The Unity of Perception

Many thanks to John Schwenkler for running the Brains blog and for inviting me to guest blog this week about my new book The Unity of Perception: Content, Consciousness, Evidence (Oxford: Oxford University Press, 2018).


Perception is our key to the world. It plays at least three different roles in our lives. It justifies beliefs and provides us with knowledge of our environment. It brings about conscious mental states. It converts informational input, such as light and sound waves, into representations of invariant features in our environment. Corresponding to these three roles, there are at least three fundamental questions that have motivated the study of perception:

Epistemology question: How does perception justify beliefs and yield knowledge of our environment?

Mind question: How does perception bring about conscious mental states?

Information question: How does a perceptual system accomplish the feat of converting varying informational input into mental representations of invariant features in our environment?

To be sure, many other questions have motivated the study of perception. To list just a few: What is the nature of the perceptual relation? What is the object of perception? How does perception guide action? What is the relation between perception and thought? But the way these questions are answered hinges on what stance is taken on the three fundamental questions.[1]

The last decade has seen an explosion of work on the mind and information questions in both philosophy of mind and cognitive science. While there has been fruitful interaction between work in philosophy of mind and cognitive science, little has been done to integrate this work with issues in epistemology. Theories motivated by addressing the mind and information questions have been developed largely independently of concerns about how perception furnishes knowledge of our environment and how it justifies our beliefs. Similarly, theories motivated by addressing the epistemology question have been developed largely independently of concerns about how perception brings about conscious mental states. To be sure, most accounts of perceptual justification rely heavily on the idea that perception justifies beliefs in virtue of its phenomenal character. However, such accounts typically take it as given that perception provides evidence and immediately proceed to addressing the question of what the relationship is between such evidence and relevant beliefs. This split between philosophy of mind and cognitive science on the one side and epistemology on the other has hindered our understanding of perception. Questions in philosophy of mind are intimately connected with questions in epistemology in particular with regard to perception: the role of perception in yielding conscious mental states is not independent of its role in justifying our beliefs and yielding knowledge. If this is right, then perceptual experience should be studied in an integrated manner.

My book develops a unified account of the phenomenological and epistemological role of perception that is informed by empirical research. So it develops an account of perception that provides an answer to the first two questions, while being sensitive to scientific accounts that address the third question. In developing such a unified account, this book aims to advance a rigorous way of doing philosophy of mind on which mental states are analyzed in light of scientific evidence while being sensitive to their epistemic, cognitive, and phenomenological role. It aims to be conceptually disciplined and empirically constrained. It is written so that it should be useful to both philosophers and scientists familiar with philosophical and scientific debates about perception; as well as those unfamiliar with these debates.

The key idea is that perception is constituted by employing perceptual capacities—for example the capacity to discriminate and single out instances of red from instances of blue. Perceptual content, consciousness, and evidence are each analyzed in terms of this basic property of perception. Employing perceptual capacities constitutes phenomenal character as well as perceptual content. The primacy of employing perceptual capacities in perception over their derivative employment in hallucination and illusion grounds the epistemic force of perceptual experience. In this way, the book provides a unified account of perceptual content, perceptual consciousness, and perceptual evidence. What unifies the account are perceptual capacities. Due to the grounding role of perceptual capacities, we can call the view developed in this book capacitism.

Such a unified account of perception opens up a new understanding of the nature of perceptual content, perceptual particularity, the phenomenological basis of evidence, the epistemic force of evidence, the origins of perceptual knowledge, the relationship between content and consciousness, as well as the relationship between consciousness and reference. Moreover, it clears the way for solving a host of open-standing problems, such as the relation between attention and perceptual knowledge, the linguistic analysis of perceptual reports, the relation between acquaintance and awareness, the rational role of perceptual experience, and the perceptual basis for demonstrative reference.

One larger aim of this book is to bring back mental capacities as a way of analyzing the mind. Despite their prominence in the history of philosophy, capacities have been neglected in recent philosophical work. By contrast, appeal to mental capacities is standard in cognitive psychology and the brain sciences. This book develops the notion of capacities in light of empirical work in cognitive psychology, neuroscience, and developmental psychology. While it is based in contemporary empirical sciences, it also harks back to a long tradition of analyzing the mind in terms of capacities. It turns out that we can use contemporary insights and tools to modernize that tradition.

Analyzing the mind in terms of capacities has many advantages, so it is surprising that recent philosophy has forsaken this approach. One central advantage is that it allows for a counterfactual analysis of mental states on three interrelated levels. On a first level, we focus on the function of mental capacities. On a second level, we focus on the mental capacities employed irrespective of the context in which they are employed. On this level, the focus is on what perception and corresponding cases of hallucination and illusion have in common. On a third level, we focus on the mental capacities employed, taking into account the context in which they are employed. On this level, the focus is on the difference between cases in which a capacity fulfills its function (perception) and cases in which it fails to fulfill its function (hallucination and illusion). I explain these terms in more detail in Chapter 2.

Let me locate capacitism within the wider philosophical landscape.

First, capacitism grounds mental states, consciousness, evidence, and content in the physical, non-mental world. As such, these features of the mind are no less amenable to scientific investigation than any other features of the world. In developing such a naturalist and physicalist view of perception, this book shows how perception is our key to the world while situating perception within that world.

Second, capacitism is an externalist account of perceptual content, consciousness, and evidence. It is an externalist account since the perceptual capacities that constitute these features of the mind function to discriminate and single out particulars in our environment. Due to their singling out function, perceptual capacities connect us to our environment. While capacitism is an externalist view, it is one that that does justice to the internalist elements of perceptual experience. In contrast to, say, orthodox versions of reliabilism, it makes room for conscious mental states that play a cognitive and epistemic role in our lives. Moreover, the capacities employed in perception can be employed derivatively in hallucination and illusion. While they do not fulfill their function when employed in hallucination and illusion, the capacities nonetheless function to discriminate and single out particulars, thereby providing a relation to how things would be were they to fulfill their function. By doing justice to the internalist elements of perceptual experience, capacitism is a modestly externalist view.

Third, capacitism is a common factor view of perception. The same perceptual capacities can be employed in perception, hallucination, and illusion. As such, they can constitute a metaphysically substantial common element. This common element grounds not only the common representational content of perceptions, hallucinations, and illusions with the same phenomenal character (Chapter 4 and Chapter 5), but also perceptual consciousness (Chapter 6) and phenomenal evidence (Chapter 7 and Chapter 8). Thus, capacitism is at its core non-disjunctivist.

Fourth, capacitism is an asymmetric account of perception, hallucination, and illusion. While, it is a non-disjunctivist account, it nonetheless holds that perception is metaphysically and explanatorily more basic than hallucination and illusion. Perception is more basic since perceptual capacities function to do what they do in perception, namely discriminate and single out particulars, and they have this function, even when employed in hallucination or illusion. So while they can be employed in hallucination and illusion, they are employed derivatively.

So capacitism walks a path between two traditional views: the common factor view and austere relationalism. The common factor view posits that a perception, hallucination, and illusion with the same phenomenal character share a common element that grounds their shared phenomenal character. Typically, the additional condition that makes for successful perception is considered to be a causal relation between the experiencing subject and the perceived object. (This approach is analogous to the epistemological view that knowledge can be factorized into belief and some additional element, say, justification.) By contrast, austere relationalism characterizes hallucination in terms of a deficiency of perception and argues that perception and hallucination do not share a common element. (This approach is analogous to the view that mere belief is to be analyzed as deficient of knowledge, but subjectively indiscriminable from it.)

With common factor views, I argue that perceptions, hallucinations, and illusions with the same phenomenal character share a metaphysically substantial common factor that grounds their phenomenal character. The common element is constituted by the perceptual capacities employed. But with austere relationalists, I argue that hallucinations and illusions can be understood only in terms of a deficiency of perceptions: there is an asymmetric dependence of the employment of perceptual capacities in hallucination and illusion on their employment in perception. I show how this way of understanding the metaphysical structure of perceptual experience can be exploited for a view of perceptual evidence.

The book has four parts. Each part develops a component of my positive account. In the rest of this week, I will devote each day to one of the four parts.


[1] No doubt, dependencies run in the other direction as well: what stance one takes on, say, the question of what the object of perception is, will affect one’s answer to the three fundamental questions. But arguably there is an asymmetry in the order of explanation between the three fundamental questions and the other questions.


Header image: Gerhard Richter, Abstraktes Bild, See

13 Comments

  1. Sounds like a fascinating book. I’d suggest that one shouldn’t separate perception from action. The utility of the unification of perception is questionable; the utility of the unification of action not so much — one can only be in one place at a time, with limits of 2 hands, etc. In brief, I feel that perception should not be separated from action; think perceptual-motor. I feel the separation of perception from action is a chronic problem in philosophy, psychology and neuroscience (I come from neuroscience).

