To be committed to the truth of a proposition is to constrain one’s options in a certain way: one may not reason as if it is false, and one is obligated to reason as if it is true. Though one is often committed to the truth of the propositions that one believes, the states of belief and commitment are distinct. For historical reasons, however, they are rarely distinguished. Distinguishing between the two states allows for a defense of epistemic deontology against (...) the charge that beliefs are not under the voluntary control of believers, and so cannot be subject to deontic evaluation. It also allows for the resolution of some disputes between naturalists and non-naturalists in epistemology, and permits us to account for obvious facts about the connection between belief and truth in a theoretically parsimonious way. (shrink)

I argue that S knows that p implies that S is properly committed to the truth of p, not that S believes that p. Belief is not required for knowledge because it is possible that one could know that there are no beliefs. Being ‘properly committed’ to the truth of a proposition is a matter of having a certain normative status, not occupying a particular psychological state. After arguing that knowledge requires commitment instead of belief, I go on to demonstrate (...) the theoretical utility of this hypothesis. (shrink)

Most explanations of the rational authority of testimony provide little guidance when evaluating individual pieces of testimony. In practice, however, we are remarkably sensitive to the varying epistemic credentials of testimony: extending trust when it is deserved, and withholding it when it is not. A complete account of the epistemology of testimony should, then, have something to say about when it is that testimony is trustworthy. In the typical case, to judge someone trustworthy requires judging them to be competent and (...) sincere. In this essay we develop an exchange-based account of testimony that shows how those who receive testimony are in a position to evaluate the sincerity of speakers. (shrink)

If a speaker selflessly asserts that p, the speaker (1) has good evidence that p is true, (2) asserts that p on the basis of that evidence, but (3) does not believe that p. Selfless assertions are widely thought to be acceptable, and therefore to pose a threat to the Knowledge Norm of Assertion. Advocates for the Knowledge Norm tend to respond to this threat by arguing that there are no such things as selfless assertions. They argue that those who (...) appear to be selfless asserters either: believe what they assert, perform a speech act other than assertion, or assert a proposition other than the one that they seem to. I argue that such counterarguments are unsuccessful. There really are selfless assertions. But I also argue that they are no threat to the Knowledge Norm. There is a good case to be made that knowledge does not require belief. And if it does not, then the fact that some selfless assertions are appropriate does not tell against the Knowledge Norm. Indeed, I argue that selfless asserters know the propositions that they assert to be true. (shrink)

If a speaker selflessly asserts that p, the speaker has good evidence that p is true, asserts that p on the basis of that evidence, but does not believe that p. Selfless assertions are widely thought to be acceptable, and therefore to pose a threat to the Knowledge Norm of Assertion. Advocates for the Knowledge Norm tend to respond to this threat by arguing that there are no such things as selfless assertions. They argue that those who appear to be (...) selfless asserters either: believe what they assert, perform a speech act other than assertion, or assert a proposition other than the one that they seem to. I argue that such counterarguments are unsuccessful. There really are selfless assertions. But I also argue that they are no threat to the Knowledge Norm. There is a good case to be made that knowledge does not require belief. And if it does not, then the fact that some selfless assertions are appropriate does not tell against the Knowledge Norm. Indeed, I argue that selfless asserters know the propositions that they assert to be true. (shrink)

I argue that the norms of assertion are engendered by the function of assertions, and that the function of assertions is, roughly, to facilitate the transmission of information from those who have it to those who need it. Assertions can play this role if they are governed by two norms. One norm is deontic in nature, and specifies the conditions under which a speaker may issue an assertion. I argue that the deontic norm permits A to issue an assertion to (...) B if and only if: (1) doing so would improve B’s epistemic position with regards to the proposition thus conveyed, and (2) the proposition conveyed is justified (for A) in a way, and to a degree, appropriate to the purposes for which B is likely to use it. The other is not deontic; it says what it is for an assertion to be good, qua assertion. This is a truth norm. Assertions ought to be true, in that an assertion is good, qua assertion, when it is true. (shrink)

