Exploring Social Water Research: Quantitative Network Analysis as Assistance for Qualitative Social Research

Abstract

This paper presents a meta-analysis of social research on water, offering a novel methodological contribution to the study of emerging interdisciplinary research fields. We propose and implement a mixed methods framework that integrates quantitative network analysis with qualitative research, aiming to enhance both to give access to new emerging empirical fields and enhance the analytical depth of empirical social research. Drawing on a dataset of publications from the Web of Science over four distinct time intervals, we identify thematic clusters through keyword co-occurrence networks that reveal the evolving structure and internal dynamics of the field. Our findings show a clear trend toward increasing interdisciplinarity, responsiveness to global events, and contemporary challenges such as the emergence of COVID-19 and the continued centrality of topics related to water management and evaluation. By uncovering latent structures, our approach not only maps the field’s development but also lays the foundation for targeted qualitative analysis of articles representative of identified clusters. This methodological design contributes to the broader discourse on mixed methods research in the social sciences by demonstrating how computational tools can enhance the transparency and reliability of qualitative inquiry without sacrificing its interpretive richness. Furthermore, this study opens new avenues for critically reflecting on the epistemic culture of social water research, particularly in relation to its proximity to applied science and governance-oriented perspectives. The proposed method holds potential relevance for both academic researchers and decision makers in the water sector, offering a means to systematically access dispersed knowledge and identify underrepresented subfields. Overall, the study showcases the potential of mixed methods designs for navigating and structuring complex interdisciplinary research landscapes.

1. Introduction

In recent years, we have witnessed the emergence of new interdisciplinary research fields, particularly in response to pressing global issues such as climate change. Within the realm of social science, social water research has evolved into a vibrant interdisciplinary domain, drawing insights from diverse disciplines such as geography, economics, ethnology, and environmental science. In connection with this, there has been a significant increase in the volume of scientific articles within social sciences addressing water-related issues. This growth presents both opportunities and challenges for researchers. Traditional qualitative methods alone struggle to fully grasp the complexity of water-related research, especially as disciplinary boundaries become less pronounced and national discourses become more interconnected. As a result, there is a growing need for, as well as interest in, innovative methodological approaches that can accommodate the vast amount of data while continuing to allow for qualitative analysis. This growing methodological interest also reflects a broader ongoing debate within the social sciences regarding the fragmentation between qualitative and quantitative research traditions. While we do not assume that mixed methods are universally superior or always appropriate, we argue that, specifically, they offer valuable opportunities for harnessing the complementary strengths of both approaches [1]. In this context, mixed methods approaches offer a promising avenue to link quantitative data with qualitative insights, providing a more comprehensive understanding of water-related social research and their societal implications.

This paper provides a previously missing methodical and exhaustive overview of the research field on water in social sciences. While the previous literature reviews often rely on qualitative or quantitative methods, this study introduces a mixed methods approach that combines quantitative keyword co-occurrence network analysis with a sociological lens. This allows for the identification of thematic clusters and research trends in a structured and reproducible way. For this aim it proposes a quantitative network analysis as part of a mixed methods research design, acknowledging the complementary strengths of qualitative and quantitative research, and applies this method to the research field on water in social science. With the lens of sociology of knowledge, this paper goes beyond a purely qualitative literature review, and the quantitative network overview facilitates a more comprehensive understanding of the vast number of papers on the topic, reducing the time needed to grasp the existing works. The aim is to reliably identify thematic fields of research that can then be further analyzed. Therefore, this paper closes the gap between the existing literature reviews with a systematic approach but without methodological control [2,3,4] on the one hand and mixed method approaches to network analysis on the other. The authors are not aware of any study that systematically combines quantitative bibliometric analysis with a qualitative interpretation of the results, bridging the gap between quantitative and qualitative literature reviews to contribute to the study of emerging interdisciplinary research fields. By doing so, this paper contributes a replicable framework for analyzing complex academic fields and provides a foundation for further qualitative deep dives into specific clusters.

Our mixed methods approach begins with quantitative network analysis based on the water literature, utilizing keyword co-occurrences to construct a network that reflects the thematic evolution of social water research. As previous studies have shown [5], the identification of the keywords used in publications can provide important information on the discourse developed in the field and its main streams and topics. We identify distinct clusters within the network, each representing thematic fields that have emerged and evolved over time. Subsequently, a first qualitative interpretation of these clusters provides insights into their dynamics and substantive content. This interpretation of clusters could serve as a starting point for further qualitative analysis through qualitative content analysis of paradigmatical articles of a cluster or other qualitative methods.

