It is often argued that inferential role semantics (IRS) entails semantic holism as long as theorists fail to answer the question about which inferences, among the many, are meaning-constitutive. Since analyticity, as truth in virtue of meaning, is a widely dismissed notion in indicating which inferences determine meaning, it seems that holism follows. Semantic holism is often understood as facing problems with the stability of content and many usual explanations of communication. Thus, we should choose between giving up (...) IRS, to avoid these holistic entailments, and defending holism against this charge, to rescue IRS. I try to pursue the second goal by analyzing certain patterns of counterfactual reasoning. Wilfrid Sellars and Robert Brandom claim that, to defend IRS, content-constitutive inferences are those counterfactually robust. While it is difficult to assess the goodness of such a view, it nonetheless entails that counterfactually non-robust inferences (which I call “modally ruled out inferences”) are not content-constitutive. If this is true, and if we take certain remarks about the grasp of concepts on board, there is a way to restrict the scope of the holism entailed by IRS to the extent of reshaping problems with the stability of content. (shrink)

Inference to the Best Explanation (IBE) is widely criticized for being an unreliable form of ampliative inference – partly because the explanatory hypotheses we have considered at a given time may all be false, and partly because there is an asymmetry between the comparative judgment on which an IBE is based and the absolute verdict that IBE is meant to license. In this paper, I present a further reason to doubt the epistemic merits of IBE and argue that it motivates (...) moving to an inferential pattern in which IBE emerges as a degenerate limiting case. Since this inferential pattern is structurally similar to an argumentative strategy known as Inferential Robustness Analysis (IRA), it effectively combines the most attractive features of IBE and IRA into a unified approach to non-deductive inference. (shrink)

The authors examined cue competition effects in young children using the blicket detector paradigm, in which objects are placed either singly or in pairs on a novel machine and children must judge which objects have the causal power to make the machine work. Cue competition effects were found in a 5- to 6-year-old group but not in a 4-year-old group. Equivalent levels of forward and backward blocking were found in the former group. Children's counterfactual judgments were subsequently examined by asking (...) whether or not the machine would have gone off in the absence of I of 2 objects that had been placed on it as a pair. Cue competition effects were demonstrated only in 5- to 6-year-olds using this mode of assessing causal reasoning. (shrink)

Functional robustness refers to a system’s ability to maintain a function in the face of perturbations to the causal structures that support performance of that function. Modularity, a crucial element of standard methods of causal inference and difference-making accounts of causation, refers to the independent manipulability of causal relationships within a system. Functional robustness appears to be at odds with modularity. If a function is maintained despite manipulation of some causal structure that supports that function, then the relationship between that (...) structure and function fails to be manipulable independent of other causal relationships within the system. Contrary to this line of reasoning, I argue that functional robustness often attends feedback control, rather than failures of modularity. Feedback control poses its own challenges to causal explanation and inference, but those challenges do not undermine modularity—and indeed, modularity is crucial to grappling with them. (shrink)

The paper presents and discusses the Hubble tension with respect to recent results in cosmology. I shall argue that the measurements from the James Webb Space Telescope and TRGB stars calibrations allow us to infer that the estimates of H0 with late universe methods are robust. Building on from robustness analysis, I conclude that the resolution of the tension cannot be expected to come from new systematics, but rather from new physics.

To study climate change, scientists employ computer models, which approximate target systems with various levels of skill. Given the imperfection of climate models, how do scientists use simulations to generate knowledge about the causes of observed climate change? Addressing a similar question in the context of biological modelling, Levins (1966) proposed an account grounded in robustness analysis. Recent philosophical discussions dispute the confirmatory power of robustness, raising the question of how the results of computer modelling studies contribute to the body (...) of evidence supporting hypotheses about climate change. Expanding on Staley’s (2004) distinction between evidential strength and security, and Lloyd’s (2015) argument connecting variety-of-evidence inferences and robustness analysis, I address this question with respect to recent challenges to the epistemology robustness analysis. Applying this epistemology to case studies of climate change, I argue that, despite imperfections in climate models, and epistemic constraints on variety-of-evidence reasoning and robustness analysis, this framework accounts for the strength and security of evidence supporting climatological inferences, including the finding that global warming is occurring and its primary causes are anthropogenic. (shrink)

I propose a distinct type of robustness, which I suggest can support a confirmatory role in scientific reasoning, contrary to the usual philosophical claims. In model robustness, repeated production of the empirically successful model prediction or retrodiction against a background of independentlysupported and varying model constructions, within a group of models containing a shared causal factor, may suggest how confident we can be in the causal factor and predictions/retrodictions, especially once supported by a variety of evidence framework. I present climate (...) models of greenhouse gas global warming of the 20th Century as an example, and emphasize climate scientists’ discussions of robust models and causal aspects. The account is intended as applicable to a broad array of sciences that use complex modeling techniques. (shrink)

