The Rhetoric of Rationalization

Medical educators are facing a critical problem: How do we accurately assess professionalism? Currently, we are beginning to understand professionalism; however, we have been relatively unsuccessful in attempts to evaluate it, and we have been unable to predict professional behavior in our students.1,2 Professionalism has been described not only as the ability to uphold the principles and values of the profession, but also as the ability to negotiate between competing values in a specific context.3 Understanding the nature of such negotiations requires asking more than “What did the student do?” or “What choice did the student make?” We must also ask, “Why did the student choose that action?” or “How did the student justify the action?”4,5 The science of rhetoric provides formal analytic tools for excavating and analyzing students' reasoning strategies.6 The particular rhetorical approach used in this study is “social constructivism,” which emphasizes how language embodies strategies for action. Social constructivist rhetoric suggests that language both reflects and shapes our reality.7 Students' stories of professional dilemmas, therefore, provide a window onto their perceptions—and constructions—of their roles in the medical world. The structure of the language used in these stories reflects the structure of students' applied logic, and therefore reflects their growing phronesis—practical wisdom and the ability to make decisions in complex situations—as they talk themselves into, and out of, professional actions.8 Such an examination of logic does not entail predicting students' intentions. Rather, looking at both how students act and their understandings of why they act provides a richer understanding of their professional development and a framework for shaping it. Method Ethics approval was obtained from institutional review boards at the University of Texas (where students' essays were generated) and the University of Toronto (where the analysis was conducted). A total of 300 essays were collected over 18 months from third-year students on an internal medicine inpatient rotation, as part of a required professionalism and ethics course. Graded as “pass” or “fail,” the assignment asked students to describe an adult case, identify conflicts, and comment, using moral and professional reasoning, on the appropriateness of its resolution. The students were assured that the assignment would remain confidential. Fifty-two papers were randomly selected from this set and their identifying features were removed to preserve anonymity. These were screened for use in the study based on the following inclusion criteria: (1) the writer described a professional dilemma, (2) the narrative was reported in the first person, and (3) the writer was both present and directly involved in decision making in the situation. Papers were excluded from the sample if: (1) writers did not describe a professional dilemma, or (2) they did not describe themselves as directly involved in decision making in the dilemma (for example, some students described situations hypothetically or generally). The thematic categories described below (Table 1) were derived inductively, through grounded theory analysis.9 Their credibility is supported by careful examination of the grammatical features that underlie students' strategic moves.10 Consistent with the grounded theory tradition, data collection and analysis were conducted in an iterative, cyclic fashion to ensure that the developing sample maximized the ability to test and revise emerging theory. Preliminary analysis was conducted by two researchers (LL and KG) on an initial batch of 15 papers (“Batch 1”), from which eight were selected as meeting the inclusion criteria. Following an analysis of argument structure and thematic content in this batch, theoretical sampling11 was conducted by screening an additional batch of 37 papers (“Batch 2”), from which 17 were added to the study sample. Theoretical sampling involves using the additional batch of papers as an opportunity to confirm and refute preliminary findings arising from analysis of the first batch in order to expand and challenge the developing theory of students' argumentative patterns. Analysis of the second batch of papers was conducted by the same two researchers. Coding structures were revised and confirmed through this process, at which point the entire sample of 25 papers was coded by a single researcher (KG). Two insider experts (DS and KS) reviewed a sample of the analyses to ensure their authenticity and representativeness.TABLE 1: Sample Application of Thematic Codes to Student PapersThe sample in this study is not the papers themselves, but rather the instances of articulated reasoning (discrete passages in which the student comments on his or her behavior in relation to the dilemma at hand). The 25 papers in the study contained 180 instances of reasoning, which recurred across all the papers. Grounded theory standards require that a data set be detailed enough to adequately represent the breadth and depth of a phenomenon; following these standards, appropriate sample size was ascertained by collecting and analyzing data until no new codes were forthcoming in the analytic structure. Through this process, these 180 instances were determined to be a sufficient sample to build a preliminary theory of how students reason through and rationalize their professional behavior. Results The papers contained between one and seven dilemmas each (average: three dilemmas per paper), suggesting that students can recognize multiple choices that may impinge on a single professional “event.” Individual students employed a variety of reasoning strategies, using between three and 13 instances of articulated reasoning each (average: seven rationalizations per paper). The numbers of strategies students used to explain whether they defined their resolutions of dilemmas as “appropriate” indicated that a single dilemma can have multiple rationalizations for