I’m working on a review of Bryan Van Norden’s book, Virtue Ethics and Consequentialism in Early Chinese Philosophy. There is a very interesting discussion in his Introduction about what he calls “the lexical fallacy,” a fallacy that he notes was first pointed out to him by Kwong-loi Shun, in conversation.
Van Norden quotes Henry Rosemont (from “Against Relativism” (1988 ) in Larson and Deutsch’s Interpreting Across Boundaries), who actually formulates and uses the lexical principle that Bryan asserts to be fallacious. Let’s call it ‘L’:
(L) “…the only way it can be maintained that a particular concept was held by an author is to find a term expressing that concept in his text. Thus we cannot say so-and-so had a ‘theory of X,’ or that he ‘espoused X principles,’ if there is no X in the lexicon of the language in which the author wrote.” (Rosemont p. 41, footnote 11).
Maybe there is something fallacious about the principle. But Van Norden’s argument for its fallaciousness seems to rely solely on apparent counterexamples. Here they are (all from Van Norden p. 22)–I’ll number them so we can refer to them in discussion:
(C1) “…it seems clear that Anaximander and Anaximenes were doing philosophy, had views about philosophy, and (in some sense) had the concept of philosophy, even though both lived before the Greek term for philosophy, ‘philosophia,’ was coined by the Pythagoreans.”
(C2) “…Aristotle claims that there are ‘unnamed virtues,’ by which he means virtues that he has a concept for but which do not have names in Greek.”
(C3) “…outside of crossword puzzle fans and those in the shoelace industry, almost no English speakers know the word ‘aglet.’ However, I submit that almost all English speakers have the concept of ‘the plastic or metal tip on the end of a shoelace.’
The principle is fallacious, Van Norden further clarifies, because it requires one-to-one lexical correspondence (pp. 22-23): “A slightly different kind of case that illustrates why the lexical fallacy is a fallacy is that of someone who has no one word for a concept because she has several words for one concept.” Bryan asks:
(Q1) “Do we speakers of English not have the concept that corresponds to the contemporary Chinese ‘wo 我’ because we use two different words that translate it in different contexts: ‘I’ and ‘me’?”
(Q2) “Do speakers of contemporary Chinese not have the concept of ‘million’ because they have to use a phrase to express it rather than one lexical item (‘bai wan 百萬,’ literally ‘hundred ten-thousands’)?
I have a few questions that I hope we can discuss here.
First, what do we think of these counterexamples? Are they obvious?
Second, what would be the reason, or argument, that explains why L is fallacious and therefore why C1-C3 are counterexamples to L?
Third, is L something of a “straw man”? Is there some more plausible lexical principle that Rosemont should be after, for his purpose (which I assume is to invite a skeptical attitude toward projects like Bryan’s that take there to be in China theories of virtue or of human rights or whatever despite, let’s suppose, there being no clear “one-to-one” translatable Chinese term). Maybe there is some lexical principle like L that does not require one-to-one lexical correspondence but instead some other lexical correspondence?
Lot’s of questions. Please feel free to comment in any sort of way…
Addendum: In Rosemont’s “Why Take Rights Seriously? A Confucian Critique” (in Leroy S. Rouner, ed., Human Rights and the World’s Religions), he apparently relies on correspondence of “concept clusters” rather than on one-to-one lexical items to make his arguments. Though I haven’t actually read that piece, it seems like something like that would provide a more plausible principle than L. Maybe someone who’s more familiar with Rosemont’s human rights discussions can fill us in…(I’m looking at you Steve Angle).
It seems to me that L is fallacious because one need not have the lexicon ‘aglet’ to have a theory of ‘aglet’. Although whether having the lexicon allows one to have a more developed theory of an aglet may be a possibility I would be open to.
FWIW, Slingerland discusses the lexical fallacy under the term ‘word fetishism’ in “Conceptual Metaphor Theory as Methodology for Comparative Religion,” Journal of the AmericanAcademy of Religion 72.1 (March 2004): 1-31. This piece seems even more directed at Rosemont as he argues that ‘rights’ can be spoken of cross-culturally. His argument is that there are deeper cognitive processes (i.e., metaphors) that are independent of language which are rooted in shared human experience.
