Inference to the Best Explanation is an unrivalled exposition of a theory of particular interest to students both of epistemology and the philosophy of science.

How do we go about weighing evidence, testing hypotheses, and making inferences? According to the model of _Inference to the Best Explanation_, we work out what to infer from the evidence by thinking about what would actually explain that evidence, and we take the ability of a hypothesis to explain the evidence as a sign that the hypothesis is correct. In _Inference to the Best Explanation_, Peter Lipton gives this important and influential idea the development and assessment it deserves. The (...) second edition has been substantially enlarged and reworked, with a new chapter on the relationship between explanation and Bayesianism, and an extension and defence of the account of contrastive explanation. It also includes an expanded defence of the claims that our inferences really are guided by diverse explanatory considerations, and that this pattern of inference can take us towards the truth. This edition of _Inference to the Best Explanation_ has also been updated throughout and includes a new bibliography. (shrink)

Active inference offers a unified theory of perception, learning, and decision-making at computational and neural levels of description. In this article, we address the worry that active inference may be in tension with the belief–desire–intention model within folk psychology because it does not include terms for desires at the mathematical level of description. To resolve this concern, we first provide a brief review of the historical progression from predictive coding to active inference, enabling us to distinguish between (...) active inference formulations of motor control and active inference formulations of decision processes. We then show that, despite a superficial tension when viewed at the mathematical level of description, the active inference formalism contains terms that are readily identifiable as encoding both the objects of desire and the strength of desire at the psychological level of description. We demonstrate this with simple simulations of an active inference agent motivated to leave a dark room for different reasons. Despite their consistency, we further show how active inference may increase the granularity of folk-psychological descriptions by highlighting distinctions between drives to seek information versus reward—and how it may also offer more precise, quantitative folk-psychological predictions. Finally, we consider how the implicitly conative components of active inference may have partial analogues in other systems describable by the broader free energy principle to which it conforms. (shrink)

I argue that inference can tolerate forms of self-ignorance and that these cases of inference undermine canonical models of inference on which inferrers have to appreciate (or purport to appreciate) the support provided by the premises for the conclusion. I propose an alternative model of inference that belongs to a family of rational responses in which the subject cannot pinpoint exactly what she is responding to or why, where this kind of self-ignorance does nothing to undermine (...) the intelligence of the response. (shrink)

Inference versus consequence , an invited lecture at the LOGICA 1997 conference at Castle Liblice, was part of a series of articles for which I did research during a Stockholm sabbatical in the autumn of 1995. The article seems to have been fairly effective in getting its point across and addresses a topic highly germane to the Uppsala workshop. Owing to its appearance in the LOGICA Yearbook 1997 , Filosofia Publishers, Prague, 1998, it has been rather inaccessible. Accordingly it (...) is republished here with only bibliographical changes and an afterword. (shrink)

Defenders of Inference to the Best Explanation claim that explanatory factors should play an important role in empirical inference. They disagree, however, about how exactly to formulate this role. In particular, they disagree about whether to formulate IBE as an inference rule for full beliefs or for degrees of belief, as well as how a rule for degrees of belief should relate to Bayesianism. In this essay I advance a new argument against non-Bayesian versions of IBE. My (...) argument focuses on cases in which we are concerned with multiple levels of explanation of some phenomenon. I show that in many such cases, following IBE as an inference rule for full beliefs leads to deductively inconsistent beliefs, and following IBE as a non-Bayesian updating rule for degrees of belief leads to probabilistically incoherent degrees of belief. (shrink)

The techniques of causal inference are widely used throughout the non-experimental sciences to derive causal conclusions from probabilistic premises. This poses a philosophical question. What in the nature of causation accounts for the possibility of causal inference? In this paper I shall seek to answer this question. I shall develop a metaphysical theory of causation designed to explain the success of causal inference techniques. I shall then indicate how this theory also accounts naturally for single-case causation and (...) the relation between causation and rational action. Finally, I shall compare the resulting picture with “interventionist” approaches to causation. (shrink)

