To begin at the beginning;—I will now suppose that I have before me several works in English, French, German, Italian, Latin, each of them purporting to be—“The History of Herodotus, translated from the Greek.” In collating these books it becomes evident that they are all derived from some one source. But it may be well to give attention to some facts at this stage of our progress.
We affirm that the Greek text of Herodotus, such as it now appears, was extant some time before the publication of the earliest printed editions. Ostensible and tangible proof of what we here allege, is afforded by the existence, at the present time, as we shall presently state, in several public libraries, of many manuscript copies of the Greek text, which, by the date affixed to them, by the character of the writing, by the appearance of the ink, and material, and by the traditionary history of some of them, are clearly attributable to different ages, from the tenth century to the fifteenth. But now if it were possible to suppose that all these copies were derived from one MS. and that one a forgery of a late date, an examination and comparison of them, and a comparison of the manuscripts with the printed editions, will furnish several special demonstrations of the point affirmed. In 1474, twenty-eight years before the appearance of the first printed edition of the Greek text, Laurentius Valla, an Italian scholar, published at Venice a Latin translation of Herodotus, purporting to have been made from the Greek. Now if, in comparing this translation with the Greek manuscripts that are still extant, it were asked which is the original, the Latin or the Greek? no one acquainted with the structure of language could hesitate in declaring for the latter; for in the Latin (as in every translation) ellipses are supplied, exegetical and connective phrases are introduced; and what is still more decisive, there are many passages in the Greek where an obvious and consistent sense is evidently misunderstood in the Latin; for Valla seems, from all his translations, to have been but imperfectly acquainted with the Greek language. In such instances the occasion of the translator’s error may often be detected; by which means incontestable proof is afforded of the fact now supposed to be questioned, namely—that the Greek is the original, and the Latin the translation. Again:—The Latin, as compared with the Greek, is deficient in many entire paragraphs, and in many single sentences. In the Greek these passages are one with the context; but in the Latin, the hiatus is either abrupt and apparent, or it is concealed by a connective sentence, evidently inserted as a link between the disjoined portions of the text. Now, when evidence like this is presented, we need not lay stress upon the traditionary history of particular manuscripts, nor upon their apparent antiquity, nor upon the genuineness of the dates affixed to them; for from the facts actually before us, we can draw only one inference. Without going further, therefore, we may conclude with certainty, that several Greek manuscripts of Herodotus were in existence some time before the publication of the printed editions; and by consequence, the averments of the first editors are confirmed, who declare that they derived their text from manuscripts—already known to the learned.
The Greek text of Herodotus was, for the first time, printed by Minutius Aldus, at Venice, September, 1502. Copies of this beautiful and correct edition, “corrected by a collation of many manuscripts,” are still extant:—it is distinguished by its retention of the forms of the Ionic dialect—a proof that the editor followed a pure and ancient manuscript, for the Ionic forms are generally lost in those copies, the text of which has passed through many transcriptions. This edition, with corrections and notes, was reprinted at Basil, in 1541, and again in 1557, by Joachim Camerarius. In 1570 the Aldine text of Herodotus was printed at Paris, by Henry Stephens, who does not profess himself to have collated manuscripts. The title-page declares that the books were “ex vetustis exemplaribus recogniti:” but in his second edition, Stephens confesses that up to that time he had not been able to procure an ancient copy by which to correct the text; he must, therefore, in the phrase just quoted, be understood to refer to the manuscripts that were consulted by Aldus. G. Jungerman, assuming the edition just mentioned as the basis of his own, in which however he made, without specification, many conjectural emendations, printed the Greek text, at Frankfort, in 1608. This was the first edition in which the text was divided into sections, as it now appears. The London edition, dated 1679, and published under favour of the name of the learned Thomas Gale, was derived, without acknowledgment, from that of Jungerman. Hitherto the editions were only successive reprints of the Aldine text; and came, therefore, all from a single source; but in 1715, an edition of Herodotus was published at Leyden, under the care of J. Gronovius, who collated the former editions with some manuscripts before unknown, or not examined. A Glasgow edition appeared in 1761; and two years later that of Wesseling, printed at Amsterdam. Some quotations from this editor’s preface will give the general reader a good idea of the method of conducting these literary labours, and of the security afforded for the purity of the text of ancient authors. Several German and Dutch editions have appeared since that of Wesseling; the most esteemed are those of Borheck, Reiz, Schaefer, and Schweighæuser. Of the laborious care bestowed by the learned editors upon these editions, the following citations from their Prefaces will give evidence. Wesseling says:—
