February 2015 Lunchtime Abstracts & Details

::: Reproducibility of Findings in Experimental Philosophy: Problems and Possible Solutions
Hamid Seyedsayamdost, Visiting Fellow
Friday, February 6, 2015
12:05 pm
817R Cathedral of Learning

Abstract: The field of experimental philosophy has received great attention, essentially for producing results that seem highly counter-intuitive. These findings, so various commentators have argued, question some of the fundamental methods used in philosophy.
We present the findings of a systematic replication attempt for a selection of results from the experimental philosophy literature. A surprisingly high number of results are not reproducible, among them some of the most cited and attention-grabbing findings of the field. Since we made our findings public, other groups have independently tested our results and confirm their robustness.
Concurrent to the problems experimental philosophy has faced, a crisis of confidence has emerged in psychology. Experimental philosophy uses the methods of psychology to study philosophical questions and so experimental philosophy provides a case study for problems of reproducibility for a larger context. Experimental philosophy, a very young field in its early stages, quickly ran into the same problems that various fields of psychology have suffered from. As the evidence from the experimental philosophy movements indicates, these problems are unlikely to be related to localized issues such as research culture or a few ‘bad apples’. As it stands, current research and publication practices keep the door open for non-replicable results.
Some of the problems that lead to non-reproducibility of published findings are discussed and possible solutions outlined.

::: How Constructive Relativity May Succeed
Allan Walstad
University of Pittsburgh, Johnstown
Tuesday, February 10, 2015
12:05 pm
817R Cathedral of Learning

Abstract: Einstein’s special theory of relativity starts with assumptions about how observations conducted in relatively moving inertial frames must compare. From these assumptions, results can be derived regarding the laws of physics in any one frame. Can we do the reverse, that is, build the correct physics entirely from considerations within a single frame and infer therefrom the results of measurements in moving frames? Can we, in this project, supply the “elementary foundations” that Einstein found lacking in his own theory, thereby converting special relativity from (in Einstein’s terminology) a “principle” theory to a “constructive” one? I demonstrate that we can go quite far toward these goals via the introduction of elementary quantum concepts and experimental results that were unavailable to Einstein in 1905 (or unrecognized in significance). One possible benefit is to resolve a recent dispute in philosophy of physics regarding the Minkowski geometry of spacetime.


::: Mechanistic Explanation and Causal Structure
Robert Richardson
University of Cincinnati
Friday, February 13, 2015
12:05 pm
817R Cathedral of Learning

Abstract: Mechanistic explanation begins with some phenomena to be explained, some systemic patterns of behavior. A mechanistic model postulates a set of parts and processes with a characteristic organization. The parts are often characterized in terms of a set of capacities. If the model accurately depicts the causal structure of the system then the model explains the systemic behavior. Mechanistic explanations apparently presuppose some account of causal explanation to fill in the mechanistic account, offering no analysis of causation. Woodward reasonably suggests that mechanistic models should be “grounded out” in causal claims best analyzed using his manipulationist account, and a variety of mechanists have embraced that suggestion.
There is a fundamental problem confronting the suggestion. The interventionist interpretation identifies as causal relations those that allow for manipulability, if not in practice at least in principle, and imposes constraints on what counts as a successful manipulation. The key question concerns the causal structure in actual systems studied – e.g., within neuroscience or biology – and whether they exhibit the sort of causal organization that a manipulationist account can accommodate. In some cases they do; yet, this is not generally true. There is typically significant feedback from ‘higher’ neural centers even to ‘early’ visual processing ‘centers.’ Simple gene regulatory networks typically involve substantial feedback, as do even the simplest metabolic models. This means that manipulations of one variable will not be possible without affecting other variables that are causally relevant. There is of course a loose sense in which ‘manipulations’ are possible to probe systemic behaviors, but in some significant range of cases these are not manipulations in the strict sense required by interventionist analyses. Nonetheless, these seem to be mechanistic models, though ones outside the manipulationist fold.

::: Causation and Default-processes
Andreas Hüttemann
University of Cologne
Tuesday, February 17, 2015
12:05 pm, 817R Cathedral of Learning

Abstract: In this paper I will explicate one important sense of causation in terms of default-processes and deviations. I will argue that this approach can account for the role causation plays in some sciences. Furthermore I will address some objections raised by Schaffer, Hall and Paul against similar accounts.

::: Accentuate the Negative: A Puzzle about the Structure of the 'Origin' Solved
James Lennox
University of Pittsburgh
Friday, February 20, 2015
12:05 pm, 817R Cathedral of Learning

Abstract: The puzzle referred to in the sub-title of this talk is this: Darwin develops the case for his theory of species origins in the first four chapters of the Origin, but in his summary at the close of chapter four he famously says that reader will need to judge the theory based on “the tenour and balance of evidence given in the following chapters”. (Darwin 1859 127)
But what immediately follows is a series of chapters that, in Darwin’s own words, present “a crowd of difficulties” that are “so grave that to this day I can never reflect on them without being staggered” (Darwin 1859 171). If that is correct, then before presenting the positive evidence for his theory in chapters 10-13, he carefully considers a staggering number of difficulties the theory faces. On the face of it, that is a strange strategy for winning converts. In this paper I argue that, beginning in the Species Notebooks, Darwin worries about how to respond to “In Principle Impossible” (IPI) objections. By tracing the restructuring of his argument between 1842 and 1859, we are able to understand the philosophical motivations behind Darwin’s decision to “accentuate the negative” in the Origin.

::: The Argument from the Good Lot: Unconceived Alternatives and 19th Century Bacteriology
Raphael Scholl, Visiting Fellow
University of Bern
Tuesday, February 24, 2015
12:05 pm, 817R Cathedral of Learning

Abstract: Laudan’s pessimistic induction rests on a brute fact about the history of science: even highly successful scientific theories have regularly turned out to be mistaken. Recently, Stanford has argued that the “problem of unconceived alternatives” can explain this brute fact and thus show us why the pessimistic induction holds. On Stanford’s view, many scientific inferences are eliminative: scientists begin by corralling a number of candidate hypotheses that may explain a phenomenon of interest, proceed to rank these hypotheses according to some standard such as explanatory power, and then accept the most successful candidate. The problem is that even if ranking is reliable, eliminative inference only leads to truth if the true hypothesis is among the candidates being ranked. Stanford shows in a series of detailed historical case studies that past scientists regularly failed even to conceive of powerful candidate hypotheses which we have since come to accept. What is more, past scientists tended to overestimate their ability to exhaust the space of likely or plausible hypotheses.
However, Stanford’s problem of unconceived alternatives does not apply to an important and widespread type of scientific inference: causal inferences in laboratory sciences. Such inferences commonly proceed from an exhaustive hypothesis space defined by the contradictories “C is a cause of E” and “C is not a cause of E”. They thus leave no room for unconceived alternatives in Stanford’s sense, although they do allow for debates about causal co-factors, intermediate steps, and alternative causes. These points will be illustrated, and their philosophical usefulness will be tested, by a detailed look at the history of infectious disease research in the 19th century — in particular, at the debates among Koch, von Pettenkofer and Pasteur on diseases such as cholera and anthrax. The historical record suggests that a principled distinction may be possible between those inferences that are vulnerable to the problem of unconceived alternatives (and thus to the pessimistic induction) and those that are not.