This essay is going to talk about psychologism and behaviourism and whether it is possible to steer a middle course between them, comparing McDowell’s and Dummett’s account of meaning.
When building a theory of meaning in order to explain in what the mastery of a language consists, the main problem, according to Heck, concerns the integration of language with the speakers’ and listeners’ conscious mental life. This integration is also the origin of all the discussions regarding psychologism and behaviourism, as it will be demonstrated in this essay.
Although this integration has been analysed in many different disciplines such as psychology, linguistics, philosophy and even cognitive sciences including informatics, it is still unclear whether or not thought precedes language.
Two accounts are considered here. One states that mind shapes language; the other states that ‘language shapes the mind’, in the sense that language and mind are involved at the same time. So, according to the former, thought precedes the linguistic expression. According to the latter, they are simultaneous. This difference is fundamental if one is to understand the difference between behaviourism and psychologism. Behaviourism dictates that linguisitic ability can be explained merely in terms of the speakers’ and listeners’ behaviour, whereas according to psychologism, the mental aspect of speech is located behind linguistic behaviour.
First of all, it is worth noting that McDowell’s aim to build a theory of name parallels Dummett’s aim to build a theory of meaning; they both want to establish a theory, of which the knowledge would suffice for understanding a language. Another shared point is that they both use the Fregean notion of sense in their theories. According to McDowell (1977, p.165), Dummett’s use of the notion of sense is precisely ‘to capture a notion of meaning which makes it true that a theory of meaning is a theory of understanding’. For his part, McDowell (1987) demonstrates that, even adopting the Fregean notion of sense it is still possible to be modest, despite of what Dummett thinks, Dummett’s conception of the theory of sense depends on the notion that it is possible to explain contents- if there are any- ‘’as from outside’’. Since the idea that it is possible to explain contents ‘as from outside’ (Dummett (1976) cited in McDowell (1987, p. 90)) is precisely what makes Dummett’s theory immodest, it follows that modesty would result in repudiating the notion of sense altogether, according to Dummett (McDowell (1987)). However, according to McDowell, Frege does not believe that it is possible to explain contents as from outside, and yet he does not repudiate the notion of sense at all (Mc Dowell, 1987, p.101). In fact, Frege is exactly the one who introduced that notion, actually.
Frege introduced ‘sense’ in opposition to ‘reference’ primarily to give an account of the bearerless names so that ‘a name without a bearer could have a sense in exactly the same way as a name with a bearer’ (Frege (1892) cited in McDowell (1977, p.172)). According to Frege, a bearerless name is a name; however it has only an intrinsic characterization and not an extrinsic one, which means that it has only a sense without a reference. Indeed, the difference between reference and sense corresponds precisely to the one between extrinsic and intrinsic characterization. The former concerns the name with the actual/external world, while the latter is concerned with the inner world of the speakers. This dichotomy between the inner world of the speakers and the external world is crucial in understanding the multiple contrasts that McDowell examines and, ultimately, the difference between behaviourism and psychologism.
Indeed, behaviourism focuses on the extrinsic characterizations, the ones in the external/actual world of the speakers and listeners. Psychologism focuses on the intrinsic ones, the ones that are in the inner world of the speakers and listeners. Psychologism employs the speaker’s mind as the protagonist while behaviourism uses the speaker’s behaviour as the protagonist. Therefore, due to the notion of sense, Frege gives an account of the mental aspect of speech, abstracting from the actual world. As McDowell (1977, p.173) states, in the Fregean view ‘thought relates to objects with an essential indirectness’, therefore ‘whether the object exists or not would then be incidental to the availability of the thought’. Reading Frege is this way, he appears to sustain the priority of thought. However, in this case the priority of thought is not over language, as stated for Grice in Heck’s paper, ‘’Use and Meaning’’, but over reality. Moreover, the apparent contradiction is that, even if Frege prefers thought over reality, he detests Psychologism. But, as McDowell (1977, p.168) states, there is a difference between psychological and psychologism: ‘phsychologism’ is a theory or tendency, which is based on the psychology of the speakers and listeners in this case. ‘Psychological’ is only an adjective that indicates a relation with psychology. Therefore, according to this distinction, Frege is able to give an account of the mental aspect of the speech act, thanks to the notion of sense, without, however, adopting psychologism; he only uses a psychological element.
