I perceive, therefore I understand – a modest claim  

By: Tamarinde Haven, MSc Mind Language and Embodied Cognition 2015, University of Edinburgh

A new proposal

In Direct Perception In the Intersubjective Context, Gallagher (2008a) introduces direct perception (DF) to account for part of social understanding. Social cognition research is predominated by theory-theory (TT) and simulation theory (ST), TT and ST agree on the fact that perception alone is insufficient to understand another’s behaviour. Instead we need to form theories based on people’s beliefs, intentions and desires (TT) or place ourselves in the other person’s shoes, thereby simulating what we would do (ST). DP is direct in the sense that it is free from deliberate cognitive inferences. It claims to be sufficient in our everyday understanding of one another. DP is informed by prior experience and holds that we perceive intentions as embodied acts in a social and cultural context, seen in the potential ways in which the perceiver can interact with these acts.

This paper is divided in three sections. First, I will argue that the main critiques of DP are based on a misunderstanding of the modesty of DP’s claims (with modest I mean moderate in estimating the amount of social understanding it explains). DP or similar views have been advocated by different authors, but my definition of DP and assessment of its critiques are based on Gallagher’s (2008a) paper, which is aimed at explicitly explaining what direct perception is and how it works. In what follows, I will emphasise some problems with the view and how DP appears to acknowledge these shortcomings. I will conclude that DP’s modesty can be seen as its strength within the broader framework of interaction theory (IT).

Context and interaction

Van Riel (2008) criticises DP based on the fact that it does not provide a full account of how we perceive intentions. According to him, intentions are most of the time not perceivable and can therefore not play a role in understanding someone’s behaviour. His example intended to highlight this is of an actor who fakes a heart attack. The actor does this so well that, based on perception alone, the perceiver will not be able to distinguish the faked heart attack from a real one, where the first is intentionally faked and the other genuine. Related to this, Jacob (2011) finds the claim rather hard to swallow that one can perceive a car and perceive intentions in equally direct ways. He investigates whether enactivist accounts can help DP perceiving (seeing the car as “driveable”), and concludes it does not always, because we can also see the car as “on sale”.

The critiques would hold if DP was equal to pure perception (only perception), but it is quite the contrary. Gallagher (2008b) replies to van Riel (2011) in emphasising the context in van Riel’s actor example. The actor who admirably well acts out the heart attack does so in the context of a stage; direct perception takes into account these contextual cues and hence is able to distinguish between the two forms of heart attacks. And even if the actor would act out this very same heart attack on the street and fool the innocent perceiver walking by, Gallagher (2008a) admits people can hide their intentions from us and DP is never said to be immune to errors. DP claims to normally suffice in everyday life (I hope you are not exposed to people faking heart attacks on a daily basis) and does so by taking into account contextual cues. Regarding the directness of object perception and intention perception, Jacobs (2011) is right in indicating a big jump. I think Gallagher (2008a) used this car example to indicate the similarities in the phenomenological experience of everyday perception, not necessarily to claim that all processes involved are equally direct. Also, DP is not just based on enactivist claims, but relies on an embodied perception where it perceives objects as objects within a context (the context of a car shop). Hence Jacob’s (2011) difficulty seems to be accounted for when one considers the modesty of DP’s claim by including context and primarily focusing on everyday experience.

Need for inferential mechanisms

Lavelle (2012) looks into the extent to which DP poses a threat for TT, and assesses their compatibility. When defining what DP is, she says: “In order for a perceptual experience to be direct, it has to be the case that there is nothing added to the contents of that perceptual experience” (Lavelle, 2012, p. 216). Lavelle’s main critique is that even if we do not need a theory to perceive other people’s intention, we need a theory of mind to interpret them. Similarly, in order to perceive my car, just perception is not enough; there is an inferential process needed to infer from basic light detection properties that the thing I perceive is in fact my car. Zawidzki (2012) states that DP cannot be a full-blown alternative to classical metapsychological theories if it cannot provide us with an answer to how the brain then does this assumingly non-inferential sense making. How do we perceive “intentions” without drawing upon representations of intentions or similar concepts (beliefs, desires, etc,)? How does the brain from its lower visual input develop complex concepts like intentions without any inferences? 1f DP cannot come up with a different explanation from current metapsychological theories, its phenomenological way of describing social cognition does not hold (Zawidzki, 2012).

Using Lavelle’s (2012) car example: if the perceiver has never seen a car before, he will need some inferential processes (he can use TT) in order to form a concept “car”. But after this concept is formed (and informs perception, the perceiver no longer needs conscious inferential processes to perceive his car. The case of novelty is discussed below, but perception goes beyond what happens in area V1 (area for visual processing containing simple and complex cells that fire in response to basic input like lines (Hubel & Wiesel, 1959). The existence of the fusiform face area can be interpreted as evidence that perceiving goes beyond detecting light properties (Sergent, Ohta. & MacDonald, 1992). DP characterises perception as updated, informed, and within a context of potential interaction. Lavelle seems to omit the fact that in ambiguous or novel situations, DP does not suffice and Gallagher (2008a) readily accepts the need for inferential processes as proposed by TT.

