1 Introduction

Ever since the publication of Kripke’s Wittgenstein on Rules and Private Language, there has been a raging debate in philosophy of language over whether meaning and thought are, in some sense, normative (Kripke, 1982). Most participants in the normativity wars seem to agree that some uses of meaningful expressions and/or some sorts of thoughts are semantically correct while others are incorrect. What they mainly disagree over is whether this, by itself, entails anything normative, for example, that they are also semantically permissible or forbidden. Many philosophers have assumed or argued that it does (Boghossian, 1989; Glock, 2019; McGinn, 1984; Millar, 2002; Whiting, 2007, 2009, 2016). Many others have argued that it doesn’t (Glüer, 2001, 2013; Glüer & Wikforss, 2009, 2015; Hattiangadi, 2006, 2007; Wikforss, 2001).

But what is it to say that a use of an expression or a thought is semantically correct? On the so-called orthodox construal, it is to say that it doesn’t amount to or result in a factual mistake, that is, in saying or thinking something false (Hattiangadi, 2006, 2007; Whiting, 2007, 2009, 2016). In slogan form: correctness consists in truth. This conception is usually taken to apply to both uses of meaningful expressions and certain sorts of thoughts and its putative normative consequences or lack thereof have been discussed to death (see e. g. Glüer & Wikforss, 2015, 2018; Miller, 2020; Whiting, 2009).

However, there is also an alternative construal on which to say that a use of an expression is semantically correct is instead to say that it doesn’t result in a distinctively linguistic mistake, that is, in misusing the expression (Buleandra, 2008; Dummett, 1991; Glock, 2019; McGinn, 1984; Millar, 2002, 2004; Moore, 1954). In slogan form: correctness is accordance with meaning. This conception is supposed to primarily apply to uses of linguistic expressions and its putative normative consequences have not been as widely discussed.Footnote 1

It is entirely natural to think that these two construals of semantic correctness are simply about different things and not in competition with each other. Semantic correctness as consisting of truth is one thing, semantic correctness as accordance with meaning is another, and both should be discussed. However, perhaps surprisingly, this is not the common view. Instead, several philosophers who subscribe to the orthodox construal have argued that the alternative construal of correctness as use in accordance with meaning doesn’t make any sense, partly because there are no clear cases of linguistic mistakes (Whiting, 2016; Wikforss, 2001).

My aim in this paper is not to argue against the orthodox construal which I think is interesting and important in its own right, though of limited applicability in the case of language use. Instead, I want to develop and defend the idea that there is a distinctively linguistic notion of correctness as use in accordance with meaning and argue that there are clear cases of linguistic mistakes. I will do this in three steps.

First, since part of the resistance to the notion of linguistic correctness as use in accordance with meaning is that it is supposed to be unclear what this amounts to, I will explicate it in more detail. I’ll start by briefly discussing the orthodox construal, only to make clear that if we are interested a distinctively linguistic notion, one that derives from the nature of linguistic meaning itself, then this is not what we have in mind. This is because the orthodox construal applies to only acts or states that are governed by a norm of truth. However, at best, only uses of declarative sentences are governed by such a norm which means the orthodox construal only applies to them. Yet, if we’re after a distinctively linguistic sense of correctness then we’re interested in something that should apply to all relevant uses of all meaningful expressions. (Sect. 2).

Second, I’ll argue that the distinctively linguistic notion of correctness makes sense if pursued from what I will call the public language perspective. On this common-sensical and entirely familiar point of view that dominates contemporary philosophy of language, linguistic meanings belong to expression-types in public languages like English, Estonian, or Esperanto. I’ll argue that this notion of linguistic meaning can be usefully explicated in terms of what the expression is semantically for doing in the language, and that this idea can further be explained in terms of its having use-conditions. To use an expression in accordance with its meaning is just to use it while being in its use-conditions. In contrast to the orthodox construal, this notion applies to all relevant uses of all meaningful expressions. (Sect. 3).

Even given something like the above explication, the skeptics have insisted that the notion of use in accordance with meaning is problematic since there are no clear cases of linguistic mistakes. One standard suggestion is that uses by speakers who are mistaken about an expression’s meaning constitute misuses (Dummett, 1986; Glock, 2019; Millar, 2002, 2004). It is sometimes objected that these should be rather reinterpreted as cases of speaking a different language or engaging in linguistic innovation. I will argue that once we distinguish between complete incompetence and being mistaken about meaning there is very little plausibility to the claim that these cases should be reinterpreted. The skeptics then claim that the appearance of a linguistic mistake is due to the presence of a desire to communicate or intentions to speak properly which are extrinsic to meaning and language (Bilgrami, 1993, 2012; Whiting, 2016; Wikforss, 2001). I’ll argue that this interesting response depends on an implicit shift to a Davidsonian individualist perspective which denies the significance of public language and operates with a radically different notion of “meaning”. I will show that from the public language perspective, the relevant intentions are intrinsic to meaning and language. Absent such intentions there is no speaking a public language like English or Estonian. The upshot is that from the public language perspective from which the notion of linguistic correctness is pursued, uses by speakers who are mistaken about meaning constitute perfectly good cases of linguistic mistakes. (Sect. 4).

2 The Orthodox Construal

Part of the resistance to the notion of linguistic correctness as use in accordance with meaning is that it is supposed to be unclear what this amounts to. My aim in this first, preparatory section, is to briefly discuss the orthodox construal, only to make clear that if we are interested a distinctively linguistic notion then this is not what we have in mind.

On the orthodox construal, to say that a use of an expression or a thought is semantically correct is to say that it doesn’t result in a factual mistake, that is, in saying or thinking something false. The paradigm cases here are the following. Suppose you apply a predicate like ‘is British’ to Gottlob, perhaps in using the declarative sentence ‘Gottlob is British’ with its meaning in English and saying that Gottlob is British. Then to say that your use was incorrect is just to say that you misapplied the predicate, applied it to something to which it doesn’t apply, and said something false. Similarly, suppose you judged that Ludwig is German. Then to say that your judgment is incorrect is just to say that it is false.

This sort of correctness applies to all acts or states that satisfy the following two conditions:

(Content):

they have representational content and truth-conditions; and

(Norm):

they are governed by a norm of truth (or one that entails truth, like knowledge)

Hence, we can talk about this sort of correctness of judgments, beliefs, and certain sorts of uses of predicates and declarative sentences, e. g. to say things or make assertions. This can be called semantic correctness in the sense of ‘semantic’ in which the word is used to talk about representation and truth-conditions in general and not exclusively about linguistic meaning. Semantic correctness on this construal amounts to nothing more than representational correctness, satisfying the relevant norm of truth.

It is paramount, in understanding this notion, that an act or state must have both of the above characteristics to be correctness-apt. Being characterized by Content alone doesn’t suffice. If we hypostasize representational contents into propositions, then we can of course say that they have truth-conditions and are true or false. But from this alone absolutely nothing follows about the correctness or incorrectness of uses of sentences or thoughts. When you perform the act of entertaining the proposition that Gottlob is British in the sense of simply calling it to mind you don’t do anything correct or incorrect no matter whether the proposition is true or false. This is because entertaining is a state that isn’t characterized by Norm. This in stark contrast with judgments and beliefs which are governed by a norm of truth or knowledge.

The orthodox construal captures a perfectly acceptable thing one can mean by ‘semantic correctness’. We might call it representational correctness. However, suppose you are interested in a distinctively linguistic notion of correctness, one that derives from the nature of linguistic meaning itself.Footnote 2 Then you have in mind a quite different notion. This is not a new claim and the standard reason given for it is that you can use a sentence linguistically correctly while still making a factual, representational mistake (Moore, 1954: 308, see also Schroeder, 2008). This is true, but I want to make the point a bit differently. Let’s start from the fact that if you thought that there is a distinctively linguistic notion of correctness then, since it’s supposed to derive from the nature of linguistic meaning, it should apply to all relevant uses of all meaningful expressions.Footnote 3 But then it’s easy to see that you must have something else in mind than representational correctness because that applies at best only to uses of declarative sentences since only these are governed by a norm of truth.

To see why, compare the following five sentences in different moods:

  1. 1.

    ‘Bertrand is British’

  2. 2.

    ‘Is Bertrand British?’

  3. 3.

    ‘What time is it?’

  4. 4.

    ‘Read the paper!’

  5. 5.

    ‘Ouch!’

