Since 2020 it has become commonplace to read in prominent medical journals that academic medicine is permeated with racism (e.g. Lewis et al., 2023). If by “racism,” such contentions are taken to imply racial animus or beliefs in racial inferiority, they would lack plausibility to many observers. The accusation of pervasive institutional racism reaches toward plausibility only if “racism” is taken to be a descriptor, not of the attitudes of individuals (or not primarily of these) but instead of the impersonal workings of academic medical institutions. Policies and procedures, however intended, the accusation might go, work to subordinate a racial group or to maintain the privileged position of another racial group through unwarranted differential treatment of group members.

In what follows, I shall consider a recent example of the accusation of systemic racism directed at academic medical institutions. The paper, recently published in Academic Medicine, is Anderson et al.’s “The Long Shadow: A Historical Perspective on Racism in Medical Education” (Anderson et al., 2023). Anderson et al.’s paper is of interest not merely on account of its indictment of contemporary medical education but for the historical context offered by the authors. They contend not only that racism is “embedded in the operating system” of present-day medical education but that that state of affairs is a natural development of the history of the United States as a mostly Christian country owing its culture and religion to Europe and the European Enlightenment. As “systemic racism” is a presently contested concept, I shall begin with a discussion of systemic racism and of alternative ways of conceiving what merits that designation. Having developed what seems to me a plausible use of the concept for purposes of this paper, I shall go on to consider the merits of Anderson et al.’s accusation as regards both contemporary United States medical education and the broader history in which that medical education system developed.

Racism, System, and Systemic Racism

Racism has, of course, been a prominent topic of academic soul searching in the past several years. It is notable that the concept has recently been a contested one. Racism traditionally was considered to be a belief in the inferiority of a given racial group, that is, a group distinguished by common biological ancestry; or discrimination and/or prejudice based upon such a belief (Merriam Company, 1967). Since the 1990s there has been ongoing debate regarding whether to maintain a “narrow scope” concept of racism, with its connotation of significant moral evil (Blum, 2002, 205) and its historical connection to the Nazi’s (Fredrickson, 2002, 162); or to expand the scope of the concept to encompass a much wider range of attitudes, practices, and social phenomena, particularly as implicating institutions and systems rather than individuals. Opponents of such “conceptual inflation” (e.g. Blum, 2002, 205) worry that the word will lose its properly condemnatory sting and meaning if it is overused. Some advocates of a “wide scope” concept (e.g. Shelby, 2014, 66) argue that the concept of racism, if it is to the do the work it needs to do, must encompass ideologies that produce unjust social relations. Others (e.g. Urquidez, 2021, 682) dispense with beliefs and attitudes altogether and define racism as simply a societal system of racial oppression.

The wide scope racism concept has been making headway in public discussion, as indicated by the 2020 revision of the definition of racism in the Merriam-Webster dictionary. The definition in the 1967 edition had by then evolved to include not only discrimination and prejudice, but also doctrines or political programs founded upon beliefs in the inferiority of racial groups.Footnote 1 The newly revised definition (Merriam-Webster, n.d.) added systemic oppression to the racist manifestations of beliefs in racial inferiority mentioned in the earlier definition. It might be contended that the change was minor, that racist systems or institutions go hand in glove with racist doctrines and programs. But the change has focused attention on perhaps the central divide in present discussion of racism, which is not whether a wide-scope definition of racism is plausible or credible; it is widely agreed that it is. To that extent, the new dictionary definition reflects an expansion of the concept of racism that has been and is occurring. What is presently in dispute is how to bring the “systemic oppression” aspect of the concept to bear; that is, in how to determine what falls under the designation of “systemic” or “structural” racism. There are two opposing poles or tendencies in this debate. In the ascendant is a radical group of scholars who adhere to what Andreas Wimmer has identified as “race-centrism”. These scholars take race to be determining of socioeconomic relations and status in the United States and (virtually) all racial disparities to the disadvantage of black people to be due to racism, past and present (Wimmer, 2015, 2186). They are opposed by other scholars of race in American society for whom racial disparities are a starting point for investigation rather than knockdown evidence of past and present-day racism as their sufficient cause. Before offering a definition of systemic racism for present purposes, I shall describe and assess these opposing positions.

Race-centrism

As might be expected, what constitutes “racism” is very much at issue between these two approaches to studying race, as is the concept of “system” as an aspect of racism. The general approach of the race-centrists is to deem American society to be a system, to observe racial inequality in multiple domains in that society, and to posit past and ongoing oppression of blacks by whites as the sufficient and exhaustive explanation of that inequality. This explanatory strategy is often evident in definitions. For example, Paradies defines racism as:

A societal system in which actors are divided into ‘races’, with power unevenly distributed (or produced) based on these racial classifications. (Paradies, 2006, 44)

Banaji, Fiske and Massey (2021) offer a recent definition of systemic racism similar in spirit to Paradies’ definition of racism:

Simply put, systemic racism refers to the processes and outcomes of racial inequality and inequity in life opportunities and treatment. Systemic racism permeates a society’s (a) institutional structures (practices, policies, climate), (b) social structures (state/federal programs, laws, culture), (c) individual mental structures (e.g., learning, memory, attitudes, beliefs, values), and (d) everyday interaction patterns (norms, scripts, habits). Systemic racism not only operates at multiple levels, it can emerge with or without animus or intention to harm and with or without awareness of its existence. (Banaji et al., 2021, 2)

On the race-centric view, the racial inequality that is a persistent feature of American society is due to racism at all levels of individual and social reality; and that racism may have no necessary connection to anything designated “racism” by older definitions of the concept. Racism for the race-centrists has become any process or mechanism that allows or furthers the persistence of racial inequality.

How do the race-centrists defend their view of racism as the sufficient and exhaustive explanation of racial inequality? They draw upon critical theory and its postmodernist connections, for many race centrists the inspiration for their scholarly work.Footnote 2 Such theory often exhibits pessimism about the possibility of objective knowledge of the social world or its history. In the view of many race-centrists, academic discourses idealizing objectivity demand debunking or deconstructing to reveal these discourses for what they are: discourses produced and employed to maintain the hegemony of white males in our social hierarchy.Footnote 3 While the race centrists do claim that their accounts of society are true, they often foreground social change as the most important goal of their work (e.g., Zuberi & Bonilla-Silva, 2008, 23). While scholarship-as-science is an unacceptably “depoliticizing” goal, scholarship as social amelioration is demanded by the moral imperative to fight oppression.Footnote 4

Race centrist scholars take a 30,000 foot view of American society;Footnote 5 they observe the racial disparities in the United States and their explanation for these (with its practical implications) straightforwardly follows from the observation—as interpretive description of inequality can suffice to reveal who is oppressed and who are the oppressors.Footnote 6 And, the voice of the oppressed is a dispositive authority for how best to understand the oppression (Zuberi & Bonilla-Silva, 2008, 332, Winant, 2015, 2182). Furthermore, racism is both reified as a macro-level social force with causal powers by race centrists; it is also taken to overshadow other candidate macro-level totalizing concepts for understanding society such as class, capitalism, or the patriarchy. As Andreas Wimmer points out, for the race-centrists, racism trumps other potential explanations of social stratification. It is a “master category” for understanding “patterns of inequality, marginalization, and difference throughout US history” (Wimmer, 2015, 2186-2188).

Race and Social World Realism

A range of scholars in history and sociology from across the political spectrum find the race-centrists’ approach to understanding the social world to be implausible. As a group these scholars resist labeling, except insofar as their approaches all presume the possibility of rigorous and impartial inquiry into the social worlds of past and present that can transcend the limitations of politics and ideology. For lack of a properly descriptive term for this scholarly approach, I shall hereafter refer to such scholars as “social world realists” (or “realists”). The fundamental divide between these scholars and the race centrists is over the possibility of objective historical and sociological inquiry, which realists affirm against the insistence of many race-centrists that any such affirmation is ideological. The ideal of scholarly objectivity must be understood carefully; realist scholars do not pretend to ideological neutrality either as an ideal or as a possibility. What they insist upon are investigative approaches to the social world that do not presume answers to causal questions before the inquiry has begun. Scholarly rigor and detachment are virtues not because they signify political neutrality but because they lay the ground for “higher levels of understanding—…higher in Nagel’s sense of being more complete, more cognizant of life’s most seductive illusion, which is that the world centers on me (or those with whom I choose to identify)” (Haskell, 1990, 134). Social world realist scholars see the possibility of differing perspectives favoring differing answers as to how best to understand social and historical phenomena, but they demand that scholarly work from any perspective answer to an ideal of impersonal truth, even if that ideal is not fully attainable. Such scholarship, in their view, will be a better starting point for moral critique and social improvement than work more overtly political that, from the realist viewpoint, has presumed from the outset what it purports to conclude from investigation.

