PREFACE.

The present work is a revised and enlarged edition of the Greek Grammar published in 1879, which was itself a revised and enlarged edition of the Elementary Greek Grammar of only 235 pages published in 1870. I trust that no one will infer from this repeated increase in the size of the book that I attribute ever increasing importance to the study of formal grammar in school. On the contrary, the growth of the book has come from a more decided opinion that the amount of grammar which should be learned by rote is exceedingly small compared with that which every real student of the Classics must learn in a very different way. When it was thought that a pupil must first learn his Latin and Greek Grammars and then learn to read Latin and Greek, it was essential to reduce a school grammar to its least possible dimensions. Now when a more sensible system leaves most of the details of grammar to be learned by the study of special points which arise in reading or writing, the case is entirely different; and few good teachers or good students are any longer grateful for a small grammar, which must soon be discarded as the horizon widens and new questions press for an answer. The forms of a language and the essential principles of its construction must be learned in the old-fashioned way, when the memory is vigorous and retentive; but, these once mastered, the true time to teach each principle of grammar is the moment when the pupil meets with it in his studies, and no grammar which is not thus practically illustrated ever becomes a living reality to the student. But it is not enough for a learner merely to meet each construction or form in isolated instances; for he may do thie repeatedly, and yet know little of the general principle which the single example partially illustrates. Men saw apples fall and the. moon and planets roll ages before the principle of gravitation was thought of. It is necessary,

iv

therefore, not merely to bring the pupil face to face with the facts of a language by means of examples carefully selected to exhibit them, but also to refer him to a statement of the general principles which show the full meaning of the facts and their relation to other principles.¹ In other words, systematic practice in reading and writing must be supplemented from the beginning by equally systematic reference to the grammar. Mechanics are not learned by merely observing the working of levers and pulleys, nor is chemistry by watching experiments on gases; although no one would undertake to teach either without such practical illustrations. It must always be remembered that grammatical study of this kind is an essential part of classical study; and no one must be deluded by the idea that if grammar is not learned by rote it is not to be learned at all. It cannot be too strongly emphasized, that there has been no change of opinion among classical scholars about the importance of grammar as a basis of all sound classical scholarship; the only change concerns the time and manner of studying grammar and the importance to be given to different parts of the subject.

What has been said about teaching by reference and by example applies especially to syntax, the chief principles of which have always seemed to me more profitable for a pupil in the earlier years of his classical studies than the details of vowel-changes and exceptional forms which are often thought more seasonable. The study of Greek syntax, properly pursued, gives the pupil an insight into the processes of thought and the manner of expression of a highly cultivated people; and while it stimulates his own powers of thought, it teaches him habits of more careful expression by making him familiar with many forms of statement more precise than those to which he is accustomed in his own language. The Greek syntax, as it was developed and refined by the Athenians, is a most important chapter in the history of thought, and even those whose classical studies are limited to the rudiments cannot afford to neglect it entirely. For these reasons the chief increase in the present work has been made in the department of Syntax.

v

The additions made in Part I. are designed chiefly to make the principles of inflection and formation in Parts II. and III. intelligible. Beyond this it seems inexpedient for a general grammar to go. In Part II. the chief changes are in the sections on the Verb, a great part of which have been remodelled and rewritten. The paradigms and synopses of the verb are given in a new form. The nine tense systems are clearly distinguished in each synopsis, and also in the paradigms so far as is consistent with a proper distinction of the three voices. The verbs in μι are now inflected in close connection with those in ω, and both conjugations are included in the subsequent treatment. The now established Attic forms of the pluperfect active are given in the paradigms. The old makeshift known as the "connecting-vowel" has been discarded, and with no misgivings. Thirteen years ago I wrote that I did not venture " to make the first attempt at a popular statement of the tense stems with the variable vowel attachment"; and I was confirmed in this opinion by the appearance of the Schulgrammatik of G. Curtius the year previous with the "Bindevocal" in its old position. Professor F. D. Allen has since shown us that the forms of the verb can be made perfectly intelligible without this time-honored fiction. I have now adopted the familiar term "thematic vowel, " in place of "variable vowel" which I used in 1879, to designate the ο or ε added to the verb stem to form the present stem of verbs in ω. I have attempted to make the whole subject of tense stems and their inflection more clear to beginners, and at the same time to lay the venerable shade of the connecting-vowel, by the distinction of "simple and complex tense stems, " which, correspond generally to the two forms of inflection, the "simple" form (the μι form) and the "common" form (that of verbs in ω). See 557-565. I use the term " verb stem " for the stem from which the chief tenses are formed, i.e. the single stem in the first class, the "strong" stem in the second class, and the simple stem in the other classes (except the anomalous eighth). Part III. is little changed, except by additions. In the Syntax I have attempted to introduce greater simplicity with greater detail into the treatment of the Article, the Adjectives, the Cases, and the Prepositions. In the Syntax of the Verb, the changes made in my new edition of the Greek Moods and Tenses have been adopted, so far as is possible in a school-book. The independent uses of

