Well then: why does logic matter? I use 'logic' here in the sense of a practical observance of good logical standards: logica utens rather than logica docens, to employ the old terms. Why should we aim at valid reasoning and consistent formulation of thought? The benefit of valid reasoning is that it never leads from truth to falsehood. Of course we do not always know that our starting-point is true: we may rather be concerned to test its truth. But as soon as we reach a false conclusion, we have acquired some further knowledge: namely, that falsehood lurks somewhere in our premises. Or again: we may be arguing with an opponent, and show by valid reasoning that premises he accepts lead to conclusions he rejects. This sort of thing is unpleasant for the victim; it seems victims have clubbed together to get the manoeuvre listed as a fallacy—the ad hominem fallacy. But ad hominem argument is not as such fallacious; if it is formally valid, the argument is not even merely eristic—for, painful as the lesson may be, to learn that your present position is indefensible is a benefit. What would be a merely eristic victory would be for an opponent to use a form of argument against you which he knew to be fallacious, but could get away with because you were too inept a logician to discern the fallacy. An opponent who uses your own premises against you ad hominem is bringing you some light; one who uses your own bad logical principles against you leaves you in your original darkness.
The benefit of consistency is that you cannot be inconsistent without being wrong about something substantive; for truth cannot be inconsistent with truth. In so far as we will to think truly, as an end,we must will to think consistently, as a means; and to opt for self-deception is, in the words of the prophet, to hew out broken cisterns that will hold no water.
There is however perhaps something to be said on behalf of inconsistency: witness the so-called Paradox of the Preface. Authors very often admit in prefaces that despite revision some errors remain in the book. If there were no error in the book before, including such a preface ensures the presence of error; the whole corpus, text plus preface, is necessarily inconsistent. No paradox thus far; the odd thing is, however, that we judge the writing of such a preface to show greater wisdom than saying in the preface something like 'I thank all my kind friends who have suggested corrections, but I remain convinced that everything I have said in this book is true'—reading such words, as I have in fact done, makes one think the author a silly fellow, although unlike his more modest rival he has not run into any inconsistency. Is it then folly to be consistent, wisdom to be inconsistent? Our judgement of the matter is in fact easily explained. Given human frailty, both texts will in fact include error. The modest author recognises this, and incurs no further error by doing so; the conceited author compounds the errors of his text by denying their existence in his preface. The modest author must indeed admit that the totality, text plus preface, is inconsistent; but inconsistency matters, as I said before, just in so far as error matters; and he is not more in error for his preface than he was already—he does not fall into some peculiarly virulent form of error by being inconsistent.
(Peter Geach, "Why Logic Matters," in Contemporary British Philosophy: Personal Statements, 4th Series, ed. H. D. Lewis, Muirhead Library of Philosophy, ed. H. D. Lewis [London: George Allen & Unwin, 1976], 86-99, at 96-8 [italics in original; footnote omitted])
Keith Burgess-Jackson, J.D., Ph.D. 