Postscript on disjunctivism and induction

A conversation with Gloria and actually teaching a session on Hugh Mellor on induction has prompted a couple of further thoughts on what a disjunctivist account of induction would look like.

1) What does a disjunctivist spin on McDowell’s remarks add to their intended use as a deconstruction of Hume’s starting point for the problem of induction?

As we would expect of a therapeutic McDowell, he himself does not aim to solve Hume’s problem but dissolve it. Hume assumes a contrast between unproblematic observation and problematic induction. But, McDowell argues, such a distinction cannot be drawn because the reports of observation are always (I want to say: always already) induction presupposing.

Consider a characteristic Humean formulation of the predicament that is supposed to invite inductive skepticism:
It may, therefore, be a subject worthy of curiosity, to enquire what is the nature of that evidence, which assures us of any.. .matter of fact, beyond the present testimony of our senses....
Taking it seriously that what is in question is testimony of our senses, we must think in terms of something content-involving-something in which, say, colors figure as apparent properties of objects. A mere wash of chromatic sensation, not referred to a supposedly perceived environment, could not count as testimony of our senses. Now my fourth Sellarsian idea can be put like this: there cannot be a predicament in which one is receiving testimony from one's senses but has not yet taken any inductive steps. [McDowell 1995: 890-1]

This looks like the bare bones of a transcendental argument. Observation is possible; observation presupposes induction; so induction is possible.

But whilst that might serve to undermine the legitimacy of a scepticism which is differentially directed against induction but not observation it does not shed light on how induction is possible. One might respond to the transcendental argument that whilst one accepts that observation stands and falls with induction, nevertheless, one is still baffled by how the latter is possible or, to put it another way, how the inductive element of observation – that which transforms it from a mere wash of chromatic sensation – is possible.

Sketching a disjunctivist picture of induction helps start to address that remaining worry. Induction is possible because one can take in, experientailly, a future-implicating pattern of events. Observation is not just of temporally unrelated time-slices. (That’s what Hume seems to rule out at the start.)

2) But doesn’t the question of what we do, experientially, take in underdetermine possible future patterns? That is, are we not in the end any better off than the Humean inductive sceptic?

To make this worry vivid requires, I think, slipping back into the assumption that all that is really taken in experientially is a set of time-slices with no implications for the future. But there is a related worry which seems legitimate: Goodman’s New Riddle of Induction. On this picture, observation reports are taken to be conceptually structured in a way which has future implications. But there remains the challenge of justifying which concepts are instantiated in past observations in the face of merely future differences.

3) What kind of future-directed structure needs to be implicit in past observations for such observations to form the basis for future predictions?

My motivation for this question stems from the contrast I have in mind between Mellor’s blankly external factors and what I take to be internal in a disjunctivist approach. On Mellor’s account, past observations warrant predictions for the future in virtue of there actually being laws which link the worldly events observed in the past and the predicted future events whether or not the subject knows this. So the question as to the nature of this link is left to the world: the link is whatever it is.

By contrast, on my sketch of a disjunctivist account, the nature of the good disjunct should be ‘internal’ to the subject’s epistemic standing and provide her with reasons for her belief. If so, it can hardly be that the subject is only justified if she has a correct account of the natural laws by the standards of a future physics. But what conception would be sufficient to justify the future-directed inference? More seems to have to be loaded onto the subject in the case of induction than in the case of object-directed perceptual thought.

4) Isn’t there a difference between the role that enduring objects play in securing the future implications of past observations and that played by recurring events?

I just want to flag this worry (suggested by G.). It might seem that the very idea of the default status of an object’s continued existence over time  is more basic than that of the continuation of a merely repeated pattern of events. I am not sure that I see this as an important distinction of kind but I may just be missing something.

Some thoughts on Hugh Mellor's 'The Warrant of Induction'

Mellor’s account of the warrant of induction in his inaugural lecture of that name turns on a comparison with observation. The idea is to get clear on the warrant of observation and then use that to motivate an account of induction. For one thing, he argues, the traditional problem stems, in part, from assuming that induction should have a warrant which even observation cannot have. Since a well grounded observation trumps a well grounded induction, that must be wrong. But his account of observation is revisionary. He says that it isn’t factive. Seeing something is so does not imply that it is so.

But observations are... fallible, a fact we obscure when we use words like ‘observe’ so that by mere definition we can’t observe what isn’t there. So I will not use ‘observe’, ‘see’, ‘hear’, etc. in that misleading way. When my looking to see if there are sparrows makes me believe that there are, I shall say that I see that there are, whether there are or not. For there might not be: our senses always can give us false beliefs and sometimes do. Observation no more entails the truth of the beliefs it gives us than induction does. How then can it warrant them? [Mellor 1988]

The question here is how observation warrants the truth of the beliefs it motivates. Mellor’s answer is that causal links enable observations to warrant beliefs because, he claims, the fact that makes a belief true is what causes the belief to be held. (Given that this is supposed to be the warrant observation provides, the causal connection must (I assume) be via a suitable sensory path.)

The claim that observation isn’t factive can now be accommodated via the further – independently plausible – claim that causation isn’t deterministic. If causation merely raises the chances of its effects (ie if it doesn’t have to make the chances of its effects = 1) then thinking that one sees a sparrow can warrant a belief in sparrows by suitably raising the chances of their being sparrows, though not to 1 and other things can also cause a belief in sparrows.

This then serves as a point of departure for thinking about the warrant of the truth of beliefs produced by induction. In this case, it cannot be that the facts that cause the beliefs are the same facts that make those beliefs true because induction isn’t a kind of direct perception of the future, for example. Rather, the facts that cause the inductive belief are the inductive base: the premises not the conclusion of the inductive inference. So what warrants the truth of the resultant belief in the conclusion?

Mellor’s answer is that the warrant is the lawlike connection between the truth of the premises and the truth of the conclusion. Providing that the world is suitably lawlike, induction is warranted. Hence a quotation at the start of the lecture from Frank Ramsey ‘we are all convinced by inductive arguments, and our conviction is reasonable because the world is so constituted that inductive arguments lead on the whole to true opinions’. But the laws need not be deterministic to provide a warrant. They just need to raise the chances of effects sufficiently.

