Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Theories Under Stress: The Evolution of Early Genetics

38 views

Published on

By analyzing a few thousand of the contributions of several biologists to the journal Nature (using the evoText project), I will map the discursive structure of the biometry-Mendelism debate, letting us understand who in the literature was in conversation with whom.

Published in: Education
  • Be the first to comment

  • Be the first to like this

Theories Under Stress: The Evolution of Early Genetics

  1. 1. THEORIES UNDER STRESS: THE EVOLUTION OF EARLY GENETICS Leeds HPS Centre Seminar, 24/5/2017 Charles H. Pence Institute for Ethical Leadership Department of Philosophy and Religious Studies
  2. 2. How can we understand the structure of scientific communities during theoretical crises?
  3. 3. THE BIOMETRY / MENDELISM DEBATE
  4. 4. color by Mads Madsden
  5. 5. Fleeming Jenkin, in 1884, from memoir by R.L. Stevenson
  6. 6. William Bateson, in 1905
  7. 7. W.F.R. Weldon
  8. 8. Karl Pearson and Francis Galton
  9. 9. From Weldon (1893)
  10. 10. Gregor Mendel
  11. 11. St. Cross Church, Holywell
  12. 12. The basic idea: (e.g., Provine 1971) • 1867: Jenkin’s review of the Origin • 1892: Bateson’s Materials • 1893: Weldon’s first biometrical work • 1901: Rediscovery of Mendel • 1906: Death of Weldon • ∼1930: Beginning of Synthesis
  13. 13. SOCIOLOGICAL STRUCTURE
  14. 14. Kyung-Man Kim, 1994
  15. 15. After Fig. 2, Kim 1994
  16. 16. New emphasis: paradigm articulators – those who “articulated the still inchoate paradigms by extending and elaborating the theory,” but without “evaluat[ing] their mentor’s theory” (Kim 1994, 35) Five of these – Darbishire, Schuster, Yule, Pearl, and Shull – converted from biometry to Mendelism between 1903 and 1910.
  17. 17. Kim’s focus: structures of education, training, and theory transmission
  18. 18. Good! But this is an active debate in the literature. Can we detect its signal there?
  19. 19. NETWORKS OF DISCOURSE
  20. 20. From previous work (Pence 2011, 2015) I knew some of this debate played out in Nature. Let’s find more. A network of around 100 biologists working on heredity published around 2,000 articles in Nature between, roughly, 1870 and 1940.
  21. 21. Aside: Check out the data! The network I will be describing can be interacted with live at: https://cpence.github.io/biometry-mendelism/ And all data is at: https://github.com/cpence/biometry-mendelism
  22. 22. (data: full network, animated network, time slices)
  23. 23. • —1884: No robust clustering, standard center-periphery network • 1885–94: Slight clustering of those involved in debate • 1895–99: Cluster of people involved in debate pulled out of broader conversation • 1900–04: Bateson and Weldon completely separate from remaining network • 1905–09: Last biometrical analysis, Pearson/Pearl working together; Weldon retreats to experimental work, dies • 1910—: Back to a cluster-free network
  24. 24. CONCLUSIONS
  25. 25. Community structure is reflected in the structure of the network of discourse
  26. 26. But! It’s not straightforward, and the networks of discourse give us interesting questions to ask about the community.
  27. 27. Paradigm “debaters?” Paradigm “warriors?” Participating in debates between paradigms pulls you out of the broader network.
  28. 28. Networks of discourse don’t sort paradigm A from paradigm B, nor do they give us Kim’s sociological structure.
  29. 29. After Fig. 2, Kim 1994, line weight proportional to edge weight in network of discourse, dashed line indicates connection present in Kim but missing in new network. Pearson-Pearl line reduced for clarity.
  30. 30. You get a variety of links across paradigms, and those connections can be difficult to describe in any other robust way.
  31. 31. Problems and Next Steps: • This is just one journal, broadly based in the UK. Can’t see Davenport’s school in the US very well. • Another siloing effect: biometricians found a new journal, Biometrika. Working on data access now. • Just one case study! Need more!
  32. 32. QUESTIONS? charles@charlespence.net http://charlespence.net @pencechp

×