This paper presents a new problem for the inference rule commonly known as Inference to the Best Explanation (IBE). The problem is that uncertainty about parts of one’s evidence may undermine the inferrability of a hypothesis that would provide the best explanation of that evidence, especially in cases where there is an alternative hypothesis that would provide a better explanation of only the more certain pieces of evidence. A potential solution to the problem is sketched, in which IBE (...) is generalized to handle uncertain evidence by invoking a notion of evidential robustness. (shrink)

It is impossible to hold patently contradictory beliefs in mind together at once. Why? Because we know that it is impossible for both to be true. This impossibility is a species of rational necessity, a phenomenon that uniquely characterizes the relation between one person's beliefs. Here, Eric Marcus argues that the unity of the rational mind--what makes it one mind--is what explains why, given what we already believe, we can't believe certain things and must believe certain others in this special (...) sense. What explains this is that beliefs, and the inferences by which we acquire them, are constituted by a particular kind of endorsement of those very states and acts. This, in turn, entails that belief and inference are essentially self-conscious: to hold a belief or to make an inference is at the same time to know that one does. An examination of the nature of belief and inference, in light of the phenomenon of rational necessity, reveals how the unity of the rational mind is a function of our knowledge of ourselves as bound to believe the true. Rational self-consciousness is the form of mental togetherness. (shrink)

The paper offers an account of inference. The account underwrites the idea that inference requires that the reasoner takes her premises to support her conclusion. I reject views according to which such ‘takings’ are intuitions or beliefs. I sketch an alternative view on which inferring consists in attaching what I call ‘inferential force’ to a structured collection of contents.

We argue that stereotypes associated with concepts like he-said–she-said, conspiracy theory, sexual harassment, and those expressed by paradigmatic slurs provide “normative inference tickets”: conceptual permissions to automatic, largely unreflective normative conclusions. These “mental shortcuts” are underwritten by associated stereotypes. Because stereotypes admit of exceptions, normative inference tickets are highly flexible and productive, but also liable to create serious epistemic and moral harms. Epistemically, many are unreliable, yielding false beliefs which resist counterexample; morally, many perpetuate bigotry and oppression. Still, (...) some normative inference tickets, like some activated by sexual harassment, constitute genuine moral and hermeneutical advances. For example, our framework helps explain Miranda Fricker’s notion of “hermeneutical lacunae”: what early victims of “sexual harassment” — as well as their harassers — lacked before the term was coined was a communal normative inference ticket — one that could take us, collectively, from “this is happening” to “this is wrong.”. (shrink)

Inferences from the absence of evidence to something are common in ordinary speech, but when used in scientific argumentations are usually considered deficient or outright false. Yet, as demonstrated here with the help of various examples, archaeologists frequently use inferences and reasoning from absence, often allowing it a status on par with inferences from tangible evidence. This discrepancy has not been examined so far. The article analyses it drawing on philosophical discussions concerning the validity of inference from absence, using (...) probabilistic models that were originally developed to show that such inferences are weak and inconclusive. The analysis reveals that inference from absence can indeed be justified in many important situations of archaeological research, such as excavations carried out to explore the past existence and time-span of sedentary human habitation. The justification is closely related to the fact that archaeology explores the human past via its material remains. The same analysis points to instances where inference from absence can have comparable validity in other historical sciences, and to research questions in which archaeological inference from absence will be problematic or totally unwarranted. (shrink)

In informal terms, abductive reasoning involves inferring the best or most plausible explanation from a given set of facts or data. It is a common occurrence in everyday life and crops up in such diverse places as medical diagnosis, scientific theory formation, accident investigation, language understanding, and jury deliberation. In recent years, it has become a popular and fruitful topic in artificial intelligence research. This volume breaks new ground in the scientific, philosophical, and technological study of abduction. It presents new (...) ideas about inferential and information-processing foundations for knowledge and certainty. The authors argue that knowledge arises from experience by processes of abductive inference, in contrast to the view that it arises non-inferentially, or that deduction and inductive generalization are enough to account for knowledge. Much AI research is hypothetical, so the importance of this book is that it reports key discoveries about abduction that have been made as a result of designing, building, testing, and analyzing actual working knowledge-based systems for medical diagnosis and other abductive tasks. The book tells the story of six generations of increasingly sophisticated generic abduction machines, RED-1, RED-2, PEIRCE, MDX2, TIPS, QUAWDS, and the discovery of reasoning strategies that make it computationally feasible to form well-justified composite explanatory hypotheses, despite the threat of combinatorial explosion. The final chapter argues that perception is logically abductive and presents a layered-abduction computational model of perceptual information processing. This book will be of great interest to researchers in AI, cognitive science, and philosophy of science. (shrink)

Consider the following three claims. (i) There are no truths of the form ‘p and ~p’. (ii) No one holds a belief of the form ‘p and ~p’. (iii) No one holds any pairs of beliefs of the form {p, ~p}. Irad Kimhi has recently argued, in effect, that each of these claims holds and holds with metaphysical necessity. Furthermore, he maintains that they are ultimately not distinct claims at all, but the same claim formulated in different ways. I find (...) his argument suggestive, if not entirely transparent. I do think there is at least an important kernel of truth even in (iii), and that (i) ultimately explains what’s right about the other two. Consciousness of an impossibility makes belief in the obtaining of the corresponding state of affairs an impossibility. Interestingly, an appreciation of this fact brings into view a novel conception of inference, according to which it consists in the consciousness of necessity. This essay outlines and defends this position. A central element of the defense is that it reveals how reasoners satisfy what Paul Boghossian calls the Taking Condition and do so without engendering regress. (shrink)

What is the connection between justification and the kind of consequence relations that are studied by logic? In this essay, I shall try to provide an answer, by proposing a general conception of the kind of inference that counts as justified or rational.

Original and penetrating, this book investigates of the notion of inference from signs, which played a central role in ancient philosophical and scientific method. It examines an important chapter in ancient epistemology: the debates about the nature of evidence and of the inferences based on it--or signs and sign-inferences as they were called in antiquity. As the first comprehensive treatment of this topic, it fills an important gap in the histories of science and philosophy.

This article generalizes the explanationist account of inference to the best explanation. It draws a clear distinction between IBE and abduction and presents abduction as the first step of IBE. The second step amounts to the evaluation of explanatory power, which consist in the degree of explanatory virtues that a hypothesis exhibits. Moreover, even though coherence is the most often cited explanatory virtue, on pain of circularity, it should not be treated as one of the explanatory virtues. Rather, coherence (...) should be equated with explanatory power and considered to be derivable from the other explanatory virtues: unification, explanatory depth and simplicity. (shrink)

Plausibly, an inference is an act of coming to believe something on the basis of something else you already believe. But what is it to come to believe some- thing on the basis of something else? I propose a disjunctive answer: it is either for some beliefs to rationally cause another—where rational causation is understood as causation that is either actually or potentially productive of knowledge—or for some beliefs to “deviantly” cause another, but for the believer mistakenly to come (...) thereby to believe that the former have rationally caused the latter. The result, I argue, is both a theoretically satisfying account of the act of inferring and a demonstration of the power of a knowledge-first approach to the philosophy of mind. (shrink)

This paper discusses the nature and the status of inference to the best explanation. We outline the foundational role given IBE by its defenders and the arguments of critics who deny it any place at all ; argue that, on the two main conceptions of explanation, IBE cannot be a foundational inference rule ; sketch an account of IBE that makes it contextual and dependent on substantive empirical assumptions, much as simplicity seems to be ; show how that (...) account avoids the critics ' complaints and leaves IBE an important role ; and sketch how our account can clarify debates over IBE in arguments for scientific realism. (shrink)

