« Without a Benevolent Spider, I Might Not Be Here Today
Main
In Which Direction Should One Write? »
September 10, 2007
A Probable Possible Prehistory of KTU 1.85
Schaeffer found four, fragmentary, instances of the Ugarit veterinary text KTU 1.85: KTU 1.85 itself (RS 17.120, the most complete and therefore the tablet used to identify and reference the text), KTU 1.71 (RS 5.300), KTU 1.72 (RS 5.285) and KTU 1.97 (RS 23.484). This post is part of a series on these tablets. You can find more, including links to the texts and translations, at the series gateway post. In this current post, I will look at the similarities and some of the differences between the texts on these tablets to determine the history of the text if such can be determined. One of the major problems in this endeavor is to sort out what may be scribal errors and what may reflect different textual ancestries among the four tablets. Pardee, 71-72, has a convenient list of variants that I will use in much of this discussion.
Summary:
Often I summarize this kind of post at the end. In this case, I'll state my conclusions right here and let the gullible and the indifferent go about their business while the skeptics, and more dedicated plow on through the details. Here is how I see the development of the text on these four tablets.
1) Some, now lost but rather extensive, list of symptoms and treatments for horses, possibly in Akkadian or some other language (Hittite or Hurrian?) but as likely in Ugaritic, was excerpted to a much smaller list, perhaps for a scribal training. When or where is not knowable on the current evidence. If there were an Akkadian (or Hittite or Hurrian) excerpt from which the Ugaritic version was translated is also not knowable on the current evidence.
2) Sometime in the history of the excerpt, or even at the time a master scribe excerpted the shorter list, what is now a somewhat garbled section 9 was formed from two (or more) sections in the original longer list.
3) From this original excerpt, two recognizable text families developed: one, consisting of KTU 1.85 and KTU 1.97, that I call "the scholar's text family," and the other, consisting of KTU 1.71 and KTU 1.72, that I call "the acropolis text family." These text family names are more for convenience than anything else. The acropolis text family is not less the product of scholars than is the scholar's text family.
4) The acropolis text family either has a single vorlage or one of the tablets was copied (or dictated) from the other or from a very near relative. It is unlikely that two student scribes produced these two tablets at the same time from a master's dictation.
5) The relationships within the scholar's text family are more complex; although KTU 1.97 could not have been copied from KTU 1.85, even via dictation; but even the other way around is not likely. It is possible to see an extra step between the vorlage of this family in KTU 1.85 that is not seen in the textual history of KTU 1.97.
All of this is very tentative and rather speculative.
Now if you want to know why I think such things, you'll need to plow through the following and it is possible you will still be shaking your head. I still shake my head at all this more than a little.
The Gory Details
Perhaps I should start with a couple of words about methodology. Everyone who has studied these tablets has seen a variety of scribal error in each of them. For this reason, I have decided to work from the largest common differences to the smaller ones. In other words, I give added significant to a missing of added section and less weight to smaller word level commonalities or differences when determining textual families. Simply put, I think two tablets that have a section that the other two do not have is more significant than a particularly verbal structure or an isolated gloss. Sometimes the smaller commonalities and differences reinforce the family divisions I am about to suggest, on other occasions they do not. I will try to explain both types of occurrences. I also think that, in the case of two of these tablets over against the other two, the archaeological context is important. This is particularly true when other similarities and differences are readily explainable. It is true that much of what follows is based on significantly underdetermine theory and each of these four tablets may have a single common origin or some set of
Two Families
My section 8 appears on two of the tablets but not on the other two. While a section 8 is represented on KTU 1.85 and very likely on KTU 1.97; there is no evidence for its presence (and evidence against its presence) on KTU 1.71 and KTU 1.72. To be sure, section 8, on both KTU 1.85 and KTU 1.97, is so fragmented that almost nothing can be made of it. On KTU 1.85 the scribe lines separating the three lines of the text of section 8 from section 7 above and section 9 below are very clear and the first few letters of each line are readable as are several letters at the end of the first line (KTU 1.85:20); all else is completely broken away or otherwise unreadable. KTU 1.97 has a clear scribe line after section 7. The two letters śś are discernable in KTU 1.97:8' after which nothing else is readable on the obverse. One can reasonably restore a w after the śś and see "horse" at about the place on the line where one would expect it. Pardee, 36, reads the letter ṣ followed by a space and before śś[w]. In this, he follows Dietrich, Loretz and Sanmarin, 112. I use Pardee's line numbering scheme for KTU 1.97. However, looking at Pardee's, 37, transcription and the accompanying photograph only a vertical wedge is readable. From KTU 1.85:20 we see that ?bd comes before śśw in section eight of that tablet. Comparing the vertical wedge in KTU 1.97:8' with the final vertical wedge in the d in KTU 1.97:2' or the b in KTU 1.97:5' and 7', I see no reason not to read the broken letter as a d with the lower wedge(s) unreadable. Note for the sake of completeness, Pardee, 36, thinks that the other choice is to read a š and does not consider a d. Dietrich, Loretz and Sanmarin, 112, see two unreadable letters in the next line of KTU 1.97 which they call line 9 but I will call 8a'. Looking closely at the photograph of those two letters it is impossible to discern what they may be. In addition, their placement on the line makes it difficult to associate them with any readable letters in KTU 1.85:21. All that one can say with certainty is that KTU 1.97 has at least two lines of text on the obverse after section 7. Turning to the reverse of KTU 1.97, we have already seen that lines 9'- 14' when collated with the extent section 9 material from KTU 1.85 and KTU 1.72 and 1.71 provide enough information for a nearly complete restoration of the section. In addition, there is enough (more than enough?) room on in the unreadable top area of KTU 1.97 obverse for the third line of section 8. Therefore, it seems most probable that KTU 1.97 as well as KTU 1.85 had complete sections 8 and 9. I am more concerned that KTU 1.97 had something more between sections 7 and 9 than I am that that lacked a section 8. However, if there was more it wasn't much more than a single line.
In addition to not having a section 8, KTU 1.71 and KTU 1.72 have two other important affinities: one linguistic, the other, the extremely close physical proximity of the locations of their discovery. First, in section 4, KTU 1.71:9 and KTU 1.72:12 both likely read ḫra, while KTU 1.85 reads yḫru. The other verbal form in the protasis of this section, yṯtn, cannot be read in either KTU 1.71 or KTU 1.72. It is possible that these texts used participles here rather than finite verbs. But who knows. Second and perhaps more importantly, Schaeffer found these two tablets within a very short distance of each other, I would guess centimeters or less, in or very near Tomb V on the acropolis of Ugarit. Both share the same topographical point, 583, with each other and with the label inscription KTU 6.12. Schaeffer discovered several other tablets nearby. I'll have more to say about this when I look at these tablets as scribal exercises. I can only surmise that all these tablets belonged to or are the product of the same master teacher and/or his apprentice scribes. KTU 1.85 and KTU 1.97 both come from some distance away and some distance from each other. Schaeffer found KTU 1.85 in the house of the master scribe Rashapabu, west and a little south of the North Palace, and KTU 1.97 in the so-called House of the Literary Tablets in the South City Trench block X.
As I indicated, KTU 1.71 and KTU 1.72 have fragmentary portions of Section 7 and Section 9 but there is no room for or any indication of Section 8 on these two tablets. From all this, I suggest an pairing with KTU 1.85 and KTU 1.97 representing a different textual tradition than KTU 1.71 and KTU 1.72. While there are several differences between the extent portions of KTU 1.85 and KTU 1.97 and at least one of those differences appears in KTU 1.72, it seems reasonable to consider KTU 1.85 and KTU 1.97 as part of the same textual family. I'll call this text family "the scholar's text family." I'll take up the differences between them later. Likewise, KTU 1.71 and KTU 1.72 appear to belong to another family. Because of the find location, I'll call this family "the acropolis text family."
Setting aside the larger issues section 9 for the moment, let's look at variations within the scholar's text family starting with the scholar's text family. As we do this, we will also look at a few "inter family" similarities that must be dealt with. KTU 1.97 is so fragmentary that there are only a few areas where differences with KTU 1.85 are apparent. They can be summarized as follows:
- KTU 1.85's preference for w k rather than k at the beginning of several sections (KTU 1.85:15 vs. KTU 1.97:2'; KTU 1.85:18 vs. KTU1.97:6' and possibly KTU 1.85:23 vs. KTU 1.97:9')
- KTU 1.85 uses w in two other places where KTU 1.97 does not (KTU 1.85:19 reads, as is common at this point in each section, a w before y[ṣ]q while KTU 1.97:7' appears to read only yṣq and KTU 1.85:27 reads mġmġ w pr where KTU 1.97:13' reads only mġm]ġ pr, here KTU 1.85 agrees with KTU 1.72)
- KTU 1.85:15 we read aḫd but KTU 1.97:2' agrees with KTU 1.72:21 in reading yiḫd.
- KTU 1.85:17 agrees with KTU 1.72.23 in reading irġn while KTU 1.97:4' reads arġn.
