Navigation – Plan du site

AccueilNuméros12-1Revue des livresEssais critiquesTwo Styles of Social Neuroeconomics

Revue des livres
Essais critiques

Two Styles of Social Neuroeconomics

Enrico Petracca
p. 95-113
Référence(s) :

Jens Harbecke and Carsten Herrmann-Pillath (eds), Social Neuroeconomics: Mechanistic Integration of the Neurosciences and the Social Sciences, London and New York: Routledge, 2021, 278 pages, 978-036727616-4

Texte intégral

Afficher l’image
Crédits : Routledge

1In a now classic in the philosophy of economics, Don Ross has distinguished two “styles” of neuroeconomics (Ross, 2008). One style, called “neurocellular economics”, is said to borrow models from economics to interpret neuroscientific data and is contrasted with “behavioral economics in the scanner”, said instead to use neuroscientific methods to trace the neural origins of behavior violating normative precepts. The new book Social Neuroeconomics edited by Jens Harbecke and Carsten Herrmann-Pillath, suggests that also in the field of social neuroeconomics there are (at least) two different styles. More precisely, the book aims to introduce a full-fledged alternative to the way social neuroeconomics has been pursued so far. The intent of this review essay is to reconstruct both the traditional and the newly introduced styles referring to them respectively as “internalist” and “externalist” social neuroeconomics. We will not limit our discussion to this distinction, though. We suggest that the book also presents, although implicitly, two further (sub)styles of externalist social neuroeconomics, one proposed by Herrmann-Pillath and the other by the same Ross with Wynn Stirling, which we name respectively “inflationary” and “deflationary” externalist social neuroeconomics. Figure 1 lays out the varieties of social neuroeconomics and serves as a scheme for our discussion.

Figure 1: The Varieties of Social Neuroeconomics

Figure 1: The Varieties of Social Neuroeconomics

1. The New Field of “Externalist” Social Neuroeconomics

The “Internalist” Received View

  • 1 The field is sometimes referred to as social cognitive neuroscience, a label that emphasizes the me (...)

2The term social neuroeconomics came into use shortly after the onset of neuroeconomics. In 2007, Ernst Fehr and Colin Camerer coined it to refer to the study of the neural origin of social behaviors such as altruism, fairness, and trust (Fehr and Camerer, 2007). The term seemed to smoothly follow the footsteps of “social neuroscience”, the subfield of neuroscience specialized in the study of social behavior (see Cacioppo et al., 2002).1 Similar to the way neuroeconomics had distinguished itself from neuroscience, social neuroeconomics presented itself as an alternative to social neuroscience for proceeding, sometimes critically, sometimes supportively, firmly within the framework of economics. But soon something happened to unsettle this seemingly linear disciplinary development. While social neuroscience has continued to be a flourishing field, made as it is of dedicated journals, scientific societies, and scholars who unhesitatingly call themselves social neuroscientists, social neuroeconomics seems to be today somehow lost, the label used sparingly (if ever) and with a whiff of embarrassment. An embarrassment that likely stems from the perceived redundancy of the term. If we want to get to the bottom of what Social Neuroeconomics tries to do, some awareness of the redundancy of Fehr and Camerer’s social neuroeconomics is essential.

3Traditional social neuroeconomics appears redundant in at least two ways. One is evidently conceptual. As economics is already thought of as a social science, the added attribute social may well read as a repetition: the term neuro-economics alone would conceptually suffice, and currently does suffice, to cover the neuroscientific study of both individual and social (i.e. interactive) behavior. Something that the term neuroscience cannot do, requiring each time a specification of its object of study. Another source of redundancy for social neuroeconomics stems from the method. The feeling one has reading Fehr and Camerer’s foundational article is that social neuroeconomics is neuroeconomics plain and simple applied to a selected (but arguably easily expandable) list of social behaviors. It claims no particular methodological status for itself, except the minimal differences related to using game-theoretic rather than decision-theoretic settings. Moreover, the feeling that the two disciplines are nearly the same is fed by social neuroeconomists’ finding that the neural circuitries of reward-evaluation involved in altruism, fairness, and trust are quite the same ones underlying reward-evaluation in non-social situations:

One emerging theme … is that social reward activates circuitry that overlaps, to a surprising degree, with circuitry that anticipates and represents other types of rewards. These studies reinforce the idea that social preferences for donating money, rejecting unfair offers, trusting others and punishing those who violate norms, are genuine expressions of preference. (Fehr and Camerer, 2007, 425)

4Or more directly: “social preferences are similar to preferences for one’s own rewards in terms of neural activation” (Fehr and Camerer, 2007, 423; see also Fehr and Krajbich, 2014). If social and non-social preferences turn out to be neurally “similar”, a separated field for studying the former may easily appear redundant.

  • 2 Authors of Social Neuroeconomics do not explicitly call Fehr and Camerer’s style this way but, as w (...)

5Preferences are so pivotal to Fehr and Camerer’s social neuroeconomics that they suffice to give this style the name “internalist”.2 Like regular neuroeconomics, internalist social neuroeconomics rejects the assumption that mental states are unobservable and interprets neural signals related to reward evaluations during social interactions in terms of social preferences. Such neural correlates of preferences are then considered necessary and sufficient for explaining behavior. All that matters to explain phenomena such as altruism, fairness, and trust is, in other words, internal to individuals’ brains in the form of social preferences: a neuro-micro-foundation of the entire idea of sociality that finds in individual brains the key to explain social outcomes. This explanatory structure sounds so familiar that we do not usually need to call it internalist; but it needs to be recognized as such when an alternative to it is presented.

