This article has its origin in a third-year psychiatry residents' general clinic, although it could have been in a medical education seminar. The problem is the same: how do you reach someone with what you have to say? (1–3) As I watched residents try to reach their patients, I started to wonder about the importance of technique. The residents were clearly "winging it." They didn't have words to describe the range of techniques available to them. In this commentary, I present a model of persuasive speech that dates back to Aristotle; I conclude by arguing that persuasive speech is intrinsic, not inimical, to the practice of ethical medicine.
Rhetoric is the art of persuasion, or, more carefully put, it is "the art or the discipline that deals with the use of discourse, either spoken or written, to inform, or persuade, or motivate an audience…" (4). For many, it is a dubious practice, the domain of the charlatan and the mountebank. Plato had Gorgias defend it ("if you have the power of uttering this word, you will have the physician your slave…") (5) in order to have Socrates criticize it as a practice without subject, but mere technique. Others, including Aristotle, Plato's student, considered argumentation necessary and universal: "For all men attempt in some measure both to conduct investigations and to furnish explanations, both to defend and to prosecute" (6, p 66). Some do it "haphazardly, others by custom and out of habit." Aristotle proposed to study it methodically; he defined rhetoric as the study of whatever is persuasive in an argument. Aristotle allowed that a speaker may be persuasive because he finds an argument by referring to events not of his making (e.g., oracles, proverbs), or because he makes an argument. Aristotle emphasized the latter and distinguished three sources of "artificial" persuasion: "Some reside in the character of the speaker, some in a certain disposition of the audience, and some in the speech itself" (6, p 74). For millennia, these modes have been referred to as ethos, pathos, and logos. Audiences will variously find one or more mode persuasive. Each audience is different, and Aristotle reminds us "persuasiveness is persuasiveness for an individual…" (6, p 76).
Many physicians feel that persuasion is too strong a description for what they find themselves doing with their patients most of the time, but only the most querulous would deny that they aim to be agents of change, even if this is limited to improving insight (and it usually is not). Freud eschewed suggestion as a treatment technique, not because it was paternalistic or presumptuous, but because he didn't think it worked. He did, however, recognize that the analyst could, and should, influence the patient in his or her self-emancipation. The psychotherapist Rollo May cut through much cant when he wrote: "[We] shall be influencing others whether they or [we] wish it or not, and [we] had best frankly recognize this (7, p 79)." Beyond recognizing that we influence others, we can benefit from knowing how we influence others, and to what ends. In simplest terms, Aristotle defined the objective of rhetoric as judgment (6, p 140). For the modern physician, concerned with honoring the patient's autonomy, this properly refers to the patient's judgment as to the best thing to do, or the best way of considering things—for the patient. It is good to remember that we can do no more than influence a patient in favor of a particular judgment, but that this is an ethical thing to do.
How we do this is the subject of the remainder of the article. I will use Aristotle's heuristic of logos, ethos, and pathos to examine how we work with patients in the clinic. Aristotle distinguished between three types of proof, which he referred to as logos, pathos, and ethos. Later rhetoricians considered the distinctions somewhat differently. Cicero, for example, distinguished between three duties of oratory: logos, to teach; ethos, to charm; and pathos, to move. Quintilian (9) merged ethos and pathos, both in contrast to logos (8).
Logos refers to an appeal for agreement based on reason, arrived at by argument and not by coercion. Logos emphasizes the propositional content rather than the performative effect of statements, as ethos and pathos do. Argumentation involves "the discursive techniques allowing us to induce or to increase the mind's adherence to the theses presented for its assent (9, p 4)." There are a number of necessary conditions for an argument to result in assent: the participants must share a common language; there must be a real intellectual contact and a willingness to engage with each other and to try to gain the adherence of the other. Agreement is more easily reached when there are shared attitudes as to what is real and what is preferable. These conditions are often not met in the clinic: there is often not a shared common language; and clinicians may not care to "find" and fully engage with the patient in order to gain assent to his or her proposals; there is often neither the modesty nor the respect for the patient required to do this. In short, we often do not argue with our patients in the sense of respectfully putting our case forward for their consideration; instead, we recommend and warn.