Regarding Zawidzki’s (2012) criticisms: DP was not intended to explain how perception works at a neuronal level. That is according to Gallagher “a problem for the neuroscientist; not for the perceiver” (Gallagher, 2008a, p. 537). Gallagher (2008a) wanted to give an account of perception from the perceiver’s point of view, not disowning the complexity of subpersonal processes at work. The critique regarding inferential mechanisms is directed at the wrong level: the brain has to make certain inferences, but DP holds social understanding in everyday life to be inference-free on a personal level. If the modesty of DP was given more consideration, Zawidzki (2012) might not have refuted it.

Regarding the question of how we perceive intentions, the framework in which DP is constituted conceptualizes intentions radically differently from classical cognitive theories (Gallagher & Varga, 2014). Borrowing concepts from action theory like Proximate intentions (intentions made according to the needs of the situation; P-intentions) and Motor intentions (intentions made according to the movement as part of an action; M-intentions) DP sees intentionality as expressed within purposeful motor behaviour (Parcherie, 2005 as by Gallagher & Varga, 2014). The perceiver is intentionally operating within a context, and his acts and aims are not to be diminished to mental states (Gallagher & Varga, 2014). For example: when I observe someone walking and placing his foot oddly and thus stumbling to maintain balance, I would not say he intended to stumble; he stumbled to prevent himself from falling (M-intention). “Actions have intentionality because they are directed at some goal or project, and this is something we can see in the actions of others.” (Gallagher & Varga, 2014, p.188). Pace those looking for pure perception: both P- and M-intentions are settled in social contexts and perceiving is embodied and framed in terms of my potential responses; I perceive as looking for opportunities to act.

Intersubjectivity first

De Jaegher (2009) attacks DP because Gallagher’s (2008a) examples tend to lose sight of the interaction between people that underlies true intersubjective understanding. Admittedly, it is easy to read direct perception as without interaction. But Gallagher (2008a) acknowledges the importance of interaction specifically in order to understand one another. The perceiver is not isolated, neither is the person he perceives. They are both constituted within a context; the acts of the other person are seen as opportunities to interact with that person according to DP. Interaction really forms an important part of DP, without which DP would not hold, so in its modesty it does embrace interaction. De Jaegher’s (2009) emphasis is on the seeming omitting of interaction, but when truly assessing DP, it is clear that DP does take interaction into account.

Clinical dilemma

Michael, Christensen and Overgaard (2014) pose a problem for DP by discussing two (sometimes seen as opposing) clinical groups: Williams syndrome and Asperger’s syndrome. Whereas children with Asperger’s syndrome are said to lack social skills, children with William’s syndrome seem to lack higher cognitive skills. But children with William’s syndrome, despite having social perceptual skills, still experience many difficulties in understanding the meaning of other people’s behaviour. The authors argue that “smart perception” alone does not seem to suffice in the intersubjective context; it needs these higher cognitive mechanisms to make sense of the social world (Michael, Christensen & Overgaard, 2014).

Once more this problem seems to fade when one considers the modesty of the DP claim. Something that claims to explain the case of everyday understanding should not necessarily work for clinical groups. Social cognition works differently in people with borderline personality disorder too, but that is no reason to reject the applicable theories about social cognition (Roepke, Vater, Preißler, Heekeren, & Dziobek, 2013). That there may be two different processes intertwined (a social and a cognitive), is a point that deserves attention in further development of DP, but clinical groups do not seem to pose a threat for DP’s claim when one acknowledges its modesty.

Remaining problems

Aside from these criticisms, which I think are based on misunderstandings of the modesty of DP, I will indicate two points I find problematic of DP based on Gallagher’s (2008a) paper. My first point regards the directness of our everyday social encounters and the related concern that DP has vision as its necessary modality. Intersubjective understanding is important in face-to-face contact, but equally important in contact over the telephone, via Facebook, WhatsApp and many other forms of media. To what extent can we rely on direct perception there? It seems difficult, because direct perception almost becomes pure perception in those cases. Even if contextual cues are given in, for example, a job interview over the phone, it still seems that we make a lot of explicit inferences when we interpret our potential employer.

Lavelle (2012) makes a similar point regarding the necessity of vision when she compares epistemic seeing to non-epistemic seeing. The former being seeing with informed beliefs, the latter without (resulting in seeing some white dots, or seeing the Big Bear in the sky at night). The emphasis is on seeing explicitly, perhaps because we intuitively equate seeing and perceiving. Whereas vision seems to be both necessary and sufficient for DP, auditory, tactile, scent and taste perception only seem to be supportive of DP. This seems to be because DP is grounded in an enactive, embodied and interactive view of perception. As described above, Gallagher and Varga (2014) explain specifically how we perceive intentions by borrowing concepts from action theory. Action theory says we understand intentions as intrinsic to the actions we observe. It thrives on a concept of an intentional person as “intentionally engaged with the world through actions… [that] involve an intentionality that is motoric and bodily.” (Gallagher & Varga, 2014, p. 188).

All examples given in Gallagher (2008a), Gallagher (2008b) and Gallagher & Varga (2014) are visual and it seems hard to argue we can “hear” the intentions in an action. We might be able to feel them, but this does not seem to happen in everyday situations. Gallagher and Varga (2014) seem to admit this, later stating that we can perceive emotions due to their embodied nature. An embodied nature is not something one can hear directly, only something one can hear of.  Direct perception thus only works with direct visual input.

My second point is whether direct perception can account for perceiving things we have never perceived before? It cannot, Gallagher (2008a) seems to suggest, and has to use inference in a TT or ST manner. In a later paper, prior experience is emphasized as the preventer of a constant need of inferencing (Gallagher & Varga, 2014). Gallagher and Varga (2014) describe the situation in which they have never perceived someone doing yoga before, and pose that person might not understand the situation by direct perception and might start to make deliberate inferences as a way of understanding why someone takes on such odd bodily poses. Clearly then, direct perception cannot suffice in novel situations; Gallagher seems to readily admit this shortcoming (e.g.: Gallagher & Varga, 2014).

Modesty as a strength

I have shown how some critiques of DP seem to be based on a misunderstanding of the modesty of its claim. After pointing out two major problems with the theory, the claim seems even more modest: DP cannot account for novel situations and relies on vision as its main modality. One could question what the use of a theory is when it is so modest. Can it really provide any insight at all? I think it can, but the worry is a grounded one. It depends on how much value you ascribe to a theory that accounts for only part of social understanding. DP provides a more parsimonious explanation of everyday social understanding. DP draws from embodied cognition and enactivism, uniting the framework behind interaction theory. De Jaegher (2009) writes in her reply to Gallagher (2009) and Hutto (2009), “the house [of interaction theory] is not up yet” (De Jaegher, 2009, p. 2). Later she says “The house is slowly being built.” (De Jaegher. 2009, p. 2), meaning that by elaborating on DP and related aspects of interaction theory, a full-blown reply to classical cognitivist approaches (TT and ST) is developed. Direct perception is one of the building blocks, admittedly, aimed at explaining a modest bit of our social understanding. All building blocks, modest or not, are needed to build a fundamental house that can stand the storm of cognitive approaches. DP’s modesty is its strength.



Gallagher, S. (2008). Direct perception in the intersubjective context. Consciousness And Cognition. (2). 535-543.

Gallagher, S. (2008). Another look at intentions: A response to Raphael van Riel’s ‘On how we perceive the social world.’. Consciousness And Cognition: An international Journal, 17(2), 553-555.

Gallagher, S., & Varga, S. (2014). Social Constraints on the Direct Perception of Emotions and Intentions. Topoi: An International Review Of Philosophy, (1), 185-199.

Hubel, D., & Wiesel, T. (1959.) Receptive fields of single neurones in the cat’s striate cortex. Journal of Physiology, 148, 574-591.

Jacob, P. (2011). The Direct-Perception Model of Empathy: a Critique. Review Of Philosophy And Psychology, (3), 519-540.

De Jaegher. H. (2009). Commentary: Social understanding through direct perception? Yes, by interacting. Consciousness And Cognition. 18, 535-542.

De Jaegher, H. (2009). What made me want the cheese? A reply to Shaun Gallagher and Dan Hutto, Consciousness And Cognition: An International Journal, 18(2), 549-5 50.

Lavelle, J. S. (2012). Theory-theory and the direct perception of mental states. Review Of Philosophy And Psychology. 3(2), 213-230.

Michael, J., Christensen. W., & Overgaard. S. (2014). Mindreading as social expertise. Synthese: An International Journal For Epistemology, Methodology And Philosophy Of Science, (5), 817-840.

van Riel, R. (2008). Extended Comment: On how we perceive the social world. Criticizing Gallagher’s view on direct perception and outlining an alternative. Consciousness And Cognition. 17, 544-552.

Roepke, S., Vater, A., Preißler, S., Heekeren, H. R., & Dziobek, I. (2013). Social cognition in borderline personality disorder. Frontiers In Neuroscience. 6, 195.

Sergent, J., Ohta, S., & MacDonald, B. (1992). Functional neuroanatomy of face and object

processing. A positron emission tomography study. Brain: A Journal of Neurology, 115, 15-36.

Zawidzki, T. (2012). Unlikely allies: embodied social cognition and the intentional stance. Phenomenology And The Cognitive Sciences, 11(4), 487-506.

Share your thoughts on this!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s