Separate Worlds? A Network Analysis of the Boundaries Between Philosophy of Science and the Sciences
Separate Worlds? A Network Analysis of the Boundaries Between Philosophy of Science and the Sciences
0 0
sunbelt_XXXVI_newport_beach
On Github mclevey / sunbelt_XXXVI_newport_beach
Separate Worlds? A Network Analysis of the Boundaries Between Philosophy of Science and the Sciences
John McLevey (University of Waterloo), Alexander Graham (University of Waterloo), Reid McIlroy-Young (University of Chicago), Kathryn S. Plaisance (University of Waterloo)Sunbelt XXXVI, Newport Beach, California
Context
Two fundamentally different perspectives on knowledge diffusion dominate debates about the disciplines.
- silos: specialized, narrow, willfully ignorant
- webs: dynamic, importing, exporting
Starting Assumption: blanket statements about how open or closed the disciplines are are unproductive.
Efforts to produce global "maps of science" are evidence of both closure (in the form of dense disciplinary clusters) and thick webs of connections across disciplinary boundaries. Disciplines are connected to varying degrees.
Our research design differs from other diffusion studies by emphasizing changes in disciplinary boundaries and reputations.
The Case
Philosophy of Science and the Sciences
Philosophers of science disagree about whether and how they should influence the sciences they study.
- the best philosophy of science is embedded and responsive
- the best philosophy of science has critical distance
- philosophy does not need to impact science to have value
- ...
Hypotheses
Positive relationship hypothesis: Net of other factors, publishing in and being cited in science journals increases the odds of being cited by other philosophers. Negative relationship hypothesis: Net of other factors, publishing in and being cited in science journals reduces the odds of being cited by other philosophers of science. Different boundaries with different sciences: Net of other factors, the positive or negative effects of publishing in any being cited in science journals varies based on the type of science a philosopher studies. Separate worlds hypothesis: Net of other factors, publishing in and being cited in science journals does not change the odds of being cited by other philosophers.Data
we made a list of philosophers of science based on publications in philosophy of science journals and dissertations we identified and collected all of their articles available in the Web of Science (n = 27,734) we identified and collected all articles citing philosophers of science available in the Web of Science (n = 140,918) we constructed a philosophy of science citation network (24,901 nodes and 40,915 edges) we counted and classified citations from non-philosophy articles to philosophy articles, creating 12 new covariatesPython package metaknowledge
http://networkslab.org/metaknowledge/https://github.com/networks-lab/metaknowledge
Co-developed by Reid McIlroy-Young and John McLevey.
ERGM
Receiving citations in the philosophy of science citation network ~
- edges
- two paths (gwdsp, $\alpha$ = .7)
- clustering (gwesp, $\alpha$ = .7)
- popularity spread (indegree)
- activity spread (outdegree)
- years since publication
- published in top-ranked journal [Y/N]
- published outside discipline [Y/N]
- number of authors
- combined author core publication score
- citation counts from 12 discipline clusters
Results
Base Model
Full Model
Discussion & Future Research
What looks like support for the different boundaries hypothesis is actually support for the separate worlds hypothesis, because the effect sizes are very small.
Willfully ignorant?
Variations in connection: proximate and distant fields
Implications for philosophers: dual careers
Thank you.
John McLevey john.mclevey@uwaterloo.ca
Alexander (Sasha) Graham a24graha@uwaterloo.ca