The Edinburgh Review: Or Critical Journal

ART. X.-The Greek Grammar of Frederick Thiersch, translated from the German, with brief Remarks, by Sir D. K. SANDFORD, M.A. of Christ-Church, Oxford, and Professor of Greek in the University of Glasgow. 8vo. Edinburgh: 1830.

T is very much to be regretted, we think, that grammar, or || the science which treats of words, and the various modifications they experience, should have so seldom been prosecuted in a scientific or rather philosophical spirit. On no subject of enquiry, indeed, has there been more pedantry, and less precision, more ostentatious dogmatism, and less real knowledge displayed, than that of grammar. Fine names have been invented, arbitrary rules accumulated, exceptions laboriously arrayed, gratuitous suppositions made, and unphilosophical shifts resorted to; while analogies have been overlooked, and the real causes and reasons of the peculiarities to be met with in language have remained unexplored. Nor has it often occurred to grammarians that ignorance might lie concealed under a circumlocution, and that a mere technical term, though it might express a fact, could not supply the want of explanation. If, for example, we meet with a dative case, where the laws of construction require a genitive; or a word used in a way which seems to violate the analogy of language; or certain unaccountable changes in the forms of words; our knowledge will not be much increased by merely telling us that the first is per schema colophonium, the second a catachresis, and the third a metaplasmus. These are terms expressive of nothing but ignorance, or rather they are the masks under which it is concealed. Men are too apt to delude themselves into a belief that they have discovered an explanation, when they have found only a name; to acquiesce in an established nomenclature, without considering the principles upon which it was originally formed; and thus to stop short in their enquiries at the very point where the real difficulty begins. The incurious are satisfied with superficial information, and indolence says it is enough. But this tendency, which so long obstructed the progress of grammatical investigation, has been at length overcome; the operations of the human mind have been anxiously examined, and carefully classified; while the principles of language, which are intimately connected with, and indeed vividly reflect, many of the most interesting mental phenomena, have been laid down with a precision and accuracy altogether unknown to the ancients. Hence the science of grammar generally has received a prodigious extension; and that of the Greek language, in particular, has been enlarged

and systematized with unexampled industry and success, especially by the scholars of Germany, who are ever foremost in the irace of improvement.

With respect to the older grammarians, the farther back we go, the more absurd and unreasonable do we find them. Having no fixed principles to guide their researches, they are perpetually differing from one another and from themselves; minute without accuracy, brief without conciseness, and full of distinctions which end in nothing but perplexity. The oldest complete grammar, that of Dionysius the Thracian, is contained in twenty-five short sections, occupying only fourteen octavo pages; yet small as it is, it abounds with minute and vexatious distinctions, which have been overlaid with more than three hundred pages of scholia, filled with that miserable trifling peculiar to grammatical annotators. The Erotemata of Demetrius Chalcondylas, and the Institutions of Aldus Manutius, possess little value besides that which they have acquired from their rarity; and the remains of Apollonius Dyscolus, of Choroboscus, Joannes Philoponus, Moschopulus, and others, are all, in a greater or less degree, of the same character with the short sectional treatise of Dionysius the Thracian. The Erotemata of Chrysoloras is in no respect superior to the work of Chalcondylas; and the Grammar of Constantinus Lascaris, though curious as being the first entire work printed with the Greek type, is a collection of bare rules, without illustrative expositions. A step in advance, however, was made by Henry Stephens, and his pupil Sylburgius, who introduced some improvements in the mode of treating the subject. The remarks of the latter on the Greek Grammar of Clenardus are full of learning, especially his Compend of Syntax; but although he did much towards the classification of the language, he left its grammar nearly as involved as he found it. Angelus Çaninius gave the first accurate account of the dialects, and Laurentius Rhodomannus reduced all the Greek nouns to three declensions; an improvement which was successively claimed by Weller, who introduced it in his Grammar, published in 1630, and by Claude Lancelot, the author of the Port Royal Greek Grammar; although the truth seems to be that Weller borrowed it from Rhodomannus, who mentions it in his Philomusus, and Lancelot borrowed it from Weller. The Port Royal Grammar is too well known to require almost any observation. Its nine books are subdivided into a multiplicity of detached rules, abounding in mistakes, and illustrated by examples taken from inferior writers. At the same time, Weller and Verwey made considerable progress towards simplification; and a great addition was also made to

VOL. LII. NO. CIV.

2 H

grammatical knowledge by Fischer's remarks upon Weller's treatise, which display much industry, and abound with new observations.