Agui and others: isn’t there a worry here that “plastic or metal tip on the end of a shoelace” has a very different sense from “aglet”? Perhaps the phrase and the term refer to the same thing, but we’re talking about what concepts a language user has. Doesn’t it matter how that concept is “presented” through language? To pull out the philosophy of language example, if I have a concept of “the evening star” does that mean I also have the concept of “the planet Venus,” if for the sake of hypothesis, I don’t have the lexical items “planet” and “Venus”?
I should say: “…if for the sake of hypothesis, I don’t have the lexical items ‘planet’ and ‘Venus’ and I don’t have the concepts that correspond to those items.”
I think none of the three counterexamples C1-3 is a prima facie counterexample to L.
(C1, Milesians): It’s unclear to me why we should think Anaximander and Anaximenes had “the concept of philosophy.”
(C2, Aristotle): Rosemont’s L says “term” rather than “word.” “Term” is commonly used in a broad way to include phrases, not just words. Aristotle has a phrase for his one nameless virtue. There is hardly room to doubt what Rosemont would say if he were asked whether his principle was about single words only.
(C3, Aglet): Rosemont’s L is not about the concepts writers have. It’s about the concepts we can have reason to think they have, on the basis of their texts. But I think there’s an important point behind this example, which is that when things are “articulated”—that is, have joints—in ways that are directly apparent and natural to notice, presumably the people involved have the relevant concepts. I’m sure the ancient Chinese had concepts for those doohickeys on a certain kind of bronze thingamajig, because it’s natural to have such a concept if those things are around.
Bryan offers Q1 (wo 我) and Q2 (million) as instantiating a single kind of case that is different from the three above. But in fact Q2 is just like C2: it’s a case of having a phrase rather than a single word (if we suppose with Bryan that bai wan 百萬 is not a word).
Q1 is different and interesting.
(Bryan asks, “Do we speakers of English not have the concept that corresponds to the contemporary Chinese”wo 我” because we use two different words …?” But again, L is not about what concepts people have; it’s about what concepts we can infer that others have on the basis of their texts. Our sense of what concepts we have isn’t relevant.)
Q1 is a case where (A) one language has a word for which another language has *no* exact synonym because instead it has two words one or the other of which can be used in each case where the first language’s word could be used. That’s different from the uninteresting case where (B) one language has two exact synonyms for a word in another language. Bryan’s account of the case fudges the difference: he describes the case as one where someone “has no one word for a concept because she has several words for one concept.” If, as Bryan says, “I” and “me” are words for one concept, then they are each words for the same concept as “wo 我”, so we have uninteresting case (A). But if the difference between “I” and “me” qualifies them as not being words for the same concept, then we have interesting case (B). And in that case I think we have to say that the languages do not by themselves justify the claim that English speakers have the concept wo我; though they come close and the difference may be, as Bryan says, unimportant.
Of all the cases Bryan cites I think the aglet is the most interesting. Of all the examples, it provokes the most interesting questions about what we mean by “concept” (must it be verbal?) and about what kind of evidence other than words or phrases there might be for someone’s having a concept (and the importance of background assumptions about how people think).
I entered #5 before seeing ##2-4. Agui’s mention of rights brings out a very important point Bryan makes: subtle differences in concepts may not be important. That is, when we say that so-and-so didn’t have a concept of XYZ, our claim strongly suggests that so-and-so didn’t have a near equivalent. We seem to be saying that XYZ was very far from so-and-so’s thoughts.
if I have a concept of “the evening star” does that mean I also have the concept of “the planet Venus,”
The issue is the relation of ‘concepts’ to ‘terms’ (at least this seems to be the way Bryan is setting it up). Must one have a term ‘aglet’ in order to have a concept ‘aglet’? I think it certainly possible that one could have a concept of it without having a term for it. Pushing the issue further one could ask about shared conceptions–how do we compare/contrast our conceptualizations of ‘aglet’? Personally I don’t have so much of a problem in this case or in the case of Venus as the ‘it’ in question is an object external to us. Conceptualizations of things may be more or less complex, but as long as they’re something ‘out there’ there seems to be a shared basis for conceptualization–the ‘it’ is part of both of our worlds.
The question of ‘philosophy’ and ‘virtue ethics’, however I find a bit more ambiguous. Philosophy and aglets, to me, are different kinds of things. The issue here perhaps becomes one of how we share conceptualizations without a clear referent the conceptualizations pertain to.