    • Scott Wagner

      To be a bit more pointed than the author in her response to you: while one should argue as you do that antiseptically separating action and perception is essentialist and distorting, one simply must get perception right before integrating it appropriately with action, to connote a conjoined ‘sensori-motor’ system. Without her a priori synthesis of the partially causal perceptual engine, it’s impossible to get where you want to with a properly situated, synthesized account of action.

      • Scott Wagner: “One must get perception right”. Well, if you accept that perception is axiomatic, yes. My argument is empirical — and largely untested. That is, when we understand perception and action better, at brain and behavior levels, they are going to be hugely interconnected. Forcing the separation will delay this understanding. The philosophical mumbo-jumbo is irrelevant to an empirical question where at present we don’t have the data.

  2. Dear John Kubie, thanks for checking in and reading my post. I agree that perception and action intersect and interact in complex ways. And I agree with you that the ways they intersect and interact has not been given enough attention in philosophy, psychology, and neuroscience. However, while they interact in complex ways, I wouldn’t say that they can’t be distinguished. I wrote some articles on the interaction between perception and action. In this book, however, I don’t focus on this aspect of perception.

    With “Unity of Perception” I mean to capture the fact that perception plays multiple roles in our lives and that an account of perception should explain these different roles (ideally) in a unified manner. The unifying element on my view are the perceptual capacities we employ. I argue that the content, phenomenal character, and epistemic force of perception are all grounded in employing perceptual capacities.

  3. Ram Neta

    Excellent blog post: I’m looking forward to more discussion of this book!
    But I’m wondering about the following thing. As I understand it, the capacitist says that the phenomenally conscious features of our perceptual states just consist in the ways in which our perceptual capacities are exercised in those states. But also, as I understand it, the capacitist says that the epistemic warrant provided by our perceptual states just consists in the ways in which our perceptual capacities are exercised in those states. Now, if she holds both of those views, then won’t she thereby be committed to the claim that the epistemic warrant provided by our perceptual states just consists in the phenomenally conscious features of our perceptual states? Accessibillists like me will find the latter claim attractive, but epistemological externalists will not: they will want to say that the epistemic warrant provided by a perceptual state can’t just be read off its phenomenal properties.

    • Dear Ram, thanks for taking the time to drop in and read my post! And thanks for the kind words.

      You’re right that the capacitist argues that phenomenal character is constituted by employing perceptual capacities (I’ll blog about consciousness on Thursday). And you’re right that the capacitist argues that the epistemic force of perceptual states stems from properties of perceptual capacities employed (I’ll blog about evidence on Friday). But despite endorsing those two claims, the capacitist doesn’t hold that the epistemic warrant provided by a perceptual state can simply be read off its phenomenal properties.

      The reason is that there is more to perceptual evidence than is reflectively accessible. I distinguish between phenomenal evidence and factive evidence.

      Phenomenal evidence corresponds to how the environment seems to one. It is constituted by the perceptual capacities employed irrespective of what environmental particulars if any they discriminate and single out.

      Factive evidence is determined by the environment to which we are perceptually related such that the evidence is guaranteed to be an accurate guide to the environment. It is constituted by the perceptual capacities employed and the particulars thereby singled out.

      I argue that when we accurately perceive our environment, we have both factive evidence and phenomenal evidence. When we hallucinate, we have only phenomenal evidence.

      So when we perceive we have more and better evidence than when we hallucinate. Thus, the capacitist doesn’t hold that the epistemic warrant provided by a perceptual state can simply be read off its phenomenal properties.

      However, factive evidence is not reflectively accessible to the perceiver (and so I’m not an epistemological disjunctivist). Since factive evidence is not reflectively accessible, the perceiver cannot tell that she is in a better epistemic position than when she perceives. (Thanks to Eva Schmidt for an excellent talk on my book at a workshop in Zürich last week that helped me think about the differences between capacitism and epistemological disjunctivism more clearly.)

      I argue that the epistemic force of perceptual experience stems from an asymmetric dependence of the employment of perceptual capacities in hallucination and illusion on their employment in perception. The rational source of both phenomenal and factive evidence lies in employing perceptual capacities that function to discriminate and single out particulars.

  4. Scott Wagner

    I really like the emphasis on capacity, the same way I hate it when other fields of the mind ignore its contours and limitations when deciding what’s going on. We’re this strange amalgam of rudimentary and complex, and this approach honors that elegantly, as a kind of error handling algorithm, if you will. I’d note in passing that your assurance of the relatively greater incorporation of capacity in neuroscience, psychology etc. is more than a little optimistic, that versions of what you’ve done need mirroring elsewhere.

