I personally prefer the Chichung Huang (Oxford Univ. Press) translation, but largely for the easy-to-use footnote apparatus that appears under each saying, rather than at the back of the book. That makes it so much easier, especially for students, for understanding a) what some of the traditional commentary is on the saying and b) what some of the opaque references to people, rituals, odes, etc. might be. I wish every translation of the Analects did that. Any opinions?
I’m partial to the Ames/Rosemont and Slingerland translations for different reasons. Slingerland’s translation scans better when reading and the Ames/Rosemont translation differs so seriously from most that I have read that it serves to remind the reader of the “philosophical” aspects of the terminology in the text in a manner which is obscured by translations which use more common coinages such as “compassion,” “benevolence,” or even “human-heartedness.”
I was not aware of the Chichuang Huang translation. Being a lover of footnotes that appear in the general vicinity of the “foot” of the referencing page, I will certainly get my hands on a copy.
Lacking the qualifications to choose between them, I’m still a bit partial to Lau (I suppose that shows my age), although I like Slingerland’s translation and, with Thomas, always consult Ames/Rosemont. I will look again at Chichung Huang’s version, having first been put off by its being characterized as a “literal” translation, as I did not think such a thing was possible (or am not quite sure what that means).
Huang, I think, is the best for undergraduates–I found it to be the most successful of the ones I’ve taught from (the others being Slingerland and Ames and Rosemont). I’ve got a sentimental connection to Lau, but I like some of the choices Huang makes better. Also, I think one advantage of the Slingerland version is that he includes some of the traditional commentary. The commentaries have not, I think, recieved enough attention from philosophers. Still, all in all I’d have to say I find Ames and Rosemont to be the most philosophically useful, and Huang to be the best overall translation.
I once used Lau exclusively, but moved to Ames/Rosemont. I think they do a better job translating the central concepts throughout, and although there are passages that I think Lau captures better, A/R allow the student to grasp “philosophically” what’s going on more effectively.
I also think the Slingerhand translation is good. He does an impressive job making the passages feel a bit “lighter” if that makes sense (A/R is a bit “heavy” at times). One thing I do not like about Slingerhand: the discussions under each aphorism. Great for me, not good for the student. When students first read the Analects, they are rightly disoriented by the style. The last thing I want, and the first thing they are looking for, is for someone to tell them what each passage means. To interact with the Analects correctly (for me) requires doing autonomous textual reconstruction.
Hi Manyul,
I hope you and others will forgive me for going on at length in answer to this question–it just so happens that I’ve spent the last six months reading back and forth among five different translations (plus the original Chinese, which I am not qualified to pronounce on but from which I can at least glean syntax, repetition of key words and phrases, etc.) I am filled with opinions and questions on this very topic, and could probably write a book, but I’ll limit myself here to what’s on top of my head. It is a funny illustration of the sheer number of Analects translations that of two sets of them, one consisting of the five which I’ve read and the other consisting of the four mentioned above, there is only one overlap. I have read 1. Dawson, 2. Lau, 3. Legge (yes, I seem to have worked backwards up to that point) 4. Leys (is there some thing about having a monosyllabic surname beginning with the letter “L” that conduces to one’s becoming a translator of the Analects?) and finally, and very recently, Brooks and Brooks (the “original” Analects.)
Here is a brief account of my impressions. Dawson, as the first one I read and the one I’ve read most, is not a translation I can describe impartially; for me I suppose Confucius will always be Dawson’s Confucius. But I think it is safe to say that Dawson’s is the most elegant, the most polished, the most pleasant Analects I’ve read. His besetting sin is to frame the sayings of Confucius in a whole bevy of conjunctions and transitions which have no counterpart in the original. A typical example: “The wise delight in water, but the humane delight in mountains. For although the wise are active, the humane are at rest. And although the the wise will find joy, the humane will have long life.” Neither “but” nor “For although” nor “and although” have any counterpart in the Chinese text! Is Dawson perverting the meaning with these words?