This paper is concerned with counterfactual logic and its implications for the modal status of mathematical claims. It is most directly a response to an ambitious program by Yli-Vakkuri and Hawthorne (2018), who seek to establish that mathematics is committed to its own necessity. I claim that their argument fails to establish this result for two reasons. First, their assumptions force our hand on a controversial debate within counterfactual logic. In particular, they license counterfactual strengthening— the inference from ‘If A (...) were true then C would be true’ to ‘If A and B were true then C would be true’—which many reject. Second, the system they develop is provably equivalent to appending Deduction Theorem to a T modal logic. It is unsurprising that the combination of Deduction Theorem with T results in necessitation; indeed, it is precisely for this reason that many logicians reject Deduction Theorem in modal contexts. If Deduction Theorem is unacceptable for modal logic, it cannot be assumed to derive the necessity of mathematics. (shrink)

A widely accepted claim about counterfactuals is that they differ from strict conditionals, that is, there is no adequate representation of them as sentences of the form   . To justify this claim, Stalnaker and Lewis have argued that some fallacious inferences would turn out valid if counterfactuals were so represented. However, their argument has a flaw, as it rests on a questionable assumption about the relation between surface grammar and logical form. Without that assumption, no consequence of (...) the alleged kind is obtained, hence the claim may be rejected. (shrink)

Counterfactuals are somewhat tolerant. Had Socrates been at least six feet tall, he need not have been exactly six feet tall. He might have been a little taller—he might have been six one or six two. But while he might have been a little taller, there are limits to how tall he would have been. Had he been at least six feet tall, he would not have been more than a hundred feet tall, for example. Counterfactuals are not just tolerant, (...) then, but bounded. This paper presents a surprising paradox: If counterfactuals are tolerant and bounded, then we can prove a flat contradiction using natural rules of inference. Something has to go then. But what? (shrink)

Abstract Timothy Williamson has recently proposed to undermine modal skepticism by appealing to the reducibility of modal to counterfactual logic ( Reducibility ). Central to Williamson’s strategy is the claim that use of the same non-deductive mode of inference ( counterfactual development , or CD ) whereby we typically arrive at knowledge of counterfactuals suffices for arriving at knowledge of metaphysical necessity via Reducibility. Granting Reducibility, I ask whether the use of CD plays any essential role in a Reducibility-based reply (...) to two kinds of modal skepticism. I argue that its use is entirely dispensable, and that Reducibility makes available replies to modal skeptics which show certain propositions to be metaphysically necessary by deductive arguments from premises the modal skeptic accepts can be known. Content Type Journal Article Pages 1-19 DOI 10.1007/s11098-011-9784-4 Authors Juhani Yli-Vakkuri, Wolfson College, Oxford University, Oxford, OX2 6UD UK Journal Philosophical Studies Online ISSN 1573-0883 Print ISSN 0031-8116. (shrink)

When people want to identify the causes of an event, assign credit or blame, or learn from their mistakes, they often reflect on how things could have gone differently. In this kind of reasoning, one considers a counterfactual world in which some events are different from their real-world counterparts and considers what else would have changed. Researchers have recently proposed several probabilistic models that aim to capture how people do (or should) reason about counterfactuals. We present a new model and (...) show that it accounts better for human inferences than several alternative models. Our model builds on the work of Pearl (2000), and extends his approach in a way that accommodates backtracking inferences and that acknowledges the difference between counterfactual interventions and counterfactual observations. We present six new experiments and analyze data from four experiments carried out by Rips (2010), and the results suggest that the new model provides an accurate account of both mean human judgments and the judgments of individuals. (shrink)

This paper is about two controversial inference-patterns involving counterfactual or subjunctive conditionals. Given a plausible assumption about the truth-conditions of counterfactuals, it is shown that one can't go wrong in applying hypothetical syllogism (i.e., transitivity) so long as the set of worlds relevant for the conclusion is a subset of the sets of worlds relevant for the premises. It is also shown that one can't go wrong in applying antecedent strengthening so long as the set of worlds relevant for the (...) conclusion is a subset of that for the premise. These results are then adapted to Lewis's theory of counterfactuals. (shrink)

As the opening line of Dickens' classic novel suggests, it is very often the case that people can imagine both better and worse alternatives to their present reality. Although Dickens was writing about events that occurred over two centuries ago, it remains just as true today that we clearly live in neither the best nor the worst of possible worlds. For instance, we can wish for the amelioration of present difficulties in the Middle East yet still take comfort in the (...) fact that the threat of nuclear war has been greatly reduced since the end of the Cold War. On a more mundane level, it is easy for us to imagine how various aspects of our lives, such as our jobs, marriages, or physical fitness, could be both better and worse. Undoubtedly due to the pervasiveness and intrinsically fascinating qualities of this phenomenon of imagining alternatives to reality, there has been a veritable explosion of research in recent years on what have been termed mental simulation and counterfactual thinking processes. This chapter speculates upon and provides evidence for the antecedents and consequences of two types of counterfactuals: upward (it could have been better) and downward (it could have been worse). The importance of studying the antecedents and consequences of counterfactual direction lie in the differential structures, emotions, intentions, inferences, and behavioral consequences of upward versus downward counterfactual thinking. (shrink)

