Appendix C: How to Collate a Latin MS

What collation is

By “collating” a MS. we mean comparing it with the received text of an author and taking note of the points in which it differs from that received text. The first thing we must do then when we take a MS. to collate is to get a copy of that critical edition of the author which is provided with the best and fullest critical apparatus. Suppose, for example, we have a MS. of Martial to collate: we must use Schneidewin's critical edition of Martial to compare it with; for a MS. of Terence we must use Umpfenbach; and so on, always taking care to state at the outset of our collation with what edition (giving the name, place, and date of the book) we are collating our MS. We set ourselves then to compare line by line, word by word, letter by letter the MS. before us with the printed edition of the author. Wherever we come upon something in the MS. different from what is in the printed edition, we write down side by side in parallel columns the reading of the printed edition and the reading of the MS. We shall find it most convenient to put the readings of the printed edition in the first or left-hand column on our page, and the readings of the MS. in the second column, the column to the right of it; for by so doing we shall be able to collate other MSS. of the same author on the same paper by merely putting additional columns to the right; and so we shall come to have on the same pages a conspectus of the readings of a number of MSS. of the same author side by side. For the same reason it is best to use the broad side of the page for writing on, so as to have room for as many additional columns, and so for the collation of as many additional MSS., as possible.

We shall have then in the first or left-hand column the readings of the received text from which any of the collated MSS. diverge, and in the columns to the right the divergent readings of the various MSS. And of course we shall have to give the reference to these words or lines that we take note of as being different in the received text and in the MSS. we have collated, so as to be able to find them when we wish to consult them in the printed edition or in any of the MSS. The reference to the reading of the printed book will be given in the ordinary way, e.g. Martial bk. xiv, epigr. i, v. 1 (Mart. xiv. 1. 1); but we cannot give a reference of this kind to the reading in the MSS., because the epigrams and lines there are not numbered as they are in our modern printed editions. We give the whereabouts of a word or sentence in a MS. in a different way, viz. by giving the number of the page on which it is found. Or rather to be strictly accurate I should substitute the word leaf, or its Latin equivalent folium, for page; for in a MS. we do not number the pages but the leaves, and what would be in a printed book pages 1 and 2 will be in a MS. leaf one, or folium primum. Page 1 is the obverse side of leaf one, and is technically known as folium primum rectum. Page 2, the reverse side, is technically known as folium primum versum; page 3 in the same way will be folium secundum rectum, page 4, folium secundum versum; and for shortness' sake we shall write them “fol. 1 r,” “fol. 1 v,” “fol. 2 r,” “fol. 2 v,” and so on. A word then which occurs on the third page of a MS. will be referred to “fol. 2 r.”

But further, the page or side of a leaf in a MS is very often written in two columns; so we shall want to state in our reference in which column the word occurs, and we shall have to add to our “fol. 2 r” something to show whether it is in the first or the second column of “fol. 2 r.” The neatest notation for the columns is by the Greek letters α and β. If our reading then be in the first column on the page we shall give the reference to it in the form “fol. 2 r α,” and we shall put this reference on the right of the column in which the readings of the MS. are given, and the reference to the reading in the printed edition on the left of the left-hand column, so as to allow the two contrasted readings to come as close as possible, that the eye may note their divergence at a glance.

This that I have described is the full style of collating a MS. Of course in most cases a much less troublesome plan will serve, viz. to jot down on the margin of our printed edition the readings of the MS. we are collating. But if we do this, unless the margin is a very wide one, it will be difficult to use the same book for the collation of more than one MS.; and it will be difficult to keep the readings of one MS. distinct from those of another, unless we write the readings of one MS. in black ink, of another in red ink, of another in purple ink, and so on; so that it may really involve as much trouble in the long run as the more detailed method of collation. Besides one has not room on the margin of a printed book to write remarks about the readings in the MS., such as whether they are written by the writer of the MS., or by some one who revised it afterwards, or by some one into whose hands it came at a much later time; and this is a thing of the utmost importance, and indeed is what makes the chief difficulty in collating a MS. accurately.

