A. Minh Nguyen

Eastern Kentucky University

Minh.Nguyen@eku.edu

Abstract. What explains first-person authority? What explains the presumption that an utterance is true when it is a sincere determinate first-person singular simple present-tense ascription of intentional state? Jacobsen offers a novel suggestion, one that combines elements from minimalism and expressivism. Minimalism holds that the disquotational schema tells us all there is about truth. Expressivism holds that self-ascriptions typically express rather than describe aspects of one’s own psychology. The novelty consists in reconciling truth-assessability with expressiveness and assigning both to self-ascriptions. According to Jacobsen, self-ascriptions each enjoy a presumption of truth because they are systematically reliable. They are systematically reliable because they are typically both expressive and truth-apt. Such self-ascriptions if sincere are certain to be true. My paper provides two objections against Jacobsen. First, the alleged typicality of expressive self-ascriptions is at best contingent. Thus Jacobsen fails to explain how a speaker’s ability to provide reliable self-ascriptions is unalienable. Second, Jacobsen’s model fails to accommodate the deeply entrenched intuition that, all else being equal, the more authority one has regarding one’s own mental states, the more knowledge one enjoys regarding their existence and character. The connection between first-person authority and first-person knowledge thus finds no place in the expressivist’s scheme.

It is widely assumed that every person enjoys special authority with respect to his own present intentional states. According to this assumption, it is necessary that, for any person, if he sincerely ascribes the presence, or absence, of a particular intentional state to his present self, then there is a legitimate presumption that what he says is true, whereas it is not necessary that, for any person, if he sincerely ascribes the presence, or absence, of a particular intentional state to another person, or to his own nonpresent self, or if he sincerely makes a claim about a certain physical aspect of the external world, then there is a legitimate presumption that what he says is true. What does it mean to say that, for any sincere determinate first-person singular simple present-tense ascription of intentional state that any competent speaker makes, there is a presumption that it is true? What it means is this. Given that a speaker with an adequate command of appropriate concepts sincerely says that he is, or that he is not, in a certain intentional state, his utterance is presumed to be true. One shall treat his utterance as if it were true unless or until one has sufficient evidence, or other epistemic grounds, to the contrary.

First-person authority raises a philosophical issue. For in general the mere fact that a property is capable of being instantiated by a subject does not confer a presumption of truth on his sincere claim that he instantiates it. That is so because in general the mere fact in question does not entitle one to assume in advance that the subject’s claim is true. It is not necessary that, if a person sincerely claims that his weight, holiday address, or retirement account number is thus and so, then there is a legitimate presumption in favor of the claim. So why is the situation any different when the properties are mental?

But the main reason why there is a problem about first-person authority is this. Unless the authority is adequately explained, it will invite skepticism about other minds. For if there is no accounting for the fact that sincere present-tense intentional-state self-ascriptions carry authority, whereas other-ascriptions do not, then it is a serious possibility that the intentional states that one ascribes to one’s present self and those that one ascribes to others are not the same sort of state. Likewise, if there is no accounting for the fact that intentional predicates such as ‘believe that p’ and ‘desire that q’ are necessarily presumed to be true of a subject when self-applied in the present indicative, whereas not so when applied to him by another, then it is a serious possibility that these terms have neither the same meaning nor the same referent in the first-person and third-person contexts.

What then explains first-person authority? What explains the presumption that an utterance is true when it is a sincere determinate first-person singular simple present-tense ascription of intentional state? Rockney Jacobsen has an answer. Jacobsen claims to have found in Wittgenstein two apparently opposing strands of thought that not only can be reconciled, but the reconciliation of which also puts one in a position to account for the systematic reliability and presumptive correctness of self-ascriptions of intentional states. The two strands of thought run as follows.

Expressivism about Self-Ascription. Expressivism is the view that self-ascriptions are typically expressions rather than descriptions of one’s own current intentional states. Self-ascriptions typically express rather than describe the intentional states denoted by the main psychological terms. In the typical case, the purpose of a self-ascription is not to convey information but to express the speaker’s attitude. In speech-act jargon, self-ascriptions are typically not assertives but expressives. There are of course exceptions. When we ascribe an intentional state to ourselves on the basis of self-observation or reflection, our self-ascription counts as a description. The same holds when we self-ascribe by relying on the testimony of someone who interprets our behavior or when we self-ascribe without specifying the content of the ascribed intentional state (e.g., "I want what you want," "I believe the same as Derrida," "I love what I hate").

Minimalism about Truth. Minimalism is the view that the Disquotational Schema á ‘p’ is true if and only if pñ tells us all there is about truth. DS embodies the concept of truth. As such our understanding of the concept is exhausted by our understanding of DS or its representative instances. An apparent corollary of minimalism is that a meaningful sentence is truth-assessable or truth-apt—i.e., apt to be assessed in terms of truth and falsity—as long as it is a suitable substitute for ‘p’ in DS. As long as it can serve as the antecedent of a conditional and admits of a negation, it is in the business of being true or false. All that is needed to have a truth-value is to exhibit certain surface syntactic features. Minimalism thus makes two related claims. First, there is nothing more to truth than what is stated or implied by DS. Second, a meaningful sentence is truth-assessable if it can embed in more complex logical constructions such as negation and the conditional.

Minimalism seems inconsistent with expressivism, however. The reason is this. (1) Self-ascriptions are suitable substitutes for ‘p’ in DS. (2) If anything is a suitable substitute for ‘p’ in DS, then it is truth-assessable. (3) So self-ascriptions are truth-assessable. (4) If anything is truth-assessable, then it has assertoric status. (5) So self-ascriptions have assertoric status. (6) If self-ascriptions have assertoric status, then they are assertions by means of which one describes the intentional states denoted by the main psychological terms. (7) So self-ascriptions are assertions by means of which one describes the intentional states denoted by the main psychological terms. (8) Assertions about the existence and character of one’s own intentional states are neither identical with nor sufficient for expressions of these very same states. (9) So self-ascriptions are neither identical with nor sufficient for expressions of the intentional states denoted by the main psychological terms.

According to Jacobsen, the "minimalist" argument against expressivism rests upon a misunderstanding. There are no good grounds for attributing (4) to minimalists. Minimalism requires only that a suitable substitute for ‘p’ in DS be truth-assessable. It does not require that truth-assessability suffice for assertoric status. Indeed, an utterance of a truth-assessable sentence meaning that p may not be an act of asserting that p. It may be stripped of any assertoric status it might otherwise have had, because the context clearly indicates that, by uttering the sentence, the speaker does not seriously represent himself as uttering a sentence he believes true. Such etiolating contexts include situations where a truth-assessable sentence is used as part of a negation, disjunction, conditional, or biconditional. Where it is employed in play acting, teaching a language, giving an example, or practicing oratory. Where it is exploited for the performance of a performative, a speech act that is not an assertive. For instance: a directive ("You have to play fair"), a commissive ("I will try my best"), a declarative ("I shall call him ‘Minime’"), or an expressive ("I am so excited"). In each of these contexts, truth-assessability and assertoric force come apart. In none of them is there any affirmation of the relevant truth-evaluable content.

The upshot of the preceding considerations is that the "minimalist" argument against expressivism fails because of its questionable attribution of a questionable claim to minimalists. As things stand, there is no good argument for the incompatibility between minimalism and expressivism. Minimalist considerations do not tell against expressivism. Self-ascriptions, though truth-assessable, are typically nonassertoric, because they are already fully employed in the business of expressing the intentional states they only appear to describe. So employed, they are dispossessed of any assertoric property they might otherwise have possessed.

Armed with minimalism and expressivism, Jacobsen offers the following explanation of first-person authority. (1) Typically a sincere utterance of ‘I believe that p’ expresses the speaker’s belief that p. (2) Something expresses the speaker’s belief that p only if the speaker believes that p. (3) So typically a sincere speaker of ‘I believe that p’ believes that p. (4) An utterance of ‘I believe that p’ is true if and only if the speaker believes that p. (5) So typically a sincere utterance of ‘I believe that p’ is true. (6) If typically a sincere utterance of ‘I believe that p’ is true, then there is a legitimate presumption that it is true. (7) So there is a legitimate presumption that a sincere utterance of ‘I believe that p’ is true. (8) This line of reasoning generalizes to self-ascriptions of other types of intentional state. (9) So first-person authority is fully explained.

The suggestion then is that first-person authority is wholly an artifact of a grammatical misunderstanding, a systematic misconstrual of self-ascriptions as self-assertions. What do we make of this suggestion? I think the suggestion has a couple of drawbacks. First, that self-ascriptions are typically expressions rather than reports is at best a contingent (and empirical) matter. The expressivist thus fails to explain how a speaker’s effortless and reliable ability to provide true self-ascriptions is a noncancelable feature.

There is no denying that some human speakers occasionally use some psychological verbs, including those of the intentional variety, in the first-person singular simple present-tense indicative active just to express the mental states denoted by the verbs. As James Urmson points out in an article aptly entitled "Parenthetical Verbs," some people sometimes employ epistemic constructions such as "I know (believe, guess, suppose, suspect, estimate, etc.) that …," not to claim that they know (believe, guess, suppose, suspect, estimate, etc.) that things are thus and so, but to state in essence that the nested statement is true, as well as "to indicate the evidential situation in which the statement is made ..., and hence to signal what degree of reliability is claimed for, and should be accorded to, the statement to which they are conjoined."

For instance, if you ask me where the cat is, then my guarded albeit sincere response "I believe the cat is on the mat" is in this case not a way of describing a certain doxastic state, but rather simply a way of claiming—i.e., expressing my belief—that the cat is on the mat. I claim this while indicating or suggesting that, even though it is not true that I am just plumping as I would apparently be if I truthfully answered, "I guess the cat is on the mat," neither do I necessarily have all the evidence I may need to guarantee the truth of the embedded statement so that the statement should be accepted if at all with the proper amount of discretion.

But the fact that some or all mature human speakers occasionally use some or all intentional predicates expressively does not mean that they all normally do. And even if they all normally do, that does not mean that it is a conceptual or necessary truth. Consider some possible world in which full-fledged self-attributing agents are exactly like mature human speakers except that they are simply incapable of uttering ‘I believe (desire, hope, fear, etc.) that p’ with the intention of expressing, as opposed to describing, their own current belief (desire, hope, fear, etc.) that p. For ease of expression, I shall call this world and its inhabitants ‘D-world’ and ‘D-worlders,’ respectively.

D-worlders never express their propositional attitudes by saying that they have them. They do not even get to possess, let alone exercise, this mode of competence. Their utterance of ‘I believe (desire, hope, fear, etc.) that p’ is invariably assertoric. It is invariably achieved in a tone of voice that suggests a deliberate matter-of-factness, accompanied by gestures that signify a solid commitment to claim-making, and so forth. Even though D-worlders inherit an incapacity for expressing their intentional states by means of self-ascriptions, they are nonetheless capable of expressing these states by alternative means such as intonation, body posture, hand gestures, gaze patterns, facial expressions, and less explicit linguistic signals. We can make D-world even more credible by stipulating that its inhabitants "have what they regard as important ideological reasons for behaving as they do, and they have, through years of training, learned to live up to their own exacting standards ... and after millions of years they begin to have children who are born fully acculturated."

There is no contradiction involved in the scenario just described. D-worlders’ lifelong incapacity for expressing their intentional states by means of avowals has a compelling genetic and ideological explanation. Nor is there any change of meaning in any intentional predicate. D-worlders believe and desire just as we do. Their attitudes are dispositions that manifest themselves in numerous ways and over a span of time. Just as ours are. The only difference is that their capacity for linguistic self-expression is on average a little bit less versatile than ours. Despite the difference, a normal D-worlder is still better at expressing himself than an otherwise normal human with a damaged language organ. But in general we do not treat such a human as incapable of thought. So the difference does not make a difference as far as meaning is concerned.

