10th Biennial ACSPRI Social Science Methodology Conference in conjunction with the Australian Social Network Analysis Conference (ASNAC)

Interdisciplinarity is commonly understood as the integration of knowledge from two or more distinct academic disciplines in scientific collaboration. Rather than treating interdisciplinarity solely as an outcome of knowledge integration, this study conceptualises it as the product of ongoing boundary negotiations among researchers from different disciplinary traditions (Gieryn, 1983). Existing studies predominantly assess interdisciplinarity at the level of publications, authors, journals, research fields, and formal organisational arrangements such as research centres, laboratories and grant scheme (Wang & Schneider, 2020). Less attention has been paid to cohesive collaborative communities formed through researchers’ informal collaborative ties, for example professional networks, conference interactions, and long-term interpersonal relationships, that may transcend organisational boundaries (Calvert, 2024). Examining such communities provides an opportunity to investigate how disciplinary boundaries are maintained, negotiated, or crossed within collaborative research.
Furthermore, existing interdisciplinary measures often implicitly assume that diversity scores are comparable across collaborations of different sizes (Porter & Rafols, 2009). However, group size constrains the maximum variety could be achieved, while group size and variety jointly constrain the range of possible diversity values. As a result, observed diversity should not be interpreted independently of the structural opportunities available to a collaboration. While existing measures quantify disciplinary diversity, they do not indicate whether a collaboration exhibits lower, expected, or higher diversity relative to its structural conditions. Consequently, they provide limited insight into whether diversity reflects structural characteristics of a collaboration, or active processes of disciplinary boundary maintenance and boundary crossing.
To address these gaps, we propose a framework that uses maximal clique as the unit of collaboration and residualised Shannon entropy as a measure of disciplinary boundary permeability. By modelling expected disciplinary diversity as a function of clique size and disciplinary variety, the framework evaluates interdisciplinarity relative to the structural opportunities available to each collaboration. This approach enables the identification of collaborative communities that exhibit higher or lower levels of boundary work than would be expected from their size and disciplinary composition. We subsequently classify maximal cliques into three types based on residualised Shannon entropy and interpret them through the lens of boundary work. Boundary-maintaining cliques have negative residual values, indicating lower observed diversity than expected.; boundary-neutral cliques have residual values close to zero, indicating diversity levels consistent with structural expectations; and boundary-crossing cliques have positive residual values, indicating higher observed diversity than expected.
To demonstrate the proposed framework, we constructed a coauthor network of 6,730 authors
from1,909 smart cities research articles published in 2022 indexed in Scopus. Maximal clique analysis identified 1,356 collaboration communities with at least 3 members. The overall residual diversity scores were negative (mean = -0.24, median = -0.49), with 58% of cliques were classified as boundary-maintaining (mean = -0.74), 24 % as boundary-neutral (mean = 0), and 18% as boundary-crossing (mean = 0.11). These findings suggest that while smart cities research is commonly characterised as interdisciplinary (Bibri, 2019), collaborations in this field largely maintained disciplinary boundaries with only a minority actively crossed disciplinary boundaries beyond what would be expected from their structural characteristics.
The proposed framework enables identification of informal collaborative communities that are observable only through co-author networks. It also disentangles structural opportunities from boundary work by distinguishing expected diversity from observed diversity. By doing so, it shifts attention from interdisciplinarity as an aggregated outcome to the ongoing processes through which disciplinary boundaries are maintained, negotiated, and crossed within collaborative communities.

One of the challenges for mixed-method research is how to integrate findings from diverse methodological traditions. Network-based methods can identify patterns of associations among actors, issues, and concepts, but provide limited insight into how these associations are negotiated and sustained. Qualitative approaches, meanwhile, can reveal meanings, yet often difficult to connect systematically across large and complex relational settings. This paper proposes to addresses this methodological challenge through a framework that integrates semantic network analysis, cartography of controversies, and boundary theory.
The framework was developed through a doctoral study of smart city development in Australia. The first component employs semantic network analysis across multiple datasets, including documents, news coverages, and codebases of government-funded and open source smart city (OSSC) initiatives. Rather than treating semantic networks as an end point, they were used as exploratory map to identify both shared conceptual spaces and conceptually distinct areas associated with different actors. Across datasets, concepts such as ‘open data’ and ‘citizen-centric’ emerged as common reference points linking government agencies and OSSC initiatives. At the same time, the networks revealed divergent thematic clusters. Government-funded projects commonly associated with sustainability, energy efficiency, public safety, and urban service delivery, while OSSC initiatives tended to emphasis democracy, accountability, food security, sustainable agriculture, and community empowerment.
These patterns of convergence and divergence informed the second component of the framework: qualitative case studies using cartography of controversies. Following statements, actors, issues, documents, technologies, and institutional arrangements over time, the case studies traced how shared concepts were interpreted and enacted. While semantic networks suggested broad agreement around ideas such as openness and citizen-centric development, qualitative tracing revealed substantial differences in meaning. For example, government agencies frequently framed citizens as service users, consultees, or sources of urban data, whit participation occurring primarily through consultation processes and sensor-enabled urban management. In contrast, OSSC organisations positioned citizens as active participants in oversight, accountability, and knowledge production. Likewise, while both endorsed open data, they framed and practiced differently. Government agencies primarily associated open data with Open Government Data (OGD) and administrative procedures, whereas OSSC organisations linked both OGD and Freedom of Information to transparency, accountability, and the strengthening of community capacity.
To integrate these findings, the paper draws on the concepts of boundary work and boundary objects. boundary object explains how concepts such as open data and citizen-centric can operate as shared reference points across heterogenous communities while remaining sufficiently flexible to accommodate divergent interpretations. Boundary work provides a means of analysing how actors construct, defend, challenge, and reconfigure the meanings attached to these concepts. Together, they offer an analytical bridge between network-level patterns and qualitative accounts of negotiation. Rather than treating quantitative and qualitative findings as separate streams of evidence, the framework uses boundary theory to explain how apparent conceptual convergence can coexist with substantive differences in interpretation and practice.
This paper contributes to methodological discussions on mixed-methods approaches in three ways. First, it demonstrates how semantic network analysis can be used to identify potential boundary objects and guide qualitative inquiry. Second, it shows how cartography of controversies can reveal the processes through which shared concepts acquire different meanings across actor groups. Third, it illustrates how boundary work and boundary object can function as integrative devices for synthesising network patterns and interpretive forms of analysis. By linking networks to negotiations, the proposed framework offers a transferable approach for researchers seeking to combine network analysis and qualitative methods in the study of collaboration, governance, and socio-technical change.