In this comment to a recent post, Bill Haines expressed frustration concerning how best to cite passages from texts like Mozi or Li Ji, and wondered if the readers of Warp, Weft, and Way might be able to uncover and then publicize some best practices. We have had some brief discussion of citing Ctext.org here, but a more general discussion would be great. Please share your thoughts!
Warp, Weft, and Way
Chinese and Comparative Philosophy 中國哲學與比較哲學
There are lots of dimensions to the question Bill has prompted me to raise, and different answers will be needed for different eras, different kinds of text, and different types of book and audience. Justin Tiwald and I thought a lot about how to cite Neo-Confucian texts in our Neo-Confucianism: A Philosophical Introduction. We decided that we would distinguish between the texts themselves, and modern editions of these texts. For example, see this opening page of our book’s Appendix 3 (“Abbreviations of Primary Sources”). Our general goal was to provide a concise reference for our notes, which could then be unpacked into a precise location in a modern edition. For examples of how the concise references look, you can look here (this page is from the companion website to our book, listing the original Chinese of all translated passages in the book).
One other thing to notice is how we handled translations. On the one hand, we take responsibility for all translations in the book. On the other hand, we are sometimes drawing on existing translations, and we also want readers to be able to consult existing translations for more context, or for alternative ways of rendering the passages. So we do our best always to cite an existing translation by writing “cf. XXXX” after the textual citation itself (where XXXX is the citation to the modern translation).
I would just like to share my thoughts as a young scholar on this issue in the context of my own work wherein I only very rarely cite translations, instead opting for my own translations and citing Chinese texts directly. I’m therefore only speaking for myself in this context.
I’m digitally inclined and my first exposure to Chinese philosophy came simultaneously with exposure to ctext.org. Therefore, when I come across a noteworthy citation to a Chinese source, my first instinct is to go to ctext as a jumping-off point for further inquiry. Because (most of) the texts on ctext have searchable concordance numbers, I’m very happy when I find these numbers in cited work. In the few instances where there are no corresponding concordance numbers on ctext, such as in the case of books 47-50 in the MZ, then I go to my print versions and try to find the passage in question (e.g., MZ 48/2 is probably early on in book 48 of the MZ), so while somewhat frustrating when ctext doesn’t have something, it’s by no means impossible to use the concordance numbers to find some quoted text in any commentary you happen to have. (That, or, just go to your university library if it has the concordances.) Because of how frequently this system works for me, I support it and use it in my own writing on the *Zhuangzi*, *Xunzi*, and *Mozi*.
A caveat though: I don’t think that ctext.org is a *substitute* for concordances, nor is it safe to simply copy-and-paste from ctext and cite only the concordance numbers. So for me, it’s important in my own citations to double check with the real thing in case there is a mismatch between ctext and the concordance. Some emendations have been made on ctext itself (e.g., http://ctext.org/mozi/book-4#n672), which are marked, in which case I flag the emendations and justifications to the extent that I follow them. My point is just that while I fully endorse citing concordance numbers *because of* the existence of ctext.org, I do not take ctext itself for gospel. Instead I see it as a quick and convenient way to get from a printed citation to the Chinese source text, from which I can consult my own commentaries and ponder a scholar’s reading of the text in question. Imperfections and lacunae on the website are inevitable, but I do think that having a standardized way of finding Chinese sources digitally is a fundamentally good thing for the field—wherever the numbers originally came from—and that following the convention could benefit everybody. Indeed, providing a convenient way to get from a standardized citation to the e-text is exactly what Donald claims to want to do on ctext’s concordance page: http://ctext.org/tools/concordance.
As for alternatives, I see two, one of which is already the case for certain texts. A number of texts already have their own numbering systems that seem well-utilized by scholars, so it’s pretty easy to circumvent concordances in favor of these reference methods. For example:
*Lunyu* – Chapter.verse “1.7”
*Daodejing* – Chapter “65”
*Mengzi* – BookVersion/chapter “1A/6”
*Xunzi* – Knoblock numbers: “23.7”, Hutton numbers “23.338”
*Lv Shi Chun Qiu* Knoblock numbers: Book/chapter.verse – “17/5.2”
These citation methods each have their own idiosyncrasies with respect to their digital counterparts on ctext.org. In general though you can’t search these numbers into ctext and get a result. In the case of LY and DDJ, I don’t think this is a huge deal since these texts are on the shorter side and it’s easy to use the standard numbers to find verses manually. The *Mengzi* is weird on ctext, since the “B” versions of each book continue the numbering from the A version rather than begin from 1. So, for example, on ctext, the first verse of book 1B has an “8” next to it, continuing from the numbering of the A version, so it basically makes all standard citations to B version books useless on ctext since the numbers don’t match.
For Xunzi, I personally cite concordance numbers since I’m doing my own translations. For citing translations, I’m not really sure if Prof. Hutton has a preference for citing his numbering system or the page number from Hutton 2014. Personally, I have been using page numbers when I reference Hutton 2014, but I’m willing to go with whatever the consensus convention (or just Prof. Hutton’s preference) may be on that issue.
For LSCQ, there is no concordance data on ctext.org, so Knoblock numbers seem to be the best alternative for now. I don’t know of a better way to cite this text, but whatever it is, I’d love for a digital equivalent of concordance numbers to be on ctext.
The second alternative I see to citing concordances is just citing page numbers from various commentaries. I don’t think this is a good idea for a number of reasons. In the first place, not everybody has the same commentaries, nor has everybody the same scholarly preference for some commentaries over others. Furthermore, in many commentaries, the commentator’s own words are quite sparse at times, the comments instead being made up of quotations from previous commentaries. This introduces layers of confusion into using commentaries to cite the original text. You could end up citing a commentary that makes emendations you disagree with, whose comments you disagree with, or whose comments are not the authors’ own, but are instead pulled from previous commentaries. I think part of the virtue of citing standardized, commentary-less concordance numbers is that it avoids this controversy by remaining neutral and standardized—even if you find the concordance edition to be mistaken (as I’m sure we all sometimes do). When textual issues *do* arise, I think it’s appropriate to cite commentaries to help illuminate the reader on why you make a particular translation or emendation choice, on the assumption of course that you’re following a commentator’s suggestion in making the emendation.
Finally, I fully agree with Prof. Angle that it’s good to refer to alternative translations, especially to supplement one’s own.
Just my two cents.
My own mindless blundering might not be wholly unique. A page here with some basics might be helpful to beginners.
