Nonconceptual Mental Content

First published Tue Jan 21, 2003; substantive revision Wed Jul 17, 2024

The central idea behind the theory of nonconceptual mental content isthat some mental states can represent the world even though the bearerof those mental states need not possess the concepts required tospecify their content. This basic idea has been developed in differentways and applied to different categories of mental state. Not all ofthese developments and applications are consistent with each other,but each offers a challenge to the widely held view that the way acreature can represent the world is determined by its conceptualcapacities.

1. Introduction

Although some of the themes in current discussions of nonconceptualcontent have surfaced at various times in recent philosophicalthinking (within the philosophy of perception, see, e.g., Dretske1969, 1981. Within philosophy of cognitive science, see, e.g., Stich1978), the notion of nonconceptual content was explicitly introducedinto analytical philosophy by Gareth Evans (Evans 1982). As part ofhis general discussion of the role of information-links in makingpossible demonstrative and other types of identification, Evansdevelops the idea that the information yielded by the perceptualsystems (including somatic proprioception) is nonconceptual. Thisnonconceptual information, he argues, is initially unconscious butbecomes conscious when it serves as input to a thinking,concept-applying, and reasoning system.

Evans is not always clear whether he understands nonconceptual contentto be a personal level or a subpersonal level phenomenon. And, infact, it seems that Evans’s conception of nonconceptual contentis in at least one important way deeply antithetical to that currentlydiscussed. Whereas much contemporary discussion of nonconceptualcontent is focused on the content ofconscious perceptualstates, it looks very much as if Evans understands perceptual stateswith nonconceptual content as beingnon-conscious until thesubject’s conceptual abilities are brought to bear on them(Campbell 2005). Nonetheless, the general idea that there might beways of representing the world independent of the thinker’sconceptual capacities has inspired other philosophers. An early,personal level application of the notion is Tim Crane’s paper onthe waterfall illusion (Crane 1988a. An illustration of the waterfallillusion, also known as the motion aftereffect illusion, can be foundhere). Crane argues that the waterfall illusion presents an experience witha contradictory content that hence cannot have a conceptualcontent—since conceptual contents must be consistent. This claimhas provoked some debate (Mellor 1988, Crane 1988b).

Early development of the notion of nonconceptual content cameprimarily from philosophers connected with the University of Oxfordand working within a broadly Fregean tradition. Christopher Peacocke,in a series of papers and then in his book on concepts (Peacocke1992), argued that the fineness of grain of perceptual experienceoutstrips the conceptual capacities of the perceiver (more on thefineness of grain argument in section4.1). He offered a detailed theoretical framework for understanding thenonconceptual content of perceptual experience, which he termedscenario content. A rather different account of the nature ofperceptual experience in nonconceptual terms was provided by AdrianCussins (Cussins 1990), who suggested that we should understand thenonconceptual content of experience in essentially ability-basedterms. Cussins argued that this personal level notion of nonconceptualcontent is naturally complemented at the subpersonal level by aconnectionist cognitive architecture. Michael Martin applied thenotion of nonconceptual content to the study of memory (Martin 1992).Recent applications of the notion of nonconceptual content havefocused on the relation between perception and action (Hurley 1998)and the analysis of conscious experience (Tye 1995, 2000).

One thesis that might be held about nonconceptual content is that athinker can represent the world nonconceptually without possessingany concepts at all. This was termed the Autonomy Thesis byPeacocke, who offers an argument against it in his (1992). Thisargument was challenged by José Luis Bermúdez(Bermúdez 1994) in favor of a notion ofautonomousnonconceptual content that can be used to explain the behavior ofnonlinguistic creatures. Peacocke was initially unconvinced by thisline of argument (Peacocke 1994), but subsequently changed his mind(Peacocke 2002). Bermúdez further deployed the notion ofautonomous nonconceptual content in exploring primitive nonlinguisticforms of self-consciousness (Bermúdez 1998). (More on theAutonomy Thesis in section6.)

The notion of nonconceptual content has also been utilized inspecifying the representational contents of subpersonal states, suchas those involved in the early stages of visual processing. Suchstates, it is suggested, have representational contents and yet it isunlikely that the subject undergoing those states has the conceptsutilized in a theoretical description of the contents of those states.(More on the nonconceptual character of subpersonal states in section4.2.) Additionally, the notion of nonconceptual content has been utilizedin the explanation of the behavior of nonlinguistic creatures or thosethat do not employ concepts. Since,ex hypothesi, suchorganisms do not have conceptual capacities, if we are to attribute tothem states with representational contents at all the only contentsavailable to us are nonconceptual representational contents. (More onthe explanation of the behavior of nonlinguistic creatures in section4.3.)

The most extensive discussion of the notion of nonconceptual contenthas been in the philosophy of perception. Many have suggested that acorrect specification of the representational contents of perceptionneed not be limited to concepts possessed by the perceiver. In otherwords, a specification of the content of perception can be sensitiveto how an organism perceptually represents its environment even thoughit utilizes concepts the organism need not possess. This claim issupported by a variety of different considerations, some of whichconcern the distinctive features of the phenomenology of perceptionand some which concern perception’s distinctive epistemic role.(See section4.1 for the debate about the nonconceptual content of perception.)

Some have objected to Evans’s original characterization of thenotion on the grounds that whether or not a correct specification ofthe representational content of some mental states is constrained bythe concepts possessed by the subject speaks only to the conditions asubject must satisfy in order to undergo such states, rather than tothe nature of the state’s representational content (e.g., Heck2000). This has led to a distinction between two forms ofnonconceptualism—“state”-nonconceptualism and“content”-nonconceptualism. The former is concerned withthe conditions one must meet in order to undergo a state with a givencontent, i.e., whether or not one must possess the concepts involvedin specifying that content, whereas the latter is concerned with thenature of a mental state’s content. Nonetheless, as we discussbelow, there is good reason to suppose that the distinction collapses,and that the two forms of nonconceptualism are mutually-entailing.(See section3 for the purported distinction between state and contentnonconceptualism.)

2. Nonconceptual content: An initial characterization

The notion of nonconceptual content is fundamentally contrastive. Inelucidating it we need to start with the notion of conceptual content.The paradigm case of a state with conceptual content is apropositional attitude such as a belief or desire. Having apropositional attitude involves standing in a certain relation to acontent (athought or aproposition). The content iswhat it is that is believed, desired, hoped for etc. Althoughpropositional attitudes are ultimately directed at certain objects,properties and/or relations (which yield theirtruth-condition, and in terms of which theirtruth-value is to be determined), it is clear that onlycertain ways of characterizing those objects/properties/relations canserve to specify the content of the relevant propositional attitude.It would be incorrect, for example, to characterize the content of mycurrent belief that my car is parked in the driveway by using theconcepts of particle physics to describe the state of affairs thatwould make it true. This would be incorrect because it completelyfails to capture how I think about the state of affairs of my carbeing in the driveway. The obvious question to ask, once thispreliminary point is in play, is ‘What constraints are imposedupon specifications of the content of propositional attitudes by therequirement to respect the way the subject thinks about thetruth-condition of the relevant attitude?’

Different theories of content will respond to this question indifferent ways, but the following is widely held to impose a minimalconstraint upon any such response.

The conceptual constraint:
Specifications of the content of a sentence or propositional attitudeshould only employ concepts possessed by the utterer or thinker.

Certain theories of content and concepts directly entail theconceptual constraint. Within a broadly Fregean tradition, forexample, the contents of propositional attitudes (and the meanings ofsentences) are taken to consist of concepts—and it is hard tosee how one can have a propositional attitude whose content is acomplex of concepts without possessing each of them.^[1]

But the conceptual constraint does not depend upon adopting anyparticular theory of content. Its plausibility stems, rather, from theconjunction of two thoughts.

In specifying what a thinker believes, what a perceiver perceivesor what a speaker is saying by uttering a certain sentence in aparticular context one has to be as faithful as possible to how thatthinker, perceiver or speaker apprehends the world.
How a thinker, perceiver or speaker apprehends the world in havingbeliefs about it, perceiving it or speaking about it is a function ofthe concepts he possesses.

This way of motivating the conceptual constraint has been explicitlyput forward by several authors who argue forcibly for the secondthesis (see, for example, Peacocke 1983, McDowell 1994a, Brewer 1999,and Noë 1999). Theorists of nonconceptual content, in contrast,accept the first constraint without the second. They hold thatspecifications of content must respect the way a thinker, perceiver orspeaker apprehends the world andbecause of this they cannotbe circumscribed by the concepts possessed by the thinker, perceiveror speaker. Theorists of nonconceptual content postulate the existenceof ways of representing the world (and hence the existence of a typeof content) that outstrip the concepts possessed by the thinker.

The conceptual constraint can be lifted in two different ways. It canbe lifted globally by simply denying that any content specificationsneed confine themselves to the concepts possessed by theutterer/thinker. The currently popular identification of propositionswith functions from possible worlds to truth values involves aglobal lifting of the conceptual constraint. Possible worldsemantics is intended to apply to all propositional attitudes and itis obvious that few believers who are not also professionalphilosophers will have any grip at all on the central theoreticalconcepts of possible worlds semantics. All content comes out asnonconceptual content in this sense (Stalnaker 1998).^[2] In terms of the two motivations identified earlier for the conceptualconstraint, this approach to specifying content appears to repudiatethe first. It holds that specifications of content need not besensitive to how the speaker or thinker apprehends the world, at leastif one thinks that such sensitivity requires a degree of aspectualitythat possible worlds semantics explicitly rejects.^[3]

It is more interesting to think about what might happen if we retainthe first motivation, but question the second, looking for ways inwhich thinkers/perceivers/speakers represent the world that are not afunction of the concepts they possess. There are three differentrepresentational domains for which such alocal lifting ofthe conceptual constraint may be plausible:

Perceptual states (see further4.1 below)
Representational states at the subpersonal or subdoxastic level(see further4.2 below)
Representational states of non-human animals and human infants whodo not seem to be concept possessors (see further4.3 below)

Of course, the plausibility of a local lifting of the conceptualconstraint will be a function of how concepts are understood. In thefollowing we make two assumptions about concepts. These assumptions donot fix a single notion of concept, but they do however determine thelogical space within which accounts can be developed of thecorrelative notions of conceptual and nonconceptual content.

The first assumption is that concepts are semantic entities ratherthan psychological entities. Concepts are constituents of contents.Attitudes towards contents are psychological occurrences, but thecontents themselves are not psychological entities. They, and theconcepts they contain, are abstract entities.^[4] The second assumption is that, although concepts are abstractentities, mastering a concept is a psychological achievement. We needa cognitive account of what it is to master a concept, even thoughconcepts are not psychological entities.

3. An important distinction: State vs. content nonconceptualism

Heck (2000) introduces an alternative conception of the debate overnonconceptual content. This conception arises from an ambiguity hefinds in Evans (1982) and other proponents of nonconceptual content.He notes that these proponents appeal to the thesis that while thebeliefs a person can have depend on the concepts she possesses, theperceptual states she can have do not. However, this thesis does notimmediately establish a conclusion about thecontents ofperception. It establishes only a conclusion about the conditions asubject must satisfy to be in a perceptualstate with somecontent. Prima facie, this conclusion appears to be compatible withthe claim that perceptual states and belief states have the same typeof content.

