Sociology and Complexity Science Blog

04/05/2025

The Atlas of Social Complexity. Chapter 28: Make Love, Not Models

QUICK OVERVIEW AND LINKS TO THE OTHER THEMES AND CHAPTERS IN THE BOOK

The first major content theme in T he Atlas of Social Complexity is Cognition, Emotion and Consciousness. This first theme includes six chapters, which I have so far blogged on. Chapter 6 addresses autopoiesis. Chapter 7 turns to the role of bacteria in human consciousness. Chapter 8 explores how the immune system, just like bacteria and cells, is cognitive – and the implications this has for our wider brain-based consciousness. Chapter 9 explores a complexity framing of brain-based cognition, emotion and consciousness. Chapter 10 explores the complex multilevel dynamics of the Self. Chapter 11 is about human-machine intelligence.

The second major content theme in The Atlas of Social Complexity is The Dynamics of Human Psychology. So far for this theme, I’ve given a basic overview, found here. I then moved on to the first theme, Human psychology as dynamical system (Chapter 13). From there I reviewed Chapter 14: Psychopathology of mental disorders ; Chapter 15: Healing and the therapeutic process; and Chapter 16: Mindfulness, imagination, and creativity.

The third major theme is living in social systems (Chapter 17). The first chapter in this section is Complex social psychology (Chapter 18). From there we move on to Collective behaviour, social movements and mass psychology (Chapter 19). Next is Configurational Social Science (Chapter 20). From there we move to the Complexities of Place (Chapter 21); followed by Socio-technical Life (Chapter 22). Chapter 23 turned to the theme of Governance, Politics and Technocracy. Chapter 24 focused on The Challenges of Applying Complexity. Chapter 25 focused on Economics in an unstable world. And, finally, Chapter 26 focused on resilience and all that jazz. Chapter 27 introduces the final theme of the book, Methods and complex causality.

The focus of the current chapter (Chapter 28) is Make Love, Not Models and explores the challenges of modelling methods today.

OVERVIEW OF CHAPTER

This chapter revisits a recurring tension in the modelling of social complexity: the allure of elegant abstraction versus the demands of real-world complexity. It begins with a seminar—an impressive model of crowd behaviour, mathematically refined, visually striking. Yet beneath its surface lies a telling absence: no data, no context, no engagement with the psychological or social dynamics of human crowds. This moment crystallises a larger issue. The chapter explores how certain strands of complexity science, particularly those rooted in the physical-computational tradition, risk prioritising form over substance. These approaches, while technically sophisticated, often assume that methods developed for natural systems can be seamlessly transferred to the social world. But social systems are not just complex—they are contextual, reflexive, and shaped by meaning, memory, and power.

Midway through, the chapter reflects on the limits of abstraction. Models such as Kauffman’s NK framework or power-law-based network theories offer valuable insights but struggle to translate into domains where variability, interpretation, and lived experience are central. Rather than dismiss these tools, the chapter calls for more grounded modelling—work that stays close to the systems it seeks to understand, blending formal methods with qualitative inquiry and participatory design. Social systems require models that support learning, not just prediction. They call for a shift from control to collaboration, from mechanical metaphors to ecological ones.

Ultimately, the chapter advocates for a modelling practice that is humble, pluralistic, and open to complexity in all its forms. It is not a rejection of mathematics but a rebalancing of method and meaning.

29/04/2025

WHO symposium on modelling and optimizing the health and care workforce: Complexity Thinking for Equitable and Resilient Healthcare Workforces

I would like to thank Tomas Zapata, Cris Scotter, and their team in the Unit of Health Workforce and Health Services in the WHO Regional Office for Europe (Copenhagen) for the opportunity to present at the WHO symposium on modelling and optimizing the health and care workforce.

