Norton, 1977), 17: "What I have called paranoic
knowledge is shown, therefore, to correspond in its more or less archaic forms to certain critical moments that mark the history of man's mental genesis, each repre- senting a stage in objectifying identification.
knowledge is shown, therefore, to correspond in its more or less archaic forms to certain critical moments that mark the history of man's mental genesis, each repre- senting a stage in objectifying identification.
Paul-de-Man-Material-Events
Of course, Webster's does offer a second meaning for the word context, "associated surroundings, whether material or mental"--a reference not to the surrounding text but to the broader reality or intentionality--but Souter dismisses this by saying, "we doubt that the broader sense applies here.
" Why?
Because "if Congress had meant to point further afield, as to legislative history, for example, it would have been natural to use a more spacious phrase, like 'evi- dence of congressional intent,' in place of 'context.
'"
The word natural, which is precisely at issue here, since we are try- ing to find out whether the statute applies only to natural persons, is here applied precisely to an artificial person, Congress, which is per- sonified as having natural intentionality. "If Congress had meant . . . " The Court's decision repeatedly relies on this type of personification: it is as though Souter has to treat Congress as an entity with intentions, even natural intentions, in order to say that Congress could not have meant to include artificial entities in its ruling. There is a personifica- tion of an artificial entity, Congress, embedded in the very project of interpreting how far the law will allow for artificial entities to be con- sidered persons.
Turning to the Dictionary Act for person and to Webster's diction- ary for context, Souter also notes that he has to define indicates. The
Anthropomorphism in Lyric and Law 215
216 Barbara Johnson
difficulty of doing so pushes him into a volley of rhetorical flourishes: "A contrary 'indication' may raise a specter short of inanity, and with something less than syllogistic force. " "Indicates," it seems, means more than nonsense but less than logical necessity. In other words, the task of reading becomes an infinite regress of glossing terms that are themselves supposed to be determinants of meaning. De Man's linguis- tic stutter returns here as the repeated effort to throw language outside itself. We could read a text, this implies, if only we were sure of the meaning of the words context and indicate. But those are precisely the words that raise the question of meaning in its most general form-- they cannot be glossed with any finality because they name the process of glossing itself.
Souter's text, in fact, is most anthropomorphic at those points where the infinite regress of language is most threatening. Congress is endowed with "natural" intentionality in order to sweep away the abyss of reference. Souter's dismissal of the prisoners' association as an "amorphous legal creature" is the counterpart to the need to reinforce the anthropomorphizability of the artificial legal creature, Congress. 17
Souter's opinion proceeds to detail the ways in which he thinks the in forma pauperis ruling should only apply to natural persons. If an af- fidavit alleging poverty is required for a person to proceed in forma pauperis, then can an artificial entity plead poverty? Souter again turns to Webster's dictionary to find that poverty is a human condition, to be "wanting in material riches or goods; lacking in the comforts of life; needy. " Souter also refers to a previous ruling, which holds that pover- ty involves being unable to provide for the "necessities of life. " It is as though only natural persons can have "life," and that life is defined as the capacity to lack necessities and comforts. "Artificial entities may be insolvent," writes Souter, "but they are not well spoken of as 'poor. '" An artificial entity cannot lack the necessities and comforts of life. Only life can lack. The experience of lack differentiates natural persons from artificial persons. To lack is to be human. In a sense, we have returned to de Man's question about mourning. Is lack human, or just a struc- ture? Whatever the case, the Court holds that associations cannot be considered persons for the purpose of the in forma pauperis procedure.
The majority was only five to four, however. In a dissenting opinion, written by Clarence Thomas, it is argued that there is no reason to re- strict the broad definition of "person" to natural persons in this case. Thomas quotes the Court's view of "poverty" as an exclusively "human condition," and comments:
I am not so sure. "Poverty" may well be a human condition in its "pri- mary sense," but I doubt that using the word in connection with an arti- ficial entity departs in any significant way from settled principles of English usage. . . . Congress itself has used the word "poor" to describe entities other than natural persons, referring in at least two provisions of the United States Code to the world's "Poorest countries"--a term that is used as a synonym for the least developed of the so-called "devel- oping" countries.
Souter has glossed the word poor as though speakers of English could only use it literally. Thomas responds by including the figurative use of poor as included within normal usage. The boundaries between natu- ral persons and artificial persons cannot be determined by usage, be- cause those boundaries have always already been blurred. In treating Congress as an entity with natural intentions, indeed, Souter has al- ready shown how "natural" the artificial can be.
At another point, Thomas takes issue with Souter's discussion of a case in which an association or corporation is considered a person de- spite strong contextual indicators to the contrary. In the case of Wilson v. Omaha Indian Tribe, 442 U. S. 653, 666 (1979), it was decided that "white person" could include corporations because the "larger con- text" and "purpose" of the law was to protect Indians against non- Indian squatters, and would be frustrated if a "white person" could simply incorporate in order to escape the provision of the law. Souter admits that "because a wholly legal creature has no color, and belongs to no race, the use of the adjective 'white' to describe a 'person' is one of the strongest contextual indicators imaginable that 'person' covers only individuals. " Justice Thomas argues that if the Court "was cor- rect in holding that the statutory term 'white person' includes a corpo- ration (because the 'context' does not 'indicate otherwise')--the con- clusion that an association is a 'person' for in forma pauperis purposes is inescapable. " Perhaps another inescapable conclusion is that despite its apparent reference to the physical body, the phrase "white person" is the name, not of a natural, but of a corporate person.
Justice Thomas refutes the reasons Souter has given for finding that artificial entities are excluded from the in forma pauperis provision, noting that there may be sound policy reasons for wanting to exclude them, but that the law as written cannot be construed to have done so. The Court's job, he writes, is not to make policy but to interpret a statute. "Congress has created a rule of statutory construction (an
Anthropomorphism in Lyric and Law 217
218 Barbara Johnson
association is a 'person') and an exception to that rule (an association is not a 'person' if the 'context indicates otherwise'), but the Court has permitted the exception to devour the rule [a nice personification]" (treating the rule as if artificial entities were excluded rather than in- cluded unless the context indicates otherwise). "Whatever 'unless the context indicates otherwise' means," writes Thomas, "it cannot mean 'unless there are sound policy reasons for concluding otherwise. '"
Permitting artificial entities to proceed in forma pauperis may be unwise, and it may be an inefficient use of the government's limited re- sources, but I see nothing in the text of the in forma pauperis statute indicating that Congress has chosen to exclude such entities from the benefits of that law.
Thus, Thomas's two conservative instincts are at war with each other: he would like the government not to spend its money, but he would also like to stick to the letter of the law.
The question of what counts as a juridical person has, in fact, been modified over time in the legal code. It was in 1871 (significantly, per- haps, at the beginning of the end of post-Civil War Reconstruction) that the so-called Dictionary Act was first passed by Congress, in which the word Person "may extend and be applied to bodies politic and corporate. " More recently, the question of fetal personhood has been debated, not only in the Roe v. Wade decision, where it was de- cided that a fetus was not a legal person, but also in Weaks v. Mounter, 88 Nev. 118, where it was decided that a fetus was a person who could sue for intrauterine injuries, but only after birth. Recently, the question of granting patents for forms of life such as oil-slick-eating bacteria or genetically altered mice has raised the question of whether a hybrid be- tween humans and close animal relatives can be patented. And also, of course, the question of the ethics and legality of cloning humans has been raised. The law has reached another crisis about the definition of "person. " In an article on constitutional personhood, Michael Rivard writes:
Current law allows patents for genetically-engineered animals but not for human beings. Humans are not patentable subject matter because patents are property rights, and the Thirteenth Amendment forbids any grant of property rights in a human being. Nevertheless, this exclusion for humans will prove impossible to maintain: within ten to thirty years, or perhaps sooner, advances in genetic engineering technology should allow scientists to intermingle the genetic material of humans and ani-
mals to produce human-animal hybrids. . . . It may soon be possible to patent--and to enslave--human-animal hybrids who think and feel like humans, but who lack constitutional protection under the Thirteenth Amendment. 18
The Thirteenth Amendment is the amendment that abolishes slavery. The constitutional protection against slavery operates as a constraint on the patent office, but it does so in a paradoxical way. The fear of re- instituting something like slavery, or property in humans, is a reaction to, but also a sign of, what must be an ongoing research goal to come as close as possible to creating the ownable, enslavable human. 19
Constitutional personhood has in fact often been defined in prox- imity to slavery. The contradiction between equal rights and chattel slavery led from the beginning to verbal gymnastics, even in the draft- ing of the Constitution itself. By not using the word slavery in the Constitution, and by revising the text of the original fugitive slave clause to refer to the legality of slavery only on the level of the states rather than of the federal government, the framers built a double inten- tionality into the very foundation of their law. Douglas Fehrenbacher, studying the egregious understanding of original intent later employed by the Supreme Court in the case of Dred Scott v. Sanford, writes of the Constitution: "It is as though the framers were half-consciously trying to frame two constitutions, one for their own time and the other for the ages, with slavery viewed bifocally--that is, plainly visible at their feet, but disappearing when they lifted their eyes. "20 A written text of law can thus contain a double intention, the trace of a compro- mise between differing opinions. No wonder interpreting the law's in- tention is so complicated. That intention can always already be mul- tiple. The distinction Justice Thomas made between interpreting the law and making policy cannot hold if the law's ambiguity allows for the possibility that the policy it governs will change.
III
The "inhuman" is not some kind of mystery, or some kind of secret; the in- human is: linguistic structures, the play of linguistic tensions, linguistic events that occur, possibilities which are inherent in language--independently of any intent or any drive or any wish or any desire we might have. . . . If one speaks of the inhuman, the fundamental non-human character of language, one also speaks of the fundamental non-definition of the human as such.
--Paul de Man, "Benjamin's 'The Task of the Translator'"
Anthropomorphism in Lyric and Law 219
220 Barbara Johnson
Only smoking distinguishes humans from the rest of the animals.
--Anonymous (quoted in Richard Klein, Cigarettes Are Sublime)
The case of Rowland v. California Men's Colony, Unit II Men's Advisory Council was ostensibly about whether a council of inmates could sue prison officials in forma pauperis to get their cigarettes back. The de- tails of the case seemed irrelevant to the question of whether an artifi- cial person has the right to sue in forma pauperis. Yet perhaps some of those details deserve note. Is it relevant that the suit to decide this ques- tion is brought by a council of inmates? The phenomenon of the in- mate civil suit has grown to the point where the case law may very well be transformed by it. In a 1995 study of inmate suits in California, it was reported that "For the last fourteen years at least, the federal courts have faced a growing caseload and workload challenge posed by inmate cases. . . . By 1992, these filings numbered nearly 30,000, and constituted 13% of the courts' total civil case filings nation- wide. "21 The majority of these suits are filed in forma pauperis. 22 The Supreme Court's decision may well have been affected by what Clarence Thomas calls "policy decisions. "
If prisoners are affecting the nature of civil proceedings, they are also, at least figuratively, affecting theoretical discussions about the na- ture of rational choice and the evolution of cooperation. The celebrat- ed "Prisoner's Dilemma" has been central to questions of self-interest and social goods since it was introduced by Albert Tucker in 1950. Max Black has even entitled his discussion of the issues raised "The 'Prisoner's Dilemma' and the Limits of Rationality. "23 Why is it that the theoretical study of rational choice has recourse to "man" con- ceived as a prisoner? Does this have anything to do with the poets' ten- dency to see the sonnet form as a prison?
And is it by chance that Rowland v. California Men's Colony, Unit II Men's Advisory Council is about cigarettes? On the one hand, it seems paradoxical that the council has to demonstrate its indigence in order to pursue its suit against the prison directors for depriving the prisoners of access to cigarettes, which in prisons function as a form of currency. On the other hand, it seems fitting that the personhood of the association is the counterpart to the humanity of the inmates, which, as common wisdom (quoted above, second epigraph) would have it, is demonstrated by the act of smoking. The prisoners would thus, in a very attenuated way, be suing for their humanity. As Richard Klein has wittily shown, smoking serves no function other than to enact a struc-
ture of desire--of human desire for self-transcendence, for repetition, for bodily experience corresponding to something other than the "ne- cessities of life" required for existence alone: in short, desire for the sublime. 24 Far from being what defines natural personhood, then, need for the "necessities of life" alone is precisely what cannot define the human.
In the article cited earlier, Rivard declares that "corporations would be presumed constitutional nonpersons," especially for liberty-related rights, unless the corporation could rebut its nonperson status by showing specific natural persons "who would be affected if the corpo- ration were denied these rights. "25 This is the opposite of the Dictionary Act, which considers a corporation a person "unless the context in- dicates otherwise. " Rivard's article is arguing for the rights of new bio- logical species who can pass the "self-awareness test" (which, in a surpisingly Lacanian move derived from Michael Dennett, he defines as wanting to be different from what one is), and he claims that corpo- rations, by their nature, do not pass this test.
But the question of the nature of corporations as persons has never been a simple one, as Rivard admits. In an article titled "The Personifi- cation of the Business Corporation in American Law" (University of Chicago Law Review 54 [fall 1987]: 1441), Gregory A. Marks out- lines in detail the history of corporate personhood. The relation be- tween corporations and the natural persons who compose them has grown more complicated over time, but in most discussions of the mat- ter, it is the "natural" person that functions as the known quantity, and the "artificial" who is either just an "aggregate" of natural persons, or a fiction created by the state, or a mere metaphor, or actually resembles (is like, to return to the Baudelairean word) a natural person in that it has a "will" of its own. Such a corporate will is a form of agency sepa- rate from that of the natural corporators, who exist behind the "veil" of the corporation. Much of Marks's article concerns the exact rhetori- cal valence of this personification:
American law has always recognized that people's activities could be for- mally organized and that the resulting organizations could be dealt with as units. Personification, however, is important because it became far more than a quaint device making it possible for the law to deal with or- ganized business entities. In American legal and economic history, personi- fication has been vital because it (1) implies a single and unitary source of control over the collective property of the corporation's members,
Anthropomorphism in Lyric and Law 221
222 Barbara Johnson
(2) defines, encourages, and legitimates the corporation as an autono- mous, creative, self-directed economic being, and (3) captures rights, ulti- mately even constitutional rights, for corporations thereby giving corpo- rate property unprecedented protection from the state. (1443)
Marks takes seriously the role of language in the evolving history of the corporation. Philosophers and legislators have gone to great lengths to minimize the rhetorical damage, to eliminate personifi- cation as far as possible, but he asserts that it is not just a figure of speech to speak of a corporation's "mind," or even its "life. " "Prac- tical experience, not just anthropomorphism, fixed the corporate mind in the management hierarchy" (1475). The corporation resembled a human being in its capacity to "take resolves in the midst of conflict- ing motives," to "will change. " Yet the analogy is not perfect. The cor- poration, for example, unlike its corporators, is potentially immortal. The effect of personification appears to derive its rhetorical force from the ways in which the corporation resembles a natural person, yet the corporation's immortality in no way diminishes its personification. When Marks says that it is "not just anthropomorphism" that under- pins the agency of the corporation, he still implies that we can know what anthropomorphism is. But his final sentence stands this presup- position on its head. Far from claiming that a corporation's character- istics are derived from a knowable human essence, Marks suggests that what have been claimed to be the essential characteristics of man (es- pecially "economic man") have in fact been borrowed from the nature of the corporation:
Personification with its roots in historic theological disputes and mod- ern business necessity, had proved to be a potent symbol to legitimate the autonomous business corporation and its management. Private property rights had been transferred to associations, associations had themselves become politically legitimate, and the combination had helped foster modern political economy. The corporation, once the de- rivative tool of the state, had become its rival, and the successes of the autonomous corporate management turned the basis for belief in an in- dividualist conception of property on its head. The protests of modern legists notwithstanding, the business corporation had become the quint- essential economic man. (1482-83)
Theories of rationality, naturalness, and the "good," presumed to be grounded in the nature of "man," may in reality be taking their no-
tions of human essence not from "natural man" but from business corporations.
Ambivalence about personification, especially the personification of abstractions, has in fact permeated not only legal but also literary his- tory. Nervousness about the agency of the personified corporation echoes the nervousness Enlightenment writers felt about the personifi- cations dreamed up by the poets. As Steven Knapp puts it in his book Personification and the Sublime:
Allegorical personification--the endowing of metaphors with the agency of literal persons--was only the most obvious and extravagant instance of what Enlightenment writers perceived, with a mixture of ad- miration and uneasiness, as the unique ability of poetic genius to give the force of literal reality to figurative "inventions. " More important than the incongruous presence of such agents was their contagious ef- fect on the ostensibly literal agents with which they interacted. 26
The uncanniness of the personification, then, was derived from its way of putting in question what the "natural" or the "literal" might be.
We have finally come back to the question of whether there is a dif- ference between anthropomorphism and personification, which arose at the end of the discussion of the essay by Paul de Man. It can now be seen that everything hangs on this question. Not only does anthro- pomorphism depend on the givenness of the essence of the human and personification does not, but the mingling of personifications on the same footing as "real" agents threatens to make the lack of certainty about what humanness is come to consciousness. Perhaps the loss of unconsciousness about the lack of humanness is what de Man was calling "true 'mourning. '" Perhaps the "fallacious lyrical reading of the unintelligible" was exactly what legislators count on lyric poetry to provide: the assumption that the human has been or can be defined so that it can then be presupposed without the question of its defini- tion's being raised as a question--legal or otherwise. Thus the poets would truly be, as Shelley claimed, the "unacknowledged legislators of the world," not because they covertly determine policy, but because it is somehow necessary and useful that there be a powerful, presup- posable, unacknowledgment. But the very rhetorical sleight of hand that would instate such an unacknowledgment is indistinguishable from the rhetorical structure that would empty it. Lyric and law are two of the most powerful discourses that exist along the fault line of this question.
Anthropomorphism in Lyric and Law 223
224 Barbara Johnson
NOTES
1. I am thinking of Richard Posner's Law and Literature (Cambridge: Harvard University Press, 1988), Richard Weisberg's The Failure of the Word (New Haven: Yale University Press, 1984), and Peter Brooks, Troubling Confessions (Chicago: University of Chicago Press, 2000). But for a legal approach that does address poetry, see the interesting discussion of Wallace Stevens by Thomas Grey and Margaret Jane Radin in the Yale Journal of Law & The Humanities 2:2 (summer 1990), as well as the more extended treatment of Wallace Stevens in Thomas Grey, The Wallace Stevens Case: Law and the Practice of Poetry (Cambridge: Harvard University Press, 1991).
2. William Wordsworth's sonnet, "Nuns Fret Not at Their Convent's Narrow Room," contains the lines, "In truth the prison, into which we doom / Ourselves, no prison is: and hence for me, / In sundry moods, 'twas pastime to be bound / Within the Sonnet's scanty plot of ground" (Selected Poetry and Prose of Words- worth [New York: Signet, 1970], 169).
3. John Keats's sonnet on the sonnet begins, "If by dull rhymes our English must be chained, / And, like Andromeda, the sonnet sweet / Fettered . . . " (The Selected Poetry of Keats [New York: Signet, 1966], 264).
4. One of several poems by Baudelaire titled Spleen describes a mood pro- duced by or analogized to a rainy day: "Quand la pluie e? talant ses immenses traine? es / D'une vaste prison imite les barreaux . . . " (Baudelaire, OEuvres com- ple`tes, vol. 1 [Paris: Ple? iade, 1975], 75).
5. For a suggestive discussion of what it means for a text to obey the law of genre, see Jacques Derrida, "The Law of Genre," in Acts of Literature, ed. Derek Attridge (New York: Routledge, 1992).
6. Paul de Man, "Anthropomorphism and Trope in the Lyric," in The Rhetoric of Romanticism (New York: Columbia University Press, 1984). Page numbers in parentheses refer to this essay.
7. United States Law Week 61:25 (January 12, 1993). Page numbers in paren- theses refer to this text.
8. This allusion to Keats's "Ode on a Grecian Urn" stands in for the premise of the compatibility of literary aesthetics with linguistic structures, and of linguistic structures with perceptual or intuitive knowledge, that de Man is often at pains to contest. See his remarks on the pedagogical model of the trivium in the titular essay of The Resistance to Theory (Minneapolis: University of Minnesota Press, 1986).
9. Friedrich Nietzsche, "Truth and Falsity in an Ultramoral Sense," in Critical Theory Since Plato, ed. Hazard Adams (Fort Worth: Harcourt Brace Jovanovich, 1992), 634-39. If the Keats poem stands as the claim that aesthetic and epistemo- logical structures are compatible, Nietzsche's text, for de Man, stands as a parody of that claim.
10. The translations are mine, for the purpose of bringing out those aspects of the poems that are relevant to my discussion.
11. "Allegory and Irony in Baudelaire," in Romanticism and Contemporary Criticism (Baltimore: Johns Hopkins University Press, 1993). This essay is part of the Gauss Seminar given by de Man in 1967.
12. Jacques Lacan, "Aggressivity in Psychoanalysis," in E? crits: A Selection, trans. Alan Sheridan (New York: W. W.
Norton, 1977), 17: "What I have called paranoic
knowledge is shown, therefore, to correspond in its more or less archaic forms to certain critical moments that mark the history of man's mental genesis, each repre- senting a stage in objectifying identification. "
13. Ibid. , 9.
14. Jacques Derrida, "Freud's Legacy," in The Postcard, trans. Alan Bass (Chicago: University of Chicago Press, 1987).
15. De Man, "The Resistance to Theory," 10.
16. United States Code (1994 edition), vol. 15, 438.
17. In a response to the present paper given at the Yale Law School, Shoshana
Felman made the brilliant suggestion that Souter would have wanted to rewrite Baudelaire's "Correspondances" as: "Le Congre`s est un temple ou` de vivants pil- liers laissent parfois sortir de confuses paroles . . . " The neoclassical, Parnassian ar- chitecture of official Washington, D. C. , and the common metaphorical expression "pillars of the community," add piquancy to this suggestion.
18. Michael D. Rivard, "Toward a General Theory of Constitutional Person- hood: A Theory of Constitutional Personhood for Transgenic Humanoid Species," UCLA Law Review 39: 5 (June 1992): 1428-29.
19. See A. Leon Higginbotham Jr. and Barbara Kopytoff, "Property First, Humanity Second: The Recognition of the Slave's Human Nature in Virginia Civil Law," Ohio State Law Journal 50:3 (June 1989): "The humanity of the slave, re- quiring that he be treated with the care due other humans and not like other forms of property, became part of the owner's property rights" (520).
20. Douglas E. Fehrenbacher, Slavery, Law, and Politics: The Dred Scott Case in Historical Perspective (Oxford: Oxford University Press, 1981), 15.
21. Kim Mueller, "Inmates' Civil Rights Cases and the Federal Courts: Insights Derived from a Field Research Project in the Eastern District of California," Creighton Law Review 1228 (June 1995): 1258-59. In the Eastern District of Cali- fornia, inmates' civil rights actions constituted nearly 30 percent of the case filings. (California Men's Colony is not in the Eastern District; it is in San Luis Obispo. )
22. Ibid. , 1276 and 1281.
23. Max Black, Perplexities (Ithaca, N. Y. : Cornell University Press, 1990). See also Robert Axelrod, The Evolution of Cooperation (New York: Basic Books, 1984).
24. Richard Klein, Cigarettes Are Sublime (Durham, N. C. : Duke University Press, 1993). Klein notes, incidentally, that Baudelaire is one of the first French writers to use the word cigarette in print (in his "Salons de 1848," 8).
25. Rivard, "Toward a General Theory of Constitutional Personhood," 1501-2.
26. Steven Knapp, Personification and the Sublime (Cambridge: Harvard University Press, 1985), 2.
Anthropomorphism in Lyric and Law 225
This page intentionally left blank
IV. The Mnemopolitical Event
This page intentionally left blank
The Politics of Rhetoric
Ernesto Laclau
Why would a political theorist like me, working mainly on the role of hegemonic logics in the structuration of political spaces, be interested in the work of a prominent literary critic such as Paul de Man? I could suggest at least two main reasons. The first is that one of the leitmotifs of Paul de Man's work has been the subversion of the frontiers separat- ing theoretical from literary disciplines, so that those dimensions that had traditionally been conceived as privative of literary or aesthetic language became, for him, actually defining features of language tout court. Against all attempts to differentiate between "appearance" and "saying," between a primary text whose message would have been me- diated by the materiality of the signs, of the figural, and a language of inquiry governed by reason, de Man had always insisted that any lan- guage, whether aesthetic or theoretical, is governed by the materiality of the signifier, by a rhetorical milieu that ultimately dissolves the illu- sion of any unmediated reference. In this sense a generalized rhetoric-- which necessarily includes within itself the performative dimension-- transcends all regional boundaries and becomes coterminous with the structuration of social life itself. Conceived at such a broad level of generality, the literariness of the literary text breaks the limits of any specialized discipline and its analysis involves something like the study of the distorting effects that representation exercises over any reference--effects that become, thus, constitutive of any experience.
Moreover, de Man himself was perfectly aware of the political and ideological implications of his approach to texts. In a famous interview with Stefano Rosso that discusses the increasing recurrence in his works of the terms political and ideological, he answered as follows:
229
230 Ernesto Laclau
I don't think I ever was away from these problems, they were always uppermost in my mind. I have always maintained that one could ap- proach the problems of ideology and by extension the problems of poli- tics only on the basis of critical-linguistic analysis, which had to be done on its own terms, in the medium of language, and I felt that I could ap- proach those problems only after having achieved a certain control over those questions. It seems pretentious to say so, but it is not the case. I have the feeling I have achieved a certain control over technical prob- lems of language, specifically problems of rhetoric, of the relation be- tween tropes and performatives, of saturation of tropology as a field that in certain forms of language goes beyond that field. . . . I feel now some control of a vocabulary and of a conceptual apparatus that can handle that. 1
As for the second reason for a political theorist to be interested in de Man's work, it has to do with something related to the political field itself. Gone are the times in which the transparency of social actors, of processes of representation, even of the presumed underlying logics of the social fabric could be accepted unproblematically. On the contrary, each political institution, each category of political analysis shows itself today as the locus of undecidable language games. The over- determined nature of all political difference or identity opens the space for a generalized tropological movement and thus reveals the fruitful- ness of de Man's intellectual project for ideological and political analy- sis. In my work, this generalized politico-tropological movement has been called "hegemony. " I intend in this essay to stress some decisive points in the work of de Man, especially in his late work, where the di- rection of his thought could be helpful in developing a hegemonic ap- proach to politics.
I
The requirements of "hegemony" as a central category of political analysis are essentially three. First, that something constitutively het- erogeneous to the social system or structure has to be present in the lat- ter from the very beginning, preventing it from constituting itself as a closed or representable totality. If such a closure were achievable, no hegemonic event could be possible and the political, far from being an ontological dimension of the social--an "existential" of the social-- would just be an ontic dimension of the latter. Second, however, the hegemonic suture has to produce a retotalizing effect, without which
no hegemonic articulation would be possible either. But, third, this re- totalization cannot have the character of a dialectical reintegration. It has, on the contrary, to maintain alive and visible the original and con- stitutive heterogeneity from which the hegemonic articulation started. How is a logic that can maintain these two contradictory requirements at the same time possible? Let us approach this question through the exploration of its possible presence in de Man's texts. We will start from the analysis of Pascal's Re? flexions sur la ge? ome? trie en ge? ne? ral; De l'esprit ge? ome? trique et de l'Art de persuader that de Man carries out in "Pascal's Allegory of Persuasion. "2
Pascal starts his study of the esprit ge? ome? trique from the distinction between nominal and real definitions--the first resulting from con- vention and being thus exempt from contradiction, the second being axioms or propositions to be proved--and asserts that the confusion between the two is the main cause of philosophical difficulties. Main- taining the separation between the two--as the geometrician does--is the first rule of philosophical clarity. However, the argument runs quickly into difficulties, as geometrical discourse includes not only nominal definitions but also "primitive terms"--such as motion, num- ber, and extension--which are undefinable but, nonetheless, fully intel- ligible. According to Pascal, these undefinable words find a universal reference not in the (impossible) fact that all men have the same idea concerning their essence, but instead, in the fact that there is a relation of reference between name and thing, "so that on hearing the expres- sion time, all turn (or direct) the mind to the same entity" (56). But, as de Man shows, this brings back the real definition into the geometrical camp itself, for
the word does not function as a sign or a name, as was the case in the nominal definition, but as a vector, as a directional motion that is mani- fest only as a turn, since the target toward which it turns remains un- known. In other words, the sign has become a trope, a substitutive rela- tionship that has to posit a meaning whose existence cannot be verified, but that confers upon the sign an unavoidable signifying function. (56)
As the semantic function of the primitive terms has the structure of a trope, "it acquires a signifying function that it controls neither in its ex- istence nor in its direction. " Ergo, "[s]ince definition is now itself a primitive term, it follows that the definition of the nominal definition is itself a real, and not a nominal, definition" (57).
This contamination of the nominal by the real definition is still
The Politics of Rhetoric 231
232 Ernesto Laclau
more visible when we move to the question of double infinitude, which is decisive in establishing the coherence and intelligibility of the rela- tionship between mind and cosmos. Here, Pascal deals with the objec- tions put to him by the Chevalier de Me? re? , according to whom--given the Pascalian principle of homogeneity between space and number--it is possible to conceive an extension formed by parts that are not them- selves extended, since it is possible to have numbers made up of units that are devoid of number. With this, the principle of infinite smallness would be put into question. Pascal's answer has two steps. He asserts, in the first place, that which applies to the order of number does not apply to the order of space. One is not a number, there is no plurality in it; but at the same time it belongs to the order of number for, given the Euclidean principle of homogeneity ("magnitudes are said to be of the same kind or species when one magnitude can be made to exceed another by reiterated multiplication"), it is part of the infinity postu- lated by that principle. On that basis, Pascal can distinguish between number and extension, but only at the price of grounding the distinc- tion in real and not nominal definitions. As de Man asserts:
The synecdochal totalization of infinitude is possible because the unit of number, the one, functions as a nominal definition. But, for the argument to be valid, the nominally indivisible number must be distinguished from the really indivisible space, a demonstration that Pascal can ac- complish easily, but only because the key words of the demonstration-- indivisible, spatial extension (e? tendue), species (genre), and definition-- function as real, and not as nominal, definitions. (58-59)
But--second step--if the order of number and the order of extension had to be separated to answer Me? re? 's objection, the rift between the two had also to be healed if the principle of homogeneity between both was to be maintained. This homology is restored by appealing, as far as number is concerned, to the zero--which, on the difference with one is radically heterogeneous with the order of number--and by finding equivalences in the order of time and motion, such as "instant" and "stasis. " This appeal to the zero, however, has dramatic consequences for the coherence of the system, which de Man describes in a passage worth quoting in full:
the coherence of the system is now seen to be entirely dependent on the introduction of an element--the zero and its equivalences in time and motion--that is itself entirely heterogeneous with regard to the system
and is nowhere part of it. . . . Moreover, this rupture of the infinitesimal and the homogeneous does not occur on the transcendental level, but on the level of language, in the inability of a theory of language as sign or as name (nominal definition) to ground this homogeneity without hav- ing recourse to the signifying function, the real definition, that makes the zero of signification the necessary condition for grounded knowl- edge. . . . It is as sign that language is capable of engendering the prin- ciples of infinity, of genus, species, and homogeneity, which allow for synecdochal totalizations, but none of these tropes could come about without the systematic effacement of the zero and its reconversion into a name. There can be no one without zero, but the zero always appears in the guise of a one, of a (some)thing. The name is the trope of the zero. The zero is always called a one, when the zero is actually nameless, "in- nommable. " In the French language, as used by Pascal and his inter- preters, this happens concretely in the confusedly alternate use of the two terms ze? ro and ne? ant. The verbal, predicative form ne? ant, with its gerundive ending, indicates not the zero, but rather the one, as the limit of the infinitely small, the almost zero that is the one. (59)
It is important to give serious consideration to this remarkable passage-- remarkable, among other things, because de Man does not pursue later in his essay all the implications of his own de? marche--for it contains, in nuce, all the relevant dimensions of the problem we are exploring. Everything turns around the role of the zero. The zero is, we are told, something radically heterogeneous with the order of number. The order of number, however, cannot constitute itself without reference to the zero. It is, in this sense, a supplement to the system that, none- theless, is necessary for constituting the latter. Vis-a`-vis the system, the zero is in an undecidable tension between internality and externality-- but an internality that does not exclude heterogeneity. The zero, in the second place is "innommable," nameless, but at the same time it pro- duces effects, it closes the system, even at the price of making it hope- lessly heterogeneous. It retotalizes the system, incurring, however, an inconsistency that cannot be overcome. The zero is nothing, but it is the nothing of the system itself, the impossibility of its consistent clo- sure, which is signified by the zero and in that sense, paradoxically, the zero as empty place becomes the signifier of fullness, of systematicity as such, as that which is lacking. The semantic oscillation between ze? ro and ne? ant that de Man observes is the result of this dual condition of the moment of closure: being an impossible object, but also a necessary
The Politics of Rhetoric 233
234 Ernesto Laclau
one. Finally, if the zero as moment of closure is impossible as an object but also necessary, it will have to have access to the field of represen- tation. But the means of representation will be constitutively inade- quate. It will give to the "innommable" a body, a name, but this can be done only at the price of betraying its true "nonbeing"; thus the tropological movement that prolongs sine die the non-resolvable dia- lectics between the zero and the one. In the words of de Man just quoted: "There can be no one without zero, but the zero always ap- pears in the guise of a one, of a (some)thing. The name is the trope of the zero. The zero is always called a one, when the zero is actually nameless, 'innommable. '"
Now, this succession of structural moments coincides, almost step by step, with the logic of hegemony that I have tried to describe in my work and which I see operating in the texts of Gramsci--to whom I will return later on. To start with, the condition of any hegemonic su- ture is the constitutive nonclosure of a system of political signification. The systemacity of a system, its closure--which is the condition of sig- nification in a system, such as Saussure's, whose identities are merely differential--coincides with the determination of its limits. These lim- its, however, can only be dictated by something that is beyond them. But, as the system is a system of differences, of all actual differences, that "beyond"--which should be heterogeneous with the system in order to fulfill its function of truly closing it--lacks the condition of a true heterogeneity if it consists in one more difference. The latter would be, in some way, undecided, suspended between being internal and external to the system. This jeopardizes the role of the "beyond" as limit and, as a result, the possibility of constituting the differences as truly intrasystemic differences. It is only if the beyond the limit has the character of an exclusion that its role as limit is restored and with it the possibility of emergence of a full system of differences. 3
However, this fullness of the system (obtained, it is true, at the price of a dialectical retrieval of its negation) has a shortcoming. For, vis-a`- vis the excluded element, all differences within the system establish relations of equivalence between themselves. And equivalence is pre- cisely that which subverts difference. So the "beyond" which is the condition of possibility of the system is also its condition of impossi- bility. All identity is constituted within the unsolvable tension between equivalence and difference.
So, as with the Pascalian zero, we arrive at an object that is at the same time impossible and necessary. As impossible, it is an empty place
within the structure. But, as necessary, it is a "nothing" that will pro- duce structural effects, and this requires that it has access to the field of representation. And, as in the dialectics between the zero and the one, this double condition of necessity and impossibility will be constitu- tively inadequate. The fullness of the system, its point of imaginary satu- ration, will be, as in the example of de Man, a nothing that becomes a something. What are the possible means of this distorted representa- tion? Only the particular differences internal to the system. Now, this relationship by which a particular difference takes up the representa- tion of an impossible totality entirely incommensurable with it is what I call a hegemonic relation.
There are only two differences between the hegemonic logic and the Pascalian dialectics between the zero and the one as described by de Man. The first is that--given the numerical nature of Pascal's case-- the zero can only be embodied by the one, while in the case of the hegemonic logic any element within the system can be a bearer of a hegemonic function. 4 The second difference is that, given de Man's in- terests, the determination of the heterogeneous character of the tropo- logical displacement from the zero to the one is the point of arrival of his analysis, whereas for a student of hegemonic logics the analysis of the exact nature of this tropological movement becomes imperative. In de Man's detotalizing discourse, what matters is to show the hetero- geneity out of which the tropological movement operates. This is also a vital part of a hegemonic analysis. But what is decisive for the latter is the determination of the partial retotalizations that the tropological movement makes possible. 5 This is the dimension that we now have to take into account. I will do it through a reference to the opposition metaphor/metonymy as presented in de Man's essay on Proust in Allegories of Reading. 6
II
The text on Proust deals, as is well known, with the discourse of young Marcel on the pleasure of reading and with the way in which such pleasure is constructed through a set of metaphoric substitutions that are, however, persuasive only through the operation of a series of con- tingent metonymic movements. De Man asserts:
The crossing of sensory attributes in synaesthesia is only a special case of a more general pattern of substitution that all tropes have in com- mon. It is the result of an exchange of properties made possible by a
The Politics of Rhetoric 235
236 Ernesto Laclau
proximity or an analogy so close and intimate that it allows the one to substitute for the other without revealing the difference necessarily in- troduced by the substitution. The relational link between the two enti- ties involved in the exchange then becomes so strong that it can be called necessary: there could be no summer without flies; no flies with- out summer. . . . The synecdoche that substitutes part for whole and whole for part is in fact a metaphor, powerful enough to transform a temporal contiguity into an infinite duration. . . . Compared to this compelling coherence, the contingency of a metonymy based only on a casual encounter of two entities that could very well exist in each other's absence would be entirely devoid of poetic power. . . . If metonymy is distinguished from metaphor in terms of necessity and contingency . . . then metonymy is per definition unable to create genuine links, whereas no one can doubt, thanks to the butterflies, the resonance of the crates, and especially the "chamber music" of the flies, of the presence of light and of warmth in the room. On the level of sensation, metaphor can reconcile night and day in a chiaroscuro that is entirely convincing. (62-63)
We see that this passage establishes the distinction between metaphor and metonymy on the basis of the two oppositions contiguity/analogy-- the dominant opposition in classical rhetoric--and contingency/necessity. As far as the first opposition is concerned, the difficulty is that the dis- tinction between analogy and contiguity is rather slippery.
The word natural, which is precisely at issue here, since we are try- ing to find out whether the statute applies only to natural persons, is here applied precisely to an artificial person, Congress, which is per- sonified as having natural intentionality. "If Congress had meant . . . " The Court's decision repeatedly relies on this type of personification: it is as though Souter has to treat Congress as an entity with intentions, even natural intentions, in order to say that Congress could not have meant to include artificial entities in its ruling. There is a personifica- tion of an artificial entity, Congress, embedded in the very project of interpreting how far the law will allow for artificial entities to be con- sidered persons.
Turning to the Dictionary Act for person and to Webster's diction- ary for context, Souter also notes that he has to define indicates. The
Anthropomorphism in Lyric and Law 215
216 Barbara Johnson
difficulty of doing so pushes him into a volley of rhetorical flourishes: "A contrary 'indication' may raise a specter short of inanity, and with something less than syllogistic force. " "Indicates," it seems, means more than nonsense but less than logical necessity. In other words, the task of reading becomes an infinite regress of glossing terms that are themselves supposed to be determinants of meaning. De Man's linguis- tic stutter returns here as the repeated effort to throw language outside itself. We could read a text, this implies, if only we were sure of the meaning of the words context and indicate. But those are precisely the words that raise the question of meaning in its most general form-- they cannot be glossed with any finality because they name the process of glossing itself.
Souter's text, in fact, is most anthropomorphic at those points where the infinite regress of language is most threatening. Congress is endowed with "natural" intentionality in order to sweep away the abyss of reference. Souter's dismissal of the prisoners' association as an "amorphous legal creature" is the counterpart to the need to reinforce the anthropomorphizability of the artificial legal creature, Congress. 17
Souter's opinion proceeds to detail the ways in which he thinks the in forma pauperis ruling should only apply to natural persons. If an af- fidavit alleging poverty is required for a person to proceed in forma pauperis, then can an artificial entity plead poverty? Souter again turns to Webster's dictionary to find that poverty is a human condition, to be "wanting in material riches or goods; lacking in the comforts of life; needy. " Souter also refers to a previous ruling, which holds that pover- ty involves being unable to provide for the "necessities of life. " It is as though only natural persons can have "life," and that life is defined as the capacity to lack necessities and comforts. "Artificial entities may be insolvent," writes Souter, "but they are not well spoken of as 'poor. '" An artificial entity cannot lack the necessities and comforts of life. Only life can lack. The experience of lack differentiates natural persons from artificial persons. To lack is to be human. In a sense, we have returned to de Man's question about mourning. Is lack human, or just a struc- ture? Whatever the case, the Court holds that associations cannot be considered persons for the purpose of the in forma pauperis procedure.
The majority was only five to four, however. In a dissenting opinion, written by Clarence Thomas, it is argued that there is no reason to re- strict the broad definition of "person" to natural persons in this case. Thomas quotes the Court's view of "poverty" as an exclusively "human condition," and comments:
I am not so sure. "Poverty" may well be a human condition in its "pri- mary sense," but I doubt that using the word in connection with an arti- ficial entity departs in any significant way from settled principles of English usage. . . . Congress itself has used the word "poor" to describe entities other than natural persons, referring in at least two provisions of the United States Code to the world's "Poorest countries"--a term that is used as a synonym for the least developed of the so-called "devel- oping" countries.
Souter has glossed the word poor as though speakers of English could only use it literally. Thomas responds by including the figurative use of poor as included within normal usage. The boundaries between natu- ral persons and artificial persons cannot be determined by usage, be- cause those boundaries have always already been blurred. In treating Congress as an entity with natural intentions, indeed, Souter has al- ready shown how "natural" the artificial can be.
At another point, Thomas takes issue with Souter's discussion of a case in which an association or corporation is considered a person de- spite strong contextual indicators to the contrary. In the case of Wilson v. Omaha Indian Tribe, 442 U. S. 653, 666 (1979), it was decided that "white person" could include corporations because the "larger con- text" and "purpose" of the law was to protect Indians against non- Indian squatters, and would be frustrated if a "white person" could simply incorporate in order to escape the provision of the law. Souter admits that "because a wholly legal creature has no color, and belongs to no race, the use of the adjective 'white' to describe a 'person' is one of the strongest contextual indicators imaginable that 'person' covers only individuals. " Justice Thomas argues that if the Court "was cor- rect in holding that the statutory term 'white person' includes a corpo- ration (because the 'context' does not 'indicate otherwise')--the con- clusion that an association is a 'person' for in forma pauperis purposes is inescapable. " Perhaps another inescapable conclusion is that despite its apparent reference to the physical body, the phrase "white person" is the name, not of a natural, but of a corporate person.
Justice Thomas refutes the reasons Souter has given for finding that artificial entities are excluded from the in forma pauperis provision, noting that there may be sound policy reasons for wanting to exclude them, but that the law as written cannot be construed to have done so. The Court's job, he writes, is not to make policy but to interpret a statute. "Congress has created a rule of statutory construction (an
Anthropomorphism in Lyric and Law 217
218 Barbara Johnson
association is a 'person') and an exception to that rule (an association is not a 'person' if the 'context indicates otherwise'), but the Court has permitted the exception to devour the rule [a nice personification]" (treating the rule as if artificial entities were excluded rather than in- cluded unless the context indicates otherwise). "Whatever 'unless the context indicates otherwise' means," writes Thomas, "it cannot mean 'unless there are sound policy reasons for concluding otherwise. '"
Permitting artificial entities to proceed in forma pauperis may be unwise, and it may be an inefficient use of the government's limited re- sources, but I see nothing in the text of the in forma pauperis statute indicating that Congress has chosen to exclude such entities from the benefits of that law.
Thus, Thomas's two conservative instincts are at war with each other: he would like the government not to spend its money, but he would also like to stick to the letter of the law.
The question of what counts as a juridical person has, in fact, been modified over time in the legal code. It was in 1871 (significantly, per- haps, at the beginning of the end of post-Civil War Reconstruction) that the so-called Dictionary Act was first passed by Congress, in which the word Person "may extend and be applied to bodies politic and corporate. " More recently, the question of fetal personhood has been debated, not only in the Roe v. Wade decision, where it was de- cided that a fetus was not a legal person, but also in Weaks v. Mounter, 88 Nev. 118, where it was decided that a fetus was a person who could sue for intrauterine injuries, but only after birth. Recently, the question of granting patents for forms of life such as oil-slick-eating bacteria or genetically altered mice has raised the question of whether a hybrid be- tween humans and close animal relatives can be patented. And also, of course, the question of the ethics and legality of cloning humans has been raised. The law has reached another crisis about the definition of "person. " In an article on constitutional personhood, Michael Rivard writes:
Current law allows patents for genetically-engineered animals but not for human beings. Humans are not patentable subject matter because patents are property rights, and the Thirteenth Amendment forbids any grant of property rights in a human being. Nevertheless, this exclusion for humans will prove impossible to maintain: within ten to thirty years, or perhaps sooner, advances in genetic engineering technology should allow scientists to intermingle the genetic material of humans and ani-
mals to produce human-animal hybrids. . . . It may soon be possible to patent--and to enslave--human-animal hybrids who think and feel like humans, but who lack constitutional protection under the Thirteenth Amendment. 18
The Thirteenth Amendment is the amendment that abolishes slavery. The constitutional protection against slavery operates as a constraint on the patent office, but it does so in a paradoxical way. The fear of re- instituting something like slavery, or property in humans, is a reaction to, but also a sign of, what must be an ongoing research goal to come as close as possible to creating the ownable, enslavable human. 19
Constitutional personhood has in fact often been defined in prox- imity to slavery. The contradiction between equal rights and chattel slavery led from the beginning to verbal gymnastics, even in the draft- ing of the Constitution itself. By not using the word slavery in the Constitution, and by revising the text of the original fugitive slave clause to refer to the legality of slavery only on the level of the states rather than of the federal government, the framers built a double inten- tionality into the very foundation of their law. Douglas Fehrenbacher, studying the egregious understanding of original intent later employed by the Supreme Court in the case of Dred Scott v. Sanford, writes of the Constitution: "It is as though the framers were half-consciously trying to frame two constitutions, one for their own time and the other for the ages, with slavery viewed bifocally--that is, plainly visible at their feet, but disappearing when they lifted their eyes. "20 A written text of law can thus contain a double intention, the trace of a compro- mise between differing opinions. No wonder interpreting the law's in- tention is so complicated. That intention can always already be mul- tiple. The distinction Justice Thomas made between interpreting the law and making policy cannot hold if the law's ambiguity allows for the possibility that the policy it governs will change.
III
The "inhuman" is not some kind of mystery, or some kind of secret; the in- human is: linguistic structures, the play of linguistic tensions, linguistic events that occur, possibilities which are inherent in language--independently of any intent or any drive or any wish or any desire we might have. . . . If one speaks of the inhuman, the fundamental non-human character of language, one also speaks of the fundamental non-definition of the human as such.
--Paul de Man, "Benjamin's 'The Task of the Translator'"
Anthropomorphism in Lyric and Law 219
220 Barbara Johnson
Only smoking distinguishes humans from the rest of the animals.
--Anonymous (quoted in Richard Klein, Cigarettes Are Sublime)
The case of Rowland v. California Men's Colony, Unit II Men's Advisory Council was ostensibly about whether a council of inmates could sue prison officials in forma pauperis to get their cigarettes back. The de- tails of the case seemed irrelevant to the question of whether an artifi- cial person has the right to sue in forma pauperis. Yet perhaps some of those details deserve note. Is it relevant that the suit to decide this ques- tion is brought by a council of inmates? The phenomenon of the in- mate civil suit has grown to the point where the case law may very well be transformed by it. In a 1995 study of inmate suits in California, it was reported that "For the last fourteen years at least, the federal courts have faced a growing caseload and workload challenge posed by inmate cases. . . . By 1992, these filings numbered nearly 30,000, and constituted 13% of the courts' total civil case filings nation- wide. "21 The majority of these suits are filed in forma pauperis. 22 The Supreme Court's decision may well have been affected by what Clarence Thomas calls "policy decisions. "
If prisoners are affecting the nature of civil proceedings, they are also, at least figuratively, affecting theoretical discussions about the na- ture of rational choice and the evolution of cooperation. The celebrat- ed "Prisoner's Dilemma" has been central to questions of self-interest and social goods since it was introduced by Albert Tucker in 1950. Max Black has even entitled his discussion of the issues raised "The 'Prisoner's Dilemma' and the Limits of Rationality. "23 Why is it that the theoretical study of rational choice has recourse to "man" con- ceived as a prisoner? Does this have anything to do with the poets' ten- dency to see the sonnet form as a prison?
And is it by chance that Rowland v. California Men's Colony, Unit II Men's Advisory Council is about cigarettes? On the one hand, it seems paradoxical that the council has to demonstrate its indigence in order to pursue its suit against the prison directors for depriving the prisoners of access to cigarettes, which in prisons function as a form of currency. On the other hand, it seems fitting that the personhood of the association is the counterpart to the humanity of the inmates, which, as common wisdom (quoted above, second epigraph) would have it, is demonstrated by the act of smoking. The prisoners would thus, in a very attenuated way, be suing for their humanity. As Richard Klein has wittily shown, smoking serves no function other than to enact a struc-
ture of desire--of human desire for self-transcendence, for repetition, for bodily experience corresponding to something other than the "ne- cessities of life" required for existence alone: in short, desire for the sublime. 24 Far from being what defines natural personhood, then, need for the "necessities of life" alone is precisely what cannot define the human.
In the article cited earlier, Rivard declares that "corporations would be presumed constitutional nonpersons," especially for liberty-related rights, unless the corporation could rebut its nonperson status by showing specific natural persons "who would be affected if the corpo- ration were denied these rights. "25 This is the opposite of the Dictionary Act, which considers a corporation a person "unless the context in- dicates otherwise. " Rivard's article is arguing for the rights of new bio- logical species who can pass the "self-awareness test" (which, in a surpisingly Lacanian move derived from Michael Dennett, he defines as wanting to be different from what one is), and he claims that corpo- rations, by their nature, do not pass this test.
But the question of the nature of corporations as persons has never been a simple one, as Rivard admits. In an article titled "The Personifi- cation of the Business Corporation in American Law" (University of Chicago Law Review 54 [fall 1987]: 1441), Gregory A. Marks out- lines in detail the history of corporate personhood. The relation be- tween corporations and the natural persons who compose them has grown more complicated over time, but in most discussions of the mat- ter, it is the "natural" person that functions as the known quantity, and the "artificial" who is either just an "aggregate" of natural persons, or a fiction created by the state, or a mere metaphor, or actually resembles (is like, to return to the Baudelairean word) a natural person in that it has a "will" of its own. Such a corporate will is a form of agency sepa- rate from that of the natural corporators, who exist behind the "veil" of the corporation. Much of Marks's article concerns the exact rhetori- cal valence of this personification:
American law has always recognized that people's activities could be for- mally organized and that the resulting organizations could be dealt with as units. Personification, however, is important because it became far more than a quaint device making it possible for the law to deal with or- ganized business entities. In American legal and economic history, personi- fication has been vital because it (1) implies a single and unitary source of control over the collective property of the corporation's members,
Anthropomorphism in Lyric and Law 221
222 Barbara Johnson
(2) defines, encourages, and legitimates the corporation as an autono- mous, creative, self-directed economic being, and (3) captures rights, ulti- mately even constitutional rights, for corporations thereby giving corpo- rate property unprecedented protection from the state. (1443)
Marks takes seriously the role of language in the evolving history of the corporation. Philosophers and legislators have gone to great lengths to minimize the rhetorical damage, to eliminate personifi- cation as far as possible, but he asserts that it is not just a figure of speech to speak of a corporation's "mind," or even its "life. " "Prac- tical experience, not just anthropomorphism, fixed the corporate mind in the management hierarchy" (1475). The corporation resembled a human being in its capacity to "take resolves in the midst of conflict- ing motives," to "will change. " Yet the analogy is not perfect. The cor- poration, for example, unlike its corporators, is potentially immortal. The effect of personification appears to derive its rhetorical force from the ways in which the corporation resembles a natural person, yet the corporation's immortality in no way diminishes its personification. When Marks says that it is "not just anthropomorphism" that under- pins the agency of the corporation, he still implies that we can know what anthropomorphism is. But his final sentence stands this presup- position on its head. Far from claiming that a corporation's character- istics are derived from a knowable human essence, Marks suggests that what have been claimed to be the essential characteristics of man (es- pecially "economic man") have in fact been borrowed from the nature of the corporation:
Personification with its roots in historic theological disputes and mod- ern business necessity, had proved to be a potent symbol to legitimate the autonomous business corporation and its management. Private property rights had been transferred to associations, associations had themselves become politically legitimate, and the combination had helped foster modern political economy. The corporation, once the de- rivative tool of the state, had become its rival, and the successes of the autonomous corporate management turned the basis for belief in an in- dividualist conception of property on its head. The protests of modern legists notwithstanding, the business corporation had become the quint- essential economic man. (1482-83)
Theories of rationality, naturalness, and the "good," presumed to be grounded in the nature of "man," may in reality be taking their no-
tions of human essence not from "natural man" but from business corporations.
Ambivalence about personification, especially the personification of abstractions, has in fact permeated not only legal but also literary his- tory. Nervousness about the agency of the personified corporation echoes the nervousness Enlightenment writers felt about the personifi- cations dreamed up by the poets. As Steven Knapp puts it in his book Personification and the Sublime:
Allegorical personification--the endowing of metaphors with the agency of literal persons--was only the most obvious and extravagant instance of what Enlightenment writers perceived, with a mixture of ad- miration and uneasiness, as the unique ability of poetic genius to give the force of literal reality to figurative "inventions. " More important than the incongruous presence of such agents was their contagious ef- fect on the ostensibly literal agents with which they interacted. 26
The uncanniness of the personification, then, was derived from its way of putting in question what the "natural" or the "literal" might be.
We have finally come back to the question of whether there is a dif- ference between anthropomorphism and personification, which arose at the end of the discussion of the essay by Paul de Man. It can now be seen that everything hangs on this question. Not only does anthro- pomorphism depend on the givenness of the essence of the human and personification does not, but the mingling of personifications on the same footing as "real" agents threatens to make the lack of certainty about what humanness is come to consciousness. Perhaps the loss of unconsciousness about the lack of humanness is what de Man was calling "true 'mourning. '" Perhaps the "fallacious lyrical reading of the unintelligible" was exactly what legislators count on lyric poetry to provide: the assumption that the human has been or can be defined so that it can then be presupposed without the question of its defini- tion's being raised as a question--legal or otherwise. Thus the poets would truly be, as Shelley claimed, the "unacknowledged legislators of the world," not because they covertly determine policy, but because it is somehow necessary and useful that there be a powerful, presup- posable, unacknowledgment. But the very rhetorical sleight of hand that would instate such an unacknowledgment is indistinguishable from the rhetorical structure that would empty it. Lyric and law are two of the most powerful discourses that exist along the fault line of this question.
Anthropomorphism in Lyric and Law 223
224 Barbara Johnson
NOTES
1. I am thinking of Richard Posner's Law and Literature (Cambridge: Harvard University Press, 1988), Richard Weisberg's The Failure of the Word (New Haven: Yale University Press, 1984), and Peter Brooks, Troubling Confessions (Chicago: University of Chicago Press, 2000). But for a legal approach that does address poetry, see the interesting discussion of Wallace Stevens by Thomas Grey and Margaret Jane Radin in the Yale Journal of Law & The Humanities 2:2 (summer 1990), as well as the more extended treatment of Wallace Stevens in Thomas Grey, The Wallace Stevens Case: Law and the Practice of Poetry (Cambridge: Harvard University Press, 1991).
2. William Wordsworth's sonnet, "Nuns Fret Not at Their Convent's Narrow Room," contains the lines, "In truth the prison, into which we doom / Ourselves, no prison is: and hence for me, / In sundry moods, 'twas pastime to be bound / Within the Sonnet's scanty plot of ground" (Selected Poetry and Prose of Words- worth [New York: Signet, 1970], 169).
3. John Keats's sonnet on the sonnet begins, "If by dull rhymes our English must be chained, / And, like Andromeda, the sonnet sweet / Fettered . . . " (The Selected Poetry of Keats [New York: Signet, 1966], 264).
4. One of several poems by Baudelaire titled Spleen describes a mood pro- duced by or analogized to a rainy day: "Quand la pluie e? talant ses immenses traine? es / D'une vaste prison imite les barreaux . . . " (Baudelaire, OEuvres com- ple`tes, vol. 1 [Paris: Ple? iade, 1975], 75).
5. For a suggestive discussion of what it means for a text to obey the law of genre, see Jacques Derrida, "The Law of Genre," in Acts of Literature, ed. Derek Attridge (New York: Routledge, 1992).
6. Paul de Man, "Anthropomorphism and Trope in the Lyric," in The Rhetoric of Romanticism (New York: Columbia University Press, 1984). Page numbers in parentheses refer to this essay.
7. United States Law Week 61:25 (January 12, 1993). Page numbers in paren- theses refer to this text.
8. This allusion to Keats's "Ode on a Grecian Urn" stands in for the premise of the compatibility of literary aesthetics with linguistic structures, and of linguistic structures with perceptual or intuitive knowledge, that de Man is often at pains to contest. See his remarks on the pedagogical model of the trivium in the titular essay of The Resistance to Theory (Minneapolis: University of Minnesota Press, 1986).
9. Friedrich Nietzsche, "Truth and Falsity in an Ultramoral Sense," in Critical Theory Since Plato, ed. Hazard Adams (Fort Worth: Harcourt Brace Jovanovich, 1992), 634-39. If the Keats poem stands as the claim that aesthetic and epistemo- logical structures are compatible, Nietzsche's text, for de Man, stands as a parody of that claim.
10. The translations are mine, for the purpose of bringing out those aspects of the poems that are relevant to my discussion.
11. "Allegory and Irony in Baudelaire," in Romanticism and Contemporary Criticism (Baltimore: Johns Hopkins University Press, 1993). This essay is part of the Gauss Seminar given by de Man in 1967.
12. Jacques Lacan, "Aggressivity in Psychoanalysis," in E? crits: A Selection, trans. Alan Sheridan (New York: W. W.
Norton, 1977), 17: "What I have called paranoic
knowledge is shown, therefore, to correspond in its more or less archaic forms to certain critical moments that mark the history of man's mental genesis, each repre- senting a stage in objectifying identification. "
13. Ibid. , 9.
14. Jacques Derrida, "Freud's Legacy," in The Postcard, trans. Alan Bass (Chicago: University of Chicago Press, 1987).
15. De Man, "The Resistance to Theory," 10.
16. United States Code (1994 edition), vol. 15, 438.
17. In a response to the present paper given at the Yale Law School, Shoshana
Felman made the brilliant suggestion that Souter would have wanted to rewrite Baudelaire's "Correspondances" as: "Le Congre`s est un temple ou` de vivants pil- liers laissent parfois sortir de confuses paroles . . . " The neoclassical, Parnassian ar- chitecture of official Washington, D. C. , and the common metaphorical expression "pillars of the community," add piquancy to this suggestion.
18. Michael D. Rivard, "Toward a General Theory of Constitutional Person- hood: A Theory of Constitutional Personhood for Transgenic Humanoid Species," UCLA Law Review 39: 5 (June 1992): 1428-29.
19. See A. Leon Higginbotham Jr. and Barbara Kopytoff, "Property First, Humanity Second: The Recognition of the Slave's Human Nature in Virginia Civil Law," Ohio State Law Journal 50:3 (June 1989): "The humanity of the slave, re- quiring that he be treated with the care due other humans and not like other forms of property, became part of the owner's property rights" (520).
20. Douglas E. Fehrenbacher, Slavery, Law, and Politics: The Dred Scott Case in Historical Perspective (Oxford: Oxford University Press, 1981), 15.
21. Kim Mueller, "Inmates' Civil Rights Cases and the Federal Courts: Insights Derived from a Field Research Project in the Eastern District of California," Creighton Law Review 1228 (June 1995): 1258-59. In the Eastern District of Cali- fornia, inmates' civil rights actions constituted nearly 30 percent of the case filings. (California Men's Colony is not in the Eastern District; it is in San Luis Obispo. )
22. Ibid. , 1276 and 1281.
23. Max Black, Perplexities (Ithaca, N. Y. : Cornell University Press, 1990). See also Robert Axelrod, The Evolution of Cooperation (New York: Basic Books, 1984).
24. Richard Klein, Cigarettes Are Sublime (Durham, N. C. : Duke University Press, 1993). Klein notes, incidentally, that Baudelaire is one of the first French writers to use the word cigarette in print (in his "Salons de 1848," 8).
25. Rivard, "Toward a General Theory of Constitutional Personhood," 1501-2.
26. Steven Knapp, Personification and the Sublime (Cambridge: Harvard University Press, 1985), 2.
Anthropomorphism in Lyric and Law 225
This page intentionally left blank
IV. The Mnemopolitical Event
This page intentionally left blank
The Politics of Rhetoric
Ernesto Laclau
Why would a political theorist like me, working mainly on the role of hegemonic logics in the structuration of political spaces, be interested in the work of a prominent literary critic such as Paul de Man? I could suggest at least two main reasons. The first is that one of the leitmotifs of Paul de Man's work has been the subversion of the frontiers separat- ing theoretical from literary disciplines, so that those dimensions that had traditionally been conceived as privative of literary or aesthetic language became, for him, actually defining features of language tout court. Against all attempts to differentiate between "appearance" and "saying," between a primary text whose message would have been me- diated by the materiality of the signs, of the figural, and a language of inquiry governed by reason, de Man had always insisted that any lan- guage, whether aesthetic or theoretical, is governed by the materiality of the signifier, by a rhetorical milieu that ultimately dissolves the illu- sion of any unmediated reference. In this sense a generalized rhetoric-- which necessarily includes within itself the performative dimension-- transcends all regional boundaries and becomes coterminous with the structuration of social life itself. Conceived at such a broad level of generality, the literariness of the literary text breaks the limits of any specialized discipline and its analysis involves something like the study of the distorting effects that representation exercises over any reference--effects that become, thus, constitutive of any experience.
Moreover, de Man himself was perfectly aware of the political and ideological implications of his approach to texts. In a famous interview with Stefano Rosso that discusses the increasing recurrence in his works of the terms political and ideological, he answered as follows:
229
230 Ernesto Laclau
I don't think I ever was away from these problems, they were always uppermost in my mind. I have always maintained that one could ap- proach the problems of ideology and by extension the problems of poli- tics only on the basis of critical-linguistic analysis, which had to be done on its own terms, in the medium of language, and I felt that I could ap- proach those problems only after having achieved a certain control over those questions. It seems pretentious to say so, but it is not the case. I have the feeling I have achieved a certain control over technical prob- lems of language, specifically problems of rhetoric, of the relation be- tween tropes and performatives, of saturation of tropology as a field that in certain forms of language goes beyond that field. . . . I feel now some control of a vocabulary and of a conceptual apparatus that can handle that. 1
As for the second reason for a political theorist to be interested in de Man's work, it has to do with something related to the political field itself. Gone are the times in which the transparency of social actors, of processes of representation, even of the presumed underlying logics of the social fabric could be accepted unproblematically. On the contrary, each political institution, each category of political analysis shows itself today as the locus of undecidable language games. The over- determined nature of all political difference or identity opens the space for a generalized tropological movement and thus reveals the fruitful- ness of de Man's intellectual project for ideological and political analy- sis. In my work, this generalized politico-tropological movement has been called "hegemony. " I intend in this essay to stress some decisive points in the work of de Man, especially in his late work, where the di- rection of his thought could be helpful in developing a hegemonic ap- proach to politics.
I
The requirements of "hegemony" as a central category of political analysis are essentially three. First, that something constitutively het- erogeneous to the social system or structure has to be present in the lat- ter from the very beginning, preventing it from constituting itself as a closed or representable totality. If such a closure were achievable, no hegemonic event could be possible and the political, far from being an ontological dimension of the social--an "existential" of the social-- would just be an ontic dimension of the latter. Second, however, the hegemonic suture has to produce a retotalizing effect, without which
no hegemonic articulation would be possible either. But, third, this re- totalization cannot have the character of a dialectical reintegration. It has, on the contrary, to maintain alive and visible the original and con- stitutive heterogeneity from which the hegemonic articulation started. How is a logic that can maintain these two contradictory requirements at the same time possible? Let us approach this question through the exploration of its possible presence in de Man's texts. We will start from the analysis of Pascal's Re? flexions sur la ge? ome? trie en ge? ne? ral; De l'esprit ge? ome? trique et de l'Art de persuader that de Man carries out in "Pascal's Allegory of Persuasion. "2
Pascal starts his study of the esprit ge? ome? trique from the distinction between nominal and real definitions--the first resulting from con- vention and being thus exempt from contradiction, the second being axioms or propositions to be proved--and asserts that the confusion between the two is the main cause of philosophical difficulties. Main- taining the separation between the two--as the geometrician does--is the first rule of philosophical clarity. However, the argument runs quickly into difficulties, as geometrical discourse includes not only nominal definitions but also "primitive terms"--such as motion, num- ber, and extension--which are undefinable but, nonetheless, fully intel- ligible. According to Pascal, these undefinable words find a universal reference not in the (impossible) fact that all men have the same idea concerning their essence, but instead, in the fact that there is a relation of reference between name and thing, "so that on hearing the expres- sion time, all turn (or direct) the mind to the same entity" (56). But, as de Man shows, this brings back the real definition into the geometrical camp itself, for
the word does not function as a sign or a name, as was the case in the nominal definition, but as a vector, as a directional motion that is mani- fest only as a turn, since the target toward which it turns remains un- known. In other words, the sign has become a trope, a substitutive rela- tionship that has to posit a meaning whose existence cannot be verified, but that confers upon the sign an unavoidable signifying function. (56)
As the semantic function of the primitive terms has the structure of a trope, "it acquires a signifying function that it controls neither in its ex- istence nor in its direction. " Ergo, "[s]ince definition is now itself a primitive term, it follows that the definition of the nominal definition is itself a real, and not a nominal, definition" (57).
This contamination of the nominal by the real definition is still
The Politics of Rhetoric 231
232 Ernesto Laclau
more visible when we move to the question of double infinitude, which is decisive in establishing the coherence and intelligibility of the rela- tionship between mind and cosmos. Here, Pascal deals with the objec- tions put to him by the Chevalier de Me? re? , according to whom--given the Pascalian principle of homogeneity between space and number--it is possible to conceive an extension formed by parts that are not them- selves extended, since it is possible to have numbers made up of units that are devoid of number. With this, the principle of infinite smallness would be put into question. Pascal's answer has two steps. He asserts, in the first place, that which applies to the order of number does not apply to the order of space. One is not a number, there is no plurality in it; but at the same time it belongs to the order of number for, given the Euclidean principle of homogeneity ("magnitudes are said to be of the same kind or species when one magnitude can be made to exceed another by reiterated multiplication"), it is part of the infinity postu- lated by that principle. On that basis, Pascal can distinguish between number and extension, but only at the price of grounding the distinc- tion in real and not nominal definitions. As de Man asserts:
The synecdochal totalization of infinitude is possible because the unit of number, the one, functions as a nominal definition. But, for the argument to be valid, the nominally indivisible number must be distinguished from the really indivisible space, a demonstration that Pascal can ac- complish easily, but only because the key words of the demonstration-- indivisible, spatial extension (e? tendue), species (genre), and definition-- function as real, and not as nominal, definitions. (58-59)
But--second step--if the order of number and the order of extension had to be separated to answer Me? re? 's objection, the rift between the two had also to be healed if the principle of homogeneity between both was to be maintained. This homology is restored by appealing, as far as number is concerned, to the zero--which, on the difference with one is radically heterogeneous with the order of number--and by finding equivalences in the order of time and motion, such as "instant" and "stasis. " This appeal to the zero, however, has dramatic consequences for the coherence of the system, which de Man describes in a passage worth quoting in full:
the coherence of the system is now seen to be entirely dependent on the introduction of an element--the zero and its equivalences in time and motion--that is itself entirely heterogeneous with regard to the system
and is nowhere part of it. . . . Moreover, this rupture of the infinitesimal and the homogeneous does not occur on the transcendental level, but on the level of language, in the inability of a theory of language as sign or as name (nominal definition) to ground this homogeneity without hav- ing recourse to the signifying function, the real definition, that makes the zero of signification the necessary condition for grounded knowl- edge. . . . It is as sign that language is capable of engendering the prin- ciples of infinity, of genus, species, and homogeneity, which allow for synecdochal totalizations, but none of these tropes could come about without the systematic effacement of the zero and its reconversion into a name. There can be no one without zero, but the zero always appears in the guise of a one, of a (some)thing. The name is the trope of the zero. The zero is always called a one, when the zero is actually nameless, "in- nommable. " In the French language, as used by Pascal and his inter- preters, this happens concretely in the confusedly alternate use of the two terms ze? ro and ne? ant. The verbal, predicative form ne? ant, with its gerundive ending, indicates not the zero, but rather the one, as the limit of the infinitely small, the almost zero that is the one. (59)
It is important to give serious consideration to this remarkable passage-- remarkable, among other things, because de Man does not pursue later in his essay all the implications of his own de? marche--for it contains, in nuce, all the relevant dimensions of the problem we are exploring. Everything turns around the role of the zero. The zero is, we are told, something radically heterogeneous with the order of number. The order of number, however, cannot constitute itself without reference to the zero. It is, in this sense, a supplement to the system that, none- theless, is necessary for constituting the latter. Vis-a`-vis the system, the zero is in an undecidable tension between internality and externality-- but an internality that does not exclude heterogeneity. The zero, in the second place is "innommable," nameless, but at the same time it pro- duces effects, it closes the system, even at the price of making it hope- lessly heterogeneous. It retotalizes the system, incurring, however, an inconsistency that cannot be overcome. The zero is nothing, but it is the nothing of the system itself, the impossibility of its consistent clo- sure, which is signified by the zero and in that sense, paradoxically, the zero as empty place becomes the signifier of fullness, of systematicity as such, as that which is lacking. The semantic oscillation between ze? ro and ne? ant that de Man observes is the result of this dual condition of the moment of closure: being an impossible object, but also a necessary
The Politics of Rhetoric 233
234 Ernesto Laclau
one. Finally, if the zero as moment of closure is impossible as an object but also necessary, it will have to have access to the field of represen- tation. But the means of representation will be constitutively inade- quate. It will give to the "innommable" a body, a name, but this can be done only at the price of betraying its true "nonbeing"; thus the tropological movement that prolongs sine die the non-resolvable dia- lectics between the zero and the one. In the words of de Man just quoted: "There can be no one without zero, but the zero always ap- pears in the guise of a one, of a (some)thing. The name is the trope of the zero. The zero is always called a one, when the zero is actually nameless, 'innommable. '"
Now, this succession of structural moments coincides, almost step by step, with the logic of hegemony that I have tried to describe in my work and which I see operating in the texts of Gramsci--to whom I will return later on. To start with, the condition of any hegemonic su- ture is the constitutive nonclosure of a system of political signification. The systemacity of a system, its closure--which is the condition of sig- nification in a system, such as Saussure's, whose identities are merely differential--coincides with the determination of its limits. These lim- its, however, can only be dictated by something that is beyond them. But, as the system is a system of differences, of all actual differences, that "beyond"--which should be heterogeneous with the system in order to fulfill its function of truly closing it--lacks the condition of a true heterogeneity if it consists in one more difference. The latter would be, in some way, undecided, suspended between being internal and external to the system. This jeopardizes the role of the "beyond" as limit and, as a result, the possibility of constituting the differences as truly intrasystemic differences. It is only if the beyond the limit has the character of an exclusion that its role as limit is restored and with it the possibility of emergence of a full system of differences. 3
However, this fullness of the system (obtained, it is true, at the price of a dialectical retrieval of its negation) has a shortcoming. For, vis-a`- vis the excluded element, all differences within the system establish relations of equivalence between themselves. And equivalence is pre- cisely that which subverts difference. So the "beyond" which is the condition of possibility of the system is also its condition of impossi- bility. All identity is constituted within the unsolvable tension between equivalence and difference.
So, as with the Pascalian zero, we arrive at an object that is at the same time impossible and necessary. As impossible, it is an empty place
within the structure. But, as necessary, it is a "nothing" that will pro- duce structural effects, and this requires that it has access to the field of representation. And, as in the dialectics between the zero and the one, this double condition of necessity and impossibility will be constitu- tively inadequate. The fullness of the system, its point of imaginary satu- ration, will be, as in the example of de Man, a nothing that becomes a something. What are the possible means of this distorted representa- tion? Only the particular differences internal to the system. Now, this relationship by which a particular difference takes up the representa- tion of an impossible totality entirely incommensurable with it is what I call a hegemonic relation.
There are only two differences between the hegemonic logic and the Pascalian dialectics between the zero and the one as described by de Man. The first is that--given the numerical nature of Pascal's case-- the zero can only be embodied by the one, while in the case of the hegemonic logic any element within the system can be a bearer of a hegemonic function. 4 The second difference is that, given de Man's in- terests, the determination of the heterogeneous character of the tropo- logical displacement from the zero to the one is the point of arrival of his analysis, whereas for a student of hegemonic logics the analysis of the exact nature of this tropological movement becomes imperative. In de Man's detotalizing discourse, what matters is to show the hetero- geneity out of which the tropological movement operates. This is also a vital part of a hegemonic analysis. But what is decisive for the latter is the determination of the partial retotalizations that the tropological movement makes possible. 5 This is the dimension that we now have to take into account. I will do it through a reference to the opposition metaphor/metonymy as presented in de Man's essay on Proust in Allegories of Reading. 6
II
The text on Proust deals, as is well known, with the discourse of young Marcel on the pleasure of reading and with the way in which such pleasure is constructed through a set of metaphoric substitutions that are, however, persuasive only through the operation of a series of con- tingent metonymic movements. De Man asserts:
The crossing of sensory attributes in synaesthesia is only a special case of a more general pattern of substitution that all tropes have in com- mon. It is the result of an exchange of properties made possible by a
The Politics of Rhetoric 235
236 Ernesto Laclau
proximity or an analogy so close and intimate that it allows the one to substitute for the other without revealing the difference necessarily in- troduced by the substitution. The relational link between the two enti- ties involved in the exchange then becomes so strong that it can be called necessary: there could be no summer without flies; no flies with- out summer. . . . The synecdoche that substitutes part for whole and whole for part is in fact a metaphor, powerful enough to transform a temporal contiguity into an infinite duration. . . . Compared to this compelling coherence, the contingency of a metonymy based only on a casual encounter of two entities that could very well exist in each other's absence would be entirely devoid of poetic power. . . . If metonymy is distinguished from metaphor in terms of necessity and contingency . . . then metonymy is per definition unable to create genuine links, whereas no one can doubt, thanks to the butterflies, the resonance of the crates, and especially the "chamber music" of the flies, of the presence of light and of warmth in the room. On the level of sensation, metaphor can reconcile night and day in a chiaroscuro that is entirely convincing. (62-63)
We see that this passage establishes the distinction between metaphor and metonymy on the basis of the two oppositions contiguity/analogy-- the dominant opposition in classical rhetoric--and contingency/necessity. As far as the first opposition is concerned, the difficulty is that the dis- tinction between analogy and contiguity is rather slippery.
