Simulation in the social sciences goes back to the 1950s (Troitzsch 2013). The agent-based approach (ABM) had been made prominent by the 1990s’ Sugarscape model (Epstein and Axtell 1996). Agents are software units that are capable of acting autonomously in a virtual environment. Autonomous action means that the actions performed by individual agents are not controlled top-down by a central processing unit, but that agents decide for themselves. In agent-based simulation models, many of these software units act and interact with each other in their virtual environment. If many of such agents are grouped together, it is possible to perform simulations that enable one to investigate social problems and phenomena. The interaction of autonomous agents enables them to grow aggregated phenomena from the bottom-up (Neumann and Troitzsch 2019). This approach enables to represent typical features of human societies that make the agent-based modelling particularly appropriate for the for social sciences. In particular it enables to represent the following features (Epstein 2006).
- Heterogeneity: In agent-based models it is possible to represent agents individually (instead of representing human beings by a “representative agent,” which is more often than not modeled as a homo oeconomicus), which implies that different agents usually have different properties.
- Bounded rationality: Agents may have limited information about their environment or may have restricted capacities for calculating possible actions or forget less recent information. Thus, actions need not be directed at optimizing but rather can be based on heuristic reasoning
However, different knowledge claims are made by different modelling approaches. These can be differentiated along the axis of nomothetic and idiographic research (Ahrweiler and Gilbert 2009). On the one hand, nomothetic modelling purposes attempt at achieving general results, typically by the means of rather simple models. The knowledge claim of these models is oriented at the ideal of mathematics to provide universally valid results independent of context. On the other hand, idiographic modelling purposes attempt at gaining insight into the particularities of a specific case, typically by utilizing more complex models.
This axis of nomothetic and idiographic knowledge claims relates to different uses of modelling. Broadly, three different uses of modelling can be distinguished (Epstein 2008, Johnson 2015): a) explanatory models in theoretical research, b) descriptive models with a more applied focus, and c) participatory models, in which researchers and stakeholders develop a model together. In explanatory models, ABM is used as a virtual laboratory for experimenting with theories (Dowling 1999). Growing macro level phenomena by local interactions of agents at the micro level (Epstein 2006) allows for exploring the implications of social science theories (Edmonds et al. 2019). Thus, they allow for making nomothetic, law like statements. On the other hand, the value of description in the process of generating scientific knowledge cannot be underestimated (Edmonds et al. 2019). Descriptive models have a more applied focus by concentrating on concrete case studies. Whereas models for theoretical exposition make universal claims of law-like statements, descriptive models represent an empirical case in space and time. Thus, they are idiographic tools. Descriptive models employ more complex micro level rules informed by the case. Typically, such models are used as scenario analysis tools (Ahrweiler et al. 2015). Much more than other modelling approaches, agent-based simulation allows for the inclusion of participating stakeholders who give input into the formalization of rule- and fact bases and can comment on intermediate results (Neumann and Troitzsch 2019). Participatory models are also based on concrete case studies, however, typically with concrete policy or real world rather than purely scientific purposes (Barreteau et al. 2014). The central feature of participatory modelling is the involvement of stakeholders in the modelling process. This can be the case at different stages of the process: Stakeholders can be involved in the development of the model, or for the verification of the agents’ rules (Moss and Edmonds 2005), as well as in the analysis of simulation results and scenarios. Participatory models can be descriptive models. However, often such models are used for representing the perspective of the stakeholders on the case, in particular for discussing different perspectives on a case. Thus, a constructivist approach is applied (Barreteau et al. 2014).
