From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

• Home
• About
• Login
• Search
• Current
• Archives

Table of contents

• Literature

1. The question whether meaning is normative has been hotly contested ever since Saul Kripke (1982) declared the relation of meaning to action to be normative. Not only is there still no consensus on the answer, an early fierce proponent of semantic normativity, Paul Boghossian, has recently argued that meaning is not normative after all. (See 1989 for the early view and 2005 for the later view.) And Kripke himself has recently been interpreted as firmly rejecting semantic normativity rather than endorsing it. (Kusch 2006, 50) Even though Kripke made it clear that the views he was advancing were inspired in him by Wittgenstein, the question whether Wittgenstein himself was a semantic normativist is rarely addressed by the participants in the debate. This is unfortunate since there is no doubt that it is Wittgenstein’s writings which initially prompted the question. One might expect that a return to the sources would shed light on an issue that has only become muddier as the years have gone by. This is what I propose to do in this paper. I shall start with Wittgenstein’s sceptical paradox, which brings out what may be considered as a trivial, certainly uncontroversial, yet, as we shall see, crucial sense of normativity. Then I shall distinguish between two more robust senses in which meaning may be thought to be normative, both of which have been the subject of dispute. I shall argue that Wittgenstein was a normativist in one of those senses. For this follows from meaning being normative in the trivial sense underscored by Wittgenstein and, more importantly, from how, according to him, normativity in this sense can be obtained.

2. What the sceptical paradox reminds us of, indeed emphasizes, is that an activity cannot be rule-governed if it can be deemed to be in accord with the rule it allegedly follows, no matter what the activity is like. (Wittgenstein 1958, #201) A fortiori, an activity cannot be a linguistic activity if can be deemed to be correct, no matter what it is like; more specifically, a linguistic expression cannot be meaningful if it can be deemed to be applied correctly, no matter how it is applied. Linguistic expressions, in order to be meaningful, must be governed by conditions of correct application. These conditions describe the semantic relations between expressions and features of extra-linguistic reality. They tell us what in the world expressions are true of, or warranted by, or what they refer to, stand for or denote. Thus, if ‘green’ means green, then ‘green’ is applied correctly to all and only green things. If ‘Kirchberg’ means Kirchberg, then ‘Kirchberg’ is applied correctly to Kirchberg and only to Kirchberg. No one doubts the platitude that meaningful expressions must have conditions of correct application. And, if this platitude deserves the label of normativity, then no one doubts that meaning is normative in this sense. Some philosophers think that there is, indeed must be, more to the paradox than a reminder of the platitude. Otherwise, they ask, why would Wittgenstein bother making such a trivial reminder? But he made it because, trivial as this condition on meaningfulness may seem, it is a condition that many traditional theories of meaning could not meet. And his reminder prompts a reexamination of the question, how can linguistic expressions be governed by conditions of correct application, which leads to the first more robust sense in which meaning may be thought to be normative.

3. According to one version of this sense, to say that meaning is normative is to say that the conditions of correctness governing the application of expressions stem from norms or rules that exist independently of any language users and which language users must by and large follow if they are to use expressions meaningfully. These norms may be provided either by abstract entities to be found in some Platonic realm, or by the natural world of essences surrounding us. But appealing to entities of these kinds is precisely what Wittgenstein argues leads to the paradox. For, to put it in a nutshell, it is only once these entities, be they abstract or natural, are regarded -- interpreted, as Wittgenstein would say -- in certain ways that they can provide some norms rather than others. But no particular interpretation is ever forced on us. So these entities can always be interpreted in such ways that the applications they allegedly govern are correct, or incorrect. The same observation is true of other entities Wittgenstein examines, such as mental pictures. Nothing, no thing, considered in itself, can provide a norm for the correct applications of a linguistic expression. Thus Wittgenstein is definitely not a normativist in this sense. (See Wittgenstein 1958, ##28-30 and 139-55) The question remains, though, what then provides linguistic expressions with conditions of correct application?

Kripke’s answer is worth mentioning here, for many philosophers have given a similar kind of answer on Wittgenstein’s behalf. According to them, what govern the application of expressions are communal norms, uses or conventions. Specifically, for Kripke, which of an individual’s applications of her expressions are correct, and thus what her expressions mean, is determined by comparing her applications to those of her linguistic community. If an individual’s applications of an expression consistently agree with those of her community fellows, then what she means by the expression is the same as what they mean by it. I do not think, however, that Wittgenstein was a normativist in the sense these remarks suggest either. For, on the one hand, appealing to communal meanings to account for the meanings of an individual’s expressions, and leaving it at that, is tantamount to evading the question what provides expressions with conditions of correct application, thereby evading the question whether meaning is normative in the sense that the conditions of correctness governing the application of expressions are themselves determined by norms. On the other hand, appealing to communal uses (as opposed to meanings) is subject to the same problem as appealing to the entities rejected by Wittgenstein. These uses too must be regarded, interpreted, in certain ways before they can provide some norms rather than others. In short, then, I do not think that Wittgenstein was a normativist in the sense that the conditions of correctness governing the application of expressions are determined either by norms that exist independently of language users and which they somehow discover, or by norms that are somehow established by a community of language users. To put it succinctly, I do not think that, for Wittgenstein, there are norms preceding meaningfulness. But the more recent debate concerning normativity has focused on what norms, if any, may follow from meaningfulness. Thus I turn to the second more robust sense in which meaning may be thought to be normative.

