Dao: A Journal of Comparative Philosophy has an annual best essay award. The award comes with a check of $1,000 and a certificate; the award winning essay will be set for free access; and a special panel will be arranged at APA Eastern Annual Meeting, whether the the check and the certificate will be awarded, together with a critical discussion of the award-winning essay.
At the beginning of each year, a selection committee of three editorial board members of the journal is formed to select three best essays published in the journal in the past year. These three essays are then sent to the whole editorial board for evaluation. The final selection of the best essay is made by a vote of the whole editorial board.
We have just finished this process of selecting the best essay published in our journal in 2011, and Edward Slingerland’s essay, “Metaphor and Meaning in Early China,” published in Dao (2011) 10:1–30, wins the award. The article is for free access at:
Here is the official citation:
This is a ground-breaking essay. Slingerland debunks a fairly common assumption that Chinese way of thinking is metaphoric, while the Western way of thinking is logical, an assumption shared by both earlier Orientalists, who claimed the superiority of the Occidental, and more recent “reverse Orientialists,” who claim the superiority of the Oriental. In contrast, using his expertise in contemporary cognitive sciences, Slingerland argues convincingly that metaphor is a universal and fundamental feature of human cognition. What makes the Chinese way of thinking unique is thus not that it is metaphoric but that early Chinese thinkers were more self-aware of the metaphoric nature of language, while modern Western thinkers are more self-deluded about what they are doing. The essay as a whole is thus original in its interdisciplinary, comparative, and philosophical natures. It is the type of work that Dao aims to promote.
For what it’s worth, I enjoyed this article. However, I wish Slingerland would engage the more philosophical literature (meaning, in part, that which is not utterly dependent on the cognitive science stuff) on analogy and metaphor, especially that which is critical of the cognitive science material. I introduce some of that literature here (should anyone be interested): http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1804987
I have only begun to come to terms with Slingerland’s paper. There is a great deal in it that I like very much. (I’ve glanced at Patrick’s piece, and it too looks wonderful.) For example, I am enthusiastic about Slingerland’s main proposal about early Chinese thought:
what is unusual about early Chinese thinkers is not that they relied upon metaphor or metaphoric blends, but rather that they devoted a great deal of conscious attention to developing vivid and consistent sets of interlocking metaphors and metaphorical blends, which makes metaphor and blend analysis a particularly crucial tool when approaching these texts. (2)
This statement is partly a large generalization about thinkers other than the early Chinese; and I have some doubt about that claim. (How many poems don’t aim to present vivid interlocking metaphors?) But as regarding early Chinese thinkers, this proposal strikes me as plausible and exciting, suggesting lots of interesting inquiries.
Two small worries: First, I gather that early Chinese thinkers neither marked a distinction between the metaphoric and the non-metaphoric, nor articulated the notion of degrees of metaphoricity. If that is true, then they can’t exactly have paid “conscious” attention to something as defined by the term ‘metaphor’. Even so, I imagine the thesis can be restated to avoid the problem.
Second, for a set of interlocking metaphors and metaphoric blends, I don’t feel very well oriented to the idea of its being “consistent.”
Hm. Maybe one part of consistency would have to be some kind of recognition in practice of a difference between literal and metaphoric levels, or among degrees of metaphor. If you have a dark coffee cup and a bright coffee cup, don’t expect them generate new little coffee cups in the cupboard on account of being female and male. And if you’re going to bury a set of coffee cups and a barista with the honored deceased, make sure they’re not real ones.
The practice of not expecting new little coffee cups and the practice of burying only fake cups, in order to amount to a practical recognition of the concept of metaphor, would have to be recognized in some way as being applications of the same idea, not just as two different matters.
It would be interesting to try to catalogue the different kinds of way early Chinese practice or theory did what we might call distinguishing the literal from various species of the metaphoric. We might thus find or work up a bunch of interesting concepts kin to ‘metaphor’.
Is it crazy to think that it would be quite a different thing to regard such practices as distinguishing the metaphoric from various species of the literal?
The key term in the above-quoted thesis about early Chinese thought is “metaphor.” Without that term—i.e., without a distinction between what is metaphoric and what isn’t—the thesis seems to lose all clarity and all interest.
But Slingerland rejects
the “Western” assumption that the literal versus metaphorical distinction really means something: that is, that there is a class of words or expressions—the “literal”— that convey an abstract, amodal [nonperspectival] meaning that, in turn, refers in some direct way to categories in the world. These “literal” meanings can then be contrasted with “metaphorical” expressions that merely coordinate or juxtapose one domain with another, but do not necessarily tell us anything about the world. (6f)
This rejection puzzles me in a few ways.
