Consciousness and Cognition xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Consciousness and Cognition journal homepage: www.elsevier.com/locate/concog
Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’ Vivian Bohl ⇑ Institute of Philosophy II, Ruhr-Universität Bochum, Germany Department of Philosophy, Institute of Philosophy and Semiotics, University of Tartu, Estonia
a r t i c l e
i n f o
Article history: Received 31 December 2014 Revised 8 May 2015 Accepted 9 May 2015 Available online xxxx Keyword: Social cognition Mindreading Direct perception Theory-theory Simulation theory Hybrid theories
a b s t r a c t This commentary argues that Gallagher’s account of direct social perception has remained underdeveloped in several respects. Gallagher has not provided convincing evidence to support his claim that mindreading is rare in social situations. He and other direct perception theorists have not offered a substantive critique of standard theories of mindreading because they have attacked a much stronger claim about the putative unobservability of mental states than most theories of mindreading imply. To provide a genuine alternative to standard theories of mindreading, the direct perception theorist needs to provide more detailed answers to the following questions: What are the criteria for distinguishing perceptual processes from non-perceptual processes? How exactly does direct social perception function on the subpersonal level? What is the content of direct social perception? How does direct perception theory relate to more recent developments in the mindreading literature? Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction In ‘‘The new hybrids: Continuing debates on social cognition’’ Shaun Gallagher briefly outlines his version of the direct social perception (DSP) account as it forms a part of the interaction theory (IT) (see e.g. Gallagher, 2004; Gallagher & Zahavi, 2012). IT is a theory of human social cognition, proposed as an alternative to standard theories of mindreading: theory-theory (TT) and simulation theory (ST). IT states that mindreading via theorizing or simulation is unnecessary in most social situations. According to the thesis of DSP, we can directly perceive some mental states of other individuals, paradigmatically their intentions and (basic) emotions. In the context of IT, DSP is understood as an embodied and enactive process in the service of social interaction. Gallagher criticizes what he labels ‘‘the new hybrids,’’ characterized as recent attempts to integrate standard theories of mindreading with IT and, in particular, with the thesis of DSP. He argues that a better model for a hybrid approach would be a pluralist one. In this commentary, I would like to point out some issues in relation to which Gallagher’s IT in general, and DSP in particular, would benefit from being developed in more detail. I suggest that more empirical research is required to back up Gallagher’s claim that mindreading occurs rarely in social situations. I raise the point that DSP-theorists have not offered a substantive critique of theories of mindreading because they have attacked a much stronger claim about the putative unobservability of mental states than most theories of mindreading actually imply. In order to provide a genuine alternative
⇑ Address: Department of Philosophy, Institute of Philosophy and Semiotics, University of Tartu, Estonia. E-mail address: [email protected] http://dx.doi.org/10.1016/j.concog.2015.05.003 1053-8100/Ó 2015 Elsevier Inc. All rights reserved.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
2
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
to standard theories of mindreading, a much more detailed explanation of how DSP works on the subpersonal level would need to be worked out. Finally, more needs to be done to relate the DSP view to recent developments in the mindreading literature. 2. How ‘‘rarely’’ do we actually mindread? Let me start with some thoughts on Gallagher’s critical analysis of the account provided by Michael, Christensen, and Overgaard (2014). Michael et al. argue that it is plausible that in social contexts high-level processes (including mindreading) and low-level processes (such as embodied social responses highlighted by IT) inform each other. Gallagher partly agrees with this proposal, saying that ‘‘IT doesn’t deny that higher-order cognitive processing may be involved in many social encounter contexts’’ (Gallagher, current issue), but he strongly resists the idea that such higher-order processing may typically involve mindreading via theorizing or simulation.1 Instead he claims that mindreading (in the sense of TT or ST) is ‘‘rarely used’’ in social encounters. Gallagher and other proponents of IT correctly point out that the assumption that mindreading is ubiquitous in human social cognition, and necessary for human social interaction, is unwarranted. Mainstream research on mindreading has been based mainly on laboratory experiments, such as false belief tasks, which are specifically designed to elicit mindreading. But such experiments cannot possibly tell us how commonly people mindread in real life. The proponents of TT and ST have arguably simply assumed that mindreading is ubiquitous in human social cognition, which has prevented empirical research on how often people actually mindread in real-life social situations. However, it is equally problematic to prematurely buy into the opposite claim, namely that mindreading is rare and hardly plays any role in human social cognition. Unfortunately, this is exactly the claim that Gallagher and the other proponents of IT are making. The problem with any claim about the frequency of mindreading is that it currently remains as speculative as the claim that mindreading is pervasive in human social cognition: at the moment we have no decisive evidence on the matter.2 We simply do not know how often and for what purposes people actually engage in mindreading in real life. So far, Gallagher has not provided compelling empirical evidence to back up his theoretical claim that in most social situations, people typically rely on low-level processes (such as DSP) and higher cognitive processes except for mindreading (in the sense of TT and ST). In sum, his claim that mindreading is ‘‘rarely used’’ remains as speculative as the claim that it is ubiquitous in human social cognition. Ascertaining the frequency of mindreading may or may not be an interesting research aim in its own right but it is clearly an important step toward specifying the role of mindreading in human social cognition. If human social cognition can in most cases be explained by reference to social interaction, DSP and non-mentalistic higher-order cognition as Gallagher suggests, it remains mysterious why the ability to attribute full-blown propositional attitudes like beliefs and desires has evolved in the first place, and what its current function is. The upshot is that it is indeed unwarranted to assume a priori that mindreading is as ubiquitous in human social cognition as standard theories of mindreading have arguably assumed, but it is equally unjustified to suppose that mindreading is as peripheral to human social cognition as the proponents of IT tend to claim. On a more positive note, by suggesting a pluralistic approach in this special issue, Gallagher seems to be moving beyond insisting that IT should be adopted as an alternative to the standard theories of mindreading. In the framework of the pluralistic approach it will hopefully be possible to learn more about the role of both, mentalistic and non-mentalistic social cognitive processes. 