The participatory approach originates from research on resource management (Pahl-Wostl 2002, Gurung et al. 2006, Campo et al. 2009, Worrapimphong et al. 2010). However, participatory approaches have also been applied for criminological tasks (Elsenbroich et al. 2017). The problem that stimulated participatory modelling was that technological solutions provided by engineering approaches proved to be insufficient to cope with the social dimensions of problems related to sustainability (Pahl-Wostl 2002). For this purpose, integrated assessment models have been developed that integrated knowledge of an issue for decision making (Rotmans 1998), inspired by the concept of post-normal science (Funtowicz and Ravetz 1993). Post-normal science emphasizes the communication of uncertainty, justification of practice, and complexity. This approach recognizes the legitimacy of multiple perspectives on an issue, both with respect to multiple scientific disciplines as well as lay men involved in the issue. For instance, Wynne (1992) analyzed the knowledge claims of sheep farmers in the interaction with scientists and authorities. In such an extended peer community of a citizen science (Stilgoe 2009), lay men of the affected communities play an active role in knowledge production, not only because of moral principles of fairness but to increase the quality of science (Fjelland 2016).
A tool specifically suited for extending the peer community is provided by participatory modelling approaches. Whereas involvement of laymen is scientifically important for capturing the details of the concrete case in the model and thus increasing the case specific validity of the model, the specific objective of the participatory approach addresses the interests of the stakeholders. For this reason, first the audience addressed by the modelling process has to be specified: For instance, the general public has to be distinguished from domain specific stakeholders in reference to their role regarding a particular issue (Pahl-Wostl 2002). During the life cycle of the modelling process it may well be that different types of audience are addressed at different stages. The general public, e.g. as citizens or voters, might play an important role during the first step of agenda setting, whereas in later stages of developing a plan for resolving the issue and the concrete implementation of measures derived from participatory modelling, domain specific stakeholder groups become important. Dependent on the issue at stake the relevant domain specific stakeholder may be either laymen, e.g. in collective resource management in fishery, or higher-level decision maker, e.g. in river control (Barreteau et al. 2014).
Many different participatory methodologies exist (Voinov et al. 2018) such as soft system methodology (Checkland and Holwell 1998), integrated assessment (Parker et al. 2002), evidence based and conceptual modeling (Moss and Edmonds 2005, Scherer et al. 2015), or companion modelling (Étienne 2014). These approaches apply different methodologies for fact-finding, process orchestration, and modeling (Voinov et al. 2018). Knowledge acquisition may involve surveys, interviews, workshops, or crowdsourcing. The process orchestration may be facilitated by methods such as role-playing games or brainstorming (Voinov et al. 2018). The most significant variety of methods exist at the stage of modelling, ranging from qualitative conceptualizations such as rich pictures (Bell et al. 2015), cognitive mapping (Novak and Cañas 2008), cultural consensus methods (Paolisso 2015), or causal loop diagrams (Sedlacko et al. 2014), to semi-quantitative models such as fuzzy cognitive mapping (Gray et al. 2014), network analysis (Prell 2012), or scenario building (Amer et al. 2013), and finally quantitative methods such as GIS maps (Sieber 2006), cost-benefit analysis (Hanley et al. 2009), Bayesian networks (Carmona et al. 2013), system dynamics (Costanza and Voinov 2003), or technically ambitious approaches such as agent-based models (Bonabeau 2002). For the co-creation of agent-based models, the engineering agent-based social simulation framework (Sieber and Klügl 2017) provides a structured approach for community modelling: In moderated focus group discussions stakeholders co-create (Mitleton-Kelly 2003) a conceptual model of the problem at hand in the framework of the unified modelling language (UML). The analysis is split into three steps of problem analysis, model scope definition, and definition of key actors and activities and the interactions among agents and between agents and objects. The result is a detailed conceptual model that subsequently can be implemented by software engineers.
