Code: TP-14

Peirception: Critical Common-sensism about Perception

I believe that we directly perceive extra-mental, external objects (tables, chairs and the like) rather than mental intermediaries (ideas, sense data or the like). But I am unsure what this belief of mine amounts to. My uncertainty derives primarily from the phrase "directly perceive," which is, despite its familiarity, too vague to be philosophically satisfying. My difficulty in seriously entertaining the idea that we don't directly perceive tables, chairs, etc. is a fine illustration of Charles Peirce's insight that the "indubitable is invariably vague." [1] Not surprisingly, the less vague and more precise I try to be about what it means to perceive something directly, the less comfortable I am with my belief. [2]

Because I believe in the direct perception of external objects, I sympathize with Susan Haack's desire to find an alternative to the less than satisfying dichotomy of, on one hand, typical realist theories of perception, which downplay the interpretative character of perception, and on the other, inferentialist theories, which deny perception's directness. And as a Peircean I am pleased by Haack's suggestion that Peirce's account of perception provides the resources for a satisfying alternative to this dichotomy. [3] I am tempted to refer to Haack's Peircean theory of perception as "Peirception." But despite the appeal this neologism may have as an essay title, it’s an inadequate name for a serious philosophical theory. Since Peirce's account of perception was one facet of his so-called Critical Common-sensism, [4] I will refer to that theory as it is articulated, amplified and adopted by Haack, not as the theory of "Peirception," but as Critical Common-sensism about perception, or simply CCS. [5]

According to Haack, the dichotomy between typical realism and inferentialism is overcome by Peirce's distinction between the perceptual judgment, the belief that accompanies a perceptual experience, and the percept, the phenomenal, interactive aspect of a perceptual experience. It is the concept of the percept that legitimates CCS’s claim to be an account of direct perception, an explanation of how perception, which is in some sense interpretative, can nonetheless be an immediate experience of external objects rather than of some mental intermediary. But as I will argue, the Critical Common-sensist concept of the percept has deep, perhaps irremediable, problems.


According to CCS, a perceptual experience (or percipuum, as Peirce called it) has two aspects: the percept and the perceptual judgment. We can think of each of these aspects separately, but neither aspect of the percipuum ever occurs apart from the other. In other words, the two elements of perception are conceptually distinct, although ordinarily they are, as Haack puts it, "phenomenologically inseparable". [6]

The percept itself has two aspects: a phenomenal aspect and a causal or reactive aspect. These characters correspond to Peirce's categories of Firstness, or quality, and Secondness, or reaction. [7] On one hand, the percept is the locus of phenomenal qualities — green, warm, garlicky, etc. Not that the percept itself is green, warm or garlicky. Rather, the percept is our experience of those qualities as they are instantiated in external objects; it is the phenomenal presentation of those qualities to us. The green shade and garlicky taste of guacamole are phenomenally presented to us in the percept. But not represented to us — the percept is a presentation, not a representation, of the external world. [8] On the other hand, a percept is also a direct causal interaction between the perceiver and the external environment.  When I perceive a bowl of guacamole, the perception involves a direct relation — a "clash", as Peirce described it [9] — between my senses and the guacamole itself. What I am perceiving is not something mental; what I am perceiving is guacamole. [10] The percept, then, is my direct perceptual interaction with external objects, an interaction that has a phenomenal aspect. CCS thus locates the directness of perception in the percept.

So the first component of CCS is:

(CCS-1) One aspect of perception is the percept, which is both a phenomenal presentation of the external world to the perceiver and a direct interaction between perceiver and external world. There is nothing representational about the percept; it does not represent the world as being any particular way; it doesn't represent the world at all; it just is.

Because CCS-1 posits that a perceiver's phenomenal experience of the world is itself a direct interaction between perceiver and external world, it seems to imply the doctrine of direct perception (DP), the claim that we directly perceive the external world rather than inner, mental objects. I say "seems to" because, as I mentioned earlier, the concept of direct perception is not nearly clear or precise enough to make the inference from CCS-1 to DP obviously valid. [11]

While CCS locates the directness of perception in the percept, it locates the representativeness of perception in the perceptual judgment, the spontaneous belief that automatically accompanies the percept. A perceptual judgment is, in Peirce's words, "a judgment asserting in propositional form what a character of a percept directly present to the mind is." [12] As Haack puts it, "[o]ur normally spontaneous judgments or descriptions of what we perceive involve interpretation, depending on background beliefs, expectation, set, as well as sensory input". [13] If I notice a bowl of guacamole on the table in front of me, I automatically and involuntarily believe that there is a bowl of guacamole there. This belief represents the world, and it represents the world as being a certain way, viz. as containing a bowl of guacamole sitting just in front of me. What's more, my belief serves as a potential explanation of the perceptual presentation of the guacamole that I experience. Why am I having a perceptual experience as of a warm, green, garlicky substance? Because there’s a bowl of guacamole in front of me (as I believe there to be). A perceptual judgment is like the conclusion of an abductive inference, although, because it is spontaneous and involuntary, it is not in fact the conclusion of an inference. [14] CCS is not a form of inferentialism; it does not imply that we infer the existence of the external world from our perceptual experience, or from beliefs about our perceptual experience.

So the second component of CCS is:

(CCS-2) Another aspect of perception is the perceptual judgment, the spontaneous and involuntary belief that is prompted by the percept. The perceptual judgment is representational, in that it represents the external world; propositional, in that it represents the external world as being a certain way and is thus capable of being true or false; and abductive, in that it serves as a potential explanation of the perceiver's phenomenal experience.

CCS-1 and CCS-2 are independent claims; it is possible for CCS-2 to be true and CCS-1 false. Even if the doctrine of the veil of perception (VP) is true, and what I directly perceive is something mental, my belief that there is guacamole in front of me would still be potentially explanatory of the specific nature of the mental items I experience when I'm really sitting in front of a bowl of guacamole.

Together, CCS-1 and CCS-2 imply that the propositional judgment is the only representational aspect of the percipuum; only the perceptual judgment, not the percept, serves a representational function in perception. I'll call this idea the representational exclusivity of judgment (REJ). [15] Peirce analyzed only the perceptual judgment, not the percept, in terms of Thirdness (representation, generality, connectedness). As Haack notes, this reflects two details of Peirce's theory: first, that percepts are non-propositional — unlike perceptual judgments, percepts make no claims and can be neither true nor false, neither certain nor uncertain [16] ; and second, that percepts are not representational at all — they don't represent anything to the perceiver; they simply are. [17]

So on the CCS model of perception, our perceptual experience of the external world is a direct interaction with that world. This seems to imply that we directly perceive the external world itself rather than representations of the world (DP). Perceptual representation is a matter of forming (potentially truth-valued) judgments about what one is perceiving, judgments that are potentially explanatory of the percept in question. Only these perceptual judgments represent the external world; percepts themselves are not representations of the external world (REJ).  If CCS is true, then Haack is right to say that we need not choose between the direct and the interpretative; perception can be both. [18]


As plausible as CCS might seem at the outset, I believe that it is beset by at least one serious problem. We can begin to understand this problem by noticing that REJ implies that the perceptual judgment, as the sole representative aspect of perception, must be the only aspect of perception capable of misrepresentation. Representation is a necessary condition of misrepresentation; only something that purports to represent x can misrepresent x. If, as REJ states, the only representation involved in the perceptual event is that of the perceptual judgment, then perceptual error, or misperception, must be a matter of false perceptual judgment or belief. The perceptual judgment is the locus of perceptual representation, and therefore the locus of perceptual misrepresentation. Our perceptions are fallible because our perceptual judgments are fallible.

But this cannot be correct. Some types of perceptual misrepresentation cannot be explained adequately by citing only the perceptual judgment. Sometimes it is the percept rather than the perceptual judgment that misrepresents the world; and this indicates that, contra CCS-1, percepts do have a representational character. The claim that every case of perceptual error is a case of false belief about what one is perceiving is plausible for many types of perceptual error. [19] But there are other cases of perceptual error, viz. hallucinations, for which it just seems wrong to say that one's belief is the sole locus of misrepresentation. Because CCS implies that perceptual error affects perceptual judgment only, the theory is committed to characterizing hallucination as a sort of belief. CCS must say, for example, that for Macbeth to hallucinate a floating dagger is for him mistakenly to believe that the space in front of his face, which is in fact empty, is occupied by a floating dagger. [20]