Integrating Social Mechanisms

6Before seeing what the alternative consists of, it is important to see why Social Neuroeconomics keeps the label social neuroeconomics at all. Isn’t it problematic, possibly redundant as just said? Well, it is not, at least given the book’s goal. After some discussion (2), the contributors to this volume have found the label convenient for a new program whose goal is to situate neuroeconomics programmatically within the realm of social science. If neuroeconomics has done just fine with conceptually covering the neuroscientific study of behavior in both social and non-social settings, it has however demonstrated very little of a social science. It is the (huge) gap with the social sciences that Social Neuroeconomics seeks to fill. To do so, it resorts to the powerful methodological framework of mechanistic integration introduced by philosopher Carl Craver. Famously, Craver (2007) depicted mechanistic integration as a “mosaic unity” in which disciplines contribute as tiles to explain the mosaic of behavior, providing constraints and opportunities for other tiles to fit in. Or emphasizing stratification, mechanistic integration may be thought of as a cake where each layer-discipline provides a different tone to the symphony of tastes.

7The bright prospect about a decade ago was that mechanistic integration would have helped neuroeconomics embrace its own hybrid nature and escape reductionism (e.g. Craver and Alexandrova, 2008). Unfortunately, mechanistic integration has since been as much preached as not pursued. In Chapter 3, Saúl Pérez-González shows that this is not due so much to laziness on neuroeconomists’ part but to deep structural differences between disciplines. If the very notion of a mechanism differs between neuroscience and the social sciences, integration resembles less of the mechanical job of matching mosaic tiles and more of the creativity of a pastry chef who tries to attune hard-to-match flavors. Caterina Marchionni and Jack Vromen focus in Chapter 2 on the idea of integration rather than on mechanisms and distinguish three alternative ways in which integration can be pursued. Integration can in their view be “minimal”, “weak”, or “strong” depending on factors such as the degree of tolerance to interdisciplinary frictions or ontological assumptions about the feasibility of integration. But neuroeconomics can also benefit, Marchionni and Vromen say, from methodological triangulation, a principle that sees progress stemming not from disciplinary integration but from disciplinary independence: the idea is that an independent origin would make mutually confirming findings more reliable and robust. Notably, the reality of social preferences is what internalist social neuroeconomists try to triangulate with other disciplines (23).

8There is a passage in Marchionni and Vromen’s lucid analysis (14) that, although crucial, I could not find stated as much clearly elsewhere in the book: the more disciplines one tries to integrate, the greater the ambition—and one can easily go on, the difficulty. The worry that too many flavors make the cake go bad is likely why social neuroeconomics has been pursued with so few ingredients. If not necessarily in terms of adding more and more layers to the cake, all authors in Social Neuroeconomics are however fully conscious that integrating social mechanisms is an ambition that requires, no matter how complicated things may get, further conceptual elaboration.


9The outcome of such further conceptual elaboration is the idea of context, presented as a crucial contribution of Social Neuroeconomics. Although its most devoted theoretical advocate is Herrmann-Pillath, the best way to introduce context is probably through Carolyn Declerck’s limpid Chapter 5 on oxytocin research. Oxytocin is a neuropeptide that in early experiments was found to increase prosocial and cooperative behavior unconditionally: the more the oxytocin (easily administered through intranasal spray) the more people’s trust (Kosfeld et al., 2005). Subsequent experiments, many of which run by Declerck and colleagues, have returned a far more complex picture of how oxytocin works. For instance, administered in an assurance game (where cooperation is a payoff dominant but risk dominated strategy), oxytocin promoted cooperation in players who had received positive social cues about other players, but had a contrary effect on those who had not received the social cues (Declerck et al., 2010). More generally oxytocin has been shown to foster parochial responses, increasing in-group cooperation and out-group competition, the latter particularly in presence of threatening social cues (De Dreu, 2012). Findings like these suggest that “oxytocin can have both prosocial and anti-social effects, depending on context” and that its role in social behavior is therefore “metafunctional”, not related to prosociality or anti-sociality per se but to greater accuracy in assessing social cues (86). Including context into the picture translates into the following conditional finding: “oxytocin facilitates approach behaviors when the environment is perceived to be safe, and … it encourages withdrawal when the environment is perceived to be threatening” (ibid.).

10For the methodological proposal of Social Neuroeconomics, these studies on oxytocin are crucial. They point to the fact that “it is impossible to treat the hormone [oxytocin] as an initial causal agent” (5). Uneventful as it might appear, this finding has the potential to wipe out single-handedly a couple of decades of internalist and reductionist research in neuroeconomics and substitute it, as of Herrmann-Pillath’s proposal, with the idea of “context as a causal mechanism” (51-52) or as “causally relevant for determining [an agent’s] behaviour” (54). We will return below to how context and its causal role are understood by Herrmann-Pillath, but let us focus for the moment on one general point related to accepting context as a determinant of behavior: context blurs the distinction between individual and social behavior. If studies on oxytocin focus specifically on social behavior and for this reason fall squarely within the canonical definition of social neuroeconomics, context does not only affect social behavior. Any behavior, even the most conceivably individual or individualistic one, can be thought of as affected one way or another by context. The true point of contention is instead how much of a context is social, that is, whether and to what extent context can be conceptually disentangled from the social dimension. Are there contexts that are not social? If yes, how? Can choice, say, of clothes or cars, be contextual but not social? If answered positively, these questions might cast again some doubts on the choice of social neuroeconomics as a label. An issue might be using the attribute social as an imperfect stand-in for context. The label “contextual neuroeconomics” would certainly be more inclusive, as it would count in authors that do not think of contexts as necessarily social, but many authors in Social Neuroeconomics think that a context is quintessentially social. Another good reason, then, for keeping on with this label.


