Having focused my undergraduate work on sociolinguistics, I was already primed to analyze healthcare language upon beginning medical school. I had spent years studying the effects of race, education, socioeconomic status, and a host of other social factors on language use and utility across various social settings, concentrating on multilingualism and dialects of American English. I knew that I wanted to use this training to delve into the art of patient communication as I transitioned into the medical field, but I wasn’t yet sure just how that aspiration might be realized. It was only a month into my undergraduate medical education, as I practiced my history-taking skills at the local free clinic, when my previous work’s applicability first became clear.

After interviewing my patient, a middle-aged Black woman, I discussed her case with the attending physician, also a Black woman. The physician spoke to me in Standard American English (SAE), the dialect of most college-educated Americans.1 (When we think of SAE, we think most basically of a speaker who the general public would claim does not have an “accent.”) When the physician and I re-entered the patient’s room together, however, she began speaking to the patient in African American English (AAE), more colloquially known as Ebonics. This was not particularly surprising, as educated Black people are often bi-dialectal, meaning they can speak both fluent SAE and AAE.2 In addition, this act of alternating between the two dialects—called “code-switching”—is highly prevalent among bi-dialectal speakers.3

As I observed the patient encounter, I noted numerous instances of the physician’s code-switching. When the physician vented her disdain for standing in long lines at restaurants, she spoke in AAE. When asking questions about the history of present illness—SAE. When inquiring about family—AAE. When giving treatment options—SAE. A pattern began revealing itself as I continued to analyze the contexts in which each dialect was utilized: AAE to connect with the patient when discussing personal matters and SAE to professionally communicate and elicit information regarding medical issues. By addressing distinct components of the patient encounter with distinct dialects, the doctor employed a strategic and meaningful use of code-switching.

I tried to discern if the physician’s code-switching had been conscious or unconscious. How did she know that her speaking in AAE would be welcomed? Had there been an unspoken language contract being negotiated under the surface to which I, as a young White male, was not attuned? How had the participants’ race, gender, and age pressured the social decision to code-switch? And finally, how could their rapport have been affected—either positively or negatively—had the physician instead spoken in a single dialect?

In free clinics during the remaining pre-clinical years of medical school and in the hospital thereafter, I continued to observe this same pattern of code-switching by numerous other physicians in a variety of care settings. And as I quickly came to understand, this strategic code-switching by doctors was not limited only to dialects, but extended also to language, tone, and lexicon.

In one instance, I observed my primary care preceptor—a White male—talking with his bilingual Spanish- and English-speaking patient. She had recently been diagnosed with type 2 diabetes, and lifestyle modification was being advised. However, the patient was hesitant to reform her habits.

As his motivational interviewing drew to a close, my preceptor asked the patient one final time if she was truly ready to change.

He reminded her, “Su es la jefe” … “You’re the boss.”

She laughed as she corrected him: “Jefa! I’m a girl!”

“Oh, right — jefa!” he echoed. “I’m so sorry!”

After a moment of adjourning small talk, my preceptor joined the patient in a hug as she affirmed her dedication to changing her lifestyle.

While doctors have been demonstrated to code-switch between languages to build rapport with their patients,4 namely by ushering into the conversation an acknowledgment of some shared race, language, or culture, I had not previously thought about how code-switching could be used to influence a patient’s adherence to a treatment plan. By switching to his unperfected second language in an act of accommodation to his patient, my preceptor humbled himself. If he had not code-switched to Spanish, would the patient have been as likely to attempt lifestyle modification? Could his act of code-switching, if implemented for the distinct purpose of charming the patient into changing her lifestyle, be considered “manipulation?” And if so, had an ethical line been crossed?

These experiences in the hospital and clinic are a good jumping-off point for a more robust contemplation of the phenomenon. For instance—beyond language and dialect, we can understand how tone is also an important consideration with regard to code-switching. While our default professional tone is largely attuned to a mentally competent adult, we still retain the ability to modify it, especially with persons on the extreme ends of the age spectrum.

Imagine an elderly woman with dementia presenting to the clinic with her daughter, her legal guardian. Many of us—with all our faults and the best of intentions—may find ourselves speaking loudly and slowly in a tone that is more patient yet paternalistic with the elderly woman. When speaking with her daughter, however, we will almost surely code-switch back to our faster paced, egalitarian, and conversationalist default tone. While pursuing optimal patient comprehension through this type of code-switching, we must be careful not to minimize the patient’s capacity; this can be insulting to the patient and hurtful for their family and fellow caregivers. Achieving respect and mutual understanding among all stakeholders in the visit must be paramount.