Hi Agui,
I guess the reason I’m happy to ascribe to my neighbors a concept of aglets is that I think aglets are salient; as a category they’re easy to notice.
I wonder whether before today I had a concept of the second centimeter of shoelace in from the base of each aglet. Better yet, substitute a unit of measurement I’ve never heard of. People for whom such a concept is important might legitimately say of me that I lack the concept, I think.
Here are some tribes:
Tribe A doesn’t keep track of the lights in the sky, and hasn’t noticed that there are bright ones near dawn and dusk.
Tribe B has noticed those two lights, which it assumes are flames on the dome of the sky. (This is maybe a little different from Bryan’s I/me case?)
Tribe C has noticed those two flames and assumes they are the same flame, traveling alongside the bonfire that is the sun. (Flames on earth sometimes travel.) The tribe calls the flame ‘Ven’.
Tribe D believes in great heavenly balls, but thinks those two are different balls.
Tribe E thinks they’re the same ball.
Tribe A has no concept of Venus. Maybe the same is true of Tribe B. Tribe C has enough of a concept of Venus to fix the name ‘Ven’ to that planet as a referent, but lacks a concept of Venus as a planet. Does Tribe D have a concept of Venus? I think tribe E has a concpet of Venus, but I’m not sure what my standard is. Maybe it has to do with the idea of shared metaphors, which Manyul reports from Slingerland in #1.
I wonder whether your point about concepts of philosophy and of virtue ethics is that there is a sense in which these entities are like badminton: they don’t exist unless the participants have concepts of them.
One of my worries about these things is that I doubt that we have clear concepts of them. ‘Philosophy’ and ‘virtue ethics’ are words that function partly as labels for somewhat miscellaneous collections of people or projects, and as putative vehicles for concepts. ‘Rights’ is a little like that too, insofar as it’s a “buzzword”.
Here’s another thing I wonder: what’s the difference between saying someone doesn’t have a concept of X, and saying she doesn’t have the concept of X?
I’ve mixed up Manyul and Agui a couple of times … sorry! Manyul first mentioned rights, and Agui reported Slingerland on metaphors.
Agui,
“The question of ‘philosophy’ and ‘virtue ethics’, however I find a bit more ambiguous. Philosophy and aglets, to me, are different kinds of things. The issue here perhaps becomes one of how we share conceptualizations without a clear referent the conceptualizations pertain to.”
Excellent point; I agree largely, though I’m not sure even having a clear referent helps in some cases, e.g. “the evening star” and “the planet Venus”–if someone in Ancient Mesopotamia refers to Venus but we can’t clearly attribute to her the concept ‘planet’ (which seems likely)–then whatever terms are available to her, it is unlikely that she actually has the concept ‘the planet Venus'(again, EVEN IF she refers to the planet Venus by something like ‘the evening star’).
I do agree that it is harder to determine whether we share conceptualizations that don’t have as referents heavenly bodies, material cultural artifacts, and the sort. Take “virtue,” for example, as referring to something like “relatively stable dispositions to act in ways that express or promote human flourishing” (I’m working from memory here–can’t walk upstairs to get Bryan’s book; the kids just fell asleep). So the question is really whether there are enough of the right concepts in early China to build the right concept cluster that is similar to “virtue.” (Aside: I think Roger Ames suggested during the Chad Hansen festival, that it was unlikely the early Chinese had the right sorts views about human psychology to attribute such ideas to them.)
I should add that there are two questions here, importantly different:
(1) CAN we find the right conceptual matter in early China to construct a notion of virtue (or whatever other notion is in question)?
(2) DO we find the notion of virtue (or whatever other notion) in early Chinese texts?
I think focusing on #2 gives rise to the temptation to some kind of lexical principle; but I’m not sure it is a bad temptation. Shouldn’t the second question be our focus? Why would the first even matter if the answer to the second is “No”?
Bryan defines “virtues” on p. 39: “Virtues are relatively stable dispositions, the possession of which contributes to leading a flourishing life. … Virtues are dispositions to think, feel, perceive, *and* act in characteristic ways.” He says flourishing is “a technical term in virtue ethics,” and defines it on p. 37f: “To flourish is to live a certain kind of life: a life characterized by the ordered exercise of one’s characteristic capacities as a human. … The claim that flourishing is the ‘ordered’ exercise of one’s capacities reflects the view that some activities are more valuable, or more worthwhile, than others.” (It’s not clear to me whether the “activities” mentioned on p. 37 are meant to encompass all four items mentioned on p. 39, or just the fourth: “think, feel, perceive, and act.”)
