son; and that his followers, ancient and mediæval, so understood him, and intended the same limitation." In recent times, the word cause becoming used almost exclusively for the causa essendi, logicians have commonly mistaken his meaning and wrongly interpreted this sophism. They define the fallacy to be the assumption without sufficient ground that one thing is the cause (causa essendi) of another. Thus, that a change in the moon is the cause of a change in the weather;— thirteen at table brings bad luck;-the dog-star, Sirius, causes the heat that prevails during his ascension." Whitefield once attributed his being overtaken by a hail-storm to his not having preached at the last town. Since many a nation having a heavy debt has prospered, therefore a national debt is a national blessing. These are clearly instances of the fallacy Post hoc ergo propter hoc, or of Cum hoc ergo propter hoc." This fallacy is merely a case of bad generalization or bad induction, and therefore, however important it may be, has no proper place in Deductive Logic. But by our recent writers it is declared to be strictly the non causa pro causa, and is introduced and exclusively discussed in this place and under this title. Now it is not only an entire deviation from the meaning of Aristotle and the scholastics thus to interpret the non causa pro causa, but also a logical blunder to include the inductive post hoc among the deductive fallacies. On the other hand, however lightly Aristotle's non causa pro causa may be esteemed, it clearly belongs to the deductive fallacies; its formal vice, since it has no middle term, being that it is quaternio terminorum.

Next to the restriction of the word cause in usus loquendi, the error was probably due secondarily to the influence of Arnauld and Aldrich, or, at least, was thereby confirmed. The former says, "The non causa pro causa is very common, and we fall into it through ignorance of the true causes of things. It is in this way that philosophers have attributed a thousand effects to nature's abhorrence of a vacuum; for instance, that vessels full of water break when it freezes, because the water then contracts, and thus leaves a vacuum, which nature cannot endure;" and so on, through a variety of illustrations."

42 airiov is fairly rendered "cause," but has the general sense of "that which is chargeable with a thing;" mostly the bad sense of "something blamable." 43 See Virgil, En. x, 273.

44 Says Cicero, "Causa ea est quæ id efficit cujus est causa. Non sic causa intelligi debet, ut, quod cuique antecedat, id ei causa sit, sed quod cuique efficienter antecedat."

45 Port-Royal Logic, pp. 251-56.

9946

Aldrich designates it "Fallacia a non causa pro causa; sive sit a non vera pro vera; sive a non tali pro tali: ut, Cometa fulsit; ergo Bellum erit. Nullo modo; nam si fuerit, aliis de causis futurum est. Hæc fallacia bene solvitur negando causam falsam; melius adducendo germanam.' Whately, under the influence mainly of Aldrich, is evidently at fault. He first accepts his mistaken view, and illustrates it. Then, dissatisfied, he guesses correctly the blunder, that logicians were confounding cause and reason; and proposes to substitute the title "Fallacy of Undue Assumption," remarking that the varieties of this are infinite." Verily; for this is merely to reason from a false premise, suppressed or disguised in any way. But such is not a logical fallacy at all, for Logic has nothing to do with the falsity of the premises. De Morgan treats the non causa pro causa very gingerly. He says, "It is the mistake of imagining necessary connection where there is none, in the way of cause, considered in the widest sense of the word." This is wide enough, truly, and might include both the right and the wrong. But his examples show that he takes the wrong view only. For instance, he quotes the statement that Saunderson had such a profound knowledge of music that he could distinguish the fifth part of a note; and then remarks, "The one who made this statement did not know, first, that any person who cannot distinguish less than the fifth part of a note to begin with, if he exhibit the least intention of learning any musical instrument in which intonation depends upon the ear, should be promptly bound over to keep the peace; and, secondly, that if Saunderson were not so gifted by nature, knowledge of music would no more have supplied the defect than knowledge of optics would give him sight." These remarks show that he had only the causa essendi in mind; for he therein denies the assumption that knowledge of music was the efficient cause of the discrimination. And so our recent English logicians generally.“,

19 48

46 Logic, Appendix, § 4, 4.

48 Formal Logic, p. 268.

47 Whately's Logic, pp. 223-33.

Hamilton, following

40 Bain makes the mistake (Logic, p. 626 and p. 675). Krug, misstates the meaning of non causa, and treats the mistaken view as a deductive fallacy. He also wrongly puts post hoc among the deductive fallacies (Logic, pp. 237-39). Mill does not use the title non causa pro causa, and omits to notice the Aristotelic species. He puts the post hoc in its appropriate place among false inductions. (See Logic, bk. v, ch. v, on “Fallacies of Generalization.") Minor writers, all that I have examined, and they are many, blunder along with passive sequacity.