Sinan Dogramaci has recently developed a view according to which the function of epistemic evaluations—like calling someone’s behavior “rational” or “irrational”—is to encourage or discourage the behavior evaluated. This view promises to explain the rational authority of testimony, by describing a social practice that promotes the coordination of epistemic procedures across a community. We argue that Dogramaci’s view is unsatisfactory, for two reasons. First, the social practice at its heart is vulnerable to free riders. Second, even if the problem of (...) free riders can be solved, it “alienates” epistemic agents from the testimony that they receive, in that, though they will accept testimony from their fellows, they will have no reason to do so. We argue that a more satisfactory view can be had if we couple the genuine insights that are to be found in Dogramaci’s proposal with the recognition that testimony is an excludible good that is often distributed according to market forces. It is this fact about test.. (shrink)

Characteristic of neo-pragmatism is a commitment to deflationism about semantic properties, and inferentialism about conceptual content. It is usually thought that deflationism undermines the distinction between realistic discourses and others, and that the neo-pragmatists, unlike the classical pragmatists, cannot recognize that truth is a norm of belief and inquiry. I argue, however, that the distinction between realistic discourses and others can be maintained even in the face of a commitment to deflationism, and that deflationists can recognize that truth is a (...) norm of belief and inquiry. If deflationism is true, realistic discourses, it turns out, are those that are inferentially integrated with a large body of other commitments, whereas those that call for an anti-realist treatment are inferentially isolated. Now, Grimm has persuasively argued that inquiry aims at achieving understanding, and that to understand something is, roughly, to grasp a large body of inferential connections in which it features. So, if he is right, realistic discourses are those in which the aim of inquiry can be achieved. This fact, together with an inferential theory of conceptual content, will, I argue, allow neo-pragmatists to recognize truth as a norm of belief and inquiry, despite their commitment to deflationism. (shrink)

I argue that, in an important range of cases, judging that one disagrees with an epistemic peer requires attributing, either to one's peer or to oneself, a failure of rationality. There are limits, however, to how much irrationality one can coherently attribute, either to oneself or to another. I argue that these limitations on the coherent attribution of rational error put constraints on permissible responses to peer disagreement. In particular, they provide reason to respond to one-off disagreements with a single (...) peer by maintaining one's beliefs, and they provide reason to moderate one's beliefs when faced with repeated disagreement, or disagreement with multiple peers. Finally, I argue that, though peer disagreement is rare, the occasions on which it does occur tend to be especially important, and the kind of response supported here is correspondingly important. In particular, how leading researchers spend their time and effort depends, in part, on how they respond to peer disagreement. And only a response of the kind supported here strikes the right balance between allowing individual researchers to freely pursue what seems to them to be worthwhile projects, and requiring that they pursue those research projects that the community of experts as a whole believes to be likely to yield significant results. (shrink)

Matthias Steup has developed a compatibilist account of doxastic control, according to which one’s beliefs are under one’s control if and only if they have a “good” causal history. Paradigmatically good causal histories include being caused to believe what one’s evidence indicates, whereas bad ones include those that indicate that the believer is blatantly irrational or mentally ill. I argue that if this is the only kind of control that we have over our beliefs, then our beliefs are not properly (...) subject to epistemic evaluation in deontological terms. I take as premises the claims (1) that acts which violate a deontic standard must be under the control of the agent that performs them, and (2) that deontic standards are deontic standards only if there is both something that it is to comply with them, and something that it is to violate them. The argument proceeds by showing that any belief which one might take to violate a deontic standard of a distinctively epistemic kind has a “bad” causal history, and so is, according to the compatibilist account, not under our control. Since these beliefs are not under our control, it follows from premise (1) that they do not violate any deontic standards of a distinctively epistemic kind. It then follows, from premise (2), that there are no deontic standards, of a distinctively epistemic kind, that govern belief. So if we have only compatibilist control over our beliefs, our beliefs are not properly subject to epistemic evaluation in deontological terms. (shrink)

Contemporary naturalism has two components. The first is ontological, and says, roughly, that all and only what the sciences say exists, really does exist. The other is methodological, and it says that only scientific explanations are legitimate explanations. Together these commitments promise a coherent picture of the world that is nicely integrated with an attractive epistemology. Despite the obvious appeal of naturalism, I would like to sound a note of caution. First, I would like to argue that naturalism's ontological commitment (...) cannot be vindicated. Not, that is, that it is false; rather, I argue that any attempt to show that it is true presupposes that it is not. Second, I argue that methodological naturalism is false. But, again, the problem is not straightforward. I will not claim that there are gaps in the explanations offered by science, such that the scientific project would be incomplete without emendation. Instead I argue that the goodness of an explanation depends, in part, on how the event to be explained is described, and that, some events, under some descriptions, call for non-scientific explanations. (shrink)

