Campbell City (part 1 of 13)

Preface

IN bringing before the public a second edition of the first volume of this Sophocles, my thanks are in the first place due to the friends who have called my attention to some of the numerous defects and errors of the former edition; especially to the Master of Balliol, to whom I now venture to dedicate this work ; to Professor Lushington, to whom I owe more than I can estimate, and wish that I could owe still more; to my friend and collaborateur in the smaller edition, Mr. Evelyn Abbott; and to my former pupils, Mr. John Masson and Mr. W. G. Rutherford, whose interest in my labours has been of great assistance to me, and without whom this volume would have been less accurate than it is. Mr. Rutherford and Mr. A. J. Ashton have prepared the index. Another friend and former pupil, Mr. Hugh Campbell, nas re-collated Par. E. ay

To the criticisms which appeared in the Cambridge Philclogical Journal 1, in Hermathena, and in the Revue Critique, I trust I shall be found to have given the attention which was merited by the position and reputation of their authors. They contained valuable suggestions on points of detail, and also raised some general questions on which it is necessary that a few words should here be said.

1 [must own to having derived benefit sometimes tempted the remark, from Dr. Kennedy's 'Studia Sophoclea,' vate, mpadrepdv pe mpodidacke, iva although his magisterial style may have

consistent. It is now a quarter of a century since he made the same objection', though certainly with less of acrimony, against the edition of Schneidewin. And the quarrel is still older, for his criticism of the German editor in this respect may be viewed as a reply to Hermann's not ungracious remark, that Elmsley would have done better to adopt a less discursive method of annotation *.

I am not aware that Schneidewin ever made rejoinder. But the grounds of his procedure are evident. He was writing not for editors but for students, whose attention ought not to be unnecessarily distracted from their author at every turn to judge between the commentator and his predecessors.

It is at once admitted that where a conjectural emendation is adopted, the name of its first author should be given. Even this, however, is not always quite a simple duty. for in O. T. 597 was conjectured independently by Musgrave and L. Dindorf with more than half a century between them. The credit of the emendation was at one time claimed for the name of Dindorf. It is now modestly ascribed to Musgrave. His words are: ' Fuit etiam cum de utkdAdovot cogitarem: sed videbatur ea vox turpem potius assentationem, quam honestam petitionem significare.' A sentence which places the editor who adopts the conjecture in this dilemma: he must ascribe it-to one who was not the first to think of it, or to one who has declined the honour. Again, has been proposed by more than one critic as a correction of in Ant. 24; but is understood by one to mean 'Just meed of virtue, by another 'What good men approve. Here are two emendations and not one only. Suppose the sense last thought of to be

Preface. 1X

It is likewise admitted that where a fresh zzderpretation 15 so recent as to leave room for doubt as to priority of authorship, it should not be adopted without acknowledgment.

But more than this in a subject so long worn as Sophocles can hardly be demanded. Originality cannot be claimed without priority, and the proof of priority may be left to the critical reader. Beyond this point the method to be followed becomes a question of discretion. Prudence of a certain kind may suggest that contemporary opinions should have a prominence out of proportion to their merit, while the ingenuities of former generations are allowed to sleep. But a higher prudence demands that the interest of the student should be steadily kept in view. And this being so, it cannot be expected that an editor besides stating his own interpretation should distinguish it from others that are similar but not the same'. Neither ought it to be required of him that in stating a known interpretation he should always count authorities in its favour. And if not only previous editions but the chief reviews of them, and by parity of reasoning, articles in the Rheinische Museum, Philologus and the Neue Jahrbiicher, together with observations made by the historians of literature and the writers on metre, are to be laid under contribution, the central work will soon be buried under the It is at least surely permissible to edit Sophocles without doing this, and one who happens to be more versed in the text than in the literature of his author may be allowed to give in his contribution without being made to stand and deliver what he knows about other men's opinions.

An edition of Sophocles is no doubt conceivable in which the plan of Elmsley's Oedipus Coloneus should be consistently carried out. It would be a work of great labour, but, in the present day, one of doubtful utility. The list of interpreters, which in Elmsley's preface occupies a dozen closely printed lines, would probably extend to as many pages, and the commentary (290 octavo pages on the single play) would have to be lengthened in proportion.

It may be perhaps suggested to draw some line of prescription ; say, at the year 1854. But we should then only be leaving out what is less in bulk and often superior in value. Thus Dr. Kennedy has made no reference to Musgrave, who certainly does not deserve to be forgotten. Few critics have been more fruitful in suggestive hints; and, in particular, by conjecturing as an emendation of O. T. 45, he put forth a view of the context of that difficult line, which has since his time been more than once maintained.

II. My observations in the Essay on the Language of Sophocles appear to one critic paradoxical, to another commonplace. These opposite objections may be left in equipoise. It would have been better worth while to show that what is there said was either (a) false, or (6) useless.

a. Dr. Kennedy does not deny 'the unsettled and transitional condition of the Greek language in its most glorious age - that of Pericles' - but thinks that I have stated this 'with undue exaggeration.'

It is possible that in handling what seemed a neglected truth, I may have here and there spoken with extreme emphasis ; and there were certainly some crudenesses, which have been removed in revision. With much of what Dr. Kennedy says about the Common-Law of Language I agree in substance; and if the general subject had been in question, I should have enlarged upon the truth that the freedom of Greek syntax, as compared with Latin, and still more with that of any modern speech, is a pervading attribute, belonging in some degree to every stage of the language. It is not meant that Aeschylus or Sophocles broke away from existing modes of expression - there are few idioms of the tragic or any other dialect for which analogies may not be found in Homer - but that in a time when language was unusually plastic each writer had his characteristic ways of moulding it, and that the ways of Sophocles were original and very subtle. 'Common- Law, however, is a less philosophical term than ' Sprachbewusstsein' (speech-consciousness), for which Dr. Kennedy makes it an equivalent - 'Sprach-geftihl' would come nearer to what we both mean - and for this very reason, that the former term conveys the notion of something fixed and outward in the place of something that is inward and free ; it puts that which forbids in the place of that which inspires. Adopting however, for the moment, Dr. Kennedy's figure of speech, let me remark that there are periods of history (1) in which the Common-Law is definite but unwritten, others (2) in which it is rendered uncertain by the introduction of new elements, and (3) that long before a digest of it has become possible, its principles and application in the usage of the courts have been growing more and more imperatively determinate.

Now, what is meant by a fixed grammatical rule? In one sense, as Dr. Kennedy truly says, there is no such thing in Greek Literature of the Classical Period. The 'Common-Law which is Greek syntax' remained uncodified 'for more than 200 years' after the time of Sophocles. This is part of what I meant to affirm. In proof of it I adverted to the recognized difference in point of regularity between a Ciceronian period, and a sentence of any classical Greek writer!. Redundant negatives, the apposition of clauses, attraction in its various modes, the return from a dependent to an independent construction, the use of the primary conjunctive in past time, these and other irregularities can hardly be said to exist in Latin'. But

+ In the present edition I have sought to obviate the effect of a too isolated treatment, by occasionally inserting between brackets examples of corresponding idioms from other Greek writers. Ox p. 62, in 11. 11, 12, and 14, 18, the two examples from Pindar have been accidentally transposed, Students are respectfully requested to correct this error.

6. Such anomalies, although not acknowledged, exist even in published writings of the nineteenth century. Thus Dr. Kennedy can speak of 'undue exaggeration,' and one of our chief contemporary masters of English style has written or at least printed as follows : - ' No event is too extraordinary to be impossible,'

I assert more than this, and more isadmitted by Dr. Kennedy. He owns indeed that 'to a certain extent we acknowledge common principles. If the rule of speech to which Demosthenes or Xenophon conformed was free and elastic as compared with that of Ciceronian Latin, it was strict and regular when compared with that to which Aeschylus, Sophocles, and Thucydides were subject. These writers, and Pindar before them, coming at a time when the language was far less fixed than in the writings of Xenophon or Demosthenes, brought in fresh elements, which had not affected either Homer or Herodotus; and the free working of their original genius interrupted in various ways the regular growth of the language towards its final shape. The result was a 'temporis partus, which disturbed the equable flow of men's thoughts and ideas 1.

Dr. Kennedy draws a just distinction in this respect between poetry and prose. But his distinction is not wholly relevant". For granting that the tragic poet 'had a large treasure-house of epic, elegiac, iambic, and lyric poetry to draw on,' his task in producing a medium for his art was not the less a new creation, in which perfect articulation could not be reached at once. Such a result may have been even retarded by the very wealth of the material which lay at the poet's command. It is true that Aeschylus had preceded Sophocles by a few years, and had already blocked out the tragic speech ; but the creative effort in question was not exhausted in him and his predecessors, nor were the conditions under which it was continued very greatly altered. In point of fact the difference in the amount of 'idiotism' between Sophocles and the Hymn to Apollo may fairly be held analogous to that between

2 Nor is the distinction by any means so absolute as it became afterwards. Herodotus has more in common with Epic Greek on the one hand and the tragic dialect on the other than Aristotle has with any poet whatever, and it would be at least as unsafe for an interpreter of Sophocles to neglect either Herodotus or Thucydides, as to omit the study of any poet, Homer and Aeschylus excepted.

