Current neuroimaging techniques with high spatial resolution constrain participant motion so that many natural tasks cannot be carried out. The aim of this paper is to show how a time-locked correlation-analysis of cerebral blood flow velocity (CBFV) lateralization data, obtained with functional TransCranial Doppler (fTCD) ultrasound, can be used to infer cerebral activation patterns across tasks. In a first experiment we demonstrate that the proposed analysis method results in data that are comparable with the standard Lateralization Index (LI) for within-task (...) comparisons of CBFV patterns, recorded during cued word generation (CWG) at two difficulty levels. In the main experiment we demonstrate that the proposed analysis method shows correlated blood-flow patterns for two different cognitive tasks that are known to draw on common brain areas, CWG, and Music Synthesis. We show that CBFV patterns for Music and CWG are correlated only for participants with prior musical training. CBFV patterns for tasks that draw on distinct brain areas, the Tower of London and CWG, are not correlated. The proposed methodology extends conventional fTCD analysis by including temporal information in the analysis of cerebral blood-flow patterns to provide a robust, non-invasive method to infer whether common brain areas are used in different cognitive tasks. It complements conventional high resolution imaging techniques. (shrink)

Just as humans can draw conclusions responsibly or irresponsibly, so too can computers. Machine learning systems that have been trained on data sets that include irresponsible judgments are likely to yield irresponsible predictions as outputs. In this paper I focus on a particular kind of inference a computer system might make: identification of the intentions with which a person acted on the basis of photographic evidence. Such inferences are liable to be morally objectionable, because of a way in which (...) they are presumptuous. After elaborating this moral concern, I explore the possibility that carefully procuring the training data for image recognition systems could ensure that the systems avoid the problem. The lesson of this paper extends beyond just the particular case of image recognition systems and the challenge of responsibly identifying a person’s intentions. Reflection on this particular case demonstrates the importance (as well as the difficulty) of evaluating machine learning systems and their training data from the standpoint of moral considerations that are not encompassed by ordinary assessments of predictive accuracy. (shrink)

There has been an intense discussion, albeit largely an implicit one, concerning the inference of causal hypotheses from statistical correlations in quantum mechanics ever since John Bell’s first statement of his notorious theorem in 1966. As is well known, its focus has mainly been the so-called Einstein-Podolsky-Rosen (“EPR”) thought experiment, and the ensuing observed correlations in real EPR like experiments. But although implicitly the discussion goes as far back as Bell’s work, it is only in the last two decades (...) that it has become recognizably and explicitly a debate about causal inference in the quantum realm. The bulk of this paper is devoted to a review of three influential arguments in the philosophical literature that aim to show that causal models for the EPR correlations are impossible, due to Bas Van Fraassen, Daniel Hausman and Huw Price. I contend that all these arguments are inconclusive since they contain premises or presuppositions that are false, unwarranted, or at least controversial. Five different common cause models are outlined that seem perfectly viable for the EPR correlations. These models are then employed to illustrate various difficulties with the premises and presuppositions underlying Van Fraassen’s, Hausman’s and Price’s arguments. In all these cases it is argued that the difficulties cut deep against these authors’ own theories of causation and causal inference. My conclusions are that causal models for the EPR correlations remain viable, that philosophical work is still required to assess their relative virtues, and that in any case the mere theoretical conceivability and empirical possibility of these models sheds doubts over Van Fraassen’s, Hausman’s and (important elements in) Price’s theories of causation and causal inference. (shrink)

In reply to Geach's objection against expressivism, some have claimed that there is a plurality of truth predicates. I raise a difficulty for this claim: valid inferences can involve sentences assessable by any truth predicate, corresponding to 'lightweight' truth as well as to 'heavyweight' truth. To account for this, some unique truth predicate must apply to all sentences that can appear in inferences. Mixed inferences remind us of a central platitude about truth: truth is what is preserved in valid (...) inferences. The question is why we should postulate truth predicates that do not satisfy this platitude. (shrink)

Discussions about the nature of essence and about the inference problem for non-Humean theories of nomic modality have largely proceeded independently of each other. In this article I argue that the right conclusions to draw about the inference problem actually depend significantly on how best to understand the nature of essence. In particular, I argue that this conclusion holds for the version of the inference problem developed and defended by Alexander Bird. I argue that Bird’s own argument (...) that this problem is fatal for David Armstrong’s influential theory of the laws of nature but not for dispositional essentialism is seriously flawed. In place of this argument, I develop an argument that whether Bird’s inference problem raises serious difficulties for Armstrong’s theory depends on the answers to substantial questions about how best to understand essence. The key consequence is that considerations about the nature of essence have significant, underappreciated implications for Armstrong’s theory. (shrink)

The philosophy of historical sciences investigates their distinct objects of study, epistemic challenges, and methodological solutions. Rethinking astronomy in this light offers a contribution. First, the methodology of historical sciences adds to a more adequate description of how astronomers study and utilize token events. Second, astronomy faces a typical difficulty in identifying traces of some past events and has developed a delicate solution. This enriches the idea of trace and suggests a methodology that relies on iterations between data-driven approaches (...) and theory-driven approaches, together with the cross-validation between multiple relevant historical events or datasets. (shrink)

Whenever the use of non-experimental data for discovering causal relations or predicting the outcomes of experiments or interventions is contemplated, two difficulties are routinely faced. One is the problem of latent variables, or confounders: factors influencing two or more measured variables may not themselves have been measured or recorded. The other is the problem of sample selection bias: values of the variables or features under study may themselves influence whether a unit is included in the data sample.

