-
22221.263789
Hypotheses about how and why animals behave the way they do are frequently labelled as either associative or cognitive. This has been taken as evidence that there is a fundamental distinction between two kinds of behavioural processes. However, there is significant disagreement about how to define this distinction whether it ought to be rejected entirely. Rather than seeking a definition of the associative-cognitive distinction, or advocating for its rejection, I argue that it is an artefact of the way that comparative psychologists generate hypotheses. I suggest that hypotheses for non-human animal behaviour are often generated by analogy with hypotheses drawn from human psychology and associative learning theory, a justifiable strategy since analogies help to establish the pursuit-worthiness of a hypothesis. Any apparent distinction is a misleading characterisation of what is a complex web of hypotheses that explain diverse behavioural phenomena. The analogy view of the distinction has three advantages. It motivates the apparent existence of the distinction based on a common inference strategy in science, analogical reasoning. It accounts for why the distinction has been difficult to articulate, because of the diversity of possible analogies. Finally, it delimits the role of the distinction in downstream inferences about animal behaviour.
-
58330.264
We say we believe that all children can learn, but few of us really believe it.” Lisa Delpit Teachers are expected to believe in the potential of every student in front of them. To believe otherwise is to give up on a central premise of the educational mission, that students can be taught. However, the people who come into the classroom have different levels of knowledge, skills, and motivations. To deny that what the student brings to the classroom matters to their potential progress is to deny empirical reality. Teachers face a tension between cultivating high expectations for student success and recognizing the limitations that a student and their circumstances impose.
-
80704.264033
In my new paper, “Severe Testing: Error Statistics versus Bayes Factor Tests”, now out online at the The British Journal for the Philosophy of Science, I “propose that commonly used Bayes factor tests be supplemented with a post-data severity concept in the frequentist error statistical sense”. …
-
124615.264048
Epistemologists have devoted an enormous amount of attention to justification. Hundreds of papers have tried to analyze the conditions under which a belief is epistemically justified; hundreds more have offered counterexamples to these analyses. Even epistemologists who look askance at conceptual analysis have found it fruitful to explore the connections between justification and other epistemic notions, such as knowledge, rationality, and evidence. Some have even suggested that justification is the central notion in epistemology.
-
140352.264061
The Inscrutable Evidence Argument targets the thesis that credences are thoughts about evidential probabilities (CTEP). It does so using cases where one knows one’s evidence speaks either strongly in favor of or strongly against a proposition, but one doesn’t know which; in such cases, it seems possible to have a middling credence in that proposition even though one doesn’t think the probability of the proposition is near 50%—contra CTEP. In this paper, I defend CTEP by conceiving of the thoughts involved differently than usual. My diagnosis of the argument turns on appreciating the difference between believing and accepting (in the sense of Bratman 1992) that a proposition has probability n, where accepting is context dependent and allows for guidance in action without commitment to truth. I develop this diagnosis in two directions, one according to which acceptances of probability-involving propositions are credences and another according to which they aren’t. Both views elude the Inscrutable Evidence Argument and are compatible with CTEP.
-
140374.264069
In a short note written in 1929, Frank Ramsey put forward a reliabilist account of knowledge anticipating those given by Armstrong (1973) and Goldman (1967), among others, a few decades later. Some think that the note comprises the bulk of what Ramsey has to say about epistemology. But Ramsey’s ideas about epistemology extend beyond the note. Relatively little attention has been paid to his reliabilist account of reasonable belief. Even less attention has been paid to his reliabilist account of reasonable degree of belief. In this paper, I spell out these aspects of Ramsey’s epistemology in more detail than has been done so far. I argue that Ramsey anticipates contemporary reliabilist accounts of justified belief and justified degree of belief. I also flesh out Ramsey’s reasons for being a reliabilist. This is worth doing since Ramsey has one of the earliest arguments for reliabilism, but it has received scarce attention. Also, Ramsey calls his reliabilism “a kind of pragmatism,” and examining the argument will help us clarify Ramsey’s pragmatist commitments and better understand his version of reliabilism. I argue that when viewed through contemporary lenses, Ramsey’s reliabilism contains revisionist elements: he’s not opposed to what we now call “conceptual engineering.”
-
140423.264077
Emotions can get things right and serve us in many productive ways. They can also get things wrong and harm our epistemic or practical endeavors. Resenting somebody for having insulted your friend gets it wrong when your friend well understood that the remark was a joke. On the other hand, if your friend is not familiar with the given cultural context and hence couldn’t quite grasp the subtly sexist nature of the joke, your resentment might not only be appropriate but also help her navigate the new social context. Hoping that your meeting with your supervisor will be productive might motivate you to prepare better but will be inappropriate if all your previous meetings were failures.
-
140451.264089
Besides disagreeing about how much one should donate to charity, moral theories also disagree about where one should donate. In many cases, one intuitively attractive option is to split your donations across all of the charities that are recommended by theories in which you have positive credence, with each charity’s share being proportional to your credence in the theories that recommend it. Despite the fact that something like this approach is already widely used by real-world philanthropists to distribute billions of dollars, it is not supported by any account of handling decisions under moral uncertainty that has been proposed thus far in the literature. This paper develops a new bargaining-based approach that honors the proportionality intuition. We also show how this approach has several advantages over the best alternative proposals.
-
140605.264098
A moral requirement R1 is said to be lexically prior to a moral requirement R just in case we are morally obliged to uphold R1 at the expense of R2—no matter how many times R2 must be violated thereby. While lexical priority is a feature of many ethical theories, and arguably a part of common sense morality, attempts to model it within the framework of decision theory have led to a series of problems—a fact which is sometimes spun as a “decision theoretic critique” of lexical priority. In this paper, I develop an enriched decision theoretic framework that is capable of model-ling lexical priority while avoiding all extant problems. This will involve introducing several new ingredients into the standard decision theoretic framework, including multidimensional utilities, de minimis risks, and the means to represent two different conceptions of risk.
-
140700.264105
Skeptical theists contend that human cognitive limitations undermine atheistic arguments from evil. One recent challenge to skeptical theism has been posed by Climenhaga (2025), who argues that if we should—as some skeptical theists argue— be agnostic about the probability of the total collection of evils we observe given theism, Pr(E|T), we should also be agnostic about the probability of theism given these evils, Pr(T|E), and therefore be agnostic with respect to God’s existence. If one is persuaded, as I am, that Climenhaga’s argument is correct, the most promising skeptical theist response available seems to be one of mitigation: concede that Pr(E|T) is not inscrutable—and thereby concede skeptical theism cannot undermine arguments from the total collection of observable evils to the nonexistence of God— but maintain that skeptical theism is still able to undermine other Bayesian problems of evil; namely, those which argue from some individual instance of observable evil to the nonexistence of God. However, as I will argue, this mitigation strategy is not viable: if Pr(Ei|T) is inscrutable, where Pr(Ei|T) is the probability of any individual instance of observable evil occurring given theism, so too is Pr(E|T) correspondingly inscrutable. Therefore, absent demonstrating Climenhaga to be incorrect, skeptical theism cannot undermine any Bayesian arguments from evil.
-
182352.264117
What is the point of inquiry? Some say that the aim of inquiring into some question is to come to know its answer; others, that the aim is to attain justified belief, epistemic improvement, or some other coveted epistemic status. Still others eschew “aim” talk altogether, and instead formulate norms governing inquiry. However, virtually all extant work on inquiry has agreed on at least this much: the aims or norms of inquiry can be specified in terms of the epistemic states of the inquirer (i.e., the agent conducting the inquiry). This paper argues that this conception of inquiry struggles to account for some central features of what is arguably the most successful form of inquiry in the modern era: scientific inquiry. We show that scientific inquiry is governed by several distinctive norms that are difficult to explain if inquiry is all about achieving epistemic benefits for the inquirer. Instead, many inquiries aim to confer epistemic benefits on others. This “inclusive” conception of inquiry has important advantages and implications.
-
599269.264124
We employ a pragmatist model of inquiry to explain how measurement in physics can solve the problem of usefulness. In spite of the fact that a variety of resources, including theory, simulation, heuristics, rules of thumb, and practical considerations contribute to the context of a specific measurement inquiry, the measurement inquiry process partially decontextualizes its results, making them useful for other inquiries. This measurement inquiry process involves a process of transformation of data we call “entheorization,” which happens in conjunction with the evaluation of uncertainty of measurement results. These uncertainty estimates then serve to define the sensitivity of the result to the aims of subsequent inquiries. On this approach, the epistemology of measurement requires treating measurement procedure, uncertainty estimation, and sensitivity to targets of inquiry as equally fundamental to understanding how measurement yields knowledge. To help understand how the abstract elements of our epistemological model of experimental inquiries are applicable to concrete episodes of measurement, we use the example of the W -boson mass measurement at the Large Hadron Collider to illustrate our arguments.
-
713251.264132
philosophical logic may also interest themselves with the logical appendices, one of which presents modal logic as a subsystem of the logic of counterfactuals. Last but not least, the work also includes an afterword that is both a severe reprimand to the analytic community for a certain sloppiness and an exhortation to all colleagues to apply more rigor and patience in addressing metaphysical issues. People familiar with Williamson’s work will not be surprised by the careful and detailed (sometimes a bit technical) argumentation, which demands careful attention from the reader. As expected, this is a most relevant contribution to an increasingly popular topic by one of today’s leading analytic philosophers.
-
748624.264139
The term algorithmic fairness is used to assess whether
machine learning algorithms operate fairly. To get a sense of when
algorithmic fairness is at issue, imagine a data scientist is provided
with data about past instances of some phenomenon: successful
employees, inmates who when released from prison go on to reoffend,
loan recipients who repay their loans, people who click on an
advertisement, etc. and is tasked with developing an algorithm that
will predict other instances of these phenomena. While an algorithm
can be successful or unsuccessful at its task to varying degrees, it
is unclear what makes such an algorithm fair or unfair.
-
921728.264146
The concept of preference spans numerous research fields, resulting in
diverse perspectives on the topic. Preference logic specifically
focuses on reasoning about preferences when comparing objects,
situations, actions, and more, by examining their formal properties. This entry surveys major developments in preference logic to date. Section 2
provides a historical overview, beginning with foundational work by
Halldén and von Wright, who emphasized the syntactic aspects of
preference. In
Section 3,
early semantic contributions by Rescher and Van Dalen are introduced. The consideration of preference relations over possible worlds
naturally gives rise to modal preference logic where preference
lifting enables comparisons across sets of possible worlds.
-
1003205.264153
The literature on values in science contains countless claims to the effect that a particular type of scientific choice is or is not value-laden. This chapter exposes an ambiguity in the notion of a value-laden choice. In the first half, I distinguish four ways a choice can be said to be value-laden. In the second half, I illustrate the usefulness of this taxonomy by assessing arguments about whether the value-ladenness of science is inevitable. I focus on the “randomizer reply,” which claims that, in principle, scientists could always avoid value-laden choices by flipping a coin.
-
1057316.264161
Accuracy plays an important role in the deployment of machine learning algorithms. But accuracy is not the only epistemic property that matters. For instance, it is well-known that algorithms may perform accurately during their training phase but experience a significant drop in performance when deployed in real-world conditions. To address this gap, people have turned to the concept of algorithmic robustness. Roughly, robustness refers to an algorithm’s ability to maintain its performance across a range of real-world and hypothetical conditions. In this paper, we develop a rigorous account of algorithmic robustness grounded in Robert Nozick’s counterfactual sensitivity and adherence conditions for knowledge. By bridging insights from epistemology and machine learning, we offer a novel conceptualization of robustness that captures key instances of algorithmic brittleness while advancing discussions on reliable AI deployment. We also show how a sensitivity-based account of robustness provides notable advantages over related approaches to algorithmic brittleness, including causal and safety-based ones.
