Joëlle Proust

Metacognition is often defined as thinking about thinking. It is exemplified in all the activities through which one tries to predict and evaluate one’s own mental dispositions, states and properties for their cognitive adequacy. This article discusses the view that metacognition has metarepresentational structure. Properties such as causal contiguity, epistemic transparency and procedural reflexivity are present in metacognition but missing in metarepresentation, while open-ended recursivity and inferential promiscuity only occur in metarepresentation. It is concluded that, although metarepresentations can redescribe metacognitive contents, metacognition and metarepresentation are functionally distinct

Le déterminisme implique que le libre arbitre n’existe pas, que nous ne pouvons pas faire autrement ; réciproquement, avoir la possibilité de faire autrement implique que le déterminisme ne s’applique pas à l’instant, s’il existe, où on l’exerce. Cependant, la question de la responsabilité rend difficile d’accepter que les agents ne puissent pas faire autrement et motive fortement à rendre compatibles déterminisme et libre arbitre ou à soutenir, dans une veine « incompatibiliste », que le cerveau humain n’est pas soumis à la loi du déterminisme. Le libertarianisme, c’est-à-dire l’idée que, chez l’agent humain, il existe des exceptions au déterminisme causal lors de l’activité cérébrale, n’est pas défendable à la lumière des découvertes récentes sur la cognition humaine. Selon le compatibilisme, un agent n’ayant pas la possibilité de faire autrement jouit encore de son Libre Arbitre. La théorie de Harry Frankfurt d’une volonté à plusieurs niveaux est étudiée de manière critique. On présente la preuve neuroscientifique d’un modèle de volonté en cascade, dans l’esprit de Frankfurt. Ce modèle permet l’émergence d’une forme de comptabilisme dans lequel l’agent peut plus ou moins contrôler ce qu’elle fait, dépendre plus ou moins de son environnement et dispose donc plus ou moins de son libre arbitre

_Questions of Form _was first published in 1989. Minnesota Archive Editions uses digital technology to make long-unavailable books once again accessible, and are published unaltered from the original University of Minnesota Press editions. In _Questions on Form_, Joelle Proust traces the concept of the analytic proposition from Kant's development of the notion down to its place in the work of Rudolf Carnap, a founder of logical empiricism and a key figure in contemporary analytic philosophy. Using a method known in France as _topique comparative_,she provides a rigorous exposition of analyticity, situating it within four major philosophical systems—those of Kant, Bolzano, Frege, and Carnap—and clearly delineating its development from one system to the next. Proust takes as her point of departure Kant's distinction between analytic and synthetic judgments. Though she makes clear that Kant drew on Locke, Hume, and Leibniz, she argues that his notion of analyticity was innovative, not simply an elaboration of something already found in their work. She shows that the analytic proposition unexpectedly (given its modest status in Kant) came to play an important part in efforts to convert problems considered "transcendental" into questions of belonging to formal logic. Ultimately, her comparison of their systems reveals that the concept of the analytic, however specific its rile in each, remains linked to a foundationalist strategy—in effect, to the transcendentalist questions Kant used when he reinterpreted the findings of his empiricist predecessors. Hence, this book's provocative claim: today's so-called logical empiricism owes much more to Kant's notion of science than to Hume's.

Against the prior view that primate communication is based only on signal decoding, comparative evidence suggests that primates are able, no less than humans, to intentionally perform or understand impulsive or habitual communicational actions with a structured evaluative nonconceptual content. These signals convey an affordance-sensing that immediately motivates conspecifics to act. Although humans have access to a strategic form of propositional communication adapted to teaching and persuasion, they share with nonhuman primates the capacity to communicate in impulsive or habitual ways. They are also similarly able to monitor fluency, informativeness and relevance of messages or signals through nonconceptual cues.

Perceivers generally show a poor ability to detect changes, a condition referred to as “Change Blindness” . They are, in addition, “blind to their own blindness”. A common explanation of this “Change Blindness Blindness” is that it derives from an inadequate, “photographical” folk-theory about perception. This explanation, however, does not account for intra-individual variations of CBB across trials. Our study aims to explore an alternative theory, according to which participants base their self-evaluations on two activity-dependent cues, namely search time and perceived success in prior trials. These cues were found to influence self-evaluation in two orthogonal ways: success-feedback influenced self-evaluation in a global, contextual way, presumably by recalibrating the norm of adequacy for the task. Search time influenced it in a local way, predicting the success of a given trial from its duration