The above scenario also does not entail any untestable hypothesis. Skepticism about other minds aside, if there were D-worlders, we would in principle be able to distinguish them from ourselves and from creatures that do not self-ascribe psychological states. For we can in principle determine whether a string of marks or sequence of sounds is a self-ascription and whether a self-ascription plays a descriptive or expressive role. Expressive self-ascriptions are expressions of the intentional states denoted by the main psychological predicates. Descriptive self-ascriptions are assertions, and so expressions of higher-order beliefs, to the effect that the self-ascriber occupies the intentional states in question. Because intentional states are dispositional states, the truth of an ascription of any of them—whether higher or lower-order, doxastic or otherwise—is answerable to behavioral manifestations that are publicly observable. But if the described scenario does not involve any conceptual incoherence, logical absurdity, meaning change, or untestable hypothesis, then D-world is possible. Since the expressivist model fails to accommodate this possibility, it fails to furnish a complete account of the authority of first-person utterances.

The expressivist model also fails to accommodate the deeply entrenched intuition that, all else being equal, the more authority one has regarding one’s own mental states, the more knowledge one enjoys regarding their existence and character. All else being equal, a person with more first-person authority has more first-person knowledge, because authority of this sort has to do with reliability and reliability has to do with justification and knowledge. But instead of accommodating the intuition, the expressivist proposal propels a disconnect between first-person authority and first-person knowledge.

Consider my counterpart in a slightly different possible world. He makes all the self-ascriptions that I make plus some expressive self-ascriptions not in my repertoire. Assuming that our shared descriptive self-ascriptions are fallible and possessed of the same degree of failure or success, my counterpart’s self-ascriptions are on average more reliable and so more authoritative than mine, because all the additional self-ascriptions that he makes are expressive and so guaranteed to be true if sincere, and this drives up the overall success rate of his self-ascriptions. But it does not mean that he has more self-knowledge. After all, expressive self-ascriptions are not factual assertions. Nor are they proper subjects of epistemic appraisal. They express no beliefs about mental states. Making them does not enlarge our self-conception, let alone self-knowledge.

We can reach the same conclusion in a somewhat different way. Suppose that, as far as truth-assessable contents are concerned, my counterpart and I make exactly the same self-ascriptions. The only difference is that, with respect to a number of self-ascriptions, his are expressive whereas mine are descriptive. As a self-ascriber, he issues more self-expressions but fewer self-descriptions than I. The total number of self-ascriptions that we each make, however, is the same. Suppose further that, whether shared or not, our descriptive self-ascriptions are equally fallible. They have the same nonmaximal rate of success. Given all this, it is clear that my counterpart’s self-ascriptions are on average more reliable and so more authoritative than mine. For, in relation to the differentiating self-ascriptions, his are expressive and so guaranteed to be true if sincere, whereas mine are descriptive with no such guarantee. But again this does not entail that he has more self-knowledge. On the contrary, it is I who enjoy the advantage. After all, I make more descriptive self-ascriptions. Such ascriptions are factual assertions. They express beliefs about my own mental states. Making them enlarges my self-conception. Under suitable conditions, it enhances my self-knowledge.

This concludes my second objection to the expressivist proposal. It is worth noting that Jacobsen has never given, or attempted to give, an argument for expressivism (or minimalism for that matter). For him expressivism is more of a pretheoretical platitude than a philosophical theory. As such it is presumed to be true. One is entitled to assume it true unless one has sufficient evidence to the contrary. Countervailing reason is thus required to override this entitlement. But so far the only significant argument against expressivism has been the one to the effect that expressive character clashes with truth-aptness. If this is a bad argument (and Jacobsen stands on good ground here), we can safely return to our pretheoretical platitude. The suggestion, however, does not fly. For the skeptic will reply that, in the absence of a cogent argument for expressivism and the presence of relevant alternatives, he is surely justified in asking how we know that expressivism holds. In any event, a return to our pretheoretical platitude seems impossible. Or at least premature. For if my twin objections stand up, the place the expressivist calls home is just not there.