But this is not an accurate account of hallucination, since it is possible for the perceptual judgment that accompanies a hallucination to be true. This is the phenomenon psychologists refer to as pseudohallucination. [21] A pseudohallucination is a hallucination that does not fool the perceiver. If I see pink elephants and believe, not that there are pink elephants in front of me, but that I'm drunk, I am not fooled by my hallucination. [22] As long as we focus on hallucinations that fool the perceiver, it is easier to accept that hallucination is nothing but false belief. But it is possible to hallucinate x and at the same time to believe that one is hallucinating x and that x isn't real. In such a case, the perceptual judgment does not misrepresent anything. Call this the argument from pseudohallucination:

1. Some perceptual misrepresentations have only a phenomenal aspect and do not affect one's perceptual judgment.

2. Therefore, the percept is, on at least some occasions, a misrepresentation of the world.

3. If x is a misrepresentation, then x is a representation.

4. Therefore, the percept is, on at least some occasions, a representation; CCS-1 is false. [23]

Full-blown hallucinations also pose a threat to CCS-1. The space directly in front of me can look two completely different ways (e.g., it can look empty, on one hand, and it can look as if it is occupied by a floating dagger, on the other) without changing at all. My experience of that area can vary wildly while the area itself stays the same. Even though my dog Murphy is lying at my feet, my percept (due to the influence of drugs, say) might be as it would be were there a dead armadillo in front of me. The opposite is true, as well. It is possible, in principle, at least, for my environment to change significantly and my perceptual experience to stay exactly the same. I may misperceive Murphy to be lying at my feet, when in fact a dognapper cum demon cum evil neuroscientist has made off with him and has given me, at the exact instant he makes my dog vanish, the experience of seeing poor Murphy still lying there. In this example, it's the external world that's changed; my perceptual experience and its underlying neurology remain the same. [24]   In sum, our perceptual experience of the external world, and the external world itself, are independent of each other. Each can change without a corresponding change in the other.

In light of these facts, CCS must be able to explain how the phenomenal and the reactive aspects of a percept can be so thoroughly independent of one another, if in fact percepts are direct interactions with the external world. In neither of the cases sketched above does the percept present the world as it really is. [25] If the percept can be so different than it ought to be given the perceiver's actual environment, then how can it be a direct interaction between perceiver and environment, as CCS-1 claims?

Call this the strengthened argument from hallucination. It resembles the traditional argument from hallucination, but it emphasizes that the independence of percept and environment is bi-directional; not only can the percept change with no corresponding change in the environment, but the opposite is possible as well. It is also not nearly as ambitious as the traditional argument, as it doesn't attempt either to disprove DP or to prove VP. Stated formally, the argument is as follows:

1. The phenomenal aspect of a percept can vary wildly without the external world itself changing at all, and thus without the interactive aspect of the percept changing at all.

2. The external world around a perceiver can vary wildly without the phenomenal aspect of the perceiver's percept changing at all.

3. Therefore, the percept is not a direct interaction between perceiver and external world; CCS-1 is false.

Although it is not inconceivable that the step from (1) and (2) to (3) is invalid, I think (1) and (2) jointly provide a very compelling reason for thinking that (3) is true. To argue convincingly that this is not the case, Haack would need to explain how both of the following can be true:

(A) The phenomenal aspect of a percept can be just as it would be were the world around the perceiver very different than it actually is.

(B) Percepts are direct interactions between perceiver and external world.

But it is not obvious that both (A) and (B) can be true. In fact, it is unclear how, if you accept (A), you can avoid giving up (B). This is because it is unclear how the phenomenal aspect of a percept, conceived as a direct interaction with real, external objects, can vary so radically when there is no corresponding variation in the percept's reactive aspect.