OVERVIEW OF WHO CONFERENCE

On 28–30 April, WHO/Europe convened a diverse group of experts for a 3-day symposium at UN City, Copenhagen, focused on advancing innovative practices in health workforce modelling and optimization. Bringing together thought leaders, policy experts, technical specialists and practitioners across sectors, this event will provide a platform to explore the complexities of workforce planning in a rapidly evolving landscape. The event was organized in partnership with the Government of Ireland, Socialstyrelsen (National Board of Health and Welfare of Sweden) and Durham University. With health systems facing acute workforce shortages, shifting service demands and the need for more flexible, resilient health services, optimizing workforce planning with data-driven strategies has never been more critical.

CLICK HERE for a PDF of my presentation

CLICK HERE for an overview of the conference

CLICK HERE for an overview of the programme

CLICK HERE to watch some of the main talks and plenaries

CLICK HERE to go to my map of the complexity sciences

CLICK HERE to see COMPLEX-IT, our interdisciplinary methods platform for nontechnical users in health policy, including systems mapping, machine learning, simulation, cluster analysis, data forecasting and data visualization.

OVERVIEW OF MY TALK:

WORKFORCE FUTURES RE-IMAGINED:

Complexity Thinking for Equitable and Resilient Healthcare Workforce(s)

Why Complexity?

I was asked to be part of the opening plenary, which set the theme for the conference, by talking about the value of the complexity sciences and systems thinking, as well as transdisciplinary modelling approaches, for addressing the current challenges facing the healthcare workforce throughout the world.

The main aim of my talk was not to offer another technocratic fix, but to provoke a deeper conversation about the paradigms shaping our approach to the healthcare workforce crisis, and why they continue to fall short – as viewed through the lens of the complexity sciences and systems thinking (See online map of the complexity sciences).

For over sixty years, the social and health sciences have documented the same litany of systemic failures: health inequities, chronic under-funding, workforce burnout, structural discrimination, and slow policy responsiveness, among others. These are not new problems. They are persistent problems — because they are complex. They are “wicked” in the truest sense: entangled in political, cultural, economic, and institutional dynamics that resist linear, reductionist solutions.

And yet, our modelling practices too often remain stuck in epistemic inertia. The same people in the same rooms using the same tools and asking the same questions, resulting in the same limited answers. The future, however, demands a different grammar — one that complexity science is uniquely positioned to offer.

We need to disruption the path-dependent inertia of our present trajectory.

As Lasse Gerrits and I outline in The Atlas of Social Complexity, disrupting the healthcare systems of different counties is a wicked problem, largely because healthcare workforces are complex socio-ecological system: historically contingent, politically contested, shaped by nested and emergent dynamics, and riddled with nonlinear feedback loops. These systems demand a rethinking of how we model, plan, and govern workforce futures.

Complexity thinking reframes healthcare systems and healthcare workforces as socio-ecological systems — nested, emergent, and historically contingent. They are not machines to be optimised but living systems to be understood, shaped, and co-evolved. In this framing, the healthcare workforce becomes a system within systems — shaped by feedback loops, power relations, and path-dependent dynamics that operate across local and global scales.

Central to my keynote was the value of case-based complexity, an interdisciplinary methodological approach grounded in the work of David Byrne and Charles Ragin (See Sage Handbook of Case-Based Methods). It offers a shift away from universal laws and one-size-fits-all projections, toward a configurational view of causality — where outcomes arise from complex interactions among social, institutional, and ecological factors.

Equifinality, multifinality, and causal asymmetry are not abstract concepts. They are essential tools for rethinking workforce modelling. What works in one country, region, or profession may not work elsewhere, and may even cause harm.

Multifinality: how similar paths lead to different conclusions.
Equifinality: how different paths can lead to the same conclusion.
Causal asymmetry: how what works in one setting might fail in another.

Health workforce ecosystems are ensembles of cases, each requiring situated, context-sensitive interventions. Complexity, in this view, is not a complication. It is a place-based contextual truth.