Heck distinguishes two different ways of drawing theconceptual/nonconceptual distinction. According to thecontent view, there is a fundamental difference between thetype of content that perceptual experiences can have and the type ofcontent that beliefs and other propositional attitudes can have.According to thestate view, in contrast, there need not bemore than one type of content. Perceptual experiences andpropositional attitudes could have contents of the same type (in bothcases content could be given in terms of sets of possible worlds, forexample). What distinguishes perceptions from propositional attitudeson the state view is that the latter are concept-dependent while theformer are concept-independent. A state-type is concept-dependent justif it is content-bearing and it is impossible for a thinking andperceiving subject to be in a token state of that type withoutpossessing the concepts required to specify the content of that tokenstate. In contrast, a subject can be in states of aconcept-independent state-type even if she lacks all or some of theconcepts required for an accurate specification of the relevant contents.^[5]

Several reasons have been proposed for consideringstate,rather thancontent, nonconceptualism as of primaryimportance in the nonconceptual debate. First, as many have noted(Byrne 2003, 2005, Speaks 2005, Crowther 2006, Heck 2007), keeping thestate/content distinction in mind suggests that much of the recentdebate purportedly aiming at thecontent of perception infact only concernsstate-nonconceptualism. Arguments for thenonconceptual content of perception usually proceed via demonstrationof the concept-independence of perception. However, this is acharacteristic of state-nonconceptualism; it does not seem to entailcontent-nonconceptualism directly. In section4.1, after reviewing the arguments that have been proposed for thenonconceptualcontent of perception, we will return to thisconcern.

A second reason offered for discussing nonconceptualism at the levelof state type rather than at the level of content is that it allowsfor a meaningful debate about nonconceptualism even if one holds thatall content is nonconceptual. Stalnaker (1984, 1998), for example,argues that all content should be understood in terms of sets ofpossible worlds. If that is the case, then the difference between thepropositional attitudes and perception should be elucidated in termsextrinsic to the type of content involved in these states. This can beachieved if we think that the propositional attitudes, but notperceptual states, are concept-dependent; that is, if we think thathaving propositional attitudes but not perceptual states requires thesubject of those attitudes to possess certain concepts. This reasoningholds equally well if one thinks that all content is conceptual. Insuch a case, one is forced to explain the difference between differenttypes of content involving states in terms extrinsic to the nature oftheir contents; the state-view is one way of explaining thisdistinction, by appeal to the capacities required of a subject if sheis to undergo such states.

An additional reason that the “state” view might be foundappealing is that it allows for the commonsensical notion that thecontents of perception can be taken at face value to produce beliefsthat havethe same content. If, on the other hand, one holdsthat the content of perception is of a fundamentally differentkind than that of belief, it is less easy to see how one canrespect the intuition that one believes the same thing that oneperceives (Byrne 2005, Speaks 2005). Furthermore, it has been arguedthat perception can play a justificatory role in the production ofperceptual beliefs only if belief and perception involve the same kindof content (McDowell 1994a, 2006, Brewer 1999, 2005). This claim hasbeen disputed (e.g., Heck 2000, Peacocke 2001a, Lerman 2010, Cahen2019. For more on the justificatory role of perception see section4.1).

The “state” view attempts to make theconceptual/nonconceptual distinction a distinction between types ofstates, those that are concept-dependent and those that areconcept-independent. Bermúdez (2007) argues that such adistinction is implausible when considered independently of thedistinction drawn by the “content” view. The difficulty heidentifies with the “state” view is that it is unclearwhat basis there might be for a distinction between concept-dependentand concept-independent state types other than the distinctivecontents of these states. That is, the “state” view owesus some account of why some states, such as the propositionalattitudes, are concept-dependent, whereas other states, such asperceptual states, are concept-independent. It seems that a naturalexplanation for why perceptual states, but not propositionalattitudes, are concept-independent is that perceptual states differfrom propositional attitudes in the type of contents that are involved(which determines, for example, the particular inferential relationsin which they can stand). This explanation is not available if we holdthat perception and the propositional attitudes have contents of thesame type.

It is open to the “state” view theorist to attempt toexplain the “state” view distinction in terms of thefunctional roles of the different state types. She might argue for thedistinction by pointing out that the functional role of belief dependson an organism’s conceptual capacities and the functional roleof perception does not depend on its conceptual capacities (since,arguably, an organism’s perceptual discriminations can outstripits conceptual capacities). However, there is a sense in which thiswould be a mere restatement of the problem rather than a solution toit, since we still need some account of the concept-(in)dependenceof these functional roles. Again, an appeal to the differentkinds of contents involved in these different states seems to be astraightforward explanation. The “state” view theoristmight also attempt to explain the “state” view distinctionby appeal to the phenomenological differences between thepropositional attitudes and perception. However, this too seemsproblematic, since it is unclear why the lack of phenomenology shouldmake a belief concept-dependent or why the phenomenology of perceptionshould make it concept-independent (Bermúdez 2007).

In a recent article, Toribio (2008) places further pressure on theplausibility of the state/content distinction. She argues thatarguments couched in “state” terms—appealing to theconditions the subject must satisfy in order to undergo states withcertain representational contents—do entail conclusions aboutthe contents of those states. The entailment is established by notingthe first motivation for the conceptual constraint, mentioned above(section 2), regarding the proper specification of content. Whenspecifying the representational content of some mental state one hasto be as faithful as possible to how the creature undergoing thatstate apprehends the world. Yet, how the creature apprehends theworld, she argues, depends on what cognitive capacities it has. Thus,a creature that lacks conceptual capacities cannot be properlyattributed states with conceptual contents. Arguments to the effectthat perception isstate-nonconceptual—that a creaturecan undergo perceptual states without having the concepts utilized inspecifying the content of those states—thus also amount toarguments that the content of those states are nonconceptual. See VanCleve (2012) and Schmidt (2015) for further arguments supporting theinterdependence of the state and content views of nonconceptualism.For responses to these arguments against the state/contentdistinction, see Duhau (2014)

4. Applying the notion of nonconceptual content

Theorists of nonconceptual content have primarily deployed the notionin the service of three different explanatory projects: (a) theproject of characterizing the content of perceptual experience, (b)the project of characterizing the content of subpersonalrepresentational states, and (c) the project of explaining thebehavior of certain non-human animals and of pre-conceptual humaninfants.

4.1 Perceptual experience and nonconceptual content

We can think about the content of perception in conceptual terms. Whensomeone has a visual experience, for example, it will seem to thatperson as if something is seen. One way of specifying the way thingsappear to the subject is in terms of the perceptual belief the subjectwould form on the basis of perception had he taken his perception atface value. Such a way of specifying the representational content ofperception gives us the propositional content of perception. Theoristsof nonconceptual content have canvassed several types of reasons forthinking that this propositional content does not exhaust therepresentational content of perception (see the entry onthe contents of perception).^[6]

First, it has been argued that the contents of perception exhibitcertain features that cannot be exhibited by the contents of thepropositional attitudes. In particular, perception is capable ofrepresenting impossible or contradictory states of affairs, as is thecase when viewing certain Escher drawings and when undergoing variousperceptual illusions. In contrast, it is argued that conceptualcontent must be consistent—one cannot have a belief with acontradictory content. Thus, Crane argues that reflecting on thewaterfall illusion shows that the contents of perception cannot beconceptual (Crane 1988a). Since one cannot undergo a state withcontradictoryconceptual contents, and yet the waterfallillusion is a perceptual state with contradictory content—theobject perceived seems to be both moving and not moving—itcannot be the case that perception involves conceptual contents (formore on this argument see, e.g., Mellor 1988, Crane 1988b, Crane 1992,and Gunther 2001).

Second, it has been argued that the content of perception is analog innature, unlike the conceptual content of propositional attitudes whichis more plausibly seen as digital. The distinction between analog anddigital representations has (for our purposes) been most perspicuouslyput by Fred Dretske (1981, Ch.6). Let us take a particular fact orstate of affairs, say the fact or state of affairs that some objects has propertyF. A representation carries theinformation thats isF in digital form if and onlyif it carries no further information abouts other than thatit isF (and whatever further facts about it are entailed bythe fact that it isF). But whenever a representation carriesthe information thats isF in analog form it alwayscarries additional information abouts. Nonconceptualistsargue that, while propositional attitudes represent the world indigital form, perceptual states represent the world in analog form.Beck (2012) argues that analog magnitude states are nonconceptualbecause they lack the recombinability necessary for concepts (asarticulated, for example, in Evans’s Generality Constraint).^[7]

Third, it has been suggested that the content of perception isunit-free (Peacocke 1986). If I perceptually represent an object asbeing a certain distance from me, I do not usually represent thatdistance in terms of a particular unit (in inches, say, as opposed tocentimeters), even though what I represent is a perfectly determinatedistance. I simply represent it as beingthat distance, wherethe content of my perception specifies the distance. Peacocke (1986,1989) argues extensively against the possibility of accommodating suchunit-free representations in purely propositional terms.

Fourth, nonconceptualists argue that the content of perception is morefine-grained than the content of propositional attitudes. Thisargument can be traced back to Evans, who asks, rhetorically,“Do we really understand the proposal that we have as many colorconcepts as there are shades of color that we can sensiblydiscriminate?” (Evans 1982, p. 229). Arguably, I canperceptually discriminate many more colors and shapes than I currentlyhave concepts for. Although I may be capable of discriminating betweentwo color chips of very similar shades of red, red₂₇ andred₂₉, not being an expert on colors I will not have theconceptsred₂₇ andred₂₉. Withmy limited conceptual repertoire, I will correctly judge both colorchips to be red. However, I will so judge on the basis of experienceswhose contents are much more specific and fine grained in a way thatcannot be captured by my conceptual capacities. (The same is true,mutatis mutandis, of shape concepts.) My ability to perceiveand discriminate the determinate shades and shapes that I do outstripsmy conceptual capacities. As such, my experiences do not seem todepend on my conceptual repertoire. Thus, in specifying the content ofperception, its accuracy conditions, we need not limit ourselves tothe concepts available to the subject—hence, perceptual contentis nonconceptual (e.g., Peacocke 1992, Heck 2000, Tye 1995, 2006.DeBellis 1995, and Luntley 2003, argue on similar lines for thenonconceptual content of music perception).

Of these, the argument from the fine-grained nature of perception hasgenerated the most discussion. Opponents of the idea that perceptualcontent might be nonconceptual argue that the fineness of grain ofperceptual experience can in fact be accommodated at the conceptuallevel. John McDowell, for example, suggests that the considerationsEvans raises have force only if we limit the types of conceptsexpected to figure in the content of perception to such generalconcepts asred₂₇,red₂₉,etc., but, he argues, there is no reason to suppose that theconceptual content of perception needs to be limited in this way.Rather, the conceptual content of perceptual experiences is given bydemonstrative concepts, such asthat shade (McDowell 1994a).^[8] He claims that our conceptual capacitiesare capable ofrepresenting colors with the same fineness of grain with which theyare perceptually represented. The information loss in the transitionfrom perception to perceptual belief is not a sign that there are twodifferent types of content in play, but should rather be understood asa transition from a more determinate type of conceptual content (e.g.,that is colored thus) to a less determinate type ofconceptual content (e.g.,that is red).

