The Last Word (Chapter Three)

The Last Word (Chapter Three)

Nagel makes two important points in this chapter. First, shows that thought is prior to language so that one cannot reduce human thought to contingent linguistic practice. Second, he shows that no naturalistic explanation of language can succeed insofar as language is a vehicle for thought and reasoning; on this front he focuses on the impossibility of grounding the intentional in the nonintentional, the meaningful in the nonmeaningful, and the normative in the nonnormative.

Section I

Nagel rejects the notion "that the social phenomenon of language is at the bottom of everything": that linguistic practice explains what our thoughts mean and what is true. First, there are plenty of cases, such as in philosophy, "where thought is often nonlinguistic and expression comes later" (p. 38). Second, we can have thoughts that are necessarily true, such as thoughts about logic, which cannot be explained because of our contingent linguistic practices (p. 39). He takes this to show that language is a tool to express thought, as diagrams can express geometric propositions, or notion can be used to express mathematical claims generally, but "is not the material out of which thoughts are made" (p. 38).

This isn't to deny contingency in language: spelling, grammar, and usage are relative to the agreement of the linguistic community. Nor does he deny that consensus among a linguistic community can determine the extension of some concepts (pp. 39, 40). But this all fall shorts of the relativist thesis that would make thought and truth dependent on language. That "and" is the English word for conjunction is a contingent matter which has no bearing on the truth of the claim p and q implies p. What exactly counts as humorous may be decided by a linguistic community, but what add two means is not dependent on such consensus (p. 40). 

Section II

Using Wittgenstein as a framework, Nagel then proceeds to discuss the impossibility of giving a naturalistic account of intentionality or meaning.[1] He says, "Thinking cannot be identified with putting marks on paper, or making noises, or manipulating objects, or even having images in one's mind - however much contextual detail (including community practice) is added to such an account." (p. 41). Such empirically discoverable details "cannot possibly explain what it is for words to have meaning" (p. 42).

Intentionality cannot be explained by the unintentional. Since language (which expresses thought) is intentional, one cannot give an external, non-linguistic analysis of it (i.e., a naturalistic account of it). (p. 42) The thought is more fundamental than any facts about the sounds I emit, the mental pictures I have, and so forth. My thought add two doesn't get its meaning from what I do (say put two rocks together) or a mental image I might form (of me pushing two rocks together) or my saying, "I'm doing addition." They don't suffice to fix the content of my thought and in fact have whatever meaning they have in virtue of me having that thought and acting on it: intentionality is unavoidable, and the intentional content of thought is primary, not something from which we could escape and get an external view point that could get us to call it into question.

Wittgenstein is right, Nagel says, in affirming "that no natural fact about me makes it true that I mean something" (p. 43). Why? He is a finite being, so any, say, behavior or physiological facts about him will be finite in number. However, the mathematical function of addition has "infinite normative implications". Behavioral evidence about me will only extend, say, to my computing integers lower than 10^N, so you can't determine that I am doing addition as opposed to some function indistinguishable from it when the integers are below 10^(N+1); call this function quaddition. (pp. 43, 44). Every sort of natural fact we could gather together would be indeterminate in this respect.

Nagel notes that, because of this, philosophers like Kripke will simply conclude that I don't mean anything (or at least determinately mean addition) when I say "plus" and that this conclusions holds for every word I use. Of course, Nagel finds this ludicrous. "We would be left without the possibility of formulating the argument for the paradoxical conclusion" (p. 44). We will still have to meaningfully use words to formulate the argument and Kripke's conclusion, for instance.

Hence, there must be some fact as to what what we mean by using some (many, most) of our words, even if we can't ground the fact that we mean such-and-such by them in natural facts about ourselves. At the end of the day, we have thoughts which we can express meaningfully and this is not something which we can explain in a nonintentional way (p. 45).