E.

g.,

Was

§ 7. The seventh class is Plurium interrogationum (rò rà #λeiw ¿pwτýμara ev tolεiv), the sophism of many questions. It is the effort to get a single answer to several questions asked in one. Pisistratus the tyrant and scourge of Athens? As he was the one, but not the other, either a yea or a nay would commit the respondent to a false position. A variation is to ask a single question, indeed, but so stated or compounded that a simple answer will assert or deny some other implied proposition. E. g., Did you take anything when you broke into my house last night? Are you the only rogue in your family? Have you quit drinking?" Have you cast your horns? From this last ancient example, the sophism is sometimes called the Cornutus. "Several questions put as one should be met at once by the decomposition of the compound question into its elements."" Obviously; as in the following example," which has long served as the standard illustration: "Menedemus, Alexino rogante, Numquid patrem verberare desiisset? inquit, Nec verberavi, nec desii." So the Royal Society savans at last solved the waggish query of Charles II: Why does not a live fish add to the weight of a bowl of water, as a dead one does? This implies two questions, which for a time the puzzled philosophers overlooked, viz. 1st, An sit? 2d, Cur sit?"

All this seems quite frivolous. The occasion for noting the sophism is to be found in the eristic method of dialectic disputation among the Greeks, which proceeds usually by question and answer, the answers being conventionally yea or nay,"-a method familiar to readers of Plato's Dialogues. The effort of the Sophist is to entrap his unwary respondent into an admission which can be turned against him as paradoxical. The following example, borrowed from Fries," is attributed, in its original form, by Diogenes Laertius (vii, § 196), to Eublides the Megarian as the inventor:

Have you lost ten counters ?-No.

Must you not have lost what you had at the beginning of the game and have not now?-Yes.

Have you ten counters now?-No.

Then you have lost ten counters, and have contradicted yourself.

But he had lost only two of the ten counters, and still had eight.

60 See Part 3d, i, § 12.

62 Originally from Diogenes Laertius, ii, 135.

51 De Soph. xxx.

53 See the hackneyed story at length in Hamilton's Metaphysics, p. 118.

"See De Soph. xvii; and Diog. Laert. bk. ii, ch. 18, § 135.

It is perhaps worthy of remark that lawyers sometimes nowadays badger unsophisticated witnesses in this way. To some compound question they demand what they call "a categorical answer," by which they mean a simple yea or nay, when either answer will entrap the witness in a self-contradiction or in other falsity. To deny the possession of a whole is not to deny the possession of a part, as in the above example and in the case of the stolen ham. To admit the existence of a certain motive (e. g., one mercenary) for an action still leaves the question undecided as to the concurrence of perhaps many other motives, and says nothing of their comparative strength.

Every question containing an ambiguous term may be viewed as double. Cicero is much puzzled to answer the question whether anything vicious is expedient." Expedient may be understood either as conducive to temporal welfare or as conducive to ultimate welfare. If the answer, in view of the latter meaning be Nay, an opponent may confute with the former meaning, saying, “But theft is certainly vicious, yet, as it may conduce to temporal welfare, it is sometimes expedient." Or if the answer, in view of the former meaning, be Yea, he may object, "But no vice can ever conduce to ultimate good, therefore nothing vicious is ever expedient."

The double question may often be construed as an incomplete, and hence false, disjunction. Thus the Cornutus may be stated, "Either you have cast your horns, or you have them still; which?" But there is a third horn omitted, i. e., or you have never had horns at all." In this form it is merely a case of a false premise.

66

The thirteen Aristotelic sophisms are comprised in the following mnemonic hexameters:

Equivocat. Amphi. Componit, Dividit, Acc. Fi.

Acci. Quid, Ignorans, Non Causa, Con. Petit. Interr.

The non causa is displaced here from the original order which is the one we have followed.

b De Off. bk. iii.

IV. EXAMPLES.

§ 1. Logic, from the time of Aristotle, became among the Grecks a profession. The acute and fun-loving Athenians especially busied themselves to invent puzzles with which to entangle and deride the stately professors; and these worthies themselves used the same means to discredit their rivals. Many of these puzzles, together with similar inventions by the scholastic logicians, have been handed down the centuries to us, discussed at every turn. As satisfactory solutions were rare, they received the title of "Inexplicabiles Rationes." They were collected, mostly from Diogenes Laertius, by Gassendi, in his Liber de Origine et Varietate Logica, and are analytically reviewed by Hegel.' Appearing generally to be a mere play of wit and acuteness, we marvel at the interest they have excited, at their celebrity, and at the importance attached to them by some of the most distinguished thinkers of antiquity. They certainly have an historical interest; and as literature makes frequent references to them, the student of Logic cannot neglect to make their acquaintance.

The disguises which sophistry may assume are innumerable. It seems to lurk most securely in the conditional forms, for these, being often very intricate, are confusing. Perhaps the most complete disguise is the dilemma, which, from its great capacity for entangled statement, was the favorite form of the Sophists, and hence is always regarded with suspicion and distrust. In some cases, however, very simple forms have proved very troublesome. We will select and examine a few of the most noted of the Inexplicables. They are known by specific names derived generally from the matter to which they were originally applied.

§ 2. The Achilles was proposed by Zeno the Eleatic, to support the leading tenet of Parmenides, the unity of all things, by showing that the identity of rest and motion is a necessary result of the contrary opinion. Probably, however, he was not serious in this argument, but intended it to retort the ridicule which had been thrown on

1See Gesch. der Philos., Werke, xvi, p. 119 sq.