Their words, for you

Several months ago, it's been so many, now, that I know not how many, wondering about the remarkable metaphor with which Davidson begins "What Metaphors Mean" ("metaphor is the dreamwork of language"), and how far he actually goes to expound upon it, and additionally remembering that I found the ending suggestive (but of what?—I couldn't remember!), I re-read just the beginning and ending of that essay. The ending is suggestive, of a kind of Romantic coextension of critical and artistic creative activity, as, in fact, is the beginning (at least, I associate it with German Romanticism). It also seems to beg to be read alongside Cavell's "Excursus on Wittgenstein's Vision of Language". (I recall being very pleased with myself for inserting WMM between two bouts of Cavell when, ages ago, I taught Philosophy & Literature; I believe I found following it all with some Danto pleasing as well—though I can no longer trace that angle.)

I had some further inchoate thoughts that I thought I might give form and existence by writing them, about—not the initiating metaphor, but rather—the putative contrast between Davidson's "brute force" theory, as as David Hills terms it, and Hills' own preference for a pretense theory. For it seemed to me—to be clear, this is before actually reading the entirety of the Davidson, or reacquainting myself with any of the actual details of the latter style of account—that the main thrust of the essay, that "metaphorical meanings" and "metaphorical truths" are the upshots, not the inputs, to understanding a metaphor, which operates by getting us (somehow—there's a reason Hills calls it "brute force") to notice certain, well, things is the normal nonspecific term to use here, but is unfortunately countable (Davidson when explaining himself adverts to pictures and cautions us from counting), so let's say that a metaphor gets us, perhaps helped along by a critic who is engaged in basically the same kind of activity, to notice the components of the metaphor in a certain way, is one that is compatible with, and perhaps even necessary for, a make-believe account—which would be relegated to filling out the "how". For how do we know what, or how, the speaker is making-believe in their metaphorizing? (How do we collaborate, in interactive, joint metaphor-making, without throwing each other off the rails, or stepping fully out of the game, and at best making observations about how one might play it—something that can also be fun, but is a different fun from that of playing the game?) If understanding a metaphor is conceived of as taking or being able to take up a game inaugurated but not wholly delimited by the metaphor-maker, cottoning on to the rules of a game already underway as it's played, or recollecting the play in tranquility, it seems all the more natural to conceive of the "metaphorical meaning" and the metaphorical truths thus recovered, if any, as the outputs of the successful, simpatico playing of the game, and successfully picking up on the game proposed or in progress would seem to be already similar to the successful noticing of what the metaphor-maker wishes to draw our attention to; it already requires sensitivity, discretion, taste, etc. Οὔτε λέγει οὔτε κρύπτει ἀλλὰ σημαίνει, and likewise for this sort of game-player; the player doesn't tell you the nature of the game, and the game, and what the players take from it, is apt to outstrip whatever the inaugurating player thought, occurrently or not. (As Hills as much as says outright, in "Aptness and Truth in Verbal Metaphor", p 145.)

But then of course I actually did re-read the Davidson, and some of Hills' papers, and am here recording some impressions, rather than attempting to make a single argument, though perhaps the chorus will point to a secret law regardless. Let's agree, you and I, to understand by these abbreviations the names that follow them: "WMM" for "What Metaphors Mean", by Davidson; "AT" for "Aptness and Truth in Verbal Metaphor", "PP" for "Problems of Paraphrase: Bottom's Dream", "M" for the "metaphor" article on SEP, and "WH" for "Τhe What and the How of Metaphorical Imagining, Part I", all by David Hills; and "MFM" and "MPOMB" for "Metaphor, Fictionalism, Make-Believe" and "Metaphor and Prop Oriented Make-Believe", respectively, by Kendall Walton. Many of the thoughts about Davidson will also advert to Hills, which is, you know, it's fine. It's not organized, but it's fine.

1. Davidson's visual emphasis. It's striking the extent to which Davidson adverts to the visual. When "he was burned up" was a live metaphor, he avers, "we would have pictured fire in the eyes or smoke coming out of the ears" (253). He denies that "associated with a metaphor is a definite cognitive content that its author wishes to convey" (262), but allows that there is something which can be called "what the author of a metaphor wanted us to see" which (definite?) thing "a more sensitive or educated reader grasps" (264). Here he's talking about the critic, hence the contrast; the critic produces a paraphrase ("in benign competition with the metaphor maker"; recall that most metaphorical paraphrasis is also metaphorical) to "make the lazy or ignorant reader have a vision like that of the skilled critic" (264; this is also very Cavellian). And of course there's this famous bit:

How many facts or propositions are conveyed by a photograph? None, an infinity, or one great unstatable fact? Bad question. A picture is not worth a thousand words, or any other number. Words are the wrong currency to exchange for a picture. (263)

I'm not really sure what to make of this, to be honest—is it predominantly in covert support of his contention that what a metaphor induces the audience to come to awareness of is non-propositional? I certainly would be hard put to assign any visual content to "metaphor is the dreamwork of language". (When I read Robinson Jeffers' "And protest, only a bubble in the molten mass, pops and sighs out, and the mass hardens", the image that comes to mind is, no word of a lie and perhaps shamefully, oatmeal as it cooks. Despite the rather blatant availability of magma!) Is it just a stand-in for the fact that he doesn't have anything to say about how the noticing of similarities or the non-propositional whatsits occurs, and falls back to a visual analogy? All he really does with it is contend that the interpretation of metaphors, like the making of metaphors, is not guided by rules. (Actually, what he says is that the interpretation of metaphors is as little guided by rules as is the making of metaphors (245). Naturally one suspects that what he means by this is what anyone would mean by it, that neither is guided by rules at all. But it's a happy fact that he put it as he did, for while the interpretation of dreams is surely not guided completely by rules that can be laid down once and for all in advance, it's also not a hermeneutic free-for-all, and it's because of the putative similarity between the interpretation of dreams, and that of metaphor, that Davidson makes this remark about rules. A point worth making, because Davidson's contention is also a point of contention between him and Hills, in PP. But the "rules of a special and difficult-to-elicit kind, offering a special and difficult-to-elicit kind of guidance" of which Hills speaks (PP, 5) don't seem obviously incompatible with mild hermeneutic heuristics that Davidson needn't fear. But we now verge on Bentham-like impenetrability.) Others were not so cautious: think of Max Black and his "system of commonplaces", which is admittedly not exactly effectively computable, but manages to say something.

