practice, so common in the ancient republics, of appointing magistrates by lot.

There is, however, this important differ- Fictitious ence; that a fictitious case which has not cases must this intrinsic probability, has absolutely no be probable. weight whatever; so that of course such arguments might be multiplied to any amount without the smallest effect: whereas any matter of fact which is well established, however unaccountable it may seem, has some degree of weight in reference to a parallel case; and a sufficient number of such arguments may fairly establish a general rule, even though we may be unable, after all, to account for the alleged fact in any of the instances. E. G. no satisfactory reason has yet been assigned for a connexion between the absence of upper cutting teeth, or of the presence of horns, and rumination; but the instances are so numerous and constant of this connexion, that no Naturalist would hesitate, if, on examination of a new species, he found those teeth absent, and the head horned, to pronounce the animal a ruminant. Whereas, on the other hand, the fable of the countryman who obtained from Jupiter the regulation of the weather, and in consequence found his crops fail, does not go one step towards proving the intended conclusion; because that consequence is a mere gratuitous assumption without any probability to support it. In fact the assumption there, is not only gratuitous, but is in direct contradiction to experience; for a gardener has, to a certain degree, the command of rain and sunshine, by the help of his watering-pots, glasses, hotbeds, and flues; and the result is not the destruction of his crops.

There is an instance of a like error in a tale of Cum. berland's, intended to prove the advantage of a public over a private education. He represents two brothers, educated on the two plans, respectively; the former turning out very well, and the latter very ill and had the whole been matter of fact, a sufficient number of

:

such instances would have had weight as an argument; but as it is a fiction, and no reason is shown why the result should be such as represented, except the supposed superiority of a public education, the Argument involves a manifest petitio principii, and resembles the appeal made, in the well-known fable, to the picture of a man conquering a lion; a result which might just as easily have been reversed, and which would have been so, had lions been painters. It is necessary, in short, to be able to maintain, either that such and such an event did actually take place, or that, under a certain hypothesis, it would be likely to take place.

On the other hand it is important to ob. Supposed cases assert serve, with respect to any imaginary case, nothing. whether introduced as an argument, or merely for the sake of explanation, that, as it is (according to what I have just said) requisite that the hypothesis should be conceivable, and that the result supposed should follow naturally from it, so, nothing more is to be required. No fact being asserted, it is not fair that any should be denied. Yet it is very common to find persons, "either out of ignorance and infirmity, or out of malice and obstinacy," joining issue on the question whether this or that ever actually took place; and representing the whole controversy as turning on the literal truth of something that had never been affirmed [See treatise on Fallacies, ch. iii. § "Irrelevant conclusion" of which this is a case.] To obviate this mistake more care must be taken than would at first sight seem necessary, to remind the hearers that you are merely supposing a case, and not asserting any fact: especially when (as it frequently happens) the supposed case is one which might actually occur, and perhaps does occur.

I can well sympathize with the contempt mingled with indignation expressed by Cicero against certain philosophers who found fault with Plato for having, in case he proposes, alluded to the fabulous ring of

Gyges, which had the virtue of making the wearer invisible. They had found out, it seems, that there never was any such ring.*

It is worth observing, that Arguments from Example, whether real or invented, are the most easily comprehended by the young and the uneducated; because they facilitate the exercise of abstraction; a power which in such hearers is usually the most imperfect. This mode of reasoning corresponds to a geometrical demonstration by means of a diagram; in which the figure placed before the learner, is an individual, employed, as he soon comes to perceive, as a sign, though not an arbitrary sign,† representing the whole class. The algebraic signs again, are arbitrary; each character not being itself an individual of the class it represents. These last therefore correspond to the abstract terms of a language.