11What said so far (integration of social mechanisms, context as a causal determinant of behavior, etc.) does not render the new strand of social neuroeconomics externalist per se. Externalism comes instead as an explicit philosophical choice made by Herrmann-Pillath and Ross and Stirling in their respective chapters. Although not exactly the same way, they both embrace the externalist viewpoint in the philosophy of mind thus making Social Neuroeconomics philosophically rare and precious. This is not the first time these authors present their externalist stance in neuroeconomics (Ross, 2011; Herrmann-Pillath, 2012), yet the hard-to-digest nature of externalism and the scarce attention it has received in economics easily make one think she is facing a first time (see Petracca, 2020). As Ross and Stirling say apropos:

It is somewhat puzzling that, particularly in the current technological environment, externalism still seems to be regarded outside cognitive science as an exotic view. It has scarcely penetrated economics at all, though it is implicitly present in some methodological reflections of Vernon Smith …, and its potential relevance to economists is explored in detail by Ross … and Herrmann-Pillath. (181)

  • 3 There are forms of externalism not specifically concerning the extension of cognitive systems (for (...)

12Externalism means in a nutshell that mental content, variously called in philosophical jargon propositional attitudes or mental representations—which economists know in the form of preferences, beliefs, expectations, and the likes—is not (exclusively) realized in the brain or neural hardware.3 In the passage above, reference to the “current technological environment” suggests that we should nowadays be quite accustomed to the idea that some mental contents, for instance pieces of memory, are realized in the technological devices we always carry in our pockets (Clark, 2008). But neither bits and bytes nor their being close to us in our pockets are necessary conditions for mind extension. Extension can be realized through a piece of paper on which we write a crucial piece of information, or, more relevantly for the topic of the book under consideration, in social practices (Gallagher, 2013). There are objects in the environment that convey or even embody “shared” mental content, such as the Statue of Liberty, which expresses societal beliefs in freedom and is used in everyday discourse to convey such beliefs. At a certain point in the book (180), it is rightly acknowledged that there is a debate among pundits of externalism about whether propositional attitudes are real or fictional entities, but in Social Neuroeconomics there is no disagreement on this point. All authors seem to think that propositional attitudes do exist and can be conveyed by any object in the vast repertoire that goes from neurons to the Statue of Liberty (although some authors, as we shall see, are more selective than others).

13Although the framework of mechanistic integration does not necessarily imply externalism, it is particularly suited to it. To see how, a little imagination can be useful: consider rotating the view of mechanistic integration from a horizontal to a vertical plane, or equivalently abandon the cake metaphor of integration and adopt the bike rack metaphor. On this view, disciplines are no longer one on top of the other but side by side, in a continuum that goes from those taking prevalently care of what goes on within the body boundary (neuroscience, psychology) to others dealing with what goes on outside of it (social sciences). This is probably a good way to represent what Herrmann-Pillath means when he says that in his view neural and population levels of explanation are connected “directly” (51, emphasis in original). “Directly” highlights that internal and external mechanisms work continuously together across the body boundary as a perfectly integrated grand mechanism.

14It is important at this point to consider that the framework of mechanist integration calls for a constitutive kind of explanation, according to which explanation for a phenomenon is provided once a mechanism for the emergence of that phenomenon is identified. On this view, constitutive explanation can perfectly accommodate externalism, as in mechanistic philosophy constitutive components of mechanisms need not be necessarily located within the body boundary. Discussing in Chapter 6 the methodology of the studies on oxytocin, Harbecke makes the interesting point that it is not at all obvious to figure out how context, a determinant of behavior (partly) outside of the body, contributes to explanation. Especially confounding is the fact that discussions of context come mostly framed in causal jargon and therefore seem to point to causal explanations, which are incompatible with constitutive explanations (in that identifying a causal nexus falls short of accounting for the constitution of a phenomenon). As Harbecke says: “it is not fully transparent whether the studies [on oxytocin, such as Declerck et al., 2010] generate causal or constitutive explanations”. But he claims, in the end, that “there are reasons to believe that a constitutive, or mechanistic explanation, is the ultimate target of the research projects represented by these scientists” (115). It is true, Harbecke says, that

Causal relationships hold between components or parts of the mechanism. But the ultimately explanatory relationship is a constitutive one of the cognitive process of prosocial decision making by a mechanism that includes OT [oxytocin] levels in the blood flow. (113)

15These remarks clarify the more general point that when context is presented in the book “as a causal mechanism” (51-52), it should be understood as a constitutive mechanism, not as a single causal relation. Also called a “causal assemblage” (58), context subsumes in a grand constitutive mechanism the web of causal relations taking place in both directions of fit across the body boundary (see also Herrmann-Pillath, 2016).

2. “Inflationary” and “Deflationary” Externalist Social Neuroeconomics

16Readers will find close collaboration between the authors of this book, to an extent that is rare in collective volumes. Cross-reference to each other’s chapter is frequent, supporting the genuine goal of presenting externalist social neuroeconomics as a “rich, innovative and coherent … paradigm” (7). The two most theoretically ambitious chapters, Herrmann-Pillath’s (Chapter 4) and Ross and Stirling’s (Chapter 9), are no exception to this collaborative attitude. They present several common and shared points, among which externalism of course, but also the assumption that propositional attitudes are real. Moreover, common is adherence to “semiotic externalism”, a position introduced by Herrmann-Pillath (2012) which seems to be shared in its essential form by Ross and Stirling: it conveys the idea that symbols, such as the Statue of Liberty, mediate between the individual and social dimension as tools for coordination and self-coordination. Symbols, in other words, would be pivotal to making sense of others in social interaction and of oneself through intra-psychic dialogue. Moreover, symbols are understood by both in a specific pragmatist and performative way congruent with Peirce’s view of semiotics (see Paolucci, 2021). (Renne Pesonen in Chapter 5 shares the same approach). One goal of this review, however, is to show that behind undisputable similarities structural differences may lurk that justify the introduction of different reference categories. There are reasons, we argue, why one would prefer to call Herrmann-Pillath’s externalist style “inflationary” and Ross and Stirling’s “deflationary” instead.