Several objections have been raised against attempts to account forthe contents of perception in terms of demonstrative concepts. First,for a capacity operative in perception to count as a genuineconceptual capacity it must be possible for that same capacity to beemployed in thought. As such, it must be the case that suchdemonstrative concepts as are arguably involved in the content ofperception are retained long enough for the possibility of theirremobilization—they cannot be reserved only for the particularinstance in which the sample is perceived. Although the appearance ofthe demonstrative concept expressible by the phrase ‘… iscolored thus’ can exploit the presence of the red₂₇color chip and account for the person’s fine-grained experienceof the chip, the demonstrative concept’s applicability cannot belimited to the presence of this sample. As McDowell says, the capacityfor employing demonstrative concepts in perception is a recognitionalcapacity (McDowell 1994a p. 57, Brewer 1999 also holds that aconceptual capacity must be available in the absence of the sample towhich it was initially applied). Kelly (2001a) calls this constraint‘the re-identification condition’ on conceptpossession.

However, the objection goes, it is evident that our capacities forperceptual discrimination far outstrip our capacities for recognition.Regardless of how short the interval between two presentations of aspecific shade it is reasonable to suppose that an organism is capableof perceiving the shade in all its fineness of grain without beingcapable of recognizing it as the shade presented earlier. There simplydoes not seem to be a dependence between our fine grained perceptualcapacities and our memory capacities. But since the recognitionaldemonstrative concepts the conceptualist appeals to do depend on suchmemory capacities (according to the re-identification condition), itcannot be the case that our having such demonstrative concepts isnecessary for our enjoying fine grained perceptual discriminations.(For discussion of empirical support for the claim that memory iscoarser grained than our capacities for perceptual discrimination, seeRaffman 1995. See also Kelly 2001a, Peacocke 2001a, 2001b, Tye 2006,Wright 2003, Dokic and Pacherie 2001 for arguments on similargrounds).

One response to this objection would be to revise (Brewer 2005) there-identification constraint on the possession of demonstrativeconcepts, or perhaps retract it altogether (Chuard 2006 exploresvarious versions of the re-identification condition and argues thatnone are necessary for the possession of demonstrative concepts).Kelly (2001a), and Dokic and Pacherie (2001), however, argue that there-identification condition is essential to concept possession.

Two potential challenges to the appeal to demonstrative concepts havebeen highlighted. The first can be termed the priority argument. Itholds that we cannot take demonstrative concepts as explanatorilybasic. The demonstrative concepts I possess are a function of mydemonstrative capacities and part of what explains my having a givendemonstrative capacity is my having certain associated experiences;the experience must be prior to the possession of the demonstrativeconcept if the former is to explain the latter (Heck 2000). Variousformulations of this argument have been developed (see Ayers 2002,Hopp 2009, Roskies 2008, 2010, and Levine 2010). For example, Levine(2010) argues that “[w]hen you see a color and think ‘thatcolor’, the seeing is prior to the demonstrating, or else youreally don’t know what you’re demonstrating… But ifthe seeing, the perceptual experience, is prior to the demonstrating,then the demonstrating can’t be what captures, or brings intoexistence, the content of that experience” (p. 191). Brewer hasprovided a response to this kind of objection, following a suggestionrejected by Heck (2000), on the grounds that the form of explanationthat we should look for is not acausal explanation, whichwould indeed require the priority of the experience, but aconstitutive one. Thus he says that “On theconceptualist view, experience of a colour sample, R, just is a matterof entertaining a content in which the demonstrative concept‘that_R shade’ figures as a constituent”(Brewer 2005, p. 221, see also Brewer 2002).

A second challenge comes from the so-called argument fromnon-veridical experience (Heck 2000). Suppose that a subject has anon-veridical experience as of a shade of color that does not exist innature and that has never previous appeared in her experience. Thenthe subject cannot have an appropriate demonstrative concept, as thereis nothing to fix the reference of that concept. For a response tothis argument see Bengson, Grube, and Korman 2011.

Recently, Settegast (2023) has suggested a novel response to thefineness of grain challenge and a defense of McDowell’sdemonstrative strategy. Underlying his reply is a hylomorphicinterpretation of McDowell’s conceptualism, according to whichperception and thought are distinguished not in terms of theircontents, but in terms of their ‘mode of givenness.’ Whilein thought and judgment Fregean senses are presented to the subjectlinguistically, in experience, it is the sensory character ofexperience that serves as the carrier of the same Fregean senses.

Before moving to other arguments that have been advanced on behalf ofnonconceptualism it is worth remarking that the conceptualist appealto demonstrative concepts in response to the fineness of grainargument is a response to a challenge to their position, rather thanan argument for conceptualism. The point of the fineness of grainargument is to put pressure on the possibility of a subject possessingall the concepts required for a correct specification of the contentsof perception. If the subjectcannot have these concepts andyet is capable of undergoing the perception, then it follows that thecontent of perception is nonconceptual. However, showing that thesubjectdoes have these concepts is not yet an argument forthe conceptualist position. For the conceptualist claim to follow,they must show not that the content of perception is in factconceptualizable by the subject undergoing the experience in all ofits fine grain, as the availability of demonstrative conceptssuggests, but rather that the subjectcannot undergo theexperiencewithout possessing these concepts. Thus even ifthe conceptualists manage to show that subjects do possess all theconcepts involved in the correct specification of the content ofperception, it still does not show that the content of perception isconceptual, it only undercuts the motivation the nonconceptualist canobtain from the fineness of grain argument (Coliva 2003,Bermúdez 2007).

A fifth argument, related to the priority argument against the appealto demonstrative concepts above, pertains to the possibility ofconcept acquisition or learning. Presumably what explains our comingto possess the observational concepts that we do is our undergoingappropriately related perceptual experiences (see, e.g., Peacocke1992, 2001a, Ayers 2004). If this is so, it cannot be the case, onpain of circularity, that undergoing a particular perceptualexperience depends on possessing the very concepts the acquisition ofwhich the experience explain. Roskies (2008) argues that if we rejectthe claim that perception with nonconceptual content plays anessential role in concept acquisition, we cannot account for ouracquisition of observational concepts without being committed to anunacceptable form of nativism. (See also Forman 2006 for a differentargument purporting to show the necessity of perception withnonconceptual content to concept learning.)

A sixth argument for the nonconceptual content of perception derivesfrom Kelly (2001b). He argues that the central reason perception mustbe nonconceptual is that the experience of a property depends on thesituation in which it is perceived and depends on the object which itcharacterizes in a way that demonstrative concepts do not. An exampleof the first sort of dependency is given by the phenomenon of colorconstancy. In this case we can see two samples of the same colorproperty illuminated differently (e.g., when looking at a wall that ispartially exposed to the sun) and experience both to be of the samecolor but because of the different illumination we experience themdifferently (one is better illuminated than the other). The problem isthat since the color is the same regardless of the illumination,demonstrative concepts such asthat color cannot besufficient to capture the experiential difference between thedifferently illuminated samples; though the experiences differ, thecontribution of the demonstrative conceptthat color to thecontent of the experience is the same, given that the propertyperceived, that color in both demonstrations, is the same.Demonstrative concepts, Kelly argues, refer to situation independentfeatures of the world (e.g., the actual color of the wall) and thusare not capable of capturing the situation dependence of perceptualcontent. The argument from the situation dependence of perceptualcontent has been subject to several criticisms on behalf of theconceptualist. One such criticism (Peacocke 2001b) is that, contraryto Kelly, it seems reasonable that the difference in experiencecan be captured by concepts regarding the illumination of thedifferent samples (see also Ablondi 2002).

Finally, there is the argument from infant/animal cognition, whichPeacocke now considers the most important motivation for thinking thatperception has nonconceptual content (Peacocke 2001a, 2001b, referringto Evans 1982, Dretske 1995, Bermúdez 1994, 1998).Considerations of phylogenetic and ontogenetic continuities in naturesuggest that animals and infants can have experiences very similar toour own. They too perceptually represent the brown tree amidst thelush green grass. However, since they lack concepts appropriate for acorrect specification of the contents of their perception, theirperceptual representation of this state of affairs must benonconceptual. If we share these perceptual capacities with suchanimals and human infants then it is reasonable to suggest that thecontent of our experiences is also, at least in part, nonconceptual(see also Schellenberg 2013, 2018). Of course, it may be the case thatour experience is richer than that of animals and that we are capableof many more perceptual discriminations than they are capable of, butto the extent that we share our perceptual representations with suchanimals, the content of our perceptual representations should beequally nonconceptual. (See more on this in section4.3)

A crucial aspect of the debate is whether theorists of nonconceptualcontent can account for the rational role of perceptual states inbelief formation. Although much of the discussion of whetherperceptual content can be nonconceptual has focused on precisely howthe manifest differences between perceptions and beliefs should bedescribed (see, e.g., Kelly 2001b, Sedivy 1996, 2006), the importantissue is whether those differences can be captured at the level ofcontent in a way that explains how perceptions can justify beliefs.McDowell’s central claim is that there can be appropriaterational relations between perceptions and beliefs only if perceptionshave conceptual contents, on the grounds that such rational relationscan hold only between conceptual states (McDowell 1994a, 2006). Byrne(2020) relies on similar reasoning in arguing that perception must beconceptual. This is a powerful challenge to defenders of nonconceptualcontent. It is far from clear, however, that they do not have theresources to meet it. This is an issue that has been much discussed inthe philosophy of perception (most recently in Lerman 2010 and Cahen2019).

Millar (1991), for example, offers a sophisticated account of how therepresentational content of perceptions can justify beliefs formed onthe basis of those perceptions, even though the relevant contents arenot conceptually individuated. One obvious strategy for the defenderof nonconceptual content would be to argue that a nonconceptualcontent can stand in logical or evidential relations (such as therelation of being consistent with, or making more probable) to anotherstate even though it is not conceptually articulated (see, e.g., Heck2000, Vision 2009). Alternatively, a nonconceptualist, followingRaleigh (2007), might push back against the thesis that justificationhas to be inferential in nature. One way of developing that strategyis proposed by Peacocke (2001a, 2004a, 2004b) who argues that therational role of a nonconceptual way of representing some X isguaranteed when the observational conceptX is partiallyindividuated by its relation to such a way of nonconceptuallyrepresenting X. So, a subject may rationally judgethat is asquare when taking her experience at face value not onlybecause her perception involves a nonconceptual way ofrepresenting a square butfor that reason, since such anonconceptual way of representing squares is part of the individuationconditions of the conceptsquare. Furthermore, the subject iscapable of reflecting on her reasons for so judging by asking herselfwhether the way things (nonconceptually) look to her is reason for herto make the judgment that she does. Brewer (2005) argues against suchproposals on the grounds that they do not manage to account for thesubject’s appreciation of her perceptionas a reasonfor a given judgment without being committed to a form offoundationalism with its various difficulties. If, however, perceptiondoes involve conceptual content, then according to Brewer, we canaccount for the subject’s appreciation of her perception as areason for a given judgment straightforwardly, since thesubject’s having a perception with conceptual content alreadyrequires that she grasps the rational relations perception with such acontent stands to other states with conceptual contents.