The Cartesian skeptic’s strategy is to tell a story about the world that is entirely consistent with all of the empirical evidence that we do, or can, have, but according to which many or all of our ordinary beliefs are false. He then suggests that, since we cannot show that his story is false, we ought to surrender those beliefs. In this paper I offer a decision-theoretic response to skepticism. Say that a cognitive attitude is a propositional attitude that may (...) be true or false. I argue that rejecting the skeptic’s story, and so retaining our ordinary opinions, will yield for us true cognitive attitudes, no matter whether the skeptic’s story is true or false, and that the best any alternative can do is yield no cognitive attitudes at all. Hence, it is rational to retain our ordinary opinions. One may be concerned that I can maintain this surprising conclusion because the cognitive attitudes at issue are not real beliefs, and do not represent the real world. I conclude the paper by arguing that this concern is misplaced. (shrink)

Normative inferentialism is a semantic theory according to which the meaning of an expression is, or is determined by, its proper inferential role. Critics of inferentialism often argue that it violates the principle of compositionality, and that it is therefore unable to explain some important linguistic data. I have two tasks in this paper: the first is to demonstrate that inferentialism, appearances perhaps to the contrary, does not violate the principle of compositionality, and the second is to explain why this (...) demonstration is unlikely to mollify critics of inferentialism. The dispute between inferentialists and their critics, I shall argue, is not about compositionality, it is a more fundamental dispute about the nature of representations. (shrink)

Much contemporary thinking about language is animated by the idea that the core function of language is to represent how the world is and that therefore the notion of representation should play a fundamental explanatory role in any explanation of language and language use. Leading thinkers in the field explore various ways this idea may be challenged as well as obstacles to developing various forms of anti-representationalism. Particular attention is given to deflationary accounts of truth, the role of language in (...) expressing mental states, and the normative and the natural as they relate to issues of representation. The chapters further various fundamental debates in metaphysics--for example, concerning the question of finding a place for moral properties in a naturalistic world-view--and illuminate the relation of the recent neo-pragmatist revival to the expressivist stream in analytic philosophy of language. (shrink)

Much contemporary thinking about language is animated by the idea that the core function of language is to represent how the world is and that therefore the notion of representation should play a fundamental explanatory role in any explanation of language and language use. The chapters in this volume explore various ways this idea may be challenged as well as obstacles to developing various forms of anti- representationalism. Particular attention is given to deflationary accounts of truth, the role of language (...) in expressing mental states, and the normative and the natural as they relate to issues of representation. The chapters further various fundamental debates in metaphysics—for example, concerning the question of finding a place for moral properties in a naturalistic world view—and illuminate the relation of the recent neo- pragmatist revival to the expressivist stream in analytic philosophy of language. (shrink)

Abstract:This paper considers the relevance of the concept of “eugenics,”—a term associated with some of the most egregious crimes of the twentieth century—to the possibility of editing human genomes. The author identifies some uses of gene editing as eugenics but proposes that this identification does not suffice to condemn them. He proposes that we should distinguish between “morally wrong” practices, which should be condemned, and “morally problematic” practices that call for solutions, and he suggests that eugenic uses of gene editing (...) fall into this latter category. Although when we choose the characteristics of future people we are engaging in morally dangerous acts, some interventions in human heredity should nevertheless be acknowledged as morally good. These morally good eugenic interventions include some uses of preimplantation genetic diagnosis. The author argues that we should think about eugenic interventions in the same way that we think about morally problematic interventions in public health. When we recognize some uses of gene editing as eugenics, we make the dangers of selecting or modifying human genetic material explicit. (shrink)

Comparing the ways in which the French cultural theorist Paul Virilio and the Australian cybernetic performance artist Stelarc criticize or defend technological cultural practices, this article argues that Virilio's ambiguous responses to avant-garde art highlight his key ideas far move clearly than his single-minded critique of 'termninal' mass-cultural practices - without any relationship to art - in Polar Inertia and Open Sky. Virlio's The Art of the Motor attacks the strategies of 20th-century technological avant- gardes for their apparent eugenicist and (...) fascist sympathies, but Virilios more recent interviews in Cyberrnonde and Voage d'hiver defend the ways in which the pre-technological avant-gardes resist the realist perspective of photography, and argue that the technological arts should offer similar resistance to the impact of cyberculture. Whereas Virilio concludes that such technological 'resistance' is probably impossible, Stelarc's recent research identifies exceptions to Virilio's rules. (shrink)

