Conceptual determination of experience, or Peacocke’s “sensations”, is shown by his own example:
…we see the array
. . .
. . .
. . .
. . .
as three columns of dots rather than as four rows. (1983)
There is no reason, prima facie, that we shouldn’t see four rows rather than three columns. This seems to suggest that concepts (of columns perhaps) are determining our “sensations”. I don’t know about the empirical research on this particular array, but I suspect that there must be someone somewhere who does indeed see four rows instead of three columns. The point is, however, that the sensations alone don’t determine the experience. Indeed, if it were just a question of sensations alone, it wouldn’t be an experience. Experience is “under an aspect”.
And the same applies to another Peacocke example One person may see the object we call a “sphere”, but not apply the concept [sphere] to it. Another person will indeed apply the concept [sphere]. More probably, he won’t apply the concept at all. The concept will belong, as it were, to the object sphere. That is, he will non-cognitively experience a sphere as a sphere. The person without the concept [sphere] will still, however, either apply a concept to the sphere, or a concept will belong to the sphere (as in the other example).
Peacocke himself says that “sensational properties do not determine representational content”No, concepts are part of the story too. However, Peacockegoes on to say that “grouping properties [the arrays into columns not rows] are sensational rather than representational”Peacocke admits that the array example “seems to suggest that we are concerned with representational, not a sensational, property: the concept of a column enters the content”The problem is I don’t understand Peacocke’s argument for saying that “grouping [is a] sensational property”. Of course “grouping” will rely on sensations, but they will not wholly determine the “grouping”. Again. Peacocke is making a distinction between “grouping” (which I think is conceptual and he doesn’t) and “sensations”. Indeed he is saying that “grouping” is not conceptual (or “representational”). Again, on this particular aspect, Peacocke doesn’t go into detail. He says himself that in “switches of aspect the sensational properties…[remain] constant”.
Martin Davies spots a dualism in Peacocke and others. He says that the idea is that “a sensational (non-representational) substrate upon which the representational superstructure…is erected.” (324)
Or
>) Representational superstructure
>) Sensational (non-representational) substrate
And, as with Block’s AC and PC, Davies quite happily accepts “non-representational properties of experience without embracing the idea of a sensational substrate”.
McGinn points out the position represented by philosophers like Peacocke. They accept
Prerepresentational yet intrinsic level of description of experiences: that is, a level of description that is phenomenal yet noncontentful… (1989)
Perhaps the problem is that Peacocke thinks in terms of “protopropositional content” (1992, pg 79) – that is, nonconceptual content. This is almost – or indeed literally – an acceptance that content prior to “propositional judgment” can’t be conceptual. Propositions almost literally make concepts. That is why “protopropositional content” is nonconceptual content. But if we don’t accept a necessary linguistic basis for concepts, we need not believe in Peacocke’s non-conceptual content.
Davies points out that Peacocke is obviously happy to accept the conceptualisations of “sensational properties”. He says that Peacocke shows us examples “pairs of experience with the same sensational properties but different representational properties” (1996). Yet Peacocke also shows us “pairs of experiences with the same representational properties but different sensational properties”. Does this, however, show us that these “different sensational properties” are nonconceptual? It seems to hint at the fact or possibility that they are not representing of anything. But I have attempted to show, based on my reading of his ‘Sensations and the Content of Experience’ paper, that this doesn’t seem to work.
At least this is the case with Peacocke’s array examples. The other cases are based on technical empirical psychological research on subjective experience, which is difficult to comment upon philosophically.
Peacocke, however, appears to make an obvious mistake. He says that a person “waking up in an unfamiliar position or place [will have] minimal representational content” . Unfamiliarity doesn’t entail lack of conceptual content. This person will still conceptualise his “unfamiliar position or place”. This only displays Peacocke’s linguistic or propositional bias. Of course this person may not be able to describe or form “propositional judgements” about the unfamiliar position or place, but that may be irrelevant to conceptual and/or representational content. When the place or position becomes “rich” with “representational content” , this may simply be a case of applying “descriptive labels…to the array” (Tye, 1990). Propositions and propositional judgements are simply additions to conceptual content that already exist. Tye says that “computational routines [“propositional judgments”] process this activity and assign an appropriate descriptive term”.
I suspect that Peacocke’s propositional or linguistic bias causes problemsfor various of his positions. However, I doubt that even Peacocke would deny his linguistic or propositional bias (though he wouldn’t accept the word “bias”). The bias itself wouldn’t cause problems in his eyes. It is simply the reality of conceptual experience, I assume, to Peacocke. (Someone like Geach was explicit about his propositional or linguistic bias vis-à-vis concepts. See his 1959.)
According to Tye, “phenomenal content” is by definition nonconceptual . He too gives us his own dualism between “phenomenal content” and belief:
a) A content is classified as phenomenal only if it is nonconceptual and poised [“for use by the cognitive centres”].
against
b) Beliefs…lie within the conceptual arena, rather than providing inputs to it.