2. (Davdison on) the inexhaustibility of metaphors. That a metaphor is inexhaustible in some sense is a commonplace, at least since Cavell's famous observation of the "and so on" that ends most attempts at paraphase. (Hills points out that Cavell referred to Empson on the "pregnancy" of metaphor, but I don't know if Empson also suggests that one could just go on forever.) This is, I think, oversold. Davidson in particular seems to oversell it here, though note the final sentence:

[I]n fact there is no limit to what a metaphor calls to our attention … When we try to say what a metaphor 'means', we soon realize there is no end to what we want to mention. If someone draws his finger along a coastline on a map, or mentions the beauty and deftness of a line in a Picasso etching, how many things are drawn to your attention? You might list a great many, but you could not finish since the idea of finishing would have no clear application. (263)

Between the beginning and the end of this short excerpt are two quite different senses in which even a successful paraphrase ought to end with "and so on"! The first part expresses the commonplace in an uncommonly strong form: not only is the metaphor endlessly productive, but we soon realize this; we have the sense quickly that we could go on endlessly. If one can't finish this task, that's only because (if we aren't simply mistaken) one's life will end first.

This one simply seems false to me. I just don't think—based on personal experience!—that we do realize this at all, in many cases, let alone "soon", and I think (cynically) that many people who say that they do realize this, even soon, would take it back if asked to actually sit down and do it for a period of, say, several hours straight. Doing this kind of thing is work, for one thing, and while Davidson suggests that one is simply spinning out things one noticed on the spot, one will quickly find, if one tries this, that in order to get the similarities out, one must first put effort in. A coastline, actually, is not a bad example, if seemingly inadvertently: there's always more detail, but only if you get closer. (Though Davidson actually refers to "running [a] finger along a coastline on a map"; I'm not really sure what one is supposed to notice in this.)

How many things can you really adduce about the sun, and about Juliet, especially keeping in mind that these things you bring in should be relevant to the scene at hand? (How does Romeo think the sun an illuminating thing with which to think about Juliet?) Only some of those things were "called to your attention"; the rest you sought out, and your seeking is apt to falter eventually.

Cavell to his credit notes that there is such a thing as "the over-reading of metaphors" ("Aesthetic Problems of Modern Philosophy", 79), and one might well think, on the one hand, that some metaphors are in fact best read not very much at all. "A great many effective similes are pretty well exhausted by the compact explanation their author promptly and explicitly supplies", says Hills (PP, 19), offering the Baconian "virtue is like a rich stone, best plain set" as an example; if we really think that this simile (rather than Bacon's purpose in making it) is exhausted by the gloss, what would we make of the alternative "virtue is a rich stone, best plain set"? What's gloss for the gloose, one wishes to say, is gloss for the glander; however much the reader's mind is set in motion—by either!—if one is interested in what the author is doing, the answer would seem to be, not much. A Bookforum review of a book by Pankaj Mishra yields such metaphors as "Burrowed in the mythic depths of society's unwritten constitution, no printed ray of rational reproof can strike them down" and "the Scot exemplifies the plaque accumulating in the brain trust of the transatlantic set"; one no doubt could follow the (geometric) "ray of rational reproof" out to infinity, but one does it, surely, to amuse oneself, not to understand Guan's point better.

On the other hand, some metaphors seem to call out for more reading than we can, in fact, provide, and Davidson's own opening metaphor is a good candidate, at least for me. Even Davidson himself doesn't quite gloss it:

Metaphor is the dreamwork of language and, like all dreamwork, its interpretation reflects as much on the interpreter as on the originator. The interpretation of dreams requires collaboration between a dreamer and a waker, even if they be the same person; and the act of interpretation is itself a work of the imagination. So too understanding a metaphor … (245)

Observe how Davidson immediately moves from "metaphor is the dreamwork of language" to the interpretation of metaphors and the interpretation of dreams. But neither the dream nor its interpretation is the dreamwork, and metaphors aren't metaphor. ("I'm going to have to stop you there. This is World of Golf. What you're describing is World of Golf Equipment. 'Golf' is an abstract noun.") Hills also: "I agree that a metaphor is like a dream" (5). That isn't the metaphor Davidson actually makes! In fact, what I was wondering, when as noted at the top I was wondering about precisely this metaphor, was whether anyone had attempted to expound on it in a way that really took its psychoanalytic element, and its abstractness, seriously, and directly, not by pivoting to the waking work of interpretation, and not even by assimilating it to "metaphors are the dreams of language". My own impression of this metaphor is that it feels evocative; it feels as if it's full of significance and meaning, but I absolutely could not get started with a paraphrase. My poor grasp of psychoanalytic theory may be part of that—and perhaps in that respect I'm one of the ignorant who need the aid of a critic. To some extent this is what Ted Cohen would call a hermetic metaphor; this is no "system of commonplaces" because "dreamwork" just isn't a commonplace concept. But, ignorant though I be, I simply don't feel as if the metaphor causes me to notice more than I could ever express; very far from it. But then, I also think that most of what people say about Romeo, Juliet, and the sun is overdoing it, relative to the scene. Even "his day begins with her"—this just doesn't seem to be of a piece with what Romeo goes on to say in his soliloquy. It is, certainly, something that could be meant by "Juliet is the sun", and it is also, certainly, a matter of dispute, perhaps not capable of being finally settled.

Which brings us to the other suggestion Davidson makes, and different it is, as an account of the seeming endlessness of paraphrase! "You could not finish since the idea of finishing would have no clear application." This is not at all saying that the noticing, the juxtaposition, the image, whatever, is infinitely rich, but rather that the activity of paraphrase (like other forms of criticism!—surely it's significant that he mentions "beauty and deftness" here) is itself open-ended and provisional. One ends with "and so on" not because one has other things in mind but doesn't want to bore the reader, and not even because one thinks one could extend the list but doesn't want to spend the time, but to signal the non-finality of the list already set down; more could be added later by a more insightful critic, and indeed some things could be removed. There isn't a once-and-for-all interpretation following which we'll be done forever; that just isn't the sort of project we're engaged in.

This is what I think Davidson ought to mean, and what I think people, in general, who emphasize the productivity of metaphor ought to mean. This is both because, as noted, I don't think the "and so on" ending a paraphrase does signal the inexhaustibility of most actual metaphors, but also because whatever I notice is whatever I notice, and whatever you notice may be something different, but we can surely disagree with each other about a metaphor, about our interpretations, which are not simply how things struck us after we took the metaphor in. And there's the matter of "what the author … wanted us to see", which the critic helps us grasp, which presumably acts as some kind of standard. (Though if we're critics after my own heart, regarding these metaphors as small-scale artworks, we won't be too fussed about the living, breathing author's opinion.) What ought I to notice? What is the best, the most satisfactory way of taking the metaphor, the one that is richest, lets us appreciate it the most deeply, illuminates its subject the most? (And—why not?—illuminates that in terms of which the subject is presented the most. Reflecting on what Churchill might have meant in calling Mussolini "the merest utensil of his master's will" is apt to make one think not only about how Mussolini is thus presented, but also and thereby about what kind of tool a "utensil" is. Someone unfamiliar with tall boots might find "Italy is a boot" instructive twice over.)

