Charles Stein discovered a paradox in 1955 that many statisticians think is of fundamental importance. Here we explore its philosophical implications. We outline the nature of Stein’s result and of subsequent work on shrinkage estimators; then we describe how these results are related to Bayesianism and to model selection criteria like AIC. We also discuss their bearing on scientific realism and instrumentalism. We argue that results concerning shrinkage estimators underwrite a surprising form of holistic pragmatism.

In these prefatory remarks, which are designed to locate my topic within the complex and wide-stretching field of Popper's thought and writings, I shall not say anything that those familiar with his work will not already know. Moreover, what I do say will take as understood many of the problems and theories, not to mention the terminology, that I shall later be doing my best to make understandable. My apologies are therefore due equally to those who know something about Popper's (...) discussions of indeterminism and of the propensity interpretation of probability, and to those who know nothing. (shrink)

Here, I consider the recent application of phylogenetic methods in historical linguistics. After a preliminary survey of one such method, i.e. cladistic parsimony, I respond to two common criticisms of cultural phylogenies: that cultural artifacts cannot be modeled as tree-like because of borrowing across lineages, and that the mechanism of cultural change differs radically from that of biological evolution. I argue that while perhaps remains true for certain cultural artifacts, the nature of language may be such as to side-step this (...) objection. Moreover, I explore the possibility that cladistic parsimony can be justified even if is true by appealing to the inference pattern known among philosophers as ‘Inference to the Best Explanation’. (shrink)

The Akaike Information Criterion can be a valuable tool of scientific inference. This statistic, or any other statistical method for that matter, cannot, however, be the whole of scientific methodology. In this paper some of the limitations of Akaikean statistical methods are discussed. It is argued that the full import of empirical evidence is realized only by adopting a richer ideal of empirical success than predictive accuracy, and that the ability of a theory to turn phenomena into accurate, agreeing measurements (...) of causally relevant parameters contributes to the evidential support of the theory. This is illustrated by Newton's argument from orbital phenomena to the inverse-square law of gravitation. (shrink)

Akaike's framework for thinking about model selection in terms of the goal of predictive accuracy and his criterion for model selection have important philosophical implications. Scientists often test models whose truth values they already know, and they often decline to reject models that they know full well are false. Instrumentalism helps explain this pervasive feature of scientific practice, and Akaike's framework helps provide instrumentalism with the epistemology it needs. Akaike's criterion for model selection also throws light on the role of (...) parsimony considerations in hypothesis evaluation. I explain the basic ideas behind Akaike's framework and criterion; several biological examples, including the use of maximum likelihood methods in phylogenetic inference, are considered. (shrink)

In this paper, I consider the relationship between Inference to the Best Explanation and Bayesianism, both of which are well-known accounts of the nature of scientific inference. In Sect. 2, I give a brief overview of Bayesianism and IBE. In Sect. 3, I argue that IBE in its most prominently defended forms is difficult to reconcile with Bayesianism because not all of the items that feature on popular lists of “explanatory virtues”—by means of which IBE ranks competing explanations—have confirmational import. (...) Rather, some of the items that feature on these lists are “informational virtues”—properties that do not make a hypothesis \ more probable than some competitor \ given evidence E, but that, roughly-speaking, give that hypothesis greater informative content. In Sect. 4, I consider as a response to my argument a recent version of compatibilism which argues that IBE can provide further normative constraints on the objectively correct probability function. I argue that this response does not succeed, owing to the difficulty of defending with any generality such further normative constraints. Lastly, in Sect. 5, I propose that IBE should be regarded, not as a theory of scientific inference, but rather as a theory of when we ought to “accept” H, where the acceptability of H is fixed by the goals of science and concerns whether H is worthy of commitment as research program. In this way, IBE and Bayesianism, as I will show, can be made compatible, and thus the Bayesian and the proponent of IBE can be friends. (shrink)

The pessimistic induction plays an important role in the contemporary realism/anti-realism debate in philosophy of science. But there is some disagreement about the structure and aim of the argument. And a number of scholars have noted that there is more than one type of PI in the philosophical literature. I review four different versions of the PI. I aim to show that PIs have been appealed to by philosophers of science for a variety of reasons. Even some realists have appealed (...) to a PI. My goal is to advance our understanding of what the various PIs can teach us about science and the threat posed by PIs to scientific realism. (shrink)

In this essay, I aim to motivate and defend a pluralistic view of epistemic rationality. At the core of the view is the notion that epistemic rationality is essentially a species of practical rationality put in the service of various epistemic goals. I begin by sketching some closely related views that have appeared in the literature. I then present my preferred version of the view and sketch some of its benefits. Thomas Kelly has raised challenging objections to a part of (...) the view, which I then present and address. I then raise some distinct worries that might linger and sketch some possible ways one might address them. (shrink)

In this paper, I argue that, other things being equal, simpler arguments are better. In other words, I argue that, other things being equal, it is rational to prefer simpler arguments over less simple ones. I sketch three arguments in support of this claim: an argument from mathematical proofs, an argument from scientific theories, and an argument from the conjunction rule.

The twin goals of this essay are: to investigate a family of cases in which the goal of guaranteed convergence to the truth is beyond our reach; and to argue that each of three strands prominent in contemporary epistemological thought has undesirable consequences when confronted with the existence of such problems. Approaches that follow Reichenbach in taking guaranteed convergence to the truth to be the characteristic virtue of good methods face a vicious closure problem. Approaches on which there is a (...) unique rational doxastic response to any given body of evidence can avoid incoherence only by rendering epistemology a curiously limited enterprise. Bayesian approaches rule out humility about one’s prospects of success in certain situations in which failure is typical. (shrink)

It is often claimed that the greatest value of the Bayesian framework in cognitive science consists in its unifying power. Several Bayesian cognitive scientists assume that unification is obviously linked to explanatory power. But this link is not obvious, as unification in science is a heterogeneous notion, which may have little to do with explanation. While a crucial feature of most adequate explanations in cognitive science is that they reveal aspects of the causal mechanism that produces the phenomenon to be (...) explained, the kind of unification afforded by the Bayesian framework to cognitive science does not necessarily reveal aspects of a mechanism. Bayesian unification, nonetheless, can place fruitful constraints on causal–mechanical explanation. 1 Introduction2 What a Great Many Phenomena Bayesian Decision Theory Can Model3 The Case of Information Integration4 How Do Bayesian Models Unify?5 Bayesian Unification: What Constraints Are There on Mechanistic Explanation?5.1 Unification constrains mechanism discovery5.2 Unification constrains the identification of relevant mechanistic factors5.3 Unification constrains confirmation of competitive mechanistic models6 ConclusionAppendix. (shrink)

The functionalist approach to kinds has suffered recently due to its association with law-based approaches to induction and explanation. Philosophers of science increasingly view nomological approaches as inappropriate for the special sciences like psychology and biology, which has led to a surge of interest in approaches to natural kinds that are more obviously compatible with mechanistic and model-based methods, especially homeostatic property cluster theory. But can the functionalist approach to kinds be weaned off its dependency on laws? Dan Weiskopf has (...) recently offered a reboot of the functionalist program by replacing its nomological commitments with a model-based approach more closely derived from practice in psychology. Roughly, Weiskopf holds that the natural kinds of psychology will be the functional properties that feature in many empirically successful cognitive models, and that those properties need not be localized to parts of an underlying mechanism. I here skeptically examine the three modeling practices that Weiskopf thinks introduce such non-localizable properties: fictionalization, reification, and functional abstraction. In each case, I argue that recognizing functional properties introduced by these practices as autonomous kinds comes at clear cost to those explanations’ counterfactual explanatory power. At each step, a tempting functionalist response is parochialism: to hold that the false or omitted counterfactuals fall outside the modeler’s explanatory aims, and so should not be counted against functional kinds. I conclude by noting the dangers this attitude poses to scientific disagreement, inviting functionalists to better articulate how the individuation conditions for functional kinds might outstrip the perspective of a single modeler. (shrink)