This epistemological shift carries profound implications. It calls for modelling practices that are not only computationally and statistically sophisticated, but also reflexive, co-productive, and ethically aware. It demands interdisciplinary platforms that integrate qualitative insight, lived experience, and participatory governance. It insists that equity is not an afterthought but an emergent property of system design — and that justice requires we see discrimination not as deviance but as a reproducible system output.

For example, Complexity resists heroic models of leadership. Instead, leadership emerges from distributed systems, collaborative processes, and shared sensemaking. The myth of top-down control must give way to adaptive, ethical governance. Going further, discrimination is not an anomaly but a systemic output. A complexity-informed equity lens reveals how racism, sexism, and other forms of power and injustices are reproduced institutionally. Equity must be built into the very fabric of modelling, not as consultation, but as co-authorship.

Ultimately, my message was simple: if we want to build health workforce systems that are resilient, equitable, and future-ready, we must let go of outdated paradigms. Complexity is not a buzzword. It is a different way of knowing and of acting. The question now is whether we have the courage to embrace it.

22/04/2025

The Atlas of Social Complexity. Chapter 27: Theme 5, Methods

The focus of the current post is CHAPTER 27: METHODS

QUICK OVERVIEW OF THEME 5

With Chapter 27 we move to the final theme of our book: methods. Theme 5 seeks to balance a rigorous critique of the methods of computational and complexity science, while also providing a clear horizon along which various advances are being made in answer to the thirteen situations. The chapters in this theme include Make love, not models (Chapter 28), Revisiting complex causality (Chapter 29), Mapping the new methods terrain (Chapter 30) and Getting philosophically real (Chapter 31). Chapter 29, in particular, includes a survey of the qualitative and interdisciplinary methods we see as most promising, which includes various case-based methods and systems mapping. Chapter 31 end the theme and the tour by grounding it all in the philosophical framing we find most useful, complex realism.

A bit more in-depth explanation:

While we have been hypercritical of the methods turn in the complexity sciences, we do acknowledge that it does mean that a lot has been happening in the field of methods, particularly in the areas of application, mixed-methods, decision making support and policy.[1] The question is, “To what extent have these advances embraced a social complexity imagination sufficient to overcome this area’s particular configuration of the thirteen situations to truly become transdisciplinary?”

The short answer is, “We still have a very long way to go”.

As a reminder, on the natural, mathematical and computational sciences side, in addition to a failure to engage the wider social sciences (Situation 2), the current methodological limitations includes: technique in the absence of theory (Situation 7), the lean toward predictive machines over learning tools (Situation 8), the minor if not entirely absent role of qualitative methods (Situation 9), and the dire sound of technicalities. On the social sciences side, it has to do with the methodological closing of the social scientific mind (Situation 10), which, while somewhat improving, still faces, particularly at the undergraduate level, the continued need for a better methods curriculum; and, in terms of staffing, the need to hire more transdisciplinary methods experts in social science and computer science departments. Despite this inexcusable lapse, many academic disciplines remain largely nonplussed.

Still, there are a handful of rather productive trajectories of methodological advance that we, as tour guides, strongly recommend pursuing for those with interested in advancing transdisciplinary methods. So far on our tour, we have highlighted (albeit somewhat indirectly) four: dynamical systems theory, complex network analysis, case-based configurational methods and complexity in evaluation.

There are other methodological avenues that we have yet to formally review, which we believe also have significant potential. They fall into three major categories – case-based complexity, systems mapping, and approachable modeling and smart methods, or AM-Smart for short – and with half of these methods being qualitative. Examples across the three include trajectory-based QCA, causal loop diagrams, participatory systems mapping, and system dynamics. Our purpose in this theme is, in part, to review these new methods – which constitutes the focus of Chapter 30 (Mapping the new methodological terrain). Our focus in Chapter 30 is primarily to set up a cafeteria approach to methods. We want to encourage readers to really explore the possibility of using, combining, or developing new and different methodological suites or interdisciplinary methodological repertoires.

