nehamas, a

18
The Postulated Author: Critical Monism as a Regulative Ideal Author(s): Alexander Nehamas Source: Critical Inquiry, Vol. 8, No. 1 (Autumn, 1981), pp. 133-149 Published by: The University of Chicago Press Stable URL: http://www.jstor.org/stable/1343210 Accessed: 12/09/2010 10:55 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ucpress. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Critical Inquiry. http://www.jstor.org

Upload: daniela-ferro

Post on 08-Apr-2015

128 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: NEHAMAS, A

The Postulated Author: Critical Monism as a Regulative IdealAuthor(s): Alexander NehamasSource: Critical Inquiry, Vol. 8, No. 1 (Autumn, 1981), pp. 133-149Published by: The University of Chicago PressStable URL: http://www.jstor.org/stable/1343210Accessed: 12/09/2010 10:55

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=ucpress.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to CriticalInquiry.

http://www.jstor.org

Page 2: NEHAMAS, A

The Postulated Author: Critical Monism as a Regulative Ideal

Alexander Nehamas

Ii n'y a pas une parole qu'on puisse comprendre, si l'on va au fond. -PAUL VALtRY

Critical pluralism, broadly stated, is the view that literary texts, unlike natural phenomena, for which there is only one correct explanation, can be given many equally acceptable, even though incompatible, interpre- tations. But the thesis that, in contrast to science, "the use ... of diverse but complementary vantages [is] not only rationally justifiable, but nec-

essary to the understanding of art, and indeed of any subject of hu- manistic inquiry" seems to me to make a virtue out of necessity and a

necessity out of fact.' Such a fact is that within sixty years of its publication, a fiction like

Kafka's Metamorphosis had already provoked 148 studies, of an astonish-

ing variety.2 This fact has been transformed into a virtue by Stanley An earlier version of this paper was read at the Western Division Meetings of the

American Philosophical Association. I am very grateful toJohn G. Bennett for his extensive and constructive comments, with some of which this version tries to come to terms. I must also thank Wayne Booth, David Carrier, Michael McCabe, Lynne McFall, and Kendall Walton for their objections and suggestions. Part of the research for this paper was sup- ported by a grant from the National Endowment for the Humanities.

1. M. H. Abrams, "A Note on Wittgenstein and Literary Criticism," English Literary History 41 (Winter 1974): 552

2. These studies are listed and discussed in Stanley Corngold, The Commentator's De-

spair: The Interpretation of Kafka's "Metamorphosis" (Port Washington, N.Y., and London, 1973); all further references to this work will be included in the text. Many more inter-

pretations of the story have, of course, been offered since. 0 1981 by The University of Chicago. 0093-1896/81/0801-0003$01.00. All rights reserved.

133

Page 3: NEHAMAS, A

134 Alexander Nehamas The Postulated Author

Corngold, who accounts for this flood of criticism by attributing it to the

very point of the story and, ultimately, to the very nature of literature.

Corngold interprets Samsa's change into what is an essentially vague, incomplete, and indescribable monster as an allegory for writing itself- an activity which, according to many recent literary theorists, is bound to result in imperfect communication, unavoidable misunderstanding, and inevitable misreading: "The negativity of the vermin has to be seen as rooted ... in the literary enterprise itself. . . . The creature ... is ... language itself (parole)-a word broken loose from the context of

language (langage), fallen into a void of meaning which it cannot signify, near others who cannot understand it" (Commentator's Despair, pp. 26, 27). If this is so, why should we be surprised that the story, like all literature, will not yield itself to a definitive interpretation?

But of course Corngold's view is reached through an interpretation which must be itself correct if it is to explain why there cannot be a correct interpretation of the story. And this paradox of method is parallel to a paradox of content. The Metamorphosis, on this view, concerns the

inability of literature to achieve perfect communication and so to receive final interpretation. This is what the story communicates. But if it suc- ceeds in communicating it, it communicates that it fails to communicate; and if it fails, since this failure is what it communicates, it succeeds!

The claim that literature can ultimately communicate only that it cannot ultimately communicate is not uncommon in recent literary the-

ory.3 I suspect that it is reached by illegitimately extending the thesis that words are polysemous or radically ambiguous. The extension is made

by assuming that if a text has a property (if, in particular, a word is

ambiguous), then it refers to that property (the word signifies ambiguity). Thus, for example, given the fact that the sentence "The green is either" is ungrammatical, Jacques Derrida infers that "it signifies an example of ungrammaticality."4 To offer another example, J. Hillis Miller assumes that the history of words is essential to their meaning and writes:

3. Jacques Derrida, for example, forcefully defends this view in Speech and Phenomena (Evanston, Ill., 1973); Of Grammatology (Baltimore, 1974); and Writing and Difference (Chi- cago, 1976). See also La Dissemination (Paris, 1972).

4. Derrida, "Signature Event Context," Glyph 1 (1977): 185. See also J. Hillis Miller, "The Critic as Host," in Deconstruction and Criticism, ed. Harold Bloom et al. (New York, 1979), p. 225. In another, brilliant essay, Miller relies heavily on the principle that if an artwork employs certain conventions, then it is also about those conventions ("The Fiction of Realism: Sketches by Boz, Oliver Twist, and Cruickshank's Illustrations," in Dickens Cen- tennial Essays, ed. Ada Nisbet and Blake Nevius [Berkeley, 1971], pp. 85-153).

Alexander Nehamas, professor of philosophy at the University of Pittsburgh, has written articles on ancient Greek philosophy, literary theory, Nietzsche, and Thomas Mann.

Page 4: NEHAMAS, A

Critical Inquiry Autumn 1981 135

The effect of etymological retracing is not to ground the word solidly but to render it unstable, equivocal, wavering, abysmal. All etymology is false etymology, both in the sense that there is always some bend or discontinuity in the etymological line, and in the sense that etymology always fails to find an etymon, a true literal meaning at the origin.5

If texts indeed consist of words so construed, it may seem to follow that

"a text never has a single meaning, but is the crossroads of multiple ambiguous meanings"6 and that therefore every text "is 'unreadable', if by 'readable' one means open to a single, definitive, univocal interpre- tation."7 This raises two questions. The first, which I shall not try to answer here, is why we must assume that a word must have had an original literal use in order to be univocal now and why we need to accept the "Rousseauistic or Condillacian law that all words were origi- nally metaphors."8 The second, to which I shall pay close attention, is why we must agree that to be "readable," a text must have a definitive interpretation-if by "definitive" we mean "unrevisable." For though the absence of an unrevisable interpretation implies that we can change our mind about what a text means, it does not imply that what a text means changes along with our mind.

Deconstructive critics begin with the realization that written texts are enormously independent of their writers and then proceed to sever altogether, at least in theory, the connection between author and text. Since writing remains "when the author of the writing no longer answers for what he has written," Derrida argues, "the text is cut off from all absolute responsibility, from consciousness as the ultimate authority, or- phaned and separated at birth."' Geoffrey Hartman traces the idea of recovering authors' intentions to the Renaissance, with its concern for establishing original texts, and argues that "the more learning and schol- arship we bring to an author, with the aim of defining his difference or individual contribution, the less certainty there seems to be of succeeding in this." Hartman concludes that "the notion of unique works of art,

5. Miller, "Ariadne's Thread: Repetition and the Narrative Line,"' Critical Inquiry 3 (Autumn 1976): 70. See also "Critic as Host," pp. 218-20.

6. Miller, "Tradition and Difference," Diacritics 2 (1972): 12. 7. Miller, "Critic as Host," p. 226. See also Paul de Man, "Nietzsche's Theory of

Rhetoric," Symposium 28 (Spring 1974): 44, and Geoffrey Hartman, "Literary Criticism and Its Discontents," Critical Inquiry 3 (Winter 1976): 205.

8. Miller, "Ariadne's Thread,"' p. 70; cf. "Tradition and Difference," p. 11. The same claim has been made at length by Derrida, "White Mythology: Metaphor in the Text of Philosophy," New Literary History 6 (Autumn 1974): 5-74. For a criticism of this view, as exhibited in the early work of Nietzsche, see Arthur Danto, Nietzsche as Philosopher (New York, 1965), pp. 37-47.

9. Derrida, "Signature Event Context," p. 181.

Page 5: NEHAMAS, A

136 Alexander Nehamas The Postulated Author

certified by the personal name of the author, fades away into nostalgia."'• Michel Foucault goes even further and claims that the author is a fiction created, more or less, by Saint Jerome, now moribund and an object of indifference: "What matter who's talking?"" In a work otherwise un-

sympathetic to post-structuralism, Jonathan Culler accepts this view when he writes, "The meaning of a sentence, one might say, is not a form or an essence, present at the moment of its production and lying behind it as a truth to be recovered, but the series of developments to which it gives rise, as determined by past and future relations between words and the conventions of semiotic systems.""2 The object of criticism cannot therefore be what the author meant by a text but what a text means in itself. Since in itself a text means what its constituents have ever meant, and since (according to deconstruction) no constituent is univocal, the text turns out to be the "crossroads" of all of its constituents'

incompatible senses. This radical pluralism is thus grounded on a view about the nature

of texts, some of the many meanings of which are exhibited, with equal plausibility, by different interpretations. But interpretations, too, are written texts, and they also need to but cannot be read. Just as every reading is a misreading, so it will be in turn misread. As Miller says,

The new turn in criticism involves an interrogation of the notion of the self-enclosed literary work and of the idea that any work has a fixed, identifiable meaning. The literary work is seen in various ways as open and unpredictably productive. The reading of a poem is part of the poem. This reading is productive in its turn. It pro- duces multiple interpretations, further language about the poem's language, in an interminable activity without necessary closure. "

But we, at least, have now been brought to the closure of the ex-

position which opened with our paradoxical reading of The Metamor-

phosis. According to this reading, writing cannot communicate; every text is misread since a reading is just an effort to impose a single coherent

meaning on the text and thus presupposes that communication has suc- ceeded. Itself an instance of this law, The Metamorphosis has generated a large number of readings; yet, "any reading can be shown to be a

misreading on evidence drawn from the text itself."'4 Every reading will

10. Hartman, "Criticism and Its Discontents," pp. 204-5; all further references to this

essay will be included in the text. 11. Michel Foucault, "What Is an Author?" Language, Counter-Memory, Practice, ed.

Donald F. Bouchard, trans. Bouchard and Sherry Simon (Ithaca, N.Y., 1977), pp. 113-38. 12. Jonathan Culler, Structuralist Poetics (Ithaca, N.Y., 1975), p. 132. This view, as I

suggest below, ultimately derives from New Criticism; see n. 16 below. 13. Miller, "Stevens' Rock and Criticism as Cure, II," Georgia Review 30 (Summer

1976): 333. 14. Ibid.

Page 6: NEHAMAS, A

Critical Inquiry Autumn 1981 137

thus be replaced, and every new reading will be in turn misread, all

circling continually around a nonexistent center, each an effort to isolate an imaginary "literal meaning at the origin," each a falcon without a falconer.

Appalled by the anarchy he takes this view to lead into, E. D. Hirsch has insisted that one of each text's many interpretations, the author's own, must be taken as canonical: "If the meaning of the text is not the author's, then no interpretation can possibly correspond to the meaning of a text, since the text can have no determinate or determinable mean-

ing."'5 Behind this view lies a theory of meaning which is ultimately derived from the work of I. A. Richards, who wrote that

[the] logical use of words with constant senses that are the same for each occurrence . . . is an extremely artificial sort of behavior. ... And the fluidity, the incessant delicate variation of the meaning of our words ... is the virtue of language for our other purposes. [It is not true] that if a passage means one thing it cannot mean another and an incompatible thing.'6

Hirsch maintains his monism in the light of, or perhaps despite, his

theory about the meaning of texts: "The nature of the text is to have no

meaning except that which an interpreter wills into existence.... A text [is] only an occasion for meaning, in itself an ambiguous form devoid of the consciousness where meaning abides.""

To this view, which bears important similarities to the approach of Hirsch's opponents, one can make, with the King in Alice in Wonderland, an easy reply: "If there's no meaning in it ... that saves a lot of trouble, you know, as we needn't try to find any." But the witticism and its wording only serve to raise the crucial question: Is a text's meaning found, or is it made? Both sides initially agree that meaning is made, that a text

15. E. D. Hirsch, Validity in Interpretation (New Haven, Conn., 1967), p. 5. On the

prescriptive nature of Hirsch's view, see Jack Meiland, "Interpretation as a Cognitive Discipline," Philosophy and Literature 2 (Spring 1978): 24-28.

16. I. A. Richards, Interpretation in Teaching (New York, 1938), p. 256. New Criticism assumed that "words . . . include at least potentially, within their appearance in a given setting, (all) the meanings they have had ... in previous contexts" (Richard Strier, "The Poetics of Surrender: An Exposition and Critique of New Critical Poetics," Critical Inquiry 2 [Autumn 1975]: 173-74). Monroe Beardsley accepts this principle in The Possibility of Criticism (Detroit, 1970), pp. 19-20. If so, however, what reason is there to think that the changing meanings of the words of a text will be subject to a single univocal interpretation, as Beardsley believes? It is this monism which deconstruction has abandoned in its claim that words actually possess, in every appearance, all the meanings they have ever had, that every passage does mean "another and an incompatible thing."

17. Hirsch, "Three Dimensions of Hermeneutics," New Literary History 3 (Winter 1972): 246. Despite some evidence to the contrary (e.g., p. 256), Hirsch generally seems to accept this radical thesis of textual indeterminacy. For considerations weighing against this thesis, see Meiland, "Interpretation" (n. 15 above), pp. 32-33, and Beardsley, Possibility of Criticism, pp. 24-26.

Page 7: NEHAMAS, A

138 Alexander Nehamas The Postulated Author

means just what it is taken to mean by its interpreters.18 Deconstruction infers that critics should therefore do self-consciously what they do in

any case, which is to make their own meaning out of every text. Hirsch, by contrast, claims that critics ought now to go on to discover the meaning which a text was made to have by its author.

There is a large gap between the monism advocated by Hirsch and the radical pluralism which follows from the writing of some deconstruc- tive critics. Within this gap are located some recent writers who argue that the meaning of a text is partly found and partly made. This ar-

gument is the basis for the limited pluralism of M. H. Abrams, Peter

Jones, and Jack Meiland, according to which the "rules of the language" to which a text belongs determine a fundamental level of meaning, independent of all points of view, given to and found by the interpreter."9 But just as the locutionary content of a sentence does not by itself de- termine what illocution that sentence is being used to perform in a

particular case, so this "central core," though it limits the legitimate overall interpretations of a text, does not exhaust its meaning. Meiland, for example, calls this fundamental level the "textual meaning" and

distinguishes it from the "literary meaning," over which critical disagree- ment occurs. He writes that "the agreed-upon textual meaning can serve as a criterion of validity for interpretations at the level of literary mean-

ing. . . . Any literary interpretation which does not cohere with basic

agreed-upon textual meaning can be ruled out as an invalid interpre- tation" ("Interpretation," p. 36).

The central difficulty with this view is that, in my opinion, it simply tends to reify whatever it is that a text's interpreters do and do not, at some particular time, agree about. Textually, Meiland writes, "Romeo and

Juliet is about a man and a woman who are in love with one another, whose families prevent their marrying, and who die due to a tragic misunderstanding" ("Interpretation," p. 35). But is this obvious because it is determined solely by the rules of English, or is it because it constitutes

18. See Hirsch, "Three Dimensions of Hermeneutics," p. 247: "If an ancient text has been interpreted as a Christian allegory, that is unanswerable proof that it can be so

interpreted." But is this proof that the text has been legitimately so interpreted? Hirsch seems to presuppose that in some sense such a reading is accurate to the text; but this seems to beg the question at issue.

19. See Abrams, "Note on Wittgenstein"; "Rationality and Imagination in Cultural

History: A Reply to Wayne Booth," Critical Inquiry 2 (Spring 1976): 447-64, esp. 457; "What's the Use of Theorizing about the Arts?" in In Search of Literary History, ed. Morton Bloomfield (Ithaca, N.Y., 1972), pp. 3-54; and "The Deconstructive Angel," Critical Inquiry 3 (Spring 1977): 425-38. Peter Jones, Philosophy and the Novel (Oxford, 1975), chap. 5, esp. pp. 182-83; all further references to this book will be included in the text. Meiland, "Interpretation," esp. pp. 29-31 and 35-37; all further references to this essay will be included in the text. See also Quentin Skinner, "Motives, Intentions, and the Interpretation of Texts," New Literary History 3 (Winter 1972): 393-408.

Page 8: NEHAMAS, A

Critical Inquiry Autumn 1981 139

such a minimal interpretation, chosen just because the critics of the text are likely to agree about it?20

The existence of a well-defined notion of literal or dictionary mean-

ing which can be of use to this view is itself problematic. Do dictionaries

give us what words must essentially mean in all their uses, or do they simply supply us with a rough guess, a coarse grid against which, but not necessarily within which, to locate individual words and phrases? Whatever the answer to this question, even if we assume that the notion of a word's literal meaning is well defined, the difficulties of this theory are far from over. The main problem is that it is not possible to identify "textual meaning" with the literal meaning of the words of which a text consists. The words' literal meaning is specified through a set of roughly synonymous words supplied by the dictionary. But the textual meaning is a summary or paraphrase, that is, an interpretation (however minimal) of what these words, given their literal meaning, are being used to do on this particular occasion.

If textual meaning is a minimal interpretation of a text, then it is not surprising that it is compatible with a number of "literary" meanings, since these now turn out to be more specific interpretations of the text. For it is clear that a number of more particular specifications of any object are compatible with a more general specification of that object, even if they conflict with one another. Something can be an item of furniture and also a chair, a chaise, or a sofa; it can be any of these and also Louis XVI, Empire, or Directory style. None of this shows that it is all of these things. Similarly, though the textual meaning of Romeo and

Juliet fails to determine a single overall reading of the play, this does not show that the play does have the many literary meanings that have been attributed to it. Compatibility with textual meaning is at best a necessary condition for validity, but this trivial fact offers no support for any sort of pluralism.

Such compatibility is at best necessary for validity because in fact we can both disagree about and revise our views of textual meaning. The object we werejust imagining may turn out not to be an item of furniture at all but a strange machine; just so, we may revise our minimal inter- pretation of Romeo and Juliet. Though we are likely to agree about textual

meaning, we cannot take this agreement for granted; textual meaning depends on substantive as well as on linguistic considerations. Are Romeo and Juliet, for example, a man and a woman or a boy and a girl? But more importantly, in many cases where our minimal and more specific interpretations are in conflict, we may choose to modify the former rather than to reject the latter. Our construal of Romeo's scream, "The time and my intent is savage-wild / More fierce and inexorable far / Than

20. Meiland is, in any case, correct that such agreement as does exist is a sufficient objection to Hirsch's thesis of the radical indeterminacy of textual meaning.

Page 9: NEHAMAS, A

140 Alexander Nehamas The Postulated Author

empty tigers and the roaring sea" has serious consequences for the nature of the misunderstanding which leads to his death.

But if textual meaning is not given, if it is also, like literary meaning, the product of revisable interpretation, have we not granted deconstruc- tion all that it wanted in the first place?2' Derrida is describing a view not unlike Meiland's when he writes that "the concept of a centered structure is in fact the concept of a free-play based on a fundamental

ground, a freeplay which is constituted upon a fundamental immobility and a reassuring certitude, which is itself beyond the reach of the free-

play.22 The "center" is for Derrida the obvious or intuitive reading of a text, Meiland's textual meaning. Derrida argues that even the most ob- vious reading is the result of interpretation and can therefore be ques- tioned, revised, or displaced.23

This is, I think, correct. Just as in scientific explanation there are no data immune to revision, so in literary criticism there are no readings impervious to question. But the fact about science does not show that

apparently competing scientific theories are incommensurable and that therefore we cannot judge between them or that each such theory con- cerns its own distinct world.24 Similarly, the point about criticism does not show that different interpretations of a text are, even if apparently incompatible, equally acceptable or that a text has as many meanings as there are interpretations of it. Readings are neither arbitrary nor self-

validating simply because they are all subject to revision. Newer readings are always guided by the strengths and weaknesses of those which already

21. Meiland is clear on the dependence of textual meaning upon interpretation (see "Interpretation," p. 36), but he thinks that it results simply from the interpretation of

physical marks as words and thus attributes to it a privileged status. 22. Derrida, "Structure, Sign, and Play in the Discourse of the Human Sciences," in

The Structuralist Controversy, ed. Richard Macksey and Eugenio Donato (Baltimore, 1970), p. 248; rpt. in Writing and Difference, pp. 278-93. Cf. Miller, "Critic as Host," p. 218: "Is the 'obvious' reading, though, so 'obvious' or even so 'univocal'? ... Is not the obvious

reading perhaps equivocal rather than univocal, most equivocal in its intimate familiarity and in its ability to have got itself taken for granted as 'obvious' and single-voiced?"

23. See Culler, StructuralistPoetics, pp. 244-45, for an elaboration of Derrida's position. The view that no part of the meaning of a text is given, which I have been supporting, bears close affinities to the approach of Stanley Fish. See, for example, "Interpreting the Variorum" (Critical Inquiry 2 [Spring 1976]: 473), where Fish attacks "the assumption that there is a sense, that is embedded or encoded in the text, and that can be taken in at a

single glance." I diverge from Fish in his inferring that meaning cannot be located in the text but in its readers' experiences. See also his "Literature in the Reader: Affective

Stylistics'" Self-Consuming Artifacts: The Experience of Seventeenth-Century Literature (Berkeley, 1972), pp. 382-427.

24. On this point, see Hilary Putnam, "Meaning and Reference," in Naming, Necessity, and Natural Kinds, ed. Stephen Schwartz (Ithaca, N.Y., 1977), pp. 119-32, and "The Mean-

ing of 'Meaning,' " Mind, Language, and Reality (Cambridge, 1975), pp. 215-72. For a different argument to this conclusion, see Larry Laudan, Progress and Its Problems (Berkeley, 1977), p. 141 ff.

Page 10: NEHAMAS, A

Critical Inquiry Autumn 1981 141

exist; and though this process may never stop, it is not for that very reason blind.

Jones has tried to supply stronger support for the pluralist thesis. He claims that interpretation, "the business of making sense of the text, of rendering it coherent," is necessarily "aspectival"; he understands

aspect as both "the point of view from which something is seen, and the

appearance or face of the object perused" (Philosophy and the Novel, pp. 182, 181). His conclusion is that since every interpretation involves a

viewpoint, and since no viewpoint (biographical, Marxist, psychoana- lytical, etc.) is privileged, different readings of a text, even if apparently incompatible, can be equally acceptable.

Now consider the following case. In The Metamorphosis, there is a

picture of a woman on the wall of Samsa's bedroom. A number of widely diverse readings of the story all take the picture as an object of Samsa's sexual interest. This unexciting fact is sufficient to show that though the

activity of interpretation can proceed from different viewpoints, its re- sults need not therefore be themselves different. Nor is it easy to show that if the results of different approaches are indeed different, then they are equally plausible. For we can, I think, produce a better (not simply a different) interpretation of the role this picture plays in Kafka's story. The text speaks of a glossy-magazine picture of "a lady done up in a fur hat and a fur boa, sitting upright and raising against the viewer a heavy fur muff in which her whole forearm has disappeared." Now Heinz Politzer describes this picture as "vulgar ... animallike"; Robert Adams thinks that it is of an "impudent salacity"; Hellmuth Kaiser claims that it portrays an "erotically active, aggressive woman"; and Peter Dow Web- ster takes this woman as an "earth-mother."25 These descriptions do not

correspond to anything in the text, but once they are casually introduced, they tend to become, for some, parts of the story itself, and the picture thus acquires an erotic content. It is a short step from this to finding sexual significance in the insect's covering the picture with his body in order to protect it from being taken from his room along with the rest of his furniture. But what we do know about the picture is that it comes from a magazine and that it is of no one in particular (which accounts, incidentally, for its sketchy description). It is an object of no character and no individuality. If anything about it is interesting, it is that while it seems to be a picture of no interest, Gregor has made a frame for it himself: this is the only productive work we know him to have done, the only thing he has actually made. What he is protecting from being taken

25. Heinz Politzer, Franz Kafka: Parable and Paradox (Ithaca, N.Y., 1966), p. 72; Robert M. Adams, Strains of Discord: Studies in Literary Openness (Ithaca, N.Y., 1958), p. 152; Hell- muth Kaiser, "Kafka's Fantasy of Punishment," Peter Dow Webster, "Franz Kafka's 'Met-

amorphosis' as Death and Resurrection Fantasy," and Corngold, "Metamorphosis of the

Metaphor," in The Metamorphosis, trans. and ed. Corngold (New York, 1972), pp. 153, 158, and 11, respectively. I quote from Corngold's translation.

Page 11: NEHAMAS, A

142 Alexander Nehamas The Postulated Author

away, by assuming a position dictated to him more by his anatomy than by his desires, is his only real creation, his only real possession. That his most expressive action has been devoted to framing and bringing into prominence an object which is not so vulgar as it is banal underscores the shallowness of Gregor's relationship to the world and the depth of his attachment to that shallowness.

Interpretation is therefore in one sense aspectival, but criticism is not for this reason less than "objective." Different ways of trying to un- derstand a text may well be equally legitimate: there probably isn't a

general argument to the effect that psychoanalytic criticism, for example, should not be practiced. But simply because an activity can be pursued in different ways, it does not follow that different results must be reached; nor that if they are, then they must be equally plausible.

Jones draws this stronger conclusion when he slips into considering interpretation no longer as an activity but as that activity's very product: "The background against which, or the viewpoint from which, we in-

terpret a text generally provides the most interesting differences between

interpretations, between the patterns of coherence different critics de- termine" (Philosophy and the Novel, p. 186). Just as "aspect" covered both

viewpoints and what is seen from them, so "interpretation" covers both the "business" of finding a pattern of coherence in a text and that pattern itself. Yet though it is necessarily true that to peruse an object we must

(geometrically, so to speak) do so through one of its appearances, faces, or aspects, what we peruse is not the appearance but the object. We cannot simply appeal to the different methods critics use in order to

justify their different readings, though we sometimes think we can be- cause we take what is true of the process of interpretation to be true of its product.

We are concerned with a pluralism of contents, not of modes or methods, with the view that the results of different approaches to a text, even if apparently incompatible, can be equally plausible parts or aspects of what the text means.26 This would be, for example, the view that

Gregor's metamorphosis stands as much for his alienation from a world of unproductive labor as for his regression to the anal stage because of an unresolved Oedipal conflict. Both views, so stated, seem plausible, and so does the pluralist position which tries to account for this ap- pearance. But if we look at the texture of these interpretations, their

plausibility ceases to be striking. The psychoanalytic reading, for ex-

ample, must construe Gregor's father's kicking the vermin when it is stuck in a doorway as an act of pure aggression; Gregor, by contrast, and quite correctly, sees it as his "salvation." The Marxist reading fails to account for the effect his sister's music has on Gregor just before his

26. Wayne Booth deals with issues generated by methodological pluralism in Critical Understanding (Chicago, 1979), but his discussion, especially pp. 284-301, extends to plu- ralism of contents.

Page 12: NEHAMAS, A

Critical Inquiry Autumn 1981 143

death. Interesting difficulties facing interpretation are usually found on this specific level; we often grant a particular reading plausibility by not

looking enough at its details.27 Actually, the proliferation of difficulties on this specific level may

make it seem again as though the deconstructive view that every reading is a misreading is correct. And, in a way, it is, since every reading can be confronted with contrary evidence. But the absence of a reading which cannot be improved, which accounts for every feature, does not make a text "unreadable": it only indicates that there is more to under- stand. There is no definite description, explanation, or theory of any- thing. And though replacement may not proceed from worse to better in every individual case, it tends on the whole to preserve good readings in order to supplant them with others that are better.

This implies that we understand The Metamorphosis better today than it was understood in the past and that we will come to understand it better in the future. And though no aspect of our understanding of the

story is given, we can on each occasion agree on the significance of some of its elements in a way which allows us to compare and evaluate, even if only tentatively, alternative readings.28 This is not obvious as long as we try to compare in general terms a Marxist, say, and a psychoanalytic reading of The Metamorphosis: on this level, to ask "Which is better?" is to ask a silly question. But the point begins to appear if we turn to the

specific, if less exciting, issues which we have been discussing here. Miller writes, correctly, that "the 'obvious or univocal reading' of a

[text] is not identical to the [text] itself"29--no reading ever is. Readings, interpretations, do not re-create or duplicate a text's meaning, they de- scribe it.3o To understand a text at all is to have an interpretation of it, and it is only in the light of one interpretation that we come to see, if and when we do, that a text can be read differently, that another inter-

pretation is better. Thus from the point of view of a particular reading, the meaning which a later interpretation will attribute to a text does not exist. Nevertheless, the later interpretation does not, in absolute terms, create that meaning: it finds a meaning which, from its own point of view, had always been there. This is so, of course, only from its own

point of view. But we cannot coherently describe this as "only a point of view" unless we produce yet another interpretation attributing to the

27. For the Marxist argument, see Bluma Goldstein, "Bachelors and Work: Social and Economic Conditions in 'TheJudgment, 'The Metamorphosis' and The Trial," in The Kafka Debate, ed. Angel Flores (New York, 1977), pp. 147-75 and 3-5. For the Freudian view, see Kaiser, "Kafka's Fantasy of Punishment," esp. p. 152.

28. Monroe Beardsley makes a similar point in his review of Booth's Critical Under-

standing in Philosophy and Literature 4 (Fall 1980): 257-65. 29. Miller, "Critic as Host," p. 224. 30. Stanley Cavell provides an excellent discussion of this issue in "Aesthetic Problems

of Modern Philosophy," Must We Mean What We Say? (New York, 1969), p. 74 ff.

Page 13: NEHAMAS, A

144 Alexander Nehamas The Postulated Author

text yet another meaning from yet another point of view (to be specified as such only by means of a further interpretation).

Meaning does not therefore reside in texts independently of all

interpretation, there to be discovered once and for all or, if we are not

lucky, to be forever lost; but this is not to say that it is fabricated. The critical monism which I advocate is a regulative ideal and identifies the

meaning of a text with whatever is specified by that text's ideal inter-

pretation. Such an interpretation would account for all of the text's features, though we can never reach it since it is unlikely that we can even understand what it is to speak of "all the features" of anything. What we do have (and that is what we need) is the notion of one inter-

pretation answering more questions about a text than another and thus

being closer to that hypothetical ideal which would answer all questions. The direction in which this ideal lies may change as new interpretations reveal features of a text previously unnoticed, rearrange the significance of those already accounted for, or even cause us to change some of our

general critical canons. And though, in this way, there may not be a

single ideal interpretation of a text toward which all of our actual inter-

pretations in fact lead, the transition from one interpretation to another can still be rational and justified. To interpret a text is to place it in a context which accounts for as many of its features as possible; but which features to account for, which are more significant than others, is itself a question conditioned by those interpretations of the text which already exist.31

To interpret a text is to place it in a context, and this is to construe it as someone's production, directed at certain purposes. A purpose is neither the end toward which motives aim, nor a text's "perlocutionary" effects, nor again a message lying behind the surface.32 Meaning is a

symbolic relation, and what an object symbolizes depends partly on which of many systems it can be construed as an element of.33 At least the choice of symbol system is an intentional act, and to appeal to intention is to appeal to a particular explanation of why a text, or one of its features, is as it is. The picture of the woman is used in The Metamorphosis

31. Putnam discusses such a view in relation to the philosophy of science and epis- temology in "Realism and Reason," Meaning and the Moral Sciences (London, 1978), pp. 123-40.

32. The first view is held by Skinner, "Motives," pp. 401-2. For the perlocution view, see Meiland, "Interpretation," p. 39 ff, and Skinner, "Motives," p. 403. For the meaning- as-message view, see, e.g., Richard Kuhns, "Criticism and the Problem of Intention,"Journal of Philosophy 57 (January 1960): for Kuhns, interpretation is the activity of "getting at the

message which may go beyond the plain literal sense" (p. 7). The notion of meaning as

message has recently been criticized by Wolfgang Iser in The Act of Reading (Baltimore, 1978), though he seems to me to conclude too quickly that meaning is "imagistic in char- acter" (p. 8).

33. This is one of the central theses in Nelson Goodman's Languages of Art (India- napolis, 1968), though Goodman avoids any discussion of intention.

Page 14: NEHAMAS, A

Critical Inquiry Autumn 1981 145

to show the banality of Gregor's life. This account is intentional in that it is teleological. But it is not thereby vulnerable to those sound arguments against appealing to intention construed as "design or plan in the au- thor's mind" or as the efficient cause of that feature.34

To interpret a text is to consider it as its author's production.35 Literary texts are produced by agents and must be understood as such. This seems to me self-evident; even deconstructive criticism generally accepts it, though it insists that the choice of agent is conventional and

arbitrary. And since texts are products of expressive actions, understand-

ing them is inseparably tied to understanding their agents. But just as the author is not identical with a text's fictional narrator, so he is also distinct from its historical writer. The author is postulated as the agent whose actions account for the text's features; he is a character, a hy- pothesis which is accepted provisionally, guides interpretation, and is in turn modified in its light. The author, unlike the writer, is not a text's efficient cause but, so to speak, its formal cause, manifested in thought not identical with it.36

A methodological constraint on this view is that the postulated au- thor be historically plausible; the principle is that a text does not mean what its writer could not, historically, have meant by it. For example, we cannot attribute to particular words meanings which they came to have only after the writer's death."3 What a writer could mean can be deter- mined by linguistic or biographical considerations but also by facts about the history of literature and the world, psychology, anthropology, and much else besides, a change in our understanding of any of which can cause us to change our understanding of the text.

Meaning therefore depends on an author's intentions even if a writer is not aware of it. Since the author's intentions depend on what the writer could have meant, a text's meaning is to that extent a thing of the past, though its understanding is itself a thing of the future.

34. See W. K. Wimsatt, Jr., and Beardsley, "The Intentional Fallacy," in Wimsatt, The Verbal Icon (Lexington, Ky., 1954), for the source of most of those arguments. Cavell ("A Matter of Meaning It," Must We Mean What We Say?, pp. 234-37) disagrees with Beardsley but offers a different account of intention.

35. This position may seem to transgress against the original New Criticism, struc- turalism (e.g., Roland Barthes, Sur Racine [Paris, 1963]), more recent theory sympathetic to the New Critics (e.g., John Ellis, The Theory of Literary Criticism: A LogicalAnalysis [Berkeley, 1974]), deconstruction, and approaches to interpretation via the activity of reading (e.g., Iser, The Act of Reading, and Fish, "Literature in the Reader" [n. 23 above]).

36. My postulated author is not unrelated to Booth's "implied author," to whom he appeals both in Critical Understanding and in The Rhetoric of Fiction (Chicago, 1961), and to Kendall Walton's "apparent artist," discussed in his "Style and the Products and Processes of Art," in The Concept of Style, ed. Berel Lang (Philadelphia, 1979), pp. 45-66. I discuss this issue in more detail in another paper, "What an Author Is" (to be presented at the MLA convention, December 1981).

37. Beardsley argues that we can attribute such meanings to texts; see Possibility of Criticism, p. 19.

Page 15: NEHAMAS, A

146 Alexander Nehamas The Postulated Author

Without Freud we would not have seen the sexual elements which are now part and parcel of Oedipus Rex. But if the Oedipal conflict is as basic to behavior as Freud thought, then the historical Sophocles, unaware of it as he may have been, could have considered it an issue. And we can

argue from this that the character Sophocles, the play's author, did consider it an issue; it is then part of the play's meaning even if we could not have realized it until this century.

We must not, by contrast, accept a view of The Metamorphosis which holds that hours on the clock correspond to years in Gregor's life and that he

should have caught the five o'clock train for work, that is, a psychic change should have occurred at the normal age of five, . .. [that is,] the formation of the superego. . . . But here it is, already six- thirty (Gregor is six and a half years old); he has missed the train or psychic energy necessary for progression.38

Kafka could not have known this highly technical, and highly doubtful, theory of development. Even if the theory were true, it simply lacks the

power and generality of the Oedipal conflict which might convince us that Kafka could have come by it on his own and that it therefore belongs to the story.

The principle of the postulated author is not sheer invention. We can find it reflected in the practice of a critic like Quentin Skinner, who refuses to read some seventeenth-century legal texts as concerned with the doctrine of the judicial review of statute because the concept of

judicial review did not arise until the next century.39 Adams, to cite a

clearly literary case, interprets the number three in The Metamorphosis as a symbol for masculinity, on the grounds that "Kafka might have learned of the association through any of several channels" though there is no evidence that he did.40 Finally, Miller supports his view that Stevens' rock, in the poem of that title, stands for literal language partly because "Stevens might even have known (why should he not have known?) the world 'curiologic'... [from the] Greek kuriologia, the use of literal expres- sions."41

Now in one sense there is something arbitrary about constructing a historically plausible figure as a text's author. In principle we could

always construct a different context and a different author and so give an unhistorical reading. This is not unlike the arbitrariness of our in-

38. Webster, " 'Metamorphosis' as Death and Resurrection Fantasy" (n. 25 above), pp. 161-62.

39. Skinner, "Motives," pp. 406-7 and n. 41. 40. Adams, Strains of Discord (n. 25 above), p. 173. 41. Miller, "Stevens' Rock and Criticism as Cure," Georgia Review 30 (Spring 1976):

10.

Page 16: NEHAMAS, A

Critical Inquiry Autumn 1981 147

terpreting representational paintings as projections of familiar Euclidean

space, since any two-dimensional figure can be construed as the projec- tion of indefinitely many alternative worlds. Just as the effort to construct such worlds could be worth making, we could always try to read a text

differently, postulating a different author and progressively refining our

conception as new readings come to affect those from which they emerged. Progress in this direction would show a text to be, in a genuine and valuable sense, polysemous. But what we actually find in criticism is a number of self-consciously partial alternatives, directed only at some of the text's features in the expectation that many more partial, non-

competing readings will emerge. This can no more support the view that each text, as a whole, has many meanings than can the claim that, given any interpretation of a text, a different one could always be con- structed. What we need in both cases is an actual reading at least as

general and powerful as the reading whose uniqueness is being ques- tioned and with which it is incompatible.42

The monism I have presented is not threatened by the existence of

many partial readings of a text since it can exploit discoveries made

through such readings in pursuing a more complete understanding of the text. Methodological pluralism is compatible with a monism of con- tent. The regulative end is to construct, for each text, a complete his-

torically plausible author-a character who may not coincide with the actual writer's self-understanding, fragmentary and incomplete as it

probably is. What a writer takes a text to mean is relevant but not telling evidence in literary criticism. Further, our construction will never be

complete: in constructing the author of The Metamorphosis, we shall have to consider his close relation, perhaps his identity, with the author of The Castle, whose precursor is Kierkegaard (another character who may ap- pear different through this connection) and who is in turn the precursor of Jorge Luis Borges and other future authors. Changes in literature and in everything that is relevant to it (in everything, that is) will change the constraints imposed upon the postulated author. And there is no reason to think that we shall ever abandon this construction, except perhaps that our interest may one day be exhausted.

Critical pluralists sometimes argue as Jones does that all interpre- tations are necessarily partial, that criticism "is always ... of a selection of properties" (Philosophy and the Novel, p. 193). If different critics were in fact necessarily concerned with different textual features, it might follow that different readings, being nonoverlapping, were equally ac- ceptable.43 But this is no more true than the claim, often made in this

42. Booth makes a similar point in more detail in Critical Understanding, pp. 168-69. 43. Though, strictly speaking, what would follow is that distinct parts of a text have

distinct meanings, not that the text as a whole has more than one meaning.

Page 17: NEHAMAS, A

148 Alexander Nehamas The Postulated Author

connection, that to see a feature differently is to see a different feature.44 Like all theorizing, interpretation is based on some features of a text but is of the text as a whole. It is therefore partial only in the trivial sense that no reading can ever account for all of a text's features, not because distinct readings are directed at distinct features. Being partial in this sense means simply that there is no final, unrevisable interpretation of

any text; but then why should we want such a reading? The aim of interpretation is to capture the past in the future: to

capture-not to recapture-first, because the iterative prefix suggests that meaning, which was once manifest, must now be found again. But the postulated author dispenses with this assumption. Literary texts are

produced by very complicated actions, while the significance of even our

simplest acts is often far from clear. Parts of the meaning of a text may become clear only because of developments occurring long after its com-

position. And though the fact that an author means something may be

equivalent to the fact that a writer could have meant it, this is not to say that the writer did, on whatever level, actually mean it.

Second, the notion of recapturing the past suggests the repetition of an earlier act of consciousness and generates the hermeneutic problem of how we can put ourselves in the position of another epoch or culture, how we can see the world as they saw it, "from within."45 Skirting most of the issues, I will simply deny that the aim of criticism is to re-create the original understanding of a text-particularly if by this we mean the

experience of an original audience. It is quite true that we can never re- create the experience of the original audience of, say, The Clouds. It is not even clear whose experience we should pursue: the common ex-

perience of all the spectators of the comedy's first performance? that of an "average" member? the experience of Socrates or of Anytus? It is

equally true that we cannot hope to re-create even our contemporaries' experience of a text. And if it is claimed that my experience of The Clouds is more similar to yours than it is to Anytus', I will agree only if you can show that our interpretations are more similar. But if this is how we construe "understanding," then capturing the Athenians' understanding of The Clouds presents no theoretical difficulties. We must find what they thought, said, and wrote about it. This is in principle public information whose loss, even if it is final, poses no logical or hermeneutical problems.46

44. Jones sometimes suggests this with his notion of the "interpreted-text"; see Phi-

losophy and the Novel, p. 193. 45. For a discussion of problems connected with the hermeneutic circle, see Anthony

Savile, "Historicity and the Hermeneutic Circle," New Literary History 10 (Autumn 1978): 49-70, and "Tradition and Interpretation," Journal of Aesthetics and Art Criticism 36 (Spring 1978): 303 and 315.

46. If it is now argued that even if we had the Athenians' interpretation, we could never really understand it, we can again apply the previous dilemma. We cannot understand their views because we cannot experience things as they did (which applies to any com- munication whatever) or because we lack background knowledge which they possessed (and which raises no problems in principle).

Page 18: NEHAMAS, A

Critical Inquiry Autumn 1981 149

Still, criticism does not aim to capture what a text's original audience

actually took it to mean but to find what the text means. We want to

develop an interpretation which will be consistent with what we know about a text's language, its writer, its original audience, its genre, the

possibilities of writing, history, psychology, anthropology, and much else. What a text means is what it could mean to its writer. But this is not what it did mean to the writer and to the text's original audience, nor need they have been able to understand it given only the articulated

knowledge of human affairs which they then had. The meaning of a

text, like the significance of an action, may take forever to become man- ifest.

Some critics believe that many texts, or parts of texts, have been

correctly interpreted once and for all.47 This, I have tried to suggest, is

unlikely. Others, fearing perhaps that a final interpretation will make the text itself dispensable, deny this possibility on the grounds that all texts are essentially ambiguous and thus always open to new readings.48 I have tried to show that we do not need to accept this latter view in order to justify what is, after all, the most basic consequence of the open- endedness of all knowledge.

As I stated earlier, though texts belong to the past, their under-

standing belongs to the future. To do just what I have said we shouldn't, let me quote, quite out of context, Sidney's "The poet ... doth grow in effect another nature." Consider this nature not as the world the text

represents but as the text itself. Each text is to our many interpretations what nature is to our many theories, and each is inexhaustible. Under-

standing a text is, in two ways, a historical enterprise: not only does it

employ history but it also unfolds in time and depends on everything we now do and will come to know about the world, which includes ourselves. Understanding a text is as easy, or as difficult, as that. In

interpreting a text we must come to understand an action, and so we must understand an agent and therefore other actions and other agents as well and what they took for granted, what they meant, believed, and what they wanted. For this reason, each text is inexhaustible: its context is the world.

47. See Savile, "Tradition and Interpretation," p. 307; Abrams, "Rationality and

Imagination" (n. 19 above), p. 457; and Hirsch, Validity in Interpretation, p. 171. 48. Iser, for example, seems to think that the dispensability of the text is a danger

which taking meaning as "imagistic" avoids; see The Act of Reading, p. 4 ff. But this fear is widespread; see Cavell's discussion of Cleanth Brooks and Yvor Winters on paraphrase in "Aesthetic Problems." Claims to the effect that what is important about literary texts is not what they mean but their "emotional impact" are prompted by such a fear.