Garth Zietsman deserves credit for addressing a subject most commentators avoid (Why is racism bad?, 6 February). Liberal societies should not place topics beyond examination by decree. The willingness to question moral orthodoxies is part of intellectual seriousness.
But precisely because the subject is combustible, it requires sharper distinctions than the article provides, and perhaps a preliminary question: what problem is being solved by mounting a defence of racism at all?
Racism, as a term, is overwhelmingly understood as a moral failing: prejudice or discrimination against persons on the basis of race. One may argue that certain race-based generalisations can be instrumentally rational under particular conditions. But to frame the discussion as a defence of “racism” (even if limited to certain circumstances) is to invite confusion between empirical observation, private prudence, and moral endorsement. It is not obvious that liberal discourse is improved by rehabilitating a word so deeply associated with injustice, even if one believes some narrow cases might be defensible.
It is uncontroversial that human beings use heuristics. We operate with probabilistic expectations about the world: we generalise. From an evolutionary standpoint this makes sense. Mental shortcuts conserve cognitive energy and may once have improved survival. Prejudice, in this narrow sense, is neither mysterious nor uniquely modern.
The question is not whether group-level generalisations exist. It is whether their existence provides moral permission to translate them into race-based judgement in individual interactions.
Demanding premise
Liberal individualism rests on a demanding premise: that the default mode of social engagement should treat the person before us as an individual rather than as a statistical proxy for a group. This is not because group differences never exist. It is because once we allow group averages to guide treatment of individuals, the epistemic and moral risks escalate quickly.
There are several layers to that risk.
First, statistical disparities do not establish causation. They certainly do not establish genetic causation. Race is an imprecise and internally heterogeneous category. Inferring stable, biologically grounded traits from socially defined group averages requires a degree of confidence that current science does not justify.
Second, even where correlations are real, they are blunt instruments. Group averages tell us nothing reliable about the individual standing in front of us. Acting on them inevitably produces false positives and false negatives. A society that normalises such reasoning will misjudge large numbers of people, even when the underlying statistics are accurate.
Third, there is a scaling problem. An individual making a private decision that affects only himself stands in a different moral and institutional position from a policymaker crafting rules for others. What may be understandable as instinctive caution becomes far more consequential when elevated into institutional practice. History shows how quickly probabilistic reasoning can harden into structured exclusion.
The article also moves too quickly in equating intimate personal preference with broader forms of exclusion. Choosing a spouse is categorically different from refusing commercial, professional, or civic interaction. Liberal societies have long recognised this distinction. The former lies within the sphere of intimate association; the latter concerns participation in public life. Collapsing the two risks blurring boundaries that protect equal standing under the law.
Indispensable
None of this requires denying uncomfortable data or suppressing inquiry. On the contrary, a commitment to free research is indispensable. If hypotheses about group differences are false, rigorous investigation will expose that. If disparities exist, understanding their causes – environmental, institutional, cultural, or biological – is essential to addressing them effectively.
But the liberal commitment to inquiry must be paired with epistemic humility. Moving from “there are statistical differences” to “racism can be neutral or good” is not a small step. It is a normative claim that carries social and institutional implications far beyond private risk assessment.
There is also a subtle danger in framing prejudice as sometimes instrumentally justified. Human beings already have a strong predisposition to think in group categories. Liberal civilisation has required centuries of discipline to curb the political consequences of that tendency. To legitimise race-based heuristics, even conditionally, risks loosening norms that protect individual standing.
Hard questions should be asked. But we should also ask whether every provocation clarifies our principles – or whether some merely blur them. A free society depends not only on open inquiry, but on restraint in how we draw conclusions from it, and in how we choose to frame those conclusions.
The deeper liberal insight is not that group differences never matter. It is that treating individuals as reducible to those differences corrodes social trust and legal equality over time. The discipline of judging the person before us individually, even when doing so requires effort, is one of the key achievements of the modern liberal order. It deserves to be protected.
[Image: Karl Magnuson on Unsplash]
If you like what you have just read, support the Daily Friend