This paper defends the idea that there might be vagueness or indeterminacy in the world itself---as opposed to merely in our representations of the world---against the charges of incoherence and unintelligibility. First we consider the idea that the world might contain vague *properties and relations*; we show that this idea is already implied by certain well-understood views concerning the semantics of vague predicates (most notably the fuzzy view). Next we consider the idea that the world might contain vague *objects*; we (...) argue that an object is indeterminate in a certain respect (colour, size, etc.) just in case it is a borderline case of a maximally specific colour (size, etc.) property. Finally we consider the idea that the *world as a whole* might be indeterminate; we argue that the world is indeterminate just in case it lacks a determinate division into determinate objects. (shrink)

1. Introduction 2. Reward-Guided Decision Making 3. Content in the Model 4. How to Deflate a Metarepresentational Reading Proust and Carruthers on metacognitive feelings 5. A Deflationary Treatment of RPEs? 5.1 Dispensing with prediction errors 5.2 What is use of the RPE focused on? 5.3 Alternative explanations—worldly correlates 5.4 Contrast cases 6. Conclusion Appendix: Temporal Difference Learning Algorithms.

The distinction between literal and figurative use is well-known and embedded in ‘folk linguistics’. According to folk linguistics, figurative uses deviate from literal ones. But recent work on lexical modulation and polysemy shows that meaning deviation is ubiquitous, even in cases of literal use. Hence, it has been argued, the literal/figurative distinction has no value for theorising about communication. In this paper, we focus on metaphor and argue that here the literal–figurative distinction has theoretical importance. The distinction between literal and (...) metaphorical needs to be captured by our account of communication because literal uses transmit information in a way that metaphorical ones do not. We argue that there is a way to explain the literal/metaphorical distinction that preserves the core of the folk-linguistic idea and gives the distinction theoretical relevance. We propose that literal uses of a word are made with the intention to conform to an established practice of use, while metaphorical uses do not so conform, but depend on this pre-existing practice. Our account can deal with data that are problematic for other theories. A further advantage is that it extends naturally to other non-literal uses of words, including metonymy. (shrink)

There is a debate about the possibility of mind-uploading – a process that purportedly transfers human minds and therefore human identities into computers. This paper bypasses the debate about the metaphysics of mind-uploading to address the rationality of submitting yourself to it. I argue that an ineliminable risk that mind-uploading will fail makes it prudentially irrational for humans to undergo it.

Postmodernism and poststructuralism challenge fundamental positions in social theory. This book sets out some of the components of a postmodern social theory of health and healing, deriving from theorists including Derrida, Deleuze and Guattari, Foucault, Cixous and Kristeva. Nicholas J. Fox observes that the knowledge of the medical profession about the body, illness and health supplies the basis for medical dominance. The body of the patient is inscribed by discourses of professional `care,' an interaction which subjectifies the patient. Fox (...) explores the character of this power - and how it may be, and is, resisted. The book illustrates with detailed examples how the organization of health care and the caring relationship itself are sites for this contestation of power. Elements of feminist theory, and Derridean concepts of diffrance and intertextuality, supply the framework for the politics and ethics of the postmodern position on health. Deleuze and Guattari's radical challenge to psychoanalysis and familial repetitions within the healer/patient contact allows a re-reading of central ideas in medical sociology. While focusing upon the possibilities of postmodern social theory, the book demands a reappraisal of issues of structure, identity and knowledge in modernist medical sociology. Modernist sociology, Fox suggests, has been complicit in the creation of `the patient,' and of 'health' and 'illness.' Written with an emphasis on accessibility, this book explores the practical consequences of postmodern theory as well as familiarizing the reader with the concepts of postmodernism. (shrink)