It’s unclear whether the brief initial claim (that the literal/metaphoric distinction doesn’t mean anything) is the claim that the distinction does not exist (i.e. the two words are meaningless when used to mark a distinction), or that it is a distinction that has no significance/importance.
The extended account of the distinction given here might be read as a concession that the distinction exists. And the distinction described in the extended account doesn’t look offhand like a trivial one: so the distinction would turn out to be both existent and important.
But the thought might be instead that the first part of the account, the definition of “literal,” is an account of something that does not exist, so that the distinction is at best unimportant.
I want to say a few things in defense of the claim that the literal-metaphoric distinction exists and is important (and thereby defend the paper’s thesis about early Chinese thought).
I disagree with Slingerland’s above account of the putative distinction between the literal and the metaphoric, the distinction that has been marked by those terms in “the West.”
FIRST, Slingerland’s account seems to cast the distinction as a distinction between kinds of word (or longer expression). But I think that’s not right. The distinction is between ways of using a word, phrase, sentence, etc.; though some sentences more likely to be used in one way than the other.
“The world is a vampire.”
“There’s a hole in Daddy’s arm where all the money goes.”
“Add two teaspoons of nutmeg.”
“The glass is half full.”
“Smith had a black heart, according to the coroner.”
But perhaps by “class of words or expressions” Slingerland meant “class of uses of a word or expression.”
SECOND, Slingerland’s account suggests that the putative distinction involves the idea that the literal differs from the metaphoric in being abstract, pertaining to categories rather than individuals. But I think the distinction has not been conceived that way. I’ll offer two sets of arguments:
(1) On the one hand, the (putatively) literal needn’t be abstract at all. For (i) a proper name can be used literally: as when one responds to a demand from Hitler with “Yes, Hitler.” If one responds in that way to a demand from someone else, one is using the name metaphorically. Indeed, (ii) if having a term to refer to something abstract is a trickier achievement for a community than having a term to refer to particular distinctive things at hand such as people and places, it will be a little harder to be literal about abstractions than to be literal about particulars.
(2) On the other hand, the metaphoric needn’t be concrete or particular. For (i) any term can be used metaphorically, even the most abstract ones. (I heard a university president tell the whole faculty that the university is a triangle—without specifying the angles, though he did distribute a diagram.) And (ii) typically what a metaphor says is that X is like Y in some abstract respect (which may or may not also be a very fuzzy respect, but that’s a different issue). If it were wholly unclear what respect the speaker had in mind, or if she had none in mind any more than any other, then I suppose the communication would fail as a metaphor.
THIRD, it seems to me that the account of metaphoric [uses of] expressions as ones that “merely coordinate or juxtapose one domain with another” is empty verbiage.
FOURTH, Slingerland’s account strongly suggests that the distinction inherently involves the opinion that the metaphoric isn’t worth much; that metaphors are likely not to say anything. That is certainly an opinion one might have, though I suspect it has never been a common opinion among educated people anywhere. And I grant that the silly opinion is logically tied to the distinction, in exactly the same way that Slingerland’s thesis about early Chinese thought is tied to the distinction: both are claims about metaphor. But the distinction itself (between two ways of using an expression) does not itself involve either the silly opinion or the exciting thesis.
(A philosopher who accepts a literal/metaphoric distinction might write that metaphoric statements don’t articulate propositions, and Smith could quote that to show that she thinks metaphors don’t say anything, don’t have meaning. But she might write nearby that she didn’t mean they don’t communicate propositions; she only meant they don’t articulate them, and by this she meant they don’t do so literally. That might be an empty claim. But her making it would show that Smith was wrong. So let’s be careful.)
( Consistently with the Silly Opinion’s never having been a common opinion among educated people anywhere, the silly opinion might have been a common opinion among philosophers for centuries, or in a school of philosophers for a few years. Philosophers say the darnedest things. Regarding a different opinion, incompatible with the silly one just discussed, Slingerland writes:
As Mark Johnson has observed (Johnson 1981b), the Western philosophical
tradition has long been characterized by a view of metaphor as philosophically
superfluous: a decorative rhetorical device expressing a thought capable of being
fully reduced to some literal equivalent, and therefore merely entertaining at best,
and potentially misleading at worst. (2)
One can’t tell from the bibliography, but I guess the reference is to a paper of Johnson’s included in a collection he edited, “Philosophical Perspectives on Metaphor.” I don’t have that handy, but I want to make a few points anyway. Slingerland’s words are ambiguous: they can be taken as a claim about the mainstream view among Western philosophers from the beginning to today, or as a claim about what some Western philosophers have held in recent decades, including today. The context strongly suggests that Slingerland is talking about the whole arc of the Western tradition. The reference to today is unambiguous in “has long,” but surely this must be a slip, as in fact the view has long not been common? The view as characterized seems to involve an obviously fallacious argument—X is possible therefore easy—which is the main thing that makes the view seem extreme; but there is a general presumption against attributing a transparent gross fallacy to a philosophical tradition or even to the hack-work in the train of a philosophical fad.