Advocates of the state/content distinction, discussed in section3, point out that although arguments for the nonconceptual nature ofperception are typically aimed at articulating the nature of thecontents of perception, that is, whether perception iscontent-nonconceptual or content-conceptual, the premises of thesearguments seem to support at most a “state” conclusion(Byrne 2003, 2005, Speaks 2005, Crowther 2006, Heck 2007). Forexample, the argument from the fine grained nature of experiencegenerally begins by arguing that one need not have the conceptsinvolved in the correct specification of the content of perception inorder to undergo a perception with that content, e.g., one need nothave the conceptred₂₇ in order to have aperception correctly specifiable in terms of such a concept. Butthough this is merely reason to hold that perception isstate-nonconceptual, the conclusion drawn from such an observation isthat thecontent of perception is nonconceptual. (Heck 2007also points this fault in the standard forms of the fineness of grainargument and proceeds to develop a form of the argument which hethinks avoids this problem and is genuinely directed at the“content” view.) However, if the distinction betweenconcept-dependent and concept-independent state types ultimately restson a distinction between the types of content involved in thesestates, as suggested in section3, then the transition from state-nonconceptualism tocontent-nonconceptualism (and from state-conceptualism tocontent-conceptualism) is a legitimate one. We would, then, have goodreason to consider the arguments above as directly relevant to thenonconceptualcontent of perception even though they oftenproceed via claims about the capacities required of a creature forhaving perceptual states.

4.2 Subpersonal computational states and nonconceptual content

The dominant paradigm within cognitive science involves postulatingrepresentational states at the subpersonal or subdoxastic levels (forphilosophical discussion see Stich 1978, Davies 1989). Examples arethe representational states implicated in tacit knowledge of the rulesof syntax. It is a fundamental tenet of a broadly Chomskyan approachto syntax that speakers are credited with tacit knowledge of a grammarfor their language and that this tacit knowledge is deployed inunderstanding spoken language. Yet when linguists give theoreticalspecifications of the syntactic rules contained within the grammarthey frequently employ concepts that are not in the conceptualrepertoire of the language-user. That is, the language-user isascribed knowledge of rules formulated in terms of concepts that hedoes not possess. A similar point holds for the representationalstates postulated in computational theories of vision such as that putforward by Marr (1980). The contents of such states are formulated interms of concepts (such as the concept of a zero-crossing) that areclearly not possessed by the average perceiver.

Why should it be thought that the language-user does not possess therelevant concepts? How could he grasp the rule if he did not possessthe concepts required to spell it out? The point is not just thatlanguage-users are not aware of the beliefs in question. Not allunconscious mental states are nonconceptual, e.g., my having thebelief that the earth revolves around the sun, often unconsciously,requires that I possess the concepts of which its content is composed.The point, rather, is that their representations of the linguisticrules are inferentially insulated from the rest of their beliefs andpropositional attitudes, in a way that is fundamentally incompatiblewith the holistic nature of conceptual contents. This is also the casefor representations at the early stages of visual processing. Therepresentations whose function is to register zero-crossings forsubsequent perceptual analysis, for example, operate independently ofthe person’s propositional attitudes, and so their content doesnot depend on the person’s conceptual repertoire.

Recently such a view of the nature of representational content hasbeen explored by Raftopoulos and Müller (2006). They suggest thatwhat makes the content of a representation nonconceptual is preciselythe fact that its content is insulated from other knowledge structuresavailable to the person. Indeed, according to Raftopoulos andMüller, “the existence of cognitively impenetrablemechanisms is a necessary and sufficient condition for nonconceptualcontent” (Raftopoulos and Müller 2006, p. 190). They reviewevidence to the effect that various early stages of visual processing,e.g., those responsible for attending to, and tracking, objects inspace and time, operate in purely bottom-up fashion. As such, theyargue, the representations underlying such capacities are produced ina conceptually unmediated way and hence their content isnonconceptual. Thus, they say, “[t]he content of the states ofzero-crossings is nonconceptual … because that content isretrieved bottom-up from a visual scene. To justify the nonconceptualcharacter of the contents of either phenomenal and subdoxastic stages[sic] … it suffices to invoke the fact that thesecontents are retrieved from visual scenes in conceptually unmediatedways” (Raftopoulos and Müller 2006, p. 215).

However, though Raftopoulos and Müller attempt to argue thatinsulation from the propositional attitudes of the person is anecessary and sufficient condition forany representation tohave nonconceptual content, this does not seem to be the case withrepresentations in general and certainpersonal levelrepresentations in particular. At least with regards topersonal level representations, Tye (1995) argues thatone’s conceptual capacities often are involved in determiningthe content of a perceptual representation, and yet that the contentof that representation is nonconceptual. This is the case, forexample, in the process of perceptually disambiguating an ambiguousfigure (e.g., the two faces/vase figure). Though the production of adisambiguated perceptual representation (e.g., experiencing the figureas a vase) might involve more than merely information recoveredbottom-up (e.g., it might employ the conceptvase), it doesnot follow that the resulting perception is conceptual. This is sosince for the content of a representation to be nonconceptual itmerely needs to be the case that onecan undergo a state withsuch representational content without possessing the concepts thatenter into a correct specification of its content, and it seemsreasonable that one can undergo the same disambiguated experience ofthe figure without possessing any concepts (and in particular, withoutpossessing the conceptvase).

A similar suggestion concerning cognitive penetrability and itspotential impact on the nonconceptual status of the penetrated statesis explored by Macpherson (2012, 2015). One of her favorite examplescomes from Delk & Fillenbaum (1965), who demonstrate thatfamiliarity with exemplars of a standardly colored category, e.g.,familiarity with (red) apples or (yellow) bananas, influences theperceived color of instances of that category. In a more recent study,Hansen et al. (2006) obtained the same effect by presenting subjectswith a picture of a banana on a uniform grey background and askingthem to adjust the color of the fruit until it appears grey. Theirresults show that “…subjects adjusted the banana to aslightly bluish hue--its opponent color--in order for it to appearneutral gray. At the point where the banana was actually achromatic,at the origin of the color space, it still appearedyellowish.” (p. 1367) In this case, it appears that one’sknowledge of bananas ‘penetrates’ one’s experienceof bananas. She calls this form of cognitive penetration, cognitivepenetration ‘lite’, where possession of some concepts maycausally influence the type of experience one undergoes. Yet sheargues that in these cases, though the token-experience of thebanana’s color is influenced by one’s prior knowledge– one undergoes a deeper yellow type of experience – it isan experience-type that one who does not have concepts pertaining tosuch colors can nonetheless undergo.

Thus, it is consistent with the notion of nonconceptual content thatthe representation in question is in fact produced in what Raftopoulosand Müller would call a conceptuallymediated way,whereby concepts in the subject’s possession influence thecontent of a personal level representation, without the content of therepresentation being specifiable in terms of concepts available to thesubject. That is, it does not seem to be anecessarycondition on the nonconceptual content of representations in generalthat they be insulated from the propositional attitudes; at least inthe case ofpersonal level representations this necessitycondition fails. Furthermore, Toribio (2014) argues that even in thecase of early vision, the specific target of Raftopoulos andMüller (2006), the claim that cognitive impenetrability is anecessary and sufficient condition on the contents of early visionbeing nonconceptual is false. (However, see Raftopoulos 2014a, 2014b,2017 for possible replies.)

Still, it is open to a critic of nonconceptual content to deny thatascriptions of content at the subpersonal level should be takenliterally. John McDowell takes this view (McDowell 1994b), suggestingthat subpersonal content is merely “as if” content (anargument for the same conclusion is offered in Searle 1990).Alternatively, it might be argued (as in Connolly 2010) that thesubpersonal information processing actually involves concepts.However, these lines of argument fly in the face not just of thepractice of cognitive scientists, but also of some importantphilosophical analyses of the notion of subpersonal representation(Burge 1986, Egan 1992, Bermúdez 1995).

4.3 Psychological explanation

In explaining the behavior of nonlinguistic and prelinguisticcreatures cognitive ethologists and developmental psychologists oftenappeal to representational states. Spelke, for example, attempts toexplain the development of the ability to perceptually organize thevisual array into unitary, persisting objects by appeal to aninfant’s capacity to form a representation of the visual surfacelayout and the presence of mechanisms following basic principles ofcohesion, boundedness, rigidity, and no action at a distance (Spelke1990). These capacities are claimed to explain the infant’srudimentary physical reasoning, underlying, for example, its abilityto track the motion of particular objects and exhibit certainexpectancy responses regarding the persistence of objects behindocclusions (as exhibited, for example, in Baillargeon’s (1987)famous drawbridge experiment, in which an infant is habituated to seea screen rotate 180°, and is dishabituated when the screen‘passes through’ an object placed behind it. Thedishabituation reflects the infant’s expectation that thescreen’s rotation will be blocked by the occluded object). Suchresearch suggests that an explanation of the infant’s behaviorshould appeal to the way the infant is perceptually aware of theworld, as segmented into objects that exhibit certain predictableregularities.

According to Spelke (1988), the content of those representationsappealed to in explaining the infant’s responses in tasks asthose mentioned above is conceptual content. The manifest behaviorreflects the infant’s incipient object concept and its capacityto utilize this concept in a primitive form of physical reasoning.However, if one has an account of what it is to possess a concept thatmakes it inappropriate to attribute mastery of the correspondingconcept to the creature whose behavior is being explained, then theappeal to representations in the explanation of its behavior is alsomotivation for the notion of nonconceptual representationalcontent.

One such understanding stresses the relation between possessingconcepts and being able to justify certain canonical judgmentsinvolving that concept, going on to argue that providingjustifications is a paradigmatically linguistic activity—amatter of identifying and articulating the reasons for a givenclassification, inference or judgment (McDowell 1994a). If this is thecase, then nonlinguistic and prelinguistic creatures will not be ableto possess concepts and those representations we attribute to them inexplaining their behavior will have to be nonconceptual. However,there is a variety of possible responses to this argument. It might beobjected, for example, that possessing a given concept simply requiresbeing able to makede facto justified judgments involvingthat concept rather than being able to justify judgments involvingthat concept. Or it might be objected that the ability to justifyjudgments involving some concept is not necessarily linguistic, sinceit is possible to identify the justification for a judgment withoutengaging in communication.

However, the argument from the need to provide psychologicalexplanations of the behavior of nonhuman animals and human infants tothe existence of nonconceptual content does not stand or fall with thethesis that concepts are necessarily linguistic. It has been arguedthat there is a distinction between two different types of thinking(Mithen 1996). Many students of the type of cognition engaged in byanimals and infants view it as being domain-specific and modular inimportant respects, best understood in terms of bodies of“knowledge” closely focused on particular aspects of thenatural and social worlds. An example of such domain-specificity isfound in the perceptual module Spelke suggests is responsible for thesegmentation of the visual array into objects according to certainbasic physical principles (Spelke 1990, 1994, Carey and Spelke 1996).These domain-specific modules are thought to have evolved separatelyand for specific purposes and are not integrated with each other(Hirschfeld and Gelman 1994). In contrast, many philosophers havesuggested that the type of conceptual thought engaged in bylanguage-users is essentially domain-general, systematic andproductive. Concept-possessors can generate an indefinite number ofnew thoughts from the concepts they possess and their thoughts obeywhat Evans (1982) has termed the generality constraint. As Evans says,“We cannot avoid thinking of a thought about an individualobjectx, to the effect that it isF, as theexercise of two separable capacities; one being the capacity to thinkofx, which could be equally exercised in thoughts aboutx to the effect that it isG orH; and theother being a conception of what it is to beF, which couldbe equally exercised in thoughts about other individuals, to theeffect that they areF” (Evans 1982, p. 75).

