Conceptualism Can’t Account for the Phenomenology of Hallucination

The argument from fineness of grain is probably the most discussed argument for nonconceptualism. (To name but a few discussants: Peacocke 1998, 2001a, 2001b; McDowell 1994, 1998, Brewer 1999, 2005, Tye 2005, Coliva 2003, Kelly 2001a, 2001b, Veillet 2014.) To account for the fine-grained phenomenal character of visual experience in terms of the concepts of the perceiver, conceptualists appeal to demonstrative concepts. So even though a subject doesn’t possess the general concept teal, say, or even the concept shade, she will be able to perceptually represent a teal dress. Two subjects who possess different concepts for the shade of the dress, for instance teal and blue-green (or this color and this shade) can still have visual experiences of the dress with the same content: this is thus. This conceptualist strategy, of capturing the fine-grained content of experience with the help of bare demonstrative concepts, can be called the ‘pure demonstrative strategy’.

I believe that this is the best strategy available to the conceptualist to counter the argument from fineness of grain. As Brewer (2005) argues, we can say that the demonstrative involved in the subject’s experience determinately picks out certain properties – for example the shade of the dress rather than its shape – because the subject can track this property and focus her attention on it. There is an attentional and tracking relation that underlies the demonstration to the shade and fixes the corresponding perceptual content. I believe that even in cases of misperception this attentional and tracking relation can help to provide a determinate perceptual content. Imagine that I am looking at a yellow dress, but to me the dress looks to be teal, where the content of my experience is this is thus. In such a case, the conceptualist might insist that the reference of the demonstrative thus involved in this purely demonstrative experience is to the shade that this demonstrative concept would track under normal conditions.

But there is a further problem lurking here, which is not about reference or determinate content, but about the phenomenology of misperception. Compare a veridical experience with a case of hallucination. In one situation, I see a teal dress lying before me; in the other I’m in a pitch-black room, but hallucinate that there is a teal dress lying in front of me. Now in the veridical situation, the conceptualist can give a very appealing account of the phenomenal character of my visual experience. She can say that I attend to the teal dress and it is the teal of the dress itself that constitutes the feel of my experience (and the shape of the dress etc.). When I hallucinate the teal dress, however, there is no teal dress present that could be the basis of my teal phenomenology. Moreover, the teal of the dress that my demonstrative that would track under normal circumstances doesn’t help. In the hallucination case, I’m in complete darkness, and the conceptualist can’t say that the teal that my demonstrative would normally track is present somehow to provide the phenomenal character. Or if she does so, it looks like she introduces a level of nonconceptual content after all. For it is then some uninstantiated property, not a concept, that constitutes the phenomenal content of the experience.

Alternatively, the conceptualist might insist that it is the modes of presentation of our pure demonstrative concepts that give rise to the phenomenal character of experience (cf. Brewer 1999, 156). Since the same demonstrative thus is involved in veridical perception and in hallucination, it can be insisted that they thereby have the same content and phenomenal character.

Here is just one problem with this strategy – it conflicts with the transparency of perceptual experience. When looking at the teal dress, it doesn’t strike me as thought it is a mode of presentation, a feature of my experience itself, that provides the teal-y feel. Rather, when I focus on my experience of the teal dress, it is the dress itself whose particular shade provides the phenomenal character of my visual experience.

Another option for the conceptualist would be to endorse direct realism / metaphysical disjunctivism (cf. Martin 2004/2009). There is a version of this view on which veridical perceptions and corresponding indiscriminable hallucinations don’t have the same phenomenal character. The subject just mistakenly thinks that they have the same phenomenology – she is unable to distinguish them. The phenomenal character of veridical perception is constituted by the objects and properties to which it relates the perceiver; since in hallucination, this relation is lacking, it doesn’t have the same phenomenal character. This move, however, is incompatible with the conceptualist claim that veridical perception and hallucination have the same conceptual content or, otherwise put, are identical appearances that the world is a certain way.

So, to do justice to the phenomenology of hallucination, the conceptualist has to abandon the view that perceptual content is conceptual – either in favor of nonconceptualism or in favor of direct realism, thereby rejecting the content view altogether. (Notably, this was done by Brewer 2006.)


Brewer, B. (1999), Perception and Reason, Clarendon Press, Oxford.

Brewer, B. (2005), ‘Perceptual Experience has Conceptual Content’, in M. Steup & E. Sosa, eds, Contemporary Debates in Epistemology, Blackwell, Malden, pp. 217–230.

Brewer, B. (2006), ‘Perception and Content’, European Journal of Philosophy 14, 165–181.

Coliva, A. (2003), ‘The Argument from the Finer-Grained Content of Colour Experiences: A Redefinition of Its Role Within the Debate Between McDowell and Non-Conceptual Theorists’, Dialectica 57, 57–70.

Kelly, S. (2001a), ‘Demonstrative Concepts and Experience’, Philosophical Review 110, 397–420.

Kelly, S. (2001b), ‘The Non-Conceptual Content of Perceptual Experience: Situation Dependence and Fineness of Grain’, Philosophy and Phenomenological Research 62, 601–608.

Martin, M. (2004/2009), ‘The Limits of Self-Awareness’, in A. Byrne & H. Logue, eds, Disjunctivism: Contemporary Readings, MIT Press, Cambridge, MA, 271–317.

McDowell, J. (1994a), Mind and World, Harvard University Press, Cambridge, MA.

McDowell, J. (1998), ‘Reply to Commentators’, Philosophy and Phenomenological Research 58, 403–431.

Peacocke, C. (1998), ‘Nonconceptual Content Defended’, Philosophy and Phenomenological Research 58, 381–388.

Peacocke, C. (2001a), ‘Does Perception Have a Nonconceptual Content?’, Journal of Philosophy 98, 239–264.