Causal Modeling Semantics (CMS, e.g., Galles and Pearl 1998; Pearl 2000; Halpern 2000) is a powerful framework for evaluating counterfactuals whose antecedent is a conjunction of atomic formulas. We extend CMS to an evaluation of the probability of counterfactuals with disjunctive antecedents, and more generally, to counterfactuals whose antecedent is an arbitrary Boolean combination of atomic formulas. Our main idea is to assign a probability to a counterfactual (A ∨ B) > C at a causal model M as a weighted (...) average of the probability of C in those submodels that truthmake A∨ B (Briggs 2012; Fine 2016, 2017). The weights of the submodels are given by the inverse distance to the original model M, based on a distance metric proposed by Eva, Stern, and Hartmann (2019). Apart from solving a major problem in the epistemology of counterfactuals, our paper shows how work in semantics, causal inference and formal epistemology can be fruitfully combined. (shrink)

This study analyses the predictions of the General Theory of Relativity (GTR) against a slightly modified version of the standard central mass solution (Schwarzschild solution). It is applied to central gravity in the solar system, the Pioneer spacecraft anomalies (which GTR fails to predict correctly), and planetary orbit distances and times, etc (where GTR is thought consistent.) -/- The modified gravity equation was motivated by a theory originally called ‘TFP’ (Time Flow Physics, 2004). This is now replaced by the ‘Geometric (...) Model’, 2014 [20], which retains the same theory of gravity. This analysis is offered partially as supporting detail for the claim in [20] that the theory is realistic in the solar system and explains the Pioneer anomalies. The overall conclusion is that the model can claim to explain the Pioneer anomalies, contingent on the analysis being independently verified and duplicated of course. -/- However the interest lies beyond testing this theory. To start with, it gives us a realistic scale on which gravity might vary from the accepted theory, remain consistent with most solar-scale astronomical observations. It is found here that the modified gravity equation would appear consistent with GTR for most phenomena, but it would retard the Pioneer spacecraft by about the observed amount (15 seconds or so at time). Hence it is a possible explanation of this anomaly, which as far as I know remains unexplained now for 20 years. -/- It also shows what many philosophers of science have emphasized: the pivotal role of counterfactual reasoning. By putting forward an exact alternative solution, and working through the full explanation, we discover a surprising ‘counterfactual paradox’: the modified theory slightly weakens GTR gravity – and yet the effect is to slow down the Pioneer trajectory, making it appear as if gravity is stronger than GTR. The inference that “there must be some tiny extra force…” (Musser, 1998 [1]) is wrong: there is a second option: “…or there may be a slightly weaker form of gravity than GTR.” . (shrink)

There is wide support in logic, philosophy, and psychology for the hypothesis that the probability of the indicative conditional of natural language, $P(\textit{if } A \textit{ then } B)$, is the conditional probability of $B$ given $A$, $P(B|A)$. We identify a conditional which is such that $P(\textit{if } A \textit{ then } B)= P(B|A)$ with de Finetti's conditional event, $B|A$. An objection to making this identification in the past was that it appeared unclear how to form compounds and iterations of (...) conditional events. In this paper, we illustrate how to overcome this objection with a probabilistic analysis, based on coherence, of these compounds and iterations. We interpret the compounds and iterations as conditional random quantities which, given some logical dependencies, may reduce to conditional events. We show how the inference to $B|A$ from $A$ and $B$ can be extended to compounds and iterations of both conditional events and biconditional events. Moreover, we determine the respective uncertainty propagation rules. Finally, we make some comments on extending our analysis to counterfactuals. (shrink)

In previous work, we studied four well known systems of qualitative probabilistic inference, and presented data from computer simulations in an attempt to illustrate the performance of the systems. These simulations evaluated the four systems in terms of their tendency to license inference to accurate and informative conclusions, given incomplete information about a randomly selected probability distribution. In our earlier work, the procedure used in generating the unknown probability distribution (representing the true stochastic state of the world) tended to yield (...) probability distributions with moderately high entropy levels. In the present article, we present data charting the performance of the four systems when reasoning in environments of various entropy levels. The results illustrate variations in the performance of the respective reasoning systems that derive from the entropy of the environment, and allow for a more inclusive assessment of the reliability and robustness of the four systems. (shrink)

It is widely held that counterfactuals, unlike attitude ascriptions, preserve the referential transparency of their constituents, i.e., that counterfactuals validate the substitution of identicals when their constituents do. The only putative counterexamples in the literature come from counterpossibles, i.e., counterfactuals with impossible antecedents. Advocates of counterpossibilism, i.e., the view that counterpossibles are not all vacuous, argue that counterpossibles can generate referential opacity. But in order to explain why most substitution inferences into counterfactuals seem valid, counterpossibilists also often maintain that (...) counterfactuals with possible antecedents are transparency‐preserving. I argue that if counterpossibles can generate opacity, then so can ordinary counterfactuals with possible antecedents. Utilizing an analogy between counterfactuals and attitude ascriptions, I provide a counterpossibilist‐friendly explanation for the apparent validity of substitution inferences into counterfactuals. I conclude by suggesting that the debate over counterpossibles is closely tied to questions concerning the extent to which counterfactuals are more like attitude ascriptions and epistemic operators than previously recognized. (shrink)

The use of counterfactuals for considerations of algorithmic fairness and explainability is gaining prominence within the machine learning community and industry. This paper argues for more caution with the use of counterfactuals when the facts to be considered are social categories such as race or gender. We review a broad body of papers from philosophy and social sciences on social ontology and the semantics of counterfactuals, and we conclude that the counterfactual approach in machine learning fairness and social explainability can (...) require an incoherent theory of what social categories are. Our findings suggest that most often the social categories may not admit counterfactual manipulation, and hence may not appropriately satisfy the demands for evaluating the truth or falsity of counterfactuals. This is important because the wide- spread use of counterfactuals in machine learning can lead to misleading results when applied in high-stakes domains. Accordingly, we argue that even though counterfactuals play an essential part in some causal inferences, their use for questions of algorithmic fairness and social explanations can create more problems than they resolve. Our positive result is a set of tenets about using counterfactuals for fairness and explanations in machine learning. (shrink)

Among philosophers working on modality, there is a common assumption that there is a strong connection between temporal possibility and counterfactual possibility. For example, Sydney Shoemaker 1998, 69 70) writes: It seems to me a general feature of our thought about possibility that how we think that something could have differed from how it in fact is [is] closely related to how we think that the way something is at one time could differ from the way that same thing is (...) at a different time. In possible worlds jargon, the ways one and the same thing of a given sort can differ across worlds correspond to the ways one and the same thing of that sort can differ at different times in the same world. Could I have been a plumber or an accountant instead of a philosopher? The answer seems to be yes -- and this goes with the fact that we acknowledge the possibility of a scenario in which someone who was exactly as I was at some point in my life undergoes a series of changes resulting in his eventually being a plumber or an accountant. In a footnote, he acknowledges that this connection between counterfactual and temporal possibility needs to be qualified in response to an objection by Randy Clarke: «The property of being the child of someone who has visited Paris is not one that one can have and then lose; but it is one that one can have in the actual world and not have in some other possible world.» (1998, 75n.9) Therefore, he restricts the inference from temporal to counterfactual possibility and vice versa to «non historical properties.». (shrink)

This paper articulates an account of causation as a collection of information-theoretic relationships between patterns instantiated in the causal nexus. I draw on Dennett’s account of real patterns to characterize potential causal relata as patterns with specific identification criteria and noise tolerance levels, and actual causal relata as those patterns instantiated at some spatiotemporal location in the rich causal nexus as originally developed by Salmon. I develop a representation framework using phase space to precisely characterize causal relata, including their degree (...) of counterfactual robustness, causal profiles, causal connectivity, and privileged grain size. By doing so, I show how the philosophical notion of causation can be rendered in a format that is amenable for direct application of mathematical techniques from information theory such that the resulting informational measures are causal informational measures. This account provides a metaphysics of causation that supports interventionist semantics and causal modeling and discovery techniques. (shrink)

I present two counterexamples to the recently back-in-favour truth-tracking account of knowledge: one involving a true belief resting on a counterfactually robust delusion, one involving a true belief acquired alongside a bunch of false beliefs. These counterexamples carry over to a recent modification of the theory due to Briggs and Nolan (2012), and seem invulnerable to a recent defence of the theory against known counterexamples, by Adams and Clarke (2005).

As it has been shaped by improvements in its tools and methods, and by its discourse with critics, I describe how the astrological research program has advanced through three stages of modelling and design limitations. Single-factor tests (for example, the many Sun-sign–only experiments that have been published) are typically underdeterministic. Multi-factor tests, unless they are very well designed, can easily become overdeterministic. Chart-matching tests have been vulnerable to confirmation bias errors until the development of a machine-based, whole-chart matching protocol that (...) has objectively produced evidence of high effect-sizes. A meta-analysis of recent results shows the rapid advancement and how to further improve the results. The value of the program is not only to corroborate the taxonomic counterfactuals of astrological “cookbooks,” but to extend their explanatory reach by the comparison of astrological postulates and inferences with philosophies in other disciplines in terms of quantifiable processes and emergent effects. (shrink)

Mental simulation refers to the imagination of alternative, counterfactual realities. This chapter provides an overview of research on simulations of the past— retrospective simulation—and simulations of the future— prospective simulation. Two major themes run throughout. The first is that both retrospective and prospective thinking are inextricably linked, relying on a mixture of episodic and semantic memories that share common neural substrates. The second is that retrospective and prospective simulation present trade-offs for the individual. On the one hand, they are functional, (...) identifying causal inferences and potential obstacles that prepare us to try harder and perform better in the future. On the other hand, they sometimes produce bias—retrospective thinking can enhance hindsight bias, whereas prospective thinking can evoke biased predictions about the likelihood of certain events as well as how one is going to feel about those events in the future. Fortunately, however, these biases can be diminished with debiasing techniques. (shrink)

