Can Complexity add anything to Critical Realism and the Morphogenetic Approach?
Margaret S. Archer J Theory Soc Behav. 2024;54:422–433 A Response – David Byrne
Again, I begin by presenting the author’s own summary of her argument:
Complexity is not ‘the same as simply complicated’ (Urry, 2005). This is because its advocates present it as a theoretical approach to explaining major aspects of the social order, usually at the macro level, whereas many social phenomena, at any level, can be full of complications (such as the incidence of road accidents) without a unifying theoretical key. Thus, the latter have a strong tendency to remain at the level of events and their study to be confined to a ‘variables’ approach, statistically combining the most variable of potentially contributory factors, without being troubled by the absence, in particular cases of one or more common contributors to accidents (such as drivers’ alcohol consumption). But ‘the Complexity Turn’ does much more than leaving empiricism behind, like Critical Realism from its earliest beginnings, and in some hands is seen as the senior partner of these two approaches. …This supposed superiority hinges upon the defining features of the Complexity approach itself, one that emphasizes instability, sudden change, far‐from‐equilibrium or non‐equilibrating systems, temporal irreversibility, and sensitivity to initial conditions typical of complex systems (Mingers, 2011). However, as with most of the so‐called ‘turns’ in the social sciences, these are features of another science or discipline which it is hoped can be adapted and imported into analysis of the social. Sometimes this is prompted by successful atural scientists themselves (such hands is seen as the senior partner of these two approaches.
A good place to start in responding to this piece is by looking at what Archer means by “complex system”. She defines this in a quote thus:
This supposed superiority (of complex realism to critical realism) hinges upon the defining features of the Complexity approach itself, one that emphasizes instability, sudden change, far‐from‐equilibrium or non‐equilibrating systems, temporal irreversibility, and sensitivity to initial conditions typical of complex systems. (Mingers 2011).
Most of this statement is wrong.
- Complex systems are not unstable – they often – indeed usually – endure for long periods of time in a much the same if not absolutely identical state.
- Changes in complex systems are not usually sudden, if by sudden is meant happen almost instantaneously. Systems generally pass through a time period of crisis – a word which never appears in the Church of Critical Realism’s accounts – which whilst much shorter than longue durée periods of relatively stability. Change when it happens is transformative in relation to the state of the system
- Temporal irreversibility does not really apply in relation to changes which do not challenge the overall intrinsic character of complex systems. The term resilience describes the ability of complex systems to “bounce back” to their predominant state from changes which do not induce a crisis condition. Transformative change is time irreversible in the sense that status quo ante is not restored.
- Complex systems are not sensitive to initial conditions. That is a characteristic of chaotic system measurement. Whether any systems beyond turbulent flow of liquids are chaotic is questionable. Sensitivity to initial conditions understood in terms of a butterfly’s wing flap initiating hurricanes is rubbish. However, there are systems, notably weather systems, where the ability to mathematically model as a basis for prediction is sensitive to very fine measurement of forces at that level and tiny differences can produce very different system states in the predictions made by the model. This is about modelling and measurement where the mathematics of chaos come into play, not about real systems in common sense of real equivalent to actual systems in Bhaskerian ontology.
To describe systems as far‐from‐equilibrium or non‐equilibrating systems is merely another way of saying they are complex. This is the one accurate point made in the passage Archer relies on as the basis of her critique.
Archer starts wrong and just keeps on going wrong. Her denouncement of the use of concepts and ways of thinking is to put it bluntly cheek from someone whose whole project was built around the importation of a metaphor – morphogenesis – from embryology. Certainly, I was first made aware of the revival – note revival – of an ontology founded around the complexity of the world by attention to this in the physical sciences represented in large part by the need to engage with particular mathematical systems, notably non-linear algebra, as a way of describing empirically observed complexity in action. Likewise, Paul Cilliers the electrical engineer was made aware of complexity by his engagement with neural net based computing systems and this led him into engaging with complexity as Paul Cilliers, the philosopher. However, that re-emergence was also in biology, not so much in terms of the work of Maturana which to my mind has led to a misleading view of complexity in the work of Luhmann and his disciples, as in von Bertalanffy’s General Systems Theory and in the more philosophical reflections of Gregory Bateson in Steps towards an Ecology of Mind. I had read widely in Biology since adolescence. When I encountered Complexity thinking it took me right back to all of Bateson, von Bertalanffy, Darcy Thompson, and Darwin himself.
Now, having gone back into the history of complexity informed ways of thinking, I know that it emerged from the response of G.H. Lewes, Darwin’s earliest important reviewer, to Darwin’s way of thinking and working. Lewes magisterial Problems of Life and Mind (1870s) is where not only the concept of emergence was first formulated but also a discussion of causation in terms of processes, relations, recursivity and the reciprocal nature of cause and effect – effects are causes – which echoes and probably was influenced by Hegel’s logic, Lewes like Engels provides us with a better way of understanding causation and causal powers. It is notable that Prigogine’s philosopher collaborator, Stengers, has been profoundly influenced by Whitehead who she describes in her Thinking with Whitehead as the main philosopher who engaged with the implications of Darwin’s way of thinking and working. I have even seen this claim made for Lakatos which is plainly absurd but actually the first philosopher onto Darwin was Lewes and he got him exactly right. Reed and Harvey’s seminal article, in the very journal in which Archer published her piece, introduced me to idea of complex realism, as they put it a fusion of the scientific ontology of complexity with the social ontology of realism. They were right then and they continued to be right, not least in Harvey’s later article, again in the Journal for the Theory of Social Behavior ‘Agency and Community: A Critical Realist Paradigm’. (2002) Journal for the Theory of Social Behaviour 32:2. This is a magisterial engagement with Bhaskar which also addresses Simmel. It is one of the best things I have ever read and the failure of Archer and Yang (of whom a very little anon) to engage with it is bizarre.
Archer is snide about Prigogine’s engagement with social science. However, she fails to cite Open the Social Sciences (1996), the Gulbenkian Report on the Future of the Social Sciences by a commission including Prigogine and chaired by Wallerstein. Wallerstein’s interest were in the tradition of Braudel and engaged with world systems. His key focus was on the transformation of the world system – in effect the creation of a world system – by the development of the maritime European empires in the 15th and 16th centuries. His focus was on colonization, globalization and the transformation of all social systems in consequence of these systemic level changes. For me this is yet another clear indication of the failure of mere critical realism to engage with macro social change and even more of the dangerous failure of it to engage with the causal powers of natural systems in relation to social systems. It was not mere imperialism which informed Wallerstein’s analyses but imperialisms engaged with the development of European capitalism, itself in very considerable part a product of the natural system impact of the Black Death in (to quote Gill Callaghan) elective affinity with Protestantism. Plague had just as great a demographic impact on the Ottoman, Indian and Chinese populations but that did not engender capitalism. It is frankly absurd to accuse complex realism of being ahistorical.
Let us get rid once and for all of Archer’s contention that complexity only deals with “temporally” irreversible processes. I put “ around temporally because the underlying structural transformational dynamics of complex systems certainly move only in one direction – forwards – through time, but that does not mean at all that during their longue durées of stability systems might move around and revert to previous states within the torus attractor. Archer says:
The point of concern here is that Complexity Theory when applied in the social sciences is dealing with ‘temporally irreversible processes’. That is what history is held to be, but how can this statement be justified? Both Buckley and I dwell upon morphogenesis (change of a system’s state) and morphostasis (working to maintain or return a system to its preceding state). This was prominent from my first book, (Archer, 1979/2013), … Nothing whatsoever warrants the presumption that the groups promoting morphogenesis will win out over those fostering a return to the status quo ante as the basic diagram makes clear; historically this depends upon the outcomes of social interaction between T2 and T3 that maybe either morphogenetic or morphostatic (at any level). (page 425)
Parsons is much better here. His account of system stability is founded around negative feedback mechanisms (not a term he uses but that is what they are) which maintain system integrity in a broadly similar state but he actually acknowledges that macro social transformation is qualitatively different from the ‘bouncing back’ characteristic of resilient systems. Positive feedback is real. No complex system is ever static but there is a radical and necessary distinction to be made between movements which do not transform and those that do. Transformation is often about the transformation of quantity into quality. And note that for Archer only groups (of humans one must assume) have agentic power. She ignores the agentic power of the natural world itself. Archer never addresses macro-social transformation in her work. Complex realism can do this. Mere morphogenesis cannot although the metaphor as part of a complexity framed programme has its uses.
I am going to deal with the rest of Archer’s diatribe by referring to her own sub-headings. First, she asserts ‘the quintessential historicity of historical realism’. For much of this section she draws on Holland (dealt with above) and in particular on his critical account of my discussion of the complex processes of causation of Tuberculosis in urban industrial societies as a whole and in my own body. I emphasize the importance of interaction, of complex and multiple causation, in both discussions. Archer summarizes Holland and adds her own take thus:
To put it bluntly, what is interacting with what and how?’ To this we can add equally bluntly, when specification is given, which of the interactants came before and which after the other. (page 427)
My discussion of TB was based around an important empirical study on Tyneside of the causes of TB by Bradbury. As he pointed out TB has one necessary cause – exposure to the TB bacillus, but many – indeed pretty well everybody – on 1930s industrial Tyneside were exposed but only some manifested a clinical disease (both my parents and stepfather all lost a sibling to TB). The prevalence of TB in industrial Tyneside was precisely a consequence of the development of a capitalist industrial society and its urban forms. I wrote about that at some length in Understanding the Urban (2001).What Bradbury’s study showed was in effect a statistical interaction in the data describing how TB manifested in individual human bodies in relation to the role of diet, housing conditions and ethnic background (being Irish for Bradbury but the Yemeni seamen in South Shields were an even more extreme case) and led to the development or absence of the clinical disease in an individual person. Industrial Tyneside was the structurally dynamic context in which TB happened. The elements surrounding diet, housing and ethnic background interacted in the individual body. The Yemenis came from a rural background and had no prior exposure at all to TB so no immune preparation and these superbly fit men (firemen trimmers – stokers to a chocolate sailor) died like flies. The Irish has one or two generations less of familial exposure and thereby less natural selection for resistance to TB. I had all of good food, good housing and natural selection – my parents despite exposure had not contracted TB. All of these things interact with each other in generating the system state of the individual human body in relation to the development of the disease. There is very little sequence in play other than the initial but also repeated process of exposure to infection other than when the disease took hold in an individual body. TB was endemic – it was always around in the whole economic and social context. A transformed context transformed both the potential for initial exposure and the likely process of development given exposure. The way Holland and Archer understand causation as cause generates effect is evident in the passage quoted. Causation is a matter of process, effect is a dynamic system state, and cause and effect are constantly interacting with each other.
Let me turn to her discussion of the importance of culture: She is blunt:
what do I find wanting? Fundamentally, the acknowledgement that culture too has a variety of structures ranging from complementarity to contradiction, that at any time and place they are not all commonly shared (Archer, 1985; Archer & Elder‐Vass, 2012), and cannot be given the frequency of their incompatible propositions (e.g. creationism and evolution) which are real contradictions. Basically, this means that people must take sides and are encouraged to do so or are often penalized for the side they take or for attempted neutrality. In so doing they are also aligning themselves with either reproduction or transformation, but when this is supressed in theorizing, then it ceases to be a triadic analysis.