Suppose, to simplify the discussion (it’s not essential), that your inferential habits are deterministic, like your mass: this habit would always make you infer that a lecture was terse, never that it wasn’t. And suppose that every lecture has some chance of being terse. Then whenever your premise (‘this is a lecture’) is true, your conclusion (‘this is terse’) has some chance of being true. And if this chance is high enough, your prediction is warranted. [Mellor 1988]

Being warranted depends on what the laws of nature are. But providing there is a sufficient natural uniformity, induction is warranted. Now putting it like this (Mellow does not) invites a comparison with versions of the problem of induction going back to Hume in which defences of induction that invoke the uniformity of nature face a problem. Since a claim about uniformity is not a ‘relation of ideas’ it must be a matter of fact. But it isn’t a directly perceivable matter of fact and thus must be the result of an inference from what can be observed. What inference? Well induction. So if the defence of induction requires appeal to the uniformity of nature, it looks to be circular. Here’s Mellor’s summary of that worry.

So far so good - provided these warrants needn’t be self-intimating. I say the law that all tree frogs are green warrants my inferring that something is green from the fact that it’s a tree frog. But suppose I must know that I have this warrant. Then I must know this law. So I must believe it, and this belief must be warranted. But the law entails the very inference which it’s meant to warrant: tree frogs can’t all be green unless this one is. So unless my inference is warranted already, my belief in the law won’t be warranted. Thus to claim that the law is what warrants this particular application of it simply begs the question of whether it’s warranted at all. [Mellor 1988]

The solution is to part company with an assumption built into Hume’s version: that justification is self-intimating or that to be justified, one must know the justification. Mellor’s solution is an externalist denial of that. One can be warranted without knowing the warrant.

This is the stock objection to contingent solutions to the problem of induction: they beg the question. And so they would if belief-warrants had to be self-intimating. But … they don’t. The law that all tree frogs are green can warrant the habit of inference which induction will then give me, just because I needn’t know that it does. I can know by induction that a frog is green without knowing that law, just as I can know that it’s green by looking at it without knowing I’m not colour-blind. I may know the law, just as I may know that I’m not colour-blind; but I needn’t. So my saying that the law is what warrants this induction doesn’t beg the question. [Mellor 1988]

The lecture conducts some further business. Because of the nature of the appeal to indeterministic or probabilistic laws in this epistemological context, Mellor reminds us that these can be thought of as objective chances with mind-independent causes and effects. He also explains the advantage of larger sample sizes in the premises of inductions and why counter-induction is generally a less good strategy than induction in a lawlike world but that no strategy triumphs in a lawless world. But the general idea is that a possibly probabilistic but generally lawlike universe warrants a form of inference which moves between nomologically related properties.

One question this prompts is the relation of warrant and truth. In the case of observation, there is a potential for a gap which isn’t there in a disjunctive account. Since ‘seeing’ a sparrow does not entail that there is a sparrow, a warranting but merely probabilistic relation between sparrows and beliefs in sparrows can coincide with the falsity of a belief. So Mellor accepts the idea that two subjects can be equally warranted in their beliefs, one can be false and the other true but the latter subject nevertheless have knowledge. That seems odd as it seems that the fact that their belief is true is, relative to their counterpart, a matter of some luck (likely though they are to be right).

A disjunctivist like McDowell can attempt to account for the fallibility of observation by conceding that whilst that faculty as a whole is fallible, its exercise on particular occasions falls into two sub-groups: actually seeing the sparrow or merely thinking one sees the sparrow. In the good disjunct, no probability is in play: the sparrow is simply there in view. So the warrant in the good disjunct is of a quite different kind to that in the bad disjunct. (The worry now is whatever is surprising about the idea that warrant and truth cannot come apart symmetrically.)

Returning to Mellor, in one kind of case of failed induction, there seems to be a similar possibility for a false but warranted induction: the unlucky case in which a probabilistic connection fails to yield the fact that fits the inductively supported conclusion. But I wonder about a case in which, for fully deterministic reasons, one causal regularity is trumped by another. The regular spinning of the Earth may be interrupted by a supernova, eg. In that case, the sun’s previous behaviour considered alone might be thought to warrant the inductive belief that it would rise the next day but the physical system as a whole suggests that such a belief is not, in fact, warranted. The former approach would require some principles of individuation of the physical system considered independently of the subject’s grasp which seems tricky. But the latter approach would undermine the possibility, in deterministic cases, of warranted but false beliefs. Whilst it might seem the exercise of epistemic responsibility, the inference isn’t in fact warranted. This seems odd.

The appeal to me, however, of Mellor’s approach is that it sidesteps the ‘internalist’ version of the problem of induction. Not requiring that one knows the principle of uniformity of nature more generally, or the local laws more specifically, blocks the circularity which Hume stresses. So is there a way of embracing that and which maps something like a disjunctivist ‘correction’ of the warrant of observation onto induction? Not quite, I think.

Note that there are opposing tensions in calling McDowell’s disjunctivist picture of seeing ‘internalist’ and ‘externalist’. It is externalist, or perhaps more precisely not internalist, insofar as the fact that one occupies the good or the bad disjunct depends on a worldly favour which is beyond what the subject can control (hence the rejection of the interiorization of the space of reasons). But it is internalist, or rather not externalist, insofar as, if one is in the good disjunct, the seen sparrow is simply there for the subject, in no sense external to their experience (contra the highest common factor approach).

At first sight, it doesn’t seem possible to play quite this game with the inductive base. One cannot simply see in the observed past behaviour of ‘sunrise’, for example, the next day’s rise, that very event. Recall that in setting up Hume's problem, the connection between past and future regularities is a matter of fact rather than a relation of ideas. But then, within the class of matters of fact, such a regularity is not directly observable and must be extracted or inferred from what is observable (past regularities). At that point, we realise that the principle needed to do this presupposes the very thing that was supposed to be the conclusion of the inference.

With Mellor in mind, we might suggest that the observed past sunrises physically necessitate the next day’s rise. But if so, it isn’t so clear what the difference is in what is available to the subject between observing past sunrises which necessitate a future sunrise and sunrises which do not. The subsequent necessitation seems to be something external to what is observed and thus external to the subject’s grasp. So the feature of the disjunctivist account of observation which merits the description 'internal or at least not external' does not apply to Mellor's view of the law that connects past to future sunrises. Such a law is not part of what the subject can take in.

In McDowell’s discussion of the testimonial transmission of knowledge, he suggests that we speak of hearing, in another’s words, that things are thus and so. There is also here a possible line of analogy in that, for one to hear that things are thus and so, the interlocutor must be reliable and that is not something one typically ensures oneself. So their reliability is not in their utterance but is a precondition for hearing in their utterance the facts.

There is, however, also a potential disanalogy to induction here. The ‘content’ that the sun will rise tomorrow – or the sun’s rise, tomorrow – is not there in the induction base to be embraced. It seems that it must be inferred from that. This might suggest that there isn’t an equivalent idea to the mix of internal and external that characterises observation in the good disjunct. The link from past to future seems, as it is with Mellor’s account, an external addition to is observed. But the problem of making the connection external is that were it ‘blankly external’, as McDowell sometimes says, then it is hard to see how it can contribute the subject’s epistemic standing.

To get something akin to the disjunctivist account of the good disjunct for observation for the case of induction, I think we need to think along these lines. The subject who bases a prediction for the future on experience of past events must not merely take the past facts as atomic.To have a bearing on the future, they see a kind pattern or ‘habit’ in the sun’s behaviour. They see in the past behaviour of the sun its ongoing, and hence future implicating, pattern. And if the pattern really is ongoing – by contrast with a kind of bad disjunct in which the seeming pattern is not really there – then the subject does have a justification which balances internal and external.

Hume’s problem of induction is generated not just from the fact that the inference from past to future is not a ‘relation of ideas’ and hence must be a matter of fact but also because it is not a matter for direct perception and hence must be an inference. But perhaps the latter distinction does not neatly divide two aspects of the inductive base: atomic facts and subsequent inference. One sees in the facts the ongoing pattern.

McDowell’s own comments on the problem of induction are aimed rather differently. He attempts to deconstruct the starting assumption which compares induction with observation unfavourably. But I wonder whether by denying the separation of observation and induction they also sustain something like a disjunctivist view of seeing the enduring pattern in the past events:

there cannot be a predicament in which one is receiving testimony from one’s senses but has not yet taken any inductive steps. To stay with the experience of colour... colour experience’s being testimony of the senses depends on the subject’s already knowing a good deal about, for instance, the effect of different sorts of illumination on colour appearances... [McDowell  1998: 411]

So the supposed predicament of the inductive sceptic is a fiction... Hume’s formulation can seem to describe a predicament only if one does not think through the idea that its subject already has the testimony of the senses and this means that scepticism about induction can seem gripping only in combination with a straightforwardly interiorizing epistemology for perception. [McDowell 1998: 412].

PS: a few more thoughts on this are here.

McDowell, J. (1995) ‘Knowledge and the Internal’ Philosophy and Phenomenological Research 55: 877-893 reprinted in McDowell, J. (1998) Meaning knowledge and reality, Cambridge, Mass.: Harvard University Press

Mellor, H. (1988) ‘The warrant of induction’  http://www.dspace.cam.ac.uk/handle/1810/3475.

Schizophrenia & the nature of delusions: to believe or not to believe?, Ecole Normale Supérieure de Lyon

I am at a two day workshop on delusions in Lyon organised by Jean-Michael Roy (pictured) and with a focus on the doxastic versus non-doxastic distinction to which he has clearly given some considerable thought. One question raised by the relation of his conception of the general framework to individual papers by others is not so much whether the distinction can be applied but whether it is the best or most helpful distinction. I'd be interested to know, however, what might replace it if it isn't.

After a very thorough taxonomy of possible positions offered by Jean-Michel, Sean Gallagher seemed surprised and faintly disturbed to be classed as occupying a kind of doxastic position. He suggested the analogy of taking a bath: in so doing, beliefs are involved but that isn't what having a bath is mainly about. So whilst his account has implications about beliefs, he wanted to resist the doxastic label. 

In his presentation, he recapitulated the account given in his paper in the Broome and Bortolotti 2009 book which criticises both top down and bottom up accounts and sketches a more general idea of delusions as the inhabiting of a different and delusional world. This, he suggested, allowed for the description of all the standard ways in which delusions have surprising properties. But he went on, beyond that paper, to present a kind of two factor biological account, plus cultural factors to shape specific content. (The second factor being a pair of localised brain mechanisms for failing to spot cognitive errors to distinguish between healthy occasional suspension of reality whilst reading novels etc. and a more permanent delusional suspension.)

I was a little unsure about how the two themes fitted together. I guess the former picture is non-doxastic in the sense that Sean's preferred  account of access to the real world is not based on intellectual, theory-theory theorised beliefs but on a richer embodied orientation. So, to put it flippantly, he also has a non-doxastic account of beliefs. But I wasn't sure what the idea of different world really added to any agglomeration of piecemeal descriptions of delusions provided by their subjects or by clinicians. And then I wasn't sure how it related to the two (plus one) model of delusions which followed. Perhaps these are simply two independent ventures: an analysis of the nature of delusion and a suggestion for their causal explanation.

Joining us on Skype, Keith Frankish presented a thumbnail sketch of his account of delusions as acceptances. Starting with a distinction between type 1 and the 2 beliefs as:

1: behavioural dispositions which are non-conscious, accessed either by actual behaviour or sometimes by imagining what one would do and strongly subject to rational norms

2: occurrent thought which are conscious, available to verbal report, weakly subject to rational norms (slips of mind etc). (They are also controlled, discrete, binary.)

Compared to these two options, delusions look more like type 2 beliefs. But they, generally, look even more like acceptances, he argued. An acceptance is the strategy (cf of scientists or lawyers) of using propositions as premises of hypotheses but with no need actually to believe them. This picture assumes reasoning is intentional, consciously controlled and binary (like type 2 beliefs). But it can be pragmatically motivated (unlike type 2 beliefs). And it can be context relative (again unlike type 2 beliefs).

On this model, type 2 belief is thus a subset of acceptances: unrestricted, evidence-sensitive acceptance or 'doxastic acceptance'.

So the claim that delusions are acceptances amounts to either:

Option 1 doxastic acceptances.

Option 2 non-doxastic acceptances.

If option 1, then they are improperly executed doxastic acceptances ie. irrational beliefs. If 2, they were never intended to be unrestricted: rational pragmatic acceptances. As indirect evidence for this idea, Keith suggested that this picture coped well with Akrasia. 

1: the action dictated by a promising policy is incompatible with some other desirable action and

2: the desire to perform this other action outweighs the desire to carry through one's premising policy.

Similarly for delusions not motivating action. And hence this suggests new questions about, not what causes someone to believe that they are dead, but what purpose does the adoption of such an acceptance for the purposes of argument serve?

I asked whether he worried about whether subjects with delusions conceived of them as acceptances. He replied that the policies which fixed an acceptance were type 1 beliefs and desires and hence non-conscious and hence subjects did not so conceive their delusion. It slowly dawned on me that this seems to suggest something like this picture. Conscious type 2 states are the result of non-conscious type 1 states which - interestingly - are nevertheless second order (not his terminology) in so far as they are about policies for dealing with contents (so beliefs about beliefs etc). So the conscious access to type 2 seems to come out if nowhere. One discovers the nature of one's type 2 state (does this include the difference between beliefs and desires as well as doxastic and non-doxastic acceptances?) by interpreting oneself (as others do). This was a price he was prepared to pay. (Standard hermeneutic worries that I may have got hold of the wrong end of the stick apply more than usually to this account of his views, I should add.)

Just now I am listening to a talk by Patrice Soom offering a formal model based on functionalism of the claim that delusions might impact on other mental properties but not vice versa. One claimed virtue is its immunity to objections to the DSM definition such as a commitment to doxastic statement, the extrinsic characterisation. It fits the Coltheart etc two factor model. It is a brave account that any unrevisable belief is a delusion but Patrice was hip to this and embraced it is an interesting prediction of the model. Still, I can't help thinking that a question has been begged in ascribing contents - eg. that they are dead - and then adding to that the subject does odd things with that content (in this case, regards it as unrevisable). I can't help being suspicious about ever using the sentence: 'Jones thinks that she is dead'.

PS: Keith has now (18/12) sent me a substantial email correcting my understanding of his view. Once I've digested it, I will post a follow-up.

Bill Fulford talking to OUP about the new Handbook of Philosophy and Psychiatry.

Bill Fulford talking to OUP about the new Handbook of Philosophy and Psychiatry.

Bootstrapping conceptual normativity?

Abstract

Both anti-reductionist and reductionist accounts of linguistic meaning and mental content face challenges accounting for learning a first language. Anti-reductionists cannot account for a transition from the pre-conceptual to conceptual without threatening to reduce the latter to the former. Reductionists of a representationalist variety face the challenge of Fodor’s argument that language learning is impossible.
This paper examines whether Ginsborg’s account of primitive normativity might provide some resources for addressing these issues. Rejecting her ‘no conception’ account of normativity in favour of a demonstrative, local conception provides one response to Fodor’s argument which is available to an anti-reductionist and at least a further hint as to how context-independent linguistic concepts can be developed from context-dependent local conceptions of how to go on.

The problem of concept learning

One of the challenges for an anti-reductionist account of linguistic meaning and mental content is making space for an account of concept learning. If, following John McDowell for example, one takes the space of reasons to answer to a distinct constitutive ideal from that of the realm of law, it is hard to see how the route from the latter to the former can be articulated. Any such articulation would threaten to provide – what the anti-reductionist denies – a reduction of the concepts of the space of reasons to those of the realm of law.

McDowell himself suggests that Wittgenstein’s phrase ‘light dawns gradually over the whole’ provides a natural metaphor for learning a first language ‘for one’s dealings with language to cease to be blind responses to stimuli: one comes to hear utterances as expressive of thoughts , and to make one’s own utterances as expressive of thoughts’ [McDowell 1998: 333]. But he suggests that this process cannot be limited to a few sentences but involves working one’s way ‘into a conception of the world’. Suggestive though Wittgenstein’s phrase is, it does little to shed light on how the process of language learning might come about so much as summarise, albeit neatly, the fact that some such process does come about.

Whilst anti-reductionists face a principled problem, the most striking recent philosophical argument about language learning comes from one of their reductionist opponents. In LOT2: The language of thought revisited, Jerry Fodor sets out a specific argument for the difficulty of accounting for concept acquisition [Fodor 2008]. Or rather, he argues that such language learning must be impossible.

Fodor’s argument has four steps:

1.       Concept learning is a rational process.

2.       The only plausible rational process is hypothesis formation and testing.

3.       But that requires the conceptual representation of the hypothesis, which presupposes possession of the concept to be learnt.

4.       So concept learning is impossible.

The first step contrasts learning as a rational process with any form of non-rational process of concept acquisition such as by surgical implantation, swallowing a pill or hitting one’s head against a hard surface. Fodor then argues that the only plausible candidate for such a rational process is a ‘process of projecting and confirming hypotheses about what the things that the concept applies to have in common’ [ibid: 132].

The argument for the third step is couched in the terms of Fodor’s ‘Representational Theory of Mind’ (RTM) and is, initially at least, restricted to primitive, that is non-definable, concepts.

