Monday, March 1, 2010

How pedantic must be critical editions?

This is a review article of:
Cristina Pecchia, Transmission-specific (In)utility, or Dealing with Contamination: Samples from the Textual Tradition of the Carakasaṃhitā in: Text Genealogy, Textual Criticism and Editorial Technique, Jürgen Hanneder–Philipp A. Maas (eds.), WZKS 52-53 (2009-2010)

Which manuscripts have to be left out of one's critical edition? The question is unavoidable, since the editor's and the readers' life is too short for a critical edition to contain all possible witnesses. Scholars usually agree that many printed editions can be just left out, since they are just re-prints of older ones, or since they reproduce the text of an older edition plus some innovations (be they conjectures or involuntary errors). Why should this not apply to manuscripts?
Pecchia applies Timpanaro's eliminatio codicum inutilium (elimination of useless manuscripts) along with the elimination of derivative manuscripts. In fact, the latter are copies of a single manuscript, which is also extant. They can be, hence, obviously left out –unless the extant manuscript they copy is now in pretty bad conditions and they are needed to restore *its* text. But derivative manuscritps are hardly the case. Much more commonly, copies are derived out of more than one exemplar, through sporadical or systematic contamination. Some of these derivative and contaminated manuscritps can also be eliminated, suggests Pecchia, although prudently. Care has to be used because contaminated manuscritps may be witnesses of lost hyparchetypes. Moreover, derivation and contamination can be ascertained only in regard to a certain section of the text –the conclusions established in regard to a segment of it do not necessarily apply to the whole text. Furthermore, contamination may be sporadical or systematic, can regard readings (if the copist has constantly before her two or more copies and *collates* them, choosing what s/he thinks to be the best reading) or portions of the text (if an exemplar is damaged or thought to be less authoritative as regards a specific segment of the text). Finally, contamination can only be decided a posteriori, after the definition of a plausible stemma. And this needs to be done after a significant part of the text (better: the whole text) has been collated, if one wants to avoid distorsions. On the other hand, Pecchia insists on the necessity of this elimination, unless (it is implied, although not explicitly said) one wants one's critical edition to be nothing but a collection of curious variant readings.

Last, Pecchia's article is a well-written and well-documented introduction to many aspects of textual criticism, quoting basically from the Italian tradition (Pasquali, Timpanaro, Segre…) and from Paul Maas' Textkritik, but also answering to criticisms addressed to them by more recent scholars.


krishna said...

Ciao Elisa!
Thanks for this interesting post. Few words just to say that recently I have red a paper by Christian Lindtner on a similar subject: "Editors and Readers", in Lama Doboom Tulku (ed. by), "Buddhist Traditions. Problems and Perspectives", Manohar, New Delhi 2001, pp. 193-204.
:-) k

elisa freschi said...

I guess it is "Buddhist Translations…" (after some struggles with Karlsruhe's Catalogue). Do you own it? It seems not to be available in any of the libraries I am acquainted with.

krishna said...

Ciao Elisa!

You guess right, it deals with Buddhist texts and translations... and yes, I do own it (if you want I can provide for you a pdf of the original paper... as soon as possible). What I find interesting in this Lindtner's work is that he - by means of few but clear concrete examples - points out a method, applicable both to Tibetan and to Sanskrit texts, for establishing the most probable reding, also by taking into account the tradition to which the text belongs (by comparisons with other texts, by knowing which works the author presumably has red, from what he quotes, etc.)
This method is really useful on account of Tibetan translations of now lost Sanskrit originals.

;-) k

elisa freschi said...

sounds interesting. I look forward to read it (whenever you have time enough to send it to me…) and to understand how 'general' his method is. In other words, does Lindtner intend to establish the correct reading in every case or only in regard to authors one is closely acquainted with?

krishna said...

