Perceptual content and relations

Philos Stud DOI 10.1007/s11098-012-0069-3 Perceptual content and relations Susanna Schellenberg Ó Springer Science+Business Media Dordrecht 2012 Bill Fish’s book Perception, Hallucination, and Illusion is a rich, sweeping, and radical development of naı̈ve realism.1 At the core of his view is the idea that in perception we are acquainted with our environment and that in virtue of being so acquainted our experience has phenomenal character. There are many radical ideas in the book, but perhaps the most radical is the idea that when we hallucinate, we have an experience that does not have phenomenal character: we mistake it to have phenomenal character and in virtue of this mistake a hallucination can be subjectively indistinguishable from a perception. The idea that hallucinations do not have phenomenal character is hard to swallow but as Fish argues, it follows from two ideas: (1) (2) the naı̈ve realist idea that phenomenal character is determined by the very scene we are acquainted with. the idea that in hallucination we are by definition not perceptually related to the scene we seem to be perceiving and so not acquainted with that scene. Fish understands the phenomenal character of an experience as its property of acquainting the subject with particular mind-independent elements of the environment. A second radical idea of Fish’s book is the idea that experience does not have content. I will focus on this thesis. I will start by getting clearer on what the thesis is that Fish rejects. Let us call the Content Thesis, the thesis that perceptual experience is fundamentally a matter of representing the world as being a certain way. With few interludes, this thesis has been orthodoxy in philosophical views about perceptual 1 Fish (2009). S. Schellenberg (&) Department of Philosophy, Rutgers University, New Brunswick, NJ, USA e-mail: [email protected] 123 S. Schellenberg experience since Kant (1781). In his famous Stufenleiter passage, Kant categorizes different kinds of representations: ‘‘The genus is representation in general (representatio). Subordinate to it stands representation with consciousness (perceptio). A perception which relates solely to the subjects as the modification of its state is sensation (sensatio), an objective perception is knowledge (cognitio)’’ (A320/B377) and categorizes perception as a representation with consciousness. Fish is not the first to question the content thesis. The thesis has been questioned by Reid (1764), by sense-data theorists, such as Russell (1913), Price (1950), Moore (1953), and most recently by austere relationalists, such as Campbell (2002), Martin (2002), Travis (2004), Brewer (2006), and Fish (2010). However although Fish is not the first and the only to question the thesis, he may well be the philosopher who has rejected the content thesis most categorically. To get to the core of how he rejects the Content Thesis, it will be helpful to distinguish it from two closely related theses. Let us call the Association Thesis, the thesis that every experience can be associated with (propositional) content in the sense that sentences can be articulated that describe how the environment seems to the subject. Let us call the Awareness Thesis, the thesis that the experiencing subject stands in an awareness relation to the content or its constituents, such that this awareness relation grounds the sensory character of the experience. The Association Thesis is so weak that it could arguably be endorsed by anyone, regardless of what other commitments they might have. It does not say that we can fully describe sensory states by expressing propositions. It states only that we can express propositions that capture at least some aspect of the sensory state of an experiencing subject. As one would expect, Fish does not deny the Association Thesis. The Awareness Thesis is a controversial version of the Content Thesis. The Content Thesis does not entail the Awareness Thesis and so we can accept the Content Thesis without accepting the Awareness Thesis. Fish denies both the Content Thesis and the Awareness Thesis. More specifically, he endorses the negative thesis that no appeal to representational content is necessary in a philosophical account of perceptual experience, in conjunction with the positive thesis that any perception essentially involves at least three components: a subject, the environment of the subject, and a perceptual relation between the subject and certain elements of her environment. One set of questions I have for Fish is what exactly the negative thesis amounts to: is the idea simply that no appeal to content is necessary to give a good account of perceptual experience? Is the idea that a representational view is less attractive than a relational view in explaining certain phenomena? Or is the idea that a representational view cannot explain certain phenomena? Fish rejects the Content Thesis in part on grounds of rejecting what he calls the local supervenience principle: ‘‘if the proximal neural conditions that occurred when one perceived an object were created in the absence of that object, the subject would have an experience with the same phenomenal character as the original perceptual experience’’ (p. 41). Fish does not aim to show that local supervenience is false. His aim is rather to show that ‘‘many of the reasons given for thinking it to be true are not as compelling as they are often assumed to be’’ (p. 42). 