Introduction

1The present paper started as a book project a few years ago. Whilst I received encouragement to develop and formalise my proposal the project got side tracked. But the theme was not forgotten, for a number of reasons. First, because the phenomenon that motivated the book project endures. More to the point, it is spreading. My specific concern is with the systematic and unquestioning tendency towards considering formal methods as superior, for somehow more scientific. That in itself need not yet be a problem, were it not ostracising researchers wishing to use and explore other (non formal) methods. Whilst the trend is particularly salient in economics (Lawson 1997; Fullbrook 2004), I am presently concerned with the spread to other social sciences [2]. In essence, I mean to include those disciplines that concern themselves with human behaviour, and I shall throughout the paper refer to these as social science disciplines. Second, questions of methodology, though central to good scientific inquiry, are too often relegated to ‘lofty’ philosophers of science. But when faced with a case of acute methodological ideology, at the expense of a more pluralist orientation, it is time to re-engage in methodological reflexivity.

2In sum, a central element of the project is to convey how the methodological tendency towards formalism has gained ground across the other social sciences. My sense is that the social scientific edifice is burning. A first concern is thus to notify the fire, second, to indicate that it is gaining ground (the phenomenon is not in fact limited to one or two disciplines, but active across the social sciences), third to convince social scientists that it should be extinguished, and fourth to encourage contesting social scientists concerned across disciplines to (re) act collectively.

3In 2007, I attended a peculiar seminar at Cambridge University. My attention was drawn not so much by the fact that the paper treated of theology, though, given my wide range of interests, that in itself could have been good reason for me to attend. Instead, it was the philosopher of sciences in me that noticed something in the title of the presentation, which seemed very odd and somewhat concerning. For the paper was titled “God, Providence and the Evolutionary Phenomenon of Cooperation”. It was presented by an eminent professor from a Faculty of Divinity. As I intuited correctly, given my relative familiarity with the abundance of game theory I had encountered in economics, the author used game theoretical outcomes, showing humans put together experimentally under certain conditions are inclined to cooperate. The researcher used such ‘evidence’ to sustain the idea that, as well as being frequently in competition and outright evil, mankind is also naturally good. She concluded that theological and Darwinian evolutionary accounts of humanity need not be taken as opposites. Attached to her presentation was an article, using game theory, published in Science titled “Five rules for the evolution of cooperation” (Nowak 2006) defending the notion that humans naturally cooperate.

4I noted in particular that, whilst formal models are by now common in economics, they still come as something of a surprise, as it did to me at the time, when used in discussions involving God. I have since joining the academic world grown increasingly aware of a tendency to adopt formal modelling that reaches well beyond economics. By formalism, I mean the practice of scrupulous adherence to prescribed or external forms (typically in mathematical or logical form). The following definitions are further offered:

5

Formalism, Noun
1 the doctrine that formal structure rather than content is what should be represented [ant: imitation]
2 (philosophy) the philosophical theory that formal (logical or mathematical) statements have no meaning but that its symbols (regarded as physical entities) exhibit a form that has useful applications
3 the practice of scrupulous adherence to prescribed or external forms [3]

6The concept is further fleshed out in the following terms by the Oxford Online Dictionary:

7

Definition of formalism, noun
1 [mass noun] excessive adherence to prescribed forms:
academic dryness and formalism

• the basing of ethics on the form of the moral law without regard to intention or consequences.
• concern or excessive concern with form and technique rather than content in artistic creation.
• (in the theatre) a symbolic and stylized manner of production.
• the treatment of mathematics as a manipulation of meaningless symbols.
  2 a description of something in formal mathematical or logical terms:
  there is a formalism which expresses the idea of superposition

Below, I give an indication of how formalism has spread across the social sciences. Besides the overt anti pluralist orientation of any potential methodological straightjacketing, my concern is with the appropriateness of the use of formalist methodology to study social scientific matter. In what follows, I briefly discuss what the limits of such methodology from a perspective that is shared by Critical Realists and the Cambridge School alike. I conclude noting that the spread of formalism is a pandemic phenomenon that spreads across the social sciences, and encourage social scientist to take seriously and re engage with methodological reflexivity in an effort to preserve an intellectual climate in the academy liable to foster the advancement of knowledge.

