Category: science-and-biology
Associated thinker(s): Stephen Jay Gould; Richard C. Lewontin
Original publication or approximate period: “The spandrels of San Marco and the Panglossian paradigm” (Gould and Lewontin, 1979); Gould’s clarification (1997)
Difficulty: Intermediate
Type: argument
Themes: adaptation; byproducts; historical explanation
Related experiments: darwins-eye; darwins-aquatic-bear
Gould and Lewontin use the mosaics in San Marco’s dome to illustrate how a present use can follow from a structure’s origin. They argue that mounting a dome on four rounded arches produces four tapering, roughly triangular spaces; the mosaics fit those spaces, but that fit does not show the spaces were created for the mosaics. Gould later clarified that the byproduct claim concerns the spaces’ number and shape, not an absence of structural utility. The use of “spandrel” for these three-dimensional pendentives has been debated.
Their biological analogy asks whether a feature’s present usefulness establishes that selection originally produced it for that use. This is an analogy and methodological argument, not a strict hypothetical experiment or a demonstration about the ancestry of any particular organism.
Does a feature’s current contribution establish why that feature originally arose?
The example encourages rival historical explanations. A structure may be selected for its present function, inherited from a prior function, or produced as a correlated consequence of development or other changes. Distinguishing those possibilities turns an appealing adaptive story into a research question rather than an automatic answer.
Gould and Lewontin criticize treating a feature’s present utility as proof that it originated through selection for that use. They do not reject natural selection: the 1979 article describes it as one important process among others and argues for a pluralistic account that also considers developmental constraints and alternative histories. Those alternatives must be assessed with biological evidence; the architectural analogy alone cannot establish how common by-products are.
Gould and Lewontin were not claiming that natural selection never operates. A feature can acquire useful functions after arising as a byproduct. The architectural terminology has itself been debated; Gould’s later clarification emphasizes the necessary number and shape of the relevant spaces. The analogy should not replace biological evidence with a verbal label.
Gould and Vrba introduced the term exaptation in 1982 for a useful character co-opted for its current role after evolving for another use or for no function. A feature may first arise as a by-product and later acquire a new use; “adaptation” and “exaptation” describe different kinds of evolutionary history. Darwin’s eye argument addresses possible adaptive sequences, while the aquatic-bear scenario highlights the gap between a possible sequence and an established history.
The full 1979 paper was read in a publicly hosted scan; its Royal Society record supplies the publication details and abstract. Gould’s 1997 paper clarifies the intended architectural analogy and responds to terminology and structural-utility objections. Gould and Vrba’s 1982 paper supports the later term “exaptation.” This entry is an analogy and methodological argument, not a strict hypothetical experiment.
Verified on 2026-11-01. Relevant passages in the full 1979 paper, Gould’s 1997 clarification, and Gould and Vrba’s 1982 definition were read. The public 1979 scan is a third-party reproduction; the Royal Society’s full text is paywalled and its online text is OCR-derived. Reflection questions and illustrative examples are original editorial material.