This raises the question of why patients do what we advise (when they do) if we have not gained their commitment through reason. As we will discuss with regard to ethos and pathos, people believe and do things for nonrational reasons. One reason is offered by the psychotherapist Rollo May, when he suggests that "…individuals allow themselves to be persuaded of an obvious untruth because it raises their prestige to believe it" (7, p 78). And, people tend to absorb attitudes and beliefs from the groups they are members of, and may feel no need to or are dissuaded from examining the rationality of their beliefs. However, if the clinician is inclined to try to reach her audience (the patient) in order to make an argument about a treatment plan, she will have to try to use a language accessible to the patient. I have listened to many residents respond to a question by a patient about their proposed treatment with explanations involving "dopamine receptors," "central alpha-2 adrenergic agonists," and the like. When studies show that few patients can even locate organs on a drawing of the body (10, 11), it is unimaginable that most patients will have any clue what receptors, second-messengers, and GABA are. Even if the clinician does try to make a respectful argument, for many patients, it will not necessarily be the argument that convinces the patient to assent to the proposal. For some, the fact that one offers an argument may be more important than the argument itself. Two modern rhetoricians noted, "To engage in argument, a person must attach some importance to gaining the adherence of his interlocutor, to securing his assent, his mental cooperation. It is, accordingly, sometimes a valued honor to be a person with whom another will enter into discussion (9)." So even if one makes the effort to offer an argument about an interpretation or a course of treatment, for some patients, it will be their emotional responses to the process rather than to the content of the argument that will determine their response.
Aristotle's second source of artful persuasion is the sense the audience develops of the speaker's character, which he refers to as "ethos." One rhetorician has described this ethical appeal of the speaker as "the persuasion exerted upon the minds and hearts of the audience by the personal character of the speaker, causing them to believe in his sincerity, his truth, his ability, his good will toward them (12, p 272)." Echoing Aristotle's suggestion that ethos is the most powerful source of belief in the speaker, one psychotherapist wrote "the eternal truism that it is what the counselor really is which exerts the influence, not the relatively superficial matter of the words being uttered. ‘What you are speaks so loudly that I cannot hear what you say’ (7)." Some aspects of one's ethical effect are not modifiable, for example, one's age (whether one is in one's youth, one's prime, or one's old age) or gender. Other aspects can be asserted through speech, manner, and "stage management," with the usual goal of establishing oneself in the minds of the audience as "possessing genuine wisdom and excellence of character (13, p 13)." Healers must satisfy specific claims in order to be considered trustworthy: 1) I am a healer; 2) I am sincere; 3) I practice a form of healing that derives its power from truth; and 4) I am making changes that will be realized in an improvement in your illness (15). The latter claim is properly validated through demonstration (i.e., the patient's condition improves); the others can only be validated indirectly through speech and behavior. The sociologist Erving Goffman has written about how the performer must take care to control the impressions that others form of him. Among the dimensions of what he calls the performer's "personal front," he distinguishes between appearance (i.e., "those stimuli that inform one of the performer's social status, whether one is engaged in ritual matters, work, informal recreation, etc.") and "manner" (i.e., "those stimuli that warn us of the interaction role the performer will expect to play in the ongoing situation (15))." Typically, the audience expects a consonance between appearance and manner. Some patients will likely experience a tension between the appearance of some residents, in terms of their clothing and grooming, and their manner as physicians (16). I have noticed that in each residency class there are one or two residents whose character distinguishes them. They appear to be recognized by their peers as individuals possessing common sense, virtue, and goodwill (6, p 141). They are accorded more latitude in their behavior and more "understanding" when they lapse, by patients and teachers alike.
