Technologies of reconstruction

1. THE KNOWLEDGE LANDSCAPE

In 2024, at the time of planning this special issue of Buildings & Cities, there was no shortage of scholarly literature about reconstruction. Quite the opposite. Over the previous decade alone, research explicitly concerned with post-disaster or post-conflict reconstruction and recovery had found ‘abode’ in more than 80 different peer-reviewed English-language journals, distributed across the built environment, social science, disaster studies, development, planning, engineering, public policy, economics, heritage, health, and peace and conflict studies.¹ Indeed, bounded literature reviews roughly across the same period reveal substantial bodies of work in almost every aspect of reconstruction (e.g. Mushtaha & Alaloul 2025; Serter & Gumusburun Ayalp 2024; Charles et al. 2021).

Besides, scholarly activities and publications were never the only location in which reconstruction knowledge resided. To state the obvious, a vast parallel literature is produced through practice. Post-disaster needs assessments, recovery and peacebuilding assessments, reconstruction plans, damage assessments, project evaluations, technical manuals, procurement documentation, policy reviews, humanitarian reports, and institutional lessons-learned exercises have been, and are being produced, by governments, the World Bank’s Global Facility for Disaster Reduction and Recovery (GFDRR), United Nations (UN) agencies, development banks, non-governmental organisations (NGOs) and other organisations after major crises. Databases such as EM-DAT provide information about disaster events and losses; the UN’s Disaster Risk Reduction (DRR) unit maintains disaster- and Sendai-related data infrastructures; ReliefWeb provides access to a large documentary record of humanitarian intervention; the World Bank maintains extensive repositories of project and policy documentation; and UN agencies and specialised organisations hold further collections concerned with displacement, infrastructure, damage assessment, heritage, housing and recovery.

What connects the academic and practice-related literature is a focus on the empirical and the particular. Study and documentation of reconstruction processes appear to organise according to two superimposing patterns: one characterised by the specialisation of the analysis, the other by the specificity of the episodes subject to such analysis.²

Within the vast data-and-information landscape thus generated, attempts at theory and synthesis are in short supply, as is comparative discussion. The broader the comparative sample, the more reconstruction tends to be reduced to a limited set of measurable variables; the more reconstruction is represented in its institutional, social, technical and political complexity, the smaller the number of cases that can practically be compared (e.g. Chrysanthakopoulos & Tagkalakis 2026; Lu & Xu 2015; Chang et al. 2010).³ Reconstruction, it could be said, is being defined through fragmented lenses and the many disciplines its various aspects can be associated with.

The analysis of book production supports this hypothesis. A survey undertaken in preparation for this special issue found that, since 2000, approximately 50 English-language scholarly books were published that substantially concerned post-disaster or post-conflict reconstruction and recovery in their various domains. The corpus includes work on housing, urban reconstruction, governance, institutional recovery, post-conflict development, cultural heritage, resilience, disaster justice, infrastructure and community recovery.⁴

Of significance is the form taken by these publications. Over half of them are edited collections, wherein the multidimensional character of the reconstruction subject is frequently addressed through the assembly of different forms of expertise or topic discussion, often—as indicated—around a common event or theme.⁵

Theory-building is scant.⁶ Within the same body of literature, only a handful of studies can be identified as primarily concerned with theoretical or conceptual explanations.⁷ Reconstruction knowledge, in other words, is produced more by accretion of discrete parts than by the integration of distinctive issues into a single, multidisciplinary explanatory framework. Rather than the object of a single discipline, reconstruction functions as a meeting place for many.

This condition has at least three institutional counterparts. The first is that, despite the quantity of the scholarship concerned with rebuilding, reconstruction has not consolidated around a major peer-reviewed journal devoted specifically to it.⁸ In this respect, the contrast with disaster studies is instructive. Disaster has progressively become an object around which an identifiable field of knowledge has organised itself, with plenty of specialised research outlets; reconstruction remains to a considerable extent an activity investigated from the perspective of other established fields (and their journals).⁹

The second aspect is that dedicated research can count on only a few physical loci. Its internationally distributed infrastructure comprises approximately 20 academic centres, institutes, laboratories, networks and research groups that identify post-disaster or post-conflict reconstruction as a substantial and explicit research theme.¹⁰ The third aspect is that there is no global repository with the reconstruction process itself as a primary unit.¹¹ Existing resources tend to contain either structured information about disasters or collections of documents produced in relation to particular events and programmes. What remains difficult to retrieve systematically and cross-reference is the relationship between the two: what was actually done after a particular disaster; by whom; under which institutional, financial and political arrangements; using what forms of technical and local knowledge; over what period; and with what physical, economic, social and political consequences.