As our epistemic ambitions grow, the common and scientific endeavours are becoming increasingly dependent on Machine Learning (ML). The field rests on a single experimental paradigm, which consists of splitting the available data into a training and testing set and using the latter to measure how well the trained ML model generalises to unseen samples. If the model reaches acceptable accuracy, an a posteriori contract comes into effect between humans and the model, supposedly allowing its deployment to target environments. Yet (...) the latter part of the contract depends on human inductive predictions or generalisations, which infer a uniformity between the trained ML model and the targets. The paper asks how we justify the contract between human and machine learning. It is argued that the justification becomes a pressing issue when we use ML to reach ‘elsewheres’ in space and time or deploy ML models in non-benign environments. The paper argues that the only viable version of the contract can be based on optimality (instead of on reliability which cannot be justified without circularity) and aligns this position with Schurz’s optimality justification. It is shown that when dealing with inaccessible/unstable ground-truths (‘elsewheres’ and non-benign targets), the optimality justification undergoes a slight change, which should reflect critically on our epistemic ambitions. Therefore, the study of ML robustness should involve not only heuristics that lead to acceptable accuracies on testing sets. The justification of human inductive predictions or generalisations about the uniformity between ML models and targets should be included as well. Without it, the assumptions about inductive risk minimisation in ML are not addressed in full. (shrink)

Considerable evidence proves that causal learning and causal understanding greatly enhance our ability to manipulate the physical world and are major factors that distinguish humans from other primates. How do we enable unintelligent robots to think causally, answer the questions raised with "why" and even understand the meaning of such questions? The solution is one of the keys to realizing artificial intelligence. Judea Pearl believes that to achieve human-like intelligence, researchers must start by imitating the intelligence of children, so he (...) proposed a "causal inference engine" to help future artificial intelligence make causal inference, pass the Minimal Turing Test, and even become a moral subject who can discern good from evil. This study attempts to provide some insights into the development of children's education from basic assumptions and construction goals of artificial intelligence, and to reflect on the causal model of artificial intelligence through children's education. (shrink)

In debates concerning the consequence argument, it has long been claimed that [McKay, T. J., and D. Johnson. 1996. “A Reconsideration of an Argument Against Compatibilism.” Philosophical Topics 24 (2): 113–122] demonstrated the invalidity of rule (β). Here, I argue that their result is not as robust as we might like to think. First, I argue that McKay and Johnson's counterexample is successful if one adopts a certain interpretation of ‘no choice about’ and if one is willing to deny (...) the conditional excluded middle principle. In order to make this point I demonstrate that (β) is valid on Stalnaker's theory of counterfactuals. This result is important and should not be neglected, I argue, because there is a particular line of objection to the revised formulations of the consequence argument that does not succeed against the original version. (shrink)

Philosophers have established that certain ethically important val- ues are modally robust in the sense that they systematically deliver correlative benefits across a range of counterfactual scenarios. In this paper, we contend that recourse – the systematic process of reversing unfavorable decisions by algorithms and bureaucracies across a range of counterfactual scenarios – is such a modally ro- bust good. In particular, we argue that two essential components of a good life – temporally extended agency and trust – are (...) under- written by recourse. We critique existing approaches to the conceptualization, op- erationalization and implementation of recourse. Based on these criticisms, we suggest a revised approach to recourse and give ex- amples of how it might be implemented – especially for those who are least well off. (shrink)

This paper concerns the recent revival of entity realism. Having been started with the work of Ian Hacking, Nancy Cartwright and Ronald Giere, the project of entity realism has recently been developed by Matthias Egg, Markus Eronen, and Bence Nanay. The paper opens a dialogue among these recent views on entity realism and integrates them into a more advanced view. The result is an epistemological criterion for reality: the property-tokens of a certain type may be taken as real insofar as (...) only they can be materially inferred from the evidence obtained in a variety of independent ways of detection. (shrink)

This note points out a conflict between some common intuitions about metaphysical possibility. On the one hand, it is appealing to deny that there are robust counterfactuals about how various physically impossible substances would interact with the matter that exists at our world. On the other hand, our intuitions about how concepts like MOUNTAIN apply at other metaphysically possible worlds seem to presuppose facts about ‘solidity’ which cash out in terms of these counterfactuals. I consider several simple attempts to (...) resolve this conflict and note they all fall short. (shrink)

This paper connects recent findings from experimental epistemology to several major themes in classical Indian epistemology. First, current evidence supports a specific account of the ordinary knowledge concept in contemporary anglophone American culture. According to this account, known as abilism, knowledge is a true representation produced by cognitive ability. I present evidence that abilism closely approximates Nyāya epistemology’s theory of knowledge, especially that found in the Nyāya-sūtra. Second, Americans are more willing to attribute knowledge of positive facts than of negative (...) facts, especially when such facts are inferred and even when the positive and negative “facts” are logically equivalent. Similar suspicions about knowledge of negative facts seemingly occur in classical Indian epistemology, suggesting that the asymmetry might not be an American quirk but instead reflect a cross-culturally robust tendency in knowledge attributions. Each of these themes—abilism and the positive/negative asymmetry—presents an exciting opportunity for further research in experimental cross-cultural epistemology. (shrink)

