Explanatory Loops and the Limits of Genetic Reductionism

Abstract

We reconstruct genetic determinism as a reductionist thesis to the effect that the molecular properties of cells can be accounted for to a great extent by their genetic outfit. The non‐reductionist arguments offered at this molecular level often use the relationship between structure and function as their point of departure. By contrast, we develop a non‐reductionist argument that is confined to the structural characteristics of biomolecules; no appeal to functions is made. We raise two kinds of objections against the reducibility claim underlying genetic determinism. First, some conceptual distinctions at the protein level cannot be captured on a genetic basis. A one‐to‐many relationship between DNA sequences and proteins emerges from them. Second, the relationship between genes and proteins is characterized by explanatory loops or reciprocal explanatory dependence. The presence of proteins is explained by the transcription from corresponding DNA sequences, and the latter is in turn accounted for by the action of proteins. By contrast, a reductive account requires a unidirectional explanatory dependence.

Acknowledgements

We are grateful for the valuable reviews of two anonymous referees for this journal. Their demands for clarification and their suggestions have done a lot to improve the argumentative line of the paper.

Notes

[1] Male humans and male chimpanzees share 98.8% of their genome, while male humans differ in 3% of the genome from their female conspecifics—as a result of their lacking a Y‐chromosome.

[2] For the separation between determinism and predictability, see Earman (1986, 7–9) and Breuer (1997, 10–14).

[3] For this conception of phenotype, see Sarkar (1998, 7–8).

[4] More recent conceptions of reduction which emphasize this demand are Kim’s approach to reduction via functionalization (Kim 1997, 284–86) and the so‐called ‘New Wave Reductionism’ (Van Eck, De Jong, and Schouten 2006, 169–70).

[5] In terms of the schematic example given before, this situation exhibits the following structure. Let the law of the allegedly reducing theory be α → β, and the laws to be reduced A₁ → B₁ and A₂ → B₂. The conceptual links are, accordingly, α = A₁ ∨ A₂; β = B₁ ∨ B₂. Under such circumstances, the derivation fails. The only relation one obtains is A₁ ∨ A₂ → B₁ ∨ B₂ which is logically weaker than the two laws to be deduced (Ruse 1976, 448).

[6] One of Kitcher’s arguments in support of anti‐reductionism briefly appeals to the occurrence of explanatory loops (Kitcher 1984, 29–30).

[7] Alternative splicing and editing are mentioned by Sarkar (1998, 156–57) as an anti‐reductionist complication of the molecular definition of genes in the functional sense. In the following sections we assess the impact of these features on genetic reductionism in general.

[8] In fact, the context dependence of genetic effects has led a number of authors to reject the metaphor of a central program altogether (Fox Keller 2000, 87, 94, 99; Burian 2004, 63).

Additional information

Notes on contributors

Martin Carrier

Martin Carrier is a Professor in the Institute for Science and Technology Studies, Bielefeld University, Germany.

Patrick Finzer

Patrick Finzer is Privatdozent for microbiology at the Univeristy of Heidelberg, Germany and Section Head Microbiology, bioscientia‐Institute for Medical Diagnostics, Moers, Germany.