Rationality has long been held to be the hallmark of what it means to be human. Consequently, the act of deductive inference—a central element of human reasoning—may be assumed to be natural. Not surprisingly, the study of formal logic has traditionally been regarded as essential for the philosophy major and recommended for many others. Yet both empirical study and pedagogical experience suggest that we deduce, on the whole, rather poorly. In fact, reasoning within formal systems seems to pose insurmountable (...) difficulties for some students. In this article, I suggest that both classroom experience and psychological research point to the possibility of a logic-related learning disability which I refer to as inference blindness. I further suggest that the dual-mechanism theory proposed by Vinod Goel offers the best way of understanding deductive reasoning and that the application of this theory suggests a preliminary hypothesis regarding inference blindness. (shrink)

The best developed formulation of closed system quantum theory that handles multiple-time statements, is the consistent (or decoherent) histories approach. The most important weaknesses of the approach is that it gives rise to many different consistent sets, and it has been argued that a complete interpretation should be accompanied with a natural mechanism leading to a (possibly) unique preferred consistent set. The existence of multiple consistent sets becomes more problematic because it allows the existence of contrary inferences [1]. We analyse (...) the conceptual difficulties that arise from the existence of multiple consistent sets and provide a suggestion for a natural set selection criterion. This criterion does not lead to a unique physical consistent set, however it evades the existence of consistent sets with contrary inferences. The criterion is based on the concept of preclusion and the requirement that probability one propositions and their inferences should be non-contextual. The allowed consistent sets turn-out to be compatible with coevents which are the ontology of an alternative, histories based, formulation [2–4]. (shrink)

The idea that knowledge can be extended by inference from what is known seems highly plausible. Yet, as shown by familiar preface paradox and lottery-type cases, the possibility of aggregating uncertainty casts doubt on its tenability. We show that these considerations go much further than previously recognized and significantly restrict the kinds of closure ordinary theories of knowledge can endorse. Meeting the challenge of uncertainty aggregation requires either the restriction of knowledge-extending inferences to single premises, or eliminating epistemic uncertainty (...) in known premises. The first strategy, while effective, retains little of the original idea—conclusions even of modus ponens inferences from known premises are not always known. We then look at the second strategy, inspecting the most elaborate and promising attempt to secure the epistemic role of basic inferences, namely Timothy Williamson’s safety theory of knowledge. We argue that while it indeed has the merit of allowing basic inferences such as modus ponens to extend knowledge, Williamson’s theory faces formidable difficulties. These difficulties, moreover, arise from the very feature responsible for its virtue- the infallibilism of knowledge. (shrink)

Hume's claim in ”Of the modern philosophy” that causal inference is implicated in an ineliminable, ”manifest contradiction” draws on a highly artificial version of an argument from perceptual relativity. Hume's statement of a ”very dangerous dilemma” draws on a mistaken argument in ”Of scepticism with regard to reason” for the conclusion that all probability, including evidence based on causal inference, reduces to zero. Contrary to Hume's own assessment, his stability‐based theory of justification has little to fear from these (...) episodes. At the same time, Hume's theory of justification requires an account of the conditions under which contradictory beliefs either continue to oscillate or lead to a conflicted resolution, rather than canceling out. Hume's treatments of the probability of causes, the direct passions, education, and contradictions involving demonstration and other belief‐forming mechanisms show that this is a matter Hume does not satisfactorily address. (shrink)

My concern here is with the possibility of defending some of R. G. Collingwood’s views regarding the roles of inference and ‘re-enactment’ in ‘doing history’. The difficulties encountered result from the fact that Collingwood never presented his views in an altogether clear and complete form, and the fact that several metaphysical and epistemological problems are involved. The specific problems dealt with are, for the most part, those which seem most easily disposed of by an adequate interpretation of Collingwood.

Children and adults with autism do worse on tests of inferences than controls. This fact has been attributed to poor language skills, a tendency to focus on detail, and poor social understanding. This study examines whether children with autism with age-appropriate language and cognitive skills have difficulties drawing inferences from academic, expository texts. Sixteen children with autism and a control group of twenty-four children were matched on language skills, nonverbal cognitive ability, and auditory and nonverbal working memory and compared on (...) their responses to questions that require inferences. The children with autism scored significantly lower on inference questions than the controls. Although language skills explain much of the variance in inference scores, diagnostic background also made a contribution. The results are discussed in the light of theories of suppression of irrelevant information and recognition of text writers’ communicative intention. (shrink)

Some philosophers find the following thesis attractive: for every logical constant C there is a set of logical rules of inference R such that a subject knows the meaning of C if and only if she accepts the rules in R. I point out some obvious but, apparently, easily forgotten difficulties concerning this thesis.

