Carl G. Hempel has often stated that inductive-statistical explanations, as he conceives them, are inductive arguments. This discussion note raises the question of whether such arguments are to be understood as (1) arguments of the traditional sort, containing premises and conclusions, governed by some sort of inductive "acceptance rules," or (2) something more closely akin to Carnap's degree of confirmation statements which occur in an inductive logic which entirely eschews inductive "acceptance rules." Hempel's writings do not seem unequivocal on this (...) issue. It is suggested that adoption of construal (2) would remove the need for Hempel's high probability requirement on I-S explanations. (shrink)

Causal assessment is the problem of establishing whether a relation between (variable) X and (variable) Y is causal. This problem, to be sure, is widespread across the sciences. According to accredited positions in the philosophy of causality and in social science methodology, invariance under intervention provides the most reliable test to decide whether X causes Y. This account of invariance (under intervention) has been criticised, among other reasons, because it makes manipulations on the putative causal factor fundamental for the causal (...) methodology; consequently, the argument goes, the account is ill-suited to those contexts where manipulations are not performed, for instance, the social sciences. The article aims to extend the account of invariance (under intervention), in a way that manipulations on the putative causal factors are not methodologically fundamental, and yet invariance remains key for causal assessment both in experimental and non-experimental contexts. (shrink)

We argue elsewhere that explanatoriness is evidentially irrelevant . Let H be some hypothesis, O some observation, and E the proposition that H would explain O if H and O were true. Then O screens-off E from H: Pr = Pr. This thesis, hereafter “SOT” , is defended by appeal to a representative case. The case concerns smoking and lung cancer. McCain and Poston grant that SOT holds in cases, like our case concerning smoking and lung cancer, that involve frequency (...) data. However, McCain and Poston contend that there is a wider sense of evidential relevance—wider than the sense at play in SOT—on which explanatoriness is evidentially relevant even in cases involving frequency data. This is their main point, but they also contend that SOT does not hold in certain cases not involving frequency data. We reply to each of these points and conclude with some general remarks on screening-off as a test of evidential relevance. (shrink)

Interventionism is a theory of causal explanation developed by Woodward and Hitchcock. I defend an interventionist perspective on the causal explanations offered within scientific psychology. The basic idea is that psychology causally explains mental and behavioral outcomes by specifying how those outcomes would have been different had an intervention altered various factors, including relevant psychological states. I elaborate this viewpoint with examples drawn from cognitive science practice, especially Bayesian perceptual psychology. I favorably compare my interventionist approach with well-known nomological and (...) mechanistic theories of psychological explanation. (shrink)

The explanatory role of natural selection is one of the long-term debates in evolutionary biology. Nevertheless, the consensus has been slippery because conceptual confusions and the absence of a unified, formal causal model that integrates different explanatory scopes of natural selection. In this study we attempt to examine two questions: (i) What can the theory of natural selection explain? and (ii) Is there a causal or explanatory model that integrates all natural selection explananda? For the first question, we argue that (...) five explananda have been assigned to the theory of natural selection and that four of them may be actually considered explananda of natural selection. For the second question, we claim that a probabilistic conception of causality and the statistical relevance concept of explanation are both good models for understanding the explanatory role of natural selection. We review the biological and philosophical disputes about the explanatory role of natural selection and formalize some explananda in probabilistic terms using classical results from population genetics. Most of these explananda have been discussed in philosophical terms but some of them have been mixed up and confused. We analyze and set the limits of these problems. (shrink)

Several metaphysical naturalists argue that the success of science, together with the claim that scientists adhere to methodological naturalism, amounts to strong evidence for metaphysical naturalism. I call this the scientific-success argument. It is argued that the scientific-success argument is similar to the no-miracles argument for realism in philosophy of science. On the no-miracles argument, the success of science is taken as strong evidence that scientific theories are true. Based on this similarity, some considerations relevant to one argument may also (...) be relevant to the other. One particular consideration is explored. The selectionist response to the no-miracles argument states that on an evolutionary model of science, in which scientific theories are accepted only after surviving a rigorous selection process, the no-miracles argument fails. The selectionist response also applies to the scientific-success argument. If scientific theories are selected for success, we do not need to explain the success of science by appealing to metaphysical naturalism. (shrink)

Seungbae Park argues that Bas van Fraassen’s rejection of inference to the best explanation (IBE) is problematic for his contextual theory of explanation because van Fraassen uses IBE to support the contextual theory. This paper provides a defense of van Fraassen’s views from Park’s objections. I point out three weaknesses of Park’s objection against van Fraassen. First, van Fraassen may be perfectly content to accept the implications that Park claims to follow from his views. Second, even if van Fraassen rejects (...) IBE he may still endorse particular arguments that instantiate IBE. Third, van Fraassen does not, in fact, use IBE to support his contextual theory. (shrink)

In “What Makes a Scientific Explanation Distinctively Mathematical?” (2013b), Lange uses several compelling examples to argue that certain explanations for natural phenomena appeal primarily to mathematical, rather than natural, facts. In such explanations, the core explanatory facts are modally stronger than facts about causation, regularity, and other natural relations. We show that Lange's account of distinctively mathematical explanation is flawed in that it fails to account for the implicit directionality in each of his examples. This inadequacy is remediable in each (...) case by appeal to ontic facts that account for why the explanation is acceptable in one direction and unacceptable in the other direction. The mathematics involved in these examples cannot play this crucial normative role. While Lange's examples fail to demonstrate the existence of distinctively mathematical explanations, they help to emphasize that many superficially natural scientific explanations rely for their explanatory force on relations of stronger-than-natural necessity. These are not opposing kinds of scientific explanations; they are different aspects of scientific explanation. (shrink)

In this paper I argue that the search for explainable models and interpretable decisions in AI must be reformulated in terms of the broader project of offering a pragmatic and naturalistic account of understanding in AI. Intuitively, the purpose of providing an explanation of a model or a decision is to make it understandable to its stakeholders. But without a previous grasp of what it means to say that an agent understands a model or a decision, the explanatory strategies will (...) lack a well-defined goal. Aside from providing a clearer objective for XAI, focusing on understanding also allows us to relax the factivity condition on explanation, which is impossible to fulfill in many machine learning models, and to focus instead on the pragmatic conditions that determine the best fit between a model and the methods and devices deployed to understand it. After an examination of the different types of understanding discussed in the philosophical and psychological literature, I conclude that interpretative or approximation models not only provide the best way to achieve the objectual understanding of a machine learning model, but are also a necessary condition to achieve post hoc interpretability. This conclusion is partly based on the shortcomings of the purely functionalist approach to post hoc interpretability that seems to be predominant in most recent literature. (shrink)

In this paper I argue that the search for explainable models and interpretable decisions in AI must be reformulated in terms of the broader project of offering a pragmatic and naturalistic account of understanding in AI. Intuitively, the purpose of providing an explanation of a model or a decision is to make it understandable to its stakeholders. But without a previous grasp of what it means to say that an agent understands a model or a decision, the explanatory strategies will (...) lack a well-defined goal. Aside from providing a clearer objective for XAI, focusing on understanding also allows us to relax the factivity condition on explanation, which is impossible to fulfill in many machine learning models, and to focus instead on the pragmatic conditions that determine the best fit between a model and the methods and devices deployed to understand it. After an examination of the different types of understanding discussed in the philosophical and psychological literature, I conclude that interpretative or approximation models not only provide the best way to achieve the objectual understanding of a machine learning model, but are also a necessary condition to achieve post hoc interpretability. This conclusion is partly based on the shortcomings of the purely functionalist approach to post hoc interpretability that seems to be predominant in most recent literature. (shrink)

In his recent book, John Norton has created a theory of inference to the best explanation, within the context of his "material theory of induction". I apply it to the problem of scientific explanations that are false: if we want the theories in our explanations to be true, then why do historians and scientists often say that false theories explained phenomena? I also defend Norton's theory against some possible objections.

Determining whether a homoplastic trait is the result of convergence or parallelism is central to many of the most important contemporary discussions in biology and philosophy: the relation between evolution and development, the importance of constraints on variation, and the role of contingency in evolution. In this article, I show that two recent attempts to draw a black-or-white distinction between convergence and parallelism fail, albeit for different reasons. Nevertheless, I argue that we should not be afraid of gray areas: a (...) clarified version of S. J. Gould's earlier account, based on a separation of underlying developmental mechanisms from the realized trait, still represents a useful approach. (shrink)

The philosophical literature on scientific explanation contains a striking diversity of accounts. I use novel empirical methods to address this fragmentation and assess the importance and generality of explanation in science. My evidence base is a set of 781 articles from one year of the journal Science, and I begin by applying text mining techniques to discover patterns in the usage of “explain” and other words of philosophical interest. I then use random sampling from the data set to develop and (...) test a classification scheme for scientific explanation. My results show that explanation and inference to the best explanation are ubiquitous in science, that they occur across a wide range of scientific disciplines, and that they are a goal of scientific practise. These explanations and inferences to the best explanation come in a diversity forms, which at least partially justifies the fragmentation of philosophical accounts. I draw two methodological lessons: first that text mining can enhance traditional conceptual analysis by establishing facts about word usage; and second that random sampling of cases can increase our confidence that a philosophical account applies in general. These empirical techniques supplement traditional philosophical methods. (shrink)

In 1958, to refute the argument known as the theoretician's dilemma, Hempel suggested that theoretical terms might be logically indispensable for inductive systematization of observational statements. This thesis, in some form or another, has later been supported by Scheffler, Lehrer, and Tuomela, and opposed by Bohnert, Hooker, Stegmüller, and Cornman. In this paper, a critical survey of this discussion is given. Several different putative definitions of the crucial notion inductive systematization achieved by a theory are discussed by reference to the (...) properties of inductive inference. The consequences of the following differences between deductive and inductive inference are emphasized: the lack of simple transitivity properties (even in a modified sense) of inductive inference, and the failure of the inductive analogue of the converse of The Deduction Theorem. The main conclusions are: (i) Hempel's original thesis may very well be right but his argument for it is unsatisfactory, (ii) theoretical terms can be logically indispensable for a non-Hempelian kind of inductive systematization, relative to both Craigian and Ramseyan elimination, (iii) Lehrer's attempt to prove the indispensability of theoretical terms for inductive-probabilistic systematization is, as a modification of Hempelian kind of inductive-deterministic systematization, unsatisfactory, and (iv) there does not seem to be much hope of escaping the conclusion (ii), if it is true, by extending the Craigian replacement programme along the lines suggested by Cornman. (shrink)

In this paper an attempt is made at developing the notion of a real and complete empirical explanation as excluding all forms of potential or incomplete explanations. This explanation is, however, no longer conceived as the proper aim of empirical science, for it can certainly be gleaned from recent epistemological publications that no comprehensive notion of a real and complete scientific explanation is likely to be constructed from within empirical science. Contrary to common understanding the empirical explanation, deductive-nomological as well (...) as statistical explanation, is considered here only as motive of scientific activities, i.e., as common aim of a transcending cooperation of scientific and non-scientific social practice. Following from this the proper aim of empirical science now consists in the development of practically relevant explanatory theories.This redetermination of the aim of scientific activities of empirical science also means criticism of the unification of deductive-nomological and statistical explanations, as it has been proposed by Wolfgang Stegmüller in his pragmatisch-epistemische Wende. For both forms of empirical explanation must be referred to fundamentally different kinds of practical relevance, the former playing a more important role in the advancement of social practice. Stegmüller's development of a comprehensive probabilistic notion of empirical explanation, as tied up to pragmatic knowledge-situations, in a way already transcends a scientifically immanent determination of it, but he seems to have stopped halfway on the road to practically relevant empirical explanations. Several insufficiencies with his probabilistic notion of empirical explanation are shown up in this paper as a consequence of his abiding by pragmatic, and not penetrating to practical, knowledge-situations. The final result of it, however, consists in a clarification and a modification of the concept of deductive-nomological explanation, originally developed by Hempel and Oppenheim. (shrink)

argue that natural selection does not explain the genotypic arid phenotypic properties of individuals. On this view, natural selection explains the adaptedness of individuals, not by explaining why the individuals that exist have the adaptations they do, but rather by explaining why the individuals that exist are the ones with those adaptations. This paper argues that this ‘Negative’ view of natural selection ignores the fact that natural selection is a cumulative selection process. So understood, it explains how the genetic sequences (...) that individuals inherit and that are responsible for their complex (and co-adapted) adaptations first arose in the gene-pool. (shrink)

In this paper, I argue that there is neither valid deductive support nor strong inductive support for Kuhn’s incommensurability thesis. There is no valid deductive support for Kuhn’s incommensurability thesis because, from the fact that the reference of the same kind terms changes or discontinues from one theoretical framework to another, it does not necessarily follow that these two theoretical frameworks are taxonomically incommensurable. There is no strong inductive support for Kuhn’s incommensurability thesis, since there are rebutting defeaters against it (...) in the form of episodes from the history of science that do not exhibit discontinuity and replacement, as Kuhn’s incommensurability thesis predicts, but rather continuity and supplementation. If this is correct, then there are no compelling epistemic reasons to believe that Kuhn’s incommensurability thesis is true or probable. (shrink)

Explaining phenomena is a primary goal of science. Consequently, it is unsurprising that gaining a proper understanding of the nature of explanation is an important goal of science education. In order to properly understand explanation, however, it is not enough to simply consider theories of the nature of explanation. Properly understanding explanation requires grasping the relation between explanation and understanding, as well as how explanations can lead to scientific knowledge. This article examines the nature of explanation, its relation to understanding, (...) and how explanations are used to generate scientific knowledge via inferences to the best explanation. Studying these features and applications of explanation not only provides insight into a concept that is important for science education in its own right, but also sheds light on an aspect of recent debates concerning the so-called consensus view of nature of science (NOS). Once the relation between explanation, understanding, and knowledge is clear, it becomes apparent that science is unified in important ways. Seeing this unification provides some support for thinking that there are general features of NOS of the sort proposed by the consensus view and that teaching about these general features of NOS should be a goal of science education. (shrink)

Quentin Smith has argued that it is logically impossible for there to be a divine cause of the universe. His argument is based on a Humean analysis of causation (confined to event causation, specifically excluding any consideration of agency) and a principle drawn from that analysis that he takes to be a logical requirement for every possibly valid theory of causation. He also thinks that all divine volitions are efficacious of logical necessity. We argue that all of these claims are (...) faulty, and that theists can resist Smith’s arguments without merely begging the question in favor of a divine cause. (shrink)

In recent years, quite a few evolutionary psychologists have come to embrace a heuristic interpretation of the discipline. They claim that, no matter how methodologically incomplete, adaptive thinking works fine as a good heuristic that effectively reduces the hypothesis space by generating novel and promising hypotheses that can eventually be empirically tested. The purpose of this article is to elucidate the use of heuristics in evolutionary psychology, thereby clarifying the role adaptive thinking has to play. To that end, two typical (...) heuristic interpretations—Machery’s "bootstrap strategy" and Goldfinch’s heuristically streamlined evolutionary psychology—are examined, focusing on the relationship between adaptive thinking and heuristics. The article draws two primary conclusions. The first is that the reliability of the heuristic hypothesis generation procedure should count no less than the conclusiveness of the final testing procedure in establishing scientific facts; nature does not always get the last word. Philosophy also counts. The second is that adaptive thinking constitutes a core heuristic in evolutionary psychology that provides the discipline with its raison d'être, but this is only possible when adaptive thinking is substantiated with sufficient historical underpinnings. (shrink)

Recently, Luk mentioned that scientific knowledge both explains and predicts. Do these two functions of scientific knowledge have equal significance, or is one of the two functions more important than the other? This commentary explains why prediction may be mandatory but explanation may be only desirable and optional.

Philosophers have recently expressed interest in accounting for the usefulness of mathematics to science. However, it is certainly not a new concern. Putnam and Quine have each worked out an argument for the existence of mathematical objects from the indispensability of mathematics to science. Were Quine or Putnam to disregard the applicability of mathematics to science, he would not have had as strong a case for platonism. But I think there must be ways of parsing mathematical sentences which account for (...) applicability of mathematics and also do not require us to believe in entities we have no evidence for, other than through reading these sentences literally. We will explore a particular way to interpret sentences of arithmetic which promises to account for their applicability without bringing in metaphysics not also brought in by science. The investigation will be limited to the arithmetic of cardinal numbers. The general strategy is to argue for the analogy between arithmetic and science, rather than to argue for one case having a particular characteristic independently of the other. (shrink)

Moral philosophers are, among other things, in the business of constructing moral theories. And moral theories are, among other things, supposed to explain moral phenomena. Consequently, one’s views about the nature of moral explanation will influence the kinds of moral theories one is willing to countenance. Many moral philosophers are (explicitly or implicitly) committed to a deductive model of explanation. As I see it, this commitment lies at the heart of the current debate between moral particularists and moral generalists. In (...) this paper I argue that we have good reasons to give up this commitment. In fact, I show that an examination of the literature on scientific explanation reveals that we are used to, and comfortable with, non-deductive explanations in almost all areas of inquiry. As a result, I argue that we have reason to believe that moral explanations need not be grounded in exceptionless moral principles. (shrink)

Wesley Salmon has advanced a new model of explanations of particular facts which requires that the explanans contain laws. The laws used in explanations (according to this model) are of the form P(A· C1,B)=p1... P(A· Cn,B)=pn. A condition imposed by Salmon on these laws is that the reference classes, i.e. A· C1... A· Cn, be homogenous with reference to the property B. A reference class A is homogenous with reference to a property B if every property which determines a place (...) selection with reference to B within A is statistically irrelevant to B in A. It is argued here that the concept of homogeneity cannot in general be satisfied in scientific explanations and that even a weaker requirement, "epistemic homogeneity," may be too strong. (shrink)

The idea that good explanations come with strong changes in probabilities has been very common. This criterion is called probabilistic difference-making. Since it is an intuitive criterion and has a long tradition in the literature on scientific explanation, it comes as a surprise that probabilistic difference-making is rarely discussed in the context of interventionist causal explanation. Specificity, proportionality, and stability are usually employed to measure explanatory power instead. This paper is a first step into the larger project of connecting difference-making (...) to the interventionist debate, and it starts by investigating whether probabilistic difference-making is contained in the notion of specificity. The choice of specificity is motivated by the observation that both probabilistic difference-making and specificity build on similar underlying intuitions. When comparing measures for both specificity and probabilistic difference-making, it turns out that the measures are not strictly correlated, and so the thesis that probabilistic difference-making is encoded within specificity has to be rejected. Some consequences of this result are discussed as well. (shrink)

Michael Strevens’ Depth: An Account of Scientific Explanation is an impressive recent contribution to the philosophical literature on explanation. While clearly influenced by several of the leading theories of the later twentieth century, Strevens’ account of explanation is firmly rooted in the causal tradition. His most notable intellectual debts in this regard owe to David Lewis, Wesley Salmon and James Woodward. Still, Strevens sees the work of these theorists as flawed in important respects, and his “kairetic account” of explanation is (...) meant to provide answers to problems his predecessors left unresolved (or poorly resolved, as the case may be). Before examining Strevens’ account in detail we should identify the more significant of these problems and briefly survey the contexts in which they arose. (shrink)

Most defenders of the new mechanistic approach accept ontic constraints for successful scientific explanation (Illari 2013; Craver 2014). The minimal claim is that scientific explanations have objective truthmakers, namely mechanisms that exist in the physical world independently of any observer and that cause or constitute the phenomena-to- be-explained. How can this idea be applied to type-level explanations? Many authors at least implicitly assume that in order for mechanisms to be the truthmakers of type-level explanation they need to be regular (Andersen (...) 2012; Sheredos 2015). One problem of this assumption is that most mechanisms are (highly) stochastic in the sense that they “fail more often than they succeed” (Bogen 2005; Andersen 2012). How can a mechanism type whose instances are more likely not to produce an instance of a particular phenomenon type be the truthmaker of the explanation of that particular phenomenon type? In this paper, I will give an answer to this question. I will analyze the notion of regularity and I will discuss Andersen's suggestion for how to cope with stochastic mechanisms. I will argue that her suggestion cannot account for all kinds of stochastic mechanisms and does not provide an answer as to why regularity grounds type-level explanation. According to my analysis, a mechanistic type- level explanation is true if and only if at least one of the following two conditions is satisfied: the mechanism brings about the phenomenon more often than any other phenomenon (comparative regularity) or the phenomenon is more often brought about by the mechanism than by any other mechanism/causal sequence (comparative reverse regularity). (shrink)