“The forms and proprieties of the Ionic dialect I have restored, wherever they could be gathered clearly from the ancient codices, and have replaced some readings which, without cause, had been rejected. Innumerable passages I have relieved from errors, yet very rarely on mere conjecture, and only in those words which the genius of the language would not admit; and in many instances have thought it enough just to point out the means of amending the text, where it is evidently corrupted.” In quoting this passage from Wesseling, Schweighæuser says, “Neither have we, except in a very few places, admitted conjectural emendations into the text; and these only where it was evident that all the readings of all the existing copies were corrupted; and where an emendation presented itself which not merely seemed probable, but which was so clear and certain as to need no argument in its favour.” Very judiciously, this editor refuses to impute to the temerity or ignorance of copyists all the variations from the Ionic forms; since it is evident that Greek writers who adopted one of the dialects, allowed themselves the liberty of occasionally using the common forms of the language: he therefore restores the ionicisms only when he has the authority of MSS. for so doing. Of Wesseling’s extreme caution, Schweighæuser thus expresses his opinion:—In this edition, excepting a few errors, easily corrected, or some cases which may be open to disputation, the learned have nothing to complain of; unless it be, that, in adopting better readings, warranted by MSS., as well as in correcting, on probable conjecture, some places manifestly faulty in all copies, the Editor was too timid—so much so, indeed, that many approved readings which he might well have admitted into the text, he ventured not to adopt. And often he preferred to leave, untouched, manifest and gross corruptions, rather than to put in their place his own emendations, or those of others, though decidedly approved by himself. As to conjectural emendations, even in those places where all the MSS. are plainly in fault, we have seen him, in his preface, ingenuously confess that he had rather be thought too cautious, than too bold: and who would not esteem, yes and admire, rather than condemn, this illustrious man, blaming his own timidity in this sort:—“In attempting to restore the language of Herodotus, I have been restrained often by more than a due timidity; but such is my nature.” This editor, in his preface, states that, having been applied to, to superintend a reprint of Wesseling’s Herodotus, he had declined doing so, unless he should be able to obtain, from the French king’s library, the loan of the MSS. of Herodotus, there preserved:—the troubles of the times preventing this, he sought for some one, residing at Paris, who would freely undertake the irksome and painful toil of collating Wesseling’s text with all those codices; and at length, by means of a learned friend, he met with a young man, a native of Greece, who executed the task of comparing the text—word by word—with the five principal manuscripts in the library, and making a separate list of the various readings in each.
From the mass of variations brought before him, the office of the editor is to select that one which most recommends itself, either by the superior authority of the codex in which it appears, or by its particular probability, or seeming accordance with the author’s style or meaning, or with the proprieties of the language. And not seldom it happens that the most inferior copies have chanced to preserve an evidently genuine reading, where the best have, as plainly, erred.—“No MS.,” an eminent critic has said, “ought to be thought unworthy of being consulted.” Yet in cases of importance, where there may be room for doubt among the existing variations, the canon must be obeyed which enjoins that, “Codices should rather be weighed than numbered.” Although discussions on subjects of this kind cannot but seem uninteresting, and even trivial to general readers—and perhaps absurd, when the gravity and strenuousness with which, sometimes, the most minute points are argued, is observed; yet it ought never to be forgotten that the credit, the purity, and the consistency of ancient literature, are very greatly promoted by the indefatigable zeal of those who devote their lives to these learned and unattractive labours.
But I now look into some of the printed editions. For instance, here is a small folio volume, in excellent style, as to type, and paper, and execution, printed in Paris, MDLXX, and edited by Henry Stephens. I have also in hand the edition edited by J. Schweighæuser, in four volumes octavo, reprinted in London, 1822; and also a more recent edition, namely—that of Professor Gaisford, in two volumes octavo. Besides these there are ten other editions of the Greek text—German, Dutch, and English. I open these several editions, at hazard—say at the beginning of the third book—Thalia: I find that they correspond, word for word, for some way on; but in the fifth line I find an unimportant variation—one form of a word is used instead of another; and further on the order of the words is a little different, but the sense is the same. Sometimes one particle or expletive is used instead of another; sometimes those expletives that barely affect the sense in any way, are omitted. Frequently the orthography of proper names is differently given in the different editions. Very rarely are these variations of so much importance as would affect the sense in a translation. But now, from the fact of the verbal identity of these editions throughout by far the larger part of them, and also from the occurrence of not infrequent, and yet inconsiderable differences, I infer, first—that they have all had a common source in some one original exemplar; and, secondly—that there have been many copyings from that first copy; and that it has been in the course of these repetitions, in which the ear, the eye, and the hand of many writers have done their part, that these departures from the author’s first copy have taken place. In a word, the printed editions have followed manuscripts; and these have undergone those chances, and those mischances which, in the ordinary course of things, must attach to a process like this, notwithstanding the care and the fidelity of those who practise it.
The next step, then, is to make search for those ancient manuscripts, or for some of them, whence these printed editions have been derived. About fifteen such manuscripts are now known, and may be inspected in public or private libraries. One of the purest of these is preserved in the French King’s library (now the Imperial) and it is thus described.—It is a parchment in folio, purchased in 1688, containing the nine books of Herodotus. This codex is by far the best of all, and appears to have been executed in the 12th century. It is distinguished by its uniform retention of the forms of the Ionic dialect—an indication of the antiquity and purity of the copy from which it was derived. The same library contains also several other MSS. of this author, which are thus described—A codex on paper, formerly belonging to the Colbertine library, containing the nine books of Herodotus: in the margin are notes of some value. This MS. was executed in 1372. A copy on paper, written in the year 1447. The negligence of the copyist is, in this instance, much to be complained of, for sometimes entire phrases are wanting. Yet it contains some readings that deserve attention. A MS. on paper, dated 1474. Besides the nine books of Herodotus, this codex contains parts of the works of Isocrates, and Plutarch, together with a lexicon of words peculiar to Herodotus. A MS., which along with extracts from several Greek authors, contains part of the first book of Herodotus, as far as c. 87. Although this codex is of late date, the extract from Clio appears to have been made from a very ancient copy. Some other codices in the same library afford also parts of our author’s work. There is a codex formerly in the Florentine library, which from the condition of the parchment, and the antique style of the writing, is manifestly of great antiquity. Montfaucon assigns it to the tenth century. This codex belongs to the same family as that of Askew, and the Medicean. Yet neither was it copied from the latter, with which, indeed, it might dispute the palm of excellence; but being derived from a more ancient source, it offers many approved readings, differing from the Medicean, where that is in fault, or where it offers no emendation of the common text. This Medicean codex is thus described in the Catalogue of the Florentine library: “Herodotus:—a very ancient codex, valuable beyond all praise. It is on paper, in quarto, well preserved: executed in the tenth century. The titles of the books are in uncial letters of gold; it contains 374 pages.” This copy was followed with a too superstitious reverence by Gronovius; yet being compelled to consult it in the public library, and under the eye of the librarian, he has not seldom mistaken its readings. A MS. of Herodotus, formerly in the library of Archbishop Sancroft, and afterwards in that of Emmanuel College, Cambridge, has been deemed of high antiquity, and great value. The libraries of Oxford contain also some codices of our author, and several are known to be in the possession of private persons. “These manuscript copies,” says Wesseling, “brought to light from various places, have not, it is manifest, originated all from one source (in modern times). Where the copy followed by Valla is torn or defective, there also the Vienna, the Vatican, and the Oxford MSS. are wanting. And in what these are remarkable, so is the Florentine. But the Medicean MS., that of Cardinal Passio, and of Askew, for the most part agree. The three first mentioned, seem to have been derived all from some one more ancient parchment, the writer of which, offended perhaps at the frequent digressions of the first book, very daringly cut them all off; and lest the hiatus should seem harsh, he skilfully fitted the parts, so as to preserve the continuity of the style. The three last, on the contrary, were derived from the copy of a transcriber better informed, who scrupled to make any needless alterations. A great number of the various readings which distinguish these MSS. are attributable to the copyists who have substituted the common forms of the language, and words better known, in the place of the Ionic forms and of obsolete words.”
All that is of any importance in proof of the genuineness and integrity of ancient books, is to know that there are now in existence several copies, evidently of older date than the first printed edition of the author; and that these copies, by their general agreement, and, not less so, by their smaller diversities, prove, at once, their derivation from the same original, and their long distance from that original; since many of these diversities are such as could have arisen only from many successive transcriptions. Beyond these simple facts, the knowledge of codices, and of various readings, is interesting to none but editors and critics.
We may now fairly assume as certain, so much as this—that the work before us—mainly such as we now have it in our hands, is an ancient work, and that it has come down to modern times in that mode of which, in the preceding chapters, we have given some account, and have adduced several instances. Our next question is this—To what age this work ought to be attributed? Or this—When did the author live and write? In obtaining an answer to this question, or to these two questions—considered as one, we must look to that succession of writers, retrogressively examined, who mention Herodotus, and his History, who describe it, and make quotations from it, or who give summaries of its contents. The proper and the most complete proof of the antiquity and genuineness of ancient books, is that which is thus derived from their mutual references and quotations. There is an independence in this kind of evidence which renders it, when it is precise and copious, quite conclusive. It is not the evidence of witnesses, who first have been schooled and cautioned, and then brought into court to do their best for the party by whom they are summoned; but it is the purely incidental testimony of unconnected persons, who, in the pursuit of their particular objects, gather up, and present to us, the facts which we were in search of. Besides—these facts have a peculiarity, which renders them eminently capable of furnishing precise and conclusive proof. A book is an aggregate of many thousand separable parts, each of which, both by the thought it contains, and by the choice and arrangement of the words, possesses a perfect individuality, such as fits it for the purpose of defining or identifying the whole to which it belongs; and if several of these definite parts are adduced, the identification is rendered the more complete. This kind of definition is moreover capable of being multiplied, almost without end; for each writer who quotes a book, having probably a different object in view, selects a different set of quotations, yet all of them meeting in the same work. We are thus furnished with a complicated system of concentric lines, which intersect nowhere—but in the book in question.
Then it is to be remembered that each of these quoting writers stands himself as the centre of a similar system of references, so that the complication of proof becomes infinitely intricate, and therefore it is so much the more conclusive. It is again involved, and so is rendered secure, by the occurrence of double or triple quotations; for example—Photius quotes Ctesias—quoting Herodotus. The proof of genuineness in the instance of a standard author, is by such means as these extended, attenuated, and involved in a degree to which no other species of evidence makes any approach.