Bavarian B—experimental edition of Book 1 from clm 14482c

Today, Ovid’s famous work is one of the touchstones of ancient literature, but we have only a handful of scraps and quota� ons to show how the earliest medieval readers received and discussed the poems—un� l the Munich Bayerische Staatsbibliothek clm 4610. This commentary, which dates from around the year 1100 is the fi rst systema� c study of the Metamorphoses, founding a tradi� on of scholarly study that extends to the present day.

This appendix contains an edition and translation of the accessus and Book 1 of the version of the Bavarian B commentary found in the manuscript clm 14482c.
This edition is termed experimental since I have employed a somewhat eclectic method to achieve my goal of interpreting a problematic text, while at the same time documenting the text in the manuscript as much as possible.
The basic principles for editing and translating the text are the same as for clm 4610. However, there are some differences because the text is more problematic than clm 4610 and there are several manuscripts containing versions of the Bavarian B commentary, which gives us an opportunity to use variant readings. In the present edition, I have chosen to incorporate variant readings only where I have judged that the text of 14482c is corrupted or hard to make sense of. A more in-depth discussion of the problems in the text will follow below.
The reason for choosing this part of the commentary to edit are threefold. Besides the simple fact that it is the first part of the text, it is also the book which received the greatest and most varied amount of explanations. Secondly, it is proportionately the longest individual part of the commentary. The commentary comments on all fifteen books of the Metamorphoses, but the accessus and Book 1 equal roughly a fifth of the entire text. Thirdly, because of the varied types of explanations as well as the amount of textual difficulties in Book 1, it is also the most difficult part of the text to edit, which almost demands that a responsible editor meets the challenge.
The reason for choosing clm 14482c is simply the fact that it appears to be the oldest version of the Bavarian B commentary, while at the same time having the longest text (as far as Book 1 is concerned). However, upon working with the text, it turned out that it contains many textual errors and problematic passages. All the other versions of the commentary also seem to contain errors, but clm 14482c would appear to stand out and, for this reason, it is uncertain if it would be wise to edit the entire commentary based on clm 14482c.
For further discussion of the Bavarian B commentary and the relationship between the different manuscripts as well as its relationship to clm 4610 see Case Study 2.

THE MANUSCRIPT
The manuscript is described in chapter 5. The following are some brief additions relevant to this edition.
The booklet 14482c consist of 23,5 folios and is found on fol. 27r-51r (l. 12). The text of the accessus and Book 1 consist of ca four folios and is found on fol. 27r-31r (l. 16).
There are no substantial marginal additions in all of 14482c. On the relevant folios, there is one short addition in margin (28r), one unidentified symbol or letters in margin (30r), three short corrections (27v, 29r, 29v), and the occasional interlinear corrections by the same or a contemporary hand. There is also a hole in the manuscripts on fol. 28, but the text is written around the hole.

Edition and translation
I employ the same principles as those used for clm 4610 as far as orthography, punctuation, mise-en-page, and translation are concerned. This also holds true for the method employed for marking editorial interventions, with the additions that in this edition I employ asterisks (*) to mark words, phrases, or entire passages I consider suspect. These passages are then provided with alternative readings when available and further elaborated on in the translation, in which the problematic passages are also marked with asterisks.

Apparatus
Besides the system used for clm 4610, this edition employs a third apparatus, which documents alternative readings for problematic passages from the other Bavarian B manuscripts. I have chosen to adopt a third apparatus and thus to separate these variant readings from the apparatus criticus for the purpose of clarity and to emphasise that the variants reported are a selection, not a complete set of variant readings to every passage in the commentary.
The text in the third apparatus is a pure transcription from the manuscripts. Abbreviated words have been expanded, but the text has not been provided with punctuation or expanded lemma. Capital letters are represented in the way they are found in the manuscripts.