For Archer without:
‘culture’ and its distinctive properties it is impossible to account for what agents think they are doing which is key to generative mechanisms, and explanations quickly become simply materialistic. (page 428)
My indifference to according to the ritual of the triadic analysis is overwhelming but there is a great deal wrong with this statement. It is of course based largely on work on complexity written nearly thirty years ago, which work was introductory, and thereby did not develop all the themes which have emerged in subsequent discussion. So far as discussion of culture is concerned Gill Callaghan and I engage with this both through the work of Bourdieu and empirically in our Complexity theory and the Social Sciences – the state of the art. Archer might have with profit looked at least at the first edition of this published in 2014 but her overall coverage of what complexity has been doing in the social sciences is meagre. What this passage demonstrates is the almost (I am saying almost but might as well say total) lack of engagement of mainstream critical realism, particularly the morphogenic current, with Historical Materialism as a philosophical project as developed inter alia by Thompson and Williams in particular in historical empirical and theoretical work. Thompson in very marked contrast to his views on Althusser was sympathetic to Bourdieu’s ways of thinking, a regard which was reciprocated. Let me do something Archer never does and attempt to synthesize what is the current state of discussion of culture in relation to complexity. I have been very much influenced here by Raymond Williams, particularly his discussion of base and superstructure in Marxism, and by Maxwell’s (2012) A Realist Approach to Qualitative Research. Archer does not reference that text but as we shall see pays no real attention at all to the actual ways in which social research is done.
Maxwell is especially adamant in asserting that: CULTURE IS NOT NECESSARILY SHARED:
‘… from the realist perspective that I’ve presented here, the concept of culture cannot be restricted to a set of shared concepts, symbols and beliefs. A culture is a system ofindividuals’ [note the plural DSB] conceptual/meaningful structures (minds) found in a given social system, and is not intrinsically shared, but participated in; … although sharing is one possible form of participation, is not the only one. Culture cannot be represented by a model on the same scale as the individual, i.e., as a “shared” set of meanings or beliefs that could be held by a single individual, but requires a model on a higher level of complexity.’ (2012 28)
Like Williams Maxwell pays careful attention to the meaning of words. Shared can mean held in common or it can mean distributed among. We need to see: ‘ … individual variation not as a deviation but as a core subject of study.’ (Maxwell 2012 30) This approach is absolutely compatible with a complex realist multi intersecting level understanding of any social phenomenon and particularly so for class
‘While we should bear in mind, in opposition to a certain mechanistic view of action, that social agents construct social reality, both individually and collectively, we must take care not to forget, as interactionists and ethno-methodologists are wont to do, that they have not constructed the categories they implement in this construction.’ (Bourdieu (1996). Understanding. Theory, Culture & Society, 13(2), 29)
For me this resonates exactly with the historical materialist account of culture developed through actual empirical engagement. Although Archer cites her study of education systems she does so only through assertion without drawing at all on its empirical content to inform her arguments. In her chapter in a handbook on culture (2005) she makes no reference to any historical materialist approach to these issues. She seems unaware of the idea of hegemony and her treatment of Bourdieu is poorly informed and a parody of his arguments. Archer was not given to drawing on empirical illustration. She did not instantiate.
Now what complex realism has done which the Church of “Critical Realism” has not is inform a considerable progamme of empirical work in which the relationships among economico-political systems and culture as lived experiences (note the plural) have interwoven trajectories which inform social agency at both the individual and collective levels. Not only is this a matter of classes and social and social movements but it also relates to the agentic powers of institutions as these are enacted in the economico-political sphere – Bourdieus’s primary causal field. We can illustrate this by considering voting behaviour in the 2024 UK General Election. This was a radical break, a transformation of quantity into quality, with the historical pattern of class voting in Britain. The trajectory had developed during the Blair years after 1997 but it accelerated in terms both of the social group support for parties and the large increase in abstention by registered voters. Labour won a massive majority on the basis of 34% of votes cast in consequence of the nature of the first past the post electoral system. It is important to remember that the UK is unusual among high income countries with the majority of people identifying as working rather than middle class. We can and should examine the motivations of people for voting as they did and for not voting at all but structures matter as they have a role in the complex and multiple causation of those very motivations. Labour’s continuation of austerity by refusing to tax incomes from wealth properly – at least on the same basis as taxes on labour incomes – or wealth itself really at all, has already led to a massive collapse in support for the party. This was coupled with a shift towards what Piketty (2017) Capital in the 21st Century called “left Brahminism” in the sense of prioritizing identities over class situation and, bluntly, bleating about white privilege in a way which systematically downplayed class relations. To explore the background to this development we need empirical work informed by an understanding of complex interwoven causal forces manifest in interwoven system trajectories at all levels, not formulaic abstractions with no material reality. Instantiation is essential for the deployment of the complexity frame of reference.