Thucydides and Herodotus. And besides the influence of earlier poetry there is also in the tragic dialect a vernacular element, which can be only very imperfectly verified - probably not the speech of the Athenian agora so much as that of other members of the Ionic race. Italian scholars are said by wandering in Tuscany to have met with living idioms that threw light upon expressions in Dante. To obtain corresponding light for Sophocles we must have had the power of wandering amongst the cities of the Delian confederacy. When we speak of the language of Sophocles then, we include the conversational parlance of his time. But to resume : -

In the Epic dialect the 'Speech-consciousness' appears as a universally acknowledged Yhemzs which no one thinks of breaking ; in the age of Pericles, it is an unwritten institute, a which, like other is found by experience to be more various and elastic than had been imagined, and to give scope for the exercise of subtlety and individual freedom. Then came the Sophists with their indispensable work, clearing the way before philosophy, and partially formulating language, before the Socratic school had formulated thought. If to the Sophists and the Socratic school combined Dr. Kennedy would ascribe the rhetorical fluency which marks the prose writers of the following century, it is not less true to say that these same teachers themselves were the exponents of a 'grammatical consciousness becoming more precise.'

It is perhaps not superfluous to remark that the object of my Essay was chiefly exegetical; viz. to account for difficulties in Sophocles by calling attention to the medium in which he worked, and to protest against the repeated attempt to make him conform to canons and external rules which are of another age. If this object is not perceived, one of two wrong impressions is apt to be formed; either that I am blaming Sophocles as a careless writer, or that I attribute to him a degree of subtlety passing into eccentricity. To do either would be of course ridiculous.

No language was ever more harmonious, more finely modulated, more exactly calculated to produce the right effect on the mind of a contemporary hearer. I was anxious to show, not that Sophocles is 'capable of saying anything 1" but that a rational account can be given of his language where it seems irregular. If we can only treat it naturally, as we would interpret the idiom of some living tongue, his way of speaking is for the most part quite intelligible even now. It is when we approach it with alien preconceptions and view it through the foreign medium of a grammar-laden consciousness, that this and much else in Greek appears crooked and obscure. We cannot doubt that the language of Shakespeare was understood perfectly by the pit at the Globe, but it was by no means an easy text for such 'correct' critics as Pope or Johnson. For us, at this distance of time, an analytical exegesis is in both cases unfortunately necessary, and we have to pull the flower to pieces before we can ascertain how it grew. Only we must bear in mind that what we are attempting to analyse is not a piece of mechanism but a work of nature®.

6. The references to the Essay which occur in the notes will show that in the application of the method as well as in the statement of it in this edition, I have, with increased caution, to a great extent held the same ground. And here a word may

pa-

be said to those who think this whole apparatus unnecessary. For great scholars and for their pupils it may well be so. Few truths about the Greek language can be otherwise than familiar to them. But if they will be at the trouble to consider the following critical note, and the theory of Sophoclean language which it implies, they will be ready to admit that my observations, even if common-place, are not altogether needless or superfluous. I quote from Nauck's edition of the Philoctetes, dated 1876, p. 143 : -

III. It has been urged as a serious objection to this edition that the Greek arguments are not reprinted here. On this subject I speak with some diffidence, knowing that the omission is felt as a defect by eminent scholars. But the point had been carefully considered.

The Greek arguments contain much valuable matter together with some things that are trivial. Precisely the same thing may be said of the ancient Scholia, and of the Anonymous Life. An edition which aims at the completeness of Boeckh's Pindar should contain all these. For an English edition conceived on the present plan, whose aim is rather that professed by Hermann in his preface to the Ajax, 'quod instituerat Erfurdtius, ut haec editio et adolescentium studiis accommodata esset, neque exclusa ab usu doctorum hominum,' it seemed more appropriate to give selections from all where they appeared most relevant, leaving the reader who wished to see any of them in their entirety to consult the

IV. The contribution offered in this volume towards the further study of the text of Sophocles is far less complete than I could have wished. My visits to the towns of Italy have (except in the case of Florence) been unavoidably hurried, and the hours per diem which I could spend in the libraries have been unfortunately limited. But in the present state of Sophoclean criticism I expect rather to be accused of having given too much space to the MSS. than too little. And indeed when the matter has been more thoroughly sifted than has yet been thought necessary, a very moderate critical apparatus may prove sufficient. Meanwhile such references to the inferior MSS. as are here given may not be without their value, if only as supplying materials for a further examination of the question of an archetype, and affording a test of that Byzantine criticism which has been thought capable of some rather notable inventions.

The classification which places the Laurentian MS. in one category and lumps together all the rest as afographa, however convenient for its simplicity, and in all probability approximately sound, at least rests on an insecure foundation so long as a considerable number of the existing MSS. remain, for any evidence that we have to the contrary, unexamined.

The following is a list of the MSS, with their supposed ages, and the Press-marks by which they are known in the various libraries. I cannot hope that this enumeration is entirely free from errors, but some pains have been taken to make it as correct as possible.

Rome

1 J am assured by Signor Bollig, Pre- 'undoubtedly of the twelfth century:' fect of the Vatican Library, that this but I have not ventured to place it so MS. is 'senza dubbio del Sec. XII,' high. It is a thin bombyx.

Cheltenham

the kindness of Signor Guidi. Vat. 1332, 920, 1363, 104. Inspected. Barberini, Minerva, Angelica. Inspected (in 186%). Naples, II. F. 34. Inspected. Ven. 468. Partially collated.

Milan

C. 24. Sup. Collation of Antigone kindly sent by Signor Antonio Ceriani, from whom I have also received, since this edition was printed, a collation of E. 103. Sup. It resembles Pal. 40.

Pal. 40. By the great liberality of Dr. Zangemeister and the other authorities I have been enabled to collate this MS. afresh in the St. Andrew's University Library *.

Oxford

Unfortunately this could not be references to this MS. are derived from done until several sheets of the text and Bothe's collation (1826). notes had been struck off. In these the

In what follows I shall attempt to show that the hypothesis of Cobet! and Dindorf?, viz. that all our other MSS. of Sophocles are derived from Laur. 32, 9 (L), is not the most probable account of the matter. To Dindorf's remark that it matters little whether we think of L, or of a MS. closely resembling L, as being the archetype, it may be replied that we want to get as near the truth as possible, and that it does matter something whether of the few good readings which make their first appearance in the thirteenth, fourteenth, or even fifteenth century, a fair proportion may be reasonably supposed authentic.

1. Allthe MSS. of Sophocles hitherto examined are proved, by corruptions common to them all, to be substantially derived irom one archetype, this dating from a time when the choric metres had been to some extent forgotten.

I say substantially derived from one, because a variant here and there may have been preserved in commonplace books and critical treatises, just as Ant. 1167, which is absent from the MSS, has been preserved to us by Athenaeus.

3. The practice of correcting MSS. from one another makes it extremely difficult to classify the MSS. of Sophocles, and the word 'family' can be used in reference to them onl: ina very modified sense.

Prep Acc Ee, Xxv

According to some recent editors, these are very numerous indeed. On the other hand, such an editor as Mr. Palmer would be inclined to defend even some of the following, which however may be taken as examples of error common to all known MSS., about which there is a general agreement. It will be enough to adduce one or two from each of the plays.

Ant. 4, 587, 966, 351, 1342, is a corruption of a gloss On or, according to Wecklein, is a gloss on Aéxpra).

The number of universal corruptions which it is found necessary, to assume has some bearing on the question of the remoteness of the archetype, which is brought nearer to L in proportion as they are believed to be more numerous, while the hypothesis of a text generally sound is more consistent with that of an earlier archetype.

Xxvi Preface

2. The immediate derivation of any MS. from L is not in question. But as a warning against rash statements it may be mentioned that in L*, besides O. T. 800, to be spoken of presently, there is a gloss 2% the text of Ant. 826, viz. 6 of which even as a gloss there is no trace in L, where the reading of the first hand, could not suggest it. Similarly the close relationship between Par. A and Ricc. 34 is modified by the misplacement in the latter of Ant. 477- 584, a passage not coinciding with a leaf of A. The same displacement occurs in the Middlehill MS. 310, which is shown by this and other peculiarities to be a descendant or cousin of Ricc. 34.