Further remarks on errors and emendations
The accessus and Book 1 in 14482c consist of 304 lines on eight and a half pages. On these 304 lines I have thus far made 96 editorial interventions. These interventions are mainly corrections of different sorts, but also include lines or passages marked as problematic (counted as one intervention) and which may, upon further analysis, turn out to consist of more errors. This means that almost a third of the lines contain an error or problematic passage. This can be compared to clm 4610, where I have made 19 interventions on the approximately 190 lines that make up the accessus and Book 1. With such a high rate of errors, it is hard to describe clm 14482c as anything else than a problematic text.
The state of the text must be kept in mind when choosing a method to edit it. I have not been able to solve every problematic passage in the text, and even if I had, I am not sure it would be wise to completely 'clean' a text such as this. On the other hand, if we return to what I said above about documenting the text and the fact that the text is highly problematic, it is possible to imagine a method where one makes no corrections at all, but simply marks the problematic passages and then comments on them in the translation. In this case, however, the translation would rapidly develop to an outright commentary of its own.
The method I have chosen seeks a compromise between these two methods.
The number of errors and the fact that they often seem to be the result of misunderstood abbreviations and letter forms would suggest that the scribe had some difficulties with the exemplar. The following is a small sample of different types of errors: There are the usual instances of words (or lines) written twice, the occasional missing word, and also the common scribal error which results from either a missing or a superfluos linea nasalis. There are many errors where short abbreviated words are concerned, for example, vel for ut or quam for quem.
There are also many cases of wrong word forms, for example, promissio for promissione, suadetur for suadent, imprimum for imprimuntur. These errors may be the result of trouble with reading the script in the exemplar or misunderstood or unrealised abbreviations.
These errors cause confusion since they result in proper words and only after careful analysis can a correction be made.
Another type of error are an error that consists of short forms of words, for example, cu for cupit, fis for fistula, fenda for fecunda, which result in nonsense words. It is unclear whether these are just errors or if they are to be considered as idiosyncratic abbreviations.
Because of the many errors and the many different types of them, it is difficult to speculate about the exemplar. One would need to make a more thorough comparison with the other manuscripts and examine letterforms in texts that may be considered possible exemplars (e.g. investigate if misinterpretations of beneventan scripts may have led to the type of errors found in clm 14482c).
In the following, I will discuss some scenarios where I have employed the method of using asterisks.
Passages marked with an asterisk are generally of three kinds: The word or passage may be suspected of being erroneous, but still present a valid reading; it may be in all likeliness erroneous but for different reasons left uncorrected; and finally it may be erroneous and corrected, but still under suspicion and thus marked with asterisks.
All three scenarios are also marked and commented upon in the translation and also provided with alternative readings from the other manuscripts if there are any.
I have not used cruxes ( †) in the edition, but some of the passages marked with asterisks may turn out to be hopeless cases and thus marked with cruxes in the end.
We will start by looking at two instances where I have decided to correct the text and not mark it with asterisk. The text is thus, in a sense, considered 'done' (although the translation still comments on these passages).

Example 1
In the accessus (lines 13-14), clm 14482c has: hi furcas subire ('these thieves') and the infinitive subire ('to go or come under'), while clm 14809 has: hic furcas subiere, which corresponds to Met. 8:700. In this case, I have chosen to correct the error. It could perhaps be argued that the reading in clm 14482c could make sense and thus should not be corrected, but because of the grammatical incongruence together with the witness from another manuscript as well as the passage in the Metamorphoses, it is sensible to correct in this case.

Example 2
The explanation to Met 1:588 found on lines 383-386 is severely corrupted. However, because it is largely based on a passage from Isidore as well as the presence of an almost identical passage in Freiburg 381, it is quite easy to correct. In this passage, the lemma reads redeuntium inpix for redeuntem iupiter. The first word is at least a real form, while the second would seem to make no sense at all. Next comes the first explanation to the word flumen where the text reads aqua depresso where in all likeliness it should be aque depressio, which is comparable in meaning to Isidore's aquarum decursus. Freiburg 381 has garbled decursus into something that looks like de confusa (by turning cur-into a con-abbreviation). On line 385, we encounter one of the most severe corruptions in the text. Isidore's prior aqua quam decursus has been turned into the unintelligible priora queque cursus (the scribe seems to have merged prior and aqua, then confused a part of aqua and quam with queque, and finally possibly dropped the prefix in decursus). The last part of this explanation contains a quotation from Virgil. Here, clm 14482c has donec flumine curuo instead of donec flumine uiuo abluero (found in Freiburg 381) from Aen. 2:717. Clm 14482c seems to have been thinking of a passage in Met. 3:342: quam quondam flumine curuo.
I have corrected all of these errors, except for the quotation from Virgil, which I do not consider an error as such as it renders an intelligible text.
These passages serve as a good example of the possibilities to detect and to correct errors when we have other texts for comparison.
Next follows two examples of instances where I have decided not to correct the text.