Pathos, Aristotle's third source of persuasion, refers to the emotional state a speaker can put the audience in that becomes conducive to his or her purposes. He understood that one's emotional state influences one's judgments about events and situations. What one feels affects what one believes. Rollo May put it: "our question is not why the person had the power to influence the other, but rather what tendencies were there in the mind of the other, probably unconscious, which made the person so ready to be influenced? There must exist some unconscious readiness to believe, some predisposition toward the influence (7, p 78)." Pascal echoed this when he wrote, "All men whatsoever are almost always led into belief not because a thing is proved but because it is pleasing (17)." Aristotle considered some emotions to be salutary (i.e., calmness, friendship, favor, pity); to this might be added joy and hope. Others are unhealthy, including Aristotle's six (i.e., anger, fear, shame, indignation, envy, jealousy), as well as demoralization and hopelessness. Positive emotions lead to hopefulness and expansion; negative emotions to despair and retreat.
A speaker will try to elicit the emotion most appropriate to his or her task. For example, a "war president" will try to move his audience toward anger, fear, outrage. On the other hand, a psychiatrist will usually try to move the patient from a negative to a positive emotional state, away from anger, fear, shame, hopelessness, to courage, calm, hopefulness (18). In the resident clinic, I have seen patients' emotional states moved in countertherapeutic ways. One resident that I observed was so intent on "laying down the law" on what he would and would not prescribe that he did not notice how frightened he made his already-anxious patient. Another was so relentless in her questioning that the patient finally exploded, saying he felt interrogated and wouldn't answer another question.
This article has suggested that Aristotle's model of rhetoric has relevance to how the residents interact with their patients. Throughout their training, they have been taught to emphasize argument in their interactions, but the sense of argument is commonly limited to explaining why they are recommending a course of action, the risks (usually limited to side effects), and why the patient should cooperate, rather than in the sense of presenting an argument for the patient to consider and responding to the patient's counterargument with a process akin to negotiation (19).
There are sound reasons why logos should remain the privileged method of contributing to patient's judgments, but it is important to recognize restrictions to the approach and to appreciate that non–reason-based forms of influence are profoundly important and that they can (and should) be consonant with a reason-based argument.
First, the restrictions to a rational appeal are not always apparent to either party, but they can include practical restrictions of time, cultural distance, and lack of a shared language (i.e., as members of different speech communities with different ways of communicating (20), which combine to subvert a shared understanding and undermine an assent based on reasoned argument).
Second, the impression that the patient develops of the physician's character is crucial. Plenty of people take their medicines not because they understand what the medicine does but because they trust the doctor who is prescribing them (20). Or, alternately, they may refuse because they don't. With this in mind, it is good to recall what Harry Stack Sullivan observed:
As regards pathos, or the emotional climate that we try to establish, we should strive to create an environment that is hopeful, remoralizing, unshaming, and calm. This means that we must balance our focus on pathology with an appreciation of the patient's efforts and strengths, and spend time—in awe—talking to those aspects.
No doubt, there will be some who reject any suggestion of the importance of persuasion in the doctor–patient relationship. This might be because of a naive belief that truth sells itself, or it may be a reflex reaction to the relentless and inescapable persuasion in the form of marketing that we are exposed to daily—or it might derive from a failure to see the difference between a physician's power over a patient and the physician's power to help a patient. The former involves domination and paternalism, but the latter consists of facilitation and assistance in the recognition of the patient's autonomy as a decision-making person. In his important reexamination of the recent priority of principles and autonomy in medical ethics, Howard Brody makes a similar distinction between "side-constraint" and "end-constraint" conceptions of (patient) autonomy (23). The former is a negative view of autonomy; it emphasizes what others should not do, and "it is silent on how they might best assist us in pursuing our interests and goals." The latter is a positive view; it describes autonomy as a morally desirable state and asserts that others may help "by enhancing [the patient's] capacity for autonomous choice and behavior and removing reversible barriers that stand in the way of her autonomy." To the degree that lack of knowledge or misunderstandings interfere with autonomous choice and recovery, an ethical physician should try to influence the person. As Brody argues, "if autonomy is to be respected, physicians must be permitted to use reasonable persuasion, just as people are permitted to do in other aspects of life where autonomy is assumed."