3. Interbreeding the standard theories of mindreading with DSP: ‘‘new hybrids’’ or old news? Whereas Gallagher finds the idea that DSP is compatible with standard theories of mindreading surprising, and labels such accounts ‘‘the new hybrids,’’ several other authors are unlikely to see anything ‘‘new’’ or ‘‘hybrid’’ about this idea (Gallagher, current issue). What is the real issue here? And what, if anything, can a DSP theorist learn from this controversy? Gallagher argues that authors like Spaulding (in press), Carruthers (current issue), and Lavelle (2012) have not properly understood what DSP amounts to. In his view, these authors interbreed DSP with theory-theory on the assumption that DSP is merely a phenomenological thesis: ‘‘DSP is compatible with an inferentialist view as long as one thinks of DSP as merely a characterization of perceptual phenomenology or phenomenal experience, and thinks that all the real inferential action of social cognition is located at the subpersonal level and is inferential.’’ (Gallagher, current issue). But why do many authors tend to read DSP in a rather thin sense as targeting mainly the perceptual phenomenology of experience? I think the main reason is that DSP has not yet been fully developed as an account of the subpersonal processing underpinning the phenomenology of DSP. Gallagher’s ambition clearly goes beyond defending DSP as just a thesis on the phenomenology of certain social experiences (e.g. apprehending other people’s intentions and emotions) – he wants to rule DSP-TT/ST hybrids out, 1 Like most authors, I prefer to use the term ‘‘mindreading’’ for mental state attributions irrespective of how the underpinning cognitive processes are specified. On my use of the term, TT, ST and DSP are all theories of mindreading. Gallagher uses the term ‘‘mindreading’’ primarily to refer to mindreading by means of theorizing or simulation. 2 Recently, Bryant, Coffey, Povinelli, and Pruett (2013) have carried out a pioneering study using the experience sampling method to investigate the frequency of mental state attributions in contrast to behavioral attributions and miscellaneous thoughts. Bohl (2014) proposes a more specific hypothesis concerning the frequency and function of mindreading, which could be tested by the method of experience sampling.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
3
and to present DSP as an alternative to TT and ST, at least for certain cases of mindreading. But in order to block attempts to interbreed DSP with the standard theories of mindreading, two important questions are yet to be answered: (1) What are the criteria for distinguishing perceptual processes from cognitive or otherwise non-perceptual processes? And: (2) What makes DSP ‘‘smart’’ on the subpersonal level? So far, Gallagher has answered these questions mainly in the negative, stipulating that DSP on the subpersonal level does not involve theoretical inferences or simulation, but he has been rather silent about what it does involve. For example, Gallagher (current issue) writes: ‘‘Perception thus involves complex, dynamic processes at a subpersonal, sensory-motor level – but these processes are part of an enactive, dynamic engagement or response of the whole organism, rather than the result of additional, extra-perceptual inferential or simulative processes.’’ His positive account has remained underdeveloped: he has not provided any detailed functional explanation of what those ‘‘dynamic processes at a subpersonal, sensory-motor level’’ are to combat most accounts of TT and ST. In cognitive science, it is common to distinguish between at least two levels of subpersonal explanation: we can describe the subpersonal processes in biological/neural terms (e.g. ‘‘mirror neurons are activated’’), but we can also provide a functional account of those processes (e.g. ‘‘mirror neurons simulate the action of another individual’’). TT and ST typically postulate functional operations at the subpersonal level to explain how mindreading works.3 Gallagher provides a personal-level phenomenological account of DSP, and he has more recently started to appeal to some processes on the neural level (such as neural plasticity) that might help to explain how DSP is implemented in the brain and how perceptual processes can be shaped by the social environment. But a more specific functional account of the subpersonal processing is yet to be spelled out. Until this is done, the option of interbreeding DSP with the standard accounts of mindreading must be regarded as open. ‘Neural plasticity’ is a general term that refers to changes in neural pathways and synapses that occur over time; it alone does not provide an answer to the question of how to distinguish perceptual from non-perceptual processes, and it does not establish whether perception is best explained in terms of computational models, predictive coding, dynamical systems or some other framework. To block the hybrid accounts, it is not sufficient to appeal to the plasticity of the perceptual system, because the hybrid theorist may argue that DSP involves not extra-perceptual theoretical inferences or simulation, but theoretical inferences or simulation within the perceptual system.
4. The (un)observability of the mental Gallagher also classifies the paper by Bohl and Gangopadhyay (2014) as one of the ‘‘new hybrids.’’ However, he seems to overlook the main point of the paper: that the critique of the unobservability assumption (UA4), which proponents of DSP often level against standard theories of mindreading, is targeted at a straw man rather than at theories of mindreading properly understood. Gallagher persists in claiming that standard theories of mindreading (TT and ST) are founded upon the false assumption that other minds are unobservable. Since this claim has been put forward by many critics of standard theories of mindreading (e.g. Gallagher, 2004, 2008a, 2008b, 2008c; Gallagher & Zahavi, 2012; Hutto, 2009; Leudar & Costall, 2009; Ratcliffe, 2007; Reddy, 2008; Zahavi, 2005, 2007; Zahavi & Gallagher, 2008; Zahavi & Parnas, 2003), it is worth recalling the main line of argument provided by Bohl & Gangopadhyay – namely, that the critique of UA largely misses the mark because several different claims are lumped together under the UA. When we tease those different claims apart, we see that most readings of the UA do not apply to theories of mindreading. At the very least, the following four readings of the UA must be distinguished: (a) The metaphysical reading: minds and mental states are entities that cannot be observed by people in ordinary circumstances. (b) The phenomenological reading: apprehending minds or mental states is never a direct (one-step) experience; people experience only meaningless physical movements and add a mentalistic interpretation to it as a second step. (c) The epistemological reading: beliefs concerning other minds/other people’s mental states can never be justified by direct perception. (d) The psychological reading: the cognitive processes for apprehending minds/mental states go beyond perceptual mechanisms.
3 To provide just a couple of examples: Nichols and Stich (2003) outline a sophisticated model of the mindreading system by postulating a set of interconnected functional units that carry out specific tasks in the system. Goldman (2006) distinguishes between high-level and low-level mindreading, and outlines different candidate mechanisms to explain how each type of mindreading works. As an explanation of low-level mindreading of emotions, he argues for a model of ‘‘unmediated resonance,’’ according to which, when one perceives another person’s facial expression, one’s own neural substrate of the corresponding emotion is triggered, classified, and attributed to the target (pp. 127–129). 4 UA is the assumption that other minds are unobservable. Gallagher labels it the unobservability principle (UP).