Peacocke, C. (2001b), ‘Phenomenology and Nonconceptual Content’, Philosophy and Phenomenological Research 62, 609–615.

Tye, M. (2005), ‘On the Nonconceptual Content of Experience’, in M. Reicher & J. Marek, eds, Experience and Analysis: Proceedings of the 27th International Wittgenstein Symposium, öbv&hpt, Vienna, 221–242.

Veillet, B. (2014), ‘Belief, Re-Identification and Fineness of Grain’, European Journal of Philosophy 22, 229–248.


  1. Hello Eva — I look forward to reading your book — I’ve only had time to skim through it for now. I just wonder about two things: (1) you seem to get much more out of the state/content distinction than I or others have. Actually, as I mention in my ’12 book, many find the distinction unhelpful and avoid or ignore the “state” view entirely, e.g. even Heck himself, Bermudez. Further, speaking about “states” as containing concepts is puzzling to me (as opposed to contents, propositions, etc.), or perhaps the state view just reduces to the content view in the end. Why exactly does the state view prove so useful in your argument? If (2) Re: fineness of grain. As you know, I spend quite a bit of time on this argument in my book (pp. 173-180) and, although I defend conceptualism, I am not a real big fan of the demonstrative concept reply. For now — re: shades of color: Why isn’t possessing, say, the general concept LIGHT RED along with the concepts DARKER or LIGHTER good enough to handle the usual paint chip cases in question? Rosenthal also talks about these “comparative concepts.” I say this despite the fact that I also don’t think that the “re-identification” condition is beyond dispute. Perhaps comparative concepts could even explain why it is difficult to re-identify a very specific shade of color in the paint chip case, i.e. the simultaneous comparative shade difference can be experienced but each shade individually over time might not. Something similar might apply to shapes, sizes, etc…

    • Eva Schmidt

      Hi Rocco, thanks for taking the time to respond to my posts!

      (1) state vs. content view
      As far as I can see, in reaction to the distinction, philosophers either limit their views to state nonconceptualism (Sainsbury & Tye 2012, Crane 2008) or work on a way of getting from states to contents (Bermúdez 2007, Heck (2007), Toribio (2008), (2011) , Van Cleve (2012)). As I argue in the book, the arguments of this latter group of philosophers are not fully convincing.

      (2) fineness of grain
      You’re right that my criticism in this post is addressed only at proponents of the pure demonstrative strategy. Actually, I’m not sure what conceptualists who say that GENERAL concepts are constitutive of perceptual experience and its content would say about the phenomenal character of hallucination. Maybe you have some suggestions?
      I’m on board with abandoning the re-identification condition. I think that we have some demonstrative concepts at least that we can employ in thought, which we possess only in certain perceptual situations (and lose them immediately after leaving those situations).
      However, I don’t think that an appeal to comparative concepts works. One of the problems of this strategy is the following: If it is granted that the subject possesses and exercises only a comparative concept for the two very similar shades (as Brewer 2005 does), then what is the concept that is involved in an experience of just one of the shades all by itself? It can’t be a comparative concept; but then what is the content of my experience of the one shade all by itself? And what does this experience have in common with my experience of the same shade in the comparative experience?
      Thanks again for your comment – I’d be happy to hear more of your thoughts on these issues.

      • Hi Eva — On the comparative concepts, it may very well be that our phenomenology is not quite the same when comparing seeing a color alone as opposed to two similar colors at the same time (especially if we don’t require “re-identification”). The same might be said for some shapes (n-sided figures), pictures of similar animals (alligator/crocodile), etc. But the comparative concepts would only be applied together with another (more general) possessed concept, e.g. DARK RED. So I may only see RED27 “as” a kind of DARK RED when presented alone, but perhaps “as” RED27 when simultaneously presented with RED28. This is oversimplifying a bit from my ’12 book but seems plausible to me.

        On conceptualism/hallucination, I’d have to think about this a bit more. I’m not a direct realist, though, and I don’t see why a conceptualist cannot hold that whatever it is (e.g. drugs) that causes one’s hallucinatory conscious experience in a given case, the experience (and content) are constituted by one’s already possessed concepts. Maybe it’s more like a vivid dream experience which would seem to require the application of already possessed concepts. And importantly, as you mention above, some of the worries you articulate do presuppose the demonstrative strategy and/or direct realism, both of which I reject.

        Best — Rocco

  2. Eva Schmidt

    Hi Rocco,

    On the first point, I’ll have to re-read the relevant passage from your book and think about it some more. One worry that comes to mind is that more than one concept is employed in the experience for the very same shade, the way you describe it. But this can’t be right since the shade is visually present to me only in one way.

    As to the second point, I have the following questions: How come concepts employed in experience give rise to this particular perceptual phenomenal character, but when the very same concepts are employed in a belief, they don’t? Why isn’t belief (mistaken belief or belief about things that I’m not confronted with) phenomenally just like perceptual hallucination?


    • Hello Eva — I guess I don’t see why more than one concept can’t be employed in order for one to experience a single shade of color. Of course, merely experiencing one shade of color is unusual and much simpler than other perceptual experiences, e.g. of a colored moving object of a given shape… That is, two or more concepts can play a combined role in having a single perceptual experience.

      Beliefs don’t have a sensory first-order component, at least in the same way that perceptual states do. Of course, in the case of perceptual hallucinations, there is no outer (or typical) cause of the first-order component but there is still at least what appears to be a sensory component, e.g. seeing pink rats on the wall due to LSD ingestion… I should think more about this though. Even if there is “cognitive phenomenology” involved in beliefs and thoughts, it is presumably not the same as perceptual consciousness. And, of course, on my view there are two levels of concepts, i.e. lower-order and higher-order.

Comments are closed.

Back to Top
%d bloggers like this: