Republican strategist Frank Luntz famously advised party leaders to emphasize the lack of ';scientific certainty' about global warming because, once the public became convinced of a scientific consensus on global warming, they would accept it and policies responding to it. But philosophical work on scientific methodology puts absolute certainty out of reach. Today almost all philosophers are fallibilists, holding that every belief is subject to correction. Does it follow that an honest report of what science has to say about any topic cannot claim that the matter is settled? I defend a more confident view: much of what science has to say about the world is settled. On this view, settled science doesn't require philosophical certainty. Instead, it is based on stable agreement and practical reliability--values central to C.S. Peirce's account of the scientific method. Defending this position requires care over changes in theory that can undermine the language in which these settled points were initially expressed. It also requires epistemic modesty, since our confidence in any claim, including observational claims, depends on a kind of induction, and could, in principle, be undermined. Interestingly, this line of thought also suggests many of the central claims the historical sciences make about our world are more stable and reliable than the more theory-dependent claims we think of as central to the foundations of physics. David Hume undermined philosophical certainty about causal laws when he argued that, no matter how many times we make observations confirming a law, it could still fail the next test. This point is sometimes confused with healthy scientific skepticism, but Hume's worry has a much broader scope than scientific skepticism. For Hume each instance of a causal relation was doubtful, no matter how similar to previous instances. But scientists are diffident about applying a well-tested law only when the circumstances differ. For instance, particle physicists regard the standard model as extremely reliable for a wide range of energies, though it is expected to fail at higher energies. Peirce's emphasis on stable agreement and practical reliability has two important implications for understanding settled science: first, it exposes our confidence in observations to Hume's worries about induction, since our confidence in observation is, on his account, grounded in regular, reliable agreement between observers. But well-established observational results are paradigmatic examples of settled science! In turn, this points toward a more general account of settled science. The successes of general relativity, quantum mechanics and the standard model in particle physics are inconclusive for the foundations of physics, which aim (ambitiously) at a unified theory of all the forces of nature. But they remain settled successes for these theories. Similarly, biology and geology have achieved stable consensus on many points, including the importance of processes such as natural selection and plate tectonics, and climate science has achieved stable consensus on many conclusions as well (Cook et al., 2013), from the processes responsible for the ';greenhouse' effect and the role of CO2 emissions in amplifying that effect, to a range of estimates for climate sensitivity, based on multiple lines of evidence, that makes the danger of continued emissions plain.
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- 1799 HISTORY OF GEOPHYSICS General or miscellaneous;
- 1704 HISTORY OF GEOPHYSICS Atmospheric sciences;
- 0485 BIOGEOSCIENCES Science policy