The Private Language Argument (I)

Abstract

This is a survey paper divided into two parts; in this issue of Crítica appears part I; it deals with what I shall call the strong line of the PLA.* In a second part —forthcoming— I shall develop further the strong line till nowadays and I shall also present the weak line.
Discussing the nature, scope and results of the PLA is very ambitious. Some philosophers see the PLA as a devastating argument against Cartesian Metaphysics and its Philosophy of Mind whereas others see only superficial linguistic platitudes in the PLA which have nothing to do with real philosophical problems. I shall review what both parties have to say and thus provide an introduction to this much debated topic.
1. A.J. Ayer’s Interpretation
Ayer sees no difficulty in a Robinson Crusoe’s acquisition and use of a PL. He could get one such language by private ostensive definitions. According to Ayer if there is a difficulty for Robinson’s private tests there will also be a difficulty for public tests.
Ayer thinks he can refute the thesis which holds that there cannot be a PL. According to J.J. Thomson, Ayer misses the target for he goes on proving two quite irrelevant thesis, namely:
(1) That there is, in fact, a language not understood by anyone besides the speaker.
(2) That sensation reports can be understood by someone else besides the speaker.
But according to Medlin, Ayer thinks that Private is a twofold concept.
Private1: a language a person uses to refer to his private experiences.
Private2: a language that ‘might indirectly convey some information to others without meaning to them exactly what it means to him’.
On Medlin’s view what Ayer tries to avoid is the implication from Private1 to Private2. If there is no implication from Private1 to Private2, then it will be possible to build a PL and there will be something else beside the expression of e.g. pain.
This much could be accepted by a defender of the PLA, but Ayer does not leave things in that state; instead of attacking the implication from Private1 to Private2 he sets forth to attack thesis (2) and confuses the whole issue.
Ayer also offered a diagnosis of the reasons why S cannot be a name. This diagnosis has pervaded a great deal of further discussion regarding the PLA. It consists in the two following conditions: S cannot be a name unless
(i) It is understood by other people.
(ii) The user could observe the object S designates.
He later came to accept the condition (i) and thus gave up his belief in a PL. But his diagnosis brought up many problems.
2. Strawson’s version
Strawson introduced the notion of Criterion in his discussion. According to him, what makes a PL private is the empirical fact that such a language refers to sensations and that sensations are private. Strawson also distinguishes between a strong and a weak thesis. The strong thesis says that no word names sensations; the weak thesis says that words name sensations only under certain conditions. The weak thesis becomes the strong one in this manner: names require identification and identification requires criteria, but self-ascription is criterionless (aeusserungen), thus there are no names for sensations. Strawson considered this an excess and recommended a more lenient thesis about criteria.
3. Malcolm’s defense
Malcolm felt that Strawson’s interpretation was misleading and mistaken. In the first place Strawson did not understand that the PLA is philosophical argument and that the PL has to be a language necessarily private. Sensation language is not a PL; it becomes private when sensations are built, as private objects. Malcolm distinguished between the two arguments: the internal one based on the notion of a rule having the form of a Reductio and the external one based on the idea of transferring one’s private mental concepts. Malcolm questions the intelligibility of such a transference and thus eliminates the external argument. According to the internal argument, the idea of a rule that couldn’t be neither correct nor incorrect is the unintelligible one and, if the notion of rule is not intelligible, the idea of private names or words won’t be either.
Regarding the criteria, Malcolm thought that Strawson read that notion in a strong behaviouristic way and thus was led to his mistaken views on sensations.
4. The Symposium between Castañeda, Chappell and Thomson
Castañeda made a strong effort trying to clarify the sense of a PL, by distinguishing between various senses of Language and Privacy. Castañeda contested the main premises in Malcolm’s argument proposing a number of candidates that could serve the purpose of determining that S was applied correctly or incorrectly; these candidates were other experiences of private objects, other words, memories, law like relationships among these private objects or inductions made from these objects. Castañeda saw no special difficulty in proposing examples of PL.
J.F. Thomson and V. Chappell co-symposiasts complained that the existence of an argument concerning PLs wasn’t clear at all. Thomson said he couldn’t understand the reasons behind the very strong premiss ‘nobody besides its professor could understand the language of sensations’. Chappell said that Castañeda did not pay attention to the type of language under discussion, namely, a language that nobody apart from his possessor could understand. Thus Castañeda committed an Ignoratio Elenchi. Besides this, Chappell objected to the above mentioned candidates use of S being correct or incorrect, since they were useless in deciding whether any use of these candidates would deprive the PL of its privacy. Further on Chappell set forth to show that even granting the use of the candidates they would provide no help at all because, for example, a recourse to other private objects would only touch one’s own conviction, not the correctness of S’s application.
5. J.W. Cook’s clarifications
Cook tried to satisfy Thomson’s misgivings and to dispell Castañeda’s confusions. In order to achieve this, he formulated an argument whose main premise stated that a necessary condition for knowing a sensation is to feel the sensation itself. This premise was the source of the philosophical privacy and so the source of the idea of a PL. if one accepts this premise the conclusion will follow that no sensation words could be taught and thus that the language of sensations is a PL. Then, Cook goes on to show the senselessness of that premise —not its falsity—. For this he uses contrasts of the ordinary language that bring out the oddness of sentences like ‘nobody could have my pains’. A sort of supersemantics seems to lie behind Cook’s charges of senselessness.
6. The dispute on Private Rules
Castañeda and J.J. Thomson objected to Malcolm’s contention that there couldn’t be private rules because these couldn’t be disobeyed; according to them there are such rules which only thinking of obeying them is to obey them, as, for example:
‘If you have pain, think that you have pain’. In these cases understanding the rule implies obeying it.
C. Ginet answered this objection noticing that “even in the case of the most easily understood sort of rule it is possible for someone to think he understands the rule when he does not”. Thus in no case thinking that one obeys a rule is to obey it.
7. PLA and Verificationism
Why —asked J.J. Thomson— cannot S be a name? Why is it impossible to follow a rule privately? All the discussion about impossibilities and senselessness is related to two philosophical theses, namely, Verificationism and Physicalism. According to the first, no sentence has meaning unless it is verifiable; according to the second, nothing counts as providing verification but what is physical. It is because PLA’s defenders hold these thesis that private naming, private rules and private checks appear as unintelligible to them. But as Verificationism is in discredit because of the charge of circularity, there is no acceptable reason to reject PLs.
8. ALP and conditions for Proposisionhood
A. Kenny tried to overcome the main obstacle of the PLA, that is to show conclusively why ‘This is S’ is not intelligible when philosophical privacy is the case. According to him, the theory of proposition laid down in the Tractatus is being used to reject ‘This is S’ as a proposition; no verificationism is involved in such rejection.
Kenny distinguishes three ‘facts’ which could serve to introduce philosophical privacy, namely, incommunicability which is epistemological (‘I cannot know your pains’) inalienability which is by ownership (‘Nobody could have my pains’) and a mixture of these two (‘I cannot know your pains because I cannot have them’). All three are dismissed and thus no philosophical privacy could be introduced. There is only one ‘fact’ which can introduce privacy, namely, secrecy, but this one doesn’t constitute a philosophical problem. So, ‘pain’ is not philosophically private and Kenny provides an explanation of Strawson’s and Casteñeda’s misinterpretations.
Kenny also rejects the idea of private names. His answer is a further elaboration of Malcolm’s and Chappell’s responses. Contra J.J. Thomson and others he claims that neither verificationism nor scepticism on Memory is required in order to reject ‘This is S’ as a proposition. ‘This is S’ is not a proposition because the three possible answers to the question ‘what do you mean by S’ fail. If it is answered “By S I mean the impression ‘S’ that I remember” it would be necessary to have a memory-copy that would have to be compared to the present impression, and two possibilities appear: they could fail to agree or not: if they cannot agree, memory is of no help because anything could do and if they can agree, memory is of no help either because something else would be needed. The third answer is this: ‘By S I mean X physical correlate’; this will do, but there won’t be any privacy left.
The failure of ‘this is S’ as a Proposition has something in common with the failure of ‘I have a pain’ as a description and with the failure of ‘I know I have a pain’ as an epistemological assertion. All these failures stem from the theory of immediate experience or acquaintance which lies deep at the source of the idea of a PL.

Enrique Villanueva

Notas a pie de página
* I shall be using PLA “Private Language Argument’ and PL for ‘Private Language’.