At best, only (1) can be used to say something and is thereby governed by a norm of truth. In contrast, (2) and (3) are used to ask questions. Acts of asking questions are not governed by a norm of truth. And even though they have answerhood-conditions, they’re not correct or incorrect dependent on receiving an answer. Similarly, (4) is used to tell someone to do something. Acts of telling someone to do something are not governed by a norm of truth either. And even though they have fulfilment-conditions, they’re not themselves correct or incorrect depending on being fulfilled. Finally, (5) is used to express pain and such acts are clearly not governed by a norm of truth nor do they involve any sorts of satisfaction-conditions. Can (2)—(5) still be linguistically misused? It’s natural to think that they can, but the orthodox construal can’t capture this.Footnote 4

To sum up, there’s absolutely nothing wrong in using ‘semantic correctness’ to talk about representational correctness, asking whether this entails anything normative and doing it under the rubric of ‘normativity of meaning’. However, it should be also clear that those who think there’s a distinctively linguistic notion of correctness have something very different in mind.

3 Linguistic Correctness

On the alternative construal, to say that a use is semantically correct is to say that it doesn’t result in a distinctively linguistic mistake, that is, in misusing the expression. In other words, it is to say that it is used in accordance with the expression’s meaning (McGinn, 1984; Millar, 2002, 2004). Part of the resistance to this notion consists in the claim that it is unclear what this amounts to (e. g. see the discussion of McGinn’s and Millar’s explications in Wikforss, 2001: 210–211 and Whiting, 2016: 227–229). Thus, our first task is to arrive at a clear conception of what this means.

The notion of linguistic correctness as use in accordance with meaning makes sense if pursued from what I call the public language perspective. This is a common-sensical and entirely familiar point of view which clearly dominates contemporary philosophy of language and is common to philosophers otherwise as different as Burge, Brandom, Dummett, Evans, Kaplan, Lewis, Millikan, McDowell, Kripke, Perry, Recanati, Searle, Soames, Stalnaker etc., the main two outliers being Chomsky and later Davidson. Its essence can be captured with the following statement: linguistic meanings belong to expression-types in public languages like English, Estonian, or Esperanto. We can elaborate on this as follows:

  1. 1.

    Types: The primary units of importance are expressions qua types and not their individual uses on particular occasions.

  2. 2.

    Meaning: An expression’s linguistic meaning in a language is what competent speakers of the language have a grasp of. It is what makes it possible for them to use the expression to speak that language. It is what language-learners aim to grasp. It is a standing, stable, context-invariant property of the expression.

  3. 3.

    Language: Languages like English, Estonian, or Esperanto are some sorts of public, communal, or, minimally, shared entities, sociolects rather than idiolects. For example, philosophers from Dummett to Burge to Kaplan would think of them as historically embedded, ongoing, rule-governed social practices (Dummett, 1991: Ch. 4, Jackman, 1999, Ridge, 2020). However, one could think of them also in a somewhat more local and temporary manner (e. g. see Armstrong, 2016).

This perspective constitutes a core take on the phenomena of meaning and language while being compatible with different ways of spelling out the details.Footnote 5

The relevant notion of linguistic meaning can be usefully explicated in terms of what the expression is semantically for doing in the language. What an expression is semantically for doing in the particular language is what its meaning enables us to use it to do, such that we couldn’t use meaningless expressions or expressions which have a different meaning to do these things.Footnote 6 For example, names and other referential expressions are plausibly for talking about particular people and objects while predicates are for expressing properties. For example, it is widely assumed that in English ‘Bertrand’ is for talking about some specific person, ‘I’ is for talking about oneself, and ‘is British’ is for expressing the property of being British. Moving to sentences, declaratives are for saying things, interrogatives for asking questions, imperatives for telling people to do things, and expressives for expressing one’s mental states.

It is a short step from the idea that expressions are semantically for doing certain things to the idea that expressions have conditions of correct use or use-conditions. For example, and please treat these just as illustrative, ‘Bertrand’ is perhaps for using while you’re thinking or referring to Bertrand and thus its use-conditions are that one has to be performing these mental acts (Hanks, 2015; Soames, 2010b). Similarly, ‘I’ is for using while you’re thinking of yourself in a first-personal or de se way, and ‘is British’ while you’re expressing the property of being British. Finally, a declarative sentence like ‘Bertrand is British’ is for using when one is doing the above acts and further predicating the property of the person, resulting in your entertaining the proposition that Bertrand is British (Soames), or judging it to be the case (Hanks). To take a different sort of example, the expressive interjection ‘Ouch!’ is for using while you’re in pain whereas ‘Oops!’ is for using when you’ve just observed a minor mishap (Kaplan MS). Finally, situational terms like ‘Hello!’ or ‘Goodbye!’ might be for using while you’re meeting someone or parting from them etc.

Given this this understanding, the most straightforward gloss on the notion of use in accordance with meaning is as follows:

Linguistic Correctness::

To use an expression in accordance with its meaning is to use it while being in its use-conditions

I think that this amounts to a perfectly clear explication of the notion and fits the bill in applying to all relevant uses of all meaningful expressions.Footnote 7

Now, even given something like the above explication, the skeptics have claimed that the notion is still problematic since there are no clear cases of linguistic mistakes. In the next section I will argue that there are.

4 Mistakes About Meaning

It is a standard suggestion that uses by speakers who are mistaken about an expression’s meaning constitute misuses (Dummett, 1986, 1994: 265, Glock, 2019: 302–303, Millar, 2004: 162–164, Moore, 1954: 309). Consider Millar:

If, though I aspire to use a word in keeping with its received meaning, I am wrong about its received meaning, then I may misuse it. If I thought ‘arcane’ meant ancient then I would be liable to use the word as if that is what it meant. (Millar, 2004: 162)

Let’s add some non-predicate examples. Imagine a student who mistakenly thinks that ‘Bertrand’ is Gottlob’s name, confusing the two philosophers. Suppose further that she tries to use ‘Bertrand’ to talk about Gottlob on a particular occasion, namely, by using it and thinking of Gottlob. If you would ask her who she’s talking about she’d tell you that she’s talking about the German guy, inventor of modern logic and so forth. The suggestion is that she’s used ‘Bertrand’ incorrectly because ‘Bertrand’ is for talking about Bertrand, and not about Gottlob.

To take another example, consider a variant of the story of the cyclops Polyphemus from Homer’s Odyssey on which he is mistaken about the fact that ‘I’ is for talking about oneself because he thinks that it is Odysseus’s name.Footnote 8 Suppose he uses ‘I’ while trying to talk about Odysseus, namely, by using it and thinking of Odysseus in screaming ‘I did it’ as a response to the question who blinded him. If you would’ve asked him to point to who he was talking about he would’ve pointed to Odysseus. The suggestion is that Polyphemus has used ‘I’ semantically incorrectly because ‘I’ is for talking about oneself, and not about others.

To take a final and in many ways the simplest example, suppose you’re going for a trip to Tallinn and I teach you the Estonian words for ‘Hello!’ and ‘Goodbye!’, ‘Tere!’ and ‘Nägemist!’. However, you misunderstand or misremember and come away thinking that ‘Nägemist’ means ‘Hello!’ not ‘Goodbye!’. You then go on to use the expression while meeting your hosts, wanting to greet them. The suggestion is that you would be misusing it because ‘Nägemist!’ is for using on parting and not on meeting.

It’s natural to think that in all of the above cases one is making a linguistic mistake. Skeptics like Daniel Whiting disagree:

...if I use a word which in English signifies the colour red to signify or speak of the colour blue, it hardly follows that I have made a mistake – perhaps I am speaking some variant of English, or have introduced a linguistic innovation, or am speaking a different language. (Whiting, 2016: 230).

The idea seems to be that we don’t have to think of the speakers in the above cases as using the relevant expressions while lacking a grip on their use-conditions, rather we could reinterpret them as speaking a different language or engaging in linguistic innovation.

There is something to Whiting’s suggestion, but we have to look at the cases in more detail and one by one (compare Burge 2003: 118–119). The issues here are complicated by the fact that ‘use’ is crucially ambiguous. In the most basic sense, to use a word is just to produce a token of it by uttering or inscribing it. Meaning need not be involved at all. In Austin’s terms, such mere uses result in phonetic acts. This can be compared with merely moving chess pieces around on the board when enacting a play for the kids, say. Rules of chess need not be involved at all. However, for our purposes the relevant sorts of uses are what I like to call uses with meaning (compare Kaplan, 1989: 602). In Austin’s terms, such uses result in rhetic or locutionary acts. These can be compared with moving a chess piece to make a move in the game.Footnote 9 It should be obvious that only the second sorts of uses can result in linguistically correct or incorrect uses.