Another observation may be made zz limine by way of caution, viz. that, as the above list shows, there is a considerable variation in the order of the plays in different MSS. And in L itself, whereas the other plays are written continuously without a break, a blank is left before the Oedipus Coloneus, which begins on a separate sheet. The same occurs in L? with regard to the Trachiniae.

The reflection occurs, that the practice of 'editing Sophocles in single plays' may have begun early, and that the history of the text may be by no means the same for all the seven.

a. O. T. 800 is omitted in L and only inserted in the margin by The omission has not been quoted from any other MS, and the line is certainly present, for I have seen it, in the following, besides the MSS. ordinarily referred to: Vat. abc, VV*V°V4 3), K, Vat. 1363, Vat., M M? Trin. pr.), Ricc. 34 pr.), Pal. 40.

On this subject see Wecklein, Ars num, 1864; R. Schneider in Neue Sophoclis Emendandi, 1869; A. Seyf- Jahrbiicher for 1877; Franz Biicheler, fert, Quaestiones Criticae, Halis Saxo- Philologische Kritik, Bonn, 1878, p. 11.

line was inserted between A.D. 1000-1250; (2) that it was inserted in a descendant of L; and (3) that notwithstanding the continued existence of the archetype, this younger MS. became the sole channel of tradition.

With regard to (1), the chance of a genuine line, not containing a having been preserved in a commonplace book or learned treatise and finding its way back into the MSS., istoo slender to be entertained, where the question is one of probability. And with regard to (3), although the remark can only be applied in strictness to the MSS. which are older than the insertion in L, yet in considering probabilities it may be allowed a wider scope.

The genuineness of the line has been acknowledged by Wecklein, and more recently by Biicheler, and will hardly be disputed by any one who feels the situation. The difference between Shak. J. C. 5. 3. 1. 103, 'I shall find time, Cassius ; I shall find time,' and 'Off with his head! So much for Buckingham, is hardly a measure of the difference between this line and anything which a Tzetzes could have produced 1.

In O. T. 896, after the words we comes this manifest interpolation in L, 'The words are written as part of the text by the first hand. It has been attempted to account for them by supposing a gloss to have been altered by a slight transposition and to have crept into the text. But no such gloss is quoted, while is hardly a natural explanation of and it is more probable that in 7 there is a corruption of some other verb. However this may be, it is somewhat strange, if L is the archetype, that while 66 has remained in every MS. of Ant. 1344, is only found in five MSS., L Pal. M* M' M® (om. Not only is it absent from the others hitherto examined,

1 The omission of single lines, which _ habit of writing in double column, the eye is a not infrequent error of the copyists returning mechanically to the same part of Sophocles, may have arisen from the the page instead of crossing over.

but in several of these there is an interlinear gloss from which the corrupt may not improbably have been derived, viz. : -

Suppose for a moment that in the archetype of L Pal. M° M? M® this gloss had partly faded, might not [yupicew | be read as The gloss, being longer than the words to be explained, may have been irregularly written and hence more illegible : -

But in this case the Trinity and other MSS. which retain the words in their original form must be inferred to have an origin anterior to the special archetype of L Pal. M? M°.

c. In Trach. 1306, Par. A gives for a slight corruption, no doubt, and one which might easily have arisen after the 10th century. But it certainly existed previously. For the scribe of L had written but corrected himself in writing by erasing the letters and completing the word thus, ad... It is reasonable to suppose that he had the choice of two readings, and that the tradition which he rejected is continued in Par. A.

ad. The explanation of the true reading is sometimes retained where the reading itself is lost. In such cases it was hardly within the scope of Byzantine criticism to draw the just inference and restore the lost word. Yet the interpretation in the Scholia on Ant. 235 exactly fits the reading which is only found in a very few MSs. of the 14th century. That dedpayuévos was known to the Scholiast is admitted by Dindorf. Does he suppose that it was revived by conjecture? Cp. Schol. on Ant. 681.

e. The unmetrical reading for in O.T. 380 is found in M M® Pal. It has not been quoted from any other MS. of the O. T., and might have been regarded as an accidental inadvertence (though Pal. 40 is remarkably free from such) but for the fact that it is also found in two Ven. 507 and Bodl. 143, which are both of the twelfth century. A presumption is thus raised that this variant is anterior to L, being derived from the text used by the original collector of the Cp. Ajax, 127-30.

jf. The theory that L is the archetype, and that all divergences from the 1st hand of L are due to conjecture, has even been pushed so far as to be applied to the marginal readings in the hands of the diorthotes or of the Scholiast. Almost every false reading is in one sense a conjecture, i.e. it is a piece of more or less clumsy guess-work on the part of the scribe. The first who in Ant. 699 wrote no doubt supposed himself to be correcting the error of a predecessor who had written perhaps with a bad up. Being of a prosaic turn of mind, a 'golden pillar' occurred to him more readily than a 'golden renown.' But there are some of these errors or conjectures that could never have been made by any one who had a sound and clearly written text before him. Who that saw the words (O. T. 134), as clearly written as they are in L, would have excogitated the variant Must not the MS. from which this reading was taken, have been derived from one in which the letters had been dropped because of their similarity to the preceding deo? But there is a further question. Who that had his choice between these two readings would select the marginal one? And yet it occurs in the text of a MS. of the 15th century, viz. Codex Laudianus, No. 54, in the Bodleian Library at Oxford. Similarly, the marginal variant for in Aj, 1056, is the textual reading of Par. F.

g. Corruptions not found in L which from their nature may reasonably be supposed to have an origin prior to the 1oth century. I assume that the following classes of errors more or less? answer this description : -

1 JT am aware that no precise line estimating the probabilities it falls to can be drawn, and that this test is one __ be considered. of very uncertain application. But in

(8) This argument may be fairly extended to a class of errors which, though not necessarily earlier than L, must have been nearly contemporary with it, or with such as the substitution of v or o for s adscript, which could not take place when the » was either subscript or dropped. E.g. in Ant. 862, M3, a late 15th century MS. (the same which gives - for has obviously a corruption of (i.e. the true reading; whereas L gives the false reading

(9) Amongst early errors might have been fairly included the wrong division of words, whether arising ex ore dictantis, © or from the habit of writing the words consecutively. Good examples are - h. Variations in later MSS. from the readings of L, or of C?2-5, which may be inferred to have an earlier origin, either (1) because of their wide currency, or (2) because unlikely to have been introduced from conjecture, or (3) for both reasons.

Xxxiv Preface

1 In O, C., Trach., Phil. the traces of the later hands in L are comparatively rare, and we are thus relieved from the doubt as to the age of C®C'. This partly compensates for the fewness of the MSS. of these plays, which arose from the fewness of readers. The scarcity both of readers and of MSs. may account for the paucity of late corrections, which, however, would be unaccountable if conjecture had been as rife in the 14th and 15th centuries as has been sometimes supposed,

Preface. Xxxv

An important distinction is indicated by some of the above examples, between those MSS. which have readings in common with L p.m. and those which agree with one of the later correctors. It will also be observed that some MSS. such as EV*M+?, while retaining some roughnesses which appear in the first hand of L, at the same time acknowledge readings which were only introduced into L by C® or And if it could be shown that when revised by C® the Medicean MS. was already in Italy, and that the MSS. in question, or some of them, were written in the East, this whole argument would be considerably strengthened.

j. It has been already seen that amongst the MSS. which are closest to Lis Pal. Gr. 40. The interpolation at O. T. 896, which is common and peculiar to L Pal. M° M? is enough to show either that all these are derived from one MS. which was distinguished by this peculiarity from others, or that the MSS. Pal. M? M?, are derived from L without modification in this particular, whereas a// the remaining MSS. are derived from a single copy of L, from which the words had been judiciously omitted. The latter hypothesis, however, is not very probable in itself, and is discredited by various circumstances, especially by the gloss appearing in Trin., etc. See above, p. xxviii.

In so far as the former of these two hypotheses as to the origin of Pal. can be confirmed, this MS. becomes a crucial instance against the theory that L is the archetype. For if a text so close to that of L is independent of it, how much more must this be true of MSS. like VAM, or of Par. AV? R, and of Par. B Vat. V7?

It is therefore worth while to adduce the following selection of readings, which have been obtained and verified directly from the two MSS. L and Pal.

> 23 3 . KOUKET e . . with Schol. . . . . . .

Xxxvi Preface

of subsequent tradition, why should Pal. at the same time (1) retain so many of the peculiarities and roughnesses of L, and yet (2) agree with C® in many readings, and, what is still more striking, (3) have roughnesses and seemingly early peculiarities of its own? Is it not more likely, on the whole, that it is derived from an archetype not very remote indeed from L, but anterior to it? The same reasoning may be applied to the case of L? M?.