Lau, after Dawson, seems heavy and even more needlessly wordy; he will never use ten words if it is possible to use twenty. For example: the famous and wonderful “學而不思則罔、思而不學則殆’ becomes ‘If one learns from others but does not think, one will be bewildered. If, on the other hand, one thinks but does not learn from others, one will be in peril.’ What justification does he have for adding “from others” which, more than Dawson’s constant padding, actually changes the meaning of the word? And why does he see the need for an “on the other hand” which is implicit in the basic structure of the sentence? (Dawson simply has “If one studies but does not think one is caught in a trap. If one thinks but does not study one is in peril.”) On style and elegance Dawson is a million miles ahead of him, and there are many times when Lau is opaque. It is possible that if at the age of 19 I had picked up Lau instead of Dawson the Analects would never have become the important book which it has been to me. One of the first really exciting sentences to me was, in Dawson, “If by keeping the old warm one can provide understanding of the new, one is fit to be a teacher.” How magnificent! Lau has: “A man is worthy of being a teacher who gets to know what is new by keeping fresh in his mind what he is already familiar with.” I don’t think I would even have paid attention.
A sort of litmus test for me is 1:7, which begins with 賢賢易色. Four simple characters! Legge says: “If a man withdraws his mind from the love of beauty, and applies it as sincerely to the love of the virtuous…” There is a lot of needless verbiage here, but there is at least no wholesale distortion. Lau says “‘I would grant that a man has received instruction who appreciates men of excellence where other men appreciate beautiful women….” This seems to me perverse. Where do the men and women come in? I know that was Zhu Xi’s interpretation–Legge’s footnotes tell us as much–but do we need to write the interpretation into the text? 色 appears all over the Analects meaning “appearance”. Why not something simple like “To value worth more than appearance”? Dawson says “If he appreciates men of quality, if he makes light of sexual attraction…” I have not read Chichuang Huang but I stood in a bookstore once and skimmed through it and, if I remember correctly, it said something like “If he prefers the company of virtuous men to that of beautiful women….” Am I alone in finding that all of this is perhaps reading too much into these four characters? And am I wrong to find the “man/woman” dichotomy perverse when there is no clear indication of it in the original?
I think, however, that it has been very stimulating to read with several translations at once; they force one to go to the original, and to imagine the reasons for the (numerous) divergences in their interpretation.
Leys is often more of a paraphrase than a translation. He is much too quick to substitute “the modern equivalent” for something in the original. Sometimes his interpretation is stimulating and of interest. None of these translations has much in the way of footnotes. Dawson has enough for comprehension; Lau doesn’t even have that, although he does have a history of the text and some other material; Leys has a lot of his own opinions and quotations from all sorts of people which are often only superficially related to the text but which often are fascinating. It’s perhaps a good book for a third or fourth translation, just to see what a thoughtful if cantankerous guy makes of the Analects; I’d never recommend it as a first.
Brooks and Brooks’ quality as a translation is overshadowed by their radical theory about the composition of the text, which I am not convinced by, although I would love to hear others’ opinions of it. Again, it was stimulating to see someone else’s often quite different interpretation of various sentences, but I didn’t like one thing. They do not translate “ren” but merely keep it as “ren”, which is fine, but then “li” they translate in all sorts of different ways, sometimes as “propriety” and sometimes as “ritual” which obscures its real centrality in the text. They have an overriding interesting in the political scene of the Warring States period, which makes a nice balance with translations which see the Analects as “written for eternity”, and it is great to learn so much historical background. But they go too far here. One runs the risk of “debunking”–of suggesting that whereas the Analects “seems” to be saying something profound, it in fact is “merely” making a reference to some local scandal.
An ideal translation, for me, should have the following elements:
1. The original Chinese. This should be the sine qua non for any serious translation.
2. Good notes, with various classical commentaries (Zhu Xi to begin with) included.
3. Discussion, whenever you interpret a line differently from other translators, as to why you interpret it that way. If Legge sees “appearance” and Dawson sees “sexuality” and you decide to say “beautiful women”, tell us what led you to that decision.
So, those are as many of the impressions of the translations I’ve read as I can think of for the moment. I would really love to hear the opinions of others as to the merits and defects of these translations and of those that I haven’t read (Slingerland, Huang, Ames and Rosemont….. )
Hi Christopher,
I can’t answer very extensively just now, but let me point you to a lengthy discussion on this blog about se 色 and the reasons one might have for either interpreting it as having to do with feminine beauty, sexual pleasure, or more simply with appearances–which of course can be important or shallow, depending on the context. Enjoy that discussion — it seems to be right up your alley.
