signs and linguistic behaviour
ABSTRACT. Halfway between linguistics and speech act theory, Edmund Husserl’s philosophy of language shares many concerns with analytic philosophers, in particular with Gottlob Frege. One concern he shares with Frege is the way in which we recognize linguistic signs as signs and how we recognize them as the particular signs they are.
The views of Frege and Husserl on the conditions that make linguistic communication possible both rely heavily on an account of the way in which occurrences of a linguistic sign are identified as occurrences of the same “sign-type.” Both attempt to describe the role of mental acts, such as intentions, in recognizing particular signs as occurrences of the same sign type, but their accounts differ in basic ways. This paper compares their respective theories.
In addition to theories of meaning and denotation, Frege also expressed views on linguistic signs and how they function in linguistic communication. According to Frege (1903, §99), signs would be useless if they could not be understood to denote the same thing at different times and in different contexts. In order to fulfil this purpose, Frege claims, the different occurrences of the same sign type must have sufficiently similar figures. Frege also believes that we cannot recognize two instances of a sign as being of the same type solely on the basis of their physical characteristics. Frege does not himself provide any concrete examples – and he considers only written signs – but he argues that considering the imperfect nature of human perception and the fact that two tokens of the same sign are seldom, if ever, exactly the same physically, we cannot rely on two signs being physically identical to decide whether they are instances of the same type – indeed, they may have quite different physical properties. Frege also thinks that abstraction cannot be used to recognize signs as instances of the same sign type, arguing that “different things cannot be made to coincide by abstraction.” (1903, §99)
But then how do we recognize signs as being of the same figure? According to Frege, in order for signs to denote the same thing at different times and in different contexts, they must be sufficiently similar. How similar? Frege answers:
… nothing else is required but that there be present the intention of producing a sign which is similar to the one that had been made previously and this need only to succeed inasmuch as the reader correctly recognises this intention. In what follows, we understand by “signs of same figure” those which, following the intention of he who writes, are supposed to be the same so as to designate the same thing. (1903 §99)1As for how we recognise that two signs have the same figure, Frege answers: because we recognise the authors’ intentions of producing similar signs in order to denote the same thing in each case, where these intentions we recognise are a particular type of mental act.
Frege thus appeals to our ability to recognize the mental acts of others as the basis of language use. But is this consistent with his notorious anti-psychologism? Given the significance of the topic in the literature on Frege, this may be an important question. It is not however my purpose to address it here. Rather, in what follows, I’ll consider whether Frege’s view that recognizing the intention of a speaker can do the job of determining the type to which a given instance of a sign belongs. To a certain extent, this depends on what Frege’s “intentions” are and Frege says little about them.
There are, I think, two plausible ways to understand Frege’s appeal to our capacity to recognise A’s intention of producing figures in order to designate things; a weak one and a strong one. On the weak interpretation, what a reader recognises is A’s intention of producing and employing a sign, that is, A’s (unspecified) intention to communicate. Pierre can recognise that Marie uses a sign only if he recognises that Marie’s intention is to communicate something. “Figures which we write or print or which, in general, are produced on the surface of a physical object”2 need not be used for the purpose of linguistic communication – Marie could be doodling – and in order to recognise them as occurrences of signs we need to recognise that they were produced with the intention to communicate. If this idea is to be taken seriously, it cannot exclude considerations of the way in which, in practical interactions, we can tell what other people are trying to do prior to any understanding of the particular signs.
Let us assume that Frege would agree to say that this idea lies behind his remarks. Would he also claim that in recognising that a graphic sign was produced with the intention of communicating, we recognize the particular type of sign it is an instance of? It seems that the ascription of a communicative intention at most restricts the scope of possible figural types to which the graphic token may belong to linguistic ones – it will neither be a sample of traditional Moroccan decorative art nor a mere doodle since those are not figures used for the purpose of communication – but it does no more than that. Frege needs something stronger.
On the strong interpretation what the observer recognises is the particular sign-type A intended to make, and so the particular thing A meant to denote by the sign. The strong interpretation does not exclude the weak one. In fact, the strong interpretation makes sense only when it is already clear to the agents involved that they are involved in a situation of communication.
Furthermore, I take it that Frege would agree with the following: Marie does not go around in the world with the purpose of denoting objects. Marie believes things about certain objects, and typically she denotes objects only insofar as she expresses or voices her beliefs (or fears, wishes, etc.) about them. That is, some of the words Marie uses denote objects, but only in so far as they are part of more complex expressions, e.g., assertions that express some of her mental states, her beliefs for instance. I take it that Frege would also agree to say that Pierre can only recognise Marie’s intention of denoting y if he recognises that she has a mental state about y’s being so-and-so.