Example – in my remark I said I had cited the HY concordances, but I was mistaken. (Googling “HY concordances” didn’t show me what those are.) In fact I cited something I was told was the proper thing to cite, and the citations took the form of strings of numbers that I have never known how to use when not in that room at that library in Hong Kong. Now Frank’s comment has made me aware of the Tools page at ctext.org – golly! – , and I see that what I had cited was the ICS concordances. The bookmarked start page I’ve been using for years at ctext doesn’t mention that Tools page, but now I see the home page does.
It seems that here the word “concordance” is being used in a sense unfamiliar to me and missing from my Compact OED. What does it mean? Tool toward a uniform convention for citations?
*
To cite Plato or Aristotle we give page and approximate line numbers from editions nobody looks at, because all editions and most translations include those numbers in the margins, enough for citers and seekers to approximate with. I’m guessing that was the intention of the “concordances”? One way to accomplish this in translations would be to have, in a footer on each page, the precise reference numbers for the beginnings of each of the translation’s paragraphs on that page. (Of course that would make it harder for someone like me to use search engines to find discussions of e.g. “1.6”.) It looks like Donald Sturgeon has, amazingly, done much of the necessary work. If there are errors they can be corrected as we go.
Prompted by an earlier comment by Paul, I’ve been looking harder at on-line resources, and I had been wondering whether one might use the numbers shown at TLS.
http://warpweftandway.com/thesaurus-linguae-sericae/
But I wonder if that site is still being curated?
For the Mencius B-books at ctext one can get used to the necessary arithmetic; but indeed beginners will be lost. Anyway some of those chapters are long, and one wants to cite more precisely than simply “1A7.”
An easy direct method for most texts would be simply to use the numbering system used at ctext, with suffixes like -a and -b for subparagraphs. That needn’t constitute approval of the text.
TLS is indeed still being curated (I am newly on their board!), but the concern you raise is valid on a more general level: I’d be very wary of citing any online text, including Ctext, because no one knows whether it will be accessible in fifty years.
The other reason why it makes no sense to cite Ctext is that absolutely no textual criticism goes into it. That’s not the point of Ctext. Ctext merely copies standard texts (that is, when it’s not corrupting them with input errors), so “citing Ctext” really just means “citing the base texts that Ctext uses.” The same is true of almost all concordances–which is why I questioned the practice of citing the HY concordances in the first place. Some concordances include an attempt at producing a standard reference text, but most do not, and in any case almost all of them are problematic.
The text that one should cite is a text by an editor who has made some effort to make sure that he or she has produced a reliable text. (That’s intentionally vague.) Sometimes a responsible scholar will have to cite more than one edition, and sometimes not–it depends on the kind of work one is doing. Variants are crucial, so the text you cite really ought to include collation notes (集校, 校勘, etc.).
I would also make a plea for citing texts that are widely distributed and a student might actually own. Yet another reason why citing the HY concordances makes no sense to me! I’d go so far as to say that American philosophers are the only people on the planet who still use the HY concordances. It’s kinda like stating GDP in 1945 dollars.
An advantage of 1A7 over 1A/7, and of 1.6 over 1:6 or 1/6, is that search engines can handle the former better. If you search for “1A/7” you just get pages that have 1A in the vicinity (or maybe not in the vicinity) of 7.
Perhaps I’m missing something, but does ctext have any of the interlineal commentaries? I fear that the commentaries are being viewed as a nicety rather than an integral part of these Chinese manuscripts.
Hi John: Thanks for your comment. Going forward, please use your first and last name (see top right for our policy).
As for the substance of your point, you’re absolutely right that we don’t want to sideline the importance of commentaries! But do you think that means whenever we cite a text, we need to cite from an edition that contains commentaries? That of course would be the case whenever one wants to cite commentaries themselves, but in other cases, it’s not as clear to me.
The “texts” were produced and then transmitted by people we call “commentators,” so we’re fooling ourselves if we think we can read a “text” without “commentary.” It’s not as though there was “text” first and then “commentary” accreting around it afterwards. That’s an entrenched misconception.
Hi Paul — yes, by all means, but I think you’re conflating “reading” the text and “citing” the text, right? The present discussion is not about whether one can read, interpret, or even identify “texts” apart from “commentaries,” but rather what practices we should employ to make reference to these texts in our own publications.
Steve, why do we “cite” a text? In order to “read” it. It’s as simple as this, I think: no edition of a classical Chinese text that lacks commentary has any scholarly value. So why would that be a useful choice for citation?
Or think of it this way: there’s no such thing as “the text.” All we have are “editions of texts.” So just be honest about which edition (or ideally editions, plural) you’re using. Citing something like the HY concordance is not being honest because all you’re really doing is citing whatever edition the HY concordance used as its baseline–and that is virtually never the best one.
For John: Yes, ctext has a number of classic commentaries, though they are integrated into the source text differently. Sun Yirang’s 墨子閒詁 is on the site, for example: http://ctext.org/mozi-jiangu and it may be accessed from the main Mozi source text page by clicking on the grey 注 icon to the left of the passage in question.
For Zhuangzi, Guo Xiang’s commentary is on the Wiki (http://ctext.org/wiki.pl?if=en&res=744703), which can be accessed from the Zhuangzi text main page in the “related texts” box at the bottom of the page: http://ctext.org/zhuangzi
The same is true of other texts. Simply go to the text’s main page on ctext and look in the “related texts” box on the lower right-hand side of the page and there are a number of commentaries to consult, many of them having been digitally transcribed.
As an aside, I find Bill’s comparison with the convention for citing Western classics enlightening: obsolete numbers from editions nobody has that just tell you which line of the text is being referenced or translated. Bekker numbers are going on 200 years old and i’m sure very few people have the original editions to use when citing Aristotle, say, but most translations include Bekker numbers anyway to make for easy access. I view concordance numbers as basically our equivalent of Bekker numbers. The fact that they’re search-able on ctext is I think crazily powerful.
Again, this is no substitute for academic due diligence and responsibly dealing with textual issues in one’s own work, when they arise, in certain scholarly contexts. The point is just accessibility. So long as it’s possible to search concordance numbers on a reputable e-text repository, I don’t consider the fact that very few people have the physical sources of these numbers to be a problem.
Thanks Frank, ctext has just become more useful for my particular project.
I just want to thank Frank for calling our attention to the fact that ctext.org can (for many texts) both display HY/ICS references, and allows one to search by these references.
Also, in response to Bill (wondering what we are meaning by “concordance,” and more generally what ICS and HY are), this and the following couple pages from Wilkinson’s Chinese History: A Manual is useful. A full concordance lists every character in a text and all the places in the text where that character appears. So these volumes have both a full edition of the text, and then the concordance itself. In the digital age, concordances are less important, but the standard reference text is still valuable.