Indeed I wonder if a kind of interpretive enterprise that Slingerland approves in this paper—“conceptual metaphor analysis … unpacking the implications of particular foundational images” (9) might also be described as working toward (more) literal analyses of metaphors. )
Those four objections are enough for now.
If Slingerland’s claim were only that the distinction as conceived by the account he gave has no meaning (a claim I address a couple of paragraphs after quoting it above), then my four objections wouldn’t oppose him. They would however argue that his claim is simply irrelevant to a discussion of the meaningfulness (existence or importance) of the distinction between the literal and the metaphoric. I think he means it to be relevant.
Immediately following the claim about the meaninglessness of the distinction, quoted earlier, Slingerland sums up his argument for that claim:
Taking empirical work on human cognition seriously—as I will argue below we should—means moving beyond this dichotomy and viewing all human language and cognition as, to a greater or lesser degree, imagistic. (7)
I suppose the idea is: All uses of language are more or less imagistic, which implies more or less metaphoric [does it?]; so no uses are purely literal; so there is no meaningful distinction between the literal and the metaphoric. In short: there isn’t a distinction because what we have instead is a scale without endpoints. That is, there is no more a distinction between the literal and the metaphoric than there is between the small and the big, or between straight wooden yardsticks and crooked ones.
A commonplace I grew up with is that most words originated from metaphors. The transition from metaphor to cliche (“dead metaphor”) to plain literality is a matter of gradual degree. (I won’t pause to say how this point is consistent with my claim above that the lit/met distinction is not a distinction among kinds of word.) So if Slingerland were using cognitive science only to show that metaphor shades perfectly gradually into literality, as a matter of degree, I think he would be arguing for nothing new or controversial. Rather the important claim is that on the less metaphoric side of the scale there is no endpoint. There’s no such thing as the purely literal, as there is no such thing as a perfectly straight wooden yardstick, or a perfectly small objection.
Now, someone who has a full understanding of the relativity of size might still bridle at the claim that there is no distinction between small and big, or that the distinction is “meaningless.” (It seems quite intelligible to say that what is distinctive about some culture’s philosophical tradition is that its books are really big, or written in really big characters.)
But that’s not the line of thought I want to pursue. I want to propose instead that in two senses the literal/metaphoric scale does have an endpoint at the literal end. For each sense, I’ll use the yardstick example as a model (a metaphor).
(A) For practical purposes, of course there is such a thing as a straight yardstick. Very small deviations are insignificant, trivial, negligible. Their unavoidability doesn’t support the claim that the everyday straight/crooked distinction for yardsticks is nonexistent or unimportant, or merely a matter of relative degree. (Other examples: a clean photocopy, a clean shirt, unscented deodorant, a large company’s balanced books, having paid one’s bills, being honest.) The same may be true of the everyday literal/metaphoric distinction, which is of course presumably the only or main form of the literal/metaphoric distinction that Westerners who write about Chinese philosophy have been concerned with.
(B) We have an idea (if I may put it that way) of perfect straightness that is not undermined by the inevitable crookedness of wood. I mean, we can do geometry, we can do proofs in geometry. I think the distinction “more/less crooked” presupposes that idea of straightness, at least as a limit. (And even if it’s just a metaphor to help us think about wooden yardsticks.) Similarly, I think Slingerland’s concept “to a greater or lesser degree imagistic” presupposes the idea of the non-imagistic at least as a limit (yes?), and I think the distinction “more/less metaphoric” presupposes the idea of the literal, at least as a limit.
That last point is far from obviously true. The straight/crooked distinction does not presuppose an idea of absolute crookedness as a limit, nor does the small/big distinction presuppose absolute bigness, nor is there a limit at either end of east/west. But, just offhand, it seems to me that on the literal/metaphoric scale, the term “literal” aims to mark a negative pole, a pole of absence or nondeviation as it were, while the idea of “metaphor” is about one kind of deviation, other kinds of deviation also being possible instead (metonymy, synecdoche!).
Now, the practical importance of everyday straightness pertains to the things we can do with it. Pure conceptual straightness is practically important too, at least partly because of the idea of performing indefinitely many operations on it. To change the subject from geometry to numbers: you can’t have arithmetic unless you have an idea of exact numbers. If numbers were never quite exact, then adding 0 many times, or adding and then subtracting 1 many times, might move you from 31 to 32 or 30. Etc. If we thought that way, we could not have astronomy, physics, accounting, economics, or computers.
