The one thing I wanted was a careful account of hinges on her view. Unlike the Norman Malcolm paper I’ve just re-read, Coliva plays down the merely animal certainty of hinges and plays up the idea that hinges are propositions, albeit set against a background of merely animal certainty. So she contrasts with Daniele Moyal Sharrock’s non-propositional and ineffabilist view, too. She connects both these rejected views at the start of the book thus:
Moreover, as to our attitude with respect to them, if we held the ineffabilist conception of hinges, whereby, failing to be propositions, they could neither be said, nor made the object of any propositional attitude, our certainty with respect to them would have to be thought of as non- propositional, non-conceptual and therefore of a merely ‘animal’ kind. (ibid: 9)
The fullest account of her own position starts from this rejection. Coliva takes hinges to be propositions, to be or to contain conceptual contents. A key claim is that this does not require that we stand in an epistemic relation to them.
Finally, it remains to characterize
our relationship to these propositions, whose role is that of norms. As Moyal-Sharrock
and other interpreters have emphasized, it is thoroughly non epistemic and
thus unlike belief, knowledge or even psychological certainty, for
Wittgenstein. What they have failed to notice, in my view, however, is that
this doesn’t entail, nor supports the view that it is animal or instinctual,
and hence utterly non-propositional. After all, it has propositions – albeit
of a normative kind – as its objects. So, what kind of propositional attitude may
it be? I think an analogy would be useful here. Think of our rules of conduct,
for example, we ought to stop at traffic lights when red. In fact we can and do
say that we accept such a norm. That is to say, we behave in accord with
it and would either reproach those who don’t, or accept to be reproached
ourselves if we didn’t. Furthermore, we teach our children to obey it. This pragmatic
sense of accepting a proposition whose nature is normative, to be
contrasted with an epistemic sense of accepting empirical propositions for
which we don’t have (enough) evidence, and with respect to which it would make
utterly no sense to talk of epistemic evidence as it has norms as its
objects, is, I submit, what Wittgenstein is alluding to when he talks of
certainty with respect to hinges qua hinge-propositions. Now, the
interesting aspect of our acceptance of hinges is that while being a pro-attitude, it is itself constitutive of epistemic rationality, and not
merely a pragmatist acceptance due to an evaluation of its expected
practical utility. (ibid: 174)
Yet, this pragmatic acceptance of rules tallies with many of Wittgenstein’s remarks which allude to our certainty as displayed in our way of acting. In particular, to acknowledge this sense of ‘accepting’ allows us not to go all the way down the path of mere ‘animal certainty’ and is compatible with saying that we accept, often implicitly and as a result of our upbringing, norms both of language and of evidential significance, and ‘instinctively’ behave in conformity with them, in ways which show no doubt. It must be stressed, moreover, that such an instinctive behaviour is not the one of the animal, but that of the human animal. That is to say, of the creature who has been drilled and taught to use language and to take part in our various epistemic practices and is thus already fully conceptual and a competent epistemic agent. If, then, there is still room for instinct here, it is one concerning our so- called second nature, not our merely first and animal one. (ibid: 174-5)
So, ‘animal certainty’ is really not the focus of On Certainty, though it is a precondition for being drilled to take part in our various language games. What On Certainty centres on are rather those propositions which, as a result of our upbringing within a community that shares a language and a form of life, we can reflectively identify as being exempt from verification and control as well as from doubt. This makes them play the role of norms, either of language or of evidential significance. Moreover, with respect to them, we do practically (as opposed to epistemically) accept them – that is to say, we behave in accord with them, in ways which know no doubt – where our doing so is, in its turn, constitutive of epistemic rationality. (ibid: 175)
So the picture is one of a background of animal certainty upon which, through education and development of a McDowellian second nature, conceptually articulated propositional certainties can be built. These are thus the object of propositional attitudes though not epistemic propositional attitudes. Various questions arise such as the nature of the attitude and how the hinges’ propositionality is realised.
With respect to the former. in the passages just quoted, Coliva sketches an initial analogy between hinges and pragmatic acceptance. But she adds that the relevant attitude to hinges is not merely a pragmatist acceptance - due to an evaluation of its expected practical utility - but is rather constitutive of epistemic rationality.
So what is the limited positive aspect of the analogy? We behave in accord with pragmatic norms and also with hinges and would reproach those who don’t. But the hinges also play a non-contingent role in the constitution of rationality. What is this role? I’m not sure. But it seems to have something to do with using – if that is the right verb – the propositional hinges for normative rather than descriptive purposes. We instruct people using hinges and we accept them as the normative standards that establish the epistemic language game, I think.
By contrast, by defending the view that hinges are propositions, though normative rather than descriptive in nature, and that our certainty with respect to them is a kind of acceptance which displays itself in our acting in accord with them, we make both certainties and certainty with respect to them effable. Of course, I think Moyal-Sharrock is right to claim that when we utter hinges qua hinges we mostly do so with heuristic purposes in mind of the kind she correctly identifies. Yet, by acknowledging their propositionality hinges become indeed sayable as such – that is to say, as norms. (ibid: 177)
The second comment – about when we utter hinges qua hinges – connects a heuristic purpose, which I think Coliva accepts, with the propositionality and hence sayability of hinges as such. Utterances can have propositional contents.
Earlier she expresses this thus:
On the contrary, I hold the view that while failing at bipolarity, they are still propositions, albeit with a normative function, rather than a descriptive one, and that we do indeed express them on various occasions: either to teach them to someone who ignored them or to remind someone of them were they to violate them, as philosophers such as Moore and a sceptic do. Contrary to other framework interpreters, most notably Moyal-Sharrock, I do think that at least the former context in which hinges are actually said is a genuine language game, by Wittgensteinian lights. (ibid: 10)
So the normative use is connected to heuristic purposes and is part of a genuine language game, not a misfiring piece of nonsense. That gives us one place for hinges: in instructing people how to play the game of giving and asking for reasons, which is based on certainties. The proposition is the content of the instructional utterance. But what is our attitude to them when they take a role in propositional attitudes which display themselves in our acting in accord with them? It seems: pragmatic acceptance. But that seems odd. It seems prima facie odd to model the certainty of hinges on mere pragmatic acceptance.
Another line of thought may help. Coliva contrasts her view with Moore’s.
when taken in those very circumstances, Moore’s propositions don’t have an empirical role, but a normative one. In the twofold sense that they are both linguistically paradigmatic judgements – that is to say, judgements that must be agreed on if the meaning of the words in them is to be determined – and epistemically paradigmatic ones. That is to say, judgements that can’t be called into question on the basis of contrary evidence, for they must stay put – however in context that might be – in order for a subject to be in a position rationally to gather any kind of evidence for other, genuinely empirical propositions. (ibid: 80)
The first part of this seems to characterise an instructive use of an utterance rather than a propositional attitude. (There is a second aspect reflected by the phrase epistemically paradigmatic to which I'll return.) But the instructive use looks to be meta-linguistic. It’s something like: call this! a ‘hand’, linking mention of a word type and a demonstrative. The related propositional attitude would seem to be an attitude to the connection between the look of a hand and the word ‘hand’. But now, surely the right attitude is knowledge? I know the meaning of the word ‘hand’. I know that this is called a ‘hand’. But perhaps this is not enough because pragmatically accepting a norm here suggests not only knowing the standrd meaning or use of ‘hand’ in English but deciding to bind oneself to that norm. I'll return to this shortly.
Coliva, however, suggests that things run deeper than either knowledge of meaning or adoption of a norm by quoting a famous paragraph from the Investigations.
If language is to be a means of communication there must be agreement not only in definitions but also (queer as this may sound) in judgments. This seems to abolish logic, but does not do so. – It is one thing to describe methods of measurement, and another to obtain and state results of measurement. But what we call ‘measuring’ is partly determined by a certain constancy in results of measurement. (PI §242)
I think that this gives content the idea that judgements about instances of hands play a role in underpinning the meta-linguistic point. (I also think that this links to the saying-showing distinction.) For Coliva, the fact that we do just judge in accord with the normative standard governing ‘hand’ is part of the logic of judgement.
To judge thus-and-so, in certain circumstances, is said, in OC, to be part of ‘our method of doubt and enquiry’. Now, everything which is part of method does in fact belong to ‘grammar’ and ‘logic’ too, in the extended, not merely linguistic sense that those terms came to have for Wittgenstein at the time of On Certainty. To judge that there is a hand in the circumstances of Moore’s proof therefore belongs to the logic of our epistemic practices because it is what must stand fast if we want to test other things – for example, the reliability of our senses – which we do in turn need to trust in order to go about forming specific empirical judgements in circumstances where it is an open question whether what we see in front of us is really a hand. So, that judgement is itself part of logic and therefore comes to have a normative role, rather than a genuinely empirical one. (ibid: 81-2)
I imagine a case in which one uses one’s hand to estimate the size of a spider – “It was as big as this hand!” – taking for granted the identity of the hand. But now I’m unsure about the nature of the propositional attitude. The hand, by contrast with the spider, is not an object of scrutiny. But still it seems bizarre to think of our attitude as being analogous to a pragmatic attitude, such as agreeing to stop at red traffic lights. We might take that towards the metalinguistic fact that we call hands ‘hands’ in English. Thus, I think we know and additionally and distinctly endorse as a practice for ourselves for successful communicative intent. As well as knowing that ‘hand’ means hand in English, we might also agree to be bound by the norm of using ‘hand’ in just this way for pragmatic reasons. We can also make sense of adopting a different policy.
But what of the unreflective, animal-based but second nature, attitude towards this thing on the end of my arm being a hand? Although I may adopt a policy of speaking in accord with English norms, it does not seem that the prior recognition of my hand as a hand - or recognising it as the same thing over time - is a matter of policy. I don’t just ‘accept’ that. No other policy seems available.
Instead it seems to be forced on one by an appreciation of one’s predicament. Other things may then follow such as the use of hands as comparison with spiders, for example, or the decision to follow regular English usage in their labelling. But we seem to be passive in the face of acknowledging what the world shows us. Why is it not simply knowledge?