At this point, Nagel introduces the problem with grounding meaningful thoughts in natural facts in a different way: "the gap between the nonnormative and the normative". Meaning, he says, entails the difference between right and wrong answers or applications, but behavioral, dispositional, or experiential facts have no such meanings. (p. 45). (Answering "5" is an incorrect answer to the problem 2+2=?; but being inclined to jump when startled is not a wrong response to being startled; so if answering "5" or "6" to that arithmetical question is just a natural response, some deep disposition we follow blindly, it can no more be wrong than jumping when startled is.)

Nagel also dismisses "the move from the terrain of truth conditions to the terrain of assertability conditions" as misguided (p. 46). By "assertibility conditions" Nagel refers to the idea that, given our linguistic practice, we are warranted to ascribe some concept to someone who seems to apply it according to common practice in finitely many cases. So, we can say that, given you answer "4" to "2+2=?" and "9" to "2+7=?", etc., you possess the concept of addition.

Two problems face this view. First, that you conform to what we all see as addition is not what it means for you to have that concept; that you have that concept (as do we) explains why you behave accordingly. Second, as noted by Nagel, what is true is not coterminous with what is assertable. Hence, you could possess a concept even if your behavior doesn't express that concept in a way observable by us. But then concept possession (and hence having thoughts whose meanings are expressible through language) is not defined by such finitely many warrantability conditions; this fails as much as the attempt to ground meaning in finitely many naturalistic truth conditions (pp. 46, 47).

Nagel closes this second section with some remarks from P. F. Strawson. Part of the quotation from Stawson provided by Nagel reads: "we do not merely experience compulsions, merely find it natural to say, in general what (we can observe that) others say too . . . rather, we understand the meaning of what we say" (p. 47). We see the meaning; thus that meaning, content, or intentionality is irreducible and not dependent on behavior, environment, etc.

Section III

Nagel begins this section by examining what Wittgenstein says about following rules. Wittgenstein seems to say that whether I follow some rule ultimately comes down to the sheer fact that I happen to act that way: "I obey the rule blindly," he says (pp. 47, 48). Nagel glossess this as: "what I am doing when I add, for example, is that I am simply producing responses which are natural to me" and unavoidable in the situation in which I find myself. To say this is what is going on when I follow a rule, such as found in addition, is to take an external view about myself, "to get outside of my arithmetical thoughts" in a way that negates the content of my thoughts (pp. 48, 49).

That external view ends all justification and grounds all meaning in a contingent fact about me (how I'd respond to stimuli), whereas Nagel would say that what makes it the case that I am doing addition is just that I grasp what addition is, so know how to apply it, and act on that knowledge: an internal view (that ends justification in the content of the concept I grasp, addition). (pp. 48, 49). 

This case, I think, well highlights the problems that follow from trying to explain how I can have thoughts in general (mean anything) or mean addition when I say "plus" in particular (pp. 49, 50). You cannot explain the intentional in terms of the nonintentional, nor ground the infinite in the finite. More on this last point presently.

Nagel believes "that our linguistic practices reach 'beyond themselves'": our possession of the concept addition, which as infinitely many implications and possible cases to which it applies, is independent of our everyday application of the word "addition" (to finitely many cases) and undergirds it.[2] Our acting in ways that our linguistic community call "adding" is not what it means for me to mean addition nor does it ground my possession of the concept addition: that I have it is because I can simply see what addition means (an explanation internal to arithmetical thought). The external view of arithmetical language will never show what I mean by the things I say or do; it can't ground the meaning or extension of addition. However, as soon as I look at things from the inside, things become clear (p. 51). 

That we mean addition, with its infinite scope, cannot be established from an external view. It cannot be ground in the fact that, from the perspective of our linguistic community we follow some rule (p. 52). Rather, "the rule-following practices of our linguistic community can be understood only through the substantive content of our thoughts" (p. 53). To generalize: the meaning of words (and rules about when to use them) may be ground in linguistic consensus, but the meanings of our thoughts expressible through language is not.

[1] A lot of what Nagel says here fits in nicely with what Feser says here.

[2] To possess the concept addition is not merely the disposition to give the correct verbal answered to certain questions; possession of the concept addition in those who have it undergirds such answered, but the concept is of far greater scope than its application to particular instances.