Suppose that at time t Lady Macbeth experiences (genuine, non-hallucinatory) percept xt, which is a direct interaction between her and her immediate environment. There is no blood on the floor in front of her, nor is her percept as if there were blood. But at t+1, she experiences (hallucinatory) percept xt+1 — she begins to hallucinate blood on the carpet. xt+1 is very different from xt in its phenomenal aspect, but not in its reactive aspect. In fact, in so far as Secondness is concerned, xt and xt+1 are identical, since Lady Macbeth's environment hasn’t changed — the carpet is still free of blood. If percepts xt and xt+1 are direct interactions between Lady Macbeth and her environment, and if that environment does not change from t to t+1, it is unclear how the phenomenal aspects of xt and xt+1 can be so very different.  The same reasoning holds with regard to scenarios in which the external world changes but the phenomenal aspect of the percept in question does not.

One possible response is to say that hallucinations don't tell us anything about the nature of genuine perception, and in particular they don't suggest that genuine perception is not of the external world. This is because genuine perception and hallucination are very different from one another, in that hallucination lacks Secondness, i.e., hallucinatory percepts lack the reactive aspect that makes genuine percepts genuine. Hallucinatory percepts are not interactions with the external world. They are thus of a psychological type that is sufficiently different from that of genuine percepts that they don't suggest anything at all about the nature of genuine percepts. [26] This suggests an addendum to CCS to the effect that hallucinatory percepts are phenomenal and non-interactive. It is its interactive character that ties the percept to the external world and that thereby makes the notion of a hallucinatory percept hard to accept. So, according to this response, we need to eliminate the reactive character from our conception of the hallucinatory percept. (A) and (B) can both be true so long as (B) is construed as referring only to genuine, non-hallucinatory percepts. [27]

I do not think this move saves CCS-1 from the strengthened argument from hallucination. Given that hallucinatory percepts are capable of being phenomenologically indistinguishable from genuine percepts, it is not clear why this characterization of hallucinatory percepts (as merely phenomenal, not also reactive) should not also apply to genuine percepts. In other words, it is not clear why we should believe that hallucinatory percepts are of such a radically different psychological type than genuine percepts that this characterization applies only to one and not to both. Further, it is unclear what a percept, lacking in Secondness and having nothing but phenomenal qualities, would be, if not something mental, an idea, or a collection of sense data. So this response seems actually to edge CCS towards a position according to which we sometimes do perceive mental items, i.e., towards a version of VP limited to hallucinatory experience. [28] [29]


My arguments are rather modest, in that they conclude only that CCS-1 is false, not that that DP is false or that VP is true. But they are important nevertheless. CCS-1 is the part of CCS that purports to secure the directness of the relation between our perceptual experience and the external world, and if my arguments are strong, then CCS cannot be an accurate philosophical theory of direct perception. As a believer in the directness perception, and as a Peircean, I am not very happy about this conclusion. Although I agree with Haack and Peirce that perception is both direct and representational, I have doubts about whether perception is direct and representational in the way CCS says it is. The theory attempts to locate the representational character of perception exclusively in the perceptual judgment. But in doing so, it fails to explain how it is that hallucinations can be anything other than false perceptual judgments, how pseudohallucinations can happen at all, or how the phenomenal and reactive aspects of the percept can be so radically independent from one another.

Neither Peirce nor Haack would not want to say that the percept itself represents the external world, since to do so is to risk giving up the claim that we directly perceive the external world. But how else to explain hallucination and pseudohallucination other than to take that risk? What is needed is a way to retain the fallible, representational nature of the percept without giving up its directness, a way to explain how one and the same psychological event can be, on the one hand, a direct interaction or "clash" between perceiver and external objects and events, and on the other, representative of those objects and events, and therefore fallible. In other words, CCS must be revised so that it no longer implies REJ, the idea that the only representational aspect of the percipuum is the perceptual judgment – it must relocate the representational character of perception, at least in part, to the phenomenal aspect of the percept.

Whether this is possible is unclear. These two characterizations of the percept — as a direct interaction between perceiver and external objects and events, and as a phenomenal event that represents the external world — are obviously in tension with one another. But the relevant concepts (direct interaction, representation, phenomenal) are far from clear enough to judge whether the percept can in fact live up to both characterizations. More work is needed to determine whether CCS can be salvaged in this, or some other, way.

For the moment, I am confident in concluding only that CCS-1 is false. The argument from pseudohallucination and the strengthened argument from hallucination each provide good reason for rejecting it. My belief that we directly perceive external objects and events is in need of better support than has been provided thus far. Whether such support can be found for this indubitable but vague belief I share with Haack, and with Peirce, is a question about "Peirception" that I must leave for another day.

(draft, August 2001)


[1] 5.446, 1905. Unless otherwise indicated, references to Peirce are to the Collected Papers, 1931-58, by volume and paragraph number.

[2] For a recent attempt at making the phrase "directly perceive" less vague, see Snowdon (1992).

[3] Haack says that Peirce's theory "is as successful an attempt as any philosopher has made to escape the confines of the false dichotomy and seize the middle ground." (1994:10)

[4] Critical Common-sensism is Peirce’s synthesis of elements from Kant's critical philosophy and Reid's common sense philosophy. For a detailed statement of Critical Common-sensism, see, e.g. Peirce, 5.438ff. Peirce explicitly states his allegiance to Reid and Kant on the matter of direct perception at 5.56 (1903), 5.444 (1905) and 8.261 (1905).

[5] Haack's most detailed articulation and defense of CCS is in Haack, 1994. Chapters 4 and 5 of Haack 1993 include a Peircean account of perceptual experience as both direct and interpretative. That account, she says, "was arrived at initially without reference to Peirce's work: but then sharpened and refined when I realized that Peirce had supplied a map to the territory through which I was stumbling." 1994:33 n.20.

[6] Haack, 1994:161.

[7] Haack gives a capsule summary of Peirce's categories (1994:14-15). For a longer account in Peirce's own words, see, e.g., Peirce, 1992:146ff.

[8] Haack puts it this way:

To say that an external object is phenomenally presented is not to say that one perceives a mental image that represents it. ... to say that the percept involves an imageNR [a non-representational image] is not equivalent to saying that perception is of an imageNR. ... In perception, one might say, one has imagesNR; but these imagesNR are not what one perceives. (1994:22-23).

[9] Peirce, 8.41, c.1885.

[10] Haack, 1996:161.

[11] What's clearer, perhaps, is that DP does not imply CCS-1; from the claim that we directly perceive extra-mental object and events, it does not follow that there is a single event in perception which is both a phenomenal presentation of the world to the perceiver and a direct interaction between perceiver and world. Other realist models of perception are compatible with DP.

[12] Peirce, 5.54, 1903.

[13] Haack, 1996:161.

[14] Haack, 1994:18. As Haack notes, Peirce characterized perceptual judgments as "the extremest case of Abductive Judgments." (5.185, 1903) A perceptual judgment is like the conclusion of an abductive inference, in that it is potentially explanatory of the relevant percept. But perceptual judgments are different than genuine abductive inferences, in that they are not voluntary, nor are they inferred from other propositions – rather, they are caused by a perceptual presentations. (Haack, 1994:18)

[15] More accurately, this corollary is implied by the conjunction of CCS-1, CCS-2, and the assumption that the percipuum consists of nothing but the percept and the perceptual judgment.

[16] CCS's claim that percepts are non-propositional, not the sorts of thing to be truth-valued, is reflected in Haack's foundherentism (her theory of justification for empirical beliefs) as the claim that experiential S-evidence, and therefore current sensory S-evidence (the perceptual states that sustain a given belief-state), is "not the kind of thing with respect to which [one] has, or needs, evidence." (1993:77) On Haack's view, perceptual states don't "say" anything about the world, i.e. they don't represent the world as being a certain way; they just are. They are therefore not the type of thing for which evidential support could be provided.

[17] Haack, 1994:15-16.

[18] This claim isn't as clear as one might like, since "interpretative" (like its synonym "interpretive") is ambiguous. It can mean, on one hand, setting forth an explanation, and on the other, setting forth (an explanatory or non-explanatory) meaning. In the first sense, perceptual judgments are interpretative in virtue of their abductive character. In the second sense, perceptual judgments are interpretative in virtue of their representational character. But both readings of Haack's claim seem correct. Perception can be direct while at the same time having a representational character; and it can be direct while at the same time having an abductive character.