If the generality constraint is taken to be an essentialcharacteristic of conceptual thought (as it is by Evans 1982, Peacocke1992, and Heck 2007, for example) then it seems to follow that many(if not all) non-linguistic creatures are not capable of engaging inconceptual thought. For illustration we can look again atinfants’ object perception. Though the infant is capable ofrepresenting its environment in a particular way that explains itsexpectations regarding object motion and object persistence, the factthat such representations are domain-specific suggests that thischaracteristic of conceptual thought is violated. In particular, thedomain-specificity of this type of cognition makes it the case thatcapacities operative in constructing object representations are notones that can be engaged generally in contexts falling beyond thedomain of operation of the module. Thus, we have a violation of thegenerality constraint with respect to these representations. Thissuggests that the kinds of capacities that are involved in theinfant’s representation of the environment are not conceptual.So, if they are correctly described as representing the world at all,their representations must be nonconceptual. See also the discussionof analog magnitude representations in Beck (2012).

Of course, the plausibility of this way of motivating the notion ofnonconceptual content is a hostage to fortune in two important senses,one empirical and one philosophical. The argument rests upon anempirical claim about the appropriate way to explain the behavior ofnon-linguistic and pre-linguistic creatures—and in particularupon the assumption that it will not turn out to be possible toexplain such behavior in non-psychological terms. But the argumentalso depends upon the thesis that the domain of behavior explicable inpsychological terms extends further than the domain of conceptpossession—and this in turn depends upon a substantivephilosophical account of what it is to possess a concept. If, forexample, possessing the concept of an F simply requires being able todiscriminate Fs from the rest of the perceptual environment and/or toact on them in a suitable manner, then it is hard to see how anyevidence that animals and young infants represent the world will notalso be evidence that they represent the world conceptually.

5. Specifying nonconceptual content

We have so far been talking about nonconceptual content in verygeneral terms. If the notion of nonconceptual content is to be useful,however, we need to have a substantive account of what nonconceptualcontent consists in and how it is to be ascribed—an account thatwill be the equivalent at the nonconceptual level of, for example, theFregean account of concepts and concept possession.

Theories of nonconceptual content can be eitherglobalaccounts orlocal accounts of content. To say that they areglobal accounts is just to say that they purport to be accounts ofcontenttout court, rather than specifications of contentcircumscribed to a particular representational domain (such as thethree domains mentioned in the section above). What is central toglobal accounts of nonconceptual content is that they deny that eventhe paradigmatically conceptual propositional attitudes ever involveconceptual contents. As a result, such accounts of content areparticularly amenable to proponents of the “state” view ofthe nonconceptual debate (discussed above, section3). If all content is nonconceptual, the nonconceptual debate can only bemeaningfully carried out at the level of state types, not at the levelof content. Furthermore, since these accounts of nonconceptual contentare fundamentally non-contrastive, they are of limited use informulating the debate about nonconceptualcontent. As aresult, in what follows we will be more interested inlocalaccounts of nonconceptual content. However, before attending to thevarious local specifications of nonconceptual content we will brieflydiscuss two popularglobal accounts of nonconceptualcontent—possible world semantics and Russellian semantics (as weshall see, in discussing Tye’s account of the content ofperception, these global views can be modified so as to apply onlylocally).

The first global account of nonconceptual content, which was alreadymentioned above, is possible world semantics (see, e.g., Stalnaker1984, 1998, Lewis 1979, 1986). According to proponents of this view,the content of a mental representation consists of a proposition,where propositions are taken to be functions from possible worlds totruth values, or alternatively the set of possible worlds in which theproposition is true. Beliefs and other propositional attitudes arethen attitudes a person might hold towards such sets of possibleworlds. For example, if one believes that Norman shot Susan, one hasthe belief attitude towards those possible worlds in which Normanactually did shoot Susan. This is merely to say that one takes theactual world to be a member of the set of possible worlds in which theproposition is true. Similarly, having a perception is having arepresentational state the content of which is the set of possibleworlds in which the perception is correct.

The second global account of nonconceptual content is theRussellian account. Proponents of this view also hold thatthe content of a mental representation should be understood as aproposition. However, they follow Russell, broadly, in arguing thatpropositions are orderedn-tuples of objects and properties.Propositional attitudes are then relations between a person and anorderedn-tuple that is the proposition that is the contentof the attitude. For example, to have the belief that Norman shotSusan is to stand in the belief relation towards the proposition<<Norman, Susan>, being shot by>. (Theories falling underthis heading differ greatly; as representatives, see Salmon 1986,Soames 1987, Perry 1979, 1980, and other various articles in thecollection edited by Salmon and Soames 1988.) Similarly, the contentsof a perceptual representation are just the objects that arerepresented and the properties that appear to characterize them. Toperceive a blue carpet is to have a representational state the contentof which is the carpet and the property blueness, that is, <carpet,blueness>. (Some Russellians, e.g., Tye 1995, 2000, 2005, hold thatonly theproperties things appear to have constitute thecontent of perception.)

Both possible world semantics and Russellian accounts of content areconsidered to becoarse grained accounts of content (incontrast to the Fregeanfine grained account of content). Forexample, it seems to follow from the possible worlds account that athinker who entertains the thought expressed by the sentences‘bachelors are unmarried men’ and ‘unmarried men areunmarried men’ is related in both cases to the same set ofpossible worlds, and hence is having thoughts with the same content.Similarly regarding the Russellian account of content. A thinker whoentertains the thought expressed by the sentences ‘Jocasta isbeautiful’ and ‘Oedipus’ mother is beautiful’is related in both cases to the same objects and properties (sinceJocastajust is Oedipus’ mother), and hence is havingthoughts with the same content. Some have argued that this implicationand other related ones are detrimental to any coarse-grained accountof content, in particular when applied to the propositional attitudes.We will not review these arguments here, since they are elaboratedelsewhere. (For more detailed discussion of these global accounts ofcontent, see the entries onstructured propositions, propositional attitude reports, andbelief.)

Turning now to thelocal accounts of nonconceptual content,we can expect that a substantive account of nonconceptual content willbe tailored to a specific explanatory task—and each of the threemotivations discussed in the previous section for introducing thenotion will lend itself to a different such account.

The most developed proposal has come from Christopher Peacocke who hasproposed a radically externalist conception of nonconceptual contentaimed explicitly at explaining the nonconceptual content of perceptualstates (Peacocke 1992). Peacocke suggests that a given perceptualcontent should be specified in terms of the ways of filling out thespace around the perceiver that are consistent with thecontent’s being correct. For each minimally discriminable pointwithin the perceiver’s perceptual field (where these areidentified relative to an origin and axes centered in theperceiver’s body) we need to start by specifying whether it isoccupied by a surface and, if so, what the orientation, solidity, hue,brightness and saturation of that surface are. This specificationgives us the way the perceiver represents the environment. The contentof that representation is given by all the ways of filling out thespace around the perceiver in which the minimally discriminable pointshave the appropriate values. The representation is correct just if thespace around the perceiver is occupied in one of those ways. Peacockecalls this type of nonconceptual content positioned scenariocontent.

However, such a specification of the content of perception does notcompletely capture the way the world perceptually appears to us. Tworepresentations may have the same positioned scenario contents and yetdiffer in the way they represent the world to us as being. As anexample, upon looking at a 45° tilted square one might perceive itas a tilted square or as an upright diamond. In both cases ofperceiving the squareas a square andas a diamondthe positioned scenario content is the same; that is, in both casesthe ways of filling out the space around the observer that areconsistent with the correctness of the scenario content are the same.However, there is still a phenomenal difference between these two waysof representing the square.

To account for this difference in the way in which the same object isperceptually represented Peacocke argues that there is an additionallayer of nonconceptual content that he calls proto-propositionalcontent, which involves objects, properties, and relations rather thanconcepts (Peacocke 1992). This level of content accounts for the wayin which the square is perceived. When it is perceived as a square oneperceives a symmetry about the bisectors of the sides, and when it isperceived as a diamond one perceives a symmetry about the bisectors ofits angles. The content is nonconceptual since the observer need nothave the concepts of symmetry, or angle, or side, that are mentionedin specifying the content of his perception.

This proposal provides an attractive way of capturing the distinctivefeatures of the phenomenology of perception highlighted in theprevious section (see4.1 above). Peacocke’s account of nonconceptual content is analog,unit-free and possesses the appropriate fineness of grain.Nonetheless, as Cahen (2019) has recently argued, such an externalistaccount of nonconceptual content, though it may capture thephenomenology of perception, is in tension with internalistcommitments pertaining to the rational role of perception. Anothersource of difficulty is that it is not so clear how such an accountcan be applied beyond the domain of perception. It may well be thatcertain subpersonal states have such nonconceptualcontents—those associated with the subpersonal underpinnings ofvision are obvious candidates. But the representational statesimplicated in tacit knowledge of syntactic theory, for example, do notfit this model.

Tye suggests a different notion of robustly nonconceptual content ofperception (Tye 2005). According to Tye, perception involvesRussellian nonconceptual content. This is supposedly in conflict withPeacocke who, in discussing the experience of perceiving the tiltedsquare, claims at one point that “We must, in describing thefine-grained phenomenology, make use of the notion of thewayin which some property or relation is given in the experience”(Peacocke 2001a, p. 240). Appealing to away in which theproperty of being a square is perceived is in conflict withTye’s coarse-grained account of the content of perception whichis required for his broader representationalist account of consciousperception (see the entry onthe contents of perception (section 7) and the entry onrepresentational theories of consciousness). Thus, Tye proposes that in perceiving the tilted square, perceptiondoesn’t represent the property of being a square in two possibleways, but rather that in addition to representing theproperty of being square it represents various other properties; whichadditional properties are represented will determine the way thesquare appears. One suggestion is that when the figure is perceived asa square, the property of being tilted is represented, and when it isperceived as a diamond the property of being upright is represented.These are represented properties of theobject, not ways inwhich theproperty, being square, is represented.

Some theorists concerned with nonconceptual content at the subpersonallevel have favored a broadly teleological account of what this contentconsists in (drawing upon the teleological theory of content developedin Millikan 1984). This approach can be extended to covercontent-bearing states such as those implicated in linguisticunderstanding (see the entry onteleological theories of mental content). The key notion here is the proper function of, for example, themechanisms underlying a particular stage in visual informationprocessing (construed normatively in terms of what those mechanismsshould do, and usually underwritten by evolutionaryconsiderations). Proper functions arerelational, where thismeans that they are defined in relation to features of theenvironment. According to teleological theories, content can bespecified in terms of relational proper functions. So, for example,the content of a state in early visual processing might be specifiedin terms of the edges in the perceived environment that that stage hasevolved to identify. Given the particular features that a processingmechanism has been ‘designed’ or ‘selected’ todetect, it is functioning correctly when it responds appropriately tothe presence of those features, and incorrectly when it responds intheir absence (for example, to a sudden contrast in light intensitynot due to the presence of an edge). Correctness conditions are fixedwith reference to evolutionary design and past performance.

Turning to the third motivation for the notion of nonconceptualcontent, one might expect the notion of nonconceptual content involvedin psychological explanation to be essentially perceptual. Someauthors have argued, for example, that the representational states ofnon-linguistic creatures are essentially perceptual and tied to thecreature’s possibilities for action and reaction in theimmediate environment (see, for example, Dummett 1993 and Campbell1994, although neither of these two authors puts the point in terms ofnonconceptual content). This thought might be developed in conjunctionwith an ability-based understanding of nonconceptual content along thelines proposed by Adrian Cussins (Cussins 1990). The central featureof Cussins’s account of experiential content is that it shouldbe understood not in terms of notions such as truth andtruth-conditions, but rather in terms of the organism’sabilities to act upon the perceived environment. What the organismperceives (the content of its perception) is a distal environmentstructured in terms of the possibilities it affords for action. Thisconception of an ability-based conception of content yields a sharpdistinction between the success-governed level of proto-thoughts andthe truth-governed level of full-fledged thought. Proto-thought doesnot involve experience of an objective world.

Other authors concerned with explaining the behavior of non-linguisticand pre-linguistic creatures have proposed applying a richer notion ofnonconceptual content that is not essentially pragmatic and that canbe deployed in a style of explanation analogous to propositionalattitude explanation. Nonconceptual contents in this richer sense areproperly assessable for truth or falsity (rather than simply pragmaticsuccess or failure) and can serve as the objects of beliefs anddesires (or proto-beliefs and proto-desires). Bermúdez (1998)offers an account along these lines developing certain aspects ofPeacocke’s scenario content. In later work he extended thisapproach to the ascription of thoughts to non-linguistic andprelinguistic creatures, developing a model based on a version ofsuccess semantics and integrating what he termed proto-logicalinferential capacities (Bermúdez 2003).

6. The autonomy of nonconceptual content

One question that arises when thinking about nonconceptual content iswhether a thinker can be in states with nonconceptual content despitenot possessing any concepts at all. That is, can nonconceptual contentbe completely autonomous of conceptual content? The possibility of anaffirmative answer to this question is important for many theoristswho wish to employ the notion of nonconceptual content to explain thebehavior of nonlinguistic and prelinguistic creatures (depending, ofcourse, on how demanding their notion of a concept is). Similarly forthose theorists who hold both that subpersonal computational statespossess nonconceptual content and that the relevant modules can existin creatures not capable of any conceptual thought.

However, Christopher Peacocke offers an argument against the autonomythesis; concluding that “… nonconceptual content is not alevel whose nature is completely explicable without reference toconceptual contents … At the most basic level, conceptual andnonconceptual content must be elucidated simultaneously”(Peacocke 1992, pp. 90–91). His argument is based on aneo-Kantian understanding of the relation between experience of anobjective world and self-consciousness. In essence he suggests that nocreature can properly be attributed states with genuine spatialcontents unless it grasps the distal environment in a minimallyobjective way. A condition on grasping the distal environment in thisway is that the organism is capable of utilizing such states informing and updating an integrated representation of the layout of itsenvironment and is thus able to reidentify particular locations withinit. The requirement of concept possession comes in because thesecapacities essentially involve an ability to represent both thespatial configuration of the environment and one’s own changingposition within that environment—and, he suggests, this would beimpossible for a creature lacking a rudimentary concept of thefirst-person.

This argument is powerful, but can be challenged. Supporters of theautonomy thesis, beginning with Bermúdez (1994), have suggestedthat the interrelated capacities to represent the spatialconfiguration of the environment and to represent one’s ownlocation within the environment need not depend on the possession of aprimitive first-person concept. Rather, these capacities can beunderstood at the nonconceptual level. In particular, Bermúdez(1995, 1998) appeals to the notion of a nonconceptual point of view toexplain the interdependence of spatial awareness of the distalenvironment and awareness of one’s own location within thatenvironment at the nonconceptual level. For further discussion of thistype of nonconceptual self-consciousness see section7.

Peacocke himself eventually accepted the Autonomy Thesis on thegrounds that the capacities mentioned in the conditions for a state tohave genuine objective spatial content do not necessitate a conceptualnotion of the first-person; indeed, he agrees with Bermúdezthat these capacities can be accounted for nonconceptually and withoutinvolvingany first-person notion (Peacocke 2002).

7. Nonconceptual content: Problems, prospects, and applications

It should be clear from the preceding sections that the basic idea ofnonconceptual content provides a promising tool for tackling a rangeof problems in the philosophy of mind and cognition. Allowing that acreature’s representational capacities can outstrip itsconceptual capacities makes it possible for philosophers and cognitivescientists to study aspects of cognition and behavior that remainoutside the scope of more traditional approaches—fromsubpersonal computational mechanisms to the psychological states ofnon-human animals and human infants to the nature of perceptualexperience.

It should be recognized, however, that there may well not be a unitarynotion of nonconceptual content applicable to these various domains.We need to distinguish between the formal notion of nonconceptualcontent (the idea of a way of representing the world that is notconstrained by conceptual capacities) and the different concreteproposals for developing this basic idea (such as Peacocke’snotion of scenario content). There is potential for serious confusionif it is assumed that a specific theory of nonconceptual contentproposed for one area can be unproblematically applied to anotherarea. It may well turn out, for example, that fundamentally differentnotions of nonconceptual content are required for subpersonalcomputational states and perceptual experiences.

Indeed, recent work on nonconceptual content has expanded beyond itsorigins in visual perception to other modalities and even beyondperception. Thus, Young (2015, 2020), drawing on a broadempirical literature, argues that olfactive representations, unlikeconceptual representations, are not concatenatively compositional. Assuch they are best understood as representational but nonconceptual.Others have moved beyond perception and have argued for anonconceptual account of the contents of emotions (see e.g., Gunther2003, Tye 2008, Döring 2009, Wringe 2015, Tappolet 2016 and 2022;for recent arguments to the contrary, see, e.g., Slaby 2020,Demmerling 2020) and possibly other affective-evaluative experiences(Mitchell 2018). Beck (2012) has argued that certain cognitive stateswith analog magnitude content (primitive representations of spatial,temporal, numerical, and other magnitudes) are best understood asnonconceptual, as they do not satisfy Evans’ GeneralityConstraint, discussed above (but see replies by Gillett 2014, Grey2014, and Beck’s further replies in his 2014). Mental imageryprovides another interesting phenomenon to which, due to itsquasi-perceptual nature, the notion of nonconceptual content seemsreadily applicable (however, see Arcangeli 2020 for an argument to theeffect that imagination, unlike perception, is importantlyconcept-dependent).

While the philosophy of perception has been the area where theconceptual/nonconceptual distinction has been most discussed, thetheory of nonconceptual content has been fruitfully employed in otherareas of philosophy. A case in point is the study ofself-consciousness (see the entry onself-consciousness). In 1998 two books independently developed theories of nonconceptualself-consciousness (Hurley 1998 and Bermúdez 1998).Bermúdez and Hurley both emphasized prelinguistic forms ofself-consciousness, arguing that linguistic self-reference andassociated modes of conceptual self-consciousness both emerge from,and need to understood in terms of, much more primitive forms ofself-specifying information. For Bermúdez these include:

information in visual and other exteroceptive perceptualmodalities specifying the perceiver’s own spatial and kinematicproperties
information via affordances that specify the perceiver’scapacities for acting upon the distal environment
information through somatic proprioception, kinesthesia, andactive touch about both the disposition of body parts and the limitsof the bodily self.
self-locating information about the subject’s changinglocation in objective space exploited in navigation involvingcognitive maps

Several of these forms of nonconceptual self-consciousness have beendiscussed more recently by Kristina Musholt (2013, 2015), who arguesthat though such forms of awareness provide necessarilyself-specifying information, they should not properly be consideredmodes of self-conscious, as they contain no explicitly self-referringcomponent. Rather than thinking of the self as being part of the(explicit) representational contents of perception and proprioception,she argues it should be thought of as being implicit in the mode ofpresentation (Musholt 2015).

Christopher Peacocke’s recent work on consciousness andself-consciousness identifies what he terms a “nonconceptualparent” of the first-person concept (Peacocke 2014). What hecalls Degree 1 of subject involvement “is exhibited by a subjectwho enjoys states with nonconceptual content that is objective, andwhich represents the subject as standing in spatial relations to otherobjects and events in the spatial world” (2014, p. 35). LikeBermúdez and Hurley, Peacocke sees Gibson’s analysis ofvision as illuminating a basic form of nonconceptualself-consciousness. Unlike them, however, he does not think thatnonconceptually self-conscious subjects are necessarily embodied.

Primitive, nonconceptual, forms of self-consciousness are alsocritical for the acquisition of the first-person concept and forgrounding thoughts whose contents have an important propertyintroduced in Shoemaker (1968). This is the property of being immuneto error through misidentification relative to the first-personconcept, defined so that properties attributed in such thoughts mustbe properties of the thinker, if they are instantiated at all. Cahenand Musholt (2017) argue that it is precisely because certainrepresentational states with nonconceptual content, e.g., perceptionand proprioception, do not contain an explicitly self-referring (orself-identifying) component, but do supply necessarily self-specifyinginformation that, while such modes of awareness are not themselvesimmune to error through misidentification, they can serve as the basisfor first-personal judgments that are so immune.

The claim that perception has nonconceptual content has furtherimplications. In his most recent book, Block (2023) utilizes thenotion in demarcating the border between perception and cognition. Itis also instrumental in supporting his own theory of consciousness, asagainst higher-order and global workspace theories of phenomenalconsciousness, as well as in supporting his influential distinctionbetween phenomenal and access consciousness.

The notion of nonconceptual content also promises to shed light on thenature of implicit cognition and procedural knowledge (know-how).Bermúdez and Cahen (2022) argue that the involvement ofrepresentations with nonconceptual contents best explains cases ofpurported implicit cognition, e.g., as involved in linguisticcompetency and in cases of know-how, such as one’s knowing howto ride a bike. In such cases, the involvement of content bearing,representational, states explains why it is correct to attribute tothe subject some piece of (linguistic or procedural) knowledge, andthe nonconceptual character of the content of those states explainswhy the subject is not in a position to articulate the knowledge shepossesses. This is why such knowledge remains implicit. Such anaccount of know-how is interesting as it carves a middle groundbetween intellectualist accounts (e.g., Stanly and Williamson 2001,Stanley 2011) who insist that such knowledge as is available to thesubject must be conceptual, which then make it unclear why it shouldremain inarticulable, and ‘ability accounts’ (e.g.,Dreyfus 2005, Noë 2005, Glick 2012, Wiggins 2012), who seeminsensitive to the explanatory significance of knowledge as involvinginternal representations. Toribio (2015) also appeals to nonconceptualrepresentations in developing her non-dispositionalanti-intellectualist account of know-how, and Ferretti (2021) arguesthat know-how involves a nonconceptual capacity to select ways ofacting.