Realist scholars are skeptical of the sort of totalizing social theory offered not only by race-centrists but by others who fix upon one aspect of the social world such as class, the means of production, gender, or colonization; who then reify concepts such as racism, classism, imperialism, or sexism as persistent macro-level social entities; and who then account for all of social reality on the basis of such a master category (Niemonen, 2010, 64–65, Reed Jr, 2016, 261–262). In the case of “systemic racism,” realists accept the cogency of the concept but construe it and its extension in the social world very differently than do race-centrists. This follows partly from their view that the macro-level is the wrong place to begin in seeking understanding of social life. Valid macro-level generalizations in history or sociology will be modest and will emerge not from simply observing the macro-level but from investigation at micro- and meso-levels.Footnote 7

To begin with the concept of racism, realists resist its detachment by the race-centrists from the traditional meaning of the word as racial animus or beliefs in racial inferiority. While accepting the possible bearing of racism on social systems, and, thus, the possible independence of systemic racism from individual attitudes, they insist that racist systemic arrangements be those that actively subordinate a group on the basis of race rather than any arrangement in which racial disparities are present and persist. Race-centrist “systemic racism” is too all-encompassing to be a useful analytic category. If almost everything that (most) whites think and do is labeled as racist, realists contend, the concept loses both analytical and moral bite. In any case, an expansion of the concept of racism is unlikely to achieve uptake in a linguistic community if it has no intelligible connection to usual linguistic usage, according to which racism and racists are vicious. Expanding the concept to encompass practices and arrangements that invidiously discriminate even if they do so absent anyone’s conscious intention is a natural emendation of usual usage; expanding it so as to ensnare any white person in United States society on the grounds that she lives among racial disparities is not.

The notion of “system” in social analysis is also restricted by realist scholars to social phenomena that conform to the basic idea of a “system”: a whole composed of parts that act upon one another in predictable ways. “System” can be predicated of families, courts, markets, hospitals, and even of collections of these, as in the case of neighborhoods, the criminal justice system, the health care system, and national arrangements of markets. At the higher level of the social world of a given society, the “system” analogy breaks down. United States society is far from a system. It more closely approximates Daniel Little’s description of the social world: “It is not a system but rather a patchwork, a mixture, an ensemble, a Rube Goldberg machine, a collage, or a jumble” (Little, 2016, 3). Order that can be discerned at this level is “the result of a large number of overlapping, independent, conditions and processes, not the manifestation of a few simple forces or a guiding system of laws” (Little, 2016, 4). Accordingly, it will not do to simply observe the macro-level, note patterns such as racial disparities, and conclude without further ado that a reified social force such as racism suffices to explain the pattern. Investigation must descend to actual social systems and determine the causal efficacy of actual social arrangements in determining social outcomes such as racial disparities in differing domains. To qualify as racist, social processes or arrangements must actively discriminate (invidiously) against a racial group. If the race-centrist picture of American society as systemically racist is to be sustained, that picture must be built up out of demonstrated racist arrangements in actually existing lower level systems rather than simply asserted. To the social world realist, the race-centrist picture of the United States is at best a hypothesis to be tested. As Wimmer contends, that picture.

…can be connected to a mechanism-based, analytical understanding of social processes whose direction can only be determined by empirical investigation and whose complexities need to be unpacked with appropriate tools. Problematizing axiomatic assumptions and probing deeper into how one could show whether or not they indeed hold up to an empirical analysis defines a series of research agendas for the future. (Wimmer, 2015, 2201)

Why the Social World Realist Concept of Systemic Racism is to be Preferred

There are compelling epistemological and moral reasons to prefer the realist view of systemic racism, according to which the concept bears upon identifiable systems that demonstrably work to invidiously discriminate against racial minorities, to the competing race-centrist concept. The epistemological reason follows from a necessary rejection of perspectival relativism about social reality. For there to be useful disputation between race-centrists and social world realists, there must be a shared presumption that truth about the social world transcending particular perspectives is attainable; that is, that objectivity about the social world is not a white male plot but a valid and necessary ideal of academic inquiry. If that is so, race-centrists must accept the realist position that impersonal disciplinary methods of inquiry are useful for answering questions appropriately put to them. Winant takes this position in his rebuttal of Wimmer. His contention is that Wimmer gets race in society wrong not because he is using illegitimate methods but because he is overusing sociological science at the expense of idiography. The realist reply would be to agree that quantitative tools and idiography indeed must be used in tandem; but to question the a priori privileging that the race-centrists accord to the voice of the disadvantaged as idiographically described by they, the race-centrists, in determining the correct account of racism in society. If a perspective-independent social world is out there to be investigated, then the race-centrists no less than social world realists need to begin its investigation with humility as to what truths about society the investigation will reveal. Society-pervading racism must be demonstrated beginning with its instances at the micro-level using disciplinary tools, not simply asserted to be present because disparities exist and, in Wimmer’s phrase, “politically engaged, emancipatory, hermeneutic, ‘good’ social science” (Wimmer, 2015, 2203) takes non-white “ordinary people” (Winant, 2015, 2182) to know their true explanation.