The approach discovers latent structures in the data through quantitative inquiry. By doing so, it seeks to increase the reproducibility of subsequent qualitative data analysis without sacrificing the analytical depth provided by human hermeneutic processes [6]. The connection ensures reliability, intersubjective validity, and full reproducibility in the analysis. The proposed sequence of techniques, drawing from both computational methods and expert substantive knowledge, maintains transparency and efficiency throughout the content analysis process.

2. Materials and Methods

2.1. Methodology

These mentioned challenges faced by qualitative research cannot be easily resolved through quantitative approaches. The exploratory capacity of qualitative methods is intricately tied to their methodological foundations, which differ significantly from those of quantitative methods, even though “all research is interpretative” [7] (p. 210).

Within the diverse methodological paradigms of qualitative research [8,9], the “social constructivism” paradigm especially emphasizes inductive reasoning. This paradigm follows the perspective of researched actors [10], enabling the exploration of previously unknown empirical phenomena. In German-speaking sociology, this paradigm is termed “reconstructive” and is closely linked to foundational works in the sociology of knowledge by Karl Mannheim and Alfred Schütz [11]. Qualitative social research, as seen through their lens, is perceived as the reconstruction of practice-related knowledge held by researched actors [12,13]. Similar perspectives are found internationally [14]. The overarching goal is to reconstruct lived experiences as the meaningful structures that are innate to the actors’ minds, guiding their everyday actions, regardless of whether the data is derived from interviews, group discussions, text analysis, or ethnological observations [15,16,17].

To preserve the strengths of qualitative social research in the face of growing amounts of complex empirical data that traditional qualitative methods struggle to comprehend, this paper proposes a mixed methods approach. This approach acknowledges the necessity of combining qualitative and quantitative methodologies to gain a comprehensive understanding of multifaceted phenomena, ensuring the continued relevance and efficacy of qualitative research in contemporary social science.

Mixed method approaches have the potential to harness the productive differences between quantitative and qualitative research [18,19,20]. While all social science research, in a broad sense, involves a blend of methods [21], mixed method approaches are grounded in the explicit acknowledgment and examination of this blending process [22] (p. 118). However, the understanding of this mixture in the history of mixed methods research is contested, questioning the precise nature of this blending [22] (pp. 119–121), and whether mixed methods should be regarded as a “third major research approach or research paradigm” [22] (p. 112) is debatable.

This paper posits that mixed method approaches are not a distinct paradigm. Instead it adopts a “dialectical stance” [20] (p. 12), following Donna Mertens. The methods are not “merged” [23] but, instead, deploy qualitative and quantitative methods accordingly to their respective strengths and put the different paradigms—postpositivsm and constructivism [24] (p. 256)—in a productive dialogue. Accordingly, within this study, quantitative network analysis provides the field-discovering systematic foundation for the interpretive integration of the quantitative results in the subsequent step, following Kuckartz’s concept of “Qualifizierung”.

We follow Betina Hollstein’s [25,26] definition of mixed method studies within network analysis by using quantitative and qualitative data, as well as quantitative and qualitative strategies of data analysis, and integrate the results into a comprehensive understanding of social water research. The inference of quantitative and qualitative analysis allows us a holistic analysis. With the quantitative network analysis, we can systematically explain the structure of social water research and, with the qualitative step of interpretative integration, we understand this structure and its development over time.

As Hollstein explains, this approach fits network analysis. It aligns with the two perspectives on network analyses, understanding networks either as structures or relations [18] (p. 3), [27] (p. 1068), [28] (p. 111). Particularly following the exploratory network analysis [28] (p. 114), the nodes of the network in this paper are not treated as independent data points (as often seen in much quantitative research) [18] (p. 2) but are understood as “meaningful relations in contextualized settings” [18] (p. 3). Such meaningful relations can, therefore, be reconstructed with an interpretative lens.

The presented mixed methods approach of quantitative network analysis and subsequent data qualification [29] provides a systematic lens into the thematic evolution and structure within the research field of water in social science. The initial step involves computational methods for pattern detection, reducing complex text to interpretable word groups. In the second step the authors analyze the data through qualitative inquiry. In general, qualification describes the conversion of quantitative into qualitative data, like translating quantitative survey results into narratives or categories. In this special case, qualification means making sense of the quantitatively generated clusters.