The notion of nonconceptual content has also been applied to shedlight on potential forms of metarepresentational cognition,particularly in nonhuman animals. Many philosophers have argued thatcertain types of metarepresentation (thinking about thinking) are inprinciple unavailable to non-linguistic creatures (e.g., Davidson1975, Bermúdez 2003). Potential counter-evidence has come frommodels of animal behavior and animal cognition that seem to dependupon attributing primitive forms of metarepresentation outside thelinguistic realm. Shea (2014) argues that the detection of rewardprediction error, which is a staple of influential models of learningin cognitive science, should be understood as a nonconceptual form ofmetarepresentation. Shea’s arguments are (sympathetically)rejected in Carruthers (2021). Carruthers, however, proposes adifferent candidate for nonconceptual metarepresentation, suggestingthat we look instead at evaluations of cognitive effort, of the sortthat occur when a creature, for example, compares an option with higheffort but high reward, to an option with lower effort, butcorrespondingly lower reward.

As we consider these applications, however, it should be reiteratedthat the notion of nonconceptual content is essentiallycontrastive—and the significance of the notion depends upon theparticular way of understanding concepts with which it is contrasted.Some psychological states that would count as nonconceptual for atheorist with a rich and demanding notion of what it is to possess aconcept would be conceptual for a theorist with a more relaxed view ofconcepts. Before we can plausibly claim to have a full understandingof the possibilities of the notion of nonconceptual content, we needto have a much clearer view than we currently have of what it is topossess a concept. As we saw earlier, on the most minimal view ofconcepts, a thinker can be credited with a concept ofFsprovided that he can discriminate things that areF fromthings that are notF. A richer view of concepts might demandthat the thinker have a full appreciation of the grounds on which onemight judge something to be anF. The most demanding view ofconcepts might require the thinker to be able to justify and defendthe judgment that something is anF. Clearly, differentlocations on this broad spectrum will generate different ways ofthinking about what is to count as nonconceptual content—as wellas different assessments of the overall significance of thenotion.

Bibliography

Anenhanced copy of the bibliography for this entry with links at PhilPapers is available.

Many of the papers listed below are reprinted in the anthologyEssays on Nonconceptual Content, York Gunther (ed.),Cambridge, MA: MIT Press, 2003.

