The basic problem of a theory of truth approximation is defining when a theory is “close to the truth” about some relevant domain. Existing accounts of truthlikeness or verisimilitude address this problem, but are usually limited to the problem of approaching a “deterministic” truth by means of deterministic theories. A general theory of truth approximation, however, should arguably cover also cases where either the relevant theories, or “the truth”, or both, are “probabilistic” in nature. As a step forward in this (...) direction, we first present a general characterization of both deterministic and probabilistic truth approximation; then, we introduce a new account of verisimilitude which provides a simple formal framework to deal with such issue in a unified way. The connections of our account with some other proposals in the literature are also briefly discussed. (shrink)

Scientists often study hypotheses that they know to be false. This creates an interpretive problem for Bayesians because the probability assigned to a hypothesis is typically interpreted as the probability that the hypothesis is true. I argue that solving the interpretive problem requires coming up with a new semantics for Bayesian inference. I present and contrast two new semantic frameworks, and I argue that both of them support the claim that there is pervasive pragmatic encroachment on whether a given Bayesian (...) probability assignment is rational. (shrink)

Bayesianism and likelihoodism are two of the most important frameworks philosophers of science use to analyse scientific methodology. However, both frameworks face a serious objection: much scientific inquiry takes place in highly idealized frameworks where all the hypotheses are known to be false. Yet, both Bayesianism and likelihoodism seem to be based on the assumption that the goal of scientific inquiry is always truth rather than closeness to the truth. Here, I argue in favour of a verisimilitude framework for inductive (...) inference. In the verisimilitude framework, scientific inquiry is conceived of, in part, as a process where inference methods ought to be calibrated to appropriate measures of closeness to the truth. To illustrate the verisimilitude framework, I offer a reconstruction of parsimony evaluations of scientific theories, and I give a reconstruction and extended analysis of the use of parsimony inference in phylogenetics. By recasting phylogenetic inference in the verisimilitude framework, it becomes possible to both raise and address objections to phylogenetic methods that rely on parsimony. 1Introduction2Problems with the Law of Likelihood3Introducing Verisimilitude-Based Inference4Examples of Verisimilitude-Based Inference Procedures4.1Parsimony inference over theories4.2Parsimony inference in phylogenetics5Conclusion. (shrink)

Imagine that one must either lose all of one’s certainty about some very important topic – about the meaning of life, for example – or a small amount of certainty about each of one’s more “mundane” beliefs – beliefs about the color of one’s socks, where one’s keys are, whether it will rain, etc. One ought to take the latter loss, no matter how many mundane beliefs are at stake. Conversely, if one had to give up a tiny bit of (...) certainty about the important topic, or all of one’s certainty about most mundane topics, one ought to take the former loss. These two judgments generate what I call the “epistemic headaches-for-lives challenge,” so called because it resembles a challenge in ethics about the relative values of headaches and lives. In this paper, I argue for two things. First, our epistemic theories should accord with these types of judgments, and so must deal with the headaches-for-lives challenge. Second, this is a serious challenge for formal epistemology. (shrink)

It seems like we care about at least two features of our credence function: gradational-accuracy and verisimilitude. Accuracy-first epistemology requires that we care about one feature of our credence function: gradational-accuracy. So if you want to be a verisimilitude-valuing accuracy-firster, you must be able to think of the value of verisimilitude as somehow built into the value of gradational-accuracy. Can this be done? In a recent article, Oddie has argued that it cannot, at least if we want the accuracy measure (...) to be proper. I argue that it can. 1Introduction2Some Nuts and Bolts3First Attempts4Oddie’s Constraint5The Good5.1Proximity over the disagreement metric 5.2Proximity over the magnitude metric 6The Bad and the Ugly 7Some More Good: The Role of Evenness of Distribution 8Some More Bad: Which Propositions to Privilege? 9Concluding Thoughts: Accuracy and Practical Value. (shrink)

It is truism that accuracy is valued. Some deem accuracy to be among the most fundamental values, perhaps the preeminent value, of inquiry. Because of this, accuracy has been the focus of two different, important programs in epistemology. The truthlikeness program pursued the notion of propositional accuracy—an ordering of propositions by closeness to the objective truth of some matter. The epistemic utility program pursued the notion of credal state accuracy—an ordering of credal states by closeness to the ideal credal state. (...) It may be tempting to think that each question that informs an inquiry must be  determinate of necessity. That is to say, there must be one correct, precise answer to it. That precise, correct answer is not only the maximally accurate answer to the question, but it also singles out the ideal credal state that is the epistemic target of the inquiry. An enormous amount of the effort has been focused on accuracy with respect to necessarily determinate questions. But often the focus of an inquiry is a question that is not necessarily determinate. Those who deem accuracy to be cognitively valuable, and fundamentally so, need an account of accuracy for indeterminate questions. Ideally such an account would cover both propositional accuracy and credal state accuracy. This is a tall order, and it is not clear that the order can be filled. (shrink)

In the general problem of verisimilitude, we try to define the distance of a statement from a target, which is an informative truth about some domain of investigation. For example, the target can be a state description, a structure description, or a constituent of a first-order language. In the problem of legisimilitude, the target is a deterministic or universal law, which can be expressed by a nomic constituent or a quantitative function involving the operators of physical necessity and possibility. The (...) special case of legisimilitude, where the target is a probabilistic law, has been discussed by Roger Rosenkrantz and Ilkka Niiniluoto. Their basic proposal is to measure the distance between two probabilistic laws by the Kullback–Leibler notion of divergence, which is a semimetric on the space of probability measures. This idea can be applied to probabilistic laws of coexistence and laws of succession, and the examples may involve discrete or continuous state spaces. In this paper, these earlier studies are elaborated in four directions. First, even though deterministic laws are limiting cases of probabilistic laws, the target-sensitivity of truthlikeness measures implies that the legisimilitude of probabilistic laws is not easily reducible to the deterministic case. Secondly, the Jensen-Shannon divergence is applied to mixed probabilistic laws which entail some universal laws. Thirdly, a new class of distance measures between probability distributions is proposed, so that their horizontal differences are taken into account in addition to vertical ones. Fourthly, a solution is given for the epistemic problem of estimating degrees of probabilistic legisimilitude on the basis of empirical evidence. (shrink)

After Karl Popper’s original work, several approaches were developed to provide a sound explication of the notion of verisimilitude. With few exceptions, these contributions have assumed that the truth to be approximated is deterministic. This collection of ten papers addresses the more general problem of approaching probabilistic truths. They include attempts to find appropriate measures for the closeness to probabilistic truth and to evaluate claims about such distances on the basis of empirical evidence. The papers employ multiple analytical approaches, and (...) connect the research to related issues in the philosophy of science. (shrink)

Elgin offers an influential and far-reaching challenge to veritism. She takes scientific understanding to be non-factive and maintains that there are epistemically useful falsehoods that figure ineliminably in scientific understanding and whose falsehood is no epistemic defect. Veritism, she argues, cannot account for these facts. This paper argues that while Elgin rightly draws attention to several features of epistemic practices frequently neglected by veritists, veritists have numerous plausible ways of responding to her arguments. In particular, it is not clear that (...) false propositional commitments figure ineliminably in understanding in the manner supposed by Elgin. Moreover, even if scientific understanding were non-factive and false propositional commitments did figure ineliminably in understanding, the veritist can account for this in several ways without thereby abandoning veritism. (shrink)

Schoenfield has constructed examples of proper inaccuracy measures that value verisimilitude (in a certain sense) in spaces of worlds equipped with a particular variety of verisimilitude metric. However, Schoenfield left it as an open question whether ‘for every space of worlds, there is a proper inaccuracy measure that values verisimilitude’. Here I answer this question in the affirmative.

Some propositions are more epistemically important than others. Further, how important a proposition is is often a contingent matter—some propositions count more in some worlds than in others. Epistemic Utility Theory cannot accommodate this fact, at least not in any standard way. For EUT to be successful, legitimate measures of epistemic utility must be proper, i.e., every probability function must assign itself maximum expected utility. Once we vary the importance of propositions across worlds, however, normal measures of epistemic utility become (...) improper. I argue there isn’t any good way out for EUT. (shrink)

Truthlikeness is a property of a theory or a proposition that represents its closeness to the truth. We start by summarizing Niiniluoto’s proposal of truthlikeness for deterministic laws, which defines truthlikeness as a function of accuracy, and García-Lapeña’s expanded version, which defines truthlikeness for DL as a function of two factors, accuracy and nomicity. Then, we move to develop an appropriate definition of truthlikeness for probabilistic laws based on Niiniluoto’s suggestion to use the Kullback–Leibler divergence to define the distance between (...) a probability law X\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$X$$\end{document} and the true probability law T\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$T$$\end{document}. We argue that the Kullback–Leibler divergence seems to be the best of the available probability distances to measure accuracy between PL. However, as in the case of DL, we argue that accuracy represents a necessary but not sufficient condition, as two PL may be equally accurate and still one may imply more true or truthlike consequences, behaviours or true facts about the system than the other. The final proposal defines truthlikeness for PL as a function of two factors, p-accuracy and p-nomicity, in intimate connexion with García-Lapeña’s proposal for DL. (shrink)

The externalist says that your evidence could fail to tell you what evidence you do or not do have. In that case, it could be rational for you to be uncertain about what your evidence is. This is a kind of uncertainty which orthodox Bayesian epistemology has difficulty modeling. For, if externalism is correct, then the orthodox Bayesian learning norms of conditionalization and reflection are inconsistent with each other. I recommend that an externalist Bayesian reject conditionalization. In its stead, I (...) provide a new theory of rational learning for the externalist. I defend this theory by arguing that its advice will be followed by anyone whose learning dispositions maximize expected accuracy. I then explore some of this theory’s consequences for the rationality of epistemic akrasia, peer disagreement, undercutting defeat, and uncertain evidence. (shrink)

Accuracy-based arguments for conditionalization and probabilism appear to have a significant advantage over their Dutch Book rivals. They rely only on the plausible epistemic norm that one should try to decrease the inaccuracy of one’s beliefs. Furthermore, conditionalization and probabilism apparently follow from a wide range of measures of inaccuracy. However, we argue that there is an under-appreciated diachronic constraint on measures of inaccuracy which limits the measures from which one can prove conditionalization, and none of the remaining measures allow (...) one to prove probabilism. That is, among the measures in the literature, there are some from which one can prove conditionalization, others from which one can prove probabilism, but none from which one can prove both. Hence at present, the accuracy-based approach cannot underwrite both conditionalization and probabilism. (shrink)

ABSTRACTSuppose that beliefs come in degrees. How should we then measure the accuracy of these degrees of belief? Scoring rules are usually thought to be the mathematical tool appropriate for this job. But there are many scoring rules, which lead to different ordinal accuracy rankings. Recently, Fallis and Lewis [2016] have given an argument that, if sound, rules out many popular scoring rules, including the Brier score, as genuine measures of accuracy. I respond to this argument, in part by noting (...) that the argument fails to account for verisimilitude—that certain false hypotheses might be closer to the truth than other false hypotheses are. Oddie [forthcoming], however, has argued that no member of a very wide class of scoring rules can appropriately handle verisimilitude. I explain how to respond to Oddie's argument, and I recommend a class of weighted scoring rules that, I argue, genuinely measure accuracy while escaping the arguments of Fallis and Lewis as well as Oddie. (shrink)

There is an ongoing debate about which rule we ought to use for scoring probability estimates. Much of this debate has been premised on scoring-rule monism, according to which there is exactly one best scoring rule. In previous work, I have argued against this position. The argument given there was based on purely a priori considerations, notably the intuition that scoring rules should be sensitive to truthlikeness relations if, and only if, such relations are present among whichever hypotheses are at (...) issue. The present paper offers a new, quasi-empirical argument against scoring-rule monism. This argument uses computational simulations to show that different scoring rules can have different economical consequences, depending on the context of use. (shrink)

This paper is about guessing: how people respond to a question when they aren’t certain of the answer. Guesses show surprising and systematic patterns that the most obvious theories don’t explain. We argue that these patterns reveal that people aim to optimize a tradeoff between accuracy and informativity when forming their guess. After spelling out our theory, we use it to argue that guessing plays a central role in our cognitive lives. In particular, our account of guessing yields new theories (...) of belief, assertion, and the conjunction fallacy—the psychological finding that people sometimes rank a conjunction as more probable than one of its conjuncts. More generally, we suggest that guessing helps explain how boundedly rational agents like us navigate a complex, uncertain world. (shrink)

Bayesianism and likelihoodism are two of the most important frameworks philosophers of science use to analyse scientific methodology. However, both frameworks face a serious objection: much scientific inquiry takes place in highly idealized frameworks where all the hypotheses are known to be false. Yet, both Bayesianism and likelihoodism seem to be based on the assumption that the goal of scientific inquiry is always truth rather than closeness to the truth. Here, I argue in favor of a verisimilitude framework for inductive (...) inference. In the verisimilitude framework, scientific inquiry is conceived of, in part, as a process where inference methods ought to be calibrated to appropriate measures of closeness to the truth. To illustrate the verisimilitude framework, I offer a reconstruction of parsimony evaluations of scientific theories, and I give a reconstruction and extended analysis of the use of parsimony inference in phylogenetics. By recasting phylogenetic inference in the verisimilitude framework, it becomes possible to both raise and address objections to phylogenetic methods that rely on parsimony. (shrink)