2.2. Data Collection

The applied sampling strategy is inspired by systematic literature review methods [30]. Possible data sources are different publication databases online, such as Web of Science, Scopus, or Google. For our analysis we choose the Web of Science database, because it is the largest database of scholarly publications currently available [31]. For this paper, articles are extracted from the Web of Science using the search term “water” within the Social Science category across distinct time intervals (2007, 2012, 2017, and 2022), to make as few preliminary decisions as possible. For each year we extract 500 articles for our analysis. The year 2022 is the year prior to the data collection to ensure that a full year is displayed in the network. Only the search term “water” is given to obtain a broad variance of articles without focusing on a certain topic. The selection of articles is guided by their relevance to the research topic. By limiting the search results as little as possible to one topic, we obtain the broadest possible spectrum of articles dealing with the major topic of water in the social sciences.

2.3. Quantitative Network Analysis

Quantitative network analysis has increasingly gained popularity in social science research, being applied to various types of data such as keywords, citations, or co-authorships. Citation and co-authorship network analyses have been used, for instance, to examine the structure and evolution of thematic research fields, such as innovation systems research [32], or to assess journal performance in the field of water research [33]. More recently, Dennis and Grady [34] applied network science tools to explore thematic overlaps and disciplinary boundaries in water-related research, demonstrating the potential of bibliometric network analysis to map interdisciplinary discourses. Citation network analyses have also been used to examine the interconnectedness of the mixed methods research community [35]. In a similar vein, the design of this study employs a quantitative network analysis approach, utilizing keyword co-occurrences as the foundation for constructing a network. We choose keywords, because the keywords are assigned by the authors themselves and thus reflect their self-positioning and affiliation to disciplinary or thematic research communities. To ensure consistency across keywords, we performed several preprocessing steps. The keywords of articles serve as nodes in the resulting network.

Subsequently, key network terminologies are employed, such as “degree” denoting the number of links or connections associated with a node. The degree of a node serves as an intuitive measure of its centrality within the network. The greater a node’s degree, the more central it is in comparison with other nodes [36]. The more central a node in the network, the more a keyword and its underlying concept serve as a central reference point for connecting the thematic aspects of a cluster. After extracting plain text data from the Web of Science, this data is transformed into a dataset using bibliometrix [37]. This dataset is then used to compile a list of keywords while standardizing various spelling variants. For that, all keywords were converted to lowercase, and lexical variants differing by only one character were identified and merged (e.g., singular/plural forms such as model/models, or spelling variations such as labour/labor). To avoid semantically distinct terms being incorrectly merged (e.g., Bali vs. Mali), the resulting matches were qualitatively reviewed and manually validated. A final manual check of all unique keywords ensured conceptual consistency before constructing the network. From these keywords, a matrix is constructed, which is subsequently transformed into a network object. To transform the data into a network object, we used the R packages igraph [38] and tidygraph [39]. Each unique keyword was represented as a node, and edges between nodes reflected co-occurrence. Based on this data, we constructed undirected unweighted co-occurrence networks. For each article, all co-occurring keywords were connected pairwise, resulting in a network in which each edge indicates that two keywords appeared together in at least one article. Multiple co-occurrences across articles were not aggregated into edge weights.

The findings are subject to descriptive analysis and the generation of plots. Each time interval is treated as an independent network. In the next step, the Louvain algorithm is applied to cluster nodes, aiming to identify thematic fields. We define clusters as groups of keywords that are more densely connected to each other than to nodes outside the group. The Louvain algorithm is chosen for community detection because of its computational efficiency and robustness in optimizing modularity. It effectively identifies densely connected groups of nodes (thematic fields in our context) without requiring prior knowledge of the number or size of communities. This makes it well-suited for analyzing large networks where the community structure is unknown. Furthermore, the identification of the largest components within the networks reveals the central nodes within each network.

2.4. Qualitative Interpretation of Clusters

After visualizing keyword clusters, the mixed methods design incorporates a qualitative interpretation. This involves considering the temporal dimension—analyzing how thematic clusters evolve over different time intervals. Additionally, the substantive dimension is explored, interpreting the content of individual clusters and discerning their distinction from one another. As part of this qualitative interpretation of the network, clusters are labeled by assigning them a heading based on the central keywords, followed by a concise summary of the thematic connection among the keywords. This qualification enriches data through contextualization, interpretation, and understanding in relation to research questions. It is important to emphasize that this qualification is not a substitute for methodically guided qualitative analysis but rather an initial exploratory approach that serves to make the quantitative data accessible and to navigate within it.