    My reference pt is mostly psych and cognition. The word ‘illusion’: are you lumping Kahneman’s (pre-action/response) bias work in as illusion, or do you see that as all interpretive, and antécédent to perception? Is illusion in your sense just getting facts wrong in call it raw perception- where does perception end and interpretation begin in your framing?

    Also, does capacity have both volume and type/qualitative implications- if so, how do those two aspects impinge differentially? My thoughts went straight to the extreme limits of throughout as we get closer to conscious evaluation.

    Thanks for the great work and clear writeup!

    • Dear Scott Wagner,

      Thanks for stopping by. And thanks for raising these excellent questions.

      The classic cases of illusion, such as, the Müller-Lyer illusion or the Titchener circles illusion, a subject perceives objects in her environment, but perceives them to have at least one property they do not in fact have, due to for example a bias of the perceptual system.

      I wouldn’t lump Kahneman’s (pre-action/response) biases as illusion, since at least some of those biases are cognitive rather than perceptual biases.

      That said, while there are clear cases of perception (seeing a shaper) and clear cases of cognition (thinking about the future), there is likely to be a fuzzy boundary between perception and cognition. One thing that is attractive about capacitism is that it can easily account for such a fuzzy boundary. Perceptual capacities can be more or less high-level. Moreover, perceptual capacities can have properties of cognitive capacities, such as concepts. For example, they can involve classifying things.

      I address these issues in more detail in the book.

      Lots to think about here.

  5. Bebhinn Donnelly-Lazarov

    Great arguments. I have a few queries relating to what you mean by ’employing’ or ‘exercising’ perceptual capacities. I have in mind the difference between looking around to see that the road ahead is clear and just being visually aware of the road. In the former case a capacity is certainly activated, employed, or exercised but the same is less clear in the latter case. Moreover my capacity to see, say, might also be ‘activated’ other than through my use of it; by some form of direct brain stimulation, for example. So, it looks like the capacity can be active, passive, or acted on, with perception present in each case. It may be that my concept of what it means for a capacity to be active is way off the mark here but I think there is something importantly different about a perception being actively ‘used’ in the sense I describe. In particular what it means to be conscious of perception in action may be something other than what it means to be conscious of perception. I tend to think that consciousness of our capacities is constituted by those capacities (and so that your mind question needs to defend the ‘bringing about’ of consciousness bit). Seeing the road is knowledge / being conscious (taking no form other than the seeing itself) that I see the road. (It just happens, because seeing is a perceptual capacity, that it also, usually, gives us a different kind of knowledge too – i.e. knowledge that the road is there.) Just some thoughts – looking forward to the book.

  6. Dear Bebhinn Donnelly-Lazarov,

    Thanks for taking the time to read and think through my post so carefully!

    I agree with you that a perceptual capacity can be activated by some form of direct brain stimulation or in some other way that does not involve any significant agency on the part of the perceiver. When I speak of employing a perceptual capacity I don’t mean to imply that any significant agency on the part of the perceiver is involved. A capacity can be employed due to brain stimulation. A capacity can be employed due to causal impact from the environment, for example, being perceptually related to a red apple that prompts one to employ the capacity to discriminate and single out red things. A capacity can be employed due to a combination of factors, such as, one being perceptually related to a hummingbird and paying close attention to the bird and all its beautiful colors.

    So I would say that in both cases you describe, one is employing perceptual capacities. But I agree with you that in the case in which one is looking around to see that the road ahead is clear one is employing a capacity in more attentive way.

    I definitely do not want to say that when we employ perceptual capacities we are conscious of employing those perceptual capacities (though perhaps in rare cases we could be conscious of employing perceptual capacities).

    If I’m having a conversation with a friend while walking down the street, I will employ lots of perceptual capacities that allow me to spot the cars and pedestrians I need to dodge, while barely paying attention to those cars and pedestrians. My attention is directed at my friend and our conversation. So I will employ only very course-grained perceptual capacities to discriminate and single out the cars and pedestrians I am passing. I am likely to barely register the colors of the cars. By contrast, I will employ significantly more fine-grained perceptual capacities to discriminate and single out the facial expressions of my friend with whom I am talking.

    On Thursday, I’ll blog about my view of consciousness. In a nutshell the idea is that perceptual consciousness is constituted by a mental activity, namely the mental activity of employing perceptual capacities. But here again, I don’t mean that this mental activity is one of which one is conscious.

    Thanks again for raising these important questions!

Comments are closed.