Edward Said’s new book is a disconcerting example of a rational position sapped by alarming faults. As a practical and theoretical critic of literature, Said sacrifices accuracy and good sense to self-indulgent carelessness. His collection of essays deals with a range of writers from Swift to Conrad, with philosophers from Plato to Derrida, and with scholars from Ibn Hazm of the eleventh century to Professor Gerald Graff. But wherever I scrutinize the reasoning or try to verify the evidence, the weakness of his accomplishment disturbs me.

The main direction of Said’s theoretical argument is that a proper understanding of literary works is impossible without a knowledge of their historical setting. A literary text, he says, is “always enmeshed in circumstance, time, place, and society” (p. 35). In taking account of these facts, Said believes, a critic should also take account of social change, particularly of the relations between the classes ruling a nation and those that produce its wealth but have small part in its government (pp. 168-173). Said suggests that the design of a literary work reveals—if only by exclusion—the social and political strains, the processes of “acquisition and appropriation,” that make a society equitable or corrupt (p.23).

Said establishes his theoretical position in his opening chapters, and cites texts by Hopkins, Wilde, Marx, and a number of novelists, to demonstrate how literary works can involve the circumstances of their own production. For Said, the essay is the proper genre for criticism—and criticism is itself an activity that works into and out of a particular social world and historical moment.

After applying his principles to the writings of Swift and Conrad, Said returns to general issues. He deals with the relation of historical cycles, or “repetition,” to originality, or change, in the evolution of culture. Relying on Vico, Kierkegaard, and Foucault for his examples, Said argues that historical patterns are not merely innate but develop from responses to experience (p. 113).

So also, when he examines the relation between theory and originality, Said treats theoretical issues as springing from the historical situation of the critic himself. He argues therefore that the concept of “textuality” has made old notions of originality untenable. The term can no longer mean merely the first instance of a phenomenon but rather “a faculty for combinatorial play” (p. 138).

In other essays, Said surveys the practice of literary criticism today, calling attention to the new preoccupation with the way a text “functions,” but also regretting the tenuous linkage between criticism and the history of institutions (p. 155). He demonstrates the ironic ambiguity of the term “leftist” as a description of critics who seem to oppose established academic practice and yet whose literary essays seldom bear on political or social action (p. 162). Here he holds up Gramsci and Raymond Williams as examples worth studying. In the weightiest theoretical essay of his book, Said compares Derrida with Foucault, deeply appreciating the achievement of both men (and finding fault with both) but finally preferring the method of Foucault.

The most informative aspect of Said’s scholarship does not, however, lie in his views on English literature or literary theory. It is, rather, his account of European Orientalism. In discussing Raymond Schwab, Ernest Renan, and Louis Massignon—all writers who tried to reinterpret the relation of European to “oriental” culture—Said sets fresh and important vistas before us.

The weakness of Said’s work is obviously not due to his choice of themes. It derives, rather, from the inadequacies of his knowledge and method. Some readers will be put off by the solemnity of Said’s attention to his own career. In the opening chapter he is perhaps too candid in relating how he came to write the pieces here assembled. Do we really wish to hear about three other books and one essay by the author? This willingness to share his impression of his own accomplishment with the reader becomes one of the less seductive aspects of Said’s rhetoric.

If I sound captious about immodesty, let me add that not only does much of the book consist of paraphrases of the work of others, but Said often leaves it uncertain how far he accepts or rejects the principles he expounds. Take the analysis of the essay as a literary form (pp. 50-53). Here Said raises ten questions, leaves them largely unanswered, and then sketches the “metaphysics” of the essay as elaborated by Lukács, with Socrates displayed as the “typical essayistic figure” (p. 52). It is a triumph of a sort to produce as a “typical essayistic figure” a man who never wrote a literary work, and whose mode of expression for his doctrines was the dialogue. One is unsure, however, whose triumph it is.

Not that Said lacks the will to tackle Socrates directly. On the contrary, he confronts but obscures the text of Plato. In the Phaedrus, Socrates draws a contrast between writing and speech. Words written in ink (en hudati melani) cannot defend themselves, and therefore cannot teach the truth effectively (Plato, 276C). Words spoken in the course of the dialectical method, however, may be planted in a fitting soul, for the speaker can suit his language to the nature of his audience (Plato, 276E-277).


Said transforms such undramatic clarity into a mysterious distinction between words “written in water” and “an intelligent word” (p. 131); and he proceeds to employ this distinction to show that for Plato, theory and originality coincide. The conclusion is more inventive than it may seem, for the word “theory” does not appear in the Greek (theoron, or “looking at,” appears in 276D).

Soon after the passage cited, Plato has Socrates talk simply about the man who can defend his written work with a knowledge of the truth as shown in spoken discussions (278C). But Said darkens the expression by claiming that Plato here brings together two sorts of men—“The lover and the lover of knowledge” (p. 132). Socrates consistently sets the mere writer against the dialectical speaker, whose faculty is essential to the true philosopher. But Said reverses the sense and declares that we may translate Socrates’ “lover of wisdom” (278D: philosophon) as ” ‘seekers after’ writing” (Said, p. 132).

Turning his attention to less ancient figures, Said misleads us again. Having a special interest in Swift, he devotes two chapters to the man and his works. It is characteristic of Said to enliven his account of an author with departures from received opinion. So we learn from a casual remark that the institution of monarchy was central to Swift’s doctrine (p. 74). It would be pleasant to hear Said illustrate this notion with Swift’s well-known couplet,

Thus think on kings, the name denotes
Hogs, asses, wolves, baboons and goats.
—“On Poetry”

Older scholars have long believed that though Swift complained bitterly about Ireland in letters to English friends, and though he sneered at the governing classes of the kingdom, he nevertheless came to identify himself with the Irish people. He certainly failed to visit England when it was easy for him to do so, and he refused opportunities to settle himself there on a good income. Scholars have also thought Swift was Irish by birth and education, and that his career as Dean of St. Patrick’s Cathedral allowed him to accumulate a fortune that could have supported a hundred modest families for a year.

But Said does not accept these facts. We learn now that the English were Swift’s fellow countrymen (p. 86), that Swift died “as alienated from Ireland as he had been, over twenty years before, from England,” and that he never accumulated “anything resembling a fortune” (p. 83).

When he discusses the writings of Swift, Said is no less independent. He celebrates the high standard of modern editions of the satirist’s work. Herbert Davis’s text of the prose, Said has discovered, is formidably scrupulous. So painstaking is this “textual orthodoxy” (to use Said’s novel terminology) that approaching Swift has become “a daunting prospect” (p. 73).

As one of the coeditors of Davis’s edition, I thank the Parr Professor of English and Comparative Literature for his good nature. I had myself supposed that Davis’s volumes abounded in errors of every description, the text of Gulliver’s Travels being peculiarly unreliable. So also the edition of Swift’s correspondence by Sir Harold Williams, in five volumes (Said finds there are seven), has seemed to me sadly inaccurate both in the text and in the notes.

When Said braves the daunting prospect and looks into the contents of the editions, he comes back with remarkable observations. In an offhand reference to Swift’s Modest Proposal, he suggests that the irony of the satire “completes itself in the reading; there is nothing to check it against” (p. 87). Yet those readers who persist to the end of Swift’s essay meet a paragraph in italics, listing economic policies against which the irony is indeed to be “checked,” for here Swift offers a positive program that interrupts the ironical tone.

Rather than limit us to a single fact, Said sometimes offers a choice between alternatives. On one page we are told that “proper words in proper places” is Swift’s definition of good style (p. 60); a little later, it becomes Johnson’s (p. 74).

On larger aspects of Swift’s art, Said remains original. For example, he has found out that Swift invariably attacks what he impersonates (p. 87). If this is so, we must believe that in Swift’s letter on preaching, the gentleman supposed to be advising a young candidate for the priesthood does not really deliver Swift’s opinions (as readers have normally thought) but rather is condemned by the author. We must now also presume that the Examiner papers, which scholars have read as supporting a Tory government, were designed to ridicule the imaginary figure whose character Swift adopts.


The social and political history of Swift’s era is blurred by Said, for he has a habit of leaping over common distinctions. So he speaks of the landed aristocracy, the Church, and the monarchy (in the early eighteenth century) as sharing the same “Tory” values (pp. 82-83). Since Anglican bishops were likely to be Whigs and parish priests Tories, since Queen Anne favored the Tories and George I the Whigs, since the great aristocrats were often hostile to the monarch and the lower clergy generally disliked King George, Said’s synthesis is a large feat of critical fantasy.

Later centuries also suffer from Said’s inventiveness, as we may see from his examinations of Renan and Conrad. The common view of Renan’s career is that he made his way by genius, against entrenched opinion. Having broken with the Church in his early twenties, he could hardly aspire to a career in the usual institutions of learning. Although he went abroad on missions of scholarly research, sent by the French government, he lived on a modest income from a position in the Bibliothèque Nationale and from the savings of a devoted sister. Only when he was nearly forty did Renan win the professorship he deserved, in the Collège de France. But he lost this on account of a notorious first lecture, in which Renan seemed to doubt the divinity of Christ; and he was quickly relegated to a sub-librarian’s place, which he quickly refused. After years as a professional author, Renan did secure official recognition and honorific appointments. But conventional biography has seen an unwillingness to compromise with the demands of the establishment as a distinguishing feature of Renan’s character.

Yet according to Said, this judgment must now be discarded. For Said, Renan was so integral a part of the French culture of his epoch that his work on Islam obtained “a far greater status and authority” than it would have got in England (p. 274). Quoting Renan on his own situation, Said comments that although the word “moi” seems lonely, “it is in fact supported by all sorts of institutions” (p. 280). Indeed, says Said, everything about Renan “exudes the authority of massive centralized institutions like schools, disciplines, missions, teams of cooperating but hierarchically arranged scientific workers” (p. 280). Not the least defect of this opinion is that Said produces little evidence for it.

So also the narrative fiction of Joseph Conrad comes under the ill-focused attention of Said. With one exception, we are told, in all Conrad’s major works, “the narrative is presented as transmitted orally” (p. 94). It certainly piques one to see The Secret Agent (not Said’s exception) thus dethroned and excluded from the list of “major” achievements, along with other tales normally praised.

After (implicitly) removing The Secret Agent from the rank normally allowed the novel by merely appreciative critics, is it not an oversight for Said to find in it the most striking image of Conrad’s idea of literary creation as a painful task (pp. 96-97)? This image is the picture of the unfortunate Stevie drawing circles. Is Said, in the manner of a French genius, making a marginal work, rather than a canonical opus, the key to an author’s art?

At moments, Said, in effect, emends evidence when it does not meet his needs. For something like this happens in his treatment of a crucial quotation from Conrad. The heedless novelist was trying to explain why he wrote in English rather than another language, and said the attempt “would be as impossible as trying to explain love at first sight.” * However, the critic wishes to show Conrad troubled by “the dissemination, reception, and perception of language” (p. 99). Said omits the words I have cited above from the long passage in which they occur, and follows up the truncated quotation with a paraphrase that supports his own argument.

As a literary theorist, Said disturbs one with his lack of precision. Where an argument begs for sharp definition and cogent reasoning, Said becomes wordy and vague. In the longest chapter of the book, he considers two theories of “textuality,” that of Foucault and that of Derrida. Although he praises and criticizes both, Said’s censure of Derrida is hardly penetrating.

A pervasive feature of Derrida’s method is the principle that we cannot appeal from any text to an immediate experience of reality: there is nothing outside “le texte.” But Said proceeds to blame the celebrated philosopher for failing to consider the “historical density, specificity, and weight” of the works which Derrida analyzes (pp. 210-211). Now in Derrida’s terms, if “historical density,” etc., are available to us, it is as further text; and in that form Derrida may indeed consider them. But otherwise, they belong to the very suppositious reality that he finds inaccessible. Said’s language is evocative and metaphorical where the utmost acuteness is required.

Said’s prose style, in keeping with his limitations as a scholar or critic, is marked by unfortunate idioms. In the language of Morningside Heights, “schema” is plural (pp. 130, 154), and “Zusammenhänge” is singular (p. 148). The correlative “both” is not parallel to “and” but to “as well as” (p. 174). “In the future” becomes “into the future” (p. 232), and “hatred for reason” becomes “hatred to reason” (p. 281). One does not effect a transfer; one “transacts” it (p. 24).

At the same time, abstractions and literary genres may take over the role of persons; and instead of an author’s feeling his work is marginal, we may find the work doing so: “What is the essay’s consciousness of its marginality to the text it discusses?” (p. 50). A work can also fail to be conscious: “Yet Foucault’s earliest work was in many ways remarkably unconscious of its owntheoretical force” (p.243).

Similarly, a situation may feel dissatisfied, as when Said says of the contemporary literary situation, “In no way does this situation have as coherent a defining characteristic as in its profound dissatisfaction with the units, the genres, the expectations of earlier times” (p. 135).

Is one pedantic to feel unhappy with such additions to the vocabulary of criticism? I trust that even those who accept Said’s doctrines will not feel bound to echo his manner of expression.

This Issue

January 19, 1984