The basic idea seems to be that virtues are dispositions to think, feel, perceive, and act in the most valuable or worthwhile ways. In other words, they are the best character traits. I heard Roger Ames’ talk at the conference for Chad, but I don’t remember what relevant views about human psychology he said the early Chinese might have lacked.
One kind of evidence that, say, Confucius had the concept “best character traits”, independent of his having those component terms, is that he occasionally offered lists whose members are in our view good candidates for “best character traits” and don’t seem to be limited by any other differentia.
Just a couple of quick, rather disjointed comments.
Rosemont’s statement may or may not be right, but I doubt it’s fallacious, in the sense of being based on faulty reasoning.
Henry says that to ascribe a concept to an ancient thinker, one must find a term (which could be a phrase, as I think Bill points out), *expressing* that concept in the thinker’s texts. Two terms can express the same concept without being exact equivalents, in the sense of sharing exactly the same pattern of use. So a one-to-one matching isn’t needed for speakers of different languages to share certain concepts. But Henry’s remark doesn’t imply that it’s needed. It implies only that you have to find *some* term–which could be “thingy” or “whatchamacallit”–expressing the concept.
Concepts can overlap partially without being identical. The Chinese “jie3” (elder sister, miss) and the English “sister” are examples. In such cases, it would be misleading to say that speakers of one or the other of the languages compared lack a concept possessed by the other. But it is also misleading to say that their concepts are the same. A careful comparison will explain both the similarities and the differences.
At least for interpretive purposes, I tend to follow Robert Brandom in individuating concepts on the basis of their inferential role. To say that two concepts are the same, then, is to claim that share the same pattern of relations to other concepts, beliefs, and actions. If this pattern of relations is significantly different, then the concepts are different, though again they may overlap. This is why concept clusters are important, as Manyul and others mention. The inferential relations between concepts in a cluster may be significantly different in different intellectual contexts.
Can someone have a “theory of X” or “view of X” without ever expressing a concept that corresponds, even partly, to our concept of X? If X refers to a general subject area—-such as “ethics” or “psychology” or even “motivation”—-then in many cases, I suppose so. But in those cases, they are likely to be using a few concepts that overlap with ours.
If X is a more specific concept, such as virtue or human nature, I’m less sure. In a few some cases, one could make a good case that a thinker has an *implicit* theory or view of X, if he says things cogently interpretable as descriptions of X without mentioning X itself. But I’d tend to question the motivation for doing this.
As an example, consider whether the Mohist core essays have a theory of human nature, despite never using a term that expresses the concept of human nature. Clearly, they have various beliefs about how humans tend to behave, both in a state of nature and in political society. But are these beliefs about human *nature*? I don’t see what grounds there could be for supporting either a yes or no answer, since the texts just don’t talk about the issue. Any motivation for attributing a theory of human nature to them is due to concerns extraneous to the aim of understanding their ideas. So I would recommend first attempting to understand them on their own terms, and only then, separately, moving on to apply that understanding to other concerns that interest us.
That’s very interesting, Chris.
I think you’re right to stress concept-clusters. If Tribe EG has Euclidean geometry and Tribe RC has only rudimentary carpentry instead, then the EG people may rightly say, “The RC word that we translate as ‘triangle’ really means something like ‘grossly triangular physical object.’” And when an EG person takes pains to teach an RC person a concept of “abstract immaterial perfect triangle,” the RC person will still in a sense lack the EG concept of ‘triangle’, because she won’t be able to use it the way the EG people do. She doesn’t have the other concepts that dovetail with it in interesting inferences. Though here we might want to distinguish between actual use in inferences and preparedness to use it in case the questions are raised.
Separately: If, when we are trying hard to think well, you and I are especially careful about definitions and about the borders of the extensions of our terms, and especially careful to make our thoughts follow the patterns of formal logic, then our thinking will tend to involve entities well-named as “concepts,” especially if we take inferential role as definitive of the shape of a concept. But suppose that when Smith and Jones are trying to think well, they are especially careful about paradigmatic cases rather than borders or definitions, and focus on kinds of thinking other than inference: association, projection, imagination. Then it will be less true of Smith and Jones that they have “concepts,” that what they’ve carefully worked up are “concepts,” even of the main things they think about. Their main thinking-tools may be best individuated not just by inferential roles, but also by other roles. Does anyone happen to know whether Brandom discusses the difference between inferential and other important roles in thinking?
