The Animal in Translation

My title straddles two of the hottest subfields in the humanities today: animality studies and translation theory.  I am not interested in producing out of the contact between these two subfields a hybrid disciplinary commodity—animal translation theory, if that is even conceivable.  I want rather to suggest a way in which animality studies and translation studies may serve to limit each other, but may also disclose aspects of that field that remain otherwise obscure.  My goal is to show how each provides for the other a way of avoiding the disabling conceptual traps set by seemingly necessary essentialisms, linguistic and species-ist.  Animality studies and translation studies and theory provide these limiting-disclosing functions for each other just where they come to rely on accounts of mediation in order to produce rules for conduct or for thought more generally.
 
In what follows I say a little more about Temple Grandin’s Animals in Translation; for Grandin’s properly philosophical bases I turn to Vicki Hearne’s work and to W.V. Quine’s.  I’ll characterize the position they share as empiricist and mean by this that Grandin, Hearne and Quine place an abstract notion of immediate stimulus at the origin of their consideration of (variously) the varieties of human attention (Grandin), the ethics of animal training (Hearne), and the indeterminacies of radical untranslatability (Quine).  I contrast the (suitably complicated) empiricist position with one I attribute to Jacques Derrida, and which I characterize as mediationist.  I mean by this that Derrida’s consideration of the “relation” between human and nonhuman animal never settles into a concept that is immediately at ease with itself.  I do not promote the latter (also to be suitably complicated below) over the former, but rather describe, first, their shared, if quite different, engagement with normativity, and, second, the productive conflict between the two.
 
I’m interested in two sorts of paradoxes.  Here’s the first.  Say I assert that my animality is what is most immediate to me.  Nothing stands between me and my circumstance as an animal; my animality is precisely that, perhaps that and only that, in respect to which I entertain no mediation, no conceptualization: it is a synonym of the facticity of my human being; my form of being in the world; my life form.  What and how I may think about my animality is subsequent to this state of affairs.  But on the other hand there is nothing about which I can think, no concept or problem more heavily overdetermined, more historically saturated than the “animal” I may or may not be; even to characterize my animality as the facticity of my human being is to position my assertion, my characterization, in a historically dense, shifting, and fractured discursive field.  Nothing is less immediate to me than my animality, which comes to me conditioned by the genealogy, the history, the differential drift of the correlative and mutually limiting concepts of the animal and the human.
 
And on the other side, the second sort of paradox, in the shape of these two assertions: everything can be translated, give or take, of course, and not just from one natural language to another, but even within a language: I have a notion, I express it in English, and beyond the specifics, beyond whatever it might mean to me in private to say the words “house” or “car” or to mention the color “red,” whatever associations I might have from early experiences with the terms “house” or “car” or with that color, you understand them well enough, we arrive at a practical exchange of information, the right or the expected things happen when I say “I have painted my house red” or “Hand me the red jar” or “I drive a red car.”  But on the other hand, nothing can be translated, properly speaking, not just between two national or natural languages but within any language.  I say “I have painted my house red,” and you understand me and say “How nice” or some such, but only because you and I have set aside the criterion of exactness or the consideration of the intentional force of my utterance, in favor of the criterion of practicality, which amounts to acknowledging that “I have painted my house red” can only be “translated” inasmuch as each term has been abstracted of its historical embeddedness for me and for you.  Both of us now say “I,” “my house,” “I have painted my house red” or “I say to you that it’s nice that you’ve painted your house red,” in a phatic, indexically neutral sense.  No expression is translatable if it is not separated (or, minimally, separable) from its circumstances of utterance, from its history, unless it is immediately abstract.  Our register is now tragic, where it was comic; by “translatable” we now mean “properly translatable,” and we have in mind something like a formal procedure that converts, without loss, sense into sense: a mathematics of translation.  A mapping.  Nothing moves from this map to that, from this space to that one there, unless it first sheds the complexities of circumstance and ascends again, utterly naked, into the idea which is its home, from which it can pass, trailing, into another location, language, map.
 
How do these two limiting descriptions of “the animal” and “translation” line up?
 
Say that we claim to be able to “read” the expression on an animal’s face, or to guess what our companion animals are thinking, or to understand the pain they feel when they are mistreated.  Such translations, we might maintain, are possible because animals belong properly and already within the field of translation, because they are sufficiently alike to human animal users of natural languages to permit analogies or outright, Aesop’s fables-ish allegorizations of animal speech.  That is why we can analogize “speaking” to animals, and perhaps more broadly speaking about them, with speaking to someone, to a human animal, whose language we don’t yet know.  Non-human animals are “in translation,” as Grandin has it, when and if the field of translation encompasses their communicative disposition, which is not linguistic.  We could say that this or that primate, call it or her Koko or Quigley, may not know what her trainer’s gestures mean to her trainer, that no non-human animal poses to itself the question of the “meaning of meaning,” and that for this reason no non-human animal can, strictly speaking, be said to “intend” this or that communication.  And yet there is communication.  Say then that two human animals who don’t share a natural language encounter each other.  Inasmuch as they are animals they share a relation to the matter of translation.  The relation need not be identical or symmetric as to the circumstances of the act of translation, but we might non-trivially maintain that as to its structure or type it is both.  For instance, it may be that I think of translation from English to Japanese under the aspect of a commercial advantage to myself, whereas my Japanese interlocutor imagines learning to translate English as a way of gaining access to a cultural sphere.  (My example squarely shoulders the coarsest of cultural clichés, quite deliberately: examples too are intra-linguistic translations; when they work to give an abstract argument concrete shape they trail, even sometimes turn upon, often-unremarked commonplaces.)  “Translation” would be a term and a practice mediated differently for my Japanese interlocutor and for me, and yet we could, perhaps, agree that what we are doing when we seek, each for her or his reason, to understand what the other is saying, is something we could both call “translation,” a limit-term or horizon or type-term that falls apart as soon as we seek to specify, each on her or his side, what it means.
 
This is controversial.  The poet, animal trainer and philosopher Vicki Hearne famously defended a vigorous trainer-to-dog anthropomorphism in her account of “How to Say ‘Fetch!’’” (55-56).  There is a crucial moment in her training of Salty, a “year-old Pointer bitch” that Hearne invents for us.  The dog has learned to “obey” the trainer, but the trainer does not yet “obey” the dog—which is to say that although the dog obeys “Salty, Sit!” the “looped thought” between trainer and dog is still unidirectional: “the flow of intention,” Hearne says, “is, as it were, one way.  In my account the dog doesn’t initiate anything yet.  She obeys me, but I don’t obey her.”  And then one day, Hearne says,
 

Salty gets my attention by sitting spontaneously in just the unmistakably symmetrical, clean-edged way of formal work.  If I’m on the ball, if I respect her personhood at this point, I’ll respond.  Her sitting may have a number of meanings. “Please stop daydreaming and feed me!”  (Perhaps she sits next to the Eukanuba or her food dish.)  Or it may mean, “Look, I can explain about the garbage can, it isn’t the way it looks.”  In any case, if I respond, the flow of intention is now two-way, and the meaning of “Sit” has changed yet again.  This time it is Salty who has enlarged the context, the arena of its use, by means of what we might as well go ahead and call the trope of projection.  Salty and I are, for the moment at least, obedient to each other and to language. (55-56)

This is lovely and striking.  For now, note that even if we are not committed to a vigorous anthropomorphism, of the sort indicated by Hearne’s injunction to “respect the personhood” of a dog, or by her attributing to the dog the use of a trope—which, if the phrase is intended seriously, indicates that Salty is able to distinguish between literal and metaphoric uses of “language,” and that the dog, communicating, installs, as the primary device in the symmetrical, intentional circuit of the language-game that both dog and trainer employ and obey, the figure of projection—even if, as I say, we are not committed to this vigorously anthropomorphic description, we have at our disposal a weaker account of communicative disposition and a weaker version of this bidirectional “flow of intention.”  This version, which Hearne characterizes accurately as “Skinnerite” but then dismisses too quickly, does the job even if we maintain that non-human animals can have no intention when they interact with human animals.  We can fairly assert that Salty or Quigley or Koko is doing something, what we would call gesturing in American Sign Language, or sitting by the dog food, or pointing to an icon or baring its teeth, and that the ape or the dog is doing this because that gesture will have an effect on her trainer—the primate or the dog is making this sign in order to have that effect, which might be the effect of producing a treat, or a kitten, or panic, from the trainer.  The word “because” here means something weaker than or at least different from verbs like “understands” and “knows,” terms we would require in order to ascribe intention to the ape or the dog.  The trainer has trained Salty or Koko or Quigley to respond in this or that way to one or another stimulus, and the ape or the dog has trained the trainer to respond in predictable ways to a counter-stimulus.  (The animal does not have to intend to train the trainer for this to occur: a response to a stimulus or a counter-stimulus occurs “because” training has happened, not because anything about the causal nature of the stimulus circle is understood or known.)  Being “in translation” in this sense means being within the closure of a stimulus circle, just as “communicating” with my Japanese-speaking interlocutor means in the first place, before any specific translations begin, acknowledging that we share a common disposition toward a vacuous or horizon-concept of translation, which may be inflected by different cultural or personal circumstances—by my economic interest in learning Japanese, or by his interest in learning English for cultural purposes.  In this restricted sense, whether what’s at issue is my conversation with my Japanese interlocutor or Salty’s attention-getting sitting or Koko’s “conversation” with her trainer, a fundamental but vacuous symmetry is installed, a common “obedience to language” where “language” is to be understood as a field of potentially translatable assertions, of assertions that all of us, the Japanese interlocutor, the trainer, the animal, myself, agree are or may be translated, are or may be understood by another about whom I can say, by means of the trope of projection precisely, that she or he or it could understand them.  And when I say that “we agree” to this, I am giving tongue to what Vicki Hearne designates when she uses terms like “spontaneous” and phrases like “unmistakably symmetrical, clean-edged way of formal work.”  What gets our attention and brings us into recognition that we “obey language” together, is “unmistakable”; it occurs “spontaneously.”  Whatever this is, it lies behind the trope of projection; it might be said to be the tenor for the trope of projection, or its schematic condition of possibility.  It is not language, but it founds the “language” to which the animal and its trainer assent, and to which interlocutors speaking recognizably different human languages also assent.
 
It is not language, but it furnishes the rules for thinking about the relation to a primary or primal language that human and non-human animals share.  It is not language, but is this foundational point, this point on which the assent to language depends, this point from which depends our spontaneous recognition that we and the non-human animals we train are beings-in-language, is it translatable?  Just what are these rules for thinking, and how are we to follow them?
 
In order to bring the question of rule-following into contact with the question of mediation in animal studies and in translation studies, let me designate two limits.  On one side, obeying a rule, being named or naming, or being trained or training, are activities related both to the immediacy of the stimulus reaction, and to the abstract translatability of concepts.  It is, in fact, impossible to separate immediacy from abstraction on this point: the presumed immediacy of stimulus reactions is the condition on which concepts are understood across languages and idioms, and vice-versa: stimulus reactions are immediate to us because they are already abstract, and fall for this reason outside of the domain of our particular interpretations.  This reciprocal arrangement is logically shaky but wonderfully secure culturally and experientially.  When Salty sits in her “immediately and spontaneously recognizable way,” I am “on the ball” if I react to her stimulus.  Under these circumstances, I am acting in a relation to my animality; I am well trained, but I am also, inasmuch as I “respect the dog’s personhood,” in a position in which my own “personhood” is being “respected” by the animal.  We’re both obedient to “language,” but my attending-to-Salty is not a rule that can be generalized.  So what sort of a rule is it that cannot be generalized?  All I can do is say, “Be on the ball” to other human animals, and by this I mean something quite empty: I mean “pay attention to the stimulus you will receive” or I mean “See, visualize,” as Grandin would have it, or I mean “Don’t mistake what is ‘immediately and spontaneously recognizable;’ don’t mistake what is unmistakable: make sure that a stimulus is a stimulus for you!”
 
The unity-and-emptiness-of-stimulus argument has any number of attractive and useful formulations, but probably the most explicit and the most famous one is found in W.V. Quine’s account of the emergence of radical translation, as he tells it in Word and Object.  Quine’s famous story has a pleasantly jokey syntax: a linguist, understood not to know the local language, meets a member of a “hitherto untouched people,” a native speaker of a language that the field linguist cannot decipher (Word and Object 25).  A first occasion for establishing the point of contact, of translation, between these two languages presents itself when a rabbit “scurries by.”  Quine calls this a “stimulus situation,” and means by this that the linguist and the speaker experience, roughly simultaneously, the stimulus of seeing the scurrying animal.  “Gavagai,” says the speaker.  Provisionally, Quine’s linguist notes down that “Gavagai” may mean “‘rabbit’ or ‘Lo, a rabbit!’”  Actually disambiguating the expression proves difficult, and impossible without “supplying native sentences for [the native] informant’s approval, despite the risk of slanting the data by suggestion.”  “For,” says Quine,
 

suppose the native language includes sentences S1, S2, and S3, really translatable respectively as ‘Animal,’ ‘White,’ and ‘Rabbit.’… How then is the linguist to perceive that the native would have been willing to assent to S1 in all the situations where he happened to volunteer S3, and in some but perhaps not all of the situations where he happened to volunteer S2?  Only by … querying combinations of native sentences….  So we have the linguist asking ‘Gavagai?’ in each of various stimulatory situations, and noting each time whether the native assents, dissents, or neither.  But how is he to recognize native assent and dissent when he sees or hears them?  Gestures are not to be taken at face value; the Turks’ are nearly the reverse of our own. (Word and Object 26)

Faced with a palpable regressive paradox, Quine’s conclusion, famously and controversially, is that in cases of radical translation, there is no way to establish definitively what “Gavagai” refers to or what it means.  The animal I name may be identified by means of a collection of descriptive attributes or predicates of identity, accidental or essential.  In this situation, which even goes beyond what Wittgenstein calls “aspect-seeing,” “Gavagai” could refer to an aspect, for instance a temporal aspect, of the rabbit-in-motion; to a part of the rabbit, the tail say; to its color; but “Gavagai” could even refer to the finger pointing.[1]  “Gavagai” could mean “I don’t know what that is.”  It could mean “food.”  Presumably the list of things the term could mean is not infinite, but probably it is not countable, either.  The animal I name may be identified by a collection of attributes that describe the animal, and the name “Gavagai” refers to one or many of them; but this naming-convention may not pertain in the discursive world of the speaker, who might be what Kripke calls a rigid-designationist rather than a descriptivist; and of course “Gavagai” might not be a noun at all.
 
Quine’s philosophical fable has elicited important glosses.  Some of his readers—those interested in exacerbating the fable’s regressiveness—have suggested that descriptive identity predicates are implied in the speaker’s gesture of pointing, and that these in turn are to be understood as collections of additional identity predicates requiring disambiguation, and so on.  They do not share the word “rabbit,” or the word “Gavagai” as a word; even the primitive ostensive function for language indicated by the speaker’s pointing is suspect, since the speaker might be saying the word “Finger!” or the phrase “I am making an indexical sign!”  Under the threat of regressions such as these, one solution is to stipulate that for there to be translation, indeed for there to be communication, a decision is called for: with sovereign assurance, one cuts through the thickening forest of predications, and settles on one, the likeliest or the most motivated.  The animal itself is manifestly not what the native speaker and the non-native linguist share when such decisions are called for, in the way that you might say that people share food, if they eat rabbit together.  The analogy to a Schmittian scenario is useful: this is the sovereign in the bush.  For the act of deciding (that “Gavagai” means one or another thing, stimulated by the presence of the rabbit; or on a different, fundamental level, that “Gavagai” is to be interpreted as a response to the stimulus of the question) is held privately by one side, on the linguist’s side, say, or on the side even of the analytic philosopher narrating the little fable.  The speaker and the linguist do not decide together, according to criteria they share, or according to what Quine calls a “manual.”  (How would they arrive at these criteria or at a shared “manual,” without a primal moment in which each side designated for the other what the criterion is for deciding what “Gavagai” means?  Or what “manual” they should share?  And wouldn’t that primal designation be subject to the same skeptical deflation as the moment when the native sees the rabbit, if that’s what he sees, and says, to the incomprehension of the linguist, “Gavagai”?  “Gavagai” could, after all, be the name of “concept” or of “convention” or “criterion” or “object” in the native tongue: our little animal-and-forest fable could also be a fable concerning the designation of entities that appear, are observed, or are produced for thought and consideration.)  Sovereign is he who decides whether there is, or is not, translation.  Sovereign is he who stipulates the criteria according to which it is to be judged whether there is, or is not, translation.  The sovereign designates the manual.
 
Or not quite.  Yes, the native speaker’s perspective necessarily drops out of the field of decisions, but on the other hand for Quine’s story-example to work, for the speaker’s perspective to drop out, then alongside the perspective of the animal itself (whatever that might be), the native speaker and the linguist must share a notion of what a “stimulus” is, or rather, they must both be obedient to the stimulus, as Vicki Hearne and Salty are obedient to language and to each other.  The words we use to describe the rabbit may be conditioned, as the word “red” is conditioned for me by my past, by my home’s color, or by my political affiliation; or as the color “white” and the noun “snow” are said to be in the Inuit tongue by the circumstances of Inuit experience.  “Rabbit” may be translatable or radically untranslatable; but the fact that there is stimulus and that the stimulus is “immediately and spontaneously recognized” as such is the condition on which there is a scene of pointing, the condition on which the question of translatability or untranslatability arises: it is the fable’s ground.  Even when by “Gavagai” we take the speaker to be saying “this is my finger,” or even if “Gavagai” means nothing—it still stimulates. “Gavagai” is a stimulus understood to arise in response to a stimulus, no matter what the first stimulus was, or what the second stimulus, the enunciation of “Gavagai,” means.  Even so, then, the conditions that Quine sets up are such that the speaker can only mean “This is my finger,” or mean nothing, or simply set about stimulating us by uttering a word, “Gavagai,” that he knows to be utterly foreign to the linguist’s lexicon—the speaker can mean or intend any of these only upon the stimulus-reaction occasioned by the scurrying rabbit.  “Gavagai,” he says, and whatever else might be at issue, we three, or four, the native speaker, the linguist, the narrating analytic philosopher, and the readers of Word and Object, “obey” the immediate and common, agreed circumstance: there is a stimulus, and it is coincident or correlative with “Gavagai.”  For the question of translation, and for the question of decision, to be posed, Quine says, there has to have been stimulus in the first place, or rather there has to be agreement that there has been a stimulus in the first place.  We recognize, “immediately and spontaneously,” that we are, as Hearne puts it, obedient to language inasmuch as we agree that there has been stimulus.  “Agreement” does not mean something like “the conscious or deliberate, common assent to a fact of some sort,” but rather something like “stipulation,” or even—this is perhaps controversial—“agreement” here means that we, the native speaker, the linguist, the narrating analytic philosopher, and the readers of Word and Object, are all trained, have all been trained, to identify a stimulus in common.  It is not required—indeed, it is excluded—that the linguist and the speaker and the rest of us have an identical description of what “stimulus” is (that is, an identical collection of predicable attributes attached to the term “stimulus”), or of how a concept works, the concept of stimulus, or of how a decision is reached.  A stimulus, the change-of-state-of-affairs, is shared by the speaker and the linguist and the rest of us and as a result we are obedient to language.  A meadow with nothing happening in it is precisely not stimulating; it’s the unmistakable and spontaneous change that’s signaled by, that is, the animal’s movement or that’s signaled by or that is the stipulation “There is stimulus,” that gives rise to the question, “Is it, the word for that which is producing the stimulus or for the stimulus itself, is it translatable or is it not translatable?”
 
“Stimulus” is neither translatable nor untranslatable between native speech and field-linguist-ese, nor a fortiori between natural languages, because, on Quine’s account, “stimulus” marks the as-yet-unpredicated spot, common to all languages, upon which all languages take the shape of syntactically-organized fields of differential predications, by means of the trope of projection.  Not the rabbit but the scurrying rabbit is the functional term here, the term on which we can stand in order, by projection, to tell our story.  I take Quine, and Hearne, and Temple Grandin, to be raising a philosophical scaffolding on the ground of empirical but abstract immediacy, and to require that both the native informant and the animal before us, the animal and the informant who bear attributes and are the subject and translator of definitional predicates, be sacrificed so as to produce the concept of the empty stimulus-function on which natural language is built, and on which decisions concerning the possibility of translating or not amongst natural languages are built.  I intend the range of senses associated with the word “sacrificed”: the native informant and the animal before us are taken in place of something else, more precious; they are offered up, propitiatory or apologetic, to a sovereign power; they are destroyed, as when we say that the stricken cattle had to be “sacrificed.”  These are different approximations to translation, as we will see.
 
I asked above regarding the foundational, void stimulus-point on which empirical philosophies stand the assent to language and balance our spontaneous recognition that we and the non-human animals we train are beings-in-language and beings obedient-to-language, whether this non-linguistic point is translatable.  This question cannot be understood in the lexicons that Hearne, and Quine, and Grandin, offer us.  To approach it, allow me now to turn to the other limiting side to the question of how rule-following meets mediation.  Consider now a philosophy critical of the naked immediacy of empirical stimulus, and of the resilient correlation between the stimulus reaction and the concept.  Does it too sacrifice the animal in translation?
 
Four animal scenes recur in Jacques Derrida’s late writing.  They are, first, the Adamic scene of the man who names the animal, who calls the animal by name (“I name you ‘Gavagai,’ my pet rabbit!”); second, the scene of the father who, required to sacrifice his son, finds, or is given, or produces, a substitute; third, the scene of the slaughterhouse, where the modern logic of industrial mass production brings about the phenomenon of the animal bred in great numbers in order to be slaughtered.  Fourth—last—is the scene of an encounter between a man, call him a philosopher, and a non-human animal whose presence offers the philosopher, or imposes upon the philosopher, what Derrida calls “the certainty that what we have here is an existence that refuses to be conceptualized [qu’il s’agit là d’une existence rebelle à tout concept].”[2]
 
There are deep and exceptionally tricky relays among these four scenes, though they do not resolve themselves into a single scene, and they do not offer us a concept, or even a concept of concept, that would allow us to group them easily (see Lawler).  (One might say that they are, like the cat that presents itself to Derrida’s gaze in The Animal that Therefore I Am, rebellious to any concept.  Derrida’s animal scenes are more like animots, animalwords, in Derrida’s splendidly strange usage, than like different animals of a single species, or different species of animal.)  For Derrida, there is never a single encounter between the philosopher, or indeed between the human animal, and another animal, human or not.   Such encounters are always at least fourfold; they take place in and are always drawn from at least four at times incompatible registers and locations—the register of Adamic naming, Abrahamic sacrifice, industrial slaughter, and the register in which the animal presents itself as, or impresses upon us that it is, irreplaceable life.  Our singular ways of thought and of expression account in part for our need to think these scenes through systematically; in our little scenes, a philosopher is never a single philosopher, both in the ways that an animal is never just an animal, and in other ways too; a philosophical scene—whether of writing, or of encounter between a non-human animal and a human—is never one.  In this manner at least the encounter between the philosopher or the philosophical human animal and the animal can never provide a “stimulus,” or indeed be a numerically discrete event like a decision.  On this score as on so many others, we would be hard-pressed to derive from Derrida’s work rules based in agreed, translatable concepts, governing decisions taken in accord with classical, individualist ethics, or regulae ad directionem ingenii.[3]  The “unshakable” certainty that this moment provides sits upon the shakiest of grounds: the “certainty” that what stands there, the being that stands before the philosopher, is an existence “rebellious to any concept” or, as David Wills translates it, “an existence that refuses to be conceptualized.”  And how could I replace that irreplaceable living being with another, or think of it alongside another being; how could I gather both of them under the sheltering, sacrificial generality of the concept, translating them (in one version of the word “translating”) one into the other?
 
In Derrida’s animal-human scenes, in each taken singly and taking the aggregate as a non-systematic collection, we come before something that rebels against any and every concept.  But what does this mean?  What can we then mean by “thought”?  By “decision”?  What sorts of rules can one build when one is unshakably certain that one faces, in the animal and in the scene and scenes in which the animal becomes present to thought, something—an existence, a word, a case—rebellious against any and every concept?  Minimally, we would be exchanging the concept of concept for the mediated and discontinuous circuit—though this is not the language that Derrida uses.  No sacrifice that does not pass through the slaughterhouse (and in this context the theologico-political value of the sacrifice acquires one order of significance in the post-industrial society); no presentation or intrusion of irreplaceable or unsubstitutable life without an Adamic exercise of mastery over it.  The discipline of philosophy, even the project of thinking what the rebellion against the concept might look like, passes through the Edenic fantasy of primal naming rights.  Adam’s task is unthinkable outside of the shame of the abject, naked philosopher-namer.  In the lexicon that we use to describe the relation among these scenes, or these classes of scenes, we lean heavily on the concepts of determination and over- and under-determination, that is, we lean heavily on the classic vocabulary of mediation.  In this lexicon, the scenes of human-animal encounter are not single scenes, units to be coordinated or composed according to rules or syntaxes: there is always more than one scene, and this plurality is the abstract condition that they obey, the condition of their relationality—in contrast to the abstract visual unity of the stimulus that we find in Quine and in Hearne.  Saying that each scene of this plural set “passes through” all the others is my way of expressing what Derrida worries as the problem of a “general singularity” or a “singular that is general,” un singulier générale or an “indeterminate generality,” une généralité indeterminée.  The expressions “an animal” and “the I” or even the pronoun “I” share this strange quality: they cannot be thought alone, and yet they are one, one scene, a singular scene, for instance the scene of the singular event of a naming or of an averted sacrifice or of the industrial slaughter of produce “animals” or the scene of the philosopher’s shame at the rebellious non-nudity of the animal.  The scenes in which “I” come before the “animal” as its namer, as the father beholden to it inasmuch as it stands in for my son, as its executioner, as the man who cannot think his concept before this animal, are general singularities or indeterminate generalities.
 
So if we may not say that the scenes of human-animal encounter about which Derrida’s late work turns resolve themselves into a concept; nor that there are systematic relays between them; nor that these scenes mediate each other; nor that they negate, determine, over-determine, or under-determine each other; then what is the rebellious, defective concept of discontinuous circuit that we may use to understand their relation?  Derrida is entirely aware of the paradoxical nature of the structure he is furnishing, a structure at work not only in the philosopher’s deployment of one or another of the human-animal scenes to this or that end—but at work also in the way that the scenes I’ve described come into relation with one another.  The term that Derrida will use to describe this sort of relation is translation, though he is using it, as I’ll suggest, in a way that’s different from the more traditional ways in which I’ve been using it to this point.  The scenes I have described translate each other, and they are untranslatable each into the other.  Scenes in which radical translation and radical untranslatability are represented, they are radically translatable and radically untranslatable amongst each other.  Everything is translatable between and among then; nothing is.
 
Let me show you what I mean.  Béliers: Le dialogue ininterrompu: entre deux infinis, le poème, is among Derrida’s last works.  It was delivered as a lecture in Hamburg in 2003, in memory of Hans-Georg Gadamer.  It consists largely of an extended analysis of a poem by Paul Celan, which Derrida refers to in the English translation by Michael Hamburger.  In this section, which I cite in Thomas Dutoit and Philippe Romanski’s English translation, “Rams,” from Sovereignties in Question, Derrida concentrates on the image of the ram:
 

There is war, and the ram, the ram made of flesh or of wood, the ram on earth or in the sky, throws itself into the fray.  … Against what does he not strike? … One imagines the anger of Abraham’s and Aaron’s ram, the infinite revolt of the ram of all holocausts. But also, figuratively, the violent rebellion of all scapegoats, all substitutes. Why me?  … The ram would, finally, want to put an end to their common world. It would charge against everything and against whomever, in all directions, as if blinded by pain. …
 
That makes for many hypotheses, and for much indecision.  That remains forever the very element of reading.  Its ‘‘infinite process.’’  Caesura, hiatus, ellipsis—all are interruptions that at once open and close.  They keep access to the poem forever at the threshold of its crypts (one among them, only one, would refer to a singular and secret experience, wholly other, whose constellation is accessible only through the testimony of the poet and a few others).  The interruptions also open, in a disseminal and non-saturable fashion, onto unforeseeable constellations, onto so many other stars, some of which would perhaps still resemble the seed that Yahweh told Abraham, after the interruption of the sacrifice, he would multiply like the stars: the abandon of traces left behind is also the gift of the poem to all readers and counter-signatories, who, always under the law of the trace at work, and of the trace as work, would lead to or get led along a wholly other reading or counter-reading.  Such reading will also be, from one language to the other sometimes, through the abyssal risk of translation, an incommensurable writing. (156-57)[4]

I would like to imagine Béliers as a reflection on the ways in which Derrida’s animal scenes come into relation with one another—that is, to read Béliers not just as a meditation on Gadamer’s career, but also as Derrida’s meditation on the ways in which his own career organizes human-nonhuman animal scenes.  On this reading, Béliers becomes an allegory of Derrida’s human-nonhuman animal scenes: its constellations, the complex singular generality of Derrida’s long visualization of the differentiated encounter between human and nonhuman animal; Béliers tells the story of the paradoxical sorts of representation these scenes turn on.
 
So note the complicated relation between numerability and dissemination in the passage I have excerpted.  In Celan’s poem and in Derrida’s analysis of the poem, the stars’ uncountability flows from two different sources.  The first is the empirical difficulty that they might pose to the project of counting: one of the two infinities of the essay’s title.  (How do we count the stars without losing count?  We capture the night sky at a moment, call it dawn, midnight, or dusk, and we count the stars upon that spangled surface above us, hoping not to lose track, hoping to count all at once, before a star sets or rises, before one shoots across the vault: a star? A planet? An asteroid? A bit of technology orbiting above?)  The other source of the stars’ uncountability flows from their double function: they are units, single stars, astronomically observable elements like the morning star or like Hesperus (neither of which, of course, is a star—and both of which, though different in name and in cultural significance, are names for the same planet), but Celan’s stars are also figures or place-holders making up a constellation, the constellation of the ram for instance.  The animal these symbolic-literal stars form is then part of the firmament but also the means by which the sacrifice on earth is interrupted: a definitive, irreplaceable part of the world-down-here.  Stars are and are not stars; their location is fixed, and yet they wander to earth on ram’s hooves; they stand above, like the starry vault Kant so prized, reducing human animality to its contingent, material qualities and signaling perhaps the sublime domain of the moral law or the absolute violence of the sovereign concept; and they also form part of culturally defined constellations.  The ram-constellation represents or is a translation of the ram-substitute in the Abrahamic tradition—and yet the poem, Derrida says, establishes each as a stand-in for the other, one ram for another, one irreplaceable form of life replacing another par figure, Derrida writes.  And at the same time this act of figurative translation between mythologies, one irreplaceable ram replaced by another from an entirely different tradition, is or must be arrested because it is improper, sacrificial, sacrilegious, as Isaac’s sacrifice is arrested by the ram who is, and cannot be, his figure.
 
Celan’s stars, Derrida’s stars, the animals in Derrida’s stars and the animals in Celan’s poem: every animal in each of the scenes that Derrida sets before us, and every scene in which a non-human animal stands to hand, every animal and every scene stands in the place of the others and is sacrificed in its place.  Each represents the others, par figure, but each is also held back from representing the other scenes or the other animals, ram for ram, a ram for a child, a cat for a philosopher, a ram for a cat.  This is not an exercise in contradiction, but an exercise in translation, in what we would have to call rebellious translation.
 
Rebellious.  The word crops up in Béliers, where it helps “One” to “imagine… figuratively, the violent rebellion of all scapegoats, all substitutes.”  And of course it has been keeping us company all along, in the shape of the decisive modifier that Derrida’s cat bears in “The Animal that Therefore I am”—that existence that refuses every concept, that rebels against every concept.  If Celan’s poem is the “figurative” image or imagination of this “violent rebellion,” then Beliers and “The Animal that Therefore I am” are its philosophical “figures.”  But what would a “rebellious” translation then be?  Let me now restate my original questions, and show you how the philosophical figure of rebellion that Derrida provides rebels against the conventional accounts of translation, of animality, of rule-following, of conceptualization, and of mediation.  How, I asked, might the question of the animal and its relation to the human, and the question of the possibility or impossibility of translation, not only serve to limit each other, but also serve to disclose, each about the other, matters that would otherwise remain obscure?  My suggestion was that we should look at the point at which each of these fields of questions depends upon an account of mediation.  I distinguished between translation and animal studies that depend upon the vacuous, visual immediacy of the notion of stimulus, under whose law translatability and untranslatability, training and personhood, abstraction and concretion are marshaled.  In contrast, non- or anti-empirical accounts of human-animal/non-human-animal relations such as Derrida’s offer a defective concept of the scene, numerically and semantically distinct, but also both singular and multiple.
 
Notice what is staked on calling this a defective concept rather than a non-concept.[5]  That Derrida’s cat, and Abraham or Isaac’s ram, embody an existence rebellious to every concept does not mean that no concepts attach to the animal these animals are—rebellion and singularity, for instance, are also concepts.  It means, however, that between the concept and the animal no immediate translation occurs.  If this is so, though, then it would seem that the word, or the concept, of “rebellion” occupies just the place that the terms “scurrying” and “stimulus” do in Quine’s argument, and that “rebellion” encodes, as the stimulus-scene does in Quine’s story and in Hearne’s argument, an imperative, a ground-rule.  In Quine’s case the command sounded something like this: “Recognize, see, in the world of the native and the field linguist but also in all possible worlds, recognize, see, rigidly, that a stimulus is a stimulus, and from that recognition, from that vision, may flow, in different, radically untranslatable or radically translatable languages, the analytically necessary notion that an animal is an animal, scurrying.”  Upon this singular recognition, or vision, the field-linguist, or Quine, or his readers, will then build their disposition toward non-human animals, and indeed toward human animals as well.  To the extent that this disposition concerns how we will behave when faced with this or that nonhuman, or human, animal, we may call it ethical, or rather, we may call it the ground of ethics (which we understand classically to entail acting according to rules we give ourselves on intuited, axiomatic grounds: this is our autonomy, our sovereignty with regard to ourselves).
 
Derrida’s ground-rule, if indeed Béliers and the scenes I have been linking do provide one, is different, inasmuch as it does not simply place the faculty of vision or sense of sight at the heart of the scene; inasmuch as this imperative is addressed to the animal but also to his, Derrida’s, reader; and inasmuch as it also serves both as a description of the cat-animal and of the addressed reader, and as its or their concept.  What results is not a version of autonomous ethics but a baldly political rule, that is, a rule that puts the axiomaticality of any ground rule, and of rule-following in general (my own rules to myself, those appear to agree on), into conflict.  “Rebel!”—this seems to be Derrida’s ground-rule.  “Rebel,” one might think, “against allowing yourself and others to be translated into a concept.”  “Find a way of thinking or of ‘existing’ that attends to this ‘rebellion.’”  The mediated sacrifice, the substitution which is and is not one, rebels against making the other animal obey one’s own concept, against translating the other’s language even figuratively, par figure, into one’s own.
 
But the ground-rule of rebellion is indeed defective.  At the end of the heroic, soixante-huitardish adventure we can glean from Béliers—where the call to rebel is sounded, as though on a ram’s horn, for all to hear—still stand the still figures of the human and his sacrificed, because domesticated, because protected, animal, serviceably and companionably trained to each other’s conduct, animals in translation each to each.  The stakes are very high.  “Rebellion,” as every classic manifesto hopes, travels; it is never just local.  It travels by example; it travels on the wings of social media, of mediatized redistribution; “rebellion” travels in translation, immediately.  “Rebellion” is what I undertake on the streets of Cairo and translate to Tripoli, or Madrid, or London, or Gaza, when I stand with others against constituted or usurping authority, in hopes of achieving a state of equity, domestic tranquility, the restitution of my land, or perpetual peace.
 
But it is not only that.  “Rebellion,” Derrida is reminding us, is untranslatable on two poetic grounds.  Rebellion, in his analysis of Celan’s poem, is also the anagram for the name of the ram, le belier—and anagrammatical logic is at war with any translation.  It is just what cannot be translated from one natural language to another.  The rigid designators “ram” or “belier,” rigid inasmuch as they have, like the word Hesperus or the word Phosphorus, the same real referent, inasmuch as they refer to the same real animal (“real” here meaning “existing as the referent for a term,” not “really existing in the world”), these rigidly designating terms are translatable, always and universally translatable.  More, though: the word bélier is the anagrammatical name for that animal’s philosophico-political function as well: Bélier is an anagram for rebellion itself—and this anagrammaticality, lingering at the point of the letter, does not refer to “rebellion,” though it produces or reveals it; it has no relation, other than the circumstance that the word’s letters can be scrambled in French to produce a rebel from a ram and a ram out of a rebel, to the concept to which “rebellion” refers.  Anagrammatical effects are at war with standard semantics: they produce semantic effects accidentally; they submit reference to the rebellious wandering of letters.  The war between and within “ram,” “bélier,” and “rebellion,” and within and between the noun “bélier” and the verb “revelle,” is the re-enactment of, or return to, a state of war—a re-bellum.  Le bélier’s rebellion against translation reveals without repeating, produces concepts without representing or figuring or referring to them, sacrifices without substituting or substitutes without sacrifice.  La rébellion révèle: rebellion reveals an ancestral war waged not between “animal” and “human” or even within the languages we might use to number and negotiate their difference, but within the concept of their difference and within the name that seems to secure the identity of each.
 
An empiricist account of the animal in translation, grounded in the single term “stimulus” or in the rigid predication “there is stimulus,” permits the field linguist and the philosopher of language and their observers and readers to make practical assertions concerning descriptive predicates.  This empiricist account permits us to make assertions regarding the substitutability of identities organized in a concept that reveals or represents a generality to which their differences can be sacrificed.  The visual, vacuous unity of the stimulus, however, is inadequate to the injunction we find in Derrida’s account—the injunction not to “respect the other animal’s personhood” when that “respect” amounts to the sacrifice of that other animal to its “personhood”; of its rebellion to its proper name; of anagrammaticality to semantics; which is to say the sacrifice of that other animal’s difference from ourselves or from the class or concept to which we say it belongs, to a difference we take to be constitutive of ourselves as “animals” imbued with human “personhood” and self-identity.  The animal is never rigidly named, or immediately named, or named only with an eye to its domination in one field or under one aspect alone: the scene of naming is also the scene of the animal’s sacrifice in a different, plausible world, or in the same one.  The scene of the animal’s naming is also the scene of its mass-production and of its mass execution.  It is likewise the scene of its blank rebellion to the philosopher’s name and against any and all concepts; and the scene of its sacrifice.
 
I would rather not leave you with the unsatisfactory sense that my argument stands on scrambling letters against the grain of one or another natural language, and then claiming that this old particularity—as old, indeed, as Lucretius’s great poem—is definitive of rebellious translation as Derrida imagines it.  I am not, and Derrida is not, looking for war and rebellion in the alphabetical seeds of things, as De rerum natura might have it.  For Derrida, building out of Celan’s poem a rebellion against translation does not entail asserting that nothing can be translated: Béliers also shows its readers how to arrange, as one might arrange stars into a constellation, the defective concept of a rebellious translation.  This rebellious translation, which is both the rebellion against conceptual translation and the practice of translation itself, does not stand or fall on what we can see literally about words: this is not a matter of finding hypograms, words under words or of finding the untranslatable unwording of words that attaches to the instance of letters alone.  To show you what I mean, allow me to return to David Wills’s exemplary translation of Derrida’s “The Animal that Therefore I Am.”  You will recall the passage I cited earlier.  “It is true,” Wills’ translation runs, “that I identify [the animal before me] as a male or female cat.
 

But even before that identification, it comes to me [il vient à moi] as this irreplaceable living being that one day enters my space, into this place where it can encounter me, see me, even see me naked [me voir, voire me voir nu].  Nothing can ever rob me of the certainty that what we have here is an existence that refuses to be conceptualized [qu’il s’agit là d’une existence rebelle à tout concept]. (9)

I stressed the word “rebellion” in these lines, which Wills renders as the “refusal to be conceptualized.”  The difference between “rebellion” and “refusal” is not trivial, because rebellion, as we have seen, returns us not to the heroic scene of an individual who asserts claims against a coercive constituted power—non serviam!—but to a state of war, to a sort of Hobbesian war of all against all.  This is not trivial, but it is easily enough fixed, and changing “rebellion” into “refusal” or correcting “refusal” with “rebellion” tells us nothing about the concepts of translation, translatability, or untranslatability that Derrida is setting before us.
 
But note what happens just where the visualizability of the scene of the animal is presented for translation.  Here, Derrida’s French marks the language’s internal translation: the re-articulation of a French expression in French, which takes place before any identification occurs, an identification of the cat as a cat of one sort or another, even of it as a cat, but which seems also to be the condition on which the French language can be identified, that is to say, seen, as French.  “[E]ven before that identification,” Derrida writes, “it [the animal] comes to me [il vient à moi] as this irreplaceable living being that one day enters my space, into this place where it can encounter me, see me, even see me naked [me voir, voire me voir nu]” (9).  Wills’s phrase is “see me, even see me naked”; Derrida, however, is stressing the homophony between the French verb “to see,” voir, and the syntactic marker of internal translation, voire, which is the equivalent, in a different sensory register, of c’est à dire.  “To see me, to see me see naked” would be one way to render Derrida’s phrase; “to see me, that is, to see me naked” would be another.  This second marks the place where the animal is the mark of internal translation, the pause where French sees itself, sees itself for itself and as itself, sees itself naked.  The first translation of me voir, voire me voir nu places the philosopher’s vision on display, the vision of one particular human animal gazing at a non-human animal, seeing himself nakedly seeing.
 
The two scenes may not be confused: everything cries out against confusing the scene of intralinguistic mistranslation with the scene of viewing, and against confusing a scene granting primacy to the instant, the undifferentiated glimpse of the vision-stimulus, with a scene in which the function of vision is disseminated, like a constellation whose stars fly from the figure they represent.  Just here, where the animal comes to see and be seen before the philosophical animal, the scene’s language places the difference between the two scenes in the reader’s eye rather than his or her ear, in the silent letter “e” that keeps voir from becoming voire.  But this is no eye we have ever seen: it is, strangely, also the language’s eye (for itself).  The entrance of another sort of sight, another voir, right here, a voir meant to keep voir from becoming voire or vice versa, only exacerbates our difficulty, since it is not clear whether this voir(e) should come to us from the subjective phenomenology of reading or from the syntax of French.  It is this, we conclude, that renders this scene of the animal in translation finally, and most radically, translatable, inasmuch as natural languages, the languages human animals write and read and speak, share the possibility of producing semantic effects, as an anagrammatical accident—including the effect of self-reference, or what might, for the human animal writing, speaking, reading, seeing itself in and by means of such language, be called self-consciousness, the attribute classically reserved for the human animal.  It goes almost without saying that these effects of self-reference are also, inasmuch as they nest ineradicably in the particulars of one or another natural language, inasmuch as they are anagrammatical effects, quite untranslatable in content.  To track, to account for, to guard and reproduce in thought and in act this war between translatability and untranslatability, this war of and within mediation and within naming, is the injunction that Derrida’s late works lay upon us, lay beside the empiricist injunction that we derive from Quine, and Hearne, and Grandin.
 
I end with two provocations.  The ethical ground comported by these two simultaneous injunctions, the empiricist on one hand, and the rebellious, mediationist imperative on the other, rule-generating and entailing, each distinctly, a practice of rule following–this divided ethical ground is the only one available to us when we seek to take account of those other animals, all those other animals, from which we draw sustenance and political life.  As for our political life: when we undertake to translate rebellion from the streets of Gaza to the streets of New York, from Madrid to Brussels, from Beijing to Cairo, we do so not only as human animals, bearers of natural or universal rights, dignities, languages, historical, biological or biopolitical qualities, but also according to the accidental rules that dilapidate those concepts.  Animals in translation: political rebellion is never just an attribute of human agents.