My dictionary seems to give priority to the latter sense. It defines "interpretative" as "Expository; explanatory." (American Heritage Dictionary, Second College Edition) Most of the time, Haack uses the words "interpretative" and "interpretive" to denote the explanatory character of the percept (e.g., 1994: 18, 31). But on one occasion, she seems to use "interpretive" to denote the percept's representational character, when she writes that "In every judgment... there is something in some degree interpretive, and hence fallible." (2000:13) Peirce's definition of "Interpretation of nature" in the Century Dictionary (1902) reads as follows: "1. Designed or fitted to explain; explaining; explanatory; 2. Inferential; implied; constructive." On that second definition, CCS holds that perceptual judgments are not interpretative, since they do not result from actual inferences. (My thanks to Cornelis de Waal for bringing Peirce's definition to my attention.)

[19] E.g., the Ponzo illusion and other psychology textbook examples of perceptual mistakes; perceiving a passing train as if it is moving and the train car you are sitting in as standing still; etc. A nice example of the Ponzo illusion can be found at (Dec. 13, 2000).

[20] Peirce himself strongly suggested this view at one point: "Hallucinations, delusions, superstitious imaginations, and fallacies of all kinds are experiences, but experiences misunderstood" (6.492, c.1896, emphasis added).

[21] See Slade and Bentall (1988:18-19). The authors note that "pseudo-hallucination" is used among psychologists in two different ways: "to refer to hallucinations that are recognised by the percipient to be non-veridical (i.e. not corresponding to a real object) and to refer to introspected images of great clarity." The conflation of the two senses is understandable, given that introspected images, even those of "great clarity," are unlikely to be mistaken for extra-mental objects. As Slade and Bentall also note, Kraupl-Taylor (1981) has suggested "perceived pseudo-hallucination" as a term for the former, and "imagined pseudo-hallucination" as a term for the latter. My argument relies on the phenomenon of perceived, rather than imagined, pseudo-hallucination.

[22] Ron Siegel describes a pseudohallucination he experienced while in a sensory deprivation tank: he saw a floating pink Buddha holding a pink balloon, but he knew at the time of the experience that what he was seeing was not real. (1992:6). Siegel and Jarvik (1975:104-105) conjecture that most drug-induced hallucinations, including the hashish-induced hallucinations reported by William James (1890:121-122), are actually pseudo-hallucinations.

Jerry Fodor appeals to pseudohallucinations, although not by that name, in a criticism of Putnam's account of perception:

Putnam prefers a 'disjunctive' account, according to which 'seeing an illusory pink elephant' is understood as 'seeming to see a pink elephant'. But he doesn't tell us how 'seeming to see' is itself to be construed unless in terms of representing. (There's a direct realist story, owing to Austin, which construes seeming to see a such-and-such as being disposed to believe that one is seeing a such-and-such. But it's a notorious non-starter. When I find myself seeming to see pink elephants, what I am disposed to believe is that I'm stewed.) (Fodor, 2000:3 of 7)

[23] The defender of CCS might respond as follows. At first, CSS may appear unable to explain where in the pseudohallucinatory percipuum error creeps in. Agreed, it does not occur in the perceptual judgment, and, as implied by CCS, neither does it occur in the percept. But appearances (as they say) are misleading: for there is in fact no misrepresentation in the pseudohallucinatory percipuum. Counter to how things may seem, there really is no misrepresentation when one knowingly experiences a hallucination. The percept is as it would be were there actually a dagger floating in front of the perceiver, but if the perceiver recognizes it as a hallucination, then there is no misrepresentation, no more so than when one rubs his eyes and "sees" phosphenes for a second or two. If you think that there is some misrepresentation here, it is only because you are assuming that the percept of the floating dagger is a representation, that it stands for the way things are in the external world. But if you don't assume this at the outset, then you won't think that there is any misrepresentation going on. When the drunk "sees" pink elephants and thinks "I'm tanked" (or when the psychonaut sees a smiling pink Buddha and thinks "It's time to leave the sensory deprivation tank"), there is no element of the percipuum that misrepresents anything — and hence no misperception. I do not find this response very convincing. It is extremely counter-intuitive to think that there is no perceptual misrepresentation going on when someone experiences a pseudo-hallucination. Of course, intuition might be wrong here, and it may well be that the defender of CCS can avoid my argument by biting the bullet and saying that pseudohallucination involves no misrepresentation. At any rate, I see no other way for CCS to avoid this argument.