This interpretation enables us to preserve the idea that Frege’s theory of sign recognition rests on certain mental states, what he calls “intentions,” speakers have – what is intentional here is Marie’s use of the sign-token for expressing her belief – without committing him to the view that Marie’s having the mental state about y’s being so-and-so, nor her using a sign to refer to y, are her primary intention when she seeks to engage in linguistic behaviour. On the strong interpretation, once Pierre has recognised that Marie intends to communicate, in order for him to determine the sign-types of the signs she uses, he needs to recognize her intention to express a particular thought.
The strong interpretation rests on the idea that if Marie utters “I’d like to eat chocolate mousse,” typically it is appropriate for Pierre to ascribe to her the desire to eat chocolate mousse. Of course, it may be the case that Pierre ascribes an intention to Marie although there is nothing intentional in Marie’s action: she may be talking in her sleep. Nevertheless, this interpretation, which introduces the idea of using the ascription of a mental state, namely, the intention to express a thought, to recognize the sign-type that a sign is an instance of, enriches the Fregean conception of linguistic communication and tackles a problem that has recently become central in speech act theory and theories of communication,3 and will be discussed at length in the following sections. However, mental state ascription does not solve the problem of determining the type to which a sign belongs but, as we will see in the conclusion, more plausibly presupposes it.
The idea the strong interpretation brings forward is that an adequate theory of language should account not only for the fact that words denote objects and concepts, but also for the fact that they “intimate” the mental states of an agent, as well for the fact that these two functions of language fulfil complementary purposes: denotation relies on intimating mental events and the intimation of mental events implies denotation. To my knowledge, this idea was first brought up in the Brentano school and was shared, with different levels of sophistication, by at least three of Brentano’s students: Marty, Twardowski and Husserl.4 Of the three, Husserl is the one who offered the clearest account of the distinction between these two functions as well as of their connection.5
In the Logical Investigations, Husserl notes that in addition to being used to denote objects and concepts and make statements that are true or false, words are also used to inform us of, or intimate, mental states of speakers or writers. In intimation, Husserl thinks that words function in a way similar to what he calls “indication.” Smoke, for instance, is called a sign or indication of fire, scars may be said to indicate a healing wound, and A’s pout will indicate A’s displeasure. Similarly, on Husserl’s account of intimation, A’s assertion that p intimates A’s belief that p. According to the Logical Investigations, intimation is similar to the relation of indication we find in the three previous examples but also differs from it in substantial ways. On the one hand, unlike a scar or smoke, but like a pout, understanding what an utterance intimates implies our recognizing some of the speaker’s mental states. But when it comes to mental state ascription there is also, according to the Logical Investigations, an important distinction to be made between bodily and linguistic behaviour.
Marie’s nodding accompanied by an avid smiling in the presence of an appetizing hypercaloric dessert will indicate to Pierre that Marie has a certain mental state, namely, the desire to eat this dessert. But the connection between Marie’s bodily behaviour and the exact content of the mental state ascribed to Marie is the result of Pierre’s interpretation, not the result of his recognizing something communicated as an intended meaning. Marie’s bodily behaviour could as well indicate her unrepentant gluttony or her hypoglycaemic condition or yet some other state of affairs. What is indicated by bodily behaviour is thus subjective: it depends on the observer’s background knowledge and assumptions, which will vary from one individual to another. But, according to the first Investigation, whenever Pierre understands Marie when she says “I would like to eat chocolate mousse,” Pierre will typically believe that she would like to eat chocolate mousse. He may or may not ascribe other mental states to her on the basis of his having understood this (or on some other basis), but if he understands what she says, he will typically ascribe at least this desire to her.
So what is the difference between the two cases? Husserl claims that a connection will be made between a person’s linguistic behaviour and her mental episode by any competent observer. The connection will furthermore be “systematic” in the following sense: from the observer’s standpoint, the utterance and the mental act have the same content. This is, as far as the Logical Investigations are concerned, what makes the difference between the way we understand speech and the way we merely interpret bodily behaviour. Thus, linguistic communication is not subject to interpretative variations the way bodily behaviour is. But what leads the observer to believe that the content of the utterance and the content of the mental act are the same?