The moral reason for preferring the realist view of systemic racism is the avoidance of moral condemnation where such condemnation is not warranted and the provision of grounds for condemnation when it is. Racism is vicious; when present it should be condemned accordingly and, if present in systemic arrangements, expunged from them. The race-centrist contention that racism is all-pervading and that all whites are complicit in it (as per Bonilla-Silva, 2021, 514), unless and until vindicated by rigorous empirical inquiry, is slander. And the corresponding antiracist conception of white moral responsibility, according to which whites buying houses, choosing school districts, making friends, or hiring employees are all racist unless in the act expressly redressing racial disparities (“inequities”) (as per Kendi, 2019a; 18; or Bonilla-Silva, 2021, 514) reduces the world and its moral navigation to a totalizing postulate and a totalizing imperative to compulsory action, with totalitarian implications.Footnote 8

Is Academic Medicine Systemically Racist?

In considering the recent barrage of accusations that academic medicine is systemically racist, and Anderson et al.’s (2023) accusation in particular, I shall proceed from a realist definition of systemic racism, as adumbrated in the previous sections. Systemic racism, I shall contend, is best thought of as arrangements or processes in an actual social system, such as individual academic medical centers or groups of these, that work to disadvantage members of a minority racial group on the grounds of group membership. For Anderson et al.’s (2023, S31) assertion that racism “is embedded in the operating system” of medical education to be vindicated, it must be shown not only that racial achievement disparities in medical education exist and persist, but that these disparities are caused by ongoing invidiously discriminatory processes.

That racial achievement disparities exist is beyond question. Comparisons of minority and white student performance as measured by test scores (Davis et al., 2013; Kleshinski et al., 2009), clerkship grades and honors (Teherani et al., 2016), and honor society membership (Boatright et al., 2017) all reveal disparities favoring white students. Anderson et al. write as if it may be taken as given that such disparities are due to structural/systemic racism in medical school assessment arrangements. But this is so only if the metrics used to rank applicants and trainees in medical school assessment during training are themselves racist: that is, the metrics disadvantage minority students not for reasons related to academic or clinical performance but for other race-correlated reasons. To demonstrate the racism of medical school assessment standards, Anderson et al., must meet the challenge articulated long ago by Randall Kennedy in response to similar accusations about standards for assessing law students:

The legitimacy of a given standard, however, cannot properly be determined wholly by reference to consequences measured by bare statistics - such as disparities between the numbers of students of color in law school and the number of professors of color. The statistics generated by a given standard may well provide a predicate for questioning it. Results indicating that a given standard disadvantages the members of one group relative to others may indicate that the standard itself needs reform. On the other hand, statistics may indicate that those who failed to satisfy the criteria in question are themselves in need of reform. Ascertaining which conclusion to reach in a particular context always requires more than statistics. It requires recourse to a complex set of normative and descriptive assumptions. (Kennedy, 1989, 1763)

A case for systemic racism in medical school assessment might seek to get off the ground with the observation that high stakes objective testing in medical training both disadvantages minority trainees and is weakly correlated with clinical performance. The suggestion might then be made that as clinical performance is the outcome of interest for medical educators, if metrics used to assess trainees both fail to correlate (much) with clinical performance and also disadvantage minorities, they are systemically racist (Slavin, 2022). Such a verdict would be too hasty. Clinical expertise (or competence) is complex and multi-faceted (Schuwirth & van der Vleuten, 2020). Measures of a proximal aspect of its development, such as knowledge, may correlate only weakly with assessments of the use of said knowledge in making sense of clinical cases on the wards. It does not follow that knowledge is not important in the development of expertise or that its achievement ought not to be measured by objective written tests. If any conclusion has emerged from the recent history of assessment in medical education, it is the importance of multiple modes of assay of the quarry to be assessed, including both psychometrically robust “objective” measurements such as written tests and more subjective ratings by those who bring clinical expertise to its assessment in trainees (Govaerts et al., 2019; Rothoff et al., 2021). The latter may be variable but need not lack validity if sampling of trainee performance is adequate and multiple raters are involved in its assessment (ten Cate & Regehr, 2019).