1. Terminological Considerations. Since Russell enunciated the principle, “Wherever possible logical constructions are to be substituted for inferred entities,” or “Wherever possible, substitute constructions out of known entities for inferences to unknown entities,” the terminological situation has become confused. Russell defined neither “construction” nor “inferred entity.” “Construct” soon came to be used for “construction,” perhaps to avoid the ambiguity whereby the latter term was used to refer to both a process and a result. But many writers now use “construction” or (...) “construct” to refer to what Russell called the “inferred entity.” This seems to be the usage of Margenau, Ramsperger, Benjamin, Bures, and many others. Again, some of the same writers in other places remain closer to Russell's usage. And I find it difficult to determine exactly the relation of the usage of some writers to that of Russell because of the differences between their contexts and Russell's. I experience this difficulty in reading, for instance, both Miss Stebbing and Werkmeister. The terminological difficulties have been evaded by some other writers who have formulated new terms to refer to Russell's inferred entities. Thus Northrop speaks of the “theoretic component” and Miller has coined the word, “interphenomena.” The terms reflecting Russell's rule in psychology have become so confused that a set of explicit definitions has recently been proposed. (shrink)

Of the many proposals for inferring genetic regulatory structure from microarray measurements of mRNA transcript hybridization, several aim to estimate regulatory structure from the associations of gene expression levels measured in repeated samples. The repeated samples may be from a single experimental condition, or from several distinct experimental conditions; they may be “equilibrium” measurements or time series; the associations may be estimated by correlation coefficients or by conditional frequencies (for discretized measurements) or by some other statistic. This paper describes two (...) elementary statistical difficulties for all such procedures, no matter whether based on Bayesian updating, conditional independence testing, or other machine learning procedures such as simulated annealing or neural net pruning. One difficulty obtains if large numbers of cells are aggregated in a measurement of expression levels from a common population of cells; the other obtains if small numbers of cells are aggregated or if samples are separately aggregated over different populations of cells. (shrink)

The difficulties facing Humean regularity accounts of laws have led some philosophers to a theory that takes laws to be necessitation relations between universals. In this paper I evaluate David Armstrong's version of this theory by considering two of its key elements: its solution to the so-called “Inference Problem” and its denial of uninstantiated universals. After considering some potential problems with each of these elements on their own, I argue that Armstrong's solution to the Inference Problem and his (...) denial of uninstantiated universals are not two independent aspects of his view. His solution to the Inference Problem depends upon his denial of uninstantiated universals. (shrink)

Inferring others’ preferences is socially important and useful. We investigated whether children infer preferences from the minimal information provided by an agent’s single action, and whether they avoid inferring preference when the action is constrained. In three experiments, children saw vignettes in which an agent took a worse toy instead of a better one. Experiment 1 shows that this single action influences how young children infer preferences. Children aged three and four were more likely to infer the agent preferred the (...) worse toy when the agent took this toy, compared with when the agent did not take either toy. Experiment 2 then shows that children consider constraints when inferring preferences from a single action. From age 5, children were less likely to avoid inferring a preference for the worse toy when the agent’s action was physically constrained. Finally, Experiment 3 provides evidence that children’s and adults’ sensitivity to constraints, when inferring preferences, is not based on a general notion of constraints, and instead depends on several specific notions. Whereas 5–6-year-olds in this experiment considered physical and socio-moral constraints when inferring preferences, they had difficulty grasping the relevance of epistemic constraints. Adults considered physical and epistemic constraints, but were not influenced by the socio-moral constraint of ownership. Together these findings contribute to a picture of cognitive development in which children are able to infer non-obvious properties on the basis of minimal concrete information, and are also sensitive to subtle changes in this information. 2016 APA, all rights reserved). (shrink)

In lieu of an abstract, here is a brief excerpt of the content:141 HUME AND CAUSAL INFERENCE Hume held that any time we reach by inference a belief about the existence or qualities of some object or event that we have not actually perceived, the inference is grounded in beliefs about causal relations holding between the object of belief and some other object or objects that have actually been experienced. I will examine here Hume's account of such (...) causal inferences. There are many places in the Treatise of Human Nature at which Hume seems to claim that causal inferences take place when and only when the percipient has had an extensive previous experience of objects of the respective types as regularly conjoined in relations of contiguity and temporal priority. It seems to me that reliance on these passages alone gives a distorted view of Hume's theory of causal inference. Contrary to appearances, Hume did not hold that extensive previous experiences of objects of the same types as properly conjoined is either necessary or sufficient for causal inference. Hume's view is more subtle and complicated, and there are important considerations of coherence and holism present in it. Getting Hume's view on this matter clear is important because it shows how he can escape some of the objections that have been laid against his theory, and it makes possible a reading of Hume as a causal realist. Hume's theory of inference provides the resources for drawing a distinction between accidental and genuine causal regularities and it is compatible with the view that causal relations (and not mere de facto regularities) hold among events in the world independently of their being perceived. The examination begins by looking at a different, but related, issue. Why does Hume think 142 that of the three types of association, only cause and effect association is capable of producing belief? Hume puts the question in the form of a possible objection to his account. For it may be said, that if all the parts of that hypothesis be true, viz. that these three species of relation are deriv'd from the same principles; that their effects in inforcing and inlivening our ideas are the same; and that belief is nothing but a more forcible and vivid conception of an idea; it shou'd follow, that that action of the mind may not only be deriv'd from the relation of cause and effect, but also from those of contiguity and resemblance. But as we find by experience, that belief arises only from causation, and that we can draw no inference from one object to another, except they be connected by this relation, we may conclude, that there is some error in that reasoning, which leads us into such difficulties. (T 107) It is important to note that Hume is talking here only about beliefs about the existence or situations of objects that have not been intuited or experienced. Beliefs about relations of ideas (e.g., that 2 + 2 = 4) are intuitions (or perhaps memories 2 of intuitions). Beliefs in the existence of things as experienced or in the previous existence of things as previously experienced automatically have the 3 vivacity of impressions or memory ideas. What Hume takes himself to need to explain is why cause and effect association allows for a transfer of this vivacity to an idea of an object or situation that has not itself been intuited or experienced and why 4 resemblance and contiguity association do not. If I see a green ball next to a red sweater, then I will normally subsequently believe that the green ball existed next to the red sweater at that time (at 143 least until my memories begin to lose vivacity). This is purely a matter of sense experience and memory. Now if I subsequently see a different red sweater (or the same sweater in a different location), I may be led to think of a green ball existing next to it purely through contiguity and resemblance association. But presumably I would not believe that a green ball existed next to the currently perceived red sweater unless I currently perceived a green ball or my belief was based on some 5 form of causal inference... (shrink)

Of the many proposals for inferring genetic regulatory structure from microarray measurements of mRNA transcript hybridization, several aim to estimate regulatory structure from the associations of gene expression levels measured in repeated samples. The repeated samples may be from a single experimental condition, or from several distinct experimental conditions; they may be “equilibrium” measurements or time series; the associations may be estimated by correlation coefficients or by conditional frequencies (for discretized measurements) or by some other statistic. This paper describes two (...) elementary statistical difficulties for all such procedures, no matter whether based on Bayesian updating, conditional independence testing, or other machine learning procedures such as simulated annealing or neural net pruning. One difficulty obtains if large numbers of cells are aggregated in a measurement of expression levels from a common population of cells; the other obtains if small numbers of cells are aggregated or if samples are separately aggregated over different populations of cells. (shrink)

Fodor & Lepore (2001) and Williamson (2003) attack the inferentialist account of concept possession according to which possessing or understanding a concept requires endorsing the inference patterns constitutive of its content. I show that Fodor & Lepore’s concern - that the conception places an exorbitant epistemological demands on possessors of a concept - is met by Brandom’s tolerance of materially bad nonconservative inferences. Such inferences themselves, as Williamson argues, present difficulties for the ‘understanding as endorsement’ conception. I show that, (...) properly understood, Brandom’s broad conception of inferential role, which encompasses social-perspectival inferential connections, has the resources to respond to Willianson’s challenge. (shrink)

The set of entities that serves as the domain for our discourse about the mind is metaphysically heterogenous. It includes processes, events, properties, modes, and states. In the latter part of the twentieth century, philosophers started to suppose that a philosophical theory of the mind should be primarily concerned with the explanation of mental states. Those states could then be mentioned in the explanations that would need to be given for mental entities of other sorts. If, for example, we had (...) a prior explanation of belief states, those states could figure in our subsequent explanation of inferences: inferences, on this approach, are to be identified with certain processes of belief revision. This states-first approach was not favoured by earlier theorists of the mind, who tended to suppose that mental events and processes are explanatorily more basic than mental states. The current states-first approach faces insuperable difficulties, which the earlier approach avoids. (shrink)

Some philosophers find the following thesis attractive: for every logical constant C there is a set of logical rules of inference R such that a subject knows the meaning of C if and only if she accepts the rules in R. I point out some obvious but, apparently, easily forgotten difficulties concerning this thesis.

This book is a clear, concise, and conceptually unified treatment of various problems, both formal and philosophical, of inductive logic and probability. Ackermann's main concern throughout the book is the problem of adducing inductive support for various hypotheses, and of deciding between two competing hypotheses which is more reasonable given the available evidence. The author begins with a general consideration of the criteria to be met by satisfactory rules of inductive inference: accordance with intuitive notions of reasonableness in simple (...) cases, sensitivity to the given evidence, capacity for unambiguous intersubjective application. Ackermann quickly disposes of the argument that valid inductive rules must be made to rest upon a "postulate of the uniformity of nature": "If past projections of regularities suddenly begin to go wrong, nondeductive rules satisfying the criteria suggested above will begin to select regularities other than the ones which have failed." This clears the way for a consideration of the general problem of evidential support or confirmation, with its well-known paradoxes and difficulties. Ackermann, like Goodman, locates the source of these in problems bearing on the semantics of scientific language, and cautiously points to Goodman's theory of projection and entrenchment as the way out of most of the muddle. Succeeding chapters deal with the notion of probability of hypotheses, statistical inference, and the application of Bayesian techniques and decision theory to science. A well-chosen bibliography completes this modest but rewarding little book.—H. P. K. (shrink)

Fodor & Lepore (2001) and Williamson (2003) attack the inferentialist account of concept possession according to which possessing or understanding a concept requires endorsing the inference patterns constitutive of its content. I show that Fodor & Lepore's concern – that the conception places an exorbitant epistemological demands on possessors of a concept – is met by Brandom's tolerance of materially bad nonconservative inferences. Such inferences themselves, as Williamson argues, present difficulties for the 'understanding as endorsement' conception. I show that, (...) properly understood, Brandom's broad conception of inferential role, which encompasses social-perspectival inferential connections, has the resources to respond to Willianson's challenge. (shrink)