The Inflationary Style

  • 4 In philosophy the attributes “inflationary” and “deflationary” are typically used in a different se (...)

17In Herrmann-Pillath’s chapter, one may sense all the strenuous and admirable effort to hold together different domains—the internal and the external—and different disciplines—from neuroscience to the social sciences—without missing one tiny bit of them along the way. They are all set to work, as Figure 4.2 shows well (55), in a grand and comprehensive mechanism that explains behavior. Inevitably, such an all-encompassing view entails the multiplication of explanatory entities, which is why we call the framework inflationary.4 For one, the framework considers an extended variety of entities as vehicles of propositional attitudes—neurons and social practices among others. As such, it adheres to an inclusive form of externalism which also makes extensive use of internalist constructs. In Herrmann-Pillath’s explanatory scheme, propositional attitudes inside agents (called “representationsIN”) work side by side with propositional attitudes external to agents (called “representationsEX”) (56-59). Both types of representation serve the purpose of explaining “actions” (differing from “behaviors” in that the latter can be described neutrally and in physical terms) that have significance at both the individual and societal levels. Eating an apple, for instance, is an act that may be related to representationsIN via memories of joyful afternoons and to representationsEX as apple-eating features as a symbol of health in TV commercials. In the form of representationsEX, symbols, cultures, and institutions stand side by side in rich and wide-ranging explanations of apple-eating (Boldyrev and Herrmann-Pillath, 2013; Herrmann-Pillath and Boldyrev, 2014). The following is a bite of Herrmann-Pillath’s grand mechanism, which shows how representationsIN and representationsEX are strictly related:

the meaning of representationsIN is the intrasomatic processes that establish mappings across various functional domains to enable and enact action: The meaning of an ‘apple’ is the whole range of potentially active sensorimotor processes that is stored in the neuronal system involving apples and the conjugated representationsEX.(56)

18Eventually, although acting in two opposite directions of fit, respectively “brain-to-world” and “world-to-brain”, representationsIN and representationsEX converge in establishing the “meaning” of actions, where meaning is presented as a “mediating concept” (56) between levels of representation. Doubling down on representations and adding a further mediating concept, Herrmann-Pillath challenges the panorama of externalisms which are usually frugal about representations (see below).

19Representations of different kinds are within the purview of different disciplines, at least before full integration is reached. This is another way in which Herrmann-Pillath’s framework is rich or inflationary: no discipline is left along the way. We can identify at least four core disciplines: neuroscience, psychology, economics, and sociology (as a standard-bearer for the social sciences). The inclusion of psychology in the pool of disciplines is sufficient to make, as we shall see, a categorical difference. Its role in neuroeconomics is much debated, if not controversial: psychology is often considered propaedeutic to establish correspondences between neural circuitries and behaviors (Padoa-Schioppa, 2008) but dispensable after that (Padoa-Schioppa and Schoenbaum, 2015). In Herrmann-Pillath’s scheme, instead, psychology is here to stay, as the ultimate ambition is to find a place within social neuroeconomics also for behavioral economics. This is part of a long-term project of Herrmann-Pillath’s that has been developed in other works of his and consists of two steps, one neurophilosophical and one conceptual. As for the neurophilosophical step, following evolutionary neuroscientist Gerald Edelman and Hayekian insights, Herrmann-Pillath advocates for a selectionist model of neural response which would provide alternative neural foundations to behavioral economics. In the selectionist model, a stimulus activates a parallel competition between neural pathways for identifying the most appropriate framing of the situation before the winning neural pathway leads to the most appropriate course of action (Herrmann-Pillath, 2021). A similar selectionist mechanism can be seen currently at work in ecological models of neural utility alternative to traditional economic ones (e.g. Cisek, 2012). The conceptual step concerns instead including a functionalist component in the framework of social neuroeconomics with the aim to explain the “why” of actions rather than merely the “how” (what mechanisms less ambitiously try to do) (61). Adding functionalism on top of mechanism interestingly permits Herrmann-Pillath to rethink the traditional idea of dual-selves in behavioral economics in terms of “dual-function” (Herrmann-Pillath, 2019), shifting the emphasis from an essentialist to a goal-directed view of neural mechanisms. But, again, the functionalist extension introduces a further conceptual layer in a framework that already admittedly deals with “messy relationships” (53).

The Deflationary Style

20We need a new category for Ross and Stirling’s style of social neuroeconomics once we realize that, contrary to Herrmann-Pillath’s firm reliance on psychology and behavioral economics, its goal is to take these disciplines out of the picture. In alternative to the quad-core framework outlined above, Ross and Stirling propose a “tridisciplinary mélange that includes sociology” (175) but takes out psychology. Those familiar with Ross’s previous works (Ross, 2005; 2014) will find this move in line with his characteristic program of de-psychologizing economics, based on the criticism

that many neuroeconomists implicitly try to transform economics into the psychology of individual valuation, choice and behaviour, thereby underweighting the discipline’s history and applications that are more consistent with the aims and objectives of a social science. (174-175)

21Ross and Stirling’s chapter is undeniably the strongest advocate that neuroeconomics is a social science—not also a social science. To the point that neuroeconomics is considered a branch of “neurosociology” (198). This indirectly leads Ross and Stirling to assume a deflationary stance that drastically reduces the number of explanatory entities. “Indirectly” means that understanding neuroeconomics as a social science comes first, and the deflation ensuing from borrowing (almost exclusively) social science explanatory tools is a consequence. Yet deflation is real, and as we are going to see it does not merely manifest itself in reducing the number of disciplines that social neuroeconomics seeks to integrate.

22Although able to capture their stance in general, the word “deflationary” (182) is used by Ross and Stirling to describe their approach to propositional attitudes in particular. They criticize internalists for understanding propositional attitudes merely as “private, internal beliefs and desires ‘written’ in ‘brain code’” (ibid.). More exactly, Ross and Stirling are not against the internal localization of propositional attitudes—representationsIN in Herrmann-Pillath’s parlance—but against understanding them along a brain-to-world direction of fit. Propositional attitudes are not something lurking in the brain that needs to be discovered, but something that from an outside direction helps reach intra-psychic coordination, when intra-psychic dialogue is used to structure our own thoughts, and social coordination, when they are used in social interaction as “social tools” (ibid., emphasis in original) to reach mutual understanding. In both scenarios, Ross and Stirling see propositional attitudes following a world-to-brain direction of fit. This is an innovative philosophical view that makes their position deflationary but not, to borrow another term from philosophy, eliminativist, for it “hardly renders PAs [propositional attitudes] of derivative or downgraded importance” (ibid.).

  • 5 The view of game theory as a “universal lexicon” for integrating the behavioral and social sciences (...)

23Ross and Stirling’s view is better understood when juxtaposed, as they explicitly do, with another view of social neuroeconomics recently proposed by Carlos Alós-Ferrer (2018). Both views converge in understanding neuroeconomics as a social discipline and in seeing game theory as the privileged way to pursue integration with the social sciences.5 Yet they differ in crucial regards, such as Alós-Ferrer’s idea that game theory should be employed to model “mindreading”, while Ross and Stirling think of it as best employed to model “mindshaping”. Mindreading is the cognitive ability to infer the true state of propositional attitudes in own and others’ minds (e.g., Nichols and Stich, 2003). As such, it presupposes internalism of propositional attitudes. When in the common interpretation of game theory players consider others’ payoffs as a way to infer their moves, they are fundamentally engaged in a process of mindreading. Mindshaping is a cognitive ability as well, but not of divinatory nature (see Zawidzki, 2013). It is about “trying to change the behaviour of others rather than concentrating on accurately modelling them for the mere sake of entertaining correct theory” (183). People mindshape each time they make strategic actions not for their intrinsic payoff value but for the indirect value of changing others’ behavior. Think, for instance, of gestures of good faith in negotiations. Ross and Stirling choose to focus on mindshaping for it is considered “the most intense form of sociality found in nature” but also for they argue, following Zawidzki (2013), that “mindreading is an ill-advised coordination strategy” (185).

  • 6 I thank Don Ross for making me aware of this point.

24It is at this point that Wynn Stirling, an electrical engineer and game theorist, provides his distinct contribution. His newly introduced approach to game theory known as “conditional game theory” (CGT) (Stirling, 2012) would be able, in the generalized and extended form presented in this chapter,6 to specifically model mindshaping. Both traditional and evolutionary game theory could not do so if mindshaping is understood in terms of “the dynamical propagation of preferences as conditioned on the strategic choices of individuals” (187). For one, evolutionary game theory focuses on strategies without postulating agents (if not as carriers of strategies), and for this reason is unable to model beliefs, which are necessarily agent-related. In Ross and Stirling’s chapter, preferences are considered “conditioned” for being influenced by other players’ preferences and for influencing them in turn, in a reciprocal conditional adjustment that goes on until mutual compatibility emerges. It is true that also in CGT players try to infer the true distribution of preferences, but

These should not be interpreted as attempts at mindreading; they are more accurately interpreted as models of a player wondering how her own preferences might turn out to strategically cohere, or not, with the distribution of other preferences in her society. (189)

25We will not speculate here on whether and to what extent CGT is able to accurately model mindshaping, nor do we go any further into discussing the prevalence of mindshaping over mindreading as a mode of socialization. We merely say that there is an incredible amount of food for thought in this chapter, and that while nourishing their thoughts readers would become acquainted with a frugal way of pursuing integration.

3. Open Questions about the Externalist Styles

26Although internally coherent and showcased in rich details, the two externalist styles above leave some open questions that we discuss in this section. Let us consider the deflationary style first. One problem inherent to deflation is that some disciplines may be underrepresented or even missing in the process. If missing was programmatic in the case of psychology, the underrepresentation of neuroscience might be less intentional. In Ross and Stirling’s chapter there is a detailed account of how propositional attitudes (understood in externalist terms) condition players’ reciprocal behavior and lead to equilibria, but not much of how the brain takes part in this process. But before going on, it is necessary to clear the ground from one possible misconception: Ross and Stirling do want to include neuroscience in the picture. This can be seen clearly in Ross (2008), where Ross criticizes “behavioral economics in the scanner” but embraces “neurocellular economics”. If Ross and Stirling had simply wanted to get rid of neuroscience, they could have adopted the view of Gul and Pesendorfer (2008), a radical externalist position that seeks to explain social behavior avoiding neuroscience completely. But Ross stands up to treating the brain as a black box: as he claimed elsewhere, “proposed as a general principle, this is a mere prejudice” (Harrison and Ross, 2010, 186). Having made that clear, readers still crave to see how neuroscience fits into deflationary externalist social neuroeconomics. Although only a few hints are given about this, the hints are nonetheless interesting. As mindshaping is supposed to be hardwired in the brain because of its ubiquity and facility, little hope is placed in the capacity of neuroimaging to disentangle it from mindreading, as both are hypothesized to use the same neural circuitry of learning. Therefore, “the recommended course of neuroscience research on mindshaping is not standard neuroeconomics based on neuroimaging data” (197). In its place, what might be called “social neurocellular economics” would pursue the arduous goal of neurally tracing the elusive world-to-brain direction of fit. One way to do so, they suggest, is to study the neural manifestations of socially induced stress; or, more interestingly, see how in interview-like situations cognitive effort is not only exerted by interviewers (as is assumed in mindreading) but also by interviewees, as mindshaping is assumed to be enacted by both roles.

