Language and Responsibility
Language and Responsibility


In this paper, I develop a general implication of a Wittgensteinian view of language, namely that all instances of language use involves an element of freedom and thus of personal responsibility. In the first section, I unfold the background for this idea while I in the second look at how we distinguish between the responsibilities we face in language.

Table of contents

    One of the important things we learn from Wittgenstein’s highly influential treatment of rule-following is that all forms of language use is normative, and that the stability of this normativity depends on a contribution from the individual language user. In the rule-following investigations, Wittgenstein focus on a range of normatively structured activities such as reading, developing a series of numbers or following the rule +2; activities that are characterised by general uniformity in the behaviour of the rule-followers. Despite the existence of such agreement, Wittgenstein wants us to resist the idea that the normativity of such activities is established by something externally to the activity itself that determines our behaviour. In §219 Wittgenstein famously discusses this way of picturing basic normativity in. He begins by drawing up the desired picture: ‘“All the steps are really already taken” means: I no longer have any choice. The rule, once stamped with a particular meaning, traces the lines along which it is to be followed through the whole of space’. The appeal of the picture lies in the fact that it portrays interpersonal uniformity as guaranteed because of an elimination of the contribution of the individual. Wittgenstein goes on to question the use of this picture. ‘– But if something of this sort really was the case, how could it help?’ I may feel as if the rule is already laid out in advance, but even if this actually is the case, this ‘in advance’ is not what I have access to; what I know is the rule and particular applications of it. Platonic rules do not do any work in my application of the rule; even if the picture of ‘rules as rail’ does. And Wittgenstein famously goes on to make us see that, instead of serving as a guarantee of the normativity of the rule and its correct application, the picture serves a different purpose: ‘No; my description only made sense if it was to be understood symbolically. – I should have said: This is how it strikes me. When I obey a rule, I do not choose. I obey the rule blindly’. What I want to emphasise here is the massive use of ‘I’ in Wittgenstein re-description of the ‘symbolic’ picture. Wittgenstein addresses the simple and automatic sense in which we follow certain basic rules, but he also emphasises that what we cannot eliminate contribution of the individual; it is always an ‘I’ that acts in this way. We cannot account for rule-following and thus for linguistic practice without the notion of an individual using language.

    Wittgenstein’s considerations on rule-following form a part of the general background of my interest, because they show how there is no domain of language where we proceed without some form of contribution on our part. Stephen Mulhall (2000) has addressed this implication of Wittgenstein’s investigation by looking at our understanding of particular concepts, and he presents two possible notions of the grammar that guides this understanding; the determinant and reflective model of grammar. The first, the determinant model, holds that all we need in order to know whether a word has been correctly applied is to have a sufficient grasp of the grammar of that word. If we encounter forms of use that do not conform to ordinary grammar, we have the choice either to dismiss this use as faulty or misunderstood, or to let this case establish a new use, that is, establish a new concept with a new grammar. The determinant approach thus suggest that the norms involved in the grammar of concepts are primary to and independent of actual instances of that use much in the same way as the rules of chess are independent of actual games of chess – a parallel that we might take to be implied by Wittgenstein’s notion of a ‘language-game’.

    Mulhall is however not satisfied that this view accurately describes what Wittgenstein is trying to show us, and he introduces another notion of linguistic understanding, the reflective approach. This approach is meant to reflect Wittgenstein’s insistence that it is often possible for us to understand quite unfamiliar or divergent uses of words, an insistence that is for example reflected in Wittgenstein’s effort to find a way of understanding the Augustinian claim, introduced at the very beginning of the Philosophical Investigations, that all words in language functions as names. According to Mulhall, Wittgenstein’s story of the builders is meant to establish a context in which this claim could be taken to describe a (admittedly very primitive) language. That is, Wittgenstein constructs a context in which it is possible for us to address the question of whether we can make sense of Augustine’s use of the concept of language as a group of names as a concept of language. That is, Wittgenstein does not accept the dichotomy presented by the determinant approach, according to which he either has to dismiss the Augustinian claim as a wrong use of the concept of language or let it introduce a new, rather different concept. In contrast, Wittgenstein is trying to make us reflect on the differences and similarities between Augustine’s view and our ordinary grammar of the concept of language.

    In this way, Wittgenstein shows us how, in Mulhall’s words, ‘the degree of resemblance needed to ground the projection of a concept of a language into this (or any) context is importantly open to individual judgement’ (Mulhall 2002: 313). Our grasp of the grammar of a concept and the criteria (in Mulhall’s Cavellian sense) that is connected to it, of course guide our assessment of such projections, but, and this is Mulhall’s point, the question of whether the norms and criteria involved in this grammar is met is open to judgement and therefore ‘ultimately rest with the individuals invited to project those criteria into this imagined context’ (ibid. 314). Even if our understanding of the grammar of a concept limits the range of uses of that we will accept as meaningful uses of that concept, any such use is also context specific in a way that makes it dependent on the understanding and imagination of the individual language-user. Mulhall sums up his point by saying that ‘any concept must be flexibly inflexible in these ways: its normativity is of a kind that enables or rather constitutes individual freedom of judgement, because its grammatical schematism is such that our projections of words are at once deeply controlled and ineliminably creative’ (ibid. 315).

    What I want to note is a consequence of this view of language, namely that the individual’s essential contribution to all forms of language use means that such use always involves an element of personal responsibility. In Mulhall’s discussion, he primary focus on our understanding of other people’s use of concepts, but we can turn the perspective round and see that his point also applies to our own uses of language. In my actual uses of language, I act from the ‘flexibly inflexible’ nature of concepts and this means that I on the hand should be able to account for how a specific context invites or allows for the use of a concept, while I on the other recognise that my use is at the same time an exercise of freedom. In talking, when I use language, I can never refer to something that will definitely settle or justify the right or appropriate use; this use ultimately also relies on my individual powers of judgement and my ability to justify such judgement.1 This means that any utterance we make, inevitable involve some form of responsibility; that is, responsibility – and possible guilt – is built into all dealings with language. Moreover, the responsibility involved in language use springs from the activity that establishes linguistic normativity, and the element of responsibility therefore cannot be eliminated.

    Even if all uses of language is subject to individual responsibility, we nevertheless does not seem to consider the question of individual responsibility equally pertinent in all cases. Typically, we do not stress the responsibility involved in uses that we are tempted to describe along the lines the lines of ideally rigid rails, while we for example are more likely both to feel and hold other responsible for their description of other people. I describe the reason why we distinguish between the responsibilities of different language uses, but I will also argue that we should not let these reason led us to an idea of essentially different ‘forms’ of linguistic responsibility.

    As the initial description of linguistic responsibility is completely general, it cannot account for the differences in the responsibility we connect to different instances of language use My suggestion is that we instead turn to Wittgenstein’s investigations of linguistic normativity, and look at the difference between the rules investigated in the section on rule-following and the uses of language where questions of responsibility becomes pertinent. One difference is the amount of agreement that we can expect within these different practices of language. We can spell out this difference if we compare basic rule-following with one of Wittgenstein’s investigations of the use of moral concepts, that is, look at the contrast between the almost uniform agreement involved in activities such as reading or doing mathematics and the notorious possibility of disagreement in morally relevant discourse. One important difference is that a part of the point of learning basic mathematics is to able to participate in a commonly shared practice that enable us to reach equivalent results. This means that when we learn to do mathematics, we also learn not to place any value on the possibility of disagreement (cf. PI §240 and Diamond 1991a: 28). If I want to be able to add two, I can only do so by accepting that I must respond in a way that minimises my individual contribution to a ‘doing the same again’, acting ‘as the rule strikes me’ or simply ‘obeying the rule blindly’. If I do not respond in this way, my application of the rule is ruled out as meaningless; I simply will not be doing mathematics.

    However, it now seems as if the question of responsibility arises in two different ways, at two different levels, we might say. First, in so far as I want to do mathematics, it is my responsibility to act in a way that is meaningful within the frame of that language-game, that is, I have a responsibility that is embedded in the language-game and tied to its purpose. Secondly, I could be doing something else, and I therefore must take responsibility for choosing to engage in this language-game rather than another. If I am the accountant of a firm where a large sum is suddenly missing from the books, I might insist that I am simply adding the numbers (that shows the deficit), and I may do so perfectly, thus living up to any responsibility connected to the language-game of mathematics, but I might be to blame for the fact that I insist on doing mathematics and not for example responding to the question of where the sum has gone missing; or at least I am to blame in so far as this is my responsibility as an accountant. We may in this way identify two forms of responsibilities, where the first is internal to the language-game that I am engaging in and the other springs from my very choice of language-game. In the example, the difference between my mathematical and my professional or moral responsibility as an accountant.

    In contrast to the case of mathematics, we do in moral discourse not consider agreement a goal in itself; a difference in purpose that reflects on our evaluations of uses of ethical concepts. When parents teach a child to use an evaluative word like ’good’, they may consider it a sign of understanding if the child starts to use the word about objects that differs substantially from the ones that was used in the teaching – even if the parents do themselves not consider these objects good. That is, to use Mulhall’s concepts, we do in ethics accept wide limits for creative use, not just of moral concepts, but of concepts in general, while it is a part of our understanding of mathematics that we accept how mathematical activity is thoroughly controlled. We could paraphrase a remark from Philosophical Investigations and say that the kind of agreement is the kind of language-game (cf. PI part II xi: 191). As we have already seen, the difference in levels of agreement that we find in mathematical and morally motivated language is not categorical, but is a matter of degree. It does not arise because our use of mathematical concepts is completely controlled, while our use of ethical concepts unfolds without restrictions; instead it results from the different forms of variation we allow in different practices, and this in an important way depends on why we engage in them, their point.