One of the most well-known participatory approaches is the so-called companion modelling (ComMod) developed at CIRAD, a French agricultural research center for international development. The term companion modelling has been coined originally by Barreteau et al. (2003) and been further developed to a research paradigm for decision making in complex situations to support sustainable development (Étienne et al 2014). While participatory approaches such as companion modelling are highly case specific and for this reason no universally valid methodology can be specified, some invariants and key elements can be identified, nevertheless. First of all, companion modelling is a process that in some cases can be very long lasting. Following Barreteau et al. (2014) the main protagonists of this process can be summarized as laymen, researchers, and technicians, and from case to case additionally supplemented by an institutional dimension of political or economic decision makers. Laymen are the local stakeholders to whom the modelling serves. Researcher are scientists specialized in the particular domain that is subject of the modelling such as land use or river regulation. Technicians are those people who actually program the model. The whole process is facilitated by so-called commodians, typically the scientists familiar with domain, who manage the interrelations of the involved actors. As most participatory methods, companion modelling is an iterative process, sometimes over several years (Constanza and Ruth 1998). A feature specific for this approach is that the model development is facilitated by role playing games, typically during stakeholder workshops, which are reflected in a subsequent debriefing.
A central feature not only, but in particular of companion models is the development of mental models that represent the perspectives of the stakeholders on an issue (Pahl-Wostl 2002, Daré et al. 2014). This reflects the constructivist perspective of the companion modelling approach (Barreteau et al. 2014). The objective of the model is not to represent an “objective” reality but to reveal the different views of different stakeholders on one and the same issue. The model serves as a boundary object that facilitates exchange of worlviews (Schmitt Olabisi et al. 2014). Eventually this might change the reference frame of norms through which the domain is constructed when different and eventually conflicting views are brought in a discourse. Thereby the co-creation of model becomes a process of knowledge construction (Radinsky et al. 2016). Argyris and Schön (2002) characterize this process by differentiating single and double loop learning, in which actors are not only learning within one reference frame to describe the world but also learn about the reference frame itself (Zellner and Campbell 2015). The objective of this constructivist approach is the facilitation of communication and social learning (Pahl-Wostl 2002, Castella 2009, Johnson 2015) with the ultimate objective of procedural justice in decision making (Joss and Brownlea 1999) and the empowerment of local stakeholders to self-govern eventually conflicting issues of their concern (Barreteau et al. 2007). However, achieving this objective is bound to critical preconditions: namely trust of the involved community in the overall process. Implementing participatory modelling needs to be legitimized in order to facilitate engagement. As inevitably not only laymen but also scientists are involved in the process, it has been recognized that legitimacy is achieved by a local anchoring of these scientists (d’Aquino 2009), whereas in case of cultural differences, a discourse open, accessible, and safe for honest discourse facilitates engagement (Voinov et al 2018). For this purpose, AI-FORA applies the concept of safe spaces for engaging local stakeholders.
Ahrweiler, P., Gilbert, N. (2009). The epistemologies of social simulation research, In Squazzoni, F. (Ed.). Epistemological aspects of computer simulation in the social sciences (pp. 12-28). New York: Springer.
Ahrweiler, P., Schilperoord, M., Pyka, A., Gilbert, N. (2015). Modelling Research Policy: Ex-Ante Evaluation of Complex Policy Instruments. Journal of Artificial Societies and Social Simulation 18 (4).
Amer, M., Daim, T. U., Jetter, A. (2013). A review of scenario planning. Futures 46: 23-40.
Argyris, C., Schön, D. A. (2002). Apprentissage organisationnel: théorie, méthode, pratique. Paris: De Boeck Université.
Barreteau et al. (2003). Our companion modelling approach. Journal of Artificial Societies and Social Simulation 6(1).
Barreteau, O., Le Page C., Perez, P. (2007). Contribution of simulation and gaming to natural resource management: an introduction. Simulation & Gaming 38(2): 185-194.
Barreteau O., Bousquet F., Étienne M., Souchère V., d’Aquino P. (2014). Companion Modelling: A Method of Adaptive and Participatory Research. In: Étienne M. (eds) Companion Modelling. Dordrecht: Springer.
Bell, S., Berg, T., Morse, S. (2015). Rich Pictures: Sustainable Development and Stakeholders–The Benefits of Content Analysis. Sustainable Development 24(2): 136-148.
Bonabeau, E. (2002). Agent-based modeling: Methods and techniques for simulating human systems. Proceedings of the National Academy of Sciences 99(suppl 3): 7280-7287.