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
4
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
Since standard theories of mindreading remain uncommitted to any particular metaphysical, epistemological or phenomenological position concerning minds/mental states, the first three readings of the UA do not apply to TT and ST. In other words, the psychological reading of the UA – the claim about the underlying cognitive processes – stands as the sole reading of the UA that actually applies to the theories of mindreading in question. Gallagher (current issue) seems to have taken this point on board, as he no longer rehearses his earlier phenomenological and epistemological arguments against the alleged UA of theories of mindreading, and admits that the phenomenology of ‘‘directly perceiving’’ other people’s mental states is compatible with TT and ST.5 However, he seems to disregard our point about the psychological version of the UA, which in turn also allows for different readings. A strong reading of the psychological UA amounts to the denial that the categorization of certain entities as ‘‘minded’’ is a purely perceptual process. In other words, it is the claim that perceptual processes make it possible to apprehend only physical objects, stripped of any psychological meaning. A medium version of the psychological UA states that perception may pick out certain salient properties that are indicative of ‘‘mindedness’’ (e.g. biological motion instantiating intentionality or agency), and thus enable recognition of some entities as in some sense ‘‘minded,’’ but denies that any particular mental state with a content can be attributed by perception alone. A weak version of the UA distinguishes between two types of mental states: those that can be attributed perceptually, and those the attribution of which requires extra-perceptual processes (theoretical inferences or simulation). As Bohl and Gangopadhyay (2014) point out: whereas critics tend to attribute the strongest psychological reading of the UA to the theories of mindreading, most versions of TT and ST actually assume the medium or the weak version. 5. The scope and content of DSP A clear example of the weak psychological UA is defended by Spaulding (in press, current issue). Spaulding argues that motor-intentions (M-intentions) and basic emotions could be directly perceived (in the sense that one can have basic perceptual beliefs about them) whereas most other types of mental states are unlikely to be directly perceived in this sense. Following Jack Lyons’ account of basic beliefs, she takes basic perceptual beliefs to be inferentially unmediated. Gallagher (current issue) seems to have missed this point, as he writes: ‘‘Accordingly, it seems as if we directly perceive M-intentions, for example, but, for theory theorists like Carruthers and Spaulding, the controlling processes are inferential and occur on a subpersonal level.’’ In other words, Gallagher interprets Spaulding as saying that the apprehension of even motor intentions and basic emotions requires extra-perceptual inferences, whereas Spaulding’s claim is the opposite. The point of Spaulding is that the perceptibility of M-intentions is a banal thesis: in her view most authors, including theory-theorists, are likely to agree to it. She suggests that a more interesting question is whether present or future-directed intentions can be directly perceived, and gives reasons for finding this unlikely. In conclusion, according to Spaulding, DSP is either the claim that we can directly perceive M-intentions and basic emotions, which most proponents of standard theories of mindreading would agree to; or it is the claim that we can also directly perceive mental states that are more loosely connected to observable behavior, such as present and/or future intentions. The latter is difficult to argue for, namely because of the many-to-many mapping between observable behavior and most mental states. An intention to do something in the future might lead to many different courses of action (or no action at all), whereas the same course of action might result from different intentions. This brings me to the more general question about the scope of DSP: what kind of mental states can we directly perceive? A serious problem for any account of mindreading is the problem of holism: behavior is connected to complex combinations of mental states. In principle, any behavioral act might be driven by almost any mental state given some combination of other mental states. Holism is a serious problem for standard accounts of mindreading (Morton, 1996, 2003; Zawidzki, 2013), but it also suggests that full-blown propositional attitudes are unlikely to be directly perceived. In any case, an important issue for any theorist of DSP is the question of what types of mental states can be directly perceived. The paradigmatic examples of mental states that are directly perceivable are intentions and emotions. But the proponents of DSP usually leave unspecified, what kind of emotions and intentions they have in mind. If Spaulding (current issue) is right in that M-intentions and basic emotions are the only plausible candidates for directly perceivable mental states then the explanatory scope of DSP seems extremely narrow. Another issue where more clarity is required is the question of what exactly the content6 of DSP is. Gallagher presents the account of Bohl and Gangopadhyay (2014) as maintaining that perception is just the registration of the sensory properties of objects, and as denying that mental states can be perceived without the help of theory or simulation.7 This is a misreading: 5 In earlier works, Gallagher has conflated the issue that theories of mindreading aim to solve (the issue of the cognitive mechanisms for mindreading) with the problem of other minds (i.e. the problem of epistemic access to other minds on the presupposition that it is possible to perceive only other people’s physical behavior, see e.g. Gallagher, 2004, p. 200). Like many other critics of theories of mindreading, he has argued that TT and ST go against the phenomenology of directly perceiving other people’s mental states (see e.g. Zahavi & Gallagher, 2008, p. 239). 6 I use the term ‘‘content’’ to refer to what is conveyed to the subject by her perceptual experience without presupposing that contents of perception have accuracy conditions. 7 Gallagher presents the account of Bohl and Gangopadhyay (2014) as a DSP-TT hybrid. At the same time, he reads the paper by Smith (2010) as supporting DSP against standard accounts of mindreading. But Smith (2010, p. 748) writes about his account: ‘‘It is no part of the present account that theory theory, or for that matter simulationism, is false.’’ Like Smith, Bohl and Gangopadhyay (2014) remain neutral to whether TT or ST are correct, and focus on outlining some phenomenological constraints that apply in the context of DSP as well as in the context of theories of mindreading. See also Gangopadhyay and Miyahara (2014) for an outline of further constraints to any theory of social perception.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
5
Bohl and Gangopadhyay remain open to the possibility that perception without theory or simulation could be up for the task of representing mental properties. As Gallagher also points out, they propose that the DSP discussion would profit from a Husserlian analysis of perception. Husserl analyses the structure of perceptual content in terms of what is ‘‘presented’’ and what is ‘‘co-presented’’ to the perceiver. He conceptualizes the co-presented aspects of physical objects as being grasped by the perceiver’s anticipations of how the presented content would change as a function of the perceiver’s engagement with the world (for example: if I walk around the chair, the now hidden profiles of the chair would gradually come to view). On Husserl’s account, perception of mental states is different from perception of physical objects. The main difference is that mental aspects are not co-presented in the same way as the absent profiles of objects are co-presented: while the co-presented absent profiles of objects can be subsequently verified by being given in presentation, the co-presented mental aspects can never be so verified. On the basis of the Husserlian account, Bohl & Gangopadhyay argue that any account of mindreading, including DSP, must reckon with the phenomenological constraint that the givenness of mental states is not reducible to the givenness of sensory properties of physical objects. They further point out that the theory-theorist could easily account for the experiential differences between object-perception and perception of mental states by referring to the differences in the underlying subpersonal mechanisms. Some authors argue that high-level properties can be (re)presented in experience because experiences can be theory-laden (see Siegel, 2015, pp. 39–40) and thus the theory-theorist can explain the difference between object-perception and mental-state-perception in terms of perception being informed by a folk-psychological theory in case of mental states. Gallagher does not buy into the idea of theory-ladenness of perception, but it remains unclear in virtue of what mental properties are perceptually given, or even what exactly those properties are and how they differ from properties of physical objects. Gallagher (current issue) explicitly agrees that the perception of mental states is not reducible to simply picking up sensory properties of physical objects. But it is less clear what exactly he takes to be given in the content of DSP. On his account, DSP is a ‘‘smart’’ enactive process in the service of social interaction. DSP is enactive in that it involves perceiving affordances for social interaction: ‘‘one perceives the other’s action or expression as an affordance for further action and interaction’’ (Gallagher, current issue). The difference between object-perception and mental-state-perception thus seems to boil down to the difference in the nature of respective affordances: whereas object-perception is a matter of perception of affordances for individual action, mental-state-perception is a matter of affordances for social interaction. Gallagher (current issue) writes: ‘‘. . .my perception of your action is already formed in terms of how I might respond to your action. I see your action, not as an objective fact that needs to be interpreted in terms of your mental states, but as a situated opportunity or affordance for my own response. If these are not mutually exclusive processes, to the extent that the enactive perception is sufficient for understanding your intentions, an extra step of mindreading is simply redundant.’’ But what is the exact relationship between perceiving social affordances and perceiving mental states as such? Insofar as Gallagher presents DSP as providing us with direct access to some of the mental states of others, he seems to imply some kind of holistic account of the content of DSP, where the other person’s mental state (e.g. intention) and the affordance for my own response are given as somehow merged into a single whole, instead of the other person’s mental state being presented separately, in an ‘‘extra step of mindreading.’’ But it is hard to see what this exactly means. The relationship between the experience of perceiving social affordances (e.g. ‘‘I can smile or frown to X’’) and that of perceiving mental states as such (e.g. ‘‘X is surprised to see me’’) requires further clarification. 6. DSP and more recent developments in mindreading research As I explained above, it is currently unclear what kind of mental states fall into the category of ‘‘directly perceivable’’ and how they figure in the content of perception. However, should it turn out that only a very small subset of mental states can be directly perceived, this does not mean that standard mindreading accounts (TT and/or ST) have won the game. Critics of standard theories of mindreading might still be right in claiming that most of the time people do not attribute full-blown propositional attitudes to other people in order to efficiently socially coordinate. In recent years, the consensus that full-blown propositional attitude attributions is a ubiquitous feature of human social cognition has been rapidly eroding. Several researchers have proposed that humans may rely heavily on some automatic low-level ability to track or register not propositional attitudes as such, but some simpler functional analogues of propositional attitudes. It is time for proponents of IT and DSP to relate to these more recent developments in the field. For example, how does the theory of teleological reasoning (Gergely & Csibra, 2003) or Butterfill and Apperly’s minimal ToM (Butterfill & Apperly, 2014) fit with DSP and IT? Would a proponent of DSP argue that tracking belief-like states is a matter of direct perception? If so, what would be the perceptual subpersonal explanation for minimal ToM? 7. Moving toward pluralism Gallagher argues that instead of building hybrids from IT/DSP and standard accounts of mindreading, a superior conciliatory strategy would be to endorse a pluralist approach. Indeed, as far as research on human social cognition is concerned, a pluralist approach looks like a promising path to take. Social cognition certainly calls for methodological pluralism: the more we are able to employ different methods to study the multifarious aspects of social cognition, the better. We have seen research on social cognition growing more methodologically pluralist over time: the methods of philosophy and psychology
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
6
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
have not only been enriched from within but also been supplemented with exciting new neuro-scientific methods. Hopefully, this trend will continue, as will a complementary influx of new methods from the social sciences. An important contribution of IT-theorists is that they have pointed out a gap in the methods for studying mindreading: namely that mindreading has for the most part been studied in observational, non-interactive contexts. Gallagher (current issue), however, proposes theoretical pluralism: the idea that there might be some truth to all theories of social cognition as long as they are taken to explain only some limited aspect of social cognition. Unfortunately, he does not provide any concrete suggestions about how to develop such a pluralist account of social cognition apart from noting that all current theories are likely to play some role in it. Neither does he discuss in detail the pluralist accounts that have been recently proposed by others, such as the account of Andrews (2008, 2012), or Fiebich (2015). The worry is that unless and until one outlines at least a broad framework within which different theories could hang together and supplement each other, the proposed pluralist approach will not get off the ground. In any case, to estimate the role of DSP in a pluralistic framework, the theory of DSP needs to be developed further to provide more detailed answers to several important questions I have pointed out here. First of all, criteria for distinguishing perceptual processes from non-perceptual processes need to be clearly spelled out. It needs to be specified in what sense the enactive perception of social interaction affordances involves the perception of mental states. And most importantly, a detailed subpersonal functional explanation of what makes DSP ‘‘smart’’ is yet to be worked out. Acknowledgments I acknowledge support received from the Volkswagen Foundation – Germany in the framework of the project ‘‘Situated Cognition. Perceiving the World and Understanding Other Minds’’ and Grants IUT20-5 and ETF9117 – Estonia. References Andrews, K. (2008). It’s in your nature: A pluralistic folk psychology. Synthese, 165(1), 13–29. Andrews, K. (2012). Do apes read minds? Toward a new folk psychology. Cambridge, MA: The MIT Press. Bohl, V. (2014). We read minds to shape relationships. Philosophical Psychology [Published online: 12 March 2014]. Bohl, V., & Gangopadhyay, N. (2014). Theory of mind and the unobservability of other minds. Philosophical Explorations, 17(2), 203–222. Bryant, L., Coffey, A., Povinelli, D. J., & Pruett, J. R. (2013). Theory of mind experience sampling in typical adults. Consciousness and Cognition, 22(3), 697–707. Butterfill, S., & Apperly, I. (2014). How to construct a minimal theory of mind? Mind & Language, 28(5), 606–637. Carruthers, P. (current issue). Perceiving mental states. Consciousness and Cognition. . Fiebich, A. (2015). Varieties of social understanding. Münster: Mentis. Gallagher, S. (2004). Understanding interpersonal problems in autism: Interaction theory as an alternative to theory of mind. Philosophy, Psychiatry, and Psychology, 11(3), 199–217. Gallagher, S. (2008a). Direct perception in the intersubjective context. Consciousness and Cognition, 17, 535–543. Gallagher, S. (2008b). Inference or interaction: Social cognition without precursors. Philosophical Explorations, 11(3), 163–174. Gallagher, S. (2008c). Understanding others: Embodied social cognition. In P. Calvo & T. Gomila (Eds.), Elsevier handbook of cognitive science: An embodied approach (pp. 439–452). London: Elsevier. Gallagher, S. (current issue). The new hybrids: Continuing debates on social cognition. Consciousness and Cognition. . Gallagher, S., & Zahavi, D. (2012). The phenomenological mind. London: Routledge. Gangopadhyay, N., & Miyahara, K. (2014). Perception and the problem of access to other minds. Philosophical Psychology [Published online: 18 March 2014]. Gergely, G., & Csibra, G. (2003). Teleological reasoning in infancy: The naïve theory of rational action. Trends in Cognitive Sciences, 7(7), 287–292. Goldman, A. I. (2006). Simulating minds: The philosophy, psychology, and neuroscience of mindreading. Oxford; New York: Oxford University Press. Hutto, D. (2009). ToM rules, but it is not OK! In I. Leudar & A. Costall (Eds.), Against theory of mind (pp. 221–262). New York: Palgrave Macmillan. Lavelle, J. S. (2012). Theory-theory and the direct perception of mental states. Review of Philosophy and Psychology, 3(2), 213–230. Leudar, I., & Costall, A. (2009). Introduction: Against ‘Theory of Mind’. In I. Leudar & A. Costall (Eds.), Against theory of mind (pp. 1–16). New York: Palgrave Macmillan. Michael, J., Christensen, W., & Overgaard, S. (2014). Mindreading as social expertise. Synthese, 191(5), 817–840. Morton, A. (1996). Folk psychology is not a predictive device. Mind, 105(417), 119–137. Morton, A. (2003). The importance of being understood. London: Routledge. Nichols, S., & Stich, S. P. (2003). Mindreading: An integrated account of pretence, self-awareness, and understanding other minds. Oxford; New York: Oxford University Press. Ratcliffe, M. (2007). Rethinking commonsense psychology: A critique of folk psychology, theory of mind and simulation. London: Palgrave-Macmillan. Reddy, V. (2008). How infants know minds. Cambridge: Harvard University Press. Siegel, S. (2015). The contents of perception. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Spring 2015 Edition). . Smith, J. (2010). Seeing other people. Philosophy and Phenomenological Research, 81(3), 731–748. Spaulding, S. (in press). On whether we can see intentions. Pacific Philosophical Quarterly. Spaulding, S. (current issue). On direct social perception. Consciousness and Cognition. . Zahavi, D. (2005). Subjectivity and selfhood. Cambridge, MA: The MIT Press. Zahavi, D., & Gallagher, S. (2008). The (in)visibility of others: A reply to Herschbach. Philosophical Explorations, 11(3), 237–244. Zahavi, D., & Parnas, J. (2003). Conceptual problems in infantile autism research: Why cognitive science needs phenomenology. Journal of Consciousness Studies, 10(9), 53–71. Zahavi, D. (2007). Expression and empathy. In M. Ratcliffe & D. Hutto (Eds.), Folk-psychology re-assessed (pp. 25–40). Dordrecht: Springer. Zawidzki, T. (2013). Mindshaping: A new framework for understanding human social cognition. Cambridge, MA: The MIT Press.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
Contents lists available at ScienceDirect
Consciousness and Cognition journal homepage: www.elsevier.com/locate/concog
Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’ Vivian Bohl ⇑ Institute of Philosophy II, Ruhr-Universität Bochum, Germany Department of Philosophy, Institute of Philosophy and Semiotics, University of Tartu, Estonia
a r t i c l e
i n f o
Article history: Received 31 December 2014 Revised 8 May 2015 Accepted 9 May 2015 Available online xxxx Keyword: Social cognition Mindreading Direct perception Theory-theory Simulation theory Hybrid theories
a b s t r a c t This commentary argues that Gallagher’s account of direct social perception has remained underdeveloped in several respects. Gallagher has not provided convincing evidence to support his claim that mindreading is rare in social situations. He and other direct perception theorists have not offered a substantive critique of standard theories of mindreading because they have attacked a much stronger claim about the putative unobservability of mental states than most theories of mindreading imply. To provide a genuine alternative to standard theories of mindreading, the direct perception theorist needs to provide more detailed answers to the following questions: What are the criteria for distinguishing perceptual processes from non-perceptual processes? How exactly does direct social perception function on the subpersonal level? What is the content of direct social perception? How does direct perception theory relate to more recent developments in the mindreading literature? Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction In ‘‘The new hybrids: Continuing debates on social cognition’’ Shaun Gallagher briefly outlines his version of the direct social perception (DSP) account as it forms a part of the interaction theory (IT) (see e.g. Gallagher, 2004; Gallagher & Zahavi, 2012). IT is a theory of human social cognition, proposed as an alternative to standard theories of mindreading: theory-theory (TT) and simulation theory (ST). IT states that mindreading via theorizing or simulation is unnecessary in most social situations. According to the thesis of DSP, we can directly perceive some mental states of other individuals, paradigmatically their intentions and (basic) emotions. In the context of IT, DSP is understood as an embodied and enactive process in the service of social interaction. Gallagher criticizes what he labels ‘‘the new hybrids,’’ characterized as recent attempts to integrate standard theories of mindreading with IT and, in particular, with the thesis of DSP. He argues that a better model for a hybrid approach would be a pluralist one. In this commentary, I would like to point out some issues in relation to which Gallagher’s IT in general, and DSP in particular, would benefit from being developed in more detail. I suggest that more empirical research is required to back up Gallagher’s claim that mindreading occurs rarely in social situations. I raise the point that DSP-theorists have not offered a substantive critique of theories of mindreading because they have attacked a much stronger claim about the putative unobservability of mental states than most theories of mindreading actually imply. In order to provide a genuine alternative
⇑ Address: Department of Philosophy, Institute of Philosophy and Semiotics, University of Tartu, Estonia. E-mail address: [email protected] http://dx.doi.org/10.1016/j.concog.2015.05.003 1053-8100/Ó 2015 Elsevier Inc. All rights reserved.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
2
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
to standard theories of mindreading, a much more detailed explanation of how DSP works on the subpersonal level would need to be worked out. Finally, more needs to be done to relate the DSP view to recent developments in the mindreading literature. 2. How ‘‘rarely’’ do we actually mindread? Let me start with some thoughts on Gallagher’s critical analysis of the account provided by Michael, Christensen, and Overgaard (2014). Michael et al. argue that it is plausible that in social contexts high-level processes (including mindreading) and low-level processes (such as embodied social responses highlighted by IT) inform each other. Gallagher partly agrees with this proposal, saying that ‘‘IT doesn’t deny that higher-order cognitive processing may be involved in many social encounter contexts’’ (Gallagher, current issue), but he strongly resists the idea that such higher-order processing may typically involve mindreading via theorizing or simulation.1 Instead he claims that mindreading (in the sense of TT or ST) is ‘‘rarely used’’ in social encounters. Gallagher and other proponents of IT correctly point out that the assumption that mindreading is ubiquitous in human social cognition, and necessary for human social interaction, is unwarranted. Mainstream research on mindreading has been based mainly on laboratory experiments, such as false belief tasks, which are specifically designed to elicit mindreading. But such experiments cannot possibly tell us how commonly people mindread in real life. The proponents of TT and ST have arguably simply assumed that mindreading is ubiquitous in human social cognition, which has prevented empirical research on how often people actually mindread in real-life social situations. However, it is equally problematic to prematurely buy into the opposite claim, namely that mindreading is rare and hardly plays any role in human social cognition. Unfortunately, this is exactly the claim that Gallagher and the other proponents of IT are making. The problem with any claim about the frequency of mindreading is that it currently remains as speculative as the claim that mindreading is pervasive in human social cognition: at the moment we have no decisive evidence on the matter.2 We simply do not know how often and for what purposes people actually engage in mindreading in real life. So far, Gallagher has not provided compelling empirical evidence to back up his theoretical claim that in most social situations, people typically rely on low-level processes (such as DSP) and higher cognitive processes except for mindreading (in the sense of TT and ST). In sum, his claim that mindreading is ‘‘rarely used’’ remains as speculative as the claim that it is ubiquitous in human social cognition. Ascertaining the frequency of mindreading may or may not be an interesting research aim in its own right but it is clearly an important step toward specifying the role of mindreading in human social cognition. If human social cognition can in most cases be explained by reference to social interaction, DSP and non-mentalistic higher-order cognition as Gallagher suggests, it remains mysterious why the ability to attribute full-blown propositional attitudes like beliefs and desires has evolved in the first place, and what its current function is. The upshot is that it is indeed unwarranted to assume a priori that mindreading is as ubiquitous in human social cognition as standard theories of mindreading have arguably assumed, but it is equally unjustified to suppose that mindreading is as peripheral to human social cognition as the proponents of IT tend to claim. On a more positive note, by suggesting a pluralistic approach in this special issue, Gallagher seems to be moving beyond insisting that IT should be adopted as an alternative to the standard theories of mindreading. In the framework of the pluralistic approach it will hopefully be possible to learn more about the role of both, mentalistic and non-mentalistic social cognitive processes. 