This paper discusses the role that appeals to theoretical simplicity have played in the debate between nativists and empiricists in cognitive science. Both sides have been keen to make use of such appeals in defence of their respective positions about the structure and ontogeny of the human mind. Focusing on the standard simplicity argument employed by empiricist-minded philosophers and cognitive scientists—what I call “the argument for minimal innateness”—I identify various problems with such arguments—in particular, the apparent arbitrariness of the relevant (...) notions of simplicity at work. I then argue that simplicity ought not be seen as a theoretical desideratum in its own right, but rather as a stand-in for other desirable features of theories. In this deflationary vein, I argue that the best way of interpreting the argument for minimal innateness is to view it as an indirect appeal to various potential biological constraints on the amount of innate structure that can wired into the human mind. I then consider how nativists may respond to this biologized version of the argument, and discuss the role that similar biological concerns have played in recent nativist theorizing in the Minimalist Programme in generative linguistics. (shrink)

There is considerable disagreement about the epistemic value of novel predictive success, i.e. when a scientist predicts an unexpected phenomenon, experiments are conducted, and the prediction proves to be accurate. We survey the field on this question, noting both fully articulated views such as weak and strong predictivism, and more nascent views, such as pluralist reasons for the instrumental value of prediction. By examining the various reasons offered for the value of prediction across a range of inferential contexts , we (...) can see that neither weak nor strong predictivism captures all of the reasons for valuing prediction available. A third path is presented, Pluralist Instrumental Predictivism; PIP for short. (shrink)

Over the past 50 years, there has been a great deal of philosophical interest in laws of nature, perhaps because of the essential role that laws play in the formulation of, and proposed solutions to, a number of perennial philosophical problems. For example, many have thought that a satisfactory account of laws could be used to resolve thorny issues concerning explanation, causation, free-will, probability, and counterfactual truth. Moreover, interest in laws of nature is not constrained to metaphysics or philosophy of (...) science; claims about laws play essential roles in areas as diverse as the philosophy of religion (e.g., in the argument from design) and the philosophy of mind (e.g., in the formulation of Davidson’s anomalous monism). In my dissertation, I consider and reject the widely-held thesis that the facts concerning laws can be reduced to the facts concerning the particular entities that the laws “govern,” and that the laws thus have no independent existence. I instead defend a version of nomic primitivism, according to which the facts about laws cannot be reduced to facts that are themselves non-nomic – i.e., to facts that do not fundamentally involve laws, counterfactuals, causes, etc. Insofar as the truth or falsity of reductionism about laws has implications for many of the problems mentioned above, I think that this result should be of interest even to those who who do not work in metaphysics or the philosophy of science. My methodology, which I lay out and defend in Chapter One, is a version of Carnapian explication. This method emphasizes the importance of articulating and maintaining clear distinctions between (1) the vague concept (or concepts) law of nature inherent in ordinary language and scientific practice and (2) the precise analyses of “law of nature” that philosophers have proposed as potential replacements for this concept. I argue that metaphysics-as-explication has clear advantages over rival conceptions of metaphysical methodology; in particular, it allows us to formulate evaluative criteria for metaphysical claims. In Chapter Two, I offer an example of how careful attention to concepts already in use can help resolve philosophical debate. Specifically, I argue that much recent literature has mistakenly assumed that there is only one concept of “law of nature” in use, while there are in fact at least two. Strong laws are the principles pursued by fundamental physics: they are true, objective, and bear distinctive relationships to counterfactuals and explanation. Weak laws, by contrast, lack at least one of these distinctive characteristics but play central roles in both the “special sciences” and in everyday life. In Chapters Three and Four, I offer extended arguments against the two most prominent versions of reductionism about laws – Humeanism and law necessitarianism. According to philosophical Humeans, the laws of nature supervene upon the non-modal, non-nomic facts concerning the behavior of particular things at particular times and places. Law necessitarians, by contrast, argue that the laws are in fact metaphysically necessary, and that which laws there are is determined by a class of primitive, modally loaded facts concerning the essences, natures, or dispositions. I argue that both of these views are mistaken insofar as they disagree with well-entrenched scientific practices and those in favor of reductionism have failed to provided sufficient reason for thinking that these practices should be revised. Much of my argument is focused on the role played by a number of supposed methodological principles, including appeals to intuition, parsimony, and methodological naturalism. While the conclusions of this dissertation are explicitly constrained to laws, many of the arguments should be of interest to those who are concerned about philosophical methodology (especially in the role of intuition in philosophical argument) or the appropriate relation between metaphysics, science, and the philosophy of science. (shrink)

Theory choice can be approached in at least four ways. One of these calls for the application of decision theory, and this article endorses this approach. But applying standard forms of decision theory imposes an overly demanding standard of numeric information, supposedly satisfied by point-valued utility and probability functions. To ameliorate this difficulty, a version of decision theory that requires merely comparative utilities and plausibilities is proposed. After a brief summary of this alternative, the article illustrates how comparative decision theory (...) affords a rational reconstruction of decisions made by exemplary scientists in two cases of theory choice: Buffon’s law and the luminiferous ether. It also offers a rational reconstruction of two cases of theory diagnosis: Mendeleev’s anomalies and the Pioneer anomaly. (shrink)

We argue that concerns about double-counting—using the same evidence both to calibrate or tune climate models and also to confirm or verify that the models are adequate—deserve more careful scrutiny in climate modelling circles. It is widely held that double-counting is bad and that separate data must be used for calibration and confirmation. We show that this is far from obviously true, and that climate scientists may be confusing their targets. Our analysis turns on a Bayesian/relative-likelihood approach to incremental confirmation. (...) According to this approach, double-counting is entirely proper. We go on to discuss plausible difficulties with calibrating climate models, and we distinguish more and less ambitious notions of confirmation. Strong claims of confirmation may not, in many cases, be warranted, but it would be a mistake to regard double-counting as the culprit. 1 Introduction2 Remarks about Models and Adequacy-for-Purpose3 Evidence for Calibration Can Also Yield Comparative Confirmation3.1 Double-counting I3.2 Double-counting II4 Climate Science Examples: Comparative Confirmation in Practice4.1 Confirmation due to better and worse best fits4.2 Confirmation due to more and less plausible forcings values5 Old Evidence6 Doubts about the Relevance of Past Data7 Non-comparative Confirmation and Catch-Alls8 Climate Science Example: Non-comparative Confirmation and Catch-Alls in Practice9 Concluding Remarks. (shrink)

The thesis of theory-ladenness of observations, in its various guises, is widely considered as either ill-conceived or harmless to the rationality of science. The latter view rests partly on the work of the proponents of New Experimentalism who have argued, among other things, that experimental practices are efficient in guarding against any epistemological threat posed by theory-ladenness. In this paper I show that one can generate a thesis of theory-ladenness for experimental practices from an influential New Experimentalist account. The notion (...) I introduce for this purpose is the concept of ‘theory-driven data reliability judgments’, according to which theories which are sought to be tested with a particular set of data guide reliability judgments about those very same data. I provide various prominent historical examples to show that TDRs are used by scientists to resolve data conflicts. I argue that the rationality of the practices which employ TDRs can be saved if the independent support of the theories driving TDRs is construed in a particular way. (shrink)

Traditionally, cognitive values have been thought of as a collective pool of considerations in science that frequently trade against each other. I argue here that a finer-grained account of the value of cognitive values can help reduce such tensions. I separate the values into groups, minimal epistemic criteria, pragmatic considerations, and genuine epistemic assurance, based in part on the distinction between values that describe theories per se and values that describe theory-evidence relationships. This allows us to clarify why these values (...) are central to science and what role they should play, while reducing the tensions among them. (shrink)

Michael Weisberg has argued that robustness analysis is useful in evaluating both scientific models and their implications and that robustness analysis comes in three types that share their form and aim. We argue for three cautionary claims regarding Weisberg's reconstruction: robustness analysis may be of limited or no value in evaluating models and their implications; the unificatory reconstruction conceals that the three types of robustness differ in form and role; there is no confluence of types of robustness. We illustrate our (...) central first claim with a case study: the application of Lotka-Volterra models to technology diffusion. (shrink)

I critically discuss a recent innovation in the debate surrounding the plausibility of rational choice theory : the appeal to evolutionary theory. Specifically, I assess Gigerenzer and colleagues’ claim that considerations based on natural selection show that, instead of making decisions in a RCT-like way, we rely on ‘simple heuristics’. As I try to make clearer here, though, Gigerenzer and colleagues’ arguments are unconvincing: we lack the needed information about our past to determine whether the premises on which they are (...) built are true—and, hence, we cannot tell whether they, in fact, speak against RCT. (shrink)

Kuhn’s famous thesis that there is ‘no unique algorithm’ for choosing between rival scientific theories is analysed using the machinery of social choice theory. It is shown that the problem of theory choice as posed by Kuhn is formally identical to a standard social choice problem. This suggests that analogues of well-known results from the social choice literature, such as Arrow’s impossibility theorem, may apply to theory choice. If an analogue of Arrow’s theorem does hold for theory choice this would (...) refute Kuhn’s thesis, but it would also pose a threat to the rationality of science, a threat that is if anything more worrying than that posed by Kuhn. Various possible ‘escape routes’ from Arrow’s impossibility result are examined, in particular Amartya Sen’s idea of ‘enriching the informational basis’. It is shown that Sen’s idea can be applied to the problem of theory choice in science. This in turn sheds light on two well-known approaches to inductive inference in philosophy of science: Bayesianism and statistical model selection. (shrink)

‘Bayesian epistemology’ became an epistemological movement in the 20th century, though its two main features can be traced back to the eponymous Reverend Thomas Bayes (c. 1701-61). Those two features are: (1) the introduction of a formal apparatus for inductive logic; (2) the introduction of a pragmatic self-defeat test (as illustrated by Dutch Book Arguments) for epistemic rationality as a way of extending the justification of the laws of deductive logic to include a justification for the laws of inductive logic. (...) The formal apparatus itself has two main elements: the use of the laws of probability as coherence constraints on rational degrees of belief (or degrees of confidence) and the introduction of a rule of probabilistic inference, a rule or principle of conditionalization. (shrink)

The existence of social learning has been confirmed in diverse taxa, from apes to guppies. In order to advance our understanding of the consequences of social transmission and evolution of behavior, however, we require statistical tools that can distinguish among diverse social learning strategies. In this paper, we advance two main ideas. First, social learning is diverse, in the sense that individuals can take advantage of different kinds of information and combine them in different ways. Examining learning strategies for different (...) information conditions illuminates the more detailed design of social learning. We construct and analyze an evolutionary model of diverse social learning heuristics, in order to generate predictions and illustrate the impact of design differences on an organism’s fitness. Second, in order to eventually escape the laboratory and apply social learning models to natural behavior, we require statistical methods that do not depend upon tight experimental control. Therefore we examine strategic social learning in an experimental setting.. (shrink)

The chapter discusses the principle of conservatism and traces how the general principle is related to the specific one. This tracing suggests that the principle of conservatism needs to be refined. Connecting the principle in cognitive science to more general questions about scientific inference also allows us to revisit the question of realism versus instrumentalism. The framework deployed in model selection theory is very general; it is not specific to the subject matter of science. The chapter outlines some non-Bayesian ideas (...) that have been developed in model selection theory. The principle of conservatism, like C. Lloyd Morgan's canon, describes a preference concerning kinds of parameters. It says that a model that postulates only lower-level intentionality is preferable to one that postulates higher-level intentionality if both fit the data equally well. The model selection approach to parsimony helps explain why unification is a theoretical virtue. (shrink)

In this paper I will argue that, in general, where the evidence supports two theories equally, the simpler theory is not more likely to be true and is not likely to be nearer the truth. In other words simplicity does not tell us anything about model bias. Our preference for simpler theories (apart from their obvious pragmatic advantages) can be explained by the facts that humans are known to elaborate unsuccessful theories rather than attempt a thorough revision and that a (...) fixed set of data can only justify adjusting a certain number of parameters to a limited degree of precision. No extra tendency towards simplicity in the natural world is necessary to explain our preference for simpler theories. Thus Occam's razor eliminates itself (when interpreted in this form). (shrink)

Philosophers of science have often favoured reductive approaches to how-possibly explanation. This article identifies three alternative conceptions making how-possibly explanation an interesting phenomenon in its own right. The first variety approaches “how possibly X?” by showing that X is not epistemically impossible. This can sometimes be achieved by removing misunderstandings concerning the implications of one’s current belief system but involves characteristically a modification of this belief system so that acceptance of X does not result in contradiction. The second variety offers (...) a potential how-explanation of X. It is usually followed by a range of further potential how-explanations of the same phenomenon. In recent literature the factual claims implied by the second variety have been downplayed whereas the heuristic role of mapping the space of conceptual possibilities has been emphasized. I will focus especially on this truth-bracketing sense of potentiality when looking closer at the second variety in the paper. The third variety has attracted less interest. It presents a partial how-explanation of X. Typically it aims to establish the existence of a mechanism by which X could be and was generated. The third conception stands out as the natural alternative for the advocate of ontic how-possibly explanations. This article transfers Salmon’s view that explanation-concepts can be broadly divided into epistemic, modal, and ontic to the context of how-possibly explanations. Moreover, it is argued that each of the three above-mentioned varieties of how-possibly explanation occurs in science. To recognize this may be especially relevant for philosophers. We are often misled by the promises of various why-explanation accounts, and seem to have forgotten nearly everything about the diversity of how-possibly explanations. (shrink)

Selectionist evolutionary theory has often been faulted for not making novel predictions that are surprising, risky, and correct. I argue that it in fact exhibits the theoretical virtue of predictive capacity in addition to two other virtues: explanatory unification and model fitting. Two case studies show the predictive capacity of selectionist evolutionary theory: parallel evolutionary change in E. coli, and the origin of eukaryotic cells through endosymbiosis.

Some think that issues to do with scientific method are last century's stale debate; Popper was an advocate of methodology, but Kuhn, Feyerabend, and others are alleged to have brought the debate about its status to an end. The papers in this volume show that issues in methodology are still very much alive. Some of the papers reinvestigate issues in the debate over methodology, while others set out new ways in which the debate has developed in the last decade. The (...) book will be of interest to philosophers and scientists alike in the reassessment it provides of earlier debates about method and current directions of research. (shrink)

Several quantitative techniques for choosing among data models are available. Among these are techniques based on algorithmic information theory, minimum description length theory, and the Akaike information criterion. All these techniques are designed to identify a single model of a data set as being the closest to the truth. I argue, using examples, that many data sets in science show multiple patterns, providing evidence for multiple phenomena. For any such data set, there is more than one data model that must (...) be considered close to the truth. I conclude that, since the established techniques for choosing among data models are unequipped to handle these cases, they cannot be regarded as adequate. ‡I presented a previous version of this paper at the 20th Biennial Meeting of the Philosophy of Science Association, Vancouver, November 2006. I am grateful to the audience for constructive discussion. I thank Leiden University students Marjolein Eysink Smeets and Lenneke Schrier for suggesting the cortisol example, and Remko van der Geest for comments on a draft. †To contact the author, please write to: Faculty of Philosophy, University of Leiden, P.O. Box 9515, 2300 RA Leiden, The Netherlands; e-mail: [email protected]. (shrink)

Akaike’s framework for thinking about model selection in terms of the goal of predictive accuracy and his criterion for model selection have important philosophical implications. Scientists often test models whose truth values they already know, and they often decline to reject models that they know full well are false. Instrumentalism helps explain this pervasive feature of scientific practice, and Akaike’s framework helps provide instrumentalism with the epistemology it needs. Akaike’s criterion for model selection also throws light on the role of (...) parsimony considerations in hypothesis evaluation. I explain the basic ideas behind Akaike’s framework and criterion; several biological examples, including the use of maximum likelihood methods in phylogenetic inference, are considered. (shrink)

Bayes' Theorem is a simple mathematical formula used for calculating conditional probabilities. It figures prominently in subjectivist or Bayesian approaches to epistemology, statistics, and inductive logic. Subjectivists, who maintain that rational belief is governed by the laws of probability, lean heavily on conditional probabilities in their theories of evidence and their models of empirical learning. Bayes' Theorem is central to these enterprises both because it simplifies the calculation of conditional probabilities and because it clarifies significant features of subjectivist position. Indeed, (...) the Theorem's central insight — that a hypothesis is confirmed by any body of data that its truth renders probable — is the cornerstone of all subjectivist methodology. (shrink)

Parsimony is a virtue of empirical theories. Is it also a virtue of philosophical theories? I review four contemporary accounts of the virtue of parsimony in empirical theorizing, and consider how each might apply to two prominent appeals to parsimony in the philosophical literature, those made on behalf of physicalism and on behalf of nominalism. None of the accounts of the virtue of parsimony extends naturally to either of these philosophical cases. This suggests that in typical philosophical contexts, ontological simplicity (...) has no evidential value. (shrink)

In this paper, I offer a new defense of scientific realism, tailored for the Akaikean paradigm of statistical hypothesis testing. After proposing definitions of verisimilitude and predictive success, I use computer simulations to show how the latter depends on the former, even in the kind of case featured in a recent argument for instrumentalism.

The paper provides a formal proof that efficient estimates of parameters, which vary as as little as possible when measurements are repeated, may be expected to provide more accurate predictions. The definition of predictive accuracy is motivated by the work of Akaike (1973). Surprisingly, the same explanation provides a novel solution for a well known problem for standard theories of scientific confirmation — the Ravens Paradox. This is significant in light of the fact that standard Bayesian analyses of the paradox (...) fail to account for the predictive utility of universal laws like All ravens are black. (shrink)

Curve-fitting typically works by trading off goodness-of-fit with simplicity, where simplicity is measured by the number of adjustable parameters. However, such methods cannot be applied in an unrestricted way. I discuss one such correction, and explain why the exception arises. The same kind of probabilistic explanation offers a surprising resolution to a common-sense dilemma.

The advent of formal definitions of the simplicity of a theory has important implications for model selection. But what is the best way to define simplicity? Forster and Sober ([1994]) advocate the use of Akaike's Information Criterion (AIC), a non-Bayesian formalisation of the notion of simplicity. This forms an important part of their wider attack on Bayesianism in the philosophy of science. We defend a Bayesian alternative: the simplicity of a theory is to be characterised in terms of Wallace's Minimum (...) Message Length (MML). We show that AIC is inadequate for many statistical problems where MML performs well. Whereas MML is always defined, AIC can be undefined. Whereas MML is not known ever to be statistically inconsistent, AIC can be. Even when defined and consistent, AIC performs worse than MML on small sample sizes. MML is statistically invariant under 1-to-1 re-parametrisation, thus avoiding a common criticism of Bayesian approaches. We also show that MML provides answers to many of Forster's objections to Bayesianism. Hence an important part of the attack on Bayesianism fails. (shrink)

Jun Otsuka’s excellent book, Thinking about Statistics - the Philosophical Foundations (Otsuka 2023) is mostly organized around the idea that different statistical approaches can be illuminated by linking them to different ideas in general epistemology. Otsuka connects Bayesianism to internalism and foundationalism, frequentism to reliabilism, and the Akaike Information Criterion in model selection theory to instrumentalism. This useful mapping doesn’t cover all the interesting ideas he presents. His discussions of causal inference and machine learning are philosophically insightful, as is his (...) idea that statisticians embrace an assumption that is similar to Hume’s Principle of the Uniformity of Nature. I discuss these topics in what follows, sometimes disagreeing with details while at other times adding ideas that complement those presented in the book. (shrink)

Scientists regularly invoke broadly aesthetic properties like elegance and simplicity when evaluating theories, but why should we expect aesthetic pleasure to signal an epistemic good? I argue that aesthetic judgements in science are best understood as a special case of affective cognition, and that the feelings on which these judgements are based are the upshots of metacognitive monitoring of the quality of our engagement with theory and evidence. Finding a theory beautiful fallibly signals that it fits well with our background (...) knowledge and that it’s neither artificially adjusted to accommodate otherwise disconfirming evidence nor fitted to noise in the data, which makes aesthetic pleasure a good heuristic for theory evaluation. (shrink)

Several philosophers and psychologists have characterized belief in conspiracy theories as a product of irrational reasoning. Proponents of conspiracy theories apparently resist revising their beliefs given disconfirming evidence and tend to believe in more than one conspiracy, even when the relevant beliefs are mutually inconsistent. In this paper, we bring leading views on conspiracy theoretic beliefs closer together by exploring their rationality under a probabilistic framework. We question the claim that the irrationality of conspiracy theoretic beliefs stems from an inadequate (...) response to disconfirming evidence and internal incoherence. Drawing analogies to Lakatosian research programs, we argue that maintaining a core conspiracy belief can be Bayes-rational when it is embedded in a network of auxiliary beliefs, which can be revised to protect the more central belief from disconfirmation. We propose that the (ir)rationality associated with conspiracy belief lies not in a flawed updating method, but in their failure to converge toward well-confirmed, stable belief networks in the long run. This approach not only reconciles previously disjointed views, but also points toward more specific hypotheses explaining why some agents may be prone to adopting beliefs in conspiracy theories. (shrink)