Example 3
In the accessus (line 15), we have the obviously erroneous statuam in de statua Pigmalionis mutata in statuam iuuenis hominis. Although in all likeliness an error, this reading is present in all the manuscripts that contain this passage. We would expect a figuram or speciem here, but since all manuscripts have statuam I have chosen to not correct this passage, but to mark it as suspect and comment on it in the translation.

Example 4
In the accessus (lines 59-60), clm 14482c has the following reading: cum intendat de re de transformatione rerum Clm 14809 would seem to have the following: cum intendit dicere de transformatione rerum.
It should be noted that the reading in clm 14809 is not entirely certain on account of the text being heavily abbreviated. However, if we assume that we have these two variants to make sense of we have two possibilities: 1. when he pays attention to the thing/the matter, to the transformation of things, 2. when he intends to speak about the transformation of things, The second alternative is quite straight forward. For the first alternative to make sense we must take the second prepositional phrase as an apposition to the first. This may be the result of the phrase originally being an interlinear gloss that has then been incorporated into the main text. It is not a smooth reading, but the commentary text is not very polished and does seem to contain incorporated interlinear glosses here and there. So, it may still be taken to be a plausible, if awkward reading. For this reason, I have marked it as being perhaps suspect, added an alternative reading in the apparatus and commented on it in the translation.
Example 3 and 4 are examples of passages where I have not corrected since the text in the manuscripts may carry the intended meaning. The final two examples revolve around two passages where I have not been able to achieve a good interpretation through judicious emendation and have thus left the text uncorrected.

Example 5
In the accessus (line 70-74), we find yet another suspect passage. With a little help from Salzburg AV4 we can detect some of the errors, and with the help of the accessus in Haun. 2008 we get some further help. We can draw the conclusion that uidelicet must be an error for utilitas, uere should perhaps be uero, a verb is missing in the cum-clause, and so on. The problem here is that neither Salzburg AV4 or Haun. 2008 offer perfect matches to clm 14482c. Besides this, the meaning is also quite obscure, which makes the difference between textual error and obscure phrasing difficult to pinpoint. For these reasons I have chosen to mark this entire passage as suspect and then offer some speculations in the translation.

Example 6
The long explanation of Met. 1:563 on lines 359-377 would deserve a case study of its own. It is the most difficult and corrupted individual explanation identified in the commentary thus far. Versions of it are present in all manuscripts, but none of them is the same and only some help can be gained by comparing them to each other. Clm 14482c has the longest explanation by far, and part of the difficulty seems to be the fact that it has incorporated interlinear glosses. This makes attempts to translate the passage into a fluent English difficult and dangerous.
On lines 367-378 laurus una anjd que media may be the result of interlinear glosses having been incorporated in the main text (una was in this scenario originally written above laurus to clarify the elementary fact that the word is feminine although it has the masculine -us ending).
There also seem to be some superfluous id est-abbreviations in clm 14482c, which makes the text difficult to understand. If we compare the following passage in two manuscripts we see that the id est is either added for some obscure reason, or it may be the result of l for lauro having been turned into the typical .i. for id est.
Clm 14482c: quia nobiles quidem id est coronabantur Clm 14809: quia nobiles quidam lauro corobantur This passage also contains many other possible errors and for the same reason, as in example 5, I have chosen not to correct, but rather to mark and to comment.
Finally, I should also mention that there are two instances where I have not been able to form a sensible interpretation of the text at all. These are found on lines 277 (Quod est) and 365 (Nam ut inde et cetera). In these cases, the Latin is preserved in the translation and commented upon in the apparatus.