Ah! Thanks!
Until a better alternative comes along, I advocate the use of book and line numbers from the Harvard-Yenching concordances to cite Mozi, Zhuangzi, and Xunzi. These numbers function like Stephanus numbering for Plato or Bekker numbering for Aristotle. They provide a compact, reliable way of identifying the exact lines referred to. (I prefer HY to ICS because the ICS numbering system produces long, cumbersome parenthetical citations.)
The point is not that the HY editions are superior critical editions. It’s that they allow precise citation. For example, Wu Yujiang’s Mozi Jiaozhu (Zhonghua Shuju, 2006) is one of the best editions, but it presents the text of each pian as one long essay, followed by endnotes. A citation giving page numbers in Wu does not inform readers exactly what lines are referred to.
Since Donald Sturgeon has already included HY numbering in many texts at Ctext.org, it’s become convenient to look things up using these numbers. Ctext.org isn’t yet complete in this regard but eventually it will be.
Future translations could adopt the practice of including parenthetical HY numbers for each paragraph or section.
To clarify, the aim of the citation is merely to identify what string of graphs in the text we are talking about. Having done that, we can go on to discuss what the best reading of the graphs might be and how we might best interpret them. Citing HY numbering does not mean we’re relying on HY as an authoritative edition. That’s not the role of a concordance.
To the contrary, I’d suggest that no edition is authoritative, and anyone working closely with pre-Han texts in Chinese would benefit from consulting and evaluating multiple editions, with accompanying commentary. In the case of Mozi, for example, these would include Wu Yujiang, Sun Yirang, and Wang Huanbiao, among others.
I’m not entirely clear. If you’re working on Wang Fuzhi’s Zhuangzi, wouldn’t it be the most convenient to cite Zhuangzi -jie and -tong? Or if you’re working on Chen Guying’s Zhuangzi, wouldn’t you likely cite Zhuangzi jinzhu jinyi? (And so on and so forth.)
Should one cite the HY in addition to whichever volume the project demands? If so, why? It doesn’t seem to help with identifying passages. (There are numerous ways to do that on a comp, such as “Ctrl-F” on any soft copy of any edition of any manuscript.) Instead, it just seems to be adding additional steps to research.(Spotting the passages in HY so you can cite HY….)
I’m likely overlooking something.
I’m curious if the conversation can be extended to Zhuangzi citation practices. I tend to cite the editions I’m using, which are currently Guo Qingfan (Zhuangzi jishi) as well as Qian Mu (Zhuangzi juanjian). (The latter is now my travelling edition.)Thus if I were to write something now, these would be the editions I cite.
Which editions should I be citing/using? Is there a demonstrably better edition that I am foolishly overlooking? Again, I’m asking with regard to the Zhuangzi. (And I mean in general. Obviously the project at hand will in many respects determine the most appropriate edition(s).)
Wang Shumin’s is very thorough and has copious references to Sima Biao’s non-extant commentary. It’s my go-to at the moment. Though hard to get, Li Mian’s is also often illuminating. (Your library might have it—mine does.)
Re: citation. My practice is to cite commentaries when the actual comments play a role in my research. If my reading of a passage relies upon an emendation or gloss made by Guo Qingfan or Qian Mu, then I’ll mention that (or, if my reading departs from a more traditional or equally plausible alternative reading, that might be worth mentioning too). Further, if I’m citing a commentator’s reading of a particular line, then I’d cite the commentator via page number.
This is just to say that I treat commentaries more like secondary sources rather than primary sources, so I wouldn’t cite the “primary” text via a commentary. Rather, I’d cite the HY concordance numbers.
I find these Zhuangzi editions useful.
– Guo: 郭慶藩，《莊子集釋》
– Wang Xianqian: 王先謙，《莊子集解》
– Wang Shumin: 王叔岷，《莊子校詮》
– Chen: 陳鼓應，《莊子今註今譯》(latest edition Taipei 2011?)
– Li: 李勉，《莊子總論及分篇評註》(little known but worth consulting)
– Qian: 錢穆，《莊子纂箋》(Zhuangzi Zuan3jian1))
Of these, only Chen’s provides a modern Chinese paraphrase. For readers who find a modern paraphrase helpful, there’s another edition posted on a few websites. It is 王世舜,《莊子譯註》. There’s also 黃錦鋐，《新譯莊子讀本》(2007), which is worth owning because it includes phonetic script (Bopomofo, 注音符號) for all the characters.
I’m happy to see I haven’t led myself too far astray lately. (I use most of these editions. Though I admittedly only have the Wang Xianqian edition because I noticed it was in virtually every bookstore in Taipei.)
My copy of Chen Guying’s edition is the single volume released by shangwu yinshuguan, which came out in 2011. Last time I was in Taiwan (about a month and a half ago) this was the newest edition I saw.
A main desideratum for a mode of citation is that it be, in general, usable by the citer’s audience. That should be taken into account. If a significant part of the citer’s audience is people who are not professional scholars of Chinese matters nor in heavy training to become such, citing commentaries won’t do—not until their line numbers have become part of the standard apparatus in published translations.
Citing Stephanus or Bekker numbers for Plato or Aristotle doesn’t involve taking a position on textual questions. The best texts for textual questions—the Oxford editions, I think—are organized to be citable by Bekker and Stephanus numbers. Such a number does not, strictly speaking, identify what string of graphs one has in mind. It just identifies a location in the text, accurate to a couple of horizontal inches in translations of Aristotle, a few more inches for Plato. I presume textual scholars of the Bible use chapter and verse in the same way; it makes sense.
In the study of Plato or Aristotle or other Western philosophers, the overwhelming majority of live interpretive issues do not involve textual issues; the readers of interpretive expositions and arguments tend to be philosophers, often with only a moderate knowledge of the original language.
Even philosophers who read Greek, when reading material in English citing Stephanus or Bekker numbers, are likely to use the numbers to look for the passages in translations, because that’s quicker. Or for the Greek they may look to a Loeb rather than an Oxford edition, because the Loebs have English on the facing page, which is helpful at least for skimming around to find something; they also report main textual variants. (Or if the citer flags a textual issue, which is rare, I think usually the scholar would say enough that the reader still doesn’t have to look to the Greek.)
To this from Paul:
Ignorant question from me:
For the main texts of early Chinese philosophy, can we identify one (or more than one) “commentary” edition such that we may reasonably suppose that that commenting party (or parties) significantly adjusted the text they had received, so that that text and that commentary are in significant the work of the same party?