123 Perceptual content and relations There are many reasons for thinking that experience has content. However, many do not establish more than the Association Thesis. Elsewhere I have argued for the Content Thesis along the following lines: P1 P2 P3 Conclusion 1 P4 P5 Conclusion 2 If a subject S is perceptually related to her environment (while not suffering from blindsight or any other form of unconscious perception), then she is sensorily aware of her environment. If S is sensorily aware of her environment, then her environment sensorily seems a certain way to her. If S’s environment sensorily seems a certain way to her, then she has an experience with content C, where C corresponds to the way her environment sensorily seems to her. If S is perceptually related to her environment (while not suffering from blindsight or any other form of unconscious perception), then she has an experience with content C, where C corresponds to the way her environment sensorily seems to her. S’s environment is either the way it sensorily seems to her or it is different from the way it sensorily seems to her. If S has an experience with content C, then C is either accurate (if her environment is the way it sensorily seems to her) or inaccurate (if her environment is not the way it sensorily seems to her). If a subject is perceptually related to her environment (while not suffering from blindsight or any other form of unconscious perception), then the content of her experience is either accurate or inaccurate.2 I called this the Master Argument for Perceptual Content. Why should we accept this argument? The first two premises are arguably not contentious. After all, it is not contentious that if we are perceptually related to our environment, then we will be sensorily aware of that environment. Moreover, it is not contentious that if we are sensorily aware of our environment, then our environment will sensorily seem a certain way to us. But why should we accept the third premise? We can recognize P3 to be true, if we recognize that there is a link between seemings and contents, more specifically the idea that the content of experience corresponds to the way the world seems to the experiencing subject given one or more sensory modes of experience. If we recognize this seems-content link, then the idea that the environment can seem a certain way to a subject without her being in a contentful mental state becomes impossible. But the idea that the environment can seem a certain way to a 2 My (2011). 123 S. Schellenberg subject without her being in a contentful mental state is precisely the idea that austere relationalism relies on. Recognizing the seems-content link is compatible with accepting that any given scene can be perceived in many different ways and it is compatible with accepting that any given perception can be articulated in many different ways. The way the environment seems to the perceiver may change from moment to moment even as her gaze remains steady. One or more proposition can be associated with every one of these particular ways that her environment may seem to her. All these propositions are equally legitimate. Nevertheless, at any given moment the world will seem to her to be a certain way. This is all that we need to establish the seemscontent link. Once one has recognized the seems-content link, only minor further commitments are necessary to establish that the way the environment seems to an experiencing subject is assessable for accuracy. In support of the fourth premise, we can say that in virtue of a subject perceiving the environment, it seems a certain way to her. The way the environment seems to a subject determines the way the environment would have to be for the content of her experience to be accurate. The environment is either the way it seems to her or it is different from the way it seems to her. In support of the fifth premise, we can say that if the environment is the way it seems to her, then the content of the experience is accurate. In all other cases, the content of the very same experience is inaccurate. So if a subject has an experience with a particular content, then this content is either accurate or inaccurate. It follows from this, together with the first two premises of the Master Argument, that if a subject is perceptually related to the environment, then the way the environment seems to her is assessable for accuracy. Together with P3, it follows that if a subject is perceptually related to the environment, then the content of her experience is either accurate or inaccurate. Now the Master Argument does not on its own establish the Content Thesis. After all, it makes no claims about whether experience is fundamentally a matter of representing. On a weak reading of the Master Argument, it establishes only the Association Thesis. So I assume that Fish would happily accept this argument. What he and other austere relationalists reject is not that some content can be associated with perceptual experience. They reject that experience is fundamentally a matter of representing. So in order to establish the Content Thesis, we need an additional argument. The following Argument for Relational Content is such an additional argument: P6 P7 123 If a subject S’s environment sensorily seems a certain way to her, then S is employing discriminatory, selective capacities that constitute the way her environment sensorily seems to her. If S is employing discriminatory, selective capacities that constitute the way her environment sensorily seems to her, then S is representing her environment in virtue of employing those discriminatory, selective capacities. Perceptual content and relations P8 Conclusion 3 P9 P10 Conclusion 4 If S is representing her environment in virtue of employing discriminatory, selective capacities, then S has a perceptual experience that is fundamentally a matter of representing her environment as being a certain way. If S’s environment sensorily seems a certain way to her, then S has a perceptual experience that is fundamentally a matter of representing her environment as being a certain way. If S is representing her environment in virtue of employing discriminatory, selective capacities, then (due to the nature of these capacities) relations to particulars are implicated in the very nature of her representational content. If relations to particulars are implicated in the very nature of S’s representational content, then S’s perceptual experience is fundamentally a matter of being related to mind-independent particulars in the environment. If S’s environment sensorily seems a certain way to her, then S’s perceptual experience is fundamentally a matter of being related to mind-independent particulars in the environment. The idea motivating premises 6–8 is that when we perceive we employ discriminatory, selective capacities by means of which we single out particulars in our environment. The relevant particulars are external and mind-independent objects, events, property-instances, and instances of relations. If we possess such a discriminatory, selective capacity, then we can employ it to single out the very type of