Dealing with corrections in the MS

Our first object in collating a MS. is to take note of what was actually written by the writer of the MS. But in almost every MS. we look at we find this has been made more or less difficult by the fact that there are readings given in the margin or written above the words in the text in different ink — I mean less faded — and in different handwriting from that of the text. These have been added by various owners of the MS. at various periods who found lines here and there unintelligible or ungrammatical, and corrected them according to their own fancy or according to the text of some other MS. of the author that they managed to get a sight of, just as people nowadays have a habit of correcting errors of printing in any book they happen to be reading. Now such variants as are mere conjectural emendations of Renaissance owners have as a rule little interest for us. We might omit to notice them altogether; but it will be safer, if we wish to make a complete collation, to give them within brackets and with a note of their being due to some one much later in date than the writer of the MS. If the reading is in the margin, we shall put this note in the form rec. in marg., that is “recent or by a later hand in the margin”; if it is written above the word in the text, we shall write rec. sup. scr.; if, as is the commonest case, it is written on the word in the text, if the word in the text has been by a line here and a curve there corrected or transformed into the new reading, so as to make it often difficult to find out what the original reading was, we shall say corr. rec., that is “corrected by a recent hand.”

Such corrections of a manifestly much later date than the MS. itself have little interest for us, unless they let us know the readings of another MS. which the owner of our MS. had, but which has now been lost. But we very often find corrections which are patently not so recent but of the same date, or nearly of the same date, as the MS.; and these are of great importance, for they must have been made by the copyist of the MS. himself or by some one who revised and corrected the MS. immediately or very soon after it was copied. When we are quite sure the correction has been made by the copyist himself, in other words, when it is merely a correction of a slip of the pen, it is unnecessary to notice the original reading, unless we are making an extremely accurate collation; for our real aim is to find out what was in the MS. which the scribe copied; and it is usually needless to make record of the fact that our scribe in a moment of forgetfulness wrote “at” instead of “et,” if the next moment he saw his error and amended it by correcting the “a” into an “e.” But where the correction is or may possibly be by a different person, of the same or not a manifestly much later date than the date of the MS., we must notice it; for it has probably been made by some one who compared the copy which the scribe had made with the original, and corrected it here and there where the scribe had departed from the original (cf. p. 41). Whenever there is a possibility of a correction being of this kind we shall record it with the word corr. before it; and after we have collated the whole MS. we shall generally find ourselves able to tell what value these corrections as a whole have, and how they came to be there.

There is one kind of correction that we cannot date, and that is an erasure. We can tell by the look of the ink in an ordinary correction whether the writing is old or recent, but we cannot tell by the look of scraped vellum whether the erasure was made on the one hand by the copyist or a contemporary, or on the other by a later owner of the MS. And still worse, we can with a little patience disengage the original reading where a word has been rewritten, but we can only rarely and with a great straining of our eyes read a word that has been erased, especially if a new word has been written in the space occupied by the erased one. An erasure then is a thing that must always be taken note of, and in collating a MS. our eyes must explore the vellum surface as well as the writing; for a scrape on the vellum means that there once was writing there. If we are fortunate enough to find out by the help of a magnifying glass what the erased letter or word was, we shall put it down as the reading of the MS., and add the emended form of the word after the note corr, ras. (i.e. rasurā); if we cannot, we must take note of the fact that there is a trace of a lost letter or word, and we must indicate the position of this erasure by stating what written word or letter it comes before or after. Thus if demo were corrected by the erasure of m into deo, and if the m could not be read, we must indicate the fact that only one letter has been erased, and that before the letter o, by putting one asterisk to represent the erased letter, thus de*o. Lastly, if a word or letter has been erased and the space has not been left vacant, but a new word or letter has been written on it, we shall give as the reading of the MS. this new word or letter, adding in ras., i.e. “standing upon or written over an erasure”; and if the new word or letter is plainly by a later hand, we shall say rec. in ras.

Where we are able to refer the various readings or corrections to distinct hands we should specify this by putting manus prima (shorter m.1) for the work of the scribe himself, manus secunda (m.2), manus tertia (m.3), and so on for the several correctors; but it is not always possible, nor indeed always necessary, to make these distinctions.