Conference paper presented at the 10th International Congress of Logic, Methodology, and Philosophy of Science, Florence, Italy (19-25 August 1995). Extrapolation here refers to the act of inferring more widely from a limited range of known facts. This notion of extrapolation, especially when applied to past events, has recently been used to formulate a pragmatic definition of truth. This paper shows that this definition has serious problems. The pragmatic definition of truth has been formulated in discussions on internal realism. In (...) this paper, the basic internal realist starting point, as expounded in the works of H. Putnam and N. Jardine, will be taken to be the claim that the distinction between judgements that are true from judgements that are false is explainable from within our judgement-forming faculty — no recourse to the external world is necessary. The paper first shows how the pragmatic definition of truth involves a counterfactual statement. It then describes how the counterfactual definition is apparently justified by an inductive argument. In the final step, this justification is shown to be unsuccessful. (shrink)

In the biomedical context, policy makers face a large amount of potentially discordant evidence from different sources. This prompts the question of how this evidence should be aggregated in the interests of best-informed policy recommendations. The starting point of our discussion is Hunter and Williams’ recent work on an automated aggregation method for medical evidence. Our negative claim is that it is far from clear what the relevant criteria for evaluating an evidence aggregator of this sort are. What is the (...) appropriate balance between explicitly coded algorithms and implicit reasoning involved, for instance, in the packaging of input evidence? In short: What is the optimal degree of ‘automation’? On the positive side: We propose the ability to perform an adequate robustness analysis as the focal criterion, primarily because it directs efforts to what is most important, namely, the structure of the algorithm and the appropriate extent of automation. Moreover, where there are resource constraints on the aggregation process, one must also consider what balance between volume of evidence and accuracy in the treatment of individual evidence best facilitates inference. There is no prerogative to aggregate the total evidence available if this would in fact reduce overall accuracy. (shrink)

Researchers often lack the necessary data to credibly estimate racial discrimination in policing. In particular, police administrative records lack information on civilians police observe but do not investigate. In this article, we show that if police racially discriminate when choosing whom to investigate, analyses using administrative records to estimate racial discrimination in police behavior are statistically biased, and many quantities of interest are unidentified—even among investigated individuals—absent strong and untestable assumptions. Using principal stratification in a causal mediation framework, we derive (...) the exact form of the statistical bias that results from traditional estimation. We develop a bias-correction procedure and nonparametric sharp bounds for race effects, replicate published findings, and show the traditional estimator can severely underestimate levels of racially biased policing or mask discrimination entirely. We conclude by outlining a general and feasible design for future studies that is robust to this inferential snare. (shrink)

The book is a philosophical refection on the possibility of mathematical history. Are poosible models of historical phenomena so exact as those of physical ones? Mathematical models borrowed from quantum mechanics by the meditation of its interpretations are accomodated to history. The conjecture of many-variant history, alternative history, or counterfactual history is necessary for mathematical history. Conclusions about philosophy of history are inferred.

It has been stated that the notion of cause and effect is one object of study that sciences and engineering revolve around. Lately, in software engineering, diagrammatic causal inference methods (e.g., Pearl’s model) have gained popularity (e.g., analyzing causes and effects of change in software requirement development). This paper concerns diagrammatical (graphic) models of causal relationships. Specifically, we experiment with using the conceptual language of thinging machines (TMs) as a tool in this context. This would benefit works on causal relationships (...) in requirements engineering, enhance our understanding of the TM modeling, and contribute to the study of the philosophical notion of causality. To specify the causality in a system’s description is to constrain the system’s behavior and thus exclude some possible chronologies of events. The notion of causality has been studied based on tools to express causal questions in diagrammatic and algebraic forms. Causal models deploy diagrammatic models, structural equations, and counterfactual and interventional logic. Diagrammatic models serve as a language for representing what we know about the world. The research methodology in the paper focuses on converting causal graphs into TM models and contrasts the two types of representation. The results show that the TM depiction of causality is more complete and therefore can provide a foundation for causal graphs. (shrink)

Many tasks in statistical and causal inference can be construed as problems of entailment in a suitable formal language. We ask whether those problems are more difficult, from a computational perspective, for causal probabilistic languages than for pure probabilistic (or “associational”) languages. Despite several senses in which causal reasoning is indeed more complex—both expressively and inferentially—we show that causal entailment (or satisfiability) problems can be systematically and robustly reduced to purely probabilistic problems. Thus there is no jump in computational complexity. (...) Along the way we answer several open problems concerning the complexity of well-known probability logics, in particular demonstrating the ${\exists \mathbb {R}}$ -completeness of a polynomial probability calculus, as well as a seemingly much simpler system, the logic of comparative conditional probability. (shrink)