particulars that the capacity is of. What happens when we hallucinate? Although the very nature of discriminatory, selective capacities is determined by functional connections between perceivers and their environment, there is no reason to think that a discriminatory, selective capacity could not be employed when one is misperceiving or hallucinating. After all one could be prompted to employ the discriminatory, selective capacities due to all sorts of nonstandard circumstances such as unusual brain stimulation or misleading distal input. If this is right, then we can employ a discriminatory, selective capacity even if a relevant particular is not present. When we hallucinate, we employ the very same discriminatory, selective capacities that in a subjectively indistinguishable perception are employed as a consequence of being causally related to external, mind-independent particulars. But since we are hallucinating, we fail to single out any particular in our environment. As a consequence, the discriminatory, selective capacities employed are baseless. They are baseless for reasons analogous to the reasons for why a concept is empty when we fail to refer. 123 S. Schellenberg What do I mean with discriminatory, selective capacities? The notion of a discriminatory, selective capacity refers to a low-level mental capacity by means of which we differentiate, single out, and in some cases type particulars.3 For example if we possess the discriminatory, selective capacity that functions to differentiate and single out red, we are in a position to differentiate instances of red from other colors in our environment and so to single out instances of red. More generally, to possess a discriminatory, selective capacity is to be in a position to differentiate and single out the type of particulars that the capacity is of, were one related to such a particular. So if we possess such a capacity, then the following counterfactual should hold: if we were perceptually related to a particular that the capacity functions to single out, then we would be in a position to single out such a particular. Like concepts, the capacities in play can be employed in different environments and in this sense are repeatable. The nature of sensory states can be understood in terms of employing discriminatory, selective capacities in a sensory mode. The relevant sensory modes are modes such as seeing, hearing, touching, smelling, or tasting. Any two experiences in which all the same discriminatory, selective capacities are employed in the same sensory mode will constitute the same sensory state. If this is right, then subjectively indistinguishable hallucinations and perceptions will share a common element, namely the discriminatory, selective capacities that the subject employs in a sensory mode. In order to give support to the ninth and tenth premises, it is crucial to take a closer look at discriminatory, selective capacities in play. The function of discriminatory, selective capacities is to differentiate and single out the type of particulars that the capacity is of. It would be unclear what it would mean to possess a discriminatory, selective capacity, the very function of which is to single out a type of particular, without being in a position to single out such a particular when perceptually related to one. If this is right, then discriminatory, selective capacities are by their nature linked to what they of in the good case. For example, if we possess the capacity to discriminate and single out instances of red from other colors, we can use this capacity to single out instances of red in our environment. Were we not in a position to use our capacity in this way, when perceptually related to an instance of red, we would not count as possessing the capacity. If we integrate the Argument for Relational Content with the Master Argument, then we can recognize that perceptual experience is fundamentally representational, insofar as perceptual content is yielded by employing discriminatory, selective capacities and the discriminatory, selective capacities constitute the experience. If perceptual content is determined by employing discriminatory, selective capacities and such capacities function to single out particulars, then relations to particulars are implicated in the very nature of perceptual content. So perceptual experience is fundamentally relational insofar as the content of experience is yielded by 3 The notion of discriminatory, selective capacities in play is well based in cognitive psychology. See in particular the work on basic visual capacities and pre-attentive discrimination in early vision, see Julesz (1981), Krummenacher et al. (2010), and To et al. (2011). 123 Perceptual content and relations employing discriminatory, selective capacities and such capacities are by their nature linked to what they of in the good case. So insofar as discriminatory, selective capacities function to single out particulars, perceptual experience is fundamentally both relational and representational. I do not see on what grounds Fish could not accept a view of the kind, given how he understands experience. After all, he will presumably allow that low-level discriminatory, selective capacities are in play when we are aware of our environment. Moreover, he will presumably allow that we can employ the very same discriminatory, selective capacities in different environments. Once one has accepted that, it follows almost immediately that experience has content. References Fish, W. (2009). Perception, hallucination, and illusion. New York: Oxford University Press. Julesz, B. (1981). A theory of preattentive texture discrimination based on first-order statistics of textons. Biological Cybernetics, 41, 131–138. Kant, I. (1781). Kritik der reinen Vernunft [Critique of pure reason] (N. K. Smith, Trans.). New York: St. Martin’s Press (1965). Krummenacher, J., Grubert, A., et al. (2010). Inter-trial and redundant-signals effects in visual search and discrimination tasks: Separable pre-attentive and post-selective effects. Vision Research, 50, 1382–1395. Schellenberg, S. (2011). Perceptual content defended. Noûs, 45, 714–750. To, M. P., Gilchrist, I. D., et al. (2011). Discrimination of natural scenes in central and peripheral vision. Vision Research, 51, 1686–1698. 123