Furthermore, the non -Fregean view associates the mental aspect of speech with the reasonableness attributed to the speaker by his/her audience (McDowell, 1977, p.172). The fundamental proposition is that the interpreter ascribes beliefs to the subject in making sense of his behaviour, in the light of reason, which means assuming that if a speaker uses a certain name it is to talk about an existing entity. This concept of reasonableness can be associated with the concept of the priority of reason, in opposition to the priority of thought that will to be analyzed in relation to Dummett’s account of the mental aspect of the speech act (Heck, 1998). Actually, Dummett uses Frege’s notion of sense in order to build upon his theory of meaning. However he criticizes the fact that ‘sense is supposed by Frege to be something objective’ (Dummett, 1975, p.65). Frege’s belief derives from his exigency of constructing a theory of sense within the framework of a realistic theory of meaning in terms of truth-conditions (Dummett, 1975, p.65), something that Dummett initially tries to do, but, after observing that some sentences are ‘undecidable’, he then realizes that a verificationist theory leads him closer to the aim.
Dummett builds a theory of meaning in order to explain philosophically in what a mastery of a language consists. However, he then realizes that the theory of meaning for the language is not sufficient, because ‘the account of what it is to have such knowledge can only be given in terms of the ‘practical ability which the speaker displays in using sentences of the language’ (Heck (2007)). This claim, which refers back to the fact that ‘the speaker will manifest his knowledge of the theory of meaning by his actual use of language’ (Heck (2007)), shows that Dummett ‘clings’ to the thesis, that Heck calls ‘the Use-Meaning Thesis’, according to which, ‘meaning is, in some way or other, determined by use’ (Heck, 2007, p.1).
The theory of meaning is divided into three parts: the first is the core theory, or theory of reference, the second is the theory of sense, and the third is the theory of force. ‘The theory of reference determines recursively the application to each sentence of that notion which is taken as central in the given theory of meaning’ (Dummett, 1975, p. 57). First of all, the central notion is truth, which is necessary in order to give an account of meaning. However the main difficulty for taking a speaker’s understanding of a sentence to consist in knowledge of its truth conditions is the fact that for some sentences it is not possible to decide whether they are true or false. This is true especially in the case of the subjunctive conditional, the past tense and quantification over unsurveyable or infinite totalities.
Therefore, after sketching a generalization of the intuitionistic theory of meaning for the language of mathematics, Dummett switches the attention to verification and falsification, because they are effective notions. ‘In a verificationist or falsificationist theory of meaning the theory of reference specifies the application to each sentence of the central notion of the theory so that ‘the speaker will manifest his knowledge of the condition for its application by his actual use of language’’((Heck (2007)). The verificationist and the falsificationist theories of meaning get closer in order to meet the requirement to incorporate into the theory of sense an account of the basis on which to judge the truth values of the sentences, because ‘they explain meanings in terms of actual human capacities for the recognition of truth’ (Dummett, 1975, p. 66).
However, they are still not sufficient to meet the requirement because they don’t provide ‘a firm foundation for a claim to know what meaning essentially is’, they only help to draw ‘an outline of the manner in which every feature of the use of a sentence can be given in terms of its meaning as specified by a recursive stipulation of the application to it of that central notion chosen’ ((Dummett, 1975, p. 67).
In addition to Dummett’s own admission to having failed to meet the ultimate requirement, McDowell accuses Dummett mainly of the fact that his notion of use of meaning/language falls into the same mistake that Quine made, which is the fact that, conceiving the language behaviouristically, he did not leave space for the rationality of the rational agents-the mental aspect of speech (Heck, 2007, p. 13).
Therefore, although Dummett’s theory seems complete and well articulated it is accused of having a ‘big’ issue. However, Heck (2007) analyses Dummett’s position in depth, in relation to McDowell’s objection. He discovers that, even if at first sight Dummett’s argument does not seem to give an account of the mental speech in language use, this is not actually true. Dummett, as well as Grice, if well analysed and developed, proposes a notion of use, which explains the role of the mental aspect of language.
Grice combines the Use-Meaning Thesis – the idea that meaning is determined by use- with the priority of thought over language, characterizing use in terms of the contents of mental states, a grasp of which needs to be understood prior to understand the actual speech act. These contents do not concern the external conditions under which the speakers make various utterances but the internal ones i.e. the psychological conditions such as beliefs, intentions etc. Therefore the Gricean conception is content-laden, but not meaning-laden, because it ‘does not help itself to any notion of linguistic content’ (Heck, 2007, p. 19). Therefore, Grice’s account provides an explanation of the mental aspect of language without accepting a meaning-laden conception, meeting precisely what McDowell aimed for.
