the readiest means of detecting the erroneous reasonings of others. For the illustration of this remark I must refer to Dr. Campbell's Philosophy of Rhetoric,* where the reader will find it fully confirmed by a train of most ingenious and refined reasoning. These associations must, of necessity, be much stronger in a language which is tied down to an analogous construction, than in one where a transpositive construction is admitted; † and it is owing to this, that we are much more easily imposed on by nonsense in Latin than in English, although we may understand both languages equally well.

Beside these considerations, it might be easily shown, that the genius of the ancient languages occasioned many more ambiguities of meaning than occur in the modern ones. In confirmation of this remark, some judicious observations are made in an Essay by the late

Vol. II. p. 93.

†The Abbé Girard was the first, according to Court de Gebelin, who introduced these two distinguishing epithets; and as the use of them has been sanctioned very generally by later French Grammarians, and as I can think of no others that appear to me to be less exceptionable, I shall continue to employ them. M. Du Marsais,

in his Treatise de la Construction Grammaticale, substitutes, instead of the epithet analogous, the word simple, or natural.* Gebelin objects to the language used both by Girard and Du Marsais, as prejudging a question which he considers as problematical, and substitutes two epithets of his own, (construction libre and construction locale,) which, in my opinion, have no advantage over them. As his criticisms, however, are always entitled to respect, I shall transcribe them in his own words.

"En donnant à la construction Françoise ou à celle de telle autre langue que ce soit, le nom d'analogue, on suppose qu'elle a plus d'analogie, de conformité, de rapport avec la nature, et qu'elle est la construction la plus parfaite et en donnant à la construction Greque et Latine le nom de transpositive, on fait entendre que celle-ci intervertit l'arrangement naturel des mots, qu'elle donne lieu à un ordre opposé à celui de la nature. On suppose encore par-là, que la nature a un ordre fixe qui lui est propre, et dont elle ne peut jamais s'écarter; qu'elle est déterminée invinciblement à suivre la même route.

"Mais ces questions ont elles été décidées ? Pouvoient elles l'être, du moins dans le tems où l'on commença à donner ces noms tranchans? ne précipita-t-on pas son jugement, d'après la différence qu'on voyoit entre ces deux sortes de constructions? et ces noms ne pouvoient-ils pas induire en erreur, en persuadant qu'en effet le Latin renversoit l'ordre de la nature auquel se soumettoient nos langues modernes ?" Tom. II. pp. 501, 502.

In answer to these objections, I have only to refer the reader to the distinction pointed out in p. 41, betwen the ordinary arrangement of words in common conversation, and the deranged collocation in rhetorical and poetical composition. In the former case, (for example, in the phrase fructum da mihi, or give me fruit) I admit there is room for disputes which may not be easily settled; but in the latter, I cannot see the possibility of any. Nobody surely can imagine the structure of one of Cicero's oratorical periods to be as natural as that of a sentence of Addison or Voltaire.

*Court de Gebelin, Tom. II. pp. 511, 512.

Professor Arthur of Glasgow ;.* whose remarks, added to those already stated, seem to authorize the general conclusion, that if, in respect of conciseness, of harmony, and of impressive arrangement, the modern tongues must yield to the ancient-in other respects, and those of far greater moment, they possess a decided superiority.

I shall conclude this subject with observing, that the modern compounded languages, though more easily acquired, furnish more difficult subjects of discussion to the universal grammarian than original languages. The difference between their structure, and that of the ancient tongues, has had a great effect in turning the attention of philosophers to grammatical disquisitions, and in this manner has contributed considerably, in the present age, to the improvement of the philosophy of the human mind.

A German gentleman, well known in the learned world, who did me the honor, more than twenty years ago, to attend some of my lectures in the University of Edinburgh, having heard one of them, in which I gave a general account of this dissertation of Mr. Smith, was so kind as to favor me in a Letter with some strictures, which appear to me unquestionably just, on the latter part of Mr. Smith's essay. "In comparing," he observes, "the ancient and modern languages, Mr. Smith ought to have expressed himself under certain limitations with regard to the latter. For the genius of the 'modern languages,' if we comprehend, under this title, those existing among the civilized nations of Europe, is very different. The German, for instance, has several striking peculiarities, which, in the strongest manner, distinguish it from others. It is, in some respects, more complicated in point of grammatical structure than the Greek or the Latin: but the most remarkable characteristic is the arrangement of words; which, though wide

Essay on the Arrangement of Ancient and Modern Language-See Arthur's Discourses on various subjects. Glasgow, 1803.

† Dr. Noehden, of the university of Goettingen, author of that highly esteemed work, entitled, a Grammar of the German Language for the use of Englishmen.

ly different from the natural order of construction, is yet limited and determined by certain rules.