Much forensic inference based upon DNA evidence is made assuming Hardy-Weinberg Equilibrium (HWE) for the genetic loci being used. Several statistical tests to detect and measure deviation from HWE have been devised, and their limitations become more obvious when testing for deviation within multiallelic DNA loci. The most popular methods-Chi-square and Likelihood-ratio tests-are based on asymptotic results and cannot guarantee a good performance in the presence of low frequency genotypes. Since the parameter space dimension increases at a quadratic rate on (...) the number of alleles, some authors suggest applying sequential methods, where the multiallelic case is reformulated as a sequence of “biallelic” tests. However, in this approach it is not obvious how to assess the general evidence of the original hypothesis; nor is it clear how to establish the significance level for its acceptance/rejection. In this work, we introduce a straightforward method for the multiallelic HWE test, which overcomes the aforementioned issues of sequential methods. The core theory for the proposed method is given by the Full Bayesian Significance Test (FBST), an intuitive Bayesian approach which does not assign positive probabilities to zero measure sets when testing sharp hypotheses. We compare FBST performance to Chi-square, Likelihood-ratio and Markov chain tests, in three numerical experiments. The results suggest that FBST is a robust and high performance method for the HWE test, even in the presence of several alleles and small sample sizes. (shrink)

What is the connection between valid inference and true conditionals? Many conditional logics require that when A is a logical consequence of B, "if B then A" is true. Taking counterlogical conditionals seriously leads to systems that permit counterexamples to that general rule. However, this leaves those of us who endorse non-trivial accounts of counterpossible conditionals to explain what the connection between conditionals and consequence is. The explanation of the connection also answers a common line of objection to non-trivial counterpossibles, (...) which is based on a transition from valid arguments to the corresponding conditionals. It also contributes to the wider project of illuminating the connections between contexts of utterance, on the one hand, and the truth-conditions of conditionals uttered in those contexts, on the other. (shrink)

Despite widespread evidence that fictional models play an explanatory role in science, resistance remains to the idea that fictions can explain. A central source of this resistance is a particular view about what explanations are, namely, the ontic conception of explanation. According to the ontic conception, explanations just are the concrete entities in the world. I argue this conception is ultimately incoherent and that even a weaker version of the ontic conception fails. Fictional models can succeed in offering genuine explanations (...) by correctly capturing relevant patterns of counterfactual dependence and licensing correct inferences. Using the example of Newtonian force explanations of the tides, I show how, even in science, fiction can be a vehicle for truth. (shrink)

Do people tend to be overconfident? Many think so. They’ve run studies on whether people are calibrated: whether their average confidence in their opinions matches the proportion of those opinions that are true. Under certain conditions, people are systematically ‘over-calibrated’—for example, of the opinions they’re 80% confident in, only 60% are true. From this empirical over-calibration, it’s inferred that people are irrationally overconfident. My question: When and why is this inference warranted? Answering it requires articulating a general connection between being (...) rational and being right—something extant studies have not done. I show how to do so using the notion of deference. This provides a theoretical foundation for calibration research, but also reveals a flaw: the connection between being rational and being right is much weaker than is standardly assumed, since rational people can often be expected to be miscalibrated. Thus we can’t test whether people are overconfident by simply testing whether they are over-calibrated; instead, we must try to predict the rational deviations from calibration, and then compare those predictions to people’s performance. I show how this can be done—and that doing so complicates the interpretation of robust empirical effects. (shrink)

Stephen Jay Gould argued that replaying the ‘tape of life’ would result in radically different evolutionary outcomes. Recently, biologists and philosophers of science have paid increasing attention to the theoretical importance of convergent evolution—the independent origination of similar biological forms and functions—which many interpret as evidence against Gould’s thesis. In this paper, we examine the evidentiary relevance of convergent evolution for the radical contingency debate. We show that under the right conditions, episodes of convergent evolution can constitute valid natural experiments (...) that support inferences regarding the deep counterfactual stability of macroevolutionary outcomes. However, we argue that proponents of convergence have problematically lumped causally heterogeneous phenomena into a single evidentiary basket, in effect treating all convergent events as if they are of equivalent theoretical import. As a result, the ‘critique from convergent evolution’ fails to engage with key claims of the radical contingency thesis. To remedy this, we develop ways to break down the heterogeneous set of convergent events based on the nature of the generalizations they support. Adopting this more nuanced approach to convergent evolution allows us to differentiate iterated evolutionary outcomes that are probably common among alternative evolutionary histories and subject to law-like generalizations, from those that do little to undermine and may even support, the Gouldian view of life. (shrink)

Semantic dispositionalism is roughly the view that meaning a certain thing by a word, or possessing a certain concept, consists in being disposed to do something, e.g., infer a certain way. Its main problem is that it seems to have so many and disparate exceptions. People can fail to infer as required due to lack of logical acumen, intoxication, confusion, deviant theories, neural malfunctioning, and so on. I present a theory stating possession conditions of concepts that are counterfactuals, rather than (...) disposition attributions, but which is otherwise similar to inferentialist versions of dispositionalism. I argue that it can handle all the exceptions discussed in the literature without recourse to ceteris paribus clauses. Psychological exceptions are handled by suitably undemanding requirements (unlike that of giving the sum of any two numbers) and by setting the following two preconditions upon someone’s making the inference: that she considers the inference and has no motivating reason against it. The non-psychological exceptions, i.e., cases of neural malfunctioning, are handled by requiring that the counterfactuals be true sufficiently often during the relevant interval. I argue that this accommodates some important intuitions about concept possession, in particular, the intuition that concept possession is vague along a certain dimension. (shrink)