students. Four categories evolved from the data, each of which was further subdivided to capture prominent variations in writers' persuasive strategies or “moves.” Narrative attitude. Narrative attitude codes depict the attitudinal tone of a student's reasoning. This attitude may appear to be authentic or crafted for the purposes of advertising a particular world view. Thirty-nine instances in the sample were coded in this category. This code divided into two further subcategories: “Deflection” and “Confession.” Deflection codes (n = 17 instances) applied to instances where the student posited reasons for behavior that served to divert attention away from other motives articulated in the narrative. One student explains, “These final events occurred on the last day of the rotation and I didn't feel that I had the emotional restraint to ‘argue’ the point with my intern.” (1.1 = Batch 1: Paper 1) The posited reasoning—that s/he will “explode” if s/he confronts his/her resident may conceal another reason suggested by the reference to “the last day of the rotation”: the student's fear that his/her evaluation will be affected if s/he confronts the resident. Confession codes (n = 22 instances), by contrast, applied to instances where students accepted blame/responsibility for their decisions. For example, one student admitted, “I did not know if I had the heart or the guts to look these two people with all these plans in their faces and tell them that Mr. X was terminally ill. So I lied to them.” (2.13) Most instances of confession were accompanied by evidence of authentic regret: for example, “I felt completely incompetent, mainly because I had missed the weakness, but also because I felt as though I had let Dr. X down.” (1.8) However, there are three instances of “crafted” confession, in which regret does not appear or does not seem authentic. An example of absence of regret is the student who reflects on concealing a poor diagnosis from a patient: “Was that decision ethical? Most likely it was not. But was it understandable? Yes it was.” (2.13) Dissociation. Dissociation codes label rhetorical strategies students use to facilitate the choice of non-involvement in professional situations. Coding revealed 71 instances of this strategy. Two kinds of dissociation were evident in the papers: “Condescension” and “Identity Mobility.” Condescension (n = 38 instances) represents the demonstration of superiority or contempt in relation to a challenging event, often as a substitution for reflection or self-assessment. Condescension represents the student as standing outside of or “above” the event in order to avoid responsibility. This particular type of dissociation is apparent in the reaction of one student who observes in horror as a hysterectomy is almost performed without the patient's consent: “I was disturbed to realize that, rather than being dismissed, this notion was being actively entertained by the group.” (1.3) Identity mobility (n = 33 instances) is the vacillation between the authority of a professional (insider) medical role and the passivity of the student (outsider) role. This code reflects a reasoning strategy based on students' dual identities as professionals and as learners. Students may use this duality as a kind of immunity when faced with challenging situations or responsibilities, such as one student who explains, “I also told her the patient that the decision was not mine, and that I would have to check with the residents and the attending. I told her this in order to keep her good graces, and also so that I could deflect her request.” (2.2) Initiative. Initiative codes denote students' willingness (or unwillingness) to act in response to professional dilemmas. Forty-six instances were coded in this category. This code subdivides into two categories: “responsibility” and “confidence.” Responsibility codes (n = 26 instances) apply to instances where students take initiative out of a sense of duty, such as when one student presents a finding at rounds against the will of the resident, explaining: “I know I could not have a clear conscience if I ignored what seemed to be a significant finding.” (2.1) They view the potential consequences of acting/not acting as a burden, and therefore make decisions that facilitate avoiding this feeling. Confidence codes (n = 20 instances) apply to instances where students take initiative based on a feeling they will be successful in a situation, or avoid initiative based on lack of confidence. If they doubt their abilities, they may hesitate to take initiative, as one student, faced with presenting a “bad” prognosis does: “I quickly left the room before my worried and fearful face gave the results away. After all, I would not know what to say or do if the patient broke down and cried.” Team membership. Team membership was used in 24 instances in the sample, and subdivides into two subcategories: “Conflict” and “Protection.” Conflict codes (n = 14 instances) apply where students act in response to sensed opposition or danger from their colleagues, as exemplified by the action of one student, who performs a procedure out of fear that refusal will result in a poor grade: “I believe that I did the right thing by following the orders given my sic by the resident despite my reservations. I do not believe that I would have faced any retribution for refusing to do the blood draw, but I know I would have been fearful of such actions.” (2.16) Protection captures students who act in the interests of the group (n = 10 instances), such as one student who concludes, “I do wish there was more I could have done but for the sake of being part of the team, I took the orders.” (1.8) Application of the thematic codes to each narrative in the data set yields a schematic of the dilemmas encountered and students' responses to these. With reference to paper 1.5, Table 1 shows (1) an articulated dilemma, (2) an illustrative excerpt, (3) the nature of the student's reasoning, and (4) the strategy code, with explanation. In the sample