The spread of formalism

8The story of the spread of formalism is by no means a straightforward one. One of the ways in which formal methods gained ground is by the venturing of economics into areas of human behaviour not traditionally broached in the discipline. But ground was also gained via non-economic routes. The first other discipline outside economics that caught my attention for its particular formal language was not theology, but philosophy, in particular that branch of philosophy referred to as analytic philosophy that had come to dominate Anglo Saxon research and teaching after the wartime period. The Penguin Dictionary of Philosophy offers the following comments under the entry of analytic philosophy:

9

The search for conceptual clarity led to a painstaking attention to detail, in contrast to the broad imaginative sweeps of grander theory. The objects of analysis were said to be concepts or propositions, but by the 1930s a linguistic turn came to be regarded as the fundamental object of analysis, and analytic philosophy was often called linguistic… Critics accuse the analytical movement of aridity and irrelevance; sympathizers see merit in its respect for rationality and its suspicion of rhetorical posturing and false profundity.

10We can contrast analytic philosophy with continental philosophy, as the latter does not insist on the use of formal methods. The spread of formalism is sufficiently important to have given birth back in 1992 to a journal entitled Mathematical Social Science. The journal describes itself on its home webpage, in the following terms:

11

The international, interdisciplinary journal Mathematical Social Sciences publishes original research articles, survey papers, short notes and book reviews. The journal emphasizes the unity of mathematical modelling in economics, psychology, political sciences, sociology and other social sciences. Topics of particular interest include the logic of individual knowledge, choice, and preference, decision under risk and uncertainty, collective choice and representation, voting and implementation, justice and inequality, theories of measurement for the social sciences, planning and organization, social networks, game theory, evolutionary theory and more generally the study of the equilibrium and disequilibrium of socioeconomic systems. Papers published by the journal are mathematically rigorous [4].

12The objective here is not to give an exhaustive account of the inroads made by formalism into the social sciences, a project that requires rather more space. It is nonetheless worth reporting further on the work of Gellert und Jablonka, specialised in education and philosophy of science respectively, as they note specifically how mathematics and formalism have come to dominate the social sciences. So much so that they edited a book titled: Mathematisation and Demathematisaton: Philosophical, Sociological and Educational Ramifications. They write:

13

Within science, there is hardly any theory, which is not formulated in mathematical terms. In sociology, psychology and education, quantitative studies are highly valued. It is hardly impossible to understand any theory of economics without a solid mathematical background. In all these fields of human activity mathematics can be regarded as the grammar of the particular scientific discourse and as a universal tool. However, mathematics being the grammar or the tool implies that the characteristics of this grammar strongly influence the development of the fields in which the use of mathematics is made. It turns out to be difficult, if not impossible, to integrate any idea that cannot be formulated in mathematical terms into an accepted body of mathematically formulated theories.

14We thus note that formalist turn is not a recent phenomenon, as specific instances of formalism can be found in political sciences, some dating as far back as the mid eighteenth century (Martin 1999), whilst we saw that analytic philosophy introduced and adopted formal logic in the Anglo Saxon academic world during the first part of the XX^th century.

The influence of economics

15

Contemporary academic mainstream economics is indeed often underpinned by ideology. But this ideology is first of all methodological in nature, being in effect the widespread cultural view that mathematics is essential to science. Incidentally I argue elsewhere not only that this ideology covers a false view in that successful natural science does not actually rest on the application of mathematics, but also that a nonmathematical economics can actually yet be a science in the sense of the successful natural sciences (see Lawson 1997, 2003 and especially 2012).

16In this passage, Lawson makes two important points: formal modelling of human behaviour constitutes 1) a methodological tendency observed in much of (mainstream) economics, 2) a form of ideological dogmatism.

