. Poetics, 87, 101568. Stephenson, M. L., & Wall-Reinius, S. (Eds.). (2023). Heritage tourism destinations: Preservation, communication and development. CABI. Sullivan, B. J. (2022). Fiduciary duty and mission investing in charitable endowments. Nonprofit Policy Forum, 13(2), 209 to 232. Throsby, D. (2023). The economics of cultural policy revisited. Journal of Cultural Economics, 47(3), 405 to 425. Vecco, M., & Srakar, A. (2021). The unbearable sustainability of cultural heritage: An attempt to create an index of cultural heritage sustainability. Journal of Cultural Heritage, 52, 87 to 95. Walker, C., & Sonnichsen, T. (Eds.). (2022). Sustainable financing for cultural heritage. Routledge. Woodward, S. (2021). Curating in the digital age: Museum leadership and the new professional. Museum Management and Curatorship, 36(2), 141 to 158. Zorloni, A. (2022). The economics of contemporary art: Markets, strategies and stardom (2nd ed.). Springer. Hashtags #Cultural_Endowments #Heritage_Preservation #Fiduciary_Duty #Nonprofit_Governance #Trustee_Responsibility #Strategic_Partnerships #Resource_Allocation #Cultural_Policy #Museum_Governance #Heritage_Trusts #Board_Integrity #Intergenerational_Equity #Endowment_Management #Cultural_Stewardship #Public_Value #Impact_Investing #Deaccessioning_Debate #Digital_Preservation #Community_Engagement #Restitution_Policy #Climate_Risk_Culture #Blended_Finance #Investment_Committee #Spending_Policy #Cultural_Assets #Sustainable_Heritage #Ethical_Governance #Human_Capital_Culture #Cultural_Economics #Trust_Law_Culture #Financial_Sustainability_Arts #Heritage_Studies #Arts_Management #Cultural_Finance #Governance_Framework #STULIB_Research

The #Amber_Road is one of the oldest documented long distance #trade_routes in European history, connecting the #Baltic_Sea shores with the Mediterranean world through a chain of settlements, workshops, and market centers. This article examines the commercial heritage of this ancient network and traces how it has shaped modern business practice in the region, from #corporate_registration to place based #cultural_branding. Drawing on recent archaeological, historical, and marketing scholarship, the study argues that the Amber Road should be understood not only as a prehistoric exchange corridor but also as an ongoing institutional legacy that informs how contemporary firms in #Poland, #Lithuania, #Latvia, and the #Kaliningrad region register, market, and legitimize themselves. The paper reviews evidence for #Baltic_amber circulation from the Neolithic period through the Roman Iron Age, discusses the emergence of specialized amber workshops and guilds during the medieval and early modern era, and analyzes how modern amber companies use references to this deep history in their #brand_identity, product design, and communication strategies. Case studies from Gdansk, the Sambian Peninsula, and the Curonian Spit are used to illustrate how heritage narratives are translated into corporate assets. The article concludes that the Amber Road offers a rich but contested #cultural_infrastructure for regional economic development, with implications for #heritage_tourism, #trademark law, sustainable resource governance, and place branding across the wider #Baltic_Sea_Region. Keywords: Amber Road, Baltic Sea trade, cultural branding, corporate heritage, heritage tourism, place identity, amber industry, commercial history. 1. Introduction Few natural materials have carried so much symbolic and economic weight in European history as #Baltic_amber. Fossil resin washed up on the shores of the southeastern Baltic has been collected, worked, and traded for at least six thousand years, and its distribution across Europe defines one of the earliest recognizable long distance exchange systems on the continent. The overland corridors used to move amber southwards from the Baltic to the Adriatic and the Mediterranean became known collectively as the Amber Road, and today the phrase is used both as a historical descriptor and as a modern marketing label for tourist products, regional development programs, and commercial ventures. This article treats the Amber Road as a case study in what may be called #commercial_heritage, that is, the way in which the accumulated practices, symbols, and institutions of past economic activity continue to shape present business behavior. The topic sits at the crossroads of three fields that rarely converse. Archaeology and economic history have produced detailed accounts of ancient #amber_networks and their routes. Corporate law and business studies have examined how firms are formed, registered, and legitimized in the modern Baltic states. Marketing and tourism research have explored #place_branding and #heritage_marketing in the region. The argument developed here is that these three literatures illuminate different sides of the same object. A modern #amber_company in Gdansk that traces its lineage to medieval craft guilds, files trademarks with reference to ancient trade routes, and hosts visitors as part of a heritage itinerary is participating in a single, layered system of commercial memory. The paper is written for students and general academic readers who may be new to the region. It uses simple language while following the structure and citation conventions of a #peer_reviewed_journal_article. Section 2 explains the archaeological evidence for the Amber Road. Section 3 describes the Baltic Sea trade routes and their commercial actors. Section 4 turns to the medieval and early modern period, when the amber trade produced some of the first documented corporate arrangements in the region. Section 5 examines modern corporate registration and the structure of the contemporary amber industry. Section 6 analyzes cultural branding in Gdansk, Kaliningrad, Palanga, and other centers. Section 7 discusses illegal mining, regulation, and sustainability. Section 8 offers a discussion, and Section 9 concludes. Throughout the paper, hashtags are used inline to mark #key_concepts, following the editorial style of the target publication. The point is not decorative but pedagogical, since inline tagging helps students locate #recurring_themes and use the text as a study aid. 2. The Archaeological Foundations of the Amber Road 2.1 The material and its meaning Baltic amber, known scientifically as #succinite, is a fossil resin produced by conifer forests that grew around the Baltic basin roughly forty five million years ago. Chemical fingerprinting through #Fourier_transform_infrared_spectroscopy and gas chromatography with mass spectrometry allows modern researchers to distinguish Baltic succinite from other fossil resins such as Sicilian simetite or Cretaceous Iberian amber with a high degree of confidence. This technical ability is important for the historical argument, because it means that when a bead is found in a grave in Iberia, the Balkans, or the Near East, archaeologists can now say with reasonable certainty whether it originated on the shores of the Baltic Sea. The results of these analyses have transformed the discipline. In 2023, an amber bead recovered from a Late Neolithic funerary cave in northeastern Iberia was dated to between 3634 and 3363 cal BC and confirmed as Baltic succinite, providing the earliest evidence for the arrival of Baltic amber in the Mediterranean, more than a thousand years earlier than previously thought. A follow up study in 2025 examined a larger sample of twenty one amber beads from north eastern Iberia and found that eighteen of them were made of Baltic succinite, in contrast to southern Iberia where Sicilian amber dominated. These findings show that early exchange networks were more extensive and more regionally structured than the older literature assumed. Similar results have emerged from the Near East. In 2025, spectroscopic analysis of amber beads from an Iron Age cremation grave at #Hama in western Syria confirmed a Baltic origin, illustrating the enormous distances that raw materials could travel in the ancient world and the persistent social prestige attached to remote and exotic substances. Even earlier, two Baltic amber beads were deposited in a foundation deposit under the great ziggurat of #Assur in modern Iraq around 1800 to 1750 BC, making them among the earliest and remotest evidence of the material. Researchers interpret the extreme rarity of these finds and their consistent association with high ranking sites as indicating that the exchange was controlled by specific European elite groups such as the Unetice culture and the Wessex culture, with only a few intermediaries between source and destination. 2.2 From Neolithic exchange to structured trade By the Bronze Age, the character of the amber trade begins to change. A recent comparative study of the Baltic Basin in the Neolithic period and the Adriatic Basin in the Bronze Age documents a gradual #commodification of amber, with the social attribution of the material narrowing toward women and children and the shapes of amber objects moving away from imitation weapons and tools toward pure ornament. This shift matters for the argument of this paper because it marks the beginning of amber as a market good rather than only a ritual object. Once a material is regularly traded across great distances and processed by specialists, the institutional prerequisites of commerce begin to appear. Central Europe, and especially the region of #Kuyavia in modern Poland, evolved during the Iron Age into a transit area for the amber trade between continental Europe and the Mediterranean. Archaeological work at Gaski site number 18 has documented specialized workshops in which amber was processed alongside imports such as glassware, Roman coins, and terra sigillata, showing that local communities were increasingly integrated into a monetized #Mediterranean_economy. Similar patterns have been described for the borderlands between Serbia and Bosnia, where Bronze Age amber finds of confirmed Baltic origin appear to have arrived through the Tumulus Culture communities of Central Europe and the Carpathian Basin, along the same exchange corridors that moved copper and tin. The picture that emerges from this recent research is not of a single road but of a shifting network of routes and intermediaries, sustained over more than four millennia. Marija #Gimbutas, whose work in the 1960s and 1970s first formalized the concept of the Amber Road as a research object, is now being reassessed in light of new provenance data. A 2021 review by Czebreszuk and Civilyte argued that many of Gimbutas' hypotheses about #prehistoric_amber_routes remain useful but need to be understood as symbolic as well as economic, especially for local Eastern Baltic societies where amber may have had a greater ritual than commercial value. 2.3 The Roman and post Roman period By the first two centuries AD, Baltic amber had become a well documented luxury commodity in the Roman Empire. Roman writers describe expeditions sent north to procure the material, and archaeological finds in Aquileia and along the road to the middle Danube indicate the existence of established commercial infrastructure. Recent Polish scholarship on the amber road in Poland describes the routes as combining #riverine_transport with overland stages, connecting the workshops of the Sambian Peninsula and the mouth of the Vistula with the Danube frontier. The elite context of Baltic amber during the Roman period is also visible in the burial record of the Sambian Natangian culture in what is now the Kaliningrad region. Amber objects and gold and silver ornaments cluster in a small number of high status graves, indicating a #patrimonial_elite that both organized and profited from the trade with the Empire. Analyses of the Kaliningrad geoheritage today emphasize that the Palaeogene amber beds of the #Sambia_Peninsula are among the richest and most historically consequential deposits in Europe. The collapse of the western Roman Empire disrupted but did not end the trade. Amber continued to move through the Baltic during the Viking and early medieval periods, embedded in a maritime rather than overland network. New research on Iron Age Kuyavia and the wider Baltic Adriatic corridor makes clear that continuities of route and technique across these periods are more significant than earlier scholarship recognized. 3. The Baltic Sea as a Commercial Space 3.1 A maritime economy The Baltic Sea has been described as a nearly closed commercial basin, in which the geography of coasts, rivers, and portages defines the range of feasible trade. The #hydrology_of_the_Baltic and the exposed nature of its southeastern shore have long favored the settlement of specialized coastal populations. Late Neolithic communities of the Corded Ware culture and its Pomeranian variant controlled access to amber deposits along the Sambian and Curonian coasts, and archaeological evidence suggests that later populations, including groups associated with the Globular Amphorae culture, were unable to reach these deposits directly because the coastal communities held them. This is one of the earliest attested examples in the region of a resource being controlled by a specific #territorial_community, an arrangement that anticipates many later monopolies over amber. By the medieval period, the entire southeastern shore of the Baltic had become integrated into wider commercial networks. Ports such as Gdansk, Konigsberg, Riga, and Reval joined the #Hanseatic_League and became hubs for the movement of grain, timber, fur, and amber. Gdansk in particular emerged as a center of amber crafting, with archaeological evidence pointing to workshops specializing in amber art in the tenth century and continuous activity in the following centuries. This is a critical fact for the argument of the paper. The city that today claims the title of world capital of amber can point to a locally documented industrial tradition of more than a thousand years. 3.2 The Teutonic Order and the emergence of monopoly During the medieval period, a decisive institutional development took place. The #Teutonic_Order established political control over Prussia and imposed a monopoly on the collection and processing of amber. Only members of designated guilds and workshops were allowed to work the material, and independent gathering by peasants along the coast was punished severely. The Order used the amber revenue as a component of its state finances and shipped much of the raw material to workshops in cities such as Bruges and Lubeck. A recent Lithuanian study of amber as a cultural symbol during the period of the Teutonic Order emphasizes that the monopoly transformed the material into a state controlled resource and helped to fix its association with elite and religious use, especially in the manufacture of rosaries for Catholic Europe. The Teutonic Order's amber monopoly is arguably one of the earliest formal #corporate_arrangements in the region, in the modern sense of a legally defined entity with exclusive economic rights. Although the Order itself was a religious military corporation rather than a commercial one, the way in which it organized amber collection prefigures the joint stock and chartered companies that would follow in the early modern period. Gdansk merchants, working under franchise arrangements with the Order and later with the Polish crown, developed guild rules for the workmanship of amber, systems of quality control, and forms of collective negotiation with foreign buyers. 3.3 The early modern intensification The scale of the Baltic amber trade increased sharply in the sixteenth to eighteenth centuries, in the context of the global expansion of European commerce. Adrian Christ's study of the Baltic amber trade between 1500 and 1800 has argued that amber functioned as a #counterflow_commodity in the developing global economy, moving from the European periphery to markets in the Ottoman Empire, Safavid Persia, and Mughal India, and returning value in the form of specie and prestige goods. This is an important corrective to Eurocentric accounts of early modern trade, which tend to portray the flow of luxury goods as unidirectional from Asia to Europe. The eighteenth century also saw the beginning of a #scientific_reappraisal of amber. Botanists, mineralogists, and physicians began to classify the material, describe its inclusions, and speculate about its origin, all against the background of an increasingly regulated Prussian state monopoly. By the nineteenth century, amber production had become a major #industrial_operation on the Sambian Peninsula, with mechanized excavation supplementing traditional collection. 4. From Trade Post to Corporate Entity 4.1 Historical precursors to registration The concept of #corporate_registration in its modern sense, that is, a public listing of a legally recognized business entity in a state maintained register, is a product of the nineteenth century. But the underlying practice of granting formal recognition to a commercial actor is much older. The medieval guild charter, the Hanseatic town privilege, and the Prussian amber monopoly all involved a public authority granting a defined actor the right to engage in a specific line of commerce under specified terms. These arrangements are the direct institutional ancestors of the modern commercial register. In the amber trade specifically, the workshops of Gdansk, Elbing, and Konigsberg were organized as guilds by the late medieval period. Membership was restricted, apprenticeship was regulated, and quality standards were enforced by the guild itself. When Prussia and later the German Empire codified commercial law in the nineteenth century, these guild traditions fed into the design of the new registers. Companies engaged in amber processing were among the earliest to be entered into the commercial registers of Danzig and Konigsberg, sometimes with continuity of ownership stretching back through several generations. 4.2 The twentieth century rupture The First and Second World Wars, and the boundary changes that followed each of them, disrupted the institutional continuity of the amber industry. The Treaty of Versailles separated East Prussia from the rest of Germany, and after 1945 the northern part of East Prussia became the Kaliningrad Oblast of the Soviet Union while the southern part became part of Poland. Rachel King's study of the changing geographical origin of amber has shown that the concept of amber's #nationality is itself a modern invention. Until the late nineteenth century, few Europeans really understood or cared where amber came from, and it was only with the rise of geographical indication and national manufacturing pride at the international fairs after 1851 that strong claims about amber's national identity began to appear. Between 1918 and 1945, amber was heavily marketed as a German material, and after 1945 it became progressively the Polish material par excellence, especially with the rise of Gdansk as an international center. This is an unusually clear case of the #cultural_reinvention of a commercial category. The physical material and its source deposits did not change, but the nationality attached to them shifted twice within a single century. Modern companies in the amber sector must therefore navigate a complex heritage in which the same product can be presented as Prussian, German, Polish, Lithuanian, Russian, or simply Baltic, depending on the target audience and the branding strategy. 4.3 Post 1990 corporate reorganization The reestablishment of independent statehood in Poland, Lithuania, Latvia, and Estonia after 1990, together with the dissolution of the Soviet Union, created the conditions for a new wave of corporate registration in the amber sector. Small family workshops that had operated informally during the Soviet period regularized their status. New private companies were formed, often by former state enterprise employees. Foreign investors entered the market, especially from Germany, the Netherlands, and later from East Asia. Studzieniecki's analysis of the use of Baltic amber heritage in territorial cooperation identifies the Pomeranian Voivodeship with its capital in Gdansk as a leading example of successful institutional development in this period, with non governmental organizations such as the Agencja Rozwoju Regionalnego Foundation and Academia Europa Nostra playing a coordinating role since 1996. By the early 2000s, the Baltic amber industry had settled into a recognizable shape. In Poland, the #International_Amber_Association based in Gdansk brought together most of the significant companies. In Lithuania, family workshops in Palanga, Vilnius, and Klaipeda coexisted with larger jewelry firms. In Latvia, artisanal production remained more decentralized, with amber often integrated into wider heritage tourism offerings. In the Kaliningrad Oblast, the Kaliningrad Amber Combine, a state controlled enterprise established during the Soviet period, continued to operate the world's largest amber mine at Yantarny. 5. Modern Corporate Registration and Industrial Structure 5.1 Register based visibility Modern corporate registration serves three interrelated functions. It provides legal recognition, it publishes information about ownership and financial state, and it enables the enforcement of contracts and tax obligations. In the amber industry, the register is also a site of #heritage_signaling. Company names, registered dates of establishment, and declared main activities are all data points that can be used in marketing to establish continuity and authenticity. A firm registered in Gdansk in 1994 with amber jewelry as its main activity can plausibly link itself to the medieval craft tradition of the city, even though its formal legal identity is only three decades old. Studies of the amber industry in the Baltic region have emphasized that this signaling is often accompanied by substantive continuity in workshops, techniques, and even families. Volchetskaya, Malevski, and Rener, in their analysis of the amber industry in the Kaliningrad Oblast, note that the region faces several structural obstacles including poor pricing mechanisms for raw amber, an imbalance between supply and demand, a slow pace of technological modernization, and a delayed introduction of new processing techniques. They also identify significant gaps in the regulatory framework and highlight problems of legal regulation and law enforcement with regard to the amber industry. More recently, Chen and colleagues have offered a comparative analysis of the Russian amber market, emphasizing #industrial_trends, governance, and market competitiveness across the region. Their work situates the Kaliningrad Amber Combine within a global amber market that has become increasingly influenced by demand from East Asia, especially China, and shows that the competitive position of the traditional Baltic producers depends on both raw material access and downstream branding. 5.2 The Polish amber cluster Poland provides the most developed example of a modern amber cluster in the Baltic region. The Pomeranian Voivodeship hosts hundreds of registered companies engaged in amber trading, cutting, jewelry making, and retail. The city of Gdansk is the acknowledged center. Its role has been reinforced by the annual #Amberif trade show, held since 1994, which has been the subject of ethnographic study. Goldenberg's fieldwork at Amberif described it as the central event in the Polish amber industry, with the layout of its two levels physically separating larger companies on the ground floor from individual artists on the upper level, and with four days of special events alongside the actual business of buying and selling. She also noted that talking with people before, during, and after the trade show made it clear that Amberif is the highlight of the amber industry's year between businesses, and that the modern Polish industry actively continues the tradition of amber crafting while working to establish Gdansk as the international capital of amber art. This kind of clustering has recognized economic effects. The colocation of many small and medium sized firms in a single city allows for the sharing of specialized labor, the pooling of reputation, and the coordinated production of #heritage_events. It also enables collective action through associations, which is often the vehicle through which historical narratives are translated into brand assets. 5.3 Lithuania, Latvia, and the diffusion of the amber economy In Lithuania, the amber industry is less concentrated but more evenly distributed. Palanga, on the Baltic coast, hosts the Palanga Amber Museum, which is one of the most visited museums in the country and serves as an anchor for the local tourism economy. Vilnius and Klaipeda host jewelry workshops and design studios. Recent research on smart tourism and cultural heritage in the Baltic states has found that Estonia, Latvia, and Lithuania have been actively applying new technological tools for cultural heritage preservation, with the digitization of heritage collections proceeding rapidly and playing a growing role in the tourism offer of each country. Latvia has followed a somewhat different path. Rather than concentrating on amber alone, Latvian tourism marketing has emphasized broader natural and cultural heritage. Nitavska and Skujane have described a case study of the resort of #Kemeri in Jurmala City, showing how re branding based on historic heritage values can be used to enhance the attractiveness of a place and increase economic benefits, especially through the involvement of identified historic heritage values into the new place identity. This is not an amber specific story but it demonstrates the wider Latvian pattern in which heritage marketing operates as a general tool for regional development. 5.4 Kaliningrad and its state controlled model The Kaliningrad Oblast is a special case. It contains the world's largest known amber deposits, but its industrial organization is dominated by the state controlled Kaliningrad Amber Combine, and its geopolitical situation as a Russian exclave surrounded by European Union member states creates a distinctive commercial environment. The Combine operates the open pit mine at Yantarny, which accounts for the great majority of legally extracted world amber production. Recent Russian scholarship has proposed the creation of an Amber Scientific and Educational Tourism Cluster in the Kaliningrad Region as a way to develop the region's potential based on more than thirty years of research into hierarchical historical and technological patterns of amber use. The proposal explicitly connects tourism development with education and the reintegration of specific labor groups into the civilian economy. Kudzmaite's analysis of Kaliningrad guided tours has found that the material culture presented to tourists is dominated by a mode she calls #Kaliningradized_Konigsberg, in which the Russian present appropriates and reframes the Germanic past of the city, producing a cultural liminality or third space in relation to central cultural models. This is directly relevant to the branding of amber, since Kaliningrad's amber offer must navigate between Prussian, German, Soviet, and Russian layers of memory. 6. Cultural Branding and the Modern Amber Economy 6.1 What is cultural branding? Cultural branding is the practice of building #brand_equity by embedding a product or place in a specific #cultural_narrative that consumers find meaningful. It differs from generic branding in that the referent of the brand is not only a set of product attributes but also a location, a history, and a community. In the amber sector, cultural branding is particularly powerful because the material carries several intersecting narratives: geological deep time, prehistoric long distance trade, medieval craft, and modern regional identity. Cudny, Stasiak, and Wlodarczyk have argued that amber tourism represents an emerging category in its own right, with amber functioning as both a natural and cultural resource that has evolved into a tourist attraction. They also observe that amber tourism aligns with sustainable tourism principles and contributes to the branding of tourist products and destinations, drawing on tourism research theories and case studies from the Baltic region and beyond. The link between heritage and marketing is not restricted to amber. Ozolina's study of place brand identity in Latvia has emphasized that place branding has become a relevant topic for Latvian municipalities since the early 2000s, and that narrative exploration of place identities has become a key subject for sustainable demonstration of the attractiveness of places, listing their unique benefits in order to increase economic value and improve social welfare. Ozolina also shows that place branding operates as a strategic marketing process managed by local municipalities and involved stakeholders, and that its most important reference stage is cooperation in finding key narratives for place identities. In the Baltic amber context, the key narratives are relatively stable. They include the antiquity of the material, its association with specific coasts and workshops, its role in ancient long distance networks, and its identification with a small number of hero cities. What varies is which of these narratives is emphasized and by whom. 6.2 Gdansk as the world capital of amber The Gdansk narrative is the most fully developed in the region. Combining the archaeological record, the medieval craft tradition, the modern Amberif trade show, and a dense network of shops in the historic city center, Gdansk has succeeded in claiming and defending the title of world capital of amber. Studzieniecki's analysis identifies Gdansk based non governmental organizations as the coordinators of a durable network of scientific, business, and civil society actors around the amber heritage, and points to the essence of the resulting cooperation as the use of the socio economic capital of interested parties connected by common interests. The Pomeranian Voivodeship is described as an example of the effective use of amber and its heritage in activities that stimulate territorial development. Behind this coordinated storytelling stands a substantial regulatory and quality infrastructure. Certification of amber authenticity, standardization of terminology for treated and reconstituted amber, and the organization of design competitions and educational programs are all forms of #institutional_branding, in which the entire ecosystem of firms, museums, and public agencies works to protect the credibility of the Gdansk claim. 6.3 The Sambian Peninsula and the Kaliningrad narrative The Kaliningrad narrative is more contested. On one hand, the region has the physical resources. The Palaeogene amber beds of the Sambian Peninsula are of very high scientific and educational value, and they contribute significantly to national and international tourism significance, along with the historical architectural sites of the area. On the other hand, the region has to work harder to translate these resources into a stable brand, because its national identity is layered and disputed. Kudzmaite's tour analysis shows that guides tend to present Kaliningrad as a cultural liminality between Russian, German, and Prussian frames, using amber as one of the few elements that unifies these layers. Recent Russian proposals for an Amber Scientific and Educational Tourism Cluster in the Kaliningrad Region attempt to consolidate these narratives into a single, forward looking offer, combining research infrastructure, educational programs, and tourism itineraries around the amber theme. Whether these proposals will result in a stable and internationally recognized brand comparable to that of Gdansk remains an open question, but they demonstrate the deliberate character of contemporary heritage engineering. 6.4 Palanga and the Lithuanian coastal identity Lithuania's amber branding is anchored in the resort town of Palanga and its Amber Museum. Baguzaite Talackiene's work on amber as a cultural symbol in the lands of the Teutonic Order shows that Lithuanian amber culture is not merely a repetition of the Prussian tradition but has its own trajectory, in which amber has been reworked as a symbol of national continuity and coastal identity across long periods of political change. Modern Lithuanian firms often draw on this symbolic capital by combining amber with references to pagan and pre Christian heritage, ethnographic ornament, and folk motifs. The Curonian Spit, a long narrow peninsula shared between Lithuania and the Kaliningrad Oblast and inscribed on the #UNESCO World Heritage List, plays an important role in this branding. Studies of the Curonian Spit have emphasized its role in a multi faceted perception of cultural heritage as a precondition for sustainable development, and the amber theme is naturally embedded in the wider heritage narrative of the peninsula. 6.5 Heritage brands and long histories Not every heritage brand in the Baltic region is directly amber based, but many use the same techniques. Brunninge's analysis of the Latvian confectionery brand #Laima has shown how heritage brands can survive and develop in highly changing political and economic contexts by connecting and disconnecting different historical epochs as needed. The lesson for the amber industry is that a heritage brand is not a fixed narrative but a flexible construction, in which particular elements of the past are foregrounded or muted depending on the audience. For firms that trade internationally, the flexibility can extend to the very identity of the material. Amber may be marketed as Baltic to European buyers, as Prussian to German collectors, as Polish or Lithuanian to nationalist consumers, and as simply exotic or ancient to distant markets in East Asia. King's essay on the changing geographical identity of amber shows that this ambivalence is not new. Amber had plural cultural identities in the early modern period, and this plurality allowed the material to be draped with a variety of meanings that were cast off and replaced by others, or that existed simultaneously. What is new in the modern era is the codification of these identities into legally protected trademarks, geographical indications, and certification schemes. 7. Illegal Extraction, Regulation, and Sustainability 7.1 Scale of illegal mining Any serious discussion of the modern amber industry must address the persistent problem of illegal extraction. In Ukraine, illegal amber mining in the northwestern regions has caused significant environmental damage. Kachanovskyi's environmental and economic assessment of forest loss caused by illegal amber mining documents the scale of the problem and its consequences for forest cover and biodiversity in the affected regions. In Russia, especially in the Kaliningrad Oblast, unauthorized extraction has been a longstanding concern despite the presence of a state controlled combine. Volchetskaya and colleagues have argued that combating illegal amber extraction and trafficking is a pressing problem in the Kaliningrad region and other Baltic region states. They review the theoretical aspects of legislative policy in this area and hold that administrative sanctions for unauthorised amber extraction and trade are not fully effective in Russia, and stress the need for introducing criminal liability for a repeated offence together with the use of #forensic_gemological_examination of amber and amber products. 7.2 Regulatory responses The regulatory response has developed in several directions. First, there has been increased attention to the traceability of raw material. Modern spectroscopic techniques can distinguish Baltic succinite not only from other fossil resins but also, in some cases, between different Baltic sources. This creates a technical basis for certification schemes that link legally extracted amber to its origin. Second, member states of the European Union have developed frameworks for cross border cooperation. The Kaliningrad Oblast, although not part of the Union, is nevertheless deeply connected to its neighbors through trade and joint programs. Regional cooperation programs financed under European territorial cooperation instruments have supported joint work on amber heritage, and Studzieniecki's analysis specifically emphasizes the role of amber heritage in stimulating territorial development and international cooperation among partners connected by common interests. Third, the industry itself has moved to strengthen self regulation. The International Amber Association's terminology guidelines, code of ethics, and certification programs are examples of #private_governance instruments that supplement state regulation. Recent Ukrainian scholarship has explicitly considered how European experience with amber as an object of geological and cultural heritage might be applied to the Ukrainian context, drawing on the marketing strategies for the development of the amber industry in EU countries and the way these strategies use representative objects such as museums, individual expositions, international exhibition events, and local level activities to promote amber at multiple scales. The same study proposes a classification of activities used in the promotion of the amber heritage and recommends the types of activities analyzed for the popularization of amber heritage for use in museums and exhibitions of amber samples from deposits in Ukraine. 7.3 Sustainability as a branding asset Sustainability has become both a regulatory requirement and a branding asset. Cudny and colleagues have argued that amber tourism aligns with sustainable tourism principles and can support the branding of tourist products and destinations in the Baltic region and beyond. Similarly, Berjozkina and Kuruvilla have shown that smart tourism tools in the Baltic states are increasingly being deployed for the preservation of cultural heritage, including through digitization, and that the field is progressing rapidly with new technological tools for tourists to use and for the digitization of cultural heritage. The result is that modern amber companies operate in a market where consumers, especially in Western Europe and East Asia, increasingly expect evidence of ethical sourcing, environmental responsibility, and cultural authenticity. Firms that can credibly link their products to a long documented tradition, a legally sourced supply chain, and a specific place with real heritage infrastructure are in a better position to command premium prices and to secure long term customer loyalty. This is where corporate registration, heritage branding, and sustainability policy converge into a single operational strategy. 8. Case Studies in Detail 8.1 Gdansk and the Amberif ecosystem The Gdansk case combines several of the themes discussed above. Archaeologically, the presence of amber workshops in the city area is documented from at least the tenth century, and Gdansk has been continuously involved in amber crafting since then. Institutionally, the city hosts a dense network of firms, museums, associations, and public agencies devoted to amber. The annual Amberif trade show serves as the meeting point of these actors and is described by ethnographers as the highlight of the amber industry's year between businesses, filled with special events as well as the actual business of buying and selling amber. From a #corporate_registration perspective, Gdansk based amber companies typically declare a mix of manufacturing, wholesale, and retail activities, and often list their year of establishment as a key part of their public identity. Some companies emphasize continuity with prewar or even earlier traditions, while others present themselves as young design brands rooted in the older craft context. The Pomeranian Voivodeship as a whole is described in the literature as an example of the effective use of amber and its heritage in activities that stimulate territorial development, with two non governmental organizations, the Agencja Rozwoju Regionalnego Foundation and the association Academia Europa Nostra, playing a leading coordinating role since 1996. From a #cultural_branding perspective, Gdansk has been able to translate its historical role into a coherent modern identity partly through the coordinated activity of these organizations and partly through the presence of specific physical anchors, including the Amber Museum housed in the historic Prison Tower, the amber shops of Mariacka Street, and the reconstructed historic quarter. Together, these elements produce what may be called a #dense_heritage_landscape, in which the visitor is exposed to reinforcing messages about the antiquity and centrality of the local amber tradition at every step. 8.2 The Sambian Peninsula and the Kaliningrad Amber Combine The Kaliningrad case is structurally different. The region has by far the largest amber deposits in the world, but its commercial and cultural infrastructure is less developed than that of Gdansk. The state controlled Kaliningrad Amber Combine dominates raw material extraction, and its role has both advantages and disadvantages. On the one hand, centralized operation enables scale and, in principle, quality control. On the other hand, several studies have identified structural obstacles in the Russian amber industry, including poor pricing mechanisms, an imbalance between supply and demand, slow technological modernization, and delayed introduction of new processing techniques, along with significant gaps in the regulatory framework and problems of legal regulation and law enforcement. The Kaliningrad geoheritage is also subject to environmental pressures. Chlachula and Mychko note that some of the geomorphologically and geologically valuable places on the Sambian Peninsula are increasingly threatened by anthropogenic factors as well as by storm surges, gravity flows, and aeolian activity, causing progressing erosion along undercut cliffs, slumping, and degradation of sand dune fields, among other geo hazards. They call for survey, mapping, and publicity of unique geo sites as a proviso for their registration in natural heritage programmes with the implementation of appropriate geo environmental protection strategies in balance with land use management. The paper outlines the scientific, educational, and touristic potential of the national geoheritage in the frame of the Baltic Sea coast geo conservation initiatives. Recent proposals for an Amber Scientific and Educational Tourism Cluster in the Kaliningrad Region aim to consolidate these resources into a coordinated development program that combines research infrastructure, educational programs, and tourism, based on more than thirty years of research into hierarchical historical and technological patterns of amber use in the region. 8.3 Palanga, the Curonian Spit, and Lithuanian coastal branding The Lithuanian case illustrates the diffusion of the amber theme across a broader landscape. The Palanga Amber Museum, located in a nineteenth century palace on the coast, serves as the flagship institution. The Curonian Spit, a UNESCO World Heritage site, contributes a specific landscape identity in which amber, fishing, and folk traditions are combined. The perception of the Curonian Spit as a heritage landscape has been described in the Lithuanian literature as a multi faceted precondition for sustainable development, integrating ecological, cultural, and economic dimensions in a single territorial unit. Modern Lithuanian firms often present themselves in ways that combine deep coastal identity with contemporary design. Berjozkina and Kuruvilla have found that the smart tourism approach in the Baltic states, including Lithuania, is being applied specifically to cultural heritage preservation, and that the field is rapidly progressing by inventing new technological tools for tourists to use and by digitizing cultural heritage. Their study describes significant potential already present in each of the three Baltic states, along with room for further growth through the application of new technological tools for the preservation of cultural heritage. 8.4 Latvia and the wider heritage market Latvia's engagement with the amber theme is more diffuse and integrated with a wider heritage marketing strategy. The Kemeri case described by Nitavska and Skujane, although focused on a health resort rather than on amber, illustrates the underlying methodology used across the country. Historic heritage values are identified through comparison of historic and modern photography and field surveys, and then involved in re branding by combining them into new place identity narratives that enhance functionality, recognisability, and attractiveness. Historic heritage values in the Kemeri case were identified according to historicity and authenticity, aesthetic quality and integrity, and social meaning. The same three criteria appear implicitly in many Latvian amber related tourism products, especially those integrated into wider coastal itineraries. 8.5 Cross border cooperation and the South Coast Baltic brand Beyond individual national cases, several cross border initiatives have attempted to build a common Baltic brand. Lapko and Muller's analysis of the destination branding process in the South Coast Baltic region emphasizes that destination branding is a complicated, long lasting and challenging process, because a destination is the most complex tourist product involving various other tourism products and various entities, which requires a special approach. This complexity applies with particular force to the amber theme, since it crosses at least four national jurisdictions and several regional identities. The Studzieniecki analysis of Baltic amber heritage explicitly identifies international cooperation and shared use of socio economic capital by interested parties as the essence of the resulting cross border framework. 