2. THE ORGANISATION OF KNOWLEDGE

What should come out of this broad-brushed picture is that if there were ever a problem facing reconstruction research, it would not be one of information scarcity but rather one of knowledge organisation. An enormous empirical record exists, which has been, and continues to be, produced for different purposes, using different vocabularies, disciplinary categories, units of analysis and criteria of success. Each study or report can be as rigorous as it can within its own specialised domain whilst remaining difficult to connect with the others. The more attention reconstruction draws, the more the resulting knowledge appears to disperse across separate disciplinary, institutional and professional constellations.

The consequence is a peculiar combination of knowledge richness and epistemological fragmentation. Reconstruction is extensively studied without necessarily becoming better understood as an aggregate. New cases generate new studies, and these studies enrich the literature; but lessons do not automatically migrate between disasters, countries, disciplines or institutions. The literature may therefore accumulate bibliographically without being equally cumulative conceptually or operationally. We may have become increasingly effective at producing and citing reconstruction knowledge, but considerably less certain about how that knowledge should be assembled, compared and transferred.

This is particularly problematic because reconstruction is not divided in reality in the same way that its scholarship is divided intellectually. Housing reconstruction affects livelihoods; infrastructure influences economic recovery; land tenure conditions rebuilding possibilities; building regulations interact with local construction knowledge; financial mechanisms affect social inequality; political institutions determine the distribution of resources; reconstruction speed affects participation; and decisions taken during recovery alter future vulnerability. What appear as separate research objects are components of the same evolving process.

The quantity of available knowledge consequently brings into focus peculiar research questions that are both internal and external to reconstruction. If reconstruction already possesses substantial empirical, disciplinary and professional knowledge, how can these different forms of knowledge be brought into usable relation? Can the relationships themselves become an object of investigation? Can conceptual and informational structures be established through which what is already known may become mutually intelligible?

3. THIS SPECIAL ISSUE

It was on this basis, and in light of the stated mission of the journal—to publish scholarly investigations on the interplay between the different scales of the built environment and the range of matters that inform such interaction—that this special issue of Buildings & Cities has been set up to focus on the complexity of post-disaster reconstruction.

In the ad-hoc vocabulary underpinning the collection, complexity does not simply refer to the number or difficulty of the problems associated with reconstruction. Instead, it describes the condition for which heterogeneous dimensions—physical, technical, social, economic, political, institutional and cultural—operate simultaneously and interdependently across different spatial and temporal scales. Decisions made in one domain alter the conditions prevailing in others, and these relationships themselves evolve as reconstruction proceeds. Complexity, in this sense, resides less in the intrinsic nature of individual components than in the relationships between them.

In its association with reconstruction, the connection between complexity and knowledge fragmentation is direct and double-edged. Knowledge fragmentation is a consequence of the complexity of reconstruction as an endeavour and the necessity to specialise the ways in which it can be approached. However, the same fragmentation renders the recomposition of its subject matter problematic.

This paradox provided the starting point for this special issue. The intention was to ask what happens when these different forms of knowledge are brought into relation: whether insights developed within particular disciplines or cases can inform work conducted elsewhere and at other scales; whether recurring relationships and constraints can be recognised across otherwise different experiences; and, ultimately, whether theory and comparison can help organise this dispersed knowledge into a more intelligible and operationally useful whole.

The agenda was therefore exploratory rather than prescriptive. It did not assume that the complexity of reconstruction could—or should—be resolved through a universal model of successful rebuilding, nor that disciplinary differences and contextual specificity ought to be overcome. The question was instead whether reconstruction could be understood as something more than an aggregation of specialised problems: as a field of knowledge and practice in which those problems acquire significance through their relationships with one another.

The call for papers translated this general concern with integration into several more specific lines of enquiry. It invited consideration of the political dimensions operating simultaneously within reconstruction; the practical difficulties of mobilising multidisciplinarity within an increasingly specialised landscape; the possibility of transferring lessons derived from intrinsically idiosyncratic experiences; and, importantly, the role of theory in reconstruction studies and practice, including its relationship with the collection and organisation of empirical data.

Contributions were therefore sought that could use the analysis, synthesis or critical examination of particular experiences to engage with these larger questions, allowing empirical perspectives to contribute to broader conceptual and theoretical development. In keeping with the ambition of disciplinary synthesis, the call also encouraged attention to be paid to the different bases from which reconstruction is studied, and to the relationships—or lack thereof—between them. This included the questions concerning how knowledge is produced and organised, how conclusions are reached and validated, how the efficacy of procedures and tools is evaluated, how lessons are articulated and transferred, and how methods and outcomes can be critically assessed and translated into operational recommendations.

Finally, rather than attempting to delimit the field in advance, the call proposed 20 topics and questions as possible points of entry into these broader concerns. These were intended to encourage contributors to look at reconstruction processes not only from within their particular disciplinary or empirical domains, but also from without: to consider how the problem under examination related to other dimensions of reconstruction and, ultimately, to the larger question of how knowledge about reconstruction might be connected.

Over 50 paper proposals were received, which eventually became published as nine peer-reviewed journal articles and six editorially reviewed commentaries (Table 1).

The collected contributions provide considerable support for the initial hypothesis (and diagnosis) concerning the state of the field: that the increasing sophistication of reconstruction knowledge is paradoxically contributing to its fragmentation, and that there may be a need for forms of synthesis capable of giving operational coherence to otherwise separate efforts.

Significantly, each contribution did so without converging upon a single interpretation of reconstruction. They all approached it through different objects: housing prototypes, temporary accommodation, engineering expertise, institutional memory, spatial justice, social capital, heritage, energy infrastructure, preparedness, supply chains and urban recovery. If the diversity might have immediately appeared to reproduce precisely the fragmentation that motivated the special issue, something else emerged as a common trait. Almost every contribution began from a relatively bounded problem, to subsequently discover that the problem could not be understood within those boundaries. Reconstruction, in other words, repeatedly forced the analysis beyond the disciplinary or sectoral territory in which it began.

The clearest theoretical articulation of this condition is provided by Lizarralde et al. Their Critical Reconstruction Theory starts precisely from the absence of a comprehensive reconstruction theory and from the repeated reinvention of reconstruction through different framings: relief, development, urban renewal, building quality, empowerment, vulnerability reduction and adaptation. Rather than seeking to identify the correct framing, they conceptualise reconstruction as a political and normative negotiation over continuity and change: what should be retained, what should be transformed and what trade-offs can legitimately be accepted. Their contribution responds most directly to the call’s question about whether theory might bring coherence to fragmented knowledge. But it also suggests that such coherence cannot be achieved by reducing reconstruction to a single objective. Theory instead becomes a means of organising relationships between competing goals.

Several other contributions arrive at a similar conclusion empirically. Rankin et al., for example, begin with the implementation of earthquake-resistant housing in Nepal, but encounter a problem that is as epistemological as it is technical. Professional engineers possess forms of knowledge essential to structural safety, but households possess different kinds of knowledge concerning materials, construction practices, domestic organisation and the cultural meaning of dwelling. The apparently simple question of how to reconstruct a safe house therefore opens onto more fundamental questions about what constitutes a house, what constitutes safety and whose knowledge is authorised to define either. Reconstruction cannot be reduced to the transmission of expertise from professional to beneficiary; it requires interaction between forms of knowledge that are differently situated.

Urushima & Yamaguchi approach knowledge from almost the opposite direction. Their historical analysis of the reconstruction of the Japanese city of Fukui demonstrates the importance of accumulated institutional knowledge: plans, standards, legislation, administrative procedures, professional experience and memories of earlier catastrophes. Reconstruction does not begin intellectually from zero after each disaster. Existing categories and instruments are mobilised and adapted to new circumstances. Yet this transfer is neither automatic nor neutral. Planning instruments inherited from earlier periods are selectively interpreted through changing political priorities and institutional relationships. In Fukui, continuity in planning technique did not necessarily imply equity in institutional response.

Taken together, both Rankin et al. and Urushima & Yamaguchi reveal an important tension. Reconstruction depends upon expertise and institutional memory, but successful reconstruction also requires the capacity to question the assumptions contained within that expertise. The problem is therefore neither simply insufficient knowledge nor excessive professional authority. Once again, it is the organisation of relationships between different kinds of knowledge.