above (Table 1), the student faces a conflict between deferring to the resident by keeping quiet and going against the resident by sharing information with the attending physician. The student chooses to provide the attending physician with findings s/he believes are important. His/her presentation of this choice as a “slip” or accident provides insight into the strategy s/he is enacting to balance (1) the desire to please the resident with (2) the sense of obligation to report the finding. The choice to tell “strains” the student's relationship with the resident, and in turn threatens the student's sense of loyalty to the medical team. To balance professional obligation with concern for team loyalty, the student takes on the narrative attitude of “deflection,” describing the action as a “slip” and implying that speaking out was a fortuitous fluke rather than an intentional betrayal. The strategy of deflection is supported by the narrative's grammatical structure, in particular its cohesive markers. Cohesive markers in texts relate what is about to be said to what has been said before, and include “additive” and “adversative” markers. Examples of “additive” markers include “and,” “or,” and “furthermore.” “Adversative” markers include “but,” “however,” and “on the other hand.”10 For instance, in paper 1.5, adversative markers reveal internal conflict in the student's reasoning: “… I decided not to bring it the finding up again, neither with him nor at rounds. It did, however, slip out of my mouth towards the end of my presentation with my attending. …” The student's use of the adversative marker “however” signals the competing alternatives facing him/her: tell or don't tell. The subsequent use of the descriptor “slip” to characterize the action of telling as accidental rather than purposeful deflects responsibility for the choice and its consequences. Discussion and Conclusions Across a wide variety of professional dilemmas, students access a narrow range of rhetorical strategies or “moves.” These strategies appear in reports of professional lapses and in reports of behavior self-identified as upstanding. In both cases, students' stories illustrated a rhetoric of rationalization, a “dancing of attitudes,”12 and strategies suggestive of writers persuading themselves, as well as their readers. Many of students' reasoning strategies created maneuvering room, foregrounding options and supplying justifications. The strategy of identity mobility is one prominent example of this. Writers skirted direct involvement or culpability by opting to identify themselves as students rather than as full-fledged medical professionals. They frequently referred to “learning” and “knowing better next time,” or to simply moving on to the next rotation. Using phrases such as “as a new third-year medical student” (2.16), students set limits on what they would and would not accept responsibility or blame for (2.15; 2.14). Prior sociologic research has identified the allotment of graded responsibility in apprenticeship,13 or “legitimate peripheral participation,”14 as both a pivotal socialization technique and a critical challenge for students. The strategy of identity mobility appears to be a generic rhetorical response to the apprenticeship role, indicating its salience in students' peripheral position, straddling the academy and the profession. The facility and appropriateness with which students used this strategy could potentially indicate their success in achieving professionalization. The rhetoric of rationalization also manifests itself in the narrative attitude of “confession.” Although many student papers contained “confession,” this strategy is usually accompanied by a modifying reason or motive. This combination of confession with rationalization suggests a kind of “backpedaling” on the part of students. They wanted to accurately present the story, and yet they hoped to cast their role in a particular light; they wanted to appear forthright without appearing completely at fault. It is important to note that rhetorical rationalization is not categorically undesirable, in students or, for that matter, in full-fledged professionals. In fact, rationalization combined with self-awareness may be a hallmark of reflective practice. In our study, some of the students' strategies could be highly functional. For instance, identity mobility may allow students to negotiate professional identities gradually, assuming more responsibility in situations for which they feel prepared, and less in those that are particularly new or daunting. Of course, the nature of the data (one essay per student at one institution) precludes generalization about individual students' reasoning patterns. And essays written in response to experiences in the clerkship setting may not reflect some important areas of professional behavior, such as community service. In addition, writing produced for a clerkship assignment might lend itself to particular rhetorical strategies while veiling others. We believe, however, that while students might misrepresent their actions and their motivations, the language patterns used in their arguments invariably provided insight into their reasoning processes. Focused on reporting what they did, the students were less likely to be conscious about the intricacies of their arguments. Students' rhetorical strategies provide insight into their storying of dilemmas and their justification of behaviors. Understanding patterns in this reasoning may assist us in anticipating common students' responses to professional dilemmas. Furthermore, once this preliminary theoretical framework is validated with data from other populations and other methods (such as discourse-based interviews16 to further probe student rationalization), patterns of rationalization may provide a link between reasoning and behaviors in the face of professional dilemmas. This understanding would help establish the parameters for a relevant and effective instructional and assessment program in professionalism.