Well, of course Lindtner - even if he does not explicitly claim it - suggests that one has to be quite acquanted with the author/s s/he treats. But naturally the simple study and/or treatment of an author does not mean "to be acquanted with" that author - at least according to Lindtner. Lindtner - among others - makes the following example on Hahn's edition of Nāgārjuna's Ratnāvalī V, 26a:
«The Sanskrit text [...] reads anuvijñaptisaṃyukto. This reading is confirmed by the tibetan version: rjes rnams rig dang ldan rtog. But it does not make very much sense in the context given. From the context it is clear that Nāgārjuna is referring to an old list of “minor faults” (kṣudravastuka). Similar lists are found in several Abhidharma works in Pāli and Chinese, and the corresponding term in Pāli is invariably anavaññatti, meaning “freedom from disregard.” The Pāli reading is not anuviññatti. So it is quite certain that Nāgārjuna did not write anuvijñapti-, but anavajñapti-. The correct reading is, incidentally, confirmed by the Chinese rendering (shun jué jué), and by Nāgārjuna'a own subsequent definition. And, naturally, it is the only one that makes sense...»
I use here this example just to make you know the kind of "method" - but probably "method" is not the right word, better would be "procedure" - to which I have spoken of. Lindtner seems to uphold, and of course I agree with him, that it is not only important to know the author "in se," but equally important is to know the cultural horizons to which the author (pro or contra) makes reference to. In this particular case the comparison with Abhidharmic texts has helped to establishing a correct reading, accepting an apparently late Chinese translation of the text, instead of the Tibetan and/or the Sanskrit ones (then Lindtner proceeds to explain in brief when and why, according to him, the mistake would have occurred.) But we have to be aware that, as far as Buddhist texts are concerned (particularly texts of late Buddhism), for a good critical edition one has to know Pāli, Sanskrit, Tibetan, Chinese, (Japanese), etc., in a comparatistic way... which is a task that only few people can accomplish (I think you could, step by step!)
All this just to point out the necessity of making reference to several sources, other that the lines of manuscripts (or xylographs in the case of Tibetan Tripitaka) of the text which one intends to edit, for establishing the right readings. These at least are Lindtner's advices.

(PS: the PDF, next week, now not at home!... ah, would it be possible to have a copy of Pecchia's essay? Recently I have had the pleasure of studing parts of the Suśrutasaṃhitā, which is the "twin text" of the Carakasaṃhitā.)

elisa freschi said...

Thanks for this very interesting comment.
Lindtner's procedure sounds convincing and compelling. I have just a couple of doubts (I'm here expressing a pūrvapakṣa and not my own siddhānta, which would be much closer to your one):

1. Could not the same result be achieved (at least in the case at hand) just through a philosophical analysis of the text? After all, in order to understand that something does not make sense, one does not need a Chinese translations. I guess you would argue that the Chinese translation and the Pāli model make one more confident about one's proposed improvement, but life is short and one cannot always counter-check all possible occurrences.
2. Knowing the cultural milieu of an author may make one choose in favour of the milieu against the author. S/he might, instead, have used an uncommon word or expression or have consciously broken with her/his own tradition.

By the way, I do not know ANY Chinese nor Japanese…I'm quite far from Lindtner's (and yours) standards!

krishna said...

Ciao Elisa,
let me say that I agree with you on both pūrvapakṣa points. Nonetheless I would here specify the following two integrations (always as pūrvapakṣa):
1. Of course philosophy could suffice for a clear understanding of a text; nonetheless, Lindtner appears to analyze the problem from a philological, not strictly philosophical, perspective: thus, when preparing a critical edition, the question is whether or not to take into account also a – for instance – Chinese translation which could IN SOME CASES be decisive for the attestation of a correct reading which explicitly confirms the editor’s suppositions. Naturally, all this depends on the doubts which assail the editor at work on a particular unclear passage: the doubts generally push the editor in deeper textual comparisons. When the text runs easily, there is no need for checking further sources (Chinese, Japanese, etc.)… as you say: life is too short! And I would add: for working in a “redundant” way.
2. Second, the cultural background of an author should be considered in a comprehensive way, i.e., in a way according to which the point of view of the author is inserted in the (its) wider cultural horizon. Therefore, the modern “analyzer” of an ancient text should, as far as possible, develop a clear understanding of both the author’s philosophy and vocabulary. But this obviously does not safe from possible misunderstandings and misreadings.
Moreover, we have also to consider that an “editio princeps” must be seen as a starting point for other philological considerations on this or that particular aspect, word or expression, which can contribute to refine the text only where and when problems arise, as for instance the limited example of Lindtner on Hahn’s edition of the Ratnāvalī (and also Lindtner’s editions, as you know, appear to need revisions :-) )
Last: my standards are not like Lindtner’s (to me too Chinese and Japanese are unknown!!)

Licenza Creative Commons
Quest' opera è distribuita con licenza Creative Commons Attribuzione - Non commerciale - Non opere derivate 2.5 Italia.