3. Results

3.1. Descriptive

To gain insight into the network structure, we initially visualize the entire networks (Figure A1, Figure A2, Figure A3 and Figure A4 in the Appendix A). The positioning of nodes is determined using the Fruchterman–Reingold algorithm, which employs a force-directed approach to place nodes so that they are evenly distributed and avoid overlap, facilitating a clear representation of the network topology. Its widespread use and intuitive visual output make it a standard choice for exploratory network analysis.

Each network was constructed based on the 500 articles for the years 2007, 2012, 2017, and 2022, reflecting a gradual increase in size. Because the size of articles stayed the same, the increase in node size means that keywords became more diverse, or more keywords were used per article. For a better overview, the density and edge count (=number of connections) are presented in

Table 1. Size, density, and edge count of the keyword co-occurrence networks in five-year intervals from 2007 to 2022.

	2007	2012	2017	2022
Size	612	711	709	750
Density	0.002139426	0.001873972	0.001948316	0.001591455
Edge Count	400	473	489	447

. The density of a network is a measure of how many actual connections (edges) exist between nodes relative to the maximum possible number of connections. It ranges from 0 to 1, where 0 indicates no connections and 1 represents a fully connected network in which every node is connected to every other node. In the context of our keyword co-occurrence networks, a higher density indicates that keywords tend to co-occur more frequently with many others, reflecting a more interconnected thematic structure.

Further analysis focuses on delineating various thematic fields within the network and understanding the relationships between keywords (Figure A5, Figure A6, Figure A7 and Figure A8 in Appendix A). To gain a comprehensive overview of the network, we examine their quantity and size, revealing insights into their distribution. In 2007, 2012, and 2017, the network largely comprised one large and otherwise largely small disconnected subgraphs, of which, in 2007, only 4 subgraphs consisted of 5 or more nodes. In 2012 and 2017, there were 8 subgraphs all larger than 5 nodes. Surprisingly, in 2022, 12 subgraphs > 5 were found in the network, whereby the main clusters were smaller on average than in 2007, 2012, or 2017. The largest subgraph comprised keywords primarily related to water management and water governance in 2007, 2012, and 2017, whereas the central theme in 2022 was COVID-19.

3.2. Identification of Thematic Clusters

Subsequently, we apply the Louvain algorithm to cluster each network. The Louvain algorithm, an efficient method for community detection, partitions the network into cohesive groups by maximizing modularity, which measures the strength of division of a network into modules or communities. Our aim here was to identify distinct structures within the network and observe how different clusters are integrated in the general keyword network. Based on this, we take a look at the five largest clusters separately (

Table A1. Keywords in the five largest clusters of each network and qualification.

2007	Keywords	Qualification
Cluster 1	Water Economics, Water, Water Shortage, Markets For Temporary Water Transfers, Globalisation, Irrigation, Data Envelopment Analysis (Dea), Pavement, Drought Risk, Urban, Residential Location, Water Framework Directive, Hydrogen Economy, Informal Water Development, Access, Water Right, Public-Private Partnerships, Public Participation, Participation, Environmental Impact, Regression Analysis, Crops, User Organizations, Central Asia, Adoption, Understandings, Potable Water, Institutional Design, Stakeholder Organisations, Foreign Investment Dependence, Social Capital, Common Agricultural Policy	Water Supply and Management: Focuses on various aspects of water supply and management, including water demand, mathematical modeling, and information, as well as economic and institutional dynamics within the water sector.
Cluster 2	Water Supply, Management, Water Demand, Hydraulic Structures, Climatic Change, Mathematical Modelling, Developing Countries, Water Conservation, Block Tariffs, Information Technology, Wsm- Dss Tool, Optimization Models, Corrosion, Water Resource, Customer, Privatization, Planning & Scheduling, Price Elasticity, Risk Assessment, Measurement, Sustainable Utilization, Shadow Price, Uniformity	Water Economics and Globalization: Emphasizes water economics, markets for temporary water transfers, and globalization
Cluster 3	Water Quality, Water Management, Participatory Irrigation Management, Bioretention, Multivariate Modeling And Prediction, Dss, Modeling, Non-Equilibrium, Heavy Metals, Ecology, Drainage, Dissolved Oxygen, Great Usuthu River, Gender, S1d_Dual, Netherlands, Action Plan	Water Quality and Management: Addresses water quality, development policy, participatory irrigation management, and ecological aspects
Cluster 4	Virtual Water In Libya, Optimization, Construction Costs, Water Distribution System, Water Security, Computation, Deterioration, Artificial Neural Network, Genetic Algorithm, Hot Water Networks, Economic Factors, Self Purification, Damage, Cross-Entropy, Pumps, Participatory Governance, Hydraulic Networks	Water Market and Policy: Focuses on water market, water transfer, water pricing, water shortages, and transboundary river management
Cluster 5	Water Market, Water Transfer, Water Pricing, Water Productivity, Water Policy, Water Allocation, Transboundary River Management, Iwrm, Botswana, Neoliberalism, Agricultural Water Productivity, Cropping Pattern, Heterogeneity, Application Efficiency, Conceptual Framework	Optimization and Virtual Water: Involves optimization, infrastructure of water supplies, virtual water in Libya, economic factors related to water, and digitalization
2012	Keywords	Qualification
Cluster 1	Computable General Equilibrium, Water Supply, Water Crisis, Performance Indicators, Water Distribution Network, Water Quality, Water Allocation, Water Distribution, Multi-Objective Optimisation, Inter-Basin Water Transfer (Ibwt), Stone-Geary, Reservoir, Shared Rivers, Agricultural Land, Scarcity, Impervious Cover, Stochastic Search Algorithms, Pretoria, Water Shortage, Water Reconciliation, Watershed Management, Modeling, Complex Network, Water Demand, Waste Load Allocation, Multiobjective Programming, Evolutionary Computation, Groundwater, Hazard, Earthquake, Non Market Valuation	Water Crisis and Distribution: Concentrates on water supply, water distribution, water shortages, multi-objective optimization, and digitalization
Cluster 2	Campylobacter, Asset Management, Alternative Solutions, Greywater Treatment, Life Cycle Assessment, Water Resource Management, Iwrm, Water Pond, Decentralization, Drought, Contamination, Decision Support System, Environment, Fuzzy Sets, Directive 80/778/Eec, Agriculture Sustainability, Identification, Drinking Water, Cross-Sectional Survey, Calnistea Catchment, Rehabilitation Planning, Optimisation, Irbm, Economic Input-Output Model, Leakage Potential, Break Prediction Model	Alternatives and Environmental Aspects: Covers asset management, alternative solutions, life cycle assessment, and sustainable agriculture
Cluster 3	Management, Global Water Governance, Water Security, Water Management, Hydrology & Water Resource, Desalination, Water Use, Sustainability, Sprinkler Irrigation, Economics & Finance, Water Administration, Australia, Water Pricing, On-Farm Water Storage, Water Vulnerability Evaluation, Flood Management, Environmental Security, Economic Models, Climatic Indices, Water Geography	Management and Water Administration: Encompasses global water governance, water security, and sustainable water use
Cluster 4	Replacement, Water Framework Directive, Water Distribution System, Valves, Water Utilities, Water Losses, Full Cost Recovery, Calibration, Cost Of Quality, Water Pipeline, Rehabilitation, Intrusion Volumes, Leakage, Water Sampling, Infrastructure, Network Analysis, Water Quality Model, Effective Supply	Water Pipelines and Distribution: Deals with water distribution systems, infrastructure, and network analysis
Cluster 5	Energy, Water, Biofuel, Competition, Bolivia, International Trade, Bayesian Belief Networks, Human Rights, Embodied Flow, Political Ecology, Blue And Green Water, Power Efficiency, Conflict, Life Cycle Cost, Ethanol, Regulation Upstream Competition, Abstraction Pricing, Newspapers	Energy and Water Conflicts: Explores the interaction between energy and water, political ecology, international trade, and social conflict
2022	Keywords	Qualification
Cluster 1	Extreme Weather Events, Climate Change, Vulnerability, Biodiversity, Multisector Dynamics, Yangtze River Economic Belt, Circular Economy Strategies, Qinghai-Tibet Plateau (Qtp), Co2 Emission, Financial Development, Urban Stormwater, Hamun Lakes, Urban Flood, Wind, Urban Planning, Eastern Mediterranean, Building Sector, Ecosystem Vulnerability, Driving Factor, Flood Hazard Index (Fhi), Agricultural Water Resource Utilization Efficiency, Water Quality, Gis, Maize Production	Extreme Weather Events and Biodiversity: Concentrates on extreme weather events, climate change, and biodiversity
Cluster 2	Urban Flooding, Urbanization And Ecological Environment, Deeplearning, Lid, Coupling Coordination Degree, Low-Impact Development, Water Resource Carrying Capacity, Sponge City, Smart Farming, Storm Water Management Model, Field Questionnaire, Cost-Benefit, Monthly Streamflow Prediction, Dam Seepage, Genetic Algorithm, Numerical Simulation, General Circulation Model, Data Analysis, Integrated Modeling, Flood Reduction, Swmm, Matlab	Urban Flooding and Digitalization: Addresses urban flooding, digitalization, and sustainable urban regeneration
Cluster 3	Circular Economy, Sustainable Urban Regeneration, Urban Stormwater Drainage Optimization, Human Nature Connectedness, Aquaculture, Environmental Assessment, Green Supply Chain Management, Green Finance, Sustainability, Digitalization, Cleaner Production, Sustainability Transitions, Sustainable Finance, Public Decision Support, Msw Supply Chain Design	Circular Economy and Sustainability: Explores circular economy, sustainable urban regeneration, sustainable finance, and green supply chain management
Cluster 4	Water Resources, Energy Efficiency In Buildings, Sustainable Development, Crop Production, Climate Smart, Food Security, Large Basin Area, Consumer, Optimal Allocation, Bioeconomy, Food Crops	Water Resources and Food Security: Covers water resources, energy efficiency in buildings, sustainable development, and food security
Cluster 5	COVID-19, Urban Water, Stormwater, Water Demand, E-Waste, Heavy Metals, Airborne Transmission, Water Supply, Nurse Wellbeing, Sustainable Management, Pandemics	COVID-19 and Sustainable Management: Focuses on COVID-19, sustainable water management, and pandemics