This book investigates the role of inference in argumentation, considering how arguments support standpoints on the basis of different loci. The authors propose and illustrate a model for the analysis of the standpoint-argument connection, called Argumentum Model of Topics. A prominent feature of the AMT is that it distinguishes, within each and every single argumentation, between an inferential-procedural component, on which the reasoning process is based; and a material-contextual component, which anchors the argument in the interlocutors’ cultural and factual (...) common ground. The AMT explains how these components differ and how they are intertwined within each single argument. This model is introduced in Part II of the book, following a careful reconstruction of the enormously rich tradition of studies on inference in argumentation, from the antiquity to contemporary authors, without neglecting medieval and post-medieval contributions. The AMT is a contemporary model grounded in a dialogue with such tradition, whose crucial aspects are illuminated in this book. (shrink)

Van Fraassen (1989) argues that Inference to the Best Explanation is incoherent in the sense that adopting it as a rule for belief change will make one susceptible to a dynamic Dutch book. The present paper argues against this. A strategy is described that allows us to infer to the best explanation free of charge.

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)

We distinguish between three philosophical views on the neuroscience of predictive models: predictive coding, predictive processing and predictive engagement. We examine the concept of active inference under each model and then ask how this concept informs discussions of social cognition. In this context we consider Frith and Friston’s proposal for a neural hermeneutics, and we explore the alternative model of enactivist hermeneutics.

Written by pioneers in the field, this practical book presents an authoritative yet accessible overview of the methods and applications of causal inference. The text provides a thorough introduction to the basics of the theory for non-time-varying treatments and the generalization to complex longitudinal data.

It has recently been argued that a non-Bayesian probabilistic version of inference to the best explanation (IBE*) has a number of advantages over Bayesian conditionalization (Douven [2013]; Douven and Wenmackers [2017]). We investigate how IBE* could be generalized to uncertain evidential situations and formulate a novel updating rule IBE**. We then inspect how it performs in comparison to its Bayesian counterpart, Jeffrey conditionalization (JC), in a number of simulations where two agents, each updating by IBE** and JC, respectively, try (...) to detect the bias of a coin while they are only partially certain what side the coin landed on. We show that IBE** more often prescribes high probability to the actual bias than JC. We also show that this happens considerably faster, that IBE** passes higher thresholds for high probability, and that it in general leads to more accurate probability distributions than JC. (shrink)

Despite decades of focused philosophical investigation, Inference to the Best Explanation still lacks a precise articulation and compelling defense. The primary reason for this is that it is not at all clear what it means for a hypothesis to be the best available explanation of the evidence. This paper first seeks to rectify this problem by developing a formal explication of the explanatory virtue of power. A resulting account of IBE is then evaluated as a form of uncertain (...) class='Hi'>inference. Overall, this paper offers a precise account and novel defense of one important version of IBE. (shrink)

An accessible and contemporary introduction to the methods for determining cause and effect in the social sciences Causal inference encompasses the tools that allow social scientists to determine what causes what. Economists--who generally can't run controlled experiments to test and validate their hypotheses--apply these tools to observational data to make connections. In a messy world, causal inference is what helps establish the causes and effects of the actions being studied, whether the impact (or lack thereof) of increases in (...) the minimum wage on employment, the effects of early childhood education on incarceration later in life, or the introduction of malaria nets in developing regions on economic growth. Scott Cunningham introduces students and practitioners to the methods necessary to arrive at meaningful answers to the questions of causation, using a range of modeling techniques and coding instructions for both the R and Stata programming languages. - -. (shrink)

Some arguments include imperative clauses. For example: ‘Buy me a drink; you can’t buy me that drink unless you go to the bar; so, go to the bar!’ How should we build a logic that predicts which of these arguments are good? Because imperatives aren’t truth apt and so don’t stand in relations of truth preservation, this technical question gives rise to a foundational one: What would be the subject matter of this logic? I argue that declaratives are used to (...) produce beliefs, imperatives are used to produce intentions, and beliefs and intentions are subject to rational requirements. An argument will strike us as valid when anyone whose mental state satisfies the premises is rationally required to satisfy the conclusion. For example, the above argument reflects the principle that it is irrational not to intend what one takes to be the necessary means to one’s intended ends. I argue that all intuitively good patterns of imperative inference can be explained using off-the-shelf formulations of our rational requirements. I then develop a formal-semantic theory embodying this view that predicts a range of data, including free-choice effects and Ross’s paradox. The resulting theory shows one way that our aspirations to rational agency can be discerned in the patterns of our speech, and is a case study in how the philosophy of language and the philosophy of action can be mutually illuminating. (shrink)

Inference, understood as a form of conscious and active belief-revision, has recently attracted much interest among philosophers of mind. Many writings on the topic depict inference as a kind of process. However, this assumption is, to my knowledge, nowhere explicitly justified or even reflected upon. In this paper, I argue that the assumption is wrong: while processes take time, it is not possible that inferences take time. Both claims are conceptual observations. It is therefore conceptually impossible that the (...) mental act of inferring is a process. (shrink)

We argue that a modified version of Mill’s method of agreement can strongly confirm causal generalizations. This mode of causal inference implicates the explanatory virtues of mechanism, analogy, consilience, and simplicity, and we identify it as a species of Inference to the Best Explanation (IBE). Since rational causal inference provides normative guidance, IBE is not a heuristic for Bayesian rationality. We give it an objective Bayesian formalization, one that has no need of principles of indifference and yields (...) responses to the Voltaire objection, van Fraassen’s Bad Lot objection, and John Norton’s recent objection to IBE. (shrink)

This book grew out of previously published papers of mine composed over a period of years; they have been reworked (sometimes beyond recognition) so as to form a reasonably coherent whole. Part One treats of informative inference. I argue (Chapter 2) that the traditional principle of induction in its clearest formulation (that laws are confirmed by their positive cases) is clearly false. Other formulations in terms of the 'uniformity of nature' or the 'resemblance of the future to the past' (...) seem to me hopelessly unclear. From a Bayesian point of view, 'learning from experience' goes by conditionalization (Bayes' rule). The traditional stum bling block for Bayesians has been to fmd objective probability inputs to conditionalize upon. Subjective Bayesians allow any probability inputs that do not violate the usual axioms of probability. Many subjectivists grant that this liberality seems prodigal but own themselves unable to think of additional constraints that might plausibly be imposed. To be sure, if we could agree on the correct probabilistic representation of 'ignorance' (or absence of pertinent data), then all probabilities obtained by applying Bayes' rule to an 'informationless' prior would be objective. But familiar contra dictions, like the Bertrand paradox, are thought to vitiate all attempts to objectify 'ignorance'. BuUding on the earlier work of Sir Harold Jeffreys, E. T. Jaynes, and the more recent work ofG. E. P. Box and G. E. Tiao, I have elected to bite this bullet. In Chapter 3, I develop and defend an objectivist Bayesian approach. (shrink)

"Correlation is not causation" is one of the mantras of the sciences—a cautionary warning especially to fields like epidemiology and pharmacology where the seduction of compelling correlations naturally leads to causal hypotheses. The standard view from the epistemology of causation is that to tell whether one correlated variable is causing the other, one needs to intervene on the system—the best sort of intervention being a trial that is both randomized and controlled. In this paper, we argue that some purely correlational (...) data contains information that allows us to draw causal inferences: statistical noise. Methods for extracting causal knowledge from noise provide us with an alternative to randomized controlled trials that allows us to reach causal conclusions from purely correlational data. (shrink)

What is involved in making an inference? This chapter argues against what Paul Boghossian calls the Taking Condition: "Inferring necessarily involves the thinker taking his premises to support his conclusion and drawing his conclusion because of that fact" (2014: 5). I won’t argue that the Taking Condition is incoherent: that nothing can coherently play the role that takings are supposed to play in inference. Instead, I’ll argue that it cannot plausibly explain all the inferential knowledge that we ordinarily (...) take ourselves to have. Moreover, I’ll argue that we don’t need it to understand the nature of inference. (shrink)

