Neuroscience is a laboratory-based science that spans multiple levels of analysis from molecular genetics to behavior. At every level of analysis experiments are designed in order to answer empirical questions about phenomena of interest. Understanding the nature and structure of experimentation in neuroscience is fundamental for assessing the quality of the evidence produced by such experiments and the kinds of claims that are warranted by the data. This article provides a general conceptual framework for thinking about evidence and experimentation in (...) neuroscience with a particular focus on two research areas: cognitive neuroscience and cognitive neurobiology. (shrink)

Selectionist evolutionary theory has often been faulted for not making novel predictions that are surprising, risky, and correct. I argue that it in fact exhibits the theoretical virtue of predictive capacity in addition to two other virtues: explanatory unification and model fitting. Two case studies show the predictive capacity of selectionist evolutionary theory: parallel evolutionary change in E. coli, and the origin of eukaryotic cells through endosymbiosis.

The issues of double-counting, use-constructing, and selection effects have long been the subject of debate in the philosophical as well as statistical literature. I have argued that it is the severity, stringency, or probativeness of the test—or lack of it—that should determine if a double-use of data is admissible. Hitchcock and Sober ([2004]) question whether this ‘severity criterion' can perform its intended job. I argue that their criticisms stem from a flawed interpretation of the severity criterion. Taking their criticism as (...) a springboard, I elucidate some of the central examples that have long been controversial, and clarify how the severity criterion is properly applied to them. Severity and Use-Constructing: Four Points (and Some Clarificatory Notes) 1.1 Point 1: Getting beyond ‘all or nothing’ standpoints 1.2 Point 2: The rationale for prohibiting double-counting is the requirement that tests be severe 1.3 Point 3: Evaluate severity of a test T by its associated construction rule R 1.4 Point 4: The ease of passing vs. ease of erroneous passing: Statistical vs. ‘Definitional’ probability The False Dilemma: Hitchcock and Sober 2.1 Marsha measures her desk reliably 2.2 A false dilemma Canonical Errors of Inference 3.1 How construction rules may alter the error-probing performance of tests 3.2 Rules for accounting for anomalies 3.3 Hunting for statistically significant differences Concluding Remarks. (shrink)

In this paper I comment on a recent paper by [Scerri, E., & Worrall, J. . Prediction and the periodic table. Studies in History and Philosophy of Science, 32, 407–452.] about the role temporally novel and use-novel predictions played in the acceptance of Mendeleev’s periodic table after the proposal of the latter in 1869. Scerri and Worrall allege that whereas temporally novel predictions—despite Brush’s earlier claim to the contrary—did not carry any special epistemic weight, use-novel predictions did indeed contribute to (...) the acceptance of the table. Although I agree with their first claim, I disagree with their second. In order to spell out my disagreement, I not only revisit Scerri and Worrall’s interpretation of crucial historical evidence they have cited in support of the ‘heuristic account’ of use-novel predictions, but I also criticise the latter on general grounds.Keywords: Periodic table; Dmitri Mendeleev; Noble gases; Use-novel predictions; Heuristic account; Ad hoc hypotheses. (shrink)

the virtues which are desirable for scientific theories to possess. In this paper I discuss the several species of theoretical virtues called 'fertility', and argue in each case that the desirability of 'fertility' can be explicated in terms of other, more fundamental theoretical virtues.

Despite the widespread use of key concepts of the Neyman–Pearson (N–P) statistical paradigm—type I and II errors, significance levels, power, confidence levels—they have been the subject of philosophical controversy and debate for over 60 years. Both current and long-standing problems of N–P tests stem from unclarity and confusion, even among N–P adherents, as to how a test's (pre-data) error probabilities are to be used for (post-data) inductive inference as opposed to inductive behavior. We argue that the relevance of error probabilities (...) is to ensure that only statistical hypotheses that have passed severe or probative tests are inferred from the data. The severity criterion supplies a meta-statistical principle for evaluating proposed statistical inferences, avoiding classic fallacies from tests that are overly sensitive, as well as those not sensitive enough to particular errors and discrepancies. Introduction and overview 1.1 Behavioristic and inferential rationales for Neyman–Pearson (N–P) tests 1.2 Severity rationale: induction as severe testing 1.3 Severity as a meta-statistical concept: three required restrictions on the N–P paradigm Error statistical tests from the severity perspective 2.1 N–P test T(): type I, II error probabilities and power 2.2 Specifying test T() using p-values Neyman's post-data use of power 3.1 Neyman: does failure to reject H warrant confirming H? Severe testing as a basic concept for an adequate post-data inference 4.1 The severity interpretation of acceptance (SIA) for test T() 4.2 The fallacy of acceptance (i.e., an insignificant difference): Ms Rosy 4.3 Severity and power Fallacy of rejection: statistical vs. substantive significance 5.1 Taking a rejection of H0 as evidence for a substantive claim or theory 5.2 A statistically significant difference from H0 may fail to indicate a substantively important magnitude 5.3 Principle for the severity interpretation of a rejection (SIR) 5.4 Comparing significant results with different sample sizes in T(): large n problem 5.5 General testing rules for T(), using the severe testing concept The severe testing concept and confidence intervals 6.1 Dualities between one and two-sided intervals and tests 6.2 Avoiding shortcomings of confidence intervals Beyond the N–P paradigm: pure significance, and misspecification tests Concluding comments: have we shown severity to be a basic concept in a N–P philosophy of induction? (shrink)