in Appendix A). For each, we aim for a more detailed description of the clusters and their associated thematic focuses and overarching categories.

We also investigate the most relevant keywords within each network based on their degree centrality [40]. Degree centrality measures the popularity or prominence of a node within the network, indicating its importance in terms of connections. Finally, we identify the most central keywords, irrespective of clustering, which are shown in

Table 2. Most popular keywords in each year. “Most popular” was defined through a degree > 5, expect for 2023, ≥5 was used to increase validity.

Year	Most Popular Words (Degree)
2007	Irrigation (13), Water Supply (11), Water (8), Water Policy (8), Water Quality (7), Water Management (7), Optimization (7), Decision Support System (7), Integrated Water Resource Management (7)
2012	Water (12), Water Footprint (12), Climate Change (11), Water Distribution System (10), Water Supply (9), Water Resource (8), Water Management (8), Drinking Water (8), Water Distribution Network (7), Water Allocation (7), Water Planning (6), Water Quality (6)
2017	Water (17), Water Footprint (17), Water Governance (11), Virtual Water (10), Water Demand (9), Water Scarcity (9), Climate Change (9), Water Conservation (8), Virtual Water Trade (8), Integrated Water Resource Management (7), Water Resource (7), Water Management (7), Water-Energy Nexus (7), Water Supply (7), Water Consumption (6), Water Distribution System (6)
2022	Climate Change (17), COVID-19 (6), Circular Economy (5), Water Resources (5), Deeplearning (5), Life Cycle Assessment (5), Water Management (5), Sustainability (5)

. This provides an initial general cross-cluster overview of relevant topics and provides insights into the enduringly central terms across different years and allows us to qualify the clusters in a comprehensible way (Table A1 in Appendix A).

3.3. Interpretation of Thematic Clusters over Time

Certain cluster topics remained stable throughout the examined period, despite partial shifts in keywords. As observed in the analysis of disconnected subgraphs, the qualification of clusters also revealed that issues related to water management and water governance were of significant research importance in 2007 (Cluster 1 and Cluster 2), 2012 (Cluster 2 and Cluster 3), and 2017 (Cluster 1 and 2). Keywords pertaining to infrastructure and digital technologies typically formed distinct clusters (Cluster 5 in 2007, Cluster 1 in 2012, and Cluster 2 in 2022), while keywords addressing ecological and social issues appeared across multiple clusters.

However, there were also changes over time. From 2007 to 2012, the research landscape witnessed a notable broadening and deepening of topics. The initial emphasis on broad themes like “Water Supply and Management” transitioned towards more specific aspects such as “Water Distribution”, “Water Allocation”, and “Multi-Objective Optimization”. This shift suggests a maturation of the field, with a growing inclination towards in-depth exploration and a finer granularity in addressing water-related challenges, as well as the social embedding of water management, as indicated by keywords such as “conflict”, “human rights” (in 2012), and “stakeholder engagement” (in 2017).

The period spanning 2012 to 2017 marked a distinct focus on environmental aspects and sustainable practices. Clusters such as “Virtual Water”, “Conflicts”, “Sustainability”, and “Social Practices” underscored a broader engagement with the intricate interactions between water, the environment, and society. Similarly, there was a shift towards keywords indicating climate change within the ecological themes accordingly. Additionally, the inclusion of themes like the “Water-Energy Nexus” and “Energy Efficiency” within the water management- and water governance-related cluster reflected an emerging awareness of the importance of efficiency in managing water and energy resources. This period demonstrated a growing recognition of the complexity of water-related challenges, with an integration of complex network analyses, stochastic search algorithms, and life cycle models.

