Quinctilian, and other rhetorical writers, have been much indebted to the Topics of Aristotle.

He was the first, as far as I know, who made an attempt of this kind: and in this he acted up to the magnanimity of his own genius, and that of ancient philosophy. Every subject of human thought had been reduced to ten categories; very thing that can be attributed to any subject, to five predicables: he attempted to reduce all the forms of reasoning to fixed rules of figure and mode, and to reduce all the topics of argumentation under certain heads; and by that means to collect as it were into one store all that can be said on one side or the other of every question, and to provide a grand arsenal, from which all future combatants might be furnished with arms offensive and defensive in every cause, so as to leave no room to future generations to invent any thing

new.

The last book of the Topics is a code of the laws according to which a syllogistical disputation ought to be managed, both on the part of the assailant and defendant. From which it is evident, that this philosopher trained his disciples to contend, not for truth merely, but for victory.

SECT. 3. Of the book concerning Sophisms. A syllogism which leads to a false conclusion, must be vicious, either in matter or form: for from true principles nothing but truth can be

justly

justly deduced. If the matter be faulty, that is if either of the premises be false, that premise must be denied by the defendant. If the form be faulty, some rule of syllogism is transgressed; and it is the part of the defender to shew, what general or special rule it is that is transgressed. So that, if he be an able logician, he will be impregnable in the defence of truth, and may resist all the attacks of the sophist. But as there are syllogisms which may seem to be perfect both in matter and form, when they are not really so, as a piece of money may seem to be good coin when it is adulterate; such fallacious syllogisms are considered in this treatise, in order to make a defendant more expert in the use of his defensive weapons.

And here the author, with his usual magnanimity, attempts to bring all the fallacies that can enter into a syllogism under thirteen heads; of which six lie in the diction or language, and seven not in the diction...

The fallacies in diction are, 1. When an ambiguous word is taken at one time in one sense, and at another time in another. 2. When an ambiguous phrase is taken in the same manner. 3. and 4. are ambiguities in syntax; when words are conjoined in syntax that ought to be disjoined; or disjoined when they ought to be conjoined, 5. Is an ambiguity in prosody, accent, or pronunciation. 6. An ambiguity arising from some figure of speech.

When

When a sophism of any of these kinds is translated into another language, or even rendered into unambiguous expressions in the same language, the fallacy is evident, and the syllogism appears to have four terms.

The seven fallacies which are said not to be in the diction, but in the thing, have their proper names in Greek and in Latin, by which they are distinguished. Without minding their names, we shall give a brief account of their nature.

1. The first is, Taking an accidental conjunction of things for a natural or necessary connection: as, when from an accident we infer a property; when from an example we infer a rule; when from a single act we infer a habit.

2. Taking that absolutely which ought to be taken comparatively, or with a certain limitation. The construction of language often leads into this fallacy for in all languages, it is common to use absolute terms to signify things that carry in them some secret comparison; or to use unlimited terms, to signify what from its nature must be limited.

3. Taking that for the cause of a thing which is only an occasion, or concomitant.

4. Begging the question. This is done, when the thing to be proved, or some thing equivalent, is assumed in the premises.

5. Mistaking the question. When the conclusion of the syllogism is not the thing that ought F

VOL. III.

to

to be proved, but something else that is mistaken for it.

6. When that which is not a consequence is mistaken for a consequence; as if, because all Africans are black, it were taken for granted that all blacks are Africans.

7. The last fallacy lies in propositions that are complex, and imply two affirmations, whereof one may be true, and the other false; so that whether you grant the proposition, or deny it, you are entangled: as when it is affirmed, that such a man has left off playing the fool. If it be granted, it implies that he did play the fool formerly. If it be denied, it implies, or seems to imply, that he plays the fool still.

In this enumeration, we ought, in justice to Aristotle, to expect only the fallacies incident to categorical syllogisms. And I do not find, that the logicians have made any additions to it when taken in this view; although they have given some other fallacies that are incident to syllogisms of the hypothetical kind, particularly the fallacy of an incomplete enumeration in disjunctive syllogisms and dilemmas.

The different species of sophisms above mentioned are not so precisely defined by Aristotle, or by subsequent logicians, but that they allow of great latitude in the application; and it is often dubious under what particular species a sophistical syllogism ought to be classed. We even find the same example brought under one species by one author,

author, and under another species by another. Nay, what is more strange, Aristotle himself employs a long chapter in proving by a particular induction, that all the seven may be brought under that which we have called mistaking the question, and which is commonly called ignoratio elenchi. And indeed the proof of this is easy, without that laborious detail which Aristotle uses for the purpose: for if you lop off from the conclusion of a sophistical syllogism all that is not supported by the premises, the conclusion, in that case, will always be found different from that which ought to have been proved; and so it falls under the ignoratio elenchi.

It was probably Aristotle's aim, to reduce all the possible variety of sophisms, as he had attempted to do of just syllogisms, to certain definite species but he seems to be sensible that he had fallen short in this last attempt. When a genus is properly divided into its species, the species should not only, when taken together, exhaust the whole genus; but every species should have its own precinct so accurately defined, that one shall not encroach upon another. And when an individual can be said to belong to two or three different species, the division is imperfect; yet this is the case of Aristotle's division of the sophisms, by his own acknowledgment. It ought not therefore to be taken for a division strictly logical. It may rather be compared to the several species or forms of action invented in law for the redress of wrongs.