Agency – Archer has a specific section on agency but begins her introduction to the topic when dealing with the historicity of realism thus:
my assertion [is] that there is a temporal separation between structure and agency and arguing that structure and agents are only analytically and not ontologically separate, meaning separable. (page 428)
She continues after her discussion of culture.
In the above two subsections, ontologically ‘structure’ and ‘agency are seen in Critical Realism as distinct strata of reality, possessing quite different properties and powers That entails examining the relationship between them. Bhaskar’s statement that ‘the causal power of social forms is mediated through agency’ (Bhaskar, 1979, p. 26). It seeks to ally ‘agency’ with ‘structure’ and ‘culture’ as bearers of emergent properties, which have temporal priority, relative autonomy and causal efficacy, in their cases, exercised towards people.
Without the inclusion of ‘culture’ and its distinctive properties it is impossible to account for what agents think they are doing which is key to generative mechanisms, and explanations quickly become simply materialistic. (page 428)
Here I go back to Dawe’s seminal article on “The Two Sociologies” and to Bourdieu on habitus. Dawe (1970) in a much cited paper argued that Sociology (and we might say related disciplines with Social in their name: Anthropology, History, Geography) could be divided into two schools of thought. There were those who interpret the word in terms of structures which create regularities and those who saw the social world as created by actions of people, whether as individuals or in some form of collectivity. Now although what people think matters Bourdieu’s account of habitus indicates that much social action, which is agentic in that it is real in its effects, is not based on conscious reflection but is more or less embodied and routinized. In a crucial posthumous publication (these can be valuable but are not always so) Bourdieu, P. (2003) ‘Participant Objectivation’ Journal of the Royal Anthropological Institute 9 2 moved to a position entirely compatible with a complexity framed historical materialism by agreeing that when there are profound changes in the dynamics of social structures – when there are great social transformations, then habitus is disrupted and can itself be transformed. What I call “the great if partial transformation” after Polyani (Byrne, D.S. Class After Industry 2019) in which societies deindustrialize whilst maintaining all the structures of market capitalism has been such a transformative rift with enormous consequences for political processes, relations and identities in post-industrial capitalism. Agency matters but to grasp how and why we need to go to the Hegelian Marx and to the historical materialist strand in Western Marxism. Social structures had causal powers far beyond how people think although that thinking can become the basis for transformative agency. As with Holland again we encounter a merely idealist framing in “critical” realism.
One of the great values of the complexity frame of reference, precisely in relation to the contemporary global polycrisis for which global heating and its consequence is a key control parameter is that it is a mode of thinking with particular relevance for the intersection of the natural and the social and for science in the sense in which science was defined by the Gulbenkian report as part of a programme for confronting this potential catastrophe. The absence of mere “critical” realism from work on the interface of the social and the ecological is both notable and desirable. What could this kind of uninstantiated maundering bring to the table?
I have decided not to waste time and energy reviewing Yang’s ‘Complexity theory for complexity reduction? – Revisiting the ontological and epistemological basis of complexity science with Critical Realism’ (J Theory Soc Behav. 2024;54:368–383) It is essentially a mashup of Holland and the points made by Archer and displays exactly the same lack of understanding of the dynamic nature of complex systems and importance of action as constitutive of reality although operating within constraints imposed by reality.
Let me conclude this attention to these pieces, which at least gave me something to do when it was too windy to cycle and too wet to garden. In reading them I have come to a very clear view on the limitations of “critical” realism. In summary this school, church might be a better word, has three fundamental faults / limitations:
- It offers no guidance for constitutive social action – the point however is to change the world and there is no engagement with transformation.
