«Theories of assertion fall into four rough categories:2 (i) Assertions are those sayings that are governed by certain norms – the norms of ...»
Forthcoming in Assertion (Eds. Brown and Cappelen), Oxford 2010
Theories of assertion fall into four rough categories:2
(i) Assertions are those sayings that are governed by certain norms – the
norms of assertion.
(ii) Assertions are those sayings that have certain effects.
(iii) Assertions are those sayings that have certain causes.
(iv) Assertions are those sayings that are accompanied by certain commitments.
The view defended in this paper - I call it the No-Assertion view - rejects the
assumption that it is theoretically useful to single out a subset of sayings as assertions:
(v) Sayings are governed by variable norms, come with variable commitments and have variable causes and effects. What philosophers have tried to capture by the term 'assertion' is largely a philosophers' invention. It fails to pick out an act-type that we engage in and it is not a category we need in order to explain any significant component of our linguistic practice.
Timothy Williamson (2000) defends a theory of type (i). He says that a theory of assertion has as its goal "[…] that of articulating for the first time the rules of a traditional game that we play" (p. 240). Among those who think we play the game of assertion, there's disagreement about what the rules are. Some think it's a single rule and disagree about what that rule is. Others think the rules change across contexts.
According to the No-Assertion view we don’t play the assertion game. The game might exist as an
object, but it is not a game you need to learn and play to become a speaker of a natural language.
Central to the No-Assertion view is the notion of a saying. I have more to say about sayings below, but for now, just think of it as the expressing of a proposition. It is a close relative of what Austin called the locutionary act. It is important to note that,
according to Austin, all illocutionary acts (e.g. assertions) are also locutionary acts:
whenever you make an assertion or ask a question, you are also performing a locutionary act, i.e. you say something. The various illocutionary speech acts are, so to speak, built on top of locutionary acts, or sayings. In other words, the need for the Thanks to Jessica Brown and John Hawthorne for extensive comments on earlier drafts. Thanks also to audiences at the Arché Philosophical Research Centre in St Andrews, Institut Nicod in Paris, LOGOS research centre in Barcelona, DIP Colloquium at ILLC in Amsterdam, and CSMN at the University of Oslo.
(i)-(iv) can be combined in various ways, see Section 3 below for elaboration. See also MacFarlane (this volume) for a related overview of theories of assertion.
notion of a speech act-neutral saying is partof the common ground between the NoAssertion view and the various theories that take assertion to be a theoretically important category. It is also common ground that if there are assertions, they are distinct from sayings – i.e. that the illocutionary act is distinct from the locutionary act. The suggestion in this paper is that we don't need the distinct category of assertion in addition to the act of saying. The act of saying (i.e. the act of expressing a proposition) combined with contextually variable norms, causes, effects and commitments can do all the explanatory work.
I argue for the No-Assertion view, primarily, by arguing against normative views of assertion, such as Williamson's. I take the normative views to be the primary alternative to the No-Assertion view. A full-fledged defence of No-Assertion would need to argue also against theories of type (ii), (iii) and (iv). Limitations of space prevent such a full-fledged defence of the No-Assertion view. However, the arguments in this paper provide, I hope, a fairly straightforward model for how to argue against views of type (ii), (iii), and (iv).
In Section 1, I say more about the No-Assertion view, in particular about what sayings are. In Sections 2 and 3, I give an overview of the different kinds of theories of assertion that one could develop and I present the key components of Williamson’s view in more detail. In Section 4, I give arguments against assertion. In Section 5, I respond to two objections.
1. The No Assertion View
In this section, I elaborate on three key components of the No-Assertion view:
(i) There are sayings
What are sayings? A good place to start is with Austin's notion of a locutionary act:
[…] the utterance of certain noises, the utterances of certain words in a certain construction and the utterance of them with a certain 'meaning' in the favourite philosophical sense of that word, i.e. with a certain sense and with a certain reference. The act of 'saying something' in this full normal sense I call, i.e.
dub, the performance of a locutionary act … a great many further refinements would be possible and necessary if we were to discuss it for its own sake.
(1975, pp. 94-5) With this as a starting point, think of an Austinian saying of p as very close, if not identical, to the act of expressing the proposition that p. As an example, take the
S can be used to express the proposition that there are blind mole-rats in Sweden (call this proposition s) because that proposition is its meaning. A speaker of English can use S to express s, and that’s what it is to use S to say that s. Of course, you don’t need to use a particular sentence, S, to say (or express) that s, it can be done in languages other than English (and, even, using other sentences of English).
For those not sure what an Austinian saying is, I suggest trying it. One way to do that is to say, in some language or other, that there are blind mole-rats in Sweden.
If you do that, then you have performed a saying, and in so doing you have expressed the proposition that s. Note that one thing you can subsequently do, having expressed s, is to refer to it with a demonstrative (as in “That’s the proposition HC asked me to express") or anaphorically (as in, “It’s an interesting thought, I have no idea whether it’s true or false, I have no evidence either way, and I didn’t mean to convey to anyone that I think it is true, I just wanted to express it"). Contrast this with simply uttering a sentence you don't know the meaning of. For those who don't speak Norwegian, try uttering, "De er mange svensker som jobber i Oslo." This sentence can be used by those of us who speak Norwegian to say that there are many Swedes working in Oslo. Those who don't speak Norwegian, can make the sounds, and so utter the sentence, but they cannot use it to say that there are many Swedes working in Oslo.
As Austin points out, "a great many further refinements would be possible and necessary if we were to discuss" the act of saying something for its own sake. (1975, p. 96) It’s not the goal of this paper to undertake that project. It is also a project that’s tangential to the arguments in this paper because the No-Assertion view and the ProAssertion views that are targeted in this paper all appeal to Austinian sayings. It is a shared notion. According to all the views I target below, the act of saying that p is part of the act of asserting that p. (I give some arguments for the saying-as-sharednotion-assumption in Section 3 below.) So one argumentative strategy that won’t work against the No-Assertion view is to say: hey, wait a minute, there are all these very tricky questions about sayings – you owe us a story about all of this.3 Well, that might be—any deep philosophical notion is surrounded by tricky questions (and the notion of expressing a proposition is about as deep as it gets)—but those questions are questions the No- and Pro-Assertion views have a joint interest in answering. The answer to those questions won't adjudicate between those views.
For now, a couple of obvious restrictions that both the No- and Pro-Assertion proponents will want to insist on: The sayings I’ll focus on involve the utterance of declarative sentences (as opposed to questions and imperatives (when these characterisations are understood syntactically4)). I also restrict sayings to complete I’m not ruling out that someone could deny that assertions supervene on sayings.
One motivation for that would be opposition to the very idea of expressing a proposition (in the thin Austinian sense). For example, philosophers such as Robert Brandom (1994) who are sceptical of the idea of sentences expressing propositions, will obviously not be attracted to No-Assertion. Note that they are also unlikely to be attracted to what I below call N-theories of assertion. It goes beyond the scope of this paper to prove that there are propositions and that we express them – I take that as a starting point here.
Linguists typically treat these as syntactic categories, well established across languages, and there is no assumption that the syntactic distinction suffices to individuate the relevant speech acts.
propositions – i.e. you don’t count as having said that p if you utter ‘If p, then q’ or “John said that p” (in so doing you have said that if p then q and that John said that p, you have not, in the intended sense, said that p).
(ii) Sayings are evaluated by contextually variable norms, none of them constitutive of the speech act5 Sayings are related to norms in much the same way as kissing and driving are. The norms that governing driving and kissing vary widely across contexts and cultures, over time, and across possible worlds. There’s no one set of norms that is essential to either activity. Kissing and driving are paradigms of non-normative activities governed by variable norms. Sayings, according to the No-Assertion view, are like that. What to say, how to say it, when to say it, whom to say it to, and the combined appropriateness of all this depends on a complex interaction of various norms, goals, and contextually variable factors. These can be weighed in a variety of ways—there need be no one correct judgement about whether a particular saying is correct, praiseworthy, or gives rise to resentment.
The kinds of norms that govern sayings, according to the No-Assertion view, are familiar – they are the kinds of norms that Grice appeals to and that various ProAssertion theorists appeal to. The No-Assertion view can appeal to any norm (and constraint) appealed to by the various Pro-Assertion views, the only difference being that the No-Assertion view doesn’t take any of these to be constitutive of a speech act type (or game).
Grice's maxims of conversation are not constitutive of the acts they govern. Grice takes them to be derived from general principles of rational cooperation. They all have analogues in "spheres of transaction that are not talk exchanges." They are norms that guide behavior, not norms that are essential to (or constitutive of) the behavior they guide. Not only are Grice’s maxims naturally taken to govern (or operate on) Austinian sayings, but, and this point is often overlooked, it is more or less impossible to think of them as governing (or operating on) normatively individuated assertions (where the norms are constitutive of the act). Consider the maxim of quality (1975, p. 41 – my emphasis): ‘Try to make your contribution one
that is true." Grice elaborates with two more specific maxims:
• Do not say what you believe to be false.
• Do not say that for which you lack adequate evidence.
How should we construe the terms ‘contribution’ and ‘say’ in these formulations?6 If we construe them as referring to Austinian sayings, things go smoothly. On that Two papers that provide excellent arguments for this component of No-Assertion, though not used in support of the claim that there’s no such thing as assertion, is Pagin (this volume), and Levin (2008).
Of course, Grice has a theory of sayings, connected to his intention based theory of content. One can be attracted to the maxims of conversation as a good description of conversational exchanges without endorsing Grice’s metaphysics of meaning. So I don’t here intend to get into Gricean exegesis – i.e. to capture his intention base notion of a saying.
construal, the maxim tells us that in order to be cooperative, you should aim to express (i.e. say) propositions that are true; and the ‘elaborations’ tell us that we should say what we believe and have adequate evidence for.
If, on the other hand, ‘say’ and ‘contribute’ are interpreted to mean ‘assert’ in the Williamsonian sense (where it is constitutive of the act that it is governed by the norm that one should assert p only if one knows p, (see Section 3 below for elaboration on this kind of view)), the maxims become necessary truths (maybe even tautologies).
On this construal, it is impossible to perform a saying not governed by the maxim of quality (because, on this construal, it is essential to the act of saying that it be governed by that norm.)7 Grice clearly didn’t intend for the maxims to be necessary truths.
The four maxims of conversations are not, according to Grice, the only ones governing talk exchanges. He emphasises that there are, “[…] all sorts of other maxims (aesthetic, social, or moral in character), such as ‘Be polite’, that are also normally observed by participants in talk exchanges […].” These aesthetic and social maxims are paradigms of norms that vary across contexts, so Grice clearly thought that the norms that govern talk exchanges are contextually variable.
Two further points about the norms that govern sayings: