Typological Thinking
Typological thinking is the mode of biological reasoning that treats species as types — ideal forms or essences — and individual organisms as imperfect copies of those types. Variation between individuals is noise. The real thing is the type, and the purpose of investigation is to identify the type by seeing through the noise of individual difference. Mayr traced the habit to Plato's theory of forms and argued that it represented the single largest conceptual obstacle to understanding evolution. Darwin's revolution required replacing typological thinking with population thinking — the recognition that populations are real and that variation within populations is the fundamental biological phenomenon.
In the AI Story
Typological thinking is not merely a historical error. It persists, in Mayr's analysis, as the default mode of much non-biological reasoning — and reappears in disguised forms whenever investigators seek ideal types that explain particular cases. In the AI discourse, the tendency is visible in the construction of representative figures — the triumphalist builder, the displaced expert, the ambivalent middle — each treated as an ideal type that organizes analysis and invites identification.
The machine learning systems themselves are structurally typological. A classifier assigns inputs to categories. A language model predicts the next token based on statistical regularities — the average behavior of tokens in similar contexts across the training corpus. The typological approximation is computationally useful; it is also a systematic erasure of the variation that distinguishes one particular instance from the statistical center.
Mayr's warning — that the type is an abstraction and the abstraction conceals the variation that matters most — applies to both the cultural discourse about AI and the technical operation of AI systems themselves. At both levels, the typological habit smooths away the specific in favor of the general, the individual in favor of the average, the irreducible uniqueness of each case in favor of the categorical sameness that is easier to think about and cheaper to compute.
Origin
Mayr's critique of typological thinking developed through the 1940s and 1950s and was articulated most fully in his 1959 address and subsequent writings. The framing became standard in the philosophy of biology, though the specific label typological thinking was Mayr's contribution.
Key Ideas
Types as Platonic essences. Treating categories as more real than the individuals within them is ancient philosophical habit, not empirical finding.
Variation as noise versus signal. The typologist treats variation as error around a true type; the populationist treats variation as the fundamental biological reality.
Discourse produces types. Public conversation about AI constructs ideal figures (builder, elegist, middle) and organizes analysis around them, losing the distributional reality in the process.
ML is typological by design. Classification and prediction systems necessarily assign individuals to types; their utility depends on the approximation being good enough for the purpose.
The cost of typology is the erasure of the particular. Whatever matters most to the individual is precisely what the typological abstraction has removed.
Appears in the Orange Pill Cycle
Further reading
- Ernst Mayr, Darwin and the Evolutionary Theory in Biology (1959)
- Ernst Mayr, The Growth of Biological Thought (Harvard University Press, 1982)
- Elliott Sober, Evolution, Population Thinking, and Essentialism (Philosophy of Science, 1980)
- Michael Ruse, Darwinism Defended (Addison-Wesley, 1982)