Clayton J. Baker, MD, CM
University of Rochester School of Medicine and Dentistry, Rochester, New York, United States


Hills like white elephants
 Photography by Vanessa p.

There is a particular type of dysfunctional communication that can occur between doctor and patient, a sort of a temporary folie-a-deux. This “collusion of non-communication” happens when a doctor-patient interview reaches a topic that one or both parties find particularly distasteful, frightening, or shameful. Seeking to avoid, or spare the other, the discomfort that might occur if the subject is openly discussed, patient and doctor convey to each other, through indirect means, their desire to avoid the sensitive subject. By both verbal and nonverbal signals, an implicit understanding is reached that the dreaded topic (which ironically is often of central importance) is off limits. Any subsequent attempt to fully address it, at least during that interaction, is aborted.

The collusion of non-communication is not limited to doctors and patients. Wives and husbands, parents and children, teachers and students, and other types of relationships can fall into this tacitly agreed-upon non-engagement. In Ernest Hemingway’s 1927 short story, “Hills Like White Elephants,” this phenomenon is subtly but very effectively illustrated between two lovers. Regardless of the relationship, the more taboo the topic, the more likely this collusion is to occur. Few topics are more problematic to discuss than abortion, which is what Hemingway’s two lovers appear to be discussing in “Hills.”

The qualifier “appear to” must be used here because, with characteristic understatement, Hemingway never allows his characters to explicitly identify the problem they are purportedly trying to resolve. The male character (identified only as “the American” or “the man”) repeatedly refers to some procedure he wants the female character (called “Jig” or “the girl”) to undergo. However, the man never names the procedure outright. Instead, he obliquely refers to it with euphemisms like “It’s just to let the air in,” and “I know it’s perfectly simple.” For her part, Jig also avoids naming it, choosing instead to derisively parrot his anodynes back at him: “Yes, you know it’s perfectly simple.”

“Hills Like White Elephants” is a unique short story that remains vivid and compelling reading today. In medical education, it can also serve as an excellent case study of the collusion of non-communication. Both by what is omitted and by what is included, the story illuminates the problems this maladaptive phenomenon creates in interpersonal relationships: how it undermines trust, engenders resentment, and thwarts effective decision-making. As such, “Hills” is a valuable text illustrating communication skills, especially surrounding difficult topics.

The narrative opens far off on the horizon, somewhere in Spain: “The hills across the valley of the Ebro were long and white.” It quickly telescopes to a little outdoor bar at a railway station, where Jig and the American sit, waiting for the train. They order drinks, bickering a bit about nothing in particular as they do so. After an implied period of silence suggested by the ordering of another round, the American tries to broach the subject on both their minds.

‘It’s really an awfully simple operation, Jig,’ the man said. ‘It’s not really an operation at all.’
The girl looked at the ground the table legs rested on.
‘I know you wouldn’t mind it, Jig. It’s really not anything. It’s just to let the air in.’
The girl did not say anything.
‘I’ll go with you and I’ll stay with you all the time. They just let the air in and then it’s all perfectly natural.’
‘Then what will we do afterwards?’
‘We’ll be fine afterwards. Just like we were before.’
‘What makes you think so?’

Back and forth they go from there. The American repeatedly downplays the seriousness of the procedure while simultaneously insisting on its importance to their happiness. Jig, weary of his cajoling, wavers, then finally agrees to go through with it because, as she says, “I don’t care about me.” He takes exception to this, claiming “I don’t want you to do it if you feel that way.” But, wittingly or not, Jig seems to have called his bluff. He professes his concern for her. They drift off topic and argue about nonspecific things. She gets up and leaves the table. When she returns, he resumes his circuitous advocacy for the (presumed) abortion, ceasing only when she threatens to make a scene. Now it is his turn to leave the table. The story ends with the train about to arrive, the American trying to comfort her, and Jig insisting enigmatically, “There’s nothing wrong with me. I feel fine.”

Especially upon first reading, the characters’ evasiveness and prevarications can frustrate the reader. Medical students assigned to read “Hills” have complained, “Why are they beating around the bush like that?” or “Why don’t they just come out and say it?” These are excellent questions, central to the story. They are also excellent points of departure for a discussion about doctor-patient communication, and potential pitfalls such as the collusion of non-communication.

Of course, there are important differences between the circumstances in “Hills” and those in a doctor-patient encounter. The nature of the relationship, the degree of intimacy between the participants, the setting – all these are different. Nevertheless, important similarities exist. There appears to be a significant knowledge gap and power differential between the American and Jig that while different in nature from a doctor-patient relationship still influences the discussion, and is subject to misuse. Most importantly, a difficult, unpleasant medical decision is in question, one requiring openness and honesty to be optimally resolved. Unfortunately, that is not how the story goes. Indeed, “Hills” serves primarily as a negative example, a cautionary tale about dysfunctional interpersonal communication under stress. However, one can learn from the wise and from the unwise, and the latter are often more memorable teachers, if we will just accept their tutelage.

How do two people, supposedly having a serious discussion about an important medical issue, tacitly agree to sidestep doing exactly that? In “Hills,” Jig and the American demonstrate some of the ways, both in things they do and things they fail to do. They never explicitly name either her condition (pregnancy) or his proposed solution (abortion), instead resorting to facile euphemisms (“They just let the air in”). Jig conveys her fear not by admitting to it, but by changing the subject repeatedly, redirecting the conversation back to irrelevancies (white elephants and beer). They both give nonverbal cues of resistance to meaningful engagement (averting their gazes, leaving the table). Their dysfunctional conversation finally comes to an oddly fitting end after a rather juvenile threat (“I’ll scream”).

What lessons might a clinician take away from this story to improve communication with patients, particularly when discussing difficult topics?  First, unlike Jig and the American, doctors must clearly define the problem in a manner that everyone understands and acknowledges. Hemingway’s characters cannot bear to say the word ‘abortion.’ How can they reasonably hope to best decide whether or not to undergo one? Analogously, a physician who must inform a patient that they have cancer should use the word ‘cancer’ – several times – to ensure that the patient leaves knowing their diagnosis. Words such as ‘tumor’, ‘growth’, or even ‘malignancy’ do not suffice for frightened patients who will grasp at any excuse to deny a dreaded diagnosis. Additionally, the appropriate, forthright use of disagreeable or loaded words often serves to normalize them, to diminish their taboo.

When Jig or the American sense resistance or resentment in the other, they respond defensively or evasively. In clinical encounters, evidence of opposition from patients signals an urgent need to explicitly discuss the patient’s attitude toward their current situation and their goals of care. No treatment is effective if the patient won’t comply with it. Such discussions, while sometimes difficult to initiate, avoid misunderstanding, and often produce subsequent dividends of openness and trust.

Unlike the characters in “Hills,” the skilled physician enumerates all possible courses of action, and systematically assesses the pros and cons of each. In modern medicine, it is rare for there to be no treatment options. However, sick, frightened, and underinformed patients often fear or assume precisely that.

The dialogue between Jig and the American ends in a standstill. Because of their mutual unwillingness to discuss the topic openly and constructively, their problem is no closer to resolution than before, and mutual ill will seems to have grown. Their collusion of non-communication renders their conversation a failure. In a doctor-patient encounter, a conclusion cannot always be reached either. However, in contrast to the example given in “Hills,” physicians can take steps to ensure that even an inconclusive discussion ends constructively. Certain simple, straightforward techniques can greatly promote mutual understanding. For example, the doctor can clearly restate the patient’s current position, then ask for confirmation that the stated impression is accurate. While this may initially make the patient uncomfortable, it signals to them that they have been heard, and that their point of view is valued. Closure can be achieved by summarizing the current situation, however inconclusive: “It looks like we won’t reach a decision today on when you’ll try to quit smoking. Think about it. We’ll discuss it again next time.”

“Hills Like White Elephants” powerfully illustrates the costs of surrendering to the collusion of non-communication. Physicians must continually be on guard against this and similar pitfalls that can render doctor-patient interactions ineffective or even counterproductive. The doctor must always be willing to name the elephant in the room.


Clayton J. Baker, MD, CM, is a practicing physician, medical teacher, and writer. Born in Iowa City in 1966, he moved to upstate New York in 1967 with his parents’ assistance. After graduating from Harvard College and McGill University and completing residency at Beth Israel Deaconess Medical Center, he practiced for four years with the Indian Health Service in Tuba City, Arizona. His writing has been published in various journals including JAMA, Annals of Internal Medicine, Academic Medicine, and The New England Journal of Medicine. He works in Rochester, New York, where he lives with his wife Jill and their children Samuel, Henry, and Audrey.