"The artificial arrangement of the parts of speech in the German language is not unworthy the attention of a philosopher it is perhaps a disadvantage in philosophical inquiries, and it might be suggested with some plausibility, that the obscurity of Kant's system is, in some degree, to be attributed to the language in which he wrote; though I am by no means decided as to this point. So much is certain, that Plattner, an eminent philosopher in Germany, conceived that artificial order of placing the parts of speech to be unfavorable to the purpose of philosophy; and that he gave a determined preference to a natural collocation of words. He went so far as to attempt to introduce the latter in opposition to the general established practice. But this is in the highest degree contrary to the habits of the people of Germany, insomuch so, that his books in which the natural arrangement of words is adopted, appear hardly legible. I have often turned from them with displeasure, and even disgust: and found it a greater labor to read and understand him, than more difficult subjects would have given me, if delivered in the usual form of arrangement."-The reader will find the subject farther prosecuted in the second edition of Dr. Noehden's Grammar.*

It is scarcely necessary to add, that this criticism of Dr. Noehden's is not meant to invalidate Mr. Smith's argument, but to suggest some necessary limitations of the terms in which it has been announced by the author. It tends, on the contrary, powerfully to support Mr. Smith's speculations; inasmuch as the German or Teutonic, falling obviously under Mr. Smith's idea of an original language, might be expected to differ in its construction from the Romanic tongues, as well as from the English, which, though it has Teutonic for its basis, has subsequently admitted largely into its composition Norman-French-itself a mixture of Latin, with the Celtic and Teutonic.

VOL. III.

*London, printed for Mawman, 1807, p. 429.

7

SECTION III.

Of Language considered as an Instrument of Thought.

ANOTHER view of language, intimately connected with the Philosophy of the Human Mind, has for its object to illustrate the functions of words considered as the great instrument of thought and of solitary speculation. In the importance of its practical applications this may justly claim the first place among the various branches of our present subject. Indeed, I do not think I should go too far, were I to assert, that if a system of rational logic should ever be executed by a competent hand, this will form the most important chapter of such a work. All, however, that I have to offer with respect to it is already exhausted in the course of my former publications; and as I am unwilling to tire my reader with repetitions, I shall here content myself with referring in a note to those passages in my works where it has happened to fall under my consideration.*

When I published my former volumes, I had not seen the ingenious Essay of Michaelis on the Influence of Opinions on Language, and of Language on Opinions.† The title is imposing, and strongly excited my curiosity; and the performance itself, though it scarcely answered the expectations I had formed of it from the great reputation of the author, may be justly regarded as an acquisition of some value to the Philosophy of the Human Mind. I was sorry, when I first read it, to find that few, if any illustrations, were taken from this branch of science, which certainly presents to a philosopher the most interesting and important of all exemplifications of

Vol. II. p.

* See Elements, &c. Vol. I. 6th edition, p. 197, et seq. pp. 412, 413. I, et seq. p. 242, et seq. Phil. Essays, 3d edition, p. 147. et seq. p. 201, et seq. p. 207, et seq. p. 226, et seq. p. 232, et seq.

† An English translation of this Essay was published at London in the year 1771, by Johnson, in St. Paul's Church-yard; but I never happened to hear of it till very lately, when a copy of it was kindly communicated to me by a friend. I had previously read a French translation. which appears to me to convey the sense of the author more clearly than the English one. The latter, however, (which we are told in the preface was revised in manuscript by the author,) is enriched with an Inquiry (by Michaelis) into the Advantages and Practicability of a Universal Learned Language, which contains some very acute and important observations.

the mutual influence which language and opinions have on each other; but, on reflection, I was led to indulge a hope, that the illustrations borrowed from sciences relating to the material world, will be turned to good account by the logicians who cultivate the science of mind; for nothing can be more evident than this, that all the conclusions of the author concerning the errors produced by the abuse of words in such sciences as botany and the other branches of natural history, must hold a fortiori in all those speculations which have the mental phenomena for their object. As this, however, is an inference not likely to occur to ordinary readers, the subject may be considered as still open to future inquirers, who, after all that has yet been said upon it, will find an ample field for original remarks, as well as for new strictures on the reasonings of their predecessors. It is a topic, indeed, which cannot be pressed too often upon the attention of philosophical students.

With the importance of this last subject, considered as a branch of logic, I am so strongly impressed, that I once intended to have brought together, and repeated in this place, the different passages from my former publications above referred to. But the dread of being tedious, has induced me to relinquish this design. Two passages alone I beg leave to transcribe, partly as they originally appeared in a different work, and may not, therefore, be known to all my readers; but chiefly as they contain some practical suggestions, of the utility of which I have long had experience. They appear to me, therefore, on both accounts, to have a claim to a place in these Elements.

"In speaking of the faculty of memory, (and the same observation may be extended to our other mental powers,) every body must have remarked, how numerous and how incongruous are the similitudes involved in our expressions. At one time, we liken it to a receptacle, in which the images of things are treasured up in a certain order; at another time, we fancy it to resemble a tablet, on which these images are stamped, more or less deeply; on other occasions, again, we seem to consider it as something analogous to the canvass of the