THE EDITOR'S PREFACE то THE SECOND EDITION. I. The Notes in this book were begun in 1856, but, from causes explained in my former Preface, not finished before 1875. The work was undertaken at the request of one who lived to see, but did not long survive, its publication. I mean my accomplished friend, the late Mr. WILLIAM LONGMAN, whose premature death was a great public as well as private loss. II. In this second edition three divisions of the former commentary (translation, vocabulary, and notes) are fused in one, and numerical reference made more distinct. This change unavoidably swells the size of the volume, which also contains an enlarged Syntax and Indices, with a verse translation of the Eclogues. III. Those who study the Virgilian Syntax should compare with it the Second Appendix to the 'Public School Latin Primer' (1878 and later), which treats concisely, but carefully, of Moods and Compound Construction. In this place I think it my duty to introduce a critical notice of the manner in which Conington's edition deals with three classes of construction; viz:— (1) Mood subordinate to actual oratio obliqua. (2) Mood subordinate to virtual oratio obliqua. (3) Mood of indirect will-speech, called in that edition by no specific name, but only termed 'oratio obliqua'; in my syntax 'petitio obliqua.' In my respect for the memory of the late Professor Conington I yield to none. I recall with affectionate admiration his wide culture, his fine genius, his poetic taste, and the amiableness of his character. But it is neither honest nor wise to praise even the dead beyond his true merits, by acknowledging him right where you know him to be wrong. 'Amicus ille: sed amicior veritas, amicior doctrinae sanitas.' Besides which, the treatment of these questions is forced upon me by the procedure of others, unless I consent to leave the cause of truth in grammar undefended. While, then, I give just honour to the many good points in Conington's notes, and to the sound judgment shown in his prefatory dissertations, I must yet say, that on questions of grammar his commentary is generally defective, often wrong. I proceed to prove this assertion in respect of the three topics named above. (1) Mood subordinate to actual oratio obliqua. We cannot suppose that C. had not observed the threefold nature of oratio obliqua (statement, question, and will-speech), as it appears in historical passages like this of Justin, v. 10: Thrasybulus, cum exercitus triginta tyrannorum fugeret, magna voce exclamat: cur se victorem fugiant? civium illam meminerint aciem, non hostium esse: triginta se dominis, non civitati bellum inferre ; where, in subordination to one verb 'exclamat,' appear (a) an oblique question, 'cur fugiant?'; (ẞ) an oblique willspeech, meminerint'; (7) an oblique statement, ' se inferre.' There seems to be no passage in Virgil, where all three constructions are thus zeugmatically conjoined; but statement and will-speech are found together occasionally, as Aen. iii. 234-5, iv. 288-94, xi. 101-5; and in none of these places is attention drawn by C. to the zeugma, nor in iv. to the subordinate subjunctives' nesciat speret.' In Ecl. vi. he takes no notice of the long construction by oblique question with uti . . . ut 31 ... 40, or of its matical point he touches in this Eclogue is the peculiar ellipse 74, which he supplies correctly. In his first volume, indeed, he has generally neglected to deal with questions of mood. In the two others he is somewhat more alive to their importance. Aen. ii. 94: me, fors siqua tulisset, Si patrios umquam remeassem victor ad Argos, Here he says: 'the pluperfect is used on account of the oratio obliqua, as in 1. 189, iii. 652, ix. 41, which confirms Donaldson's opinion that the so-called futurum exactum is really only the perf. subj.'1 This note is strange. Since oratio obliqua affects mood more than tense in a subordinate verb, we might suppose pluperfect a lapse for pluperfect subjunctive, if the context did not shew tense to be in the writer's mind rather than mood. As the other two passages which C. cites (also cited by me in Virg. Syntax) likewise have a pluperf. subj. (fuisset), this adds to the probability that he was thinking of tense more than of mood. At iii. 652, he merely refers to ii. 94, though in that place the construction (huic me quaecumque fuisset addixi) is not actual but virtual oratio obl., of which, however, Conington had no clearly defined notion. On ix. 41, indeed, where the oratio obliqua is of the form will-speech (neu auderent), he does say correctly that 'fuisset' in the oratio recta would be fuerit. The conclusion in his mind was 1 The opinion said here to be confirmed is quite untenable. All tenses of the Thought-mood may indeed on occasion convey the idea of futurity: but this does not prove that the Fact-mood has no tense of its own conveying exact futurity, while instances may be cited by the hundred to prove that it has. It is more philosophically true to complete the cycle of tenses alike in each mood thus: therefore the right one as to the effect of oratio obliqua upon mood, but incorrectly expressed in two of the places cited. Aen. ii. 229: scelus expendisse merentem Laocoonta ferunt, sacrum qui cuspide robur Laeserit et tergo sceleratam intorserit hastam. On this place Con. writes: 'laeserit and intorserit rather than laesisset-intorsisset, because of ferunt.' This is both vague and incorrect. Of mood he really says nothing, leaving the orat. obl. to be implied: but he imputes the use of perf. instead of plup. to the tense of the principal verb. And here he is wrong, as a place in the same book (ii. 433) proves: Testor, in occasu vestro nec tela nec ullas Vitavisse vices Danaum et, si fata fuissent Here the principal verb (testor) is pres., the infinitives 'vitavisse, meruisse,' are past: yet the subordinate verb (fuisset) is plup. not perf. Thus we see the error. The tense in the former place was not because of ferunt'; but because the infin. 'expendisse' is a pres. perf., 'has paid the forfeit' (being an event immediately preceding), while, in the latter, 'vitavisse, meruisse' are aoristic of time indefinitely past, that I shunned,' 'that I deserved. The finite tense corresponding to the former was thus, 'laeserit' (has injured), to the latter, 'fuissent' (had been). Again, therefore, we find an incomplete grammatical aesthesis. Aen. iii. 581: Et fessum quotiens mutet latus, intremere omnem &c. Here he writes: 'the subj. seems to be required by the oratio obliqua, although mutat is the first reading of Pal. and Med.' Better had he written, is absolutely required' for the scribes of Pal. and Med. were as ignorant on such points as some moderns are. Aen. vii. 272: hunc illum poscere fata. He says, 'hunc illum esse quem fata poscunt,' betraying unfamiliarity with the laws of suboblique construction: he should have written 'poscant.' Aen. vii. 427: Haec adeo tibi me, placida cum nocte iaceres, His note is: 'Cum iaceres connected with fari, and so marking not the time when Juno gave the commission, but the time when the commission was to be exercised.' This is true, but inadequately explained for the connection with fari has little to do with it: iubet me fari cum iaceas' would be as correct as iussit me fari cum iaceres.' 'Iussit' determines tense, 'iussit fari' mood. Aen. vii. 766: 6 Namque ferunt fama Hippolytum, postquam arte novercae caeli venisse sub auras &c. Here C. simply and truly says: the subj. is accounted for by the oratio obliqua.' It is curious, however, that here 'occidisset, explesset,' would be more normally correct: but as Virgil could not get in the former word, he compromises the tenses by making them aoristic. In the remaining places cited by me (V. S.) under this head (which are twenty in number), C. makes no remark about mood. (2) Mood in subordination to virtual oratio obliqua. (See p. 664, and Appendix II. to 'Public School Latin Primer.") Of this subject Conington evidently had no definite knowledge, but only vague glimpses, here and there appearing. Nor could he gain more than this from such a Syntax as Madvig's; the use of which in English schools and colleges. |
