The problem of dogmatism

Oskari Kuusela on why Wittgenstein rejected theories

A distinctive feature of Wittgenstein’s philosophy is his rejection of philosophical theses and theories. Instead he comprehends philosophy as an activity of clarification. How he understands the contrast between this activity and philosophical theorising, however, is not immediately obvious and constitutes a disputed topic among his readers. Apparently symptomatic of this unclarity is that many of Wittgenstein’s interpreters in fact attribute various philosophical theories to him either explicitly or implicitly, against their own self-understanding. Either way, this constitutes a problem.

To attribute theories to Wittgenstein is to characterise his work as inconsistent, as containing a contradiction between his methodological statements about philosophy and his actual philosophical practice. Beyond scholarly concerns, to attribute theories to Wittgenstein is to miss out on the possible benefits of rethinking the nature of philosophy with him. More specifically, he claims to have found a strategy for avoiding dogmatism in philosophy, a problem he sees as intimately connected with philosophical theories. The problem of dogmatism thus understood might also be seen as one central reason why philosophy remains enmeshed in dispute, and doubts persist about its value.

Part of the difficulty of understanding what exact purpose Wittgenstein’s rejection of theories and theses serves is that he doesn’t explain as clearly as one could hope for in his published work what he means by philosophical theories or theses. Thus, for example, his rejection of theorising has been taken to mean that one shouldn’t hold any positive views about the objects of philosophical investigation. For many – presumably, including those who attribute theories to him – this would mark the end of philosophy, rather than a new beginning.

On the other hand, given that this interpretation is rather straightforwardly based on the assumption that the only way to have a positive view about something in philosophy is to have a theory, one might be suspicious about it as a way to understand a suggestion about how to philosophise without theories. And indeed, as a closer inspection of Wittgenstein’s manuscripts and typescripts reveals, by philosophical theories or theses he means something quite specific. What is at stake then, arguably, is not a rejection of philosophy altogether. Rather, Wittgenstein is to be seen as responding to certain well-specified problems pertaining to the practice of philosophy, and as suggesting certain methodological adjustments as their resolution.

What Wittgenstein understands by the traditional conception of philosophical theories or theses is that they are statements about essences. Customarily, a thesis of essence is understood as a statement about a necessary characteristic of an object of investigation, such a characteristic being one anything must have in order to qualify as having a particular essence or as falling under a particular concept. Accordingly, a thesis of essence, if correct, holds universally for all objects that share an essence or fall under a relevant concept. For example, if it is of the essence of humans that they are rational beings, then in order for a being to count as human (in the full sense of the word) he or she must be rational. Correspondingly, Wittgenstein characterises a theory of the essence of wishing as one “which would have to explain every single case of wishing”.

Although there might be other types of statement that can be called “philosophical theses”, theses of essence in this sense are representative, to a great extent, of what has been understood by a thesis in the philosophical tradition. (They capture by and large what might be called “metaphysical philosophy”.) It is problems pertaining to theses or theories of this kind that Wittgenstein is concerned to address.

The dogmatism of theories and theses of the above kind might be characterised in the following way. In making a statement about what all cases possessing a particular essence or falling under a concept must be, a philosopher runs the risk of doing injustice to the actual manifoldness of cases that fall under the concept or are to be identified as possessing a certain essence. More concretely, given that a philosophical analysis of something can (as a matter of fact) be based on an examination of a limited number of examples only, in making her claim about what all cases must be the philosopher is at risk of coming to confer the characteristics of the examples she has actually examined onto the rest of the cases. Thus, however apt her characterisation of those actually studied cases is, when extended to cover every possible case that falls under the relevant concept, the characterisation is in danger of becoming a prejudice and a dogma that prevents her from seeing the objects of investigation as they really are.

Importantly, the difficulty is not merely a psychological one, pertaining to philosophers’ powers to convince themselves about the correctness of their own theories. Rather it is, so to speak, a structural one, concerning the kind of statement a philosopher is supposed to make. The danger of doing injustice to some cases here has everything to do with the philosopher aspiring to make a statement that applies to all cases without exception. It is this aspiration that makes it so difficult for her to acknowledge cases that, although they don’t fit the thesis, should be recognised as falling under the concept in question. For, from the traditional point of view, to acknowledge such exceptions is, basically, to accept that the theory has been refuted by counter examples (requiring revision, at least), or to admit that (as it stands) the theory has mere empirical validity, capturing some instances but not their common essence.

Accordingly, the worry about the dogmatism of philosophical statements is not that the philosopher might be generalising carelessly – and that she should base her assertions on a more representative sample of cases as scientists are expected to do. This would be the wrong kind of corrective move because empirical generalisations cannot, as a matter of principle, support the kind of universal, exception-less statements that theses of essence are.

But it would also not be correct to take the critique to be that there is something inherently wrong with philosophers analysing individual cases. What is problematic is the combination of such analyses with a particular assumption about the unity of concepts, according to which conceptual unity depends on the presence of certain features necessarily shared by all instances falling under a concept.

Granted this assumption, an analysis of a single case could indeed reveal something that holds necessarily of all cases falling under a concept. For, if all cases must share the same essential features, then, by revealing the essence of one case one has revealed the essence of them all.

Wittgenstein, however, questions this assumption about conceptual unity. The unity of a concept, according to him, need not be based on the presence of a feature/features common to all cases falling under it. The problem with philosophical theses then is that they take for granted a problematic assumption about concepts or essential features. This is the source of the problem of dogmatism.

Consider Wittgenstein’s famous remark according to which “the meaning of a word is its use in the language”. On a widely held interpretation, the remark states a necessary condition for the meaningfulness of words. Wittgenstein’s view, it is said, is that the use of a word in a language is constitutive of its meaning and determines what meaning a word has, whereby by “use” one usually understands an established, rule-governed use. A word then has a meaning only insofar as it has such a rule-governed use in a language, or its employment is parasitic on such uses.

Thus understood, Wittgenstein’s remark about meaning emerges as a statement about a necessity pertaining to the use of language. It aims to capture a rule of language that the use of the word “meaning” is claimed to be governed by. But to state that language use conforms to the rule “the meaning of a word is its use” is to claim that all instances of meaningful uses of words fall under the concept of meaning as defined by this rule, i.e. that there are no meaningful uses of words that aren’t rule-governed or parasitic on rule-governed uses. But, problematically, if one accepts the earlier characterisation of philosophical theories or theses, Wittgenstein’s statement about meaning clearly qualifies as a philosophical thesis. To state that having an established rule-governed use is necessary for the meaningfulness of words is to make a statement about what all the cases falling under the concept of word-meaning must be. Here the problem of dogmatism arises again.

Ultimately, Wittgenstein’s rejection of philosophical theories and theses is to be understood as part and parcel of a methodological innovation that consists of rethinking the role of philosophical statements, and is meant to release us from the problem of dogmatism. This methodological adjustment can be explained by reference to his method of stating rules for language use with the purpose of making it perspicuous. Philosophical definitions can be taken as examples of such rules. Notably, however, this method of the tabulation of rules is only one method of philosophical clarification for Wittgenstein, not the method.

A key insight behind this method is that, instead of constituting the object of the philosopher’s investigation about which statements are made, rules are to be seen as the philosopher’s means or mode of presenting language use. More precisely, according to Wittgenstein, rules (such as the one about meaning as use) articulate conceptions, pictures or models that can be employed in the clarification of language use. But such models are not deployed with the purpose of making any claims about how language is actually used, asserting that language use conforms to the rule exactly and in all respects.

Instead, the models are employed in the capacity of what Wittgenstein calls an “object of comparison”. This is to use a model to clarify specific aspects of the use of expressions by comparing the actual use of language with the model and paying attention to both similarities and differences between the model and the reality described (i.e. actual language use). Thus, no claim is made that all cases falling under a concept or identified as having a particular essence must be as the model presents them. Rather than being treated as a fact about the cases falling under a concept that they all necessarily possess a certain characteristic, it is a fact about the philosopher’s mode of presentation that it presents such and such as a necessary characteristic of those cases.

Here it is important that the goal of the activity of clarification, as Wittgenstein conceives it, is to resolve certain particular philosophical problems. This means that the correctness of philosophical models is relative to particular elucidatory tasks.

For example, the conception of meaning as use seems able to resolve certain philosophical problems relating to the concept of meaning. To name one, it does away with the need to postulate abstract objects that numerals stand for, which philosophers are sometimes led to postulate. But that doesn’t mean that one should, therefore, always regard meaning as something determined by the use of an expression. Arguably, there are aspects of the meaningful use of words which don’t fit the model of meaning as use, and Wittgenstein too recognises this.

For instance, the model doesn’t take into account the relevance of tone for meaning, although the way something is said can affect meaning in various ways. Notably, tone can convey meaning also in the absence of any linguistic conventions, as illustrated by the ways we use sound to convey emotion. If so, the model of meaning as use doesn’t capture the actual use of the word “meaning” as it is. Crucially, however, employing the model as an object of comparison one is free to use other models to describe such different cases. An object of comparison does not commit one to any exclusive theses about meaning.