This article provides a speech act analysis of ‘crime-enacting’ provisions in criminal statutes, focusing on the illocutionary force of these provisions. These provisions commonly set out not only particular crimes and their characteristics but also their associated penalties. Enactment of a statute brings into force new social facts, typically norms, through the official utterance of linguistic material. These norms are supposed to guide behaviour: they tell us what we must, may, or must not do. Our main claim is that the (...) illocutionary force of such provisions is primarily ‘world-creating’, i.e. effective, or declarational, rather than directive. We assume that directive illocutionary force is either direct or indirect, showing that provisions need not contain the linguistic items that make for direct directives and that according to standard tests no indirect directive is present. A potential counter-argument is that any utterance serving to direct behaviour is necessarily a directive. We show that this behaviour-directing property is shared by some clear non-directives. (shrink)

Economics takes an individualistic approach to human behaviour. This is reflected in the use of 'contingent valuation' surveys to conduct cost benefit analysis for economic policy evaluation. An individual's valuation of a policy is assumed to be unaffected by the burdens it places on others. We report a survey experiment to test this supposition in the context of climate change policy. Willingness to pay for climate change mitigation was higher when richer individuals were to bear higher costs than when, as (...) is usual, no explicit information was provided about cost distribution. This result is inconsistent with the usual interpretation of contingent valuation data. It also suggests that the data may be biased indicators of policy acceptance. Additional survey questions suggest that a collective mode of reasoning is common. (shrink)

A common theme in the historiography of Greek ethics says that modern ethics is characterized by imperative notions such as ‘duty’—and with a Judeo‐Christian notion of imperatives or commands issued by god—whereas ancient ethics supposedly deals mainly with ‘attractive notions such as ‘good’ and ‘virtue’. This thought is often juxtaposed with the idea that imperative notions betoken a conflict between one's duty and one's good, because an imperative seems to be required only to command people to do what they do (...) not wish to do, which is felt to be inappropriate to the supposed eudaimonism of Greek ethics. In fact Greek ethics makes substantial use of imperative notions, and does not attempt to show that attractive notions are more basic or preferable in ethics. (shrink)

Thus far it has been shown that Greek ethics is not as different from modern ethics as is commonly held, and that we cannot oppose a harmonizing Greek ethical outlook with a modern view that involves a conflict between happiness and adherence to ethical standards. Greek ethics has universalistic features—though they are different from the egalitarian characteristics of modern positions and do not focus on the notion of benevolence in the way that modern ethics does—and it mostly distinguishes self‐referential and (...) non‐self‐referential aims. Moreover, there is not in any case such a thing as a monolithic ethical view of ‘the Greeks’, and the idea of a straightforward contrast between ‘ancient’ and ‘modern’ ethics is best avoided. (shrink)

One of the main vehicles for the reconciliation of individual and social happiness that has supposedly been characteristic of Greek ethics is the concept of the polis. In the Hegelian tradition it has been thought that the Greeks reduced all norms and values to standards laid down by and for the city‐state, and that this fact made it possible for them to hold that the well‐being of an individual is entirely compatible with the well‐being of his fellow‐citizens and of the (...) city‐state as a whole. However, the Greeks were familiar with and attached substantial weight to non‐polis‐based norms and values, even universalist ones like the Golden Rule, and made ample conceptual room for conflict between the happiness of a citizen and that of his community. Moreover, Greek attitudes on this issue were variegated, not uniform. (shrink)

This introductory article is structured around the following themes: it begins with a brief overview of some important works that have paved the way for the present discussion. This is followed by a critique of the concept of “experience” and the philosophies based on it, that was first presented by feminist thinkers Joan Scott and Judith Butler in the 1980’s. The question this debate poses to the discussions in this book is whether focusing on experience is still a philosophically viable (...) option. After that, the views of Edmund Husserl – often described as “the father of phenomenology” – are presented on the particular themes of motherhood and pregnancy, as it is often overlooked that he had anything to say on the topic. Then follows a short outline of the structure of the experience of pregnancy, and also the modest suggestion that pregnancy should be seen not only as “split subjectivity” but also as a specific mode of phenomenological “in-between.” Thereafter the question is taken up whether pregnancy as a philosophical topic might also affect the methodological core of phenomenology. The article ends with a speculative outlook towards certain themes that have developed as a consequence of thinking pregnancy philosophically. (shrink)

This book advocates a radical shift of concern in philosophical, historical, and sociological studies of the sciences, and explores the consequences of such a shift. The historically-oriented first part of the work deals with the ways in which ranges of questions become real and cease to be real for communities of inquirers. The more philosophically-oriented second part of the work introduces the notion of absolute reality of questions, and addresses doubt about the claims of the sciences to have accumulated absolutely (...) real questions. It is argued that recent studies in the sociology and social history of the science pose strong challenges to the sciences by revealing how appeals to authority, vested interests, and rhetorical and aesthetic sensibilities play substantial roles in the practices of the sciences. The final chapter defends the pragmatic stance of the work, and of its companion, The Fortunes of Inquiry, and draws morals about the roles of criticism and reflection in the philosophy of science and in the sciences themselves. (shrink)

A clearly written, authoritative introduction to Taylor's work.