- KTU 1.85:17 reads ydk while KTU 1.97:4' reads td[kn].
I will take these up in order.
1) KTU 1.85 has a stronger preference to begin each section with a w than any other of the four tablets does. This leads me to believe that in the writing of KTU 1.85 or its vorlage, still within the scholar's text family, the a master dictated the text to the (student) scribe and either the master or the student added the ws. It is of course possible, but in my view less likely, that this phenomenon is the result of dittography. Regardless of the source of these extra ws, they do make it unlikely that KTU 1.85 was the source of KTU 1.97.
2) The w before yṣq in KTU 1.85:19 occurs in every section at this point on every tablet except KTU 1.97:7'. I take its absence in KTU 1.97:7' to be scribal omission.
3) I am not sure how aḫd came to be in KTU 1.85:15. One explanation is that it is more "original" than is yiḫd and that yiḫd in KTU 1.97:2' and KTU 1.72:21 are dittography from the beginning of section 5, which does indeed have yiḫd. Since it is not clear exactly what either word means, the issue cannot be resolved on semantic grounds. There is not a whole lot of help from grammar either. When a section clearly has a finite verb in this position, it has the form yCCC that I take to be indicative yqtl form. This is the form one would expect after k meaning "when" or "if." There is an issue of how verbs, at least in this text, were vocalized. Although I do not agree that the verbs in the apodoses need be passive, one should nonetheless consult Cohen and Sivan, 44, of the various verbal forms. In any case, the common verbal form would favor yiḫd over aḫd in KTU 1.85:15. But sections 2 and 3, if these are verbs at all, are either qal perfects or participles. In my view, the syntax favors participles but this is far from certain. aḫd could be either a qal perfect or a qal participle. On balance, I tend to think that aḫd is an error; how it came about, I do not know. It may reflect some intermediate text between the parent of the scholar's text family and KTU 1.85 that KTU 1.97 does not perpetuate but this is well beyond any evidence.
4) The problem of irġn in KTU 1.85:17 in agreement with KTU 1.72.23 against arġn in KTU 1.97:4' is another one of those difficult issues that bridges the two families as I have suggested them. Again, we are faced with a word of unclear meaning regardless of the first vowel. One might note the spelling mad in KTU 1.14 II:35 where one would expect the common word for "much", mid. But the issue in KTU 1.14 II:35 is likely related to some scribal confusion as to how to express an aleph that closes a syllable (sometimes called a vowelless aleph). If that were the case then it would not seem to have any application in our text. One might also note that "my brother" is written aḫy in KTU 1.6 II:12, iḫy in KTU 2.41:17 and uḫy in KTU 1.4:17. Personal names with a/iḫ show the same confusion (aḫmlk, iḫmlk). Explanations vary. My own guess is that arġn is a scribal error for irġn. It may be that there was uncertainty as to the pronunciation of an unfamiliar technical term. I will have more to say about this when I discuss these tablets as scribal exercises in a future post.
5) Based on its frequent occurrences in the other sections 2, 3, 4, 5, 7 and 8 in this same grammatical position, one can reasonably assume that ydk is original in KTU 1.85:17 and td[k(n)? ] in KTU 1.97:4' is some kind of error. Again, I will have more to say about this when I discuss these tablets as scribal exercises. But like the aḫd verses yiḫd issue discussed above this may be evidence, no matter how limited, of an intermittent textual variation between KTU 1.85 and KTU 1.97 but if so it indicates an even more complex textual history within the scholar's text family. One should also consider tdkn in section 10 of KTU 1.72:39. However, this last section is problematic for other reasons. Two of the versions (KTU 1.85:32 and KTU 1.71:25, the sections is broken away from KTU 1.97) lack a verb at this point. I will argue below that the missing verb is an artifact of the common ancestor of both families and that it is related to the complex problems of section 9.
Other, minor, differences between KTU 1.85:32 and KTU 1.97 can easily be explained as scribal errors. I will deal with a couple differences between KTU 1.71 and KTU 1.72 below.