Many classically valid meta-inferences fail in a standard supervaluationist framework. This allegedly prevents supervaluationism from offering an account of good deductive reasoning. We provide a proof system for supervaluationist logic which includes supervaluationistically acceptable versions of the classical meta-inferences. The proof system emerges naturally by thinking of truth as licensing assertion, falsity as licensing negative assertion and lack of truth-value as licensing rejection and weak assertion. Moreover, the proof system respects well-known criteria for the admissibility of inference rules. Thus, (...) supervaluationists can provide an account of good deductive reasoning. Our proof system moreover brings to light how one can revise the standard supervaluationist framework to make room for higher-order vagueness. We prove that the resulting logic is sound and complete with respect to the consequence relation that preserves truth in a model of the non-normal modal logic _NT_. Finally, we extend our approach to a first-order setting and show that supervaluationism can treat vagueness in the same way at every order. The failure of conditional proof and other meta-inferences is a crucial ingredient in this treatment and hence should be embraced, not lamented. (shrink)

In a situation in which several explanations compete, is the one that is better qua explanation also the one we should regard as the more likely to be true? Realists usually answer in the affirmative. They then go on to argue that since realism provides the best explanation for the success of science, realism can be inferred to. Nonrealists, on the other hand, answer the above question in the negative, thereby renouncing the inference to realism. In this paper I (...) separate the two issues. In the first section it is argued that a rationale can be provided for the inference to the best explanation; in the second, that this rationale cannot justify an inference to realism. The defence of the inference rests on the claim that our standards of explanatory power are subject to critical examination, which, in turn, should be informed by empirical considerations. By means of a comparison of the realist's explanation for the success of science with that of conventionalism and instrumentalism it is then shown that realism does not offer a superior explanation and should not, therefore, be inferred to. (shrink)

Predictive processing and active inference have become influential frameworks for understanding perception, action, emotion, and consciousness. Yet their philosophical interpretation remains contested. Should they be understood as internalist theories of neural representation, or can they be reinterpreted as accounts of embodied and socially situated sense-making? This article argues for the latter by developing a processual-perspectivist account of enactive inference. Building on the notion of predictive engagement, the paper interprets active inference not as the secluded operation of a (...) brain-bound predictive system, but as an embodied, affectively regulated, and socially mediated process of world-involvement. It further argues that consciousness is best understood as a dynamic stabilization of perspective formation, grounded in affective regulation, predictive openness, corrective constraint, and neurodynamic viability. The article proposes a triadic framework of processual perspectivism: the first-person perspective of embodied affectivity, the second-person perspective of interpersonal and we-forming co-regulation, and the third-person perspective of neurodynamic description. This framework integrates enactive accounts of sense-making, recent work on affective criticality in active inference, and phenomenological analyses of second-person and we-experience. The result is a neurophilosophical account in which mind is not merely predicted by the brain, but enacted through living, feeling, interacting, and neurodynamically regulated processes of perspective formation. (shrink)

In his seminal Inference to the Best Explanation, Peter Lipton adopted a causal view of explanation and a broadly Millian view of how causal knowledge is obtained. This made his account vulnerable to critics who charged that Inference to the Best Explanation is merely a dressed-up version of Mill’s methods, which in the critics’ view do the real inductive work. Lipton advanced two arguments to protect Inference to the Best Explanation against this line of criticism: the problem (...) of multiple differences and the problem of inferred differences. Lipton claimed that these two problems show Mill’s method of difference to be largely unworkable unless it is embedded in an explanationist framework. Here I consider both arguments as well as the best Millian defense against them. Since the existing Millian defense is only partially successful, I will develop a new and improved account. As an integral part of the argument, I show that my solutions to the problems of multiple and inferred differences offer new insight into Lipton’s main case study: Ignaz Semmelweis’s discovery of the cause of childbed fever. I conclude that the method of difference can overcome Lipton’s challenges outside an explanationist framework. (shrink)