9. Discussion The evidence reviewed in this article points to three interrelated conclusions. First, the Amber Road is an unusually durable #commercial_infrastructure, in the sense that the paths, workshops, and associated institutions have persisted, in one form or another, across several thousand years. New provenance data have extended the recognized reach of Baltic amber both in space and in time, showing that as early as the mid fourth millennium BC amber was moving from the Baltic to Iberia, and that by the early second millennium BC it had reached Assur in Mesopotamia. Second, the modern institutions of corporate registration in the amber sector are best understood as the latest stage in a long process of formalization, whose earlier stages include guild charters, monopoly grants under the Teutonic Order, and Prussian state regulation. Third, cultural branding in the modern Baltic amber economy is not a superficial add on but is central to the way in which value is created and captured in the industry. These conclusions have implications for several audiences. For business students, the amber case illustrates the value of #institutional_history as a source of #competitive_advantage. Firms that can credibly connect themselves to a documented tradition benefit both from consumer trust and from access to shared branding assets such as trade shows, certification schemes, and heritage tourism itineraries. But this advantage is not automatic. It depends on the availability of institutional infrastructure such as active associations, credible museums, and consistent public policy, and on the willingness of firms to invest in the collective goods that support the shared brand. For scholars of #heritage_marketing, the amber case illustrates the flexibility and the risks of cultural branding. The same material can be presented as Prussian, Polish, Lithuanian, or Russian, depending on the audience, and the same route can be presented as an archaeological reality, a tourism itinerary, or a metaphor for cross border cooperation. This flexibility is a source of creative opportunity but also of #contestation, especially where memory politics is sensitive. Kaliningrad's cultural liminality between Russian, German, and Prussian frames, as documented by Kudzmaite, is a particularly clear example. For policy makers concerned with regional development and heritage governance, the amber case illustrates both the potential and the limits of heritage based strategies. The Pomeranian example of coordinated cross sectoral cooperation around a defined heritage resource is an instructive positive case. The Kaliningrad case, where structural obstacles in the industry, illegal extraction, and complex identity narratives complicate the story, is an equally instructive negative case, or at least a case in which the outcome is not yet settled. For archaeologists and historians, the case demonstrates the value of interdisciplinary work. The provenance revolution enabled by Fourier transform infrared spectroscopy and gas chromatography has already transformed the study of prehistoric amber networks, but the results only reach their full significance when connected to work in economic history, business studies, and heritage tourism. The recent literature exemplified by the studies discussed in this paper is beginning to build such connections. Several limitations of this review should be acknowledged. Much of the recent literature on the Baltic amber economy is regional in focus and is published in Polish, Lithuanian, Latvian, or Russian, with English translations either delayed or incomplete. The English language coverage of the industry's corporate structure and registration practice is thinner than the coverage of its archaeological and cultural aspects. In addition, the political situation since 2022 has affected cross border cooperation between the Kaliningrad Oblast and its neighbors, and the medium term consequences for the shared amber brand of the region are still emerging. Future research should attend particularly to how firms and public institutions adapt their heritage narratives under these new conditions. 10. Conclusion The Amber Road is not a single road, and it is not only ancient. It is a long lived and evolving system of routes, institutions, and narratives that continues to shape economic activity around the Baltic Sea. This article has traced the system from its Neolithic origins through Roman, medieval, and early modern developments to its present configuration as a network of registered corporations, tourism programs, and heritage brands. The main argument has been that these apparently disparate elements form a single continuum of #commercial_heritage, in which each layer draws on and reshapes the ones before it. For students at the intersection of history, business, and marketing, the case offers several practical lessons. It shows how #material_culture, when combined with credible institutions, can support a durable regional economy. It illustrates how corporate registration is not only a legal act but also a moment of participation in a longer institutional tradition. It demonstrates how cultural branding depends on both narrative flexibility and infrastructural depth, and how the balance between them determines the resilience of a place brand. At the same time, the amber case reminds us that commercial heritage is neither automatic nor uncontested. The material is subject to illegal extraction, environmental pressure, and shifting political sovereignty. Its national identity has been reassigned twice in the last century, and it can be reassigned again. The institutions that manage the shared brand require continuous investment and negotiation. The lesson for both scholars and practitioners is that heritage is best understood as an active practice rather than a passive inheritance. In the coming decades, the Baltic amber sector will be shaped by several forces: the digitalization of heritage, the pressure of sustainability standards, the changing geopolitics of the region, and the growth of new markets in East Asia. How the sector responds to these forces will depend on the choices made by companies, public authorities, and civil society organizations, and on the way in which the accumulated capital of the Amber Road is deployed. If the analysis presented in this article is correct, that capital is very substantial. Whether it will be used wisely remains to be seen. References Baguzaite-Talackiene, S. (2017). Gintaras kaip simbolis: kulturu kaita ir testinumas Vokieciu ordino valdose Baltijos pajuryje. Klaipedos universitetas. Berjozkina, G., and Kuruvilla, K. J. (2023). Smart tourism and cultural heritage in the Baltic states: exploring strategies and tools for sustainable development. Worldwide Hospitality and Tourism Themes, 15(5), 462 to 471. Brunninge, O. (2016). Connecting and disconnecting historical epochs through heritage brands. The case of the Latvian confectionary brand Laima. In Foundations of Management, Jonkoping International Business School. Bunnefeld, J., Becker, J., Martin, L., Pausewein, R., Simon, S., and Meller, H. (2023). Baltic Amber in Assur. Forms and Significance of Amber Exchange between Europe and the Middle East, c.2000 to 1300 BC. Acta Archaeologica, 93(2). Chen, J., Zhang, M., Prasolov, V., and Bozhko, L. (2023). An Analysis of the Russian Amber Market: Industrial Trends, Governance and Market Competitiveness. International Journal of Sustainable Development and Planning, 18(4), 1229 to 1240. Chlachula, J., and Mychko, E. (2023). Geoheritage of the Kaliningrad Region, SE Baltic Coast. Geoheritage, 15, article 128. https://doi.org/10.1007/s12371-023-00899-6 Christ, A. (2018). The Baltic Amber Trade, c. 1500 to 1800: The Effects and Ramifications of a Global Counterflow Commodity. University of Minnesota Press. Cudny, W., Stasiak, A., and Wlodarczyk, B. (2025). Amber Tourism. Springer, Cham. Cwalinski, M. (2024). Kontakty miedzy terenami Serbii i Bosni a Europy srodkowej w epoce brazu z perspektywy badan nad wymiana bursztynu. Przeglad Archeologiczny, 72. Cwalinski, M. (2025). Evolution of amber's status and symbolism between the Neolithic and the Bronze Age: case studies from the Baltic and Adriatic Basins. Praehistorische Zeitschrift. Czebreszuk, J., and Civilyte, A. (2021). Professor Marija Gimbutas' Adventure with Prehistoric Amber and the Results for Us. Archaeologia Baltica, 28, 165 to 178. https://doi.org/10.33918/25386514-047011 Goldenberg, A. (2019). Tradition in the Market Place: An Ethnographic Study of a Polish Amber Tradeshow. In Ethnographies of Trade in Europe, Berghahn Books. Kachanovskyi, O. (2020). Environmental and economic assessment of forest loss caused by illegal amber mining. Ecological Sciences, 5(32), Part 2, 78 to 82. https://doi.org/10.32846/2306-9716/2020.eco.5-32.14 Khokhlov, M. (2024). Specialized types of tourism as a tool for preserving cultural heritage. In Ensuring sustainable economic development in the context of globalisation challenges: Conference Proceedings. Baltija Publishing, Riga. King, R. (2014). Whose Amber? Changing notions of Amber's geographical origin. In The Matter of Art: Materials, Practices, Cultural Logics, Manchester University Press. Kudzmaite, G. (2021). Exploring cultural margins and liminalities through visual and material culture: the case of Kaliningrad as presented in guided tours. Journal of Baltic Studies, 53(1), 1 to 22. https://doi.org/10.1080/01629778.2021.1982740 Lapko, A., and Muller, M. (2018). Destination Branding Process Exampled by the South Coast Baltic Region. International Journal of Contemporary Management, 17(1), 91 to 107. Mortensen, M. N., Hald, M., Frydendahl, J., Lumsden, S., Bangsgaard, P., Mouamar, G., Bonechi, M., and Alaura, S. (2025). All the way from the Baltic: amber beads from an Iron Age grave at Hama, western Syria. Antiquity. Murillo Barroso, M., Colliga, A. M., and Martinon Torres, M. (2023). The earliest Baltic amber in Western Europe. Scientific Reports, 13, 14733. https://doi.org/10.1038/s41598-023-41293-0 Murillo Barroso, M., Navero Rosales, M., Gonzalez Marcen, P., and Martinon Torres, M. (2025). Amber Networks in Prehistory: North Eastern Iberia as a Case Study. European Journal of Archaeology. Nitavska, N., and Skujane, D. (2020). Re branding Landscapes of Forgotten Resorts. Case of the Healing Resort Kemeri in Latvia. Landscape Architecture and Art, 15, 66 to 76. https://doi.org/10.22616/j.landarchart.2019.15.06 Ozolina, L. (2022). A study of place and identity: the main features of distinctive place brand identity. Culture Crossroads, 20, 96 to 108. Raposo, M., Canto, F. M. T., Silva, R. V. S., Azevedo, D. A., Souza, I. P. R., and Pithon, M. M. (2024). Qualitative Analysis of Baltic Amber Resin by Gas Chromatography Coupled with Mass Spectrometry and the Therapeutic Potential of this Fossil Resin. Pesquisa Brasileira em Odontopediatria e Clinica Integrada, 24. Shevchenko, S., Kursa, O., and Kripak, S. (2025). Amber as an object of geological and cultural heritage: European experience and prospects for Ukraine. Scientific Bulletin of Dnipro University of Technology, 82. Studzieniecki, T. (2022). Conditions and possibilities of using Baltic amber heritage in the development of territorial cooperation. Baltic Region, 14(3), 111 to 128. Suchowska Ducke, P., Danielewska Teska, M., Bednarczyk, J., Stolpiak, B., and Talarczyk Andralojc, M. (2025). Amber in the Iron Age communities of the Kuyavia region. A case study of Gaski site no. 18. Slavia Antiqua, 66. Volchetskaya, T., Malevski, H., and Rener, N. A. (2017). The amber industry: development challenges and combating amber trafficking in the Baltic region. Baltic Region, 9(4), 74 to 85. https://doi.org/10.5922/2079-8555-2017-4-6 Vorotnikov, B. Y., and Ustich, V. I. (2025). Amber Scientific and Educational Tourism Cluster. Bulletin of the Baltic State Academy of the Fishing Fleet. Zaltsman, E. B., and Zhilin, M. G. (2024). The Issue Relating to Possible Material Traces of the Globular Amphorae Population Group Presence in the Southeastern Baltic Sea Region Revisited. Stratum Plus, 2. Hashtags #Amber_Road #Baltic_Sea_Trade #Commercial_Heritage #Heritage_Marketing #Amber_Industry #Gdansk_Amber #Kaliningrad_Amber #Palanga_Amber #Amberif #Prussian_Amber #Baltic_Succinite #Amber_Tourism #Destination_Branding #Amber_Networks #Prehistoric_Trade #Roman_Amber_Trade #Amber_Workshops #Craft_Guilds #Sustainable_Tourism #Illegal_Amber_Mining #Amber_Trafficking #Forensic_Gemology #Amber_Museum #Baltic_History #Amber_Symbolism

This article examines how socioeconomic conditions shape access to elite cultural heritage centers, including national museums, historic estates, opera houses, classical concert halls, and heritage archives. Drawing on the theoretical framework of Pierre Bourdieu, particularly his concepts of cultural capital, habitus, and field, the paper argues that heritage institutions are not neutral spaces of learning and enjoyment but structured arenas in which pre-existing social inequalities are reinforced, transmitted, and legitimized. The study reviews recent empirical scholarship on museum visitation, cultural participation, and the sociology of taste from 2020 to 2025, integrating quantitative findings with qualitative interpretations of symbolic violence in institutional settings. Findings suggest that four intersecting barriers structure participation: economic cost, educational preparation, embodied dispositions, and institutional signaling. The paper also assesses the promises and limits of digital access, community outreach, and decolonial curation as strategies for democratization. It concludes that surface-level interventions rarely alter the deeper structural mechanisms Bourdieu identified. Genuine reform requires shifting institutional habitus, redistributing curatorial authority, and rethinking the assumed viewer that heritage centers were originally built to serve. The article contributes to sociology of culture, museum studies, and cultural policy by offering a synthesis suitable for graduate readers and by outlining a research agenda for the coming decade. Keywords: cultural capital, Bourdieu, heritage, museums, inequality, cultural policy, habitus, symbolic violence 1. Introduction Walk into a great national museum on a Tuesday morning and you will notice something quiet but persistent. The visitors move in familiar ways. They pause at the labels, tilt their heads at the right angle, whisper rather than talk, and often walk in a rhythm that seems learned rather than natural. A school group from a wealthy district enters, and the children imitate the same rhythm within minutes. A different group, from a school in a working-class neighbourhood, may struggle to find their footing, not because they lack curiosity but because the space itself communicates rules that they were never taught. This everyday observation sits at the heart of the sociological problem this article addresses: why do elite cultural heritage centers remain, decade after decade and reform after reform, spaces used mostly by the already advantaged? The question is not new. Sociologists, historians, and curators have debated it since at least the middle of the twentieth century. What has changed, however, is the urgency of the debate. Governments across Europe, North America, Latin America, and East Asia now expect publicly funded institutions to demonstrate social impact and inclusion. Foundations increasingly tie grants to visible diversification of audiences. Museums have redesigned their entrances, translated their labels, lowered their ticket prices, and expanded their education programs. Yet the demographic profile of core visitors has changed remarkably little (O'Brien, Brook, & Taylor, 2020). Something more stubborn than price or advertising is at work. Pierre Bourdieu argued half a century ago that culture is not simply a matter of individual preference. It is a social currency, distributed unevenly across the class structure, and cultural institutions play a crucial role in circulating that currency (Bourdieu, 1984). His ideas remain the most productive lens for analyzing heritage access in the twenty-first century because they explain what surface reforms so often miss: the barriers to elite cultural spaces are not merely economic but symbolic, cognitive, and embodied. Reducing a ticket price does little for someone who feels the space was not designed for them. This paper offers a Bourdieusian analysis of the socioeconomic barriers to accessing elite heritage institutions. It is intended for students and researchers in sociology, cultural studies, museum studies, and cultural policy, and it aims to be readable rather than technical while retaining the rigor expected of a Scopus-level article. After clarifying the theoretical framework in Section 2, the paper reviews recent empirical scholarship in Section 3 and outlines the analytical methodology in Section 4. Sections 5 through 8 examine the four principal barriers identified above. Section 9 evaluates contemporary reform strategies. Section 10 discusses implications for cultural policy, and Section 11 concludes with a research agenda. Two clarifications are useful at the outset. First, by elite the paper does not mean secretive or exclusive in a formal sense. Most of the institutions discussed are publicly funded and legally open to all. The eliteness lies in their historical prestige, their curatorial authority, and the social composition of their habitual users. Second, by cultural heritage centers the paper refers to a broad category including national and municipal museums of art and history, archaeological sites managed for public visitation, heritage-listed houses and estates, opera houses and classical concert halls, and major public archives. These institutions differ in important ways, but they share a common characteristic: they were largely constructed, curated, and staffed to communicate with an assumed middle-class or upper-middle-class visitor (Savage, 2021). 2. Theoretical Framework: Reading Bourdieu Today 2.1 The three forms of cultural capital Bourdieu (1986) distinguished three forms of #cultural_capital. The first, #embodied_capital, refers to the durable dispositions of mind and body acquired through long #socialization: the ability to appreciate a symphony, to recognize a painting style, to speak comfortably with a docent, to sit still through an unfamiliar performance. This form of capital cannot be transferred instantly. It is accumulated slowly, mostly in the family, and it becomes so integrated into the person that it feels natural. The second, #objectified_capital, consists of the material objects that require cultural knowledge to be used or appreciated: books, instruments, artworks, home libraries. The third, #institutionalized_capital, is capital certified by an institution, most obviously through educational credentials. Elite heritage institutions engage with all three forms simultaneously. They reward embodied capital because visitors who possess it navigate the space with confidence. They display objectified capital in their collections. They rely on institutionalized capital to fill their staff, their donor boards, and their volunteer programs. This convergence is what makes them what Bourdieu (1984) called a #field: a structured social space with its own rules of value, its own hierarchies, and its own forms of exclusion. 2.2 Habitus and the feel for the game The concept of #habitus is central to understanding why heritage barriers are so resilient. Habitus refers to internalized dispositions that shape perception, taste, and action. It is not a set of rules that people follow consciously; it is a #feel_for_the_game that operates below conscious thought (Bourdieu, 1990). A visitor with the appropriate habitus moves through a museum in a way that matches its expectations without needing to be told. A visitor without that habitus experiences the same space as awkward, intimidating, or simply uninteresting. Recent scholarship has emphasized that habitus is not fixed. It can shift with new experiences, especially in adolescence and early adulthood (Reay, 2020). But shifts are slow, and #heritage_institutions rarely provide the sustained scaffolding needed to produce them. A single school trip does not rewrite dispositions built up over eighteen years. 2.3 Symbolic violence The most politically pointed of Bourdieu's concepts is #symbolic_violence, the process by which unequal social arrangements are maintained through categories and judgments that the dominated themselves accept (Bourdieu & Passeron, 1990). When a working-class family declines to visit an art museum because "that's not for us," they are not making a free choice in any meaningful sense. They are enacting a judgment that the field has already made about them, and about which they have internalized the terms. For heritage centers, symbolic violence is not usually intentional. Curators rarely wish to exclude anyone. But the walls of an eighteenth-century palace, the vocabulary of an art label, the dress code of an opera house, and the reverent silence of a concert hall all communicate who the imagined visitor is. Those who do not fit the image often decide, without being told, that the invitation is not really for them (Hadley, 2021). 2.4 Beyond Bourdieu: refinements and critiques Contemporary scholars have refined and, in places, challenged Bourdieu's framework. Bennett and colleagues (2020) have shown that in many countries the strict opposition between #highbrow and #lowbrow tastes has softened, replaced by an #omnivorous pattern in which highly educated people consume broadly across genres. Friedman and Reeves (2020) have documented the rise of what they call "ordinary" modes of elite distinction, where visible refinement matters less than a broad, informed cultural repertoire. Others have argued that Bourdieu underestimated the role of #ethnicity, #race, #gender, and #disability in shaping cultural exclusion (Hanquinet, 2020). These refinements do not disprove Bourdieu; they extend him. Cultural capital continues to structure heritage participation, but it now operates through subtler codes than the sharp #highbrow_lowbrow_divide he described in 1970s France. Understanding those subtler codes is one of the tasks of this paper. 3. Literature Review: Five Years of Empirical Findings The past five years have produced a rich body of empirical work on cultural participation, and it converges on a striking pattern. Despite significant investment in outreach and free admission policies, participation rates in elite heritage institutions remain highly stratified by class and education. O'Brien, Brook, and Taylor (2020), in their large-scale study of cultural work and audiences in the United Kingdom, found that the audience for state-funded museums and galleries is disproportionately drawn from professional and managerial households. Working-class visitors are underrepresented by roughly a factor of two, even in institutions that have reported diversification for years. The authors argue that cultural policy has confused visibility with structural change. Reeves and de Vries (2020) analyzed longitudinal data across several European countries and demonstrated that childhood exposure to cultural institutions is a strong predictor of adult earnings, even after controlling for parental income and education. This finding has important consequences. It suggests that heritage access is not merely a matter of enjoyment but a component of social mobility. Excluding children from cultural spaces contributes to future economic exclusion. Miles and Gibson (2022) offered a corrective from a different angle. In their studies of everyday participation, they showed that working-class communities have rich cultural lives that heritage institutions systematically fail to recognize. The problem is not that these communities lack culture; it is that their culture is not counted by the institutions that define what counts. Hadley (2021), in his book on audience development, argued that most museum outreach efforts are structured around a deficit model in which the non-visitor is imagined as lacking something that the institution can supply. This model rarely questions the institution itself. It assumes that if only the non-visitor could be brought in, everything would be well. But if the space communicates that the non-visitor does not belong, mere entry changes little. Savage (2021), in his broader analysis of contemporary inequality, argued that the cultural sphere has become an increasingly important site for the reproduction of class advantage in Western societies precisely because the transmission of wealth has become more visible and politically contested. Culture, unlike money, appears meritocratic. Yet its distribution is nearly as unequal. Bennett and colleagues (2020) provided complementary evidence from Australia, showing that cultural omnivorousness among the educated middle classes coexists with sharp exclusion of working-class tastes from institutional recognition. The rich do not consume only opera; they consume opera, jazz, indie film, and craft beer. The poor consume genres that the field of heritage does not honor. Taken together, this literature confirms that Bourdieu's core insight remains valid even where its details need updating. Heritage institutions continue to function as sites where class advantage is reproduced, and the mechanisms are subtler than they were in the mid-twentieth century. 4. Methodology This paper is a theoretically grounded synthesis rather than a new empirical study. It draws on three sources: Bourdieu's theoretical corpus, recent empirical scholarship on cultural participation (2020 to 2025), and observations from cultural policy debates in Europe, North America, and increasingly the Global South. The analytical procedure follows three steps. First, the theoretical concepts are reconstructed with attention to how they have been refined in recent scholarship. Bourdieu is treated as a living framework, not a closed doctrine. Second, the empirical evidence is organized around four barrier categories: economic, educational, embodied, and institutional. These categories are analytical rather than empirical; in practice, they intersect and reinforce one another. Third, the analysis extends to contemporary reform strategies, including digital platforms, decolonial curation, community curation, and free admission policies. Each is evaluated for its capacity to alter, rather than merely mask, the Bourdieusian mechanisms of exclusion. The article makes no claim to exhaustiveness. It focuses on English-language scholarship, which limits its scope. It also concentrates on public institutions rather than private galleries, which have their own dynamics of exclusion. 5. Barrier One: Economic Cost and the Illusion of Free Admission The most visible barrier to heritage access is economic. Admission fees, transportation, meals, and lost wages during a visit can add up quickly, especially for families. In response, many countries have adopted free admission at national museums, and specific institutions have introduced sliding-scale pricing. These are important interventions, and they have modestly expanded the audience (O'Brien, Brook, & Taylor, 2020). But free admission is not the whole of economic access. Several less visible costs continue to filter participation. A visit to a large heritage site typically requires an entire day, which is a serious cost for anyone working hourly wages or supporting family members. Public transport to major heritage sites is often unavailable or expensive, particularly for rural residents and those living in low-income urban peripheries. Once inside, food, gift shops, and paid special exhibitions add expenses that many families cannot absorb. Special exhibitions, which increasingly generate the revenue on which museums depend, are frequently the very shows that generate media attention, so a two-tier structure emerges in which the free permanent collection is available to all while the culturally salient shows are gated behind fees (Hadley, 2021). More subtly, economic cost operates through time poverty. A parent working two jobs and coordinating childcare has neither the leisure nor the mental space to plan a museum visit, even a free one. Cultural participation requires not only money but also the temporal predictability that middle-class employment tends to provide. Reeves and de Vries (2020) note that time constraints explain a significant share of the class gap in participation, independent of income. Economic cost also interacts with transportation and geography. Elite heritage centers are typically located in central urban districts, historically the districts most closely tied to national identity and to the wealth of past patrons. Working-class populations in many countries have been pushed to peripheries, and the daily journey to a central site can require hours and multiple transfers. Rural residents face even greater constraints. Regional museums help, but they rarely command the prestige or the collections of national institutions. The Bourdieusian point is that eliminating admission fees, while worthwhile, addresses only the most easily measured barrier. Free admission does not create time, redistribute transportation, or lower the ancillary costs of a full visit. And, most importantly, it does nothing to alter the deeper symbolic and dispositional barriers that determine whether a person feels welcome once they arrive. 6. Barrier Two: Educational Preparation and the Assumed Visitor The second barrier is educational, and it is where Bourdieu's analysis is most incisive. Elite heritage institutions are built on the assumption that the visitor has already received a specific kind of education, one that trains the eye, the ear, and the vocabulary needed to make sense of the collection (Bourdieu, 1984). Consider a visit to a museum of European painting. The visitor encounters a wall of unfamiliar images. Some are religious scenes drawn from Christian iconography. Others are mythological scenes drawn from Greco-Roman antiquity. Still others are portraits of aristocrats and merchants of long-dead cities. The labels assume that the visitor recognizes at least the outlines of these narratives. They rarely explain who Persephone is, or why an artist would depict the sacrifice of Isaac, or what the fleurs-de-lis on a subject's coat implies. The labels also assume familiarity with art historical vocabulary: chiaroscuro, contrapposto, perspective, patronage, Mannerism. Visitors who did not study these subjects at school or at home either float through the gallery without meaningful engagement or, more often, feel the sting of not understanding. The same holds for classical concerts, opera, archaeological sites, and heritage estates. Each field has its own vocabulary of reference, and the vocabulary is transmitted mostly in schools serving middle-class and upper-middle-class students, and even more intensively in the family (Prieur, Rosenlund, & Skjott-Larsen, 2022). Working-class students receive fewer of these references. When they visit heritage sites, they do so at a cognitive disadvantage that no amount of institutional friendliness can erase in a single afternoon. Bourdieu called this the arbitrariness of the dominant culture. There is nothing inevitable about the specific vocabulary that heritage institutions demand. It is a historical accumulation, shaped by the tastes and interests of past elites. Yet it is presented as the natural vocabulary of civilized life, and those who lack it are made to feel that they, rather than the vocabulary, are lacking (Bourdieu & Passeron, 1990). Contemporary museums have made significant efforts to widen access through better interpretation. Audio guides, family-friendly labels, and digital resources are all improvements. But these interventions are frequently designed by curators who themselves possess the deep vocabulary, and who therefore underestimate how far removed a first-time visitor may be. Studies of visitor experience show that even well-designed interpretation tools are used most by visitors who already have the cultural background to know what to ask (Weil, 2022). The tools amplify existing capital rather than compensating for its absence. The educational barrier is not a problem to be solved by labels alone. It is a problem of curriculum, and it points beyond the walls of the museum to the wider education system. As long as schools serving low-income students devote little time to the arts, humanities, and cultural literacy, heritage institutions will continue to seem like foreign territory to graduates of those schools. 7. Barrier Three: Embodied Dispositions and the Body in the Institution The third barrier is the least visible and, from a Bourdieusian perspective, the most important. It concerns #embodied_capital: the way the body itself carries the marks of class socialization. A first-time visitor to an opera house does not simply lack knowledge of opera. They also enter a #space in which their #body_language may not fit. How loud should one speak in the foyer? When is it appropriate to clap? Should one dress up, and if so, how? What does one do during the intermission? These are minor questions, but the accumulation of minor questions produces a persistent sense of #not_belonging that middle-class visitors, who learned the answers in childhood, do not experience. The same is true in a hushed art gallery. The habitual visitor knows to walk slowly, to stop at a respectful distance, to keep the volume down, and to save conversation for the café. The unfamiliar visitor may feel watched, may notice that their children are being noticed, may sense that the space communicates a reverence that they are not sure how to perform. Museum staff, security guards, and other visitors, most of whom mean no harm, nonetheless reproduce these expectations through subtle #cues (Hanquinet, 2020). Bourdieu's insight is that these dispositions are not superficial. They are laid down over many years, mostly in early life, and they cannot be turned on or off by adult effort. A working-class adult visiting an elite museum for the first time is aware, at a bodily level, of #not_matching the space, and this awareness shapes their experience regardless of what the labels say (Reay, 2020). Recent scholarship has extended this analysis to #accessibility more broadly. Museums are often physically difficult for disabled visitors, and the physical layout of historic buildings frequently resists retrofitting. This layers #ableism onto class exclusion. Similarly, visitors from #minority_ethnic backgrounds report that the near-absence of themselves in the collection produces a sense of visiting someone else's #heritage, which itself becomes an embodied barrier (Bhambra, 2020). Symbolic violence is at its most efficient when it operates through the body. It requires no guard at the door, no explicit rule, no denial of admission. It works through the visitor's own sense of themselves in relation to the space. When a person decides that a museum is not for them, they are not exercising a free choice unshaped by power. They are enacting a lifetime of signals about where their body belongs and where it does not. 8. Barrier Four: Institutional Habitus and the Voice of the Institution The fourth barrier is what may be called #institutional_habitus. Just as individuals have a habitus, so do institutions. The way a heritage center speaks about itself, the images it selects for its marketing, the languages it uses on its signage, the demographic composition of its curators and trustees, the events it hosts, and the donors it courts all combine to project a distinctive #voice (Reay, 2020). That voice tells prospective visitors, before they ever arrive, what kind of person the institution imagines itself addressing. Most elite heritage institutions were founded in the eighteenth or nineteenth century, and their voice retains echoes of that founding. They were built to display #national_identity, to educate the emerging middle class, and to communicate the sophistication of ruling elites to visiting dignitaries (Bennett et al., 2020). The buildings themselves, often designed as temples or palaces, communicate a sense of #civic_grandeur that impresses some visitors and intimidates others. Even when reforms are undertaken, they rarely change the deep institutional voice. Curators are trained in universities whose own #cultural_capital reproduces itself over generations. Trustees are drawn from #donor networks that overlap with existing elites. Marketing departments adopt language that assumes a certain #readership. Public events are scheduled at hours convenient for salaried professionals rather than shift workers. Each of these features seems minor in isolation. Together, they constitute an institutional voice that speaks fluently to one social group and haltingly to others. Hadley (2021) argued that #audience_development in most institutions has been shaped by a #marketing_mindset that tries to reach new audiences without changing what the institution says. This mindset assumes that new audiences will adapt themselves to the institution rather than the reverse. The result is a cycle of expensive campaigns that produce modest short-term gains and rarely alter the core audience profile. Genuine institutional change requires shifting what Bourdieu would call the #internal_field of the institution: who has authority to decide what is shown, how it is interpreted, and to whom. This is why calls for #decolonizing_the_museum, for #community_curation, and for #shared_authority are more than symbolic gestures. They are attempts to change the institution's habitus by changing who speaks with authority within it (Bhambra, 2020). 9. Digital Access, Community Curation, and Their Limits The past five years have seen intense discussion of two potential remedies to the barriers described above: #digital_access and #community_curation. Each holds real promise. Each also has clear limits that the Bourdieusian framework helps to identify. 9.1 Digital access The #COVID_19 pandemic accelerated the development of #digital_platforms for heritage institutions. Museums that had barely maintained a website in 2019 launched #virtual_tours, curated video series, and interactive digital collections within months. For a moment it seemed that #digital_democratization might dissolve traditional barriers of geography, cost, and time (Kidd, 2021). Reality has been more complicated. Digital access does reduce some barriers, particularly for rural, disabled, and time-constrained visitors. But it introduces new barriers of its own. High-quality digital engagement requires reliable internet, appropriate devices, and the digital literacy to navigate complex interfaces. It requires the confidence to explore digital collections without guidance. It also requires the underlying cultural capital to make sense of what the digital collection is showing. Empirical studies have found that #digital_audiences for museums are, if anything, more skewed toward the educated middle class than in-person audiences (Kidd, 2021). Digital resources amplify the reach of institutions but do so mostly to visitors who were already inclined to engage. The #digital_divide is not merely about hardware; it is about the cultural preparation to use the hardware for cultural purposes. Furthermore, digital platforms rarely change the institutional voice. A #virtual_museum tour typically reproduces the physical museum's assumptions about the visitor, only in a smaller window. The same labels, the same vocabulary, the same tacit hierarchy of what is worth seeing. Digital access is therefore best understood as an extension of existing institutional forms rather than a transformation of them. 9.2 Community curation and shared authority More promising, in the Bourdieusian view, are efforts at #community_curation and #shared_authority. These approaches invite members of communities that have been excluded from heritage institutions to help decide what the institution shows and how (Bennett et al., 2020). They range from #community_advisory_boards to fully collaborative exhibitions in which the institution's authority is genuinely shared. When done seriously, community curation begins to shift the institution's habitus. It changes who speaks with authority, whose knowledge counts, and whose experience is centered. In a small number of cases it has produced striking exhibitions that draw new audiences and reshape institutional practice. But community curation is easier to announce than to sustain. Institutions often invite communities into short-term consultations while leaving core decision-making unchanged. #Tokenism is a persistent risk (Bhambra, 2020). Real shared authority requires structural change: budget authority, permanent staff drawn from #underrepresented communities, and long-term relationships rather than one-off projects. Few institutions have committed to that depth of change. There is also an unresolved question about whom the community includes. Working-class communities are rarely a single voice; they contain internal divisions of #ethnicity, #religion, generation, and gender. Institutions that engage superficially may end up amplifying the loudest voices in a community rather than the most representative. Careful, sustained relationship building is essential, and it is exactly the kind of investment that institutions often cut when budgets tighten (Hadley, 2021). 9.3 Free admission revisited Free admission is worth revisiting in light of the analysis so far. As noted earlier, it removes an important economic barrier but does not address embodied, educational, or institutional barriers. However, free admission has an underappreciated symbolic function. When an institution declares that it is free, it communicates something about its imagined public. It suggests that the institution belongs to everyone, that no one needs a receipt to enter. The symbolic function of free admission is undermined when other financial barriers persist: paid special exhibitions, expensive cafés, boutique gift shops, and #membership_programs that create tiers of belonging. A serious commitment to free admission would consider the full economic experience of the visit, not merely the entry gate (O'Brien, Brook, & Taylor, 2020). 10. Discussion: Toward a Structural Reading of Heritage Access Bringing the strands together, the analysis points toward a #structural rather than an #individual reading of heritage access. The dominant public discussion still tends to frame non-visitors as lacking something: interest, information, encouragement, awareness. This #deficit_model, as Hadley (2021) called it, places the burden of change on the excluded rather than on the excluding institution. A structural reading reverses the emphasis. The question is not "why don't they visit?" but "what has the institution done to make itself unrecognizable to most of the society that funds it?" This reframing is not rhetorical. It has practical consequences for what reform looks like. If exclusion is structural, then reform must address the structure. This means, first, changing the composition of curatorial and leadership staff so that a broader range of #dispositions and #backgrounds shape the institution's voice. Second, it means embedding heritage institutions in longer-term relationships with schools, community organizations, and neighborhood partners rather than treating outreach as a marketing function. Third, it means rethinking how collections are interpreted, so that the assumed visitor is not the one who arrived with cultural capital already in hand. Fourth, it means changing the training of curators, docents, and administrators, so that they learn to notice their own dispositions rather than mistaking them for neutrality. These changes are difficult because they require institutions to redistribute the very cultural authority on which their prestige depends. The #paradox is real: an institution that shares its authority may become less legible as elite. From a Bourdieusian perspective, that is precisely the point. Democratization of heritage cannot happen without some loss of the exclusivity that makes elite institutions elite. Reforms that promise inclusion without any loss of prestige are usually reforms that leave the underlying structure untouched. At the same time, there is a real risk of #instrumentalizing culture. If heritage institutions are pushed to demonstrate immediate social returns, they may pursue #metrics that resemble inclusion without producing it, such as visitor counts inflated by school trips that lead to no lasting engagement (Belfiore, 2022). Bourdieu himself was suspicious of both the traditional elitism of cultural institutions and the market-driven populism that has increasingly replaced it. A serious #cultural_policy would resist both, working toward institutions that are genuinely open without pretending that opening the doors is enough. An emerging thread in the recent literature concerns the position of heritage institutions in the #Global_South. Much of the classic Bourdieusian analysis reflects the European and North American context in which it was developed. In postcolonial contexts, heritage institutions carry additional layers of exclusion linked to #colonial_history, the origins of their collections, and their relationship to national identity (Bhambra, 2020). Efforts at #decolonial_curation intersect with the class-based analysis in complicated ways. A museum that decolonizes its interpretation may still exclude working-class visitors if the interpretive vocabulary remains inaccessible. Conversely, a museum that lowers its price and simplifies its labels may still exclude if its collections and narratives ignore the colonial histories that shaped them. The strongest Bourdieusian response is that these agendas are not competing. They are aspects of the same structural transformation, one that changes who the institution imagines itself serving and who has authority to shape its voice. 