- It offers no real basis for a methodological programme in the social and interfacing ecological natural sciences. Yang says: ‘As a philosophy of science, CR does not prescribe a methodology’ (page 376) but then proceeds to outline one in terms of: ‘an understanding of causality in terms of the powers and mechanisms of particular things’. There is no sense here of process or relation. Archer does not mention the term in her piece. Holland in reviewing my arguments never develops any account of how a methodological programme relates to the actual conduct of social research other than to assert that there is as basis in the “critical” realist programme for reconciling contradictory findings from qualitative and quantitative investigation. This is stuck in a mode of verification which does not correspond to dynamic reality.
- In consequence of the “critical” realism’s lack of a coherent methodological programme as a basis for actual research, there is a distinct lack of real empirical work which is based on the approach. The domain in which it is asserted that “Critical Realism” can and has informed a programme of substantive politically relevant research is health but my reading of Scambler and Scambler’s useful article on: ‘Theorizing health inequalities: The untapped potential of dialectical critical realism’ (2015) Social Theory & Health Vol. 13, 3/4, 340–354 (which provides a clear account Bhaskar’s thinking), to my mind does not establish any real difference between his formulation and that which can be derived from the Historical Materialist tradition in Marxism itself. John Rex in Key problems of Sociological Theory (1962) identified the importance of understanding of process and relation as Engels took this position from Marx. I have no problem with the Scamblers’ Bhaskar but I don’t see what this adds to an action oriented historical materialism. However, Bhaskar cannot be held responsible for the church constructed in his name and his account of the layers of the real, actual and empirical, causally interconnect layers of reality as a whole, is helpful. Hence, Reed and Harvey’s development of complex realism as a synthesis. As with all syntheses it adds to the elements synthesized by moving beyond Bhasker’s assertion of the openness of social systems, to recognise that the systems which matter – social, natural and their intermingling in the socio-ecological, and not just open but also complex. There is some good empirical work with transformative potential done by people who identify as critical realists. Sayer’s, (one of the clearest exponents of the realist progamme), Why we can’t afford the rich (2016) is an excellent and well informed book, but realism does not appear in the index and it is not based around a realist account as such. There are a number of ethnographic studies (although it seems rather more accounts proposing the method for ethnography), primarily in health and organizational contexts, which deploy a critical realist framing. These generally refer to the value of setting accounts of social action in structural contexts and the importance of understanding causation as all of multiple, complex and contextual. All this is fine but what complex realism adds to the mix is an understanding of and emphasis on the importance of systems and their relations. So, for example, no account of social action in health systems in the 2020s should fail to take account of the structural inter-relationships among health systems, the nature of austerity in relation to fiscal systems, and the imperialist pressure of capital and capitalists for privatization of them. You will not develop an adequate explanation of what is going on in a UK hospital without attention to the implications of the Private Finance Initiative (PFI) and the consequence of the loss of 40% of UK hospital beds over the last twenty years. Equally the war waged by rich elites on the taxation of incomes from wealth since Thatcher and the utter refusal of governments to consider taxing stocks of wealth themselves weakens the resource base for funding adequate health care provision. When attention is paid to interwoven systems, then this becomes apparent – the very thing that contra Archer complex realism adds to mere realism.
I might have written these two pieces up as an article and submitted them to a journal but that process, speaking as a former editor of a journal and reviewer for many, takes time and contributes to the intellectual masturbatory tendencies of the contemporary academy. The book by Gill Callagham, Emma Uprichard and myself on Global Crises: Complexity based research and practice for social transformation has just been published by Policy Press and my main focus will be on developing scenarios as a basis for criticizing the meager efforts of public policy and planning in Scotland and the North East of England.
I will return to the memoir of a life in methods in the near futur.
David Byrne
Complexity to inform issues of methodology, taxation and politics in a context of crisis 
Leave a comment