Campo, P. C., Mendoza, G. A., Guizol, P., Villanueva, T. R., Bousquet, F. (2009). Exploring management strategies for community-based forests using multi-agent systems: A case study in Palawan, Philippines. Journal of Environmental Management 90(11): 3607–15.
Carmona, G., Varela-Ortega, C., & Bromley, J. (2013). Participatory modelling to support decision making in water management under uncertainty: two comparative case studies in the Guadiana river basin, Spain. Journal of environmental management 128: 400-412.
Castella, J.-C. (2009). Assessing the role of learning devices and geovisualisation tools for collective action in natural resource management: Experiences from Vietnam. Journal of Environmental Management 90(2): 1313–9.
Checkland, P., Holwell, S. E. (1998). Information, Systems and Information Systems: Making Sense of the Field. Chichester: John Wiley and Sons.
Constanza, R., Ruth, M. (1998). Using Dynamic Modeling to Scope Environmental Problems and Build Consensus. Environmental Management 22(2): 183-195.
Costanza, R., Voinov, A. (Eds.)(2003). Landscape simulation modeling: a spatially explicit, dynamic approach. New York: Springer-Verlag.
D’Aquino P. (2009). La participation comme élément d’une stratégie globale d’intervention: the self-incrementing management approach. Cahiers Agricultures 18 (5): 433-439.
Daré, W., Van Paassen, A., Ducrot, R., Mathevet, R., Queste, J., Trébuil, G., Barnaud, C., Lagabrielle, E. (2014). Learning about interdependencies and dynamics. In M. Etienne (Ed.). Companion Modelling. A Participatory Approach to Support Sustainable Development, (pp. 233-262). Dordrecht: Springer.
Dowling, D. (1999). Experimenting on theories. Science in Context 12 (2): 261-273.
Edmonds, B. et al. (2019). Different modelling purposes. Journal of artificial societies and social simulation 22(3).
Elsenbroich, C., Anzola, D., Gilbert, N. (2017)(eds.). Social dimensions of organized crime. Cham: Springer.
Epstein, J. (2006). Generative social science. Studies in agent-based computational modeling. Princeton, NJ: Princeton University Press.
Epstein, J. (2008). Why model? Journal of Artificial Societies and Social Simulation (11/4).
Epstein, J., Axtell, R. (1996). Growing artificial societies: Social science from the bottom up. Cambridge: MIT Press.
Étienne, M. (2014)(Ed.). Companion Modelling: A Participatory Approach to Support Sustainable Development. Dordrecht: Springer.
Fjelland, R. (2016). When Laypeople are Right and Experts are Wrong: Lessons from Love Canal. International Journal for Philosophy of Chemistry 22(1): 105–125.
Funtowicz, S., Ravetz, J. (1993). Science for the post-normal age. Futures 31(7): 735-755.
Gray, S. A., Gray, S., Cox, L., Henly-Shepard, S. (2013). Mental modeler: A fuzzy-logic cognitive mapping modeling tool for adaptive environmental management. Proceedings of the 46th International Conference on Complex Systems, 963-973.
Gurung, T. R., Bousquet, F., and Trebuil, G. (2006). Companion Modeling, Conflict Resolution, and Institution Building: Sharing Irrigation Water in the Lingmuteychu Watershed, Bhutan. Ecology and Society 11(2): 36.
Johnson, P. (2015). The SWAP model: Policy and theory applications for agent-based modelling of soil and water conservation adoption. Surrey: Doctoral thesis.
Joss, S. Brownlea, A. (1999). Considering the Concept of Procedural Justice for Public Policy and Decision-Making in Science and Technology. Science and Public Policy 26 (1999): 321-330.
Mitleton-Kelly, E. (2003). Complexity research approaches and methods: The LSE complexity group integrated methodology. In Keskinen, A., Aaltonen, M., Mitleton-Kelly, E. (Eds.). Organisational Complexity (pp. 56–77). Tutu Publications. Finland Futures Research Centre, Turku School of Economics and Business Administration, Turku, Finland.