But Hemsterhuys far outstripped his predecessors by the boldness and originality of his views, no less than by the learn ing and sagacity with which he supported them. Availing himself of some hints thrown out by Scaliger and Vossius, and probably influenced by considerations drawn from the peculiar structure of the Oriental tongues, he was led to conclude that the primary verbs consisted of two or three letters, from which all the other forms and inflexions were derived; and that, by skilful decomposition, the root or elementary part might, in every case, be determined. Plausible arguments may be urged in favour of this etymological theory, which was received as a great discovery by Valcknaer, Rhunken, Lennep, Albert Schultens, Everard Scheide, the Bishop of St David's, and others; but notwithstanding all this weight of authority, it seems to us, we confess, to be radically unsound. Much of the Greek language is of Asiatic origin; a considerable portion of its vocabulary is pure Sanscrit; the whole of its inflexions and conjugations have been modelled upon the sacred language of India. Greek and Sanscrit answer to each other as face answers to face in a glass. But, in Sanscrit, the roots or elementary parts are of posterior formation; they are the work of grammarians alone-mere technical elements obtained by arbitrary resolution, not primary or original forms, convertible into new species of words by the artifices of inflexion and conjugation. They are not natural roots, and consequently can have had no share in the original formation of the language. They are significative by consonants alone; and for this reason differ diametrically from Greek roots, which are significative or determinable by vowels only. The theory of Hemsterhuys is therefore wholly inapplicable to all that portion of the Greek which is incontestably of Asiatic origin; and there would be no great difficulty in showing that it is equally so to the remainder. But whatever objections may be taken to the speculation of Hemsterhuys, it certainly tended to stimulate enquiry, and produced many collateral investigations of the greatest importance to the general science of grammar. In Hermann's celebrated treatise De Emendanda ratione Græcæ Grammaticæ, there is much to gratify the lovers of philosophical discussion as applied to the subject of Greek grammar; and although it may be true that he trusted too much to metaphysical principles and the universal nature of speech,' it seems at least equally so, that his example has operated powerfully on the minds of his learned countrymen, and encouraged them to undertake and

execute those valuable works on the subject which have recently appeared in Germany, and which reflect so much credit on the transcendent scholarship of that country.

According to a distinction laid down by Lord Bacon, there are two kinds of grammar, the literary and the philosophical; the former treating of the analogy of words to one another-and the latter, of the analogy subsisting between words and things. The one of these, it is obvious, may be kept quite distinct from the other, and, for mere practical purposes, it may be desirable that this distinction should be observed. But, on the other hand, it can scarcely be doubted that the most perfect grammar is that in which the literary is tempered with the philosophical, and in which, without diverging into speculative generalizations, principles are employed to connect and classify facts. This is the aid which science naturally lends to practice, and which practice repays, by contributing in its turn to extend the boundaries of science. Accordingly, most of the valuable Greek grammars which of late years have been published in Germany, appear to have been constructed on what may be called the composite principle; in other words, their authors have sought to combine philosophical views of general grammar with perspicuity of arrangement, and fulness of exemplification, in as far as regards their immediate object; at the same time endeavouring, with more or less success, to render the one subservient to the other. Of this class of works, the most distinguished are the grammars of Matthiæ, Buttmann, and Thiersch. Matthiæ is clear, full, and comprehensive; his views are generally sound, and his system of syntax is admirable. Buttmann is more profound, but less luminous and practical, than Matthiæ; and his ample erudition is not always digested in the best possible order. Thiersch is superior to both in the essential requisites of correctness and philosophical precision; while, as exhibiting an historical analysis of the Greek language, his work is perhaps unrivalled. It has indeed been said, that this work is less a grammar of the classical language as it appears in the mass of writers, than of that earlier form of it which is called the elder, the Homeric, or the epic dialect; and it is doubtless true, that the learned Professor has treated of the versification and dialect of Homer in all their varieties, with infinite care and elaboration; considering, probably, that a thorough knowledge of these is, as the translator remarks, indispensably necessary for those who desire to comprehend in their whole depth and compass, the 'Grecian tongue and literature.' But the scope of the work is by no means so confined as this observation would imply; for it is only necessary to glance at the table of contents, to see that the grammar of Professor Thiersch, like all that are mean for

elementary instruction, treats first of the common dialect-then, somewhat largely, of the Homeric dialect while all that remains to be said of the other dialects is comprised in the appendix. In the part of the work which treats of construction, and which has not yet appeared in an English dress, he has perhaps drawn his examples too exclusively from Homer: but this defect, we are assured, will be remedied in the translation, which will comprise a complete system of Greek syntax from the Homeric down to the Hellenistic dialect, and thus render the work, in its English form, the most comprehensive and valuable that has yet appeared.

We have already remarked, that philosophical precision is the distinguishing character of Professor Thiersch's grammar. As a specimen of this, we shall extract the eighty-fifth section, in which the author treats of the tenses of verbs.

yd bob

1. We consider objects either as now being, or as having been, or as hereafter to be affected by their properties, and hence divide time into three parts, the present, past, future.

2. If we consider the three times (xpovos, tempora) in relation to one another, other distinctions of time appear to attach themselves to those above enumerated, and we may, putting all together, discriminate each particular time as incomplete, complete, or about to be completed.

3. Hence we may distinguish,