27Now consider some open questions about the inflationary style where again the ground should first be cleared from one possible misconception. There is nothing judgmental in the word inflationary: it is intended as a neutral term that merely emphasizes the cumulative or additive approach to integration. If anything, it actually highlights the boldness of taking on all the frightening challenges that integration brings along. It comes only as a matter of fact that, following Marchionni and Vromen’s insight, the more challenges are faced the more questions are left open. And Herrmann-Pillath seems to be well aware of this. Consider his attempt to hold together not only the mechanistic and functionalist approaches as seen above, but also the dynamical one. As he says:

the potential of mechanistic philosophy and methodology for rendering the externalist position into an empirically productive research methodology of social neuroeconomics … cannot be taken for granted, as … the dynamical systems view is often seen as staying in tension with the mechanistic approach, in the same vein as classical functionalist positions. (52)

28Herrmann-Pillath is honest to alert the reader that integration of the dynamical approach is difficult, maybe the hardest thing to do in his framework if the dynamical systems view is to be fully taken on.

29Dynamical systems have had a distinctive history in social neuroeconomics. There has actually been an entire research program under the label of social neuroeconomics based on dynamical systems (e.g., Oullier, Kelso and Kirman, 2008). What is perhaps even more relevant for our discourse is that this is an explicitly externalist program, representing a rare case of externalist social neuroeconomics before Social Neuroeconomics. Social neuroeconomics based on dynamical systems not only maintains that brains work differently during social interaction (something that, recall, Fehr and Camerer seem to deny) but also that other brains constitute social interaction, this latter being a clear externalist claim. What does it mean? Experiments employing hyperscanning techniques (i.e. able to record multiple brains simultaneously) have shown that during social interaction brains synchronize their respective frequencies such that synchronized frequencies become neural markers of social coordination (e.g., Tognoli et al., 2007; Dumas et al., 2010). On this view, other brains become an essential part of the synchronized system and the very idea of synchronization would have no sense if it did not include them as constitutive components. This is a full-fledged case of externalist constitution. Although Herrmann-Pillath mentions this literature (49-50), specific connections with it are not drawn nor do they seem urgent. Herrmann-Pillath’s idea of dynamical systems seems to be less technically dependent on the mathematics of differential equations, phase transition, and attractors, and more concerned with the intuition that neural and cognitive phenomena come into being over time and through interaction. In many passages, this is made clear claiming that what the framework is about are processes, not outcomes (instead of representation and context, the framework is said to be about “representing” and “contextualization”, see 57; 62). But this less formal understanding of dynamical systems does not seem to fit either with those dynamical approaches that in the philosophy of mind are known under the name of enactivism (e.g. Di Paolo and De Jaegher, 2012; Gallagher et al., 2013). For one, although they both find some inspiration in American pragmatism, they do not share other pivotal intellectual references, enactivism being mainly based on phenomenology and ecological psychology. One open question, then, is how the inflationary view relates to the broader panorama of dynamical systems approaches in cognitive science and neuroscience.

30Speaking of dynamical systems, another open question concerns this time both the inflationary and deflationary styles. Rather than an open question, what we shall discuss here is a possible alternative, inspired by dynamical systems, to how the two styles conceive of propositional attitudes. As noted, they both assume the reality and explanatory centrality of propositional attitudes, although Herrmann-Pillath sees them working in both directions of fit and Ross and Stirling consider them only along the world-to-brain direction. As discussed above, such frugality makes Ross and Stirling’s position deflationary but, importantly, not eliminativist. Supporters of dynamical systems, however, usually advocate uncompromising eliminativism: propositional attitudes are taken “radically” out of the explanatory picture (Chemero, 2011). Ross and Stirling rightly mention in their chapter some strands of externalism that, contrary to their view, consider propositional attitudes as “fictitious constructs” (180). But if fictitiousness refers to the ontological status of propositional attitudes, externalist eliminativism rather considers them as explanatorily unnecessary. How, then, is dynamical systems social neuroeconomics eliminativist about propositional attitudes (see, Favela, 2014; Kiverstein and Miller, 2015)? Consider again the studies on brain synchronization in social interaction (e.g., Tognoli et al., 2007; Dumas et al., 2010). It is not at all obvious how synchronized neural signals would map onto propositional attitudes of individuals in these studies. If any such mapping were posited, it would rather seem a cumbersome assumption (Gallagher, 2017). Synchronized neural frequencies certainly correlate with behavioral variables of social coordination, but they are not cognitive representations of coordination in any obvious sense. In other words, the fact that people with synchronized neural frequencies coordinate does not at all justify reading those frequencies as manifestations, say, of individuals’ prosocial thoughts.