Ian Hacking instigated a revolution in 20th century philosophy of science by putting experiments (“interventions”) at the top of a philosophical agenda that historically had focused nearly exclusively on representations (“theories”). In this paper, I focus on a set of conceptual tools Hacking (1992) put forward to understand how laboratory sciences become stable and to explain what such stability meant for the prospects of unity of science and kind discovery in experimental science. I first use Hacking’s tools to understand sources (...) of instability and disunity in rodent behavioral neuroscience. I then use them to understand recent grass-roots collaborative open science initiatives aimed at establishing stability in this research area and tease out some implications for unity of science and kind creation and discovery in cognitive neuroscience. -/- . (shrink)

I offer an explicit account of the underdetermination thesis as well as of the many challenges it poses to scientific realism; a way to answer to these challenges is explored and outlined, by shifting attention to the content of theories. I argue that, even if we have solid grounds (as I contend we do) to support that some varieties of the underdetermination thesis are true, scientific realism can still offer an adequate picture of the aims and achievements of science.

Given the plethora of competing logical theories of validity available, it’s understandable that there has been a marked increase in interest in logical epistemology within the literature. If we are to choose between these logical theories, we require a good understanding of the suitable criteria we ought to judge according to. However, so far there’s been a lack of appreciation of how logical practice could support an epistemology of logic. This paper aims to correct that error, by arguing for a (...) practice-based approach to logical epistemology. By looking at the types of evidence logicians actually appeal to in attempting to support their theories, we can provide a more detailed and realistic picture of logical epistemology. To demonstrate the fruitfulness of a practice-based approach, we look to a particular case of logical argumentation—the dialetheist’s arguments based upon the self-referential paradoxes—and show that the evidence appealed to support a particular theory of logical epistemology, logical abductivism. (shrink)

The use of neuroscientific evidence in criminal trials has been steadily increasing. Despite progress made in recent decades in understanding the mechanisms of psychological and behavioral functioning, neuroscience is still in an early stage of development and its potential for influencing legal decision-making is highly contentious. Scholars disagree about whether or how neuroscientific evidence might impact prescriptions of criminal culpability, particularly in instances in which evidence of an accused’s history of mental illness or brain abnormality is offered to support a (...) plea of not criminally responsible. In the context of these debates, philosophers and legal scholars have identified numerous problems with admitting neuroscientific evidence in legal contexts. To date, however, less has been said about the challenges of evaluating the evidence upon which integrative mechanistic explanations that bring together evidence from different areas of neuroscience are based. As we explain, current criteria for evaluating such evidence to determine its admissibility in legal contexts are inadequate. Appealing to literature in the philosophy of scientific experimentation and theoretical work in the social, cognitive and behavioral sciences, we lay the groundwork for reforming these criteria and identify some of the implications of modifying them. (shrink)

Optogenetic techniques are described as “revolutionary” for the unprecedented causal control they allow neuroscientists to exert over neural activity in awake behaving animals. In this paper, I demonstrate by means of a case study that optogenetic techniques will only illuminate causal links between the brain and behavior to the extent that their error characteristics are known and, further, that determining these error characteristics requires comparison of optogenetic techniques with techniques having well known error characteristics and consideration of the broader neural (...) and behavioral context in which the targets of optogenetic interventions are situated. (shrink)

A common intuition about evidence is that if data x have been used to construct a hypothesis H, then x should not be used again in support of H. It is no surprise that x fits H, if H was deliberately constructed to accord with x. The question of when and why we should avoid such “double-counting” continues to be debated in philosophy and statistics. It arises as a prohibition against data mining, hunting for significance, tuning on the signal, and (...) ad hoc hypotheses, and as a preference for predesignated hypotheses and “surprising” predictions. I have argued that it is the severity or probativeness of the test—or lack of it—that should determine whether a double-use of data is admissible. I examine a number of surprising ambiguities and unexpected facts that continue to bedevil this debate. (shrink)

Descriptive accounts of the nature of explanation in neuroscience and the global goals of such explanation have recently proliferated in the philosophy of neuroscience and with them new understandings of the experimental practices of neuroscientists have emerged. In this paper, I consider two models of such practices; one that takes them to be reductive; another that takes them to be integrative. I investigate those areas of the neuroscience of learning and memory from which the examples used to substantiate these models (...) are culled, and argue that the multiplicity of experimental protocols used in these research areas presents specific challenges for both models. In my view, these challenges have been overlooked largely because philosophers have hitherto failed to pay sufficient attention to fundamental features of experimental practice. I demonstrate that when we do pay attention to such features, evidence for reduction and integrative unity in neuroscience is simply not borne out. I end by suggesting some new directions for the philosophy of neuroscience that pertain to taking a closer look at the nature of neuroscientific experiments. (shrink)

According to conventional wisdom, local gauge symmetry is not a symmetry of nature, but an artifact of how our theories represent nature. But a study of the so-called theta-vacuum appears to refute this view. The ground state of a quantized non-Abelian Yang-Mills gauge theory is characterized by a real-valued, dimensionless parameter theta—a fundamental new constant of nature. The structure of this vacuum state is often said to arise from a degeneracy of the vacuum of the corresponding classical theory, which degeneracy (...) allegedly arises from the fact that “large” (but not “small”) local gauge transformations connect physically distinct states of zero field energy. If that is right, then some local gauge transformations do generate empirical symmetries. In defending conventional wisdom against this challenge I hope to clarify the meaning of empirical symmetry while deepening our understanding of gauge transformations. I distinguish empirical from theoretical symmetries. Using Galileo’s ship and Faraday’s cube as illustrations, I say when an empirical symmetry is implied by a theoretical symmetry. I explain how the theta-vacuum arises, and how “large” gauge transformations differ from “small” ones. I then present two analogies from elementary quantum mechanics. By applying my analysis of the relation between empirical and theoretical symmetries, I show which analogy faithfully portrays the character of the vacuum state of a classical non-Abelian Yang-Mills gauge theory. The upshot is that “large” as well as “small” gauge transformations are purely formal symmetries of non-Abelian Yang-Mills gauge theories, whether classical or quantized. It is still worth distinguishing between these kinds of symmetries. An analysis of gauge within the constrained-Hamiltonian formalism yields the result that “large” gauge transformations should not be classified as gauge transformations; indeed, nor should “global” gauge transformations. In a theory in which boundary conditions are modeled dynamically, “global” gauge transformations may be associated with physical symmetries, corresponding to translations of these extra dynamical variables. Such translations are symmetries if and only if charge is conserved. But it is hard to argue that these symmetries are empirical, and in any case they do not correspond to any constant phase change in a quantum state. (shrink)

I argue that the Bayesian Way of reconstructing Duhem's problem fails to advance a solution to the problem of which of a group of hypotheses ought to be rejected or "blamed" when experiment disagrees with prediction. But scientists do regularly tackle and often enough solve Duhemian problems. When they do, they employ a logic and methodology which may be called error statistics. I discuss the key properties of this approach which enable it to split off the task of testing auxiliary (...) hypotheses from that of appraising a primary hypothesis. By discriminating patterns of error, this approach can at least block, if not also severely test, attempted explanations of an anomaly. I illustrate how this approach directs progress with Duhemian problems and explains how scientists actually grapple with them. (shrink)

According to a standard view, predictions of new phenomena provide stronger evidence for a theory than explanations of old ones. More guardedly, a theory that predicts phenomena that did not prompt the initial formulation of that theory is better supported by those phenomena than is a theory by known phenomena that generated the theory in the first place. So say various philosophers of science, including William Whewell (1847) in the 19th century and Karl Popper (1959) in the 20th, to mention (...) just two.Stephen Brush takes issue with this on historical grounds. In a series of fascinating papers he argues that generally speaking scientists do not regard the fact that a theory predicts new phenomena, even ones of a kind totally different from those that prompted the theory in the first place, as providing better evidential support for that theory than is provided by already known facts explained by the theory. (shrink)

Our paper studies the anatomy of the discovery of the Higgs boson at the Large Hadron Collider and its influence on the broader model landscape of particle physics. We investigate the phases of this discovery, which led to a crucial reconfiguration of the model landscape of elementary particle physics and eventually to a confirmation of the standard model. A keyword search of preprints covering the electroweak symmetry breaking sector of particle physics, along with an examination of physicists own understanding of (...) the discovery as documented in semiannual conferences, has allowed us an empirical investigation of its model dynamics. From our analyses we draw two main philosophical lessons concerning the nature of scientific reasoning in a complex experimental and theoretical environment. For one, from a confirmation standpoint, some SM alternatives could be considered even more confirmed by the Higgs discovery than the SM. Nevertheless, the SM largely remains the commonly accepted account of EWSB. We present criteria for comparing degrees of confirmation and expose some limits of a purely logical approach to understanding the Higgs discovery as a victory for the SM. Second, we understand the persistence of SM alternatives in the face of disfavourable evidence by borrowing the Lakatosian concept of a research programme, where the core idea behind a group of models survives, while other aspects adapt to incoming data. In order to apply this framework to the model landscape of EWSB, we must introduce a new category of research programme, the model-group, and we test its viability using the example of composite Higgs models. (shrink)

The main aim of this paper is to revisit the curve fitting problem using the reliability of inductive inference as a primary criterion for the ‘fittest' curve. Viewed from this perspective, it is argued that a crucial concern with the current framework for addressing the curve fitting problem is, on the one hand, the undue influence of the mathematical approximation perspective, and on the other, the insufficient attention paid to the statistical modeling aspects of the problem. Using goodness-of-fit as the (...) primary criterion for ‘best', the mathematical approximation perspective undermines the reliability of inference objective by giving rise to selection rules which pay insufficient attention to ‘accounting for the regularities in the data'. A more appropriate framework is offered by the error-statistical approach, where (i) statistical adequacy provides the criterion for assessing when a curve captures the regularities in the data adequately, and (ii) the relevant error probabilities can be used to assess the reliability of inductive inference. Broadly speaking, the fittest curve (statistically adequate) is not determined by the smallness if its residuals, tempered by simplicity or other pragmatic criteria, but by the nonsystematic (e.g. white noise) nature of its residuals. The advocated error-statistical arguments are illustrated by comparing the Kepler and Ptolemaic models on empirical grounds. ‡I am grateful to Deborah Mayo and Clark Glymour for many valuable suggestions and comments on an earlier draft of the paper; estimating the Ptolemaic model was the result of Glymour's prompting and encouragement. †To contact the author, please write to: Department of Economics, Virginia Tech 3019 Pamplin Hall (0316), Blacksburg, VA 24061; e-mail: [email protected]. (shrink)

Kuhn maintains that what marks the transition to a science is the ability to carry out ‘normal’ science—a practice he characterizes as abandoning the kind of testing that Popper lauds as the hallmark of science. Examining Kuhn's own contrast with Popper, I propose to recast Kuhnian normal science. Thus recast, it is seen to consist of severe and reliable tests of low-level experimental hypotheses (normal tests) and is, indeed, the place to look to demarcate science. While thereby vindicating Kuhn on (...) demarcation, my recasting of normal science is seen to tell against Kuhn's view of revolutionary science. (shrink)

In this paper, we examine possible formalisations of the riskiness condition for hypothesis testing. First, we informally introduce derivability and riskiness as testing conditions together with the corresponding arguments for refutation and confirmation. Then, we distinguish two different senses of confirmation and focus our discussion on one of them with the aid of a historical example. In the remaining sections, we offer a brief overview of the main references to the risky condition in the literature and scrutinise different options for (...) formally capturing riskiness; we show why none of them works. We conclude with some remarks about the relation between derivability and riskiness and claim that riskiness essentially involves a contextual, pragmatic component that eludes a complete formal reconstruction. (shrink)

Philosophers of science have recently begun to pay more attention to scientific practice, moving away from the discipline’s focus on theories. The creation of the Society for Philosophy of Science in Practice in 2006, as well as the emergence of scholarship on experimental practice (e.g. Sullivan 2009; 2010; 2016) as well as on the tools scientists use to construct explanations and theories (e.g. Feest 2011) all point to a disciplinary shift towards a more practice-conscious philosophy of science. In addition, scholars (...) are realizing the potential social relevance of philosophy of science and identifying the obstacles that stand in the way of realizing said potential (e.g. Douglas 2010). The aim of this dissertation is to propose a novel descriptive framework for philosophy of science, one that rests on the idea that scientific practice should be described as accurately as possible so that philosophical discussions of science rest on a solid descriptive base and can more easily be of use in engaging users and practitioners of science. The dissertation is organized along the stages of developing what I call the dynamic-iterative model of scientific practice. In chapter 1, I clarify the project’s philosophical background and outline six conditions of adequacy on a descriptive model of science. These conditions allow me to evaluate the three descriptive models reviewed in chapter 2. Chapter 3 combines Nickles’ problem-solving view of science (1981) with core insights from the heuristics and biases literature into a workable version of the dynamic-iterative model. I then apply the model to the case of place cell research in chapter 4 as a test of the model’s usefulness and applicability. In chapter 5 I evaluate the model with respect to my six conditions and assess its strengths and weaknesses. The dynamic-iterative model is a flexible descriptive framework that not only allows for more complete philosophical analyses of cases than other current models, but also supports important practical applications. (shrink)

"Partial realism" is a common position in the contemporary philosophy of science literature. It states that the "essential" elements of empirically successful scientific theories accurately represent corresponding features the world. This thesis makes several novel contributions related to this position. Firstly, it offers a new definition of the concept of “empirical success”, representing a principled merger between the use-novelty and unification accounts. Secondly, it provides a comparative critical analysis of various accounts of which elements are "essential" to the success of (...) a theory, including structural realism and the divide et impera strategy. A novel account of essentialness, entitled the “empirically successful sub-theory "account", is defended. Thirdly, it is argued that the realism/anti-realism debate should put to the side metaphysical questions and focus instead on partial realism's commitment to the continuity of science. Because this commitment lacks metaphysical implications, it is referred to as "deflationary realism". Anti-realists cannot reject deflationary realism as a matter of a priori principle; its overall viability can therefore only be assessed by a careful examination of the history of science. Finally, another consequence of partial realism, named "partial rationalism", is defended. Partial rationalism states that, in cases where several competing theories have been suggested, scientists are rational just in case they accept the essential elements of each of the scientific theories on offer. This novel position subverts the existing literature on scientific "revolutions", as it sometimes demands that scientists devise a synthesis between competing scientific theories, instead of "choosing" only one. The philosophical points defended in this thesis are illustrated and supported by case studies from the history of science, including Fresnel’s wave theory of light, the Copernican revolution, the "neo-Darwinian synthesis" in evolutionary biology, the "prion revolution" in molecular biology, the miasma theory of disease, and the chemical revolution. (shrink)

Neuroimaging technology is the most widely used tool to study human cognition. While originally a promising tool for mapping the content of cognitive theories onto the structures of the brain, recently developed tools for the analysis, handling and sharing of data have changed the theoretical landscape of cognitive neuroscience. Even with these advancements philosophical analyses of evidence in neuroimaging remain skeptical of the promise of neuroimaging technology. These views often treat the analysis techniques used to make sense of data produced (...) in a neuroimaging experiment as one, attributing the inferential limitations of analysis pipelines to the technology as a whole. Situated against the neuroscientists own critical assessment of their methods and the limitations of those methods, this skepticism appears based on a misunderstanding of the role data analysis techniques play in neuroimaging research. My project picks up here, examining how data analysis techniques, such as pattern classification analysis, are used to assess the evidential value of neuroimaging data. The project takes the form of three papers. In the first I identify the use of multiple data analysis techniques as an important aspect of the data interpretation process that is overlooked by critics. In the second I develop an account of inferences in neuroimaging research that is sensitive to this use of data analysis techniques, arguing that interpreting neuroimaging data is a process of isolating and explaining a variety of data patterns. In the third I argue that the development and uptake of new techniques for analyzing data must be accompanied by changes in research practices and standards of evidence if they are to promote knowledge generation. My approach to this work is both traditionally philosophical, insofar as it involves reading and analyzing the work of philosophers and neuroscientists, and embedded insofar as most of the research was conducted while engaged in attending lab meetings and participating in the work of those scientists whose work is the object of my research. (shrink)

Predictivism is the view that successful predictions of “novel” evidence carry more confirmational weight than accommodations of already known evidence. Novelty, in this context, has traditionally been conceived of as temporal novelty. However temporal predictivism has been criticized for lacking a rationale: why should the time order of theory and evidence matter? Instead, it has been proposed, novelty should be construed in terms of use-novelty, according to which evidence is novel if it was not used in the construction of a (...) theory. Only if evidence is use-novel can it fully support the theory entailing it. As I point out in this paper, the writings of the most influential proponent of use-novelty contain a weaker and a stronger version of use-novelty. However both versions, I argue, are problematic. With regard to the appraisal of Mendeleev’ periodic table, the most contentious historical case in the predictivism debate, I argue that temporal predictivism is indeed supported, although in ways not previously appreciated. On the basis of this case, I argue for a form of so-called symptomatic predictivism according to which temporally novel predictions carry more confirmational weight only insofar as they reveal the theory’s presumed coherence of facts as real. (shrink)

Maxwell claimed that the electrostatic inverse square law could be deduced from Cavendish's spherical condenser experiment. This is true only if the accuracy claims made by Cavendish and Maxwell are ignored, for both used the inverse square law as a premise in their analyses of experimental accuracy. By so doing, they assumed the very law the accuracy of which the Cavendish experiment was supposed to test. This paper attempts to make rational sense of this apparently circular procedure and to relate (...) it to some variants of traditional problems concerning old and new evidence. (shrink)

In recent years, several scientific disciplines have been undergoing replication crises, and in response, preregistration has been offered as a solution to replicability problems. In this paper, I will draw connections between this new focus on preregistration and an older debate in the philosophy of science, namely predictivism—the thesis that predictions are epistemically superior to accommodations. Specifically, I shall argue that predictivism justifies preregistration. As it turns out, predictivists of all stripes have subtly different reasons to support preregistration. This unity (...) is significant because firstly, preregistration proponents often seem to be implicitly committed to stronger versions of predictivism, but strong predictivism has long been deemed untenable by philosophers of science. Furthermore, the efficacy of preregistration in dealing with Questionable Research Practices (QRPs) like p-hacking and HARKing is at best contentious, which blocks a straightforward empirical justification of preregistration. Although empirical validation of preregistration—which need not be based on preventing QRPs—will eventually be required, it would be nice to have a principled justification for preregistration while we await the empirical evidence. I will also argue that preregistration offers something in return to predictivism: the former bolsters the latter by serving as a counterargument to accommodationism—the antithesis of predictivism. (shrink)

This article argues that common intuitions regarding (a) the specialness of ‘use-novel’ data for confirmation and (b) that this specialness implies the ‘no-double-counting rule’, which says that data used in ‘constructing’ (calibrating) a model cannot also play a role in confirming the model’s predictions, are too crude. The intuitions in question are pertinent in all the sciences, but we appeal to a climate science case study to illustrate what is at stake. Our strategy is to analyse the intuitive claims in (...) light of prominent accounts of confirmation of model predictions. We show that on the Bayesian account of confirmation, and also on the standard classical hypothesis-testing account, claims (a) and (b) are not generally true; but for some select cases, it is possible to distinguish data used for calibration from use-novel data, where only the latter confirm. The more specialized classical model-selection methods, on the other hand, uphold a nuanced version of claim (a), but this comes apart from (b), which must be rejected in favour of a more refined account of the relationship between calibration and confirmation. Thus, depending on the framework of confirmation, either the scope or the simplicity of the intuitive position must be revised. 1 Introduction2 A Climate Case Study3 The Bayesian Method vis-à-vis Intuitions4 Classical Tests vis-à-vis Intuitions5 Classical Model-Selection Methods vis-à-vis Intuitions 5.1 Introducing classical model-selection methods 5.2 Two cases6 Re-examining Our Case Study7 Conclusion. (shrink)

According to the comparative Bayesian concept of confirmation, rationalized versions of creationism come out as empirically confirmed. From a scientific viewpoint, however, they are pseudo-explanations because with their help all kinds of experiences are explainable in an ex-post fashion, by way of ad-hoc fitting of an empirically empty theoretical framework to the given evidence. An alternative concept of confirmation that attempts to capture this intuition is the use novelty criterion of confirmation. Serious objections have been raised against this criterion. In (...) this paper I suggest solutions to these objections. Based on them, I develop an account of genuine confirmation that unifies the UN-criterion with a refined probabilistic confirmation concept that is explicated in terms of the confirmation of evidence-transcending content parts of the hypothesis. (shrink)

In her 1996 book, Error and the Growth of Experimental Knowledge, Deborah Mayo argues that use- (or heuristic) novelty is not a criterion we need to consider in assessing the evidential value of observations. Using the notion of a “severe” test, Mayo claims that such novelty is valuable only when it leads to severity, and never otherwise. To illustrate her view, she examines the historical case involving the famous 1919 British eclipse expeditions that generated observations supporting Einstein's theory of gravitation (...) over Newton's. My plan here is to defend use-novelty as a valuable methodological principle. I begin by exposing a weakness in Mayo's criticism of use-novelty. Remedying this weakness re-establishes the worth of use-novelty under specific conditions; in particular, heuristically novel data are to be preferred, as I will say, “prima facie”. Armed with this revised version of use-novelty, I re-examine the history of the eclipse experiments and offer an interpretation of this episode that to an extent—and contrary to Mayo—restores the mildly heretical, Earman/Glymour evaluation of this episode offered in their (1980). I conclude by responding to criticism of my assessment of Mayo's work. (shrink)

Scientific theories are developed in response to a certain set of phenomena and subsequently evaluated, at least partially, in terms of the quality of fit between those same theories and appropriately distinctive phenomena. To differentiate between these two stages it is popular to describe the former as involving the accommodation of data and the latter as involving the prediction of data. Predictivism is the view that, ceteris paribus, correctly predicting data confers greater confirmation than successfully accommodating data. In this paper, (...) I take issue with a variety of predictivist theses, argue that their role for issues of confirmation is extremely limited, and attempt to account for the appeal that predictivism has enjoyed. Introduction Temporal Predictivism Heuristic Predictivism Weak Predictivism 4.1 Inference to better theories 4.2 Inference to better methods Arguments for Strong Heuristic Predictivism 5.1 Best explanations argument 5.2 Conditional support 5.3 Unique explanations Increased Explanatory Unification 6.1 Explaining what other theories can't 6.2 Contrived hypotheses 6.2 Strength and simplicity Conclusions CiteULike Connotea Del.icio.us What's this? (shrink)

In this dissertation, I examine a view called ‘Epistemic Structural Realism’, which holds that we can, at best, have knowledge of the structure of the physical world. Put crudely, we can know physical objects only to the extent that they are nodes in a structure. In the spirit of Occam’s razor, I argue that, given certain minimal assumptions, epistemic structural realism provides a viable and reasonable scientific realist position that is less vulnerable to anti-realist arguments than any of its rivals.