vi

the moods are given before the dependent constructions, except in the case of wishes, where the independent optative can hardly be treated apart from the other constructions. The Potential Optative and Indicative are made more prominent as original constructions, instead of being treated merely as elliptical apodoses. The independent use of μή in Homer to express fear with a desire to avert the object feared is recognized, and also the independent use of μή and μή οὐ in cautious assertions and negations with both subjunctive and indicative, which is common in Plato. The treatment of ὥστε is entirely new; and the distinction between the infinitive with ὥστε μή and the indicative with ὥστε οὐ is explained. The use of πρίν with the infinitive and the finite moods is more accurately stated. The distinction between the Infinitive with the Article and its simple constructions without the Article is more clearly drawn, and the whole treatment of the Infinitive is improved. In the chapter on the Participle, the three classes are carefully marked, and the two uses of the Supplementary Participle in and out of oratio obliqua are distinguished. In Part V. the principal additions are the sections on dactylo-epitritic rhythms, with greater detail about other lyric verses, and the use of two complete strophes of Pindar to illustrate that poet's two most common metres. The Catalogue of Verbs has been carefully revised, and somewhat enlarged, especially in the Homeric forms.

The quantity of long α, ι, and υ is marked in Parts I., II., and III., and wherever it is important in Part V., but not in the Syntax. The examples in the Syntax and in Part V. have been referred to their sources. One of the most radical changes is the use of 1691 new sections in place of the former 302. References can now be made to most paragraphs by a single number; and although special divisions are sometimes introduced to make the connection of paragraphs clearer, these will not interfere with references to the simple sections. The evil of a want of distinction between the main paragraphs and notes has been obviated by prefixing N. to sections which would ordinarily be marked as notes. I feel that a most humble apology is due to all teachers and students who have submitted to the unpardonable confusion of paragraphs, with their divisions, subdivisions, notes, and remarks, often with (a), (b), etc., in the old edition. This arrangement was thoughtlessly adopted to preserve the numbering of sections in the Syntax

vii

of the previous edition, to which many references had already been made; but this object was gained at far too great a cost. I regret that I can make no better amends than this to those who have suffered such an infliction. A complete table of Parallel References is given in pp. xxvi.-xxxv., to make references to the former edition available for the new sections.

I have introduced into the text a section (28) on the probable ancient pronunciation of Greek. While the sounds of most of the letters are well established, on many important points our knowledge is still very unsatisfactory. With our doubts about the sounds of θ, φ, χ, and ζ, of the double ει and ου, not to speak of ξ and ψ, and with our helplessness in expressing anything like the ancient force of the three accents or the full distinction of quantity, it is safe to say that no one could now pronounce a sentence of Greek so that it would have been intelligible to Demosthenes or Plato. I therefore look upon the question of Greek Pronunciation chiefly as it concerns the means of communication between modern scholars and between teachers and pupils. I see no prospect of uniformity here, unless at some future time scholars agree to unite on the modern Greek pronunciation, with all its objectionable features. As Athens becomes more and more a centre of civilization and art, her claim to decide the question of the pronunciation of her ancient language may sometime be too strong to resist. In the meantime, I see no reason for changing the system of pronunciation² which I have followed and advocated more than thirty years, which adopts what is tolerably certain and practicable in the ancient pronunciation and leaves the rest to modern usage or to individual judgment. This has brought scholars in the United States nearer to uniformity than any other system without external authority is likely to bring them. In England the retention of the English

viii

pronunciation of Greek with Latin accents has at least the advantage of local uniformity.

Since the last edition was published, Allen's new edition of Hadley's Grammar has appeared and put all scholars under new obligations to both author and editor. The new edition of Monro's Homeric Grammar is of the greatest value to all students of Homer. Blass's new edition of the first quarter of Kühner is really a new work, abounding in valuable suggestions. From the German grammars of Koch and Kaegi I have gained many practical hints. I am also greatly indebted to many letters from teachers containing criticisms of the last edition and suggestions for making it more useful in schools, too many indeed to be acknowledged singly by name. Among them is one from which I have derived special help in the revision, a careful criticism of many parts of the book by Professor G. F. Nicolassen of Clarksville, Tennessee. Another of great value came to me without signature or address, so that I have been unable even to acknowledge it by letter. I must ask all who have thus favored me to accept this general expression of my thanks. Professor Herbert Weir Smyth of Bryn Mawr has done me the great service of reading the proofs of Parts I. and II. and aiding me by his valuable suggestions. His special knowledge of Greek morphology has been of the greatest use to me in a department in which without his aid I should often have been sorely perplexed amid conflicting views. All scholars are looking for the appearance of Professor Smyth's elaborate work on the Greek Dialects, now printing at the Clarendon Press, with great interest and hope.

WILLIAM W. GOODWIN.

Harvard University,

Cambridge, Mass., June 30, 1892.

ix