3. Interbreeding the standard theories of mindreading with DSP: ‘‘new hybrids’’ or old news? Whereas Gallagher finds the idea that DSP is compatible with standard theories of mindreading surprising, and labels such accounts ‘‘the new hybrids,’’ several other authors are unlikely to see anything ‘‘new’’ or ‘‘hybrid’’ about this idea (Gallagher, current issue). What is the real issue here? And what, if anything, can a DSP theorist learn from this controversy? Gallagher argues that authors like Spaulding (in press), Carruthers (current issue), and Lavelle (2012) have not properly understood what DSP amounts to. In his view, these authors interbreed DSP with theory-theory on the assumption that DSP is merely a phenomenological thesis: ‘‘DSP is compatible with an inferentialist view as long as one thinks of DSP as merely a characterization of perceptual phenomenology or phenomenal experience, and thinks that all the real inferential action of social cognition is located at the subpersonal level and is inferential.’’ (Gallagher, current issue). But why do many authors tend to read DSP in a rather thin sense as targeting mainly the perceptual phenomenology of experience? I think the main reason is that DSP has not yet been fully developed as an account of the subpersonal processing underpinning the phenomenology of DSP. Gallagher’s ambition clearly goes beyond defending DSP as just a thesis on the phenomenology of certain social experiences (e.g. apprehending other people’s intentions and emotions) – he wants to rule DSP-TT/ST hybrids out, 1 Like most authors, I prefer to use the term ‘‘mindreading’’ for mental state attributions irrespective of how the underpinning cognitive processes are specified. On my use of the term, TT, ST and DSP are all theories of mindreading. Gallagher uses the term ‘‘mindreading’’ primarily to refer to mindreading by means of theorizing or simulation. 2 Recently, Bryant, Coffey, Povinelli, and Pruett (2013) have carried out a pioneering study using the experience sampling method to investigate the frequency of mental state attributions in contrast to behavioral attributions and miscellaneous thoughts. Bohl (2014) proposes a more specific hypothesis concerning the frequency and function of mindreading, which could be tested by the method of experience sampling.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
3
and to present DSP as an alternative to TT and ST, at least for certain cases of mindreading. But in order to block attempts to interbreed DSP with the standard theories of mindreading, two important questions are yet to be answered: (1) What are the criteria for distinguishing perceptual processes from cognitive or otherwise non-perceptual processes? And: (2) What makes DSP ‘‘smart’’ on the subpersonal level? So far, Gallagher has answered these questions mainly in the negative, stipulating that DSP on the subpersonal level does not involve theoretical inferences or simulation, but he has been rather silent about what it does involve. For example, Gallagher (current issue) writes: ‘‘Perception thus involves complex, dynamic processes at a subpersonal, sensory-motor level – but these processes are part of an enactive, dynamic engagement or response of the whole organism, rather than the result of additional, extra-perceptual inferential or simulative processes.’’ His positive account has remained underdeveloped: he has not provided any detailed functional explanation of what those ‘‘dynamic processes at a subpersonal, sensory-motor level’’ are to combat most accounts of TT and ST. In cognitive science, it is common to distinguish between at least two levels of subpersonal explanation: we can describe the subpersonal processes in biological/neural terms (e.g. ‘‘mirror neurons are activated’’), but we can also provide a functional account of those processes (e.g. ‘‘mirror neurons simulate the action of another individual’’). TT and ST typically postulate functional operations at the subpersonal level to explain how mindreading works.3 Gallagher provides a personal-level phenomenological account of DSP, and he has more recently started to appeal to some processes on the neural level (such as neural plasticity) that might help to explain how DSP is implemented in the brain and how perceptual processes can be shaped by the social environment. But a more specific functional account of the subpersonal processing is yet to be spelled out. Until this is done, the option of interbreeding DSP with the standard accounts of mindreading must be regarded as open. ‘Neural plasticity’ is a general term that refers to changes in neural pathways and synapses that occur over time; it alone does not provide an answer to the question of how to distinguish perceptual from non-perceptual processes, and it does not establish whether perception is best explained in terms of computational models, predictive coding, dynamical systems or some other framework. To block the hybrid accounts, it is not sufficient to appeal to the plasticity of the perceptual system, because the hybrid theorist may argue that DSP involves not extra-perceptual theoretical inferences or simulation, but theoretical inferences or simulation within the perceptual system.
4. The (un)observability of the mental Gallagher also classifies the paper by Bohl and Gangopadhyay (2014) as one of the ‘‘new hybrids.’’ However, he seems to overlook the main point of the paper: that the critique of the unobservability assumption (UA4), which proponents of DSP often level against standard theories of mindreading, is targeted at a straw man rather than at theories of mindreading properly understood. Gallagher persists in claiming that standard theories of mindreading (TT and ST) are founded upon the false assumption that other minds are unobservable. Since this claim has been put forward by many critics of standard theories of mindreading (e.g. Gallagher, 2004, 2008a, 2008b, 2008c; Gallagher & Zahavi, 2012; Hutto, 2009; Leudar & Costall, 2009; Ratcliffe, 2007; Reddy, 2008; Zahavi, 2005, 2007; Zahavi & Gallagher, 2008; Zahavi & Parnas, 2003), it is worth recalling the main line of argument provided by Bohl & Gangopadhyay – namely, that the critique of UA largely misses the mark because several different claims are lumped together under the UA. When we tease those different claims apart, we see that most readings of the UA do not apply to theories of mindreading. At the very least, the following four readings of the UA must be distinguished: (a) The metaphysical reading: minds and mental states are entities that cannot be observed by people in ordinary circumstances. (b) The phenomenological reading: apprehending minds or mental states is never a direct (one-step) experience; people experience only meaningless physical movements and add a mentalistic interpretation to it as a second step. (c) The epistemological reading: beliefs concerning other minds/other people’s mental states can never be justified by direct perception. (d) The psychological reading: the cognitive processes for apprehending minds/mental states go beyond perceptual mechanisms.
3 To provide just a couple of examples: Nichols and Stich (2003) outline a sophisticated model of the mindreading system by postulating a set of interconnected functional units that carry out specific tasks in the system. Goldman (2006) distinguishes between high-level and low-level mindreading, and outlines different candidate mechanisms to explain how each type of mindreading works. As an explanation of low-level mindreading of emotions, he argues for a model of ‘‘unmediated resonance,’’ according to which, when one perceives another person’s facial expression, one’s own neural substrate of the corresponding emotion is triggered, classified, and attributed to the target (pp. 127–129). 4 UA is the assumption that other minds are unobservable. Gallagher labels it the unobservability principle (UP).
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
4
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
Since standard theories of mindreading remain uncommitted to any particular metaphysical, epistemological or phenomenological position concerning minds/mental states, the first three readings of the UA do not apply to TT and ST. In other words, the psychological reading of the UA – the claim about the underlying cognitive processes – stands as the sole reading of the UA that actually applies to the theories of mindreading in question. Gallagher (current issue) seems to have taken this point on board, as he no longer rehearses his earlier phenomenological and epistemological arguments against the alleged UA of theories of mindreading, and admits that the phenomenology of ‘‘directly perceiving’’ other people’s mental states is compatible with TT and ST.5 However, he seems to disregard our point about the psychological version of the UA, which in turn also allows for different readings. A strong reading of the psychological UA amounts to the denial that the categorization of certain entities as ‘‘minded’’ is a purely perceptual process. In other words, it is the claim that perceptual processes make it possible to apprehend only physical objects, stripped of any psychological meaning. A medium version of the psychological UA states that perception may pick out certain salient properties that are indicative of ‘‘mindedness’’ (e.g. biological motion instantiating intentionality or agency), and thus enable recognition of some entities as in some sense ‘‘minded,’’ but denies that any particular mental state with a content can be attributed by perception alone. A weak version of the UA distinguishes between two types of mental states: those that can be attributed perceptually, and those the attribution of which requires extra-perceptual processes (theoretical inferences or simulation). As Bohl and Gangopadhyay (2014) point out: whereas critics tend to attribute the strongest psychological reading of the UA to the theories of mindreading, most versions of TT and ST actually assume the medium or the weak version. 