A comparable argument emerges around participation, community and social capital. These concepts have often been presented as antidotes to technocratic reconstruction. Yet the contributions to this collection resist the assumption that local participation is inherently progressive. Ubesingha et al. show that social capital can provide trust, solidarity and informal accountability, but can also reinforce exclusion, corruption, dependency and gender inequalities. Particularly revealing is the comparative breadth of their cases, which range from high-income Italy to low-income Mozambique and several Asian economies with very different institutional structures. Reconstruction systems vary from highly formal state-led arrangements to donor-dependent and informal ones. Nevertheless, recurring ethical problems appear across these different contexts.

Gasseloğlu & Gonçalves make a related argument through spatial justice. Participation is only one component of justice, alongside distribution and recognition. A reconstruction programme may therefore be participatory while remaining distributively unequal, or distribute resources widely while failing to recognise particular groups and forms of experience. Their attempt to bring these dimensions together responds particularly closely to the special issue’s ambition to examine whether distinct conceptual domains can be made to inform one another within a systemic view of reconstruction.

Paidakaki’s longitudinal analysis of New Orleans after Hurricane Katrina shows these tensions unfolding politically over two decades. Reconstruction becomes a contested field involving government, markets, housing organisations, advocacy coalitions and communities operating at different institutional scales. The ability of community organisations to influence recovery depends not simply upon their local legitimacy but upon their capacity to establish horizontal coalitions and vertical relationships with institutions. The local and institutional therefore cannot be treated as alternatives. Reconstruction takes place through their actual and effective interaction.

One consequence of these papers is to make the search for ‘best practice’ problematic. While the call for papers explicitly asked whether the abundance of specialised knowledge permits the pursuit of best disaster-response practices overall, the contributions displayed considerable caution. They do not imply that lessons cannot be transferred, but they repeatedly show that the effects of a practice depend upon the institutional, social and material system in which it is inserted.

Bolchover & Mundle demonstrate this clearly in Kathmandu. Their alternative infill-housing prototype is not simply evaluated as an architectural object. Its possible adoption depends upon regulation, finance, construction skills, household preferences and established technological conventions. A technically plausible and economically competitive building solution cannot transform reconstruction independently of the system through which buildings are commissioned, approved, financed and produced. The problem of architectural innovation consequently becomes a problem of institutional and socio-technical change.

Akdede et al. make a remarkably similar point through temporary housing. The question of what happens to temporary accommodation after its initial use might appear primarily a matter of design for reuse. Their analysis instead reveals the importance of land ownership, tenure, institutional policy, maintenance, user adaptation and the physical condition of settlements. What begins as a circularity problem becomes a governance and property problem. In both papers, therefore, the physical object leads outward into the larger reconstruction system.

This pattern strongly supports the original proposition that reconstruction knowledge becomes misleading when divided too securely between disciplinary jurisdictions. Architecture cannot be isolated from finance and regulation; engineering cannot be separated from cultural understandings of dwelling; circular design cannot be separated from land tenure; community participation cannot be separated from distributions of power.

Jiang’s analysis of post-war Rome extends this argument into heritage. The reconstruction of damaged monuments was simultaneously a technical conservation problem, an exercise in international cultural cooperation and a component in the redefinition of Italian national identity after Fascism. Heritage reconstruction consequently operated at several scales and through several systems of meaning at once. Restoring an architectural object could also participate in rebuilding political legitimacy and redefining relationships between national and international institutions. Here particularly clearly, reconstruction of the past becomes part of the construction of a future.

The commentaries reinforce these conclusions while introducing perspectives grounded more directly in practice. Potangaroa et al. draw attention to material supply chains and practitioner knowledge, reminding readers that reconstruction is not only a problem of governance, justice or epistemology. It is also an act of production. Materials, machinery, and labour must be mobilised and transported through environments whose productive and logistical systems may themselves have been damaged. Their intervention exposes another consequence of disciplinary fragmentation: a reconstruction strategy may appear convincing from the perspectives of policy or design while being materially impossible to implement. The productive capacity of the territory is therefore not incidental to reconstruction but one of its fundamental conditions.

Binetti makes a related argument at the infrastructural scale. Energy reconstruction cannot sensibly be treated as an autonomous sector because energy enables virtually every other component of recovery: healthcare, water supply, communications, housing reconstruction, economic activity and the return of displaced populations. Conversely, rebuilding energy systems depends upon transport infrastructure, finance, security, technical expertise and institutional capacity. Moreover, allocation decisions may be politically contested and can reinforce rather than resolve divisions created by conflict. Energy thus provides an unusually clear example of reconstruction as a system of interdependencies rather than a collection of parallel sectoral projects.

Haddad & Sakr extend the question of interdependence to governance, particularly in contexts where the institutions expected to coordinate reconstruction are themselves fragile. Their commentary shows that reconstruction cannot be treated simply as the implementation of projects within a pre-existing administrative framework: the capacity to establish priorities, coordinate actors, allocate resources and maintain accountability depends upon institutional relationships that may have been weakened by conflict or disaster and may already have been deficient beforehand. Reconstruction consequently becomes inseparable from the reconstruction of governance itself. Yet this also introduces a potential circularity. Effective recovery requires institutional capacity, while the recovery process may simultaneously be expected to restore or strengthen that capacity. Coordination, accountability and trust therefore emerge not merely as conditions for reconstruction but as possible outcomes of it, placing institutional reconstruction alongside physical and social recovery as part of the same process.

The contributions also destabilise the temporal boundaries of reconstruction. MacAskill & Dalziel’s emphasis on preparedness and Comerio’s call to ‘Build Better Before’ both suggest that the capacity to reconstruct is substantially produced before destruction occurs. Building regulations, institutional relationships, financing arrangements, professional capacity and infrastructure investment determine the options available afterwards. The conventional division between mitigation, preparedness, emergency response and reconstruction may therefore itself constitute a form of knowledge fragmentation. The call for papers asked how short-, medium- and long-term priorities can be coordinated and whether mitigation policies should serve as gauges for rebuilding. These commentaries push the question further: reconstruction may have to be understood as part of a continuum of territorial governance extending both before and long after the destructive event.

Al-Harithy’s reading of Lebanon similarly challenges a neat beginning and end to reconstruction. Repeated cycles of destruction and rebuilding demonstrate that the replacement of physical assets cannot necessarily be equated with urban recovery. Reconstruction can occur while social vulnerability, political fragmentation and displacement persist. Post-conflict rebuilding may consequently reproduce conditions that contribute to subsequent crises. The object of reconstruction, in this interpretation, cannot simply be the building or even the city as physical fabric: it must include the capacity of urban communities to recover and sustain social life.

This is where the editorial decision not to separate a priori post-disaster and post-conflict reconstruction appears reasonable. The contributions do not suggest that the two conditions are equivalent. Warfare involves intentional destruction, political settlements, disputed sovereignty and continuing security threats in ways that differ substantially from most hazard-related disasters. Yet once attention moves from the cause of destruction towards the organisation of reconstruction, significant commonalities emerge. Both require decisions concerning resources, land, housing, infrastructure, displacement, institutional authority, expertise, participation, time and future risk. The value of bringing them together lies, therefore, not in collapsing their differences, but in asking which reconstruction problems are recurrent and which remain determined by the specificity of the destructive event. This is precisely the comparative problem anticipated in a passage of the call: how to capture ‘the essential whilst isolating the peculiar’.

On this point, however, the collection also exposes the limits of the original ambition. It confirms the fragmentation of reconstruction knowledge more convincingly than it demonstrates how that fragmentation can be overcome. Most contributions remain recognisably rooted in particular disciplinary traditions and empirical objects. There is no common dataset or agreed set of variables through which a Nepalese housing programme, post-war Japanese planning, New Orleans advocacy networks, Lebanese urban recovery and energy infrastructure after war can readily be compared. The call’s questions concerning comparative methodology and the possibility of constructing comparable reconstruction datasets therefore remain largely unanswered. Similarly, the potential contribution of computational methods and large-scale data analysis receives little attention.

These absences could themselves be informative, by indirectly suggesting the absence of (or the difficulty in defining) the overall conceptual infrastructure necessary to integrate the field’s accumulated empirical knowledge. But they also raise the possibility that integration should not mean the creation of a single unified reconstruction paradigm. Indeed, at least four broad orientations can be recognised across the collection. A critical-normative position, represented most clearly by Lizarralde et al., Gasseloğlu & Gonçalves and Paidakaki, which understands reconstruction through justice, power and the negotiation of transformation. An institutional position is visible in Urushima & Yamaguchi, Haddad & Sakr and MacAskill & Dalziel, which emphasises governance, coordination, preparedness and institutional capacity. A situated epistemological position, especially apparent in Rankin et al. and Ubesingha et al., underscores negotiation between professional, institutional and local knowledge. Finally, a socio-technical and productive position, explicit in Potangaroa et al. and implicit in the housing and infrastructure contributions, insists that reconstruction ultimately depends upon material resources, labour, technologies, firms and supply systems.