1 Pal. consistently writes for pis, and commonly treats as an enclitic. It has a habit of sharpening the accent at the end of a line even where the sense is continued, and also (in common with MP) of occasionally doubling the accent on some monosyllables, such as and (sic), apparently to mark an emphasis. Cp. Gardthausen Griech. Palaeogr. p. 287. The order of words is marked in some places to assist construing with a, y, etc. In this and other MSS. there is a strong tendency to omit the augment, as in O. T. 1523, for 1454, for 743, for

This argument will be further strengthened, should it ever be ascertained, as is not impossible, that the revision of L by C® was subsequent to the arrival of the MS. in Italy, whereas most of the MSS. above named were written in the East. See above, p. xxxvi.

The portion of Cod. Pal. gr. 40 which contains the Ajax, Electra, and Oedipus Tyrannus of Sophocles, is a carefully written MS. of about the middle of the 14th century (according to Mr. H. O. Coxe) with but few erasures or corrections. As we have seen, it agrees very closely with L, more closely, perhaps, than any other MS. of these plays, but contains readings which were only introduced into L by hands of the 14th and 15th centuries, and some also of which there is no trace in L, but which are found in other MSS. and are in some cases of a high degree of verisimilitude. It is no doubt possible to account for these facts by the hypothesis that Pal. is in the main derived from a copy of L not later than the 12th century, into which, or into descendants of which, corrections were afterwards introduced. But the facts adduced on pp. XXVI-XxXxiil, supra, point to a different hypothesis, which is, antecedently, at least quite as possible, namely, that Pal. is derived in common with M? from a not very remote predecessor of L, which contained not only the readings that are peculiar to these three MSS, but also several of those which found their way into the chief MS. either at a later period or not at all. This does not exclude the possibility of some readings which are common to Pal. with other MSS. of the 14th century having an origin subsequent to L. But if this be the true account of the matter, it follows a fortiori, that other MSS. which do not contain the readings peculiar to Pal. with L and M?, but do contain some of those found in Pal. that are foreign to L, acknowledge a source that is prior to the special archetype of L, Pal., and This argument might be extended by including in the narrower list other MSS, such as 122, or TI, or M, or V, which point to an archetype, higher than that of L Pal. M?, and yet having lost some readings which may reasonably be referred to a still more remote origin. When this line of reasoning is compared with the preceding argument, pp. xxiv ff, they are found to be confirmatory of one another, for it was there seen that, even in MSS. of the 15th century, readings are found which can hardly be accounted for on any supposition but that of the misreading of early MSS. On the whole therefore, although the subject is not one admitting of demonstration, it may be held as probable that in using, as we cannot help doing, some of the inferior MSS. to supplement the evidence of L, we are relying not on Byzantine conjecture, but on what is substantially a much earlier tradition.

Even some of the readings attributed to Triclinius are found beyond the range of the Triclinian MSS. commonly so called; and although the industry of Triclinius is now referred, on good evidence!, to the first quarter of the 14th century, yet it seems improbable that such MSS. as V and 15, which are ambiguous between the 13th and 14th centuries, can have been affected by it. The cases of Vat. and Paris B are more doubtful. It appears safest on the whole to constitute a family of MSS. including T and Farn., but also B Vat. V7, and partially related to V*, Ven. 617, and even to L? and V. This last MS. is older than Triclinius, and may be found useful for the text of Sophocles, as Kirchhoff has found it to be for that of Euripides. Triclinius would appear to have had access only to an inferior text similar to that of Vat. and Paris B, and to have used conjecture freely, although to a less extent than he is commonly reputed to have done. Another distinct family is that of which Paris A is the type, including Vat. a, which is perhaps the earliest MS. except L, and having a text in some respects superior to that of L, By a curious fortune, this good text prevailed in the earliest

1 The Colophon of the Venetian MS. Wattenbach and Velsen, Exempla Codd. of Hesiod, signed by Triclinius, and Gr. Minusc. Tab, XXI. having a date equivalent to 13:6.

printed editions, the Juntine editors probably having access to L, and Aldus having printed chiefly from V°, a MS. somewhat later than A, and similar to it, but not derived from it ; whereas Turnebus introduced the Triclinian recension, which prevailed until Brunck selected A for preference, and Elmsley in 1825 called attention to the supreme excellence of L. Another MS. of the A family is R. 34, of which the Middlehill MS. is a late transcript or congener. Paris E, again, is closely related to A and V%, but is independent of them. It has been already shown that L, Pal., M7, M®, L?, A are related to each other. Beyond this it is difficult to go. Close relationship may be found between some later MSS., as between the Glasgow MS. and Paris D, but such discoveries are of slight value; it is more important to point out that some MSS., such as VM, Trin., Pal., A, even when corrupt, are comparatively naive and uninterpolated, whereas even the text of A and V® (as well as of 1.3 Vat. T) is not free from the suspicion of scholastic emendation, not from the hand of Triclinius certainly, but it may be of Moschopulus, or of Thomas Magister. In these naive MSS, if anywhere, we may expect to find amongst a heap of rubbish an occasional waif of earlier tradition.

In the matter of orthography I have for the most part followed the received custom, as, for instance, in writing for the termination of the 2nd person singular of the present and future middle and passive, and not with the MSS. This and some other constant habits of the chief MS, such as that of writing I have dropped from notice. In some things I have, however, ventured to return to the MSS., as in writing for yrwpoiut, as the difference of readings makes it difficult to ascertain the usage of the tragedians here. The troublesome custom of sharpening the accent before every stop, however slight, has no MS. authority, and, by whomsoever introduced, can hardly be defended. It is, however, the current usage, and I have adhered to it.

In describing the metres I have partially and with some hesitation made use of terminology which has of late become current in Germany. Since Rossbach and Westphal, following hints given by Boeckh, propounded their theory of the Greek metres, and especially since the edition by Westphal of the fragments of Aristoxenus and other rhythmical writers, a considerable change has taken place in the manner of viewing this whole subject. The tendency has been to dwell rather on the continuous than the discrete aspect of metrical phenomena, and to treat the laws of metre as inseparable from those which govern music and rhythm. In particular the notion of variations of time has been transferred from music to metre. Without wishing to be understood as committing myself to any positive judgment on questions that are still undecided, - _perhaps hardly capable of decision, - I have adopted the following expressions, which must be here explained.

1. Syncope of Thesis. Where two metrically accented syllables come together, it was formerly usual to speak of an antispastic effect: - an iambus, for instance, and a trochee following it being supposed to pull opposite ways. Instead of this the rhythm is now said to be modified by 'syncope' or suppression of the intervening thesis. Thus the Cretic rhythm may be described as a series of trochaic dipodies with the second thesis of each syncopated or suppressed ; a bacchius as a di-iambus with syncope of the second thesis, etc.

2. Lyrical Dactyls. According to a doubtful theory the dactyls in logaoedic verse are each of them equivalent in time to a trochee, much as a triplet may be occasionally introduced in ordinary music without altering the time. Such a foot is called a 'lyrical' or 'cyclic' dactyl

3. Irrational Syllables. Where a long syllable according to the usual view takes the place of a short one, as in the alternate theses of an iambic trimeter or the like, it is now supposed not to have really occupied the full time of the long syllable in arsi, but a time intermediate between that of the long and short syllable, and to have had an effect corresponding to that of a Ritardando or Rallentando in music. Such a syllable is called in the language of the rhythmical writers, an irrational time or syllable.

4. Extension of Long Syllables. The interval or 'rest' which is imagined to have taken the place of the syllable suppressed by syncope is otherwise conceived as an extension of the preceding arsis, or, in other words, as a 'pause' on the accented note. The sign - or 4 in the schemes of metre in this edition may be interpreted as representing either of these. Those who disbelieve in the existence or importance of such metrical phenomena may treat the sign as simply equivalent to the usual mark of a long syllable.

In the division of the lines I have not followed recent innovations but have adhered pretty closely to the system approved by Boeckh; not breaking up the longer rhythms, nor attempting to mark the subdivisions of them, which the student who has an ear and has given attention to the subject will be able to realise for himself. Other things being equal, I have preferred to avoid the division of single words between two lines.

The citations from Pindar follow Boeckh's arrangement of the lines: the same which is adopted by Dissen and Mommsen. The tragic fragments are quoted from Nauck's edition. Other references will be found (I trust) to agree with the editions at present most in use.

In concluding this long preface I have still to acknowledge my obligations, chiefly for kindness with respect to the MSS., to the custodians of various public libraries; above all, to my friend Antonio Ceriani, the learned prefect of the Ambrosian Library at Milan: to my friend H. O. Coxe, Bodley's Librarian at Oxford: to Dr. Neubauer of the same Library: and to the Librarians at Florence, Venice, Rome, Paris, Heidelberg, Leipsig, and elsewhere. Nor must I forget to record my special obligations to Mr. Bywater of Exeter College, Oxford.

