Kegar Thus, for instance, a perfect expansion of the Sherlock Holmes stories would not have to be counted as a true description of the world, despite its coherence. Subsequent security is security which a belief acquires as a result of its contribution to the coherence of the set. So as the system grows, that link can become stronger. Classic Problems and Contemporary Responses. We have no independent reason to prefer to retain highly observational beliefs in preference to theoretical ones.
|Published (Last):||1 October 2012|
|PDF File Size:||17.69 Mb|
|ePub File Size:||15.96 Mb|
|Price:||Free* [*Free Regsitration Required]|
In the last two chapters we have begun to treat our beliefs as a kind of interrelated theory, and the problem has been how the beliefs are related. There are of course many aspects of this question which we have not examined, but we have found reason to reject one answer to it.
This is the view that the relation is crucially asymmetri- cal; that there is an asymmetrical distinction between evidence and theory under which evidence confirms and disconfirms theory in a way in which theory cannot confirm or disconfirm evidence. Foun- dationalism offers such a structure in its assertion that the direction of justification is all one-way, and in its claim that there are some comparatively fixed points in the structure, the basic beliefs.
The notion of inference from fixed points clearly embodies the relevant asymmetries. The notion of inference itself is asymmetrical.
It is possible to infer B from A without being able to infer A from B. The notion of coherence, on which a more completely holistic theory is based, is intended to be symmetrical. All coherentists agree that consistency is a necessary condition for coherence. Bradley added Bradley, , pp. We shall see why soon. But consistency and completeness were not enough; they did not capture the feeling that a coherent set stuck together or fitted together in a special way.
To capture this, classical coherentists use the notion of entailment p entails q iff, given p, q must be true. Brand Blanshard wrote that in a fully coherent system "no proposition would be arbitrary, every proposition would be entailed by the others jointly and even singly, no proposi- tion would stand outside the system" Blanshard, , vol. But this account of coherence in terms of mutual entailment is disputed. Ewing suggested that it would be sufficient that each member of a coherent set be entailed by all the rest Ewing, , p.
Indeed, can we make sense of the idea of a system within which each member entails all the rest? Instead of answering this question directly, we can move towards it by considering an objection to any use of the notion of mutual entailment as the central element in a coherent set. That notion, as Blanshard uses it, is symmetrical enough.
But entailment as tradi- tionally understood is not a matter of degree. And this is not just because it becomes more complete; completeness can hardly be a virtue in itself. And we cannot rely on the point that the relations of entailment only hold between members of a complete set, because this would not really capture the sense in which we aim, in expanding our belief-set, to make it more coherent. Other problems with the appeal to entailment are explored in Rescher, ,ch.
So if we are to have a coherence theory of justification, we need to give a good sense to the idea that justification can grow. An alter- native account of coherence, offered in Lehrer and Sellars , defines a coherent set as one which is consistent, complete and mutually explanatory. The idea here will be that, as the set increases in size, we can hope that each member of it is better explained by the rest. Explanations can improve in quality; this accounts for the growth of justification.
And the notion of mutual explanation is clearly symmetrical, in the required sense. Two comments could be made on this account of the coherent as the mutually explanatory.
And it is just as well, because there is no really clear notion of completeness available here. We might perhaps suppose that a complete set contains every proposi- tion or its contradictory.
In the same way, we have no clear idea of a perfect explanation, a point from which things cannot be improved. So doubts about completeness make me happy to leave it out of the definition of coherence. Other reasons will emerge later. The second point is that coherence is a property of a set of beliefs, not of the members. The set is coherent to the extent that the members are mutually explanatory and consistent. This will be important in what follows.
It may seem, then, that our account in terms of mutual explana- tion is an improvement on that which appeals to entailment to tie the coherent set together. And it is the basis of our remark above that entail- ment is not a matter of degree. But Blanshard does not conceive of entailment in this way, as we would expect of anyone who is a holist in the theory of meaning. For him, entailment only occurs within a system; and since the system determines the meanings of p and of q, it determines the strength of the link between p and q.
So as the system grows, that link can become stronger. There is anyway an obvious intuitive link between entailment, as Blanshard sees it, and explanation. To explain q by appeal to p is to show why q should be true, given p. The explanation works to the extent that it shows that, given p, q must be true.
And like entailment, explanation should be viewed holistically rather than atomistically. So at the end of the day our two accounts of coherence collapse into each other. Before we turn to the coherence theory of justification we need first to consider the coherence theory of truth; the two are closely connected. Propositions are true to the extent that there is a coherent set of which they are members.