Under the head of Invented Example, a Fable and distinction is drawn by Aristotle, between illustration. Parabolè and Logos. From the instances he gives, it is plain that the former corresponds (not to Parable, in the sense in which we use the word, derived from that of Parabole in the Sacred Writers, but) to Illustration; the latter to Fable or Tale. In the former, an allusion only is made to a case easily supposable; in the latter, a fictitious story is narrated. Thus, in his instance above cited, of Illustration, if any one, instead of a mere

* Atque hoc loco, philosophi quidam, minime mali illi quidem, sed non satis acuti, fictam et commenticiam fabulam prolatam dicunt a Platone : quasi vero ille, aut factum id esse, aut fieri potuisse defendat. Hæc est vis hujus annuli et hujus exempli, si nemo sciturus, nemo ne suspicaturus quidem sit, cum aliquid, divitiarum, potentiæ, dominationis, libidinis, caussa feceris-si id diis homini busque uturum sit semper ignotum, sisne facturus. Negant id fieri posse. Quanquam potest id quidem; sed quæro, quod negant posse, id si posset, quidnam facerent? Urgent rustice sane: negant enim posse, et in eo perstant. Hoc verbum quid valeat, non vident. Cum enim quærimus, si possint celare, quid facturi sint, non quærimus, possintne celare, &c. (Cic. de Off b. iii. c. 9.)

The words, written or spoken, of any language, are arbitrary signs; the characters of picture-writing or hieroglyphic, are natural signs.

allusion, should relate a tale, of mariners choosing a steersman by lot, and being wrecked in consequence, Aristotle would evidently have placed that under the head of Logos. The other method is of course preferable, from its brevity, whenever the allusion can be readily understood and accordingly it is common, in the case of well-known fables, to allude to, instead of narrating, them. That, e. g. of the horse and the stag, which he gives, would, in the present day, be rather alluded to than told, if we wished to dissuade a people from calling in a too powerful auxiliary. It is evident that a like distinction might have been inade in respect of historical examples; those cases which are well known, being often merely alluded to, and not recited.

Fable and tale.

The word “ fable" is at present generally limited to those fictions in which the resemblance to the matter in question is not direct, but analo. gical; the other class being called novels, tales, &c. Those resemblances are (as Dr. A. Smith has observed) the most striking, in which the things compared are of the most dissimilar nature; as is the case in what we call fables; and such accordingly are generally preferred for argumentative purposes, both from that circumstance itself, and also on account of the greater brevity which is, for that reason, not only allowed but required in them.* For a fable spun out to a great length becomes an allegory, which generally satiates and dis gusts; on the other hand, a fictitious tale, having a more direct, and therefore less striking resemblance to reality, requires that an interest in the events and persons should be created by a longer detail, without which it would be insipid. The fable of the Old Man and the Bundle of Sticks, compared with the Iliad, may serve to exemplify what has been said: the moral conveyed by each being the sane, viz., the strength acquired by union, and the weakness resulting from division; the lat

vice.

A novel or tale may be compared to a picture; a fable, to a de

ter fiction would be perfectly insipid if conveyed in a few lines; the former, in twenty-four books, insudportable.

Of the various uses, and of the real or apparent refutation, of Examples, (as well as of other arguments,) I shall treat hereafter; but it may be worth while here to observe, that I have been speaking of Example as a kind of Argument, and with a view therefore to that purpose alone; though it often happens that a resemblance, either direct, or analogical, is introduced for other purposes; viz. not to prove any thing, but either to illustrate and explain one's meaning, (which is the strict etymological use of the word Illustration,) or to amuse the fancy by ornament of language: in which case it is usually called a similie: as, for instance, when a person whose fortitude, forbearance, and other such virtues, are called forth by persecutions and afflictions, is compared to those herbs which give out their fragrance on being bruised. It is of course most important to distinguish, both in our own compositions and those of others, between these different purposes. 1 shall accordingly advert to this subject in the course of the following chapter.

CHAP. III.—Of the various use and order of the several kinds of Propositions and of Arguments in different

cases.

satisfaction

§ 1. The first rule to be observed is, that Arguments it should be considered, whether the princi- of Confuta pal object of the discourse be, to give satis- tion and of faction to a candid mind, and convey instruction to those who are ready to receive it, or to compel the assent, or silence the objections, of an opponent. For, cases may occur, in which the arguments to be employed with most effect will be different, according as it is the one or the other of these objects that