A benefit of Wittgenstein’s method then appears to be that it leads to a certain increase in the flexibility of philosophical thought. Philosophers are released from a particular kind of asceticism which is a consequence of the requirement of their having to make the kind of sweeping statements that traditional theses of essence are. More specifically, the employment of rules as objects of comparison makes possible what one might call multidimensional grammatical descriptions, i.e. simultaneous descriptions of the uses of language from various points of view. Aspects of language that can be captured in philosophical theses only on pain of contradiction can be captured by using definitions as objects of comparison, because the employment of one definition as an object of comparison does not exclude the use of another one in the same capacity, unlike two different philosophical theses about a subject matter. An example is Wittgenstein’s characterization of language as an instrument designed for a particular (external) purpose and his characterization of its rules as arbitrary, i.e. not determined by any independently given purpose. Comprehended as theses these characterisations contradict and exclude each other.

Wittgenstein’s rejection of philosophical theories doesn’t therefore mean that he, or whoever adopts his method, couldn’t have any positive views about the objects of philosophical investigation. It merely means not presenting those views in a dogmatic manner, as theses that all relevant cases must fit. Importantly, there is nothing in principle that prevents one from making the kind of novel use outlined here of traditional philosophical theories too. This might then enable one to avoid problems to which those theories may lead when put forward as exclusive accounts of relevant matters. In a way, Wittgenstein’s approach, therefore, allows one not to take sides in philosophical disputes and to take onboard whatever might be correct in the traditional theories.

In this sense his philosophy is not anti-metaphysical, implying a rejection of what has been said in the metaphysical tradition of philosophy. His approach is, nevertheless, non-metaphysical in the sense that it constitutes a rejection of traditional theorising about essences and the essential features of things. To adopt his approach is, ultimately, to understand the role and status of philosophical statements in a new way.

Oskari Kuusela is lecturer in pthhilosophy ae t University of East Anglia and author of The Struggle Against Dogmatism: Wittgenstein and the Concept of Philosophy, Oskari Kuusela (Harvard University Press)

Leave a comment ?


  1. Readers of this article may find in my essay on “The Wittgenstein Enigma” an unaccustomed approach to Wittgenstein’s thought. The essay is on my blog:
    and also appears in my latest book, The Sphinx and the Phoenix (2009).

  2. An Article about Wittgenstein « Issues in Contemporary Philosophy - pingback on November 11, 2009 at 4:42 pm
  3. This is an excellent and clear introduction to what is likely to be the most radical and striking aspect of the later Wittgenstein’s philosophy. Also, it seems to me a rather promising approach to reformulate the whole issue in the terminology of ‘dogmatism’.

  4. The danger, it seems to me, is to interpret Wittgenstein’s rejection of dogmatism as a sort of aspectival relativism (as Gordon Baker does). Commenting on the famous meaning is use passage in the Investigations, Hutchinson and Read seem to head in that direction:

    Indeed, we are not, it seems, obliged to define meaning in terms of use even in a “large class of cases,” we merely “can” for our purposes. (Phil Hutchinson and Rupert Read, “Toward a Perspicuous Presentation of “Perspicuous Presentation'”, p. 141)

    This is a bit like infering from the fact that not all games involve winning that we are not obliged to define a large class of games as involving winning.

  5. @N.N.: I think you’re quite right. Of course we’re not “obliged” to define meaning in terms of use but we’d better do so if we want to explain to someone what a word means.
    Also, ‘therapy’ seems to me somewhat misconstrued in the exteme therapeutical usage. Stress the personal aspect as much as you like, if you look at actual therapies, you’ll see how much they actually have in common. That is to say, if I’m convinced (or “liberated”) by a particular philosophical therapy (e.g. the ‘private language’ treatment) than there’s no reason to suppose that it won’t work for you too. In fact, it makes a lot of sense to expect that it will simply because we share our language.
    However, I didn’t notice Kuusela proposing a kind of “aspectival relativism”. I haven’t read the (whole) book though.

  6. N.N.: Your criticism of me/Hutchinson is not warranted. What on Earth would it be to ‘define a large class of games as involving winning’? It is the strangest thing I ever heard of. Either you are defining ‘game’, or you are not.
    And if you are, the question arises of the purpose for which you are offering the definition.
    That is all that Phil and I, following Baker, following Wittgenstein, suggest. I have yet to hear an effective argument against the suggestion.