17The latter observation is entirely consistent with critical realist thinking originally developed by Bhaskar, and later by Lawson, and others identified with the Cambridge School. For formalist methodology carries the ontological presupposition of a world in which events operate by means of constant mechanisms that work in isolation, and can be concretely and/or abstractly reproduced. To the extent that the economy is inherently social, that is, it depends at least in part on (free) human practice in being (re) produced by such practice, economic science deals with an open world that is by and large neither constant, nor isolatable, nor artificially reproducible. In such a context, non-formal methods appear preferable to those social scientific scholars who share an ontology of an open social realm (a point to which we return below).

18Traditionally, economics measures and studies things like national income, wages, inflation, and so on. Over the last few decades, however, economic scholars have sought to measure and formalise more and more aspects of human behaviour, including the motivations behind savings, but also motivations behind suicide, addiction or marriage. This trend is a form of economic imperialism over, or colonisation of, subjects traditionally left to other disciplines. I prefer here to use the term abduction, in the sense of kidnapping, as defined by the Oxford Online Dictionary [5]:

19

Definition of abduction Noun [mass noun]
1 the action of forcibly taking someone away against their will: they organized the abduction of Mr Cordes on his way to the airport [count noun]: abductions by armed men in plain clothes
(in legal use) the illegal removal of a child from its parents or guardians: the man is also accused of the attempted abduction of another youngster
2 Physiology the movement of a limb or other part away from the midline of the body, or from another part. The opposite of adduction (see adduct1).

20The illegal removal from the parents well captures these developments in economics. We find for example economic research programmes using theories coming out of neuro scientific scholarship (Martins 2011). There is a sense in which such findings that are fathered (are mothered) by neuroscientists, and kidnapped to be taken into the unfamiliar environment of economic research and theory, to be there transformed to often quasi unrecognisable form.

21The process of abduction tells of economics branching out to study non economic subjects using formal methodology. There is an additional way in which economics gains influence, and this is where and when the typical methodological practices are copied and imported into other disciplines by political scientists, psychologist, anthropologist, historians, sociologists, and so on. The process is tied to a commonly held belief that formalism brings about neutrality, and objectivity, where these are associated with good scientific practice (Drechsler 2011; Hudson 2010). What is more, there is an implicit association being made by social scientists, both in and outside of economics, that precise quantitative measures create a form of exactness, factuality, and an ability to predict, and that it cannot be achieved without the use of numbers. Such an assumption of course calls for an examination of the assumed relation between measurement, exactness (or precision), fact and prediction (Morgan 2012).

22Economics is the discipline that is perhaps best known for its emphasis on formalism. Yet, the discipline is in crisis. As Davis comments, “the most sophisticated economic models of the world economy—premised on a mechanistic vision of markets—failed to predict, and still cannot explain, the recent crisis and turmoil in the world economy” (Davis 2010, p. 237). To the extent that other social sciences (increasingly) adhere to formalism, the current crisis in economics should serve to caution those in other disciplines wishing to emulate the discipline.

23Note that, whilst economics is hailed, by economists and non-economists alike, as the most scientific of the social sciences, other disciplines did not systematically take their cue from economic scholarship to engage in formalism. The formalist methodological turn was also directly influenced by a general awe of mathematics and the natural sciences. Instead, economics continues to be a point of reference for its alleged ability to emulate the hard sciences, even where other disciplines have found their own path to formalism.

Methods and ontology

24It will be clear to the reader that, if I am concerned with the spread of formalism, both in and outside of economics, covering more and more areas of human behaviour, it is because I think that there is a problem with the particular methodological trend. Let me elaborate.

25Pluralism, including methodological pluralism, is part and parcel of the advancement of, both social and natural, science (for a full discussion of the merits and limits of methodological pluralism and interdisciplinarity, see Bigo and Negru 2008, 2011). The freedom to explore and choose one’s research tools without being ostracised is essential to good science. And though few academics would argue differently, we observe in practice that such freedom is not always everywhere granted, as adherence to methodology may and does in places determine the possibility to successfully climb the PPP—promotion, publication, prizes—ladder.