Here is a consideration that is perhaps more compelling than just "this is what I think he ought to mean". Davidson mostly speaks of metaphor, reasonably enough, as a "device", as simile is a device, and paronomasia, litotes, and syllepsis are devices. But he also, at the beginning and end, speaks of it as an artwork, a creative production somewhat more august than those other things. And so it is fitting if his position about metaphor and its interpretation at least could be applied to artworks and their interpretation more generally. (It is likely a fault of an account of metaphor if it makes metaphor too hard to connect to art in general.) And that's what I think taking him this way enables us to do; I think it's not for nothing that one gets a distinct whiff of Cavell in the final paragraph.

3. Davidson on metaphor's propriety I think the real point of WMM is this:

… what we attempt in 'paraphrasing' a metaphor [is] to evoke what the metaphor brings to our attention. I can imagine someone granting this and shrugging it off as no more than an insistence on restraint in using the word 'meaning'. This would be wrong. The central error about metaphor is most easily attacked when it takes the form of a theory of metaphorical meaning, but behind that theory, and statable independently, is the thesis that associated with a metaphor is a definite cognitive content that its author wishes to convey and that the interpreter must grasp if he is to get the message. This theory is false as a full [???] account of metaphor, whether or not we call the purported cognitive content a meaning. (262)

That is, not the stuff about "bringing to our attention" or even "nonpropositional contents" or limitless object of attention or the visual metaphor, but a more purely negative point: a metaphor does not involve a definite content that the maker has in mind and the appreciator must grasp or err, as if the point is to smuggle some thought out in disguise. No definite content, such that it is to be recovered by an audience. This is, I think, a somewhat modest claim, and one that's rather plausible in the case of grander, more gravid metaphors, especially if we think that one of the things implicit in the idea that the author "wishes to convey" a "definite" content is the further idea that the content is in some sense had in advance of the uttering by the author. In which case, it is also, I think, pretty plausible in instances of plain, non-metaphorical speech, too. There's a reason we often "work out" an idea by talking, or by writing; drafting isn't just a process of hitting on the aptest formulation of something that was formulated correctly all along in the language of thought but one of thinking in its own right. Hills's "oracular utterances"—which, I confess, I don't know if he's actually discussed in print, though he does use the phrase in AT—would be a good example, too.

But of course metaphors are present not only in on-the-fly utterances but in the final, worked-over product as well, in the sorts of works that are so worked that the author can account for the being of most of the bits. It remains plausible that the biographical author doesn't intend the fullness of what an audience defensibly finds in the metaphor (as in other cases of artistic interpretation!), but that author presumably has some idea why the thing's there and what it's supposed to be doing in the work and to/for the audience. (Nevertheless, that understanding could come after it's written down, in a fit of inspiration!) One can press the point by noting that Davidson himself is at pains to deny that "metaphor is confusing, merely emotive, unsuited to serious, scientific, or philosophic discourse"; rather, "metaphor is a legitimate device not only in literature but in science, philosophy, and the law" (246), domains in which, at least in their public accounts of themselves, vague suggestiveness is not favored.

One must hope, too, that Davidson wouldn't agree that whatever I happen to notice is equally correct, though since his gloss on paraphrase is just as a restatement of what one in fact does notice one may be hard put to pin that on him. One wishes to say that Lawrence, emerging from the desert and not yet having had his lemonade, overhearing Romeo and concluding that Juliet is merciless, something to flee from, not to be faced but feared, and whatnot, would be getting not only Romeo but Romeo's utterance all wrong, notwithstanding that a different person differently situated could use those same words to that different effect. We can account Lawrence's interpretation incorrect without recourse to a definite thing Romeo meant in the same way that we can account incorrect wildly off interpretations of any artistic production: the correct one is so much more satisfactory, coheres so much better with Romeo's tone when he speaks, what he does and says before and after, fits better with other things we'd like to say about or to him. But this is a rather third-personal way of carrying on (and it occurs to me to wonder, now, if radical interpretation in general isn't a rather third-personal way of carrying on, in addition to whether metaphor, aside from the occult noticing stuff, really even has to be that special for Davidson); at any rate, if we meant to understand how metaphor can legitimately be used in serious, scientific, or philosophic discourse despite the lack of a definite cognitive content its author wishes to convey, we may not be much helped. Maybe we are: does that legitimacy require that kind of standard of correctness? Perhaps it's enough that, in fact, the author can correctly predict, often enough, that such and such a metaphor will be taken in such and such a way. Perhaps the life of a metaphor is that their makers are their first audience, who trust that their audiences, in turn, will notice in them what they noticed; all whirling in the same being, not transmitting specific messages one to the other.

4. (Hills on) Davidson on paraphrasability as such. Hills calls Davidson an "opponent of paraphrase"; Davidson explicitly denies the possibility of what he chooses to term "paraphrase" as applied to metaphor:

I agree with the view that metaphors cannot be paraphrased, but I think this is … because there is nothing there to paraphrase. Paraphrase, whether possible or not, is appropriate to what is said: we try, in paraphrase, to say it another way. But if I am right, a metaphor doesn't say anything beyond its literal meaning … This is not, of course, to deny that a metaphor has a point, nor that that point can be brought out by using further words. (246)

But, on the other hand, while Davidson objected to shrugging off his denial that metaphors mean something as "an insistence in restraint in using the word 'meaning'" (262), I can't see any reason not to think of this as an insistence on restraint in using the word "paraphrase", since Davidson is quite clear that "so-called paraphrase" has a "legitimate function" (264). It just isn't properly so called, because Davidson has quite definite ideas about what "paraphrase" can mean. But it's hard for me, anyway, to see how even on Davidson's construal of the whole affair such an effort doesn't fall under "a bringing out with further words of the point of some initial words", and what's wrong with that as a capacious understanding of "paraphrase"? It's not as if it's a particularly load-bearing bit of jargon, or as if the meaning Davidson brings in is so clearly the sole legitimate reason.