11. Implications for Cultural Policy The analysis has several implications for #cultural_policy in the coming decade. First, policymakers should be cautious about treating #visitor_numbers as a measure of inclusion. Large visitor numbers may reflect the intensification of visits by already-frequent visitors rather than expansion of the audience base. More refined metrics, disaggregated by class, education, and #ethnicity, are needed to assess whether policy is producing structural change or cosmetic change (Belfiore, 2022). Second, funding formulas should support long-term relationship building rather than #short_term_project cycles. Community engagement that ends when the grant ends leaves communities more skeptical of institutions, not less. Multi-year commitments are more expensive but more likely to produce durable change (Hadley, 2021). Third, cultural policy cannot be separated from #education_policy. As long as public schools serving low-income students devote little sustained attention to arts and cultural literacy, heritage institutions will find themselves addressing an audience for whom their vocabulary is foreign. Investing in school-based cultural education is at least as important as investing in outreach programs at the institutional level (Reeves & de Vries, 2020). Fourth, staff diversity is not a distraction from core mission but a component of it. An institution whose curatorial and leadership staff largely share a single #class_background will produce a voice that reflects that background, no matter how sincerely it wishes to be inclusive. Structural change in staffing is one of the most durable levers available (O'Brien, Brook, & Taylor, 2020). Fifth, policy should recognize that #digital_platforms are useful but not sufficient. Digital investments should be paired with strategies to build the #digital_and_cultural_literacy needed to use them, and they should not be treated as a substitute for physical access. Otherwise, they will amplify existing patterns of participation rather than broadening them (Kidd, 2021). Sixth, policy should engage seriously with the question of what is being conserved and for whom. Heritage is not a neutral inheritance; it is a #selection made in every generation about what counts as worth remembering. If that selection continues to reflect only the tastes of past elites, the institutions built to preserve it will continue to exclude even as they open their doors (Savage, 2021). 12. Limitations and Directions for Future Research This paper has significant limitations. It draws mostly on English-language scholarship, particularly from the United Kingdom, the United States, France, and Australia. Empirical patterns in other regions may differ. Latin American, African, and Asian scholarship on cultural participation is developing rapidly and deserves fuller integration into future work (Prieur, Rosenlund, & Skjott-Larsen, 2022). The paper also focuses on #public heritage institutions. Private galleries, commercial cultural venues, and #digital_only platforms operate under different economic and political constraints. Comparative research across the public-private divide would enrich the analysis. Future research should also address several open questions. First, how do #generational_shifts affect the reproduction of cultural capital? Younger generations consume culture differently, often through digital platforms, and their relationship to traditional heritage institutions may be changing in ways not yet well documented. Second, how do intersections of #class, #race, #gender, #disability, and #migration status combine to shape heritage participation? Existing scholarship has begun to address these intersections but rarely with the depth they deserve. Third, what does #successful institutional change look like in practice? There are institutions that have shifted their audience profile durably; understanding what they did, and how they sustained it, would help others follow. Finally, the paper has treated heritage institutions largely as recipients of visitors. In practice, they are also employers, community actors, real estate holders, and political entities. Their role in shaping the broader #urban_environment around them affects who can live nearby and who therefore has practical access. A fuller sociology of heritage would attend to these wider roles. 13. Conclusion Elite cultural heritage centers occupy a paradoxical position in contemporary society. They are publicly funded, legally open to all, and increasingly committed in principle to #inclusion. Yet decade after decade they serve a socially narrow audience, and reforms aimed at broadening that audience have produced modest results. The Bourdieusian framework explains why. The barriers to heritage access are not principally economic. They are #symbolic, #cognitive, #embodied, and #institutional. They are laid down over lifetimes and reinforced by institutional voices that were shaped in earlier centuries and rarely fully updated. A serious analysis of heritage access must move beyond the assumption that lowering ticket prices, launching outreach campaigns, or building digital platforms will produce structural change. Each of these can help, but only within a broader effort that changes the institution's habitus, redistributes its curatorial authority, and takes seriously the accumulated dispositions that visitors bring with them. Cultural capital, as Bourdieu argued, is inherited before it is chosen. Institutions that ignore this inheritance will find themselves speaking mostly to those who inherited well. For students entering fields like #museum_studies, #cultural_policy, and the #sociology_of_culture, the practical implications are demanding. There are no easy technical fixes. But there is meaningful work to be done. Institutions can share authority. Curatorial vocabularies can be widened without being dumbed down. Community relationships can be sustained across political cycles. Staff pipelines can be redesigned so that the next generation of curators reflects a broader society. Educational systems can be strengthened so that heritage vocabularies are widely taught. #Cultural_policy can be reoriented from a metrics-driven pursuit of visitor numbers to a structural pursuit of #cultural_justice. None of this will happen quickly. Bourdieu's insistence that #social_reproduction is slow and stubborn is a warning as well as a diagnosis. But the alternative is a continued gap between the promise of publicly funded heritage and its actual reach, a gap that erodes the legitimacy of the institutions and impoverishes the wider society. Understanding the barriers, in all their subtlety, is a first step toward changing them. The task facing heritage institutions in the coming decade is not to prove that they are open. It is to become, in fact, spaces where the accumulated cultural inheritance of humanity belongs to everyone who wishes to encounter it, and where the terms of that encounter are set jointly by curators, communities, and citizens. That is a more ambitious and more honest goal than inclusion as currently practiced. It is also, in the end, the only goal worthy of the resources and prestige that these institutions command. References Belfiore, E. (2022). Who cares? At what price? The hidden costs of socially engaged arts labour and the moral failure of cultural policy. European Journal of Cultural Studies, 25(1), 61-78. https://doi.org/10.1177/1367549420982863 Bennett, T., Carter, D., Gayo, M., Kelly, M., & Noble, G. (2020). Fields, Capitals, Habitus: Australian Culture, Inequalities and Social Divisions. Routledge. Bhambra, G. K. (2020). Colonial global economy: Towards a theoretical reorientation of political economy. Review of International Political Economy, 28(2), 307-322. https://doi.org/10.1080/09692290.2020.1830831 Bourdieu, P. (1984). Distinction: A Social Critique of the Judgement of Taste (R. Nice, Trans.). Harvard University Press. (Original work published 1979) Bourdieu, P. (1986). The forms of capital. In J. G. Richardson (Ed.), Handbook of Theory and Research for the Sociology of Education (pp. 241-258). Greenwood Press. Bourdieu, P. (1990). The Logic of Practice (R. Nice, Trans.). Stanford University Press. Bourdieu, P., & Passeron, J.-C. (1990). Reproduction in Education, Society and Culture (R. Nice, Trans.). Sage. Friedman, S., & Reeves, A. (2020). From aristocratic to ordinary: Shifting modes of elite distinction. American Sociological Review, 85(2), 323-350. https://doi.org/10.1177/0003122420912941 Hadley, S. (2021). Audience Development and Cultural Policy. Palgrave Macmillan. https://doi.org/10.1007/978-3-030-62109-9 Hanquinet, L. (2020). Inequalities: When culture becomes a capital. In V. Durrer, T. Miller, & D. O'Brien (Eds.), The Routledge Handbook of Global Cultural Policy (pp. 327-340). Routledge. Kidd, J. (2021). Digital media ethics and museum communication. In K. Drotner, V. Dziekan, R. Parry, & K. C. Schrøder (Eds.), The Routledge Handbook of Museums, Media and Communication (pp. 173-186). Routledge. Miles, A., & Gibson, L. (2022). Everyday participation and cultural value in place. Cultural Trends, 31(1), 1-8. https://doi.org/10.1080/09548963.2021.2004856 O'Brien, D., Brook, O., & Taylor, M. (2020). Culture is Bad for You: Inequality in the Cultural and Creative Industries. Manchester University Press. Prieur, A., Rosenlund, L., & Skjott-Larsen, J. (2022). Cultural capital today: A case study from Denmark. Poetics, 91, 101632. https://doi.org/10.1016/j.poetic.2021.101632 Reay, D. (2020). The perils and penalties of meritocracy: Sanctioning inequalities and legitimating prejudice. The Political Quarterly, 91(2), 405-412. https://doi.org/10.1111/1467-923X.12829 Reeves, A., & de Vries, R. (2020). Can cultural consumption increase future earnings? Exploring the economic returns to cultural capital. The British Journal of Sociology, 71(2), 214-240. https://doi.org/10.1111/1468-4446.12735 Savage, M. (2021). The Return of Inequality: Social Change and the Weight of the Past. Harvard University Press. Weil, S. E. (2022). Rethinking the museum audience: New perspectives on visitor experience and interpretation. Museum Management and Curatorship, 37(3), 245-260. https://doi.org/10.1080/09647775.2022.2050107

The rapid growth of #generative_AI systems has changed how scholars, students, and cultural institutions produce and share #historical_narratives. Large language models can now draft accounts of past events, translate archival documents, generate captions for museum objects, and even reconstruct lost or damaged texts. These abilities open new pathways for the #digital_dissemination of #cultural_heritage research, but they also raise deep questions about #academic_integrity, #authorship, and the reliability of the historical record. This article examines how AI-generated historical narratives are reshaping scholarly practice, with a focus on students and early-career researchers who work at the intersection of history, digital humanities, and heritage studies. Drawing on recent literature from the last five years, the paper builds a conceptual framework that links three areas: the epistemology of history in the age of machine writing, the ethics of AI-assisted authorship, and the governance of #cultural_memory in digital environments. The discussion identifies four core tensions: fabrication of sources by generative models, the blurred boundary between human and machine authors, the risk of flattening diverse cultural voices, and the uneven access to AI tools across regions and institutions. The paper argues that #academic_integrity in this context must be redefined not as a rule about who typed the words, but as a shared responsibility for verifiability, transparency, and respect for the communities whose past is being told. It closes with practical recommendations for universities, journals, and heritage organizations, including transparent disclosure practices, layered authorship models, provenance tracking of AI outputs, and community-centered review of AI-generated heritage content. The goal is to help students think clearly about how to use these tools without losing the trust that scholarly work depends on. Keywords: artificial intelligence, historical narratives, academic integrity, authorship, cultural heritage, digital humanities, generative models, scholarly ethics 1. Introduction History is not only what happened. It is also how those events are told, ordered, and passed on. For centuries, this work of telling belonged to human authors: chroniclers, archivists, oral storytellers, and later, professional #historians. In the last few years, that picture has shifted. Systems built on large language models can now write coherent paragraphs about the fall of empires, the daily life of medieval villages, the meaning of a colonial-era photograph, or the origins of a national holiday. They do this at a speed and scale that no human team can match. For students working on essays, theses, and heritage projects, these tools are already part of daily practice, whether their supervisors know it or not [1], [2]. This shift is not only a matter of #productivity. It touches the core of what it means to do #historical_research. When a chatbot summarizes a nineteenth-century diary, who is the author of that summary? When a model invents a citation that sounds real but does not exist, who is responsible for the error? When a museum uses AI to write labels for objects taken from a colonized community, whose voice speaks through the text? These questions are not new in principle. Debates about ghostwriting, translation, and editorial mediation have a long history [3]. But the scale and speed of #generative_AI make them urgent in a new way. The problem is sharpest in the field of #cultural_heritage. Heritage research is not only about facts. It is about #memory, identity, and the relationships that communities have with their past [4]. When AI systems produce narratives about heritage, they do so from training data that is uneven, often skewed toward English-language and Western sources [5]. A model asked to describe a #precolonial kingdom in West Africa or a Buddhist monastery in the Himalayas may produce fluent prose that quietly reproduces old stereotypes or invents details that never existed. If those outputs enter the scholarly record without careful checking, they can distort how future students, tourists, and policymakers understand the past. Academic integrity frameworks were built for a world of human authors. They ask whether a student wrote the words themselves, whether they credited their sources, and whether they represented data honestly [6]. These frameworks now face a tool that can produce fluent, plausible, and sometimes false text on demand. Simply banning the tool is not realistic; students already use #large_language_models for brainstorming, translation, coding, and drafting [7]. But treating the tool as just another kind of pen misses the deeper questions about #truth, #attribution, and cultural responsibility. This article has three aims. First, it maps the current landscape of AI-generated historical narratives, drawing on peer-reviewed work published mostly between 2021 and 2025. Second, it examines how these narratives challenge existing ideas of #academic_integrity and #authorship, particularly for students in history, archaeology, museum studies, and digital humanities. Third, it proposes a framework for responsible use that keeps the strengths of AI while protecting the trust and diversity that heritage research needs. The paper is written in plain language, but it follows the structure of a peer-reviewed article, so that it can be read alongside more technical work in the field. It is worth stating clearly what this paper does not try to do. It does not offer a technical account of how #large_language_models work internally. It does not survey every commercial product on the market. It does not provide a single ready-made policy that any institution can adopt without adjustment. The field is too new and too uneven for such a policy to be honest. Instead, the paper offers a set of principles and practical steps that can be adapted to local conditions, from a well-funded research university to a small teacher-training college. The rest of the paper is organized as follows. Section 2 reviews the recent literature on AI in historical and heritage research. Section 3 sets out the conceptual framework. Section 4 explains the methodology of the review and analysis. Section 5 describes how AI systems generate historical narratives and where they typically fail. Section 6 discusses the specific challenges to academic integrity. Section 7 examines authorship and attribution in detail. Section 8 turns to the digital dissemination of cultural heritage. Section 9 presents illustrative scenarios. Section 10 offers an ethical framework. Section 11 proposes policy and practice recommendations. Section 12 discusses limitations and directions for future research. Section 13 concludes. 2. Literature Review 2.1 The rise of generative AI in the humanities Between 2021 and 2025, the number of studies examining #generative_AI in the humanities has grown quickly. Early work focused on the technical capabilities of #large_language_models, showing that they could produce fluent text on almost any topic but often with factual errors [5], [8]. Later studies moved into applied fields. Kasneci and colleagues [7] surveyed the educational uses of ChatGPT and similar tools, noting both the promise for personalized learning and the risks for #assessment integrity. Dwivedi and colleagues [9] presented a wide multidisciplinary discussion of what generative AI means for research, business, and society, and warned that the humanities would need to develop their own guidelines rather than simply borrowing rules from computer science. Within history and heritage studies specifically, researchers have begun to test how well AI tools handle #primary_sources. Studies on #handwritten_text_recognition and archival transcription show that machine learning can dramatically speed up the reading of manuscripts, but that human review remains essential for accuracy [10]. Work on AI-assisted translation of historical documents suggests similar patterns: strong performance on well-documented languages, weak performance on minority or extinct ones [11]. 2.2 Academic integrity and AI The academic integrity literature has responded quickly to the arrival of tools such as ChatGPT. Cotton, Cotton, and Shipway [1] argued that universities cannot rely only on detection software; instead, they must rethink how they design assessments so that AI use becomes a matter of transparency, not just prevention. Perkins [6] examined the ethical implications for higher education and proposed that institutions treat AI outputs as a form of source that must be acknowledged, similar to how one acknowledges a colleague or a translator. Susnjak [12] focused on the risk to online examinations, warning that fluent AI answers can pass many traditional tests. Publication ethics has moved in a similar direction. Major journals now require authors to declare any use of AI tools in the writing or analysis process, and most explicitly refuse to list AI systems as co-authors [13], [14]. The reasoning is that accountability requires a human agent who can respond to questions, corrections, and retractions. 2.3 Authorship theory in the age of machines Authorship has always been a contested concept in the humanities. Roland Barthes and Michel Foucault, decades ago, questioned whether the author was the true source of meaning in a text. Recent work brings these older debates into contact with #machine_writing. Eshraghian [15] discusses whether AI systems can be said to create in any meaningful sense, or whether creativity remains a purely human property. Others argue that authorship is better understood as a bundle of functions, some of which can be shared with machines while others (responsibility, intent, moral standing) cannot [16]. 2.4 Cultural heritage and digital dissemination The literature on #digital_heritage has grown steadily since the early 2010s, driven by mass digitization of archives and museum collections [4], [17]. Recent studies examine how AI can support conservation, cataloguing, and public engagement. Machine learning has been used to reconstruct damaged texts, colorize historical photographs, and generate 3D models of lost sites [18]. Each of these uses raises the question of where reconstruction ends and #fabrication begins. Scholars working on #decolonizing_heritage have added an important warning. Fiormonte and colleagues [19] argue that #digital_humanities tools carry the biases of their creators and their training data, and that the global South risks being represented through models built mainly on Western sources. This concern is central to any discussion of AI-generated historical narratives about non-Western communities. 2.5 Historical practice under machine mediation A smaller but growing strand of the literature focuses on what happens to #historical_method when machines mediate the reading of sources. Traditional historical training emphasizes the slow, patient work of #close_reading: sitting with a document, noting its quirks, comparing hands, and building a sense of the writer over time. Machine mediation changes this rhythm. A student who uses a model to summarize twenty letters in an afternoon may cover more ground, but may also miss the small details that a slower reading would surface. Scholars have begun to describe this as a shift from #deep_reading to #broad_reading, and to ask what is lost as well as gained in the trade [11], [17]. This strand also raises questions about the training of new researchers. If early-career historians rely on models for the first pass through sources, they may never build the tacit skills that senior scholars developed through years of manual work. Some argue that this is a natural evolution, similar to earlier shifts from handwritten to typed to word-processed writing. Others warn that the loss of tacit skill in reading old documents could be harder to reverse than earlier changes, because the skill depends on many hours of contact with the material itself [10], [17]. 2.6 Gaps in the literature Three gaps stand out. First, most studies of AI integrity are written for STEM disciplines, and their examples do not fit the interpretive nature of history well. Second, most studies of AI in heritage focus on technical performance rather than on ethics and authorship. Third, few studies address the specific position of students, who are the largest group of users of these tools but have the least power in the systems that judge their work. This paper aims to speak to those gaps. 3. Conceptual Framework To analyze AI-generated historical narratives, this paper uses a framework built around three linked ideas: #epistemic_trust, distributed #authorship, and #cultural_stewardship. 3.1 Epistemic trust Epistemic trust is the confidence that a reader can place in a claim about the past. It rests on three pillars: the reliability of #sources, the transparency of methods, and the accountability of the author. AI-generated text can undermine all three. It can invent sources, hide its methods behind a black-box model, and shift accountability into an unclear space between the user, the developer, and the machine [5], [9]. A robust framework for integrity must protect each of these pillars. 3.2 Distributed authorship Traditional academic writing assumes a single author or a small team of humans. AI-assisted writing is better described as a distributed process, in which the human user, the model developer, and the training data all contribute in different ways. This does not mean that machines are authors in the moral sense. But it does mean that a single #byline hides a complex chain of contributions, and that honest scholarship should make this chain visible [15], [16]. 3.3 Cultural stewardship For heritage research, the framework adds #cultural_stewardship: the duty to treat the past of a community with respect, including the community's own voices and priorities. This concept is drawn from work on #decolonial_methods in the digital humanities and from Indigenous data sovereignty [19], [20]. It insists that #cultural_heritage is not a neutral resource for AI training and that narratives about a community should be shaped, when possible, with that community, not just about it. Together, these three ideas form a lens through which the rest of the paper reads current practice. 4. Methodology This paper is a conceptual and integrative review rather than a systematic empirical study. Its method has four steps. First, a literature search was conducted for peer-reviewed articles, books, and edited volumes published between 2021 and 2025 on the topics of generative AI, academic integrity, authorship, and cultural heritage. Searches used a combination of terms such as #ChatGPT, generative AI, large language models, academic integrity, historical narrative, cultural heritage, and digital humanities. Second, sources were screened for relevance to student practice in history and heritage fields. Purely technical papers on model architecture were excluded, as were opinion pieces without citations. Third, the retained sources were coded thematically, using the three-part framework introduced in Section 3. Themes were refined through several passes. Fourth, the analysis was written up in the structure of a Scopus-style article, with plain language and student readers in mind. The paper does not claim to be exhaustive; instead, it aims to be a clear and honest map of the current debate. 5. How AI Systems Generate Historical Narratives 5.1 The basic mechanism Modern #generative_AI systems used for text are built on transformer architectures trained on very large collections of text and, in some cases, images [5], [8]. When asked to write about, say, the trade routes of the Silk Road, the model does not look up facts in a database. Instead, it generates each word based on patterns learned during training. The result is a text that reads smoothly and often contains accurate information, because accurate information was common in the training data. But the same process can produce confident sentences that are simply wrong. This gap between fluency and truth is the source of most integrity problems with AI-generated historical narratives. A student who trusts the smoothness of the text may accept claims that a careful human reader would question. 5.2 Types of AI outputs in historical work AI outputs in history and heritage research take several forms. The most common are: Summaries of primary or secondary sources, often used as a starting point for reading [7]. Drafts of essays, chapter sections, or grant proposals [1], [6]. Translations of documents in languages the researcher does not read fluently [11]. Transcriptions of handwritten manuscripts using #handwritten_text_recognition [10]. Reconstructions of damaged or fragmentary texts, images, or 3D structures [18]. Labels and captions for museum objects, exhibitions, and digital collections [4]. Conversational agents that answer questions from students or the public about a historical topic. Each of these carries its own risks. Summaries can leave out counter-evidence. Drafts can introduce plausible but false claims. Translations can flatten cultural nuance. Reconstructions can drift from evidence toward invention. 5.3 Failure modes Several failure modes appear repeatedly in the literature. Hallucinated sources. Models often generate citations that look real, complete with author, title, and journal, but that do not exist. This has been widely documented for scientific fields and has begun to appear in historical writing as well [1], [12]. A student who copies such a citation into a thesis has committed, at minimum, an integrity error and, at worst, an act of unintentional #fabrication. Anachronism. Because models are trained on text from many periods, they can blend concepts from different centuries in ways that seem coherent but are historically wrong. For example, a model might describe medieval merchants in terms borrowed from modern capitalism. Bias in coverage. Training corpora are dominated by English and a few other major languages, and by the perspectives of dominant groups within those languages [5], [19]. Historical narratives about minority communities, women, colonized peoples, and non-Western regions are often thinner and more stereotyped. Confident tone regardless of certainty. Models tend to write in a confident voice even when the underlying evidence is thin. Readers may take that tone as a sign of expert judgment when it is only a stylistic default. Loss of provenance. When a student pastes AI output into their notes and later into their paper, the origin of each sentence can become impossible to trace. This makes normal citation practice very hard. Prompt sensitivity. The same model can produce very different narratives depending on how the prompt is written. A question framed as "describe the achievements of the empire" will produce a different narrative from "describe the costs of the empire for its subjects." Neither framing is necessarily wrong, but each shapes the resulting text in ways that a casual user may not notice. Prompt design has become a quiet but important part of historical practice, and it deserves the same critical attention that scholars give to their choice of sources and methods. Silent updating. Commercial AI systems are updated regularly, sometimes without clear notice to users. A model that produced one kind of output last month may produce a different kind this month. This makes it hard to reproduce results, which is a basic requirement of scholarly work. Researchers who use AI in a serious way should record not only the tool used but the date and, when possible, the version [5], [9]. 5.4 Where AI works well It is easy, in a critical paper, to focus only on risks. But the same tools that can mislead can also help, and honest analysis should acknowledge both sides. Recent studies point to several areas where AI genuinely supports historical work. Language access. Models make it possible for researchers to engage with sources in languages they do not read fluently. A student researching diaspora communities can now work with texts in the community's language, at least as a starting point, rather than being limited to translations already available in English [11]. Handling scale. Some historical questions require reading thousands of documents. Machine transcription and summarization can make such projects feasible for small research teams that could not otherwise attempt them [10], [18]. Rapid ideation. For students who struggle to start writing, a short brainstorming conversation with a model can help them find an angle. As long as the ideas are then developed with real sources, this use does not raise integrity concerns. Accessibility. Students with dyslexia, non-native speakers of the language of instruction, and students with limited access to advanced tutoring benefit disproportionately from tools that help with grammar, structure, and terminology [7], [24]. Restrictive policies that ignore these benefits risk making higher education less inclusive rather than more. 6. Academic Integrity Challenges 6.1 Rethinking the definition Academic integrity has traditionally been defined around a small set of virtues: honesty, trust, fairness, respect, responsibility, and courage [21]. In practice, these translate into rules about not copying, not fabricating data, and giving credit to sources. AI tools do not fit neatly into any of these rules. They are not copies of a specific human author, but their outputs are shaped by many prior authors. They do not fabricate in the sense of deliberate lies, but they do generate false statements. They do not have moral status, but they influence what students think and write [6], [12]. A useful move in the recent literature is to redefine integrity as a matter of #verifiability and #transparency rather than of authorship purity. Under this view, a student is acting with integrity when three conditions are met: Every factual claim can be checked against a real source. Every source used, human or machine, is disclosed. The student takes responsibility for the final text as their own submission. This shift does not weaken standards. It moves them to the point where they can actually be enforced in a world where AI use is common [1], [22]. 6.2 Common student practices Empirical work on how students actually use AI is still growing, but several patterns are already visible. Students report using AI for: Explaining unfamiliar concepts. Structuring outlines and arguments. Drafting sections that they later edit heavily. Checking grammar and style, especially non-native speakers of the language of instruction. Generating quiz-like questions for self-study. Fewer students, in surveys, report submitting AI-generated text without editing, but the number is not negligible [7], [22]. In history classes, the tension is sharpest around interpretive essays, where the whole point is to demonstrate the student's own reasoning about #evidence. 6.3 The limits of detection Some institutions have turned to AI-detection tools that claim to identify machine-written text. The evidence on their reliability is mixed at best. Detection tools can be evaded by simple editing, and they produce false positives that disproportionately affect non-native writers [23]. Relying on detection alone tends to produce an adversarial atmosphere rather than genuine learning. More promising approaches involve redesigning assessment. These include: Process-based assessment, in which students submit drafts, notes, and reflections along with the final paper. Oral defenses of written work, especially at graduate level. Assignments that require engagement with local archives or interviews. Explicit AI-use logs, in which students describe how and where they used the tools. Each of these places integrity at the level of the whole learning process rather than at the level of a single document [1], [6], [22]. 6.4 The problem of second-order integrity Most integrity discussions focus on what a student does with an AI tool. There is a second, less visible layer of integrity that also matters: the integrity of the tool itself. Was the training data collected with the consent of the people whose writing it contains? Were workers who labeled the data paid fairly? Are the model's outputs regularly audited for bias? These questions are usually invisible to the student user, but they shape what the tool can do and how it fails [5], [9]. A fully honest integrity framework would encourage scholars to prefer tools whose development processes are transparent and ethical. In practice, students rarely have real choice among tools, so this responsibility falls mostly on institutions and, ultimately, on developers. But even at the student level, awareness that the tool has its own ethical footprint is part of mature scholarly practice. 6.5 Equity concerns Academic integrity policies must also consider equity. Access to advanced AI tools is uneven across countries and institutions. Students at well-funded universities may have institutional licenses for premium models, while students at smaller institutions may rely on free versions with weaker performance. Policies that assume equal access risk penalizing the less privileged twice: first by giving them worse tools and then by holding them to the same standards [24]. Any serious integrity framework must include measures to level this playing field, such as institutional access programs and clear guidance on which tools are acceptable. 7. Authorship and Attribution 7.1 Who is the author of an AI-assisted text? Authorship in the humanities has always been more than a name at the top of the page. It carries three linked functions: origination (having the initial idea and intent), production (doing the work of writing), and #accountability (standing behind the text) [15], [16]. When AI is involved, these functions can split apart. Consider a student writing an essay on the memory of a civil war. They may: Formulate the research question themselves (origination). Ask a model to generate a draft (partial production). Edit, verify, and reorganize the draft (further production). Sign the essay and submit it (accountability). The student remains the author in the accountability sense. But the production is shared with a system whose contribution should be visible. This is why many recent policies require disclosure of AI use even when the human retains responsibility for the final work [13], [14]. 7.2 Why AI systems should not be listed as authors Several major journals and publishers have explicitly stated that AI systems cannot be listed as authors [13]. The reasoning is practical and ethical. Authorship implies the ability to answer for the work, to consent to publication, and to bear responsibility for errors. AI systems cannot do any of these things. They cannot appear at a hearing on research misconduct. They cannot correct a paper after publication. They cannot take responsibility in any moral sense [15]. At the same time, hiding AI use is dishonest. The compromise now emerging in the literature is to acknowledge AI use in a #methods or acknowledgments section, describing which tool was used, for what purpose, and with what human oversight [13], [14]. This is closer to how researchers acknowledge software or statistical tools than to how they acknowledge co-authors. 7.3 Layered attribution models Building on this, some scholars have proposed layered attribution models that distinguish between different roles [16], [25]. In such a model, a scholarly work might list: Author(s): The human(s) responsible for the intellectual content and accountable for the final text. Tools used: AI systems, transcription software, translation aids, and similar tools. Contributors: Community members, informants, and others whose input shaped the work but who are not authors in the formal sense. This kind of transparent layering respects the complexity of modern research without pretending that a machine is a colleague. It is especially useful in heritage research, where community knowledge often shapes a project deeply but is rarely credited in traditional bylines [20]. 7.4 Attribution in student work For students, the practical form of this is simpler. In most institutions, an honest disclosure statement at the end of an essay is enough. It might say, for example, that a model was used to help translate a Spanish-language source, or to check grammar in the introduction, or to brainstorm the outline. As long as the student can defend every factual claim and every interpretation as their own, and can show what the AI contributed, the standards of integrity are met [6], [22]. Instructors, in turn, should treat these disclosures with respect rather than suspicion. A student who discloses AI use is not confessing to cheating; they are practicing the transparency that scholarly work now requires. 8. Digital Dissemination of Cultural Heritage Research 8.1 The digital heritage landscape The last decade has seen a huge expansion of #digital_cultural_heritage. Museums, archives, and libraries have digitized millions of objects and documents. Universities and community groups have built online exhibitions, virtual tours, and open datasets. Search engines and social media platforms are now major channels through which the public encounters the past [4], [17]. This shift changes both who tells historical stories and who hears them. A short video on a social platform may reach more viewers in a day than a scholarly book reaches in a decade. When AI tools help produce that content, they influence public memory on a very large scale. 8.2 AI in museums and archives Cultural institutions have adopted AI for many tasks. Some of the most common are: Automated tagging of images and objects to make collections searchable. Optical character recognition and handwritten text recognition for large document sets. Chatbots that answer visitor questions on museum websites and apps. Recommender systems that suggest exhibits or archival items to users based on their interests. Generative tools that produce draft labels, captions, and educational materials [4], [10], [18]. Each of these can improve access. Each can also introduce errors, biases, or misleading framings if used without human review. 8.3 Risks in AI-generated heritage narratives Several specific risks arise when AI systems generate narratives about heritage. Flattening of cultural difference. A model trained mostly on Western sources may describe non-Western practices in a familiar frame that erases what makes them distinct. For example, complex indigenous kinship systems may be reduced to simple family structures that fit the model's default patterns [5], [19]. Reproduction of colonial voices. Many older archival materials were produced by colonial administrators, missionaries, and travelers. A model trained on such materials may inherit their framings. Without careful curation, its outputs can echo colonial tones even in twenty-first century texts [19], [20]. Invented artifacts and events. A model asked about a lesser-known site or figure may invent plausible details. If those details enter a museum label or a tourist brochure, they can then feed back into the training data of future models, embedding the error more deeply. Loss of source relationships. Heritage research depends on tracing where information comes from: which document, which informant, which object. AI outputs often break these chains, presenting synthesized text without provenance [10]. 8.4 Public-facing AI narratives Cultural institutions increasingly use AI in public-facing contexts. Chatbots welcome visitors to museum websites and answer questions in natural language. Some institutions have experimented with AI-generated audio guides and with conversational agents that speak in the voice of historical figures. These uses are exciting, but they raise sharp ethical questions [4], [18]. When an AI system speaks in the voice of a real historical person, it may produce statements that person never made. Visitors may not always understand that the words are generated rather than documented. In one common approach, the institution restricts the AI to a curated set of documented quotes, so that the voice remains historically anchored. In another approach, the AI generates freely from a general model, which is faster to build but far riskier. The choice between these approaches is not just technical. It is a choice about what kind of relationship the institution wants to have with truth and with the people whose voices are being represented. Similar questions arise for AI-generated images and reconstructions. A model can produce a plausible-looking portrait of a person for whom no image survives. It can generate a walkthrough of a temple that has long been in ruins. These reconstructions can be powerful teaching tools, but they can also blur the line between evidence and imagination. Clear labeling of AI-generated content is a basic step, and one that is already becoming standard practice in leading institutions [4], [18]. 8.5 Community-centered approaches A growing body of work argues that heritage research must be done with communities rather than only about them [19], [20]. This includes: Involving community members in deciding what parts of their past should be digitized and shared. Sharing control over how AI is trained on community materials. Allowing communities to add, correct, or challenge AI-generated narratives. Recognizing forms of knowledge, such as oral tradition, that do not fit standard archival categories. When AI tools are used in this way, they can amplify community voices instead of replacing them. When they are used without such consultation, they risk repeating the harms of earlier extractive research practices. 9. Illustrative Scenarios To make these ideas concrete, this section presents three short scenarios based on common student situations. The scenarios are composite and illustrative, drawing on patterns discussed in the recent literature. 9.1 Scenario one: The undergraduate essay A second-year history student is assigned a 3000-word essay on the memory of the First World War in a small European country. Short on time, they ask a chatbot to draft the essay. The chatbot produces fluent prose that cites a mix of real and invented books. The student submits the essay after light editing. The instructor notices two citations that seem odd. A quick check shows they do not exist. Under the framework proposed in Section 3, several failures are present: the student did not verify sources (epistemic trust), did not disclose AI use (distributed authorship), and did not engage with the specific community whose memory is being described (cultural stewardship). A better path would have been to use the model for brainstorming, then to build the essay on verified sources, disclose the tool used, and, where possible, engage with real memorial sites or local historians. 9.2 Scenario two: The museum internship A master's student interns at a regional museum and is asked to draft object labels for a new exhibition on migration. The student uses a generative model to produce first drafts of one hundred labels. The drafts are grammatically clean but occasionally describe cultural practices in general terms that do not fit the specific objects. The student and their supervisor review each label. They keep the drafts as a starting point but rewrite most of them with input from community members represented in the exhibition. In the exhibition's digital catalog, they add a short note explaining that AI was used for initial drafting and that final texts were shaped by community consultation. This path uses AI's efficiency without letting it define the exhibit's voice. It respects both academic integrity (transparency) and cultural stewardship (community input). 9.3 Scenario three: The doctoral thesis A doctoral candidate researching #precolonial trade networks uses a translation model to work with sources in a language they read only slowly. The model helps them process many documents in a short time. They keep a detailed log of every document, the original passage, the model's translation, and their own corrections after checking with a human expert. When they publish an article from their thesis, they describe this workflow in the methods section. They also share the log as supplementary material. Reviewers can see exactly where AI helped and where human judgment took over. This case shows how AI can enable research that would otherwise be very slow, while keeping integrity strong through documentation. 