4. Meaning is normative, in this sense, in that statements about the meaning of expressions, in effect, about their conditions of correct application, entail prescriptions or obligations about how to use the expressions. The debate here is two-fold. First, there is the question whether the prescriptions entailed are categorical or hypothetical. Categorical prescriptions tell speakers what to do (what they should or may do) with the expressions regardless of the goals speakers want to achieve by using them. Hypothetical prescriptions tell speakers what to do with the expressions depending on the desires they have in using them. E.g., if I want to tell the truth, I should apply ‘green’ only to green objects (provided, of course, that I mean green by ‘green’). Second, there is the question whether, if those prescriptions are merely hypothetical, the normativity that belongs to meaning is of an interesting or genuine variety, that is, a variety that distinguishes it from that which applies to any fact, including any natural fact.

It is hard to believe that anyone has ever seriously subscribed to the claim that meaning is categorically normative (though see Boghossian 1989, 533), where this means that, in order to mean something by an expression, a speaker ought to use it correctly “quite independently of what she wants to do.” (Hattiangadi 2006, 228). If this were the case, it would follow, absurdly, that no one can ever tell a lie. Softening the claim by saying that one has a prima facie semantic obligation to use expressions correctly is of no help (contra Whiting 2000). I may have a prima facie moral obligation to do so, which could be overridden by mere desires only at the cost of making me immoral. But my alleged semantic obligation surely could be so overridden, at no semantic cost whatsoever (except of course that I would have wrongly described a state of affairs). All that immediately follows from the fact that meaningful expressions have conditions of correct application is that statements about the meaning of expressions imply hypothetical obligations. To repeat, they tell speakers how to apply their expressions given the desires they have. Now no one denies that meaning is normative in this hypothetical sense. The question is, does it follow from this that meaning is genuinely normative? Is everything not potentially normative in this sense? Take the favourite analogy used by those who deny that meaning is genuinely normative. Thus compare the hypothetical obligations implied by statements about the meaning of expressions with other means/end prescriptive statements such as, “If I want to stay dry, I should go outside only if it is not raining”. Obviously, that facts about the weather dictate how I should behave, given my desires, does not make these facts genuinely normative. Similarly, it is argued, that statements about the meaning of my expressions dictate how I should use them, given my desires, does not make meaning genuinely normative. (See, e.g., Boghossian 2005, 207) I beg to differ. I think there is an important disanalogy between hypothetical prescriptions involving the weather and those involving meaning.

Facts about the weather do not always dictate how I should behave, say, when planning to go out; they may become irrelevant, as in the case where I no longer care about staying dry. But facts about linguistic expressions, i.e., their conditions of correctness, always dictate how I should behave when intending to produce a meaningful utterance. Indeed, they dictate my linguistic behaviour regardless of what my specific desire is, that is, not independently of my desire, but regardless of whether my desire calls for a correct application or for an incorrect one. Thus, depending on my desires, I should apply expressions in certain ways, correctly or not, and this is obviously dependent on what their conditions of correctness are to begin with. What this brings out is the claim that statements about the meaning of expressions always imply hypothetical prescriptions that, unlike those implied by statements about the weather, speakers must take into account. And this, it seems to me, does bring out a sense according to which meaning may be deemed to be genuinely normative. For, according to this sense, if none of the hypothetical prescriptions that flow from statements about the meaning of my expressions has application to me, then I do not mean by them what the statements say they mean; the statements become false. On the other hand, if none of the hypothetical prescriptions that flow from statements about the weather conditions has application to me, this in no way affects the truth-value of those statements. The weather conditions do not change; they just become irrelevant. This suggests that normative implications about how to use expressions are essential to meaning; they indeed follow from expressions having conditions of correct application. They are part of what it is for expressions to mean what they do. As E.H. Gampel has put it, meaning facts are “essentially such as to guide action”. (1997, 229) Other facts which are truly only contingently normative are guides only because we happen to take them as guides. But meaning facts are the facts they are because they guide us in certain ways, because they have normative implications.

Was Wittgenstein a normativist in this sense? He ought to have been, since, if I am right, this sense of normativity follows simply from the claim that meaningful expressions have conditions of correct application. What needs to be stressed, though, is that it follows from this because of a lesson Wittgenstein has taught us perhaps better than anyone else. This is that nothing to which meaning could be thought of as reducible, or in terms of which it could be thought of as explainable, could on its own provide the conditions of correctness that govern the application of expressions. Only meaning facts, i.e., meaningful expressions, wear their conditions of correct application, and hence their normativity, on their sleeves. Any other fact has to be regarded, interpreted, in some way or other in order to become normative. But meaning facts cease to be meaning facts if their normativity is not intrinsic to them, and they become different meaning facts if their normative implications change. Thus by returning to the Wittgensteinian sources of the normativity debate, we are reminded that it all started with the “trivial” recognition that meaningful expressions have conditions of correct application. As Wittgenstein laboured to show, there is much difficulty in discovering what will provide those conditions. Indeed it turns out that nothing short of meaning will do. This is why Wittgenstein could not be a normativist in the first robust sense but has to be one in the second sense. What this further suggests to me is that the claim, so often heard, that if meaning is genuinely normative then it cannot be naturalized is mistaken. Rather, it is because meaning cannot be naturalized that its normativity is intrinsic to it.

Refbacks

• There are currently no refbacks.