Consider any concept that you’re prepared to accept as primitive, the concept GREEN as it might be. Then ask ‘What is the hypothesis the inductive confirmation of which constitutes the learning of that concept?’ Well, to acquire a concept is at least to know what it’s the concept of ; that is, what’s required of things that the concept applies to. So, maybe learning the concept GREEN is coming to believe that GREEN applies to (all and only) green things; it’s surely plausible that coming to believe that is at least a necessary condition for acquiring GREEN. Notice, however, that (assuming RTM) a token of the concept GREEN is a constituent of the belief that the concept GREEN applies to all and only green things. A fortiori, nobody who lacked the concept GREEN could believe this; nobody who lacked the concept GREEN could so much as contemplate believing this. A fortiori, on pain of circularity, coming to believe this can’t be the process by which GREEN is acquired. [ibid: 137-8]

And hence, he argues, no primitive concepts can be learnt. (He goes on to lift this restriction but I will ignore that move here.)

Fodor’s RTM is implicit in the way he sets out the third stage: that a token of the concept is a constituent of the belief about the extension of the concept. This reflects his idea that concept possession is explained by inner vehicles of content: mental representations. But even without RTM, it is plausible to argue that the capacity to have such a belief – however realised – presupposes possession of the concept, ascribed at the level of the person, and so the argument floats free of Fodor’s particular views.

The principled problem of describing a process of concept learning from an anti-reductionist perspective and Fodor’s specific argument that concept learning is impossible presents a two-fold challenge. Can Fodor’s argument be blocked and if so can the materials used to do that shed light on concept learning even from an anti-reductionist perspective?

Ground rules

In what follows, I will make two substantial but related assumptions. First, that concept learning is normative. Second, that it is, in some sense, a rational process.

The view that concept learning is normative might follow from the view that concept use is normative. Such a view has been espoused by Blackburn, McDowell and others. But it has recently been contested by philosophers also responding to Kripke’s argument for meaning scepticism which starts – or, at least, seems to start – by assuming that meaning is normative and then using this to undermine its factual base. McDowell’s response is to agree that the claim that meaning is normative is itself philosophically innocuous and thus seek to show how that escapes Kripke’s argument. Anti-normativists such as Hattiangadi agree that meaning is connected to a notion of correctness but deny that correctness need be a normative notion. For example if R states a rule for the correct use of a term t which applies in virtue of features f

R             (x)(t applies correctly to x ↔ x is f)

then Hattiangadi argues that this ‘simply states the correctness conditions of an expression; it does not tell me what to do’ [Hattiangadi 2007: 223]. A mere descriptive sorting does not appear to raise the problems for an account for meaning that the claim that it is normative appears to.

One of the key arguments deployed by anti-normativists is that meaning yields no clear prescriptive norm in itself. Correctness conditions, for example, do not prescribe that a true (or more broadly correct) use should be made unless also combined with a prescriptive norm that one ought to speak the truth (or more broadly correctly). And that additional norm seems too strong for specifically semantic normativity.

Nevertheless, however that debate pans out, a rational process of concept acquisition or learning does look to be normative. In this context, correctness conditions are not merely a neutral way of sorting subsequent utterances but rather constitute the aim or goal of developing linguistic competence. (This is not to say that learning must always involve practising with only true utterances, customary though that may be. Parents may, and in my experience often do, teach correct use indirectly be deliberately calling out the wrong names of objects as a kind of game.) If an anti-normativist wishes to argue that the relevant prescriptive ought applies not directly in virtue of the rules of correctness of words but an adoption of those rules as the goal of concept learning, so be it.

The second assumption is that concept learning is, in some sense, a rational process. McDowell calls the acquisition of a first language a matter of being ‘cajoled’ [ibid: 333]. It is possible that being the recipient of such cajoling is not so much a rational response as being brutely changed in such a way that one can become a rational subject and make subsequent rational responses. But I will assume that it is possible to say something about rationality of the proto-linguistic responses of a subject in such a position.

My stalking horse will be Hanna Ginsborg’s account of ‘primitive normativity’.

Ginsborg on primitive normativity

Ginsborg’s account of primitive normativity is – like many of the anti-normativists - designed as a response to Kripke’s meaning scepticism [Kripke 1982]. As Ginsborg interprets the dialectic, Kripke assumes that grasp of what one meant by a word in the past sets the standard for the correctness of one’s current use of it. Part of what justifies giving the answer ‘125’ to the question of what ‘57 plus 68 equals’ is what one has previously meant by ‘plus’. Kripke then mounts a sceptical attack on how we can now know what we did earlier mean. Perhaps by ‘plus’ we really meant quus which tracks the plus function for past usage but not the two current numbers.

In response, Ginsborg denies that to claim that one ought to say ‘125’ one needs first to establish that one previously meant addition. That one ought to say ‘125’ is independent of any assumption about past meanings. She claims: ‘I maintain that there is a sense in which you ought to say “125,” given the finite list of your previous uses, independent of what meaning, if any, those uses expressed’ [Ibid: 232-3]. Given this context, ‘primitive normativity’ is a normativity independent of, and prior to, grasp of meaning. It is located below the level of facts about meaning though still irreducibly normative. She sets out an example of a child who is able to recite numerals and has learnt to count up in twos conuting from ‘40’ with ‘42’. Ginsborg characterises a conceptual-normativist account of the child’s saying ‘42’ as follows:

[T]he child says “42” after “40” because she recognizes, although without being able to put that recognition into words, that she has been adding two and that 40 plus two is 42. Her sense of the appropriateness of what she is saying thus derives from her recognition that it fits the rule she was following: a rule which she grasps, even though she is unable to articulate it. [ibid: 238]

On this higher level view, the correctness of the move – saying ‘42’ – depends on gasping a rule governing it. Primitive normativity involves less than that. But, at the same time, it involves more than the merely reliable dispositional reactions of a suitable trained parrot. By contrast with such a parrot, that the child does not respond ‘blindly’ to her circumstances.

Even though she does not say “42” as a result of having grasped the add-two rule, nor a fortiori of having “seen” that 40 plus two is 42, she nonetheless “sees” her utterance of “42” as appropriate to, or fitting, her circumstances. [ibid: 237]

So even though the child lacks full blown conceptual mastery, she has a sense of appropriateness, fitting or belonging which merits the label ‘normativity’. The parrot lacks any such sense and hence is merely governed by dispositions not norms.

Ginsborg gives a second example of the kind of middle level behaviour. She describes a child sorting coloured objects before she has acquired determinate colour concepts.

As she puts each green object in the designated box, it is plausible that she does so with a sense that this is the appropriate thing to do. She takes it that the green spoon “belongs” in the box containing the previously sorted green things and that the blue spoon does not, just as the child in the previous example takes 42 and not 43 to “belong” after 40 in the series of numerals. But her sense of the appropriateness of what she is doing does not, at least on the face of it, depend on her taking what she is doing to accord with a rule which she was following, for example, the rule that she is to put all the green things in the same box. For her grasp of such a rule would presuppose that she already possesses the concept green. [ibid: 235]

There are two sorts of general consideration to support the idea of some sort of primitive normativity. One relates to Ginsborg’s specific dialectical context. Mere dispositions will not provide a satisfactory response to Kripke’s sceptical argument whilst full blown conceptual normativity will be vulnerable to his original argument. (This is not to say that these are clear cut but they provide a rationale for attempting to articulate a middle ground.)

The other relates back to Fodor’s argument. Mere dispositions seem to leave too much of a gap still to cross to explain how a dispositional stage might be an intermediary en route to conceptual mastery. By contrast, invoking full blown conceptual mastery to characterise the counting child is to provide no answer to the question how basic concepts can be learnt.

Nevertheless, there are two options for characterising primitive normativity based on two distinct things Ginsborg says. She says of the counting child both that:

1: ‘she lacked any conception of what her saying “42” after “40” had in common with her having said “40” after “38”’ [ibid: 234 italics added]

but also:

2: ‘it seems plausible to imagine her insisting, with no less conviction than a child who was able to cite the add-two rule, that “42” was the right thing to say after “40”: that it “came next” in the series, or “belonged” after 40, or “fit” what she had been doing previously’ [ibid: 234 italics added]

The former states that the counting child has no conception of what one move has in common with a previous move. The second allows for the possibility of some conception that the next move fits or belongs (ie does have something in common) with the previous one in context. The latter allows for a conception albeit a local one. Which does Ginsborg hold?

‘No conception’ primitive normativity

There is reason to think Ginsborg believes in the more radical, minimal version. One suggestive passage runs:

The utterance, from [the counting child’s] point of view, is not appropriate to the context in virtue of its conforming to a general rule which the context imposed on her, for example, the add-two rule. Rather, she takes it to be appropriate to the context simpliciter, in a way which does not depend for its coherence on the idea of an antecedently applicable rule to which it conforms. [ibid: 234-5]

Now one way to interpret the phrase ‘antecedently applicable rule’ is a context-independent general specification of a rule. In the context of a mathematical series, that is a plausible way of cashing out full blown conceptual normativity. And hence its rejection might allow for a merely demonstratively specified local conception of the demands of a rule. On this alternative view, whilst the child does not have a general conception of what it is to add two, cannot grasp its relation to other aspects of arithmetic for example, she can, nevertheless, recognise in some particular context that saying ‘42’ is the right move.

But the phrase ‘antecedently applicable rule’ might equally be taken to mean, and hence to rule out, any conception of a rule. If so, the context imposes a sense of what move belongs with previous moves, of what next move is right, independently of any conception the child has of what she is doing. The way the quotation continues adds to this latter impression:

This is not to deny that the normativity depends on any facts about the context, since the appropriateness of “42” depends on her having recited that particular sequence of number words. But it is to deny that her claim to the appropriateness of “42” depends on her recognition of a rule imposed by the context in virtue of the relevant facts, or a fortiori on her recognition of “42” as a correct application of the rule. [ibid: 235]

This suggests a picture according to which facts about the context external to the child’s conception of the demands of the rule nevertheless make normative demands on her. The context of having counted up to 40 makes saying ‘42’ appropriate independently of her conception of what she is doing. ‘42’ belongs to what has gone before, is thus normatively connected to it, but she does not recognise that this is the demand that the rule makes in the context.

A second passage provides a distinct argument for the ‘no conception’ view of primitive normativity on the assumption that even a local conception of what the next move is requires some grasp that this is relevantly the same as previous moves.

[T]he child’s recognition of similarity is not sufficient to account for her taking herself to be going on appropriately. She must not merely take herself to be going on the same way; she must also take it that going on the same way is the appropriate thing to do in the context, which is to say that she must grasp a rule with a content like go on the same way or do the same thing you were doing before. We are thus left with the problem of how to account for her grasp of this rule... [ibid: 240]

The argument is that grasp of sameness is insufficient for knowing how to continue. One would need to grasp the further rule that one should go on in the same way, that this is what the relevant normative demand is.

There is something to this worry. There seems little prospect of factoring grasp of a rule into grasp of sameness plus grasp that sameness is what one ought to aim at. Wittgenstein stresses that agreement is internal and relative to the particular rule [cf Wittgenstein 1953 §224]. Thus grasp of the rule and grasp of what agrees with it and hence what is relevantly the same in virtue of according with the rule goes hand in hand.

But this point applies equally to what Ginsborg does make explicit: that the child grasps that the next move fits, belongs or is appropriate to the context. Those notions are equally insufficient for going on correctly. (A rule could dictate that the next move should stand out from, rather than fitting, what has gone before.)

It seems on balance, however, that Ginsborg does subscribe to a minimal ‘no conception’ version of primitive normativity (I will return to a further strategic reason why this is so shortly). Further, she seems not to be alone. In a passage in which she discusses how little may be necessary for rule following, Julia Tanney considers the conceptual possibility of rule following without the ability to cite higher level rules, or to repeat the performance or without training. She comments:

[I]f we agree with the thought that someone might be able to solve Rubik’s Cube even if she had never been trained by anyone, then this gives us a reason to reject the idea that there must be an internal connection between the rules that govern an activity and the individual who makes the moves. We can say that it is sometimes enough to credit someone with playing the game if she acts in accordance with the rules. Knowledge (implicit or otherwise) has dropped out of the picture. To insist that someone cannot solve the puzzle unless she somehow conceives the rules (even if she cannot articulate them, even to herself) and acts in the light of her conception of the rules is simply dogmatic. What would justify such insistence? If this person were suddenly entered in a contest and produced the cube with the colours in the right places, we would not withhold the prize because she merely acted in accordance with, but did not follow, the rules. Acting in accordance with the rules is solving the puzzle in certain cases. [Tanney 2013: 85-6]

On this account, having rejected a number of potentially necessary substantial claims as in fact unnecessary for rule following, Tanney concludes that, in the right context, mere accord with a rule constitutes rule following. Further, this does not seem to be merely a claim about the epistemology of the ascription of rule following – which, indeed, in the right context, apparent accord warrants the further ascription of intentional rule following – since Tanney connects it to the rejection of an internal connection between rules and agent. An epistemological interpretation, by contrast, is consistent with maintaining that accord in performance is evidence for such a connection, amounting to the grasp of the rule by the agent. Instead, and in response to a number of bogus explanations of rule following which fail because they presuppose precisely the abilities they purport to explain, Tanney offers a kind of deflationary approach. The failure of cognitivist explanations of rule following leads to a rejection of cognition. ‘To insist that someone must conceive the rules somehow – even if what it would be for her to conceive these rules is inaccessible to us – is misguided; it fails to explain anything’ [ibid: 86].

Despite this support, the ‘no conception’ version of primitive normativity faces a key objection. It severs the connection between primitively rule-governed behaviour and intentional action and the blurs the distinction between mere accord with a rule and intentionally following it. It is this distinction which marks the difference between a merely dispositional parrot, whose behaviour may accord with a rule available to a third person description, and a human subject with some sense of her new moves fitting or belonging with what went before, some sense of normative correctness.

But there is no need to get into such difficulty if the aim is merely to fit an intuitive description of the phenomenology of the child’s early performance in, as we might say, counting in twos or grouping by colour. The middle ground between dispositional accord with a rule and full-blown conceptual normativity is not the primitive normativity of someone with no conception of what she is doing but with a merely local conception. Such a conception is not tied to the local context of counting or sorting objects brutely or merely externally in virtue of an ascription of rule-accord by an observer. Rather, it is expressed by the demonstrative judgements of the child and her capacity to demonstrate and explain by example what fits with what she has been doing.

This idea runs counter to one of Ginsborg’s explicit claims: ‘I maintain that there is a sense in which you ought to say “125,” given the finite list of your previous uses, independent of what meaning, if any, those uses expressed’ [Ibid: 232-3 italics added]. On the local conception this is wrong. Correctness is tied via a local conception to what a speaker’s past utterances expressed even if the speaker is unable to offer a context-independent linguistic codification of her actions as instances of following the plus-two rule or the sorting of green objects. Her conceiving of her actions might not extend very far up the natural numbers (eg beyond 100) or to cover darker or lighter shades of green (by contrast with the vivid colours of children’s toys). So it is potentially doubly local: expressible only in some particular context of practical demonstration (by contrast with context-free linguistic codification) and covering only some particular instances and thus not actually extensionally equivalent to our concepts of plus two, or green but rather a primitive version of them.

I suggested that there is a further strategic reason why this view is unavailable to Ginsborg. She deploys the idea of primitive normativity as a novel response to Kripke’s sceptical argument. Her aim is to sidestep the arguments Kripke deploys against any justification one can currently offer for knowing what one meant in the past by one’s words, thus to undermine a standard of correctness for current use. Primitive normativity has, for those strategic purposes, thus to be independent of any conceptual conception. A local conception is, however, a form of conceptual conception and its expression in a pattern past, finite examples is just as much subject to Kripke’s argument as a full blown linguistic concept. That is to say, that is not part of a new defence of meaning against Kripke’s argument.

‘Local conception’ primitive normativity and language learning

Primitive normativity guided by a local conception of what a speaker is doing promises a partial answer to the initial two-fold challenge of describing language learning. One aspect of Fodor’s challenge was to sketch a rational mechanism for concept acquisition (hence learning). On the assumptions that a) the only plausible option is hypothesis formation and testing and b) hypothesis formation presupposes the very conceptual mastery in question, no rational mechanism seems possible.

Primitive normativity guided by and expressive of a local conception is a plausible intermediary between mere dispositional accord with rules and full blown linguistic mastery. The intermediate stage involves testing the hypothesis that a new linguistic concept expresses a content previously grasped in some local demonstrative manner. As suggested above, the grasp of a full blown linguistic concept may requires the piecemeal extension of a more primitive, merely local conception of a rule. But there may be some gradations of understanding between having no and a first language.

Of course, the very idea of an essentially situation-dependent conceptual understanding does not fit within the basic idea of Fodor’s representational theory of mind according to which content always has an inner vehicle. So no such middle ground is available to Fodor himself. Thus for anyone uneasy with Fodor’s innativism, his argument against language learning remains a powerful reductio of his representationalism. But the idea does provide a way to address the version of his argument mentioned at the start framed in terms of prior concept possession but agnostic about Fodor’s account of inner vehicles.

What of anti-reductionism? A local conception offers only partial progress here. By contrast with Ginsborg’s own account of primitive normativity, the idea that normativity always presupposes that the subject has some albeit local conception according to which she acts provides no middle ground between the ‘space of reasons’ and the ‘realm of law’ or the ‘manifest image of man in the world’ and the ‘scientific image’. Even a local conception belongs in the space of reasons or the manifest image. So it cannot be part of a route into those spaces from without. But it does help put a little flesh on the bones of the idea that ‘light dawns gradually over the whole’.

‘The whole’ need not merely be understood to be the gradual acquiring of a world view, as Wittgenstein describes in On Certainty (from where the phrase comes) [Wittgenstein 1969 §141]. It can also include mastery of primitive albeit still conceptually structured rules. This is no account of how conceptual normativity can be bootstrapped from non-conceptual dispositions but does suggest how more complex and abstract concepts cane be developed from more primitive local forms.

