COLLINGWOOD’S PECULIARITY
Such a lone explorer was R. G. Collingwood. One reason why his explorations of the logic of questions and answers were neglected is that they were unrelated to orthodox formal logic and seemed - at least to him - to be in conflict with formal logic.
He explicitly suggested, in open conflict with all other logicians, that the meaning of a statement is not a constant proposition but a variable which depends on the question it comes to answer. Even the truth-value of a statement, he was bold enough to assert, can vary from question to question. Nowadays we can go further. The meaning of a question, David Harrah now contends, itself depends on our total background-knowledge. The idea of dependence on total background-knowledge is Popper’s. In his memorable review of the works of Harrah and Belnap, in the Australasian Journal of Philosophy of 1964, Hamblin, himself a former pupil of Popper’s, suggests that the whole of Harrah’s use of information theory is thereby swayed from the classical Shannonian use closer to that of Kemeny and Carnap, and nearly arrives at Popper’s position.Now, if Collingwood is right, then, to some measure, the meaning of an answer depends on the question it answers; and if there is the smallest measure of truth in Harrah’s claim that the meaning of a question depends on background-knowledge, then, not only information theory, but also logic, has to alter quite radically. What is amazing, though, is that even so, much of the logic of questions can be developed within the old system - contrary to Collingwood’s expectations. (Even much of the relativisation of answers to different bodies of background-knowledge has been developed quite classically. See Aqvist, op. cit., final section.)
The question remains very simple: is there a compelling reason to accept Collingwood’s claim (with some modifications)? The authoritative assessment of Collingwood, unfortunately, is that of Alan Donagan, as expressed in his comprehensive The Later Philosophy of R.
G. Collingwood (1962). Donagan is doubtless right in his criticism of Collingwood. The criticism is very similar to what goes on between present-day erotetic logicians. No matter what Harrah or Belnap think of Aqvist’s criticisms, these are often devastating, no one in his senses will consider them the touchstone for erotetic logic. This is perhaps the chief difference between a loner and a team: a criticism which in a team is a stimulant for further work permits us to ignore the loner in good conscience. But our real reason for ignoring the loner is that we do not share his preoccupations, that we are left cold by his questions and quests.Collingwood’s questions and quests, however, refused to lie dormant for long. They kept cropping up, and even in diverse places. Whenever an attempt was made to integrate our background-knowledge (to use Bunge’s idiom), relating questions or interests, Collingwood’s questions came close to the surface. I shall not dwell on these topics, each being sufficient for a separate study. Let me only mention two conspicuous items. The one is Chomsky’s theory of ambiguity: every sentence is ambiguous and may be read differently depending upon where we put the emphasis, and the resolution of the ambiguity depends on context. The second, or perhaps even the same, idea comes from Gestalt psychology, just as a picture may look different depending on the observer’s frame of mind, the viewpoint from which he approaches the picture, so a sentence may have different meaning depending on the question preceding it.
These are most important and glaring instances. But chiefly we have, philosophically, discovered that research is the pursuit of questions, some fruitful, others not. This is another topic which Collingwood thought important and tried to interest people in, to no avail. (Even Donagan, in 1962, failed to sift the grain from the chaff.)
Briefly, Collingwood believed that science is certitude, and so he believed in induction. He also believed that science consists in putting questions to nature. The combination of these two ideas is what he has called the Baconian method - quite incorrectly, since it is Whewell’s, not Bacon’s, methodology which Collingwood was following. So Collingwood rightly asked nature good questions and put his hypotheses to test, but wrongly went on to verify his answers. Hence, his colleagues concluded, and Donagan concurs (p. 200), Collingwood’s method is defective. Hence, concludes Donagan, Collingwood’s view of questions in science is erroneous. It is particularly here that Popper’s theory offers a tremendous boost to Collingwood.
VII.