Or, Paul, are you just saying the texts were adjusted by that kind of person?
*
(I think philosophical thinking is an important part of interpretive (and therefore I guess of textual) scholarship on philosophical texts in general, though I suspect interpretive scholarship of Chinese materials has been badly hobbled by the tendency to require that published papers say something philosophically interesting.)
Bill, I’m not sure that I understand the question, but I’m having a hard time thinking of a received text that was NOT what you call “adjusted.”
Here, I’ll treat you guys to a section from an unpublished guide I have for Ph.D. students at Penn:
The first layer of commentary for most classical texts is the chu 注/註, or “[principal] commentary.” Authors of chu were usually very early transmitters of the text. Many more classical chu existed in antiquity than are extant today. This is because the chu of a small number of transmitters were typically singled out by posterity as authoritative. Editions of the text would thenceforth be published only with the canonical chu; other chu would survive in fragments, if at all. But ancient chu were usually not neutral. Authors tended to write chu in order to express their personal understanding of the text, which was often idiosyncratic and creative. The chu could come to represent an entire tradition, with its own world-view, its own glosses, and, not infrequently, its own version of the text itself. We speak, for example, of the Wang Pi version of the Lao-tzu, the Mao version of the Shih-ching, etc. There can be crucial differences among various textual traditions. Therefore, comprehensive variorum editions—called “collected collations” (chi-chiao 集校) and the like—are indispensable.
Moreover, some knowledge of the mechanisms of manuscript transmission will help one understand why texts were associated with particular chu. Warring States and early imperial manuscripts were largely produced for audiences that were either familiar with typical bureaucratic formulae (in the case of administrative documents) or already knew by heart the material being recorded (in the case of canonical or philosophical texts); manuscripts were, consequently, written with economical and underdetermined graphs. Insiders, who perhaps learned the texts under the guidance of an authoritative teacher, knew when their community read the graph tui 兌 as shuo/shui 說 and not, say, t’o 脫 (to cite a common example of graphic underdetermination), but to outsiders, such codicological conventions naturally left the text open to a multiplicity of interpretations. Redactors who transformed manuscripts into the received texts that we have today did their best to eliminate this type of ambiguity by adding (or regularizing) semantic classifiers (such as the 言 in 說), thereby making the text intelligible to a larger number of readers, but also inevitably narrowing the range of possible interpretations. This process of specifying the meaning of the underdetermined manuscript was the basis of the composition of chu; as texts circulated ever more widely among readers who, unlike the ancients, had no specialized knowledge or authoritative teacher to guide them orally, explanatory chu came to be regarded as indispensable.
Apologies for the Wade-Giles Romanization. I do it as a very weak attempt to forestall plagiarism.
Hi Paul, thank you very much!
I apologize for my unclarity.
Regarding this argument about reading (not about citing):
I was thinking that a necessary condition of that argument’s making sense for a given text would be this:
That we have, for that text, at least one commentary such that we have reason to believe that the party who composed that commentary had, themselves, quite substantially altered the text from what they had received.
For then that party would be in significant part an author of that text, that version of the text. The text would not be a creature independent of that particular commentary. We might think:
(1) “I shouldn’t imagine I can take the Shijing as evidence of what anyone thought or did before this commentator made his edition, or even that the poems we have are worthy as literary works in their own right; I should just read the text+commentary as a work in its own right”
– or at least –
(2) “In reading the Shijing as evidence of what people thought or did before this commentator made his edition, I must (a) try to suss out what changes this commentator is likely to have made, or at least (b) make arguments that are not hostage to a few fine points of the text”
So I was asking if that (italicized) necessary condition is indeed true of “the texts”– or at least of many of them. I think you’re saying Yes, for the main early texts.
Though I’m not sure how much adjustment (from what that very commentator received) you’re saying it’s reasonable to believe there was for the typical text in the best editions we have, and that matters to the force of the argument, as does our reason for interest in the text.
It seems to me that insofar as the necessary condition is met (a question of degree), thoughts (1) and (2a) would argue against the usual sort of variorum edition, in favor of presenting (even translating) distinct commentaries (text+commentary) in parallel columns or facing pages. That’s what one would have to read.
The point is that the manuscript with the zhu is so thoroughly shaded by interpretive decisions, and not just in the comments and glosses. I repeat, and not just in the comments and glosses. Thus, the wish to get to a pre-Qin “text” while shedding the zhu is a fantasy. And given the manner of composition, it hardly makes sense to speak of “texts” at all.
For instance, it makes sense to talk about Zhuangzi qua Guo Xiang’s Zhuangzi, but it makes no sense to talk about Zhuangzi simpliciter (that is, the “text” as it was prior to the zhu).
Scholars that think in terms of the “text” as essential and the comments as a nicety are failing to understand the complexities of manuscript culture.
One is free to isolate the comments and glosses and then focus on the parables, fables, aphorisms, &c. within a given received manuscript. But given the dearth of material evidence and the nature of manuscript culture, one cannot be claiming to do much more than supplying their own narrative about how the various, fables, parables, &c. cohere.
Although, sometimes the results of this storytelling are as interesting as the glosses and comments of one or another edition. Personally, I’ve become more interested in what historically self-identified Daoists (e.g. Xuanxue, Chongxuan, &c.) have to say, rather than my contemporary academics. That’s not a slight, it’s just that my goal is to engage *the whole* genealogy of Zhuangzi exegesis.
Although, Zhuangzi might not be the best example, as there is evidence that the first seven chapters cohered across different editions in the Wei-Jin period. But even still, that’s way after the fact.
John has explained my position better than I could! Looking forward to meeting you when I come to Singapore in August.
Thanks John!
To this:
My quantitative question was only about the material other than the comments and glosses.
So shaded … that what?
The long answer sounds to me like what I called “(2)” above. Is that it?
Maybe so, but what I was trying to ask was, if we have one or two best zhu editions for some text, how much sense can it maybe still make (for one or another purpose) to focus on the “texts” we find there, noting variant readings if there’s more than one zhu, but largely ignoring the glosses and other remarks. I don’t mean: not being critical about how the text might divide into mutually independent sections, or have lost big or little bits etc. That’s a completely separate issue. We are critical that way when texts seem to merit it. And if we have multiple zhu for one text, with lots of variants between them, then it makes sense for scholars to pore over the zhu to see whether some variants are not worth other readers’ attention.