These positions are broadly complementary, but not seamlessly so. Their tensions are real. Speed can conflict with participation. Strong institutions can increase implementation capacity while constraining local agency. Professional standards can reduce physical vulnerability while displacing locally valued practices. Transformative reconstruction can correct earlier inequalities but can also destroy valued continuities. Market mechanisms can mobilise resources rapidly while exacerbating spatial inequality. Local social networks can strengthen recovery while reproducing exclusion.

The significance of the collection may lie precisely in exposing these seemingly difficult-if-not-impossible-to-resolve tensions. All the contributions seem to lead towards a natural thesis: that reconstruction theory should aim for the dialectical composition of its multiple intervention dimensions rather than their unification into a whole. The role of theory could then be the provision of a conceptual structure through which different forms of knowledge would become mutually intelligible: a structure capable of identifying recurrent variables such as time, scale, authority, expertise, resources, production, context and justice, while allowing their relationships and relative importance to vary between circumstances.

Seen in this light, the collection substantially confirms the premise of the call whilst reformulating its implied solution. Fragmentation is indeed a challenge; yet synthesis cannot mean dissolving disciplinary differences into overarching normative protocols, however inclusive, ecumenical and efficient these may seem. Reconstruction requires frameworks capable of connecting partial understandings without eliminating the specificity from which their in-depth explanatory value derives.

The strongest collective conclusion may therefore be a confirmation of what was suggested at the outset through the notion of complexity: reconstruction is not one problem imperfectly divided among different disciplines: it is a phenomenon constituted by the interaction of multiple problems. Physical rebuilding, social recovery, institutional reconstruction, economic production, cultural continuity and political transformation are not parallel processes that can simply be added together after separate analyses. They continually condition one another.

This refines the kernel of the theoretical challenge, which is to understand how the partial forms of knowledge involved in reconstruction can be composed in ways that permit reciprocal awareness, conceptual coordination and, ultimately, coherent action, all the while retaining their differences. In this sense, the contributions do not merely confirm the complexity anticipated by the call. They make its original question more demanding: the task is not simply to organise what is known about reconstruction, but to organise it in ways capable of revealing and interrogating the multiplicity of relationships through which reconstruction actually operates.

4. TECHNIQUES AND TECHNOLOGY

One way of clarifying the mechanics of such a process could be that to draw attention to the distinction between technique and technology. Technique can be understood as the translation of knowledge into organised practical action. As such, it necessarily requires preparation: the systematic mobilisation of theoretical knowledge, empirical evidence and accumulated experience to provide a basis for problem appraisal and decision-making. This requirement is evident across virtually every domain of reconstruction. Indeed, each contribution to this collection contains, explicitly or implicitly, a technical argument: a critical account of how something was done, how it might have been done differently or how it should be done in future.

Yet the techniques of reconstruction are not the same thing as a technology of reconstruction. Here the term ‘technology’ can usefully be recovered from an older sense, closer to the systematic treatment of technē through the logos after it: not technology as an assemblage of devices, but as the systematic organisation of knowledge concerning techniques and practical action (Schatzberg 2018). The distinction is useful precisely because it shifts attention from individual techniques to the intellectual means through which techniques can be understood, compared, developed and related to one another. A technology of reconstruction, in this sense, would not consist of a universal set of procedures. It would consist of a systematic discourse capable of placing the different forms of knowledge informing reconstruction into relation.

This offers another way of representing both the challenge identified by the collection and a possible strategy for addressing it. The objective would not be to produce technocratic formulas capable of determining reconstruction independently of context, irrespective of the ideological positions supporting them. Instead, it would be to develop a genuinely discursive technology of reconstruction: an open structure through which the knowledge embedded in different fields of reconstruction practice could be examined comparatively, brought into relation and combined where useful.

Theory has practical implications that also need consideration. How could a technical discourse of such magnitude and breadth be constructed and maintained; by whom; and on the basis of what evidence?