Our other goal, which forms the majority of chapters in Theme 5, is more philosophical. Since the beginning of the tour, we have rallied against a certain way of doing social complexity, which for lack of a better phrase, is grounded in a ‘naturalising’ or physical-computational science stance. While we have variously explained our concerns with this approach, we have not formally addressed the underlying epistemologies driving it. Here is a point to remind readers: we are not opposed to any method per say, as much as we are opposed to how they are used. Our concerns are one of approach. Our philosophical objective in Theme 5 is to have an in-depth conversation with the methodologies we find the most problematic. We do this in two ways. We will riff (in the spirit of jazz improvisation) on a variety of epistemological issues and concerns across chapters 28 through 30, comparing and contrasting different ways of methodologically thinking about causal complexity. We end by outlining our own position, Chapter 31, which is grounded in a complex, critical realism.

[1] See, for example, Barbrook-Johnson, P., & Carrick, J. Combining complexity-framed research methods for social research. International Journal of Social Research Methodology. 2021: 1-14; Gilbert, N., et al. Computational modelling of public policy: Reflections on practice. Journal of Artificial Societies and Social Simulation. 2018: 21(1).

09/04/2025

The Atlas of Social Complexity. Chapter 26: Resilience and all that Jazz

The first major content theme in The Atlas of Social Complexity is Cognition, Emotion and Consciousness. This first theme includes six chapters, which I have so far blogged on. Chapter 6 addresses autopoiesis. Chapter 7 turns to the role of bacteria in human consciousness. Chapter 8 explores how the immune system, just like bacteria and cells, is cognitive – and the implications this has for our wider brain-based consciousness. Chapter 9 explores a complexity framing of brain-based cognition, emotion and consciousness. Chapter 10 explores the complex multilevel dynamics of the Self. Chapter 11 is about human-machine intelligence.

The focus of the current post is CHAPTER 26: RESILIENCE AND ALL THAT JAZZ

QUICK OVERVIEW OF CHAPTER

Environmental sciences in a book about social complexity? Certainly. The study of ecologies and natural systems has generated a wealth of insights about complexity that cannot be ignored by social scientists. Importantly, the distinction between social and natural or physical systems is analytical only. In reality, social life is deeply embedded in all sorts of ecological processes. It is common to think of this embedding or coupledness in terms of socio-ecological systems. This part of the tour surveys the complexity of that coupledness, e.g., in terms of coevolution, and discusses its analytical implications for the study of social complexity. We will spend ample attention to the concept of resilience because it originates from the study of coupled socio-ecological systems, and in many ways has gained more traction than other concepts from the complexity sciences.

The idea that the natural state of the earth is being put under pressure as a consequence of humankind’s exploitation has a long history. While there may have been times when smoking chimneys were regarded as signs of improvement during the Industrial Revolution,[1] such romanticised images of societal progress have long since disappeared. The environmental crisis that has risen since then, as exemplified in e.g., the break-down of the ozone layer, acid rain, and, more recently, rising worldwide average temperatures, is clear for all but the most sceptical. The study of the environment and its reciprocal relationship with human activities continues to attract considerable scholarly attention. Importantly, this scholarly work has contributed significantly to the complexity canon. Consider, for example, how system dynamics modelling aids to understanding the feedback loops driving environmental change;[2] how the concept of coevolution articulates the reciprocity in coupled socio-ecological systems;[3] or how hysteresis helps understanding why it is incredibly difficult to restore ecological systems to a previous stable state once toppled into a new state.[4]

The inclusion of this intellectual strand in the Atlas is threefold. Firstly, it is one of the fields that was quick to recognize that the earth, the environment, the atmosphere, the ecosystems and their niches, etc. are best understood as complex systems.[5] In terms of a human example, fertilizers and climate change can impact the vegetation of a grassland (biotic) and the quality of the minerals and soil (abiotic) upon which the ecosystem depends, which leads to changes in biodiversity. This is why an ecological community is systemic and complex.

