Wednesday, October 16, 2019

Are There Degrees of Assent? (Re-Post)

This is a slightly revised version of a post first published in 2007.

*****

In ECHU IV.15.2, John Locke introduces the notion of 'degrees of assent'. On the basis of this notion he formulates his claim in IV.16.1 that the degrees of assent should be regulated by the degrees of probability. This is a fairly common view in modern philosophy; you can call to mind, for instance, Hume's famous saying, "The wise man proportions his belief to the evidence," which Hume means quite literally: the degree of belief should match the degree of evidence. Is it true, however?

J. H. Newman, in his An Essay in Aid of a Grammar of Assent, argues that it is not, by arguing that there are no degrees of assent. I'll look a bit at this argument. Some terminological issues, first. Newman makes a distinction between assent and inference as ways in which claims or propositions can be accepted. The idea is that inference is conditional: it is the connecting of claims to each other. In inference one accepts something's following as a conclusion rather than in itself. Assent has to do with the acceptance of the claim itself; and the question of the degrees of assent is whether this acceptance mirrors or echoes inference or not. This is because the connection of a thing with its evidence is precisely what inference is. Newman distinguishes two sorts of assent; they differ not in degree of assent but in the vividness or forcefulness of the thing to which one assents. In notional assent one assents to something taken notionally or abstractly; in real assent, which Newman also calls belief, one assents to it taken as psychologically powerful and living. What distinguishes notional from real assent is that when a proposition is assented to really, it is taken as something personal and important, whereas a proposition notionally assented to is not.

Now, to Newman's argument.

1) First, we need to see how assent is distinguished from inference at all. Newman argues that Locke's claims would in effect make assent indistinguishable from inference, or, at best, a superfluous echo. So the degrees-of-assent view, as put forward by Locke, raises the question of whether there is properly any such thing as assent at all. If, however, there is reason to think assent something distinct from inference, this suggests that Locke's view is wrong. So Newman suggests six points on this issue.

1a) Assents may endure without the presence of the inferential acts that originally elicited them. We may assent to something long after we have forgotten why we did so.

1b) Without tangible reason, assents may fail while the inferential acts that originally elicited them endure.

1c) Sometimes assent is not given in spite of strong and convincing arguments.

1d) There are many cases in which the arguments, good as they may be, nonetheless do not incline us toward the position at all. We see this in implicit and explicit uses of 'burden of proof': 'burden of proof' essentially throws all the weight on one side. In such cases we do not assent a little bit in proportion to the evidence; we simply do not assent at all.

1e) Moral and psychological motives may hinder assent to logical conclusions. We can infer against our will, but, as the saying goes, 'a man convinced against his will is of the same opinion still.'

1f) Even in the province of mathematics assent and inference must be contrasted. Simple mathematical arguments often command assent straightway; but we are often more cautious about complicated mathematical arguments. Likewise, party feeling and the like have often slowed the acceptance of mathematical conclusions. Likewise, a mathematician might well make it a rule not to accept a mathematical argument that has not been corroborated by someone else, even if he is fairly sure that the argument is sound. And so forth. Even demonstration is not ipso facto assent.

Thus there is good reason to believe that inference and assent are distinct acts of mind that can be made apart from each other. This does not mean that there is no connection between the two; but one may well be had without the other.

2) If inference and assent are distinct, however, we can ask how they are distinguishable. There are some basic prima facie reasons to think that the categorical, unconditional nature of assent is precisely what distinguishes the two:

2a)Given that inference is conditional and non-categorical, we have some reason to think that assent would be unconditional and categorical.

2b)Further, If assent is acceptance of something as true, and if we do not accept conditionally what we think to be true, then we have another reason for thinking assent to be categorical, straightforward adherence to the claim itself.

2c)Further, inference is always inference, even when demonstrative. It makes sense to say that assent gives categorical and absolute recognition to conclusions held by demonstrative inference, beyond the accepting of them merely inferentially, i.e., merely in virtue of their linkage with other propositions. If assent is categorical in such a case, however, we would need a good reason to think it is not categorical in other cases.

But we can go further and ask whether assent is conditional in the concrete as well as in abstract, demonstrative cases. There are, Newman thinks, many cases in which we accept claims without really assenting to them. For instance, we suspect them to be true, we conjecture that they might be true, we presume on the basis of what we know that they are true, we conclude they are true, etc. But these are not assents but inferences; they involve propositions insofar as they are connected with other propositions evidentially, and need not be considered assent at all. We can, of course, assent to the probability of something's being true or false; but this is not a degree of assent, but an assent to a degree of probability, which is not the same. In assent, we simply accept the claim; and it seems difficult to find any case of assent in which any 'degrees' that may not be more accurately attributed to things beside the assent itself.

3) So, looking at the facts, we find many cases in which we infer without assenting, and no cases in which it is clear that the assent is conditional; we also can find many cases in which it is clear that the assent is unconditional, even though the cases admit of no reasoning more certain than the probabilistic. And what is more, it seems to be, by a sort of unanimous witness of the human mind, a strong evidence that there is no medium, no tertium quid, between assenting and not assenting. What sorts of unconditional assents in probabilistic matters can we find? Here's Newman's list:

3a) that we exist;
3b) that we have an individuality and an identity all our own;
3c) that we think, feel, and act;
3d) that we have a present sense of good and evil, of right and wrong, of true and false, of beautiful and hideous, whatever account we may have of this;
3e) that such-and-such happened yesterday or last year;
3f) that of many things we are ignorant;
3g) that of many things we are in doubt;
3h) that of many things we are not in doubt;
3i) that our own self is not the only being existing;
3j) that there is an external world;
3k) that the world is a system with parts and a whole, a universe carried on by laws;
3l) that the future is affected by the past;
3m) that the earth is a globe;
3n) that the regions of the earth see the sun by turns;
3o) that vast tracts of the earth are land or water;
3p) that there are really existing cities on definite sites, like London, Paris, Florence, and Madrid;
3q) that, unless something has happened to them like an earthquake or terrible fire, these cities are today much what they were yesterday;
3r) that we had parents (despite having no memory of our birth);
3s) that we shall die (despite the uncertainty of the future);
3t) that we cannot live without food (despite never having tried);
3u) that the world of men has a history that precedes our time considerably (despite our not having experienced it);
3v) that there have been rises and falls of states, great men, wars, revolutions, art, science, literature, religion;
3w) that our intimate friends are not being treacherous to us;
3x) sometimes that someone is hostile and unjust to us;
3y) that we have sometimes been cruel or unkind to others, or that we have sometimes been ungenerous to those who loved us;
3z) that that we have moral weaknesses and that our wealth, health, position, and good fortune can be precarious;
3aa) that we have such-and-such physical weaknesses or flaws;
3ab) that such-and-such food or medicine is good for us;
3ac) that such-and-such food or medicine would harm us;
3ad) that we have made such-and-such mistakes, or gone through such-and-such major turning points, or had such-and-such successes;
3ae) perhaps that we have a sense of the presence of a Supreme Being;
3af) perhaps our religious beliefs;

And so forth. And in these cases, despite the fact we cannot prove them with perfect certainty, in our practical lives we simply accept them as true. We may be wrong in accepting some of them; but no one feels guilty for just simply assenting to them despite this possibility. No one is worried that their unhesitating and unqualified assent to the claim that so-and-so is their mother is not something they can back up with rigorous demonstration. Now, unless there is something fundamentally irrational about the way we think, simple assent to non-demonstrated conclusions cannot be irrational. There is no reason why we should not simply accept that Great Britain is an island, even if we cannot prove it beyond the shadow of any doubt. And Locke himself effectively admits this in several places.

4) Now, people sometimes accept the degrees-of-assent view because of some ways in which we tend to talk about the subject. So what is really going on in these cases?

4a) Take claims of modified assent, qualified assent, presumptive assent, prima facie assent, or half-assent. Newman explains many of these as indicating the difference between notional assent and real assent; in some cases we are not genuinely assenting at all; some are indications of the conditions that need to be put on the claim before we will assent to it.

4b) Sometimes we talk about conditional assent; and this, again, means we think we will assent under certain contingencies or conditions.

4c) Sometimes we are simply talking about the circumstances of assent.

4d) When we use phrases like 'firm assent' or 'weak assent', but again it can be shown that either these are differences in that to which we are assenting, or in the concomitants or circumstances of assent (e.g., our feelings about what we are assenting to, or how much it sparks our imagination).