Aesthesis, the classical term for sensing and perceiving, is at the heart of innumerable problems that plague global society. The purpose of this article is to open a conversation on aesthesis. We survey the roots and relevance of aesthesis as a direct albeit contested relation and engagement with the world and with Others. From its pre-Socratic origins, aesthesis has been both a pragmatic, somatic concept, prompting a re-evaluation of the distinction between experience and abstraction. We trace its ongoing repression from (...) Plato through ‘western’ theories of formal Aesthetics. Drawing on a relational interpretation of Protagoras’ aesthesis, we argue that modern pragmatists and radical empiricists, as well as more contemporary critics of the ‘colonization’ of aesthesis (Mignolo and Vasquez) by formal Aesthetics recognize and develop the relational and ethical aspects of aesthesis. We consider the role of the body, affect, and of the intangible or virtual qualities of aesthesis. The ethics of obligations (Weil) in the polis (Arendt) shows how aesthesis informs politics despite its repression in favour of moral and legal norms. We argue this is relevant to contemporary crises such as xenophobia and ecocidal climate warming. (shrink)

There is increasing evidence for epigenetically mediated transgenerational inheritance across taxa. However, the evolutionary implications of such alternative mechanisms of inheritance remain unclear. Herein, we show that epigenetic mechanisms can serve two fundamentally different functions in transgenerational inheritance: (i) selection-based effects, which carry adaptive information in virtue of selection over many generations of reliable transmission; and (ii) detection-based effects, which are a transgenerational form of adaptive phenotypic plasticity. The two functions interact differently with a third form of epigenetic information transmission, (...) namely information about cell state transmitted for somatic cell heredity in multicellular organisms. Selection-based epigenetic information is more likely to conflict with somatic cell inheritance than is detection-based epigenetic information. Consequently, the evolutionary implications of epigenetic mechanisms are different for unicellular and multicellular organisms, which underscores the conceptual and empirical importance of distinguishing between these two different forms of transgenerational epigenetic effect. (shrink)

Fodor’s radical concept nativism flowed from his view that hypothesis testing is the only route to concept acquisition. Many have successfully objected to the overly-narrow restriction to learning by hypothesis testing. Existing representations can be connected to a new representational vehicle so as to constitute a sustaining mechanism for a new representation, without the new representation thereby being constituted by or structured out of the old. This paper argues that there is also a deeper objection. Connectionism shows that a more (...) fundamental assumption underpinning the debate can also be rejected: the assumption that the development of a new representation must be explained in content-involving terms if innateness is to be avoided. -/- Fodor has argued that connectionism offers no new resources to explain concept acquisition: unless it is merely an uninteresting claim about neural implementation, connectionism’s defining commitment to distributed representations reduces to the claim that some representations are structured out of others (which is the old, problematic research programme). Examination of examples of representational development in connectionist networks shows, however, that some such models explain the development of new representational capacities in non-representational terms. They illustrate the possibility of representational development that is not explicable-by-content. Connectionist representations can be distributed in an important sense, which is incompatible with the assumption of explanation-by-content: they can be distributed over non-representational resources that account for their development. Rejecting the assumption of explanation-by-content thereby opens up a more radical way of rejecting Fodor’s argument for radical concept nativism. (shrink)

This paper advocates explicitness about the type of entity to be considered as content- bearing in connectionist systems; it makes a positive proposal about how vehicles of content should be individuated; and it deploys that proposal to argue in favour of representation in connectionist systems. The proposal is that the vehicles of content in some connectionist systems are clusters in the state space of a hidden layer. Attributing content to such vehicles is required to vindicate the standard explanation for some (...) classificatory networks’ ability to generalise to novel samples their correct classification of the samples on which they were trained. (shrink)