5. The scope and content of DSP A clear example of the weak psychological UA is defended by Spaulding (in press, current issue). Spaulding argues that motor-intentions (M-intentions) and basic emotions could be directly perceived (in the sense that one can have basic perceptual beliefs about them) whereas most other types of mental states are unlikely to be directly perceived in this sense. Following Jack Lyons’ account of basic beliefs, she takes basic perceptual beliefs to be inferentially unmediated. Gallagher (current issue) seems to have missed this point, as he writes: ‘‘Accordingly, it seems as if we directly perceive M-intentions, for example, but, for theory theorists like Carruthers and Spaulding, the controlling processes are inferential and occur on a subpersonal level.’’ In other words, Gallagher interprets Spaulding as saying that the apprehension of even motor intentions and basic emotions requires extra-perceptual inferences, whereas Spaulding’s claim is the opposite. The point of Spaulding is that the perceptibility of M-intentions is a banal thesis: in her view most authors, including theory-theorists, are likely to agree to it. She suggests that a more interesting question is whether present or future-directed intentions can be directly perceived, and gives reasons for finding this unlikely. In conclusion, according to Spaulding, DSP is either the claim that we can directly perceive M-intentions and basic emotions, which most proponents of standard theories of mindreading would agree to; or it is the claim that we can also directly perceive mental states that are more loosely connected to observable behavior, such as present and/or future intentions. The latter is difficult to argue for, namely because of the many-to-many mapping between observable behavior and most mental states. An intention to do something in the future might lead to many different courses of action (or no action at all), whereas the same course of action might result from different intentions. This brings me to the more general question about the scope of DSP: what kind of mental states can we directly perceive? A serious problem for any account of mindreading is the problem of holism: behavior is connected to complex combinations of mental states. In principle, any behavioral act might be driven by almost any mental state given some combination of other mental states. Holism is a serious problem for standard accounts of mindreading (Morton, 1996, 2003; Zawidzki, 2013), but it also suggests that full-blown propositional attitudes are unlikely to be directly perceived. In any case, an important issue for any theorist of DSP is the question of what types of mental states can be directly perceived. The paradigmatic examples of mental states that are directly perceivable are intentions and emotions. But the proponents of DSP usually leave unspecified, what kind of emotions and intentions they have in mind. If Spaulding (current issue) is right in that M-intentions and basic emotions are the only plausible candidates for directly perceivable mental states then the explanatory scope of DSP seems extremely narrow. Another issue where more clarity is required is the question of what exactly the content6 of DSP is. Gallagher presents the account of Bohl and Gangopadhyay (2014) as maintaining that perception is just the registration of the sensory properties of objects, and as denying that mental states can be perceived without the help of theory or simulation.7 This is a misreading: 5 In earlier works, Gallagher has conflated the issue that theories of mindreading aim to solve (the issue of the cognitive mechanisms for mindreading) with the problem of other minds (i.e. the problem of epistemic access to other minds on the presupposition that it is possible to perceive only other people’s physical behavior, see e.g. Gallagher, 2004, p. 200). Like many other critics of theories of mindreading, he has argued that TT and ST go against the phenomenology of directly perceiving other people’s mental states (see e.g. Zahavi & Gallagher, 2008, p. 239). 6 I use the term ‘‘content’’ to refer to what is conveyed to the subject by her perceptual experience without presupposing that contents of perception have accuracy conditions. 7 Gallagher presents the account of Bohl and Gangopadhyay (2014) as a DSP-TT hybrid. At the same time, he reads the paper by Smith (2010) as supporting DSP against standard accounts of mindreading. But Smith (2010, p. 748) writes about his account: ‘‘It is no part of the present account that theory theory, or for that matter simulationism, is false.’’ Like Smith, Bohl and Gangopadhyay (2014) remain neutral to whether TT or ST are correct, and focus on outlining some phenomenological constraints that apply in the context of DSP as well as in the context of theories of mindreading. See also Gangopadhyay and Miyahara (2014) for an outline of further constraints to any theory of social perception.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
5
Bohl and Gangopadhyay remain open to the possibility that perception without theory or simulation could be up for the task of representing mental properties. As Gallagher also points out, they propose that the DSP discussion would profit from a Husserlian analysis of perception. Husserl analyses the structure of perceptual content in terms of what is ‘‘presented’’ and what is ‘‘co-presented’’ to the perceiver. He conceptualizes the co-presented aspects of physical objects as being grasped by the perceiver’s anticipations of how the presented content would change as a function of the perceiver’s engagement with the world (for example: if I walk around the chair, the now hidden profiles of the chair would gradually come to view). On Husserl’s account, perception of mental states is different from perception of physical objects. The main difference is that mental aspects are not co-presented in the same way as the absent profiles of objects are co-presented: while the co-presented absent profiles of objects can be subsequently verified by being given in presentation, the co-presented mental aspects can never be so verified. On the basis of the Husserlian account, Bohl & Gangopadhyay argue that any account of mindreading, including DSP, must reckon with the phenomenological constraint that the givenness of mental states is not reducible to the givenness of sensory properties of physical objects. They further point out that the theory-theorist could easily account for the experiential differences between object-perception and perception of mental states by referring to the differences in the underlying subpersonal mechanisms. Some authors argue that high-level properties can be (re)presented in experience because experiences can be theory-laden (see Siegel, 2015, pp. 39–40) and thus the theory-theorist can explain the difference between object-perception and mental-state-perception in terms of perception being informed by a folk-psychological theory in case of mental states. Gallagher does not buy into the idea of theory-ladenness of perception, but it remains unclear in virtue of what mental properties are perceptually given, or even what exactly those properties are and how they differ from properties of physical objects. Gallagher (current issue) explicitly agrees that the perception of mental states is not reducible to simply picking up sensory properties of physical objects. But it is less clear what exactly he takes to be given in the content of DSP. On his account, DSP is a ‘‘smart’’ enactive process in the service of social interaction. DSP is enactive in that it involves perceiving affordances for social interaction: ‘‘one perceives the other’s action or expression as an affordance for further action and interaction’’ (Gallagher, current issue). The difference between object-perception and mental-state-perception thus seems to boil down to the difference in the nature of respective affordances: whereas object-perception is a matter of perception of affordances for individual action, mental-state-perception is a matter of affordances for social interaction. Gallagher (current issue) writes: ‘‘. . .my perception of your action is already formed in terms of how I might respond to your action. I see your action, not as an objective fact that needs to be interpreted in terms of your mental states, but as a situated opportunity or affordance for my own response. If these are not mutually exclusive processes, to the extent that the enactive perception is sufficient for understanding your intentions, an extra step of mindreading is simply redundant.’’ But what is the exact relationship between perceiving social affordances and perceiving mental states as such? Insofar as Gallagher presents DSP as providing us with direct access to some of the mental states of others, he seems to imply some kind of holistic account of the content of DSP, where the other person’s mental state (e.g. intention) and the affordance for my own response are given as somehow merged into a single whole, instead of the other person’s mental state being presented separately, in an ‘‘extra step of mindreading.’’ But it is hard to see what this exactly means. The relationship between the experience of perceiving social affordances (e.g. ‘‘I can smile or frown to X’’) and that of perceiving mental states as such (e.g. ‘‘X is surprised to see me’’) requires further clarification. 6. DSP and more recent developments in mindreading research As I explained above, it is currently unclear what kind of mental states fall into the category of ‘‘directly perceivable’’ and how they figure in the content of perception. However, should it turn out that only a very small subset of mental states can be directly perceived, this does not mean that standard mindreading accounts (TT and/or ST) have won the game. Critics of standard theories of mindreading might still be right in claiming that most of the time people do not attribute full-blown propositional attitudes to other people in order to efficiently socially coordinate. In recent years, the consensus that full-blown propositional attitude attributions is a ubiquitous feature of human social cognition has been rapidly eroding. Several researchers have proposed that humans may rely heavily on some automatic low-level ability to track or register not propositional attitudes as such, but some simpler functional analogues of propositional attitudes. It is time for proponents of IT and DSP to relate to these more recent developments in the field. For example, how does the theory of teleological reasoning (Gergely & Csibra, 2003) or Butterfill and Apperly’s minimal ToM (Butterfill & Apperly, 2014) fit with DSP and IT? Would a proponent of DSP argue that tracking belief-like states is a matter of direct perception? If so, what would be the perceptual subpersonal explanation for minimal ToM? 7. Moving toward pluralism Gallagher argues that instead of building hybrids from IT/DSP and standard accounts of mindreading, a superior conciliatory strategy would be to endorse a pluralist approach. Indeed, as far as research on human social cognition is concerned, a pluralist approach looks like a promising path to take. Social cognition certainly calls for methodological pluralism: the more we are able to employ different methods to study the multifarious aspects of social cognition, the better. We have seen research on social cognition growing more methodologically pluralist over time: the methods of philosophy and psychology
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
6
V. Bohl / Consciousness and Cognition xxx (2015) xxx–xxx
have not only been enriched from within but also been supplemented with exciting new neuro-scientific methods. Hopefully, this trend will continue, as will a complementary influx of new methods from the social sciences. An important contribution of IT-theorists is that they have pointed out a gap in the methods for studying mindreading: namely that mindreading has for the most part been studied in observational, non-interactive contexts. Gallagher (current issue), however, proposes theoretical pluralism: the idea that there might be some truth to all theories of social cognition as long as they are taken to explain only some limited aspect of social cognition. Unfortunately, he does not provide any concrete suggestions about how to develop such a pluralist account of social cognition apart from noting that all current theories are likely to play some role in it. Neither does he discuss in detail the pluralist accounts that have been recently proposed by others, such as the account of Andrews (2008, 2012), or Fiebich (2015). The worry is that unless and until one outlines at least a broad framework within which different theories could hang together and supplement each other, the proposed pluralist approach will not get off the ground. In any case, to estimate the role of DSP in a pluralistic framework, the theory of DSP needs to be developed further to provide more detailed answers to several important questions I have pointed out here. First of all, criteria for distinguishing perceptual processes from non-perceptual processes need to be clearly spelled out. It needs to be specified in what sense the enactive perception of social interaction affordances involves the perception of mental states. And most importantly, a detailed subpersonal functional explanation of what makes DSP ‘‘smart’’ is yet to be worked out. Acknowledgments I acknowledge support received from the Volkswagen Foundation – Germany in the framework of the project ‘‘Situated Cognition. Perceiving the World and Understanding Other Minds’’ and Grants IUT20-5 and ETF9117 – Estonia. References Andrews, K. (2008). It’s in your nature: A pluralistic folk psychology. Synthese, 165(1), 13–29. Andrews, K. (2012). Do apes read minds? Toward a new folk psychology. Cambridge, MA: The MIT Press. Bohl, V. (2014). We read minds to shape relationships. Philosophical Psychology [Published online: 12 March 2014]. Bohl, V., & Gangopadhyay, N. (2014). Theory of mind and the unobservability of other minds. Philosophical Explorations, 17(2), 203–222. Bryant, L., Coffey, A., Povinelli, D. J., & Pruett, J. R. (2013). Theory of mind experience sampling in typical adults. Consciousness and Cognition, 22(3), 697–707. Butterfill, S., & Apperly, I. (2014). How to construct a minimal theory of mind? Mind & Language, 28(5), 606–637. Carruthers, P. (current issue). Perceiving mental states. Consciousness and Cognition. . Fiebich, A. (2015). Varieties of social understanding. Münster: Mentis. Gallagher, S. (2004). Understanding interpersonal problems in autism: Interaction theory as an alternative to theory of mind. Philosophy, Psychiatry, and Psychology, 11(3), 199–217. Gallagher, S. (2008a). Direct perception in the intersubjective context. Consciousness and Cognition, 17, 535–543. Gallagher, S. (2008b). Inference or interaction: Social cognition without precursors. Philosophical Explorations, 11(3), 163–174. Gallagher, S. (2008c). Understanding others: Embodied social cognition. In P. Calvo & T. Gomila (Eds.), Elsevier handbook of cognitive science: An embodied approach (pp. 439–452). London: Elsevier. Gallagher, S. (current issue). The new hybrids: Continuing debates on social cognition. Consciousness and Cognition. . Gallagher, S., & Zahavi, D. (2012). The phenomenological mind. London: Routledge. Gangopadhyay, N., & Miyahara, K. (2014). Perception and the problem of access to other minds. Philosophical Psychology [Published online: 18 March 2014]. Gergely, G., & Csibra, G. (2003). Teleological reasoning in infancy: The naïve theory of rational action. Trends in Cognitive Sciences, 7(7), 287–292. Goldman, A. I. (2006). Simulating minds: The philosophy, psychology, and neuroscience of mindreading. Oxford; New York: Oxford University Press. Hutto, D. (2009). ToM rules, but it is not OK! In I. Leudar & A. Costall (Eds.), Against theory of mind (pp. 221–262). New York: Palgrave Macmillan. Lavelle, J. S. (2012). Theory-theory and the direct perception of mental states. Review of Philosophy and Psychology, 3(2), 213–230. Leudar, I., & Costall, A. (2009). Introduction: Against ‘Theory of Mind’. In I. Leudar & A. Costall (Eds.), Against theory of mind (pp. 1–16). New York: Palgrave Macmillan. Michael, J., Christensen, W., & Overgaard, S. (2014). Mindreading as social expertise. Synthese, 191(5), 817–840. Morton, A. (1996). Folk psychology is not a predictive device. Mind, 105(417), 119–137. Morton, A. (2003). The importance of being understood. London: Routledge. Nichols, S., & Stich, S. P. (2003). Mindreading: An integrated account of pretence, self-awareness, and understanding other minds. Oxford; New York: Oxford University Press. Ratcliffe, M. (2007). Rethinking commonsense psychology: A critique of folk psychology, theory of mind and simulation. London: Palgrave-Macmillan. Reddy, V. (2008). How infants know minds. Cambridge: Harvard University Press. Siegel, S. (2015). The contents of perception. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Spring 2015 Edition). . Smith, J. (2010). Seeing other people. Philosophy and Phenomenological Research, 81(3), 731–748. Spaulding, S. (in press). On whether we can see intentions. Pacific Philosophical Quarterly. Spaulding, S. (current issue). On direct social perception. Consciousness and Cognition. . Zahavi, D. (2005). Subjectivity and selfhood. Cambridge, MA: The MIT Press. Zahavi, D., & Gallagher, S. (2008). The (in)visibility of others: A reply to Herschbach. Philosophical Explorations, 11(3), 237–244. Zahavi, D., & Parnas, J. (2003). Conceptual problems in infantile autism research: Why cognitive science needs phenomenology. Journal of Consciousness Studies, 10(9), 53–71. Zahavi, D. (2007). Expression and empathy. In M. Ratcliffe & D. Hutto (Eds.), Folk-psychology re-assessed (pp. 25–40). Dordrecht: Springer. Zawidzki, T. (2013). Mindshaping: A new framework for understanding human social cognition. Cambridge, MA: The MIT Press.
Please cite this article in press as: Bohl, V. Continuing debates on direct social perception: Some notes on Gallagher’s analysis of ‘‘the new hybrids’’. Consciousness and Cognition (2015), http://dx.doi.org/10.1016/j.concog.2015.05.003
