Imágenes de página
PDF
ePub

must be disproved on other grounds than the bare fact of their long concealment. Velleius Paterculus is not quoted till the time of Priscian, 400 years after he wrote, and then comes a silence of 900 years till Aventinus. Pithæus, the first editor of Phædrus' in 1596, could find no reference to his author before Avienus, a period of 400 years. But such silence, as Bentley remarks, cannot pass for a negative proof that there were no such authors.

'If they were read,' he says, 'the readers of them were not writers themselves, so as to let posterity know that they had read them. Although the writers of the barbarous ages do not speak of those authors, they do not say anything to imply that they had no existence. If they do say anything amiss which could be corrected from them, it is to be imputed only to their own ignorance, or laziness, that they would not search for them.' †

To this, with some authors, must be added the scarcity of manuscripts obtainable. Cicero's Letters are not mentioned between the time of Lupus of Ferrieres in the ninth century and Petrarch's discovery, an interval of silence very improbable had there been any number of copies then existing. But, exceptis excipiendis, with the great body of classical authors the evidence afforded by tradition is abundant. Throughout the Dark Ages the continuity of knowledge was, however slenderly, maintained, and the remains of ancient literature, taken as a whole, contain within themselves the proof of the genuineness of each part. For the purposes, however, of emendation this class of testimony is subject to restrictions. Did the writers quote from memory, or from a manuscript in their possession? If from memory, they might easily have

Santen (pref. ed. Amsterdam, 1780) threw doubts upon the value of the discovery. Poggio, says Jovius, found Quintilian in salsamentarii 'taberna.' The 'Ethiopics' of Heliodorus were published by Opsopæus from a book given to him by an Hungarian soldier, which he had picked up in the pillage of King Corvinus' library (pref. ed. princ. 1534). Disraeli (Cur. of Lit. p. 8) gives, among other instances of discovered MSS., an anecdote of Sir Robert Cotton, who was said to have found his tailor on the point of cutting up an original copy of Magna Charta. See also Bentley (Dissert. ad Phal. p. 375) for pretended discoveries, including that of the Revelation of St. James, supposed to have lain buried in Spain from the time of the apostle. Unluckily the MS. was written in modern Spanish, a difficulty, however, which Aldrete accounted for by the proleptic exercise of the 'gift of tongues!'

* Botfield's Prefaces, pp. 627-9.

† Dissertation on Phalaris, p. 367, ed. 1817.

6

forgotten the exact words; if from a manuscript, that manuSo far, indeed, from such script might have been corrupt. quotations, even when professedly literal, contributing to prove the purity of the earliest existing copies, or their identity with the original, they reveal, in some cases, the remote introduction of error. 'It is not easy in any other way,' says Mr. Munro, 'to explain the agreement of Macrobius and Nonius (fourth century) with the archetype of all existing MSS. of Lucretius in many indisputable corruptions.'* Cicero's cration pro Flacco appears to have been imperfect in the same century from another reference in Macrobius; and the writings of Solinus, Martianus Capella, and Isidorus contain quotations from Pliny with flagrantly corrupted words.† Columella, who quotes several passages from Cicero, makes him mistranslate Xenophon, but probably he merely had Cicero before his eye and put his thoughts into his own language. And in the fifth century Cassiodorus misquotes from Tacitus, though his words quodam Cornelio scribente'-a conjectural reading gives quondam-argue a slender acquaintance with his author. Indeed, lax and inaccurate quotations were the rule; and from the seventh to the thirteenth century, though allusions to the classics more frequently occur, they are for the most part indirect and unauthenticated. Orosius refers to Tacitus in the fifth century, and may have had direct access to his works; but Freculphus, Bishop of Lisieux, in the ninth, as well as John of Salisbury-whose references to Cicero's Epistles are quote second-hand from from Macrobius and Quintilian

Orosius; and Adam of Bremen is content with the readings of Rudolphus, a monk at Fulda in the ninth century, the only one apparently of the German or Italian historians of the Middle Ages who had consulted Tacitus' writings.§ Of Lucretius few traces are to be found in that interval; and, according

Introduction to Notes, p. 313.