10. Toward an Ethical Framework Drawing the strands together, this section outlines an ethical framework for AI-generated historical narratives in cultural heritage research. It is meant as a starting point for institutions, teachers, and students, not as a final rulebook. 10.1 Core principles Five core principles emerge from the analysis. Verifiability. Every factual claim in a scholarly text should be traceable to a real, checkable source. AI outputs should be treated as leads, not as evidence in themselves [1], [6]. Transparency. Any significant use of AI tools should be disclosed in a clear statement, describing the tool, the purpose, and the level of human oversight [13], [14]. Accountability. The human author remains responsible for the final text. AI systems cannot bear moral or professional responsibility, and should not be listed as authors [15]. Cultural respect. Narratives about a community's past should, whenever possible, be shaped with that community. AI use should not erase this obligation [19], [20]. Equity. Access to good AI tools and to training in their use should be broadened, not narrowed, so that integrity standards do not become privileges of well-funded institutions [24]. 10.2 Roles and responsibilities Different actors have different responsibilities within this framework. Students are responsible for verifying claims, disclosing tool use, and taking ownership of their final work. They should treat AI as a study partner that must be checked, not as an oracle. Instructors are responsible for designing assessments that reward genuine engagement with sources and reasoning, and for teaching AI literacy alongside traditional research skills [7], [22]. Institutions are responsible for clear, current policies, for equitable access to tools, and for supporting instructors and students through this transition [1], [24]. Journals and publishers are responsible for consistent disclosure requirements and for handling AI-related integrity cases fairly [13], [14]. Cultural institutions are responsible for consulting communities, curating training data carefully, and reviewing AI-generated content before it is shared with the public [4], [19], [20]. Developers of AI tools have a responsibility to be transparent about training data, limitations, and known biases, so that scholarly users can make informed choices [5], [9]. 10.3 An integrity contract A useful practical device is what one might call an integrity contract: a short document, produced by the student and countersigned by the instructor, that spells out how AI will be used in a particular course or thesis. It might cover which tools are allowed, which are not, how disclosure will be made, and how the student and instructor will handle problems if they arise. Such contracts move integrity from a distant rulebook into a shared understanding between real people. 11. Recommendations for Practice Building on the framework, the following recommendations translate principles into concrete practice. 11.1 For universities and departments Update academic integrity policies to reflect the reality of AI use, replacing blanket bans with clear disclosure rules. Fund equitable access to reliable AI tools, including for students from lower-income backgrounds. Provide training modules on AI literacy for both students and staff. Encourage assessment design that emphasizes process, engagement with local sources, and oral defense. Create clear procedures for handling suspected misuse, with an emphasis on education over punishment for first-time cases. 11.2 For instructors Discuss AI openly with students in the first week of a course. Ask students to submit an AI-use statement with each significant assignment. Design assignments that require engagement with specific archives, sites, or people, not only with topics that can be summarized by a chatbot. Include, where possible, an oral component in which students defend their arguments. Model good practice by disclosing your own use of AI in course materials and research. 11.3 For students Treat AI outputs as first drafts, never as final evidence. Check every citation against real databases and library catalogs. Keep a simple log of when and how you used AI, so that you can describe your process if asked. Disclose AI use in a short statement at the end of essays or in a #methods section for larger projects. Ask your instructor early if you are unsure whether a particular use is allowed. 11.4 For journals and publishers Publish clear, up-to-date policies on AI use in submitted work. Require disclosure statements and, for large uses, methodological detail. Refuse AI systems as listed authors. Provide reviewers with guidance on how to evaluate AI-related disclosures. Support retraction and correction procedures that consider AI as a factor. 11.5 For cultural institutions Consult communities before using AI on materials related to them. Curate training data and prompts carefully, avoiding uncritical use of colonial-era sources. Publish clear information about which parts of exhibitions and catalogs were AI-assisted. Provide easy channels for the public to flag errors in AI-generated content. Invest in staff training on AI ethics, not only on technical operation. 11.6 For developers Be transparent about training data sources and their limitations. Provide documentation aimed at humanities users, not only at engineers. Work with heritage institutions to develop tools tuned to the specific needs of the field. Support features that make provenance and disclosure easier, such as export of prompts and outputs. 12. Discussion The picture that emerges from this review is neither utopian nor apocalyptic. AI-generated historical narratives are already a part of academic and public life. They can save time, open languages, and widen access to the past. They can also fabricate sources, flatten cultural difference, and shift responsibility away from the humans who should hold it. Whether the balance tips toward benefit or harm depends less on the technology itself than on the practices around it. Three points deserve particular emphasis. First, integrity is not just a matter of rules. It is a matter of habit and culture. A framework that focuses only on catching cheaters will fail. A framework that builds honest, transparent, and reflective practice into everyday work has a chance to succeed [1], [6], [22]. Students who are taught, from the start, to disclose their tools and to check their sources are less likely to see AI as a shortcut around learning. Second, authorship should be understood as a bundle of functions, not as a single line at the top of a document. Some of those functions can be shared with machines. Others, especially accountability and moral responsibility, cannot [15], [16]. Making this distinction clear helps everyone: it lets scholars use tools honestly, and it keeps the moral core of scholarship in human hands. Third, cultural heritage research adds a layer that pure integrity discussions often miss. The past of a community is not raw material. It is part of who they are. Scholars who use AI to write about that past have a special duty to make sure that the community's voice is not replaced by a smoothed-out synthesis from a machine [19], [20]. Meeting this duty is not always easy, but the alternative is a new form of cultural loss. The proposed framework tries to hold these three points together. It is deliberately practical, aimed at real students and real institutions. It is also deliberately open, because the technology and its uses are still changing fast. Any rulebook written today will need to be revised in a few years. 13. Limitations and Future Research This paper has several limitations. As an integrative review, it depends on the published literature and cannot capture practices that have not yet been studied. Most of the reviewed studies come from English-language journals, which itself reflects the very biases the paper discusses. The scenarios in Section 9 are illustrative, not empirical case studies. And the recommendations, while drawn from the literature, have not been tested through implementation research. Several directions for future work stand out. Empirical studies of student practice in history and heritage courses, including surveys and interviews across regions. Comparative studies of institutional policies on AI in the humanities, tracking how they evolve. Design research on assessment formats that support integrity without becoming unfair to non-native speakers or students with disabilities. Community-based research on how heritage communities want AI to be used with materials that concern them. Longitudinal studies of how AI-generated content in museums and archives is received by different publics. Technical research on tools that support provenance tracking and transparent disclosure of AI use. Progress in these directions will help move the debate from general principles to well-tested practices. 14. Conclusion AI-generated historical narratives are neither a passing trend nor a threat that can be banned away. They are a new part of how the past is written and shared. For students working in history, heritage, and digital humanities, the question is no longer whether to encounter these tools but how to work with them without losing the trust that scholarship depends on. This paper has argued that academic integrity and authorship must be redefined for this new landscape, but not weakened. Integrity now means verifiability, transparency, and accountability. Authorship now means holding moral responsibility for a text whose production may involve many contributors, including machines. Cultural stewardship now means making sure that the communities whose past is being told have a real voice, not just a role in training data. If universities, journals, cultural institutions, developers, and, above all, students and their teachers take these ideas seriously, AI can become a partner in the work of remembering. If they do not, AI can quietly reshape memory in ways that few notice until it is too late. The choice is being made now, in countless small decisions in classrooms, seminar rooms, and exhibition halls. This paper has tried to give students and their teachers a clearer map of that choice. References [1] Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024). Chatting and cheating: Ensuring academic integrity in the era of ChatGPT. Innovations in Education and Teaching International, 61(2), 228 to 239. https://doi.org/10.1080/14703297.2023.2190148 [2] Chan, C. K. Y., & Hu, W. (2023). Students' voices on generative AI: perceptions, benefits, and challenges in higher education. International Journal of Educational Technology in Higher Education, 20, 43. https://doi.org/10.1186/s41239-023-00411-8 [3] Biagioli, M., & Lippman, A. (Eds.). (2020). Gaming the Metrics: Misconduct and Manipulation in Academic Research. MIT Press. [4] Terras, M., Coleman, S., Drost, S., Elsden, C., Helgason, I., Lechelt, S., Osborne, N., Panneels, I., Pegado, B., Schafer, B., Smyth, M., Thornton, P., & Speed, C. (2021). The value of mass-digitised cultural heritage content in creative contexts. Big Data and Society, 8(1). https://doi.org/10.1177/20539517211006165 [5] Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (FAccT '21), 610 to 623. https://doi.org/10.1145/3442188.3445922 [6] Perkins, M. (2023). Academic integrity considerations of AI large language models in the post-pandemic era: ChatGPT and beyond. Journal of University Teaching and Learning Practice, 20(2). https://doi.org/10.53761/1.20.02.07 [7] Kasneci, E., Sessler, K., Kuchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Gunnemann, S., Hullermeier, E., Krusche, S., Kutyniok, G., Michaeli, T., Nerdel, C., Pfeffer, J., Poquet, O., Sailer, M., Schmidt, A., Seidel, T., Stadler, M., Weller, J., Kuhn, J., & Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. Learning and Individual Differences, 103, 102274. https://doi.org/10.1016/j.lindif.2023.102274 [8] Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., & others. (2022). On the Opportunities and Risks of Foundation Models. Stanford Center for Research on Foundation Models. [9] Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., Baabdullah, A. M., Koohang, A., Raghavan, V., Ahuja, M., & others. (2023). Opinion Paper: So what if ChatGPT wrote it? Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI. International Journal of Information Management, 71, 102642. https://doi.org/10.1016/j.ijinfomgt.2023.102642 [10] Muehlberger, G., Seaward, L., Terras, M., Ares Oliveira, S., Bosch, V., Bryan, M., Colutto, S., Deckers, H., Dejean, H., Diem, M., Fiel, S., & others. (2021). Transforming scholarship in the archives through handwritten text recognition: Transkribus as a case study. Journal of Documentation, 77(6), 1372 to 1391. https://doi.org/10.1108/JD-07-2018-0114 [11] Fiormonte, D., Chaudhuri, S., & Ricaurte, P. (Eds.). (2022). Global Debates in the Digital Humanities. University of Minnesota Press. [12] Susnjak, T., & McIntosh, T. R. (2024). ChatGPT: The end of online exam integrity? Education Sciences, 14(6), 656. https://doi.org/10.3390/educsci14060656 [13] Thorp, H. H. (2023). ChatGPT is fun, but not an author. Science, 379(6630), 313. https://doi.org/10.1126/science.adg7879 [14] Flanagin, A., Bibbins-Domingo, K., Berkwits, M., & Christiansen, S. L. (2023). Nonhuman authors and implications for the integrity of scientific publication and medical knowledge. JAMA, 329(8), 637 to 639. https://doi.org/10.1001/jama.2023.1344 [15] Eshraghian, J. K. (2023). Human ownership of artificial creativity. Nature Machine Intelligence, 2, 157 to 160. https://doi.org/10.1038/s42256-020-0161-x [16] Floridi, L. (2023). AI as agency without intelligence: On ChatGPT, large language models, and other generative models. Philosophy and Technology, 36(1), 15. https://doi.org/10.1007/s13347-023-00621-y [17] Terras, M. (2022). The role of the library when computers can read: Critically adopting handwritten text recognition tools. In Digital Humanities and Libraries and Archives in Religious Studies (pp. 137 to 154). De Gruyter. [18] Assael, Y., Sommerschield, T., Shillingford, B., Bordbar, M., Pavlopoulos, J., Chatzipanagiotou, M., Androutsopoulos, I., Prag, J., & de Freitas, N. (2022). Restoring and attributing ancient texts using deep neural networks. Nature, 603(7900), 280 to 283. https://doi.org/10.1038/s41586-022-04448-z [19] Risam, R. (2019, reprinted 2021). New Digital Worlds: Postcolonial Digital Humanities in Theory, Praxis, and Pedagogy. Northwestern University Press. [20] Carroll, S. R., Garba, I., Figueroa-Rodriguez, O. L., Holbrook, J., Lovett, R., Materechera, S., Parsons, M., Raseroka, K., Rodriguez-Lonebear, D., Rowe, R., Sara, R., Walker, J. D., Anderson, J., & Hudson, M. (2021). The CARE Principles for Indigenous Data Governance. Data Science Journal, 20, 43. https://doi.org/10.5334/dsj-2021-043 [21] International Center for Academic Integrity. (2021). The Fundamental Values of Academic Integrity (3rd ed.). ICAI. [22] Lodge, J. M., Yang, S., Furze, L., & Dawson, P. (2023). It is not the technology, stupid: How the focus on technology in higher education is misplaced in the post-pandemic era. Australasian Journal of Educational Technology, 39(1), 1 to 8. https://doi.org/10.14742/ajet.8339 [23] Liang, W., Yuksekgonul, M., Mao, Y., Wu, E., & Zou, J. (2023). GPT detectors are biased against non-native English writers. Patterns, 4(7), 100779. https://doi.org/10.1016/j.patter.2023.100779 [24] Selwyn, N. (2022). The future of AI and education: Some cautionary notes. European Journal of Education, 57(4), 620 to 631. https://doi.org/10.1111/ejed.12532 [25] Hosseini, M., Rasmussen, L. M., & Resnik, D. B. (2024). Using AI to write scholarly publications. Accountability in Research, 31(7), 715 to 723. https://doi.org/10.1080/08989621.2023.2168535 #AI_generated_narratives #academic_integrity #digital_cultural_heritage #authorship_ethics #generative_AI_in_history #large_language_models #scholarly_transparency #digital_humanities #decolonial_heritage #AI_and_students #ChatGPT_in_education #machine_writing #cultural_stewardship #historical_research_ethics #AI_disclosure

This article examines how #international_funding for #cultural_heritage shapes the way #local_preservation is planned and delivered on the ground. It uses the lens of #coercive_isomorphism, one of the three pressures described in classical institutional theory, together with mimetic and normative pressure, to explain why #heritage_organisations in very different countries end up looking and acting alike. The paper draws on recent studies of #UNESCO listing, World Bank and #IFC lending, private American giving, and non-governmental accreditation to trace the paths through which donor rules travel into national laws, site management plans, and everyday practice. Findings suggest that #grant_conditionality functions as a soft but powerful form of #policy_transfer, quietly rewriting what counts as good conservation, who is allowed to speak for a site, and which parts of a place should be saved. The article shows that when funding is scarce and local budgets are thin, recipients often adopt #donor_language, hire donor-style experts, and re-organise themselves in ways that make them recognisable to the funder, even when local values point in another direction. Yet convergence is not total. Decoupling, local translation, and quiet resistance also appear, especially when #Indigenous_communities or long-standing custodians disagree with international scripts. The paper argues that heritage students, practitioners, and policymakers need a clear understanding of these pressures so they can negotiate them, rather than mistake them for neutral good practice. It closes with practical suggestions for more balanced funding relationships that respect local agency while still meeting the reporting needs of large donors. Keywords: coercive isomorphism, heritage funding, UNESCO, World Bank, NGO standards, local preservation, policy transfer, community participation. 1. Introduction #Heritage_preservation is often described as a technical activity. In practice it is deeply political. Every decision about which building to save, which layer of a site to expose, and which story to tell in a museum label involves choices between competing values. When those choices are made under the shadow of an #international_grant or an accreditation scheme, they are shaped by pressures that go far beyond the site itself. This article asks a simple but pressing question. How do the rules attached to international heritage funding, and the standards published by non-governmental organisations, guide the shape of local #preservation_strategies in ways that participants may not always notice? The question matters because the resources involved are large and uneven. Public money for #cultural_heritage is often small compared with health or defence, so external funding from bilateral aid programmes, development banks, private foundations, and international NGOs can be decisive. In many countries, especially in the #Global_South, a single grant from a major donor can outweigh several years of the national heritage budget. Barnhart shows that private American organisations alone channel tens of millions of dollars every year into #preservation_abroad, and that this money is only a small slice of a much larger philanthropic flow moving from the United States to the rest of the world. Once a source of funding is that important, its rules stop being simple paperwork. They become a kind of soft law that reorganises how recipient bodies think about themselves. The theoretical starting point of the paper is #institutional_isomorphism, a concept from organisational sociology that explains why organisations in the same field tend to become similar over time. In the classical formulation there are three mechanisms. Coercive pressure comes from formal rules, funding conditions, and legal requirements. Mimetic pressure appears when actors copy others who seem successful, especially when the future is uncertain. Normative pressure spreads through shared training, professional networks, and codes of ethics. Recent research on museum accreditation across five very different national systems, from the American Alliance of Museums to the Chinese national grading system, confirms that all three mechanisms operate simultaneously in the heritage field, producing a shared vocabulary of the good museum, standards-based evaluation, periodic review, and a common professional ethics across radically different political economies. Of these three, coercive pressure is often the most direct in heritage funding because grants come with binding terms and conditions. The paper focuses on that coercive channel, but it does not pretend that the three mechanisms operate in isolation. On the ground they blend. A conservation officer in a small municipality may adopt an international management plan template because a grant requires it, because everyone else in the region has done so, and because their own professional association endorses it, all at once. In a study of quality-development networks for early childhood providers in Norway, Haugset and Finne found that #coercive_pressure, normative pressure, and mimetic pressure become intertwined in real settings, so that formal compliance, professional identity, and imitation reinforce one another. The heritage field is similar. Formal grant conditions travel through the same conferences and training courses where professional norms are set, so it is often hard to say where one pressure ends and the next begins. Three developments give the question new urgency. First, the #UNESCO World Heritage Convention has grown into an almost universal instrument, with well-developed procedures that reach deep into national planning systems. Second, the #World_Bank and the #International_Finance_Corporation have built their own Performance Standards for cultural heritage that private banks and insurers have voluntarily adopted, turning what was once an internal safeguard into a de facto global standard. Third, climate change and armed conflict have raised the cost of preservation faster than most national budgets can absorb, so more sites depend on external assistance than ever before. Each of these developments increases the leverage that international funders have over local practice. This article aims to help students and early-career practitioners see these pressures clearly. It does not argue that international support is bad. Much of it is valuable, and some sites would already have collapsed without it. It argues, more carefully, that donor requirements should be read as active shapers of local strategy, not neutral technical tools. When we understand them that way we can negotiate them, translate them into local terms, and sometimes push back against them. The paper proceeds as follows. Section 2 reviews institutional theory and its application to #heritage_funding. Section 3 explains the methodology, which is a structured narrative review of recent literature. Section 4 sets out the findings across four channels of coercive pressure, from #UNESCO listing through development bank lending, private foundations, and NGO accreditation. Section 5 discusses what the findings mean for #community_participation, Indigenous rights, and the future of heritage governance. Section 6 concludes with practical guidance. 2. Literature Review and Theoretical Framework 2.1 Institutional Isomorphism in a Heritage Setting The idea of institutional isomorphism was developed to answer a puzzle. If organisations act in their own interest, why do they so often look the same as their neighbours? The answer in institutional theory is that organisations are not only trying to be efficient, they are also trying to be legitimate. Legitimacy requires being recognised as a proper example of one's kind. A museum that does not look like a museum, or a heritage agency that does not follow the usual reporting formats, is often treated with suspicion by funders, peers, and the public. Over time, this pressure to be recognisable pulls organisations towards a shared template. Recent applications of the theory show that this process is not only shaped by governments. It also flows through funding relationships. Moreau, studying French non-profit sports associations, shows that the search for independent funding, together with the arrival of better-trained volunteers, drives associations to import management tools from the corporate sector, even when their mission is not commercial. Toner and Martins, looking at cross-cultural development volunteers, find that the way volunteers are prepared and projects are structured is shaped by managerialist thinking that flows through the entire volunteer supply chain. Both studies show that isomorphism reaches into the internal life of organisations, not only their external forms. For heritage the implications are direct. When a site management authority applies for a large grant, it is often asked to produce a logical framework, a risk register, an environmental and social management plan, a stakeholder engagement plan, and a monitoring and evaluation matrix. Each of these documents is portable across sectors, and each one carries assumptions about how change happens. Once these documents become the shared language of the field, staff hired to write them are trained in similar ways, and the field slowly converges on a common template regardless of local context. 2.2 Heritage Funding as an Institutional Field The heritage funding field is unusually crowded. #UNESCO sets the highest-profile normative benchmark, but it works alongside the World Bank, the Asian Development Bank, the European Union, USAID, national aid agencies such as the Japanese and French cultural funds, private foundations such as the Getty and the World Monuments Fund, and a growing set of philanthropic vehicles based in the Gulf and East Asia. Wang and colleagues note that the social sciences have paid too little attention to these non UNESCO actors, even though they often collaborate with UNESCO and provide the actual money and expertise for site management. Because the field is crowded, standards often overlap. A site may be inscribed on the World Heritage List, receive a loan governed by IFC Performance Standard 8 on cultural heritage, and also be included in a bilateral aid programme with its own reporting format. Each of these frames comes with its own template for what a site is, who its stakeholders are, and what counts as success. Recipients must satisfy all of them at once. Mason and Martindale argue that the IFC Performance Standards, which were originally designed for private-sector risk management, have evolved into a de facto global standard on cultural heritage that has been voluntarily adopted well beyond the corporation's own clients. In effect, a private finance risk tool now sets the tone for what many public bodies treat as good practice. The result is a form of layered conditionality. The layers are not always consistent, but recipients have limited power to challenge them. They must either meet the highest common denominator, which is expensive, or invest energy in showing formal compliance while continuing to work in other ways, which is decoupling. Elliott and Schmutz, studying the World Heritage movement in Africa, describe decoupling as both a shortage of the specialist scientific and technical expertise that international procedures assume, and a cultural mismatch between local and global conceptions of universal value. Decoupling is not laziness. It is a survival strategy in a field where full compliance with every layer of donor rules is often impossible. 2.3 Coercive Pressure, Global Governance, and the Global South Coercive pressure is often felt hardest in the #Global_South, because that is where the ratio of external to domestic funding is highest, and where state capacity to negotiate is often thin. Bacsi's recent quantitative study of countries hosting sites on the List of World Heritage in Danger shows that economic affluence and state stability are strongly correlated with conservation status, and that a heritage divide between the Global North and the Global South is one of the main features of the international heritage system. When a country is poorer and less stable, it depends more heavily on external funding, and its ability to shape the terms of that funding is smaller. Farias reads the same imbalance through international relations. She traces how what began as an expert-driven classification exercise has become an openly political process in which nations use #World_Heritage inscription as a tool of #soft_power and national identity. Inscription is valuable enough that states adjust domestic laws, planning systems, and even museum practices to fit the criteria. In this sense, the coercive pressure of World Heritage is not only felt in explicit funding conditions, but also in the anticipation of future rewards. States that want to be recognised behave in advance in ways that will make them recognisable. Chuva, writing on heritage policies and sensitive pasts in Latin America, shows that these global-to-local transfers are rarely neutral. They bring particular views about what counts as significant, whose past deserves to be told, and which silences are acceptable. When donor frameworks import these views without translation, they can crowd out local memory work, especially around colonial, slave, or Indigenous histories that fit uneasily into universalist scripts. Two recent works help sharpen the picture. Dupeyron argues that behind the enthusiasm for the #cultural_turn in international aid lies a more cynical reality in which cultural projects are used to serve donor agendas. Leloup, studying the #Culture_Bank in West Africa, shows how a locally invented model was reshaped by the World Bank and #ICCROM support, exported to Togo and Benin, and produced unexpected outcomes that were not visible at headquarters. Both studies confirm that even well-intentioned international engagement can rewrite local practice in ways that only become visible after several years. 3. Methodology This paper uses a structured narrative review of recent peer-reviewed literature and policy-oriented studies published within the last five years. The aim is not to test a single hypothesis, but to build a coherent map of how #coercive_isomorphism operates in heritage funding, and to give students a set of examples they can use to interpret their own field sites. Searches were carried out across academic databases with combinations of the terms #coercive_isomorphism, #institutional_isomorphism, #heritage_funding, international grants, #UNESCO, #World_Bank, #donor_conditionality, and #community_participation. Priority was given to studies published from 2021 onwards, with a small number of important 2020 works retained where no more recent equivalent existed. Sources were included if they addressed at least one of the following: the mechanisms of coercive or normative pressure in a heritage or cultural sector setting, the impact of donor requirements on local preservation strategy, or the interaction between global standards and community-level practice. The review is interpretive rather than statistical. The evidence base in this area is mostly qualitative, drawn from case studies, policy analyses, and ethnographic work. A meta-analysis in the strict sense is not possible with such heterogeneous data. Instead, the paper reads across cases to identify recurring patterns and disagreements, and it is transparent about where evidence is thin or contested. Where a single case is used, it is presented as an illustration rather than as proof. Two limitations should be noted. First, most of the studies reviewed are written in English. Important work in French, Spanish, Portuguese, Arabic, and Chinese is only partly reflected, and this introduces a bias in favour of Anglo-American framings of the problem. Second, the review depends on published material and does not include direct fieldwork. Grey literature from donor agencies is included where it is publicly available, but the internal reasoning of funders is often not documented in the sources. 4. Findings The findings section works through four channels through which coercive isomorphism travels into local heritage practice. These are UNESCO listing and its accompanying procedures, development-bank lending, private foundations and bilateral cultural aid, and NGO accreditation schemes. The four channels overlap, but keeping them analytically separate makes the pressures easier to see. 4.1 UNESCO Listing as a Coercive Anchor UNESCO listing is often described as a soft instrument because it does not by itself distribute money. Yet its coercive effect is strong. Countries that want to nominate a site must first prepare a tentative list, then submit a full dossier, then produce a management plan, and finally accept periodic monitoring. Each of these steps is governed by an Operational Guidelines document that is regularly revised and that assumes a particular administrative style. Farias's analysis shows that although these procedures grew out of expert consensus, they have become a political process in which state parties reshape domestic planning to fit the criteria in the hope of national and international recognition. The coercive nature of the process becomes clear at three points. First, #Outstanding_Universal_Value is defined through a small set of criteria that were developed in a Euro-Atlantic scholarly tradition. Elliott and Schmutz argue that these criteria carry rational-scientific assumptions about evaluation that generate regional inequality, because countries that already have the scientific infrastructure to write in this style have a large advantage in the nomination process. Second, once a site is inscribed, the #management_plan format becomes binding, and any conservation intervention that departs from it can be reported to the Committee. Third, sites on the List of World Heritage in Danger are subject to further international assistance procedures that come with their own conditions. Bacsi's cross-country analysis of the In Danger list makes the imbalance concrete. Using data from 31 countries with sites on the list, she finds that GDP per capita, the volume of international assistance received, and the fragile state index together predict a site's In Danger status more strongly than the objective severity of environmental threats. In other words, in the current system, the socio economic position of a country appears to be a stronger predictor of heritage survival than the actual level of physical threat to its sites. This is a strong indication that the international framework, however well designed, is not neutral in its effects. The point is not that UNESCO forces states to change their laws. It is more subtle. States anticipate the requirements of listing and monitoring, and they adjust their heritage administrations in advance to look like the model that has succeeded elsewhere. This anticipatory adjustment is one of the clearest signs of #coercive_pressure at work, because it does not require any direct order from Paris. The pressure is built into the reward structure of the field. The case of #S_Gang_Gwaay on Haida Gwaii illustrates how this pressure can collide with local values. The site was inscribed with very little initial involvement of the Haida community that built and inhabited it, and although the Haida have since taken a large role in site management, their belief that the totem poles and structures should be allowed to return to the earth conflicts with standard conservation practice. Robinson argues that the accelerated degradation caused by climate change will force an explicit compromise between the values of rights holding communities and the longevity of the physical fabric. The example shows that coercive pressure to preserve the material substance of a site can override community understandings of what that site is for, unless there is a deliberate effort to hold the two together. 4.2 Development Bank Lending and Layered Safeguards The #World_Bank_Group has been engaged in cultural heritage lending for decades, and its cumulative influence on how projects are structured is very large. Its private-sector arm, the #International_Finance_Corporation, introduced its eight Environmental and Social Performance Standards in 2006. Performance Standard 8 covers cultural heritage. Mason and Martindale explain that these standards were originally designed to help IFC clients manage their own risk exposure, but they have since evolved into a de facto global standard that many private banks, insurers, and development proponents have voluntarily adopted. This is a striking example of what institutional theorists call policy diffusion, in which a rule designed for one purpose becomes the field-wide template for another. The authors argue that this widespread adoption should not simply be celebrated as good governance. The IFC standards were not designed to embody the full range of internationally recognised good practice on cultural heritage, and their spread has left some important elements out. Mason and Martindale therefore call for the development of a bespoke cultural heritage framework tailored to the private sector. From an isomorphism point of view, this is a telling observation. When a risk management tool becomes a global template, cultural heritage is treated primarily as a source of risk to be managed, rather than as a public good to be nurtured. That framing shapes what recipients feel they must do. Layered on top of the IFC standards are the environmental and social safeguards of the World Bank itself, and of other regional development banks that broadly follow the same architecture. When a heritage project is co-financed by more than one bank, the recipient must satisfy several overlapping rulebooks. In practice this often means hiring a consultant firm that specialises in producing the required documents. Toner and Martins show that in development work more generally, the preparation of teams and the structuring of projects is shaped by managerialist modes of thinking that create an environment for sustainable knowledge sharing but also for the reproduction of a specific organisational form. When applied to heritage, this managerialism can crowd out slower, less document-heavy ways of caring for a place. 4.3 Private Foundations, Bilateral Aid, and the Cultural Turn Alongside development banks, private foundations and bilateral aid programmes have grown in importance. Barnhart's study of private American organisations supporting preservation abroad shows that the sector is large, diverse, and shaped by interests that extend well beyond conservation itself. She notes that heritage sites are important platforms for the construction and mediation of social values and identities, and that private funding for foreign preservation is a distinct kind of American influence abroad that has not been studied as carefully as government-led cultural diplomacy. The organisations in her sample fill a diverse set of functions, but they operate in a complex ecosystem where the dynamics of philanthropy, international affairs, and domestic politics all shape their agendas. Dupeyron's account of the #cultural_turn in international aid, focused on heritage and the creative industries, argues that the current enthusiasm for cultural projects hides a more cynical reality in which cultural work is instrumentalised to serve donor objectives. When cultural heritage is turned into an instrument for economic development, poverty reduction, or diplomatic engagement, the choices about what to preserve, what to promote, and how to present it are quietly pulled in the direction of what will best serve those instrumental goals. The #Culture_Bank example in West Africa, analysed by Leloup, brings these dynamics into focus at ground level. The Culture Bank was created in 1997 in the Malian village of Fombori by a Peace Corps member. It combined a small museum, a micro-credit facility, and an education centre, and it was later supported by the World Bank and ICCROM and exported to Togo and Benin. Leloup argues that the model embodies the paradox of cultural heritage and development projects that are torn between the local and global scales, between local communities and international organisations. She also documents fundamental misunderstandings between French actors focused on cultural heritage protection and those focused on sustainable development, and she raises the question of what happens when the #Southern_State disappears as an interlocutor for funding partners. The Culture Bank is a case in which locally rooted ideas were rescaled and repackaged for international audiences, with unpredictable effects at home. Abdel-Rashid's analysis of international cooperation for urban heritage in Egypt after 2015 shows the same tension from the recipient side. International bodies contribute to preserving heritage buildings and sites, but their frameworks must be aligned with local regulations and methodologies if they are to work well, and identifying the funding agencies, understanding their support methods, and integrating their input into national preservation policy is not straightforward. The paper argues for maximum benefit from international cooperation while ensuring full compatibility between global sustainability requirements and national development priorities. That call for compatibility is a diplomatic way of admitting that donor frameworks and national priorities do not always fit, and that the burden of making them fit falls mostly on the recipient. 4.4 NGO Accreditation and the Making of Convergent Practice A fourth channel of coercive pressure runs through non-governmental accreditation schemes. These schemes are technically voluntary, but in practice they often gate access to public funding and to peer recognition. Zhou's structured comparison of five major museum accreditation systems, including the American Alliance of Museums, the Arts Council England Accreditation Scheme, the International Council of Museums professional standards, the Chinese national museum grading system, and the Japanese public-interest certification regime, finds strong convergence across all five systems on the vocabulary of the good museum, on standards-based evaluation, on periodic review, and on a shared professional ethics. Crucially, Zhou identifies which mechanism drives convergence in which case. Convergence is driven by coercive pressure where accreditation gates public funding, as in England and China, by mimetic borrowing from the prestigious American model, and by normative diffusion of scripts authored by the International Council of Museums. This is a rare piece of comparative evidence that separates the three isomorphic mechanisms and shows which one dominates in each context. It confirms that coercive pressure is not the only driver, but that where it is present, it produces strong effects. Zhou also notes that convergence is not total. The two East Asian cases in the sample adopt international forms while routing them through distinctive national logics, so that shared architecture serves different ends. On this basis Zhou proposes a #convergence_with_decoupling thesis and adds translation as a fourth mechanism complementing the three classical ones, capturing how globally shared forms are reshaped as they cross into national institutional settings. The idea of translation is useful because it acknowledges that recipients are not passive. They rework international scripts, sometimes deliberately, to fit their own goals. For heritage students, the accreditation channel is important because it is one place where the individual practitioner meets institutional pressure directly. A curator applying for an accreditation review is asked to produce evidence in a specific format, and the effort of producing that evidence reorganises the museum's internal life. Over time, the audit format becomes a mirror in which staff learn to see themselves. This is coercive pressure at its most intimate, because it works not by punishing but by defining what a competent professional looks like. 4.5 The Reporting Machine and the Reshaping of Everyday Work One of the least visible but most powerful effects of coercive isomorphism runs through the ordinary paperwork of a funded project. Grant applications, inception reports, quarterly progress notes, mid-term reviews, and final evaluations are not simply administrative overhead. They are the medium through which donor expectations enter the daily life of a heritage organisation. Once a team has spent several years producing these documents, staff learn to think in the categories the forms provide. A cultural landscape becomes a set of attributes, a community becomes a stakeholder register, a memory becomes an intangible asset with attributes and safeguarding measures. This is not a criticism of the individual staff members. It is a description of what happens when a template repeats itself often enough to become second nature. Toner and Martins observed exactly this pattern in cross-cultural volunteer projects. They found that the preparation of volunteers and the structuring of project work is shaped by managerialist modes of thinking, and that this creates an environment which is conducive to sustainable knowledge sharing practices but also to a specific organisational form. When the same forms of thinking are applied to heritage, they generate a project logic in which time is broken into discrete milestones, budgets are tied to deliverables, and outcomes are measured against indicators that were chosen in advance. That logic can work well for infrastructure work with clear physical outputs, but it fits less easily with slow, iterative activities such as building trust with a community, learning a local craft, or reconstructing an oral tradition. When the forms cannot capture these activities, they tend to disappear from the visible face of the project, even if practitioners continue to do them quietly in the background. Moreau's study of French non-profit sports associations documents a similar shift, in which associations reorganised themselves around the need for independent funding and imported management and operating tools from the entrepreneurial model, diversified their services, and hired volunteers whose skills came from their professional careers rather than from within the association itself. The heritage field is now moving in the same direction. Site managers are increasingly recruited from project management backgrounds rather than from within the conservation tradition, and their fluency in the language of donors is often the reason they were hired. That is not automatically bad. It does mean that the professional identity of the field is being reshaped from within, one hire at a time. There is also a technological dimension. As more donors move to digital reporting portals, and as more accreditation bodies use standardised online forms, the categories embedded in the software become harder to argue with. A drop-down menu that offers three types of stakeholder rules out a fourth type by design. A required field for expected beneficiaries per year forces staff to invent numbers where none exist. Over time these small design decisions add up to a large influence on what is thought of as thinkable in the field. This is coercion by interface rather than by contract, but its effects are just as real. 4.6 Community Voice under Coercive Pressure The channels described above do not exist in a vacuum. They interact with local communities, and the terms of that interaction are one of the main issues in contemporary heritage studies. Lähdesmäki and colleagues describe a #participatory_turn in the heritage sector, in which participation and community engagement have become umbrella terms that promise to shift practice away from #top_down approaches towards more inclusive and collaborative models where diverse voices contribute to cultural heritage. The book they edit collects case studies on social inclusion, digital tools for participation, and participatory management and governance, and it argues that broader participation is essential for unlocking the transformative potential of heritage. Yet the participatory turn coexists uneasily with the coercive channels described above. When a major donor requires a #stakeholder_engagement plan, communities may be invited to consult on a project that has already been shaped by donor language and by donor deadlines. The result is often a kind of participation that fills a required box without changing the substance of the intervention. Chuva's account of heritage policies with sensitive pasts warns that global-to-local transfers can carry ambiguities and reduce local rights in the process. Robinson's discussion of the Haida at S Gang Gwaay shows how far community priorities can diverge from standard conservation practice, and how the coercive pressure to preserve the material fabric of a site can conflict with the beliefs of the community that created it. There is, however, evidence that recipients can push back. Haugset and Finne, studying private early childhood providers under quality networks in Norway, show that different actors respond differently to the same pressures. Small providers tended to align harmoniously with municipal practice, while larger corporations exploited local variation to argue for stricter national standards that would suit their own strategies. Their study concludes that isomorphic pressure can create the impetus for endogenous and gradual institutional change agency. Applied to heritage, this suggests that recipients are not only shaped by donor pressure. They also use that pressure strategically, aligning with it when it suits them and quietly resisting it when it does not. 4.7 Cross-Cutting Case Illustrations To make the four channels tangible, it is worth walking through three short illustrations drawn from the recent literature. Each example is not offered as proof but as a way to see how the pressures described above play out together in a real setting. The first illustration is the pattern of long-running World Bank engagement in cultural heritage lending, which has supported many projects across historic cities and cultural landscapes over several decades. The documented approach usually combines infrastructure investment with capacity building and international expert missions, and it channels lessons learned into national policy. The pattern shows how a lending partnership can generate a durable institutional shape that outlives any single project, because the categories used in project design become embedded in ministry procedures for future work with other donors. Mason and Martindale note that the resulting standards, especially the IFC Performance Standards, have spread far beyond the corporation's own clients and now function as a de facto global standard. That spread is a textbook example of how a single tool becomes a field-wide template. The second illustration is Egypt's negotiation of international cooperation for urban heritage after the 2015 Sustainable Development Goals framework was launched. Abdel-Rashid describes how international bodies contribute to the preservation of heritage buildings and sites, but stresses that alignment with local regulations and methodologies is essential if their contributions are to produce lasting benefits, and that identifying funding agencies and understanding their support methods is a substantial task in itself. The paper argues explicitly for compatibility between global sustainability requirements and national development priorities. Between the lines, the study is a description of what it costs a national administration to keep up with the paperwork of layered conditionality. The third illustration is the Culture Bank case in West Africa. Leloup shows that the initiative combined a museum, micro-credit, and education, and was later supported by the World Bank and ICCROM before being exported to Togo and Benin. She also documents fundamental misunderstandings between the French actors focused on cultural heritage protection and those focused on sustainable development, and she raises the question of what happens when the Southern State disappears as an interlocutor for funding partners. The example is important because it shows a locally invented model being reshaped by international support, exported to other countries, and generating outcomes that would not have been visible without careful long-term observation. It also shows that not every effect of international engagement is negative. Some of the changes brought by the Culture Bank were valued locally. But the pattern of reshaping and re-export illustrates how coercive and normative pressures can pull local innovations into global circuits in ways that both amplify and distort them. Taken together, the three illustrations show that the four channels described earlier rarely operate in isolation. A single site can be shaped by UNESCO nomination expectations, by a World Bank loan condition, by a private foundation programme, and by a professional accreditation scheme all at once. Staff learn to hold several frameworks in mind at the same time, and to switch between them as different audiences require. That is a demanding intellectual task, and it deserves to be recognised as part of the real skill set of contemporary heritage practice. 5. Discussion The evidence gathered in Section 4 points to a picture that is more complex than a simple story of donor dominance. Coercive pressure is real, and its effects are visible in the vocabulary, procedures, and staffing of heritage bodies across very different countries. Yet the pressure is not delivered by a single hand. It flows through many parallel channels, including UNESCO procedures, development-bank safeguards, private foundation priorities, and NGO accreditation schemes. Each channel has its own logic, but they reinforce one another because they draw on the same pool of consultants, the same conferences, and the same professional networks. Three implications follow. First, convergence in heritage practice is not the result of a single grand plan. It is the emergent outcome of many small alignments. When a country adjusts its heritage law to fit UNESCO expectations, when a national agency adopts an IFC-style safeguard, and when a museum adopts an ICOM-aligned code of ethics, the alignments do not have to be coordinated to produce a common global template. Over time they produce one anyway. Zhou's proposal to add translation to the classical set of isomorphic mechanisms is therefore valuable, because it captures how recipients rework the imported material as they align with it. Second, the effects of coercive isomorphism are not evenly distributed. Bacsi's quantitative analysis suggests that poorer and less stable states carry the highest cost of a system where socio-economic position and #state_fragility predict heritage survival more strongly than the severity of environmental threats. Elliott and Schmutz add that the same procedures that produce convergence also produce regional inequality, because they assume a scientific and technical infrastructure that many African countries do not fully possess. Any honest defence of the current system has to account for this uneven burden. Third, community voice is at once the biggest promise and the weakest link in the current architecture. The literature on participation is optimistic and rich, and Lähdesmäki and colleagues show that a range of practices exists for shifting decisions towards communities and away from top-down authorities. But when community participation is folded into a donor-required document, the shift can be nominal rather than real. Robinson's discussion of S Gang Gwaay shows how a genuine #Indigenous_voice can point conservation practice in a direction that the international framework was not built to accommodate. A serious response to coercive pressure has to protect the space where such voices can shape the terms of engagement, not just the details of implementation. There are also implications for how we teach heritage. Students in conservation, museum studies, and heritage management often learn technical skills in workshops that assume the international framework as background. The framework is presented as best practice, and the political history behind it is often glossed over. This paper argues that students would benefit from being taught the framework as one option among several. That does not mean rejecting international standards. It means treating them as historically situated tools that have their own biases and limits. A further point deserves attention. Coercive pressure is often experienced by recipients as ambivalence rather than as opposition. They are grateful for the funding, they respect the expertise that comes with it, and they see the standards as broadly reasonable. At the same time they know that the standards impose a shape on their work that does not always fit local circumstances. Nardella and colleagues, in older but still cited work on Aksum in Ethiopia, described this as the paradox of only foreigners being able to do certain kinds of technical assistance, and the same phenomenon shows up in the current literature. Recipients are not simply resisting or complying. They are living with the tension. The discussion should also acknowledge the positive side of institutional pressure. Zhou's finding of a shared professional ethics across five very different systems is not only a story of homogenisation. It is also a story of global dialogue and mutual learning, in which practitioners in one country can trust that a peer in another country is working to a comparable standard. Chirikure's older but still influential warning about power imbalance at UNESCO World Heritage sites reminds us that the alternative to shared standards is not necessarily local flourishing. It can also be patronage, corruption, or neglect. The task is not to abolish shared standards, but to make them more responsive to the places they claim to serve. It is also worth being honest about the limits of the framework used in this article. Institutional isomorphism was developed to explain why organisations in the same field come to resemble one another, but it does not fully explain why some resemble each other more than others, or why some organisations manage to stay outside the pull of the field. The heritage sector contains many small community groups, religious foundations, and Indigenous custodianship arrangements that do not neatly fit the isomorphic model. These groups sometimes ignore international standards, sometimes selectively adopt parts of them, and sometimes build parallel systems that speak only to their own constituencies. A more complete theory of the field would need to account for these actors rather than treating them as leftovers from a modernising process. Recent work on participation and community engagement points in this direction, but there is still room for a more explicit theoretical account of how non-isomorphic actors survive in a field otherwise dominated by donor logic. Another area that deserves further attention is the relationship between coercive isomorphism and gender. Heritage funding decisions are often made in committees where women, especially women from recipient countries, are underrepresented. The kinds of heritage that these decisions favour, such as monumental architecture and grand landscapes, tend to reflect a particular gendered view of what is worth saving. Domestic craft traditions, textile knowledge, food practices, and oral histories that are often carried by women are less easily captured in the categories used by major donors. Chuva's work on sensitive pasts hints at this problem when it notes that global-to-local heritage policy carries ambiguities and can constrain rights, but the gender dimension deserves more explicit treatment in future studies. This paper cannot fill that gap, but flagging it is important. A further theme that would repay attention is the changing role of the Chinese state and other emerging donors in the international heritage system. As Wang and colleagues note, the social sciences have largely neglected donors other than UNESCO, including the World Bank, the Asian Development Bank, the European Union, USAID, and many others, which often collaborate with UNESCO or run their own bilateral and multilateral cultural initiatives. In the last decade, Chinese cultural aid and investment in heritage sites along the Belt and Road, together with growing programmes from Gulf states, have added new layers to the pattern. Whether these new donors will reproduce the same conditionality logic, or whether they will create genuinely different templates, is one of the open questions of the field. Zhou's finding that Chinese national museum grading operates as a coercive gatekeeper for public funding while still borrowing from the American model suggests that convergence with the existing framework is at least partial, even where political systems differ sharply. There is a policy dimension too. If coercive pressure is a fact of the field, then policymakers and donor staff should ask how it can be exercised more accountably. Mason and Martindale's call for a bespoke cultural heritage framework tailored to the private sector is one attempt to move in that direction. Farias's observation that heritage listing has become a #soft_power tool for states suggests that any reform of the framework will have to negotiate with governments that use listing for their own ends. Reform of the international system will therefore be slow, but small changes at the level of individual donor policies could make a difference, especially if they leave more room for translation and community-led design. A last observation concerns climate change. As Robinson's work on S Gang Gwaay makes clear, climate impacts are accelerating the degradation of many sites, and the coercive pressure to preserve their material fabric may not be sustainable in the new conditions. Bacsi's finding that fragile states with In Danger sites depend on international assistance to survive climate shocks suggests that donor conditionality will become more, not less, important in the coming decade. If that is true, then the design of donor conditions is a central heritage question for the 2020s and 2030s. It cannot be left to specialists in finance. There is also a research agenda that follows from the discussion. Comparative studies that pair sites receiving similar levels of international funding but embedded in different national systems would allow analysts to separate the effects of coercive pressure from those of local political culture. Longitudinal studies that follow the same site over ten or twenty years would show how much of the shape imposed by early donor engagement survives after the funding stops. Ethnographic studies of the offices where grant proposals are actually written would open up the black box of how donor language is learned and reproduced. Each of these designs is demanding, but each would help refine the picture given here, which is necessarily painted in broad strokes. Teaching also deserves a note. Heritage curricula in many universities still present international frameworks primarily as tools, and only rarely as objects of analysis in their own right. A course module dedicated to the political economy of heritage funding, drawing on the literature reviewed here, would help students recognise the pressures they will face as professionals. Practical exercises could ask students to read a real grant application, identify which parts of the text speak in donor language and which in local voice, and rewrite one section in a way that preserves the local voice while still meeting the donor's needs. This kind of exercise builds a skill that graduates will use for the rest of their careers. 6. Conclusion This article set out to show how #coercive_isomorphism operates in heritage funding, and how international grant requirements and NGO standards shape local preservation strategy. It has argued three points. First, coercive pressure travels through many parallel channels, from UNESCO listing through development bank lending and private foundations to NGO accreditation, and these channels reinforce one another because they share personnel and vocabulary. Second, the resulting convergence is real but uneven, with poorer and more fragile states carrying a heavier burden, and with community voice at risk of being reduced to a required box on a donor form. Third, recipients are not passive. They translate international scripts, sometimes align with them strategically, and sometimes resist them, especially when local or Indigenous values point in a different direction. For students, the practical takeaways are straightforward. When you read a heritage project document, notice the vocabulary. Ask which parts of the text are borrowed from international templates and which parts speak in a local voice. Ask who wrote the document and for whom. When you plan a project of your own, treat the donor's framework as a partner to be negotiated with, not as neutral background. Where possible, build in space for #community_voice at the design stage, not only at the implementation stage. For practitioners, the lesson is that legitimacy work matters, but it is not the same as good heritage work. Aligning with donor expectations may bring the money in, but if the alignment goes too deep it can hollow out the very thing the money was meant to protect. Balance is difficult, but it can be pursued deliberately by keeping #local_knowledge alive in the internal life of the organisation, even as the outer face of the organisation speaks the donor's language. For policymakers, the paper points to a need for more careful design of #donor_conditions. Bespoke cultural heritage frameworks, as proposed by Mason and Martindale, would be a step in the right direction, as would more explicit space for community-led criteria in UNESCO and development-bank processes. Bacsi's evidence that socio-economic position drives heritage outcomes as strongly as physical threat should be a wake-up call for anyone who assumes that the current system is neutral. Heritage is fragile. It depends on many small acts of care by many different people over long periods of time. International funding is often necessary to sustain those acts of care, but it also carries the risk of quietly redirecting them. Understanding the pressures at work is the first step to keeping the balance right. This paper has tried to give students a starting map. The next steps belong to the practitioners, communities, and policymakers who use it. References Bacsi, Z. (2026). The wealth gap in World Heritage: Economic disparity and state fragility as factors of World Heritage preservation. Sustainability and Heritage Studies. https://doi.org/10.3390/su18010001 Barnhart, D. (2020). The many faces of preservation funding: An examination of private American organizations supporting preservation abroad. Journal of Preservation Studies, 8(2), 45 to 72. Chuva, M. (2022). Heritage policies and sensitive pasts: Between ambiguities and rights from global to local. Portuguese Literary and Cultural Studies, 36 and 37, 106 to 121. Dupeyron, A. (2020). The cultural turn in international aid: Impacts and challenges for heritage and the creative industries. Journal of Cultural Economy, 13(6), 691 to 706. https://doi.org/10.4324/9780429279256 Elliott, M., and Schmutz, V. (2016). Diffusion and decoupling in the world heritage movement: Exploring global and local tensions in Africa. Poetics, 59, 45 to 62. Farias, D. B. L. (2022). UNESCO's World Heritage List: Power, national interest, and expertise. International Politics, 59(6), 1091 to 1109. https://doi.org/10.1057/s41311-021-00351-y Haugset, A. S., and Finne, H. (2023). Governing early childhood education and care quality development among diverse private ECEC providers in Norway. Education Inquiry, 15(4), 601 to 618. https://doi.org/10.1080/20004508.2023.2280291 Lähdesmäki, T., Turunen, J., Terian, A., and Garcia-Bardidia, R. (Eds.). (2025). Engaging communities in cultural heritage. Routledge. Leloup, M. (2020). The Culture Bank in West Africa: Cultural heritage and sustainable development. African Studies Review, 63(4), 812 to 831. Mason, A. R., and Martindale, A. (2023). Rethinking cultural heritage in the International Finance Corporation Performance Standards. Advances in Archaeological Practice, 11(3), 269 to 282. https://doi.org/10.1017/aap.2023.20 Moreau, D. (2021). Mimetic isomorphism in non-profit organisations: Sports associations in the Nord Pas-de-Calais departments. Societies, 11(3), Article 100. https://doi.org/10.3390/soc11030100 Robinson, S. (2025). S Gang Gwaay and World Heritage preservation in a time of climate change. International Journal of Heritage Studies, 31(4), 411 to 427. Toner, J., and Martins, J. (2021). Institutional isomorphism in collaborative, cross-cultural, project-based development work: An inquiry into the knowledge sharing behaviour of volunteers. Journal of Knowledge Management, 25(9), 2189 to 2207. https://doi.org/10.1108/JKM-08-2020-0620 Wang, S. L., Peycam, P. M. F., Hui, Y. F., and Hsiao, H. (Eds.). (2020). Heritage as aid and diplomacy in Asia. ISEAS Publishing. Zhou, Y. T. (2026). Institutional isomorphism in global museum quality assurance: A comparative analysis of five accreditation systems. Museum Management and Curatorship, 41(2), 145 to 168. Abdel-Rashid, S. E. (2020). The role of international cooperation in promoting sustainable development and supporting conservation and development policies at the national level. International Journal of Architectural Research and Planning of Sustainable Development, 4(1), 55 to 72. #coercive_isomorphism #heritage_funding #institutional_theory #UNESCO_World_Heritage #World_Bank_safeguards #IFC_Performance_Standards #NGO_accreditation #donor_conditionality #policy_transfer #community_participation #Global_South_heritage #cultural_diplomacy #preservation_strategy #legitimacy_pressure #decoupling_in_heritage

The last decade has seen an unprecedented loss of #built_heritage due to armed conflict, terrorism, climate stress, and unplanned urban growth. In response, a new field of practice has emerged that uses #photorealistic_3D modeling to bring back destroyed monuments in the digital space. This article reviews the main methods now used for such work, including terrestrial and aerial photogrammetry, terrestrial laser scanning, unmanned aerial vehicle surveys, historic building information modeling, neural radiance fields, and 3D Gaussian splatting. It also examines the ethical questions that these methods raise, such as authenticity, transparency, community consent, digital colonialism, and the long term care of digital assets. Case studies from Palmyra, Aleppo, Mosul, Bamiyan, Notre Dame, Kaesong, and the Sumy region of Ukraine are used to show how technical choices carry ethical consequences. The article argues that #digital_restoration is not a neutral technical act. It shapes memory, identity, and future physical restoration. Practitioners therefore need clear standards for #data_transparency, source documentation, and community involvement. The paper closes with practical recommendations for students and early career researchers working at the intersection of #digital_heritage and architectural conservation. #Sustainability, #FAIR_data, and #CARE_principles are highlighted as guiding frames for future work. Keywords: photorealistic 3D modeling; destroyed heritage; photogrammetry; laser scanning; neural radiance fields; HBIM; digital twin; ethics of restoration; digital colonialism; authenticity 1. Introduction Cultural heritage sites are being destroyed at a faster rate than at any earlier point in modern history. The list of losses in only the last decade includes the Temple of Bel and the Arch of Triumph in Palmyra, the Al Nuri Mosque and the Al Hadba minaret in Mosul, large parts of the Old City of Aleppo, the roof and spire of Notre Dame de Paris, and hundreds of civic and religious buildings across Ukraine. Each of these losses carries a wound that goes far beyond stone and wood. Buildings hold #collective_memory, they anchor identity, and they act as physical proof that a community existed in a particular place and time. When a monument is destroyed, that anchor is torn out. This is one reason cultural sites are so often targeted in war and terrorism. Attacks on symbolic buildings and cities have long been used to erase memory and to assert power, which is why safeguarding the architectural record has become an urgent global task. At the same time, the tools available for #digital_documentation have advanced in ways that would have seemed impossible only twenty years ago. Anyone with a smartphone, a light drone, or a modest budget for a laser scanner can now record a monument at a resolution that once required national mapping agencies to organize. What was true for the photogrammetric survey of Abu Simbel in the 1960s, when a temple was documented and relocated as the Aswan Dam rose, can today be done by a small team with off the shelf equipment. The tools have become smaller, faster, and cheaper, which has opened the door to #community_documentation and volunteer led projects around the world. This article looks at how these tools are used, and how they should be used, for the #digital_restoration of #destroyed_heritage. The central question is not only what can be done, but what should be done. #Photorealistic_reconstruction is powerful. A well made virtual copy can carry emotional weight, teach architectural history, guide physical rebuilding, and preserve knowledge for future generations. But a photorealistic image can also mislead. It can hide uncertainty, present speculation as fact, and be used to advance political or commercial goals that do not serve the community that lost the site. The very same technique can produce a scholarly reference model or a marketing image, and the difference is not always visible from the outside. The article is written for students and early career researchers who are entering this field, and it is intended to help readers see both the technical and the ethical picture at once. It is organized into six main parts. Section 2 sets the context by explaining what is meant by lost architecture and by digital restoration. Section 3 reviews the current #technological_methodologies, from classic photogrammetry to the latest #neural_radiance_fields and #Gaussian_splatting. Section 4 presents case studies where these tools have been applied. Section 5 turns to ethics and discusses authenticity, transparency, digital colonialism, community consent, and stewardship. Section 6 offers a discussion of common tensions and trade offs. Section 7 gives conclusions and practical recommendations. The goal throughout is to describe a mature #research_practice that treats technology as a means and not as an end. 2. Background and Scope 2.1 What counts as lost architecture Lost architecture is a broader category than it may first appear. It clearly includes buildings that have been completely destroyed, such as the Temple of Bel in Palmyra, which was demolished in the summer of 2015 during armed conflict and is now studied almost entirely through digital reconstructions built from surviving images. It also includes buildings that suffered partial destruction but whose original state can only be seen through digital methods, such as Notre Dame de Paris after the 2019 fire, or the Al Hadba leaning minaret in Mosul, destroyed in 2017. Beyond conflict, the category includes structures lost to fire, earthquake, flood, and slow decay, as well as buildings demolished for urban development. A recent Polish study of the Koprowka villa in Celestynow illustrates this last case, where a culturally significant property disappeared from the village and was reconstructed as an HBIM model from LiDAR data of the site, preserved window frames, archival photographs, and documents. The category also includes buildings that were designed but never built, or that existed only in early states and were later transformed. These belong to the wider question of #digital_reconstruction as a scholarly practice, and they raise many of the same ethical questions as physically destroyed buildings. 2.2 What counts as digital restoration Digital restoration is a family of practices, not a single method. At one end it includes precise geometric documentation of surviving fragments. At the other end it includes speculative visualization of buildings that no longer exist. In between lie many hybrid workflows that combine survey data, historic photographs, drawings, and expert interpretation. Recent scholarship stresses that the word restoration should be used with care. As one team argues, the expansion of digital heritage practices has blurred the boundaries between restoration, reconstruction, simulation, and imaginative projection, and this blurring has weakened the philosophical basis of restoration as a discipline. The same authors turn to Cesare Brandi's classical theory of restoration, and especially the idea of #unita_potenziale or potential unity, to argue that even without physical intervention a digital restoration must meet the same ethical and philosophical standards as any other form of restoration. This position matters for how we design and communicate digital models. A photorealistic image that mixes measured survey with speculative modeling is not automatically wrong, but it must be honest about what is known and what is guessed. A study of the three pagodas at the Mireuksa Temple site in Korea makes this point directly. It concludes that authenticity cannot be defined only by material survival but must also be judged through the ethical and transparent communication of historical uncertainty, and it proposes a framework of #memory_authenticity for cases where architectural evidence is fragmentary or lost. 2.3 Why the field is urgent The pace of loss is not slowing. Cultural heritage worldwide is at rising risk from #climate_change, armed conflict, urbanisation, illicit trafficking, and the broader forces of globalisation, and these threats produce irreversible loss of both tangible and intangible assets. In parallel, the technology for capturing sites is improving quickly, which means that every year of delay is a year of missed opportunity. Practitioners in the digital humanities describe the situation as a race against time to train and equip enough people to record irreplaceable sites before more of them are lost to disaster or conflict. For students entering the field, the practical message is simple. There is no shortage of urgent work. 3. Technological Methodologies This section reviews the main technologies now used for #photorealistic_3D reconstruction of destroyed heritage. No single method is best for every case. The right choice depends on what survives, what images and drawings exist, what accuracy is needed, and what will be done with the model. 3.1 Close range and terrestrial photogrammetry Photogrammetry is the oldest of the digital methods and remains the most widely used. It works by extracting geometry and texture from a set of overlapping photographs. Modern software based on structure from motion and multi view stereo pipelines can turn hundreds of photographs into a dense point cloud, then into a textured mesh. The same textures can be transferred back onto the geometry to produce a #photorealistic_model. For destroyed sites, terrestrial photogrammetry has a special role because it can work with historical images taken before destruction. In Palmyra, for example, photogrammetric workflows were built from crowdsourced tourist photographs collected from social media and web search engines, and these were then processed through advanced photogrammetric pipelines to build 3D models of monuments that no longer exist. In Ukraine, following the 2022 full scale Russian invasion, photogrammetry has been used to build 3D models of landscapes and lost cultural heritage objects in the Sumy Region, using archival maps and imagery to design a reliable base for reconstruction and analyzing the integration of photogrammetry with LiDAR, thermal imaging, artificial intelligence, and augmented reality. Photogrammetry remains the standard in restoration work because it produces reliable and high resolution textures. Its main limits arise with reflective, transparent, or textureless surfaces, where it can fail to recover a stable geometry. In such cases it needs to be combined with other methods. 3.2 Aerial and UAV photogrammetry Unmanned aerial vehicle photogrammetry has transformed what a small team can record. UAV based imagery supports high resolution mapping, 3D modeling, monitoring, risk assessment, and preservation, and it has become a cost effective way to document large and complex sites. In the workflow, an operator plans a flight path, the UAV captures overlapping images, and the images are processed through structure from motion and multi view stereo to produce dense point clouds, digital surface models, and orthophotos. Case studies across cultural heritage sites in Greece show how UAV based 3D modeling can be applied to sites of varying complex geometry, and how the resulting data can be integrated with LiDAR, remote sensing imagery, GIS, and BIM. For endangered or already damaged sites, aerial photogrammetry is often the only realistic way to capture roofs, tall towers, and inaccessible zones. It has been used at churches in the Yaroslavl region of Russia, where UAV imagery combined with geodetic measurements and architectural survey produced dense point clouds and polygonal models of interior and exterior spaces at scales of 1 to 50 or 1 to 200. 3.3 Terrestrial laser scanning and LiDAR Terrestrial laser scanning, often abbreviated as TLS, and mobile LiDAR complement photogrammetry with high geometric accuracy. A TLS instrument fires millions of laser pulses per second and records the return time and intensity, producing a dense colored point cloud that can capture facades, interiors, and complex ornament with millimeter accuracy. A recent practice based guide argues that TLS plays a pivotal role in capturing detailed and accurate representations of cultural heritage, and it addresses gaps in streamlined practice through case studies, expert interviews, and a best practice guide grounded in field experience. The same guide notes that TLS data sets integrate well with BIM, virtual reality, and augmented reality, which supports preservation, tourism, education, and interpretation. Integrated surveys that combine TLS with high resolution imagery and GIS have been applied to endangered architectural heritage in Iraq. Work on the Shanasheel house in Al Basrah, known for its intricately decorated wooden facades, uses TLS to record facades within a 3D environment and then integrates the results with GIS platforms for spatial data management and risk assessment. In Kurdistan, the Mudhafaria minaret in Erbil has been reconstructed by photogrammetric analysis from hundreds of photographs, producing a detailed 3D model that supports both conservation and historical documentation. For destroyed structures where nothing physical remains to scan, LiDAR has an indirect role. It can be used to document the site as it now stands, capture surviving foundations or fragments, and provide a #geospatial_anchor onto which reconstructions from photogrammetry and archival sources can be aligned. In the case of Koprowka in Poland, LiDAR survey data of the empty site was combined with preserved window frames, archival photographs, and documents to build an HBIM model of a villa that was demolished twenty years earlier. 3.4 Neural radiance fields and 3D Gaussian splatting Neural radiance fields, known as NeRF, and 3D Gaussian splatting are the most recent additions to the toolkit. Both belong to the family of #neural_rendering methods and both offer photorealistic scene reconstruction from unstructured image sets, which makes them attractive for heritage projects that rely on mixed image archives. NeRF represents a scene as a continuous volumetric function learned by a neural network, while 3D Gaussian splatting represents a scene as a large collection of oriented 3D Gaussians that can be rendered very quickly. Recent comparative studies show a mixed picture. A comparison in Palermo between NeRF and 3D Gaussian splatting for a noble chapel in the Cemetery of Santa Maria dei Rotoli used both source algorithms and the Luma AI web application and concluded that the methods are especially interesting for the visualization of heritage starting from unconventional acquisitions such as video. A comparative evaluation across sculptures, design objects, and architectural buildings suggests that NeRF and 3D Gaussian splatting are promising alternatives to photogrammetry precisely in cases where reflective, transparent, or textureless surfaces limit classical pipelines. NeRF and Gaussian splatting are also being explored in underwater archaeology, where visibility and access are limited, for reconstructions of the ancient city of Dwarka in Gujarat. For destroyed heritage, the appeal of neural rendering is that it can produce visually convincing images from crowdsourced photographs even when calibration is imperfect. The risk is that photorealistic appearance can outrun geometric accuracy. A model that looks like a photograph is not necessarily correct in scale or in the placement of details. Users of these methods need to test the model against measurable ground truth wherever possible. 3.5 Historic Building Information Modeling Historic Building Information Modeling, known as HBIM, extends conventional BIM into the heritage domain. HBIM enables digital mapping, documentation, analysis, and management of historic architecture, and it supports the integration of survey data such as laser scanning and photogrammetry. A recent review describes HBIM as an emerging field that extends BIM to preservation, management, and analysis of heritage structures, and stresses the importance of good data selection and management for both geometric and non geometric information such as historical and architectural knowledge. The same review notes that HBIM has moved rapidly toward heritage preventive preservation, and highlights the need for data interoperability, standardized protocols, and real time structural analysis if HBIM is to reach its full potential. HBIM has been applied directly to lost heritage in conflict zones. A study of the Al Hadba minaret in Mosul used freely available web based images processed in three photogrammetry packages, scaled with onsite measurements of the surviving base, and imported into a BIM authoring environment to produce a #HBIM_model. The work revealed challenges that recur in every conflict zone, including sparse documentation, conflict between historical sources, constraints on crowdsourced data, and their effects on point cloud quality, and it highlights HBIM as an accessible baseline approach for cultural recovery in resource constrained settings. HBIM is not only a modeling framework. It is also a container for knowledge. Levels of detail, digitization methods, and software choices all shape what future researchers and conservators can do with the model. Introductions to HBIM tools stress that historic components are heterogeneous, complex, and irregular, and that BIM libraries built for new construction cannot be applied directly to heritage. Tools and workflows must be adapted or reinvented for HBIM. 3.6 Digital twins A digital twin is more than a static model. It is a living digital counterpart of a physical site, linked to data streams such as environmental sensors, structural monitoring, and visitor flows. Digital twins are increasingly used to preserve ancient and historical sites for architectural restoration, and recent projects have used AI assisted drones with LiDAR photogrammetry and AI driven path planning to build high resolution models of archaeological sites in several countries. Projects on structural integrity protection show how digital twins can support long term conservation planning for heritage buildings. The Notre Dame de Paris digital twin project shows the depth of what is possible. After the 2019 fire, a multidisciplinary task force gathered documentary and analytical resources on the cathedral and worked on an innovative digital ecosystem to bring together, analyze, and correlate multiple levels of reading toward the production of new knowledge about the cathedral, its transformations over time, its materials, and the alterations produced by the fire. The project treats the twin not as a display but as a research instrument that supports scientists from many fields, and it shows how #semantic_awareness can be built into 3D data through knowledge engineering and shape analysis. More recent work extends the digital twin idea further. A framework called the Memory Twin builds on Holistic Historic Building Information Modeling and enriches digital heritage environments with narrative, emotional, and community driven content, integrating high fidelity visuals with Paradata, Metadata, Data, and intangible cultural values. This work is explicitly aligned with the FAIR and CARE principles and treats digital twins as a paradigm shift from static documentation toward participatory, value driven preservation. 3.7 Crowdsourced imagery and low cost photogrammetry For destroyed sites where no professional survey exists, crowdsourced imagery has become essential. A study on mass photogrammetry frames this as a democratised approach that turns the art of photogrammetry into a more mechanical practice, so that anyone with inexpensive equipment and basic knowledge can achieve acceptable results, and it reports the experience of acquiring more than 300 3D models from over 25 priority sites and museums in East Asia. Similar work has focused on digitizing city scale heritage in Erbil in the Kurdistan Region of Iraq, using smartphone cameras paired with Agisoft Metashape and structure from motion photogrammetry to produce reconstructions that match professional grade equipment to within a few millimeters for artifacts and a centimeter or two for architecture. These low cost pipelines democratize access to preservation and empower local communities, researchers, and educators. For Aleppo, one of the world's oldest cities, a Syrian and Italian collaboration used photographic tours from before the war together with post war surveys to document walls, towers, mosques, and minarets, and to prepare for physical restoration. The reconstruction plan for the Great Umayyad Mosque minaret uses old photographs to help identify stones from the pre war structure and to place them back in their original positions, an example of how #digital_documentation can guide anastylosis and physical restoration. Crowdsourcing raises technical challenges. Public images vary in resolution, angle, exposure, and metadata. They may cover only tourist facing views and miss less photogenic elements. When lens distortion, calibration data, and time stamps are missing, geometric accuracy is harder to guarantee. This is one place where NeRF and Gaussian splatting show promise, since they can tolerate a wider range of image conditions than classical pipelines, but it also means that quality control and validation are needed before such models are used for restoration decisions. 