This insight takes us to the second reason why this theme should be explored in the Atlas: it is useless to think of ecosystems as existing independently from complex social systems. Humans and their environmental are fully intertwined.

The third reason follows from the previous two: of all concepts and methodologies gathered in the complexity canon, it is those from environmental sciences that – on the whole – appear to resonate strongly with a wider audience in their application to pressing societal problems.

One of those rich aspects is resilience.

Resilience, it must be stressed again, is a concept used just as much outside of social complexity or the study of SES, as it is within these fields.[6] Somewhat predictably, there are multiple and competing definitions of resilience.[7] This is not only because different applications ask for different conceptualizations, but it is also because the concept has been around long enough to be refined (but not discarded).

Arguably the most well-known version is the one introduced by Crawford Stanley Holling in 1973. In this seminal paper[8], he argues that resilience is present when systems bounce back after having been put under strain.

Clearly, there is more to it than this simple working definition. What, exactly, is a socio-ecosystem’s internal structure? How can one know a system’s limits, and, knowing that coupled systems are multifaceted instead of monoliths, how many limits does a system have? What are the timescales for short-term and long-term recovery? And, importantly, under what conditions can a system be said to have been restored to its original state? These are tough questions, but the last one is especially difficult to answer when it comes to social systems.

The main point is this: definitions and operationalizations of resilience are much harder than the general idea of resilience. It is very difficult to move from that general idea (that works as long as it is being kept simplistic and abstract) to concrete assessments of such resilience (that quickly can become something in the eye of the beholder).

The power of the concept is that it can be understood intuitively and that it offers an overarching concept that can host several other important aspects of complex systems, including regime shifts, punctuated equilibrium and hysteresis.[9] A such, it captures many aspects that we surveyed in this Theme 4 of the Atlas – and so, we suggest all readers engage this topic as it is not standalone. Instead, it is deeply tied to everything else we have surveyed on our tour.

[1] Jürgen Osterhammel, Die Verwandlung der Welt: Eine Geschichte des 19. Jahrhunderts, 4. Aktual. edition (München: C.H.Beck, 2009).

[2] D. H. Meadows, ‘Whole Earth Models and Systems’, CoEvolution Quarterly, 1982, 98–108.

[3] M. A. Gual and R. B. Norgaard, ‘Bridging Ecological and Social Systems Coevolution: A Review and Proposal’, Ecological Economics 69, no. 4 (2010): 707–17.

[4] Marten Scheffer and Stephen R. Carpenter, ‘Catastrophic Regime Shifts in Ecosystems: Linking Theory to Observation’, Trends in Ecology & Evolution 18, no. 12 (December 2003): 648–56.

[5] S. A. Levin, ‘Ecosystems and the Biosphere as Complex Adaptive Systems’, Ecosystems 1, no. 5 (1998): 431–36; Gregg Hartvigsen, Ann Kinzig, and Garry Peterson, ‘Complex Adaptive Systems: Use and Analysis of Complex Adaptive Systems in Ecosystem Science: Overview of Special Section’, Ecosystems 1, no. 5 (1 September 1998): 427–30; C. S. Holling, ‘Understanding the Complexity of Economic, Ecological, and Social Systems’, Ecosystems 4, no. 5 (1 August 2001): 390–405.

[6] Tuna Taşan-Kok, Dominic Stead, and Peiwen Lu, ‘Conceptual Overview of Resilience: History and Context’, in Resilience Thinking in Urban Planning, ed. Ayda Eraydin and Tuna Taşan-Kok, GeoJournal Library (Dordrecht: Springer Netherlands, 2013), 39–51; Ran Bhamra, Samir Dani, and Kevin Burnard, ‘Resilience: The Concept, a Literature Review and Future Directions’, International Journal of Production Research 49, no. 18 (15 September 2011): 5375–93.

[7] Patrick Martin-Breen and J. Marty Anderies, ‘Resilience: A Literature Review’, 2011.

[8] C. S. Holling, ‘Resilience and Stability of Ecological Systems’, Annual Review of Ecology and Systematics 4 (1973): 1–23.