Hypothetical case: Suppose I read the main old zhu edition of some text and I find the glosses and remarks highly uninteresting. (That’s how I’ve personally found most of the little commentary I’ve looked at, all on the Analects, and why I haven’t looked at great quantities. Of course I could be completely wrong, and you report that at least for some other texts the commentaries are quite interesting. Good to hear. So call mine a hypothetical case, addressing an issue of principle.) And suppose I find the “text” within the edition intriguing. Must I throw out the text with the commentary? Sure, I grant that it’s not what it was, but it’s not entirely not what there used to be.
And maybe there was no “the Zhuangzi” before the zhu. But it looks like there were those chapters or paragraphs, or something very similar. If they were originally philosophically interesting, that means there was some kind of deep order to the ideas, and that’s the kind of thing that can be seen and debated and used in textual criticism. (Alongside philology etc.) Anyway philosophers often think that’s the case. I guess the zhu writers thought so too.
And if one doesn’t find the deep order in a text or parts of it, then too bad for that “text,” by all means.
My sense is that regarding the “texts” that is the remarks attributed to Confucius in roughly the first ¾ of the Analects, the debate is very young yet. There hasn’t been much focus on precisely that text, and much of the discussion has taken as a core premise what I think is a demonstrable falsehood (the focus on the family by Confucius-as-there-represented). My sense is that the discussion isn’t anywhere near mature enough to justify the kind of despair that one could imagine excuses people’s “supplying their own narrative” and aiming to make it look to the general reader like interpretation of Confucius. Of course you haven’t defended that, John.
Lately at WW&W it seems that when people talk about the hopelessness of original texts, usually the example is the Zhuangzi. I don’t know how representative it is.
I gather that for some important texts, or big chunks of them, the earliest forms we have are not commentary editions?
Re: What I meant by supplying their own narrative.
To stick with Zhuangzi, the manuscript is so underdetermined that it’s normal for a key notion or phrase to be read in entirely contrary ways. Chengxin 成心, for instance, has been read as a positive attribute (see Hanshan Deqing)and as a negative attribute (see Cheng Xuanying or Qian Mu).
Many key notions in the Zhuangzi have been subject to such contrary readings over the past two millennia. Yet many of these contrary readings are compelling and philosophically robust in their own ways, the contradictions notwithstanding. Given the underdetermined nature of the manuscripts, there’s no determining which reading is “true to the text” (which in and of itself seems to be a senseless expression in this context).
I’m skeptical that modern scholars have ahistorical and non-question-begging tools to change this trend anytime soon. (That is, the trend of commentators supplying interpretive glosses that contradict one another.) And even if they did discover such tools and principles of “good interpretation,” there is unfortunately no pre-Qin manuscript for such tools and principles to be applied to (at least with regard to Zhuangzi).
I’m not sure why this longstanding state of affairs would be cause for despair.
Re: The choice of Zhuangzi.
I think the case is even more extreme with the Laozi, as it is an aphoristic collection. (See, for instance, Alan Chan’s ‘Two Visions of the Way,’ which compares Heshang Gong and Wang Bi, and then compare this with Xiang’er.)
But that evidence is overdetermined, yes? I mean, I gather that for the Zhuangzi in particular it wouldn’t be entirely surprising to find, when we dig up the original audiotapes, that they leave us with similar questions regarding some key terms. The text kind of aims at that sometimes?
I don’t know the force of that without knowing whether the different readings center on different sections of text. Maybe it’s true of the Zhuangzi in the strongest way: the same chapter gives distinct evidence of two contrary profound and complex viewpoints, (and not as e.g. characters in a contest of ideas aiming to challenge both and put nothing of the kind in their place, such as we find in e.g. Plato’s Protagoras, Kant’s Antinomies, or Kierkegaard’s Either/Or). I think that would show the Zhuangzi to be an unusual text, not that philosophical judgment & creativity can’t help us read behind the materials we have from early China. (Granted, it’s possible for a good philosopher to elide two different positions, thinking of one or the other with shifts in context or in the exigencies of argument. The difference could be such that much later readers are highly sensitive to it. But it sounds like that’s not the kind of thing you’re talking about.)
By despair I meant the abandonment of a particular hope. What I had in mind specifically was the hope of finding in the Analects (the “texts” we have—mainly the standard and the Dingzhou I think) much more than a bag of sayings from here and there, nothing that could help us in any historical inquiry about matters prior to the commentaries, if we don’t give much or any attention to the glosses. Because I agree with you in not expecting quick conclusions (of which that despair would be an example); I think we’ve only just begun to use the tools we have.
Weren’t you saying the state of our evidence about the Zhuangzi material does justify the analogous despair regarding the Zhuangzi?
Bill wrote: “when we dig up the original audiotapes …”
This is going to have to be my last comment on this thread, because (a) it’s clear that everybody has long since made up his or her mind about how to cite primary sources, which was the original topic, and (b) Bill, you just don’t seem to be trying to understand.
THERE ARE NO ORIGINAL AUDIOTAPES. THERE NEVER WERE ANY. It’s not as though some sage named Zhuangzi wrote or dictated a “text” that was then gradually corrupted by various copyists over the centuries. That is essentially a Lachmannian model, which was devised for medieval European literature and has scant applicability to classical Chinese texts. (There are doubts as to whether it’s valid even for medieval European texts, but that’s a different matter for a different board.)
What we have instead of an Urtext is a redaction by Guo Xiang (or the roughly contemporaneous person whose material Guo Xiang pilfered, if, like some historians, you think Guo Xiang was nothing but a plagiarist) compiled out of pre-existing snippets of material that he thought belonged together in a book to be called Zhuangzi.
There is good indirect evidence that there was a book (or books) approximating Zhuangzi before Guo Xiang (because other texts, starting conspicuously with Huainanzi, start to quote and allude to it in stable ways), but otherwise we’re almost completely in the dark as to the contents.
And that book antedating Guo Xiang was itself compiled out of pre-existing snippets (according to one theory, this was at the court of the Prince of Huainan, but nobody knows). Some redactor, presumably living in the Western Han this time, brought these snippets together, juxtaposing them as he pleased, probably also emending them as he pleased, and certainly glossing them as he pleased (since, remember, the underlying manuscripts would have been severely underdetermined), because he thought he was doing the world a favor by giving us a redacted and coherent “Zhuangzi.”
Here’s the main point: this is a perfectly typical case. The only peculiar complexity lies in the fact that Guo Xiang redacted a previous redaction, but otherwise this is how MOST texts were compiled. Yanzi chunqiu. Xunzi. Mengzi. Yuejue shu. Laozi of course (because the Guodian manuscripts have blasted the silly fantasy that the received text was written by a wise old man named Master Lao). Really, you have to think long and hard to come up with an example that DOESN’T fit this model.