The Problem of Section 9
Using, as I have been doing, KTU 1.85 as the primary reference, I now turn to section 9, KTU 1.85:23-29, with glance that the problems of section 10, KTU 1.85:30-32. In the post in which I offered a translation of section 9, I noted that the composite text was reconstructed, following Pardee, from bits and pieces of three of the four tablets (KTU 1.85:23-29, the most complete, KTU 1.97:9-14, KTU 1.72:27-35 and part of a single word from KTU 1.71:22). What little overlap there is, mostly between KTU 1.85 and KTU 1.72, shows no meaningful differences between the scholar's text family and the acropolis text family. However, because of the difficulties of the reconstruction, it is possible that differences between the families and within the families are lost to the broken parts of the tablets. What is obvious is that this section is itself composite in someway or other. For reference, my translation of the section is as follows:
When a horse cries out (coughs?), one crushes together a measure of juniper (oil?) and the fruit of ʿṮRB seed and bitter almond,
(which is) like gd,
(which is) like ṮMRG (some kind of plant)
and MǴMǴ (some kind of plant) and a measure of NNI and abukkatu-plant resin and a measure of heliotrope (?) and MǴMǴ and lettuce seed
(which is) like donkey (?)IRĠN plant,
and pours it into its nose.
If you want to see the Ugaritic text, please go to my post in this section. Anyone who has read the other sections, even casually, will notice that section 9 is not only much longer but also scrambled in some ways. Cohen and Sivan, 52ff, attempt to reconstruct it as two sections, the first extending between lines 23 and a completely conjectured 24A, the second section extending from a conjectured 24B to 29. While their reconstruction is possible, I think, because of the repeat of MǴMǴ, the first sub-section must have extended at least through line 25. But I see high probability of recovering the original text, which no doubt consisted of two or more sections, from the extant tablets.
It is the case is that a vorlage of both families already had the garbled merged section. In my view, there are two identifiable stages of development prior to either of our text families. The first one involved the confounding of two or more smaller sections and the second involved an attempt to make sense of it by adding tmṯl, (which is) like, before certain ingredients to explain the otherwise incomprehensible text. It is possible that the merged section arose because the protases (the symptoms) in the once individual sections were the identical. I further speculate that the first step may have happened when scribes excerpted what became the vorlage of both text families from a much larger collection.
I believe one sees a related problem in my section 10. There is clearly something wrong with the lack of a verb (likely ydk or tdkn) at the end of the first part of the "Components and preparation" portion of the apodosis in both KTU 1.85:32, and KTU 1.71:28. KTU 1.72:39 appears to have fixed the problem. While it would be hard to prove, I believe that the more immediate vorlage of both text families was missing the verb ydk or tdkn. My best argument is that it is harder to explain how two scribes both dropped it than it is to explain how a single scribe dropped it and then other scribes reproduced the same mistaken text (at least) twice and corrected it once. I even wonder, it this text might have gone through a stage where a scribe copied something that he or she did not understand beyond a certain superficial level. Despite its formulaic style, the vocabulary of this text is highly technical. It is possible that one or more of the scribes that copied it had no better understanding of these words than we do. If so, problems like those seen in sections 9 and 10 are inevitable.
Proto KTU 1.85 as an Excerpt
Several times, I have mentioned that I believe that the one of the vorlagen of both text families was a now lost much larger list of symptoms and cures. Of the existence of such a larger text one can never be certain but two things, one external and the other internal, lead me to this belief. First, as I have indicated elsewhere, structurally similar lists of "instructions," broadly speaking, such as wisdom proverbs, law codes, omens and other professional instruction manuals often run hundreds of sections long, often over several tablets. I think it odd that this professional manual would only run ten or so sections. Second, the likelihood that section 9 resulted from the merger of two or more sections in some original points directly to a larger list of treatments. It was not uncommon for master teacher to excerpt larger texts to make them more manageable both pedagogically and physically for their students. Student practice tablets tended to fit in the hand. That would be a fair description of the size of each of the four extant tablets that I have been discussing. The largest, whose measurements we can know with any certainty, KTU 1.85, is 14.5 cm high by 10.5 cm wide, just right for a student's hand.
As I said at the outset, all this is very speculative. It is disconcerting that even with such direct evidence it is so difficult to be certain about the textual relationship of these four tablets.
References:
Dietrich, Manfried, Oswald Lorenz, and Joaquín Sanmartin, The Cuneiform Alphabetic Texts from Ugarit, Ras Ibn Hani and Other Places (KTU:second, enlarged edition), Abhandlungen zur Alt-Syrien Palästinas (ALASP), 8, Münster: Ugarit-Verlag, 1995
Pardee, Dennis, Les Textes Hippiatriques, Ras Shamra-Ougarit II, Paris: Editons Recherche sur les Civilisations, 1985
Posted by Duane Smith at September 10, 2007 7:38 PM | Read more on Ugarit |
Trackback Pings
TrackBack URL for this entry:
http://www.telecomtally.com/mt/mt-tb.cgi/2248
Comments
Post a comment
Please read Abnormal Interest's Comments Policy.