Now, completely incompetent speakers, those who have no grip on the expression’s meaning nor even any thoughts about what it might be, can’t even try to use an expression with its meaning. For example, suppose I teach you how to pronounce the Estonian sentence: ‘Lumi on valge’ without telling you what it means (and suppose you also don’t form any hypothesis about what it means). It should be clear that in such a situation you can’t even try to use it with its meaning, try to speak Estonian (compare Austin, 1962: 97, Burge, 2003: 118). If you’d utter the sentence then perhaps we should indeed reinterpret you as innovating or speaking a different language.Footnote 10

But our cases are not cases of complete incompetence, but of being mistaken about the meaning. I teach you to say ‘Nägemist!’, but you misunderstand or misremember what it means. Is your predicament like complete incompetence in that you can’t even try to speak Estonian? No. It’s natural to think that in such a situation you can at least try to speak Estonian. You do that by using the sentence with the intentions to participate in the practice. Your use thereby goes beyond a mere use and is a use with meaning, and if you’re not in the use-conditions then you count as misusing the expression. And in this case there seem to be no independent reasons to reinterpret you as not trying to speak Estonian but engaging in linguistic innovation or speaking a different language. That is clearly not what you yourself think you’re doing! Similarly, it’s hard to see what reasons there are to think that speakers who are mistaken about what ‘arcane’ means, whose name ‘Bertrand’ is, and what ‘I’ is for talking about, should be reinterpreted. After all, what the speakers actually do in such cases if their error is pointed out to them is simply adjust their understanding and usage (Burge, 2003: 131, compare Dummett, 1986: 462).

There is a standard response to this which was first forcefully made by Bilgrami:

What sense of norm, then, do I accept? It emerges form the fact that individuals intend to speak like others in the community speak and they intend to speak as they have in the past rather than waywardly. This means that individuals do intend to speak in a way that is natural to describe as, and that they themselves describe as, correctly. … we have a norm that is perfectly sensitive to the social linguistic practices that surround an individual. But the pragmatic explanations underlying the norm make it clear that it is … an extrinsic norm. It says “Speak like others do, if it pays to do so”. Or more specifically it might say “I ought to use words as others do, if I want to be easily understood.” (Bilgrami, 1993: 134–135, compare Davidson, 1992: 261, 1994: 9, Wikforss, 2001: 211)

The basic idea behind the rejoinder is that the appearance that the above cases constitute misuses and the reasons why we adjust our understanding and usage are due to the presence of intentions to speak like others do. Such intentions are extrinsic to language and meaning. As applied to our central case of ‘Nägemist!’, you have a desire to be understood and an intention to speak like others do, and it is only because of these extrinsic factors that your use while not being in the use-conditions counts as some sort of a mistake. But it’s not a linguistic mistake or misuse in the relevant sense.Footnote 11

Similarly, consider how Whiting continues the passage I quoted above:

Of course, if I’m intending to speak (‘proper’) English, then I am making a mistake insofar as I am failing to execute my intention. But any norm that would deliver the verdict that such a mistake has been made would, once more, be instrumental. (Whiting, 2016: 230)

Like Bilgrami, Whiting seems to say that the appearance that the above cases constitute misuses depends on the presence of intentions to speak “properly”. Again, such intentions are extrinsic to language and meaning. As applied to our case of ‘Nägemist!’, you have an intention to speak “proper” Estonian and that’s why your use while not being in the use-conditions counts as a mistake.

I must confess that it took me years to understand this line of thought. I finally got it when I realized that it relies on an implicit rejection of the public language perspective and subscription to the Davidsonian individualist persrpective which denies the significance of public language and operates with a different notion of “meaning”. Let me elaborate.

The rejoinder relies on the Davidsonian view that we have an independent power to imbue our uses of words with “meaning” by using them with meaning-intentions (see Davidson, 1986, Bilgrami, 1993: 142, Bilgrami, 2012: 113, for useful discussion see Glüer, 2011: Ch. 2, Glüer, 2013: 341–342, also Camp, 2013). From this point of view, the primary units of importance are particular uses of expressions, that is, token utterances and inscriptions. The role of intention is to imbue these particular uses with “meaning” on that occasion (Glüer) or context-locally (Camp). Thus, when we use an expression, we can choose to imbue the particular use with the same “meaning” that it has in the shared “language”, understood in terms of an overlap in dispositions to use, or with a different “meaning”. This depends on whether we have a further desire or intention to speak like others do or “properly”. On this view, what happens when you utter ‘Nägemist!’ in our case is that you imbue your utterance with a “meaning” to the effect that you are greeting your hosts. But then the utterance by itself doesn’t constitute a mistake. It only constitutes a mistake given an extra, optional desire or intention to speak like others do (Bilgrami) or “properly” (Whiting). From such a Davidsonian perspective there is no need for the notion of a distinctively linguistic mistake, one that derives from the nature of linguistic meaning itself, nor is there any room for it since “meaning” is a one-off property of uses.