Chris writes: “In a few some cases, one could make a good case that a thinker has an *implicit* theory or view of X, if he says things cogently interpretable as descriptions of X without mentioning X itself. But I’d tend to question the motivation for doing this.”
What kinds of bad motivation for such claims do you think are especially likely, Chris? I think one good reason for investigating the implicit views of e.g. Confucius on topic X is that we think X is one of the real central questions of ethics and we want to learn what we can about ethics from Confucius. Another is that we are setting out Confucius’ views for nonspecialists (which is a leading ulterior aim of interpretive investigation).
Chris writes: Clearly, [the Mohists] have various beliefs about how humans tend to behave, both in a state of nature and in political society. But are these beliefs about human *nature*?”
By “human nature” I think I mean how humans in general tend to behave in the main sorts of context they might encounter, so my answer to your question would be Yes, at least if the kinds of political context the Mohists discuss are representative, in their view, of the range of main possibilities.
Chris writes: “Any motivation for attributing a theory of human nature to them is due to concerns extraneous to the aim of understanding their ideas. So I would recommend first attempting to understand them on their own terms, and only then, separately, moving on to apply that understanding to other concerns that interest us.”
I guess you would agree that that sort of method can be taken too far.
I don’t see why the motivation has to be extraneous. The aim of *my* understanding their ideas has to involve interpreting them to *me.* And my interpretive effort is going to have to involve questions I can raise, as part of my effort to stretch my imagination to receive new ideas.
I’m not sure what to make of the other counterexamples van Norden gives, but I think (Q1) is no good. The distinction between ‘I’ and ‘me’ is just grammatical. If English-speakers lack the concept ‘wo’, then Cicero lacked the concept expressed by the English word ‘person’ (or any other English noun), since he would have used ‘homo’, ‘hominem’, etc. in different grammatical contexts. But this seems implausible.
But Chris’s comment got me thinking about another possible counterexample. My brother’s wife’s sister tells me that there’s a single word in Spanish to describe my relation to her. If I were to write a book about my extended family, I would probably never use a term that refers to the abstract relationship we stand in to each other (e.g., “sibling of a sibling-in-law”). Would the absence of such a term in my book be good reason to think that I lack the relevant concept?
Excellent point about ‘I’ and ‘me,’ David. I think with the “sibling of a sibling-in-law” case, this would be like the artifact case…if you have shoelaces and tips on shoelaces, then you likely have the concept of an aglet even without the term. So I would say that if we had evidence of marriage-like relationships in your social context and sibling-like relationships that were recognized in the same context, then we could infer that you had a concept of the sibling of a sibling-in-law. There would, I take it, be a substantive anthropological question to be answered in both cases, maybe more so in the marriage-like relationship case than the sibling case.
Your example raises an interesting further point. It might be possible to construct the right concept cluster in the object culture to be able to attribute to it a concept cognate to one of our own. But that points to at least this hypothesis one might make about the relationship between concepts and a lexicon: if a lexicon lacks a specific term for a concept for which our lexicon has one, it is likely that the concept plays a smaller role, if any at all, in the discourses of that other lexicon than of our own. For example, “relation-on-one’s-mother’s-side” is marked in Korean by the term ‘weh’ (wai 外, in Chinese); so, one’s-grandmother-on-one’s-mother’s-side has a specific term: ‘weh-harmoni’ (vs. simply ‘harmoni’ for father’s side grandmother). In English there isn’t a specific term for weh-harmoni but obviously the concept can be constructed. But doesn’t that at least show the concept plays a smaller role, if at all, in English-speaking cultures? Anyway, it’s a hypothesis. I’m sure someone will think of good counterexamples…
David, I’m inclined to agree with you about Q1 (I/me). Certainly if we’re considering saying things like “The Laputans had no concept of X” we should not worry about such possible fine points as whether ‘I’ and ‘me’ express different concepts. I suppose that’s the point Bryan is making with the counterexample.