Moss, S., Edmonds, B. (2005). Sociology and simulation. Statistical and qualitative cross-validation. American journal of sociology 110(4): 1095-1131.
Neumann, M., Troitzsch, K.G. (2019). Social Simulation. In SAGE Research Methods. DOI: 10.4135/9781526421036764890
Novak, J. D., Cañas, A. J. (2008). The theory underlying concept maps and how to construct and use them. Technical Report IHMC CmapTools.
Pahl-Wostl, C. (2002). Participative and Stakeholder-Based Policy Design, Evaluation and Modeling Processes. Integrated Assessment 3(1): 3-14.
Paolisso, M. (2015). Understanding Culture and Environment Dynamics Using Cultural Consensus Analysis. In Ruth, M. (Ed.). Handbook of Research methods and Applications in Environmental Studies. Cheltenham, UK : Edward Elgar Publishing.
Parker, P. et al. (2002). Progress in integrated assessment and modelling. Environmental Modelling & Software 17(3): 209–217.
Radinsky, J., Milz, D., Zellner, M., Pudlock, K., Witek, C., Hoch, C., Lyons, L. (2017). How planners and stakeholders learn with visualization tools: using learning sciences methods to examine planning processes. Journal of environmental planning and management 60(7): 1296-1323.
Rees, M. (2017). Black holes are simpler than forests and science has its limits. AEON, 1 December.
Prell, Ch. (2012). Social network analysis: history, theory, and methodology. London: Sage.
Rotmans, J. (1998). Methods for IA: The Challenges and Opportunities Ahead. Environmental Modelling and Assessment 3 (1998): 155-180.
Scherer, S., Wimmer, M., Markisic, S. (2013). Bridging narrative scenario texts and formal policy modelling through conceptual policy modelling. Artificial intelligence and law 21(4): 455 – 484.
Schmitt Olabisi, L., Blythe, S., Ligmann-Zielinska, A., Marquart-Pyatt, S. (2014). Modeling as a tool for cross-disciplinary communication in solving environmental problems. In O’Rourke, M., Crowley, S., Eigenbrode, S.D., Wulfhorst, J.D. (Eds.). Enhancing communication & collaboration in interdisciplinary research (pp. 271-290). Thousand Oaks, CA: Sage.
Sedlacko, M., Martinuzzi, A., Røpke, I., Videira, N., Antunes, P. (2014). Participatory systems mapping for sustainable consumption: Discussion of a method promoting systemic insights. Ecological Economics (106): 33-43.
Sieber, P., Klügl, F. (2017). What software engineering has to offer to agent-based social simulation. In Edmonds, B., Meyer, R. (eds.). Simulating social complexity. A handbook 2nd edition (pp. 81-117). Heidelberg: Springer.
Sieber, R. (2006). Public participation geographic information systems: A literature review and framework. Annals of the association of American Geographers 96(3): 491-507.
Stilgoe, J. (2009). Citizen Scientists: Reconnecting Science with Civil Society. London: Demos.
Troitzsch, K. G. (2013). Historical introduction. In B. Edmonds & R. Meyer (Eds.), Simulating social complexity (pp. 13–21). New York, NY: Springer.
Voinov, A. et al. (2018). Tools and methods in participatory modelling. Selecting the right tool for the job. Environmental modelling and software 109: 232-255.
Worrapimphong, K., Gajaseni, N., Le Page, C., and Bousquet, F. (2010). A companion modeling approach applied to fishery management. Environmental Modelling & Software, 25(11): 1334–1344.
Wynne, B. (1992). Misunderstood misunderstanding: social identities and public uptake of science. Public Understanding of Science 1(3): 281–304.
Zellner, M.L., Campbell, S. (2015). Planning for Deep-Rooted Problems: What Can We Learn from Aligning Complex Systems and Wicked Problems? Planning Theory and Practice 16(4): 457-478.