Clark Glymour

Jaegwon Kim's Mind in a Physical World is an argument about mental causation that provides both a metaphysical theory and a lucid commentary on contemporary philosophical views. While I strongly recommend Kim's book to anyone interested in the subject, my endorsement is not unconditional, because I cannot make the same recomendation of the subject itself. Considering arguments of Davidson, Putnam, Burge, Block, and Kim himself, I conclude that the subject turns on a variety of implausible but received arguments, and that a useful study of mental causation cannot be divorced from scientific details of cognitive psychology, physics, and neuroscience

Various proposals have suggested that an adequate explanatory theory should reduce the number or the cardinality of the set of logically independent claims that need be accepted in order to entail a body of data. A (and perhaps the only) well-formed proposal of this kind is William Kneale’s: an explanatory theory should be finitely axiomatizable but it’s set of logical consequences in the data language should not be finitely axiomatizable. Craig and Vaught showed that Kneale theories (almost) always exist for any recursively enumerable but not finitely axiomatizable set of data sentences in a first order language with identity. Kneale’s criterion underdetermines explanation even given all possible data in the data language; gratuitous axioms may be “tacked on.” Define a Kneale theory, T, to be logically minimal if it is deducible from every Kneale theory (in the vocabulary of T) that entails the same statements in the data language as does T. If they exist, minimal Kneale theories are candidates for best explanations: they are “bold” in a sense close to Popper’s; some minimal Kneale theory is true if any Kneale theory is true; the minimal Kneale theory that is data equivalent to any given Kneale theory is unique; and no Kneale theory is more probable than some minimal Kneale theory. I show that under the Craig-Vaught conditions, no minimal Kneale theories exist

Rather than attempting to characterize a relation of confirmation between evidence and theory, epistemology might better consider which methods of forming conjectures from evidence, or of altering beliefs in the light of evidence, are most reliable for getting to the truth. A logical framework for such a study was constructed in the early 1960s by E. Mark Gold and Hilary Putnam. This essay describes some of the results that have been obtained in that framework and their significance for philosophy of science, artificial intelligence, and for normative epistemology when truth is relative.

The ultimate focus of the current essay is on methods of “creative abduction” that have some guarantees as reliable guides to the truth, and those that do not. Emphasizing work by Richard Englehart using data from the World Values Survey, Gerhard Schurz has analyzed literature surrounding Samuel Huntington’s well-known claims that civilization is divided into eight contending traditions, some of which resist “modernization” – democracy, civil rights, equality of rights of women and minorities, secularism. Schurz suggests an evolutionary model of modernization and identifies opposing social forces. In a later essay, citing Englehart’s work as an example, Schurz identifies factor analysis as an example of “creative abduction”. The theories of Englehart and his collaborators are reviewed again in the current essay. Published simulations and standard statistical desiderata for causal inference show the methods Englehart used, factor analysis in particular, are not guides to truth for the kind of data Schurz recognizes as common in political science. Recent work in statistics, philosophy and computer science that makes advances towards such methods is briefly reviewed.

A central theme in western philosophy was to find formal methods that can reliably discover empirical relationships and their explanations from data assembled from experience. As a philosophical project, that ambition was abandoned in the 20th century and generally dismissed as impossible. It was replaced in philosophy by neo-Kantian efforts at reconstruction and justification, and in professional statistics by the more limited ambition to estimate a small number of parameters in pre-specified hypotheses. The influx of “big data” from climate science, neuropsychology, biology, astronomy and elsewhere implicitly called for a revival of the grander philosophical ambition. Search algorithms are meeting that call, but they pose a problem: how are their accuracies to be assessed in domains where experimentation is limited or impossible? Increasingly, the answer is through simulation of data from models of the kind of process in the domain. In some cases, these innovations require rethinking how the accuracy and informativeness of inference methods can be assessed. Focusing on causal inference, we give an example from neuroscience, but to show that the model/simulation strategy is not confined to causal inference, we also consider two classification problems from astrophysics: identifying exoplanets and identifying dark matter concentrations.

Magnani, Lorenzo (2001), Abduction, Reason, and Science: Processes of Discovery and Explanation. New York: Kluwer Academic/ Plenum Publishers. Magnani. Lorenzo, and Nancy Nersessian (eds.) (2002), Model-Based Reasoning: Technology, Science, Values. New York: Kluwer Academic/ Plenum Publishers. Joseph E. Earley, Sr. (ed.), Chemical Explanation: Characteristics, Development, Autonomy, Annals of the New York Academy of Sciences, vol. 988. New York Academy of Sciences (2003), 370 pp., $130.00 (cloth).

In "The Epistemology of Geometry" Glymour proposed a necessary structural condition for the synonymy of two space-time theories. David Zaret has recently challenged this proposal, by arguing that Newtonian gravitational theory with a flat, non-dynamic connection (FNGT) is intuitively synonymous with versions of the theory using a curved dynamical connection (CNGT), even though these two theories fail to satisfy Glymour's proposed necessary condition for synonymy. Zaret allowed that if FNGT and CNGT were not equally well (bootstrap) tested by the relevant phenomena, the two theories would in fact not be synonymous. He argued, however, that when electrodynamic phenomena are considered, the two theories are equally well tested. We show that it is not FNGT and CNGT which are equally well tested when the electrodynamic phenomena are considered, but only suitable extensions of FNGT and CNGT. Thus, there is good reason to consider FNGT and CNGT to be non-synonymous. We further show that the two extensions of FNGT and CNGT which are equally well tested when electrodynamic phenomena are considered (and which could be considered intuitively synonymous) not only satisfy Glymour's original proposed necessary condition for the synonymy of space-time theories, they satisfy a plausible stronger condition as well

Convergent realists desire scientific methods that converge reliably to informative, true theories over a wide range of theoretical possibilities. Much attention has been paid to the problem of induction from quantifier-free data. In this paper, we employ the techniques of formal learning theory and model theory to explore the reliable inference of theories from data containing alternating quantifiers. We obtain a hierarchy of inductive problems depending on the quantifier prefix complexity of the formulas that constitute the data, and we provide bounds relating the quantifier prefix complexity of the data to the quantifier prefix complexity of the theories that can be reliably inferred from such data without background knowledge. We also examine the question whether there are theories with mixed quantifiers that can be reliably inferred with closed, universal formulas in the data, but not without

A general principle for good pedagogic strategy is this: other things equal, make the essential principles of the subject explicit rather than tacit. We think that this principle is routinely violated in conventional instruction in statistics. Even though most of the early history of probability theory has been driven by causal considerations, the terms “cause” and “causation” have practically disappeared from statistics textbooks. Statistics curricula guide students away from the concept of causality, into remembering perhaps the cliche disclaimer “correlation does not mean causation,” but rarely thinking about what correlation does mean. The treatment of causality is a serious handicap to later studies of such topics as experimental design, where often the main goal is to establish (or disprove) causation. Much of the technical vocabulary of the language used in research design textbooks consists in euphemisms for specific causal relations, e.g, “latent variable,” “intervening variable,” “confounding factor,” etc. The multiplicity of terms used to refer to causation results in confusion and, in effect, may hinder understanding of the basic principles of research design.

The ability to identify the mineral composition of rocks and soils is an important tool for the exploration of geological sites. Even though expert knowledge is commonly used for this task, it is desirable to create automated systems with similar or better performance. For instance, NASA intends to design robots that are sufficiently autonomous to perform this task on planetary missions. Spectrometer readings provide one important source of data for identifying sites with minerals of interest. Reflectance spectrometers measure intensities of light reflected from surfaces over a range of wavelengths. Spectral intensity patterns may in some cases be sufficiently distinctive for proper identification of minerals or classes of minerals. For some mineral classes, carbonates for example, specific short spectral intervals are known to carry a distinctive signature. Finding similar distinctive spectral ranges for other mineral classes is not an easy problem. We propose and evaluate data-driven techniques in two stages: first, evaluating algorithms to identify which components are probably present in a given rock; second, trying to improve this classification by automatically searching for spectral ranges optimized for specific classes of minerals. In one set of studies, we partition the whole interval of wavelengths available in our data into sub-intervals, or bins, and use a genetic algorithm to evaluate a candidate selection of subintervals. As an alternative to these computationally expensive search techniques, we present an entropy-based heuristic that gives higher scores for wavelengths more likely to distinguish between classes. Results are presented for four different classes, showing reasonable improvements in identifying some, but not all, of the mineral classes tested.

By combining experimental interventions with search procedures for graphical causal models we show that under familiar assumptions, with perfect data, N - 1 experiments suffice to determine the causal relations among N > 2 variables when each experiment randomizes at most one variable. We show the same bound holds for adaptive learners, but does not hold for N > 4 when each experiment can simultaneously randomize more than one variable. This bound provides a type of ideal for the measure of success of heuristic approached in active learning methods of casual discovery, which currently use less informative measures

Words, Thoughts and Theories argues that infants and children discover the physical and psychological features of the world by a process akin to scientific inquiry, more or less as conceived by philosophers of science in the 1960s (the theory theory). This essay discusses some of the philosophical background to an alternative, more popular, “modular” or “maturational” account of development, dismisses an array of philosophical objections to the theory theory, suggests that the theory theory offers an undeveloped project for artificial intelligence, and, relying on recent psychological work on causation, offers suggestions about how principles of causal inference may provide a developmental solution to the “frame problem”.

Using Gebharter’s representation, we consider aspects of the problem of discovering the structure of unmeasured submechanisms when the variables in those submechanisms have not been measured. Exploiting an early insight of Sober’s, we provide a correct algorithm for identifying latent, endogenous structure—submechanisms—for a restricted class of structures. The algorithm can be merged with other methods for discovering causal relations among unmeasured variables, and feedback relations between measured variables and unobserved causes can sometimes be learned

After reviewing theoretical reasons for doubting that machine learning methods can accurately infer gene regulatory networks from microarray data, we test 10 algorithms on simulated data from the sea urchin network, and on microarray data for yeast compared with recent experimental determinations of the regulatory network in the same yeast species. Our results agree with the theoretical arguments: most algorithms are at chance for determining the existence of a regulatory connection between gene pairs, and the algorithms that perform better than chance are nonetheless so errorprone as to be of little practical use in these applications.