Still, in the abstract, I wonder whether ‘I’ and ‘me’ or ‘homo’ and ‘hominem’ don’t express slightly different concepts. Do ‘dog’ and ‘of a dog’ express different concepts? Do ‘of a dog’ and ‘to a dog’? ‘Dog’ and ‘dogs’? In a language in which every “noun” is declined through five or six or eight cases, it makes sense to have the concept of “words” or “lexical items” that span all five or six or eight, and thus have no unique spelling. I speak of such a lexical item when I say “‘Homo’ in Latin can have any of six cases.” (Is it six?) Does such a lexical item express a concept? Could it be the same concept as the English ‘man’? I find these questions puzzling.
Manyul, I agree. It does seem that if a person’s lexicon lacks an item for X, then the concept of X is probably playing a smaller role in her “discourse.” Though I’d want to add three points: (a) The point becomes more interesting and stays true if we substitute ‘thinking’ for ‘discourse’. (b) the idea of a lexicon’s having or lacking something applies more to individual words than to phrases, and (c) e.g. early Chinese philosophical culture does not stress, to anything like the extent that ours does, the creation of a new term whenever we want to use a new idea, so your proposal may be more true of us than of others.
In a paper of mine I offer the fact that Confucius uses the word ‘harmony’ only twice in the Analects (13.23, 16.1) as some evidence that harmony isn’t a leading concept in his thought. I wonder whether that’s right.
Chris, I wondered above about the difference between “having a concept of X” and “having the concept of X.” One way to characterize the difference is this: I suppose two tribes might each have “a concept of Venus” without having the same concept, but two tribes can’t have “the concept of Venus” without having the same concept. Brandom’s standard is presumably for determining whether concept A and concept B are the same concept, not for determining whether they are concepts of the same thing. Is that right? I suppose that when we ask whether the early Chinese had a/the concept of truth, what we’re interested in is whether they were equipped to have views about truth: whether they readily referred to it.
Chris, another point on the them-first method: if I’m studying the views of the living, then real immersion is a live possibility. But where what I’m studying is a few snippets of text from a culture that is itself very poorly evidenced, I have to do a lot of reasoning to work toward interpretations, and that means relying more on my ideas.
Manyul, it seems to me that what makes it the case that those of us who lack the word ‘aglet’ nevertheless have the concept is not that aglets or shoelaces are artifacts. For most us they are just things we see around, not things we make. I think the reason we have the concept is something like this: that aglets are saliently similar to each other and saliently different from other things. The same applies to such concepts as ‘sibling’. Of course, the words we do have will make a difference to what categories we don’t have single words or short phrases for will be salient to us.
We have other kinds of signs than words: maps, logos, graphs, charts, models, etc. So did the early Chinese. I wonder how far that’s relevant to “concepts.” The other signs might not be as relevant to inference as words are, but they are relevant to reference.
Hi, I’m not a philosopher, and haven’t read much (er, any?) Chinese philosophy (outside this blog!), so this comment might go a bit wide of the mark. But I think I’ve got two connected points to make.
First, in Rosemont’s original statement it says: “the only way it can be maintained that a particular concept was held by an author is…”
Now this is describing an explicitly historical project. Rosemount is talking about using evidence in texts to attempt a reconstruction of an historical personage.
(Though it is possible he’s using the word author ambiguously, as Chris Fraser does above: “Any motivation for attributing a theory of human nature to them is due to concerns extraneous to the aim of understanding their ideas.” As it stands, Chris’s assertion is untrue. If we want to understand the ideas of “them”, the Mohists, then we must indeed consider whether they had a theory of human nature. But I think I’m right in interpreting “their ideas” as “the ideas expressed in their extant texts”. This convention of using an author’s name to refer to a set of texts must be very carefully negotiated when addressing this issue.)
Reconstructing a person is a larger project than just understanding a text, because a text will never express the entirety of a person. There must always be inferences, and the assumptions of the reconstructor will inevitably play a role. It is not merely a philosophical project, but involves at the very least history and psychology (individual psychology for an individual author, group psychology for a group of authors).
As it is not (just) a philosophical project, Chris “would question the motivation for doing this.” I assume the answer lies in the difficulty of early Chinese texts and the uncertainty over authorship and text relationships. In this field philosophy necessarily advances hand in hand with historical research.
The second point is that I wonder if the question isn’t being put in a slightly backwards way here. In the original post, Manyul Im suggests that Rosemont’s purpose is “to invite a skeptical attitude toward projects like Bryan’s that take there to be in China theories of virtue or of human rights”.
Not having read any of this literature, I’m guessing here, but presumably Van Norden has found some evidence in the texts which leads him to reconstruct authors who do have theories of virtue or human rights. If Rosemont’s criticism is that these theories are not explicit in the texts, then it’s really irrelevant, because what Van Norden is doing is not describing the texts; he is attempting to reconstruct a plausible human mind that would have created the texts we have. (Correct me if I’m wrong on this!)
So it seems likely that Rosemont isn’t doing this. What he’s doing is looking for evidence of an absence – evidence that the minds which created these texts did not have the theories suggested by Van Norden. The question is not, under what circumstances can we say that X had a theory of Y? The question is, under what circumstances can we say that X did not have a theory of Y?
If this is true, then Rosemont’s assertion is incorrect as it baldly stands. I find no mention of human rights in the works of Delia Smith (Britain’s Martha Stewart), but this does not imply that she lacks the concept. So what Rosemont could be saying is something like this: ‘lots of ancient Chinese philosophers wrote about topics which are intimately related to the modern ideas of virtue and rights. If they had such concepts, they definitely would have made them explicit in the texts.’
This obviously rests on a psychological theory which may need to be made explicit and compared with the kind of psychological theories that Van Norden is using in his reconstruction of ancient Chinese psyches.
Thanks to Manyul for blogging so elegantly. It’s a pleasure to read this blog, even though I understand very little of it!
Phil,
Thanks for the nice compliments. I think you’re right that Rosemont’s real point in introducing L is to try to show that the Chinese or the early Chinese do not or did not have certain concepts. With ‘rights,’ it seems like Rosemont often argues something more like: the contemporary Chinese political and intellectual context can do well enough without the concept of rights and might do better without it, even if they “have” the concept–understand it, are competent with it, etc. I think anyone who tries to defend some version of L with respect to the *early* Chinese would be motivated by the same aim as you suggest–e.g. trying to show that the early Chinese did *not* have a concept of virtue, or perhaps more plausibly, a concept of morality.
“Evidence of absence,” as you so astutely put it, is very difficult to show in any exhaustive way, at least in the respect. To show evidence of textual absence of something in a text is not knockdown evidence of conceptual absence of the same thing, say, in the thoughts of the author, authors, or audience of the text.
Herbert Fingarette, by the way, was taken to task using this very point by many, for his argument in chapter 2 (“A Way without a Crossroads” ) of *Confucius: the Secular as Sacred* that suggested that Confucius lacked any real theory of blameworthy or praiseworthy choice largely because of the absence of any clear discussion of such choice in the Analects. “How could the absence in the text *prove* absence in Confucius’ thought?” the critics wondered.
On the other hand–and here I’m repeating what I always thought Fingarette’s real argument was–if a text has *conspicuous* absences of something, i.e. when it seems like the concept would be “called for” or “natural,” or if the concept would or should be an available part of a solution to a problem, then that gives us a good reason to suspect that the concept does not exist–is not available–for the author(s) of the text. I confess that I make this kind of move with respect to ‘universals’ in trying to understand why Gongsun Longzi and his interlocutor argue the way that they do about whether a white horse is a horse. I confess it, but I’m not sure it’s really a hermeneutic sin. But I’m open to edification.
I dug up Rosemont’s actual text. Principle L comes in the fourth of a list of four points, making up a very long footnote. In the third point he reports Fodor’s distinction between lexical and phrasal concepts (roughly: concepts expressed by a single word or only by a phrase), and says of his own paper, “it has seemed appropriate in this text to take ‘concept’ as Fodor’s ‘lexical concept’ and to use it as synonymous with ‘term’ …”. The fourth point is presented as an amplification of the “importance” of this terminological distinction in the study of e.g. ancient Chinese philosophy, where texts are the only evidence we have.
But it’s unclear whether in the fourth point (including L), Rosemont is using ‘concept’ in the way endorsed in point three. If he is, L is wholly trivial; if he is not, L seems implausible (to recap the dilemma Phil laid out in #19). It rather looks as though he hasn’t made up his mind. Indeed, it’s hard to see what could motivate using ‘concept’ to mean ‘term’ except a desire to fudge the distinction.