There may, that is, be many metaphysical problems which bear little more in common than that they are regarded as metaphysical puzzles. To this extent, "metaphysical" may be like "interesting" or "popular" or "taboo", i.e. the term may describe something extrinsic, a way of our regarding the thing described, rather than any intrinsic feature of the thing itself. It may well be that there is no way other than by giving examples to explain what is to be regarded as a metaphysical puzzle.
If this is true, there should be no cause for alarm. For if this is true, metaphysics is no worse off in this regard than is, for example, mathematics. There is no single determinate feature, other than tradition, which makes some puzzle or some technique a mathematical one. When certain persons at the end of the sixteenth century set their minds to developing what has come to be known as algebra, many mathematicians did not know what to make of the newly developing techniques and body of knowledge ([112], 122-6). Was algebra, or was it not, mathematics? Or, again, at the turn of the twentieth century, as set theory was being developed at the hands of a few mathematicians, it was being roundly condemned by others ([112], 203-4). Was set theory a genuine part of mathematics or not? Mathematicians, without of course ever taking a vote on the question, but rather just by their accepting and using algebra and set theory, collectively decided (not discovered!) that these new techniques and their attendant concepts were to be regarded as mathematical. Similarly today there is a debate among physical anthropologists whether 'forensic archaeology' is a bona fide discipline alongside forensic anthropology.9 There is no court of appeal to address one's questions to, to settle such disciplinary {page 19} disputes. Nowhere is it authoritatively written what is, and what is not, to count as being mathematical; nowhere is it written what is, and what is not, to count as falling within the sphere of forensic anthropology. And – to make the point I am driving at – nowhere is it, nor could it be, authoritatively written what is, and what is not, to count as being a metaphysical problem.
This is not to say, however, that anything and everything is eligible for being regarded as being a metaphysical problem. No more so than is everything eligible for being regarded as being a mathematical or an anthropological problem.
To what extent, then, can we say what a metaphysical problem is, or put another way, what metaphysics itself is? There is no simple answer. The scope of metaphysics changes somewhat from generation to generation (remember the quotation [p. 6] from Collingwood, speaking of history); it may even change from philosopher to philosopher. I think it would be foolhardy to attempt to give anything like a definition or some formula whose application would give a verdict: "Yes, this is a metaphysical problem"; or, "No, this is not a metaphysical problem". To learn what metaphysics is, or better, what sorts of problems philosophers regard as being metaphysical problems, one should look into a variety of philosophical books. And in doing that, one will quickly discover that a great many, remarkably diverse, problems are regarded as being metaphysical ones.
This essential vagueness must be terribly unsatisfactory for the newcomer.