Document Type


Date of Award



Conrad, Joseph, Criticism and interpretation

Degree Name

Doctor of Philosophy (PhD)


English, General Literature, and Rhetoric

First Advisor

Robert Kroetsch

Second Advisor

Bernard Rosenthal

Third Advisor

William B. Stein


The first critics of the novel—some supposing themselves practical, some simply Puritanical—tended to find that new form frivolous or even pernicious. Consequently, certain early novelists attempted to circumvent condemnations soon standard by concealing their novels beneath the artifice of other more accepted forms of discourse, the diary or epistle, for example. Or, like Fielding, they could maintain that their new creation was essentially a modified version of an established genre. Thus Joseph Andrews becomes a comic epic in prose. Other early novelists, however, pursued a different course. They argued the ethical validity of their work by asserting that they presented “truth.” The novel, they insisted, served as a mirror, or even a corrective lens, in which reality was effectively imaged. And so such books as Moll Flanders, Roderick Random, or the first American novel, The Power of Sympathy, all begin with the prefatory fiction that the fiction to be presented is fact.

Clearly, our earliest novelists proceeded most into the realms of fiction when they advanced this mimetic justification for their work. As Frank Kermode has argued, the novel, unlike a treatise, must reduce and transform fragments from the outside world in order to create a different new world, the universe of fiction. While humans live perpetually in medias res, novels present an illusion of completeness and imply that life, like an Aristotelian plot, has three distinct parts: a discernible beginning, a comprehensible middle, a definite end. Therefore, despite theoretical asides to the contrary, fiction has always—by its nature—violated the demands of “realism” when that term is used in its strictest sense. Unlike reality, novels end. Fiction—just as early critics forewarned—“has to lie” (Kermode 140).

But the novel’s intrinsic inability to imitate the larger world in any exact fashion has, perhaps, a more fortunate consequence. Novelists have long been self-consciously concerned with the ramifications of invention. Writers as different in time—but not in technique—as Laurence Sterne, on the one hand,and John Barth, Donald Barthelme, and Jorge Luis Borges, on the other, fictionally assess the complex relationship between possible fact and probable fantasy, between realistic content and artistic form. In other words, these experimental novelists explicitly focus on a tension implicit in all fiction. The novel must be both a mirror and a model; it must reflect aspects of life which can be recognized and comprehended by the reader but it also must give the experience portrayed within the work a cohesion and organization seldom seen in the flux of unstructured experience. And this tension demonstrates the naiveté of those who insist that novels be simply true. “A mirror carried along a high road,” Stendhal's phrase, which certainly does not describe Stendhal’s novels, would give us only a formless series of images, not a work of art at all.

There is, then, in any fiction, a necessary conflict between argument and art. Consequently, it is not surprising that only the greatest writers can be “true to life” and true to the intricate design of their novels at one and the same time. Yet even these novelists frequently have problems with endings, for, in the ending, the demands of probability and the demands of pattern are most in conflict. Probability, as already noted, dictates that experience does not neatly conclude. Pattern, particularly Aristotelian pattern, requires that it must. Only then is the audience left “calm of mind, all passion spent” (Milton 558). Experience, moreover, must not only end, it must terminate in such a fashion as to validate the work’s essential design. Even such great novels as Crime and Punishment and Middlemarch do not fully resolve this difficulty. Both Raskolnikov’s eventual reformation and Dorothea’s second marriage can be seen as deriving more from the final exigencies of plot than from the ultimate promptings of revealed character.

As Frank Kermode, throughout much of The Sense of an Ending, argues, every novelist, upon approaching the conclusion of his story, must confront the gulf that separates story from life and character from person. Moreover, whether he concludes with some summary tag (“they lived happily ever after”) or with an ambiguous and uncertain situation (the protagonist at a crucial crossroad with both reader and character uncertain as to which path will be—or should be—taken), the novelist still implicitly addresses the larger question of the appropriateness of any ending. If the final reassuring tag does not resolve the basic conflicts of the novel as a whole (a charge sometimes leveled, for example, at Dickens’s revised ending for Great Expectations), it hardly provides a real, satisfactory conclusion. Conversely, in a novel centering on indecisiveness and confusion, a continuing uncertainty might be more "complete" than some final choice made by an anti-hero protagonist.


My basic point is simple. If, as Friedman and others argue, fictional endings must express predominant views about man’s place in his society and world, how then can apparent exceptions (those briefly discussed as well as others not mentioned) be explained? If we assume that eighteenth-century Englishmen were somehow closed, what do we do with Fielding who could produce both open and closed works? Clearly, the relationship is not as direct as Friedman would suggest. Earlier centuries may not be so patterned, so monolithically the same, as this critic’s arguments imply. Even if it took the twentieth century to fully appreciate the intricacies of Moby-Dick, the eighteenth century still greeted each new volume of Tristram Shandy with considerable excitement. What I am really arguing is that writers such as Sterne, Brontë, Thackeray, and Melville were not mere freaks, “sports” who were born decades or centuries ahead of their time and who consequently wrote with a brilliance and complexity that only contemporary readers can recognize and appreciate. To suggest this is to oversimplify and to do these earlier novelists and the time in which they wrote an injustice, one that Richter has rightly corrected.


E. M. Forster’s criticism of a certain kind of novel which follows a too “rigid pattern” might here be recalled. Such a book, he notes, “may externalize atmosphere, spring naturally from the plot, but it shuts the doors on life and leaves the novelist doing exercises, generally in the drawing room” (234). One obvious implication of this observation is that critical rigidity can also lead to rather pointless mental calisthenics. Thus, although the following chapters are substantially indebted to theories on endings and analyses of different endings advanced by the various critics whom I have quoted in my introduction, I hope to avoid what seems a common problem. As I believe I have shown, a priori assumptions about the necessary nature of conclusions mar the work of these scholars and often lead them to overlook the complexity of the individual books that they assess. Consequently, I am not primarily concerned with the problem of deciding if Conrad’s endings are “open” or “closed,” even though I make extensive use of such terminology. Indeed, Richter’s “closed” book—closed on the deeper level he implies with that term—could well be Friedman’s “open” one.

In one sense, then, it is probably impossible to prove, to the satisfaction of all critics, that Conrad is predominantly a writer of “open-ended” or “closed” books or that he is the transitional figure Friedman posits. But that problem does not at this point seem crucial. Instead of attempting to fit an assessment of Conrad’s endings into the procrustean bed of a specific terminology and corresponding ideology, it is more productive to analyze how his endings actually work. How does Conrad terminate a book? I will answer this question by examining the five major novels to show how the final sections of each especially illustrate the larger meanings suggested in the work as a whole. In brief, the primary purpose of this study is to demonstrate just how carefully Conrad structured his conclusions and how a careful assessment of a specific ending also argues a general interpretation of the book thus concluded.