Ablondi, F., 2002. Kelly and McDowell on perceptual content.Electronic Journal of Analytic Philosophy, 7, URL =http://ejap.louisiana.edu/EJAP/2002/Ablondi.pdf.
Arcangeli, M., 2021. The conceptual nature of imaginative content.Synthese, 199(1): 3189–3205.
Ayers, M., 2002. Is perceptual content ever conceptual?Philosophical Books, 43(1): 5–17.
–––, 2004. Sense experience, concepts, andcontent—objections to Davidson and McDowell. In R. Schumacher(ed.),Perception and Reality: From Descartes to the Present,Paderborn: Mentis.
Baillargeon, R., 1987. Object permanence in 3.5- and 4.5-month-oldinfants.Developmental Psychology, 23(5): 655–664.
Beck, J., 2012. The generality constraint and the structure ofthought.Mind, 124: 563–600.
–––, 2014. Analogue magnitudes, the GeneralityConstraint, and nonconceptual thought.Mind, 123(492):1155–1165.
Bengson, J., Grube, E., & Korman, D. Z., 2011. A new frameworkfor conceptualism. Noûs, 45: 167–189.
Bermúdez, J. L., 1994. Peacocke’sargument against the autonomy of nonconceptual content.Mind andLanguage, 9: 402–418.
–––, 1995. Nonconceptualcontent: From perceptual experience to subpersonal computationalstates.Mind and Language, 10: 333–369.
–––, 1998.The Paradoxof Self-Consciousness, Cambridge MA: MIT Press.
–––, 2003.ThinkingWithout Words, New York. Oxford University Press.
–––, 2007. What is at stakein the debate about nonconceptual content?PhilosophicalPerspectives, 21(1): 55–72.
Bermúdez, J. L., and Cahen, A., 2022. Implicit cognition inrelation to the conceptual/nonconceptual distinction. In J. R.Thompson, (Ed.),The Routledge Handbook of Philosophy and ImplicitCognition, pp. 69–78, New York, NY: Routledge.
Block, N., 2023.The Border Between Seeing and Thinking.Oxford University Press.
Brewer, B., 1999.Perception andReason, Oxford: Oxford University Press.
Brewer, B., 2002. Précis of Perception and Reason, andresponse to commentator (Michael Ayers).PhilosophicalBooks, 43(1): 1–4, 18–22.
–––, 2005. Perceptual experience has conceptualcontent. In E. Sosa and M. Steup (eds.),Contemporary Debates inEpistemology, Oxford: Blackwell.
Burge, T., 1986. Individualism andpsychology.Philosophical Review, 95: 3–46.
Byrne, A., 2003. Consciousness and nonconceptual content.Philosophical Studies, 113: 261–274.
–––, 2005. Perception and conceptual content. InE. Sosa and M. Steup (eds.),Contemporary Debates inEpistemology, Oxford: Blackwell.
–––, 2020. Concepts, belief, and perception. InC. Demmerling and D. Schröder (Eds.),Concepts in Thought,Action, and Emotion: New Essays, New York, NY: Routledge.
Cahen, A., 2019. Nonconceptual apprehension: The problem ofperceptual reason-giving.Synthese, 196(6):2355–2383
Cahen, A. and Musholt, K., 2017. Perception, nonconceptualcontent, and immunity to error through misidentification.Inquiry, 60(7): 703–723
Campbell, J., 1994.Past, Space and Self, Cambridge MA: MIT Press
–––, 2005. Information processing, phenomenalconsciousness, and Molyneux’s question. In J. L. Bermúdez(ed.),Thought, Reference, and Experience, Oxford: ClarendonPress.
Carey, S. and Spelke E. S., 1996. Science and core knowledge.Philosophy of Science, 63(4): 515–533.
Carruthers, P. 2021. Explicit nonconceptual metacognition.Philosophical Studies, 178(7): 2337–2356.
Chuard, P., 2006. Demonstrative concepts withoutre-identification.Philosophical Studies, 130(2):153–201.
–––, 2007. The riches of experience.Journalof Consciousness Studies, 14(9–10): 20–42.
Clark, A., 2001. Visual experience and motor action: Are the bondstoo tight?Philosophical Review, 110: 495–519.
Coliva, A., 2003. The argument from the finer-grained content ofcolour experiences: A redefinition of its role within the debatebetween McDowell and non-conceptual theorists.Dialectica:International Journal of Philosophy of Knowledge, 57(1):57–70
Connolly, K., 2011. Does perception outstrip our concepts infineness of grain?Ratio, 24: 243–258.
Crane, T., 1988a. The waterfallillusion.Analysis, 48: 142–147.
–––, 1988b. Concepts inperception.Analysis, 48: 150–153.
–––, 1992. The nonconceptual content ofexperience. In T. Crane (ed.),The Contents of Experience,Cambridge: Cambridge University Press.
Crowther, T. M., 2006. Two conceptions of conceptualism andnonconceptualism.Erkenntnis, 65: 245–276.
Cussins, A., 1990. The connectionistconstruction of concepts. In M. Boden (ed.),The Philosophy ofArtificial Intelligence, Oxford: Oxford University Press.
Davidson, D., 1975. Thought and talk. In S. Guttenplan (Ed.),Mind and Language. Oxford, Oxford University Press.
Davies, M., 1989. Tacit knowledge and subdoxastic states. In A. George (ed.),Reflections on Chomsky, Oxford: Blackwell.
DeBellis, M., 1995.Music and Conceptualization,Cambridge: Cambridge University Press.
Delk, J. L. and Fillenbaum, S., 1965. Differences in perceivedcolour as a function of characteristic color.The American Journalof Psychology, 78(2): 290–293.
Demmerling, C., 2020. Emotions and the conceptual space of humanlife. In C. Demmerling and D. Schröder (Eds.),Concepts inThought, Action, and Emotion: New Essays, New York, NY:Routledge.
Dokic, J. and Pacherie, É., 2001. Shades and concepts.Analysis, 61(271): 193–202.
Döring, S., 2009. The logic of emotional experience:Noninferentiality and the problem of conflict without contradiction.Emotion Review, 1(3): 240–247
Dretske, F., 1969.Seeing and Knowing, Chicago:University of Chicago Press.
–––, 1981.Knowledge andthe Flow of Information, Cambridge, MA: MIT Press.
–––, 1995.Naturalizing the Mind,Cambridge, MA: MIT Press.
Dreyfus, H., 2005. Overcoming the myth of the mental: howphilosophers can profit from the phenomenology of everyday experience.Proceedings and Addresses of the American PhilosophicalAssociation, 79: 43–49.
Duhau, L., 2014. Perceptual nonconceptualism: Disentangling thedebate between content and state nonconceptualism.EuropeanJournal of Philosophy, 22(3): 358–370.
Dummett, M., 1993. Origins of AnalyticalPhilosophy, London: Duckworth.
Egan, F., 1992. Individualism, computation andperceptual content.Mind, 101: 443–459.
Evans, G., 1982.The Varieties of Reference, Oxford:Oxford University Press.
Ferretti, G., 2021. Anti-intellectualist motor knowledge.Synthese, 198: 10733–10763.
Fodor, J., 1975.The Language of Thought, New York:Crowell.
–––, 1987.Psychosemantics: The Problem ofMeaning in the Philosophy of Mind, Cambridge MA: MIT Press.
–––, 1998.Concepts: Where Cognitive ScienceWent Wrong, New York: Oxford University Press.
–––, 2008.LOT 2: The Language of ThoughtRevisited, Oxford: Oxford University Press.
Forman, D., 2006. Learning and the necessity of non-conceptualcontent in Sellars’s “Empiricism and the Philosophy ofMind”. In M. P. Wolf and M. Lance (eds.),TheSelf-Correcting Enterprise: Essays on Wilfrid Sellars, New York:Rodopi.
Gillett, G., 2014. Concepts, consciousness, and counting bypigeons.Mind, 123: 1147–1153.
Glick, E., 2012. Abilities and know-how attributions. In J. Brownand M. Gerken (Eds.),Knowledge Ascriptions (pp.120–139). Oxford: Oxford University Press.
Gray, D. 2014. Counting-ish creatures and conceptual content.Mind, 123: 1141–1146.
Gunther, H. Y., 2001. Content, illusion, partition.Philosophical Studies, 102: 185–202.
–––, 2003. Emotion and force. In Y. Gunther(ed.),Essays on Nonconceptual Content. Cambridge MA: MITPress
Hamlyn, D. W., 1994. Perception, sensation and non-conceptualcontent.Philosophical Quarterly, 44: 139–153.
Hansen, T., Olkkonen, M., Walter, S., & Gegenfurtner, K. R.,2006. Memory modulates color appearance.Nature Neuroscience,9: 1367–1368
Heck, R. G., 2000. Nonconceptual content and the space of reasons.Philosophical Review, 109: 483–523.
–––, 2007. Are there different kinds of content?In J. Cohen and B. McLaughlin (eds.),Contemporary Debates in thePhilosophy of Mind, Oxford: Blackwell.
Hirschfeld, L. A. and Gelman, S. A.,1994.Mapping the Mind: Domain-Specificity in Cognition andCulture, Cambridge: Cambridge University Press.
Hopp, W., 2009. Conceptualism and the Myth of the Given.European Journal of Philosophy, 17: 363–385.
Hurley, S., 1998.Consciousness inAction, Cambridge MA: MIT Press.
Kelly, S. D., 2001a. Demonstrative concepts and experience.The Philosophical Review, 110(3): 397–420.
–––, 2001b. Thenon-conceptual content of perceptual experience: Situation dependenceand fineness of grain.Philosophy and PhenomenologicalResearch, 62(3): 601–608.
Laurier, D., 2004. Nonconceptual contents vs. nonconceptualstates.Grazer Philosophische Studien, 68: 23–43.
Lerman, H., 2010. Non-conceptual experiential content andreason-giving.Philosophy and Phenomenological Research, 81(1): 1–23.
Levine, J., 2010. Demonstrative thought.Mind andLanguage, 25: 169–195.
Lewis, D. K., 1979. Attitudes de dicto and de se.Philosophical Review, 88: 513–543.
–––, 1986.On the Plurality of Worlds,Oxford: Blackwell.
Luntley, M., 2003. Nonconceptual content and the sound of music.Mind and Language, 18(4): 402–426.
Macpherson F., 2012. Cognitive penetration of colour experience:Rethinking the issue in light of an indirect mechanism.Philosophyand Phenomenological Research, 84 (1): 24–62.
–––, 2015. Cognitive penetration andnonconceptual content. In J. Zeimbekis & A. Raftopoulos (eds.),The Cognitive Penetrability of Perception: New PhilosophicalPerspectives. Oxford: Oxford University Press, 2015.
McDowell, J., 1994a.Mind and World, Cambridge MA:Harvard University Press.
–––, 1994b. The content of perceptualexperience.Philosophical Quarterly, 44: 190–205.
–––, 2006. Conceptual capacities in perception.Reprinted in J. McDowell (ed.),Having the World in View: Essayson Kant, Hegel, and Sellars. Cambridge, MA: Harvard UniversityPress, 2009.
Marr, D., 1980.Vision, Boston:Houghton Mifflin.
Martin, M., 1992. Perception, concepts, and memory.ThePhilosophical Review, 101: 745–763.
Mellor, D. H., 1988. Crane’s waterfallillusion.Analysis, 48: 147–150.
Millar, A., 1991.Reasons and Experience, Oxford: OxfordUniversity Press.
Millikan, R. G., 1984.Language, Thought and Other BiologicalCategories, Cambridge MA: MIT Press.
Mitchell, J., 2018. On the non-conceptual content ofaffective-evaluative experience.Synthese.https://doi.org/10.1007/s11229-018-1872-y.
Mithen, S., 1996.The Prehistory of the Mind, London:Thames and Hudson.
Musholt, K., 2013. Self-consciousness and nonconceptual content.Philosophical Studies, 163(3): 649–672.
–––, 2015.Thinking about Oneself: FromNonconceptual Content to the Concept of a Self. Cambridge MA: MITPress
Noë, A., 1999. Thought and experience.AmericanPhilosophical Quarterly, 36: 257–265.
–––, 2005. Against intellectualism.Analysis, 65: 278–290.
Peacocke, C., 1983.Sense and Content, Oxford: OxfordUniversity Press.
–––, 1986. Analoguecontent.Proceedings of the Aristotelian Society, 60:1–17.
–––, 1989. Perceptual content. In J. Almog, J.Perry, and H. Wettstein (eds.),Themes from Kaplan, New York:Oxford University Press.
–––, 1992.A Study ofConcepts, Cambridge MA: MIT Press.
–––, 1994. Nonconceptualcontent: Kinds, rationales and relations.Mind and Language,9: 419–429.
–––, 1998. Nonconceptualcontent defended.Philosophy and Phenomenological Research,58: 381–388.
–––, 2001a. Does perceptionhave a nonconceptual content?Journal of Philosophy, 98:239–264.
–––, 2001b. Phenomenology and nonconceptualcontent.Philosophy and Phenomenological Research, 62(3):609–615.
–––, 2002. Postscript toPeacocke 1994. In Y. Gunther (ed.),Essays on NonconceptualContent, Cambridge MA: MIT Press.
–––, 2004a. Interrelations: Concepts, knowledge,reference and structure.Mind and Language, 19(1):85–98.
–––, 2004b.The Realm of Reason,Oxford: Oxford University Press.
–––, 2014.The Mirror of the World:Subjects, Consciousness, and Self-Consciousness. Oxford: OxfordUniversity Press
Perry, J., 1979. The problem of the essential indexical.Noûs, 13: 3–21.
–––, 1980. Belief and acceptance.MidwestStudies in Philosophy, 5: 533–542.
Raffman, D., 1995. On the persistence of phenomenology. In T.Metzinger (ed.),Conscious Experience, Munich: ImprintAcademic Verlag.
Raftopoulos, A. and Müller, V. C., 2006. The phenomenalcontent of experience.Mind and Language, 21(2):187–219.
Raftopoulos, A., 2014a. The cognitive impenetrability of thecontent of early vision is a necessary and sufficient condition forpurely nonconceptual content.Philosophical Psychology,27(5): 601–620.
–––, 2014b. Nonconceptual content: A reply toToribio’s “Nonconceptualism and the cognitiveimpenetrability of early vision”.PhilosophicalPsychology, 27(5): 643–651.
–––, 2017. Cognitive penetration lite andnonconceptual content.Erkenntnis, 82: 1097–1122.
Raleigh, T., 2017. Against an inferentialist dogma.Synthese, 194: 1397–1421.
Roskies, A. L., 2008. A new argument for nonconceptual content.Philosophy and Phenomenological Research, 76:633–659.
–––, 2010. ‘That’ responsedoesn’t work: Against a demonstrative defense of conceptualism.Noûs, 44: 112–134.
Salmon, N., 1986.Frege’s Puzzle, Cambridge MA: MITPress.
Salmon, N. and Soames, S., (eds.), 1988.Propositions andAttitudes, New York: Oxford University Press.
Schellenberg S., 2013. A trilemma about mental content. In S.,Joseph (ed.),Mind, Reason, and Being-in-the-world.London: Routledge.
–––, 2018.The Unity of Perception: Content,Consciousness, Evidence. Oxford: Oxford University Press.
Schmidt, E., 2015.Modest Nonconceptualism. Dordrecht:Springer.
Searle, J., 1990. Consciousness, explanatory inversion andcognitive science.Behavioral and Brain Sciences, 13:585–642.
Sedivy, S., 1996. Must conceptually informedperceptual experience involve non-conceptual content?CanadianJournal of Philosophy, 26: 413–431.
–––, 2006. Nonconceptual epicycles.EuropeanReview of Philosophy, 6: 33–66.
Settegast, S., 2023. Fineness of grain and the hylomorphism ofexperience.Synthese, 201(6): 1–29.
Shea, N., 2014. Reward prediction error signals aremeta-representational.Noûs, 48: 314–341.
Shoemaker, S., 1968. Self-reference and self-awareness.Journal of Philosophy, 65: 555–567.
Slaby, J., 2020. A challenge to perceptual theories of emotion. InC. Demmerling and D. Schröder (Eds.),Concepts in Thought,Action, and Emotion: New Essays, New York, NY: Routledge.
Soames, S., 1987. Direct reference, propositional attitudes andsemantic content.Philosophical Topics, 15: 47–87.
Speaks, J., 2005. Is there a problem about nonconceptual content?Philosophical Review, 114: 359–398.
Spelke, E. S., 1988. The origins of physical knowledge. In L.Weiskrantz (ed.),Thought without language, Oxford: ClarendonPress.
–––, 1990. Principles of object perception.Cognitive Science, 14(1): 29–56.
–––, 1994. Initial knowledge: six suggestions.Cognition, 50: 431–445.
Stalnaker, R., 1984.Inquiry, Cambridge, MA: MITPress.
–––, 1998. What might nonconceptual content be?In E. Villanueva (ed.),Concepts (Philosophical Issues,Volume 9), Atascadero: Ridgeview.
Stanley, J. 2011.Know How. Oxford: Oxford UniversityPress.
Stanley, J., and Williamson, T. 2001. Knowing how.The Journalof Philosophy, 98: 411–444.
Stich, S., 1978. Beliefs and subdoxastic states.Philosophy of Science, 45: 499–518.
Szabo, T. G. and Hawthorne, J. (eds.), 2006.PerceptualExperience, Oxford: Oxford University Press.
Tappolet, C., 2016.Emotions, Values, and Agency. Oxford:Oxford University Press.
–––, 2022.Philosophy of Emotion: AContemporary Introduction, 1st edition. New York, NY:Routledge.
Toribio, J., 2008. State versus content: The unfair trial ofperceptual nonconceptualism.Erkenntnis, 69(3):351–361.
–––, 2014. Nonconceptualism and the cognitiveimpenetrability of early vision.Philosophical Psychology,27(5): 621–642.
–––, 2015. Opacity, know-how states, and theircontent.Disputatio, 7: 61–83.
Tye, M., 1995.Ten Problems of Consciousness, CambridgeMA: MIT Press.
–––, 2000.Consciousness, Color andContent, Cambridge MA: MIT Press.
–––, 2005. On the nonconceptual content ofexperience. In M. E. Reicher & J. C. Marek (eds.),Experienceand Analysis, Proceedings of the 27th International WittgensteinSymposium, Vienna: Öbv & hpt.
–––, 2006. Nonconceptual content, richness, andfineness of grain. In T. G. Szabo and J. Hawthorne (eds.),Perceptual Experience, Oxford: Oxford University Press.
–––, 2008. The experience of emotion: Anintentionalist theory.Revue Internationale de philosophie,243: 25–50.
Van Cleve, J., 2012. Defining and defending nonconceptual contentsand states.Philosophical Perspectives, 26 (1):411–430.
Vision, G., 2009. Fixing perceptual belief.The PhilosophicalQuarterly, 59(235): 292–314.
Wiggins, D., 2012. Practical knowledge: knowing how to and knowingthat.Mind, 121: 97–130.
Wright, W., 2003. McDowell, demonstrative concepts, andnonconceptual representational content.Disputatio, 14:39–54.
Wringe, B., 2015. The contents of perception and the contents ofemotions.Noûs, 49(2): 275–297.
Young, B. D., 2015. Formative non-conceptual content.Journalof Consciousness Studies, 22: 201–214.
–––, 2020. Smell’s puzzlingdiscrepancy: Gifted discrimination, yet pitiful identification.Mind and Language, 35(1): 90–114.

Academic Tools

How to cite this entry.
Preview the PDF version of this entry at theFriends of the SEP Society.
Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

Bibliography ofThe Nature of Contents in PhilPapers.

Open access to the SEP is made possible by a world-wide funding initiative.
The Encyclopedia Now Needs Your Support
Please Read How You Can Help Keep the Encyclopedia Free

Browse

About

Support SEP

Mirror Sites

View this site from another server:

USA (Main Site)Philosophy, Stanford University

Info about mirror sites

Library of Congress Catalog Data: ISSN 1095-5054

Movatterモバイル変換