Although the theory of meaning Husserl puts forward in his Logical Investigations has similarities with Frege’s theory of sense and reference, it also differs from it in certain important respects. The differences are brought to the fore when we compare their views on the role of mental state ascription and sign recognition. According to the Logical Investigations, what explains Pierre’s understanding of Marie’s utterance is a complex network of relations between:
(i) Marie’s mental state p
(ii) Marie’s utterance of the sentence p
(iii) the thought p´ (caused by Marie’s utterance) in Pierre
(iv) the abstract content of (i) and (iii), i.e. that p
Marie’s utterance ‘I would like to eat chocolate mousse’ intimates a certain mental state, a desire she has. This desire to eat chocolate mousse has a “quality” (roughly, its propositional attitude, in this case, a desiring) and a “matter” (what she desires). The matter of Marie’s desiring is the aspect of it that makes it about eating chocolate mousse. As the diagram shows, the desire, and so its matter, is different from the meaning (i.e., “content”) of the expression. It is also not a cause of Pierre’s understanding of Marie’s utterance.
In uttering ‘I would like to eat chocolate mousse,’ Marie causes Pierre to hear the sentence, which causes him to understand the sentence, i.e., to have a corresponding mental state that is however not itself a desire. Husserl, however, remains undecided as to the explanatory role of causal relations in his theory of language perception and understanding. What is explicit however is that what explains Pierre understanding Marie, according to Husserl, is the fact that the mental state Pierre has upon hearing Marie’s utterance (upon perceiving the sounds she produces) and the mental state Marie has have the same content. And the two thoughts have the same content because their respective matters are instances of the same objective meaning.
So in the Logical Investigations, the possibility of communication is explained by there being objective entities, meanings (“contents”) which certain aspects of mental states, the matter, are instances of, and that can be instantiated in different speakers simultaneously or in the same speaker at different times. Note that the semantic properties of the sign (meaning and denotation) are not the same as the mental properties of the thought (having a matter, intending an object). What is original in Husserl is the way he reconciles the two sets of properties, semantic and mental: the mental state, by virtue of its matter, is an instance of the meaning. Husserl’s view clearly commits him to some conception of objective meaning, so his semantics is anti-psychologistic – yet he still holds that mental states are intrinsically involved in language.
But what, exactly, is the role that the intimation of mental states plays in this model? Although Husserl is adamant that the relation between the meaning of an utterance and the matter of the thought the utterance intimates is systematic and that the latter is invariably involved in the former, the nature of this relation is not made clear. Rather, his theory of meaning in the Logical Investigations, that thoughts instantiate meanings, leaves the question of this relation’s nature unanswered.
Husserl was not satisfied with the theory of meaning he put forward in the Logical Investigations, especially with the minimal role claimed for intimation in his explanation of linguistic communication. Later, in his unpublished Sign and Expression (1913-14), he explored a different conception of how we come to recognise others’ utterances as signs and act upon that recognition. In this later theory, his explanation of how we recognize the meaning of what a speaker says becomes a special case of a more general explanation of the interpretation of (psycho-physical) actions. In this explanation, intimation plays a central role. Husserl describes linguistic behaviour as a complex and sophisticated form of voluntary bodily movement that must be perceived as voluntary in order to be understood. This involves an elaboration of the idea that one of the conditions of linguistic understanding is the recognition of communicative intentions: Husserl now says that Pierre’s perceiving Marie’s linguistic behaviour as voluntary has a “motivational” effect on Pierre. Husserl introduces a technical term to designate this motivational effect: Marie’s voluntary action seeks to engage Pierre to perform a corresponding voluntary (psycho-physical) action.
On Husserl’s new view, when Marie asserts “Alonzo is an admirable logician,” she is not simply producing sounds meant to be perceived by Pierre and that may cause him to have a certain thought; rather, she is seeking to engage him to do something. Behaviour and context will typically provide an important part of the information Pierre needs in order to find out what Marie wants from him. This will restrict his range of possible responses. (Questions, commands, etc. will function analogously.) For instance, on the basis of Marie’s tone, Pierre can recognize whether she is asking him a question, or ordering him around or just stating something. (Husserl 1913-14b, 90) Typically, what she minimally seeks to engage Pierre to do when she asserts something to him is that he co-believe whatever she herself believes. So if Marie asserts “Alonzo is an admirable logician” to Pierre, then, excluding instances of irony, sarcasm, theatrical productions, etc., what she seeks to engage Pierre to do is acquiesce and co-believe that Alonzo is an admirable logician. Of course, Pierre is in no way compelled by Marie’s demand. He is free to believe or not to believe that Alonzo is an admirable logician. (He can, for instance, doubt it, deny it, etc.)