-
1118598.264169
In this paper, we argue that a perceiver’s contributions to perception can substantially affect what objects are represented in perceptual experience. To capture the scalar nature of these perceiver-contingent contributions, we introduce three grades of subject-dependency in object perception. The first grade, “weak subject-dependency,” concerns attentional changes to perceptual content like, for instance, when a perceiver turns their head, plugs their ears, or primes their attention to a particular cue. The second grade, “moderate subject-dependency,” concerns changes in the contingent features of perceptual objects due to action-orientation, location, and agential interest. For instance, being to the right or left of an object will cause the object to have a corresponding locative feature, but that feature is non-essential to the object in question. Finally, the third grade, “strong subject-dependency,” concerns generating perceptual objects whose existence depends upon their perceivers’ sensory contributions to perception. For this final grade of subject-dependency the adaptive perceptual system shapes diverse representations of sensory information by contributing necessary features to perceptual objects. To exemplify this nonstandard form of object perception we offer evidence from the future-directed anticipation of perceptual experts, and from the feature binding of synesthetes. We conclude that strongly subject-dependent perceptual objects are more than mere material objects, but are rather a necessary combination of material objects with the contributions of a perceiving subject.
-
1402670.264178
In the last fifth of their interview, Adelstein and Huemer discuss my views. I now respond point by point. Adelstein:
And it just almost feels like there's something different going on when Bryan Caplan does moral reasoning than when I do. …
-
1485723.264186
A decade ago, Effective Altruism got an early taste of bad PR when someone at an EA Global conference was widely reported as enthusing that EA was “the last social movement the world would ever need,” or words to that effect. …
-
1742959.264194
Levy’s Upward Theorem says that the conditional expectation of an integrable random variable converges with probability one to its true value with increasing information. In this paper, we use methods from effective probability theory to characterise the probability one set along which convergence to the truth occurs, and the rate at which the convergence occurs. We work within the setting of computable probability measures defined on computable Polish spaces and introduce a new general theory of effective disintegrations. We use this machinery to prove our main results, which (1) identify the points along which certain classes of effective random variables converge to the truth in terms of certain classes of algorithmically random points, and which further (2) identify when computable rates of convergence exist. Our convergence results significantly generalize earlier results within a unifying novel abstract framework, and there are no precursors of our results on computable rates of convergence. Finally, we make a case for the importance of our work for the foundations of Bayesian probability theory.
-
1821242.264201
As I write in the spring of 2025, we are in the midst of a crisis in the United States. The crisis is economic, social, political, and legal. One dimension of this crisis is the attack on higher education by the Trump administration. To date, this attack has included: o Cuts to funding for existing federal grants to higher education o Substantive content restrictions on applications for new grants o Deporting, or canceling visas of, international students and scholars without due cause o Denial of entry into the United States of international scholars traveling for academic or research activities o Increases to the amount of overhead that universities must pay to support federal grants o Threats to increase endowment tax on universities from 1.4% to as much as 35% o Closure, or severe cuts to funding, of libraries, museums, and archives.
-
1839358.264209
There are four well-known models of fundamental objective probabilistic reality: classical probability, comparative probability, non-Archimedean probability, and primitive conditional probability. I offer two desiderata for an account of fundamental objective probability, comprehensiveness and non-superfluity. It is plausible that classical probabilities lack comprehensiveness by not capturing some intuitively correct probability comparisons, such as that it is less likely that 0 = 1 than that a dart randomly thrown at a target will hit the exact center, even though both classically have probability zero. We thus want a comparison between probabilities with a higher resolution than we get from classical probabilities. Comparative and non-Archimedean probabilities have a hope of providing such a comparison, but for known reasons do not appear to satisfy our desiderata. The last approach to this problem is to employ primitive conditional probabilities, such as Popper functions, and then argue that P(0 = 1 | 0 = 1 or hit center) = < 1 = P (hit center | 0 = 1 or hit center). But now we have a technical question: How can we reconstruct a probability comparison, ideally satisfying the standard axioms of comparative probability, from a primitive conditional probability? I will prove that, given some plausible assumptions, it is impossible to perform this task: conditional probabilities just do not carry enough information to define a satisfactory comparative probability. The result is that of the models, no one satisfies our two desiderata. We end by briefly considering three paths forward.