In lieu of an abstract, here is a brief excerpt of the content:Reviewed by:Insight and Inference: Descartes's Founding Principle and Modern PhilosophyTom SorellMurray Miles. Insight and Inference: Descartes's Founding Principle and Modern Philosophy. Toronto: University of Toronto Press, 1999. Pp. xviii + 564. Cloth, $120.00.This book reopens the question of the correct interpretation of 'cogito, ergo sum,' and considers the significance of Descartes's first principle for Western philosophy up to and including the twentieth century. The gist of Miles's (...) interpretation is captured in the first three words of his title. The cogito is both insight and inference. How is it an inference, exactly? In the course of a process of selective attention to elements of thought, the cogito brings each person to the explicit knowledge of the fact of his or her existence (49). We have an implicit grasp of what the cogito makes explicit; and the passage from implicit to explicit knowledge amounts to a kind of inference. It is like an Aristotelian induction (cf. chapter 20).Inference is the topic of Part Three. Part Two concerns certainty. The cogito is certain in Meditation Two not in the light of the thought that God is veracious, but in the light solely of reflection on one's actually occurring thoughts (47). As for what Cartesian thoughts are, Miles seems to suggest that they are rather like what Husserl gets out of his phenomenological reduction: "intentionally-structured or object-directed acts of thinking" that are the sole immediate objects of consciousness (45). Miles rejects, correctly I think, the attribution to Descartes of any of the usual forms of representationalism.The diagnosis in Part Two of the certainty of the cogito seemed to me to be plausible on its face. It is in line with, although much more elaborate than, some reasonably familiar "psychologistic" interpretations clearing Descartes of the charge of circularity in the reasoning from clarity and distinctness to God's existence and nature and back again. (See A. Kenny, Descartes: A Study of his Philosophy [New York: Random House, 1968], chapter 8, and J. Tlumak, "Certainty and Cartesian Method" in M. Hooker, ed., Descartes: Critical and Interpretive Essays [Baltimore: Johns Hopkins University Press, 1978], 40-73. See also L. Loeb's article on the Cartesian Circle in J. Cottingham, ed., The Cambridge Companion to Descartes [New York, Cambridge University Press, 1992], 200-235.) To the extent that Miles departs from these interpretations, it seems to be by trying to base the authority of the cogito on the immutability of God (215-18). But this move seems ill-advised. Not only does it land Descartes in difficulties over the interpretation of pieces of scripture in which God changes his mind; it may also be in tension with the idea of omnipotence. On the whole, however, there is much in Part Two to admireófrom the interpretation of the simple natures (chapter 10), to the differences between scientia, persuasio, and notitia (chapter 11).Part One is less satisfactory, because Miles seems intent on reading Descartes through Husserl. Although he does not evade the objection that this approach is anachronistic (91-93), Miles does reinforce the impression, already created in the 55-page introduction, that he is unsure which interpretive perspective to use to control the book. Is he writing primarily as a twentieth century reader attracted to phenomenology, or is he someone who is keen to penetrate the seventeenth century and earlier and let the meanings of the early modern period infuse the text? Miles leaves the reader uncertain. In Part One we have a Husserlian Descartes. In the Introduction, the sixteenth and seventeenth century perspectives supply the preferred interpretation of [End Page 122] Descartes's metaphysics as something more or less traditional—concerned with the most certain, most universal, and the highest kind of being, the kind of being studied by theology.It is hard not to be put off by Miles's very swift dismissals of mainstream Anglo-American interpretations of Descartes's metaphysics as having some special connection with his physics (27ff). Can this line of interpretation really be refuted by quoting passages in which Descartes says that metaphysics is more basic and more certain than physics? Surely not. In choosing the formulation... (shrink)

According to Douglas Walton, the concept of presumption relates to both logical and dialogical components. Logically, a presumption is the conclusion of a presumptive defeasible inference. Dialogically, the function of a presumptions to shift the burden of proof to the respondent in order to move the dialogue forward when the proponent, due to an objective lack of evidence, cannot present a sufficiently persuasive proposition. Presumptive status, assigned only at the argumentation stage of dialogue, is provisional: a particular presumption stands (...) until further counter-indications are found. This paper points out some possible difficulties pertaining to Walton’s dialogical model of presumption. Firstly, once a presumption's typical practical function is made precise, Walton’s idea of presumption starts to lose its conceptual cohesion, thereby becoming applicable in the opening stage of dialogue, as well. Secondly, cohesion is lost also in practical contexts, since Walton's "seat belts" example might indicate that presumptions are not evidentially defeasible if the practical stakes are raised high enough. Most importantly, contextual evaluation may reveal that presumptions do not shift the burden of proof in any specific or genuine way. (shrink)

A basic problem of daily life is determining who owns what. One way that people may solve this problem is by relying on a ‘first possession’ heuristic, according to which the first person who possesses an object is its owner, even if others subsequently possess the object. We investigated preschoolers’ use of this heuristic in five experiments. In Experiments 1 and 2, 3- and 4-year-olds inferred that an object was owned by the character who possessed it first, even though another (...) character subsequently possessed it. Two-year-olds also showed this bias, but only when the object was placed between the characters when children were asked about ownership. Experiment 3 ruled out the possibility that children’s bias to select the first possessor results from a tendency to select the character first associated with the object. Experiment 4 showed that 3- and 4-year-olds have difficulty disregarding the first possession heuristic, even when provided with evidence that the character who first possessed an object is not its owner. But Experiment 5 found that children can disregard the heuristic in at least some situations. These five experiments suggest that the first possession heuristic guides children’s ownership inferences. The findings provide the first evidence that preschoolers can infer who owns what, when not explicitly told, and when not reasoning about objects with which they are personally acquainted. (shrink)