Husserl’s point is that to understand what Marie asserts and to act on it, Pierre must at least recognise that she has produced the sounds to arouse in him the disposition to believe as she does. More generally, if a person did not recognize a speaker’s intention to communicate, and in fact, recognize the speaker’s intention to communicate a particular thought, and further recognize that the speaker seeks to engage him to produce a mental act (belief, desire, doubt, etc.) corresponding to one of the speaker’s, he could not ascribe to the speaker the relevant mental states and so would not understand her. Without such intimation there is no understanding.
In Husserl’s manuscripts on Expression and Sign, the recognition of the communicative intention raises an important question: how does Pierre come to recognise which belief (answer, response, etc.) Marie demands from him, e.g., how does he come to recognise that she seeks to engage him to believe that Alonzo is an admirable logician and not that Alonzo is a rather pitiful rhymester? The Husserlian answer would be the following: Pierre recognises what belief Marie seeks to engage him to have because she has made available precisely that which is necessary for his being able to have the belief that Alonzo is an admirable logician, namely, a token of ‘Alonzo is an admirable logician,’ which intimates this thought to Pierre.
A full explanation of this idea would require us to look in detail into Husserl’s colossal ontology of language. The framework, however, is already at hand in the Logical Investigations.6 In the first Investigation, Husserl asserts that one cannot have the belief that, e.g., Alonzo is an admirable logician, outside of language, i.e., one cannot have the belief that Alonzo is an admirable logician independently of a graphic or auditory complex of a certain type being available to one through perception or imagination. As he will put it later, one cannot “undress” the belief from its linguistic clothing and retain the “naked” thought.7 Husserl’s position does not imply that all mental states are language dependent: perceptions, for instance, are not. But in the case of the belief that Alonzo is an admirable logician, the marks or sounds being available is indispensable. It is the recognition of the sign’s type that ultimately enables the observer to determine what a speaker says.
Contrary to what Frege suggests in the Grundgesetze 2, on Husserl’s view, mental state ascription is not the only thing that enables us to determine what a speaker says. Indeed, what precise mental state a person has in a certain situation of communication can often only be determined on the basis of the type of the linguistic signs she uses. Although Husserl does not have a full-fledged answer to the question of how this is achieved, what he has to say interested the linguists of his time. Roman Jakobson, for example, explicitly pursued the Husserlian semiotic project. Combining linguistics and speech act theory, logic and psychology, Husserl’s philosophy of language shares many concerns with contemporary analytic philosophy.
1 My translation. „Wenn nämlich die Zeichen nur den Zweck haben, der Verständigung der Menschen untereinander... zu dienen, braucht beim Schreibenden nur die Absicht vorhanden zu sein, ein dem früher gemachten ähnliches Zeichen herzustellen, und das braucht nur soweit zu gelingen, dass der Lesende die Absicht richtig erkennt. Wir wollen im Folgenden unter « gleichgestalteten Zeichen » solche verstehen, welche nach der Absicht des Schreibenden gleichgestaltet sein sollen um dasselbe zu bezeichnen“.
2 Cf. Frege 1903, §98
3 See, for instance, Kemmerling 2002
4 Cf. Marty 1873, Twardowski 1894
5 My remarks in this section are based on Investigations 1 and 5 in Logical Investigations
6 For Husserl’s ontology of language, see Lapointe 2004
7 Husserl 1913-14a, A I 18, p.5-44
Frege, Gottlob, 1884, Grundlagen der Arithmetik, Stuttgart, Reclam, 1987
––––––, 1903, Grundgesetze der Arithmetik 2, Jena, Verlag Hermann Pohle
Husserl, Edmund, 1913-14a, Manuscript A I 18, 5-44, Die Konstitution des Ausdrucks im Sprechen und Verstehen
––––––, 1913-14b, Manuscript A I 17 II, 90, Der Ton als Anzeige für die Satzart
––––––, 1984, Logische Untersuchungen, Husserliana Band XIX/1-2, Den Haag, Nijhoff
Kemmerling, Andreas, 2002, “Expressing an Intentional State,” in Speech Acts, Mind, and Social Reality, Dordrecht, Kluwer
Marty, Anton, 1873, Ursprung der Sprache, Würzburg, Stuber
Lapointe, Sandra, 2004, “Husserl et Frege sur la sémiotique des langages algorithmiques,” Recherches husserliennes, vol. 22, pp. 111-144
Simons, Peter, 1995, “Meaning and Language,” in Barry Smith and David Woodruff Smith (eds.), Cambridge Companion to Husserl, Cambridge, CUP, pp. 106-137
Twardowski, Kazimierz, 1894, Zur Lehre vom Inhalt und Gegenstand der Vorstellung, Wien
Department of Philosophy
Kansas State University
Manhattan, KS 66506
lapointe@ksu.edu