It is often suggested that disagreement among scientific experts is a reason not to trust those experts, even about matters on which they are in agreement. In direct opposition to this view, I argue here that the very fact that there is disagreement among experts on a given issue provides a positive reason for non-experts to trust that the experts really are justified in their attitudes towards consensus theories. I show how this line of thought can be spelled out in (...) three distinct frameworks for non-deductive reasoning: namely, Bayesian Confirmation Theory, Inference to the Best Explanation, and Inferential Robustness Analysis. (shrink)

This article reviews research on counterfactual, social, and temporal comparisons and proposes a Reflection and Evaluation Model (REM) as an organizing framework. At the heart of the model is the assertion that 2 psychologically distinct modes of mental simulation operate during comparative thinking: reflection, an experiential (“as if”) mode of thinking characterized by vividly simulating that information about the comparison standard is true of, or part of, the self; and evaluation, an evaluative mode of thinking characterized by the use of (...) information about the standard as a reference point against which to evaluate one’s present standing. Reflection occurs when information about the standard is included in one’s self-construal, and evaluation occurs when such information is excluded. The result of reflection is that standard–consistent cognitions about the self become highly accessible, thereby yielding affective assimilation; whereas the result of evaluation is that comparison information is used as a standard against which one’s present standing is evaluated, thereby yielding affective contrast. The resulting affect leads to either an increase or decrease in behavioral persistence as a function of the type of task with which one is engaged, and a combination of comparison-derived causal inferences and regulatory focus strategies direct one toward adopting specific future action plans. (shrink)

A minimal essentialism (‘intrinsic biological essentialism’) about natural kinds is required to explain the projectability of human science terms. Human classifications that yield robust and ampliative projectable inferences refer to biological kinds. I articulate this argument with reference to an intrinsic essentialist account of HPC kinds. This account implies that human sciences (e.g., medicine, psychiatry) that aim to formulate predictive kind categories should classify biological kinds. Issues concerning psychiatric classification and pluralism are examined.

Morgenbesser's Coin is a thought experiment that exemplifies a widespread disposition to infer counterfactual independence from causal independence. I argue that this disposition is mistaken by analysing a closely related thought experiment.

The standard model of cosmology is acclaimed in physics as accurate, robust, well-tested, our best scientific theory of the cosmos, but it has had serious anomalies for a while, including the Hubble tension, anomalous galaxies, and the completely unexplained nature of dark energy and dark matter. And lurking behind it all is the lack of a unified theory: General Relativity (GR) and quantum mechanics (QM) are inconsistent. Now startling new observations by the James Webb Space Telescope (JWST) in 2022 (...) of the early universe present the strongest challenge yet to the standard model, and whispers have started that this shows there is something wrong with the fundamental theory, General Relativity itself. This would be a crisis for cosmology. But haven’t they tested this theory already, and shown it is correct? How could it turn out wrong at this late stage? Here we compare the standard cosmology with an alternative fundamental theory, that has a strikingly different overall cosmological behavior: a simple cyclic expansion function. It is simple and deterministic. There are only two or three general parameters. The interesting result is that this alternative cosmology: (A) closely matches the expansion observed and modelled through the CDM standard model, now going back to red-shifts of 5-15; and (B) it also predicts unexpected early galaxy formation now being reported by the JWST. The point here is not to try to prove this alternative theory however, but rather show how it compares to the conventional cosmology. This show us clearly how weak the empirical evidence for the standard model really is against a counterfactual fundamental theory. Some results established in science are robust against theory change, but we find the standard cosmological model and the implications drawn from it are not robust at all. (shrink)

Despite its many advantages as a metaethical theory, moral expressivism faces difficulties as a semantic theory of the meaning of moral claims, an issue underscored by the notorious Frege-Geach problem. I consider a distinct metaethical view, inferentialism, which like expressivism rejects a representational account of meaning, but unlike expressivism explains meaning in terms of inferential role instead of expressive function. Drawing on Michael Williams’ recent work on inferential theories of meaning, I argue that an appropriate understanding of the pragmatic role (...) of moral discourse—the facilitation of coordinated social behavior—suggests the kind of inferences we should expect terms with this function to license. I offer a sketch of the inferential roles the moral ‘ought’ plays, and argue that if we accept that the relevant inferential roles are meaning-constitutive, we will be in a position to solve the Frege-Geach problem. Such an inferentialist solution has advantages over those forwarded by expressivists such as Blackburn and Gibbard. First, it offers a more straightforward explanation of the meaning of moral terms. It also gives simple answers to at least two semantic worries that have vexed contemporary expressivists—the “problem of permissions” and the commitment to “mentalism”, both of which I argue are problems that don’t get traction with an inferentialist approach. I conclude by considering ways in which this approach can be expanded into a more robust semantic account. (shrink)