For this reason, I think Hills's argument against Davidson in PP misfires: "If we had a valid objection to paraphrasis along these Davidsonian lines", he writes about the coastline passage, "we'd have a parallel and equally valid objection to ecphrasis … [but as] in the case of ecphrasis, the fact that the idea of finishing lacks clear application in no way entails that the idea of starting is in the same boat" (30). But Davidson doesn't object to the enterprises of ecphrasis or (so-called, for him) paraphrasis as such. The objection is that this doesn't give you anything which he's willing to countenance as "the meaning", that "definite cognitive content that its author wishes to convey". Hills does think that paraphrasis is an attempt to recover or discover the (metaphorical) truth-conditions and in that sense the metaphorical meaning (but not "the definite cognitive content the author [wished] to convey"!), but that's a further thesis about paraphrasis, not the entry fee for those who want to tolerate the activity. You really have to lean in to the "exchange" in "Words are the wrong currency to exchange for a picture": no matter how many words you hand over, you never get the whole picture in return (despite what's said about Lichtenberg's Commentaries on Hogarth's Engravings), just so many partial views. If you think that you have to lean in to "wrong currency" and wish to argue that there's a conceptual mistake here somewhere, you do still have to confront the fact that Davidson doesn't argue against the practice of paraphrasis (and presumably wouldn't against ecphrasis), but against the name, and a mistake one might make in thinking about what one gets out of it.

5. Hills on metaphorical meaning: a preliminary. Hills thinks, for various reasons, that we ought to recognize a distinctive metaphorical meaning, and category of metaphorical truth; he thinks doing so, moreover, could lead to wide-ranging revisions in the way we think about language and truth in general. I'm not so sure about that, despite agreeing with him about many of the more particular claims he makes along the way, such as these:

[(a)] An ambitious paraphrase of an ambitious metaphor can quickly outrun both what the speaker could plausibly have had in mind at the outset and what listeners could reasonably be expected to gather from his words on the spot. Even the least ambitious metaphors have a knack for lending themselves to manifestly unintended construals … [(b)] what's at stake in getting an ordinary intertexual allusion is often merely (some part of) what we should gather from someone else's words, [but] what's at stake in getting a metaphor is what we should understand by the words … what truth-conditional content we should assign the words themselves.

This suggestion conflicts with … [(c)] the thought that language use is invariably a matter of getting something across or getting something down—of communicating or recording an already formed thought.

(AT, 145–6; similar thoughts crop up in the concluding section of PP, and PP §17.)

Well, of those, at the ones I've labeled (a) and (c), anyway. (b) strikes me as more questionable, and it strikes me that it should strike Hills as less than fully established, too, because he also, elsewhere, speaks specifically of metaphorical truth, and once you've made that allowance, haven't you allowed that we can keep assigning the non-metaphorical truth conditions the way we used to? And it would seem that (b) is vulnerable to a Davidson-like objection: Hills's isn't, by his own telling, what he calls a "semantic twist" account; "infant" in "Tolstoy was a great moralizing infant" isn't imbued with a special nonce literal meaning. For us to appreciate what Mann's metaphor is about, we have to take "infant" as referring to infants, and "Tolstoy" to Tolstoy, and all the other words as having all their usual denotations. Hills suggests further that "if we think of metaphorical truth values as determined by metaphorical sentence contents" (AT 146–7), then we can further break down the latter into metaphorical expressions and start doing compositional semantics on metaphors. It is not obvious to me that even for Hills we should satisfy the premise: is that how a pretense theory operates? It seems a much more holistic affair than that. (Perhaps this is the gist of (d) in PP, 28.) He states that "to take words metaphorically is to assign content to them twice over" (147), but that leaves it open whether we're actually assigning meanings wordwise rather than to "the words" as a whole. (How does one assign content twice to the individual words in "the work is the death mask of its conception"? Doesn't it rather invite one to think about the creative and intellectual process generally, its completion or culmination generally, satisfactions and finality? Paraphrasis rarely amounts to saying "take these two words with these new significances".)

It likely will not be necessary to point out how like Davidson on metaphor (c) on language-use in general is. Davidson thought that a metaphor's not communicating a thought formed in advance was grounds for saying the metaphor didn't mean anything beyond what it literally means; Hills doesn't, but again, insofar as Hills speaks of a metaphorical meaning for metaphors, with their own truth conditions, it isn't clear to me how much these positions actually conflict. If we assign contents twice over—and we have to do that; we shouldn't abandon our initial interpretation of the "twice-apt metaphor" (PP, 40) Scottie makes in Vertigo (which is also his initial interpretation): you shouldn't keep souvenirs of a killing; it's bad hygiene—we can keep our account of the first content and the question of how the second content should affect our theorizing in general is still open. (One reason we shouldn't simply dismiss the literal meaning of Scottie's utterance is that the aesthetic value of the metaphor he expresses with it is enhanced by the fact that he conveyed, if that's the right word, that idea by a double meaning. And we can expect the initial single meaning to be important wherever there's a question of oracularity or "brainstorming", I suspect.)

6. Hills on metaphorical meaning: disagreement. One reason—I think—that Hills wishes to recognize a distinctive metaphorical meaning is related to the Frege-Geach problem, though I can't remember if he's put it this way in print, and hence am relying on my memory, which is over a decade old (I have discovered, to my surprise and distress, that I can no longer find, and probably not longer have, the notes Hills distributed to the members of his seminar on metaphor that I took in … 2005). But, basically, if someone says to you, "if music be the food of love, play on", how are you to take that, and what are you to do? Play on, or no? But since I can't remember to what end he introduced this thought—it may also have been to caution against a too-narrow view of the forms metaphor takes—and since I am also not sure how to integrate it with what I (with all my limitations!) understand to be his understanding of metaphorical truth anyway, perhaps we should let it lie. Somewhat more explicitly, though, in AT, he brings up the pattern of disagreement with metaphors. One thing you can do with a set of words is convey an idea which is not literally entailed by those words, as in, famously, conveying the idea that so-and-so is no good at philosophy by putting in your letter of recommendation nothing more than "so-and-so is punctual and has very neat handwriting". But what if you think so-and-so is just great at philosophy?

We need to consider the specific verbal forms that agreement, disagreement, and questioning can plausibly take in different cases. In general, when I get one thing across by saying something else, my listener can't agree with, disagree with, or question, the thing I get across by using the standard devices for assenting to, dissenting from, or challenging the something else—what my words actually (literally) say. (AT, 127)

You can't say, "yes he is!" or "that's just not so" to "so-and-so is very punctual" and thereby disagree with the conveyed thought that so-and-so is bad at philosophy (by asserting the contrary, in the first case, or simply by denial, in the second). But those who think that Juliet's not really all that and reply "bosh!" to Romeo's utterance "Juliet is the sun" aren't denying that she's literally the sun, are they? They're denying whatever they understand Romeo's metaphor to mean:

So it would appear that Romeo's meaning gets lodged in Romeo's words in a way that Grice's meaning … never gets lodged in Grice's words. The words of Romeo's utterance … get taken so as to express a thought they wouldn't express if they were taken literally—one which may be true or false or indeterminate in its truth value, one to which we are free to respond in ways that are appropriate only to thoughts that speakers have actually put into words. (AT, 127, emphasis added.)

When the Juliet-skeptics say "bosh", are they literally denying? Or metaphorically? If they said outright "Juliet is not the sun", that would be their continuation of the metaphor by means of denial, just as surely as another speaker's having said "arise, fair sun", instead of Romeo's having said it, would be their continuation of the metaphor by means of extension. The truth value Hills speaks of is presumably a real one, real truth or real falsity or real indeterminacy, not a pretend one, but if it appears that a meaning with such a value "gets lodged in Romeo's words" because one can issue a denial, then hadn't that better be a real denial, not a pretend one? (If it's a pretend denial, and a truth value that lives only as long as the pretense does, what do we do with "if music be the food of love, play on"?) It seems that Hills is thinking of the denial as the denial not, as I'm tempted to put it, of the metaphor, but of what the metaphor means—Romeo's meaning, that gets into his words, which mean something else. (How distant we are from the concluding sections of AT! Just a page later Hills supposes that by "semantic success" we might "mean something like conveying a determinate, determinately true proposition" (AT 128): an ambitious metaphor doesn't seem to have much hope for semantic success. And what's this about Romeo's thought?)

7. A distinction between two kinds of paraphrase of metaphor. The sort of paraphrase we tend to see, when the example is Romeo's speech, is what you might call one that illuminates the metaphor by extending it: it shows one, so to speak, how the instant metaphor works, but showing one how one might operate with the system the metaphor implies or suggests. This is what Davidson is on about when he says that "the critic is, so to speak, in benign competition wit the metaphor maker. The critic tries to make his own art easier or more transparent in some respects than the original, but at the same time he tries to reproduce in others some of the effects the original had on him" (WMM, 264); this takes taste, sensitivity, and creativity, just as does formulating a good metaphor—one that might require the aid of a critic—in the first place. This also seems to fit with the sort of prop-orientation, and game-likeness, that Walton and Hills discuss. Light, perhaps, will dawn gradually over the whole; you'll pick up the knack of how to talk about A in terms of B, and in so doing you'll get what the point of the original metaphor was all along. A notable feature of this kind of paraphrase is that the sentences it produces tend also to be metaphors. My suspicion is that this is what folks mostly are thinking of when they think of "paraphrase" in this specific philosophical-metaphorical context.

Here's an example, from William James, of another kind of paraphrase, adduced by Hills in PP, in the section on "authorized paraphrase" (which contains many examples of, mostly, this kind of paraphrase). As in his text, the italics are added indicate which part is the metaphor (of interest):

Habit is thus the enormous flywheel of society, its most precious conservative agent. It alone is what keeps us all within the bounds of ordinance, and saves the children of fortune from the envious uprisings of the poor. It alone prevents the hardest and most repulsive walks of life from being deserted by those brought up to tread therein.

Aside, perhaps, from the bit after the comma in the first sentence, the rest, while it may contain other metaphors, doesn't really contain more metaphors that show us how to think about society with reference to flywheels, as might have been the case (with some damage to accuracy) had James written that habit absorbs and renders smooth smooth the boisterous activity of the unruly poor. Or compare this other example of a metaphor "fully prepared to paraphrase itself" (PP, 24): "You're the cream in my coffee / You're the salt in my stew / You will always be / My necessity / I'd be lost without you". Imagine that you've been asked to paraphrase the metaphor—just the one in the second line will do—with Cavell's paraphrase of Romeo's as your example. You would never come up with the remaining three lines, just as Cavell would never have accepted those lines as an adequate paraphrase of Romeo's metaphor. It's not hard to continue in the extension line: you, the salt in my stew, are that thing that brings out the flavor of life, makes everything else better, even makes everything else more itself. Things are dull, lackluster without you; you are that unnoticed in itself thing that makes everything else worth noticing at all (is this taking it too far? Perhaps, but perhaps not; salt is a background player).

This second sort of paraphrase—the sort that finds "you're my necessity" a paraphrase of "you're the salt in my stew"—is a telling sort of paraphrase, similar to the way we explain an idiom by just telling someone "'kicked the bucket' means 'died'" (Cf PP 7; 25). The first is not. I mention this because of Hills' contention (in AT) that "metaphor is at once fully aesthetic and fully semantic" (157) and that "paraphrase undertakes to display (approximately and in part) what would make Romeo's metaphor, taken precisely as metaphor, come out true—in other words, what would make it come out metaphorically true" (125). The distinction is of interest in that the second sort of paraphrase provides an exit, as it were, from the metaphor: one could say that "you're the salt in my stew" is metaphorically true if "you will always be my necessity" is true according to the way it should be taken (hyperbolically, not literally, but also not metaphorically). But the first doesn't: "you're the salt in my stew" isn't metaphorically true if "you make things more vivid" is literally true (much less if "you bring out the flavor of other things" is literally true). Not only is that still a metaphor, it's a metaphor that really ought to be understood in the context of its being an explication or extension of saline culinary metaphor; the vividness here is that of an enhanced quiddity, not a greater intensity. And this is clearer in the case of the even more obviously solar paraphrases of "Juliet is the sun". "You're the salt in my stew" is metaphorically true if those other things exhibit the sort of truth apt for them, but that's just metaphorical truth again.

8. Further fictional truths; "Juliet just plain is the sun" Hills and Walton both chide Elisabeth Camp and Catherine Wearing for imputing to the them too robust a, or the wrong kind of, conception of what sort of imagining of Juliet as the sun/pretending Juliet to be the sun Romeo putatively does and we putatively are called on to join him in. Hills: "Once Romeo's words are uttered (they reason), once it is at least defeasibly fictional that Juliet is the sun, it is likewise at least defeasibly fictional that certain of the sun's actual properties belong to her 'by implication'" (WH, 27); they impute to the fictionalist several theses which neither he nor Walton hold, among them that "once a metaphor is sprung on us, we are invariably called on to imagine true what the metaphor literally says—that the primary subject is the secondary subject" and that "we'll find ourselves compelled to imaginatively ascribe to the primary subject whatever real properties of the secondary subject this signal of ours calls to mind" (28). Walton emphasizes repeatedly throughout MFM that the pretense theory is prop-oriented, and concerned with "conditional principles of generation, which determine what is or would be fictional should the prop possess certain properties" (4), stating that "insofar as make-believe is prop oriented, we are usually not concerned with implied fictional truths … there is no point in even raising the question of whether, if it is fictional that Bill is a bulldozer, this fictional truth implies that fictionally Bill is enormous, clanking, and diesel guzzling" (5–6).

It is not too difficult to see why someone might think that we are called on to imagine that the primary subject is the secondary subject: "In what spirit do Romeo and his listeners entertain the thought … that Juliet just plain is the sun? … In saying 'Juliet is the sun', Romeo pretends that she just plain is exactly that" (AT, 147). "We come to suspect that Romeo is imagining his new love to just plain be exactly that … His words serve to signal an understanding on his part … under the rules of the game that he and any suitably attuned listeners are playing together, he and they are to imagine Juliet to be the sun—imagine her to just plain be exactly that" (M; this phrase also occurs in WH, 24). Whether we do or not is another matter, but isn't that (just plain) what we are called on to imagine?—Whatever that actually means. I find it pretty hard to make out. One might be forgiven for thinking that imagining that she "just plain is" the sun means imagining that the temperature on her surface (her skin?) is many thousands of degrees: for isn't the sun just plain quite hot? ("A paraphrase of Romeo’s metaphor would specify features of Juliet in virtue of which the proposition that she is just plain is exactly that, the sun, comes out fictional" (WH, 25): if it's only properties x, y, and z in virtue of which Juliet counts as the sun, that doesn't mean that once she is the sun in the game, she lacks further solar properties. If she does lack those further properties, then she isn't "just plain" the sun—nicht wahr? She's sunlike in suchlike respects.) Or in general, for thinking that such imagining does mean imputing to Juliet properties derived first from the sun, even if we're constraining those properties to what we think Romeo had in mind.

Hills and Walton occasionally use the example of "Italy is a boot" to illustrate prop-oriented make-believe; this is, I take it, no more a metaphor than is "France is hexagonal" (that is, not a metaphor), but that's ok; it's meant to illustrate a certain kind of make-believe, not metaphor, though metaphor is also supposed to be prop-oriented make believe. Imagine someone saying that, cued by "think of Italy as a boot", you were to imagine that Italy "just plain is exactly that"! The suggestion strikes me, frankly, as absurd; I don't even know where to begin with it. But let's think about this. One way to proceed with "Italy is a boot" is to, say, look at Italy, and imagine that the north of Italy is a pant leg, tucked in to the top of the boot and bloused, and that Sicily and Sardinia are pebbles or clods of dirt casually kicked by the wearer. This, I take it, is truly prop-oriented; one considers the prop—Italy and goes from the initial prompt and further facts about it, facts which fulfill unstated conditional principles, to further pretend facts. This can be amusing, if somewhat empty. That's not the sort of thing Hills brings up, though: "turn Italy for the time being into an improvised representation of a boot, and you have a readymade scheme for locating particular Italian cities in relation to each other, deriving from established ways of thinking and talking about boots and their component parts. Games of this second sort are prop oriented" (WH, 22).

Now imagine that I, someone who knows something about the geography of Italy, am telling you, someone who knows its general shape and location in the Mediterranean but nothing about its internal articulation, but who is familiar with the parts of footwear, about the regions of Italy. "Think of Italy as a boot", I say. "Calabria comprises, more or less, the toe and vamp". (Or, definitely following Walton, "Crotone is on the vamp": MPOMB, 40) When I say this, I am relating the parts of Italy to the parts of a boot. But you can't do that; you don't know what the parts of Italy are. I'm instructing you. You understand it by relating the parts of a boot to the parts, which you are thereby learning about, of Italy. Surely it is an "implied fictional truth" that Calabria is near the heel of Italy?

Hills chastises Camp and Wearing for "a lack of curiosity about how clear cases of make believe actually work" (WH, 27). Hills himself is perceptive and interesting whenever he elucidates a metaphor, which makes it all the stranger that he excludes facts about the sun as having a role to play in understanding the metaphor, or playing the game (in opposing the second thesis I mentioned above, or in the enumeration on WH, 24). The prop-oriented order of things is that facts about the prop generate truths in the game in accordance with some principles of generation; facts about Juliet—her importance to Romeo, for instance—generate fictional truths about Juliet-the-sun does—nourishes him, for instance. There is no metaphorical truth that Juliet-the-sun dazzles or blinds or burns or becancers Romeo because there is no corresponding literal truth that Juliet the person damages him through exposure. (Er, at least, that isn't what Romeo believes!) The logical order is that the properties of the prop come first. But this is not the interpretive order! Especially when the prop is a cipher limned mostly by metaphor, you have to start from the facts about whatever it's said to be, and use those to guide your insight into the prop. (We don't know much else about "you" than that you're the top.) And even when we do have a pretty good idea about the prop, starting from the other end can produce insights one wouldn't otherwise have had. Yes, the facts about the prop must jibe with whatever you come up with in this way—it wouldn't be an insight, otherwise—but that's hermeneutic holism for you; we have to make sense of this Juliet being dazzling or dangerous, if we think the dangers of being too long in, or staring too long at, the sun are relevant to the metaphor.

It's less surprising that Walton excludes "facts about the sun"-type facts, and Walton's example is helpfully blunt:

Our [Walton's and Hill's] view is that in interpreting (2) [the sentence "Bill is a bulldozer"] the hearer recognizes a game with conditional principles of generation, but does not necessarily take it to be fictional that Bill is a bulldozer, let alone work out what would follow—either what fictional truths this one implies, or what would actually be the case if Bill were a bulldozer, or what else one would imagine about Bill if one imagined (pretended) him to be a bulldozer. What the speaker asserts, truly or falsely, is just that Bill has properties such as to make it fictional that he is a bulldozer.

The hearer will take it to be fictional that Bill is a bulldozer if she thinks the speaker's claim is true. (7)

The thing about this is that although Walton refers to what might happen "as the game proceeds" (7), it's unclear whether, or why, the game would proceed at all; it seems singularly uninteresting. In fact, the real game here would seem to be this: the speaker has in mind certain conditional rules of generation, and makes a claim that follows from those rules and an unstated premise, and hearers starts from what they believe about the speaker's beliefs about the Bill in the utterance and attempt to discover both the rules and the premise: in virtue of what might Bill be bulldozer? In other words, it's a riddle.

What distinguishes "Italy is a boot [sc. roughly, in respect of shape]" from "Bill is a bulldozer"? It seems as if we could eke out the latter with "in respect of insensitivity to obstacles" and be done with it, and if Bill and bulldozers actually are or are believed to be insensitive to the would-be obstacles each encounters, then are we not finished? Walton says that "the hearer will take it to be fictional that Bill is a bulldozer if she thinks the speaker's claim [that Bill is implacable] is true", but why would she do that? "Pointing out what fictional truths the props generate directly should the speaker’s claim be true suffices to call attention to the features of the prop we are interested in" (7). But the hearer's going on, once she has cottoned on to the principles of generation and had her attention brought to the interesting features of the prop, to take the generated truths to be fictional, or to participate in an established, ongoing game, seems entirely optional, on this presentation. (It also seems somewhat optimistic, or perhaps pessimistic, depending on your view of what's going on: it seems, that is, to presume that the features of the prop the speaker is interested in are not really all that obscure, wanting only the making of the metaphor to bring them into the light. It seems unambitious.) What would be the point? Even when Walton does entertain the extension of the metaphor, he imagines the initial speaker having initially said something more involved:

She might, extending the metaphor, describe Bill as an “enormous, clanking, diesel guzzling bulldozer,” thereby calling attention not only to Bill’s determination and stubborness, but also, let’s say, to his manner in meetings—his huffing and puffing, pounding the table, rising threateningly from his seat and clinking loudly on glasses to get attention, often after having had too much to drink. (8)

Walton's description of his own position is curiously one-directional, and this, I think, accounts for the curiously clue-like character of metaphors as he describes them. And the overall riddle-like character makes it hard to see how paraphrase could be open-ended, or even called for: there would seem to be one principle of generation. But if you allow facts about bulldozers to suggest new ways of conceiving of Bill, then you can begin to see how there might be something new afoot, and for that matter how the metaphor might actually be illuminating. Rather than just "what is it about Bill that makes so-and-so say he's a bulldozer?" you get "what does thinking about Bill and bulldozers together suggest?". And—this is why I mentioned Hills's skill earlier—this is what suggestive, successful paraphrasis actually does do. The proof of this is that one is able to say so much about Juliet, or the "you" who's the salt in "my" stew, or the creative process, at all: one thinks about the sun and sees the propriety of suggesting that Romeo's day begins with Juliet, salt and the propriety of saying that when you're in love, the world itself seems heightened, death masks and the propriety of saying—as explication if not endorsement—that the completed thing is a still reminder, a concretized something, of something that was once dynamic, and seemed as if it could be anything.

It may be objected that it's not the case that Romeo's day fictionally begins with Juliet because it actually begins with the sun, but rather because of some presumably literally stated fact about Juliet that suits her, in the putative game, to be the sun in this respect also. Thus we still do not really care about a derived truth; it's still all principles of generation. That may be. But if the idea of a "game" isn't itself simply a prop that is meant to aid our thinking about metaphor—then we must take seriously how it's actually played. We already know that, since a good, ambitious paraphrase (I take it that paraphrase is simply another form of participation) can outstrip anything that the maker could plausibly have thought at the time or that anyone listening could plausibly have understood at the time, that not all the principles of generation could have been objects of the players' thoughts. I find the suggestion that they're discovered distasteful. They are rather put in place, newly installed in an evolving game, by the player saying "Romeo's day begins with Juliet"; this is after all, as most ambitious paraphrases are, another metaphor. (Discovery would also suggest that one could get it wrong in paraphrasis, not because the paraphrase is unsatisfying, inconsistent, or even merely inferior to another, but because it depends on a principle that wasn't there to be discovered in the sense in which those which are discovered all. But this seems wrongheaded to me; the test for the interpretation is our satisfaction with its deliverances, given its success in hanging together with whatever else we've said, and that's enough.) And if that player has gotten there by asking "what, in the sun, is there in a beloved?", then the game is concerned with derived truths, and in a sense the fictional truths are fictionally true because of their being derived.

9. Hills on metaphorical meaning: intertextual allusion; oracular utterance. The more, or so it seems to me, one emphasizes that "metaphor … often strikes us as inspired or oracular" (AT, 145) or the possibility of a metaphor's proper or fruitful interpretation far outstripping anything that was or could have been meant, or understood, on the spot, the less plausible, and more significant, is the claim that the upshot of paraphrase is "what we should understand by the words … what truth-conditional content we should assign the words themselves", as distinct from "what we should gather from" them (ibid., italics in original), which is striking, because Hills makes these claims more or less back to back. The latter is Hills's assessment of the stakes in intertextual allusion (as you can see, I'm circling material somewhat):

The actual or merely suspected presence in one text of an allusion to earlier ones invites us to scour our cultural memories for appropriate earlier texts, our search being guided in part by how salient various earlier texts are for the community being addressed, in part by how satisfying in promises to be to read the new text in the light of this or that earlier one—with the upshot that an author may sometimes inadvertently allude to a book he's never heard of. But what's at stake in getting an ordinary intertextual allusion is often merely (some part of) what we should gather from someone else's words. While if I'm right about the status of paraphrase, what's at stake in getting a metaphor is what we should understand by the words, what we should take the words themselves to mean in this particular context, what truth-conditional content we should assign the words themselves. (ibid.)

(This may be too indirect to really constitute an allusion to Cavell's discussion, in "A Matter of Meaning It", of discussion of Fellini's possibly intended allusion to Philomel in La Strada, but that discussion is certainly not irrelevant.) Why would we say that someone can have alluded inadvertently? The fact that Hills refers to a "text" here is suggestive: it positions the case not as one in which, in the flow of conversation, someone says something which could have been construed as an allusion if one of the participants were to focus on it, but as one in which the words (spoken or written) are already the object of scrutiny as such. What's important is that one stands apart from the words and considers them as artistic interpretanda, things that can mean in multiple ways, things constructed from which we are called on to gather something. This isn't the normal transparency with which we interact with written or spoken words! Here are these things, the words, and this thing, the text; it's an object for contemplation in its own right, in which we can find significance. It's a bit of verbal art, basically, or something we're prepared to count as such; the reason the "normative influence of actual or inferable speaker intentions" is lessened is that we construe the thing as something for which the historical intending of a person is not considered relevant. The "status of paraphrase" Hills mentions is, I believe, that "we try to enjoy [metaphors] in order to understand them"; paraphrase is the vehicle of the enjoyment and delivers the understanding. (I think. I could easily have missed a more explicit bit earlier on.)