From the public language perspective this gets things completely backwards. First, those who subscribe to this perspective balk at the idea that we have an independent power to imbue our uses of words with meaning.Footnote 12 At best, one can use words while having things in mind or speaker mean things a la Grice. Second, as was emphasized above, on this picture the primary units of importance are expressions thought of as types and not particular uses. It is these that have linguistic meanings in languages like English or Estonian. The role of intention is not to imbue the use with meaning, but to make it the case that the use is a use of the expression-type which has a meaning and thereby activate the meaning that is already there (compare, again, Kaplan, 1989: 602). In other words, when we use an expression we can either merely use it or try to use it with its meaning in the particular language and thereby participate in the relevant practice (compare Reimer, 2004: 328–329). What happens when you utter ‘Nägemist’ in our case is that you try to speak Estonian and try to use the sentence with its meaning. Your use thereby goes beyond a mere use and is a use with meaning, but since you’re mistaken about what the meaning is, you do so while not being in the use-conditions and thereby count as misusing it. The crucial role is played by the intention to use the expression with its meaning in Estonian, and this intention is intrinsically related to language and meaning. Absent such intentions there is no speaking a public language like English or Estonian. But once we have this intention in place, no further extrinsic desire or intention to speak like others do or “properly” is relevant or required to deliver the verdict that the above use is a misuse.

Let’s sum up. The skeptics claimed that the notion of linguistic correctness is problematic since there are no clear cases of linguistic mistakes. I have discussed the standard suggestion that uses by speakers who are mistaken about an expression’s meaning are misuses. It is sometimes objected that these should be rather reinterpreted as cases of speaking a different language or engaging in linguistic innovation. I’ve argued that once we distinguish between complete incompetence and being mistaken about meaning there is very little plausibility to the claim that these cases should be reinterpreted. The skeptics then claim that the appearance of a linguistic mistake is due to the presence of a desire to communicate or intentions to speak properly which are extrinsic to meaning and language. I have argued that this depends on an implicit shift to a Davidsonian individualist perspective which denies the significance of public language and operates with a radically different notion of “meaning”. And I have argued that from the public language perspective the relevant intentions are intrinsic to meaning and language.

One might wonder whether this doesn’t ultimately show that the viability of the notion of distinctively linguistic correctness as use with accordance to meaning depends on whether the public language versus Davidsonian individualist perspective is correct. Quite so! The problem is that this is not at all how the dialectic is generally presented in the literature. Those skeptical of linguistic correctness tend to make it seem like there are no clear cases of linguistic mistakes, no matter one’s background perspective.Footnote 13 This is emphatically not the case and the notions of public language, linguistic correctness, and linguistic mistake form a package deal. If one wants to call the latter into question one should tackle the public language perspective head on. In sum, from the public language perspective from which the notion of linguistic correctness is pursued, uses by speakers who are mistaken about meaning constitute perfectly good cases of linguistic mistakes.Footnote 14

5 Conclusion

Where does this leave the debate over normativity? Is it plausible, on this construal of linguistic correctness, that correct uses are also, in some sense, permissible? As indicated above (see fn. 7), thinking of meaning in terms of use-conditions is by itself compatible with both normative and non-normative construals. One can think of it in terms of rules of use and permissibility, but also in terms of conventional regularities in use or overlap in dispositions to use. Of course, the people coming from the public language perspective like Burge, Dummett, Kaplan, Searle etc. have most frequently thought of them in terms of rules of use. That seems to suggest that for the concept of linguistic meaning that has been under discussion, normativism is relatively plausible. However, ultimately, this will depend on which of the three indicated ways of thinking about it is best, and this can only be settled by weighing the costs and benefits of the three packages as wholes.

Be that as it may, it is important to understand that even if linguistic meaning is normative, it would be normative only in the way that law, etiquette, and games are. In Derek Parfit’s terms, the notion of normativity in play here is that of rules and not of (authoritative, genuine) reasons—what more recently has come to be called formal versus authoritative normativity (Baker, 2017, Finlay, 2019: 204–208 Parfit, 2011: 144–146). This sort of normativity isn’t by itself in any way reason-giving or authoritative like prudence or morality are usually taken to be. But the thesis of normativity of meaning is therefore no less interesting. Even formal normativity sets strong constraints on theories of meaning. To come back full circle to Kripke’s discussion, it seems to rule out dispositionalism all by itself.Footnote 15