The vulnerable point of trainee assessment for discrimination issuing from racism is not the succession of objective tests that trainees undergo. These tests, involving the minutiae of basic and clinical science, almost certainly do not contain questions assaying culture or community-specific knowledge unrelated to professional work that might advantage white over black students.Footnote 9 The vulnerable point is instead the subjective ratings of faculty and more senior trainees that afford the possibility of race-specific bias creeping into assessment. Such bias, if it is in fact present, would indicate direct (“taste based,” in economic parlance) rather than systemic racism, as raters would be using racially neutral rating rubrics in a racially biased fashion.

Is such racism at work in the racially discrepant outcomes of the assessment of students on the wards in medical schools? As a longtime educator in this context, I think it unlikely; but any plausible answer to the question will be the result of rigorous investigation rather than opinion or anecdote. Fortunately a way forward to answering the question is suggested by economists’ use of outcome tests to assess bias in decision-making (Canay et al., 2023). The idea is that if a group is being disadvantaged by gatekeeper decisions that do not reflect group-level differences relevant to the task performance being assessed, such discrimination will be reflected in post-decision outcomes across groups. If, for instance, clinical faculty demand a higher level of performance from black medical students than from whites for awarding honors in a clerkship, black students at the margin of honors awards will be performing better than similar white students. Presuming there are other measures of clinical performance that are not biased against black students, such as the written clerkship exam, these can be used to detect discrimination in faculty global ratings. Presuming a general positive correlation between faculty global ratings and the clerkship written exam, comparing black student performance on the written exam among those at the honors margin with that of white students at the same margin will show similar group scores if honors awards are fair. If blacks are being discriminated against, black students at the honors margin will likely have higher written exam scores than their white counterparts.

Outcome tests have not, to my knowledge, been brought to bear on the question of medical student assessment. Yet, some such investigation is necessary for the proper evaluation of accusations of racism in medical student assessment, as made by Anderson et al., and others. Pending the results of such investigation, it is plausible to suppose that the results of medical student assessment reflect medical student performance and that group differences in assessment outcomes reflect group differences in academic preparation for professional work. Anderson et al. suggest that present assessment metrics be replaced by an exclusive reliance on mastery based assessment, in which trainees all receive a single rating of “pass” to proceed to the next level of training or to be judged competent to practice. Such competency judgments are important at the margins of passing and failing trainees but must not become the only mode of trainee assessment.

Mastery based assessment is a crucially important development reflected in the “competency” movement that has swept over medical education in the past twenty years. There is no question that it has enabled the focus of educators on a critical function of medical training: the judgment that trainees have reached thresholds of competence to progress to ascending stages in responsibility for patient care and eventually to readiness for independent medical practice. Medical education perhaps serves no more important role for society than in assuring that trainees are ready for practice by the time they complete their training. However, judging threshold levels of competency is not the only task of medical training and as educators have focused more closely on defining and assessing clinical competence, they have become increasingly aware of dimensions of clinical work that are not captured by the competence construct but that are important elements of the educator’s task.

Operational judgments of competence, such as “can perform a complete and appropriate assessment of a patient,” may underestimate the importance of context in abilities to perform a task. Trainees who can appropriately assess some patients may fall short in assessing others (Eva et al., 2016). Attempts at defining elements of complex tasks, the presence of which can be taken to signify competence fall prey to the fallacy of supposing that such building blocks of competence add together to the thing itself; whereas in reality the ability to skillfully use the building blocks of competence in clinical work is something beyond their mere additive possession (Huddle & Heudebert, 2007). Furthermore, judging trainees to be competent for particular tasks risks omitting important aspects of expertise that transcend specific tasks—such as “tolerance of ambiguity, agility in the face of complexity, and habits of mind such as curiosity, innovation, and a commitment to life-long learning” (Hodges, 2010). Focusing on minimal levels of adequacy of clinical work ignores variability of performance in the passing range and the importance of encouraging not merely minimal competence but proficiency (Eva et al., 2016). While competency judgments are important for standard setting, adequate assessment cannot do without global ratings of clinical work that can capture more nuanced aspects of expertise and convey judgments of lesser or greater approach to an ideal of excellence (ten Cate & Regehr, 2019).

In sum, Anderson et al.’s accusations of racism in medical school assessment are not substantiated. Their suggested remedy for racism, an exclusive focus on threshold competence in assessment of trainees, would impoverish medical education and its assessment. And limiting assessment to certifying threshold competence would not suffice to eliminate racism from medical education if racism is present in the subjective dimension of clinical assessment—as subjective judgment is an inescapable component of threshold competency judgments.

The “Long Shadow;” the History of Western Medicine and of the European Christian West as a Tale of Racial Oppression

Unlike the story they offer of performance disparities, the account of medical education’s history offered by Anderson et al. breaks new ground. To my knowledge, no one before them has tied Christianity, Europe, the Enlightenment, 19th century racial “science” and twentieth century medical education reform into one comprehensive story of racial enmity and oppression. This dark master narrative is presented as an established truth in the early sections of Anderson et al.’s paper (2023). It is sourced to 9 scholars (Anderson et al., 2023, notes 7-15) of varied disciplines, who provide individual links from which the chain of this story is forged.

One of the nine, African American Studies scholar Terence Keel, ties pre-modern Christianity to post-Enlightenment racial science through the work of Professor of Religion Denise Kimber Buell, whose revisionist interpretation of early Christianity reads “ethnic reasoning” into the New Testament and the church fathers (Buell, 2005). In Keel’s hands, such reasoning becomes itself a historical actor ranging across time and place (Hamm, 2019) to tie Christianity to ostensibly non-Christian post-Enlightenment science (Keel, 2018). It is Anderson et al. themselves who link this larger story to the more particular stories of United States medical educational reform, thus fathering the racial sins of 19th and 20th century organized medicine upon European civilization.

Far from being established truth, this broad picture is deeply controversial both in its overall sweep and in its details. Denise Kimber Buell and Geraldine Heng are able to portray New Testament language about Christian identity as racial and Christian ill-treatment of Jews in the Middle Ages as racism through expansive definitions of racial thinking and racism. In contrast to common usage according to which “race” denotes common biological ancestry, Buell’s “ethnic reasoning” expands to encompass early Christian talk about Christians being a chosen people, as are the Jews (Buell, 2009, 119–120). Such language is certainly present in early Christian authors but it is not at all clear that it will bear the weight of the racial sense Buell wishes to attribute to it. She gets to that sense by redefining racial thinking as involving ways of classifying people not only fixed but also fluid—so that ancient talk of peoples (using several different words that do not translate to race: ethnos, gens, laos) can be painted as racial in our modern sense of the word. The contention that the significance of such talk is to signal racist potentialities in Christian thinking is implausible to many scholars (e.g. Rubies, 2017, 45; Gruen, 2020, Ch. 11-12) and is in stark contradiction to usual theological readings of many New Testament texts (La Du Toit, 2020). Geraldine Heng (2018) uses a similar definitional strategy to place Medieval Christian ill-treatment of Jews squarely under the category of racism. Rather than its usual definition as racial animus or a belief in racial inferiority, racism for Heng is:

…one of the primary names we have…attached to a repeating tendency, of the gravest import, to demarcate human beings through differences among humans that are selectively essentialized as absolute and fundamental, in order to distribute positions and powers differentially to human groups.” (Heng, 2018, 3)

In short, the move is conceptual engineering; take all of the myriad ways in which human beings have identified themselves in deep seated ways as in-groups over against out-groups, of which racial distinctiveness is one such way, and declare them all to be “racism”. Once this move is made, group antagonisms from the dawn of time to the present may be said to manifest racism. But this prompts the question raised by William C. Jordan early in the debate over the origins of racism. “Is every hatred a form or variant of racism?” (Jordan, 2001,170) Jordan makes the point that historians have generally preferred to understand the past on its own terms rather than to homogenize it into the political categories of the present. Critical race studies scholars such as Heng might reply that failing to use “race” to characterize past religious or culture-inflected hatreds would be to be untrue to “important strategic, epistemological and political commitments (Heng, 2018, 23). The past is to be brought to lend its weight to the right side of present day political battles; even if, perhaps, the past so mobilized is such as would be unrecognizable to those who actually lived in it.

The heavy lifting in Anderson et al.’s argument that the roots of 19th and early 20th century scientific racism are to be traced, not merely to Christianity, but to a white European Christianity, is performed by Terence Keel’s Divine Variations: How Christian Thought Became Racial Science (Keel, 2018). Keel follows Buell in asserting that Christianity is not fundamentally universalist but is instead essentially tainted by “racial reasoning.” He then ties Christianity to the work of Enlightenment natural historians doing early work on human taxonomy, in particular, Johann Friedrich Blumenbach. The argument proceeds: Blumenbach was influenced by Christianity. Blumenbach is “the father of modern racial science”. As modern racial science bore fruit in the eventual development of eugenics and scientific racism, these latter are best understood as natural growths from the Enlightenment and its Christian roots.

The argument is curious in that it ties together huge and complex historical developments that most historians have wanted to view as opposed to one another or, at the least, very different. What are Keel’s grounds for positing an organic connection between Christianity and Blumenbach’s racial typology? Keel proceeds by drawing a parallel between (his version of) Blumenbach’s genealogical scheme and Martin Luther’s (supposed) racial reasoning about human origins. He argues for a connection between these not on grounds that there is any evidence of actual connection but because they purportedly served a similar function in the worldviews of two different historical eras. Keel is here appealing to a 20th century German philosopher, Hans Blumenberg. Blumenberg thought that intellectual history sometimes proceeded by a continuity of questions accompanied by change in accepted answers to these questions. If a religious answer to a question became obsolete and was jettisoned, the question still had to be answered. Its “answer position” was vacant and demanded filling. The secularized answers to questions previously answered by religious views, Blumenberg thought, might best be seen as “reoccupation of answer positions that had become vacant” (Blumenberg, 1985, 65). It is in this highly attenuated form that a connection exists, in Keel’s view, between Christianity and 18th century racial classifications of humanity (Keel, 2018, Ch. 3).

Keel’s Blumenbach and his Luther are both questionable versions of those historical figures (Hamm, 2019, 240–243; Junker, 2019). But, the larger problem with his argumentative strategy is the notion of historical connection it implies. Blumenberg’s contention that important questions and thus “answer positions” remain the same across historical epochs is eminently questionable (Gordon, 2019, 164–167). But even if that were granted, the suggestion that suitedness to an “answer position” is enough to deem different ways of thinking kin to one another is implausible. If it were so, any intellectual development at all could be praised or blamed as the source or outcome of another. The Enlightenment becomes simply the Inquisition’s next phase. Atheism is simply a another stage in Christian thinking. Day does not simply follow night, it is night in another form. Historical continuity is established at the price of draining that continuity of any actual content (Hamm, 2019, 243–244).

Denise Kimber Buell and Geraldine Heng have not established that racism or racial reasoning were or are essential to Christianity. And Terence Keel has not shown that 18th century racial classifications were “profoundly Christian”; or that scientific racism owes anything to Christianity, directly or indirectly. The final suggestion in Anderson et al.’s story, that American medical education was and is steeped in a racism that descended from these sources is similarly unfounded. That American medicine was entangled with slavery, segregation and other social ills of the past does not suffice to show that these ills were essential to its character. The problematic move made by these scholars is the inference from racist Christians, Europeans, imperialists, whites, or physicians—which there undoubtedly have been in the past and are, unfortunately, present today—and racism as a core or essential feature of their being as members of these groups. What these scholars offer us is not history as traditionally understood, but history as theory; in this case, critical theory, which has unfortunately, increasingly infiltrated the medical literature (Huddle, 2022). Its aim is to put a form of life—what attackers and defenders variously refer to as western civilization, western liberal culture, or liberal humanism—in question. The means of doing so is to take the form of life at issue and essentialize it: claim, that is, that whatever its manifestation in the world may happen to be, its real, perhaps hidden essence, the tendency or force lying at its heart, is oppression. The variety of oppression varies with the brand of critical theory brought to bear, Marxist, feminist, postcolonial, or, in this case, race-based. On the latter view, Europeans, Christians, whites, and imperial functionaries do not exhibit racism sporadically or idiosyncratically; racism is part of the core of who they are. Their history can thus be told as a master narrative of racial oppression varying in form but not in essential character through successive historical epochs.

The difficulty with any such master narrative is that it is brought to the interrogation of the past rather than derived from such interrogation. Its truth claims are axiomatic rather than empirical. Rather than regarding the past as a realm in which the master narrative can be compared with what actually happened, the past is (just) to be constructed so as to display the narrative. And, the narrative is a weapon to be wielded in a present day political struggle. What is at stake here is our understanding of the relations of past and present. Do we approach the past first as a realm of antecedents of some salient aspect of the present and make history one long story of progress or regress to the successes or predicaments of our own day? Or, do we seek the past in its alterity, recognizing that kinships and antagonisms of other times, while perhaps superficially similar to those of today, have to be understood on their own terms? This is an old dispute among historians and until the past 20 years or so, it would have been regarded as long settled. “Presentism” or “whig history” (Butterfield, 1931), as the first approach became known, was viewed as unacceptable in history writing. Historians should seek (first) not to grade their sources but to understand them. Or, so it was widely presumed.

What the history presented by Anderson et al. reveals is a recent rebellion against realist historical orthodoxy, brought to the field from literary and postcolonial theory. Social world realist historians conceive of people moving through the stream of time speaking and acting in the idiom of their times and places, seizing upon ideas and conceptions available in a given context and using these according to chosen purposes of making arguments or demarcating in and out groups or changing the world in some fashion. The contrasting strategy of critical theory is to see people of the past as conduits for hypostasized metaphysical forces such as racism, imperialism, capitalism or colonialism, forces which are what they are whenever or wherever manifested in the human world. Presentism has returned with a vengeance to (some) historical study (Goldman, 2023). Whether the results are good history is not simply an established fact, as Anderson et al. contend, but instead a deeply contested matter.

It is unsettling to see huge and complex disputes in the humanities and social sciences reduced to dogmatic certainties offered as established truth in respectable medical journals. Anderson et al.’s piece is, however, not an isolated case and reflects a new and worrisome tendency in the medical literature. Medical journals have always been vulnerable to hubris in publishing pieces and taking stances on social and political issues beyond the purview of professional expertise. That reflects an all too common tendency among physicians to assert professional authority in the absence of warrant for doing so. But, in the recent past we have seen an explosion of race-centrist accusations and assessments of racism in medicine in major medical journals without any hint of an accompanying awareness of the empirical literature in history, sociology, and economics that would expose the implausibility of these accusations—that is, absent the deeply problematic theoretical framework of the race-centrists. This flowering of “overcritique” (Kilminster, 2013) in major medical journals is especially troubling given the high standards of scientific rigor generally evident in the science published alongside the accusations of medical racism. The explanation is in part a salutary moral awakening over race in our public discourse since George Floyd’s death, that has affected academic medicine as it has affected most other sectors of our society. This awakening could be of great benefit in our struggle to erase health care and health disparities; or it could impede that struggle if the causes of the disparities are misunderstood as a result of dogmatism masquerading as empirical inquiry. The proliferation of articles such as “The Long Shadow” in the medical literature exposes a need for more rigorous review of journal submissions on topics of great societal moment by authors whose expertise is primarily medical. If medical authors are to venture beyond medicine to analyze society, medical journal editors must ensure that they do so according to the canons of scholarship not of medicine merely, but of (traditionally realist) history, sociology, and economics. Editors need to broaden their net of reviewers to include non-physicians able to assess historical offerings such as “The Long Shadow” properly. If such articles are to be published, for which a case can certainly be made, they should be accompanied by counterpoint or editorial comment clarifying the larger scholarly context of the controversial positions argued for in such articles. Failure to do this will open medical journals to the charge of partisan hackery, as is now an increasingly common perception (Daniels, 2019; Editorial Board, 2020).

Conclusion

American medical education is systemically racist and very naturally so, as the Western civilization out of which it emerged is, at its core, racist. This assertion, made by Anderson et al. and many others, follows from an engagement with the past and present informed by a particular theoretical stance, which I have described as “race-centrism” informed by critical theory. My argument is that critical theory’s stance toward knowledge about the world is at best in tension with incoherence and, possibly, at its core incoherent. If objectivity about the social world is a false ideal, then race-centrism fares no better as an explanatory account of that world than does the empirical inquiry into the past and present pursued by the social world realist scholars who reject it. Each is merely a discourse arbitrarily preferred by its adherents. If, on the other hand, there is a perspective-independent social world out there to be grappled with, and the possibility of sound knowledge as a result, then investigation of that world and resulting explanatory accounts must be judged not by their political implications, but instead by the canons of disciplinary inquiry impugned by critical theory as ideological. This is the insight that social world realist scholars of race in American society bring to their confrontation with race-centrism in history and sociology. A full weighing of the respective merits of these competing approaches to scholarship is, of course, too tall an order for an analysis of a paper such as Anderson et al.’s, although I have suggested that there are important reasons to doubt the cogency of race-centrism. What I have shown, if my argument succeeds, is that the race-centrist condemnation of medical school assessment methods, and the history out of which those methods purportedly developed, as elaborated by Anderson et al., will not withstand critical scrutiny.

Academic physicians are in the midst of a re-thinking of medical trainee assessment and of ways in which to lessen the racial performance disparities such assessment reveals. Attempts to improve assessment should be cognizant of the complexity of clinical competence and expertise and of the importance of multiple modes of assessment for interrogating both. Such attempts must be accompanied by redoubling efforts to prepare minority trainees prior to medical education and to help them during medical training to succeed. These efforts will be aided by careful study of performance disparities and of their causes. Such study ought not to be pre-empted by premature declarations that their cause is racism or by impoverishing remedies such as focusing on competence to the exclusion of expertise.