Manyul–I spent some time this morning looking over the discussion you link to and the discussions in the other blogs to which you link there; I haven’t had time to read all of them thoroughly but will do so soon. I enjoyed reading the discussion immensely, learned a lot from it, but most of all I was amazed. A few months ago I began a rather long essay about the Analects which included a few pages on this very question; my arguments there actually coincide with yours (although without your erudition) in several points. Even more surprising is that the “researches” (if they can be called that) which I describe in my essay took place in January or February, about the same time when your discussion was taking place. Anyhow, for what it’s worth, here’s what I wrote. It’s too longand a bit repetitive, only a rough draft, and, here, out of context, but its connection to your discussion is so close that I cannot resist sharing it:
The second translation-related question to which I should like to turn my attention has to do with a passage which puzzled me and seemed somehow out of place from the first time I read the Analects. In Dawson’s translation 1:7 mentions “making light of sexuality”. This was both mysterious–what does “make light of” mean in this context?–and somehow incongruous: sexuality isn’t mentioned very much elsewhere; it is not one of the themes of the book. (A Chinese friend who opened my copy of Dawson’s translation immediately noticed that passage–it is, after all, on the first page–and rejected it as obviously wrong: “This is a serious book”, he declared; “it does not talk about sex”.) When I came to the Lau translation I found the following: ‘I would grant that a man has received instruction who appreciates men of excellence where other men appreciate beautiful women’. This struck me as strangely different from Dawson (who renders the whole phrase as “If he appreciates men of quality, if he makes light of sexual attraction…”) For one thing, Dawson does not make clear the connection between what for Lau are the two halves of one thought. Secondly, while they render the first half very similarly, they diverge rather bizarrely for the second, for if for many people there is clearly some connection between “sexuality” and “beautiful women” the two concepts are not entirely co-terminous. I turned to the Chinese text and was amazed when I figured out that only four characters accounted for what both Dawson and Lau made into such full phrases: 賢賢易色. On the first half of the statement I can remark that to any doctrine according to which each word in the original language should be translated faithfully with one and only one corresponding word in the second language (a doctrine which I largely hold myself) this sentence, like the first sentence of the Tao Te Ching, presents a fairly formidable challenge. 賢賢 could be translated as “to value that which is of value” or “to value the valuable” or “to know the worth of the worthy” but there is something unnatural about all such constructions and one can see here where neither Dawson nor Lau made the attempt, merely translating the first 賢 as “appreciate”.
As for 易色, the following preliminary thoughts occured to me. 易 occurs in the common modern Mandarin word 容易, “easy”, and also in the title, 易經, of the Book of Changes. (One of my students said: “The yi in rong yi is the same as the yi in yi jing, but the yi jing isn’t easy!”) As for 色, it means “color” in Mandarin. This is not immediately helpful, but it did give me an idea. Many years ago I noticed that the Yukio Mishima novel whose title is translated into English as “Forbidden Colors” becomes, in French, “Les Amours Interdits”. How did English “colors” become French “loves”? I questioned a Japanese friend who said that while the basic meaning of the Japanese word is indeed “color”, it has the symbolic or metaphorical meaning of the erotic. (There is a conundrum for the translator!) The English translator had chosen the surface, the French the metaphorical meaning; neither could do both. My immediate question now was–was the Japanese word used in the title of Mishima’s novel, by any chance, the same as this character? Thanks to the wonders of the internet and Wikepedia, a search of less than a minute produced the answer. It was; the title in Japanese is 禁色. (Wikipedia also confirms the double meaning of the character: ‘色 in this case means “erotic love”, although it can also mean “color”‘.) Now it would be very wrong to make a leap from the meaning of a Kanji in modern Japanese to its meaning in ancient China, but at least this was a clue and a suggestive one, since it at least lends some confirmation to Dawson’s word “sexuality”.
I do not remember whether I immediately began searching through the Analects for other occurances of 色 or whether I noticed them
independently, while looking for something else, but at any rate I soon noticed several other passages in which the character occurs. 1:3 reads: ‘巧言令色、鮮矣仁’ and is translated by Dawson: “Clever words and a plausible appearance have seldom turned out to be humane”,
and by Lau as “‘It is rare, indeed, for a man with cunning words and an ingratiating face to be benevolent.” Then there is 2:8, which begins: “子夏問孝。子曰‘色難。…” and is translated by Dawson as “Zi Xia asked about filial piety. The Master said: ‘It is the demeanor that is difficult…'” and by Lau as “Tzu-hsia asked about being filial. The Master said, ‘What is difficult to manage is the expression on one’s face….'” Furthermore, 8:4 has Master Zeng say: “君子所貴乎道者三、動容貌、斯遠暴慢矣、正顏色、斯近信矣、出辭氣、斯遠鄙倍矣。” Dawson has: “The things which a gentleman values in the Way are three: in transforming his demeanor he banishes violence and rudeness, in composing his expression he keeps close to sincerity, and in the style of his utterances he banishes courseness and impropriety.” It is interesting that the word demeanor, which we have just seen Dawson use above to translate 色, is here being used to translate another word; 顏色 becomes expression. Lau has: “There are three things which the gentleman values most in the Way: to stay clear of violence by putting on a serious countenance, to come close to being trusted by setting a proper expression on his face, and to avoid being boorish and unreasonable by speaking in proper tones.” It seems that (the beautiful women aside) Lau is consistent in using some form of the word “face” in his translations of 色. Finally, there is 9:18: ‘吾未見好德、如好色者也’ which Dawson translates as “I have never come across anyone who admires virtue as much as he admires sexual attraction.” Lau has ‘I have yet to meet the man who is as fond of virtue as he is of beauty in women.’
As I thought about all of these sentences I saw no obvious reason why the two passages, 1:7 and 9:18, in which Dawson and Lau agree in reading a sexual meaning in 色, should not be read in the same way as all the others. The contrast between appearance and inner worth is a major theme of the whole Analects. In the ongoing and highly nuanced discussion of the relationship between culture and substance to which I have already devoted a great deal of space in this essay, in the repeated warnings against eloquence, glibness of speech, and what Dawson calls “plausible” appearances, and in the frequent distinctions between the material correctness of ritual and the expression of humaneness and reverence which is ritual’s function, it occurs on virtually every page of the book. By contrast sexuality is not a major theme of the book and, without the two passages just quoted, does not in fact occur at all. Is there not something artificially imposed here? Why not translate 1:7 as “If he values worth more than appearances…” and 9:18 as “I have never seen anyone who loved virtue as much as loved appearance”? Or, if the translator feels a need for some stronger word, more fitting of something which tempts men to love it more than virtue, why not simply say “beauty”? After all, erotic attraction is only the extreme end of a whole spectrum of cases in which men are led by appearance to distort their values. There are many people who are not likely to be of erotic interest to us but whose politeness, charm, eloquence, or otherwise pleasing appearance may lead us to a preference which the inner value of their ideas or character would not justify; conversely there are those whose rudeness of demeanor may mask inner worth. How many people, choosing a dog or cat for a pet, have found themselves lured from sober considerations (which animal’s temperament will suit my home?) by the cuteness of a particular candidate? To how many books have I not paid sufficient attention simply because their style was not calculated to please? To how many books have I paid too much attention because the style was riveting? Was this not Johnson’s complaint against the Essay on Man: that the mediocrity of its doctrine was given false importance in the minds of many readers by the dazzling verse in
which it was couched?
The point here is that to translate 色 as “sexual attraction” or “the beauty of women” has a limiting, to translate it as “appearances” or at least simply as “beauty” a universalizing tendency. The latter includes the erotic as one element in a whole category of phenomena; the former excludes all the others. One notices that, as usual in the Analects, the case being made, if we look at an assortment of sentences together, is far from being simple or obvious. If 1:3 puts pleasing appearance in opposition to humaneness, 2:8 and 8:4, by contrast, suggest that one’s appearance–one’s ‘demeanor’ or ‘facial expression’–are in fact important. The goal in 8:4 seems to be that of the Phaedrus: “May the inner and the outer be as one.” As I have remarked earlier in the case of ritual, it often seems as if Confucius has in mind two sorts of enemies: one the one hand, those who would say “it is the inner worth that counts; culture, formal ritual correctness, politeness–these things are of no importance;” and, on the other, those whose inordinate stress on correctness and formality would at least seem to suggest a hollow or superficial lack of concern for substance. By contrast, Confucius seems to say, if the performance of ritual is informed by humaneness and reverence, then outer beauty and solemnity and correctness should arise as a natural product and expression of that inner substance. In this, as I have said, Confucius is an antidote to the prevailing assumption of so many people who, in the last fifty years, have at least seemed to act on the assumption that surface roughness and rudeness is a badge of inner sincerity and rectitude, and who seem to suspect politeness or formality as evidence of falseness. If I am repeating things that I have already said at length earlier, it is because these passages which I am discussing now seem to extend that theme into the field of everyday life and domestic relationships.
In this way I became highly suspicious of the Dawson and Lau translations of 1:7 and 9:18. They introduced a discordant element into the text of the Analects and obscured the connection between these passages and several others in which the same character occurs. My suspicions were confirmed when, discovering the wonderful 溫 故 知 新 website (named interestingly after my favourite Analects passage, as discussed above)(http://afpc.asso.fr/wengu/wg/wengu.php?l=intro) I first began looking at the translations of Legge and Couvreur. For 1:7 the former gives: “If a man withdraws his mind from the love of beauty, and applies it as sincerely to the love of the virtuous…”; the latter, « Celui qui, au lieu d’aimer les plaisirs, aime et recherche les hommes sages…” Each of these is far broader and more catholic in its application than either Dawson or Lau. For 9:18, the translations are ” “I have not seen one who loves virtue as he loves beauty.” and « Je n’ai pas encore rencontré un homme qui aimât la Vertu autant que l’éclat extérieur. » Each of these corresponded very nicely to the sense I had already been developing about these passages. “L’éclat extérieur” seemed particularly felicitous (perhaps in the way that only French words can in the ears of native speakers of English) and again opens the door to a whole galaxy of situations in which men may be led from love of virtue by the temptations of appearances, whereas Dawson and Lau close the door to all but one narrow one.
And so I was prepared to make a sort of cause of this issue, all the more so since neither Dawson nor Lau deigns to give anything like serious footnotes, by which I mean ones which would explain why they made the translation choices they did whenever those choices were not obvious. Then I noticed a footnote to Couvreur: “D’aucuns traduisent « l’éclat extérieur » par « les femmes » (cf A. Lévy) , ou bien « l’instinct charnel» (cf A. Cheng). Le terme chinois admet ces nuances diverses.” This at least begins an explanation, but I still felt justified in the opinion which I had reached by two facts: first, that, whatever he says in his footnote, Couvreur has chosen the broadest, not the most limiting sense; secondly that “l’eclat exterieur” includes “l’instinct charnel” which in turn includes “les femmes”, each term being more limiting than the previous; without a reason for doing otherwise, should a translator not choose the broadest, not the narrowest interpretation?
What the 溫 故 知 新 website does not include are the copious and scholarly footnotes which Legge, alone among the translators I have seen, provides in his work, so it was to these that I now turned, forsaking the internet for the moment and turning to the archaic method of walking to a library. What I found did not entirely support my theory, for he says:
“色 has a different meaning from that in the 3rd chapter. Here it means ‘sensual pleasure.’ Literally rendered, the first sentence would be, ‘esteeming properly the virtuous, and changing the love of woman,’ and great fault is found by some, as in 四書改錯, XIII. i, with Chu Hsi’s interpretation which I have followed; but there is force in what his adherents say, that the passage is not to be understood as if the individual spoken of had ever been given to pleasure, but simply signifies the sincerity of his love for the virtuous.”
Yet I note two things. While notes of this sort bring us many miles above the level of Dawson or Lau–they indicate to the student the world of Confucian scholarship, and point the way to deeper knowledge, rather than behaving as if this translator were the first to attempt the task, as if his interpretation came from revelation or intuition, and as if the task of the reader, in analyzing the translator’s choices, were that of reinventing the wheel–still, he does not give very much in the way of arguments supporting the assertion that “色 has a different meaning from that in the 3rd chapter. Here it means ‘sensual pleasure.'” What arguments are advanced by Chu Hsi and by those who find great fault with him and by his adherents, we are not told. But what speaks loudest to me, at least, is that while, like Couvreur, he gives in a footnote the narrower possible meanings–“sensual pleasure”, “the love of woman”–in his actual translation he gives the most general interpretation, “beauty”. It is as if Legge and Couvreur, while feeling compelled to mention the traditional interpretation in a footnote, in fact even while being themselves convinced by this interpretation, yet do not feel justified in intruding it into their actual text when nothing in the bare Chinese character of the original allows them to do so; later translators, it would seem, were not so circumspect.
The final translation to which I turned, as I have said, is that of Simon Leys, who for 1:7 says “a man who values virtue more than good looks….”, which sounds to me like a happy middle, and for 9:18 “I have never seen anyone who loved virtue as much as sex”, which sounds vulgar. Nothing new was to be found here. But then I noticed a passage of the Analects which I had previously overlooked and which, alas, goes some way towards overturning the theory I had so painstakingly groped towards. It is 16:7: 孔子曰‘君子有三戒、少之時、血氣未定、戒之在色、及其壯也、血氣方剛、戒之在鬥、及其老也、血氣既衰、戒之在得’: The gentleman is to be on guard, in his youth against 色, in middle age against 鬥, and in old age against 得. The whole passage reminds us so strikingly of the three beasts in Dante and of the whole moral tradition which lies behind them that the translator who has Dante in mind should be on guard against importing into Confucius things which may not be there. At any rate the three vices are translated by Legge as “lust”, “quarrelsomeness”, and “covetousness”; by Lau as “the attraction of feminine beauty”, “bellicosity”, and “acquisitiveness”; Couvreur has “les plaisirs du sens”, “les querelles”, and “la passion d’acquérir”; Dawson gives us “matters of sex”, “matters of contention” and “matters of acquisition”; finally Legge (who whatever his faults at least sees that the pithy sentences of Confucius should not be shrouded in the verbiage to which so many of his predecessors seem to have been addicted) has “lust”, “rage”, and “rapacity”, which sounds as if it was taken directly from an early commentator on Dante.
Does this passage alter the situation somewhat? It at least militates against any uniform translation of 色 as appearance: we cannot say that in his youth the gentleman must be on guard against appearance; but, then again, 9:18 had already done that. The Dantescan parallel (and notice we are talking, not even about something explicitly said by Dante, but only about a common but not undisputed allegorical interpretation of the three beasts in Canto I) lends some weight to someone who believes in universal ideas but cannot by itself decide the matter; in fact, since the parallel with Dante is suggested not primarily by Confucius’ text but by the suggestive translations of Legge and Leys, those translations cannot help us if we are to go beyond them to the text itself. Again, Couvreur helps us here. “Les plaisirs du sens” certainly includes sex but it also includes the delirium induced by wine and music and perfume and flowers and soft garments and all sorts of other things by which an incipient 君子 might be distracted in his pursuit of virtue and learning.
That, then, is where I leave it; past this point the amateur can do no more and it is time for the serious scholar of language to take over. I leave with the same tendency to prefer the broader over the narrower interpretations both because that is the path of caution, and because it seems to fit more neatly into the whole mental world of the Analects.
Finally, I feel permitted to ask a question. Even if we admit the narrower sense of “sexuality” rather than “beauty” or “sensual pleasures”, it does not follow that “sex” and “woman” are synonomous terms. Legge and Couvreur, thought their actual translations
are the friendliest to my view, seem in their footnotes to assume that ‘sensual pleasure’ and ‘the love of woman’, ‘l’instinct charnel’ and ‘les femmes’ were synonomous or interchangeable. Lau takes the far more drastic step of importing this tendency wholesale into his
translations, for where Dawson and Leys may see sex, he insists on the very specific terms ‘beautiful women’ (1:7), ‘beauty in women'(9:18), and ‘the attraction of feminine beauty'(16:7). Is there any justification for this? One suspects that some of these writers were only dimly aware of the existence of people for whom sexual attraction does not take an exclusively feminine form, and I would not be unwilling to attribute this view to Confucius as well, but I do feel compelled to ask whether there is any justification in etymology, in the nature of the character 色 itself or in anything else in the relevent passages, for this second leap, not only from appearances in general to sexuality, but from sexuality in general to “woman”. There may well exist some justification but no translation which I have seen has furnished any. It would be ironic if a process of inquiry that began with the particular Mishima novel to which I alluded above were to end here!
Hi Christopher; that is a remarkable coincidence. I don’t read French, so your remarks about the French translation add something interesting to the discussion. Thanks for posting the comment. For your question at the end: I think ‘love/lust for a woman’ and ‘sensual pleasure’ are connected by translators because se 色 is related, of course, to appearance and the tacit implication seems to be that it is through visual contact with the beauty of women that the love/lust is evoked, assuming a male perspective to be reflected exclusively in the discourse, of course — not an implausible assumption. The homoerotic perspective is difficult, if not impossible, to find until much later historically, I think — I haven’t done any extensive research about this. In any case, I would suspect that se 色 as ‘love/lust/sensual pleasure’ picks up the heterosexual male perspective through dominant usage by people who write from that perspective. (If anyone out there knows more about this and cares to comment, we’d be appreciative.)