The first difficulty concerns the collation of empirical knowledge. If reconstruction is a situated practice, the relationships between the domains through which it operates cannot simply be deduced in the abstract. They must also be reconstructed empirically from particular cases. Theory can identify relationships worth looking for, but practice provides the evidence through which their forms, strengths and consequences become visible. The movement is therefore necessarily reciprocal: theory provides categories and questions through which cases can be interrogated, while cases test, qualify and reorganise the theoretical relationships proposed.

The scale of this exercise is formidable. Reconstruction has occurred in innumerable historical, geographical, political and economic circumstances, and new instances continue—unfortunately—to be added. Their records are dispersed across scholarly publications, government documents, project reports, evaluations, technical specifications, financial records, legislation, institutional archives, and the knowledge retained by communities and practitioners. Much of this material is incomplete; some has never been systematically recorded; some remains inaccessible or protected. Even where information exists, assembling it across cases and domains in a form permitting comparison represents a task the scale of which may exceed the capacities of conventional scholarly methods.

A second difficulty concerns not the existence of information but its release. Reconstruction knowledge is not produced within a neutral economy. Access to data permits their elaboration; elaboration contributes to expertise; and expertise has value within professional and consulting markets. Knowledge accumulated through projects may consequently remain proprietary, institutionally bounded or professionally guarded. Transforming such information into publicly accessible knowledge—a public good—can thus impose costs on actors whose competitive position partly derives from privileged access to experience and information. The fragmentation of reconstruction knowledge is, from this perspective, not simply an epistemological consequence of disciplinary specialisation. It may also be sustained by the institutional and economic conditions under which expertise is produced and exchanged.

Transforming such information into openly accessible knowledge consequently raises a familiar public-good problem. Knowledge has long been recognised as possessing important public-good characteristics, while the institutional and commercial conditions of its production can create incentives for restricting access to it (Stiglitz 1999). In reconstruction, such incentives may be particularly significant where expertise and competitive advantage derive partly from privileged access to accumulated project experience.

These two problems point in different directions but lead to the same impasse. An enormous and continuously expanding body of potentially relevant evidence is too dispersed and heterogeneous to be systematically assembled through conventional scholarly labour alone. At the same time, an information economy restricts the circulation of at least part of that evidence and therefore limits the possibility of turning accumulated experience into cumulative public knowledge. A technology of reconstruction would have to confront both conditions.

It is here that the question raised in the call concerning the possible role of computation resurfaces in a different form. The issue is not whether computational techniques—and artificial intelligence (AI) in particular—might improve individual reconstruction tasks. The more consequential question for the argument developed here is whether they might assist in constructing the knowledge infrastructure of reconstruction itself: if computational methods could help locate, collate, classify, and relate the immense and heterogeneous empirical record of reconstruction in ways that conventional scholarship cannot readily achieve.

This would imply a rather different role for AI from that recently associated with its application to disasters.¹² Rather than asking primarily whether AI can predict damage, optimise resource allocation, classify buildings or accelerate particular technical operations, the question becomes if it can help in the understanding of what has collectively been learned from reconstruction. Its potential contribution would lie not in replacing theory, expertise or situated judgement, but in extending their empirical reach: making it possible to interrogate a body of experience, the scale, dispersion and heterogeneity of which have hitherto made systematic comparison extremely difficult.

Such an undertaking would, however, return immediately to the theoretical problem with which this discussion began. Data cannot by themselves constitute meaningful knowledge. They require a conceptual structure through which their significance and relationships can be established (Rowley 2007; Floridi 2012). Before reconstruction experiences can be compared computationally, decisions must be made about what defines a reconstruction case, what entities and processes should be recorded, what relationships matter, what constitutes an outcome, and how differences of time, scale, authority, resources, production, context and justice should be represented. AI may therefore provide a means of organising empirical knowledge at a scale that conventional scholarship cannot easily manage, but it cannot supply the conceptual structure through which that knowledge acquires meaning. That remains the work of theory.

The proposition is consequently circular in a productive sense. Theory identifies the relationships through which reconstruction might be understood; computational methods make it possible to search for and organise evidence concerning those relationships across an expanding universe of cases; empirical comparison tests and complicates the theoretical structure; and the revised structure, in turn, changes the questions asked of the evidence. The possible contribution of AI would thus reside not in producing a unified theory of reconstruction, but in making the dialectical composition of its fragmented knowledge empirically manageable.