The “Game of the Rule” is easy enough: I give you the beginning of a sequence of numbers (say) and you have to figure out how the sequence continues, to uncover the rule by means of which the sequence is generated. The game depends on two obvious constraints, namely (1) that the initial segment uniquely identify the sequence, and (2) that the sequence be non-random. As it turns out, neither constraint can fully be met, among other reasons because the relevant (...) notion of randomness is either vacuous or undecidable. This may not be a problem when we play for fun. It is, however, a serious problem when it comes to playing the game for real, i.e., when the player to issue the initial segment is not one of us but the world out there, the sequence consisting not of numbers (say) but of the events that make up our history. Moreover, when we play for fun we know exactly what initial segment to focus on, but when we play for real we don’t even know that. This is the core difficulty in the philosophy of the inductive sciences. (shrink)

A well-known difficulty that affects all accounts of laws of nature according to which the latter are higher-order facts involving relations between universals (the so-called DTA accounts, from Dretske in Philosophy of Science 44:248–268, 1977; Tooley in Canadian Journal of Philosophy 7:667–698, 1977 and Armstrong (What is a Law of Nature?, Cambridge University Press, Cambridge, 1983)) is the Inference Problem: how can laws construed in that way determine the first-order regularities that we find in the actual world? Bird (...) (Analysis 65:147–55, 2005) has argued that there is no solution to the Inference Problem which is consistent with both categorical monism (that is, the view that all natural properties are categorical) and basic tenets of Armstrong’s account of the laws of nature. This paper shows that, given Armstrong’s view about laws as first-order structural universals whose instantiation ‘produce’ nomic regularities and under specific plausible metaphysical assumptions concerning nomic relations which are consistent with a broadly construed DTA approach to laws, there is no extra difficulty regarding the Inference Problem in a categorical monistic context besides the ones that beset structural universals in general. (shrink)

Previous research showed that people can make trait inferences from single behaviors described in sentences, without either intentions to do so or awareness of having done so. This suggested that these inferences might be automatic. By definition, “automatic” cognitive processes occur without intentions or awareness, without effort, and without using capacity-limited cognitive processing resources. Winter, Uleman, and Cunniff attempted to manipulate available cognitive capacity by varying the difficulty of the concurrent cognitive task. This did not affect unintended trait inferences, (...) suggesting that these are automatic by this criterion. But they had no direct measure of available cognitive capacity. In the present study, we added a probe reaction time measure of capacity to their procedure and extended the range of task difficulties. Earlier findings of trait inferences without intentions or awareness were replicated, but there was also evidence that the concurrent task interfered with the trait inference process. Hence, although trait inferences can be “spontaneous” , subjectively effortless, and difficult to disrupt with a concurrent task, they are not entirely automatic because they do use capacity-limited resources. These results also confirm that the cued-recall evidence for spontaneous trait inferences reflects important encoding and not merely retrieval phenomena. (shrink)

Earlier in this volume, Wesley Salmon has given a characteristically clear and trenchant critique of the account of non-demonstrative reasoning known by the slogan `Inference to the Best Explanation'. As a long-time fan of the idea that explanatory considerations are a guide to inference, I was delighted by the suggestion that Wes and I might work together on a discussion of the issues. In the event, this project has exceeded my high expectations, for in addition to the intellectual (...) gain that comes from the careful study of his essay, I have benefited enormously from the stream of illuminating emails and faxes that Wes has sent me during our collaboration. Doing philosophy together has been an education and a pleasure. Salmon's essay would place Inference to the Best Explanation beyond the pale of acceptable philosophical accounts of inference. According to Salmon, Inference to the Best Explanation has serious internal difficulties and compares very unfavourably with Bayesian approaches to these matters. My aim in the following remarks is irenic. I hope to show that a number of the claimed difficulties either are not really difficulties or are avoidable. In some cases, the avoidance will require a mild reinterpretation of the account that lies behind the slogan `Inference to the Best Explanation'; in others, it will require admitting limits to the scope of the account. For I accept at the outset that Inference to the Best Explanation cannot possibly be the whole story about the assessment of scientific hypotheses. For me, the interesting idea is simply that we sometimes decide how likely a hypothesis is to be correct in part by considering how good an explanation it would provide, if it were correct. This is the idea of explanatory considerations providing a guide to inference, and this is the idea that I will here promote. (shrink)

Bell's theorem is expounded as an analysis in Bayesian inference. Assuming the result of a spin measurement on a particle is governed by a causal variable internal (hidden, “local”) to the particle, one learns about it by making a spin measurement; thence about the internal variable of a second particle correlated with the first; and from there predicts the probabilistic result of spin measurements on the second particle. Such predictions are violated by experiment: locality/causality fails. The statistical nature of (...) the observations rules out signalling; acausal, superluminal, or otherwise. Quantum mechanics is irrelevant to this reasoning, although its correct predictions of experiment imply that it has a nonlocal/acausal interpretation. Cramer's newtransactional interpretation, which incorporates this feature by adapting the Wheeler-Feynman idea of advanced and retarded processes to the quantum laws, is advocated. It leads to an invaluable way of envisaging quantum processes. The usual paradoxes melt before this, and one, the “delayed choice” experiment, is chosen for detailed inspection. Nonlocality implies practical difficulties in influencing hidden variables, which provides a very plausible explanation for why they have not yet been found; from this standpoint, Bell's theoremreinforces arguments in favor of hidden variables. (shrink)

Recent work on Hume's Theory of Perception has shown that Hume takes the appearance of impressions to vary according to the ideas under which they are subsumed. In this paper, I argue that the vulgar position in the section where he discusses the Inference from Constancy is characterised by an ideal primordial state of mind where impressions are directly encountered without being subsumed under any idea. In particular, impressions which are not subsumed under the idea of a perception do (...) not appear to the mind as impressions. Read in this light, Hume's Inference from Constancy is immune to the difficulties which have commonly been raised against it. (shrink)

Some researchers on binary choice inference have argued that people make inferences based on simple heuristics, such as recognition, fluency, or familiarity. Others have argued that people make inferences based on available knowledge. To examine the boundary between heuristic and knowledge usage, we examine binary choice inference processes in terms of attribute substitution in heuristic use (Kahneman & Frederick, 2005). In this framework, it is predicted that people will rely on heuristic or knowledge‐based inference depending on the (...) subjective difficulty of the inference task. We conducted competitive tests of binary choice inference models representing simple heuristics (fluency and familiarity heuristics) and knowledge‐based inference models. We found that a simple heuristic model (especially a familiarity heuristic model) explained inference patterns for subjectively difficult inference tasks, and that a knowledge‐based inference model explained subjectively easy inference tasks. These results were consistent with the predictions of the attribute substitution framework. Issues on usage of simple heuristics and psychological processes are discussed. (shrink)

According to inferentialism about self-knowledge of desire, the basic way in which we come to know what we want is through inference. In this paper, I argue that in a wide range of cases of knowing one’s desire, inference is insufficient. In particular, I look at two inferentialist models, one proposed by Krista Lawlor and the other by Alex Byrne and look at the challenges that they face in securing safe self-ascriptions. In response to these difficulties, I argue (...) that we can explain how inferentially based self-ascriptions can be safe when we consider the agent’s role in sustaining their desires through attending and elaborating on the content of desire through imagination. (shrink)

In this paper I critique Peter Lipton’s attempt to deal with the threat of Bayesianism to the normative aspect of his project in Inference to the Best Explanation. I consider the five approaches Lipton proposes for reconciling the doxastic recommendations of Inference to the Best Explanation with BA’s: IBE gives a ‘boost’ to the posterior probability of particularly ‘lovely’ hypotheses after the Bayesian calculation is performed; IBE helps us to set the likelihood of evidence on a given hypothesis; (...) IBE helps us to set the prior probabilities of hypotheses and evidence; IBE guides us in determining which evidence is relevant to a given hypothesis IBE functions as a heuristic for otherwise difficult Bayesian calculations. I agree with Lipton in rejecting . However, I then go on to point out difficulties for , , , and , all of which Lipton provisionally accepts. As far as is concerned, the explanationist and the Bayesian both fall silent in the same situations. In the final analysis, seems to be moot. devolves on and , and since I reject both of them it is a bad option. And should be considered – if at all – only in unimportant cases; in vital ones, BA is clearly better. I then propose a sixth way in which IBE and BA could be seen as complementary. Yet this suggestion relegates IBE to a secondary, supporting rôle vis-à-vis BA. I then question whether this auxiliary status is the most the explanationist can hope for. (shrink)

Erdelyi does us all a great service by his customarily incisive discussion of the various ways in which our field tends to neglect, confuse, and misunderstand numerous critical issues in attempting to differentiate conscious from unconscious perception and memory. Although no single commentary could hope to comprehensively assess these issues, I will address Erdelyi’s three main points: How the dissociation paradigm can be used to validly infer unconscious perception; The implications of below-chance effects; and The role of time. I suggest (...) that significant progress on construct validity issues is possible; below-chance effects are part of a more general bidirectional phenomenon, very likely unconscious, and do not threaten absolute subliminality; and practice/learning effects pose potential difficulties for time-based dissociation paradigms. (shrink)

In this paper I consider the relationship between the spatial properties of visual perceptual experience and the sense-datum inference. I argue that the sense datum inference should be accepted if spatial properties are not merely intentionally present in such experiences. This result serves to underline the seriousness of the difficulties that are presented to direct realism by a particular class of illusory spatial experiences based on the geometry of visual perceptual experience. In light of these considerations I argue (...) that it is a constraint on direct realism that the underlying spatial character of visual perceptual experiences must be the same as that of the physical world. (shrink)

Researchers currently working on relational reasoning typically argue that mental model theory (MMT) is a better account than the inference rule approach (IRA). They predict and observe that determinate (or one-model) problems are easier than indeterminate (or two-model) problems, whereas according to them, IRA should lead to the opposite prediction. However, the predictions attributed to IRA are based on a mistaken argument. The IRA is generally presented in such a way that inference rules only deal with determinate relations (...) and not with indeterminate ones. However, (a) there is no reason to presuppose that a rule-based procedure could not deal with indeterminate relations, and (b) applying a rule-based procedure to indeterminate relations should result in greater difficulty. Hence, none of the recent articles devoted to relational reasoning currently presents a conclusive case for discarding IRA by using the well-known determinate vs indeterminate problems comparison. (shrink)

