12/18/11

Charles Sanders Peirce on A Definition of Logic


From C.S. Peirce, 'Chronological Edition', CE 1, pp. 169-170. Charles Sanders Peirce, "Harvard Lectures 'On the Logic of Science'", (1865),'Writings of Charles S. Peirce: A Chronological Edition, Volume 1, 1857-1866', Peirce Edition Project, Indiana University Press, Bloomington, IN, 1982.


"Some reasons having now been given for adopting the unpsychological conception of the science, let us now seek to make this conception sufficiently distinct to serve for a definition of logic. For this purpose we must bring our 'logos' from the abstract to the concrete, from the absolute to the dependent. There is no science of absolutes. The metaphysical logos is no more to us than the metaphysical soul or the metaphysical matter. To the absolute Idea or Logos, the dependent or relative 'word' corresponds. The word 'horse', is thought of as being a word though it be unwritten, unsaid, and unthought. It is true, it must be considered as having been thought; but it need not have been thought by the same mind which regards it as being a word. I can think of a word in Feejee, though I can attach no definite articulation to it, and do not guess what it would be like. Such a word, abstract but not absolute, is no more than the genus of  all symbols having the same meaning. We can also think of the higher genus which contains words of all meanings. A first approximation to a definition, then, will be that logic is the science of representations in general, whether
mental or material. This definition coincides with Locke's. It is however too wide for logic does not treat of all kinds of representations. The resemblance of a portrait to its object, for example, is not logical truth. It is necessary, therefore, to divide the genus representation according to the different ways in which it may accord with its object.

"The first and simplest kind of truth is the resemblance of a copy. It may be roughly stated to consist in a sameness of predicates. Leibniz would say that carried to its highest point, it would destroy itself by becoming identity. Whether that is true or not, all known resemblance has a limit. Hence, resemblance is always partial truth. On the other hand, no two things are so different as to resemble each other in no particular. Such a case is supposed in the proverb that Dreams go by contraries, -- an absurd notion, since concretes have no contraries. A false copy is one which claims to resemble an object which it does not resemble. But this never fully occurs, for two reasons; in the first place, the falsehood does not lie in the copy itself but in the 'claim' which is made for it, in the 'superscription' for instance; in the second place, as there must be 'some' resemblance between the copy and its object, this falsehood cannot be entire. Hence, there is no absolute truth or falsehood of copies. Now logical representations have absolute truth and falsehood as we know 'à posteriori' from the law of excluded middle. Hence, logic does not treat of copies.

"The second kind of truth, is the denotation of a sign, according to a previous convention. A child's name, for example, by a convention made at baptism, denotes that person. Signs may be plural but they cannot have genuine generality because each of the objects to which they refer must have been fixed upon by convention. It is true that we may agree that a certain sign shall denote a certain individual conception, an individual act of an individual mind, and that conception may stand for all conceptions resembling it; but in this case, the generality belongs to the 'conception' and not to the sign. Signs, therefore, in this narrow sense are not treated of in logic, because logic deals only with general terms.

"The third kind of truth or accordance of a representation with its object, is that which inheres in the very nature of the representation whether that nature be original or acquired. Such a representation I name a 'symbol'.

Cap tip Jon

RECCMENDATION Richard Gordon Quantum Touch

The Slow as Molasses Press