And I may take occasion to lament the loss of men who have done so much for Sophocles, as Otto Jahn, Gustav Wolff, and William Linwood. Lastly, as an excuse for inevitable shortcomings, I would - a a Coon > a na? f. 2 N prefer this plea

Words marked with an asterisk are conjectural. Words marked with an obelus are probably corrupt. Words within square brackets are of doubtful authority.

Letters bracketed are erased in the MS. Letters marked with a line are written over an erasure. The dots (...) represent erased letters that are illegible.

On The Language Of Sophocles

Tue language of the Greek writers of the fifth century B.c. exhibits a degree of variety, which contrasts equally with the uniformity of Epic poetry and the comparative sameness of Xenophon and the Athenian orators. In no period of any literature do there exist so many linguistic differences between writers who were almost contemporary. These are not simply matters of dialect, and are not adequately described as peculiarities of style. They appear also in a tentative use of words and grammatical constructions which could only have been allowed even to great original writers in a comparatively early stage of literature. They are characteristic of the time when the most copious and plastic of Western tongues was growing into new forms under a manifold creative impulse, given by minds possessed of extraordinary genius, and stimulated to intense activity by changes of national and political life. A wider experience than had found utterance in the recitations of the rhapsodists, or in the simpler tones of the earlier lyric and gnomic poetry, and the growth of reflection, giving at once more concentration to thought and also the power of analysing it, required new vehicles of expression. And the first attempt to frame these in the different kinds of Greek literature led to many novelties of language, some of which did not permanently retain their hold. Each writer had his own tendencies, resulting in anomalies peculiar to him: and in an undeveloped but active state both of speech and thought, these licences passed unquestioned and easily found acceptance.

The limits of such divergence were not yet determined by grammatical rules. The race that gave grammar to Europe were by this time not unconscious of the complex structure of their language', but this consciousness was still indefinite, and was rather an incentive than a hindrance to invention. It is not easy for us, who often speak as well as write 'like a book,' to imagine what the art of composition must be in such a period of transition, when every considerable writer is a creative artist, and moulds the materials of common speech and early literature to his purpose according to the bent of his individual genius.

The age of Dante or of Luther occurs once only in the life or history of a language. The poetic faculty does, indeed, whenever it appears, exercise a sensible influence in re-creating or modifying the speech through which the poet works, and such a subtle master as Virgil or Goethe, or our own Shelley, will even at a late period produce new combinations, though commonly without violating the established usage, now fixed by a much longer tradition. But the working of the creative force in the adult state of literature gives no adequate conception of the power with which the same spirit operates in a time of transition, when the old materials are partially fused, and re-cast in other forms. At such moments the original instinct of language, which is never quite dormant, bursts forth with new energy, giving tise to phenomena which are intermediate between the first formation of language and a grammatical rule.

It is of some assistance to the English student of this subject to think of the second or Elizabethan stage of English literature, when the comparatively unreffecting art of Chaucer's age was succeeded by another style to which the revival' of literature, last felt in England, and the influence of the Reformation had given birth. New thoughts craved new expression, and the stream of utterance was quickened and disturbed by the influx of ideas. Hence arose peculiarities and diversities of language. Great liveliness and subtlety appear together with a certain want of accuracy. Shakespeare often throws words together in a way that would be impossible now, and it may seem no less strange that Spenser and Shakespeare should have been writing at the same time in England, than that Herodotus and Sophocles, or Sophocles and Euripides, should have been contemporary Greeks. In making this comparison, however, it must be borne in mind that logical analogies are more powerful in Greek than in English. The simplicity of ancient art extends also to grammar and diction. :

The period of expansion and growth in the Greek language began before Simonides, and the last phase of the process (which, however, can nowhere be regarded as complete) appears in the refinements of Plato. The disturbing influence is least visible in the transparent style of Herodotus, who retains in the new form of prose much of the equability and stateliness of Epic narrative. 'Though contemporary with Sophocles, he in fact represents an intermediate era, that of the somewhat.stationary and complacent civilization of the Ionic cities, which was continued after their political independence was impaired. He preserves, even more largely than theEpic poets did, the 'paratactic' structure, or parallelism, which belongs to a very early phase of syntax. This and other archaic traces in him may be partly due to the oral character of his materials and partly to the logographers who preceded him. His work is the conclusion of a period of continuous growth amongst the Eastern settlements of the Hellenic race; and while the study of him throws much light both on the language and ideas of Attic tragedy, it is a light that falls on them from behind.

But on the western side of the Aegean there had arisen from the beginning of the century, partly under the influence of the colonies acting on the mother states, a series of writers who used language more artificially, chief amongst whom were Pindar in lyric poetry, Aeschylus and Sophocles in tragedy, and Thucydides the historian. Their art was tentative and irregular, often deserting conventionality with great boldness, and yet straitened by the exigencies of a new endeavour. Under many different forms of ingenuity and refinement, or of abruptness and condensation, we recognise in all of them the struggling of the same creative effort. The tragedians worked under conditions peculiar to themselves, and have a certain external community of character. 'There grew up, as was natural, a sort of common dialect amongst those who were writing the same species of poetry for the same audience. Yet even amongst the tragic writers the variety is at least as remarkable as the resemblance. There are differences in their mode of using grammatical constructions and in their choice of words, besides marked differences of style, which can be traced even in the fragments. 'There is nothing in tragedy like the uniformity of Epic diction. The Epic commonplace is parallel to the rhetorical monotony of later writers *.

In order to a complete handling of this subject it would be necessary to consider in the first place the general characteristics of the language of Greek tragedy; distinguishing (1) its relation to the Epic dialect and to the language of Herodotus, (2) the infusion of gnomic and (3) of lyric elements which it contains, (4) the traces of vernacular conversational idiom to be inferred from the comparison of Thucydides, the early orators, and the old Comedy, (5) the effects of the effort towards concentration, observable especially in the frequent use of abstract for concrete, of the cognate accusative, of cases without prepositions, and of the poetical plural.?

The scope of the following pages is more limited. I have endeavoured to collect into one view some cf the chief idioms which occur in Sophocles, including those which have given most trouble to the

1 With the gradual adaptation of language to the drama may be compared the first use of language by the fathers of Greek philosophy, which is also a sort of experiment, and, like the language of tragedy, is intermediate between poetry aud prose.

2 Not much light is to be obtained from the remaining fragments of the predecessors of Aeschylus. It is permissible, however, to conjecture that Choerilus had set the example of using rugged metaphor, while Phrynichus had the melodious smooth- ness which old-fashioned lovers of poetry in the time of Aristophanes attributed to him, but that in other respects the early dramatists departed little from the simplicity of gnomic verse. On the other hand, of the contemporaries of Sophocles, Achaeus of Eretria seems to have delighted in forced and obscure expressions, and Ion of Chios, as we can partly judge even from the few lines of his poetry which remain, indulged like Sophocles, though in a feebler manner, in experiments of diction, commentators; and I have arranged them under heads, the greater number of which would be found equally applicable to other Greek authors, especially to the Attic authors of the same period. It would be absurd to represent Sophocles as an eccentric writer. But he made subtle use of a material full of subtlety, which was plastic to his touch, _and the resulting lines are more difficult to trace than those which have been impressed by some generally accepted mould.

An obstacle to right interpretation is removed when it is once for all acknowledged that canons, however useful, cannot be strictly applicable to language that is still in process of natural growth, and in which primitive zazve/é may be looked for side by side with novel abstractions and modes of thought.

In attempting to describe the linguistic peculiarities of Sophocles, I propose to consider, first, grammatical constructions; and secondly, the use of words: and shall begin by characterising the structural peculiarities of the Greek language in general, during the transitional period of which I have already spoken.

§ 1. In the writers of this period the power of reflecting on language appears rather as a subtle perception of analogies than as a rule of grammar. The mind is feeling after a limit which is not yet clearly prescribed. The genius of logic, inherent in the Greek language from the first, is awakened into fresh activity and has a transforming power over style. The forms of antithesis and inference already begin to complicate expression, and to react on thought, which becomes more self-conscious and less direct and simple.

In this process, which was part of the more general transition from poetry to prose, something was probably due to the change which began some generations earlier, but was still hardly completed, from the habit of oral to that of written composition. The author, who now sat pen in hand with the written page before him, must have become increasingly conscious of analogies of language and modes of thought. This generalized consciousness would be felt and used by him as a new power. Although the ultimate effect of a written literature is to induce a still greater fixity, yet by breaking in upon the conventional sameness of oral composition, the use of writing for a time combined with other causes to bring in a new kind of variety and individuality in the treatment of language. The working of this reflective and logical inventiveness appears chiefly in the opposite forms of analysis and concentration.

The increasing frequency of needlessly minute distinctions, belonging rather to the language than to the thought, the fondness for interrogative and negative expressions, the antithetical structure of whole paragraphs, the perpetual use of the forms of hypothesis, annines and inference, are evidences of the analytical tendency.

