This paper is an attempt to put the work of the past several decades on the problems of implicit learning and unconscious cognition into an evolutionary context. Implicit learning is an inductive process whereby knowledge of a complex environment is acquired and used largely independently of awareness of either the process of acquisition or the nature of that which has been learned. Characterized this way, implicit learning theory can be viewed as an attempt to come to grips with the classic (...) epistemological issues of knowledge acquisition, representation and use. The argument is made that the process, despite its seeming cognitive sophistication, is of considerable evolutionary antiquity and that it antedates awareness and the capacity for conscious control of mentation. Various classic heuristics from evolutionary biology are used to substantiate this claim and several specific entailments of this line of argument are outlined. (shrink)

In recent decades it has become increasingly clear that a substantial amount of cognitive work goes on independent of consciousness. The research has been carried out largely under two rubrics, implicit learning and implicit memory. The former has been concerned primarily with the acquisition of knowledge independent of awareness and the latter with the manner in which memories not readily available to conscious recall or recognition play a role in behavior; collectively these operations comprise the essential functions of the cognitive (...) unconscious. This paper reviews the recent history of work on these issues, identifies some of the problems confronting researchers, and introduces a theoretical framework based on principles of evolutionary biology within which to view them. The argument is that the cognitive unconscious, despite its apparent sophistication, is of considerable evolutionary antiquity and antedates conscious cognitive systems by a considerable amount of time. Various entailments of this evolutionary perspective are explored including such issues as phylogeny and ontogeny of implicit processes, the robustness of the implicit functions as displayed by the capacity to resist disruptions from psychological and neurological disorders, the relationship between implicit cognition and intelligence, and individual differences in implicit cognitive abilities. (shrink)

A number of ways of taxonomizing human learning have been proposed. We examine the evidence for one such proposal, namely, that there exist independent explicit and implicit learning systems. This combines two further distinctions, (1) between learning that takes place with versus without concurrent awareness, and (2) between learning that involves the encoding of instances (or fragments) versus the induction of abstract rules or hypotheses. Implicit learning is assumed to involve unconscious rule learning. We examine the evidence for implicit learning (...) derived from subliminal learning, conditioning, artificial grammar learning, instrumental learning, and reaction times in sequence learning. We conclude that unconscious learning has not been satisfactorily established in any of these areas. The assumption that learning in some of these tasks (e.g., artificial grammar learning) is predominantly based on rule abstraction is questionable. When subjects cannot report the rules that govern stimulus selection, this is often because their knowledge consists of instances or fragments of the training stimuli rather than rules. In contrast to the distinction between conscious and unconscious learning, the distinction between instance and rule learning is a sound and meaningful way of taxonomizing human learning. We discuss various computational models of these two forms of learning. (shrink)

69 Thompson-Schill, S.L. _et al. _(1997) Role of left inferior prefrontal cortex 59 Buckner, R.L. _et al. _(1996) Functional anatomic studies of memory in retrieval of semantic knowledge: a re-evaluation _Proc. Natl. Acad._ retrieval for auditory words and pictures _J. Neurosci. _16, 6219–6235 _Sci. U. S. A. _94, 14792–14797 60 Buckner, R.L. _et al. _(1995) Functional anatomical studies of explicit and 70 Baddeley, A. (1992) Working memory: the interface between memory implicit memory retrieval tasks _J. Neurosci. _15, 12–29 and cognition (...) _J. Cogn. Neurosci. _4, 281–288 61 Bäckman, L. _et al. _(1997) Brain activation in young and older adults 71 Petrides, M. (1994) Frontal lobes and behavior _Curr. Opin. Neurobiol._ during implicit and explicit retrieval _J. Cogn. Neurosci. _9, 378–391. (shrink)

Linguists take the intuitive judgments of speakers to be good evidence for a grammar. Why? The Chomskian answer is that they are derived by a rational process from a representation of linguistic rules in the language faculty. The paper takes a different view. It argues for a naturalistic and non-Cartesian view of intuitions in general. They are empirical central-processor responses to phenomena differing from other such responses only in being immediate and fairly unreflective. Applying this to linguistic intuitions yields an (...) explanation of their evidential role without any appeal to the representation of rules. Introduction The evidence for linguistic theories A tension in the linguists' view of intuitions Intuitions in general Linguistic intuitions Comparison of the modest explanation with the standard Cartesian explanation A nonstandard Cartesian explanation of the role of intuitions? Must linguistics explain intuitions? Conclusion. (shrink)