  7. Hello. Just a quick comment. Presumably it’s important to be clear what’s relativism & what’s not in this connection. Normally, I take it, the term is used in connection with truth, so that what is true is e.g. said to be relative to some conceptual framework, and so on. Consequently you get a situation of the impossibility of communication between frameworks, or have to accept that both p & -p can be true at the same time. Which is a problem. And so on. A central point of mine in the Dogmatism book is that it’s misleading to think of e.g. W’s grammatical statements as true/false statements to begin with. Rather they articulate modes of presentation or ways of looking at things (models to be used as objects of comparison). Now the important point is that one mode of presentation doesn’t exclude another like the truth of p excludes the truth of -p (unless you’re in the relativist… erm… mess). So, you don’t end up in the same problematic situation as with relativism about truth. (I wouldn’t call this relativism at all to keep things clearer. To be honest I’m not sure what ‘relativism of aspects’ would mean, i.e. what it would be to be relativist about aspects.) But is it then that we can just come up with modes of presentation or ways of looking till cows come home & philosophy becomes a pointless game of comparison (since anything can be compared with anything)? Well, there the point is that very few comparisons or ways of looking at things can solve our philosophical problems for us, i.e. make things perspicuous & help us see how things work. So, it doesn’t follow that philosophy is watered down to some uninteresting game of coining comparisons.

  8. Sorry. That’s a massive comment. I’m a totally inexperienced blogger.

  9. Dr. Read,

    The analogy with winning was poorly chosen. Rather than explain it, I would rather drop it.

    My objection to your claim that we are not (according to Wittgenstein) obliged to define meaning in terms of use is that it (in my opinion) misunderstands Wittgenstein’s use of ‘can.’

    Consider the following passage from the Grammar:

    “I want to say the place of a word in grammar is its meaning.

    But I might also say: the meaning of a word is what the explanation of its meaning explains.

    The explanation of the meaning explains the use of the word.

    The use of a word in the language is its meaning.

    Grammar describes the use of words in the language.

    So it has somewhat the same relation to the language as the description of a game, the rules of a game, have to the game.

    Meaning, in our sense, is embodied in the explanation of meaning.” (pp.59-60)

    Wittgenstein “wants to say” that the meaning of a word is its place in grammar. But he “might” also say that the meaning of a word is what an explanation of its meaning explains. By his use of “might” is Wittgenstein suggesting that the meaning of a word might not be what the explanation of meaning explains? Apparently not, because he says meaning “in our sense” is what the explanation of meaning explains. What Wittgenstein seems to be doing is presenting different definitions of ‘meaning’ that amount to the same. An explanation of meaning explains the use of a word. But the use of a word ‘situates’ it in grammatical space. Thus, we can use either description because they are equivalent.

    It seems to me that Wittgenstein’s us of “can” in the Investigations passage is better understood along these lines. In a large number of cases, we can say that meaning is use. In the same cases, we could also say that meaning is location in grammatical space. But even if we say the latter, the former is still correct.

  10. Dr. Kuusela,

    I borrow this sense of ‘relativism’ from Gordon Baker:

    There seems no good reason to think (or even to think that Wittgenstein thought) that the subservience of descriptions to forms of description, and hence the ‘relativism’ of acknowledging the possibility of different forms of description, comes to an end at the frontiers of empirical discourse! (“Philosophical Investigations section 122: neglected aspects,” pp. 55-6)

    “In claiming the importance of perspicuous representations for ‘our form of representation’, Wittgenstein seems to have been explicitly subscribing to a form of relativism…” (Ibid., p. 59)

    As I understand Baker (see my post here), there is no such thing as a correct description of the grammar of an expression (e.g., ‘meaning,’ ‘ostensive definition’). One description is as good as another, and we can adopt or discard it as we please.

  11. your magazine is excclent

  12. @N.N. The bit left out from the quote from Philosophical Grammar:
    “If, on the other hand, by the word “meaning” we mean a characteristic sensation connected with the use of the word then the relation between the explanation of a word and its meaning is rather that of cause and effect”

    Seems to suggest one might take meaning as Wittgenstein does. The contrast, however, is with the classical Lockean-like views of meaning. My sense is that he is suggesting the two views as the only possibilities. What he does throughout the Investigations is present his & the classical view and then show some philosophical confusion occurs in taking the latter as a genuine possibility. I would read the passage that way, as a setup, rather then as suggesting any view of the meaning of ‘meaning’ yields the same conclusion.

  13. The problem of dogmatism « Fire EXIT - pingback on January 2, 2010 at 8:23 am
  14. The problem of dogmatism at - pingback on January 2, 2010 at 9:14 pm
  15. Couldn’t the same thing be accomplished through the use of stipulative definitions? That is, “By T I mean any object possessing characteristics x, y, z.” Doing so avoids the problem of essentialism while providing a normative “object of comparison” for dealing with Wittgensteinian radial categories.

    I think it is the gambit that Paul Grice chose in his later years.

Leave a Reply

Trackbacks and Pingbacks: