SOURCE: Lamb, Robert Paul. “Hemingway and the Creation of Twentieth-Century Dialogue.” Twentieth Century Literature 42, no. 4 (winter 1996): 453-80.

[W]hile one can do nothing about choosing one's relatives, one can, as artist, choose one's “ancestors.” … Hemingway [was] an “ancestor.”

—Ralph Ellison (140)

In July 1961, the Saturday Review devoted a special memorial issue to Ernest Hemingway, in which writers and critics from around the world paid tribute to the recently deceased author and attempted to assess his impact on their own national literatures. Although the Hemingway mystique was given heavy emphasis, many contributors also spoke to his artistic influence. The exiled Spanish political philosopher Salvador de Madariage observed that “Hemingway's manner of writing, his direct, simple, yet forceful prose” had “exerted an undoubted influence on the new generation of Spanish novelists” (18). From Italy, novelist Carlo Levi credited Hemingway's art as fundamental “in determining the character and mode of thought of our time” (19). And Alan Pryce-Jones, the former editor of the Times Literary Supplement, asserted that there was “not a living writer in England who has been unaffected by the laconic speed of his dialogue, the subtle revelation of character that lies behind a spoken phrase” (21). Today, such claims remain undisputed; most critics take for granted that Hemingway's techniques have profoundly influenced subsequent generations of writers across the boundaries of nationality, gender, race, ideology, sexual orientation, class, religion, and artistic temperament.¹

Pryce-Jones ventured that Hemingway's art, especially his innovative dialogue, might “turn out to be his enduring memorial as a writer, whatever his fascination as a man” (21). However, in the years since his death, Hemingway criticism has focused more on the biographical, thematic, and cultural content of his work than on his narrative techniques, and while it is true that his prose style has been exhaustively analyzed and countless passages of his dialogue read for content, there exists not one single systematic or even a sustained analysis of his art of dialogue. The following essay attempts to redress that neglect. Through a close examination of passages from three stories, written between 1923 and 1927, it will show how Hemingway evolved the techniques that would change the nature of twentieth-century fictional dialogue. The passages are drawn from “Indian Camp,” in which he for the first time employed the characteristic devices that distinguish his dialogue; “A Canary for One,” in which he elevated banality in speech to the level of art through the extension of repetition to dialogue; and “Hills Like White Elephants,” in which he blurred the line between fiction and drama, allowing dialogue an unprecedented constructive role in a story's composition. The essay concludes by assessing the historical and aesthetic significance of Hemingway's revolution in the writing of dialogue.

MINIMUM SPEECH AND MAXIMUM MEANING: THE FUNCTIONS OF MODERN DIALOGUE

… Hemingway is the one who had the most to do with my craft—not simply for his books, but for his astounding knowledge of the aspect of craftsmanship in the science of writing.

—Gabriel García Márquez (16)

In “Notes on Writing a Novel,” Elizabeth Bowen cut to the crux of exactly why modern dialogue is so difficult to write. She observes that it must imitate certain “realistic qualities”: spontaneity, artlessness, ambiguity, irrelevance, allusiveness, and erraticness. Yet, behind the “mask of these faked realistic qualities,” it must be “pointed, intentional, relevant. It must crystallize situation. It must express character. It must advance plot” (255). It must, in other words, be truly verisimilar—like reality, but not an actual transcription of reality itself. “Speech,” Bowen goes on to say, “is what the characters do to each other”; aside from a few extreme physical acts, it is “the most vigorous and visible inter-action of which characters … are capable” (255). Consequently, speech “crystallizes relationships. It should, ideally, [be so] effective as to make analysis or explanation of the relationships between the characters unnecessary” (255). Although dialogue is not generally effective as a means of exposition, of conveying necessary information (what invariably occurs at the beginning of a play, and that takes all of the considerable artifice of the theater subsequently to overcome), it can express present relationships and, by implication, their past as well. But to do so effectively requires great talent; dialogue must imply subtly, suggestively, and never through direct statement. Usually, the way characters say something is more important than what they say.

Bowen further observes that each sentence spoken by a character must display either “calculation” or “involuntary self-revelation” on their part (256). Most good dialogue, I should hasten to add, displays both of these processes, for in fiction, as in life, it is virtually impossible not to be, to some degree, self-revelatory (no matter whether “self” is conceived of as socially constructed, dialogical, or autonomous and coherent). Generally, she states, characters should “be under rather than over articulate,” and what they “intend to say should be more evident, more striking (because of its greater inner importance to the plot) than what they arrive at saying” (256). Robie Macauley and George Lanning agree, noting that “speech, as a way of characterization, moves forward by means of partial concealment, partial exposure” (78), because what characters say may be the result of inner conflictedness, or they may be saying what they think the other person wishes to hear. In speech, they may become aware of their own confusion, or something the other person does might make them modify their original intention. They may become more confused as they speak, may end up saying the opposite of what they started to say, may even wish that they had not spoken at all. In short, all of the myriad complexities that inhere in real-life dialogue inhere as well in fictional dialogue, the one great difference being that in fiction there is an author who exercises some control over what is being expressed (or incompletely expressed, as the case may be). Dialogue, therefore, demonstrates not only communication but, more importantly, the limits of communication between characters as well.

Hemingway, with his deceptively simple dialogue, managed to capture these dynamics of real-life speech. The earliest example can be found in his first story masterpiece, “Indian Camp,” written between November 1923 and February 1924. In the deleted opening, young Nick Adams, left alone at night by his father and his Uncle George, succumbs to a vague existential dread and fear of death, and summons the men back from their fishing by firing three shots. So far, the story is about as badly written as a story can be. Nick's feelings have been directly and unsuggestively stated by an awkward psychologizing narrator; the narrative's action sequences are confused and confusing; the characters are stilted; and so many flashbacks have occurred in so brief a span of text that the unities of time, place, and action have been badly compromised.

But the fragment closes with three consecutive passages of dialogue that serve an enabling function in leading the story and its fumbling author into the realm of art. The first passage, a conversation between the two men on the lake after they hear the shots, serves no purpose—it is senselessly repetitive and oddly stagnant; it does not advance the plot; it uses unnecessary identification tags; the two characters talk alike; and the dialogue fails to fit their personalities. The next passage, in which the two men return to the camp, is a huge improvement. Each character has one speech; each speech is in character; and each reveals calculation and involuntary self-revelation as the father demonstrates his concern for his son, Nick wishes to persuade the men that there was good reason to call them back but reveals his embarrassment, and Uncle George expresses contempt and reveals his cruelty. The final passage is almost technically flawless:

In the morning his father found two big basswood trees that leaned across each other so that they rubbed together in the wind.

“Do you think that was what it was, Nick?” his father asked.

“Maybe,” Nick said. He didn't want to think about it.

“You don't want to ever be frightened in the woods, Nick. There is nothing that can hurt you.”

“Not even lightning?” Nick asked.

“No, not even lightning. If there is a thunder storm get out into the open. Or get under a beech tree. They're never struck.”

“Never?” Nick asked.

“I never heard of one,” said his father.

“Gee, I'm glad to know that about beech trees,” Nick said.

(“Three Shots” 15)

Hemingway effectively locates the scene with a precise, concrete sentence. The use of the word found (instead of saw) is suggestive. Dr. Adams has deliberately sought out a forest noise for his son, either to console him by offering a palpable reason for Nick's fear on the previous night, or, equally revealing, because he believes in his son despite the evidence and George's opinions. The ensuing dialogue accords with the speakers' personalities. Dr. Adams both respects and wishes to encourage his child's autonomy. He asks Nick if he thinks it was the trees he heard. The child, still ashamed of his earlier behavior, is warily noncommittal. Sensing his son's reticence, Dr. Adams tactfully directs the conversation away from the particular embarrassing incident onto the general topic of how nothing in the woods “can hurt you.” The “you” here refers to Nick, of course, but it also implies the more general sense of “one.” The tactic succeeds. Because his father has addressed Nick's fears indirectly, the boy no longer feels ashamed, and his curiosity causes him to engage in conversation. The subsequent dialogue about lightning and beech trees allows the two characters to settle into the security of a father-as-teacher/son-as-learner pattern of behavior. It produces a sort of catharsis, with Nick's final speech coming across almost as a sigh of relief.

By the end of the deleted opening, then, Hemingway is using dialogue to crystallize relationships, express character, and advance plot. Furthermore, in their speeches the characters display both calculation and involuntary self-revelation. Lastly, the author is turning away from narrative commentary; experimenting with omission (the events of the previous night that underlie the final scene are referred to but never explicitly mentioned); and using highly verisimilar simple discourse. In the story that follows, “Indian Camp” proper, he will employ all of these techniques and extend to dialogue, for the first time in fiction, all the devices of indirection, juxtaposition as a way of having meaning derive from proximity, irony, miscommunication, and compression.²

The final passage of dialogue in “Indian Camp” is particularly illustrative. In the story, Nick and his Uncle George accompany Dr. Adams on an unanticipated visit to an Indian camp where a pregnant Indian woman is suffering from a difficult labor. Nick is forced to assist while his father performs an emergency cesarean operation with fishing equipment and no anesthetic. After the successful operation, Dr. Adams's preoccupation turns to egocentrism, causing George to resent him. Then he discovers that the woman's husband, confined to a bunk over his wife due to a foot injury, has taken his life by cutting his throat during the operation. The discovery deflates the doctor, who suddenly concerns himself with his son's welfare; he takes Nick from the shanty, leaving George behind to await the authorities. The last passage of dialogue reads:

“I'm terribly sorry I brought you along, Nickie,” said his father, all his post-operative exhilaration gone. “It was an awful mess to put you through.”

[Q1] “Do ladies always have such a hard time having babies?” Nick asked.

[A1] “No, that was very, very exceptional.”

[Q2] “Why did he kill himself, Daddy?”

[A2] “I don't know, Nick. He couldn't stand things, I guess.”

[Q3] “Do many men kill themselves, Daddy?”

[A3] “Not very many, Nick.”

[Q4] “Do many women?”

[A4] “Hardly ever.”

[Q5] “Don't they ever?”

[A5] “Oh, yes. They do sometimes.”

[Q6] “Daddy?”

[A6] “Yes.”

[Q7] “Where did Uncle George go?”

[A7] “He'll turn up all right.”

[Q8] “Is dying hard, Daddy?”

[A8] “No, I think it's pretty easy, Nick. It all depends.”

(“Indian Camp” 18-19)

In his first speech, Nick's father admits his mistake. His use of the diminutive “Nickie” suggests that he is now concerned with Nick's anxieties, but in trying to comfort his son he also betrays his own feelings of guilt. His apology is not for Nick's having seen the dead Indian (which could not have been anticipated), nor for his thoughtlessness in having Nick attend the horrifying operation (which was his only irresponsible act). Instead, he apologizes for bringing Nick along in the first place (his least questionable decision), which undercuts the apology by passing over his truly unconscionable act. What he had previously termed a “little affair” (the operation), he now calls an “awful mess”—an understatement that covers all the events Nick has witnessed (including his father's paternal inadequacies) and, by its euphemistic nature, continues to diminish the apology. The last phrase of the statement shows Dr. Adams looking at these events from Nick's perspective (“to put you through”) in order to console him, but his guilty feelings are manifest in his use of an inert construction; a more direct admission of culpability would have been: “I put you through an awful mess.” By his need to assuage his own guilt, then, the doctor's apology is involuntarily self-revelatory: He is still mainly concerned with his own needs, not his child's.

If the doctor's initial 17-word speech is revealing, the ensuing eight questions and answers are a marvel of indirection, miscommunication, suggestiveness, and compression. Nick has conflated all of the events he has witnessed and therefore asks about the operation, although, by the end of the passage, it will become clear that what he really wants to know about is the probability of death (either his father's or his own). His father, however, is obsessed with the suicide and so, for all his newfound sensitivity toward his son and Nick's careful attention to his father, the two characters miscommunicate throughout the conversation.

Nick's first question elicits a somewhat detached response from his father, whose thoughts are elsewhere. Dr. Adams can draw on his medical knowledge to answer the question; the repetition of “very” and the understated “exceptional” give the impression of a considered, dispassionate reply. Nick's second question, however, directly presents the mystery at the heart of the story: “Why did he kill himself, Daddy?” The use of “Daddy,” which Nick previously employed when frightened by the woman's screams, suggests the anxiety beneath his outwardly calm demeanor. But his father does not notice Nick's anxiety, and he cannot, in any case, satisfactorily answer the question. His profession has equipped him to answer medical queries, not psychological ones. He answers honestly—he does not know—but he also senses that his reply is not enough to satisfy Nick, so he follows it with an explanation that is strategically vague: “He couldn't stand things, I guess.”³

From the moment Nick asks about the suicide, the dialogue takes a notable turn. The words “I don't know,” coming from so proud a man who has just performed with such competence under extreme duress, resonate with the doctor's deep sense of confusion, guilt, and deflation. For the rest of the conversation, Nick will focus on death and suicide, and he will ask questions that either cannot be answered or that his father is not in the mood to address. Dr. Adams's inability to answer these questions, and the shock he feels over what has happened, will force him back into the self-absorption he displayed during the operation. And yet, although his answers to his son's questions will be brief and somewhat perfunctory, they will have an oddly calming effect on Nick, relieving his son's anxiety. Even though the characters miscommunicate, the failure to communicate has an ironically successful result.

Nick's questions are relentless. When he catches his father in an inattentive reply, as in the answer to question 4, he immediately issues a follow-up question that reveals his dissatisfaction with his father's response. His father's preoccupation can be glimpsed in the laconic, vague nature of his replies (“Not very many,” “Hardly ever,” sometimes,” “He'll turn up,” “It all depends”). The sixth reply (“Yes.”) is particularly revealing. The absence of a question mark indicates that his father's inflection is declarative rather than interrogative; Nick has gotten his attention, but he remains lost in his own thoughts. When Nick then asks about Uncle George, a subject that his father is especially not interested in, the reply is again unspecific. The addition of the phrase “all right” resonates with the two times the expression was previously used in the story: Nick's response to his father's postoperative inquiry about how Nick liked being an intern (17) and a sarcastic remark George made about Dr. Adams being a “great man, all right” (18). Nick's use of the expression was intended to satisfy his father but revealed his lack of enthusiasm. George's use both intended and revealed his resentment of his brother. Here, Dr. Adams wishes to satisfy his son, but he unintentionally reveals his own lack of enthusiasm for answering any more questions as well as his own resentment toward George, whose earlier sarcasms anticipated the doctor's present feelings of inadequacy.

Nick, of course, cannot comprehend death. He can only feel it as absence. His first glimpse of death was in the context of the Indian father's withdrawal from life. Therefore, the anxieties he expresses in the passage concern absent fathers. The four questions he asks ending with “Daddy?” make manifest the subject of these anxieties (including his sixth question, which is intended to bring his mentally absent father back into the conversation). In his second question he asks why the Indian baby's father killed himself and receives an uncomforting but honest reply. What he really wants to know is whether he is safe from suffering the same fate as the Indian baby boy. So he continues his inquiry by a more circuitous route. His next two questions, about the frequency of male and female suicide, are unconsciously self-referential—he wants to know about his own father and mother—and the responses are comforting. But the real point of the last three questions is revealed only in their juxtaposition. The sixth question is about Nick's sense of his own father's mental absence; the seventh question is about Uncle George's physical absence, here serving as a displacement of Nick's anxiety over his own father's mortality; and the eighth question, read in the above context, might be about the probability of Nick's father's death. Ironically, his father misunderstands the question, in light of his own concerns, to be about whether the act of dying is difficult to face, and his answer is unintentionally chilling: “… it's pretty easy.” Even more ironically, however, the words do not matter, for, as the doctor says, it all “depends.”

Just what it depends on is revealed in the subsequent final two paragraphs, the second of which reiterates the images of the first: “In the early morning on the lake sitting in the stern of the boat with his father rowing, he felt quite sure that he would never die” (19). The first part of the sentence, an objective correlative for Nick's sense of immortality, placed in juxtaposition with Uncle George's absence as a representation of death, triumphs over it. Ambiguities and ironies compound. If the antecedent to the final “he” is Nick's father, a less likely possibility but one that Hemingway purposely leaves open, then all the miscommunication between the two and his father's disquieting responses have inadvertently comforted Nick. On the other hand, if the antecedent is Nick, the much more likely possibility, then another irony is created by the disjunction between Nick's sense of his own immortality and the reader's knowledge that it is otherwise.⁴ Moreover, it is the final turn in the screw of the passage's indirection, for it means that what Nick was really asking about all along concerned his anxieties about his own finitude, not his father's. Which means that what the whole story has been about was not the cesarean operation, the Indian's suicide, or the probability of Dr. Adams's death, but Nick's first encounter with ontological shock, the numbing realization of one's own mortality, which Hemingway omitted from the story when he discarded the opening pages, only to treat it at the end by, characteristically, having Nick deny it. And all of these matters, I might add, are compressed into just a few “simple” sentences of enormously suggestive dialogue in which two characters thoroughly miscommunicate in such subtle ways that readers of the passage, for the past seven decades, have assumed that the two were communicating clearly.

BANALITY INTO ART: THE USES OF REPETITION

As a writer I was astonished by Hemingway's skill. … I have never understood, to this day, how Hemingway achieved his powerful dialogue. … [W]hat Hemingway offered … was not dialogue overheard, but a concentrate of it, often made up of superficially insignificant elements—mere fragments of everyday phrases, which always managed to convey what was most important.

—Ilya Ehrenburg (20)

In the remaining stories he wrote to complete In Our Time, and in The Sun Also Rises, Hemingway continued to employ the innovative dialogue techniques of “Indian Camp.” Dialogue crystallized situation, expressed character, and advanced plot, and speeches were marked by calculation and involuntary self-revelation. Verisimilitude in dialogue was achieved mostly by indirection, banality, simplicity of diction, and pervasive miscommunication, while relevance was produced through irony, juxtaposition, and compression. But although repetition was a significant part of his prose style, he had not yet developed a technique for employing it meaningfully in dialogue.

By early 1926, however, Hemingway was working toward applying the principles of repetition that he had learned from Gertrude Stein and James Joyce. Stein's theory of repetition was designed to provide synchrony in, and remove linearity from, narrative, as the three justifications she advanced for repetition—beginning again, using everything, and the continuous present—make clear (516-22). Her influence on Hemingway's use of repetition can be seen in his passages of description, in his attention to surfaces in passages of free indirect discourse, and, unfortunately, in his somewhat amateurish attempts to imitate his mentor in the repetitions of “Up in Michigan,” “Mr. and Mrs. Elliot,” and even “Cat in the Rain.” For Joyce, repetition was employed chiefly for rhetorical poetic effects, as in the famous final paragraph of “The Dead.” Joyce's influence would be most discernible in Hemingway's pet technique of gathering up selected words and phrases from a paragraph and repeating them in a different order in a summary sentence at the end of that paragraph (e.g., the opening paragraph of “In Another Country”).⁵

In extending repetition to dialogue, however, Hemingway put it to unprecedented use. He managed to capture the repetitive, rambling nature of real-life speech while still exercising the selectivity that is necessary in fiction. In Bowen's terms, he made his dialogue seem irrelevant while remaining perfectly relevant. Most people in real life repeat themselves endlessly when they speak, anxious lest their auditors not catch every last detail and all intended meaning. Speech is also replete with repetitive trivial exchanges. On the other hand, the closer people are, the longer and deeper the history of their relationship, the more they tend to speak to each other in a kind of shorthand that would make their conversation incomprehensible to an outsider (or reader). As Edith Wharton states, “all that is understood between [people] is left out of their talk” in real life (73). Thus, if characters in fiction “have to tell each other many things that each already knows the other knows[,]” then the only way “to avoid the resulting shock of improbability” would be to water down the dialogue with so many irrelevant commonplaces that the reader would grow bored and frustrated (73). Wharton's own solution to the problem was to resort to summary treatment or to interlace her dialogue with narrative, which enabled her to control dialogue through narratorial access to the consciousnesses of her characters. But Hemingway found a way out of the dilemma that enabled him to rely heavily on dialogue, which, given the advantages of scenic treatment in the highly compressed modern short story, contributed to his achievement in the genre. By repeating phrases, words, sounds, and even cadences, he made his dialogue seem repetitive, while the different contexts of these repetitions changed their meanings and kept the dialogue pointed and relevant.

Although repetition can be found in The Sun Also Rises and in such stories as “An Alpine Idyll” and “The Killers,” both completed in the spring of 1926, its use was purely for the sake of verisimilitude, and it did not advance the plot. A typical example from the former occurs when the characters refer to their skiing: “‘You oughtn't to ever do anything too long.’ / ‘No. We were up there too long.’ / ‘Too damn long,’ John said. ‘It's no good doing a thing too long’” (“Alpine” [“An Alpine Idyll”] 111). Here, the repetition seems mimetic, but it must be restricted to a very brief passage or else it would interfere with the movement of the narrative (even though the idea of “doing a thing too long” is directly relevant to the story's theme). Likewise, the dialogue repetitions in “The Killers” are limited mainly to the ways in which the Chicago gunmen mimic their captives, which expresses their truculence and gives the conversations an ominous tone but does not play a part in advancing the plot.

Several months after finishing these stories, however, in August, Hemingway sat down to write “A Canary for One,” a sadly neglected masterpiece in which he introduced his new technique of repetition in dialogue. In the story, the narrator and his wife, both Americans, are returning from the southern coast of France to begin their separation (a fact foreshadowed in the story but withheld from the reader until the final sentence). The action of the narrative takes place in a train compartment, which they share with an annoying, elderly American woman who is bringing a canary home to her daughter. The story, disguised until midpoint as a third-person narrative, focuses on the behavior of the woman and on the passing scenery as the narrator tries not to think about Paris and the impending separation. The American woman, who monopolizes the story's action, is obnoxious, xenophobic, self-centered, domineering, and hard of hearing (in both the literal and figurative senses). She iterates her fear of a possible train wreck and reveals that she has ended her daughter's romance with a man in Vevey, Switzerland, because he was a foreigner, an act that has devastated her daughter. Just before the conclusion of the story, as the train pulls into the Paris station, the American woman and the narrator's wife engage in their second and last fully reported dialogue, which ends with the narrator's second and final speech. In the passage, Hemingway employs every technique in his arsenal, including repetition, to construct what is arguably the finest and most complex piece of dialogue he would ever write:

“Americans make the best husbands,” the American lady said to my wife. I was getting down the bags. “American men are the only men in the world to marry.”

“How long ago did you leave Vevey?” asked my wife.

“Two years ago this fall. It's her, you know, that I'm taking the canary to.”

“Was the man your daughter was in love with a Swiss?”

“Yes,” said the American lady. “He was from a very good family in Vevey. He was going to be an engineer. They met there in Vevey. They used to go on long walks together.”

“I know Vevey,” said my wife. “We were there on our honeymoon.”

“Were you really? That must have been lovely. I had no idea, of course, that she'd fall in love with him.”

“It was a very lovely place,” said my wife.

“Yes,” said the American lady. “Isn't it lovely? Where did you stop there?”

“We stayed at the Trois Couronnes,” said my wife.

“It's such a fine old hotel,” said the American lady.

“Yes,” said my wife. “We had a very fine room and in the fall the country was lovely.”

“Were you there in the fall?”

“Yes,” said my wife.

We were passing three cars that had been in a wreck. They were splintered open and the roofs sagged in.

“Look,” I said. “There's been a wreck.”

The American lady looked and saw the last car.

(“Canary” [“A Canary for One”] 106-07)

On the surface, the conversation seems superficial, repetitious, and awkward—just what one might expect from strangers in such circumstances. It is also a study in miniature of how to use indirection in dialogue. There are two sorts of indirection taking place here. First, the conversation appears to be about Vevey and the American lady's daughter, but it is really about the conflicted emotions experienced by the separating couple. Second, the passage amply demonstrates the wisdom of the Joseph Conrad/Ford Madox Ford “unalterable rule” regarding the rendering of “genuine conversations”: “no speech of one character should ever answer the speech that goes before it” (Ford 200-01). As Ford put it, such “is almost invariably the case in real life where few people listen, because they are always preparing their own next speeches” (201). But beneath the verisimilar surface there is calculation and revelation, character is expressed, and plot is advanced (the passage provides the emotional climax to any rereading of the story). Moreover, even the repetition here is not merely verisimilar (as in “An Alpine Idyll” and “The Killers”); by repeating words in different contexts, Hemingway changes their referents and meanings. These semic qualities accrue to the repeated words and gather force each time the word reappears.

The narrator and his wife are facing forward. He is looking out the window toward Paris and the future. His wife is looking at the American lady with whom she is speaking, and the American lady is looking backward to the rear of the train and the past. Because they are nearing the station and the external scenes he observes remind him of his approaching loss, and because he hears the two women speaking of matters that concern him, the narrator listens carefully to the conversation and, for only the second time in the story, reports it fully.

The American lady makes her comment about the exclusive virtue of American husbands in conjunction with the narrator's apparently quotidian act of getting down the bags, an act that seems to correspond to her views on American husbands but that is given ironic relevance in juxtaposition with what must surely be the narrator's sense of it as a physical step toward the separation of the couple's possessions. When the American lady repeats her observation, the wife asks about Vevey, partly to change the painful course of the conversation. Yet, by choosing to divert it with talk of Vevey, she involuntarily betrays her desire to talk about the once happy past. The American lady predictably takes the question about Vevey as a cue to talk about her daughter, and the wife goes along on that tack. But instead of conversing about the canary that the American lady has just mentioned, the wife is irresistibly drawn to asking about the nature of the broken love affair that stands, for her, as a sign of her own impending separation. When the American lady then tells of the Swiss with whom her daughter fell in love, she twice mentions Vevey, causing the narrator's wife, in a moment of weakness and out of a desire to turn from the sign of her unhappy future to the memory of her happy past, to utter the enormously understated “I know Vevey” and to reveal that it was the site of her honeymoon. From that moment on, the narrator's wife will try to hold onto Vevey and the past. At the same time, the narrator experiences her attempts, by dwelling on Vevey, to ward off the painful emotions caused by their ever-nearing separation. He, in turn, tries unconsciously to hold onto their married status by using, in his wife's remaining five speeches, the identification tags “said my wife,” although these are obviously unnecessary for the purposes of identifying the speaker.

The American lady, who could not care less about the couple's honeymoon, swiftly shifts the conversation back to her daughter's love affair and inadvertently reveals that she feels somewhat defensive, perhaps even guilty, about what she has done. As in a previous dialogue, she uses the phrase “of course” to justify her actions. But the narrator's wife is no longer interested in the unhappy daughter. She drops even the amenity of talking about the daughter and continues her spoken reverie on Vevey. Here an extraordinary event occurs. The American lady, who throughout the story has been totally oblivious to all around her, realizes that the narrator's wife wants to talk about Vevey. For the rest of the conversation she actually focuses on what the narrator's wife wants to talk about, and she responds with questions about the honeymoon and with statements that relate to what the wife says.

Part of the emotional impact of the conversation derives from the fact that if the narrator's wife makes an impression strong enough to pierce the self-absorption and alter the behavior of the American lady, then it must be quite a strong impression indeed. The wife also emerges from her near anonymity to become the center of the scene, a transformation heightened by the drum roll of “said my wife” tags supplied by the narrator. And when the wife's speeches are stitched together, they are emotionally compelling in and of themselves: “I know Vevey. We were there on our honeymoon. It was a very lovely place. We stayed at the Trois Couronnes. Yes. We had a very fine room and in the fall the country was lovely. Yes.”

When the American lady starts replying to what the narrator's wife is saying, she slips subtly into the present tense: “Isn't it lovely?” But then, in asking about the honeymoon, she returns to the past tense: “Where did you stop there?” The wife changes the verb to “stayed” (avoiding the primary meaning of “stopped”) and gives the name of the hotel. “Trois Couronnes” is deliberately chosen by Hemingway for its literary allusiveness; it is the same hotel in which Henry James's “Daisy Miller” takes place. Both stories present three American travelers, and in both stories Americans are robbed of their innocent illusions (although in “Canary” we are presented with the aftermath of the characters' initiation). “Trois Couronnes” means “three crowns,” but both James and Hemingway, with their excellent command of French, may have known that it also means “three fool's caps.” If the narrator, so closely modeled on Hemingway himself, also knows the double meaning of “Trois Couronnes,” then perhaps he is aware of the irony of the name as a reflection on the three inhabitants of the compartment.

As noted, the conversation is manifestly repetitious, which invests it with verisimilitude as the two women repeat, in various contexts, each other's phrases. The American lady, who seems incapable of meaningful conversation, is only able to parrot what the narrator's wife says. And the wife, lost in her memories, latches onto phrases used by the American lady that she herself finds, in the grip of these memories, meaningful. But, as also noted, the repetition serves a dual purpose, without which the entire passage, however mimetic, would be vitiated, as beneath the banal surface the repeated words and phrases expand in meaning because of the changing contexts in which they appear.

For instance, the words fall, Vevey, and lovely are each used four times, and the word fine twice. The American lady tells the narrator's wife that she and her daughter left Vevey two years ago “this fall.” Moments later, responding to the information that the couple had been in Vevey on their honeymoon, she says “[t]hat must have been lovely” but follows by saying that she did not know that her daughter would “fall in love with” the Swiss, changing the original meaning of fall. The wife agrees that “[i]t was a very lovely place,” slightly changing the referent of lovely from honeymooning in Vevey to Vevey itself. A second implied meaning accrues to Vevey here: that it was a place where one could “fall in love.” The American lady then agrees with the wife who has just agreed with her, but she puts her statement in the present tense—”Isn't it lovely?”—changing the referent from Vevey past to Vevey present, and calls the Trois Couronnes “a fine old hotel.” The wife then gathers up the repeated words and phrases and sums up her sense of the conversation: “We had a very fine room and in the fall the country was lovely.” In her sentence, the meaning of fine changes from “prestigious” (revealing the American lady's values) to “nice” or “lovely” (indicating the wife's values); fall once more refers to a season (although it still echoes with the previous sense of “to fall in love”); and lovely describes Vevey in the past tense (conflating lovely, fine, falling in love, the room, the countryside, and Vevey—but locating it all in the past). When the American lady then asks if the couple was “there in the fall” and the wife replies “Yes,” the conversation that began with the wife asking when the American lady left Vevey is brought full circle. Its focus has, by subtle increments, shifted from the American lady and her daughter in the present (“this fall”) to the American couple in Vevey in the past (all emphases mine).

The narrator listens carefully, the bags at his feet, looking out the window. Perhaps he too is being lured back into the past by the circular, mesmerizing conversation. But then he sees the wrecked train. When the American lady asks if they were in Vevey in the fall, his wife says yes, but now they are passing three wrecked cars. The narrator, in his second and final speech, calls their attention to the present—“Look”—and announces, “There's been a wreck.” Just as in an earlier speech, his statement seems commonplace but actually reveals his hostility and resentment: toward the American lady, the dissolution of his marriage, and the painful reliving of the happy past.

When the narrator points out the wreck, in 5 syllables totaling a mere 20 letters, his statement serves 6 functions (a remarkable example of dialogue compression). First, he indicates the literal wreck that has occurred. Second, the wreck is the physical realization of the fears about a train crash that the American lady has expressed throughout the story. Third, the couple's marriage, which the narrator's wife has been reliving, is a wreck. Fourth, the three people, like the three cars they are passing, are also wrecks (the wreck symbolizes the three characters as well as the couple's marriage). Fifth, the narrator, by his statement, wrecks the women's conversation. Lastly, since that conversation has been a reenactment, of sorts, of their previously happily married state, he has perhaps repeated in the present (especially since the story is based on the real-life first marriage that Hemingway wrecked) what he had done in the past. Certainly, his speech seems to “crystallize relationships.” What he has said is pretty much the equivalent of “Shut up!”

DRAMA INTO FICTION: BLURRING THE GENRES

[Hemingway's] is an obscuring and at the same time a revealing way to write dialogue, and only great skill can manage it—and make us aware at the same time that communication of a limited kind is now going on as best it can.

—Eudora Welty (90)

With the publication of “A Canary for One,” in April 1927, Hemingway's technical innovations in writing dialogue were complete. But the question of just how far he could push his new art—to what extent dialogue could carry a whole story—remained. The answer came a month later when he wrote “Hills Like White Elephants,” a story consisting almost entirely of dialogue. A woman and a man sit outside a bar at a railway station in Spain waiting for a train to Madrid—and they talk. Unlike the characters of “A Canary for One,” these characters know each other well and thus speak in the sort of shorthand that Wharton observed would make a conversation unintelligible to an outsider. Therefore, the very premise of the story forced Hemingway to construct relevance for the reader from what should have seemed nonsense. In addition, he made his task even more difficult by omitting the actual subject of their conversation (she is pregnant, and he wants her to have an abortion); as countless critics have noted, abortion, pregnancy, and babies are never once mentioned in the story. Lastly, as if to increase the challenge, he forswore any sort of narrative commentary or access to any character's consciousness; the nondialogue is completely neutral and, with the exception of one key symbol near the end, contributes nothing toward the reader's making sense of the dialogue.

Following a brief description of the physical setting, the characters have their first exchange: “‘What should we drink?’ the girl asked. She had taken off her hat and put it on the table. / ‘It's pretty hot,’ the man said. / ‘Let's drink beer’” (“Hills” [“Hills Like White Elephants”] 39). The woman's first speech will turn out to be revealing of her character; she lacks a sense of autonomy, is possessed of precious little will, and looks to her partner to make the decisions. His reply is even more revelatory. Although she is perfectly willing to have him make the decisions, whether about drinks or abortions, he needs to believe that she is actually taking part in the decision-making process at the same time as he prevents her from doing so. Here he succeeds and is able to get her to request the beers he so obviously wants, not by directly saying that he wants them but by merely making a statement about the weather. Such successful manipulation conveys the dynamics of their relationship in the present and, we may assume, in the past as well. The first exchange, so easy to overlook, tells us virtually all we need to know about these two characters: It crystallizes their relationship and their situation; it expresses their characters; and it encapsulates the ensuing plot. It also amply demonstrates Bowen's other main points about dialogue—that it is what characters do to each other, and that it contains calculation and involuntary self-revelation—all beneath a seemingly banal, spontaneous, and utterly artless surface.

The remainder of the story leading up to the climactic exchange plays out the couple's problems as they discuss, in an extremely veiled and shorthand manner, her pregnancy and the nature of their relationship. Much of the conversation is so obscure that on the literal level it can be comprehended only in light of the entire story. For instance, early in the text when it is clear that they are having a conflict but not what that conflict is about, she says that her Anis del Toro tastes like licorice, and he seems to respond innocuously: “That's the way with everything.” She agrees, but adds: “Everything tastes of licorice. Especially all the things you've waited so long for, like absinthe.” He is caught short by her statement and can only weakly reply, “Oh, cut it out” (40). To anyone but them, the conversation is about alcohol that tastes like licorice, whereas it is really about her desire to have a baby.

Or is it? Later in the story, she seems amenable to having the “simple operation,” as he terms it, if that will make everything all right between them, by which she means if he will respond to her when she makes such statements as the hills “look like white elephants” (41). However, such a response is beyond him—it would entail a capacity to see the world through her eyes and not just his own—and so he tries to distract her by saying that if she has the operation they will be “fine afterward” (41). At the same time, he undercuts his promises even as he protests his love for her, as in the following passage:

“And if I do it you'll be happy and things will be like they were and you'll love me?”

“I love you now. You know I love you.”

“I know. But if I do it, then it will be nice again if I say things are like white elephants, and you'll like it?”

“I'll love it. I love it now but I just can't think about it. You know how I get when I worry.”

(41, emphasis mine)

Although he ostensibly says what she wants to hear, the way he says it reveals more than what he says. Not only does he avoid answering her questions, the juxtaposition of “I love you” and “I love it” speaks volumes about his true feelings.

Throughout the story, he uses language to cloak his desires in a “logic” that can assault her language of metaphorically expressed desire. Since he cannot understand her language, she either resorts to mimicking his own words and phrases back at him: “things will be like they were” (41); or responding with passive aggressive self-abnegation: “Then I'll do it. Because I don't care about me” (41); or merely negating his statements: “We can have everything.” / “No, we can't.” / “We can have the whole world.” / “No, we can't.” / “We can go everywhere.” / “No, we can't. It isn't ours anymore.” / “It's ours.” / “No, it isn't” (42). Finally, she turns to silence to avoid further verbal battering. Only he won't stop talking.

If she wants respect and understanding even more than she wants the baby, it is equally clear (if the dialogue is carefully read) that what he wants is not just for her to have the abortion but also to acknowledge that she wants to have it, that is, to feign volition, thus absolving him from any responsibility for the actions he demands. His motive is manifest in an utterance halfway through the story that he repeats, in various forms, six more times: “Well … if you don't want to you don't have to. I wouldn't have you do it if you didn't want to” (41). Her responses to these attempts at verbal manipulation range from asking him if he wants her to have the operation (making him assume the responsibility), to asking whether he'll love her if she does (forcing a concession for agreeing to the abortion), to saying that she'll do it because she doesn't “care about” herself (a passive aggressive counterattack), to asking him if they can stop talking about it (avoidance). The one and only action she will not perform is to allow him to coerce her into pretending that the abortion is her own decision.

The following passage, which begins with the sixth variation of his trademark utterance (here introduced with particular insistence), takes place after her weary plea that they “maybe stop talking” (42) and is the emotional climax of the story:

[1] “You've got to realize,” he said, “that I don't want you to do it [1] if you don't want to. I'm perfectly willing to go through with it [2] if it [3] means anything to you.”

[2] “Doesn't it [4] mean anything to you? We could get along.”

[3] “Of course it [5] does. But I don't want anybody but you. I don't want any one else. And I know it's [6] perfectly simple.”

[4] “Yes, you know it's [7] perfectly simple.”

[5] “It's [8] all right for you to say that, but I do know it [9].”

[6] “Would you do something for me now?”

[7] “I'd do anything for you.”

[8] “Would you please please please please please please please stop talking?”

[9] He did not say anything but looked at the bags against the wall of the station. There were labels on them from all the hotels where they had spent nights.

[10] “But I don't want you to,” he said, “I don't care anything about it [10].”

[11] “I'll scream,” the girl said.

(43)

The techniques employed in the above passage, so representative of the text as a whole, as well as of Hemingway's art of dialogue in general, should, by now, be manifest. The gender-based miscommunication in which the man's assertive declarative statements are parried by the woman's mimicking of him (paragraphs 2 and 4), by her questions (2, 6, 8), and by her request that he stop talking at her (8) that finally explodes in frustration (11) are typical of their entire conversation. Also typical is the way he makes a general statement, the content of which is intended to pacify her (“I'd do anything for you.”), that is revealed as a lie when she subsequently asks for “something” specific (8) and is refused (10). Moreover, he again reveals his hypocrisy through juxtaposition when he initially employs “perfectly” to modify his willingness to have the baby (1) and in his very next speech uses it to modify the supposed simplicity of having an abortion (3). That juxtaposition does not go unnoticed by her, as her mimicry indicates (4).

But the most remarkable aspect of the passage is Hemingway's full-blown employment of repetition. The repetition of key words like want and perfectly and polysyllabic words that have one syllable in common (anything, anybody, any one, something) creates a powerful verisimilitude, but the contexts in which these words are used keep the dialogue relevant. Even more extraordinary are the various uses of the word it in achieving these dual purposes of dialogue. It is used 10 times, but the antecedent/referent continually changes. The first time, it refers to having the abortion, the second time to having the baby, and the third through fifth times either to having the baby or to the baby itself. The sixth time, it again refers to having the abortion; the seventh time, it refers to having the abortion or, perhaps, to her sense of their entire situation; the eighth time, it is an expletive (there is no antecedent); and the ninth time, it refers to the “knowledge” that having an abortion is simple. These uses of it not only mirror the shorthand manner by which people in real life refer to matters that they both understand (or think that they understand), it also creates the ironic ambiguity that makes for relevance. By subsuming such radically incompatible antecedents within one pronoun, Hemingway demonstrates the process by which, in dialogue, all hope of communication may become impossible.

After the woman's emotional request for the man to stop talking (paragraph 8), Hemingway allows himself two brief sentences of nondialogue that aptly sum up the man's real attitude toward his mate. Looking at the bags with labels that symbolize his desire to make her into a purely sexual object that would leave him unencumbered by the responsibilities of love and family (and mutual respect), the man tries one last time his verbally violent sentence (in a truncated form) in order to coerce her into “choosing” the abortion of her own “free will” (paragraph 10). In the sentence he uses the word care, which he had previously used in insisting that he “cared” about her. Here, he says that he doesn't care about “it,” which in its tenth and final incarnation has come to conflate the abortion, the baby, the entire conversation, and (in juxtaposition with her as “object” in the previous paragraph 9) the unsubjugated, nonsexual part of the woman's self. Her reply, and ours? “I'll scream.” Then, there is one last irony: “‘I'd better take the bags over to the other side of the station,’ the man said.” A good place for that particular symbol! And there is one last moment of possible triumph as she smiles at him in the knowledge that she has not relinquished her last small shred of autonomy. With her final speech—“I feel fine. … There's nothing wrong with me. I feel fine” (44)—Hemingway leaves his talking couple to an ambiguous fate and brings his dialogue experiment to a close.⁶

A NEW DIALOGUE FOR A NEW GENRE: HEMINGWAY'S LEGACY

Hemingway systematized a treatment of dialogue in a manner now scarcely possible to appreciate, so much has the Hemingway usage taken the place of what went before.

—Anthony Powell (110)

What exactly was it that “went before” Hemingway's revolutionary innovations in fictional dialogue? For Henry James, whose theory and practice of dialogue were exemplary in the late nineteenth century, fictional dialogue was purely complementary; its proper and only function was to be “directly illustrative of something given us by another method” of presentation (“London” 1404). The idea that dialogue could crystallize situation or advance plot was, to James, ludicrous, and any attempt to have dialogue undertake such a “constructive office” he termed “suicidal” (“Balzac” 137, “London” 1404). In fact, even the notion of “really constructive dialogue, dialogue organic and dramatic, speaking for itself, representing and embodying substance and form” was an “abhorrent thing,” appropriate to the theater but never to fiction (“Preface” 1127, “Balzac” 137). Moreover, James, for whom mimesis was a paramount goal of fiction, did not believe that direct speech was capable of being mimetically reproduced, and so he called for writers to recognize “the impossibility of making people both talk ‘all the time’ and talk with the needful differences” (“London” 1404).

Hemingway, of course, owed a great deal to James, whose own dialogue served as a powerful model for the young author in its indirection, ambiguity, and portrayal of communication as veiled, partial, and difficult.⁷ But James's theories of fiction—despite his movement toward increased dramatization, foreshortening, and the effacement of the narrator—derived from his work in the nineteenth-century novel of manners. The sine qua non of that genre was the dense depiction of social texture and the representation of lapse of time, and dialogue for purposes other than illustration interfered with these fictional aims (“Howells” 505-06). Hemingway's prosaics, however, derived from an entirely different genre—the emerging modern short story—and they were influenced as well by his experiences in journalism, by the impressionism of Stephen Crane, Stein's theories of repetition, Joyce's complex dialogue patterns, and the imagism of Ezra Pound. The modern short story's demands for radical compression, which led to the need for a high degree of suggestiveness and implication, eliminated the portrayal of social texture and duration that lay at the heart of the novel of manners. These generic demands enabled Hemingway, perhaps even compelled him, to rely on and further compress dialogue, allowing it to assume a greater responsibility in fictional composition than ever before, even to the point of removing almost completely the minimal narrative commentary without which, James had felt, fiction would cease to be fiction and would cross over into drama. James, of course, could not have anticipated the new genre and its heavy reliance on direct speech, and even some of Hemingway's most illustrious contemporaries would find his dialogue-laden stories unseemly. Virginia Woolf, for one, criticized such dialogue in Jamesian terms in her review of the stories in Men Without Women (8). But Dorothy Parker, reviewing the same volume, fully understood that Hemingway's style was “far more effective … in the short story than in the novel” and that the new genre demanded radically different techniques of construction and representation (93).⁸

The Jamesian novel of manners was written for a pre-Freudian audience, and it treated the romantic egoist within a fully developed social world. Consequently, it focused on the consciousnesses of characters capable of rich perception, feeling, and self-awareness. Dialogue was akin to a game of chess, played by sophisticated characters highly skilled and inordinately sensitive to the slightest nuance. Although a great deal of miscommunication occurred, much communication took place as well. The Hemingway short story, on the other hand, portrayed transient modern individuals cut loose from their social moorings. His characters, no less capable of feeling, are much less articulate, sophisticated, skilled in the strategies of speech, and consciously self-aware. But their conversations, as I have tried to show, are every bit as complicated as those of James's characters because, however limited their consciousnesses may be, they possess the same complex unconscious motivations as any human characters.

It should go without saying that one need not be a highly developed, self-aware, social creature to be worthy of consideration, either in literature or in life. Yet, it bears noting, because from the start—with Lee Wilson Dodd's comments on Hemingway's attention to “certain sorts of people” with “oddly limited minds, interests, and patterns of behavior” (323); Wyndham Lewis's famous essay on Hemingway as a “dumb ox” whose characters are dull witted and bovine; and D. S. Savage's nearly obscene assessment of Hemingway's art as representing “the proletarianization of literature; the adaptation of the technical artistic conscience to the subaverage human consciousness” (14-15)—Hemingway's detractors have invariably based their attack on his donnée: his inarticulate characters and their class-bound cultural limitations. Such sentiments are, of course, deplorable, and one would expect today's multicultural and class-sensitive critics to denounce them with alacrity as, at the very least, elitist. But literary criticism has always had an ambivalent attitude toward the uneducated classes, defending them in principle but finding itself viscerally repelled by fiction that features such characters not as a type of “noble savage,” but from the inside in all their awkwardness and crudity. Hemingway—cosmopolitan, multilingual, bookish, and intellectual though he was—did not turn from such characters. In finding ways to allow them to speak, the writer who once jokingly referred to himself as “the Henry James of the People” (“Hadley” 556) fashioned new techniques that brought them fully into the pages of fiction.⁹

Admirers of Henry James are not often found appreciating Ernest Hemingway.¹⁰ The former portrays intellectually rich worlds that critics find delightful; the latter deals with unsophisticated characters who have difficulty expressing their thoughts and emotions. Moreover, in some of today's so-called “cutting edge” critical discourse, these characters are conflated with a caricatured author who seems to represent much of what is most pernicious in the unrevised canon: the physically imposing, bullying, bearded, cigar-smoking, misogynist, racist, heterosexist, white man who hunts, fights, fishes, and fornicates (and, what is worse, writes endlessly about it). Such criticism, of course, is extremely unfair. As Toni Morrison has recently observed, “it would be irresponsible and unjustified to invest Hemingway with the thoughts of his characters” (85).¹¹ Even more to the point, she cautions against judging “the quality of a work based on the attitudes of an author or whatever representations are made of some group” (90). Morrison's advice is especially appropriate for Hemingway scholarship since, the various ideological agendas of critics notwithstanding, Hemingway's fiction continues to influence writers of widely diverse backgrounds and ideologies. For whenever we consider technique—without which, as Arnold Isenberg reminds us, there can be no art¹²—his presence is unavoidable. If nothing else, to go from James's use of dialogue to Eudora Welty's and Raymond Carver's, one must pass through the texts of Ernest Miller Hemingway who, during a period of 3[frac12] years, completely altered the function and technique of fictional dialogue and presented it as one of his many legacies to twentieth-century literature.

Notes

1. One need only mention Hemingway's influence on authors as different as Isaac Babel, Sean O'Faolain, Dorothy Parker, Evelyn Waugh, Graham Greene, William Faulkner, Elio Vittorini, Nathalie Sarraute, Albert Camus, Heinrich Böll, Eudora Welty, Cesare Pavese, Ralph Ellison, J. D. Salinger, Ann Beattie, Raymond Carver, John Munonye, Norman Mailer, Philip Roth, Edna O'Brien, Gabriel García Márquez, and Toni Morrison.

2. Lest I appear to be belaboring the obvious, it is important to repeat that I am aware that critics have thoroughly examined Hemingway's employment of these devices in his narrative prose and that they have also explored the genealogy of these techniques in terms of influences. But narrative prose and direct speech are very different elements of fiction, and no one has ever shown how Hemingway extended these techniques to dialogue. In other words, it is not my intention merely to rehash imagist poetics or Hemingway's “tip of the iceberg” theory but to provide the first systematic examination of his dialogue techniques.

3. Dr. Adams's reply is suggestive. He consciously means to say that the Indian was emotionally unable to go on living, but he employs an idiom with the word stand, reminding the reader of the Indian's injured foot that prevented him from joining the other Indian men outside the shanty and forced him to be a silent witness to his wife's ordeal. The Indian father's injury is a sign of his helplessness, and perhaps functions in Dr. Adams's speech as a subtle indication of the helplessness that the story's other father now feels.

4. Hemingway often employed such referential ambiguity in his dialogues, as I will later show, in order to reproduce the ways in which people miscommunicate (a technique he may have learned from reading Henry James's The Awkward Age). He also, on occasion, used it for an ironic effect. For instance, just before Dr. Adams discovers the suicide, he announces: “Ought to have a look at the proud father. They're usually the worst sufferers in these little affairs” (18). The intended referent of “proud father” and “worst sufferer” is the Indian father, but the statement is unintentionally self-referential, since Dr. Adams is the only father in the story who is proud (albeit in another sense), and he is, at that moment, the most insufferable of the characters. “Little affairs” is meant as an understated reference to the brutal operation, but it also refers to the baby and, on a metatextual level, to the story itself.

5. For a variation on the technique, see the sixth paragraph of “A Canary for One,” in which Hemingway takes four phrases from the second sentence and rewords and repeats them in a different order in the last sentence. Although Stein exerted a powerful influence on Hemingway's prose, repetition as a literary technique was very much “in the air” after the war. Hemingway had been experimenting with it as early as 1920, two years before he met or read Stein, in his journalism for the Toronto Star Weekly. See Reynolds, Young Hemingway, pp. 191, 213-14.

6. Their fate is all the more ambiguous because we do not know if there is any significance to his taking the bags to the other side of the station. For a complementary reading of the dialogue in “Hills” that explores how the two characters' inability to communicate derives from gender-based linguistic, ontological, and epistemological differences, see the excellent article by Smiley. My only point of contention with Professor Smiley is that I see the man's verbal acts as consciously manipulative and sinister, while she views them as a social construction of which he is mainly unaware and therefore somewhat victimized. I also think that the narrative's action hinges on the woman's refusal to feign volition in the abortion decision, in effect defending what little autonomy she has left, rather than on her desire for love and/or the baby. I do, however, fully agree that the rhetoric of both characters is strongly gender based and very representative, and that gender lies at the core of their relationship and its difficulties.

7. It was Hemingway's most influential mentor, Ezra Pound, who compelled the young writer into the works of Henry James, an instruction abetted by the fact that Hemingway's first two wives were James enthusiasts. Although Hemingway alternately ridiculed and praised James and was loath to acknowledge any influence (which was his usual response to any writer who had truly mattered in his development), and although the impact of James on Hemingway would not be fully felt until the late 1940s and 1950s when Hemingway was desperately trying (and failing in his attempts) to write Jamesian novels, nevertheless early in his career James's texts had shown him, as Reynolds has observed, that “the significance of … dialogue appears frequently in the white space between the lines” and that “it is what the characters do not say that is highlighted by their conversation” (Paris Years 30).

8. Regarding the purposes of fictional dialogue, there is a clear distinction between the views of James, Wharton, and Woolf, who were primarily novelists, and of Bowen, Hemingway, and Welty, who were completely at home in the modern short-story genre. The notion that dialogue should be limited to an illustrative function was especially at odds with the new genre. Bowen addressed the critical point of difference:

   Each piece of dialogue must be “something happening.” Dialogue may justify its presence by being “illustrative”—but this secondary use of it must be watched closely, challenged. Illustrativeness can be stretched too far. Like straight description, it then becomes static, a dead weight—halting the movement of the plot.

   (“Notes” 256)

   Bowen, who greatly admired James, felt that James's short stories, for all their virtuosity, represented a “dead end” in the genre's development that could neither be imitated nor advanced upon. See her introduction to The Faber Book of Modern Short Stories (1936), reprinted in Collected Impressions, p. 39. For a discussion of the peculiar demands of the modern short-story form and why James found these and the consequences of compression inhospitable, see my “Observations on Hemingway, Suggestiveness, and the Modern Short Story,” The Midwest Quarterly 37.1 (Fall 1995): 11-26.

9. John Ciardi reminds us: “What counts, as I see it, is the way in which the GIs of World War II lived and died with Hemingway dialogue in their mouths. … Their language was not out of Hemingway but out of themselves. Yet it justified his power as nothing else could” (32).

10. Despite the fact that they were both prominent in the old canon and were often linked thematically, James and Hemingway have always seemed to appeal to different audiences, as the famous exchange of letters in 1955 between Leon Edel and Philip Young, at the time the two leading critical champions of James and Hemingway, respectively, makes clear. For their exchange, see Folio 20 (Spring 1955): 18-22; reprinted in Hemingway: A Collection of Critical Essays. Robert P. Weeks, ed. Englewood Cliffs: Prentice, 1962. 169-74.

11. Although Morrison explores the ways in which the Africanist presence influences the form and content of Hemingway's texts, she is quick to add that “there is no evidence I know of to persuade me that Hemingway shared [the racist views of one of his characters]. In point of fact there is strong evidence to suggest the opposite” (86).

12. Isenberg observes:

    Art is nothing if it is not control. But we control only those of our acts whose outcome we foresee; and we foresee no result unless we have been over the ground before. It is technique, therefore, that gives direction to impulse and marks the difference between art and caprice.

    (58)

I wish to thank several wonderful friends and scholars who have, in different ways, inspired this essay and/or contributed to its final form: Warner Berthoff, Kermit Vanderbilt, Marc Dolan, Debra Jacobs, Wendy Flory, Dick Thompson, and Patsy Yaeger.

Works Cited

Bowen, Elizabeth. “Notes on Writing a Novel” [1945]. In Collected Impressions. New York: Knopf, 1950. 249-63.

Ciardi, John. “The Language of an Age.” Saturday Review 44 (29 July 1961): 32.

Dodd, Lee Wilson. “Simple Annals of the Callous.” Book review of Men Without Women, by Ernest Hemingway. Saturday Review of Literature 4 (19 Nov. 1927): 322-23.

Ehrenburg, Ilya. “The World Weighs a Writer's Influence: USSR.” Saturday Review 44 (29 July 1961): 20.

Ellison, Ralph. “A Rejoinder” [1964]. In “The World and the Jug.” In Shadow and Act. New York: Vintage, 1972. 107-43.

Ford, Ford Madox. Joseph Conrad: A Personal Remembrance. Boston: Little, 1924.

Hemingway, Ernest. “An Alpine Idyll.” In Men Without Women. [1927]. New York: Scribner's, 1970. 109-15.

———. “A Canary for One.” In Men Without Women. 103-108.

———. Letter to Hadley Mowrer. 25 Nov. 1943. In Ernest Hemingway: Selected Letters 1917-1961. Ed. Carlos Baker. New York: Scribner's, 1981. 554-56.

———. “Hills Like White Elephants.” In Men Without Women. 39-44.

———. “Indian Camp.” In In Our Time. [1925]. New York: Scribner's, 1970. 15-19.

———. “Three Shots.” In The Nick Adams Stories. Ed. Philip Young. New York: Scribner's, 1972. 13-15.

Isenberg, Arnold. “The Technical Factor in Art.” Journal of Philosophy 43.1 (1946): 5-19. Rpt. in Aesthetics and the Theory of Criticism: Selected Essays of Arnold Isenberg. Eds. William Callaghan et al. Chicago: U of Chicago P, 1973. 53-69.

James, Henry. “The Lesson of Balzac” [1905]. In Henry James: Literary Criticism: French Writers, Other European Writers, The Prefaces to the New York Edition. Ed. Leon Edel. New York: Library of America, 1984. 115-39.

———. “London Notes” [1897]. In Henry James: Literary Criticism: Essays on Literature, American Writers, English Writers, the Preface to the New York Edition. Ed. Leon Edel. New York: Library of America, 1984. 1387-1413.

———. “Preface to The Awkward Age” [1908]. In Henry James: Literary Criticism: French Writers, Other European Writers, The Prefaces to the New York Edition. Ed. Leon Edel. New York: Library of America, 1984. 1120-37.

———. “William Dean Howells” [1886]. In Henry James: Literary Criticism: Essays on Literature, American Writers, English Writers, The Preface to the New York Edition. Ed. Leon Edel. New York: Library of America, 1984. 497-506.

Levi, Carlo. “The World Weighs a Writer's Influence: Italy.” Saturday Review 44 (29 July 1961): 19.

Lewis, Wyndham. “Ernest Hemingway: The ‘Dumb Ox.’” In Men Without Art. London: Cassell, 1934. 17-40.

Macauley, Robie, and George Lanning. Technique in Fiction. 2nd ed. New York: St. Martin's, 1987.

Madariaga, Salvador de. “The World Weighs a Writer's Influence: Spain.” Saturday Review 44 (29 July 1961): 18.

Márquez, Gabriel García. “Gabriel García Márquez Meets Ernest Hemingway.” New York Times Book Review (26 July 1981): 1, 16, 17.

Morrison, Toni. Playing in the Dark: Whiteness and the Literary Imagination. Cambridge: Harvard UP, 1992.

Parker, Dorothy. “A Book of Great Short Stories.” Book review of Men Without Women, by Ernest Hemingway. The New Yorker 3 (29 Oct. 1927): 92-94.

Powell, Anthony. Messengers of Day. Vol. 2 of To Keep the Ball Rolling: The Memoirs of Anthony Powell. New York: Holt, 1978.

Pryce-Jones, Alan. “The World Weighs a Writer's Influence: England.” Saturday Review 44 (29 July 1961): 21.

Reynolds, Michael. Hemingway: The Paris Years. Oxford: Basil Blackwell, 1989.

———. The Young Hemingway. Oxford: Basil Blackwell, 1986.

Savage, D. S. “Ernest Hemingway.” In Focus Two. Eds. B. Rajan and Andrew Pearse. London: Dobson, 1946. 7-27.

Smiley, Pamela. “Gender-Linked Miscommunication in ‘Hills Like White Elephants.’” The Hemingway Review 8.1 (1988): 2-12. Rpt. in New Critical Approaches to the Short Stories of Ernest Hemingway. Ed. Jackson J. Benson. Durham: Duke UP, 1990. 288-99.

Stein, Gertrude. “Composition as Explanation” [1926]. In Selected Writings of Gertrude Stein. Ed. Carl Van Vechten. New York: Vintage, 1972. 513-23.

Welty, Eudora. “Looking at Short Stories” [1949]. In The Eye of the Story: Selected Essays and Reviews. New York: Vintage, 1979. 85-106.

Wharton, Edith. The Writing of Fiction. 1925. New York: Octagon, 1966.

Woolf, Virginia. “An Essay in Criticism.” Book review of Men Without Women, by Ernest Hemingway. New York Herald Tribune Books 9 (9 Oct. 1927): 1, 8.