In this paper I criticize the most significant recent examples of the practical knowledge analysis of knowledge-how in the philosophical literature: David Carr [1979, Mind, 88, 394–409; 1981a, American Philosophical Quarterly, 18, 53–61; 1981b, Journal of Philosophy of Education, 15(1), 87–96] and Stanley & Williamson [2001, Journal of Philosophy, 98(8), 411–444]. I stress the importance of know-how in our contemporary understanding of the mind, and offer the beginnings of a treatment of know-how capable of providing insight in to the use (...) of know-how in contemporary cognitive science. Specifically, I claim that Carr’s necessary conditions for know-how fail to capture the distinction he himself draws between ability and knowing-how. Moreover, Carr ties knowing-how to conscious intent, and to an explicit knowledge of procedural rules. I argue that both moves are mistakes, which together render Carr’s theory an inadequate account both of common ascriptions of knowledge-how and of widely accepted ascriptions of knowledge-how within explanations in cognitive science. Finally, I note that Carr’s conditions fail to capture intuitions (heshares) regarding the ascription of know-how to persons lacking ability. I then consider the position advocated by Stanley & Williamson (2001), which seems avoid Carr’s commitments to conscious intent and explicit knowledge while still maintaining that “knowledge-how is simply a species of knowledge-that" (Stanley & Williamson, 2001, p. 411). I argue that Stanley and Williamson’s attempt to frame a reductionist view that avoids consciously occurrent beliefs during exercises of knowledge-how and explicit knowledge of procedural rules is both empirically implausible and explanatorily vacuous. In criticizing these theories I challenge the presuppositions of the most pervasive response to Ryle in the philosophic literature, what might be described as “the received view." I also establish several facts about knowing-how. First, neither conscious intent nor explicit representation (much less conscious representation) of procedural rules are necessary for knowing-how given the theory of cognition current in cognitive science. I argue that the discussed analyses fail to capture the necessary conditions for knowledge-how because know-how requires the instantiation of an ability and of the capacities necessary for exploiting an ability—not conscious awareness of purpose or explicit knowledge of rules. Second, one must understand knowledge-how as task-specific, i.e., as presupposing certain underlying conditions. Conceiving of know-how as task-specific allows one to understand ascriptions of know-how in the absence of ability as counterfactual ascriptions based upon underlying competence. (shrink)

In this paper, I submit that it is the controlled part of skilled action, that is, that part of an action that accounts for the exact, nuanced ways in which a skilled performer modifies, adjusts and guides her performance for which an adequate, philosophical theory of skill must account. I will argue that neither Jason Stanley nor Hubert Dreyfus have an adequate account of control. Further, and perhaps surprisingly, I will argue that both Stanley and Dreyfus relinquish an account of (...) control for precisely the same reason: each reduce control to a passive, mechanistic, automatic process, which then prevents them from producing a substantive account of how controlled processes can be characterized by seemingly intelligent features and integrated with personal-level states. I will end by introducing three different kinds of control, which are constitutive of skilled action: strategic control, selective, top–down, automatic attention, and motor control. It will become clear that Dreyfus cannot account for any of these three kinds of control while Stanley has difficulty tackling the two latter kinds. (shrink)

Naturalistic thinking and knowing, the tacit, experiential, and intuitive reasoning of everyday interaction, have long been regarded as inferior to formal reason and labeled primitive, fallible, subjective, superstitious, and in some cases ineffable. But, naturalistic thinking is more rational and definable than it appears. It is also relevant to design. Inquiry into the mechanisms of naturalistic thinking and knowledge can bring its resources into focus and enable designers to create better, human-centered designs for use in real-world settings. This article makes (...) a case for the explicit, formal study of implicit, naturalistic thinking within the fields of design. It develops a framework for defining and studying naturalistic thinking and knowledge, for integrating them into design research and practice, and for developing a more integrated, consistent theory of knowledge in design. It will (a) outline historical definitions of knowledge, attitudes toward formal and naturalistic thinking, and the difficulties presented by the co-presence of formal and naturalistic thinking in design, (b) define and contrast formal and naturalistic thinking as two distinct human cognitive systems, (c) demonstrate the importance of naturalistic cognition in formal thinking and real-world judgment, (d) demonstrate methods for researching naturalistic thinking that can be of use in design, and (e) briefly discuss the impact on design theory of admitting naturalistic thinking as valid, systematic, and knowable. (shrink)

The German discussion about experience-guided work has led to the question of how work experience can be regarded within the process of designing technical artifacts. This paper offers a solution for the area of skilled maintenance work. Some considerations about the nature of experience and about the problems skilled workers have in aquiring work competences within computer aided production environments are introduced in order to illustrate the design philosophy: A decision-support-system is described which stimulates workplace learning by enabling previous and (...) present users of the system, including maintenance staff and shop floor workers, to exchange and build on experience. (shrink)

This article explicates the interaction between implicit and explicit processes in skill learning, in contrast to the tendency of researchers to study each type in isolation. It highlights various effects of the interaction on learning (including synergy effects). The authors argue for an integrated model of skill learning that takes into account both implicit and explicit processes. Moreover, they argue for a bottom-up approach (first learning implicit knowledge and then explicit knowledge) in the integrated model. A variety of qualitative data (...) can be accounted for by the approach. A computational model, CLARION, is then used to simulate a range of quantitative data. The results demonstrate the plausibility of the model, which provides a new perspective on skill learning. (PsycINFO Database Record (c) 2012 APA, all rights reserved). (shrink)

A proposed empirical criterion for task susceptibility to introspective awareness distinguishes cognitive processes of which one cannot be aware from those of which one can be aware. The empirical criterion for task susceptibility to awareness effects proposes that there are tasks which cannot be affected by awareness of the rules constituting the tasks. These criteria were applied to research programmes in rule-learning in which past studies in the area of learning without awareness were included as well as current research in (...) implicit learning. The principal question addressed in these studies is whether or not rule-learning can occur without awareness. An historical review showed that rule-learning occurred in tasks which were both susceptible and insusceptible to introspective awareness and to awareness effects. Accordingly, it has been proposed that rather than attempt to decide theoretically and empirically between the opposing hypotheses—that of “cognitive learning” on the one hand, which assumes that awareness is a necessary condition for rule-learning, and that of “automatic learning” on the other, which assumes direct, automatic and unconscious processes—efforts should rather be directed toward developing a theoretical approach which is based on both conscious and unconscious processes. However, an approach of this kind encounters severe problems, such as the generation of contradictory predictions, which result from the employment of several incongruent and irreconcilable models of explanation. The criteria for task susceptibility offer a way out of these difficulties. (shrink)

Complete URL to this document: http://srsc.ulb.ac.be/axcWWW/93-Principles.html.

This article addresses issues in developing cognitive architectures--generic computational models of cognition. Cognitive architectures are believed to be essential in advancing understanding of the mind, and therefore, developing cognitive architectures is an extremely important enterprise in cognitive science. The article proposes a set of essential desiderata for developing cognitive architectures. It then moves on to discuss in detail some of these desiderata and their associated concepts and ideas relevant to developing better cognitive architectures. It argues for the importance of taking (...) into full consideration these desiderata in developing future architectures that are more cognitively and ecologically realistic. A brief and preliminary evaluation of existing cognitive architectures is attempted on the basis of these ideas. (shrink)

Although many studies have provided evidence that abstract knowledge can be acquired in artificial grammar learning, it remains unclear how abstract knowledge can be attained in sequence learning. To address this issue, we proposed a dual simple recurrent network model that includes a surface SRN encoding and predicting the surface properties of stimuli and an abstract SRN encoding and predicting the abstract properties of stimuli. The results of Simulations 1 and 2 showed that the DSRN model can account for learning (...) effects in the serial reaction time task under different conditions, and the manipulation of the contribution weight of each SRN accounted for the contribution of conscious and unconscious processes in inclusion and exclusion tests in previous studies. The results of human performance in Simulation 3 provided further evidence that people can implicitly learn both chunking and abstract knowledge in sequence learning, and the results of Simulation 3 confirmed that the DSRN model can account for how people implicitly acquire the two types of knowledge in sequence learning. These findings extend the learning ability of the SRN model and help understand how different types of knowledge can be acquired implicitly in sequence learning. (shrink)

The cognition of music, like that of language, is partly rooted in enculturative processes of implicit and incidental learning. Musicians and nonmusicians alike are commonly found to possess detailed implicit knowledge of musical structure which is acquired incidentally through interaction with large samples of music. This paper reports an experiment combining the methodology of artificial grammar learning with musical acquisition of melodic structure. Participants acquired knowledge of grammatical melodic structures under incidental learning conditions in both experimental and untrained control conditions. (...) Subsequent analysis indicates a large effect of unsupervised online learning in the experimental and control group throughout the course of the testing phase suggesting an effective ongoing learning process. Musicians did not outperform nonmusicians, indicating that musical expertise is not advantageous for the learning of a new, unfamiliar melodic system. Confidence ratings suggest that participants became aware of the knowledge guiding their classification performance despite the incidental learning conditions. (shrink)

A number of ways of taxonomizing human learning have been proposed. We examine the evidence for one such proposal, namely, that there exist independent explicit and implicit learning systems. This combines two further distinctions, between learning that takes place with versus without concurrent awareness, and between learning that involves the encoding of instances versus the induction of abstract rules or hypotheses. Implicit learning is assumed to involve unconscious rule learning. We examine the evidence for implicit learning derived from subliminal learning, (...) conditioning, artificial grammar learning, instrumental learning, and reaction times in sequence learning. We conclude that unconscious learning has not been satisfactorily established in any of these areas. The assumption that learning in some of these tasks is predominantly based on rule abstraction is questionable. When subjects cannot report the “implicitly learned” rules that govern stimulus selection, this is often because their knowledge consists of instances or fragments of the training stimuli rather than rules. In contrast to the distinction between conscious and unconscious learning, the distinction between instance and rule learning is a sound and meaningful way of taxonomizing human learning. We discuss various computational models of these two forms of learning. (shrink)

It is not rare in philosophy and psychology to see theorists fall into dichotomous thinking about mental phenomena. On one side of the dichotomy there are processes that I will label “unintelligent.” These processes are thought to be unconscious, implicit, automatic, unintentional, involuntary, procedural, and non-cognitive. On the other side, there are “intelligent” processes that are conscious, explicit, controlled, intentional, voluntary, declarative, and cognitive. Often, if a process or behavior is characterized by one of the features from either of the (...) above lists, the process or behavior is classified as falling under the category to which the feature belongs. For example, if a process is implicit this is usually considered sufficient for classifying it as “unintelligent” and for assuming that the remaining features that fall under the “unintelligent” grouping will apply to it as well. Accordingly, if a process or behavior is automatic, philosophers often consider it to be unintelligent. It is my goal in this paper to challenge the conceptual slip from “automatic” to “unintelligent”. I will argue that there are a whole range of properties highlighted by the existing psychological literature that make automaticity a much more complex phenomenon than is usually appreciated. I will then go on to discuss two further important relationships between automatic processes and controlled processes that arise when we think about automatic processes in the context of skilled behavior. These interactions should add to our resistance to classifying automaticity as unintelligent or mindless. In Sect. 1, I present a few representative cases of philosophers classifying automatic processes and behaviors as mindless or unintelligent. In Sect. 2, I review trends in the psychology of automaticity in order highlight a complex set of features that are characteristic, though not definitive, of automatic processes and behaviors. In Sect. 3, I argue that at least some automatic processes are likely cognitively penetrable. In Sect. 4, I argue that the structure of skilled automatic processes is shaped diachronically by practice, training and learning. Taken together, these considerations should dislodge the temptation to equate “automatic” with “unintelligent”. (shrink)

In recent years, a debate concerning the nature of knowing-how has emerged between intellectualists who claim that knowledge-how is reducible to knowledge-that and anti-intellectualists who claim that knowledge-how comprises a unique and irreducible knowledge category. The arguments between these two camps have clustered largely around two issues: intellectualists object to Gilbert Ryle's assertion that knowing-how is a kind of ability, and anti-intellectualists take issue with Jason Stanley and Timothy Williamson's positive, intellectualist account of knowing-how. Like most anti-intellectualists, in this paper (...) I will raise objections to Stanley and Williamson's account of knowing-how and also defend the claim that ability is necessary for knowing-how attributions. Unlike most discussions of knowing-how, however, I will return to more Rylean considerations in order to illustrate that any intellectualist account of knowing-how, not simply Stanley and Williamson's preferred variety, will fail because it will be unable to account for fundamental differences in the knowledge required to instantiate an ability and the knowledge involved in propositional thought. (shrink)

Why are linguistic intuitions good evidence for a grammar? In 'Intuitions in Linguistics' ([2006a]) and Ignorance of Language ([2006b]), I looked critically at some Chomskian answers and proposed another one. In this article, I respond to Fitzgerald's 'Linguistic Intuitions' ([2010]), a sweeping critique of my position, and to Culbertson and Gross' 'Are Linguists Better Subjects?' ([2009]), a criticism of one consequence of the position. In rejecting these criticisms, I emphasize that the issue over linguistic intuitions concerns only metalinguistic ones. And (...) I argue that my critics, like many others, make too much of the distinction between grammaticality and acceptability intuitions. (shrink)