I can certainly understand that philosophers are not interested in these critical details. I can even sympathize with their desire to be able to cite classical Chinese texts in the same effortless way that the real power-brokers in their departments cite Plato and Aristotle. But what I can’t ever accept without objection is the idea that the “text” can be divorced from its history of transmission, that “commentary” is just the various niveaux of crap that accumulated around the text, because that’s a fallacy and it impacts your interpretations. (If it didn’t impact your interpretations, I’d concede that my criticisms are pedantic. But it does.)
You guys simply have to abandon the assumption (and all the corollaries that come with it) that your texts were written by the great philosophers whose names they bear. Here, I’ll offer a prize: who is the first Chinese philosopher who can plausibly be said to have written the book we now possess? OK, I’ll answer: it’s Lu Jia. I win the prize. (Honorable mention if you said Xunzi or Han Feizi, because you’d be nearly right, but the problem is that someone else redacted the received versions. Thanks for playing.)
Enjoy the beautiful weekend, guys.
According to Karla Mallette in European Modernity and the Arab Mediterranean: Toward a New Philology and a Counter-Orientalism, University of Pennsylvania Press, 2011, the Lachmannian method is still highly regarded by textual critics. (p. 9)
Of course. We’ve been over that. Tapes of what exactly, I left open. I did consider adding details about their diverse sources etc. etc. to forestall that kind of worry about what I meant, but then I thought that would be overkill. And starkly irrelevant to the little argument I was making. It was a hypothetical case to make a point of principle. (Tapes?!)
Actually now I think I may simply have missed John’s point and addressed an argument he wasn’t making. I don’t know.
I do cling to the idea that the Analects might be some kind of evidence about what Confucius probably thought or didn’t think, because I have an argument to make (rather, to finish revising) about that. I think my project can get by on pretty weak assumptions about authenticity. But there are fallback projects if needed. Paul, you could think of my project as broadly supportive of your pet peeve here.
I’ve been struggling toward understanding and engagement here (same thing); but also hesitating to agree to some things, especially before understanding them or their basis. Until Paul’s latest comment I just wasn’t clear on what was being said in general, and in particular what was being implied about the text I’m interested in. I explained why.
Nobody has an obligation to answer my questions, (e.g. “I gather that for some important texts, or big chunks of them, the earliest forms we have are not commentary editions?”), but I might be a useful proxy for the recalcitrant philosophers one might like to convince or engage with, even the ones who are unlike me in that they work in academic departments or believe the Chinese philosophers wrote those books.
Not to me. But sure this has all been a digression.
I’ve learned a great deal from this thread about the available helpful ways to cite things, and if I can be of assistance toward making a general reference page on that I’d be happy to.
“I gather that for some important texts, or big chunks of them, the earliest forms we have are not commentary editions?”
Hi Bill. I can think of three examples of texts whose oldest versions “bypass” commentaries.
Excavated texts such as the Guodian and Mawangdui manuscripts. There might be commentary material that we just don’t recognized, or there might be accretions juxtaposed in with them that are quite late, but these texts for the most part got no 注.
The Later Mohist *Canons*. Nobody really understood how to put these together and understand the systematic alterations made by scribes over the years until the 20th century. Chris Fraser has an excellent introductory essay in Graham’s *Later Mohist Logic, Ethics, and Science.* It’s basically an accident that we have them and in some ways they too are “excavated texts” even though we’ve had them all along.
The Lv Shi Chun Qiu was commissioned, written, and assembled during a fairly precise period, although some material probably comes from older sources and isn’t entirely original. In fact, parts of book 28 of the Zhuangzi appear in the LSCQ with minor differences. The oldest commentary was by Gao You from the Eastern Han dynasty, while the LSCQ was a late Warring States text.
Knowing the history of texts and being careful about assuming too much systematicity in the original texts is important. However, I still think we can learn about Warring States thought by studying the “masters” texts.
Thanks Frank! I guess the Analects too would be a text for which our earliest version of a big chunk of it is not a commentary edition. That all makes me wonder about the whole hypothesis about how these texts were gradually written through successive commentaries, based I guess on a significant number of examples of known rewriting, and general ideas about the culture.
Hi Bill,
I think in general accretional theories of textual composition tend to assume that the accretion process itself began before commentaries came along. So when people like the Brooks’s say that there are certain parts of the Analects or Mozi that seem older than others, presumably they see this as predating commentaries. As for an extant, commmentary-less edition of the Analects, I don’t know of one.
As far as texts being primarily the result of later commentary writing and rewriting, I don’t really take things that far.
Re: “As far as texts being primarily the result of later commentary writing and rewriting, I don’t really take things that far.”
I think there’s an ambiguity here. Are you talking about (a) the manuscripts we currently have (e.g., Zhuangzi zhu or Zhuangzi shu) or (b) manuscripts that are earlier yet no longer extant (e.g. a pre-Qin edition of Zhuangzi)?
Because if you’re talking about the latter in terms of the former, then you’re making a number of assumptions: that ‘b’ existed, that ‘b’ existed in a form that can be inferred from the contents of ‘a,’ that we’re in a better position to interpret ‘b’ in light of ‘a’ than earlier commentators (who sometimes even had access to earlier redactions), that the stories in ‘b’ represented a coherent philosophy (or more egregiously, that this philosophy represents a “school”), and so on.
One could claim that critical scholars are wrong, and that these sorts of assumptions are reliable. But even then I’d argue that the project appears strange. The more detailed a philosophy one teases out of ‘a’ to interpret ‘b,’ the more assumptions required and the less plausible it becomes. Why not just focus on ‘a’?
Butting in, in response to John’s last paragraph:
(Critical scholars are indeed wrong about many things, according to critical scholars. This young field is shot through with disagreement, and it’s almost impossible for any one person to have all the necessary intellectual tools: historical, linguistic, philosophical, etc.)
Every step in empirical inquiry is imperfect. Each one makes assumptions and thus introduces a risk of error. We might infer that the more advanced the project, the greater the risk. Progress is always and increasingly tenuous. Why go any distance to ge wu if it’s risky? On the other hand, sometimes things fall together and the result of progress can be something more certain and stable than what one had before.
Specifically: I think a historian who is generally concerned with time t would have a distinct reason for interest in something just because it might reflect time t rather than t+1.
But I think philosophers have reason to care about older versions of a text too, to the degree that there can be grounds for hope—based partly on a philosopher’s interpretive tools, to be sure—that the text or some identifiable part of it can be attributed to an individual or small group distinct from the commentator responsible for the later text. (For example, the materials attributed to Youzi in several old collections—similar to each other in philosophy and rhetoric, and different in those respects from e.g. the rest of the Analects.) To this point I posted some relevant argumentation a few months ago – it’s Section 6 of a long comment Manyul helped me put up. (The larger context was the question whether interpretation of old Chinese philosophical texts, or any texts at all, should aim toward finding the views of an author.)
Hi John:
All I’m saying is that I think there’s pre-Qin material in the Zhuangzi. I don’t think there needs to have been a pre-Qin edition of the Zhuangzi for that to be the case. Because I think there’s pre-Qin material in the Zhuangzi, I don’t think that the whole text was written by (“was primarily the result of”) later commentaries, e.g., Guo Xiang.
I think the thread of the conversation has been lost.
I still don’t see how that warrants treating the parables, fables, aphorisms, &c. in a manuscript while neglecting the comments and and glosses. And, in turn, that we are warranted in citing a concordance as a common coordinate among scholars. And, in turn, why HY would be chosen as that concordance.
The concordance does not better represent an earlier redaction of a manuscript. Sure some of the stories might have been circulating in some form or other, but that does not warrant associating them with Zhuangzi prior to the Han. The concordance instead represents a distortion of a received redaction.
Hi John,
Lost – yes but not hopelessly, I hope. Here’s a go—which partly involves stating and answering things I think you weren’t saying, just to try to keep things clear.
I think Frank’s “As far as texts being primarily the result of later commentary writing and rewriting, I don’t really take things that far” meant only this and nothing more:
(1) he thinks the texts we have took shape in significant part before commentaries—which is not to say that they weren’t actively or tendentiously gathered, sifted, shuffled and rewritten before any commentator got to them; nor does the line we quoted from him take a position on how much the earliest ancestral materials may have been scrambled before any commentator saw the results and applied his spatula. And
(2) he thinks the amount of scrambling done by commentators once they got to work was not so much that we should regard commentators individually or collectively as the primary authors of the “texts” as we have them.
Do you agree with (1)? (2)?
I think it was a slightly separate point of his that the rewriting and reshuffling (by or before commentaries) is sometimes or often not so thorough that the resulting material has to be called entirely post-Qin and cannot serve as evidence of what was done and thought in pre-Qin times.
But I imagine we can agree that the final judgment on these matters has to wait upon the success or failure of many particular interpretive efforts.
Right; that crazy idea wasn’t on the table.
Frank had answered: “All I’m saying is that I think there’s pre-Qin material in the Zhuangzi.”
I think nobody was thinking it would warrant that.
Someone might think this:
(A) Suppose you want to talk about putative pre-Qin material that you think is present in or at least quasi-present in and evidenced by a bit of what we call the Zhuangzi. If you identify locations within that bit by the page and line numbers of a concordance of the Zhuangzi, you are thereby associating that material with the Zhuangzi; whereas if you identify those locations by the page and line numbers of a commentary on the Zhuangzi, you are not.
But it seems to me offhand that citing the commentary runs the greater risk of false associations (if only because the commentary brings in more material to be potentially wrongly associated), though the absolute quantity of the risk is completely negligible in all cases. And even if it were large it would be easily counteracted.
Another thought might be this:
(B) The scholarly community on any of the big old texts includes a bunch of people who cite the text and not a commentary edition, who are mistaken and/or unconcerned about the integrity and history of the text, and therefore neglecting the importance of commentaries and the fact of scrambling. That should stop.
My sense is that in this thread that point has been agreed on all sides from the outset; or at least, nobody has been assuming that it isn’t true. It hasn’t been at issue.
But suppose the main reason for a policy of having people cite only commentaries is to correct that error about how to approach texts, and suppose that in order to get people to see the reason to cite commentaries you have to get them to see that that approach is an error; you have to get them to abandon that error. It would follow that the process of convincing people to cite commentaries would remove the main reason for the policy of citing commentaries, at least for the convinced parties. Meanwhile there are also huge disadvantages to that policy, discussed above.
But I’ve been addressing things you didn’t say, John. You did say this:
John, I’m guessing your thought is the following, but I didn’t even think of it until now:
(C) Granted, those who cite Stephanus or Bekker numbers for Plato or Aristotle don’t imagine that they are thereby accepting all (or any) of the editorial decisions of Stephanus or Bekker, who might even have died before the best old manuscripts we have were discovered for some texts. The best current scholarly editions use these numbers without the slightest concern for the editorial opinions of Mr S and Mr B. Still, there is a general assumption that for Plato and Aristotle, that manuscript culture preserved, with only very minor variations, something Plato wrote, or something contemporaneous or nearly contemporaneous with Aristotle. That’s what makes sense of the concept of “correcting” a received text. Now, the whole concordance project, because it makes small editorial decisions that go against the commentary’s “text,” is based on the assumption that the situation for the texts of early China is the same as for Plato and Aristotle. Whereas in fact the idea of a making a slight correction to the commentary’s text is a fundamentally mistaken idea. The commentary’s text is what it is, and it is not a close approximation to some other thing, describable but lost. Correcting that text, even on the basis of something like the Dingzhou bamboos, would be like correcting a bit of Matthew on the basis of a parallel passage in Mark. The concordances embody this mistake, in two ways: (a) by making slight revisions, and (b) by presenting the text as though it were a whole work whereas in fact it is a kind of excerpt, like the concluding couplets of all of Shakespeare’s sonnets. To cite them is to endorse that mistake. So we should not cite them; we should cite the commentaries.
So here’s my response. I’ll forgo nibbling around the margins of the premise today; rather I’ll simply accept the whole premise (except for the illustrative comparison within (b)), for the sake of argument about the conclusion.
First, we’re only talking about what kind of numbering system to use to identify locations in the text. No system is even a candidate for current use unless it can be practical for the bulk of the intended readership of the citing paper to use the numbers to find the location in the text (in one language or another). Therefore, I would think, for many (most?) texts and intended audiences, no concordance and no commentary fits the bill.
Second, citing a concordance can’t make sense unless one’s intended readers have access to some version of the text that uses those numbers. But for the great majority of texts and intended audiences, the bulk of the audience will not have access to the concordances themselves, in any medium. Therefore, when we’re talking about citing concordances, we’re basically talking about using those numbers to find locations in other editions of the text. We’re talking about texts for which the concordance numbers have a life outside of the concordances. If the numbers have no life beyond the concordances themselves, citing them isn’t much use (except in the case of a text for which the concordance edition is in fact the most readily available edition; I don’t know if there are such texts).
In that sense using a concordance number is like using a surname that an ancestral freed slave took from the master family. It embodies a false idea in a way, but it’s what we have; and we don’t have to believe the false idea to use the name.
I’m under the impression that most commentary editions do not come with ready sets of numbers by which one can identify a location in the “text” part to within 10 or 20 characters. (I’m under the impression that in many cases the finest-grained set of numbers is the page numbers assigned by the modern publisher, which seems an ephemeral system.) Such editions are missing one of the key desiderata for a system of citation, but that’s correctable. One could publish an edition of the commentary with such numbers. And then the question arises: why not use the numbers from some concordance, the one that has been most widely used for citations previously? Offhand it seems to me that that’s what one should do (with suffix numbers of course to refer to the comments, if they’re bulky enough that it wouldn’t suffice to say “comment after 11.3.8” or “11.3.8c” or whatever. And then the result would be a situation not unlike what we now have for Plato and Aristotle: the definitive editions of the text are published with a numbering system born from an obsolete edition, a system that has a life quite independent of any particular edition or kind of edition.
For a text where there has been no significant practice of citing by any concordance, there would be no particular reason to use the concordance numbers as the numbers in an adequately numbered edition of the commentary. And maybe there aren’t many old texts that have been often cited by concordance numbers. I don’t even know.
It would make sense to cite a location in a text by using the numbers in an adequately numbered edition of a commentary, if either (a) the edition is widely available, or (b) its numbers have been printed into at least one other edition that is widely available, as Bekker and Stephanus numbers are.
Meanwhile the question remains: how do I tell my readers what line of the Liji I’m talking about? What’s the best practice, for this summer? Ctext hasn’t indexed the Liji to any concordance.
(I see scholars citing the Liji by numbers and I don’t know what the numbers mean. Often the scholar doesn’t say. With Bekker numbers you know what they are even if no edition of the Nicomachean Ethics appears in any bibliography or footnote. If you never saw such numbers before, it doesn’t matter; just pick up any translation and there they are. But for the Liji, no.)
I’m starting to suspect that there is a massive gulf between historical research and philosophical research. ‘1’ and ‘2’ are *very* controversial. (Not controversial in that Frank and others believes them, but that the critical scholarship supports them.)
Fortunately, historians have made an effort to bridge the gap** between philosophers and historians. I (in my delusions of grandeur) like to think I’m trying to bridge the gap from the other side (however imperfectly).
**An obvious recent case of an historian writing explicitly for philosophers is Michael Nylan’s essay, “Academic Silos, or, What I Wish Philosophers Know about History,” from Sor-hoon Tan’s anthology on methodology (Bloomsburg, 2016).
Thanks for the bridge work, John!! – I hope the historians aren’t building a bridge to nowhere – I’ve now read what I could of Nylan’s piece (via googlebooks, my main access). That should get mentioned on the guidance page I’m working on. Other things?
I gather from your first two sentences that you think I’ve suggested that (1) and (2) are modest or uncontroversial claims (I didn’t mean to; my “only” was just for the sake of precision), and you think this view reflects what interpreters in philosophy depts think (I don’t know what they think, except maybe a few people on the Analects).
I have not yet seen anyone deny (1) on this thread; I thought Paul denied it above but Frank made me look again and now I don’t see that he did.
Oh, but oops, (1) is ambiguous. Or my words miss what I meant. Here’s a rewrite:
(1a) he thinks the texts we have are descendants of materials that became bulky and may have been much reworked before commentaries—which is not to say that the line we quoted from him take a position on how much the earliest ancestral materials may have been scrambled before any commentator saw the results and applied his spatula. And
‘1’ and ‘2’ aren’t only controversial, they are also problematic. And this is relevant to how one approaches and cites these ancient manuscripts.
* * *
To be clear, I don’t mean to suggest that all philosophers (or people working in philosophy departments) make problematic historical assumptions. I have a traditionalist view insofar as I view my senior scholars as more knowledgeable and wise than myself, so I don’t think I should name names in a public setting. After all, these senior scholars have been at this for a long time, and I might change my mind and I might be wrong.
(Still, I will say this: I have a theory that the reason that the University of Michigan produced so many top scholars is that they covered both historical and philosophical fronts. I like to think that with scholars like Loy Hui Chieh,Scott Cook, Sor-hoon Tan, Bryan Van Norden, Lo Yuet Keung, and Lisa Raphals that NUS likewise covers many fronts. But, as with everything else in life, there’s the delightful possibility that I’m completely wrong.)
For the Analects I was thinking of the Dingzhou and Pyongyang bamboos:
https://www.academia.edu/1760535/Confucius_Sayings_Entombed_On_Two_Han_Dynasty_Bamboo_Analects_Manuscripts
Oh wow that’s cool! I must have missed them! thanks for the reference!
For a quick initial view of how the Dingzhou bamboos might be significantly different, use Amazon’s “Look Inside!” to search for “Dingzhou” in Ames&Rosemont’s Analects.
So the question is how, in one’s paper, to identify for the reader the location in a text of the string or passage one is talking about, so that readers can have a look. The commentary proposal, as I understand it, is that one should cite page and line numbers from some edition of one or another commentary.
Of the arguments raised for using commentaries (and of the objections to the contrary proposals), each is explicitly answered above (before the digression) in ways that strike me as ample. (And there has been no reply addressing the answer.)
With one exception. I mean one argument for commentaries has not really been addressed. It is an argument not explicitly presented above but maybe sort of suggested.
That is: for these texts (or most or some of them), the text as an object in its own right is a fiction, so it would be positively misleading to use a mode of citation that gives the impression that the text is an object in its own right.
That’s a legitimate concern. (The “digression” above is somewhat relevant to it.)
But suppose one shares the concern and agrees that it is very important. Still it seems to me there are other ways to address it in one’s paper than by a mode of citation that tends to impede the understanding or evaluation of one’s argument. So one should prefer those other ways.
To that, can think of two replies, defending using the commentaries.
(a) One might think that citing the commentaries amounts to administering a kind of kick toward anyone who doesn’t have them, and the accumulation of kicks would push the scholarly community in the right direction, e.g. toward what (b) looks for.
(b) One might intend one’s defense of commentaries not as an argument about what we should do now, but rather as an argument about what system should be implemented in the long run. We should have bilingual editions of commentaries, or we should have editions in Chinese and editions in translation that build in a common numbering system.
To (b), I have no objection. That would be dandy.
To (a), which concerns current practice, it seems to me that a more powerful force for attention to commentaries would be vigorous, intelligible and productive activity in the various relevant fields—-textual scholarship, philosophical interpretation, history, etc. For this reason I think (b)’s proposed means might well work against its proposed end.
At the end I mean (a)’s proposed means.