We argue for a naturalistic account for appraising scientific methods that carries non-trivial normative force. We develop our approach by comparison with Laudan’s (American Philosophical Quarterly 24:19–31, 1987, Philosophy of Science 57:20–33, 1990) “normative naturalism” based on correlating means (various scientific methods) with ends (e.g., reliability). We argue that such a meta-methodology based on means–ends correlations is unreliable and cannot achieve its normative goals. We suggest another approach for meta-methodology based on a conglomeration of tools and strategies (from statistical modeling, (...) experimental design, and related fields) that affords forward looking procedures for learning from error and for controlling error. The resulting “error statistical” appraisal is empirical—methods are appraised by examining their capacities to control error. At the same time, this account is normative, in that the strategies that pass muster are claims about how actually to proceed in given contexts to reach reliable inferences from limited data. (shrink)

I document some of the main evidence showing that E. S. Pearson rejected the key features of the behavioral-decision philosophy that became associated with the Neyman-Pearson Theory of statistics (NPT). I argue that NPT principles arose not out of behavioral aims, where the concern is solely with behaving correctly sufficiently often in some long run, but out of the epistemological aim of learning about causes of experimental results (e.g., distinguishing genuine from spurious effects). The view Pearson did hold gives a (...) deeper understanding of NPT tests than their typical formulation as accept-reject routines, against which criticisms of NPT are really directed. The Pearsonian view that emerges suggests how NPT tests may avoid these criticisms while still retaining what is central to these methods: the control of error probabilities. (shrink)

Most theories and hypotheses in psychology are verbal in nature, yet their evaluation overwhelmingly relies on inferential statistical procedures. The validity of the move from qualitative to quantitative analysis depends on the verbal and statistical expressions of a hypothesis being closely aligned – that is, that the two must refer to roughly the same set of hypothetical observations. Here, I argue that many applications of statistical inference in psychology fail to meet this basic condition. Focusing on the most widely used (...) class of model in psychology – the linear mixed model – I explore the consequences of failing to statistically operationalize verbal hypotheses in a way that respects researchers' actual generalization intentions. I demonstrate that although the “random effect” formalism is used pervasively in psychology to model intersubject variability, few researchers accord the same treatment to other variables they clearly intend to generalize over (e.g., stimuli, tasks, or research sites). The under-specification of random effects imposes far stronger constraints on the generalizability of results than most researchers appreciate. Ignoring these constraints can dramatically inflate false-positive rates, and often leads researchers to draw sweeping verbal generalizations that lack a meaningful connection to the statistical quantities they are putatively based on. I argue that failure to take the alignment between verbal and statistical expressions seriously lies at the heart of many of psychology's ongoing problems (e.g., the replication crisis), and conclude with a discussion of several potential avenues for improvement. (shrink)

an observation to formulate a theory, it is no surprise that the resulting theory accurately captures that observation. However, when the theory makes a novel prediction—when it predicts an observation that was not used in its formulation—this seems to provide more substantial confirmation of the theory. This paper presents a new approach to the vexed problem of understanding the epistemic difference between prediction and accommodation. In fact, there are several problems that need to be disentangled; in all of them, the (...) key is the concept of overfitting. We float the hypothesis that accommodation is a defective methodology only when the methods used to accommodate the data fail to guard against the risk of overfitting. We connect our analysis with the proposals that other philosophers have made. We also discuss its bearing on the conflict between instrumentalism and scientific realism. Introduction Predictivisms—a taxonomy Observations Formulating the problem What might Annie be doing wrong? Solutions Observations explained Mayo on severe tests The miracle argument and scientific realism Concluding comments. (shrink)

Several leading topics outstanding after John Earman's Bayes or Bust? are investigated further, with emphasis on the relevance of Bayesian explication in epistemology of science, despite certain limitations. (1) Dutch Book arguments are reformulated so that their independence from utility and preference in epistemic contexts is evident. (2) The Bayesian analysis of the Quine-Duhem problem is pursued; the phenomenon of a "protective belt" of auxiliary statements around reasonably successful theories is explicated. (3) The Bayesian approach to understanding the superiority of (...) variety of evidence is pursued; a recent challenge (by Wayne) is converted into a positive result on behalf of the Bayesian analysis, potentially with far-reaching consequences. (4) The condition for applying the merger-of-opinion results and the thesis of underdetermination of theories are compared, revealing significant limitations in applicability of the former. (5) Implications concerning "diachronic Dutch Book" arguments and "non-Bayesian shifts" are drawn, highlighting the incompleteness, but not incorrectness, of Bayesian analysis. (shrink)

“What did the President know and when did he know it?”Senator Howard Baker, Watergate hearings, 1973Why do scientists accept or reject theories? More specifically: why do they change from one theory to another? What is the role of empirical tests in the evaluation of theories?This paper focuses on a narrowly-defined question: in judging theories, do scientists give greater weight (other things being equal) to successfulnovel predictionsthan to successful deductions of previously-known facts? The affirmative answer is called the “predictivist thesis” (Maher (...) 1988).It is primarily philosophers who are interested in this question, and they have treated it mostly as a normative or logical problem. Can the writings of historians of science tell us how scientists have treated novel predictions in the past? Until recently historians have rarely addressed this point. (shrink)