Having justification for inferring a conclusion from some premises is independent of seeing how to make that inference. But if you do make the inference, then your belief in the conclusion is justified based on those premises. These natural thoughts about inference and justification generate two puzzles. One concerns logical omniscience, and the other concerns inferences to correct conclusions via fallacious rules. In this paper I introduce and develop what I call the Inferential Parsing Hypothesis as a (...) way of resolving these two puzzles. The hypothesis recruits an analogy between inferential processes and syntactic parsing processes to motivate novel epistemic theses about the nature of inferential basing. (shrink)

Information about the structure of a causal system can come in the form of observational data—random samples of the system's autonomous behavior—or interventional data—samples conditioned on the particular values of one or more variables that have been experimentally manipulated. Here we study people's ability to infer causal structure from both observation and intervention, and to choose informative interventions on the basis of observational data. In three causal inference tasks, participants were to some degree capable of distinguishing between competing causal (...) hypotheses on the basis of purely observational data. Performance improved substantially when participants were allowed to observe the effects of interventions that they performed on the systems. We develop computational models of how people infer causal structure from data and how they plan intervention experiments, based on the representational framework of causal graphical models and the inferential principles of optimal Bayesian decision‐making and maximizing expected information gain. These analyses suggest that people can make rational causal inferences, subject to psychologically reasonable representational assumptions and computationally reasonable processing constraints. (shrink)

Many epistemologists take Inference to the Best Explanation (IBE) to be “fundamental.” For instance, Lycan (1988, 128) writes that “all justified reasoning is fundamentally explanatory reasoning.” Conee and Feldman (2008, 97) concur: “fundamental epistemic principles are principles of best explanation.” Call them fundamentalists. They assert that nothing deeper could justify IBE, as is typically assumed of rules of deductive inference, such as modus ponens. However, logicians account for modus ponens with the valuation rule for the material conditional. By (...) contrast, fundamentalists account for IBE with an ill-defined set of relations that happen to furnish their favorite set of inductive inferences. To our eye, this seems a little too convenient—there is too much room for ad hoc, just-so stories about the “striking” correspondence between our explanatory and inductive practices. We will argue that the (explanatory) pluralism adopted by the leading theorists of the best explanation—philosophers of science—undermines fundamentalism. Section 1 clarifies fundamentalism’s key tenets. Section 2 presents pluralism’s challenge to fundamentalism. Section 3 considers a potential fundamentalist reply to this challenge. Sections 4 through 6 canvass the leading candidates for developing this fundamentalist reply, showing each to be unsatisfactory. (shrink)

This article discusses how inference to the best explanation can be justified as a practical meta - argument. It is, firstly, justified as a practical argument insofar as accepting the best explanation as true can be shown to further a specific aim. And because this aim is a discursive one which proponents can rationally pursue in — and relative to — a complex controversy, namely maximising the robustness of one’s position, IBE can be conceived, secondly, as a meta - (...) argument. My analysis thus bears a certain analogy to Sellars ’ well - known justification of inductive reasoning ; it is based on recently developed theories of complex argumentation. (shrink)

An influential suggestion about the relationship between Bayesianism and inference to the best explanation holds that IBE functions as a heuristic to approximate Bayesian reasoning. While this view promises to unify Bayesianism and IBE in a very attractive manner, important elements of the view have not yet been spelled out in detail. I present and argue for a heuristic conception of IBE on which IBE serves primarily to locate the most probable available explanatory hypothesis to serve as a working (...) hypothesis in an agent’s further investigations. Along the way, I criticize what I consider to be an overly ambitious conception of the heuristic role of IBE, according to which IBE serves as a guide to absolute probability values. My own conception, by contrast, requires only that IBE can function as a guide to the comparative probability values of available hypotheses. This is shown to be a much more realistic role for IBE given the nature and limitations of the explanatory considerations with which IBE operates. (shrink)

The mainstream philosophical approach to inference, which insists on sentence-like representations and a linguistic capability, excludes non-human animals as possible agents capable of making inferences. However, an abundance of studies show that many animal species exhibit behaviors that seem to rely on some kind of reasoning. My focus here are the transitive inference tasks, which most species solve quite successfully. These findings put pressure on the mainstream views, and still lack a convincing explanation. I introduce the concept of (...) affective representations, which satisfy the semantic and structural requirements for participating in inferences. I also argue for a broader conception of inference, and show how we can apply this view to explain the results of the transitive inference studies. Finally, I suggest it is more useful to think of flexibility of thought in terms of a continuous range, rather than a dichotomy of flexible vs. inflexible. (shrink)