    To look at the idea of the point of language use, we can draw on the number of places, where Wittgenstein discusses this difference between language-games by involving the idea of the purpose (‘Zweck’) of a word or a language-game (se for example PI 345, and LWPP I 890). Wittgenstein often opposes the idea that we may meaningfully talk of the purpose of language; what he wants, is instead to show us that we have a multitude of purposes in using language, and that such purposes are part of what determines meaningful use. Moreover, Wittgenstein links such purposes with the idea of a central or essential use of a word, for example in a 1949 version of recurring remark. ‘Non & ne --- They have the same purpose, the same use – with one qualification. So are there essential and non-essential differences among the uses? The distinction does not appear until we begin to talk about the purpose of a word’ (LWPP II 2, cf. LWPP I 384-5). To talk of the purpose of words may help us to distinguish between what is essential to our understanding of that word, and what is not. In general, to understand a statement a person makes, we need to have some grasp of what she wants to do in presenting this statement. Understanding her purpose is an integrated part of understanding the use she makes of her words. We find the point in Wittgenstein’s remarks, when he talks interchangeable about the purpose and the use of words. (LWPP I 291, 138, 326).

    However, if our understanding of the use of a word is connected to the purpose of using that word, then understanding draws on a very wide range of considerations about what it is meaningful to do, what is important etc. If someone spoke in a manner that revealed that he had very different purposes with his use of words, we would not just think that he had had a peculiar facon de paler, we would have much more general concerns. Thus Wittgenstein continues, ‘--- We might think it strange. “He doesn’t play our game at all” – one would like to say. Or even that this is a different type of man’ (ibid., my italics). When we try to understand what other people are saying, we draw on our general understanding of what they could want to do with their words, what the function or purpose could be. That is, words or language-games have different purposes because they fill out different roles in our lives. There are two implications of this. The first is that even the simplest uses of language connect to an elaborate understanding of what a human being is (see also Crary 2007). Secondly, the norms or criteria that guide our language-games are shaped to accommodate the purpose we have in engaging in them.

    The second implication means that the purpose of engaging in a particular language-game is part of what accounts for the differences in the responsibility, we attribute different uses of language. If we thought that an important part of the purpose of mathematics was to voice our convictions, then it would be impossible to do mathematics. Instead, we consider agreement a part of this purpose and this means that part of what we accept when we learn to do mathematic is that it only places a very restricted and well defined set of responsibilities both on ourselves and others. The reasons why we engage in morally relevant language use is very different, and I will venture the claim that one such purpose is exactly to voice our convictions, of value for example. Moreover, if this is right, then we in ethics value the possibility of speaking our mind higher than we value the possibility of reaching agreement on particular matters. That is, in order to be able to voice our own moral considerations, we allow that the statements of a wide variety of such considerations are understandable moral uses of language, and in doing so, we also allow for the possibility of widespread moral disagreement. In morally relevant language use we share a purpose that can be said to involve a shared acceptance of the possibility of widespread disagreement. That is, the possibility of disagreement – and the existence of such disagreement – in ethics does not reflect the failure of our present moral status, but is an integrated part of the grammar of morally relevant uses of language. Moreover, ´because of the possibility for disagreement, each of us, when engaging in ethical language use, undertake the responsibility that we should be able (at least in principle) to supply or describe the context that invites our particular use of words. This is so because the context is neither laid out in advance nor necessarily commonly shared, and this means that the responsibility connected to morally significant uses of words becomes much more far reaching than the responsibility we undertake when engaging in mathematics.

    The important question now becomes whether we can uphold the distinction between settled purposes of particular language-games and the purposes of individual language users, for example whether we can distinguish between the mathematical and the professional responsibility facing our accountant. The very idea of a purpose seems however, to make it impossible to uphold such a distinction. Our individual purpose in engaging in particular instances of language use determines what is done in that use in a way that makes it impossible to uphold the idea of a general and independent purpose of separate language-games. This means that we assume responsibility not only of our particular uses of language, but of entire language-games, and maybe even of language as such.


    1. Baz, Avner (2003), “On When Words are Called for: Cavell, McDowell, and the Wording of the World”, Inquiry 46:4; 473-500.
    2. Crary, A. (2007), Beyond Moral Judgment, Cambridge Mass.: Harvard University Press
    3. Diamond, C. (2000), ‘Ethics, Imagination and the Attractions of Wittgenstein's Tractatus’ in A. Crary and R. Read (eds.): The New Wittgenstein, London: Routledge.
    4. Smith, M. (1994), The Moral Problem, Oxford: Blackwell
    5. Wittgenstein, Ludwig (1982), Last Writings on the Philosophy of Psychology I, Oxford: Blackwell 1982. [LWPP I]
    6. Wittgenstein, Ludwig (1992), Last Writings on the Philosophy of Psychology II, Oxford: Blackwell. [LWPP II]
    7. Wittgenstein, Ludwig (1953/2001), Philosophical Investigations, translated by G. E. M. Anscombe, Oxford: Blackwell. [PI]
    Avner Baz poses a similar concern against the conception of language found in McDowell’s writings, see Baz 2003.
    Anne-Marie S. Christensen. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.