(In computers that learn by experience, which often become too complex for us to untangle, some of the functionality sometimes comes from physical side-effects of the hardware. This kind of deviation from a calculus needn’t be problematic, except maybe when we’re designing weapons we can’t afford to test, or working out whether CERN will make a black hole …)
The abstract science that literalness gives us is logic. (And math I guess.) Consider: “A rat is a kind of test tube. A toddler is a kind of rat. Therefore a toddler is a kind of test tube.”
Logic is not perfectly reliable in practice, because we are made of wood. But it is important.
Philosophy is a funny thing. It’s where we get most confused, and so in a way have most need of the crutch of logic; but it’s an area where our ideas are especially metaphoric, so logic is in a way least helpful.
I guess most Western philosophers have all along aimed to make reliable crafts or sciences about their topics: that is, to figure out the basics (of this or that) well enough that lesser minds, or minds less wholly given over to the enterprise, could be well enough oriented toward further inquiry that they could proceed in it and then start certifying each other, correctly, as reliable authorities for the rest of us, even if they stay wood.
It seems to me that that’s a main way that Western thought has managed to be cooperative, collective, and thus to make progress. That is, I propose, by working out how to break down various kinds of problem and idea (abstractions, or chemistry, or history) into bits that travel well: that can be understood, thought-about, discussed, investigated and answered without having to look outside the sentences to decide what is the respect in which X is supposed to be like Y. In brief: we go from House to diagnostic Q/A software.
How about that?
In illustration of point a.FIRST above:
In criticizing Slingerland’s paper above, I said the literal use of language (or at least the ideal of the literal use of language) makes logic possible. I was a little confused; I was to some extent making the same mistake I had (in effect) accused him of making: the mistake of confusing (a) the distinction between metaphorical and literal (i.e. standard) uses of a word or phrase, and (b) the distinction between imagistic and, uh, conceptual thinking or reference: I mean thinking that succeeds in referring to definite individuals, classes, features. I’m not sure I presented my thought in enough detail to make it actually depend on my confusion … for there is a case to be made that logic depends on at least having the ideal of the second term in each of the two distinctions. I’ll explain…
First I’ll lay out the two distinctions in a simple way that just ignores all the profound, plausible, famous worries about language and/or thought, and never mind the point that one term of the distinction shades gradually into the other.
To use a word literally is to use it in the standard conventional way, with the standard conventional meaning (however vague and/or imagistic that might be). That’s what “literally” means. To use the same word metaphorically is to use it to say something else instead, something constructed using the literal meaning as a reference-point. If applying word W to a thing standardly means that the thing is L, then to use W metaphorically of a thing is to say that the thing is like L — in some respects one expects the audience to guess. There is no standard or conventional answer to the question, “Like L in what respects?,” because if (or insofar as) there were a standard answer, that use of W would then be a literal use, not a metaphorical use. It would be a literal use even if the mechanisms of the thinking involved were imagistic.
The core meaning of the word ‘metaphor’ is to refer to a kind of figure of speech. That is, it’s a way of using the literal meaning of a word or phrase to construct a new and different local meaning. Thus what gets used metaphorically is words and phrases. But we sometimes speak of using things as metaphors: using water as a metaphor, for example. In that extended use of ‘metaphor’, the metaphorical does not contrast with the literal (though I suppose we could say, metaphorically, that one is using water “literally” if one is drinking it).
Now, logic arguably depends on the idea of literal (and therefore non-metaphorical) use of terms, in that logical manipulations depends on the assumption that the sentences we use have standard meanings, meanings insulated from influence by the context (what statements are next to it in the argument, what the weather is like when the argument is being made, etc.).
How important that point is in practice, and how resistant it is to being exploded by profound plausible famous worries about language and thought, depends on what logic is for. The main thing it is for, I think, is to help us collaborate in thinking: to help the different people collaborate and to help the different parts of the same person’s sensibility and opinion/set collaborate. Resistance to variation in the meanings of symbols seems pretty central to the whole idea of logic. So I think the idea that logic (at least as an ideal) depends on the literal use of language (at least as an ideal) is of great and robust practical importance.
Let’s turn to the other distinction, between imagistic thinking and thinking that succeeds in referring to definite individuals/classes/features. It’s obvious enough that logic (at least as an ideal) depends on the latter kind of thinking (at least as a kind of ideal). What isn’t so obvious is the relation between the two terms of the distinction. In the previous distinction, one term was in its very concept a departure from the other. But in the present distinction there seems to be no neat relation between the terms. Offhand it looks as though imagistic thinking is defined by a kind of mechanism of representation (obviously not the only mechanism; recall Peirce’s tripartite distinction between representing by similarity, by causal correlation, and by convention). And offhand it looks as though the other kind of thinking is defined by a kind of success, not by a kind of mechanism.