Built environment as interface: a relation-based framework for the intersections between built, biotic, social, and health processes during COVID-19 and beyond

The paper addresses how current COVID-19 measurement, modeling, and data practices inadequately capture the complex intersections among built, biotic, social, and health processes. It poses the overarching question: to what extent do existing practices fail to represent these intersections, and how can conceptual foundations be reworked to embed the causal relevance of built environments? The authors present a two-part agenda: (1) a diagnostic analysis of representational failures—missing data on race/ethnicity and social determinants, lack of frameworks to include built environment moderators, and overreliance on shallow surrogate measures; and (2) a prescriptive program advocating relation-focused concepts (context, nudge, affordance, interface) to better model and visualize how built environments regulate multi-process pathways. The introduction situates the work within interdisciplinary efforts, highlights counterintuitive causal relations (e.g., density vs. connectivity; social beliefs shaping transmission), and argues for parameterizing models by context while avoiding reduction of the built environment to single variables.

The paper reviews multiple literatures spanning public health, urban planning, environmental sociology, philosophy of science, systems biology, and behavioral science. It synthesizes: (a) integrative health concepts (One Health, EcoHealth, ecosystem and planetary health) and critiques their anthropocentric and Eurocentric tendencies and lack of fine-grained socio-ecological process detail; (b) extensive research linking built environments with ecological factors, social structures, and health outcomes (e.g., Hamlin & Sheard; Frumkin; Cooper; Gruebner; Pinter-Wollman); (c) evidence on air pollution and COVID-19 outcomes, with gaps in accounting for race/ethnicity, lifetime exposure, and local built variations; (d) debates on density vs. connectivity in explaining transmission and mortality; (e) behavioral ‘nudge’ literature and its limitations due to context-dependent social norms, culture, reactance, and individualism; (f) perspectivism in philosophy of science to frame selective scientific representations; and (g) calls for transdisciplinary, participatory approaches and ‘healthy publics’ that foreground relations and community engagement.

This is a conceptual and theoretical analysis that develops a relation-based representational framework for embedding built environment causal roles in COVID-19 science. Methods include: (1) diagnostic critique of scientific practices (epistemological, ontological, methodological) exposing three interlinked problems: measurement outcome gaps, representational framework absence, and surrogate measure overuse; (2) comparative robustness and perspectival analyses to illustrate how different representations capture partial system aspects; (3) conceptual development and synthesis of ‘context’, ‘nudge’, ‘affordance’, and ‘interface’ to characterize built environment moderating/mediating functions; (4) use of diagrammatic and visual heuristics (wiring diagrams, Waddington-style landscapes, hypersurfaces) to model multi-variable interactions and built environment as a dynamic interface; (5) integration of case-based evidence and small-data studies (e.g., airflow transmission tracing, sanitizer placement effects, density/connectivity analyses) to exemplify contextual moderation; and (6) proposing methodological pathways for GIS–EHR integration and community-participatory measurement (‘storying processes’) to co-generate variables and improve data granularity. No primary empirical datasets are analyzed; instead, the paper assembles cross-disciplinary evidence and offers representational tools and methodological recommendations.

- Identified three intertwined problems in COVID-19 scientific practice: 1) Measurement outcome problem: Missing or inadequate data on race, ethnicity, and social determinants; failure to capture built environment moderators in EHRs and surveillance. 2) Representational framework problem: Lack of conceptual infrastructure to embed causally relevant built environment processes operating at multiple scales and intersections with social/biotic factors. 3) Surrogate measure problem: Overreliance on shallow proxies obscures upstream multifactorial causes and interactions, leading to misrepresentation. - Demonstrated context sensitivity and counterintuitive relations: high-density settings can have low transmission when mediated by social beliefs/compliance; connectivity may outperform density as a predictor; cultural and belief systems modulate behaviors and transmission. - Showed limitations of ‘nudge’ frameworks: nudge effectiveness is mediated by social norms, culture, self-construal, reactance, and inequities; nudges can backfire or exclude groups (e.g., mask-related discrimination against Asian Americans). - Advanced ‘affordance’ as a superior relational concept: built contexts create differential capacities for action and health, including ‘built environment burden’ (e.g., disproportionate pollution exposure) and inhibition of beneficial biotic access. - Proposed the built environment as an ‘interface’ that reorganizes multi-process causal landscapes, deregulating some factors and leaving others unaffected; introduced 3D landscape and hypersurface visualizations to represent multidimensional interactions and emergent properties. - Highlighted methodological avenues: integrate GIS-derived built environment variables into EHRs; begin with complex landscape representations before reduction; develop community-participatory measurement platforms to surface hidden variables and inequities (e.g., Shingle Mountain case). - Selected data points cited to illustrate arguments: • A 1 µg/m³ increase in chronic PM2.5 exposure associated with a 9% (95% CI: 6–12%) increase in COVID-19 mortality (Coker et al., 2020). • In-line-of-vision hand sanitizer placement increased physician use to 53.8% vs. 11.5% at doorway-adjacent placement (p=0.0011) (Birnbach et al., 2010). • Airflow case evidence indicates droplets can travel ~21 feet in ~5 minutes under certain indoor conditions (Kwon et al., 2020).

The relation-based framework addresses the central research question by reframing built environments from single variables to dynamic interfaces that co-modulate biotic, social, and health processes. By diagnosing measurement, representational, and surrogate problems, the paper shows why standard models miss emergent properties and inequities (e.g., accumulated exposure, environmental racism). Introducing affordances and interfaces clarifies how built contexts differentially promote or inhibit health capacities and behaviors, offering a path to more accurate, context-aware models (e.g., parameterizing optimal contact rates by cultural risk perceptions and connectivity). Visual landscape metaphors operationalize complex multidimensional relations, guiding model reconstruction rather than mere parameter filling. The proposed community-participatory measurement extends epistemic scope to include locally salient variables, improving validity and equity. Overall, these developments enhance explanatory power, align interventions with real contextual dynamics, and make public health practice more responsive and just.

The paper contributes a relation-centered representational framework that reconceives the built environment as an interface regulating multi-process pathways among built, biotic, social, and health systems. It specifies three core problems (measurement outcomes, representational frameworks, surrogate measures), critiques over-simplistic reliance on nudges, and advances ‘affordance’ and 3D landscape/hypersurface visualizations to capture emergent, context-dependent causality. Practically, it recommends integrating GIS-based built environment variables into EHRs, reconstructing models to reflect contextual dynamics, and enabling community-participatory measurement to uncover unacknowledged variables and inequities. Future research should (i) empirically operationalize interface/affordance variables across contexts, (ii) test and refine multidimensional landscape models, (iii) develop socio-technical platforms for community data generation and storying processes, and (iv) build transdisciplinary collaborations that center affected communities in representing complexity.

- Conceptual and illustrative scope: The framework is primarily theoretical and relies on synthesized evidence and exemplars rather than new empirical datasets; proposed visualizations (landscapes/interfaces) need empirical operationalization and validation. - Data availability and integration: Implementing GIS–EHR integration and capturing lifetime exposure or fine-grained built context variables may face data gaps, standardization challenges, and privacy concerns. - Context dependence: Affordances and interface effects are inherently context-specific; transferring parameters across settings risks misrepresentation without local calibration and community input. - Practical and technological constraints: Community-participatory measurement requires equitable socio-technical infrastructures, governance, and resources that may be unevenly available. - Interdisciplinary coordination: Achieving the advocated transdisciplinary synthesis and participatory practices can be hindered by institutional silos, funding structures, and methodological fragmentation.