At this point the agenda of this introduction is substantially complete. The reasons for the special issue have been laid out, the place of its contributions within this plan has been outlined, and a preliminary exploration of the possibilities for their future theoretical correlation has been started. This does not mean, however, that the work is done. If the idea of a global reconstruction knowledge base retains any validity, the practical conditions of its organisation must eventually be addressed: who maps the knowledge involved, who collects it, who maintains it, who administers access to it and under what institutional authority.

5. TOWARDS AN ONTOLOGY OF RECONSTRUCTION

The argument developed here suggests that the first object of such an effort should not, in fact, be AI itself. Before computational systems can organise reconstruction knowledge, some agreement is required about what there is to organise. A more immediate task would therefore be the development of an open ontology of reconstruction: a provisional conceptual specification of the entities, processes and relationships through which reconstruction cases could be described and compared. Its purpose would not be to establish another normative reconstruction model, nor to determine in advance which forms of practice are correct. Rather, it would provide a common relational structure through which heterogeneous evidence could be located and connected while preserving the contextual and disciplinary differences on which its meaning depends.¹³

Developing such an ontology would itself require collective action. No single discipline, professional group or institution could reasonably claim authority over a phenomenon the defining characteristic of which would be precisely the interaction of different forms of knowledge. Nor would it seem desirable for its conceptual infrastructure to be controlled by commercial providers of computational technology. The task would more appropriately belong to an international, multidisciplinary and publicly accountable initiative involving researchers, practitioners, governments, international agencies and representatives of communities with direct experience of reconstruction. Existing institutions concerned with recovery could provide the organisational setting rather than requiring the creation of another body.

There is already a basis for such an initiative. The International Recovery Platform (IRP), for example, exists specifically as a global partnership for strengthening knowledge and exchanging experience concerning recovery, rehabilitation and reconstruction, while the World Bank’s GFDRR has developed common methodologies for recovery frameworks and accumulated substantial bodies of recovery documentation (IRP 2026; GFDRR 2024). The existence of such structures demonstrates that the international organisation of reconstruction knowledge is neither institutionally unprecedented nor practically inconceivable. What is missing is a mechanism designed specifically to convert dispersed and discrete reconstruction experience into a cumulative and relational body of knowledge.

A feasible first action could consequently be relatively modest: the establishment of an international working group charged with developing and testing a reconstruction ontology against a deliberately heterogeneous sample of well-documented cases. Rather than beginning by attempting to capture the universe of reconstruction experience, the exercise could establish whether cases differing in cause, location, scale, political economy and institutional setting can be represented through a common relational framework without losing the characteristics that make them distinct. The resulting structure could then provide the basis for an open, progressively expanding corpus in which reconstruction episodes, rather than documents, become the principal units of organisation.

Only at this stage would AI assume the role envisaged here. Computational systems could assist in locating relevant material, extracting information from heterogeneous records, connecting evidence to the categories and relationships established by the ontology, identifying similarities and differences across cases, and revealing absences or contradictions requiring human interpretation. The resulting infrastructure would need to preserve the provenance of the information it contains and permit its conceptual structure to remain open to revision as new cases, perspectives and theoretical questions emerge.

Responsibility for such a system would consequently need to be distributed, but its knowledge infrastructure should arguably remain a public good. International organisations could provide continuity, convening power and access to existing recovery records; national and local governments could contribute official documentation; universities and research networks could develop and critically revise its conceptual structure; practitioners and communities could contribute forms of knowledge rarely captured in institutional records; and private organisations could provide both data and computational expertise. What should remain publicly governed are the ontology, the standards through which evidence is described and related, its provenance requirements, and, as far as legitimate restrictions permit, access to the resulting knowledge base.¹⁴

The immediate policy implication is therefore perhaps less ambitious than the technological possibilities might suggest. Rather than calling for the development of an AI for reconstruction, the first requirement is to create the intellectual and institutional conditions under which computational technology could usefully assist reconstruction knowledge. Developing and testing an open ontology would be a concrete place to begin. If successful, it could provide the missing interface between the accumulated experience of reconstruction, the theoretical structures through which that experience is interpreted, and the technical preparation through which knowledge is returned to practice. In the process, it may help to contest collectively and intellectually the banality, let alone the vulgarity, of the aggressive destruction and reconstruction proposals tragically populating the present time and age.

AI DECLARATION

No artificial intelligence (AI) was used in the researching or writing of this editorial.