Concluding Remarks

31Before Social Neuroeconomics, social neuroeconomics had the appearance of a struggling, redundant, sometimes even intimately embarrassed field: it flaunted its social nature twice but in the end explanations were social not even once. Social Neuroeconomics raises the crucial question that either the field is fully social or it is not. Dealing with the social means, at the first level of approximation, seeing neuroeconomics as capable of integrating different disciplines via mechanisms and of structurally including the social sciences, in addition to or in place of other disciplines, in the disciplinary mélange. But as the book shows well, mechanistic integration is hardly per se the panacea: piling up disciplines does not save the day; if anything, it complicates things. What Social Neuroeconomics does differently from other attempts at mechanistic integration, and this is its winning move, is to embrace externalism. The “implicit reductionism” (174) of neuroeconomics is overcome not merely by counting the social sciences in but by understanding what takes place outside of the brain as a constitutive part of the explanatory system. As an implication of this, the contextual/social dimension becomes an unavoidable, constitutive part of any strand of neuroeconomics.

32Dealing with externalism, and particularly with extended cognitive systems, is complicated. The idea of an “externalist neuroeconomics” appears to many as an oxymoron, and as such with no real chance of dethroning folk psychology-inspired neuroscience (see also Petracca, 2020). To overcome this problem, social neuroeconomists could find inspiration from what has happened in social neuroscience, the cognate field, where spokespersons for both internalism and externalism have gathered around a table, given reciprocal legitimacy, and laid out frankly their deep differences, even finding residual space for agreement (De Jaegher, Di Paolo, and Adolphs, 2016). Comparing the styles of social neuroeconomics closely one against the other and in their varieties, as we have tried to do in this review, can help better appreciate why the research program of Social Neuroeconomics is a possible game-changer for the entire neuroeconomics enterprise.

Haut de page


Alós-Ferrer, Carlos. 2018. A Review Essay on “Social Neuroscience”: Can Research on the Social Brain and Economics Inform Each Other? Journal of Economic Literature, 56(1): 234-264.

Boldyrev, Ivan A., and Carsten Herrmann-Pillath. 2013. Hegel’s “Objective Spirit”, Extended Mind, and the Institutional Nature of Economic Action. Mind and Society, 12(2): 177-202.

Cacioppo, John T., Gary G. Berntson, Ralph Adolphs, C. Sue Carter, Martha K. McClintock, Michael J. Meaney, ... and Shelley E. Taylor (eds). 2002. Foundations in Social Neuroscience. Cambridge, MA: MIT Press.

Chemero, Anthony. 2011. Radical Embodied Cognitive Science. Cambridge, MA: MIT Press.

Cisek, Paul. 2012. Making Decisions through a Distributed Consensus. Current Opinion in Neurobiology, 22(6): 927-936.

Clark, Andy. 2008. Supersizing the Mind: Embodiment, Action, and Cognitive Extension. New York: Oxford University Press.

Craver, Carl F. 2007. Explaining the Brain: Mechanisms and the Mosaic Unity of Neuroscience. New York: Oxford University Press.

Craver, Carl F., and Anna Alexandrova. 2008. No Revolution Necessary: Neural Mechanisms for Economics. Economics and Philosophy, 24(3): 381-406.

De Dreu, Carsten K. W. 2012. Oxytocin Modulates Cooperation within and Competition between Groups: an Integrative Review and Research Agenda. Hormones and Behavior, 61(3): 419-428.

Di Paolo, Ezequiel A., and Hanne De Jaegher. 2012. The Interactive Brain Hypothesis. Frontiers in Human Neuroscience, 6: 163.

De Jaegher, Hanne, Ezequiel Di Paolo, and Ralph Adolphs. 2016. What does the Interactive Brain Hypothesis Mean for Social Neuroscience? A Dialogue. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1693): 20150379.

Declerck, Carolyn H., Christophe Boone, and Toko Kiyonari. 2010. Oxytocin and Cooperation under Conditions of Uncertainty: the Modulating Role of Incentives and Social Information. Hormones and Behavior, 57(3): 368-374.

Dumas, Guillaume, Jacqueline Nadel, Robert Soussignan, Jacques Martinerie, and Line Garnero. 2010. Inter-brain Synchronization during Social Interaction. PloS One, 5(8): e12166.

Favela, Luis H. 2014. Radical Embodied Cognitive Neuroscience: Addressing “Grand Challenges” of the Mind Sciences. Frontiers in Human Neuroscience, 8: 796.

Fehr, Ernst, and Colin F. Camerer. 2007. Social Neuroeconomics: the Neural Circuitry of Social Preferences. Trends in Cognitive Sciences, 11(10): 419-427.

Fehr, Ernst, and Ian Krajbich. 2014. Social Preferences and the Brain. In Paul Glimcher and Ernst Fehr (eds), Neuroeconomics: Decision-Making and the Brain, Second Edition, 193-218. San Diego: Academic Press.

Gallagher, Shaun. 2013. The Socially Extended Mind. Cognitive Systems Research, 25: 4-12.

Gallagher, Shaun. 2017. Enactivist Interventions: Rethinking the Mind. New York: Oxford University Press.

Gallagher, Shaun, Daniel Hutto, Jan Slaby, and Jonathan Cole. 2013. The Brain as Part of an Enactive System, Behavioral and Brain Sciences, 36(4): 421-422.

Gintis, Herbert. 2007. A framework for the unification of the behavioral sciences. Behavioral and Brain Sciences, 30(1): 1-16.

Gul, Faruk and Wolfgang Pesendorfer. 2008. The Case for Mindless Economics. In Andrew Caplin and Andrew Schotter (eds), Foundations of Positive and Normative Economics, 3-42, New York: Oxford University Press.

Harrison, Glenn, and Don Ross. 2010. The Methodologies of Neuroeconomics. Journal of Economic Methodology, 17(2): 185-196.