Such is Newman's argument, stripped down a bit. H. H. Price in his Gifford Lectures, Belief, argues against Newman on this point. Unfortunately, Price's argument, although in places insightful, never really takes the trouble to fix what Newman's argument actually is, and so is a bit all over the place. Price claims that if Locke's position is false, "our human condition must be both more miserable and more intellectually disreptuable than we commonly suppose" (p. 133). He glosses the former in a later passage:

It would be more miserable, because we so often need to be able to assent to propositions on evidence which is far less than conclusive; and therefore we need to be able to assent to them with something far less than total or unreserved self-commitment, if we are to have any guidance. (p.155)

But one may assent to something without "total or unreserved self-commitment"; many notional assents are of this sort, because they do not introduce the right sort of concomitants (the additional feelings and imaginations beyond assent that make what we are assenting to seem more than a mere abstract claim). Contrary to Price's account of Newman, Newman does allow we can assent to something with less than total or unreserved self-commitment; but the issue of self-commitment goes way beyond assent, since (as one might guess from the name alone) it involves the whole person. And we may assent without committing ourselves entirely to the truth of that to which we are assenting. I may categorically assent to the claim "Great Britain is an island" without thereby committing myself to live or die in defense of that truth, and without committing myself to staking my life on the truth of the claim. But nothing in this requires that the assent itself be tentative or conditional; I don't merely accept "Great Britain is an island" insofar as it is evidentially linked with such-and-such other propositions. I simply accept it as true. I can think of other propositions to link it to, of course; but this is not involved in my assent to it.

Price further glosses his claim:

If Newman were right, our situation would also be more intellectually-disreputable than we commonly suppose. In such circumstances, where we have evidence which is less than conclusive, only two alternatives would be open to us: either complete suspense of judgement, or else an assent of the all-or-nothing ('unconditional') sort, which would be unreasonable, because nothing short of conclusive evidence could justify it. (p. 155)

But why in the world would one accept this claim? Locke himself, for instance, does not; Newman's argument isn't devoted to criticizing Locke alone, but this is one of the strengths of Newman's argument if taken as a criticism of Locke. Locke admits that there are cases in which it can be perfectly reasonable to accept as certain what we cannot prove to be certain; and he thinks, rightly, that demanding otherwise is unreasonable. Most of us do not have a demonstration showing with conclusive evidence that the sun will rise tomorrow; however, we assent to the claim unconditionally. And, indeed, isn't it more reasonable to think this reasonable than to think that we could only assent to the claim "in an intellectually reputable" way if we could rigorously demonstrate it? Price goes on to say:

When our evidence for a proposition, though not conclusive, is favourable, or favourable on balance when any unfavourable evidence there may be is taken into account, we can assent to that proposition with a limited degree of confidence; and we can then conduct our intelletual and practical activities 'in light of' the proposition, though not without some degree of doubt or mental reservation. (p. 156)

Newman has no problem with this, and it does not show that assent comes in degrees. All it shows is that a claim's relation to other claims admits of degrees, and we can recognize that. Newman allows that we can assent to something's being only-so-probable; but again, this is not the same thing as assenting-only-so-much. You can assent categorically and simply to something's being only-so-probable, an act 'in light of' that proposition. Likewise, there is nothing to prevent one from acting in light of inferences, i.e., things accepted only in virtue of their relation to another conclusion, even one does not assent to them. (E.g. a person may infer that God exists if such-and-such is true, but not assent to it, and still may hedge his bets in matters where God is concerned, for precisely the reason that such-and-such could conceivably turn out to be true.) So Price is not giving Newman's view the credit for the flexibility it actually has.

In short, Price has it exactly backwards. Our human condition is only miserable and intellectually disreputable if Locke's claims are true. Locke's claims, so far from being essential for being reasonable, make unreasonable things that are obviously reasonable. If they were true, we would all be breaking Locke's rule about assent every day, in a myriad of situations. We couldn't help ourselves, first, because it would be impossible for us to follow Locke's claim, and, second, because it would be unreasonable for us to do so. Further, we have very little evidence on which to base the claim that assent comes in degrees; so if Locke's rule should be accepted, we should not hold his first claim very strongly. And this is the best that could be said assuming that assent comes in degrees. In actual fact, we have reasons to think assent does not come in degrees, despite being associated with some things that do. And further, no one has ever given any reason to believe that assent does come in degrees beyond some vague appeals to ways of talking that can be explained more fruitfully in other ways. It seems the only real question is "Should we assent or should we not?", not "To what degree should we assent?"