[24] The possibility of this sort of scenario had not occurred to me before I read the first chapter of John Foster's The Nature of Perception.

[25] I feel naturally inclined to say that the percept gets the world wrong — an indication of how dissatisfied I am with the response to the argument from pseudohallucination sketched in note 23.

[26] This is reminiscent of disjunctive analyses of perceptual experience, such as those defended by Hinton (1973), Pitcher (1970) and, perhaps, Putnam (2000). For a strong criticism of such theories, see chapter six of Robinson (1994).

[27] It is questionable whether Peirce himself would have accepted this picture. We can think about the two components of the percipuum as if they were separable, but on Peirce's view they are not separable in fact — they never occur apart from one another. And even had Peirce not been explicit about this, his view of the categories more generally, outside the context of their application to perception, strongly suggests that he would not accept a picture of a percept, even a non-genuine percept, as being characterized by Firstness only. For example:

The typical ideas of firstness are qualities of feeling, or mere appearances. The scarlet of your royal liveries, the quality itself, independently of its being perceived or remembered, is an example, by which I do not mean that you are to imagine that you do not perceive or remember it, but that you are to drop out of account that which may be attached to it in perceiving or in remembering, but which does not belong to the quality. For example, when you remember it, your idea is said to be dim and when it is before your eyes, it is vivid. But dimness or vividness do not belong to your idea of the quality. They might no doubt, if considered simply as a feeling; but when you think of vividness you do not consider it from that point of view. You think of it as a degree of disturbance of your consciousness. The quality of red is not thought of as belonging to you, or as attached to liveries. It is simply a peculiar positive possibility regardless of anything else. (8.329, 1904, emphasis added)

So if the CCS model is to remain true to it's originator, it seems that it cannot avoid modeling non-genuine perception as an interaction between perceiver and... well, something. But neither does Peirce seem to have wanted to imply that non-genuine perception is an interaction with non-ego. Peirce's view was that it is the interaction between self and non-self that explains why genuine perception is involuntary and, sometimes, surprising. Hallucinatory percepts lack this interaction. At one point he suggests that their "unreality" just is this "lack of insistency." (8.149, 1901)

[28] From a Peircean point of view, it weighs against this proposed defense of CCS that some hallucinatory percepts persist through time. Peirce thought of persistence as resistance against change and thus as a hallmark of Secondness (5.45, 1903). So he would not accept the characterization of all hallucinations as lacking Secondness. A more important mark against this defense is that it gets the facts about hallucination wrong: at least some hallucinatory perceptions are interactions with one's environment. Suppose that Macbeth is perceiving normally (genuinely) at time t and begins to hallucinate a dagger at t+1. Macbeth's perceptual interaction with his environment does not cease altogether at t+1 -- he still perceives, and genuinely so, the room he is in, the clothes he is wearing, and other physical objects. There is, of course, no real dagger floating in front of him, so his percept of his environment at t+1 is not an interaction with, among other things, a dagger. But it is still an interaction with, among other things, the room he is in, the clothes he is wearing, etc. Even in the more radical cases of hallucination, the isolation chamber sorts of cases, CCS can maintain that there is still some sort of interaction with one's environment taking place. When one is floating in an isolation chamber, the clash between one's senses and the extra-mental world may be as muted as possible, but it is, CCS may still claim, a clash, an interaction, just the same. One may see an utterly black visual field at time t and a floating and grinning pink Buddha at t+1, but the clash does not cease at t+1. What changes from t to t+1 is just the phenomenal aspect of the percept, not it's element of interaction. So I don't think this response saves CCS-1 from the strengthened argument from hallucination.

[29] CCS might attempt to reconcile (A) and (B) by saying that when I am hallucinating, I am perceiving in a certain way, the way I would be perceiving were there actually a dagger floating in front of me. When Lady Macbeth hallucinates a blood stain, there is no red patch that she actually sees, no red sense data with which she is interacting. Rather she is "perceiving redly," i.e.,  perceiving part of the non-red carpet in a red manner. Her perception is a direct interaction with her environment — the object of her perception is the floor itself, not a mental representation of the floor. It's just that the activity of perception occurs in such a way that the floor appears to her to have a red patch on it.

But this "adverbial" approach doesn't save CCS-1 from the strengthened argument from hallucination. First, adding an adverbial element to CCS is no less ad hoc than the disjunctive approach sketched above. Given that some hallucinations are phenomenologically indistinguishable from genuine perceptions, and given that there is something that I perceive when I experience a genuine (non-hallucinatory) perception,  there seems to be no compelling reason for claiming that I am not perceiving something when I hallucinate. But set aside this general objection to the adverbial approach. What is more relevant to the current discussion is that the adverbial approach, even if it were true, would not help to reconcile (A) and (B).  The strengthened argument from hallucination does not depend on the claim that when I hallucinate a F object, there must actually be some F object (physical or mental) that I am perceiving. The premises of the argument are compatible with the adverbial theory. The point of the argument is not that we perceive sense data when we hallucinate. The point is, rather, that the phenomenal aspect of perception cannot be one and the same thing as an interaction with the external world, since (according to the argument) the phenomenal character of perception is radically independent of its interactive character. Describing the phenomenal aspect of perception in adverbial terms does not get around this fact. So "going adverbial" would not save CCS-1 from this argument.

This proposed response is clearly inspired by the adverbial theory of perception. See, e.g., Chisholm (1957). I take it that adverbialism itself has been refuted by Jackson (1975; 1977, chapter three). Tye (1984) has attempted to defend the theory, but Robinson (1994) argues persuasively that Tye's defense fails.


Chisolm, R. (1957). Perceiving: A Philosophical Study. Ithica, NY: Cornell University Press.

Fodor, J. (2000) "A Science of Tuesdays," review of The Threefold Cord: Mind, Body and World by Hilary Putnam (Columbia University Press, 1999), in London Review of Books 22, July 20, 2000, online at .

Haack, S. (1993). Evidence and Inquiry. Oxford: Blackwell.

   ___. (1994). "How the Critical Common-sensist Sees Things," Histoire, Epistemologie, Langage, 16, pp.9-34. Also forthcoming in New Studies on the Philosophy of C.S. Peirce, ed. Delaney, C.F., Fordham University Press.

   ___. (1996). "Reflections on Relativism: From Momentous Tautology to Seductive Contradiction," Philosophical Perspectives, 10: Metaphysics, ed. James E. Tomberlin, Oxford: Blackwell Publishers. Reprinted in Manifesto of a Passionate Moderate, Chicago: University of Chicago Press, 1998, 149-166. Page references are to the reprint.

Hinton, J. (1973). Experiences. Oxford: Clarendon Press.

Jackson, F. (1975). "On the Adverbial Analysis of Visual Experience," Metaphilosophy 6: 127-35.

   ___. (1977). Perception: A Representative Theory. Cambridge: Cambridge University Press.

James, W. (1890). The Principles of Psychology, volume 1. New York, NY: Henry Holt.

Peirce, C. (1931-58). Collected Papers, Hartshorne, C., Weiss, P. and Burks, A., eds., 8 vv., Cambridge, MA: Harvard University Press. References by volume and paragraph number.

   ___. (1992).  Reasoning and the Logic of Things, Ketner, K., ed., New York, NY: Harvard University Press.

Pitcher, G. (1970). A Theory of Perception. Princeton, NJ: Princeton University Press.

Robinson, F. (1994). Perception. London: Routledge.

Siegel, R. (1992). Fire in the Brain: Clinical Tales of Hallucination, New York, NY: Dutton.

Siegel, R. and Jarvik, M. (1975). "Drug-induced Hallucinations in Animals and Man," in Hallucinations: Behavior, Experience, and Theory, ed. by R. Siegel and L. West, New York, NY: John Wiley and Sons, 81-161.

Slade, P. and Bentall, R. (1988). Sensory Deception. A Scientific Analysis of Hallucination, Baltimore, MD: Johns Hopkins University Press.

Snowdon, P. (1992). "How to Interpret ‘Direct Perception’," in The Contents of Experience. Essays on Perception, ed. Tim Crane, Cambridge: Cambridge University Press.

Tye, M. (1984). "The Adverbial Approach to Visual Experience," Philosophical Review 93: 195-225.