I analyze the frame problem and its relation to other epistemological problems for artificial intelligence, such as the problem of induction, the qualification problem and the "general" AI problem. I dispute the claim that extensions to logic (default logic and circumscriptive logic) will ever offer a viable way out of the problem. In the discussion it will become clear that the original frame problem is really a fairy tale: as originally presented, and as tools for its solution are circumscribed by (...) Pat Hayes, the problem is entertaining, but incapable of resolution. The solution to the frame problem becomes available, and even apparent, when we remove artificial restrictions on its treatment and understand the interrelation between the frame problem and the many other problems for artificial epistemology. I present the solution to the frame problem: an adequate theory and method for the machine induction of causal structure. Whereas this solution is clearly satisfactory in principle, and in practice real progress has been made in recent years in its application, its ultimate implementation is in prospect only for future generations of AI researchers. (shrink)

The proper function of the heart is pumping the blood. According to what we call the type etiological view, this is because previous tokens of the type HEART were selected for pumping the blood. Nanay :412–431, 2010) argues that the type etiological view is viciously circular. He claims that the only plausible accounts of trait type individuation use proper functions, such that whenever the type etiological view is supplemented with a plausible account of trait type individuation, the result is a (...) view that uses proper functions to explain proper functions. We refine this objection, and argue that Nanay at most establishes a potentially benign definitional circularity. However, we show that the type etiological view’s reliance on types nevertheless generates a vicious regress. Hence the type etiological view is false. We reject dispositional and modal alternatives to the type etiological view for the reason that they either cannot accommodate malfunction, or do so at the cost of proliferation; and then formulate a novel token etiological view that overcomes both problems because it makes no reference to trait types. (shrink)

Different why-questions emerge under different contexts and require different information in order to be addressed. Hence a relevance relation can hardly be invariant across contexts. However, what is indeed common under any possible context is that all explananda require scientific information in order to be explained. So no scientific information is in principle explanatorily irrelevant, it only becomes so under certain contexts. In view of this, scientific thought experiments can offer explanations, should we analyze their representational strategies. Their representations involve (...) empirical as well as hypothetical statements. I call this the “representational mingling” which bears scientific information that can explain events. Buchanan’s thought experiment from constitutional economics is examined to show how mingled representations explain. (shrink)

The Generality Problem is widely recognized to be a serious problem for reliabilist theories of justification. James R. Beebe's Statistical Solution is one of only a handful of attempted solutions that has garnered serious attention in the literature. In their recent response to Beebe, Julien Dutant and Erik J. Olsson successfully refute Beebe's Statistical Solution. This paper presents a New Statistical Solution that countenances Dutant and Olsson's objections, dodges the serious problems that trouble rival solutions, and retains the theoretical virtues (...) that made Beebe's solution so attractive in the first place. There indeed exists a principled, rigorous, conceptually sparse, and plausible solution to the Generality Problem: it is the New Statistical Solution. (shrink)

The aim of this paper is to develop an adequate version of the D-N theory of explanation for particular events and to show how the resulting D-N model can be used as a tool in articulating a regularity theory of causation and an analysis of the truth conditions for counterfactual conditionals. Starting with a basic model that is largely the product of other workers in this field, two new restrictions are formulated in order to construct a version of D-N explanation (...) that does not yield the counterintuitive results that have plagued all earlier versions. An additional condition is then developed that is indispensable for utilizing the D-N model as a tool in formulating a regularity theory of causation. Finally, it is shown how a suitable model of potential D-N explanation facilitates the formulation of a theory of counterfactual conditionals. (shrink)