Which differences are significant

What sort then of divergences from the received text should be noticed? Are we in a Latin MS., e.g., to notice such a thing as quamquam instead of quanquam, coena instead of cena, etc.? In most cases, unless the MS. is of such an age that its spellings have some authority, and unless our collation is to be an extremely minute one, there is no need for burdening ourselves with a mass of useless details like that. Only we must state at the outset of our collation exactly what divergences of reading we do not profess to take note of. In most Latin MSS., for example, we need not record such divergences as h omitted or inserted; uu and uo; e and ae; c, p, t for ch, ph, th; y and i; ti and ci; f and ph; c and qu; oe and e or ae; the separation or fusion of words, e.g. sed et for sedet or sedet for sed et; capitals and small letters; -is and -es in the plural; -em and -en in acc. sing. of Greek names; quidquid and quicquid; quamquam and quanquam; and so on, because they do not really testify to the spellings of the original from which our MS. was copied. Although the copyist saw quamquam in the MS. before him, it is quite possible that he might prefer to write quanquam, or febus instead of Phoebus. But it is inadvisable to lay down any general rule as to what divergences are not worth noting. That is a question to be decided according to the character of the MS., the state of the text of the author, the aim of our collation, and other considerations. The one essential thing is that we state clearly at the outset what variants our collation does not profess to take notice of.

It remains to mention some other formulae that we want, such as om. when a word or line has been omitted, bis when it has been repeated by mistake, transp. when the order of two words or lines has been transposed. The rest can easily be picked up from any critical edition of a classical author. Without wasting time on enumerating them all, I shall pass on to what is more important, viz. the way to describe a MS.; for at the outset of our collation we must give an account of the MS. we are collating. If there is a catalogue of MSS. in the library, one may copy the description given of the MS. there; if not, one must make the description for oneself, and in this way. First we give the title, home, and press-mark of the MS., e.g. Codex Martialis Oxonii in bibliothecâ Bodleianâ Add. MSS. 12345. Then must follow an account of the material of which the MS. is composed, whether vellum (membran.) or paper (chart.), of its size, whether folio, quarto, or octavo, of its date, of the no. of its leaves, and of its state of preservation. Thus: codex chart(aceus) in ivto majori, saec. xi. in(euntis), foll. 100, pict(uris) orn(atus), mut(ilus). Then an account of its contents: continet Martialis Epigrammatum libros i-x, xii - xiv; desunt lib. xi et Spectacula. So much is absolutely necessary, and we may add any details about the MS. that we think should be mentioned.

Goal is the reconstruction of the original

The thing to be remembered in our description of the MS. and in all our collation is that our aim must be to find out the readings and the character of the original MS. from which our MS. was copied. The textual emendation of an author becomes a hopeful thing as soon as we are able to classify the MSS. of the author, to say that so many have all been copied from one archetype, so many from another, so many from a third, and so on; when we can reduce the host of existing MSS. of the author to two or three groups, and determine from the readings of the individual MSS. the text of the two or three archetypes from which they have all come. The number of fifteenth-century MSS. of our author may appear unwieldy at first, but it is possible in time so to manipulate them as to elicit from them the readings of, let us say, three ninth-century MSS. which have been lost to us, but which we can reconstruct, as it were, can put together piece by piece, from the traces which the later MSS. supply. And that is where the advantage appears of minute collation, of noting each and every divergence or peculiarity of a MS. In tracing the pedigree of a MS. these minutiae are extremely helpful. The mere similarity of text is not sufficient of itself to enable us to tell that one MS. has been copied from another. But if a peculiarity in the reading of one can be explained from some little accidental circumstance of another, such as the cases mentioned on p. 65, we have tangible proof of the connexion of the two. The safest rule to follow in collating a MS. is therefore to take note of as much as our time will permit, to deem unworthy of notice as little as possible. The fact that a scribe first wrote at, then immediately corrected it to et (p. 122 above) may, for example, indicate some peculiarity of his original, whether that the form of its a was like an e, or that it had at corrected in the margin or elsewhere to et, or that its reading was at, while et is an emendation due to the scribe of the copy. It may equally be due to a mere clerical error that has no such significance.

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 United States License.

An XML version of this text is available for download, with the additional restriction that you offer Perseus any modifications you make. Perseus provides credit for all accepted changes, storing new additions in a versioning system.

hide Display Preferences
Greek Display:
Arabic Display:
View by Default:
Browse Bar: