waste

Things Organized Neatly

[I wrote this several months back, after reading Alphabetical Diaries, but since I had gotten to that book after it had been out for a while, no venue of which I was aware (important proviso) that seemed as if it might be interested in something about it had not already published something about it, so I just let it sit around. But on the occasion of the return of The Clock to MoMA, I thought I might take it back out and put it here.]

What sort of thing is Sheila Heti's Alphabetical Diaries (henceforth AD)? What does it do? It is hardly tempting to say that it is, with its lack of any temporal or narrative structure, a diary, though its matter is diaristic. Let us ask, to start, what it is like; what may it usefully be set alongside?

We may say: it is a work of accumulation, in which the pileup of detail, the first not necessarily connected to the next, both carries one along and limns a subject to whom all these details occur, and resembles in this respect texts like Lucy Ellman's Ducks, Newburyport or David Markson's Wittgenstein's Mistress (which also ought to be set alongside each other), or his later so-called Notecard Quartet, in whose first book, This Is Not a Novel, "Writer" announces:

A novel with no intimation of story whatsoever, Writer would like to contrive. ...

Plotless. Characterless. Yet seducing the reader into turning pages nonetheless.

In all of them the line "Nonlinear. Discontinuous. Collage-like. An assemblage" occurs. Quite so.

The "collage" of AD was not quite so freely assembled as Ellman's or Markson's, because of the fairly severe constraint under which Heti composed it, both as regards what was permitted to occur in it and how those things can be ordered. We may also say: AD is an appropriative work composed of bits and pieces of another work, like the centos that rearranged individual lines of Homer or Virgil to narrate the life of Jesus. (To say nothing of the many, many examples that do not restrict themselves to a single source.) And, as the title tells us, these bits occur in alphabetical order. (A computerized alphabetical order, so that we get, for instance, the sequence "Alone in a room. Alone. Alone. Alone. Alone.", because the period (ASCII 46) comes after the space (ASCII 32). Were I, with my human prejudices, to have alphabetized these same five sentences, the first would have been last, the punctuation disregarded. Nevertheless it must be admitted that that order in the text is better, or at any rate less bathetic.) Here one may be tempted to think of the Oulipo and the extreme constraint-based work it and its affiliates have produced: Walter Abish's Alphabetical Africa, Georges Perec's Les Revenentes, or Christian Bök's Eunoia. But very few Oulipian constraints, even the mind-bendingly restrictive ones employed by someone like Doug Nufer (consider Never Again), determine the material nearly so thoroughly as Heti's does, and those that do, such as N+7, tend toward the trivial. (They are so to speak organizational procedures rather than compositional constraints.) The combination of the restricted source text with the rule of alphabetization, with Heti's contribution to the final form of the text being one of selection after the mechanical procedure has been carried out, rather than creation of material according to a constraint, is less like Oulipo than it is like, say, total serialism. (Merely requiring alphabetically-ordered sentences would be another affair and one could imagine Nufer writing an alphabetical novel from scratch.) This is, it's true, a pretty rough-and-ready sort of distinction, which may not admit of a decent precisification; the basic thought, though, is that Oulipian practices attempt to spur on the creation of fresh material, even if the range of permissible material is quite restricted. But Heti does not create any new matter in Alphabetical Diaries.

What remains to Heti is selection, or, if you like, erasure of what is not to be included, and in a textual mode, emphasizing the latter, we might put it alongside something like Jonathan Safran Foer's Tree of Codes or Tom Phillips's A Humument, though for Heti the unit of preservation is not words in whatever spatial arrangement they may have on a page but entire sentences: having alphabetized her roughly 500,000 words of sentences, Heti then cut away roughly 440,000, or 88%, of them. Her creativity is curatorial: after the choice of source text and rule has been made, this operation of selection/deletion represents the sole intervention remaining to her.

We need not, however, limit ourselves to literary works, and may place AD productively alongside Christian Marclay's collage film The Clock (henceforth TC), also an appropriative work of accumulation subject to a strict constraint. Marclay of course did not limit himself to one film as his source, but while he and his assistants had the entirety of available film as possible sources his selections were, in the event, fairly classical (that is to say, TC consists largely of well known American and European films, along with a few TV shows: Columbo and The X-Files both make appearances). He also had greater freedom in placing the scenes relative to each other: all the 12:05 scenes must precede every 12:06 scene, but they can be freely ordered within the minute.

TC is ordered chronologically, not alphabetically—something that would not be totally absurd in film; Arst Arsw does not really reward viewing in full, but it does exist, and one could imagine it being done better. Like AD it has its characteristic beats: in Heti we get clusters around possessive adjectives and pronouns; around question words; around names; around now and then, if and should. I will forms a thicket of injunction and prediction within the long I section. In Marclay all the hallmarks of the generic day are observed: breakfasting in the morning; shootouts a little past noon (Charles Bronson's Harmonica, in Once Upon a Time in the West, even observing to Henry Fonda's Frank that they're a little past the canonical time); filing out of school at three and the office at five. All things in due course, and even Hell itself awaits its appointed hour to breathe out its contagion to this world. Since in each case one knows where one is in the work's progression these landmarks can be anticipated and serve to generate tension and release. It is not that the works are not so absorbing that one pays attention to where one is, marking time until the next thing happens, the next unit begins, so much as that being absorbed in these works is, in part, being aware of the progression of time and letter.

David Velasco, writing in Artforum, called TC a "remarkably bad" piece of cinema that is nevertheless compelling as a "paracinematic—and, indeed, aesthetic—experience". I am not sure that AD is remarkably bad as literature, but it is compellingly paraliterary in the way it directs attention not even to "language" but to letters and the alphabet. To be aware that the Ps cannot last all that much longer and that Q must be coming up soon, and then to be amused that the Q chapter is all of one sentence long, is not the sign of distraction that "surely this chapter must be ending soon" would be in another sort of book: awareness of its literally literal qualities is awareness of its substance.

In both works plot and reference are lost in the magical principles of adjacency and resemblance; one experiences narrativity without narrative, in which the formal devices that create the feel of coherence are laid bare. Even without Marclay's sensitive sound editing purely visual features knit disparate scenes together: a character looks to one direction, and is met in the "reverse" shot by the answering gaze of a character in an entirely different film, and it all flows together. Someone opens a door and enters a building—cut—someone else steps forth, exiting a different one, and the logic of crossing a threshold prevails. One person looks down at another's wristwatch. What AD lacks in the aural it gains in anaphora: it is astonishingly difficult, as one reads, to remember that the pronoun "he" in one sentence does necessarily refer to the previous sentence's male name, or corefer with its "he":

Then he said, what is your name? and his eyes flickered across the alley, paved in stones. Then he spat on me. Then he went to get a cigarette and I could smell him smoking.

Who knows if the questioner, the spitter, and the smoker are the same person? "I" and "me" in two or three successive sentences presumably do all refer to the forensically identical person, but they could be years and miles apart.

I watched through the window as Vig left the garden and went into the street and started writing on his BlackBerry. I welled up with tears. I went back to my room where my clothes were.

In TC, it is generally pretty obvious when two sequences come from the same film; here, one may wonder whether even seemingly perfectly matched pairs of sentences had, originally, anything to do with each other:

Does he die, drowning, trying to kiss himself? Does he stare forever at his image?

Perhaps these were written just like this, one after the other, only to be separated by scads of "does he really"s, finally to be placed back in order when those intervening sentences were deleted. Perhaps they had, as composed, nothing to do with each other, written years apart—surely a person is allowed to think about Narcissus twice. Yet these terse little narratives go down easily, and stand out, as they wouldn't in a conventionally composed text, for their being comprehensible together. While they could have come together by the mere operation of alphabetization, their seemingly purposive adjacency is one of several reminders that while AD, like TC and like Markson's quartet, may be an "assemblage" in the mere sense of being made out of preexisting material, it is also assembled in the sense of being put together with purpose, not just thrown into an alphabetical heap. Not for nothing does it begin with six sentences starting "a book that…" ("a book that is a game" is the last); not for nothing does "I once" so niftily follow "I never" without a gap; not for nothing is the section of the "I" chapter beginning with the bare first-person pronoun set off from the rest of the text by some decorous vertical space. Heti's elimination of 88% of the text did not, one must presume, proceed haphazardly, and, just as does Markson's, her text seems to drop hints about itself:

All the elements of the world, everything I encounter and that other people encounter, can be put in a book.

Don't forget that although you aren't telling a story, you must still do what stories do, which is lead the reader through an experience.

It will not be fiction, and it will not tell a story, and there will be no characters, and you will not worry about the voice or the way it is written, just about what you are saying.

To quote Markson quoting Dizzy Dean: it ain't bragging if you can do it.

What are we to make of AD's selection, its constructedness? Catherine Lacey, writing in Bookforum, may well be right that Heti did not redact as she did "to remove the embarrassing bits", since after all "vulnerability has long been Heti's compass as a writer", but we need not think that she was engaged in Calvinistic deceptions to think that the fact of her selections is significant. (Anyway, do you know what Heti considers embarrassing, what wounding? Maggie Nelson once told a seminar that she considered it more revealingly personal to discuss her writing process face to face with them than to describe, for the benefit of strangers, anal sex in the opening of The Argonauts. The possibility that Heti is actually increasing the proportion of embarrassing or wounding bits—if that's her thing, why not go all out?—is not broached.) Here at last it seems to matter somewhat that text which serves as AD's source is Heti's own diary: there would seem to be a real life to be glimpsed, at a fair few removes, in these pages.

Velasco again on TC and its (rather more obvious) constructedness writes that "Every movie has its own temporal grammar, and Marclay typically gives us just enough of a film to reveal its particular speed or pacing … he makes salient the idiosyncrasies of movie time", the time of its individual component movies and the time of movies in general. This way of putting what Marclay is up to suggests that he's accurately conveying to us something about, say, Otto Preminger's Laura, or about time in movies—something that was anyway true—and one of the things that a rearrangement of existing material, as in TC, promises to do is make clear salient sympathies and connections that would otherwise remain hidden. There are facts about movie time that could also be understood independently of TC, but which it brings out in a compact way. But the same technique can be employed, as in Markson, to create an impression of sympathies and connections, the entries in their sequence adumbrating a consciousness which is a pure intentionality. There's no substantial Novelist of whom we might say, as we can of movie time, that we might have learned the same things about him some other way. There's just the panoply of notes, with the occasional wry comment, out of which we create a character. "And so the chorus points to a secret law", as Goethe put it, but in this case the chorus legislates by its pointing.

Wittgenstein, in his Remarks on Frazer's Golden Bough, called the kind of presentation that puts one thing alongside another to transmit an understanding of "just the connections" übersichtlich, generally translated "perspicuous" but used by him somewhat eccentrically to denote a display that enables multiple perspectives, showing us multipolar relationships and the overall shape of a thing. "It is equally possible", he writes shortly before invoking Goethe's chorus himself, "to see the data in their relation to one another and to gather them into a general picture without doing so in form of a hypothesis concerning temporal development", the way, for instance, a diary might show a person's evolution in narrative time. The decidedly atemporal organization of AD seems to achieve just such a perspicuous representation: each of the individual male figures in the book receives, naturally, his own section for his first name: Lars, Hanif, Pavel, Lemons, a sort of composite, cubist portrait; the "he" section displays from all angles The Heti Man. "I will" gives a compact representation of her resolutions and suspicions. Overall, the radical organization of the text seems to present us with Heti's sensibility in an uncommonly direct fashion.

Sort of. David Markson really existed, of course—but it's not called This Is Not a Memoir and the figures of the quartet, Novelist, Writer, Reader, Author, whose subjectivities are the hidden principles of the collage, aren't David Markson. Sheila Heti really exists, and it is called Alphabetical Diaries. But we aren't, for all that, confronted with Heti's first-order sensibility.

Diaries are already a scene of construction, memorialization, and sense-making; potentially of self-deception and self-justification. They are not naïve records of the events of a person's life, written with, as it were, an innocent pen, not because their authors are necessarily trying to distort things but because there is no innocent pen. Still, Heti's was a diary of events many of which not only Heti experienced, about which intersubjective agreement is possible, and the full text, recording ten years of her life, would, moreover, presumably enable its reader to put together more than its temporally various authors realized they were revealing. One would learn, if not about Heti and her life, at least about Heti the diarist and her perspective. This might be even more the case with an unredacted alphabetized diary: who knows what might be revealed about Heti the diarist in (one imagines) hundreds of pages beginning "I "?

We have, however, a much reduced text, Heti the redacter's version of Heti the diarist's version of Heti the person's life. (It's possible that if we had the full text to compare with the redacted version we would find it to be just more of the same, that Heti's editing is really just stripping out redundancy. But why assume that's the case?) If we try to look Heti face to face in this text we will see her through two glasses, darkly. Why so much as think that that's what's happening? AD is built out of Heti's diary, and the material is, unsurprisingly, convincingly diaristic in precisely the way that diaristic and epistolary novels aren't—full of the grain and strangeness that someone cutting a diary from whole cloth wouldn't think to invent. But it isn't a diary, nor is it even an excerpt of a diary; she leads us through an experience as if of revealing hidden connections through an objective, semantically indifferent rule of arrangement, but she has constructed the impressions that we will thereby get by selecting what is there to be connected to what. Not, indeed, to deceive us, since we could only be deceived if we thought we were learning about Heti as she lives and breathes, but to construct characters: Lars, Hanif, Claire. And the chief one, who isn't named. Call her Diarist.

I could have sworn I posted this in 2019

The fact that no one seems to be relating Ducks, Newburyport to Wittgenstein's Mistress, despite the fact that each seems to consist of linked statements or allusions to facts that gradually detail the main character's state of mind and even a narrative of sorts, the fact that speakers of English seem to have a relationship to the grammar of their language marked primarily by ignorance and awe, the fact that one sees this in for instance the fetishization of the German compound word over the English space-separated compound noun, the fact that in the beginning, sometimes I left messages in the street, the fact that the structure of long sentences in three modern novels would in fact be an interesting thing to pursue, the fact that the long sentences of Krasznahorkai are unlike the long sentences of Bernhard are unlike the long sentences of Ellmann (are unlike the long sentences of, say, James, Sebald, or Murnane), the fact that the sentences themselves but also their reception is worthy of examination, the fact that the mere length of a sentence is rarely the most interesting thing about it, presuming it's interesting at all, the fact that Markson's myriad of short sentences effects a similar pileup of fragment and detail, the fact that nevertheless the small dot, the line break, and the capital letter do seem to make a difference for the reader, the fact that repetition and length can vary independently, the fact that a long sentence (or a short one) can ease the reader along, hinder and obstruct the reader, or be something of a neutral medium through which the reader passes neither eased nor obstructed, the fact that, to risk repetition myself not merely in headwords, and to risk moreover a bit of over-cuteness, one oughtn't be so dazzled by the factivity of punctuation that one overlooks its activity, that is to say, a sentence's word count is of less interest to the critic than what the sentence does with its dimensions, the fact that Frank Sinatra was reported to have had quite a long sentence but, according to someone, I can't remember who, who read it herself, he wasn't a good writer, the fact that the name that comes to mind regarding that anecdote is Jayne Mansfield's, the fact that I have no particular interest in pursuing the question of that critic's identity further at this time, the fact that when I began to compose the list entry three before this one, I had had an idea about how to continue this sentence, the fact that now, having composed both that entry, and the one one before *this* one, that is, the one in which I took up the topic of the topic I had had in mind, I no longer recall what it was, so that I can't be confident that it was, as occurred to me when I began writing in this present vein, to concern the letter Gerald Murnane once wrote to a reviewer about that reviewer's misidentification, in Murnane's opinion, of a run-on sentence, or other grammatical chimera, as a single properly formed sentence, the fact that the sentence under dispute in that letter had been praised by the reviewer on the grounds of its length, which as I have already stated if perhaps not quite argued is a silly reason to praise a sentence, the fact that, despite this just-voiced opinion of mine, Murnane describes and reproduces the letter in an essay entitled "In Praise of the Long Sentence", the fact that you may read the essay, and the letter, yourself, the fact that by Murnane's lights, and to be sure my own as well, this present series of clauses, is not a sentence but rather simply one clause laid after another with no particular grammatical relationship between them that would serve to unify them into a single sentences, the fact that, as Facebook reminded me, On This Day a year or two ago I offered a false etymology of "congeries" purporting to derive it from "congee", explaining that congee is a porridge in which the grains of rice remain discrete, as indeed do any inclusions in the porridge, so that one may say that it is a collection of grains but not a whole, the fact that "congeries" is a delightful word, in my opinion, the fact that I feel very much as if I could go on practically forever like this, such is the fertility and capacity of ever-productive all-encompassing immodest brainpan, the fact that at the same time I do not feel as if I could go on practically forever like this, because I've gone on for a bit already and am uncertain whether this will actually amuse anyone, the fact that a nontrivial amount of my blog activity is oriented toward the prospective amusement of myself or others rather than anything more, I don't know, laudable, not to say that the amusement of oneself or others is *not* laudable, but also not neglecting the fact that it is not, perhaps, using the medium to its fullest capacity, the fact that it is also, of course, not using it in a detrimental way, the fact that not using something poorly is not really the same as using it well or grounds for praise, the fact that, ideally, this sentence would come to a graceful end rather than simply stopping, the fact that a main verb is as yet wanting—all these things seem to be worthy of consideration.

On Agnes Callard's Aspiration

I started writing something about Aspiration, a book I believe to be wholly without merit, a while back, and while normally I would just post such a thing here, it is far too long to do so (I know many things here have been quite long, but this is a new low for me). So you may find it here.

On Lydia Tár's EGOT

If "is Lydia Tár a real person" was the question about Todd Field's movie that sparked the most mirth at the end of 2022 (alongside, naturally, handwringing about The Youth's putative inability to recognize a fiction film, absent superheroics), "is the ending of Tár a dream sequence" was the question that sparked the most scorn. The idea that it is, or at least might be, was given its most thorough airing in an article by Dan Kois, who, collating evidence of strange doings, sights, and sounds throughout the film, claims that the film is "a kind of ghost story, in which we're so deeply embedded in Lydia Tár's psyche that nearly everything that appears onscreen is up for debate", and never more so than in its final third.

Plenty of folks on social media (by which I mean Twitter, by which I mean Twitter as I experience it) were on hand deride Kois and others pushing that line for taking a shallow, puzzle-box approach to a work of art, in which the point of engagement with art is finding out the objectively correct answer to a question, the question being, usually, not terribly interesting.

How curious, then, the reactions that came in in January, when a few interviews with Field appeared in which he was called on to answer questions about Tár, which he mostly, to his credit, phrased as speculation, but sometimes answered definitively, like Rowling making her little pronouncements about what her characters get up to in their off hours: to Michael Schulman for The New Yorker, whose questions were bafflingly irrelevant to the film, he said that Tár won her Tony for work with Ivo van Hove; to Kate Aurthur for Variety, which ought to know better, he said that Tár actually never met, much less studied under, Leonard Bernstein. To these revelations there was no such derisive response; on the contrary, at least a few people were delighted to have their suspicions about Bernstein "confirmed". But—what attitude toward the film as a work of art lies in the impulse to ask such questions, or treat Field as in a position to confirm any such thing? For that matter, what attitude is it, really, that takes a work of art as a puzzle?

Certainly, there are works of art which contain puzzling elements, and getting straight about them is an unobjectionable, if perhaps not always necessary, part of appreciating them. It is not for nothing that Perec's preface to Life a User's Manual discusses the skillful creation of jigsaw puzzles, the imaginary dialogue between puzzle-maker and puzzle-solver involved in deciding what gets cut, or fit, where. We can regard Primer, for instance, as in part an exercise in constructing an extremely minimal syuzhet from which a much more involved and ramified fabula can be extrapolated; that those who attempt to do so arrive at substantially the same series of undepicted events is a testament to the skill with which it's put together. We confront something like a puzzle whenever we begin to suspect that a narrator is not reliable; the suspicion is an invitation to figure out how far and in what respect the narrator ought to be mistrusted: what's being concealed, what misrepresented? At times the text's own invitation to treat itself as offering up a mystery to be solved may become quite overt; think of some of Gene Wolfe's stories, perhaps "Seven American Nights". Such an invitation may always be declined. But taking it up needn't be "treating an artwork like a puzzle" in a derogatory sense; it could be the prelude to a deeper appreciation. If it seems too cold-bloodedly rational—if you prefer, perhaps, to let the feel of the work wash over you—recall that it was the arch-Romantic Schlegel who not only reminded us that a text teaches its readers how it is to be read, but also mocked his contemporaries thus: "If some mystical art lovers who regard every criticism as a dissection and every dissection as a destruction of pleasure were to think logically, then 'wow' would be the best criticism of the worthiest work of art".

Regarding an artwork as a puzzle in the derogatory sense can't simply mean, for instance, noticing, and trying to interpret, the apparitions in some shots in Tár. It means something more like treating the work not merely as being, in some respects, puzzling, and trying to make coherent sense of those aspects in light of the whole, but as being exhausted by those aspects that present a puzzle, and approaching it solely in those terms: it has a solution and the point is to find it; once you've found it, you're done and can move on. Once the jigsaw puzzle has been finished, who pauses to appreciate, in any sense, the image they've assembled? That's not the point. The point was to figure it out. It becomes mysterious, with this attitude, why anyone should hold art or particular works of art in any particularly high esteem, why the encounter with art is so important in anyone's life, how it could offer intellectual and emotional satisfactions beyond those (which are real!) of finding the solution.

Kois actually made sure to say explicitly that a work of art is not a puzzle box, but since he doesn't really do anything with the result of his analysis, one can forgive his accusers somewhat for ignoring that declaration. (A much superior piece in Collider by Martin Millman was concerned with many of the same aspects of the film as Kois, but used its observations to make thematic points.) Having put the clues together that, he thinks, show the unreality of the concluding sections of the film, he ends the piece. What are we to make of this newly assembled picture of Tár? Who knows—that doesn't seem to have been the point. But what Kois is up to, the relation that the puzzle-solving attitude bears to the artwork-as-puzzle, is still preferable to what Schulman was up to in The New Yorker.

A well made puzzle might be extraordinarily difficult, or fairly straightforward. It might require obscure or specialized knowledge; it might require very little beyond what the audience of potential solvers to which it addresses itself can be presumed to have. If well made, though, it should be self-sufficient, first, in the sense that it should contain within itself what is necessary to solve it. This will naturally be relative its audience; the more difficult linguistics olympiad puzzles are apt to stump those with no linguistics training, and if you've never heard of an épée some lazy crossword clues may simply be beyond you. These are, if you like, hermetic (as Ted Cohen called certain jokes hermetic), but nothing in them is secret; nothing is hidden. You may not know the thing that solves it, but you could have known it. You may need to appeal to the creator for hints, but if the puzzle is really well constructed it shouldn't be necessary for everyone to do so, in order to make progress; it should be tractable by at least some solvers. It should be self-sufficient in the further sense that if you have solved it, you should be able to recognize it as solved, to defend and explain your solution to the satisfaction of another. Obviously, sometimes this task is fairly trivial: these previously entangled rings are now separate; what more do you want? But that's not always the case. There really is a puzzle about what's going on in "Seven American Nights" and one can't simply present one's answer and expect it to be accepted without showing how it works—how it is a solution that fits satisfactorily with the whole. One may think that there has to be a single answer to a puzzle, but there's no reason to insist on that. Crossword constructors occasionally make puzzles whose single right answer will be determined by an event that hadn't yet taken place when the puzzle was published; the outcome of an election or sports game, say. Such a puzzle could easily be made to enjoy multiple solutions by changing the schematic clue "victor in such-and-such a contest" to "contender in such-and-such a contest". This would be a neat if perhaps not very interesting trick. But it is useful to cancel the idea that interest in puzzles implies an aversion to multiplicity. The puzzling events of "Seven American Nights", or for that matter of Tár, may admit of multiple interpretations without that meaning one can't regard them as well made puzzles. One wouldn't be able to tell that someone else was wrong just because they had a different answer, but one would expect that other person, as well as oneself, to show how each answer is an answer, how it fits together with the information given and how it answers whatever one takes the question to be. The essential point is that in taking something to be or pose a puzzle, one takes the process of finding and justifying a solution as something it is in one's power, at least in principle, to do. If the solver must appeal to the author to find out if they've answered it correctly, or must appeal for a hint because the puzzle is simply missing information, it's at best flawed. Gollum was right to object to "what have I got in my pocket": that's no riddle.

Interest in a puzzle, or something treated as a puzzle, is interest in something public. Some people may be better positioned to make a go of it, but in principle anyone else could be in the same position; no one has a special kind of relationship to the puzzle that makes them more authoritative than anyone else. Even creators of puzzles are well positioned only by knowing the same kind of thing that successful solvers will know; they're familiar with these things because they made the puzzle what it is. But their knowledge is of the puzzle, and they can even learn things about it they didn't already know; there may be other routes to a solution than they had envisaged. (Think of the Gordian Knot.) Moreover, whatever hints they give or announcements about the solution they may make are answerable to the puzzle; if the creator announces later "the solution is such-and-such", that had better actually fit with the starting points and declared goal of the puzzle. If they say "here's a hint: …", that hint had better be such as to help someone make progress. If, that is, the puzzle is "what actually happens at the end of Tár, and in particular, is it all a dream of Lydia Tár's?", and Todd Field himself announces that it is, then a fair rejoinder is "on what basis might anyone be able to tell that?"; if he says "you can tell from careful attention to a, b, and c", then a, b, and c had better actually support his contention, in the first place, and it's open, in the second, for anyone else to say "when you consider x, y, and z as well, this other interpretation of a, b, and c becomes compelling, which points away from 'it was a dream' solution toward this other one". There may be another solution, for all Field's goals when (as we are pretending) he constructed the puzzle. Whatever else may come with regarding a work of art as a puzzle to be solved, it at least takes the work, or some aspect of it, as the object of interest, and it takes the audience as able to do the solving.

What sort of attitude does someone have who asks whether Lydia Tár really knew Bernstein, or how she won her Tony? Michael Schulman, who wrote the New Yorker piece about her EGOT, referred to his confirmed-by-Field conjecture as "CANON", in all caps, on Twitter, suggesting that the basic orientation emerges from fandom, and it does seem to have a certain fannish deference and passivity. (Of course fandom has its own forms of activity, but with respect to what really happened, what's canonical, it shows a great deal of deference to whoever seems to be running the show.) But how does Field come to be in a position to confirm that Lydia Tár worked with Ivo van Hove? Nothing of the sort is suggested in the movie. If Tár were a real person, we could imagine that Field is just better acquainted with her than we are, but, alas, she isn't a real person, and there would seem to be no more to know about her than one can learn from the film, where she has the only existence she does have. If Field has an authority here any attentive viewer does not, then what the artwork is cannot simply be the film anyone can view attentively; Lydia Tár's existence must extend beyond the film. How else could Field know more about her than we do?

As with puzzles, Field could simply be more reliable as a guide to what's depicted in the film, both in the sense of what can be gleaned from it and in the sense of what's on screen to notice in the first place, than the average viewer, in virtue of his more intimate acquaintance with it. Even professional reviewers, who might be supposed to be good at watching movies and getting their gist, get basic facts of plot wrong surprisingly often; maybe it's better to put our trust in someone who's got a firmer grasp on the whole affair. But that isn't why someone asks Field whether Tár really knew Bernstein, or worked with van Hove. Those questions presuppose that Field has some special, esoteric knowledge, knowledge that we just can't get except by asking. After all, didn't he invent Lydia Tár, and everything we do see? (Ignoring for the moment the fact that movies are not created by just one person.) Why not suppose that he also invented a few more things, things that just didn't make it on screen, for one reason or another? It hardly strains the imagination. Perhaps he charted her career in great detail. Tár exists onscreen, yes, but also in Field's imagination, pages of notes, a private archive, whatever. We all see what's on the screen, which is public, but that only gives part of the story—whatever Field imagined in service of, or along the way to, creating the film also counts. And only he knows those things.

If you think that Field is uniquely authoritative and can confirm speculation about things that are only suggested, or not even suggested, in the course of the film, then the film, the only thing the public has full access to, is incomplete, eked out by something inaccessible to the rest of us. The epistemological deference we grant Field has, alas, metaphysical import; sitting in the theater, we audience members are even more like the wretches in Plato's cave than the projectionist's art suggests. It's not that understanding the film requires it to be supplemented by further things that are in principle also accessible to the public, even if not actually known, as in the case of the hermetic puzzle; that which completes it, that Field uniquely has knowledge of, are Field's own private thoughts and fancies, of which the thing we see is but the imperfect, incomplete manifestation. An artist who thinks of their work along these lines has in a sense never published it—they wish to retain a proprietary interest in it, keep it on a leash, rather than sending it on its way to become what it will be in the eyes of its public. Such an artist asserts the primacy of what they wanted to do over what they actually did. If the audience of a work thinks of it along these lines, it abdicates its role as appreciator and interpreter, wishing not only that there should be an answer, but someone who can give it the answer, and in fact setting things up so that someone has to give it the answer. For critical appreciation and assessment as an independent activity is all but impossible.

That conclusion may seem melodramatic when faced with trivial examples—guessing games about awards of no moment at all in the film. (Whether Tár was a fabulist all along about Bernstein is of greater interest.) There we have something firmly in "what have I got in my pockets" territory, but which is, at least, beside the point; potentially playful in the same way "who would win in a fight, Batman or the Predator?" is. (Which—this is admittedly petty—makes it all the more unseemly that Schulman chose to use his time with Field to pursue those questions.)

But the deference shown to the creators of works goes beyond the trivial and extraneous. It is easy, for instance, to find defenses against the obvious antisemitic reading of They Live claiming not that the reading is the crude, unilluminating application of a template whereby any tale of a sufficiently large conspiracy deceiving and exploiting the masses is congruent to any other, but that it's incorrect because John Carpenter says it's not what he meant—an act of bald ipsedixitry unproductive of critical insight, in addition to being somewhat flimsy if one thinks the availability of an antisemitic reading is important: "he may not have thought of it before", as Stanley Cavell wrote in a similar context, "but he had better think of it now." (I should add that I don't know in what spirit Carpenter himself has said these things.) Our knowledge of the film that we can watch contains the same potential for interpretation; what's changed is our knowledge of the film-cum–Carpenter's mind, and we are asked to disregard our lying eyes in favor of Carpenter's assurances.

Granted, we will not accept just anything from the creator of a work—"actually, this was funny" will not fly. (Or so one hopes. There do exist statements by artists about their work that amount to an announcement that their joke made you laugh, actually.) And perhaps claims that such and such a reading is, or is not, definitively correct, would also not fly, if they seemed to go against too much of what anyone can see in the public face of the work. But even regarding the creator as having the power to settle open controversies, or add, post-publication, to the store of facts, has a deep impact on how we can relate to an artwork.

My point is not that we've known since Barthes, or Foucault, or Cavell, or Nehamas, or or the New Critics, or the Romantics, or whatever style of critical autonomy you like, that this sort of attitude is wrong. (Perhaps, to pick up on its apparent fannish sympathies, it has reasonable application in the case of long-running serialized works with a master plan.) My point is that it's stultifying, and stultifying in a way that looking at a work of art as a puzzle is not. It makes our interest in a work of art into a sort of pretend appreciation. No matter how searching or sophisticated a reading, how compelling, productive, or interesting it is to others, it is always in principle hostage to someone else, the author or whoever it may be, coming along and saying, "no, that's not right"—and there's an end on it, because they know. It's disappointing if someone regards a work of art as a puzzle to be solved, because such an interest in art seems trivializing. It seems incompatible with the significance many find in their engagement with art in reducing such engagement to the finding of clever solutions, after which the work can be set aside. But at least such a person thinks that they can find the solutions. They have not mystified the nature of the artwork, or made it someone's private property. Someone who thinks Todd Field, uniquely, can say whether Lydia Tár really studied with Leonard Bernstein still thinks there's a single answer to be had to that question, but they don't even think they can find it out.

POST SCRIPT

As it happens, while I was sitting around deciding what to do with this here piece of writing you've nearly finished, Aaron Bady, on Twitter, posted a link to a paper (in some kind of draft state) on Teju Cole's Open City. It's good and I recommend it. I mention it here because, in it, Bady makes use of Cole's personal statements, about the novel, about its reception (praising Alyssa Rosenberg's review), and about rape culture in general (ie not in connection with his novel), and I couldn't help but feel, as I read the pertinent sections, as if the apparently most important uses were in fact the most extraneous. (Partly this is done in service of arguing for against the identification of Cole and Julius, the novel's narrator, in which some critics, eg James Woods, indulged, and I have no objection to that, but that aim can be (and is!) prosecuted by producing a reading of Julius's character in the novel, and contrasting it with the real-life Cole, and setting them aside each other, not by taking Cole as privileged insight into the novel. Another citation of Cole has him suggesting a "plausible" (NB!) way of thinking about the novel as "a series of visits to [Julius's] psychiatrist", a suggestion redeemed by its fruitfulness, not its source.) This apparent extraneousness is especially interesting since Bady not only quotes Cole but also casts in a negative light the "New Critical impulse to avoid the 'intentional fallacy' [and] reluctance to give Teju Cole's opinion of his novel any interpretive weight", and connects that impulse to the impulse that led critics to essentially (and astoundingly) ignore even the possibility that Julius is a rapist.

The specific opinion most important statement of Cole's is that Julius did, in fact, rape Moji, as Moji asserts, an assertion that Julius more or less (as I remember it, anyway) ignores. One reason to be a little cautious here—to embrace the most anodyne possible version of an "intentional fallacy", the reminder that what one intends to do is not always what one does in the end—is apparent in the part of the interview Bady quotes: the book would end with "three vicious thwacks of the hammer, and then a soft exit to strings. I'm attracted, in art, to things that trouble the complacency of the viewer or reader", Cole says. But—manifestly, since otherwise there would be no need for Bady's paper—the complacency of many reviewers, at least, was not troubled. Bady remarks that Rosenberg hits a false note when she says that the ending of the novel "forces" us to reevaluate Julius. Cole, too, seems to have hit a false note; he may have intended to trouble our complacency, but this intention seems not to have been brought off. The hammer thwacks weren't vicious enough; maybe the hammer wasn't even really there. But this is after all an intention about the effect the ending should have, not an intention about what takes place within the world of the narrative, however implicit it may be; perhaps that makes the difference.

Suppose that Cole had said, of Moji's accusation, not "it's absolutely true" but "no, it's false" or "It's inconclusive on purpose" or nothing, declining to address the issue. And then suppose that we read Bady's paper, starting from the part beginning "in the remainder of this essay", wherein he develops the reading that "Julius works to forget [his rape of Moji], through his narrative, in ways which can be (and must be) read back into the novel". Would that reading be less convincing? It doesn't depend on taking Cole at his word regarding Moji's assertion. Indeed, although Bady says that "the novel's frustratingly limited first-person narration does not allow us to answer with any confidence" the question whether Julius raped Moji, in my eyes his reading allows us to conclude that the narration does allow us to answer the question with confidence in the affirmative: it is an argument that the affirmative answer, read back into the novel from the beginning, is present in the novel from the beginning, that while Julius does not narrate his consciousness of his rape as such, it nevertheless pervades and shapes his actions and his narration: "the reading is present in the novel, a pattern of references and associations which organize the stream of Julius's consciousness." (What, one has to wonder, do the critics who think it up in the air whether Julius raped Moji or not, or those, if there any, who think that he simply did not, think the episode is doing in the novel? Is it a prank? A warning—look, these accusations can come even for the learnèd flaneur?) Such a reading also reveals the book as better—more artfully and cunningly constructed—than it would otherwise seem; this is a dimension of its attraction.

It is ironic that Bady, in (aptly) describing his reading as "felicitous", ie, as one which "simply works or does not", contrasts that view of readings with a view of fiction as "a constative utterance", "a set of propositions about an objective reality which can either be confirmed or denied", after having previously cast suspicion on the "New Critical … reluctance to give Teju Cole's opinion of his novel any interpretive weight". Absent a compelling, felicitous reading, Teju Cole's opinion of his novel could only play the role of confirming or denying a supposition about what has happened within it, precisely as if it were an objective reality of which he has knowledge. But in the presence of such a reading, what need have we for his opinion in particular? It's not that there's a reason not to give Cole's opinion any weight: it deserves all the weight it can support. But he supports it the way we all do.

Gefahr im Ziel

Like, I suppose, most of The Drift's readers, I stand ever at the ready to believe that whatever Obama's been up to recently is an expression of Pure Ideology. When you've got an easy target, though, your work has to be correspondingly more rigorous; if you're criticizing as wrong someone your audience is already apt to think is wrong, you risk appearing self-satisfied or ungenerous. If you yield to the temptation of being breezy, pointing out this or that (and merely pointing it out) as if as a reminder to a reader who's already in the swim, you risk appearing supercilious where you should be incisive.

[Kristine McDvitt Tompkins and Douglas Tompkins]'s donation vastly expanded Chile’s national parklands, according to The New York Times, “enlarging the area of protection for pumas, condors, flamingos, and endangered deer species.” But nearby farmers have organized a resistance to the rewilding project. They complain that the rise in puma populations is threatening their herds, which they depend upon for their livelihood. Our Great National Parks has a short attention span, however: there’s no time to provide this context before moving on to glamor shots of an Andean condor chick.

Alas, "The Audacity of Nature Docs" has an even shorter attention span, not even pausing the explain how this is context. I am in no position to dispute that it's a fact, and even a fact that has something to do with the park in Chile. But a fact is not an analysis; an accumulation of facts is not context. The fact is presented as if its significance is obvious on its face (it's bad, I guess; a strike against the park), so that we, the knowing ones, have only to be told that the documentary doesn't include it to tell that we've got one up on it, but what is its significance? Who are the nearby farmers? What should we make of their claims? And why weren't there enough pumas around to trouble them before?

When Yellowstone was created, when, as Adams reminds us, it was inhabited by "several indigenous tribes that served as … effective caretakers of the land", wolves lived there too. Then for a while they didn't, or at least did only in negligible numbers: settlers—farmers who complained that the wolves were threatening their herds—were among those whose influence led to a pretty successful program of extermination. They're back now, though, and get this: farmers ("ranchers" might be more apt, but I'm following Adams's usage) complain about them. What should we make of their complaints?* (To what extent does their existence in their present form depend on the absence of wolves in the intervening years?) I personally do not think that it's a mark against the wolf programs; I am rather inclined to see them as an obstacle. Should I think the same of the Chilean farmers/ranchers? (It's also not clear that wolves were reintroduced with the aim of restoring Yellowstone to a "mythical wilderness [from] a time when humans did not exist", an image  that may well be the documentary's ideal of a national park but is not exactly current in conservation, rather than restoring an endangered predator to an ecosystem from which it had been removed, which is not quite the same thing.)

I have no idea how far  an analogy between the complainants about wolves and the complainants about pumas can be carried; I haven't got the context. But the outlines seem suggestive. We might ask what the wolf population in the American West would have been like if the indigenous peoples living in Yellowstone hadn't had the treaties they violated. (It's quite possible that that wouldn't have been enough, but who knows.) We do know that Native Americans participated in the reintroduction of wolves and have the authority to manage wolf populations on the lands they currently occupy. Adams closes by asking, "what would happen if, rather than kicking people off their land to convert it into more tourist destinations and reserves for future land exploitation, the conservation industry followed the lead of people who are already practicing preservation?" Well might one ask! Would there be pumas?

*What, for that matter, should we make of their occupation? These are cattlemen, after all.

I doff my assless chaps one leg at a time just like the next intermillennial yobbo

Picking up Saint Sebastian's Abyss, I knew that it contained passages in a style highly imitative of Thomas Bernhard's, a distinctive style which I had recognized immediately on seeing an excerpt from the book, but I did not know that the book was written exclusively in that style, though perhaps to a lesser degree in its closing chapters, unless as I read, I thought, I had simply gotten used to it by then, and as I read, I thought increasingly that this stylistic fidelity came with a certain detriment to the book, which, had it appeared in a world in which Bernhard had never existed or written, or in which he had written only theater pieces, and vitriolic and yet simultaneously friendly letters to his editor, and excoriations of Austria, and Vienna, and Viennese society, but had not written his novels, his immediately recognizable ranting monologues of eccentrics returning over and over to themes and phrases and obsessions, would be something absolutely first-rate, but which could in this world, as I read, I thought, in which Bernhard had existed, and had not contented himself with theater pieces, and testy exchanges with his editor, and condemnations of Austria, and Vienna, and Viennese society, but had also written the novels on which his fame rests, at least in the English-speaking world, only be seen as a kind of Bernhard lite, as something which continually invites comparison to Bernhard but which cannot bear up under those comparisons which it brings upon itself, specifically it invites comparison to The Loser, a novel about two friends one of whom is traveling in the days surrounding the other's death, the other an Austrian who has a lung condition, friends who have left practice for theory whose encounter with a fictional or at least fictionalized artist of surpassing ability derails their lives and fractures their friendship, and this comparison is something unbearable for the novel, as I read, I thought, for the personages in the novel are flimsy, which is not to say that Bernhard trafficked in the fully fleshed out character of literary psychological realism, he didn't, his obsessive, culture-mad personages were caricatures, but this does not save Saint Sebastian's Abyss, for its personages are then caricatures of caricatures, especially the narrator, and especially Schmidt, whose comical European-æsthete snobbery is too familiar to bite, whereas Bernhard's personages are weird, and Bernhard's personages, too, dwell in a familiar world, or rather, since Bernhard's personages for the most part do not interact with any world familiar to me, they dwell in their world in a way which seems compatible with the world familiar to me, even their obsessions are reconcilable with the world familiar to me, for instance Glenn Gould, the most important piano virtuoso of the century, as the narrator of The Loser said, I thought, even though he is not accurately represented in The Loser, we can understand his being the s u b j e c t of an obsession, and the c o n t e n t of their obsession are such as we can comprehend, whereas Saint Sebastian's Abyss, the greatest painting of all time, as the narrator and Schmidt both said, I thought, is so sketchily described that their obsession is not so comprehensible, but that at least must be forgiven, because only failure could attend the author who attempted to describe plausibly a painting that the reader could understand a personage being obsessed with as the greatest of all time, but also the c o n t e n t of their obsession, the books they write, seem to be the work of cranks, and but the novel presents them as not cranks, which is another way that Bernhard's personages are and these personages are not compatible with the world familiar to me, because Bernhard's personages most likely are cranks but they are supported by family wealth which exempts them from having to be anything other than cranks, being cranks is permissible to them, and perhaps even to Schmidt, whereas the personages of Saint Sebastian's Abyss are both scholarly and popular successes, the books of the narrator have made him wealthy, and presumably the books of Schmidt have made him wealthy, and they both possess acolytes and followers, and indulgent editors at presses, and the ability to place articles with ease, they are discussed on television, are celebrities, in other words, on the strength of their scholarly writings on Saint Sebastian's Abyss, the painting, and the two other paintings by Count Hugo Beckenbauer, and no other paintings, and one result of this is that the world of the novel seems strange to me, ich kann mich in diesem Buch nicht finden, and the satirical suggestion that their obsession is, well, kind of silly and crankish lands less well, because they are making a go of it, after all, and perhaps history will strip them naked as it did Saint-Saëns, as indeed Schmidt does to the narrator, but that episode cannot prove that they were silly or crankish, because it still occurs on their own terms, whereas the books of Bernhard's cranks do not make them famous, bring them no riches and no scholarly esteem, in fact the chief attribute that attach to the works of the personages in Bernhard's novels is that they are not finished, whereas in Saint Sebastian's Abyss they are almost all finished and are plentiful, and this too makes their being obsessive cranks less consistent with the world as it is familiar to me, which is not to say that I do not think that there are no successful cranks or for that matter frauds in this world, because I do think that, but we have nothing really to judge the personages of Saint Sebastian's Abyss by, because their works are necessarily merely sketched, except perhaps their utter sincerity about art, but I see no reason to call that anything worse than unworldly, and the general suspicion that so many books on just one topic must be cranky, and certainly the narrator's book speculating about what Beckenbauer's lost paintings may have been like seems unscholarly in topic, but what of it, shall just that one book condemn him, and so the idea as expressed in a blurb that the book "flays art of its pieties of and pretensions" does not land, because the book takes place in a pretense world, unless the blurber simply means that the personage of Beckenbauer is dissolute, but what of it, Caravaggio was a murderer and not even fictional, so why would a sex-addled painter who is fictional be any worse?

(And what, for the love of god, did that other blurber see in it, or in Krasznahorkai, to connect them?)

Two or three confusions about five or six pages of Bernard Williams

This is just an attempt to work through some questions I had on reading Ethics and the Limits of Philosophy, pp 143–148 and thereabouts, about "ethical knowledge" and its possibility in a fictive hypertraditional society (its salient tradition being its lack of reflection), in particular concerning the way that the "possibility[] of the insightful but not totally identified observer[] bears on an important question, whether those who properly apply ethical concepts of this kind [ie, thick ones] can be said to have ethical knowledge" (142)—how the observer is implicated in there being ethical knowledge in such a society, and how reflection is involved, and how reflection is said to destroy ethical knowledge. (When I say "work through some questions" I mean something like "express the questions somewhat combatively".) The last, in particular, now seems much stranger than I recall having found it on my first reading of the book, long ago; one might easily have thought both beforehand and, taking Williams at his word, afterward, that "reflection might destroy knowledge[] because thick ethical concepts that were used in a less reflective state might be driven from use by reflection, while the more abstract and general ethical thoughts that would probably take their place would not satisfy the conditions of propositional knowledge" (167). I say "taking Williams at his word" because he introduces that sentence with the phrase "earlier I said". But earlier he did not say that; what he said earlier was much weirder. (Or more carefully: earlier he said that, but disconnected from the context in which he says it, which requires him to have said something much weirder.)

But more on that anon; first, the argument that there's ethical knowledge in the first place in such a society (the one topic leads directly into the other, anyway). Broadly, it seems to be like this: First, those who have knowledge exhibit certain features: they believe the judgments they make, and the judgments are true, and the truth isn't happenstance but rather their practices track the truth. This is presumably meant to be a sufficient set of conditions on being knowledge(able). Second, practitioners of thick-ethical-concept discourses believe the judgments they make in using those concepts, and their use of the concepts varies as the circumstances vary. Lemma: If the judgments they make using those concepts are true, then they have, and the judgments constitute, knowledge. Third, one reason to think that the judgments are not true fails, and another reason fails if the people being talked about are sufficiently unreflective about what they're doing. Conclusion: truth and knowledge, at least in certain scenarios.

First, this is not exactly a constructive argument. That is: if you were concerned about how these practices might involve knowledge, or in what sense the practitioners can be said to know that so and so is such and such, you wouldn't really be enlightened here. (How could you be? We're talking about people whose practices you're on the outside of, and Williams talks only about "the boy is F", so he couldn't really try to get you inside.) When an argument by elimination succeeds, the result is seen to be unavoidable, but not necessarily in an illuminating way: it couldn't be these other things, so it must be this remaining thing. (That isn't necessarily a knock on it, though I personally did find it to be, as can sometimes be the case with arguments by elimination, a trifle trickily unsatisfying. Nor is it necessarily to say that a different kind of argument in this kind of case might even be possible; I'm willing to believe that in more philosophical cases than one would like to acknowledge you just sort of have to limn the thing from the outside.)

Second, if you're making an argument by elimination, you really have to make a case that you're eliminating all possibilities but those you wish to establish. Williams doesn't make such a case, and I don't think he is eliminating what's necessary. (I'm also not convinced he should get to claim the "tracks the truth" requirement; the brief argument there is simply that if any of their talk with their thick concept is true, then the fact that they vary their talk with varying circumstances, and correct each other, etc., shows that they track the truth. I suppose there's some consideration like: we only know what the content of these terms even is from the various ways they employ them, so what they do must come out to be tracking the truth, if there's truth. But it seems so easily won!.)

Williams confronts directly two possibilities for why the claims may not be truthful. One of them is this: if the observer is able to say "so and so's statement 'X is F' is true", then the observer ought to be able to just say "X is F" directly, but, Williams says, "he is not prepared to do that, since F is not one of his concepts" (143). What he seems to mean by this is that "F" is not one of his words: he explicates the disquotational principle involved with reference to slang terms for entities of a school, and the square teacher who understands but does not use the slang surely has concepts involved. To be somewhat more concrete: I, not being of the youth, or part of Twitch culture, etc., would not say that something is "based" rather than "cringe", or "poggers". But I might judge (more readily in the former than the latter case) that someone else was speaking correctly or incorrectly in using them. This discrepancy doesn't seem harmful to the possibility that things might truly be based (bzw. cringe), because what holds me back isn't their not being "my concepts" in the sense of my not knowing my way about them (though I might judge that I know my way about them to the extent of judging them correctly when used but not using them correctly myself in novel situations, as when a person has enough of a language to assess a translation but not enough to speak it fluently), but simply their not being my words to use; they belong to another population and I would feel or look silly speaking them myself. (Williams actually gives the disquotation principle as "A cannot correctly say that B speaks truly in uttering S unless A could also say something tantamount to S" (143; bold added), and that "tantamount" seems as if it actually gives one an out completely.)

It's an reasonable question, I think, why Williams is confident that the observer would be willing to quote the term in question. Most actually existing people do not observe the analytic philosopher's niceties regarding use and mention of terms quite as rigorously as the philosophers do, especially around fraught terms. If F is not the observer's concept in a strong way, then his ability to judge about its use in another's statement is called into question; if it's not his term in a strong sense, then he might reasonably be skittish in uttering it even within a quotation. But this is more to the point with regard to the possibility I think Williams is leaving unaddressed. Another interesting question is why Williams subsequently characterizes the issue here as one in which "their notions [are] so different from the observer's that he could not assert what they asserted" (145), which really does seem to put this in territory different from the slang example and raise the question of how the observer is judging the truth of the quoted statements afresh.

The second argument that claims involving the terms in question are not true that Williams considers is false, "not because they can be mistaken in ways that the locals themselves could recognize, but because an entire segment of the local discourse may be seen from outside as involving a mistake" (145). One thinks, perhaps, witch discourse, or of "magic" more generally, of which Williams says:

"magic, at least, is a causal conception, with implications that overlap with scientific conceptions of causality. To the extent this is so, magical conceptions can be seen from the outside as false, and then no one will have known to be true any statement claiming magical influence … the problem is that their statements [] imply notions similar enough to some of [the observer's] for him to deny what they assert. (145)

That is, the system of scientific concepts the observer can find in magic-talk overlap sufficiently with the scientific concepts with which the observer is familiar that the observer can relate the two and condemn the former as at root mistaken.

In the ethical as opposed to theoretical context, Williams puts it thus:

the locals' statements imply something that can be put in the observer's terms and is rejected by him: that it is right, or all right, to do things he thinks it is not right, or all right, to do. Prescriptivism sees things in this way. The local statements entail, together with their descriptive content, an all-purpose ought. We have rejected the descriptive half of that analysis—is there any reason to accept the other half? (145–6)

It is in answering that question that the issue of zermalmende Reflexion arises, but before we get there—isn't there something odd about this? If the observer finds the claim to be false, must it be because of such a prescriptivist split? Compare this example from the next chapter:

[M]embers of a culture that does not admit human sacrifice encounter members of another that does. They conceptualize differently the ritual killings, but this does not mean that the first group, if horrified, are laboring under an anthropological misunderstanding. It is, as they might put it, a deliberate killing of a captive, which is enough for their ethically hostile sentiments to extend to it. (158)

Indeed, they might be laboring under a crystalline anthropological understanding (mightn't they?) in which they very well understand what the other folks are getting at when they talk about the sacrifices but think it's all just erroneous. Or—and one really must wonder whence comes the temptation to set all these encounters in imagined ethnographic encounters with or between "primitive" peoples—consider something like the following. (And here I confess that it's not with the greatest comfort that I'll use (or even mention) the term that follows, but am doing so because I think it's more useful to have a concrete example than to talk of a "headman" saying "F", and because I don't think that endless circumlocutions would really have helped.) I believe that I am, and that most people I know are, capable of judging the aptness, in use, by the standards of those who use it full-throatedly, of terms such as "slut". I wouldn't use it myself, and I wouldn't even be thrilled to quote its use and assess the quotation, but I think I know what people are about when they do use it, and I would and do reject its application in general, as involving a whole complex of patriarchal nonsense which I also wish to reject. It certainly seems as if this rejection involves my deeming any particular application of it wrong, not because it's misapplied but because I think it can't be correctly applied. I don't believe that in adopting such an attitude I must adopt the prescriptivist analysis; certainly, since the basis of the rejection is another set of "thick" values set against patriarchy, it doesn't seem that it boils down to a description plus a thin "universal moral notion" (146). (One can also ask how it is that I come to be able to judge regarding uses of the term without inhabiting the system of values of which it is a native. But since I do do so—or allege that I do—if the only account thereof is the prescriptivist one, well, so much the worse for Williams.) But Williams only considers that one might call the application of the rejected thick terms systematically wrong from a thin, rather than a competing thick, position. So it isn't clear to me that the argument by elimination covers the needed ground.

My first thought about the observer unwilling to say in his own person "X is F" was that it would be this kind of case—unwilling to say it because to use of F is to take part in an odious-to-the-observer scheme. But it can't be that—the rest of the discussion wouldn't add up, and anyway, the observer would be unlikely to mention it in such a blasé fashion or to judge the sentence using it true. But I don't see how this kind of rejection of the thick concept maps on to the division into a pure description plus a thin "right" or "wrong".

It's about this that Williams does bring in the role of reflection:

The basic question is how we are to understand the relations between practice and reflection … in relation to this society, the question now is: Does the practice of the society, in particular the judgments that members of the society make, imply answers to reflective questions about that practice, questions they have never raised? … There are two different ways in which we can see the activities of the hypertraditional society. … One of them may be called an "objectivist" model. According ot this, we shall see [them] as trying … to find out the truth about values … We shall then see their judgments as having these general implications [as we see claims about magic as having implications about cause more generally]. On the other model we shall see their judgments as part of their way of living, a cultural artifact they have come to inhabit … we shall take a different view of the relations between that practice and critical reflection. We shall not be disposed to see the level of reflection as implicitly already there. (147)

Note that the judgmental observer of just a minute ago doesn't need to care about this; that observer can seemingly say "well yes they aren't trying to get at abstract truths about patriarchy (or whatever) but their practices evince patriarchy and they are incorrect in their ground-floor statements because of that".

Why is this relevant? Well, consider the practitioners of magic; they seem, or have seemed to many ethnologists, to be engaging in causal reasoning, but with faulty premises, but because they share a topic with what we know, we are in a position to deny what they assert (as Williams had said, p 145). (Must the practitioners of magic be reflective about it, or is it enough that we read off causal implications? It seems as if it must be the former, given the distinction between the objectivist and non-objectivist readings of practices, but, in this domain at least, why?) They're up to something similar enough to what the observer knows for the observer to get his conceptual-schematic hooks in, is, it seems, a very crude way to take it. In the ethical case: "if we take the nonobjectivist view … various members of the society will have knowledge … But on the objectivist view they do not have knowledge, or at least it is most unlikely that they do, since their judgments have extensive implications" (148) which, one expects Williams to continue, impinge on a domain of the observer's knowledge, which enable the observer to deny what they assert (or, as one could more simply put it, which are false). This would parallel the discussion at the introduction of this possibility for denying knowledge: "magical conceptions can be seen from the outside as false … the local criteria do not reach to everything that is involved in such claims" (145), and so on. But that isn't how Williams goes on; he goes on to say "… which they have never considered, at a reflective level, and we have every reason to believe that, when those implications are considered, the traditional use of ethical concepts will be seriously affected" (148). Why do we have every reason to believe this? What are the implications (Williams of course can't tell us, because the only term he's given is the schematic "F")? We don't say that the practitioners of magic have got it wrong because, if they continue their investigations into causality, the practice of magic will be unsustainable (if we did say that their practice would collapse with continued investigation, we'd say that it would happen because they've got it wrong), nor, of course, did Williams suggest that we do. It is an open question in any given case whether mere reflection on the use of a thick ethical concept will, because of whatever further implications it has, tend to undermine it, or anyway, shouldn't that be what someone skeptical of ethical convergence ought to think? (There seems, too, to be something odd about the claim "you don't have knowledge now because, in the future, you (or your descendants!) will find this practice foreign". By the end of this section Williams is saying what he will later say he said: if "reflection characteristically disturbs, unseats, or replaces those traditional concepts [then] reflection can destroy knowledge" (148), but right here he is saying that reflection proleptically already has destroyed knowledge even before any disturbance, unseating, or replacement has happened.) In actual courses of reflective thought sometimes the wildest nonsense ends up being confirmed and the distressing implications just remain unconsidered or are explained away, and often, also, the reflection that destroys must be carried out at a rather high level of sophistication (we don't all have Williams's skill). At the same time, it seems that Williams can't quite say here, as he does in the case of magic, that the observer, because he can see the that the observees are engaged in a practice of trying to get it right that connects with his like practice, can bring his greater knowledge to bear and say that they're wrong on that account, because the observer, who one presumes is also reflective, shouldn't on Williams's account have that greater knowledge. But it's only something like that, as far as I can tell, that justifies his denying them knowledge now, merely because of the kind of life they have, before the corrosion typical of reflection has set in.

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?)

My pettiest post ever

For someone whose professional career was based on her persnickety checking and correction of putatively minor details of punctuation and whatnot, Mary Norris demonstrates a surprising willingness to be fast and loose with her descriptions in Greek To Me. Here, for instance, is something she says about the letter chi, χ, as it appears, transliterated, in English words:

Speakers of English often have trouble pronouncing words with ch in them—melancholy, chalcedony, chiropodist, chimera—because ch also represents the sound in such common English words as church, chicken, and cheese. (You could say our alphabet is imperfect.)

Speakers of English have trouble pronouncing such words as those in "rødgrød med fløde", in at least a double sense: first, when confronted with the written words, speakers of English are unlikely to even know how to begin; they will find it difficult to go from a written representation to an audible performance. Second, even after hearing a Dane speak out the phrase or one of its components, a speaker of English will find it difficult to repeat the performance back, producing the same sounds again.

Speakers of English certainly do not have trouble pronouncing the sounds /k/ or /tʃ/, which is completely unsurprising, since both of those are common sounds in English. Speakers of English may occasionally be wrong in how they pronounce a word containing "ch", as I was, for a long time, with "chalcedony", but that, surely, does not mean that such speakers have trouble pronouncing them. They pronounce them with ease and confidence, just wrong. I would never have characterized myself, in the period in which I was incorrect about "chalcedony", as having trouble with its pronunciation, and I doubt that anyone else would have, either, even someone who knew its correct pronunciation. Perhaps speakers of English are often mistaken about "chiropodist" (though Norris gives us no real reason to believe that, either), but do they often have trouble?

Are we Aristotle's enemies?

Are we Aristotle's Enemies? is the title of a paper Agnes Callard gave at the 2011 Pacific APA, with commentary by Michael Thompson, which I would dearly love to have heard, both because I suspect he had interesting things to say about it in general and because of a particular part of the paper where he's mentioned only to be set aside, in what I think was an intellectual lapse—not because he deserves to be treated more thoroughly in himself, but because putting aside a view simply because only one person holds it, when it might offer a way out of the difficulty you're proposing, strikes me as irresponsible.* At any rate, I didn't hear either her delivery of the paper or his delivery of his response, because I was speaking elsewhere at the same time. But she did kindly send me a copy of the paper.

Probably unsurprisingly, the paper does not explicitly answer its titular question, but it's hard to avoid the conclusion that the intended answer is yes. I do not think that, in general, enemyship is a symmetrical relation, but it does seem as if in this case it is: if we are Aristotle's enemies, then Aristotle is reciprocally ours. (Aristotle being dead, it makes more sense for us to think of him as our enemy, since we still interact with him in the form of his writings and influence, than for us to think of ourselves as his enemy, since he is, as mentioned, dead, if there is to be an enemy on the scene at all.) At the very least, if we are not his enemies, or he ours, we may not be able to be friends: The cost of understanding Aristotle might be that we can no longer read him in the amiable spirit to which we are accustomed. The crux is the concept of natural slavery, which goes deep in his philosophy, or rather, comes from deep in his philosophy: his views about slavery … follow directly from his views about the nature of action and how it is bound up with choice. These views, or rather, the underlying basis for these views, makes for a deep difference between Aristotelian ethics and anything we could, today, call ethics.

I don't much like the first-order discussion of the paper, for reasons that I will relegate to the paragraphs following the last occurrence of a doubled asterisk in this post.** I'm more interested in the way she talks about Aristotle. She does not downplay his commitment to slavery as an acceptable or even good and just phenomenon; instead, she invokes it at several times, and even concludes with it, saying that the difference between Aristotelian ethics and modern rights ethics explains why we’re against slavery, and why Aristotle wasn’t. The discussion of slavery is not actually necessary to establish the putative difference (it's present in the elaboration of what she takes to be Aristotle's position, but more as a description of Aristotle on slavery than as anything that furthers the discussion of the theoretical basis of his ethics). It is, as she advertises in her introduction, a consequence of the deeper view on action, choice, and "moral sources" that she lays out. (Or at least, it's advertised as a consequence. It doesn't actually follow from anything she presents as his views on choice, etc.!—we get his thesis that slavery is actually good for those who are lesser, but that doesn't follow from his views on choice, etc.; it's something additional.) Its function in the paper is to drive us to answer its titular question in the affirmative.

She does not attempt to minimize but amplify the moral significance of Aristotle's commitments; she does not, for instance, suggest that he ended up with a view on action, choice, etc., that underwrites the idea that some of us are less than fully accomplished persons because of the stratification of and presence of slavery in ancient society, as if he lacked the wit to envision alternate social arrangements (something that would be a truly spectacular suggestion, both because his predecessors were able to do so, as Bryan W. Van Norden points out in this generally good piece, and even Euthyphro was able to frame the thought that one can do wrong to a manual laborer—and because a world of slavery and of the subjugation of women and manual laborers was not exactly unknown to modern thinkers). Perhaps for the same reason, she isn't really interested in why Aristotle thinks the things he does at the empirical level, either—she mentions his assessment of Dacians, but it isn't really all that interesting to her.

What, if we believe all the arguments in the paper, would we do with Aristotle? Would his ethics be salvageable—might there be something to the vision of megalopsuchia, "great deeds" (her translation of eupraxia), and being tall? Well, maybe, but the list of specific things Aristotle thinks virtue to consist in is not very interesting, except insofar as it still lives alongside other visions of the good and noble in our lives; the interesting thing would be the account of the acquisition of virtue, the nature of deliberation, etc., and on our present supposition, those are all vitiated by their leading directly to approving slavery. (It would be even worse for Aristotle, here, if we did think that he had arrived at his account simply because he looked around and saw a world of injustice and subordination and couldn't but echo it. If we knew in advance that that was why he thought what he did, we might be interested in the details as a kind of curiosity—what explanation for what he saw did this famous empiricist come up with?—but given that we'd see the explanatory project as misguided from the jump, how could it rise above being a curiosity?) Would it be instructive as a contrast class? Perhaps, but Callard has no reference to Aristotle in the development of her account of what she takes our current (oddly univocal) beliefs to be. We don't, evidently, need him to learn for ourselves about our own commitments. And we might well wonder if idly debating about such things, in the abstracted manner characteristic of many contemporary philosophers, in which the nature of the phenomena being discussed is generally covered over, doesn't tend to damage our souls, to put it in terms that an ancient philosopher would recognize: for there is a far greater risk in buying teachings than in buying food [because] you cannot carry teachings away in a separate container. You put down your money and take the teaching away in your soul by having learned it, and off you go, either helped or injured. Ironically, it's those who get hotheaded about such things that keep themselves best.

---

What are we to make of Callard's much later declaration that Aristotle is not our enemy? I don't think the piece in which the claim appears is very good—in fact, I think it's quite bad. To pick up the thread most recently laid down above, at one point she says that dangerousness [of speech], I have been arguing, is less a matter of literal content than messaging content. In fact, she has not argued that at all; to argue that, one would have to argue not only that messaging content is dangerous—which she doesn't do—but also that literal content is not—which she also doesn't do. (She points out that what she calls messaging content, which is not actually about the content but the use of an utterance, has an aim other than the conveying of a truth via the content of the utterance. She calls this, absurdly, extra-communicative, as if communication were primarily a matter of conveying truths via utterances whose contents are those truths. I'm so happy for you! would be, by her lights, at least on dangerous ground vis-a-vis messaging content, since the point of saying such a thing is not, in the first instance, to inform its hearer of its speaker's emotional state. Oh, yay! is even worse in that regard, and I now pronounce you man and wife might give her much to ponder. None of these things is plausibly "dangerous", think what you will about the institution of marriage.) Her concern with messaging content is in any case ironic, since there is no serious way to take her piece except as a bit of messaging content, given the context of its publication and, especially, that there is no remote threat of Aristotle's being cancelled, however one might understand that term. Are we really to imagine that Callard chose to write this piece simply to unburden herself of some things she holds to be true and to inform us that she holds them to be true, and that's it? It hardly seems worth the effort! It is an attempted intervention in something it doesn't mention, one in which being able to pose as being a reasonable, civil person who just wants to hear the other person out is itself a position of power. It is too bad that sometimes on particular occasions simple utterances of things that even are truths will convey further things, as when someone who truly has been living in a cave says all lives matter, not realizing the further significance doing such a thing has. The solution to this is not to insist on an impossible nightmare of strictly literal utterances, but sensitivity and discretion. But these things aren't to be extended time and time again to the same subjects.

*Much turns on the idea that A wrongs B and B is wronged by A are different albeit reciprocally entailing facts; that there are active facts and passive facts, rather than one fact differently expressed. (Active and passive are particularly unfortunate. Consider this bats description of rights (also a rather bats scenario in general):

… a right makes someone a threat. If B was going to prevent someone from φ-ing, and he had to choose between A, who has a right to φ, and C, who doesn’t, A’s right to φ would be an intelligible ground of his choosing to act on C. B cannot act on A with impunity—that is, he cannot act on A without thereby doing something wrong. A’s right might, then, exert a certain pressure on B’s treatment of A. Notice that B’s experience of this pressure, insofar as he does experience it, will be independent of A’s exercising his right. This pressure is the pressure of a threat: should B act on A (rather than C) he knows that A holds the trump card that will make him into a wrongdoer.

On this description, the same pair of sentences about A and B having wronged/been wronged bye each other could be replaced by these, to describe the same reality: A is made vulnerable to B and B gains the upper hand over A.)

**For one thing, it seems to be internally inconsistent; I don't see how you can simultaneously say that enslaving those who are natural slaves is good for them in the way she says it can be (bringing them into some connection with the human good) and also say that only the naturally free can be moral patients, capable of being wronged or done well by. And she identifies without argument, and without even seeming to realize that these are different sorts of things, the idea that someone has "passive virtue", that is, someone who responds virtuously without the involvement of choice or deliberation (because the source of the action is outside the agent), and the idea that someone can be a "moral patient", that is, that someone can be wronged or done well by. (By "without argument" I mean to include the fact that she doesn't cite anything from Aristotle that would support his holding this position, whether he argues for it or not. Of course the "moral source" language is not his.) There is certainly nothing prima facie in the idea of someone's being capable of reacting rightly to the sight of someone needing urgent help to connect it with their being capable of being wronged by being denied an agreed-upon share of some goods, for instance. I take it that someone is wronged in cases of pleonexia! (An uncharitable person would consider this motivated by little more than a pun.) Later she will say that, for Aristotle, what’s wrong is what the unjust persons does, by choice, to his fellow citizen. But she hasn't equipped us to understand why it has to be done to a fellow citizen. And this is actually the heart of the claim that the defense of slavery falls out of Aristotle's understanding of action and choice, as far as I can tell—that, for instance, it would be inconceivably un-Aristotelian to object to slavery on the grounds that a just person wouldn't do that to another human, even if you accept the idea that person is a status only some humans have. After all, there are lots of things that the virtuous person doesn't do even though the patients of the action aren't "moral sources" (for instance, torturing cats); the virtuous person doesn't do them because to do them is base or vicious.