The effort of concentration appears in the more complex structure of sentences, and their connection with one another, and in some forms of inference, as well as in the condensation of words and phrases. This concentrating energy seems generally to accompany the poetic instinct in a reflective age, and is a common source of obscurity in the highest literature. For obvious reasons, it has a peculiarly fitting sphere for its operation in the language of tragedy.

A lesser effect of this tendency to reflect on language is the constant attempt to vary and extend the range of idioms, which, ceasing to be used instinctively, are caught up by the fancy of the mind and carried by a sort of metaphor into new applications. 'The sentences 'are not made, but grow.' General analogies often prevail over particular uses, and often two currents of logical association meet in one. Hence arise double constructions, changes of construction, and what may be called feeble or imperfect constructions, the government of one word by another being often suggested rather than actually determined. Where language has been fully developed, as in the studied periods of Cicero, the beginning of a sentence is always waiting for the end: the design of the whole is prior to the execution of the parts. But in the stage of language we are considering, the order and coherence of words and clauses are natural rather than grammatical, The meaning breaks upon us as we go along: the antecedent does not 'trammel up the consequence,' but the apodosis is continually modified by the interference of some intervening word. These are the peculiarities rather than the common characteristics - which appear with increasing frequency from Homer downwards - the exception,not the rule. It is not of course meant that all the uses of words and constructions in Sophocles or the greater part of them exhibit these qualities, but that there is sufficient to enable us clearly to distinguish between the style of Lucian or Xenophon and that of Sophocles.

The acknowledgment of such an unfixed and growing condition of language may be thought to introduce unnecessarily into the study of Greek syntax an element of uncertainty. And it must be admitted that the attempt to trace constructions beyond grammatical rules is one which may easily degenerate into fanciful subtleties. The study of these varying analogies is more difficult than the ascertainment of a fixed rule, just as curvilinear is more difficult than rectilinear measurement, or as motion is harder to calculate than position. But without attempting in some way to determine the nature of their language, apart from the definitions of a later age, the writers of the transition period can never be thoroughly understood. To require in _ Sophocles or Thucydides the exact articulation of perfectly grammatical writing would be no less mistaken than to interpret a speech in Coriolanus or Cymbeline as if it were a paragraph of Pope or Johnson. Many conjectural emendations which have been invented to satisfy this unreasonable requirement are found, when rightly weighed, to disturb the real harmony of the composition.

The grammatical consciousness is ever becoming more precise and imposing narrower limits on the use of language. A modern English writer could hardly confuse the singular and plural verb, as was done continually in the time of Shakespeare'; nor pass from the second person to the third, nor put 'who' in the objective case, nor omit the nominative, nor interchange the transitive and intransitive verb, nor with the same freedom employ a noun or adverb as a verb, or a verb as a noun", nor speak with Milton of the ' unexpressive nuptial song.' Even Shelley is hardly permitted, except by his latest editor, to use 'loved more' as an equivalent for 'felt more love' so as to 'include or involve a noun in its cognate verb' (Forman's Shelley, vol. 1]. _p. 270). Nor could the cultivated Greek of the time of Dionysius, Plutarch, or Longinus, have endured in a contemporary the apparent solecisms which they were quick to observe in Thucydides and other writers of the earlier age. On the other hand, the anomalies of the ante-grammatical writers have an important relation to the abstractions of grammar. They are the symptoms of the first stage of the reflective process which gradually made such experiments impossible, just as the tentative reasonings and even the fallacies of the early philosophers, which could have no existence after the Organon was written, were the necessary antecedents of the Organon.

In Grammar, as in other sciences, the generalizations and distinctions which have been adopted often prove inadequate to the subtlety of the process to be analysed. 'They are sufficiently applicable to the state of language which they help to perpetuate, though even here they cannot stop the flow of change, but have a less definite relation to the fluctuating analogies of the period in which grammatical structure is a growing framework not yet finally set. The regular grammarian is apt to be baffled or misled in applying his principles to the interpretation of a writer lke Thucydides or Sophocles. Yet their so-called irregularities, arising as they do out of the same instinct from which the regular and acknowledged forms of the language have taken their rise, must admit of at least approximate explanation. And there are two ways in which the validity of such a process may be tested: (1) the study of the history of constructions, giving the key to some analogies which the later classification has tended to obscure: (2) if at first sight the method of such interpretation may seem sceptical and fanciful, it is not therefore to be rejected, if supported by a sufficient number of instances either from the same writer or from others. The mere number of such exceptions will prove that they are not merely capricious, but have a rule or principle of their own. It might also tend to prevent vagueness and confusion in such an inquiry, to distinguish between idioms current at the time and idioms of the particular writer. This distinction, however, although real andimportant, cannot be always accurately observed ; partly from our ignorance of the spoken lauguage, and partly because we often cannot tell with certainty how much is due to the genius of a writer, and how much to the habit of his age. In the unfixed state of the Attic dialect, some forms or modes of expression, such as the use form in s was derived from the Northern For other instances of Elizabethan English dialect. The confusion existed, freedom, see Mr, Abbott's ' Shakspearian and the necessity of clearing it determined Grammar,' of the article for the relative, may have been still current in common speech which were afterwards thought to be peculiar to an Epic or Lyric style. And there are probably others which, though passing out of use, were familiar to the hearers of poetry, and from which the

'modern' poet. occasionally selected. In the manner of such selection, and in the modification of the selected forms, the characteristic qualities of an author are sure to appear. 'This is seen chiefly in the general colouring of the language and the choice of words, but is also perceptible in peculiarities of grammatical structure.

§ 2. Sophocles is distinguished in point of syntax from other contemporary writers chiefly by his noble sense of harmony, and by the combination of extreme refinement with moderation and simplicity. He is, to borrow a phrase from Aristotle, : - the first equable writer of the reflective age; for Herodotus, as already noticed, is essentially of an earlier type. The incipient logical analysis, and ethical and political thought, and also the love of condensation, which have such strange effect in the intricate or knotted constructions of Thucydides, are present in Sophocles, and are often in him a cause of uncertainty, but never of harshness. He has neither the ruggedness of Aeschylus, nor the strained ingenuity of Pindar. Like all the greatest writers, he can only be partially studied by the light of his age: he has Gnaacien dics of his own, which are best seen by comparing him with other great masters of language in different ages. In the absolute subordination of every tone and emphasis to the dominant feeling, and the perfect adaptation of the language to the degree of tenderness or sternness, of passion or coldness, which is required by the situation, he resembles Virgil and Goethe, and those parts of Shakespeare in which his exuberant invention 15 most under control. This power of moderation adds greatly to the effect of pathos, in which Sophocles is unsurpassed.

§ 8. The peculiarities of structure in the lyric passages of Sophocles are not essentially different from those in the dialogue. The influence of the earlier lyric poetry is more perceptible in them, just as reminiscences of Homer occur chiefly in the long speeches, and especially in narrative. But both the Epic and Lyric elements, as well as that of common speech, are present everywhere, and are so blended by the poet's art that any broad line of distinction would be illusory. Even the Doric dialect, the traditional vehicle of lyric expression, is only partially assumed, and hardly interferes with the harmony of the general effect. In the following digest of idioms it has been thought sufficient to denote, by the letters 'lyr.' prefixed, examples taken either from the choral odes or from other lyrical passages. . We may therefore here observe the chief particulars in which these differ' from the senarii in point of syntax. They are mostly differences.of degree. The spirit of fanciful association is more. active, and' more apt to usurp the place of conventional use. The concentrating effort is -more intense, recalling in some passages the spirit of Hebrew poetry. There is more of rapid and abrupt transition, more of substitution and inversion, of redundance and deficiency. The article is often omitted; prepositions are often added pleonastically. The stress of lyric feeling is combined with the elevation of poetic thought. And the language varies from great simplicity to great complexity, and from the most perfect serenity to the most highly-wrought excitement.

Comparative philology has made familiar the law of decay and regeneration in language, by which inflexions gradually lose their meaning and are replaced by other elements. 'This process is accompanied by the tendency to analytical expression, which makes the meaning of terminations and prefixes more explicit by the use of auxiliary words. 'The history of the case-endings affords a striking exemplification of this law. In English, as in most modern languages, there are at most two oblique cases, .the possessive and objective, and of the latter the pronouns alone retain the inflexion. The possessive expresses only one relation, and the objective can hardly express more than one without leaning on a preposition. Sanskrit has the accusative, genitive, dative, instrumental, locative, and ablative. The Greek language retains traces, in the locative adverbs, of an almost equal number of cases, and each of the three remaining cases expresses a variety of relations, which in the earlier literature were more vividly conceived than in the more conventional usage of later times. It would be instructive to study the gradations of the change from Homer downwards, through the earlier and later Attic, the common dialect and the Byzantine period, to modern Greek. The dying vigour of the old forms would be seen to revive and grow again from time to time. 'Experiments' more or less conscious sprang up and withered, before the prevailing habit was finally determined. Opposite tendencies are seen working together: the preposition being sometimes inserted as well as sometimes omitted unexpectedly. In this process, as in every important modification of language, the poetic instinct plays an active part. More conscious than other men of implicit meanings and connections of words, the poet gives new life to the decaying forms. All poetry abounds with novel modes of speech, from which common language selects what is most in accordance with prevailing tendencies and most suitable for use. That the case-endings still retained much of their force in the old Attic is evident from the language of Thucydides. 'The poet seizes this instrument of condensed expression and strives to multiply it, partly by reverting to earlier modes, but chiefly through analogical and metaphorical applications. The exact intention in each instance is not always easily determined. It was no doubt felt but could not have been explained by a contemporary Greek.

Even in English, where the limits of the experiment are so narrow, there have been similar attempts to extend the use of the cases. Thus the remote object is often expressed without a preposition by a second objective case which thus performs the function of a dative, and there are many instances in the Elizabethan poets of a sort of dattvus ethicus. 'A' would manage you his piece thus,' 'A good sherris-sack .. ascends me into the brain,' etc.

§ 6. The question as to the origin of inflexions has little direct bearing on the present subject. Although the theory that derives the cases from relations of place is most in accordance with the analogy of language - -which seems to proceed, like the other phenomena of thought, from without inwards, - and is probably true, yet this account of the cases, in the later state of language which is alone known to us, affords hardly any help in determining the multifarious and anomalous relations which in the course of ages have grown up around them. And could it be shown that the function of each case was in the beginning simple and definite, and not rather tentative and vague, yet they have not reached the position in which we first find them without many cross-influences which are hard to investigate. Reflections from other relations, the exigencies of growing expression, and the re-action of different uses upon each other, have woven a sort of tangle, in which the eye can hardly follow any branch to its origin. Hence, while the distinction between external and internal relations may afford a principle of classification, which accords in some degree with changes actually observed, this is not to be supposed equivalent to the distinction between earlier and later modes in the period under discussion. 'There are further differences which it may be useful to bear in mind. In attempting to analyse the uses which are actually found, it is important to recognize the difference between the metaphysical analogy which has suggested certain constructions, and the rule of usage to which they tend to conform, and, thirdly, the generalization of them which is made by the grammarian. And it may be observed, in passing from this part of the subject, that many things are determined by use, quite regularly, of which no metaphysical explanation can be given, and which can hardly be reduced to a general principle.

The genitive has two main significations, which are clearly distinguished, although in some instances they pass into each other, the ablative and the attributive. The former is gradually discarded, except with prepositions, and the case more and more assumes the character, to which it has been thought to owe the name denoting the kind to which a thing belongs, and thus tending always to express the wider of two conceptions. 'The connection between the governing noun, and the noun in the genitive by which it is included or determined, is sometimes so indefinite as to be extremely difficult to analyse.

It will be convenient to arrange under the above headings the characteristic or exceptional uses of the genitive in Sophocles. It may be observed that nearly all of them are retained in later Greek with the support of prepositions.

8. a The use of the genitive to denote the origin of motion is retained by Sophocles (in common with the other tragedians) from the earlier poets.

O. T. 151. lyr. ras 'That camest from the golden shrine of Delphi to glorious Thebes" O. 572. And from what land I came. El. 78. | And from the doors I seem to perceive the cry of some hand-maid moaning within.' Here the genitive marks the place from which the sound is heard to come. The notion 'within' is added afterwards. Ib. 324. .. | Xpuoddcuw . | . . Ib. 1139. map- Bapos - ' Nor out of the blazing fire did I gather, as is meet, the sad burden.' Ant. 1008. érnxero - The dank humour wasted off the thigh-bones.' Aj. 1148. | Tr. 781. éexpaivee - ' Rains forth the white marrow from the hair.' in composition here assists the construction.

O. T. 808. | From his chariot watching me as I was passing by he smote me full upon the head with the double goad.' The genitive is to be connected equally with and and in its relation to both is éyov. So the locative adverb = Tr. 938.

2. The notion 'from,' easily passes into that of separation from, which occurs in a few passages: e.g. in O. T. 1522. pou - ! Take away from me.' Ib. 762. Out of sight, away from this city.' 'The case is partly governed by in composition. There is probably an extension of the same analogy in El. 159. lyr. 78a | Happy, in that his youth is hidden out of the way of sorrow.' Cp. however infr. y. 3. 6.

1. The possessive genitive, 'of' or 'belonging to,' describes or defines the word with which it is in regimen. It has sometimes the force of an adjective rather than of the case of a noun.

a. O. 447. The help of kindred,' i.e. the help which kindred gives. El. 162. lyr. Bnpari - With auspicious movement prospered of Zeus.' Ant. 153. lyr. Baxytos - ' Bacchus who belongs to Thebes and shakes the ground,' or, 'Bacchus, the earth-shaking God of Thebes.' O. C. 945. | . . The unholy union of a child (with a parent)' 19. And Night with her stars is vanished,' (i. O. C. 1325. és (where, however, there seems to be a reflection from the ablative use) - ' I lead the dauntless army of Argos (or from Argos) to Thebes.' Phil. 43. véorov - ' A foraging expedition.' Ib. 104. Opéc0s - ' Boldness relying upon strength.' Aj. 356. lyr. vaias téxvas - ' Ye whose aid consists in maritime skill,'

O. T. 411. And so I will not be enrolled the client of Creon. O.C. peyi- Come, Athens, the city called by the name of mightiest Pallas.' Ant. 738. od rod Is not the city held to be a possession of the ruler?' Phil. 73. rod

(1) With os; Ant. 256. . . But a few thin grains of dust, as from one who shuns the curse:' © or Tr. 767. | rovos The robe clung fast, moulded to his form in every limb, as by the hand of a statuary.'

Ant. 114. lyr. With plumage white as snow.' Ib. 423. | sharp cry, like that of the bird in the bitterness of sorrow.'

Tr. lyr. Never allows the longing of her eyes to rest, or wipes away her tears ;' Ba. i.e. 'the longing that her eyes feel.' Cp. O. C. 729.

. 6S00- - ' Mindful fear for her husband regarding his journey.' Where may be partly governed by and is added in explanation of In O. T. 1200. lyr. is genitive of the object. O.C. 466. Phil. 17. A chance of sitting in the sun.' sometimes made to indicate an indefinite connection, which no single word exactly supplies. This may be viewed as the most abstract meaning of the case, in which all specialty of meaning is lost.

O. C. 436. When I desired this, I found no helper' Ant. 1194. és | --- For why in addressing thee should I use soft words, wherein J shall hereafter be found to have spoken falsely?' O.C. 647. Evvovcias - A gift, in respect of (or, implied in) your dwelling with us.' Ant. 1265. lyr. Bovhevpdroy - ' Alas, misery in regard to my counsels!' or, 'Alas, me! for the results of my counsels.' The expression seems to be a confusion between avorABa and In O. T. 728. the genitive seems to be in a double construction; (1) genitive of relation after and (2) genitive of cause with (Cp. p. 15. 1. 4) 1078. lyr. Ib. 1096. lyr. SO with the apparent ellipse of [Cp. Hadt. iv. 142. | Of whom you oftentimes sent me secret messages, how that you would yourself appear as the punisher. Where the construction is aided by Phil. 439. El. 317. ; ; O. C. 661. | And though they have been emboldened to speak high words of fetching you away.' Ib. 513. lyr. | Respecting that unhappy sorrow that came without remedy. Aj. 771. Ant. 1182.

O. T. 1260. --- As if following some guide.' O. 1558. Sc. dvros - ' With no one of his friends for guide.'

On The Language

O. T. 1006. That 1 might get some advantage from you, when you come home.' gov genitive of cause: gen. abs. E]. 1485. yap | He of mortals who delays death, when mortals are entangled in woes. Tr. 703. | Baxxias dunéhkov - ' As from the rich draught of the blooming vintage, when this is poured to . earth from Bacchus' vine.' .

El. 341. Phil. 3. .. Ant. 38. xaxn - ' Degenerate daughter of a noble house.' O. C. 1323. rod | Begotten by adversity ;' 'heir to adversity. Ib. 1378. | toad' Since ye were such sons of a blind father.'

Tr. 765. | aivarnpa - ' And when from the dread sacrifice blazed the flame mingled with blood" O.'T. 1012. Lest you should contract a pollution derived from your parents?' Ib. ggt. And what is there coming from her that gives you cause to fear?' Tr. 203. 4. | viv 'As we now enjoy a glad vision unhoped for by me to which this report gives rise.' Ant. 1017. | ydvov - ' Infected through dogs and birds devouring of the unhappy fallen son of Oedipus.' Where the gen. is derivative after and also partitive with Bopas.