Consider an example of this phenomenon, which also happens to be an allusion (not a textual allusion) that sparks a bout of metaphoric thought: Scottie, in Vertigo, says to Judy, "There was where you made your mistake, Judy. You shouldn't keep souvenirs of a killing." Hills says of this utterance (as I mentioned before) that it's a "twice-apt metaphor", but it isn't, straightforwardly, a metaphor at all, not the way "Juliet is the sun" is. Nothing is said to be anything it isn't; there isn't even a Pound-like laying-alongside. If it's construed as a metaphor metaphorical (Hills refers to its metaphorical understanding, to taking it metaphorically) it's because Scottie, and presumably Judy, as it were hearing the echo of Scottie's statement, realize that "souvenir" could also have been used as a way of alluding to Scottie himself, and appreciating the justice of alluding in that way. But Scottie doesn't say he's a souvenir. (To point back to the talk of denial above: if Judy said "but I didn't keep you", or "but you aren't a souvenir", both of those remarks would be completely in-bounds given what they've both taken up from their sudden revision of their understanding of the import of Scottie's words. But isn't it clear that her saying either of these things would represent a taking up of what the identification Scottie makes, a participation in it, and a denial of its propriety?) That his words could have been used the way that he comes to accept them—that is, that "souvenir" could have referred to him—comes as a surprise to him—that's why it's "oracular"—and it's also something that he can only realize once he's said it—that's why it's an oracular utterance. Utterance, thinking out loud, is a defamiliarizing move; it objectifies the uncertain by putting it into some single specific form, which can then be thought about. I don't think "thinking out loud" is metaphorical; I think it's a mode of thinking itself, not merely the verbalization of an already formed thought, and one of its uses is that it forces the forming, because you have to actually say something. It's tempting to think that the alternative is a mentally indistinctly represented distinct, fully formed thought, but I think the truth is more often that it's just an inchoate mess in there until it's put into shape. And once it's put into shape you can take up an attitude to it, even be surprised by it. (It's a neat trick, but not of course dispositive, that an utterance is—and I only really realized this because of the defamiliarizing move of encountering its cognate, Äußerung, in German—a making-outer, an outering.) The utterance can be "oracular", can say more than you knew and say something back to you, because once objectified it can be inspected, have questions put to it. (This is also why bits and pieces of phrases can be repurposed; one can see in them and their parts new potentials. "Moneybags must be so lucky" is the title of a book, and the "must" is the must of "it must be nice to …". But it's extracted from a sentence of Marx in which the "must" conveys a necessity under which Moneybags stands: he is required to be so lucky as to find someone who …. "Now is the winter of our discontent" is ungrammatical if we try to read it as Shakespeare meant it.) But this is getting a bit far afield.

Let us consider what Hills says about Scottie's utterance, taken literally and taken metaphorically:

Taken literally, the utterance is an explanation of how Judy was found out … it's reckless to keep an object that connects you so directly to a crime in which you are implicated, and it's unhealthy to even want to renew one's memories of such a crime and one's role in it … Taken metaphorically … you shouldn't renew your connection to another person (me) in a way that turns him (me) into a cooped-up, immobilized bit of private property, simply for the sake of his capacity to help you recall a stretch of the past that you would both of you be better off trying to forget. (PP, 40)

What's interesting, to me, about these paraphrases is that the first concerns souvenirs of a killing and the second concerns a person being a souvenir; only the "better off trying to forget" part of the second connects up with the souvenir being of a killing, but really, you shouldn't keep people around like that for anything. Given that Scottie speaks of "souvenirs" generally, we can actually get by with taking everything he utters literally!

Why is it that recognizing a textual allusion helps us with what we should gather from some words, rather than what they mean? One simplistic thought is that an intertextual allusion is metatextual (whereas a metaphor is a first-order affair): it doesn't show up in, as it were, the words themselves, but when we recognize it it helps us frame the context of the words, by relating one text to another. The recognition realigns our approach. But it doesn't change what the words mean (one could imagine actually cases in which the presence of an allusion helps one settle an ambiguity, but whatever). Scottie's realization that he too can be called a souvenir doesn't change what "you shouldn't keep souvenirs from a killing" means, and doesn't render it a metaphor, any more than "thou art the man" renders Nathan's story a metaphor; I don't even think it has a metaphorical construal, given the generality. But Scottie and Judy's shared realization of the aptness of Scottie-as-souvenir does add an edge to the statement, and make it more of a reproach. (It's puzzling that earlier (18–19), discussing Tracy Lord's explanation of "yare" in The Philadelphia Story, Hills actually says that it's simply not a metaphor—even though Tracy is clearly thinking of both ships and herself by the time she finishes. We can even imagine that this possibility only occurs to her partway through, at which point it's too compelling to give up.) We can't even say for sure that there's a metaphorical presupposition, "Scottie is a souvenir"; since it's unarticulated, couldn't it be a simile?

10. Their words, for you.. In AT Hills speaks of the truth-conditional content of the words; in PP (28), our words (in expressing the worry that "what our words mean [being] a matter of what we inferably intend to accomplish by means of them seems to be under threat"). When Scottie's caught up short by the words he's uttered, what they suggest to him isn't anything he meant; when I suggest that Juliet is Romeo's sun also in that he should consider limiting his exposure to her, I don't take myself to be expanding on something he meant. Do I particularly care, in paraphrasing on Benjamin's death mask aphorism, what Benjamin thought he was about it making it? I do not. Neither do I particularly care, in "now is the winter of our discontent", that Shakespeare's "is" is an auxiliary verb (the main verb is "made"); the words are there, free to suggest what they will, as I am free to make of them what I will. One might nevertheless believe that even if a paraphrase of a particular metaphor doesn't get at what the maker meant, i.e., what the maker had in mind already, it still gets at what the metaphor means, and what the metaphor means already: as if we were trying to recover the living face from the mask in death. (Paraphrase captures "the meaning and content already accruing to a metaphorical utterance" (PP, 26; emphasis added. Only we can't quite state that meaning—except tautologously—we have to limn it paraphrastically.) Whence accrues this meaning? It can't be from the game to which the metaphor-maker invites one, because the invitation can't be to a game which is already onto the meaning (since it exceeds their grasp), but it also can't be from the game's free progression, because that's meaning that's made as the game goes—it isn't what already accrued. And the "already" suggests that the games, though we may play them in any number of moods, actually have a point: they ought to be uncovering this meaning. So it's odd that Hills refers to "the multiplicity of acceptable nonsynonymous paraphrases of the same metaphor" as something that "suggests that if paraphrases really gave metaphorical meanings or contents, metaphors would be much more ambiguous than we ordinarily suppose them to be" (PP, 27), because if there' a content that already accrues, that multiplicity ought to suggest that many of them are more or less wrong, n'est-ce pas? They're acceptable because the already-accrued content isn't available. If, on the other hand, the fact that we accept them is supposed to suffice by itself—if we can do without a single secret meaning they hopefully approximate and are judged against—then why talk about a meaning—akin to the literal meaning (close enough kin that the weird properties of metaphorical meaning form a first assault on popularly accepted features of literal meaning)—that already accrues to the metaphor? We talk about the meaning of a poem, but we don't think that the necessity of tact and critical intelligence for sussing it out forms a challenge to compositionality. Why think this for the meaning of a metaphor, especially when the processes by which we arrive at each seem to be, basically, the same? (Couldn't Davidson point to the actual descriptions of how metaphorical understanding proceeds and say, "yes, it's like that", while keeping the rest of his largely negative project intact?)