References

Fodor, J. (2008) LOT2: The language of thought revisited. Oxford: Oxford University Press

Ginsborg, H. (2011) ‘Primitive Normativity and Skepticism about Rules’ The Journal of Philosophy 108: 227-254

Hattiangadi, A. (2006) Oughts and Thoughts: Rule-Following and the Normativity of Content. Oxford: Oxford University Press

Kripke, S. (1982) Wittgenstein on rules and private language. Oxford: Blackwell

McDowell, J. (1998) Meaning knowledge and reality. Cambridge, MA: Harvard University Press.

Tanney, J. (2013) Rules, Reason and Self Knowledge. Cambridge, Mass.: Harvard University Press

Wittgenstein, L. (1969) On Certainty. Oxford: Basil Blackwell.

Notes for a talk on the meaning of life

I am giving a short 20 minute talk some time between 7:45 and 8:30am tomorrow in Scholars on campus on the meaning of life. It is one of those things that non-philosophers find disappointing about academic philosophy that this subject doesn't come up all that often and I must say that I feel rather fraudulent.

Exploring the Meaning amongst the Doing

Or: What we talk about when we talk about ‘the meaning of life’

1: The problem of talking about ‘the meaning of life’

Douglas Adams, Deep Thought, 42 and the ‘Ultimate Question of Life, The Universe, and Everything’.

We already know the absurdity of that being the right answer. And yet we also realise that there is something very odd about the question.

Philosophers’ quick ways with such things: declare such issues meaningless via tests of meaning such as the verification principle. Too quick.

2: Making sense of our actions

We explain our actions differently from other happenings. We shed light on / make sense of / justify them.

Two forms of explanation. “Why are you gathering kindling?”.
Either: “I am building a fire”.
Or “I want to build a fire.”
The latter may better fit a scientific explanation but the former is basic.

Rational explanation.

3: Action explanation iterates

“Why are you building a fire”. “I’m cooking / I want to cook a meal”.
“Why are you cooking a meal?”…

We make sense of individual actions by putting them into a broader context. Without limit?...

4: The danger of contemplation

The iteration of practical action explanation stops with what is not (<> cannot be) called into question: eg a local conception of a life. (Links to a sense of identity.)

But at 4am:
“Why am I – do I live as – a chef?”
“Although I am a chef, should I be one?”

For example, for anything I have been told to do by an authority, at 4am I can ask, should I listen to that authority? Eg, the law. Nothing written down actually compels. Nothing is intrinsically compelling. We have to agree to be so bound.

5: The meaning of life?

The problem: without some yet broader context, any project / conception of life could always be questioned.

Hence we want a context for our actions which itself needs no further explanation / justification: the Meaning of Life itself.

But a) the more from a local, assumed conception of a life – eg. being a chef – to a completely general one, the more possibilities are open and hence the stronger the selective justification needs to be. What general conception of a Meaning of Life would enable the derivation of all the different local conceptions of life which, as a matter of fact, turn out about right? (Would it also apply to life on Mars?)

And b) (as above) we have no model for an intrinsically compelling conception of what we ought to do.

6: The moral

The meaning of a life is found only within some local conception of a life. There are limits to disengaged contemplation, explanation and justification: the meaning is in one sense ineffable.

We can only explain the attraction of a local conception of a life to those who already share similar sensibilities.

We demonstrate it. The meaning is in the doing.

At the Philosophy at 40 at Anglia Ruskin conference

Having only taught a single module (on Wittgenstein) for a couple of years here, I am something of an interloper at the conference celebrating 40 years of philosophy at Anglia. But it is interesting to arrive to see a hugely developed campus and to sit in The Lord Ashcroft building in rather a pessimistic round table discussion of the future of universities (and the future of philosophy). There is a shared agreement that there has been a collapse in the idea of a common good and hence the ground rules for financing higher education (and for that matter, energy supply, the railways, the postal service etc).

In that context, one particular paper from the morning stood out. Mike Wilby gave a paper on natural  inter-subjectivity which took as a starting point the contrast between chimps and small children taking part in cooperative activity. Whilst chimps have a sensitivity of fellow chimps' perceptions in competitive behaviour (subordinate chimps only taking food when they see that the dominant chimp cannot see it) they do not in cooperative behaviour. By contrast, small children are able to play cooperatively from the age of nine months or so.

Wilby took his challenge to be to give an account of essentially shared or mutual mental states in the face of incredulity from the likes of John Searle or Peter Strawson. Considering a case in which two people must cooperate to catch a rabbit, he suggested that, like an individual case, there would need to be some account of a development from a general prior intention (to catch some rabbit or other) to an particular or object-dependent intention-in-action to catch that! rabbit. In the individual case, the obvious intermediate is a perception of a particular rabbit.

Wilby's argument was that none of the three states in the cooperative activity could plausibly be reduced to an individual account. An individualistic version of the prior intention would have to be something like: I intend that we catch a rabbit. But I can only intend my own actions. At most I can intend to make it come about that we catch a rabbit. But such individual intending to make a joint action come about might not be cooperative (Bratman's 'mafia objection') and hence does not capture the cooperative example at hand.

There are also problems also with modelling joint attention in individualistic terms. Wilby suggested a kind of never ending escalation from:

I perceive the rabbit.

I perceive the rabbit and I perceive that you perceive the rabbit.

I perceive the rabbit and I perceive that you perceive the rabbit and I perceive that you perceive that I perceive the rabbit etc etc

But none of these closes off what mutual attention seems to achieve. And there are similar problems with the intention-in-action. Given additionally the complex individualistic model would have to be grasped by the nine month old children who play cooperatively this is surely all better accounted for by the idea of genuinely mutual mental states such as prior intentions, joint attention and mutual intentions-in-action.

I rather liked all this, not least because it provides a further argument against the reductionism of representational theories of mind.