Title:
The productive polysemy of scientific language
Abstract:
Scientists, philosophers and historians frequently observe—and alternately lament or praise—that many scientific terms lack a general definitions. Instead, terms like “gene”, “attention” or “force” are polysemous: they have multiple related meanings none of which provides general application conditions for all contexts in which researchers use these terms (Cartwright 1999, Wilson 2006, Rheinberger and Müller-Wille 2018, Anderson 2023, Taylor 2023).
In this talk I defend the communicative and epistemic value of polysemous terms which lack general definitions. My argument combines cognitive linguistics with a patchwork approach to scientific concepts. Information-theoretic arguments suggest that mapping multiple meanings to the same word form decreases uncertainty in communication systems where speakers can decode the relevant meaning from context (Piantadosi et al. 2012). Empirical corpus studies additionally show that word forms that are short and easy to process are more polysemous than longer and more difficult ones (Brochhagen and Boleda 2022).
Dovetailing this research are patchwork analyses of “gene”, “force” and “attention” . They provide constraints under which polysemous terms are not only efficient for communication, but also epistemically productive. Researchers broaden existing local uses or “patches” as long as the techniques they involve remain reliable, the domains to which they are applied are homogeneous, and the properties they target are significant to reach epistemic goals (Haueis 2024). For example: physicists broadened Newton’s use of “force” from collisions on earth to celestial mechanics; classical geneticists broadened Mendel’s use of “gene” by extending recombination techniques from animals to plants. By contrast, continuum mechanics creates a novel “force” patch where Newtonian techniques proved unreliable (e.g. plastic deformation). Molecular biologists created a new patch of “gene” when these techniques proved unreliable to target related phenomena like protein synthesis (Müller-Wille and Rheinberger 2018, chs. 5 and 6).
The upshot of the argument is twofold. The cognitive linguistic view suggests the
empirical hypothesis that that scientific terms which are easier and shorter (e.g. “gene”, “force”) are more polysemous than longer and more difficult ones (e.g. “endoplasmatic reticulum”). The patchwork analysis suggests that scientists can advance their understanding (e.g. of collision or inheritance phenomena) by stitching multiple patches of together without any of the patches providing general definition of the concept in all contexts.
References:
- Anderson, B. (2023). Stop paying attention to “attention.” WIREs Cognitive Science, 14(1), e1574. https://doi.org/10.1002/wcs.1574
- Brochhagen, T., & Boleda, G. (2022). When do languages use the same word for different meanings? The Goldilocks principle in colexification. Cognition, 226, 105179. https://doi.org/10.1016/j.cognition.2022.105179
- Cartwright, N. (1999). The Dappled World. Cambridge University Press.
- Haueis, P. (2024). A generalized patchwork approach to scientific concepts. BJPS 75(3), 741–768. https://doi.org/10.1086/716179
- Piantadosi, S. T., Tily, H., & Gibson, E. (2012). The communicative function of ambiguity in language. Cognition, 122(3), 280–291. https://doi.org/10.1016/j.cognition.2011.10.004
- Taylor, H. (2023). Attention as a patchwork concept. EJPS, 13(3), 36. https://doi.org/10.1007/s13194-023-00538-5
- Rheinberger, H.-J., Müller-Wille, S. (2018). The gene: From genetics to postgenomics. The University of Chicago Press.
- Wilson, M. (2006). Wandering significance: An essay on conceptual behavior. Clarendon Press.