In another exemplification of tonal code-switching, we can envision a patient interview with a young woman who has brought her school-aged son to the clinic with her. Although the physician will largely converse in her default professional tone, it is easy to imagine her—when inquiring about the patient’s lifestyle—turning to the child, stooping down low, raising the pitch of her voice, and asking, “Has mommy been eating her broccoli at home?” This is a complex linguistic interaction for a number of reasons. It engages the child in the adult conversation taking place, solicits information about the patient from another source, involves linguistic accommodation in the form of tonal code-switching, and may even cheekily imply that the physician trusts the child’s response more than his mother’s. Depending on the patient’s interpretation of the interaction, this could be perceived as either a lighthearted tactic to liven up the interview or an unwanted diversion of attention away from the patient and a marker of lack of trust.

A final type of code-switching that is routinely practiced in the medical setting is a lexical one—word choice. In fact, this is a lauded skill that medical students are taught and physicians work to perfect throughout their careers. As medical professionals, we interact with patients spanning the spectrum of educational achievement. With oftentimes very little prior social information about a patient, we are charged with tailoring our explanations of complex medical states, pathophysiologic mechanisms, and pharmacologic treatment modalities to their individual level of understanding. We can imagine how to appropriately code-switch to a different vocabulary to accommodate these different patients.

For instance, we might simplify our explanations more dramatically for a patient who was unable to complete high school than for a patient who graduated college. We would likely choose a more sophisticated lexicon to use with a patient who works in the medical profession. And at some intermediate level, we may only temper the words we choose to use with a patient who, for example, holds a Ph.D. in a social science; that is, we may speak at a high level with regard to vocabulary and sentence construction while still limiting our use of more esoteric medical jargon. As in the previous example of the patient with dementia, we must toe the line between ensuring that the patient achieves complete understanding and preventing them from becoming offended by our attempts to adapt to their level of language comprehension through our word choices.

Discussions regarding how doctors both currently utilize code-switching and, perhaps, should utilize code-switching are far from being fully fleshed out, as the application of linguistic inquiry into code-switching in medicine is still in its infancy. Considering the limited literature and the case studies and thought experiments we have here considered, however, I believe there are several real benefits that can come from a physician code-switching in the medical interview. Most readily apparent is its ability to help establish or build rapport. Folks appreciate feeling a sense of kinship with their primary care providers, and code-switching to highlight a common bond in the patient-provider relationship is one means of achieving this. Similarly, it can engender trust in the partnership. If the patient sees a willingness of the physician to accommodate to their manner of speaking, they may come to view the doctor as more genuine and have faith in the diagnoses and plans they formulate. Code-switching can also enhance a patient’s understanding. If the physician speaks in their patient’s dialect, language, tone, or vocabulary, it is easy to imagine how that patient’s comprehension of their illness or condition might increase. This, combined with the aforementioned strengthened rapport and engendered trust, leads to a potential increase in patient adherence—a central tenet of quality patient care.

However, there is also one very real risk inherent in the decision to code-switch: offending our patients. If this happens, the best-case scenario is that the rest of the visit becomes tense and awkward. The worst-case scenario involves a complete breakdown in communication and possible termination of the patient-physician relationship. To avoid these outcomes, I believe code-switching is most safely and effectively performed within three parameters.

Firstly, if we are employing another language, dialect, or cultural norm in the practice of code-switching, we should have a legitimate claim to or affiliation with that language, dialect, or cultural norm. Consider when and how a privileged White male, for instance, might code-switch with his patients in contrast to a Black female. We must take stock of our personal identities and understand that there’s a very fine line between engaging with a culture and co-opting or appropriating it.

Secondly, our word choice, tone, and body language should all convey clear and honest intentions with our patients: to improve their care experience by practicing humility and compassion, accommodating their preferences, and building a genuine bond with them. Code-switching should not be done with the intention of getting our patients to think we are “cool” or manipulating them into following our treatment plans. Likewise, we should not code-switch only to showcase our foreign language skills or proficiency in impressions.

Lastly, code-switching must be done only within the strict realm of professionalism. When assuming a manner of speaking disparate from our default professional one, it can be a slippery slope wherein we find ourselves becoming too casual. For instance, a physician code-switching his lexicon to use vulgarity, no matter how accepting of this practice the patient might seem, can never be acceptable. Patients have very clear expectations for how their healthcare providers should speak to them, and code-switching should not be used as an opportunity or excuse to practice outside of these professional limits.

With these considerations and an increased knowledge of the prevalence, risks, and potential benefits of code-switching, I encourage us as medical professionals to pay closer attention to how we and our colleagues utilize this linguistic tool in the healthcare arena. Assessing its frequency and purposes of use in our practices lays a groundwork from which necessity to change can be evaluated. If we decide to begin or continue to code-switch, we must do so with care. And finally, we should consider foremost the individuality of our patients and their perceived comfort with the practice of code-switching. Given the history and currently tensing state of race relations in our nation and the importance of cultural competency in primary care, this linguistic self-evaluation is not only relevant, but wholly exigent.