Metaphysicians often claim that some metaphysical theory should (or shouldn’t) be believed because it exhibits (or fails to exhibit) theoretical virtues such as simplicity. Metaphysicians also sometimes claim that the legitimacy of these sorts of appeals to theoretical virtues are vindicated by the similar appeals to theoretical virtues which scientists make in scientific theory choice. One objection to this methodological move is to claim that the metaphysician misdescribes the role that theoretical virtues play within science. In this paper I defend (...) the metaphysician’s use of theoretical virtues against this objection. (shrink)

Physical laws are strikingly simple, although there is no a priori reason they must be so. I propose that nomic realists of all types (Humeans and non-Humeans) should accept that simplicity is a fundamental epistemic guide for discovering and evaluating candidate physical laws. This principle of simplicity clarifies and addresses several problems of nomic realism and simplicity. A consequence is that the oft-cited epistemic advantage of Humeanism over non-Humeanism disappears, undercutting an influential epistemological argument for Humeanism. Moreover, simplicity is shown (...) to be more tightly connected to lawhood than to mere truth. (shrink)

Evaluative studies of inductive inferences have been pursued extensively with mathematical rigor in many disciplines, such as statistics, econometrics, computer science, and formal epistemology. Attempts have been made in those disciplines to justify many different kinds of inductive inferences, to varying extents. But somehow those disciplines have said almost nothing to justify a most familiar kind of induction, an example of which is this: “We’ve seen this many ravens and they all are black, so all ravens are black.” This is (...) enumerative induction in itsfullstrength. For it does not settle with a weaker conclusion (such as “the ravens observed in the future will all be black”); nor does it proceed with any additional premise (such as the statistical IID assumption). The goal of this paper is to take some initial steps toward a justification for the full version of enumerative induction, against counterinduction, and against the skeptical policy. The idea is to explore various epistemic ideals, mathematically defined as different modes of convergence to the truth, and look for one that is weak enough to be achievable and strong enough to justify a norm that governs both the long run and the short run. So the proposal is learning-theoretic in essence, but a Bayesian version is developed as well. (shrink)

In this article, I will provide a critical overview of the form of non-deductive reasoning commonly known as “Inference to the Best Explanation” (IBE). Roughly speaking, according to IBE, we ought to infer the hypothesis that provides the best explanation of our evidence. In section 2, I survey some contemporary formulations of IBE and highlight some of its putative applications. In section 3, I distinguish IBE from C.S. Peirce’s notion of abduction. After underlining some of the essential elements of IBE, (...) the rest of the entry is organized around an examination of various problems that IBE confronts, along with some extant attempts to address these problems. In section 4, I consider the question of when a fact requires an explanation, since presumably IBE applies only in cases where some explanation is called for. In section 5, I consider the difficult question of how we ought to understand IBE in light of the fact that among philosophers, there is significant disagreement about what constitutes an explanation. In section 6, I consider different strategies for justifying the truth-conduciveness of the explanatory virtues, e.g., simplicity, unification, scope, etc., criteria which play an indispensable role in any given application of IBE. In section 7, I survey some of the most recent literature on IBE, much of which consists of investigations of the status of IBE from the standpoint of the Bayesian philosophy of science. (shrink)

Several philosophers and psychologists have characterized belief in conspiracy theories as a product of irrational reasoning. Proponents of conspiracy theories apparently resist revising their beliefs given disconfirming evidence and tend to believe in more than one conspiracy, even when the relevant beliefs are mutually inconsistent. In this paper, we bring leading views on conspiracy theoretic beliefs closer together by exploring their rationality under a probabilistic framework. We question the claim that the irrationality of conspiracy theoretic beliefs stems from an inadequate (...) response to disconfirming evidence and internal incoherence. Drawing analogies to Lakatosian research programs, we argue that maintaining a core conspiracy belief can be Bayes-rational when it is embedded in a network of auxiliary beliefs, which can be revised to protect the more central belief from disconfirmation. We propose that the irrationality associated with conspiracy belief lies not in a flawed updating method, but in a failure to converge toward wellconfirmed, stable belief networks in the long run. This approach not only reconciles previously disjointed views, but also points toward more specific descriptions of why agents may be prone to adopting beliefs in conspiracy theories. (shrink)