Under this head or some other must be included the genitive of that from which a change is made. El. 1262. Aj. 208, 9. | WE

Tr. 267. | How he was maltreated by a freeman.' Ib. 112. lyr. Bopéa | (cp. 1. ¢c.) - * AS one may see waves driven by the unwearied south-west or north-east wind.' El. 344. Aj. 807.

O. T. 1478. | And in reward for bringing them may the deity guard thee better than he has guarded me.' El, 626. ddvéeis - ' For this insolence you shall not escape.' Ant. 1074. Aj. 1116. | Fr. 630 (N.) .. Also expressing purpose - Phil. 107. kK...

the verb the subject or object of the action is put in the genitive. Comp. the Horatian instance, 'Sanguine viperino cautius vitat' (==cautius quam vitat sanguinem viperinum).

in flight a swifter foot than (is plied by) stormswift steeds.' The genitive sometimes follows a verb in which some notion of comparison is implied.

. The genitive of nouns signifying a region or space has sometimes an indefinitely partitive meaning (as in the common use with adverbs) - ' somewhere in,' ' anywhere in.'

O. C. 694. lyr. oiov .. yas eraxove - ' Such a thing as I do not hear reported of (as growing) anywhere in Asia.' Aj. 659. Having dug it in earth where none may see' (where the order shows that is only joined

922. You know not whither, or to what thoughts you are borne away. Ib. 1174. Where speech is impossible, what shall I say?' In Aj..731. the partitive form of expression which is similar to &c., has been retained, but the meaning is lost. 'And strife having run to the furthest extreme.' But see below, 5 @.

6. The notion of hitting, lighting upon, and aiming at, may also have arisen out of the partitive use. Hence the genitive after &c. The same analogy affords an explanation of the genitive in such cases as the following : -

El. 869. Neither having received any burial.' (Cp. the Homeric .. 1]. 67.) El. 1451. yap They found a loving hostess at their journey's end.'

O. C. 1519. Undisturbed by age.' Ib. 1521. Without the hand of a guide.' El. 1002. mos Ib. 1214. .. O.C. 1383. Unfathered by me.' Ib. 1722. lyr. --- 'Exempt from liability to ill' Ib. 1719. lyr. 'Who have lost a father". O. T. 885. lyr. 'Without fear of justice' Tr. 23. - ' Untroubled at the sight.' El. 232. lyr. Ib. 243. lyr. i.e. yoveas - ' From honouring my parents.' course (epayapévn). The genitive, as distinguished from the accusative, marks that the limit was touched but not passed. (Cp. supr. p. 3. 4)

5. The genitive of time is not peculiar to Sophocles. It describes a time which is not exactly defined. | Hdt. iii. 134. Thue. v. 14.

6. There is a use, not easily brought under any of the preceding heads, in which the genitive and the governing noun are coextensive, and only present different aspects of the same notion. By this means both words receive an emphasis.

O.T. 1474. guotv - Ant. 1209. Bons - - *Lhere comes about him the dim sound of a lamentable cry.' El. 758. onodod - 'The poor dust, his mighty frame." O.C. 923. - 'Hapless suppliant beings.' Ib. 1029. The great insolence which is manifested in your present daring.' Phil. 81. vicns - ' Victory as a possession.'

The dative, whether arising out of an original locative case or not, has three main significations, which cannot be further analysed in the existing language.

a. 1. The simply locative case, expressing 'in or at a place,' is no less frequent in the Tragedians than in Homer. Sophocles here as elsewhere condenses language by omitting the preposition, and preserving the independent meaning of the inflexion.

O.T. 20. Gaxei - 'Is seated in the agora.' Ib. 1266. | El. 174. lyr. Ib. 313. He now chances to be in the fields.' Tr. 100. lyr. Situate in either mainland.' (Cp. the Epic use of with the dative).

seasons come round" Ib. 198. For if Night at her close leave aught' (?). El. 194. Sad was thy voice at his return.'

Phil. 715. In a period of ten years.' Phil. 721. --- the fulness (or multitude) of many months.' Ib. 598. |

O.C. 580. --- time you shall learn, not, I ween, at the present time.' Where means a limit of time, and a point of time.

force of the accusative with and may be regarded as intermediate between the dative of place='in' (a), and the dative of reference=' to'

eround. Ant. 1236. (i. 6. 'He pressed the sword into his side. Aj. 240. (== ---- - Having bound him 'to' (or 'at') a pillar.' O.C. 483 Ib. 1690. pornos (?) - ' Let cruel Hades take me down to be with my father. Cf. the pregnant use of ev in O. T. 656, 7.

Tr. 940. airiav) - ' Cast her into an evil reproach.' Ib. 597. reoei - ' You will never fall into disgrace.' (?)

2673. lyr. | per For whom beforetime we gave our long assiduous labour.' In Aj 37. | came on my way as a guardian watchful over thy chace, the dative depends on the whole phrase And in 235. lyr. there is the further notion of addition.

a. The simplest form of this use occurs in El. 496. lyr. | | That we shall never see a portent coming without harm to the murderers and their accomplices.' :

6. This is sometimes so used as to be equivalent to with the genitive. El, 442. | --- Consider then whether the dead who is in his sepulchre is likely to receive this present kindly at her hand" = In El. 226-8. lyr. rive | ; | tive ; the dative seems to have nearly the force of (See under 1) Cp. the Scotch idiom 'to ask at.'

Tr. 140. lyr. . Without thought for his children." El]. 1066. lyr. Péya - Oh rumour that reachest mortals beneath the earth.' |

Ant. 859. lyr. re | Our lot,' and so 'the lot which fell on us,' where we must supply from

O.C. 962. | rod otdparos - ' Who hast uttered to my hurt.' Tr. 541. Kadkovpevos - ' Our trusty Heracles, reputed a good man' (ironical). In the first person singular this has often a precatory force (as in Greek generally). El rogo. . eyOpav - ' I would pray that you may live higher than your enemies." O.T. 1512. viv But I would have you pray this prayer. El. 144. ri Ib. 887. | In the second person, oo: nearly=the particle ro. O.'T. 708. | K.T.A, made an outcast. Phil. 1030. madkac - * And am long dead, so far as you are concerned;' Aj. 1128. am dead already, as concerns him. El. 1152. am dead in thee,' i.e. Death has destroyed the life that I had in thee. Ib. 226. i.e. (But see above, 19,

f. Dative of opinion = 'in the sight of, 'in the judgment of.' (Cp. esp. Eur. Phoen. 495, 6. | kal

O. T. 8. 6 Oedipus, whom all account renowned.' O.C. 1446. dvdéia For all may see that ye deserve.no ill' Tr. 1071. Ant. go4. And yet those who consider well will say I did honour thee.'

El. 1295. viv 6da - ' We will still the laughter of our enemies by this attempt.' O.T. 494. lyr. | OidirdAa - ' 1 shall assail with proof the popular fame of Oedipus.' O. C. 1160. Urging what request by sitting there?' Aj. 946. lyr. | | By this cry of grief, you have given voice to the silent work of the cruel Atreidae.' Phil. 494. | atvrév- - ' Often through those who came hither I moved him to make the voyage.' And with a curious ellipse or inversion, Ib. 1191. lyr. yroua - To do what (do you bid us) by a change of mind (in you)?'

1212. cr&a - ' Do I not justly mourn?' Ant. 107. lyr. Bavra ravoayia - Him who came in panoply.' Ib. g6o. lyr. | Oedy - ' That he was mad to tempt the god.' Ib. 1003. Tearing each other in bloody fray.' Phil. 758. yap This wandering pain is come again after a pause.' 'Tr. 596. mpaoons - ' If you do under concealment even what will not bear the light." O.C. 548. lyr. O. T. 1339. lyr. Aj. 1094. dv yovaiow. In two instances this dative, by a somewhat forced construction, takes the place of a predicate. O. T. 908. lyr. (1. 6. adyduevos) - ' And nowhere is

Apollo manifestly held in honour,' (where, however, the construction is assisted by in O.C. 278. (i.e. Cause to be in no way respected.) In 47. may be explained as dative of manner, to which the participle is added epexegetically.

The dative of the cause or reason - ' by reason of' - which, as the generalized use of this case, may be compared with the genitive of relation (supr. p. 13), is very frequent in Sophocles, and occurs in Thucydides, e.g. vi. 31 sub fin. 6 Te ots Ib. 33. TLY@Y

(1) Cause. Ant. 335. lyr. Goes, borne by the Southern blast.' Ib. 589. lyr. mvoais - - ' Courses o'er the dark depth, carried by the winds from Thrace.' Ib. 1234. Missed his father, who fled forth.' O. T. 517. Tending to injury by words or deeds.'