26Second, once we allow for a plurality of methods, we should be able to examine the merit of any particular method, a process that is just as much a condition of intellectual progress, as is the freedom to explore different tools in the first place. In the theological example taken above, possible lines of inquiry include the following questions: do the insights of the researcher, and more generally, of scholars resorting to formal modelling, (not) pre-exist the models used? What indeed does the introduction of models add to initial propositions? Do they further understanding of the phenomena being studied? Do they make the argument more convincing? If so, why?

27What constitutes good science, and the limit of the use of formal models in social scientific endeavour, is a vast topic of research. It is the position of the Cambridge School, one that resonates with the conception of good science advanced by critical realist Roy Bhaskar, that I present here (Groff 2008; Al-Amoudi and Willmott 2011). The Cambridge School is associated with an ontologically oriented project under the auspices of the Cambridge Social Ontology Group (or CSOG) that was instituted a few years ago, including in its membership scholars within reach of Cambridge (UK). The ontology it explores and seeks to refine and develop has roots in the original ontological contribution made by Bhaskar: critical realism. Critical realism is a meta-physical framework that, since Bhaskar’s initial work, has been understood and interpreted in varying ways. It has further led to the production of a range of substantive theories, so that what that body has come to stand for is neither uniform, nor contradiction free. In particular, critical realism is a metaphysical framework that does not prejudge the specific form substantive theory takes on.

28In his seminal text The Possibility of Naturalism (1979), Bhaskar, advances, amongst other things, that the purpose of science is to succeed in explaining the phenomena under study, where, incidentally, prediction is by no means essential to such a move (Bhaskar 1975). He writes:

29

theoretical explanation proceeds by description of significant features, retroduction to possible causes, elimination of alternatives and identification of the generative mechanism or causal structure at work.

30Clearly, the identification of possible causes requires the use of appropriate methods (Bhaskar 1975). One’s choice of method should, Bhaskar argues, in turn depend on the nature of the object under study, that is, on the object’s ontology. In the natural world events can (in some cases) be isolated and reproduced. The researcher is in such cases able to artificially close systems, in which mechanisms are isolatable, and outcomes predictable. If one accepts, by contrast, that the social world is open, where no such control and isolation can be (re) produced, then it follows that social matter does not lend itself well to formalism and prediction. The distinction between open and closed systems I am making here is one that is coherent with the Cambridge School (Bigo 2006), and with that body of thought referred to as critical realism.

31The association of science good with prediction has a long history. Lawson explores the relation between economics, on the one hand, and mathematics, formalism and positivism on the other. Below he is critical of prediction per se:

32

It is often suggested (for example by Descartes, Leibniz, Whewell, Duhem and Popper) that a theory which successfully anticipates certain events is to be preferred to one which is formulated to accommodate those same events after they have been reported.

33By contrast, it is suggested, that if formal models give the argument apparent weight, they do not necessarily add anything substantial to the insight gained without the model (Lawson 2012b; Willmott 1993, 1996, 2005; Contu and Willmott 2005; Curtis 2008; Bigo 2008).

34As noted, formalism and quantitative measures are associated with precision and neutrality. Yet there is nothing to say that 1) a numerical measurement is more exact than a non-numerical one, 2) the alleged exactness should constitute better science, and 3) predictions are part of the best scientific outcomes (Jackson 2009; Morgan 2012).

35Or worse, whilst formal models may carry an air of respectability, they possibly also obstruct rather than aid our analysis of social scientific phenomena. Echoing such critique, it is suggested that certain phenomena, and perhaps especially human behaviour, are best captured by the use of prose, rather than mathematical formulae (Willmott 2005). Some advance that words can limit our best understanding of reality, and that we can resort to other sorts of (non-linear and non verbal) methods and logics (Martin and McIntyre 1994; Johnson in this number).

36Critics of formal modelling believe that what is gained in elegance is lost in realisticness (true to life-ness) (Lawson 1997). Because untrue assumptions are made, false elements are introduced into the picture. When, for example, fixed tastes for economic agents are being assumed, tastes are imagined to be independent from agents’ changing personal circumstances, when no such independence exists (Horst 2000). Alternatively, for it is possible to focus on elements of the whole, without assuming things other than they are (in what is a process of abstraction). A medical doctor can for example focus on and analyse the mechanisms of the liver, without assuming away (isolating) the dynamic reality of the effects of the other organs on the liver.

37If the social scientists is an open system, formalism may not be the best way to uncover the dynamics of society, whether under its economic, psychological, historical, sociological, or philosophical aspects. Johnson in this number already suggests that aspects of reality that can be more fully apprehended through lived experience, as for example proposed in cybernetics through the use of technology and play.

38In sum, it is not because economics, or any other science, engages in formal modelling, that it is necessarily more scientific [6]. To quote Bruno Latour, “the illusion [is] that the essence of scientificity lies in the formal properties of their [the social sciences] representational systems, their theories” (Latour 1999, p. 12). The above discussion leads us consider that social scientists may be inspired to tread with caution, when considering the adoption of formalism in their quest for scientific progress.

Contesting voices

39As noted, unlike the economics discipline, most other social sciences are not dominated by formalism. There still mostly exists a form of methodological pluralism, a tolerance for a variety of methodological approaches to co exist, with noted the exception of the classical brand of analytic philosophy, now re-named “exact philosophy”, which is bound by a formal methodological straightjacket of its own.

40I have opted to dedicate this section to give ample space to the direct testimonials of voices contesting formalism. From within economics, we find a growing number of economists on the margins engaged in philosophical and methodological reflexivity, pointing to the limits of formalism in economics. The number of papers published in (heterodox) economic journals, such as the Cambridge Journal of Economics, Economic Thought, The World Economic Review to name a few, are a testimony to the growing discontent with formalism in economics (Lawson 2003; Mirowski 2004; Hudson 2010; Drechsler 2011; Bigo 2008; Bigo and Negru 2008, 2011; Martins 2011; Morgan 2012).

41I wish to give space to three voices in particular: three eloquent critiques, by non-economists, of the formalist turn. They discuss the place of formalism in economics, and question the appropriateness of the methodology in social science more generally. Crucially, all suggest a consideration of, when not a turn to, other methods. First, Robert Nadeau, professor at George Mason University, notes formalism in economics grew out of XIX ^th century physics, and comments on how the methodology is inappropriate to come to grips with both economic and environmental challenges in virtue of its “unscientific axiomatic assumptions”. He writes:

42

In [neoclassical]economic textbooks, the creators of this theory are credited with transforming the study of economics into a rigorously mathematical scientific discipline. But what is not widely known is that neoclassical economic theory was created by substituting economic constructs derived from classical economics for physical variables in the equations of a soon-to-be outmoded mid nineteenth century theory in physics. The mathematical formalism that resulted from these substitutions was predicated on unscientific axiomatic assumptions that remained essential unchanged in subsequent extensions and refinements of neoclassical economic theory. And this explains why the mathematical formalism used by contemporary practitioners of neoclassical economic theory effectively precludes the prospect of implementing scientifically viable economic solutions for a broad range of very menacing environmental problems.

43Second, Simon Duncan, Professor in Comparative Social Policy at Leeds University, and member of Care, Values and the Future of Welfare (CAVA), is especially critical of the phenomenon in the economic literature that I earlier identified as abduction. He writes:

44

In neo-classical economics it is assumed that individuals, as separate economic agents whose ‘preferences’ or ‘tastes’ are already given, make selfish cost-benefit type analyses in order to maximise their personal utility(…). While on occasion the ‘taste’ for ‘leisure’, ‘welfare’ or ‘pleasure’ more generally is included in the definition of utility, this is usually operationally defined in monetary terms (…). Using these assumptions, mathematical models based on differential calculus and the technique of constrained maximisation can symbolise rational actors working on the principle of marginal utility. In turn, this mathematics is seen to allow the natural-science like prediction of behaviour. Neo-classical economics is thus claimed as a universal theory potentially applicable to most of human life (…). While other social science disciplines (…) remain essentially peripheral to the explanation of social behaviour.