We look at two fundamental logical processes, often intertwined in planning and problem solving: inference and update. Inference is an internal process with which we uncover what is implicit in the information we already have. Update, on the other hand, is produced by external communication, usually in the form of announcements and in general in the form of observations, giving us information that might not have been available (even implicitly) before. Both processes have received attention from the logic (...) community, usually separately. In this work, we develop a logical language that allows us to describe them together. We present syntax, semantics and a complete axiom system; we discuss similarities and differences with other approaches and mention how the work can be extended. (shrink)

Abstract In his Aquinas Lecture 1992 at Marquette University, Ernan McMullin discusses whether there is a pattern of inference that particularly characterizes the sciences of nature. He pursues this theme both on a historical and a systematic level. There is a continuity of concern across the ages that separate the Greek inquiry into nature from our own vastly more complex scientific enterprise. But there is also discontinuity, the abandonment of earlier ideals as unworkable. The natural sciences involve many types (...) of inference; three of these interlock in a special way to produce “retroductive inference,” the kind of complex inference that supports causal theory. (shrink)

A surge of empirical research demonstrating flexible cognition in animals and young infants has raised interest in the possibility of rational decision-making in the absence of language. A venerable position, which I here call “Classical Inferentialism”, holds that nonlinguistic agents are incapable of rational inferences. Against this position, I defend a model of nonlinguistic inferences that shows how they could be practically rational. This model vindicates the Lockean idea that we can intuitively grasp rational connections between thoughts by developing the (...) Davidsonian idea that practical inferences are at bottom categorization judgments. From this perspective, we can see how similarity-based categorization processes widely studied in human and animal psychology might count as practically rational. The solution involves a novel hybrid of internalism and externalism: intuitive inferences are psychologically rational (in the explanatory sense) given the intensional sensitivity of the similarity assessment to the internal structure of the agent's reasons for acting, but epistemically rational (in the justificatory sense) given an ecological fit between the features matched by that assessment and the structure of the agent's environment. The essay concludes by exploring empirical results that show how nonlinguistic agents can be sensitive to these similarity assessments in a way that grants them control over their opaque judgments. (shrink)

Explanation is asymmetric: if A explains B, then B does not explain A. Tradition- ally, the asymmetry of explanation was thought to favor causal accounts of explanation over their rivals, such as those that take explanations to be inferences. In this paper, we develop a new inferential approach to explanation that outperforms causal approaches in accounting for the asymmetry of explanation.

This article considers the prospects of inference to the best explanation as a method of confirming causal claims vis-à-vis the medical evidence of mechanisms. I show that IBE is actually descriptive of how scientists reason when choosing among hypotheses, that it is amenable to the balance/weight distinction, a pivotal pair of concepts in the philosophy of evidence, and that it can do justice to interesting features of the interplay between mechanistic and population level assessments.

To learn about real world phenomena, scientists have traditionally used models with clearly interpretable elements. However, modern machine learning (ML) models, while powerful predictors, lack this direct elementwise interpretability (e.g. neural network weights). Interpretable machine learning (IML) offers a solution by analyzing models holistically to derive interpretations. Yet, current IML research is focused on auditing ML models rather than leveraging them for scientific inference. Our work bridges this gap, presenting a framework for designing IML methods—termed ’property descriptors’—that illuminate not (...) just the model, but also the phenomenon it represents. We demonstrate that property descriptors, grounded in statistical learning theory, can effectively reveal relevant properties of the joint probability distribution of the observational data. We identify existing IML methods suited for scientific inference and provide a guide for developing new descriptors with quantified epistemic uncertainty. Our framework empowers scientists to harness ML models for inference, and provides directions for future IML research to support scientific understanding. (shrink)