Notice, however, that the theory does not identify truth with coherence. It gives no sense to the notion of a true set. Instead, it defines truth for members of sets.
A pro- position is true iff it is a member of a coherent set. Propositions cannot be coherent, in the required sense, and sets cannot be called true unless they are members of larger sets. However, the theory does purport to offer a definition of truth. It does not restrict itself to telling us what circumstances would justify us in taking a proposition to be true. It might do this by claiming that we are justified in believing that p is true to the extent that doing so would increase the coherence of our belief-set.
The coherentist does make this claim; he does offer a criterial account of truth, a theory about what are the criteria for truth. But he also offers an account of what truth itself is, a definitional account. The two accounts are supposed to fit together, as we shall see. Many philosophers who have shown an interest in the coherence theory of truth have disputed the view that the theory offers a defini- tion of truth on the grounds that, taken that way, the theory is manifestly false e.
Russell, It is manifestly false because no matter how tight our account of coherence we shall have to admit that there may be more than one coherent set of propositions.
Nothing in the notion of coherence, as defined, gives us any right to say that there is a unique most coherent set. But it is obviously the case that there can be at most one complete set of truths. So truth cannot be defined in terms of coherence alone.
If there is more than one theory equally effective in handling the evidence, what are we to say about the different theories? Can we perhaps say that they are all true, or that all their members are true?
It seems that we cannot. If our different coherent sets are all of them verging on complete, if they constitute complete but different descriptions of the world, how can we admit that all the parts of these different descriptions of the world are true? Surely if the descriptions are different, they are competing, and the prize they are competing for is the prize of truth.
Hence only one of these competing sets can contain nothing but truths, and the coherence theory of truth is wrong. This objection to the coherence theory of truth is standard. We can call it the plurality objection. Brand Blanshard writes , vol.
But if intended to represent the coherence theory as responsibly advocated, it is a gross misunderstanding. Blanshard is arguing that the plurality objection fails to appreciate the empiricist character of his coherentism. For he takes it, as do other coherentists such as Bradley, that there is only one coherent set, and that this set is distinguished from all rivals by being empiri- cally grounded. This is so, according to Bradley, because of the very aim of thought and enquiry, which is to discover the most systematic ordering of our experience Bradley, , p.
My object is to have a world as comprehensive and coherent as possible, and, in order to attain this object, I have not only to reflect but perpetually to have recourse to the materials of sense. I must go to this source both to verify the matter which is old and also to increase it by what is new. And in this way I must depend upon the judgements of perception. What these coherentists are saying is that the enterprise is to start from the data of experience and to construct a set of beliefs around those data which will order the data in the most systematic coherent way.
To do this we may need to reject some of the data, but we cannot reject them all because our very aim is to make sense of what we have as data. This appeal to the need for an empirical grounding manages to exclude all the more fanciful putatively coherent sets of proposi- tions from our reckoning. Thus, for instance, a perfect expansion of the Sherlock Holmes stories would not have to be counted as a true description of the world, despite its coherence.
But unfor- tunately, even when we have ruled out all such coherent sets, there will be more than one remaining. For nothing in the appeal to the need to order the data of experience can make it the case that there need be one most systematic ordering.
This is, after all, just what the underdetermination of theory by evidence amounts to. So the plurality objection still has teeth. The right defence against the plurality objection is offence. We should ask whether there is any other theory of truth, any other account of what truth is, which fares better.
It emerges quickly that none of the standard theories of truth have the desired consequence that there can only be one set of truths. Certainly the traditional opponent of the coherence theory, the correspondence theory, faces the same difficulties.
Correspondence theories try to erect an account of truth upon the undeniable remark that for a proposition to be true is for it to fit the facts. But as long as facts and true proposi- tions are kept separate from each other, what is there to prevent there being two distinct sets of propositions which "fit the facts" equally well? But perhaps the plurality objection still has a point. After all, the coherentist must admit that the competing theories are all true since they are all equally coherent , while the correspondence theorist can say that one is true and the others false.
The correspon- dence theorist has this advantage because he says that there is something beyond and distinct from the competing theories, the world, which can make it the case that one is true and the rest false.
Introduction to Contemporary Epistemology
JONATHAN DANCY AN INTRODUCTION TO CONTEMPORARY EPISTEMOLOGY PDF