Sir Francis Galton writes: “Those who find a difficulty in understanding how a feebly felt mental action can vanquish a strong desire, will find the difficulty vanish if they consent to assume a physiological and not a psychical standpoint. The gain is as great as viewing the planetary system after the fashion of Copernicus, instead of that of Ptolemy. There is nothing contrary to experience in supposing that conflicting physiological actions may be perceived with a distinctness quite disproportionate (...) to their real efficacy. We may compare the conflict between faintly perceived activities of one kind and clearly perceived activities of another kind, to that between troops dressed in a uniform scarcely distinguishable from the background with others clad in staring scarlet. We must be content to admit that our consciousness has a very inexact cognisance of the physiological battles in our brain, and that the mystery why apparently weak motives of one class should invariably get the better of apparently strong motives of another class, lies wholly in the word ‘apparently’. In short, that the appearances of their relative strength are deceptive”. (shrink)

I argue that Goodman's puzzle of grue at least poses no real challenge about inductive inference. By drawing on Stove's characterisation of Hume's characterisation of inductive inference, we see that the premises in an inductive inference report experienced impressions; and Goodman can be interpreted as posing a real challenge about inductive inference only if we treat an epistemic subject's observations more as logical contents and less as experienced impressions. So, even though the grue puzzle was effective (...) against its stated logicist targets, it is not thereby an enduring difficulty regarding experience's ability to impart epistemic justification via inductive evidence. (shrink)

In this essay, I home in on the difficulties with group-to-individual (G2i) inferences in neuroscience and how they impact the legal system. I briefly outline how cognitive shortcutting can distort legal decisions, and then turn my attention to G2i inferences, with a special focus on issues of intellectual disability and brain damage. I argue that judges and juries are not situated to appreciate the nuances in brain data and that they are required to make clinical decisions without clinical training. As (...) a result, they effectively ignore those responsibilities and simply decide cases in virtue of what they already believe to be true. How judges actually make decisions in highstakes criminal cases is troubling, but they are also hamstrung in a variety of ways. (shrink)

Explanationist strategies for defending epistemological scientific realism make heavy use of a particular version of inference to the best explanation known as the no-miracle argument. I consider ESR to be a genuinely philosophical—non-naturalistic—thesis which contends that there are strong arguments to believe in some non-observational claims made by scientific theories that are partially observationally correct. In this paper, I examine the grounds of the strength of these arguments from what I call a contemplative perspective which focuses on the end (...) products, i.e. theories, of the scientific activity as opposed to the pragmatist view which considers science to be primarily an activity. I briefly rehearse the main difficulties of the no-miracle argument and of inference to the best explanation in general. I argue that a convincing defence of ESR should be based on the empirically ascertained reality of causal connections between theoretical entities which possess properties that are in principle observable and the results of measurements or observations. The knowledge of those causal connections may well deliver an—even the best—explanation of the appearances. But belief in the existence of some unobservable entities is mainly justified by their empirically attested causal role, not on their possible explanatory function. (shrink)

The problem of valid induction could be stated as follows: are we justified in accepting a given hypothesis on the basis of observations that frequently confirm it? The present paper argues that this question is relevant for the understanding of Machine Learning, but insufficient. Recent research in inductive reasoning has prompted another, more fundamental question: there is not just one given rule to be tested, there are a large number of possible rules, and many of these are somehow confirmed by (...) the data — how are we to restrict the space of inductive hypotheses and choose effectively some rules that will probably perform well on future examples? We analyze if and how this problem is approached in standard accounts of induction and show the difficulties that are present. Finally, we suggest that the explanation-based learning approach and related methods of knowledge intensive induction could be, if not a solution, at least a tool for solving some of these problems. (shrink)

Epistemologists have known for two-and-a-half centuries that there are serious difficulties surroundingnon-demonstrativeinference. The best-known problem,theproblem of induction, was first diagnosed by Hume in theTreatise. In our own century, several more problems were added, e.g., by Hempel —the paradox of the ravens—and by Goodman —the “new,” or exacerbated, problem of induction. But an even greater blow lay ahead: within the decade after Goodman's problem appeared, Gettier was to publish his famous challenge to the traditional analysis of knowledge which, again, underscored how (...) problematic inductive inferences are. (shrink)

In, Jeremy Avigad makes a novel and insightful argument, which he presents as part of a defence of the ‘Standard View’ about the relationship between informal mathematical proofs and their corresponding formal derivations. His argument considers the various strategies by means of which mathematicians can write informal proofs that meet mathematical standards of rigour, in spite of the prodigious length, complexity and conceptual difficulty that some proofs exhibit. He takes it that showing that and how such strategies work is (...) a necessary part of any defence of the Standard View.In this paper, I argue for two claims. The first is that Avigad’s list of strategies is no threat to critics of the Standard View. On the contrary, this observational core of heuristic advice in Avigad’s paper is agnostic between rival accounts of mathematical correctness. The second is that that Avigad’s project of accounting for the relation between formal and informal proofs requires an answer to a prior question: what sort of thing is an informal proof? His paper havers between two answers. One is that informal proofs are ultimately syntactic items that differ from formal derivations only in completeness and use of abbreviations. The other is that informal proofs are not purely syntactic items, and therefore the translation of an informal proof into a derivation is not a routine procedure but rather a creative act. Since the ‘syntactic’ reading of informal proofs reduces the Standard View to triviality, makes a mystery of the valuable observational core of his paper, and underestimates the value of the achievements of mathematical logic, he should choose some version of the second option. (shrink)