what the private language argument is and isn't

We still, 71 years after the Philosophical Investigations was published, find ourselves surrounded by versions of its 'private language argument' (§§243ff) which are both hopeless arguments and hopeless interpretations. To put this right, I've set out below how the principal hopeless PLA tends to go, and then present how I think the main argument in those sections actually does go. (Arguing the interpretative case isn't my task here. But perhaps the interpretative principle of charity alone should take us some of the way there...)

LW annoyed at the bad PLAs 
Hopeless PLA

To provide the stage-setting for this version of the PLA, let's start by elaborating and deploying a philosophical conception of mind as 'inner'. To construe mind as inner in this respect is to hold to the following two claims. We can i) usefully be said to know, perhaps through something we'll call 'introspection', the 'goings on' in our own conscious minds (we know what we at least seem to see; we know the sensations we feel; we know what thoughts we have; etc). By contrast we don't really know what's going on in others' minds. And ii) the relationship between our ('outer') expressive, behavioural, linguistic lives and our ('inner') mental lives is but contingent. That's to say, it doesn't enter into our very concept of (say) "pain" that someone in pain will be disposed to act in certain ways (squeal, recoil, etc).

Now, with that conception in place, without in any way questioning it, imagine the following. I want to develop a new concept for a new detail within my inner life that has struck me. I inwardly focus on, say, a new sensation, and at the same time pronounce "this, and what is just like it, shall henceforth be called an "S"".

Wittgenstein now comes along and insists "but you haven't thereby given "S" a meaning". ... Why? ... Well, this inner baptism has none of the necessary consequences of genuine definition, and in any case lacks the requisite contextual character to support the making of genuine definitions. ... But, erm, why think this? ... Well, if it were a genuine definition then there would later be such a thing as using it right or wrong. ... Well, ok, and why can't I do that here? ... It's because there's no way of verifying whether that later use is correct. (We assume the earlier sensation has died away so it's not like checking that the sensation in your right knee is the same as the S in your left.) Perhaps you misremember, etc. And, like I also said, the requisite context is also missing: we normally only get going with ostensive definition once we've got a whole bunch of pre-existing ways of relating to the item in question. It's these ways, and not any momentary act of semantic fiat, that gives the above "and what is just like it" any content; without them we are without a way of determining just what is here at issue, namely what shall count as "just like it". Result: we must give up this philosophical conception of mind as inner.

I think it shouldn't be too hard to dispose of these arguments! The first imports an implausibly verificationist premise. Why should the fact that I can't verify that I've correctly used "S" mean that my use is meaningless? The second can just be denied. Ok, sure, there's often a bunch of stage setting in the language, but given how nicely everything is laid out before me in the internal world at the time of baptising S with "S", we can just dispense with it here. And there shall in any case always be an issue, in any definition at all, of the scope of proper judgements of what is and what isn't sufficiently and relevantly similar to the initially ostended item.

Result: 'the private language argument' fails. We can after all hold onto our conception of mind as inner. And anyway, abandoning it looks like abandoning ourselves to behaviourism. That's to say, to acknowledge a more than contingent relationship between S and the behaviour S evinces (for a pain sensation, say: my avoidance, my wince, my yelp, etc.) is to collapse the having of S into these behaviours. And that seems to leave us with no inner life at all!

Plausible PLA

See: he's a little happier now

Now needless to say I don't think any of that is Wittgenstein's argument. Is there a better one to be had though? Well, here's what I think he's really saying:

Let's begin by accepting the following two claims about the inner. One: it's 1) a fact about any genuine sensation discourse that there's no such thing as me being wrong that I've got a toothache. It can't seem to me that I've toothache, but actually I've got butterflies in my stomach. Misleadingly put: if it seems to me I've got toothache, then I'm right that I've got toothache. Better put: there's no such thing as going right or going wrong here. For sure, if I say "I've got toothache" when I've got toothache then I've said something true. But correctness and truth are different concepts: "I've got toothache" is not an expression of a judgement I make about how things are; it's instead a direct articulation of how things are. (It's judgements that are right or wrong, and the class of true or false statements is broader than that of those which express judgements.) In saying it I don't voice my belief that I've got toothache; I voice the toothache itself. (If we can't directly voice, express, sensations etc. without expressing our beliefs about them - if every voicing is the expression of a judgement about how we find things with ourselves - then how come I can simply voice, express, these beliefs? Or am I (cue the infinite regress...) voicing, expressing, my belief that I have a belief ... that I have a toothache?) Unless, of course, I'm pretending - in which case I'm saying something false (but still not expressing a false belief that I've got toothache; I don't believe it at all!) Two: we will also temporarily accept the second of the above-claims from the 'hopeless PLA': that ii) inner and outer are only contingently related, so that we shan't get to the meaning of our mind discourse through thinking about a minded being's expressive behaviour.

Now, to return to the 1) first of these two latter claims about the inner: we might characterise this as: subjectivity repels normativity. (It thwarts the establishment or maintenance of that which to be properly itself must be truly characterisable in terms such as 'correct' and 'incorrect'.) If it were the case that I could be wrong that I've got a certain S, then 'S' can't be the name of a sensation. (The private linguist might deny this, of course, but then they've got even bigger problems with getting their own topic properly into view than the anti-private language argument can help with.) But the private linguist is committed to setting up a normative practice - since to give a meaning to 'S' is to make for the possibility of correct and incorrect uses of the term. It's here, then, that we have our problem. (Much of what's contained in the later 'private language argument' paragraphs of the Investigations is in fact an ongoing tussle between a desperate interlocutor who somehow hopes that, despite what at root he acknowledges to be the subjectivity (ie constitutive lack of normativity) of the inner, a fantasy of normativity can nevertheless make for at least a kind of normativity in his discourse about the inner. Wittgenstein's thankless task is to patiently bring him back down to earth.)

Wittgenstein's own solution is to give up the private linguist's conception of mind as 'inner'. (This has literally nothing to do with now relying on the judgement of others as well as one's own; we don't get to normativity by piling up collectives of subjectivities.) First, I'm not usefully said to know that I'm in pain (when I am in pain), which is to say: this really is no epistemic achievement of mine. At most we have here someone insisting they do know how to use the word 'pain', or someone fending off a too-knowing, too-intrusive, interlocutor who rides roughshod over their first person authority. And my irreproachability regarding my avowals of my sensations has nothing to do with their manifest apparentness in something (wrongly) called 'introspection'. My irreproachability here isn't to do with my being definitely correct - we are not, remember, in a domain in which talk of correctness or incorrectness finds a place. It's rather to do with the fact that, since I'm not here even in the judgement business, my avowals of my sensations, so long as they truly are that and not dissimulations, of their nature give voice to my sensations themselves. Second, the relation between sensations and sensation behaviour (avoiding that which hurts me, wincing and groaning when hurt, etc) is not a contingent one. To be disposed to react thus is part and parcel of what it is to experience sensations like this.  

But how, then, are we to avoid a collapse into behaviourism? Well, this really is the beauty of Wittgenstein's positive reconceptualisation of what makes for an actual 'inner' life. For this conception not only involves the above-articulated recapture of the immanence of sensation in sensation's declaration, thereby preventing the collapse into an alienated conception of sensation declarations as voicings of judgements of what we notice within. But it also enables us to have our cake and eat it when it comes to grasping both i) the genuine immanence of mind in action, sensation in expression, etc., and ii) the possibility of having sensations without reacting as expected. And in truth it's just this conjoint necessity-and-flexibility which gives our central concepts of mind their whole character. So, Geoff, you know, has an ongoing and severe leg pain. And yet his seemingly confident stride is as before, he's no more inclined to rest the leg, he doesn't get it seen to, and so on. Is this possible? Is it straightforwardly intelligible? The behaviourist will say it's impossible and unintelligible; the pundit of the (corrupt, above-articulated) conception of the inner ought to find it possible and straightforwardly intelligible; but my enlightened reader will I hope find it possible yet not straightforwardly intelligible. For, I suggest, it only becomes truly intelligible to us when we see that the disposition to act in the ways Geoff declines to act must be overridden. (Mary, it turns out, has promised him £5000 if he ignores and feigns the absence of the pain.) In fact it's in this way that we regain for ourselves a non-corrupt conception of the inner: the actual inner of which we ordinarily, pre-philosophically, speak is the domain of our secret, unexpressed, conscious and unconscious and semiconscious, longing and shame, hope and dread.


  1. One of the most illuminating treatments of the private language argument I have read, and certainly unbeatable on a per word basis! Thank you!


Post a Comment

Comment here!

Popular Posts