[9] Joana Figueiredo and Henrique M. Pereira, ‘Regime Shifts in a Socio-Ecological Model of Farmland Abandonment’, Landscape Ecology 26, no. 5 (1 May 2011): 737–49; Marjolein Sterk, Ingrid A van de Leemput, and Edwin THM Peeters, ‘How to Conceptualize and Operationalize Resilience in Socio-Ecological Systems?’, Current Opinion in Environmental Sustainability, Sustainability governance, 28 (1 October 2017): 108–13.

08/04/2025

Advancing a social complexity imagination to disrupt place-based health complexities -- IRSPM Conference Presentation

It was great to present with my colleague Jonathan Wistow at the International Research Society for Public Management (IRSPM) Conference, hosted by the University of Bologna, Italy.

For those who attended our presentation or are just interested in what we discussed, here is an overview with links to a PDF of our presentation and our methods platform, COMPLEX-IT. See, also, this blog post by my co-author, Jonathan Wistow, as background for our argument.

ABSTRACT

Advancing a social complexity imagination to disrupt place-based social complexities: Health inequalities and multiple conjunctural causation

The complexity and persistence of health inequalities pose significant challenges to public management and service delivery, affecting both their understanding and response strategies (Wistow et al., 2015). For Salway and Green (2017), part of the problem is the miss-specification of system boundaries in public policy that fails to go far enough up the causal chain. Policy and research have neglected the complex causal factors that produce the socio-economic inequalities in which health inequalities are (re)produced, (Doyal with Pennell, 1979, Navarro, 2009, and Schrecker 2017), which leads us to extend a theoretical exploration of the issue, meeting a key aim of this panel. One potential solution is to employ a social complexity framework and corresponding methods that embrace rather than simplify these persistent complexities (Wistow et al., 2015).

We develop a ‘social complexity imagination’ as a form of creative destruction designed to drive science and its relationship with practice (Castellani and Gerrits, 2024). We do so by framing multi-level and multi-actor governance system responses to the deeply embedded and ‘wicked challenge’ of health inequalities within debates about how the social contract and political economy influences place-based social complexities (i.e., trajectories of unequal social and economic outcomes) and the scope and capacity for policy systems to respond to these (Wistow, 2022). Health inequalities are a key measure of the unequal and deeply embedded trajectory of places in a post-industrial society like England and require disruption to the social contract in order to enable policy and public management to be more effective.

As demonstration, we analyse a longitudinal (pre and post) place-based dataset of all English local authority administrative areas. This was designed around a social determinants of health (SDH) perspective (see, Marmot et al., 2020) and Public Health Outcomes Framework (Office for Health Improvement & Disparities, n.d.) for England alongside some additional health service and place-based contextual data. The dataset covers a wide-range of factors relating to social complexities and trajectories of place.

In terms of methodological innovation, we will use the COMPLEX-IT platform to explore place-based complex conjunctural causation around two significant health inequalities outcomes – life expectancy and healthy life expectancy, which represent systems outcomes at the level of local authorities (as places and policy systems). COMPLEX-IT is a case-based, multi-methods platform employing the tools of computational social science to facilitate exploring and analysing complex data and support applied social inquiry (Schimpf and Castellani 2022). COMPLEX-IT provides a bespoke suite of techniques: cluster analysis, machine learning, data visualisation, data forecasting, case-based scenario simulation and, case-based systems mapping.

Through the findings and analysis, we identify elements of health inequalities that are amenable to local complexity-informed public management policies and service delivery and those that are less tractable to interventions at the local level alone and, therefore, require intervention higher up the causal chain via national place-sensitive policy and interventions. Through the methodological exploration of the issue, we contribute to the second aim of the panel and use this to extend our theoretical exploration, which contributes to the first aim.

CLICK HERE for the PDF of our Presentation

CLICK HERE to explore our methods platform, COMPLEX-IT

Translate