Several early writers contain passages from Pliny which cannot be found in any existing MSS., including the best. But Sillig conjectures with reason that most of these quotations were fictitious, made to give a spurious authority to their writings. Pliny being the standard author on Natural History throughout the Middle Ages, the prevailing ignorance and the probable scarcity of MSS. offered special temptations to this kind of fraud.

Orelli's Cicero, ed. 1828, vol. iv. part ii. p. 472. But Cicero's Timæus' shows the lax way in which he translated from his Greek originals.

§ Ritter in his edition of Tacitus has given an admirable review of the early historical notices of his author contained in other writing.

[ocr errors]

to Mr. Munro, Honorius of Autun, in the twelfth century, merely quotes him second hand. Direct reference alone can prove the existence of the work quoted at the time the writer made his reference. At the same time quotations appear to have suffered comparatively little from the process of transcription.' If they serve to reveal in some cases the extent of our losses, they have rescued much of classical literature from oblivion: they have been freely used by scholiasts; and Wesseling, among others, while complaining of the indifference of previous editors of Herodotus to this source of information, acknowledges its aid in supplying lacunæ and confirming doubtful readings.

But the evidence of age, sufficient, as we have seen, to establish, in conjunction with other lines of proof, the connexion of existing copies with their lost originals, has a more immediate bearing upon their authority. Inasmuch as corruptions must increase with the process of transcription, it is a mere truism to assert that the nearer you approach the author's lifetime the purer will be the text, there being less chance, in proportion, of the multiplication of error by time. And, accordingly, broadly speaking, the oldest manuscripts are the best. But this rule is subject to restrictions the later we advance; and its inconclusiveness is shown by the fact, to which we shall hereafter advert, that no existing manuscript, however ancient, is free from corruption of some kind. It is an axiom in collation that a younger copy will often supply the defects of one of higher antiquity:

'Some manuscripts (says Mr. Taylor) as late as the thirteenth or even fifteenth century, afford clear internal evidence that, by a single remove only, the text they contain may claim a real antiquity, higher than that even of the oldest existing copy of the same work. For those older copies sometimes prove, by the peculiar nature of the corruptions which have crept into the text, that they have been derived through a long series of copies; while perhaps the text of the more modern manuscript possesses such a degree of purity and freedom from all the usual consequences of frequent corruption, as to make it manifest that the copy from which it was taken was so ancient as not to be distant from the time of the first publication of the work.' (Pp. 14-5.).

These considerations plainly show that the authority of manuscripts must be estimated by the relation of parts to the whole, not by isolated copies. Codicibus bonis, says Madvig, obtemperandum est, non serviendum. The age of the text rather than of the copyist is the thing for critics to ascertain; and Erasmus justly rebuked the prevailing vice of editors in his

[ocr errors]

day when he remarked nimis imperiti est hominis, libros ' annorum numero æstimare, ac non potius rerum indicio.' Even mere fidelity of transcribing, however meritorious, is not evidence of the internal goodness of a manuscript; for it might have been most faithfully copied from a corrupt original. The merit of antiquity, in brief, consists in the groundwork it affords for explaining the subsequent introduction of error. We shall have occasion, later on, to observe how the neglect of these simple principles by early critics and first editors led to two opposite faults, each showing an ignorance of the relative value of manuscript authority-the one, an exclusive adherence to some vetus exemplar 'surdum plerumque oraculum,' as Wolff remarks, nisi constanter consulentibus-the other, an indiscriminating regard for mere numerical testimony, irrespective of its source.

[ocr errors]

The aggregate authority of existing manuscripts depends upon their relation to the archetype, either known or inferred; but even that point at which a common parentage can be assumed, denotes an interval, more or less in proportion to their antiquity, from the authors' original. We can only sketch in outline the leading principles of such an inquiry; and it will be convenient, for that purpose, to distinguish their application to three separate epochs which appear to us to govern the distribution of existing materials. Without assuming too arbitrary a division of time, the first may be taken to extend from the date of classical authors to the seventh century, the nadir of the human mind in Europe, and the period when the stock of manuscripts may be said to have first reached its minimum.

pro

As might be expected, few classical manuscripts of this epoch exist, though, fortunately for their preservation from decay, they were written on the most durable materials. Some there are, indeed, whose probable age exceeds that of the oldest biblical MSS. extant. Of the Greek copies of the Scriptures, the four most ancient, which we give in their bable order of precedence, occupy different periods in the fourth century. These are the famous Codex Sinaiticus, unearthed by Tischendorf in 1859; the Codex Frederic Augustus, his earlier discovery in 1846; the Codex Alexandrinus, which was presented to Charles I. by Cyril, patriarch of Alexandria, and is now in the British Museum; and the Codex Vaticanus of the Italians. To the fifth century are assigned the Codex Ephrem of the French, a palimpsest, and the Codex Beza, found by Beza in the monastery of St. Irenæus at Lyons, and given by him in 1581 to the University of Cam

bridge. As regards the Hebrew manuscripts of the text, if we assume with Mr. Forsyth that the subscription on the Pentateuch roll brought from Darghestan is an accurate index of its date, we have nothing more ancient than A.D. 580. It is enough merely to observe that this represents an interval of about a thousand years from the time of Ezra, one of the latest of the Old Testament writers. In this respect the majority of classical manuscripts have the advantage. It is true that the oldest known copy of Herodotus does not go beyond the tenth century. Between Eschylus and Sophocles and their earliest archetypal codex, the parchment of the tenth century in the Laurentian Library, the gap is still longer; and the Townley MS. of Homer, which is probably the oldest complete copy of that poet, is as recent as the thirteenth century. On the other hand there are some, among the earliest copies of other authors, where this interval assumes far more modest proportions. Putting aside some few fragments that are supposed to be older than the Christian era, perhaps the oldest classical manuscript of importance is the Vatican palimpsest of Cicero's De Republicâ discovered by Mai-one of the latest, if not the most valuable of the results first disclosed in the last century by that new source of ancient learning, and carried back by competent judges to

[ocr errors]

* Marsh's notes on Michaelis, vol. ii. p. 720. Dr. Whiston, with suspicious minuteness, assigned it to A.D. 130, thirty years after the death of St. John, and Dr. Kipling, who published a facsimile in 1793, thought this estimate not impossible.' Compare Nouveau Traité de Diplom., vol. iii. p. 37, where the learned Benedictines are more sceptical.

f Dr. Lowth, in reviewing the results of Kennicott's earlier researches, has not omitted to explain how abundantly this defect of age, as regards the formation of the text, is compensated by the authority of ancient versions made 'from different languages, in much earlier times, ' and from manuscripts in all probability much more correct and perfect 'than any now extant.' (Prelim. Dissert. to Isaiah, p. xliv.)

Now in the British Museum (Cat. MSS. N.S., part ii. p. 37, Lond. 1841). Mr. Cureton edited in 1851 a Syriac palimpsest of about 3,000 lines, which he ascribes to the fifth century. He mentions three other fragments of remote date:-1. The Ambrosian MS. at Milan, discovered by Mai, but containing only 800 lines; 2. A papyrus roll in the possession of Mr. Banks, with 678 lines of the last book (Philol. Mus. 1832, p. 177); and 3. A MS. found in the hand of a mummy at Monfalout, with 306 lines of B. xviii. (Athenæum, No. 1141, p. 913). Zonaras relates that in the library at Constantinople, founded in the fourth century, there was a roll of dragons' gut, 100 feet in length, with the entire Iliad and Odyssey written in golden letters.

« AnteriorContinuar »