This is almost foundationalist in that “phenomenal content” is seen as input to be worked upon later (if only split seconds later). Beliefs, on the other hand, are outputs. And if this is truly foundationalist, then a) above would be “the Given”. It certainly seems foundationalist when Tye makes the epistemic point that “phenomenal content” is “poised for use by the cognitive centres”. That is, it comes before, epistemically, such use.
Tye makes the strange point that “phenomenal content” is “representational”. How can this be so? How can something represent “that there are such and such co-instantiated locational and nonlocational features” without concepts?
References
Block, Ned, ‘Begging the Question Against Phenomenal Consciousness’, from Behavioral and Brain Sciences, 15, 1992, 205-206
- ‘On a Confusion About a Function of Consciousness’, from Behavioral and Brain Sciences, 18, 1995
Church, Jennifer, ‘Fallacies or Analyses’, from Behavioral and Brain Sciences, 18, 1995, 251-252
Davies, Martin, ‘Externalism and Experience’, from Philosophy and Cognitive Science: Categories, Consciousnessand Reasoning, 1996
Kant, Immanuel, Prolegomena To Any Future Metaphysics
Peacocke, Christopher, A Study of Concepts, Cambridge, Mass.:MIT Press, 1992
- ‘Sensation and Content of Experience’, from Sense and Content, Oxford: Clarendon Press, 1983
Searle, John, ‘Who is computing with the brain?’, from Behavioral and Brain Sciences 13, 1990
Tye, Michael, ‘A Representational Theory of Pains and Their Phenomenal Character’, from Philosophical Perspectives, Vol.9, 1990, 223-239
Tye, Davies, Peacocke and Block on Animal Concepts
Finally, what do contemporary philosophers have to say on these subjects? Take Tye.
I may agree with Tye when he says that
Having the concept F requires, on some accounts, having the ability to use the linguistic term ‘F’ correctly. (1990)
But it all depends on what precisely the concept F is. If it’s [infinity] then yes indeed we would probably need “to use the linguistic term”. But not if the concept is [cat], and [cat] is not tied to “cat”. Of course I’m using the word “cat” within the square brackets because it’s shorthand for whatever constitutes, say, a dog’s concept [cat]. So perhaps I should symbolise it [C]. The problem with this that people wouldn’t know what object the dog’s concept is about. So I use a linguistic and English term within the brackets.
If Tye had said
Having concepts require the ability to use linguistic terms correctly.
He would have been, I think, incorrect.
Following on from that sentence, Tye says
On other accounts, concept possession requires the ability to represent in thought and belief that something falls under the concept…
There’s no problem with the above, as long as “thought”, “belief” and “falls under a concept” are not taken sententially or linguistically. And there’s no obvious reason why they should be. A dog must think that the rattling dog-chain means that it will be going for a walk. It will believe that a walk will be forthcoming. And the dog-chain must fall under the concept [dog chain] or, again, instead of using the English word “dog chain” which will be unknown to the dog, it will fall under the concept [C], where “C” simply symbolises whatever constitutes the dog’s concept.
On the word “thought” itself. Some philosophers are sceptical about animal thought. Take this strange and unelaborated description of a monkey:
…while we may be prepared to say that it knows [that it’s safe up a tree], we may be less happy to say that the monkey thinks that it is safe.
Prima facie, how can the monkey know without thinking? This writer includes reasoning, believing, reflecting, calculating and deliberating as examples of thought. I think monkeys do all these things. At first I hesitated with the word “calculating”. But only because I over-sophisticated the term by thinking in terms of abstract mathematical calculations. However, since my topic is concepts, I can’t go into these broader areas of thought. An interesting question remains however. Is it animals’ lack of concepts that exclude them from all these cognitive states? Or they have no concepts because they can’t think?
Perhaps what is motivating the idea of “non-conceptual content” is that animals have “experiences” but don’t necessarily deploy concepts. Martin Davies writes:
…the experiences of …certain creatures, who, arguably are not deployers of concepts at all. (1996)
And later:
…a creature that does not attain the full glory of conceptualised mentation, yet which enjoys conscious experience with non-conceptual content…
All the above depends on what animal we are talking about and what Davies means by the word “concept”.Indeed it is hard to fuse “experience” and “non-conceptual content” together in the first place, and not just from a quasi-Kantian position on experience and concepts. If Davies is talking about floor lice, what he says may well be correct. If he’s talking about dogs, monkeys, etc, I’m not so certain (the intermediary animal cases are, as ever, vague).
Is the fact that animals don’t use a language or a human language causing the bias against animal concept use? Davies doesn’t say. If one takes a Fodorean “language-infested” view of “mentation”, then one would probably agree with Davies. If one were a non-computationalist or non-Fodorean, one may question the linguistic bias of Davies’s position.
An even more explicit example of this linguistic bias can be seen in Christopher Peacocke:
The representational content of a perceptual experience has to be given by aproposition, or set of propositions, which specifies the way the experience represents the world to be. (1983)
Peacocke, in the above, is distinguishing “representational content”, which is propositionally specifiable, from pure unconceptualised sensations. And, again later, Peacocke displays his linguistic or sentential bias:
The content of an experience is to be distinguished from the content of a judgment caused by the experience.
So not only do we have a dualism of “sensations” (the “contents of experience”) and “judgement” (which is “caused by the experience”), we have the specifically intellectualist and probably linguistic term “judgment”. Not even Kant meant something propositional or sentential by his “judgment” in his philosophy of mind and world. But Peacocke, judging by what he wrote earlier in this paper, meant applying “a proposition, or set of propositions” to the experience by “judgement”.
Perhaps if everything weren’t so language-infested Peacocke would have more of a case. If “representational content” is “given by a proposition”, then perhaps so too with concepts .
Indeed Peacocke states his own dualism, or troilism, explicitly:
…we need a threefold distinction [of experience] between sensation, perception, and judgment…
Peacocke, in note 3, quotes another philosopher to back-up his case:
… “sensation, taken by itself, implies neither the conception nor the belief of any external object…Perception implies an immediate conviction and belief of something external"…
The reasoning behind Peacocke and Davies’s position may be that if animals, or certain animals, are always non-conceptual creatures, then we too, on a Darwinian perspective, may start off as non-conceptual creatures. Not just as infants, but as adults too. That is, pure phenomenal consciousness is common to both humans and animals. However, as adults (and humans), phenomenal consciousness and sensations are later conceptualised.
There is a dualism (another one) here between phenomenal consciousness and conceptual consciousness. But if Davies is being Darwinian in accepting that we share phenomenal consciousness with animals, why can’t he be equally Darwinian by accepting that – some – animals share concepts with us? Why should concepts be sentence-shaped?
Is this a disguised foundationalist dream once again? Davies himself makes a distinction between
a)perceptual content is the same kind of content as the content of judgement and belief… (1996)
And, alternatively
b)…perceptual content is a distinct kind of content, different from belief content.
This is dualism in a pure form. Passage a) is very Davidsonian. Judgements/beliefs and perception are as one. Passage b), on the other hand, gives us “uninterpreted” content, separate, we may say, from “all schemes and science”. Of course, I wouldn’t necessarily use the word “science” or even “schemes”, but I would say: Separate from all concepts.
Ned Block also makes his own distinction between “representation” and “intentional representation” (note 4, 1995). He says that that an animal has an experience that is “representational”. However, it is not an “intentional representation”. This is how Block makes the distinction:
#) Intentional representation = “representation under concepts”
##) Representation= “representation without any concepts”
The mistake Block makes appears to be so obvious that I read his passage over and over again to see what was going on. The mistake is simply this. The animal in question “doesn’t possess the concept of a donut or a torus”. We can accept that. However, the animal may “represent space as being filled in a donut-like way”. Again, that is acceptable. So, yes, this animal won’t have our concept [donut] or our concept [torus]. However, it may have its own concept [C] of the donut and likewise its own concept [C] of the torus. This is why Block allows the animal representations. That is, it “represents space as being filled in a donut-like way without any concepts”. This experience has “representational content without intentional content”. So, apparently, it has representations because its experience is of something “donut-like”. But it isn’t intentional or conceptual because it doesn’t have our concept [donut]. But it may, as I said, have its own. This appears to be obviously wrong. It displays a linguistic bias and basis for all concepts. And therefore excludes, by definition, all animals from having conceptual content of the said experience. But, logically, it would mean that a fellow human being without the concept [donut] would only have a representation of the donut, not an intentional representation of it.
Even someone who has the concept [donut] must have experienced the donut under other concepts before he applied the concept [donut]. And not just the basic Kantian concept [object] or [thing]. These are true atomic concepts which are the building blocks of later concepts. However, before the object we call a “donut” fell under the concept [donut], and after it fell under the concept [object] or [thing], other concepts would have been applied or belonged to the donut. For example, [white thing], [round thing], [small round thing], etc. Even the animal, without the concept [white thing] etc, would possibly have its own alternative non-linguistic or non-human alternatives.
I also have a problem with Block’s use of the term “representation”. You can only represent something as something. Or it is a representation of something. Therefore one needs concepts, not necessarily linguistic ones, of something and concepts of a thing as something.
The problem here may be accounted for by what Block says himself (again in note 4). He says that “P-consciousness is not an intentional property”. I agree. He also says that “P-conscious content cannot be reduced to or identified with intentional content”. Again, I agree. He also qualifies these distinctions by saying that “intentional differences can make a P-conscious difference”. He also says that “P-consciousness is often representational”. However, he is still hinting at something that I don’t accept. That PC can exist without intentional or representational (that is, conceptual) content. The distinctions he makes are possibly real and worthwhile. However, PC is like a finger which can’t exist without a hand. And the hand, in this case, is conceptual content (concepts). Of course a finger is distinct from a hand, but, as yet, I haven’t seen a functioning finger without a hand.
(Note. I have taken on board Block’s philosophy, but his comments on the psychological empirical research in this paper are harder to decipher.)