Moving forward from 2017 to 2022, the research landscape exhibited a heightened focus on digitalization and climate change. The increasing presence of topics such as “Digitalization” and “Deeplearning” indicated a rising importance of technologies like artificial intelligence and the role of digitalization in water management. Simultaneously, the emergence of the “Circular Economy” cluster, coupled with research on “COVID-19”, as well as the growing importance of keywords like “Climate Change”, “Sustainability”, and “Extreme Weather Events” in multiple clusters, highlighted a growing emphasis on sustainable resource use and resilience to external shocks. The interplay of these themes suggests a research landscape attuned to the evolving challenges posed by global phenomena, with a particular interest in leveraging digital technologies for water management and adapting to the impacts of climate change.

This offers a comprehensive understanding of the evolving network dynamics and thematic concentrations over time, essential for the analysis of underlying trends and patterns within the scholarly domain.

4. Discussion

The evolving landscape of social water research reflects several key trends. Firstly, there is a growing tendency towards interdisciplinary research, evident in topics such as the “water and energy nexus”, “social practices”, and the “circular economy”. This integration of diverse disciplines signifies a concerted effort to comprehensively tackle complex challenges within the water sector.

Secondly, the shifting cluster topics indicate an adaptation to societal changes. For instance, the inclusion of “COVID-19” reflects responses to current global events, while discussions on “urban flooding” and “sponge cities” highlight societal reactions to the impacts of climate change. Moreover, the emergence of “digitalization” underscores technological advancements and evolving research needs.

Thirdly, while similar content is not always linked within clusters, there remains a consistent centrality of management and evaluation themes, which underscores its significance within social water research. Topics such as water pricing and virtual water management are dispersed across various clusters, often interconnected with privatization, ecological concerns, agriculture, and digitalization.

Furthermore, the evolution from broad explorations to more specific inquiries, coupled with an increasing complexity in research methodologies, reflects the maturation of the field. The shift towards environmental sustainability, heightened awareness of water and energy efficiency, and the recent focus on digitalization and resilience underscore the dynamic nature of research efforts in addressing contemporary water-related challenges and may also indicate the shift from the prevention of water crises to the adaptation to ongoing crisis-ridden water relations.

As mentioned, the central focus on management and evaluation practices remains prominent, with an exception noted in 2022. Additionally, there is an emergence of new topics independent of the central subgraph, such as digitality and climate change in 2007, and sponge cities, urban impacts, and AI in 2022, which introduce new subjects into the discourse of social water research. These small subgraphs act as forerunners, as seen with climate change in 2007, which gained significance in the primary network after 2012. Notably, COVID-19 stands out as an exception within this trend. Furthermore, there is a noticeable increase in the number of disconnected subgraphs, which indicates the growth of the relevance of water across various parts of the field of social sciences. As the network grows larger and more complex, social water research emerges as an interdisciplinary field, prompting considerations of whether 2022 marks a turning point or remains an exception to this trend.

5. Conclusions

Our study responds to the increasing demand for innovative methodological approaches to tackle the expanding domain of interdisciplinary research fields like social water research, propelled by pressing global challenges such as climate change. Our proposed mixed methods approach, anchored in quantitative network analysis, allowed us to structure the initially opaque terrain of an interdisciplinary research field such as social water research and its extensive volume of scientific articles, illuminating thematic areas and dynamics. This quantitative and exploratory approach has allowed us to uncover developments and subdivisions within the emerging field of social water research that may have been overlooked by disciplinary boundaries. By laying a foundation for access to this field, this approach could pave the way for further qualitative studies. For instance, central or representative articles within clusters could be identified and subjected to qualitative content analysis, or key authors could be approached for expert interviews. Furthermore, the methodology aligns well with a critical realist paradigm, enhancing the interpretive depth and reproducibility of qualitative research. The quantitative analysis uncovers latent structures in the data, which is then explored qualitatively to provide deeper insights. This approach not only maintains but also enhances the analytical depth provided by qualitative methods, while ensuring reliability, intersubjective validity, and reproducibility in the analysis. Our study therefore exemplifies how quantitative analysis can lay the groundwork for qualitative exploration, showcasing the value of mixed methods approaches in advancing interdisciplinary research fields.