Herrmann-Pillath, Carsten. 2012. Towards an Externalist Neuroeconomics: Dual Selves, Signs, and Choice. Journal of Neuroscience, Psychology, and Economics, 5(1): 38-61.

Herrmann-Pillath, Carsten. 2016. Constitutive Explanations in Neuroeconomics: Principles and a Case Study on Money. Journal of Economic Methodology, 23(4): 374-395.

Herrmann-Pillath, Carsten. 2019. From Dual Systems to Dual Function: Rethinking Methodological Foundations of Behavioural Economics. Economics and Philosophy, 35(3): 403-422.

Herrmann-Pillath, Carsten. 2021. Evolutionary Mechanisms of Choice: Hayekian Perspectives on Neurophilosophical Foundations of Neuroeconomics. Economics and Philosophy, 37(2): 284-303.

Herrmann-Pillath, Carsten, and Ivan Boldyrev. 2014. Hegel, Institutions and Economics: Performing the Social. New York: Routledge.

Kiverstein, Julian, and Mark Miller. 2015. The Embodied Brain: Towards a Radical Embodied Cognitive Neuroscience. Frontiers in Human Neuroscience, 9: 237.

Kosfeld, Michael, Markus Heinrichs, Paul J. Zak, Urs Fischbacher, and Ernst Fehr. 2005. Oxytocin Increases Trust in Humans. Nature, 435(7042): 673-676.

Lau, Joe, and Max Deutsch. 2019. Externalism About Mental Content. In Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy.

Nichols, Shaun, and Stephen P. Stich. 2003. Mindreading: An Integrated Account of Pretence, Self-Awareness, and Understanding Other Minds. New York: Oxford University Press.

Oullier, Olivier, J.A. Scott Kelso, and Alan P. Kirman. 2008. Social Neuroeconomics: A Dynamical Systems Perspective. Revue d’Économie Politique, 118(1) : 51-62.

Padoa-Schioppa, Camillo. 2008. The Syllogism of Neuro-economics. Economics and Philosophy, 24(3): 449-457.

Padoa-Schioppa, Camillo, and Geoffrey Schoenbaum. 2015. Dialogue on Economic Choice, Learning Theory, and Neuronal Representations. Current Opinion in Behavioral Sciences, 5: 16-23.

Paolucci, Claudio. 2021. Cognitive Semiotics. Minds, Meaning and Cognition. Dordrecht: Springer.

Petracca, Enrico. 2020. Neuroeconomics beyond the Brain: Some Externalist Notions of Choice. Journal of Economic Methodology, 27(4): 275-291.

Ross, Don. 2005. Economic Theory and Cognitive Science: Microexplanation. Cambridge, MA: MIT Press.

Ross, Don. 2008. Two Styles of Neuroeconomics. Economics and Philosophy, 24(3): 473-483.

Ross, Don. 2011. Neuroeconomics and Economic Methodology. In John B. Davis and D. Wade Hands (eds), The Elgar Companion to Recent Economic Methodology, 61-93, Cheltenham: Edward Elgar.

Ross, Don. 2014. Philosophy of Economics. New York: Palgrave Macmillan.

Stirling, Wynn C. 2012. Theory of Conditional Games. New York: Cambridge University Press.

Tognoli, Emmanuelle, Julien Lagarde, Gonzalo C. DeGuzman, and J. A. Scott Kelso. 2007. The Phi Complex as a Neuromarker of Human Social Coordination. Proceedings of the National Academy of Sciences, 104(19): 8190-8195.

Zawidzki, Tadeusz W. 2013. Mindshaping: A New Framework for Understanding Human Social Cognition. Cambridge, MA: MIT Press.‬‬

Haut de page


1 The field is sometimes referred to as social cognitive neuroscience, a label that emphasizes the mediation of cognition between the social and neural dimensions. Dropping the attribute cognitive is sometimes done for simplicity or to emphasize that other dimensions, such as affectivity, mediate between the social and the neural.

2 Authors of Social Neuroeconomics do not explicitly call Fehr and Camerer’s style this way but, as we shall see, the term is implied every step of the way

3 There are forms of externalism not specifically concerning the extension of cognitive systems (for an overview, see Lau and Deutsch, 2019). As they are not directly relevant to our discussion, we will not consider them here.

4 In philosophy the attributes “inflationary” and “deflationary” are typically used in a different sense, which has to do with the truth of a statement or a concept. We resemanticize these terms in the context of disciplinary integration, although the philosophical and reassigned meanings may occasionally converge throughout the discussion.

5 The view of game theory as a “universal lexicon” for integrating the behavioral and social sciences can also be found in Gintis (2007). Thanks to Ivan Boldyrev for pointing this out to me.

6 I thank Don Ross for making me aware of this point.

Haut de page

Table des illustrations

Titre Figure 1: The Varieties of Social Neuroeconomics
Fichier image/png, 10,0k
Haut de page

Pour citer cet article

Référence papier

Enrico Petracca, « Two Styles of Social Neuroeconomics »Œconomia, 12-1 | 2022, 95-113.

Référence électronique

Enrico Petracca, « Two Styles of Social Neuroeconomics »Œconomia [En ligne], 12-1 | 2022, mis en ligne le 01 mars 2022, consulté le 09 août 2022. URL : ; DOI :

Haut de page

Droits d’auteur


Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0

Haut de page
  • Logo Association Œconomia
  • DOAJ - Directory of Open Access Journals
  • Revue soutenue par l’Institut des sciences humaines et sociales du CNRS
    CNRS - Institut national des sciences humaines et sociales
  • OpenEdition Journals
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search