Dummett, on the other hand, reaches the same objective but in a different way, since he rejects the priority of thought over language since he argues that it is ‘at best useless to appeal to a speaker’s possession of particular concepts in attempting to explain what it is for her words to express them’ (Heck, 2007, p. 19). Heck (2007) states that, even if he rejects the priority of thought to language in terms of the contents of sentences which need to be grasped prior to our understanding of them, he still give an account of the role of the mental aspect because he sustains the priority of reason to any linguistic capacity. This admission permits ‘to characterize use in terms of psychological notions (…) and yet refuse to appeal to any prior capacity to entertain thoughts with the very contents expressed by the utterances’ (Heck, 2007, p.19).
Therefore, Dummett’s theory, which is at the same time content-free and non –behavioristic, if well interpreted, can give an account of the role of the mental aspect in the use of language and confute McDowell’s view, according to which, the only way of ‘registering the role of mind’ in our use of language is by describing use in terms of the contents of speech acts’’(McDowell (1998) cited in Heck (2007, p.13)). For McDowell the only way to account for the mental aspect of speech is to accept that use is content-laden. Dummett rejects the idea that the meaning of a sentence can be taken to be the thought it is used to express.
Developing Dummett’s account in the light of McDowell’s objection, Heck (2007) shows that Dummett’s theory remains intact to his attacks, and, in particular, that McDowell’s conviction that ‘Dummett’s refusal to countenance a meaning-laden notion of use is driven by a flawed epistemology of understanding’ (Heck, 2007, p.2) is wrong. This flawed epistemology of understanding, which can also be called ‘the sense-datum conception of understanding’ (Heck, 2007, p.3) considers understanding as the listeners/speakers’ reconstruction of the meaning from simply facts about what sounds were produced under what circumstances, which are the only facts immediately available to them when they experience someone making an utterance (Heck, 2007, p.2). Therefore, according to McDowell, Dummett’s belief that experience is content-free is what makes his notion of use not content –laden. However, Heck (2007) proves that if Dummett’s notion of use is not content-laden is not because he thinks that experience is content-free, but because he does not attribute to the speakers/listeners the ability to grasp the content of the sentence prior to understanding, as stated above. Therefore, according to Heck, Dummett, despite McDowell’s accusations, is successful in giving an account of the mental aspect of speech.
McDowell (1987), in his paper ‘’In defence of Modesty’’, gives an account of the mental aspect of speech, while drawing a distinction between behaviourism and psychologism in relation to modesty and immodesty –or full-bloodedness- with regard to Dummett’s theory of meaning, thus sustaining the validity of modesty. The main difference between modesty and immodesty is that ‘Modesty, in relation to meaning, is the denial that we can ‘hope to give an account of the concepts expressible by the primitive vocabulary of a language’’ (McDowell, 1997, p.105). According to the immodest-or full-blooded-theory of sense sustained by Dummett, however, it is possible to do that ‘as from outside’. McDowell (1987) claims that, even embracing behaviourism -as immodesty requires- and trying to avoid psychologism, it is not possible to locate the mental aspect of speech anywhere except behind linguistic behaviour, therefore falling into psychologism. On the other hand, according to McDowell, to accept modesty would avoid this contradiction, because as the outward aspect of linguistic behaviour is essentially content-involving, the mind’s role in speech is part of what one presents to others, not something behind it. Moreover, McDowell underlines the fact that this outward aspect is ‘presentable’ only to those who already understand that specific language in which the thought is formulated.
This last specification is crucial because if it is necessary to already understand the language used in carrying the content-involving outward aspect of a certain linguistic behaviour, it is not possible to characterize language use ‘as from outside’ that language employed, which was Dummett’s aim. According to McDowell (1987, p. 90), trying to characterize language use ‘as from outside’ would only generate an endless recursive system of languages. Giving a theory of language in a distinct language-object language- and then ‘as from outside’ that language in a second language-the background language-, it would then be necessary to give a theory of language ‘as from outside’ that second language and continue like this endlessly. This system would be only helpful to the contents expressible in the object language, but, according to McDowell, ‘Dummett’s idea is that a proper theory of meaning for a language would be formulated ‘as from outside’ content altogether’ (McDowell, 1987, p. 91). In McDowell’s opinion, then, Dummett’s idea is fallacious.
The problem of recursion is connected with the one of circularity, which is one of the main obstacles facing the construction of a theory of meaning. Circularity is well shown by a passage in which Dummett explains the concept ‘square’ using that same concept. If the ‘explicans’ term -the term which explicates- contains the ‘explicandum’ one-the one which is supposed to be explicated- then the argument generated is a question begging, because, instead of answering the question, it uses that very same concept which it was supposed to explain. This applies to the theory of meaning because, even using only one language and so giving a theory of language as from inside that language, it is still very probable to come up against circularity. Indeed, a theory of meaning itself lies on the ultimate presumption of giving a full meaning to the concept of meaning, which is, in my opinion, intrinsically ‘question begging’, and that is why I agree with McDowell’s opinion that giving a theory of meaning as from outside is not possible, or at least not in a non-recursive and not question-begging way.
In fact, in another paper called ‘Dummett on truth conditions and meaning’ John McDowell (1985, p.356) states that this ‘alleged’ circularity applies only to languages with undecidable sentences i.e. sentences of which it is not possible to decide the truth of falsity. Because if all sentences were effectively decidable, according to McDowell, it would be easily possible to equate understanding a sentence with the practical ability of using ‘the ‘ex hypothesi’ available effective means for deciding whether the sentence is true or false, and to accept or reject it according to the results’, despite Dummett’s conviction that understanding a sentence consists in a theoretical knowledge. Moreover, McDowell (1985, p.356) adds, even if there are no effective means for attributing truth or falsity to a sentence, it is still possible to accept or reject the sentence in a way that is rational in the light of how one understands it. The complete impossibility to judge the truth or falsity of a sentence would undermine the fact that the sentence has a sense at all. Therefore, McDowell confutes Dummett’s idea that undecidability of some sentences undermines the entire theory of meaning, because he thinks that understanding a sentence does not consist in a theoretical knowledge, but in the practical ability to judge a sentence, which is implied by its having a sense.
Even while trying to steer between psychologism and behaviourism the problem of circularity is the first difficulty that McDowell encounters. Still using the example of the concept ‘square’ he tries to avoid circularity by saying that ‘it is not the words of Dummett’s suggested explanation by which one is guided in using the word ‘square’, but what they express’ (McDowell, 1987, p. 95). However, he then realizes that saying this, ‘in connection with the propositions of the theory of meaning’, would lead him into the idea that we can ‘strip off the linguistic clothing and penetrate to the pure naked thought beneath’, in accordance to the conception of language as a mere code (McDowell, 1987, p. 95). This would support the psychologistic view, therefore undermining McDowell’s initial intent of avoiding psychologism. However, a way to support Dummett’s theory, according to McDowell, would be to say that, since ‘square’ is used only in ‘the first intention-that is, never inside a content-specifying ‘’that’’ –clause’, the concept ‘square’ is only employed but not ‘displayed in its role as a determinant of content’ (McDowell, 1987, p.91). In this sense a full-blooded theory would describe a practical capacity to acquire the concept, employing that concept ‘as from outside its role of determinant of content’ (McDowell, 1987, p.91). Therefore, that would lead to the conclusion that a full-blooded theory of meaning is constructed ‘’as from outside’’ content and concepts. However, the idea that a truth-conditional ‘core’ would need to be formulated ‘as from outside’ content is confuted by McDowell’s (1987, p.92) conviction that the truth conditions are derived precisely from the fact that they specify the content for actual or potential assertions, therefore they cannot be given ‘as from outside content’. Therefore, even trying to justify Dummett’s belief that it is possible to give an account of meaning ‘as from outside’ would ultimately lead us to the conclusion that it is not true, unless begging the question, in this case of what truth conditions are. In conclusion, according to McDowell, despite of Dummett’s belief, it is not possible at all to explain content ‘as from outside’, in accordance with Brentano, because, as we have demonstrated above, it would lead either to recursion or to question-begging.
Moreover McDowell thinks that it is possible to steer a middle course between psychologism and behaviourism, respecting a modest account of meaning. The reason why Dummett rejects a modest account is because he wants to give an account of meaning as from outside content, since he conceives understanding of language as ‘lying open to view’ (McDowell’s (1987, p.92)). On the other hand, accepting modesty would seem to fall into psychologism because it would involve the conception of language as a code for thought, but McDowell demonstrates that this is not true. The only reason why modesty would imply the psychologistic conception of language as a code ‘depends on the assumption that content must be capturable ‘’as from outside’’(McDowell’s (1987, p.92)). However, rejecting this assumption, modesty does not imply the psychologistic conception of language anymore. Therefore, it is possible to have a modest account of meaning and at the same time give an account of the mind’s role in speech, avoiding falling into psychologism.
In conclusion, having given an account of the mind’s role in the speech act and having steered a middle course between psychologism and behaviourism, I think that, despite of Heck’s attempts to defend Dummett’s account from McDowell’s attacks, McDowell’s argument is more convincing, especially considering his stance on circularity. I believe that the aim to characterize meaning ‘as from outside’ in a non-recursive and non-circular way is impossible. On the other hand, it is possible to give an account of meaning using some primitive concepts, and that is why a modest account is preferable to an immodest one, as McDowell sustained against Dummett.