-
2004396.264217
This is probably an old thing that has been discussed to death, but I only now noticed it. Suppose an open future view on which future contingents cannot have truth value. What happens to entailments? …
-
2011014.264224
While correct as far as it goes, this standard picture can encourage an overly sharp distinction between scientific activities and ethical deliberation. Far from entering only at the policy-making stage, ethical judgments often shape scientific research itself. This is most obvious in the choice of research questions. The choice of what to study ultimately affects what knowledge can be brought to bear in real-world decisions, including consequences for which (and whose) decisions can be made with the benefit of scientific insight.
-
2184789.264236
In the philosophy of religion, ‘de jure objections’ is an umbrella term that covers a wide variety of arguments for the conclusion that theistic belief is rationally impermissible, whether or not God exists. What we call ‘modal Calvinism’ counters these objections by proposing that ‘if God exists, God would ensure that theistic belief is rationally compelling on a global scale’, a modal conditional that is compatible with atheism. We respond to this modal Calvinist argument by examining it through the lenses of probability, modality, and logic – particularly, we apply analytical tools such as possible world semantics, Bayesian reasoning, and paraconsistent models. After examining various forms of the argument, we argue that none can compel atheists to believe that serious theistic possibilities worth considering would involve the purported divine measure.
-
2184879.264247
Alejandro Bortolus, Chad L. Hewitt, Veli Mitova, Evangelina Schwindt, Temitope O. Sogbanmu, Emelda E. Chukwu, Remco Heesen, Ricardo Kaufer, Hannah Rubin, Mike D. Schneider, Anne Schwenkenbecher, Helena Slanickova, Katie Woolaston, Li-an Yu
Determining appropriate mechanisms for transferring and translating research into policy has become a major concern for researchers (knowledge producers) and policymakers (knowledge users) worldwide. This has led to the emergence of a new function of brokering between researchers and policymakers, and a new type of agent called Knowledge Broker. Understanding these complex multi-agent interactions is critical for an efficient knowledge brokering practice during any given policymaking process. Here we present 1) the current diversity of knowledge broker groups working in the field of biosecurity and environmental management; 2) the incentives linking the different agents involved in the process (knowledge producers, knowledge brokers and knowledge users), and 3) the gaps, needs and challenges to better understand this social ecosystem. We also propose alternatives aimed at improving transparency and efficiency, including future scenarios where the role of artificial intelligence (AI) technologies may become predominant in knowledge-brokering activities.
-
2184937.264264
Diagnosing patients with disorders of consciousness involves inductive risk: the risk of false negative and false positive results when gathering and interpreting evidence of consciousness. A recent proposal suggests mitigating that risk by incorporating patient values into methodological choices at the level of individual diagnostic techniques: when using machine-learning algorithms to detect neural evidence of responsiveness to commands, clinicians should consider the patient’s own preferences about whether avoiding false positives or false negatives takes priority (Birch, 2023). In this paper, I argue that this proposal raises concerns about how to ensure that inevitable non-epistemic value judgments do not outweigh epistemic considerations. Additionally, it comes with challenges related to the predictive accuracy of surrogate decision-makers and the decisional burden imposed on them. Hence, I argue that patient values should not be incorporated at the level of gathering evidence of consciousness, but that they should play the leading role when considering how to respond to that evidence.
-
2356281.264273
I confess that, when I allow myself to think about it, I am amazed that I understand so little about what it is we philosophers do. I believe I can distinguish good philosophical work from bad—I can recognize when philosophy is done well—but I do not have a clear understanding of what it is that I am recognizing, and when I try actually to say what our discipline does, my remarks turn out to be naive and crude, more like the groping efforts of a beginning student than like the contributions of an advanced scholar to the field. …
-
2444102.26428
I argue that the epistemic aim of scientific theorizing (EAST) is producing theories with the highest possible number and degree of theoretical virtues (call this “TV-EAST”). I trace TV- EAST’s logical empiricist origins and discuss its close connections to Kuhn’s and Laudan’s problem-solving accounts of the aim of science. Despite TV-EAST’s antirealist roots, I argue that if one adopts the realist view that EAST is finding true theories, one should also endorse TV-EAST. I then defend TV-EAST by showing that it addresses the challenges raised against using the “aim of science” metaphor and offers significant advantages over the realist account.