45Third, in a paper titled, Outgrowing physics envy: Reconceptualizing social research, David Nichols at the Department of Family and Child Sciences, Florida State University, makes a compelling case for social scientists to critically consider accepted notions of good science, and the kinds of methods adopted in its name:

46

There is a pervasive sense of unease among social scientists concerning the status of social research. This unease is rooted partly in a false dichotomy between objectivity and subjectivity and a belief that an idealized positivist version of classical physics should be the model for all sciences. Experimental methodology is one of many valid ways of obtaining knowledge and carries with its use a particular set of problems, particularly when social phenomena are studied.

47We can thus observe a resistance to the pandemic formalist turn in social scientific research. Though dispersed, scholars contesting formalism are found across the ‘soft’ scientific disciplines. The proportion of scholars not espousing formalism is of course difficult to assess. Nonetheless, I suggest that the project of identifying scholarship that does not (wish to) conform to formalism exclusively (methodologically heterodox research) to be a useful one.

48It is certainly worthwhile to establish more securely the weight of the contributions produced by the methodological heterodoxy. When economists, for example, exile themselves in history, politics or sociology departments, as well as in business schools, their very dispersion can give the (false) impression that, outside the economics as practised in economics departments, not much research is being done. Finding out where the heterodoxy practices and publishes serves to correct misperceptions of this kind. Such a project should thus aid dispel beliefs (held by formalists and non-formalists alike) that scholars not conforming are fewer than they really are.

49A key project then would be to more firmly establish that contesting voices exist, as exemplified by the passages above that demonstrate the concern with the ground formalism is gaining across the social sciences. The next step would be for scholars concerned to join efforts. My hope is indeed that ‘soft’ scientists may grow to contest the formalist turn, collectively, in favour a greater pluralism. My paper, in highlighting the (so far little noted and noticed) commonality of plight across the social scientific heterodoxy, seeks to alert those less aware, and further foster a common movement towards methodological reflexivity and emancipation.

Concluding comments

50The objective of this paper has been to share my concern with the (continued) rise of formalism in the social sciences. We saw economics has certainly played a (passive and active) role in inspiring other social scientists to follow their methodological orientation. My aim has further been to alert scholars across the social science at large of the phenomenon, and to take a critical stance, by re-engaging in methodological reflexivity. Indeed, when “practioners of a scientific discipline pay no heed to methodological questions, the discipline is destined to stagnation and possible degeneration” (Davidsen 2008, p. 55). I can but encourage researchers from different disciplines to recognise the dangers of methodological ideology, as constituting a form of dogmatism that is anti pluralist and anti intellectual. For the noted rise in formalism has in the cases of disciplines such as economics and some branches of philosophy, led to diminished intellectual freedom, given the limits imposed to explore and adopt different methodologies.

51In the end, I am not taking a stance against formalism per se, but against the privileging of formalism, as the most appropriate methodology to advance social scientific scholarship. Such privileging deserves to be at least open to question. We in fact saw that formalism may not be wholly appropriate, given the nature (the ontology) of social scientific matter. A most eloquent summary of the state of affairs comes from Austrian economist, Fritz Machlup (1902–1993). He describes the feeling of inferiority that plagues social scientists, and their lack of criticality towards the spread of formalism to the detriment of other methods in the following terms:

52

Good “scientific method” must not proscribe any technique of inquiry deemed useful by an honest and experienced scholar. The aggressiveness and restrictiveness of the various methodological beliefs which social scientists have developed — in subconscious attempts to compensate for their feelings of inferiority vis-à-vis the alleged “true scientist” — are deplorable. Attempts to establish a monopoly for one method, to use moral suasion and public defamation to exclude others, produce harmful restraints of research and analysis, seriously retarding their progress.

53Machlup’s comments, made over fifty years ago, are still highly relevant. As ‘heterodox workers’ are dispossessed of their methodological capital, the despair and frustration experienced by many social scientists, there is potentially fertile soil for halting a formalist pandemic, and bring about greater methodological emancipation. In particular, I humbly urge and encourage social scientific researchers to re-engage individually and collectively in methodological reflexivity in an effort to preserve intellectual integrity, and foster better research practices in the academy.