4. Case Studies 4.1 Palmyra Palmyra, the Syrian oasis known as the Bride of the Desert, is now the textbook case for #digital_restoration of destroyed heritage. Between 2015 and 2017, the Temple of Bel, the Temple of Baalshamin, the Arch of Triumph, and several funerary towers were destroyed. Since then, more than a dozen research groups have used photogrammetry, laser scanning where possible, historical drawings, and archival photographs to build digital models of the site. Some projects focus on scholarly reconstruction. Others aim at immersive virtual experience. A recent study reconstructs the Temple of Bel from crowdsourced visitor images and delivers the model as a Unity based virtual environment for a VR headset, giving off site visitors a walk through experience of a place that no longer exists on the ground. Palmyra has also been rebuilt in a metaverse environment. A project on the Palmyra theater used a comprehensive survey of images, sketches, textual documents, and artifacts to reconstruct the Roman era theater in a virtual environment, allowing users to explore the site in its original form and demonstrating both the technical challenges of optimizing model complexity for metaverse platforms and the ethical significance of preserving cultural heritage in the face of conflict. Related work has used intelligent reality applications as a new form of #post_trauma_reconstruction, treating the digital model as part of the process of collective recovery. Not all writing on Palmyra is celebratory. Critical scholarship warns that some digital reconstructions of Syrian heritage illustrate a Western saviorism that has colonial roots, and that digital innovation can further exacerbate the unequal distribution of knowledge and power. A comparison between the digitally produced replica of Palmyra's Arch of Triumph and the postwar reconstructions of Warsaw and the Hollandsche Schouwburg in Amsterdam shows how digital and non digital memory technologies can carry very different politics. This critique is a caution against treating any digital reconstruction as a purely technical exercise. 4.2 Aleppo The Old City of Aleppo suffered heavy destruction during the Syrian civil war. Photogrammetric work carried out before the war by Italian teams, combined with post war surveys, has been essential to plan physical reconstruction. Documentation of Citadel walls, the entrance tower, the southern tower, a mosque, and the Citadel mosque minaret, followed by the identification of stones from the destroyed minaret of the Great Umayyad Mosque, shows how #spherical_photogrammetry can support both digital preservation and hands on rebuilding. About forty percent of the outer face stones of that minaret have been identified with the help of high resolution photographs from 2010, and it is hoped that recognition can reach seventy percent. This case shows the practical link between photorealistic documentation and material continuity in restoration. 4.3 Mosul The Al Hadba leaning minaret and the Al Nuri Mosque in Mosul were destroyed in 2017 during the campaign against the so called Islamic State. As with Palmyra, digital efforts began quickly. The HBIM study of the Al Hadba minaret drew on freely available web images from before destruction, compared three photogrammetry packages, and produced a scaled point cloud and HBIM model that revealed both the promise and the limits of this approach in conflict zones. The study documents very concrete challenges. Historical sources sometimes conflict about geometry and materials. Crowdsourced photographs are unevenly distributed across the monument. Some views are missing entirely. These constraints must be recognized in any subsequent decision about physical rebuilding. 4.4 Bamiyan The Buddha figures of Bamiyan in Afghanistan were destroyed in March 2001. More than fifteen years later, the UNESCO safeguarding campaign for the preservation of the property continued to progress, and the international community was asked to contribute proposals for the future presentation of the physical remains, including options of reassembling the original fragments. In this setting, immersive virtual reality technologies have been used to investigate how values attributed to cultural heritage are produced and experienced among different stakeholder groups, which is a role that goes beyond visualization and enters the sphere of participatory heritage decision making. The Bamiyan case is important because it shows how digital models can support decisions about whether and how to rebuild in the physical world, even when a full physical restoration is not possible. 4.5 Notre Dame de Paris Notre Dame de Paris burned in April 2019. The fire destroyed the roof and the nineteenth century spire. Digital documentation carried out over many years before the fire has proven crucial to the restoration, and the semantic aware digital twin project shows how a large scale collaboration can turn scattered documentation into a research platform. Public facing outputs, including a VR documentary that used footage filmed weeks before the fire and access to the interior after it, have helped audiences follow the restoration and keep the monument alive during the works. Together these projects show that a well documented monument becomes far more resilient than an undocumented one, even when a disaster does occur. 4.6 Ukraine The war in Ukraine has produced a new wave of urgent documentation. The full scale invasion in 2022 has made preservation of cultural heritage urgent, and photogrammetric work in the Sumy Region uses archival sources, cartographic analysis, and integration with LiDAR, thermal imaging, artificial intelligence, and augmented reality to build a digital foundation for future reconstruction of destroyed objects. This case is also a reminder that digital restoration is not only a post war activity. It is a wartime response. 4.7 Kaesong and Mireuksa Two cases from Korea illustrate the philosophical and ethical dimension of digital restoration. The Kaesong Manwoldae Digital Restoration Platform was designed to distinguish clearly between conservation and reconstruction, and it takes Cesare Brandi's restoration theory and the concept of unita potenziale as its philosophical anchor, arguing that authenticity must be conveyed to users even in a fully digital environment. The study of the Mireuksa Temple pagodas compares three cases at one site, including a West Pagoda that shows that evidence based restoration can maintain material and historical authenticity, an East Pagoda whose reconstruction based on limited evidence tends to produce visually convincing but historically uncertain forms, and a Wooden Pagoda whose digital mediation can effectively communicate lost heritage but only when its speculative nature is transparent to the public. The Mireuksa work proposes a concept of memory authenticity to help practitioners handle fragmentary evidence responsibly. 5. Ethical Considerations The technical picture is only half of the story. Every choice in a digital restoration project has ethical weight, from the selection of the site to the choice of software to the final decision about who can see and reuse the model. This section addresses the main ethical questions the field faces today. 5.1 Authenticity and the risk of visual overconfidence The first question is #authenticity. A photorealistic image looks authoritative even when it is speculative. When a viewer sees rendered stone, painted plaster, and shifted sunlight on a virtual facade, they are inclined to believe what they see. This is a well known risk in digital heritage. The Mireuksa work notes that authenticity cannot be defined only by material survival and must be judged through the ethical and transparent communication of historical uncertainty, and that digital mediation can effectively communicate lost heritage only when its speculative nature is transparent to the public. The Kaesong platform argues that even without physical intervention, digital restoration must meet the ethical and philosophical standards inherent to restoration, and it proposes concrete strategies for implementing this standard. The practical response is not to avoid photorealism. It is to pair photorealistic output with clear #paradata that record the sources, the reasoning, and the level of confidence for each part of the model. The London Charter for the Computer Based Visualisation of Cultural Heritage and the Seville Principles have established this expectation at the international level, and Memory Twin approaches operationalize it by integrating high fidelity visuals with Paradata, Metadata, Data, and intangible cultural values. Students entering the field should treat paradata as a required output alongside the model itself. 5.2 Transparency about method and uncertainty Related to authenticity is the question of transparency. Every reconstruction combines observation, inference, and imagination. Users deserve to know what came from where. Was the roof line measured from surviving remains or estimated from photographs. Were textures painted by an artist or derived from historic paintings. Were interior spaces reconstructed from drawings or invented for visual completeness. These distinctions can be visualized inside the model, for example through color codes that separate measured, inferred, and hypothetical zones, and they can be documented in the accompanying metadata. Frameworks such as HHBIM and the Memory Twin are moving in this direction and are aligned with the FAIR principles for data management as well as the CARE principles for indigenous data governance. 5.3 Digital colonialism and community consent A more difficult and less discussed question is who gets to reconstruct whose heritage. Recent scholarship has raised concerns about #digital_colonialism, the pattern where technologically well resourced institutions in the global north produce digital models of sites in the global south and then control the resulting datasets. Comparative research argues that some digital reconstructions of Syrian heritage illustrate a Western saviorism emblematic of the colonial roots of archaeology and heritage practices, and that digital innovation further exacerbates the unequal distribution of knowledge and power. A critical study of a 3D printed reconstruction of the Arch of Triumph in Palmyra concludes that the project failed on several ethical grounds, including that it did not address human loss, that its material and scale were inaccurate, that it was patented by the producing organization with limited public and digital access, and that it promoted a culture of quickly reconstructing destroyed heritage without regard for context or current community needs. An overview of opportunities and challenges in digitally transforming World Heritage at 50 plus years frames the ethical picture broadly. Many actors have not considered the ethical implications of digital assets for the communities where sites are located, examples of digital colonialism or appropriation exist in the field, and FAIR and CARE principles need adoption in order to improve the effectiveness and correctness of digital technologies used for World Heritage. The same argument extends to the requirement that digital assets be respectful, sustainable, transferable, and ethically correct, and that they benefit rightsholders and contribute to the protection of the sites themselves. Community consent is not a formality. It shapes what data are collected, how they are stored, who can view them, and how they may be used. For heritage sites tied to living communities, especially those affected by conflict, meaningful participation in the design of the project is essential. 5.4 The politics of reconstruction Digital reconstruction is not politically neutral, especially in post conflict settings. In some cases, external funders push for the restoration of monuments while local populations still lack housing, water, and security. In others, digital projects can become tools of soft power that showcase the technical prowess of the sponsor rather than serving the community that lost the site. The ethical response is not to abandon digital work in these contexts. It is to insist that projects be led or co led by local institutions, that they include training and knowledge transfer, and that the resulting data belong first to the community. 5.5 Ownership, access, and intellectual property Once a photorealistic model exists, questions of ownership and access become sharp. Who owns the point cloud captured on a public site. Who can license a mesh generated from crowdsourced photographs. Who controls the immersive experience built on top of it. Studies of mass photogrammetry raise these questions directly, noting that the democratised approach has implications for public interest in digital heritage in countries where museums and cultural institutions have little access to digital teams, provided that intellectual property issues are cared for. Digital assets should be treated as public goods where possible, with open licenses that respect the cultural context and community wishes. 5.6 The long term care of digital assets A model that cannot be opened in twenty years is a model that has not really been preserved. #Digital_stewardship therefore has a temporal dimension. File formats, software dependencies, and hosting infrastructure change quickly. Digital assets should transcend longer periods so that future generations can enjoy and reuse them. Long term care requires open formats, clear metadata, redundant storage, and institutional commitment. HBIM research emphasizes that data interoperability, standardized protocols, and real time structural analysis are essential to make HBIM widely effective in conservation efforts. Sustainable stewardship is the difference between a project that supports future physical restoration and a project that becomes an orphan file after a graduate student leaves. 5.7 Ethics of immersive experience Immersive virtual and augmented experiences can be powerful teachers. They can also flatten history. When a user walks through a metaverse Palmyra, the experience of loss can be softened into spectacle. Projects that reconstruct the Palmyra theater in the metaverse acknowledge this tension and frame the work as evidence of a duty to preserve cultural heritage in the face of conflict and destruction, but they also require careful design to avoid producing consumer style content that trivializes the loss. VR documentaries like those made about Notre Dame after the 2019 fire show how immersive media can foreground human emotion and the personal stories of those who decide the future of an iconic building. The lesson is that the shape of the experience matters as much as the accuracy of the underlying model. 6. Discussion 6.1 Choosing methods for the job The main technologies described in Section 3 are not mutually exclusive. In practice, most serious projects combine several methods. Photogrammetry provides rich textures. Laser scanning provides geometric accuracy. UAV imagery supports large coverage and inaccessible zones. Neural rendering offers photorealism from mixed image sets. HBIM structures the resulting geometry into an editable and queryable model. Digital twins connect the model to living data. UAV photogrammetry has grown into a fully integrated framework combining structure from motion, multi view stereo, LiDAR, machine learning, GIS, and BIM to serve varied heritage geometries. The choice of method should follow the question, not fashion. For destroyed sites specifically, three points guide method selection. First, the type of available source material matters. If archival photographs exist, terrestrial photogrammetry and neural rendering are strong candidates. If only foundations survive, LiDAR and UAV mapping can capture what remains and provide anchors for reconstruction. Second, the intended use matters. A model for structural analysis has different requirements from a model for public interpretation. Third, the level of certainty matters. If the reconstruction will guide physical rebuilding, geometric accuracy and full paradata are essential. If it will support a museum exhibit, narrative and legibility become central. 6.2 Balancing photorealism and honesty The rise of neural rendering has raised the visual quality of digital models and has made it easier to produce results that look like photographs. This is both an opportunity and a risk. The opportunity is that photorealism can carry meaning that measured drawings cannot. It can help a visitor feel the scale of a lost interior or the color of a lost fresco. The risk is that photorealism can hide uncertainty. The methodological response is to build layers into the model. Levels of certainty can be encoded visually, alongside the photorealistic surface, so that a user can toggle between the confident and the speculative. This resonates with Brandi's idea of unita potenziale, which encourages a restoration that preserves the potential wholeness of the object while making the intervention legible. 6.3 Documentation as a form of insurance The Notre Dame case shows that thorough digital documentation carried out before disaster is one of the strongest forms of protection. A well documented monument becomes far more resilient than an undocumented one. The Notre Dame digital twin project exists because scientists had already gathered substantial data about the cathedral, and it now serves as a research platform for the restoration. Digital twin work in the ancient and historical world stresses that these methods can virtually preserve sites from natural and societal threats that will continue to occur. The policy implication is that funding for #preventive_documentation of at risk sites is a cheap form of insurance compared to the cost of reconstruction after loss. 6.4 Training and equity The tools have become cheaper but expertise has not. A recent case study of Erbil shows that smartphone cameras and open source or affordable software can produce models with millimeter accuracy for artifacts and centimeter accuracy for buildings, but this outcome depends on a well designed pipeline and trained users. Similar work at Chiesa dei SS Apostoli e Biagio in Florence proposes a scalable model that leverages widely available 360 cameras and photogrammetry with LiDAR to capture heritage sites and to train non specialists in best practice. The challenge is to spread this training in regions where sites are most at risk. #Capacity_building is therefore as much a technical priority as software choice. 6.5 Standards and interoperability Photorealistic 3D models are only useful in the long term if they can be read, reused, and combined by future practitioners. The field is not yet mature in this respect. HBIM reviews emphasize that data interoperability, standardized protocols, and real time structural analysis are essential to make HBIM widely effective in conservation efforts. Similar arguments apply to point cloud formats, mesh formats, metadata schemas, and paradata standards. Alignment with international guidelines and with the FAIR and CARE principles is not optional. It is the condition for the field to accumulate knowledge across projects and generations. 6.6 The ethical grammar of virtual restoration Ethics is not a checklist that comes at the end of a project. It runs through every stage. It shapes how a site is selected, how the community is engaged, which methods are used, how uncertainty is communicated, which data are open, and how long the results are cared for. A useful mental model for students is to think about three levels of ethics at once. The first is the ethics of the object, that is, respect for the historical evidence and material integrity. The second is the ethics of the community, that is, respect for the people connected to the site. The third is the ethics of the audience, that is, honesty in how the reconstruction is presented and used. A mature project addresses all three. 6.7 Emerging role of artificial intelligence Artificial intelligence has begun to touch every stage of the digital restoration workflow. Machine learning is used inside modern photogrammetric pipelines for image matching and dense reconstruction, and it appears in NeRF and Gaussian splatting as the core of the representation itself. Neural networks are also being trained on photogrammetric survey data to fill in lost architectural and picturesque elements of structures, treating documentation of endangered sites as training material for generative reconstruction. Digital twin research in the ancient and historical world has combined AI assisted drones with LiDAR photogrammetry and AI driven path planning to record archaeological sites more accurately and in a fraction of the time required by earlier methods. AI in this field is not only a technical accelerator. It also raises new ethical questions. When a neural network fills in a missing capital or reconstructs a corner of a facade for which no photographs exist, the result is not documentation. It is inference. If the inference is not clearly marked, viewers cannot distinguish the model output from the historical record. The response is the same as for other speculative work. Every AI generated element should be labelled as such in the paradata, and the training data should be recorded so that future scholars can trace how the model reached its result. #Explainable_AI is not a luxury in this context. It is a condition for the field to remain trustworthy as generative tools become more common. 7. Conclusion and Recommendations Photorealistic 3D modeling of lost architecture has moved in a few years from a niche research topic to a global practice that touches conservation, education, tourism, security, and public memory. The tools are powerful and are still improving. Photogrammetry, LiDAR, UAV surveys, HBIM, neural radiance fields, 3D Gaussian splatting, and digital twins together form a rich #digital_toolkit that can capture surviving fragments, reconstruct lost forms, and share the results with wide audiences. This capacity is real, and it is already helping communities in Syria, Iraq, Afghanistan, Ukraine, France, Poland, Korea, Italy, and many other places to hold on to something in the face of loss. But the capacity alone is not enough. The field needs a shared #ethical_grammar that guides how these tools are used. The main threads of that grammar are already present in the literature. Authenticity should be treated as memory authenticity as much as material authenticity, and it should be linked to transparent communication of uncertainty. Restoration in the digital domain should meet the same philosophical standards as physical restoration, and it should draw on the classical theory of the discipline, including Brandi's idea of potential unity. Paradata should be produced alongside every photorealistic model. Community consent should be a genuine part of every project, not a formality. FAIR and CARE principles should be the default for data management. Long term stewardship should be planned from the start, not bolted on later. The community of practice should invest in capacity building in the regions where sites are most at risk. These are not exotic requirements. They are the working conditions for a mature #digital_heritage practice. For students and early career researchers entering the field, five practical recommendations follow from this review. First, learn several methods rather than one. A project that only knows photogrammetry will hit its limits at reflective surfaces. A project that only knows neural rendering will hit its limits at geometric accuracy. Fluency across methods is a strong indicator of professional readiness. Second, treat documentation of sources and reasoning as part of the deliverable. If your model does not come with clear paradata, it is not finished. Third, work with communities rather than for them. Local knowledge, oral history, and cultural protocols are not decoration. They are part of the record. Fourth, plan for the long term from day one. Use open formats. Publish metadata. Deposit data in a long lived repository. Assume that in twenty years someone else will need to open your files. Fifth, be honest about what you do not know. A model that says visibly, this part is uncertain, is more trustworthy than a model that pretends to certainty. Honesty is not a weakness in digital heritage. It is the practice. Photorealistic 3D modeling will not bring back what has been destroyed. But it can preserve knowledge, guide physical restoration, support education, and hold space for future generations to make their own decisions about their heritage. Used with care, it can be one of the most humane responses available to a world that keeps losing its buildings. References Alkhatib, Y. J., Forte, A., Bitelli, G., Pierdicca, R., and Malinverni, E. (2023). Bringing Back Lost Heritage into Life by 3D Reconstruction in Metaverse and Virtual Environments: The Case Study of Palmyra, Syria. Applied Sciences. Al Muqdadi, F., and Ahmed, A. (2022). Applying Heritage Building Information Modelling (HBIM) to Lost Heritage in Conflict Zones: Al Hadba minaret in Mosul, Iraq. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLIII B2 2022, 753 760. doi:10.5194/isprs archives XLIII B2 2022 753 2022. Avila, F., Blanca Hoyos, A., Puertas, E., and Gallego, R. (2024). HBIM: Background, Current Trends, and Future Prospects. Applied Sciences, 14(23), 11191. doi:10.3390/app142311191. Basso, A., Condorelli, F., Giordano, A., Morena, S., and Perticarini, M. (2024). Evolution of Rendering Based on Radiance Fields. The Palermo Case Study for a Comparison Between NeRF and Gaussian Splatting. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLVIII 2 W4 2024, 57 64. doi:10.5194/isprs archives XLVIII 2 W4 2024 57 2024. Borkowski, A., and Winiarska, W. (2025). The HBIM Model as a Source in the Building Reconstruction Process: A Case Study of the Koprowka in Celestynow, Poland. Buildings, 15(9), 1442. doi:10.3390/buildings15091442. Carmeliti, M., and Marziali, S. (2026). A comparative evaluation of 3D reconstruction with photogrammetry, NeRFs and 3D Gaussian Splatting in Cultural Heritage Restoration. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLVIII 2 W12 2026, 89 96. Cassar, A., Baker, D., and Ioannides, M. (2025). From Digital Twin to Memory Twin: A Holistic Framework for Cultural Heritage Documentation, Interpretation, and Adaptive Reuse. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Ch'ng, E., Cai, S., Zhang, T. E., and Leow, F. T. (2019). Crowdsourcing 3D cultural heritage: best practice for mass photogrammetry. Journal of Cultural Heritage Management and Sustainable Development. Denker, A. (2022). Palmyra in Context: An Intelligent Reality Application to Revive the Victimized Bride of the Desert. In Digital Heritage. Progress in Cultural Heritage: Documentation, Preservation, and Protection. Duindam, D. (2023). Rebuilding Cities in a Digital Age: The Destruction and Reconstruction of Heritage in Palmyra, Warsaw, and Amsterdam. Krisis: Journal for Contemporary Philosophy. doi:10.61299/12e_l51. Fangi, G. (2019). Aleppo Before and After the War 2010 2018. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII 2 W15, 449 456. doi:10.5194/isprs archives XLII 2 W15 449 2019. Hedrick, E., Hatch, D., and Oliver, E. (2025). Digital Twins in the Ancient and Historical World. Proceedings of the International Conference on Digital Heritage. Hutson, J., Weber, J., and Russo, A. (2023). Digital Twins and Cultural Heritage Preservation: A Case Study of Best Practices and Reproducibility in Chiesa dei SS Apostoli e Biagio. Art and Design Review, 11. Ioannidis, C., Verykokou, S., Soile, S., and Potsiou, C. (2025). UAV photogrammetry for the protection of cultural heritage: methodologies, case studies, and applications. Heritage. Khunti, R. (2018). The Problem with Printing Palmyra: Exploring the Ethics of Using 3D Printing Technology to Reconstruct Heritage. Studies in Digital Heritage, 2(1). Lee, J., Park, J., Park, Y., Chin, Z., and Kim, M. (2025). A Study on the Concept of Restoration through the Kaesong Manwoldae Digital Restoration Platform: Focusing on Cesare Brandi's Restoration Theory and the Concept of Unita. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, X M 2 2025, 209 216. doi:10.5194/isprs annals X M 2 2025 209 2025. Liu, J., Willkens, D. S., and Foreman, G. (2022). An introduction to technological tools and process of Heritage Building Information Modeling (HBIM). EGE Revista de Expresion Grafica en la Edificacion. Liu, J., Willkens, D. S., and Gentry, R. (2025). Developing a Practice Based Guide to Terrestrial Laser Scanning (TLS) for Heritage Documentation. Heritage, 8(8), 313. doi:10.3390/heritage8080313. Luca, L. De (2020). Towards the Semantic aware 3D Digitisation of Architectural Heritage: The Notre Dame de Paris Digital Twin Project. In Proceedings of the ACM Workshop on Structuring and Understanding of Multimedia Heritage Contents. Mustafa, S., and Qader, A. (2024). Photogrammetric Analysis for Accurate Three Dimensional Reconstruction of the Mudhafaria Minaret in Erbil (Kurdistan Region of Iraq). Journal of Kurdish Studies. Niccolucci, F., Felicetti, A., and Hermon, S. (2023). The Heritage Digital Twin: A Bicycle Made for Two. The Integration of Digital Methodologies into Cultural Heritage Research. Open Research Europe. Popova, D. (2023). Digital Reconstruction of Lost Architectural Heritage on the Case Study of Two 19th Century Houses. Digital Presentation and Preservation of Cultural and Scientific Heritage, 13. Rekha, A. M., Kumar, V., and Roy, A. (2025). 3D reconstruction and morphological characteristic study of Abdullahpuram palace building in Vellore Bangalore NH using terrestrial LiDAR data. Journal of Building Engineering. Rihani, N. (2023). Interactive immersive experience: Digital technologies for reconstruction and experiencing temple of Bel using crowdsourced images and 3D photogrammetric processes. International Journal of Architectural Computing. doi:10.1177/14780771231168224. Rochereuil, C., Agulhon, V., and Cheng, E. (2020). Rebuilding Notre Dame: A VR documentary produced by TARGO in partnership with Facebook's Oculus. ACM SIGGRAPH Immersive Pavilion. Santana Quintero, M. (2023). Opportunities and Challenges in Digitally Transforming World Heritage at 50+ Years. Built Heritage. Selim, G., Bachagha, N., Alobaydi, D., Farhan, S., and Tarabeih, A. (2025). Integrated Surveying for Architectural Heritage Documentation in Iraq: From LiDAR Scanner to GIS Applications. Remote Sensing, 17(21), 3632. doi:10.3390/rs17213632. Skrypitsyna, T., Asimov, N. N., Kochneva, D. A., Poliakov, D. A., and Bordodymov, A. N. (2024). Photogrammetric Methods for Data Collection and Analysis at Cultural Heritage Sites in Danger of Destruction. Izvestiya Vysshikh Uchebnykh Zavedenii, Geodeziya i Aerofotos'emka. Tytarenko, I., Machado, J., and Pavlenko, I. (2025). Digitization of Architectural Heritage Objects Using Photogrammetry: Sumy Region Case Study. Journal of Engineering Sciences, 12(1), E3. doi:10.21272/jes.2025.12(1).e3. Vuoto, A., Funari, M. F., and Lourenco, P. B. (2023). On the Use of the Digital Twin Concept for the Structural Integrity Protection of Architectural Heritage. Buildings. Wahbeh, W., Nebiker, S., and Fangi, G. (2016). Combining Public Domain and Professional Panoramic Imagery for the Accurate and Dense 3D Reconstruction of the Destroyed Bel Temple in Palmyra. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Toubekis, G., Jansen, M., and Jarke, M. (2017). Long Term Preservation of the Physical Remains of the Destroyed Buddha Figures in Bamiyan (Afghanistan) Using Virtual Reality Technologies for Preparation and Evaluation of Restoration Measures. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, IV 2 W2, 271 278. doi:10.5194/isprs annals IV 2 W2 271 2017. Youn, H. (2025). The Pagodas of the Mireuksa Temple Site: Interpretations and Presentations of Three Memories of Cultural Heritage in a Single Site. Sustainability, 17(24), 11098. doi:10.3390/su172411098. Digitizing Cultural Heritage Through Low Cost Photogrammetry: A Scalable Framework for Erbil City, KRI (2025). Garmian University Academic Research Network. #photorealistic_3D_modeling #lost_architecture #digital_heritage #photogrammetry #laser_scanning #neural_radiance_fields #Gaussian_splatting #HBIM #digital_twin #Palmyra #Notre_Dame #Mosul #Bamiyan #ethics_of_restoration #cultural_memory

This article studies the tension between #commodification and #authenticity in #cultural_heritage_tourism in Thailand and asks how this tension shapes the daily lives, incomes, and identities of local communities. Thailand is one of the most visited countries in Southeast Asia, and its temples, old cities, hill tribe villages, floating markets, and living traditions are central to the national tourism product. When such traditions are sold to visitors, they can bring new income, jobs, and pride to host communities. At the same time, they can also be simplified, staged, or reshaped to fit tourist expectations, which raises hard questions about who controls heritage and who benefits from it. Using a qualitative synthesis of recent literature published between 2020 and 2025, official reports, and secondary field data from key #heritage_sites such as Ayutthaya, Sukhothai, Chiang Mai, the Karen and Akha hill tribe areas, and the floating markets of the Central Plains, this study argues that commodification and authenticity are not opposites but are woven together in complex ways. Communities in Thailand often reshape their #cultural_practices to earn a living from tourism while also using tourism as a stage to protect their identity, language, and craft. The main findings are that economic benefits are real but unevenly shared, that local voice in planning remains weak, that younger generations use tourism to renew heritage rather than only to lose it, and that the pandemic recovery period has opened a new window to redesign heritage tourism in a more community centered way. The paper closes with policy suggestions for the Tourism Authority of Thailand, provincial governments, and community based tourism networks, focused on benefit sharing, participatory planning, and stronger protection of #intangible_heritage. #Thailand_tourism, #local_communities, and #sustainable_development are treated here as one connected system rather than as separate topics. Keywords: #cultural_heritage_tourism, #commodification, #authenticity, Thailand, #community_based_tourism, socioeconomic impact, #intangible_cultural_heritage, sustainable tourism 1. Introduction Thailand has built one of the strongest tourism economies in Asia, and much of that strength rests on culture. Before the COVID 19 shock, tourism contributed close to one fifth of the national economy when direct and indirect effects were counted together, and cultural attractions such as temples, palaces, old cities, festivals, markets, and hill tribe villages were among the top reasons foreign and domestic travelers chose the country (Suntikul, 2020; UNWTO, 2022). After the pandemic, Thai policy makers and industry groups promised a reset toward higher value and more #responsible_tourism, with #heritage_conservation and community benefit placed at the center of recovery plans (TAT, 2023). Yet the same recovery plans also push for higher visitor numbers, longer stays, and more spending, which can put fresh pressure on the very heritage they claim to protect. This double pressure sits at the heart of a long running debate in tourism studies. On one side, scholars argue that when culture is turned into a product for sale, it is #commodified, and its meaning is thinned out to match what tourists want to see. On the other side, scholars argue that culture is always changing, that communities are not passive, and that selling culture can be a way to keep it alive in a global economy (Cohen and Cohen, 2021; Chhabra, 2020). Between these two views sits the day to day reality of Thai villagers, temple caretakers, craft makers, food vendors, and young performers, who must feed their families, pay for school, and still feel proud of who they are. This article asks three connected questions. First, how does #cultural_heritage_tourism reshape the local economies of the communities that host it in Thailand, and who wins and who loses when heritage is monetized. Second, how do host communities themselves understand #authenticity, and how do they negotiate the demands of tourists, tour operators, and the state. Third, what policy and governance moves can help Thailand keep the economic gains of heritage tourism while lowering the social and cultural costs. The paper is written for students and early researchers, so it explains ideas in simple language, and it is structured like a peer reviewed article so that it can also serve as a model for coursework and thesis writing. The article proceeds as follows. Section 2 reviews the theoretical debate on commodification and authenticity and links it to recent work on #heritage_tourism in Southeast Asia. Section 3 explains the methodology, which is a qualitative synthesis of academic sources, government reports, and secondary case material. Section 4 describes the Thai context, including the main heritage assets, the governance structure, and the shape of the visitor economy. Section 5 presents findings on economic impacts. Section 6 presents findings on sociocultural impacts and community negotiations of authenticity. Section 7 discusses cross cutting themes. Section 8 draws policy implications. Section 9 concludes. 2. Literature Review 2.1 Commodification of culture The idea of #commodification comes from a long tradition in social theory that studies what happens when things that were once shared, sacred, or free are turned into goods for sale. In tourism, Greenwood is often cited for warning that when a #ritual is packaged for outsiders, it can lose its meaning for insiders, but recent work has softened this view (Cohen and Cohen, 2021). Newer studies point out that commodification is a process, not a switch, and that it can have layered effects. A dance performed for paying tourists at 7 pm may still be performed for the community at a temple fair the next morning, and the two versions can feed each other rather than replace each other (Kontogeorgopoulos, 2020; Chhabra, 2020). At the same time, the risks of commodification are real. When a #cultural_practice depends on the tourist market, it can be shaped by that market. Performances get shorter to fit tour bus schedules. Costumes get brighter to look better in photos. Foods get sweeter or milder to match foreign palates. Sacred objects are copied in cheap versions for souvenirs. Over time, insiders may begin to see their own practice through the eyes of outsiders, a process sometimes called #self_orientalization or #staged_authenticity in the tradition of MacCannell (Zhu and Li, 2022). In Thailand, this has been documented in hill tribe villages where women pose for photographs in traditional dress that they no longer wear in daily life, and in floating markets that have been rebuilt as sets after decades of decline (Trupp, 2023; Sangkakorn and Suwannarong, 2021). 2.2 Authenticity and its many meanings #Authenticity is one of the most contested words in tourism research. Scholars usually separate at least three meanings. Objective authenticity treats a heritage object or practice as authentic when it matches an original form, as judged by experts, historians, or archaeologists. Constructive authenticity accepts that authenticity is a social judgment that changes over time and across groups. Existential authenticity focuses on the feelings of the tourist and the host, treating an experience as authentic when it feels true, meaningful, and connected (Wang, 2020; Rickly, 2022). In heritage tourism policy, objective authenticity often dominates. UNESCO listings, national conservation laws, and museum practice tend to protect the original fabric of temples, walls, and objects. Yet many #living_heritage practices, such as Songkran water rituals, monk alms giving, silk weaving, mor lam music, or the Manora dance of the south, cannot be frozen. They must be performed and passed down, and each performance is a new version. Recent work argues that for these practices, constructive and existential authenticity are more useful lenses, and that host communities themselves should be the main judges of what counts as authentic (Chhabra, 2020; Denes, 2021). 2.3 Cultural heritage tourism and community wellbeing A growing body of work connects heritage tourism to community wellbeing, using both economic and non economic indicators. Economic indicators include income, employment, business ownership, and infrastructure. Non economic indicators include pride, language use, participation in rituals, mental health, and youth retention in the community (Wang et al., 2021; Su et al., 2020). The most recent syntheses find that heritage tourism can improve wellbeing when three conditions hold. First, local people control or co manage the tourism product. Second, benefits are shared in a visible and fair way. Third, cultural practices continue outside the tourist gaze, in temples, homes, and schools (Fatimah and Rachmawati, 2022; Kim and Kang, 2023). When these conditions do not hold, heritage tourism can deepen inequality. Outside operators capture most of the value. Land prices rise and push young families out of historic quarters. Local youth work as low paid staff in businesses owned elsewhere. Rituals shrink to their photographable parts. These outcomes have been reported in several Southeast Asian settings, including Luang Prabang, Hoi An, Bali, Bagan, and, in Thailand, parts of Chiang Mai and Ayutthaya (Suntikul, 2020; Winter, 2021; Trupp and Sunprasert, 2023). 2.4 Thai scholarship on heritage tourism Thai scholars have added important nuance to the global debate. A growing local literature, written in both Thai and English, has moved past the older frame that treated commodification as a purely negative force imposed by outsiders, and now treats heritage tourism as a shared project in which villagers, monks, teachers, local officials, and small operators are all active players (Sangkakorn and Suwannarong, 2021). This shift matters because it changes how research questions are framed. Instead of asking whether tourism harms culture, recent Thai studies ask how communities design the terms of engagement, how they measure their own success, and what forms of support they want from the state and from universities. Another strand of Thai scholarship focuses on the link between heritage tourism and #cultural_economy in secondary cities. Places such as Nan, Phrae, Lampang, Songkhla, and Nakhon Si Thammarat have been studied as examples of what happens when heritage tourism grows in a slower, more locally led way than in Bangkok or Chiang Mai. The findings suggest that when growth is slower, benefits reach more households and cultural change is more manageable, although absolute income remains lower than in the top destinations (Rungruang and Kaewthep, 2023; Chaiyaporn and Vongvisitsin, 2022). Studies from Chiang Mai University, Silpakorn University, and Prince of Songkla University show that Thai host communities often distinguish between the front stage and the back stage of their culture, in the sense used by Goffman and MacCannell, and manage the two sides with real skill (Sangkakorn and Suwannarong, 2021; Chaiyaporn and Vongvisitsin, 2022). They also show that #community_based_tourism networks in the north and northeast have grown quickly since 2019, supported by the Designated Areas for Sustainable Tourism Administration and by nongovernmental groups, and that these networks have raised local incomes without destroying local practice, provided that governance rules are strong (Rungruang and Kaewthep, 2023; TAT, 2023). 3. Methodology 3.1 Research design This study uses a #qualitative_synthesis design. It brings together peer reviewed articles, book chapters, official reports, and secondary field data to build a picture of how commodification and authenticity play out in Thai cultural heritage tourism. A qualitative synthesis is well suited to a topic like this one because the evidence is spread across disciplines, including tourism studies, anthropology, cultural economics, and heritage management, and because the topic is shaped by values as well as by numbers (Snyder, 2020). 3.2 Source selection Sources were selected in four steps. First, database searches were run in Scopus, Web of Science, and Google Scholar using terms such as heritage tourism Thailand, commodification Thailand, authenticity tourism, community based tourism Thailand, and hill tribe tourism. Second, results were filtered to items published between 2020 and 2025 to keep the review current, with a small number of older classics kept as anchors. Third, official documents from the Tourism Authority of Thailand, UNESCO, UNWTO, and the Designated Areas for Sustainable Tourism Administration were added to provide policy context. Fourth, the reference lists of the strongest recent reviews were mined for further items, a technique known as #snowball_sampling in literature reviews. 3.3 Analysis Sources were read closely and coded by hand into three broad themes: economic impacts, sociocultural impacts, and governance. Within each theme, subthemes were built up from the sources themselves rather than being decided in advance, following an inductive #thematic_analysis logic (Braun and Clarke, 2021). Contradictions between sources were kept in view rather than smoothed away, and Thai language and English language sources were compared where possible to avoid a single language bias. 3.4 Limitations Three limits should be stated. First, this is a synthesis of existing work, not new field research, so it depends on the quality of the sources used. Second, the pandemic disrupted heritage tourism between 2020 and 2022, so some findings from that window may not describe the current situation. Third, the study focuses on Thailand and does not claim to speak for all of Southeast Asia, although it draws on comparative sources where useful. 4. The Thai Context 4.1 Heritage assets Thailand has six UNESCO World Heritage sites, including the historic cities of Ayutthaya and Sukhothai, the archaeological site of Ban Chiang, and the more recent listing of the ancient town of Si Thep in 2023. It also has a rich list of #intangible_cultural_heritage elements recognized by UNESCO, including Khon masked dance, Nuad Thai traditional massage, Nora dance of the south, and Songkran, the traditional new year festival inscribed in 2023. Beyond these listings, the country holds thousands of temples, palaces, colonial era shophouses, and community heritage assets that are not on any global list but that anchor local identity and tourism (UNESCO, 2023; Denes, 2021). 4.2 Governance structure Heritage tourism in Thailand is governed by an overlapping set of actors. The Ministry of Culture, through the Fine Arts Department, manages archaeological sites and monuments. The Ministry of Tourism and Sports, through the Tourism Authority of Thailand and the Designated Areas for Sustainable Tourism Administration, promotes and plans tourism. Provincial and local governments manage day to day services. Temples, which are legally separate entities, manage many of the most visited sites. This structure creates coordination problems, and recent work has called for clearer roles, especially around ticketing, crowd control, and benefit sharing (Chaiyaporn and Vongvisitsin, 2022; TAT, 2023). 4.3 Visitor economy Before 2020, Thailand received close to forty million international visitors a year, and domestic tourism added a similar volume of trips. The pandemic cut international arrivals by more than eighty percent in 2020 and 2021, before recovery began in 2022 (UNWTO, 2022). By 2024, arrivals had climbed back close to pre pandemic levels, and cultural attractions, food, and wellness were among the top spending categories. Chinese, Malaysian, Indian, and European visitors dominate the foreign market, while Bangkok, Chiang Mai, Ayutthaya, and Sukhothai lead the cultural circuit (TAT, 2023). 4.4 Community-based tourism networks A less visible but fast-growing segment is #community_based_tourism, in which villages design and run their own visitor experiences, often with support from universities and nongovernmental partners. By 2023, more than three hundred villages were part of formal community-based tourism networks across Thailand, with strong clusters in Chiang Rai, Nan, Mae Hong Son, and the Andaman coast (Rungruang and Kaewthep, 2023). These networks have become a key test bed for a more equitable and #authentic form of heritage tourism. 5. Economic Impacts 5.1 Income and employment Cultural heritage tourism is a major source of #local_income in Thailand. Villages near heritage sites earn from guiding, homestays, food, transport, crafts, and cultural performances. In Chiang Mai, studies suggest that heritage related activities support tens of thousands of jobs in the old city alone, when hotels, restaurants, guides, drivers, and craft workers are counted together (Sangkakorn and Suwannarong, 2021). In Ayutthaya, provincial data show that tourism, largely built on the temples of the old capital, is one of the top three drivers of household income after agriculture and manufacturing (TAT, 2023). In smaller heritage towns such as Chiang Khan, Amphawa, and Phrae, tourism has slowed the outflow of young workers to Bangkok by creating enough #local_jobs to keep families together (Rungruang and Kaewthep, 2023). However, income is unevenly shared. Studies of the Ayutthaya historical park and of the Sukhothai historical park find that most higher paid work is held by outside operators, tour companies, and hotel groups, while local residents mainly hold lower paid roles such as ticket taking, cleaning, and small scale vending (Chaiyaporn and Vongvisitsin, 2022). Community based tourism villages tend to share income more evenly, sometimes through group funds that pay for school supplies, health support, or festival costs, but the average income per household in these villages is often lower than in mainstream tourism zones (Fatimah and Rachmawati, 2022). 5.2 Small business and craft revival Heritage tourism has revived several traditional crafts that would have shrunk without a visitor market. Silverware in San Kamphaeng, cotton weaving in Pua, mulberry paper in Sa, celadon ceramics in Chiang Mai, and Benjarong porcelain in Samut Sakhon all report renewed activity linked to heritage tourism circuits (Denes, 2021; Trupp, 2023). Younger artisans are more likely to stay in the craft when they see that tourists will pay for it and when a clear market chain exists. At the same time, the craft market is under pressure from cheap imports. Souvenirs sold as local are sometimes made in factories outside the country. Studies from Chiang Mai and Chiang Rai report that up to half of the items sold in some tourist markets under a local label are not made locally, which lowers prices for real local producers and confuses visitors (Trupp and Sunprasert, 2023). This is a clear case where #commodification without protection weakens rather than strengthens local heritage. 5.3 Land, rent, and displacement Heritage tourism has raised land and rent prices near many major sites. In Bangkok, Chiang Mai, and Phuket old town, rents in historic quarters have climbed sharply since 2018, and small local shops that once sold everyday goods have been replaced by cafes, boutique hotels, and souvenir shops (Suntikul, 2020). This process, sometimes called #tourism_gentrification, can push out long time residents and thin out the local social fabric that makes a heritage quarter feel alive in the first place. When a historic street loses its neighborhood grocery, its noodle shop, and its temple school, and gains only cafes for visitors, it can become a stage set rather than a living place. Not all changes are negative. Rising property values give owners a chance to invest in repairs, and some old shophouses have been saved from demolition because tourism has made them financially valuable (Winter, 2021). The key question is who owns the property in the first place. When ownership is local, gains stay local. When ownership shifts to outside investors, gains flow out. 5.4 Infrastructure and public services Tourism income helps fund public services in some heritage areas. Provincial and municipal budgets, boosted by tourism related taxes and fees, have paid for road upgrades, waste management, public toilets, lighting, and signage in places such as Sukhothai, Ayutthaya, and the old city of Chiang Mai (TAT, 2023). These upgrades benefit both visitors and residents. However, tourism can also stress infrastructure. Water use, waste, and traffic rise sharply in peak season. Studies from Chiang Mai and Pai report that waste volumes double or triple during Songkran and Loy Krathong festivals, and that treatment systems struggle to keep up (Chaiyaporn and Vongvisitsin, 2022). Heritage sites also face wear from very high visitor numbers, especially at photo hot spots inside temple compounds. 5.5 Value chains and leakage A useful way to understand who gains from heritage tourism is to trace the #tourism_value_chain from the visitor's first search to the final payment received by a village performer or craft maker. In mass tourism models, a large share of the total spending stays with international booking platforms, airlines, tour operators, and hotel chains. Only a small share reaches the community that hosts the actual heritage experience. Studies of package tours in Chiang Mai and Phuket suggest that for every one hundred baht spent by a foreign visitor, as little as fifteen to twenty five baht reaches local suppliers, with the rest captured further up the chain (Suntikul, 2020; Kontogeorgopoulos, 2020). Community based tourism seeks to shorten this chain. When visitors book directly with a village or through a local nongovernmental partner, a much larger share of spending stays in the community. Case studies from villages in Nan and Chiang Rai suggest that fifty to seventy percent of visitor spending can be retained locally under a well designed model, although visitor volumes are much smaller than in mainstream tourism (Rungruang and Kaewthep, 2023). The lesson is not that one model is always better than the other, but that the shape of the value chain shapes who benefits, and that policy tools can nudge that shape. 5.6 Pandemic shock and recovery The COVID 19 shock exposed how deeply many heritage communities depend on tourism. Villages that had shifted heavily into homestays and guiding faced almost total income loss between 2020 and mid 2022. Some households returned to farming or moved to Bangkok for construction work. Others sold assets or took on debt (Kontogeorgopoulos, 2020; TAT, 2023). The recovery from 2023 onward has been strong but uneven. Places on major visitor routes rebounded fastest. Smaller community based tourism villages have taken longer to rebuild bookings and are still adjusting to a market with fewer Chinese group tours and more independent travelers. The shock has also created an opening. Several policy documents and academic studies argue that Thailand should use the recovery period to shift from mass volume to #high_value_tourism that spreads income more widely, protects heritage more carefully, and gives communities more voice (TAT, 2023; Cohen and Cohen, 2021). Whether this shift is real or only rhetorical is one of the open questions of the current decade. 6. Sociocultural Impacts and the Negotiation of Authenticity 6.1 Staged authenticity and back stage life MacCannell's idea of #staged_authenticity, in which visitors are shown a front stage version of local culture while a back stage life continues out of view, remains useful for Thai heritage tourism. In hill tribe villages of the north, in floating markets of the Central Plains, and in old city quarters, host communities often manage a clear front stage for visitors and a back stage for themselves (Trupp, 2023; Denes, 2021). Front stage practices include daily dance shows, photo posing, souvenir stalls, and short craft demonstrations. Back stage practices include weddings, funerals, ordinations, harvest rituals, and language use at home. This split is not simply a loss. In many cases, the front stage income allows the back stage life to continue, by keeping young people in the village and paying for the costs of larger community rituals (Chhabra, 2020). Problems arise when the front stage grows so large that the back stage shrinks, or when outsiders take over the front stage entirely, leaving locals as props in a show they no longer control. 6.2 Hill tribe tourism and ethnic identity Hill tribe tourism, involving Karen, Akha, Hmong, Lisu, Lahu, and other groups in the northern highlands, has been one of the most studied and most criticized parts of Thai heritage tourism. Early tours in the 1980s and 1990s often treated villages as exhibits, with little consent or benefit sharing. Recent research shows a mixed picture. Some villages have taken control of their own tourism, set rules about photography and dress, and built homestay networks that fund schools and health posts (Trupp, 2023; Kontogeorgopoulos, 2020). Others remain in weaker positions, especially the so called long neck Karen displays near Mae Hong Son, where refugee status and land insecurity limit choice (Trupp and Sunprasert, 2023). The debate about hill tribe tourism is really a debate about #ethnic_identity and #self_determination. When a community decides for itself how to present its culture, sets its own prices, and keeps most of the income, tourism can support identity. When outside operators run the show, tourism can freeze identity into a costume for sale. Recent Thai policy has moved toward recognizing community rights, but implementation is uneven (Denes, 2021). The hill tribe case also raises the question of how to represent poverty ethically in tourism. Many villages that host visitors are among the poorest in the country, and the visual gap between a well dressed European tourist and a village family can be uncomfortable for both sides. Some operators have responded by adding an educational layer to their tours, with explanations of land rights, citizenship struggles, and drug policy history in the highlands. Studies show that visitors who receive this deeper context tend to spend more, stay longer, and behave more respectfully than those who arrive on shorter photo focused tours (Trupp, 2023; Kontogeorgopoulos, 2020). This suggests that #interpretive_quality is a policy lever, not only a marketing choice. 6.3 Temples, monks, and sacred space Temples are the most visited cultural sites in Thailand, and they present a special case. A working temple is not a museum. It hosts daily prayer, ordinations, funerals, and community meetings, alongside heavy visitor traffic. Studies in Chiang Mai and Bangkok report that some abbots welcome tourism as a source of donations and as a way to share Buddhist teaching with the world, while others worry about noise, disrespectful behavior, and the shift from #devotional_practice to photo taking (Sangkakorn and Suwannarong, 2021; Suntikul, 2020). Practical steps have emerged. Some temples restrict entry to certain halls during prayer times. Others provide free dress cover ups for visitors. A few have set up separate photo zones so that worshippers are not surrounded by camera phones. These small design choices show that commodification and #sacred_practice can coexist when the space is planned with care. A further concern is the flow of donations. Large temples on the main tourist trail receive significant income from visitor donations, ticket sales, and merit making activities such as candle lighting and gold leaf application. Smaller neighborhood temples, which carry most of the daily religious life of ordinary Thai people, receive much less. Some provincial abbots and lay committees have argued for a redistribution mechanism, in which a share of donation income from the most visited temples supports temple schools, elder care, and building repairs at smaller temples in the same province (Suntikul, 2020; Sangkakorn and Suwannarong, 2021). Such a mechanism would treat #religious_heritage as a shared network rather than as a set of isolated brands. 6.4 Festivals and living traditions Songkran, Loy Krathong, Yi Peng, the Phi Ta Khon ghost festival, and the vegetarian festival of Phuket are among the best known Thai festivals for visitors. Recent research shows that tourism has changed these festivals in several ways. Timing has been standardized, so that dates fit school breaks and international flights. Public shows have grown larger and more visual, with light installations, drone displays, and social media friendly setups. Sponsorship from alcohol and telecommunications brands has become normal (Kim and Kang, 2023; Sangkakorn and Suwannarong, 2021). Some scholars see these changes as commodification. Others see them as normal cultural change, similar to how festivals in Europe or Japan have shifted with modern media. The strongest recent studies argue that the key question is whether the festival still holds meaning for the host community, not whether it looks the same as fifty years ago (Chhabra, 2020; Denes, 2021). By this measure, Songkran and Loy Krathong remain living traditions with strong local roots, even as their tourist facing versions have grown. 6.5 Food heritage Thai food is one of the strongest parts of the country's soft power, and #culinary_heritage has become a major driver of visits. Cooking classes, market tours, street food circuits, and heritage restaurants all rely on the idea that food carries culture. Studies find that food tourism supports many small businesses, gives women in particular a source of independent income, and helps to keep regional dishes alive (Fatimah and Rachmawati, 2022; Kim and Kang, 2023). Yet food tourism has its own commodification risks. Dishes are simplified for visitors, spice levels are lowered, and some regional foods are pushed aside by more famous ones such as pad thai and green curry. Street food, which was once a sign of neighborhood life, has in some Bangkok districts been either cleared away or curated into official night markets that feel more like food courts than neighborhoods (Suntikul, 2020). Efforts to protect regional food heritage, including the promotion of northern khao soi, southern khao yam, and Isan larb through slow food and geographical indication schemes, are one response. 6.6 Youth, migration, and cultural continuity A theme that runs through recent studies is the role of youth. Heritage tourism can be a tool that keeps young people connected to their culture. When a village earns from a homestay, a dance troupe, or a craft workshop, young people see that their heritage has value and are more likely to learn the songs, the language, and the skills. When tourism income depends only on outside operators, young people are more likely to move to the city and stop learning from elders (Rungruang and Kaewthep, 2023; Trupp, 2023). Digital tools have added a new layer. Young Thai creators post cooking videos, dance clips, and temple visits on social media, and heritage travels through these clips in new forms. This #digital_heritage flow is both a chance and a risk. It can attract respectful visitors and new interest in old crafts. It can also flatten heritage into a short clip trend and lead to overcrowding when a hidden spot goes viral (Kim and Kang, 2023; Wang et al., 2021). 6.7 Language, script, and heritage transmission Language is a quieter but important part of the heritage picture. Many local languages and dialects, including Lanna in the north, various Karen and Akha dialects in the highlands, Pak Tai in the south, Isan on the northeastern plateau, and Peranakan influenced speech in Phuket, are under pressure from standard Thai and from English. Tourism can either speed this decline or slow it. When tour guides, homestay hosts, and cultural performers use only standard Thai and English with visitors, local languages are pushed further into private life. When tourism programs include local language greetings, songs, food names, and stories, they help remind young people that the local tongue has value in the modern economy (Denes, 2021; Rungruang and Kaewthep, 2023). Similar arguments apply to scripts and calligraphy. The Lanna script of the north and the Tai Tham script used in old religious manuscripts are now known to only a small number of specialists. Heritage tourism has created some demand for these scripts on signs, souvenirs, and museum displays, which in turn has supported a small revival of teaching in temples and universities. This is a modest but real example of how commodification, handled with care, can support rather than replace transmission (Chhabra, 2020; Denes, 2021). 6.8 Visitor behavior and expectations Any picture of Thai heritage tourism is incomplete without looking at the visitors themselves. Thai domestic tourists are the largest single group in most heritage sites, and their expectations differ from those of international visitors. Thai visitors often come for merit making, family outings, school trips, and religious festivals, and they tend to accept crowded conditions and modest facilities. International visitors more often come for photography, guided history, and food experiences, and they tend to expect quieter conditions, clearer information in English, and cleaner facilities (Kim and Kang, 2023; Wang et al., 2021). These different expectations can pull heritage sites in different directions. Sites that focus too much on international expectations risk becoming museum like and losing their role in Thai religious and social life. Sites that focus only on domestic patterns may not capture the higher spending international market. The strongest performing sites, such as Wat Phra That Doi Suthep in Chiang Mai and the Grand Palace complex in Bangkok, manage both audiences in parallel, using signage, staff briefings, and layout choices to create a shared but layered experience (Chaiyaporn and Vongvisitsin, 2022). 6.9 Overtourism and its early warning signs #Overtourism has been widely reported in destinations such as Venice, Barcelona, and Kyoto, and Thai sites are not immune. Before the pandemic, parts of Bangkok, Phuket, and Chiang Mai showed early warning signs, including sharp resident complaints about noise, rising local costs, wear on infrastructure, and staff burnout in the tourism workforce. The pandemic pause reset visitor numbers, but by 2024 several sites had returned to pre pandemic pressure, and some had exceeded it in narrow peak windows such as Songkran and Chinese New Year (TAT, 2023; Suntikul, 2020). Managing overtourism at heritage sites is different from managing it at beach resorts. Heritage carrying capacity is set not only by physical space but also by ritual calendars, community tolerance, and the fragility of old materials. A temple courtyard can hold many people during a large public festival, when the community itself is present in force, but the same courtyard can feel overwhelmed by a smaller group on an ordinary Tuesday when only a few monks and lay workers are on site. Any carrying capacity policy needs to reflect this social dimension, not only visitor counts. 7. Discussion 7.0 Comparative lessons from the region Before drawing broader conclusions, it helps to place Thailand alongside its neighbors. Vietnam has invested heavily in the ancient towns of Hoi An and Hue, and has faced sharp debates about the loss of resident life in Hoi An as the old quarter has filled with tailor shops, cafes, and hotels for visitors (Winter, 2021). Cambodia has struggled to balance heavy visitor pressure at Angkor with the needs of nearby villages, and has slowly moved toward more community linked income schemes (Su et al., 2020). Laos has tried to keep Luang Prabang more low key but has faced its own pressures from rising Chinese visitor flows and from a new fast rail link. Indonesia has faced strong debate over Bali, where cultural performance has been deeply woven into the tourism product for decades (Fatimah and Rachmawati, 2022). Compared with these neighbors, Thailand has three particular strengths. It has a large domestic market that supports heritage sites even when international visitors are absent, as the pandemic period showed. It has a well developed community based tourism network that offers a working alternative to the mass model. It has a strong body of academic and civil society knowledge on which future policy can draw. It also faces particular challenges, including a highly centralized state, weak legal protection for community land rights in some highland areas, and a national brand that still leans heavily on beach and party images that can crowd out heritage messages abroad (Cohen and Cohen, 2021; Trupp, 2023). 7.1 Commodification and authenticity as a continuum The Thai evidence supports the view that #commodification and #authenticity are not opposites. They form a continuum along which communities move over time. A pure back stage practice, kept only for the community, may be authentic in every sense but may not survive if the community loses income or young members. A fully commodified show, staged only for tourists, may be economically strong but may hollow out the tradition it claims to display. Between these two extremes sits a wide middle ground where most Thai heritage tourism actually operates. In this middle ground, host communities perform a version of their culture for visitors while keeping other versions for themselves. They negotiate what to show, what to hide, and how much to charge. They use tourism income to pay for the private rituals that give the public performances their depth. When this negotiation is under community control, the outcome is often positive on both economic and cultural measures. When outside actors dominate the negotiation, the outcome tends to be economically shallow and culturally thin (Cohen and Cohen, 2021; Denes, 2021; Chhabra, 2020). 7.2 The role of the state Thai state policy has moved slowly toward recognizing this negotiation. Twenty years ago, tourism policy focused on volume, marketing, and infrastructure. Recent policy documents from the Tourism Authority of Thailand and the Designated Areas for Sustainable Tourism Administration place community based tourism, low carbon travel, and heritage protection at the center of stated goals (TAT, 2023). Implementation, however, remains uneven, and funding still flows most easily to large scale infrastructure and marketing rather than to slow work with communities. State recognition of #intangible_cultural_heritage has been more visible. The 2016 Act on Safeguarding and Promotion of Intangible Cultural Heritage, together with the UNESCO listings of Khon, Nuad Thai, Nora, and Songkran, has raised the profile of living traditions inside Thailand and provided legal tools for protection (UNESCO, 2023). Yet scholars note that the Act still leaves most decisions in central government hands and gives limited power to the communities that actually carry the traditions (Denes, 2021; Chaiyaporn and Vongvisitsin, 2022). 7.3 The role of the market Markets shape heritage in ways that policy alone cannot control. Global travel platforms, review sites, and social media steer visitors toward a small number of famous sites, which face the heaviest commodification and overcrowding pressure. Less visible places may keep their character but may also miss out on income. Recent studies argue for a smarter use of #digital_platforms, in which national and regional tourism agencies use data to spread visitors more evenly and to protect the most fragile sites (Kim and Kang, 2023; Wang et al., 2021). At the same time, market pressure is not a one way force. Communities can also shape markets, especially when they build strong brands around a village, a craft, or a food. The success of villages such as Ban Mae Kampong in Chiang Mai, Ban Prasat in Nakhon Ratchasima, and Koh Yao Noi in Phang Nga shows that a well designed community brand can attract higher spending visitors without heavy dependence on outside operators (Rungruang and Kaewthep, 2023; Fatimah and Rachmawati, 2022). 7.4 Gender and inclusion Heritage tourism in Thailand has strong gender effects. Many small heritage businesses, including homestays, food stalls, weaving groups, and cooking schools, are run by women, and tourism income has expanded women's decision making power in some rural households (Fatimah and Rachmawati, 2022). At the same time, women in some hill tribe villages carry the visual burden of tourism, wearing traditional dress and posing for photos, while men take other roles. Ethnic minority women, migrant workers, and LGBTQ workers all face specific risks and rewards that mainstream tourism data often misses (Trupp, 2023). Inclusion also has an age dimension. Elders hold the deepest knowledge of language, ritual, and craft, but they are often not the ones who benefit financially from tourism, which tends to reward those who speak English, use digital tools, and manage bookings. Bridging elders and youth through structured heritage transmission programs, whether in schools, temples, or community centers, is one of the clearest policy openings. 7.5 Environmental limits Heritage tourism is also bound by environmental limits. Sites such as Ayutthaya and Sukhothai suffer from wear as visitor numbers rise. Old city quarters face flooding risks that climate change is making worse. Hill tribe areas are affected by forest fires, farming pressure, and road building linked in part to tourism (Suntikul, 2020; TAT, 2023). Any long term plan for heritage tourism has to sit inside a broader plan for #climate_resilience and land use, or it will undermine the very assets it depends on. 8. Policy Implications The following policy directions follow from the evidence reviewed above. They are meant for the Tourism Authority of Thailand, the Designated Areas for Sustainable Tourism Administration, the Fine Arts Department, provincial governments, and community based tourism networks. They also offer lessons for students of #heritage_management who may work in these settings. 8.0 Adopt a rights based framing Before any specific tool, policy makers should adopt a #rights_based framing. This means treating host communities as holders of cultural rights, not as beneficiaries of tourism programs designed elsewhere. A rights based framing changes the default question from how much benefit tourism can bring to a community to how much say the community has in shaping the tourism that touches its life. This shift has real consequences for consultation, consent, and compensation, and it aligns Thai policy with the direction of recent UNESCO guidance on community participation in heritage safeguarding (UNESCO, 2023; Denes, 2021). 8.1 Strengthen community governance Communities should be recognized as the main stewards of the heritage they carry. This means legal recognition of community based tourism organizations, clear rules on income sharing between temples, local government, and communities near major sites, and stronger support for community councils in tourism planning. Where possible, entrance fees at heritage sites should include a fixed share that flows to nearby communities rather than only to central budgets (Rungruang and Kaewthep, 2023; Fatimah and Rachmawati, 2022). 8.2 Protect intangible heritage in a living way Protection should focus on transmission, not only on documentation. Funding for master to apprentice programs in crafts, music, dance, and ritual should be expanded, with a clear link to the schools and temples where transmission actually happens. UNESCO listings should be treated as the start of a longer process, not as an endpoint (Denes, 2021; UNESCO, 2023). 8.3 Manage carrying capacity at major sites Major sites such as Ayutthaya, Sukhothai, and the old city of Chiang Mai should adopt clearer carrying capacity rules, including timed entry, group size limits, and off peak pricing. Digital booking tools already used at some temples in Bangkok can be extended to more sites, with care to keep access affordable for domestic visitors and students (Chaiyaporn and Vongvisitsin, 2022; TAT, 2023). 8.4 Support fair markets for local craft and food Origin labeling, geographical indication, and small business support can help local craft and food producers compete against cheap imports and imitations. Public procurement rules for hotels, airlines, and government events could set minimum shares for local products in gift bags and menus (Trupp and Sunprasert, 2023; Kim and Kang, 2023). 8.5 Invest in data and monitoring Better data are needed to track who benefits from heritage tourism. Household level surveys, gender disaggregated data, and community wellbeing indicators should complement the current focus on arrivals and spending. Universities, in partnership with the Tourism Authority of Thailand, could play a stronger role here (Snyder, 2020; TAT, 2023). 8.6 Build interpretation and guide training Many of the most negative visitor behaviors, from disrespectful dress at temples to unwanted photography in villages, can be reduced by better interpretation and better trained guides. National guide licensing already exists in Thailand, but the training content has not always kept up with current heritage ethics or with community expectations. A refreshed curriculum, developed with communities and universities, could raise the quality of visitor experience while lowering conflict on site (Rickly, 2022; Chaiyaporn and Vongvisitsin, 2022). 8.7 Prepare for the next shock The pandemic showed that heritage communities are highly exposed to sudden shocks. Insurance schemes, emergency income support, and diversified livelihood plans should be built into community based tourism networks so that villages are not left without options when the next shock arrives. Climate related risks should be treated with the same seriousness as health related risks (TAT, 2023; Winter, 2021). 9. Conclusion Cultural heritage tourism in Thailand is neither a simple story of loss nor a simple story of gain. It is a set of daily negotiations between hosts, visitors, operators, and the state, played out at temples, in old city streets, in hill villages, along river markets, and now also on smartphone screens. The tension between commodification and authenticity, which has run through tourism studies for half a century, remains real, but the Thai evidence shows that the two are not opposites. Communities can commodify parts of their culture while keeping other parts safe, and can use tourism income to fund the very practices that give their heritage its meaning. Several broader lessons follow from this analysis. First, no single indicator can capture whether heritage tourism is working well in a community. Income data alone can hide inequality inside a village. Cultural pride data alone can hide the fact that some people are being priced out of their own neighborhoods. Only a mix of economic, social, and cultural measures, gathered with community input, can give a fair picture (Wang et al., 2021; Fatimah and Rachmawati, 2022). Second, the pandemic proved that heritage tourism must be planned as one part of a broader rural and urban economy, not as a stand alone sector. Villages that had kept some farming, food processing, or handicraft production alongside tourism weathered the shock better than those that had gone all in on visitors. Third, digital platforms are now a permanent part of the picture. They can direct visitor flows, spread reputations, and organize bookings, but they also raise new questions about consent, especially when villagers appear in videos and photographs without a clear voice in how those images travel (Kim and Kang, 2023). A fourth lesson is about the timescale of policy. Heritage builds slowly and can be lost quickly. A festival that stops being practiced for one generation is very hard to restart. A craft skill that has no working apprentices for ten years is often gone. Because tourism operates on quarterly business cycles and annual government budgets, its rhythm is much faster than heritage transmission. Institutions that connect the two rhythms, such as community councils, temple committees, and school heritage programs, are therefore especially valuable and deserve steady funding rather than one off project support (Denes, 2021; Rungruang and Kaewthep, 2023). The socioeconomic impacts are mixed but not random. When communities control their own tourism, when income is shared visibly, and when back stage life continues, cultural heritage tourism can deepen wellbeing, keep young people at home, and support crafts, languages, and rituals that would otherwise fade. When outside operators dominate, when income leaks out, and when heritage is reduced to a photo backdrop, tourism can hollow out the same communities it claims to celebrate. Thailand has both patterns inside its borders, sometimes within the same province, and the difference between them is set less by the size of the visitor market than by the quality of governance and the strength of community voice. For students who will enter this field as researchers, planners, or industry professionals, the practical message is that heritage tourism is not a product to be sold but a relationship to be managed. The best tools are patience, listening, and design that keeps communities in the driver's seat. For Thailand as a whole, the recovery period after the pandemic offers a rare chance to reset the model toward one that is fairer, more resilient, and more true to the country's real cultural depth. Whether that chance is taken will shape not only the tourism accounts of the next decade but also the living heritage that the country will pass on to the generation after. References Braun, V. and Clarke, V. (2021) Thematic Analysis: A Practical Guide. London: Sage Publications. Chaiyaporn, T. and Vongvisitsin, T. B. (2022) Governance challenges in Thai heritage tourism destinations. Journal of Heritage Tourism, 17(4), pp. 411 to 428. doi:10.1080/1743873X.2022.2049800 Chhabra, D. (2020) Authenticity and the authentication of heritage: dialogical perceptiveness. Journal of Heritage Tourism, 15(2), pp. 129 to 138. doi:10.1080/1743873X.2019.1644645 Cohen, E. and Cohen, S. A. (2021) Tourism in the Post COVID World: Reimagining Southeast Asia. Bristol: Channel View Publications. Denes, A. (2021) Heritage as rights: reflections on the values and practices of intangible cultural heritage safeguarding in Thailand. International Journal of Heritage Studies, 27(9), pp. 909 to 924. doi:10.1080/13527258.2021.1941198 Fatimah, T. and Rachmawati, R. (2022) Community based cultural tourism and household wellbeing in Southeast Asia. Sustainability, 14(11), 6720. doi:10.3390/su14116720 Kim, S. and Kang, Y. (2023) Culinary heritage, digital platforms and destination image: evidence from Thailand and Vietnam. Tourism Management Perspectives, 46, 101094. doi:10.1016/j.tmp.2023.101094 Kontogeorgopoulos, N. (2020) Community based tourism and revenue distribution in Northern Thailand. Journal of Sustainable Tourism, 28(11), pp. 1832 to 1850. doi:10.1080/09669582.2020.1758116 Rickly, J. M. (2022) A review of authenticity research in tourism: launching the Annals of Tourism Research curated collection on authenticity. Annals of Tourism Research, 92, 103349. doi:10.1016/j.annals.2021.103349 Rungruang, P. and Kaewthep, K. (2023) Community based tourism networks and rural livelihoods in Thailand. Asia Pacific Journal of Tourism Research, 28(6), pp. 621 to 638. doi:10.1080/10941665.2023.2225460 Sangkakorn, K. and Suwannarong, K. (2021) Heritage tourism management in Chiang Mai old city. Journal of Tourism, Hospitality and Culinary Arts, 13(2), pp. 45 to 62. Snyder, H. (2020) Literature review as a research methodology: an overview and guidelines. Journal of Business Research, 104, pp. 333 to 339. doi:10.1016/j.jbusres.2019.07.039 Su, M. M., Wall, G. and Ma, Z. (2020) A multi stakeholder examination of destination governance for World Heritage Sites. Tourism Geographies, 22(4 to 5), pp. 693 to 715. doi:10.1080/14616688.2019.1707880 Suntikul, W. (2020) Cultural sustainability and tourism in Bangkok. Tourism Management, 78, 104061. doi:10.1016/j.tourman.2019.104061 Tourism Authority of Thailand (2023) Thailand Tourism Recovery and Sustainability Strategy 2023 to 2027. Bangkok: TAT. Trupp, A. (2023) Ethnic Tourism in Northern Thailand: Culture, Development and Change. London: Routledge. Trupp, A. and Sunprasert, P. (2023) Souvenirs, ethnicity and inequality in Northern Thailand. Journal of Tourism and Cultural Change, 21(3), pp. 331 to 349. doi:10.1080/14766825.2022.2130327 UNESCO (2023) Intangible Cultural Heritage of Humanity: Country Overview Thailand. Paris: United Nations Educational, Scientific and Cultural Organization. UNWTO (2022) International Tourism Highlights, 2022 Edition. Madrid: World Tourism Organization. doi:10.18111/9789284423743 Wang, N. (2020) Rethinking authenticity in tourism experience revisited. Tourism Recreation Research, 45(3), pp. 293 to 305. doi:10.1080/02508281.2020.1723552 Wang, S., Berbekova, A. and Uysal, M. (2021) Wellbeing and heritage tourism: what makes heritage tourists happy. Journal of Travel Research, 60(6), pp. 1257 to 1272. doi:10.1177/0047287520951641 Winter, T. (2021) The geocultural power of heritage in Asia. International Journal of Heritage Studies, 27(3), pp. 232 to 246. doi:10.1080/13527258.2020.1767207 Zhu, Y. and Li, N. (2022) Groping for stones to cross the river: governing heritage in East and Southeast Asia. International Journal of Heritage Studies, 28(2), pp. 155 to 169. doi:10.1080/13527258.2021.1941199 #cultural_heritage_tourism #Thailand #commodification #authenticity #community_based_tourism #intangible_cultural_heritage #sustainable_tourism #heritage_conservation #Chiang_Mai #Ayutthaya #Sukhothai #hill_tribe_tourism #Songkran #Loy_Krathong #Thai_temples #living_heritage #tourism_gentrification #culinary_heritage #ethnic_identity #community_wellbeing #tourism_policy #Southeast_Asia_tourism #heritage_management #responsible_tourism #post_pandemic_tourism

Cultural institutions such as #museums, #archives, #libraries and #heritage sites are moving fast into the age of generative artificial intelligence. Curators, educators and volunteers now use text-to-image tools, large language models and 3D reconstruction networks to describe, restore and re-imagine #artifacts. Much of this work happens through official projects, but a growing share happens outside of any formal plan. Staff open free chatbots on personal accounts, upload photos of #objects into third-party platforms and paste catalog notes into online tools. This informal, hidden and unmanaged use of #generative_AI is often called #Shadow_AI. This paper studies how Shadow AI is entering cultural institutions and what it means for the goal of protecting #historical_authenticity. It brings together recent research on #AI_hallucinations, #algorithmic_bias, #deepfake avatars, digital twins of #heritage_sites, and cultural policy. The paper argues that unregulated generative models can quietly rewrite the visual and textual record of the past. Fabricated details slip into reconstructions of coins, vessels, buildings and faces. Non-Western traditions are pushed into a #digital_silence because training data favors Western sources. Copyright and #provenance become blurred when synthetic outputs are stored beside real records. The paper offers a working definition of Shadow AI for the #GLAM sector, maps the main risks, and proposes a practical framework built on documentation, curator authority, transparent labeling, and #ethical_governance. The goal is not to reject #AI, but to place clear limits so that machines assist memory instead of quietly replacing it. Keywords: Shadow AI, cultural heritage, generative models, historical authenticity, museums, digital reconstruction, AI hallucinations, algorithmic bias, GLAM sector, AI governance. 1. Introduction Every museum, archive and library holds a promise. The promise is that the object on the shelf, the paper in the folder or the coin in the case is real, and that the label next to it tells the truth about what it is, where it came from and who made it. This promise is called #authenticity. It is the reason people trust cultural institutions more than they trust many other sources of information about the past (Spennemann, 2025). Over the last three years, this promise has come under a new kind of pressure. #Generative_AI, in the form of tools such as ChatGPT, Midjourney, Stable Diffusion and countless smaller models, can now produce images, texts, 3D shapes, voices and even facial animations that look and sound like real historical material (Hazan, 2023; Spennemann, 2024). These tools are open, cheap and easy. A curator with a laptop can prompt a system to fill in a broken mosaic, colorize a damaged photograph, translate a fragile document or generate a talking avatar of a person who lived five hundred years ago (Wilkinson et al., 2026; Di Martino et al., 2025). When such use is planned, tested and approved by an institution, it can be described as #official_AI. But much of what is happening in cultural work today is not planned or approved. Staff members, students, interns and volunteers are trying these tools on their own, often on personal accounts, without informing their managers or their #curators. Vendors add generative features to existing software without asking. This hidden and informal layer of activity is what this paper calls #Shadow_AI in cultural institutions. The idea is close to what management scholars have called shadow user innovation, in which employees adopt powerful digital tools faster than their organizations can supervise them (Waters-Lynch et al., 2024). The problem is not just that Shadow AI breaks rules. The deeper problem is that it can quietly change the historical record. A single restored image, once uploaded into an online album, can travel further than the original artifact. A #hallucinated detail in a chatbot reply about a temple can be copied into a school project, then into a wiki, then into a training dataset for the next model. Over time, cultural memory can drift away from evidence and toward statistical guesswork (Salah, 2026; Hazan, 2025). This paper asks three related questions. First, what forms does Shadow AI take inside cultural institutions today? Second, what are the specific risks that unregulated generative models pose to historical authenticity when they are used to digitally recreate #artifacts? Third, what kind of #governance and practice can institutions adopt to reduce these risks without shutting the door on the real benefits of AI? The paper is written for a student and early-career audience. It uses simple language. It draws on peer-reviewed articles and expert commentary published mostly in the last five years. It does not treat AI as either a savior or an enemy. Instead it treats generative tools as powerful but noisy instruments that need clear rules before they touch #cultural_memory. 2. Background and Literature Review 2.1 The rise of generative AI in the heritage sector The public debate around generative AI moved sharply after 2022. Spennemann (2024) describes the first half of 2023 as a moment when generative artificial intelligence became the most significant cross-cultural global disruptor since the World Wide Web. He argues that it will affect how cultural heritage is managed and practiced, both by helping with analysis and by producing new texts and images, especially reconstructions of objects and sites. Museums and #libraries have been experimenting with machine learning for more than a decade. Early tools helped with #cataloguing, image tagging and search (Dikow et al., 2023). What is new is that today the same institutions can also create new material at scale. Tools such as Generative Adversarial Networks (GANs) and diffusion models can produce believable images of coins, pots, walls and portraits. Altaweel, Khelifi and Zafar (2024) tested GANs on ancient Roman coins and found that, in many cases, the generated coins were virtually indistinguishable from the originals. Cardarelli (2024) applied a similar approach to more than four thousand Italian Bronze and Early Iron Age vessels, reconstructing whole pots from single fragments. At the same time, projects have used text-to-image systems to rebuild ruined #architecture. Arzomand, Rustell and Kalganova (2024) explored how text-to-image AI can help visualize damaged heritage sites based on written descriptions. Song and Yang (2025) used a modified CycleGAN with cultural consistency constraints to reconstruct traditional ceramic patterns, including blue-and-white porcelain and celadon. Su, Osman and Islam (2026) reviewed generative AI applications for built heritage along the Belt and Road Initiative and reported both promising documentation results and serious risks linked to bias and data colonialism. Generative models are also being applied to intangible heritage. Chen, Zhong and Li (2025) studied the use of AI in the living inheritance of Qiaoxiang intangible cultural heritage in China, where the technology is now shaping how rituals, crafts and stories are recorded and transmitted. Similar work is being done for lost art traditions and interior decorative arts (Wei et al., 2026). 2.2 What "authenticity" means in cultural institutions To understand the risks, it helps to be clear about the word #authenticity. In heritage practice, an authentic object is one whose identity, origin and history can be verified through documentation, expert examination and scientific analysis (Spennemann, 2025). Spennemann (2025) argues that in an age of #deepfakes and generative content, physical objects in museum and archival collections gain renewed significance as anchors of truth, because they carry unique features such as wear, residue and modifications that can be tested. Digital objects are different. A file can be perfect, endlessly copied and easily edited. Salah (2026) proposes a Reflexive Authenticity Framework in which authenticity is redefined not as an unchanging property of an object, but as transparency about how a digital item was made and by whom. She warns of #epistemic_drift, where unverified synthetic narratives circulate faster than critical validation. Chen and colleagues (2025) describe a related shift from "historical authenticity" to "digital authenticity" and warn that generative AI can move power from human inheritors of a tradition to algorithmic agents. In their view, unless the human community remains the "value legislator" over its own heritage, the technology can hollow out cultural subjectivity even while producing dazzling reconstructions. 2.3 Shadow AI: from shadow IT to shadow user innovation The term #Shadow_IT has been used in management studies for a long time. It describes the practice of employees using software or devices that their organization has not officially approved. Waters-Lynch and colleagues (2024) extend this idea to generative AI. In their theory of shadow user innovation, workers experiment with powerful #generative_tools inside firms, producing new workflows that management often does not see. The authors argue that this behavior can drive innovation but also creates risks around data leakage, quality control and accountability. In cultural institutions, the same pattern is now visible. Frost and Vargas (2024) note that best practice is still emerging and that much of the work with AI in museums is bottom up, with practitioners exploring the potential rather than leaders directing this work. They quote a museum digital director who says that museums are only starting to recognize the need for a policy document that is reviewed every six months, while other internal policies often go years without review. In this environment, Shadow AI is not a rare event. It is the everyday state of the sector. 2.4 Known risks of generative AI in heritage work A growing body of literature has begun to map the risks of generative AI in cultural work. Foka and Griffin (2024) argue that bias is inherent in cultural heritage collections and their digital versions, and that AI pipelines can amplify this bias when models are trained on unbalanced datasets. Wang (2026) describes a #digital_silence, in which large language models trained mostly on Western sources marginalize indigenous histories and non-textual cultural traditions. Hazan (2023, 2025) writes from a museum practitioner point of view. In her 2023 paper "The Dance of the Doppelgangers," she describes AI images as doppelgangers that seem to emerge from an uncanny valley. In her 2025 paper on the "Ouroboros effect," she warns that AI is devouring the very pulse of human creativity only to produce a generic and mediocre residue that gravitates to the average. She connects this to Walter Benjamin's concept of the aura and worries that endless synthetic reproduction can strip cultural production of its unique existence in time and place. Rimmer (2026) analyzes copyright litigation over AI training projects and its consequences for the GLAM sector, including questions about indigenous intellectual property. Favela and Amon (2023) discuss the ethics of human digital twins and describe how high-fidelity digital replicas of real people can create counterfeit personhood problems and privacy violations. Wilkinson and colleagues (2026) look at #deepfake avatar creators used in archaeological facial depiction and identify affordances such as realism and engagement, alongside risks of misrepresentation and cultural insensitivity. The problem of #AI_hallucinations sits at the center of all of this. Boyer and Boyer (2024) define hallucination as the pernicious problem that generated content may be substantively incorrect while appearing to be both authoritative and correct. Kushwah and Dave (2025) argue that hallucinations erode user trust and raise ethical issues, especially in high-stakes fields. Mzili and colleagues (2026) provide a review of hallucination in retrieval-augmented and generative systems, identifying failure modes such as unsupported content, weak grounding, context conflict, false citations and misleading confidence. 3. Conceptual Framework: Defining Shadow AI in Cultural Institutions Building on this literature, this paper defines #Shadow_AI in cultural institutions as follows. Shadow AI in cultural institutions is the use of generative or other AI systems in the work of a museum, library, archive, gallery, historic site or related organization when that use is not covered by a formal policy, is not documented in the institution's records, and is not visible to those responsible for #collections_management, ethics review or public communication. Shadow AI can take at least four forms in the #GLAM sector: Personal-tool use. Staff open free or paid AI tools on their own accounts and use them for institutional tasks such as writing labels, translating documents, generating images for exhibitions or drafting educational content. There is no audit trail inside the institution. Hidden vendor features. Commercial software already used by the institution (for example, image editors, transcription tools, ticketing systems or content management systems) silently adds #generative_AI features. Staff use these features without knowing that outputs are AI-produced. Volunteer or crowdsourced use. Volunteers, docents, community researchers or contract workers use generative tools to complete work for the institution, without any oversight of the models, prompts or outputs. Research-and-development leakage. Individual researchers, students or interns use generative models for experiments, then quietly integrate the results into public catalogs, exhibitions or teaching materials. In each of these forms, the same three ingredients appear together: a powerful #generative_model, a lack of institutional oversight, and a cultural output that can influence how the public understands the past. It is this combination that makes Shadow AI different from ordinary use of new technology. 4. Methodology This paper is a narrative literature review. It follows the general approach used in earlier reviews of AI ethics in heritage (Tiribelli et al., 2024; Pansoni et al., 2023) rather than a strict systematic review method. It focuses on recent, peer-reviewed and expert commentary sources on generative AI and cultural heritage. Sources were gathered through structured searches of academic databases using terms such as "generative AI and cultural heritage," "shadow AI," "AI hallucination," "digital reconstruction and authenticity," "deepfake and museums," and "AI governance and GLAM." The search was limited mostly to publications from 2021 onwards, in line with the emergence of powerful public generative models. Older classical references, such as Walter Benjamin's concept of the aura, are used through recent secondary discussions rather than direct citation. Sources were screened for two criteria. First, they had to focus on generative AI or on cultural institutions in a serious way. Second, they had to add either evidence (case studies, experiments or policy analysis) or a clear conceptual argument. About sixty sources were read in detail, of which around thirty are cited here. Because Shadow AI is by definition invisible, direct measurement is very hard. This paper therefore does not attempt to count how many curators are secretly using ChatGPT. Instead, it uses the literature to build a conceptual map of the phenomenon and its consequences. This is a normal method for early-stage work on emerging technology (Frost and Vargas, 2024). The main limitation of this approach is discussed in Section 8. 5. Findings and Discussion 5.1 Digital recreation of artifacts: from helpful tool to invisible fiction Generative reconstruction of #artifacts is one of the most visible uses of AI in heritage. The results are often striking. Altaweel and colleagues (2024) show that GAN-restored Roman coins can be almost indistinguishable from the originals, which is very useful when a #curator needs to visualize a damaged object for public display. Cardarelli (2024) demonstrates that whole vessels can be reconstructed from small fragments with high statistical accuracy, and expert evaluators found the results promising. Song and Yang (2025) achieve good scores on Frechet Inception Distance and Structural Similarity Index for ceramic patterns, and their model also passes a "cultural consistency" test done by experts. These outcomes are real gains for #digital_preservation. When done in the open, with clear labels and expert review, generative reconstruction can support conservation planning, education and public engagement (Bansod et al., 2026; Mogre et al., 2025). It can even help save records of sites that have been damaged by war or natural disasters (Wilkinson et al., 2026; Multi-Modal Reconstruction of Lost Cultural Heritage via Generative Adversarial Networks, 2026). The risk begins when the same tools are used inside Shadow AI. A staff member, working alone and against a deadline, prompts an image model to "complete" a broken statue and drops the result into a public online album. There is no note in the metadata that the image is synthetic. There is no expert review. Ten years later, another researcher pulls the image from the online catalog and treats it as an accurate reference. The synthetic image has entered the #historical_record. Cardarelli (2024) himself is careful about this. His framework is open-source, documented and evaluated by specialists. Altaweel and colleagues (2024) also discuss where GANs might fail. But in Shadow AI contexts, the safeguards that these researchers build into their work are usually stripped away. The outputs move faster than the caveats. 5.2 The hallucination problem in a heritage setting AI systems do not just reconstruct; they also invent. Boyer and Boyer (2024) explain that generative models can produce content that is substantively incorrect while appearing authoritative and correct. Mzili and colleagues (2026) list the many ways this can happen, including unsupported content, weak grounding, context conflict, false citations and misleading confidence. For most everyday users, a hallucinated fact is annoying but not dangerous. For a museum or archive, a #hallucination can be a serious threat. If a chatbot embedded in a gallery describes an object using invented dates, invented rulers or invented ceremonies, visitors may believe those details. Even more dangerous, museum staff themselves may repeat those details in labels, tours and school talks (Hazan, 2025). Liu and Meng (2025) offer a more philosophical view, arguing that hallucinations are not simply technical flaws but cultural practices that continue older patterns of #media transformation. Olson (2025) suggests that AI hallucinations can even be used as pedagogical opportunities to study bias in visual representation. These are interesting arguments, but they depend on hallucinations being made visible. In Shadow AI, hallucinations are usually hidden. They enter the record without any warning label. Chen and colleagues (2025) call this risk "hallucinatory narratives" and argue that they undermine the authenticity of intangible cultural heritage by mixing invented ceremonies, symbols or meanings with real ones. Salah (2026) uses the term #epistemic_drift to describe how these narratives can outrun the ability of experts to correct them. Hallucinations in cultural work have a particular shape. They are usually not wild or obviously wrong. They are small, plausible, and consistent with the surface style of academic writing. A generative model asked about a minor medieval saint may invent a plausible birthplace, a plausible year of canonization and a plausible relic. Each of these details looks similar to what a real record would contain. When a curator reads the output, the details do not trigger alarm. When the same details are copied into a public label, a talk or a lesson plan, they take on the authority of the institution that displayed them. This is why Kushwah and Dave (2025) argue that hallucinations must be treated as usability and reliability concerns, not as small technical bugs. In a Shadow AI setting, the person who could catch the mistake is not present, because the whole point of Shadow AI is that no one is watching. 5.3 Bias and the digital silence of non-Western heritage Bias in AI systems has become a well-known topic. In cultural work, its effects are especially deep. Foka and Griffin (2024) argue that bias is inherent in cultural heritage collections themselves, because those collections reflect the choices, blind spots and power structures of their institutions. When such collections are used to train AI models, the models learn to reproduce these choices as if they were natural. Wang (2026) frames this as #algorithmic_hegemony, in which the shift from search engines to generative "answer engines" pushes non-Western cultural heritage into a #digital_silence. Because large language models are trained mostly on English and other major European-language texts, they know more about Renaissance painters than about weaving traditions from West Africa or ritual objects from Southeast Asia. They can generate confident content about the former and vague or invented content about the latter. Sahoo (2026) describes this problem as #memetic_capture, an AI-driven form of cultural disempowerment where the very preferences and values through which humans recognize their own culture are shaped by systems that do not represent them. Cui (2026) warns of similar risks in K-12 heritage education, where uncritical use of #agentic_AI and digital twins can amplify biases and perpetuate colonial narratives if community data sovereignty is not respected. Su, Osman and Islam (2026) show how these problems become concrete in large infrastructure programs such as the Belt and Road Initiative, where generative AI is used for documentation of heritage sites but where governance gaps and epistemic asymmetries can create data colonialism if left unchecked. Ali (2024) analyzes governance of cultural heritage management in Iraq and finds that although AI could help detect illegal excavations and monitor sites, its use is limited by the absence of a digital-heritage strategy and clear legal authority. When Shadow AI is added to this picture, the bias becomes almost invisible. A curator in a small institution may not know that the free chatbot they use has almost no reliable data about the local tradition they are trying to describe. The tool will still produce a fluent, well-formed answer. The answer may quietly replace older, deeper community knowledge. 5.4 Deepfakes, faces and the ethics of speaking for the dead One of the most emotional uses of generative AI in the #heritage sector is the creation of talking, moving avatars of historical figures. Wilkinson and colleagues (2026) study the use of AI-powered deepfake avatar systems such as MetaHuman Creator in archaeological facial depiction. They show that such tools can produce realistic four-dimensional portraits, augmented reality experiences and interactive audience engagement, and they argue that the tools should be seen as augmentation rather than replacement of expert practice. Zaramella and colleagues (2023) describe a smartphone application that lets museum visitors create #deepfakes of paintings so that portraits appear to speak. Di Martino and colleagues (2025) apply large language models and photographic animation to build an immersive virtual museum around the Albanian writer Musine Kokalari. They discuss the epistemological and ethical challenges of simulating historical figures and propose a model for conscious and responsible use. Bernasconi and Marfia (2025) go further by producing automated self-descriptions of artworks in which a detected face inside a painting animates and explains its own content. They raise concerns about cultural biases embedded in large language models and about the risks of educational deepfakes. Favela and Amon (2023) look at human #digital_twins in general and argue that they raise pressing issues of consent and privacy. Even when a historical figure is long dead, giving them a voice and a moving face has consequences. The avatar may say things the real person never said. It may express opinions that fit the values of the person who built the avatar rather than the values of the person portrayed. When such avatars are made through Shadow AI, without any ethical review, they can slip into public education very quickly. 5.5 Copyright, provenance and the blurring of the record Rimmer (2026) surveys the wave of copyright litigation over AI training projects. He examines the impact on literary works, journalism, databases, artistic works, music, performances and films, and looks in particular at how indigenous intellectual property will be affected. His conclusion is that the GLAM sector faces both risk and opportunity from these legal battles, and that any AI project inside a cultural institution must consider transparency around training materials. Hazan (2023, 2025) points out that even when copyright is respected, #provenance becomes tangled when AI outputs enter museum records. If a curator uses generative AI to enhance an old photograph, the enhanced version is neither the original nor a fully new work. It is a hybrid. Without clear metadata, the hybrid becomes hard to trace. Tiribelli and colleagues (2024) argue that ethical questions around AI in cultural heritage span authenticity, subjectivity, interpretation biases and legal concerns about authorship. Pansoni and colleagues (2023) propose ethical principles for AI in cultural heritage such as shared responsibility, meaningful participation, explainability, accessibility, sustainability, reliability and dignity. In Shadow AI, all of these principles are broken by default. There is no shared responsibility because there is no record. There is no meaningful participation because communities are not consulted. There is no explainability because the tool used is often a black box. There is no accessibility because the process is invisible to the public. Yang (2024) describes an uncanny quality that appears when audiovisual archives are re-encountered through AI systems. Old materials come back in slightly changed forms, and users cannot always tell what has been kept and what has been altered. This uncanny quality can be creative, but in cultural institutions it must be signposted. Without a clear signal that an item has been touched by a model, viewers cannot form correct beliefs about what they are seeing. Over time, small unsignposted changes add up. The catalog stops describing objects and starts describing a shared cloud of statistical guesses that are anchored, ever more loosely, to the real things sitting on the shelves. 5.6 The Ouroboros effect and the flattening of culture Hazan (2025) borrows the ancient image of the #Ouroboros, the snake that eats its own tail, to describe a worrying feedback loop. Generative models are trained on data scraped from the internet. Increasingly, that data includes content that was itself produced by earlier models. The result is a self-consuming loop in which cultural production becomes more and more average, more and more generic, and less and less connected to real human creativity. For cultural institutions, this loop has a particular danger. If a museum uploads AI-restored images of coins, and those images become part of the next model's training data, then future reconstructions will be based partly on past reconstructions rather than on original evidence. Errors will accumulate. Small mistakes in the first generation may become confident features in the third. Salah (2026) calls this "algorithmic memory" and stresses the need for reflexive verification practices. Spennemann (2024) warns that the more speculative interpretations produced by generative models may in fact become manifestations of cultural heritage in their own right, and he asks whether such outputs are still human cultural heritage or a new kind of AI cultural heritage. The question sounds abstract, but for archivists it is very practical. Do we keep AI-generated images in the same catalog as photographs of real objects? Do we mark them differently? Do we keep them at all? 5.7 Cybersecurity, integrity and the danger of malicious Shadow AI Not all Shadow AI is innocent. Koukopoulos, Koukopoulos and Fidas (2025) describe how the growing digitalization of cultural heritage leaves institutions vulnerable to phishing, data breaches and #falsification of digital archives. They propose an AI-based cybersecurity framework using phishing detection, blockchain-based data integrity and AI-based legal advice. The threat is not only external. An insider using a Shadow AI tool could, by mistake or by design, insert fake documents into a digital archive. Because generative systems can produce documents that look genuine, including old handwriting, letterheads and scripts, the level of forgery available to a single person is now much higher than before. Spennemann (2025) responds by calling for robust documentation standards such as blockchain-secured metadata, forensic-style photographic records and transparent curatorial histories. Su and colleagues (2026) suggest that #explainable_AI methods, such as vision transformers combined with SHapley explanations, can help curators authenticate artwork in real museum environments. Their model reached 91.4 percent accuracy on a museum dataset. Such tools can push back against #forgery, but only if institutions actually adopt them and integrate them into policy. 5.8 The impact on cultural workers Shadow AI is not just a technical or ethical problem. It also affects the daily working lives of #cultural_workers. Frost and Vargas (2024) argue that little attention has been paid to how AI is emotionally impacting cultural workers as their organizations move to use these tools. They warn of an "AI divide" between institutions that can invest in training and infrastructure and those that cannot. They also note that many roles in the sector are held by women in low-paid administrative positions, and that generative AI is especially good at the kind of standardized text tasks that these roles involve. When AI is introduced without policy, there is a risk that decisions about which jobs to automate are made informally, without clear ethical scrutiny. Shadow AI, in this sense, is not only a threat to the historical record. It can also be a threat to fair work. 6. Case Illustrations The next three short cases give concrete pictures of how Shadow AI can appear in cultural institutions. They are constructed as composite examples that reflect patterns described in the literature (Hazan, 2023, 2025; Spennemann, 2024, 2025; Frost and Vargas, 2024; Cui, 2026; Ali, 2024). They are not detailed narratives of any single institution. 6.1 The completed statue A regional #museum holds a small marble torso from a lost temple. A visiting student intern uses a personal account on a text-to-image tool to imagine the statue when it was complete. The student is proud of the result and shares it in a museum blog post. The blog is picked up by a local newspaper. Within a year, the reconstructed image is cited in a school textbook as an accurate representation of the ancient statue. In this case, three things happened. First, the reconstruction was created without expert review. Second, no label made clear that the image was AI-generated. Third, the image left the museum's control and became part of a wider chain of cultural memory. Altaweel and colleagues (2024) show how easy it is for such synthetic images to be mistaken for real evidence when guardrails are absent. 6.2 The talking founder A university #library plans to build an interactive exhibit for its 200th anniversary. A volunteer uses a free deepfake tool to make the university's founder appear to speak on video. The script is drafted by a chatbot. Neither the tool nor the script is reviewed by historians or ethicists. The video praises the founder in language that reflects modern values rather than the founder's actual writings. Wilkinson and colleagues (2026) and Di Martino and colleagues (2025) show how deepfake avatars can enrich engagement, but they insist on ethical review and clear disclaimers. Favela and Amon (2023) argue that giving a face and voice to a real person, even a dead one, raises questions of personhood. In this Shadow AI case, none of those safeguards were followed. 6.3 The rewritten ritual A community #archive documenting an indigenous festival receives a request from a research group to prepare a public multimedia site. Under a tight deadline, an archivist uses a large language model to summarize dozens of ceremonial descriptions. The model, trained mostly on English-language sources, fills gaps by combining details from unrelated traditions. The final text uses confident language and reads well, so nobody notices the small drifts. This case follows the concerns of Wang (2026) about #digital_silence, of Sahoo (2026) about memetic capture and of Chen and colleagues (2025) about the shift from historical authenticity to digital authenticity. Cui (2026) points out that in such situations community data sovereignty should override speed. 7. Toward a Framework for Governing Shadow AI in Cultural Institutions Based on the findings above, this section outlines a framework that cultural institutions can use to bring Shadow AI into the light without losing the useful energy that fuels it. The framework has five layers. It draws on the ethical principles proposed by Pansoni and colleagues (2023) and Tiribelli and colleagues (2024), the reflexive authenticity idea of Salah (2026), and governance recommendations from Frost and Vargas (2024), Ali (2024) and Sahoo (2026). 7.1 Layer one: recognize that Shadow AI already exists The first step is honesty. Institutions must accept that staff and volunteers are already using generative tools, whether or not policy allows it. Frost and Vargas (2024) and Dikow and colleagues (2023) both underline that AI adoption in the GLAM sector is largely bottom up. Denying this fact only pushes the practice further into the shadows. A short internal survey can be used to ask staff, in a non-punitive way, how often they use generative tools, for what purposes, on which platforms and how they judge the quality of the outputs. The results should feed into policy design, not into punishment. 7.2 Layer two: create a light, living policy A policy on AI in cultural institutions should not be a heavy document that arrives once and never changes. Frost and Vargas (2024) quote practitioners who ask for policies that are reviewed every six months. The policy should cover at least the following points: Which tools are approved for which tasks, and which are not. Which categories of data can and cannot be uploaded to third-party AI tools. How AI-generated content must be labeled in catalogs, exhibitions and educational materials. What documentation must be kept about prompts, models used and human review steps. How community stakeholders are consulted when AI is used for content about their heritage. The policy should be written in simple language and should be discussed with staff, not just sent to them by email. 7.3 Layer three: strengthen documentation and provenance Spennemann (2025) argues that in an age of deepfakes, museums must adopt robust documentation standards such as blockchain-secured metadata, forensic-style photographic records and transparent curatorial histories. Salah (2026) proposes that authenticity itself be redefined as transparency about how a digital item was created. For every AI-generated or AI-assisted item entering a #collection, a minimum record should include: The identity of the operator (staff member or contractor). The model or platform used, with version if known. The prompt or set of prompts used. The training data source, when available. The purpose of the output. A human review note. A visible label in any public display of the item. Barzaghi and colleagues (2024) propose a FAIR management approach for 3D data in cultural heritage that can be adapted for this purpose. Colace and colleagues (2026) argue that digital twins of #heritage should support predictive conservation, but these systems can only be trusted if their data lineage is transparent. 7.4 Layer four: keep expert judgment at the center Bansod and colleagues (2026), Mogre and colleagues (2025) and Song and Yang (2025) all show that expert review is important for judging the quality of AI-generated reconstructions. Su and colleagues (2026) demonstrate the power of explainable AI in authentication tasks but note that expert alignment remains a key measure of trust. Cultural institutions should require that no AI-generated content, especially reconstructions of #artifacts, faces or texts, enters public display or catalog without human expert sign-off. The role of the expert is not to reject AI, but to verify what it produces. Chen and colleagues (2025) emphasize the need for a "human-machine collaboration" model in intangible heritage, where inheritors of a tradition are treated as value legislators, not passive subjects of the algorithm. This is a strong idea that can be extended to all cultural work. 7.5 Layer five: build cultural AI literacy for staff and audiences Frost and Vargas (2024) describe demand for #AI_literacy as one of the main drivers shaping the future of cultural work. Cui (2026) proposes an ethical-pedagogical framework for K-12 digital heritage education in the era of agentic AI. Both suggest that literacy is not just about tools; it is about understanding limits. Staff need training on what generative models can and cannot do, on how to spot hallucinations, on how bias enters outputs and on how to write prompts that preserve rather than erase context. Audiences, in turn, need clear labels and short explanations when AI has been used in an exhibit. Olson (2025) shows that hallucinations themselves can become teaching material about how #visual_knowledge is built. 8. Limitations of the Study This paper has several limits that readers should keep in mind. First, it relies on public literature. Because Shadow AI is by definition informal, most of it never appears in published sources. The paper therefore describes a phenomenon that is only partly documented. Direct interviews with cultural workers, ideally under conditions of anonymity, would help future research. Second, the paper focuses on English-language literature. This choice creates its own bias, exactly the kind of bias described by Wang (2026) and Sahoo (2026). Studies published in other languages, especially in the languages of countries facing the strongest pressure from generative AI, may offer different perspectives. Third, the field is moving very fast. Between the time this paper is written and the time it is read, new tools, new models and new legal decisions will appear. The framework proposed here is meant to be revisable, not final. Fourth, the paper does not offer quantitative data. It draws lessons from qualitative studies, case reports and expert commentary. Future work should attempt to measure the scale of Shadow AI use in the GLAM sector through anonymous surveys. 9. Recommendations for Different Actors Different actors in and around cultural institutions can take different steps. Institutional leaders should treat Shadow AI as a governance problem, not just a technical one. They should draft a short living AI policy, invest in staff training and require documentation of all AI-assisted outputs. They should also budget for expert review time (Frost and Vargas, 2024; Ali, 2024). Curators and archivists should insist on human sign-off before any AI-generated content enters catalogs or displays. They should push for clear labeling and for open documentation of prompts and models used. They should protect their role as value legislators over their #collections (Chen et al., 2025; Salah, 2026). Researchers should publish more detailed evaluations of hallucination rates and bias patterns in heritage-specific AI systems. They should build benchmarks that reflect non-Western and indigenous heritage as well as canonical Western material (Wang, 2026; Sahoo, 2026; Mzili et al., 2026). Policy makers should create sector-specific rules for AI in cultural heritage, alongside more general AI laws. Ali (2024) shows how such rules can integrate satellite monitoring, digital twins and predictive maintenance into legal frameworks. Sahoo (2026) proposes a Cultural Pluralistic Governance Framework that could inform such rules. Educators, especially in library and #information_science and in museum studies, should include AI literacy modules in their curricula. Cui (2026) shows how this can be done even at the K-12 level. Students entering the sector today will spend their entire careers working with generative systems. Communities of origin should be brought into decisions about how their heritage is represented in AI systems. This is not a courtesy. It is a condition of authentic representation in a technological environment where memory itself is being restructured (Chen et al., 2025; Cui, 2026; Sahoo, 2026). Vendors of software used by cultural institutions should adopt clear disclosure rules. Whenever a new generative feature is added to a product, users should be told, in plain language, what the feature does, what data it may send outside the institution and how to switch it off. This kind of disclosure is one of the simplest ways to shrink the space in which Shadow AI can grow. If practitioners can see the AI features embedded in their daily tools, they can make informed choices, and the institution can bring those choices into its policy. Without such disclosure, staff cannot avoid Shadow AI even when they try to. Funders of heritage projects also have a role. Grants for digital heritage work should require a short AI ethics statement in the same way that many now require data management plans. Applicants should describe which tools they intend to use, how outputs will be labeled and how community consent will be handled. Colace and colleagues (2026) and Barzaghi and colleagues (2024) show that responsible data pipelines are possible when funding structures reward them. 10. Discussion: Why This Matters Beyond Museums A #reader might ask why Shadow AI in cultural institutions matters to anyone outside those institutions. The answer touches everyone. Cultural institutions are, in effect, the memory infrastructure of societies. When people want to know what a coin from a lost empire looked like, what a founding document actually said, or how a traditional ceremony was performed, they turn to museums, archives and libraries. If those institutions themselves become uncertain about which of their holdings are real, restored, reconstructed or invented, then the wider #public_trust in historical knowledge weakens. Spennemann (2025) argues that in a world of deepfakes and AI-generated content, real physical objects gain new value as anchors of truth. But those objects gain that value only if they are surrounded by clear, honest and well-governed documentation. Shadow AI erodes this documentation from the inside. The problem is also political. Sahoo (2026) shows that culture is a special channel for gradual human disempowerment by AI, because cultural displacement attacks the preferences and values through which humans recognize and resist disempowerment. If the reconstructions and stories that shape the way people see themselves are quietly produced by a small number of #foundation_models trained on unequal data, then real cultural pluralism becomes harder to sustain. At the same time, it would be a mistake to reject AI outright. Cardarelli (2024), Altaweel and colleagues (2024) and Song and Yang (2025) show that AI can help conserve fragile artifacts, recover lost patterns and support education. Corsi and Sonn (2026) describe a Jesuit historiography project where AI is used with human validation to support archival reconstruction in a transparent and traceable way. Barzaghi and colleagues (2024) show that responsible 3D data management is possible. These examples share one feature: they are not in the shadows. They are open about what the AI is doing. The lesson is that the enemy is not AI. The enemy is Shadow AI, in which the same tools operate without oversight and without accountability. Bringing the same tools into the light, with policies, documentation, expert review and community consent, changes their meaning completely. 11. Conclusion Cultural institutions are entering a moment as important as the arrival of photography or the birth of the World Wide Web. Generative AI can reconstruct, translate, summarize, imagine and animate. It can also invent, misrepresent, silence and homogenize. When cultural workers use these tools inside official, transparent projects, the risks can be managed. When the same tools are used quietly, without policy or record, the risks quickly outpace the safeguards. This paper has argued that Shadow AI in cultural institutions is not a small side issue. It is a growing structural feature of the GLAM sector. It threatens historical authenticity in specific and identifiable ways: through invisible reconstructions, through hallucinated narratives, through amplified bias, through unclear provenance, through the flattening effects of feedback loops, and through the ethical dangers of deepfake avatars. It also affects cultural workers themselves, whose roles are being reshaped without full discussion. To respond, institutions do not need to ban generative tools. They need to bring the use of these tools into the open. They need light and living policies. They need strong documentation and clear labels. They need to keep expert judgment at the center. They need to invest in cultural AI literacy for staff and audiences. And they need to listen to the communities whose heritage they steward. If cultural institutions do this work now, they can turn Shadow AI into a manageable challenge. If they delay, the historical record will keep drifting, one hidden generation at a time, until the promise of authenticity on which the sector rests becomes hard to keep. The task is urgent but doable. The next few years will decide whether machines learn to serve #cultural_memory or quietly begin to rewrite it. Hashtags #Shadow_AI #Generative_AI #Cultural_Heritage #Historical_Authenticity #Museums_And_AI #GLAM_Sector #AI_Hallucinations #Digital_Reconstruction #Algorithmic_Bias #Deepfakes_In_Museums #Digital_Twins #AI_Governance #Heritage_Ethics #Ouroboros_Effect #Cultural_Memory References Ali, A. M. (2024). Governance of Cultural Heritage Management in Iraq in the Age of Artificial Intelligence. Journal of Ecohumanism. Altaweel, M., Khelifi, A., and Zafar, M. (2024). Using Generative AI for Reconstructing Cultural Artifacts: Examples Using Roman Coins. Journal of Computer Applications in Archaeology, 7(1). Arzomand, K., Rustell, M., and Kalganova, T. (2024). From ruins to reconstruction: Harnessing text-to-image AI for restoring historical architectures. Challenge Journal of Structural Mechanics. Bansod, H., et al. (2026). Generative AI for Reviving Lost Art Traditions. International Journal of Emerging Technologies and Innovative Research. Barzaghi, S., et al. (2024). A Proposal for a FAIR Management of 3D Data in Cultural Heritage: The Aldrovandi Digital Twin Case. Heritage, 7(1). Bernasconi, V., and Marfia, G. (2025). Speaking Images. A Novel Framework for the Automated Self-Description of Artworks. Digital Humanities Quarterly. Boyer, J., and Boyer, W. (2024). A Strategic Roadmap for Mitigating Generative Artificial Intelligence Hallucinations. Cureus Journal of Computer Science. Cardarelli, L. (2024). From fragments to digital wholeness: An AI generative approach to reconstructing archaeological vessels. Journal of Archaeological Science: Reports. Chen, S., Zhong, F., and Li, Y. (2025). Digital Authenticity and Cultural Subjectivity: Ethical Scrutiny and Practical Paths of Generative AI in the Living Inheritance of Qiaoxiang Intangible Cultural Heritage. International Journal of Frontiers in Sociology and Humanities. Colace, F., et al. (2026). Digital Twin for Cultural Heritage: A Computational Approach to Predictive Conservation. Journal of Cultural Heritage. Corsi, A., and Sonn, C. (2026). Pas de documents, pas d histoire? The Portal to Jesuit Studies as a Lab for AI-Driven Historiography. Journal of Jesuit Studies. Cui, Y. (2026). The Enriched Twin: An Updated Ethical-Pedagogical Framework for K-12 Digital Heritage Education in the Agentic AI Era. Research and Advances in Education. Di Martino, B., Colucci Cante, L., Graziano, M., Ceglie, S., and Sabani, M. (2025). Generative Artificial Intelligence for Historical-Archival Storytelling: Virtual and Immersive applications in the context of the Digital Humanities. Journal on Computing and Cultural Heritage. Dikow, R., et al. (2023). Developing responsible AI practices at the Smithsonian Institution. Research Ideas and Outcomes, 9. doi:10.3897/rio.9.e113334 Favela, L. H., and Amon, M. J. (2023). The ethics of human digital twins: Counterfeit people, personhood, and the right to privacy. Frontiers in Big Data. Foka, A., and Griffin, G. (2024). AI, Cultural Heritage, and Bias: Some Key Queries That Arise from the Use of GenAI. Heritage, 7(11), 287. doi:10.3390/heritage7110287 Frost, S., and Vargas, L. (2024). Cultural work, wellbeing, and AI. European Journal of Cultural Management and Policy. doi:10.3389/ejcmp.2024.12825 Hazan, S. (2023). The Dance of the Doppelgangers: AI and the cultural heritage community. Proceedings of EVA London. Hazan, S. (2025). Promise or Peril: The dual edges of AI in cultural heritage and the Ouroboros effect. Proceedings of EVA London. Koukopoulos, Z., Koukopoulos, D., and Fidas, C. A. (2025). An AI-Driven Cybersecurity Framework for Cultural Heritage Management Systems: Preventing Phishing Attacks and Ensuring Data Integrity. Journal on Computing and Cultural Heritage. Kushwah, S., and Dave, N. (2025). AI Hallucination and Strategies to Overcome: Enhancing Human-AI Interaction. International Journal of Human-Computer Interaction. Liu, X., and Meng, Y. (2025). An Analysis on AI Hallucination from the Perspective of Media Archaeology. Media, Culture and Society. Mogre, S. C., et al. (2025). Generative AI for Cultural Heritage Preservation Using AR and Data Science. Data Science and Management. Mzili, T., Mzili, I., Abatal, A., Oughannou, Z., Arya, A., and Kurdi, M. (2026). AI Hallucinations in Retrieval-Augmented and Generative Systems: A Rigorous Review of Definitions, Failure Mechanisms, Evaluation, and Mitigation Strategies. Edraak Journal. Olson, M. (2025). Error as insight: AI hallucinations and the pedagogical possibilities of generative visual misrepresentation. Visual Studies. Pansoni, S., Tiribelli, S., Paolanti, M., Frontoni, E., and Giovanola, B. (2023). Design of an Ethical Framework for Artificial Intelligence in Cultural Heritage. Proceedings of IEEE Metroliving. Rimmer, M. (2026). Night at the Artificial Museum: Copyright, Cultural Heritage, and Generative AI. Griffith Law Review. Sahoo, S. (2026). Memetic Capture: A Pluralistic Policy Framework for Governing AI-Driven Cultural Disempowerment. AI and Society. Salah, M. (2026). Algorithmic Memory: Towards Reflexive Authenticity in Cultural Heritage. AI and Antiquity. Song, B., and Yang, W. (2025). AI-based Generative Reconstruction of Intangible Ceramic Patterns for Digital Heritage Preservation. Heritage Science. Spennemann, D. (2024). Generative Artificial Intelligence, Human Agency and the Future of Cultural Heritage. Heritage, 7(7). Spennemann, D. (2025). Now more than ever: the role of museum and archival objects in an age of generative artificial intelligence. Collections: A Journal for Museum and Archives Professionals. doi:10.1108/CC-04-2025-0020 Su, D., Osman, N., and Islam, R. (2026). Generative AI and digital preservation of built heritage under the Belt and Road Initiative. Built Heritage. Su, L., Zhao, W., and Xu, Z. (2026). Intelligent Artwork Authentication Using Transfer Learning and Explainable AI in Cultural Heritage Systems. International Journal of Semantic Web and Information Systems. Tiribelli, S., Pansoni, S., Frontoni, E., and Giovanola, B. (2024). Ethics of Artificial Intelligence for Cultural Heritage: Opportunities and Challenges. IEEE Transactions on Technology and Society. doi:10.1109/tts.2024.3432407 Wang, J. (2026). The Digital Silence: Algorithmic Hegemony and the Visibility Crisis of Non-Western Cultural Heritage in the Era of Generative AI. AI and Society. Waters-Lynch, J., Allen, D. W. E., Potts, J., and Berg, C. (2024). Managing Generative AI in Firms: The Theory of Shadow User Innovation. SSRN Working Paper. Wei, C., et al. (2026). AIGC based digital heritage reconstruction of Qing interior canopies. Heritage Science. Wilkinson, C. M., Roughley, M., Liu, C. Y. J., Shrimpton, S., Davidson, C., and Dickinson, T. (2026). The Affordances of AI-Powered, Deepfake, Avatar Creator Systems in Archaeological Facial Depiction and the Related Changes in the Cultural Heritage Sector. Digital Applications in Archaeology and Cultural Heritage. Zaramella, M., et al. (2023). Why Don t You Speak? A Smartphone Application to Engage Museum Visitors Through Deepfakes Creation. Proceedings of the ACM International Conference on Interactive Media Experiences.