*PSA 2016, symposium on “Data in Time: Epistemology of Historical Data” organized by Sabina Leonelli, 5 November 2016* *See published version: "Radiocarbon Dating in Archaeology: Triangulation and Traceability" in Data Journeys in the Sciences (2020) - link below* Archaeologists put a premium on pressing “legacy data” into service, given the notoriously selective and destructive nature of their practices of data capture. Legacy data consist of material and records that been assembled over decades, sometimes centuries, often by means and for purposes (...) long since discredited or superseded. The primary strategies by which archaeologists put the data to work for new purposes are, I argue, secondary retrieval, recontextualization, and experimental modelling. I focus here on a particularly telling and complex example of secondary retrieval: the extraction of new data from old by means of radiocarbon dating. This is by no means a straightforward process of retrieving physical samples from legacy data to which 14C techniques can be applied that can, on their own, decisively settle chronological questions. When Libby’s post-war radiocarbon revolution got under way, it was expected to establish an absolute chronology that would render obsolete the local and relative chronologies on which archaeologists had long relied. Transformative though it has been, bringing these tools of physical dating to bear on archaeological problems has been a long, tortuous process, now described as proceeding through two subsequent radiocarbon revolutions. The second was an extended process of calibration by which 14C chronologies were corrected and refined, often against the very lines of evidence they were meant to displace. The most recent, a “pragmatic Bayesian” approach to archaeological dating, is motivated by concern that, no matter how much it is refined, radiocarbon dating cannot on its own resolve the chronological problems that archaeologists address; the challenge, its advocates argue, is to ‘fully integrate archaeological information with 14C dating in order to address archaeologically relevant timescales and episodes’. This is a genre of “robustness” reasoning that illustrates its epistemic risks as well as its appeal. As recent philosophical debate makes clear, it depends on appeals to the convergence of independent lines of evidence that may have more rhetorical than epistemic force and that may be spurious. Drawing on this philosophical literature I identify a set of conditions that must be met if these risks are to be avoided, all of which are an explicit focus of debate in cases of contestation about and reconciliation of chronologies based on legacy data. (shrink)

A stalwart view in the philosophy of science holds that, even when broadly construed so as to include theoretical auxiliaries, theories cannot make direct contact with observations. This view owes much to Bogen and Woodward’s influential distinction between data and phenomena. According to them, data are typically the kind of things that are observable or measurable like "bubble chamber photographs, patterns of discharge in electronic particle detectors and records of reaction times and error rates in various psychological experiments". Phenomena are (...) physical processes that are typically unobservable. Examples of the latter category include "weak neutral currents, the decay of the proton, and chunking and recency effects in human memory". Theories, in Bogen and Woodward’s view, are utilised to systematically explain, infer and predict phenomena, not data. The relationship between theories and data is rather indirect. Data count as evidence for phenomena and the latter in turn count as evidence for theories. This view is becoming increasingly influential, Stathis Psillos and Mauricio Suárez ). In this paper I argue contrary to this view that in various significant and well-known cases theories do make direct contact with the help of suitable auxiliaries. (shrink)

It is sometimes held that facts confirm a hypothesis only if they were not used in the construction of that hypothesis. This requirement of "use novelty" introduces a historical aspect into the assessment of evidence claims. I examine a methodological principle invoked by physicists in the experimental search for the top quark that bears a striking resemblance to this view. However, this principle is better understood, both historically and philosophically, in terms of the need to conduct a severe test than (...) in terms of use novelty. Nevertheless, a historical factor remains in the assessment of some evidence claims. (shrink)

In this paper it is suggested that Freud's 'tally argument' (Grunbaum 1984) is not best interpreted as a risky claim concerning the efficacy of psychoanalytic therapy, but as a risky claim concerning the implications of theoretical psychoanalytic explanations of the efficacy of psychoanalytic therapy. Despite the fact that Freud never empirically established that these implications hold, the 'tally argument' does draw attention to a critical distinction that is too often neglected in contemporary empirical studies of psychoanalysis and other forms of (...) psychotherapy: between empirical evaluations of the efficacy of psychotherapy and empirical evaluations of theoretical explanations of the efficacy of psychotherapy, and the different forms of comparative enquiry relevant to each. It is argued that the contemporary neglect of this critical distinction, in conjunction with the common negative conception of placebo control treatments in psychotherapy research, has led to the epistemic impoverishment of experimental studies of the various professional psychotherapies. In consequence, although there is good empirical evidence for the efficacy of psychoanalysis and other forms of professional psychotherapy, there is no good empirical evidence for theoretical psychoanalytic explanations of the efficacy of psychoanalysis, or for traditional theoretical explanations of the efficacy of other forms of professional psychotherapy. (shrink)