Within the evolving landscape of social water research, this approach allows to identify several key trends. Interdisciplinary research is increasingly prevalent, adapting to societal changes and technological advancements. Themes such as the “water and energy nexus”, “social practices”, and “circular economy” highlight the integration of diverse disciplines. Within our study, we could observe the emergence of a truly interdisciplinary discourse of social water research within the observed time period. Water management is still the most important topic within this field, but its discussion is increasingly interconnected with questions of digitalization, sustainability, and climate resilience. Further qualitative research could explore whether this emergent interdisciplinary discourse leads to a new epistemic community and epistemic culture that in consequence constitutes social water research as its own scientific discipline. The centrality of water management within our network analysis suggests that this epistemic culture could be characterized by the perspectives of applied science close to the viewpoints of water management and governance, which could lead to the marginalization of more critical approaches.

The responsiveness of the field to global events, like the inclusion of “COVID-19”, and the focus on technological advancements, such as “digitalization”, underscore the dynamic nature of social water research and its close connection to applied science and governance. While still relying on fundamental research from other disciplines, social water research is a highly adaptive field that is deeply shaped by the moving realities of water.

Our study reveals a consistent focus on management and evaluation practices, with themes such as water pricing and virtual water management dispersed across various clusters. The persistent centrality of these themes underscores their significance within social water research. Additionally, new topics, such as digitality, climate change, sponge cities, urban impacts, and AI, are emerging, indicating the field’s ongoing evolution and responsiveness to contemporary challenges.

Without further qualitative research, like content analysis of key articles or expert interviews within the field, a comprehensive interpretation of these trends still has to wait, but the interpretative integration highlights the importance of understanding water as an interrelated topic and therefore points towards concepts of hybridity [41,42], non-passive materialities [43,44], complex understanding of waters [45,46], and the rising interconnectedness of social complexity [47,48].

However, it is essential to acknowledge the limitations of our proposed method. The selection of articles was rudimentary, relying solely on the presence of the word “water” in titles, keywords, or abstracts. Consequently, certain areas of social water research, such as studies on marshes, wetlands, or glaciers where the term “water” may not be explicitly mentioned, might have been overlooked. This is important because, for specific questions within social water research not only but primarily linked to climate change effects, it reproduces the tendency to underestimate the importance of such border phenomena as wetlands or glaciers. To address this limitation, a preliminary exploratory qualitative research step is recommended to capture relevant keywords of the respective field comprehensively, thus broadening the scope of the data.

Furthermore, our reliance on the Web of Science for article selection introduces bias inherent in its selection criteria, which may favor certain aspects of the social sciences over others and may exhibit international disparities. This potential bias should be reflected in further studies, especially because water-related social issues often show important geographically uneven distributions that are linked to global power structures [3]. However, we believe that these selections mirror the evolving interdisciplinary and international fields like that of social water research, as the visibility of publications in platforms like Web of Science increases the likelihood of their integration into emerging fields.

It is important to note that our chosen method does not delve into the content of the articles; rather, it focuses on identifying emerging subfields within a broader interdisciplinary research field through self-selected keywords. While this approach may not pinpoint particularly innovative or insightful articles, it serves the purpose of recognizing nascent scientific domains. For the identification of specific articles, more tailored research questions may be necessary. Therefore, our work suggests several trajectories for further exploration. By refining our research questions and selecting representative articles for qualitative interpretation within each cluster, we can deepen our understanding of specific thematic fields and uncover latent ones. Additionally, further research could, for example, focus on path dependencies within the thematic evolution of interdisciplinary research fields, shedding light on the historical trajectories that have shaped their current landscapes.

In essence, the proposed step of employing a mixed method design thrives on the subsequent qualitative analysis. It is through this qualitative lens that we can delve deeper into the intricacies of social water research and glean richer insights from the identified clusters and networks. By combining qualitative research with quantitative network analysis, we are able to navigate the expanding domain of interdisciplinary social water research. Our study thus shows the value of mixed methods approaches in an advancing interdisciplinary research field, exemplifying how quantitative analysis can lay the groundwork for qualitative exploration.

Beyond a better understanding of social water research, this study could also provide crucial insights for decision makers within water management. For water management is essential to recognize a vast variety of dimensions such as “infrastructure sustainability”, “economic and financial performance”, “environmental sustainability”, and “corporate and sectoral governance” [49] (p. 43). Identifying thematic clusters could help to avoid disciplinary limitations and support the recognition of important knowledge from different areas of the field of social water research, especially if they are located outside of the central thematic cluster(s) that are structured around management and evaluation practices.