This articles examines three ways in which the connection between semantic and pragmatic representations of a single action can be tightened up in order to remedy the puzzle of deviant causation. A first move consists in making the feedback process, i.e. the dynamics of the relationship between both representational components, a central element in the definition of an action. A second step brings in the action-effect principle, emphasizing the teleological relation of each pragmatic representation type with its external effects. A final step consists in elucidating the constitutive character of demonstrative reference for the contents of working memory states

Dans son compte-rendu de mon livre, Les Animaux Pensent-ils?, Machery objecte que l'évolution n'étant ni hiérarchique ni linéaire, il n'et pas justifié de proposer une analyse hiérarchique des représentations. Je réponds à cette objection, en montrant qu'on peut en effet distinguer des types de représentation par leurs propriétés sémantiques et computationnelles. On peut reconnaître le caractère anagénétique du développement de la cognition sans pour autant légitimer une conception hiérarchique et continuiste de l'évolution des espèces

This chapter examines whether, and in what sense, one can speak of agentive mental events. An adequate characterization of mental acts should respond to three main worries. First, mental acts cannot have pre-specified goal contents. For example, one cannot prespecify the content of a judgment or of a deliberation. Second, mental acts seem to depend crucially on receptive attitudes. Third, it does not seem that intentions play any role in mental actions. Given these three constraints, mental and bodily actions appear to have a significantly different structure. A careful analysis of the role of normative requirements, distinguishing them from instrumental reasons, allows the distinction between mental and bodily forms of action to be clarified. Two kinds of motives must be present for a mental act to develop. The first kind is instrumental: a mental act is performed because of some basic informational need, such as the need to "remember the name of that play". The second kind is normative: given the specific type of mental action performed, there is a specific epistemic norm relevant to that act. These two motives actually correspond to different phases of a single mental act. The first motivates the mental act, i.e. makes salient the corresponding goal. The second offers an evaluation of the feasibility of the act, on the basis of its constitutive normative requirement. Conceived in this way, a characterization of mental acts avoids the three difficulties mentioned above. The possibility of pre-specifying the outcome of epistemic mental acts is blocked by the fact that such acts are constituted by strict normative requirements. That mental acts include receptive features is shown to be a necessary architectural constraint for mental agents to be sensitive to epistemic requirements. Finally, the phenomenology of intending is shown to be absent in most mental acts; the motivational structure of mental acts is, rather, associated with error- signals and self-directed doubtings. Mental acts need to be recognized as a natural kind of action meant to normatively control and enhance cognitive efficiency according to current processing needs

This chapter aims to relate two fields of research that have been rarely – if ever – associated, namely embodied communication and metacognition. Exploring this relationship offers a new perspective for understanding the relationship between self-knowledge and mindreading. "Embodied communication" refers to the process of conveying information to one or several interlocutors through speech and associated bodily gestures, or through gestures only. It is prima facie plausible that embodied communication crucially involves metacognitive interventions. Let the term ‘conversational metacognition’ refer to the set of abilities that allow an embodied speaker to make available to others and to receive from them specific markers concerning his/her "conversing adequacy". The hypothesis explored is that embodied communication in humans involve metacognitive gestures. Examples are offered from manual gesturing and from orofacial expressions. A final discussion examines the respective roles of altruistic and Machiavellian pressures in conversational metacognition.

Given disagreements about the architecture of the mind, the nature of self-knowledge, and its epistemology, the question of how to understand the function and scope of metacognition – the control of one's cognition - is still a matter of hot debate. A dominant view, the self-ascriptive view (or one-function view), has been that metacognition necessarily requires representing one's own mental states as mental states, and, therefore, necessarily involves an ability to read one's own mind. The self-evaluative view (or two-function view), in contrast, takes metacognition to involve a procedural form of knowledge that is generated by actually engaging in a first-order cognitive task, and monitoring its success. The comparative and developmental arguments supporting, respectively, each of these views are discussed in the light of Hampton's operational definition of metacognition. New arguments are presented in favor of the two-function view. Recent behaviorial and neuroscientific evidence suggests that metacognitive assessment relies on dedicated implicit mechanisms, which are wholly independent, and indeed dissociable, from theory-based self-attribution. The two-function view is claimed to be the best interpretation of these findings.

Growing suspicions were raised however that an exclusively language-oriented view of the mind, focussing on the characterization of anhistorical, static mental states through their propositional contents, was hardly compatible with what is currently known of brain architecture and did not fare well when confronted with results from many behavioral studies of mental functions. My aim in what follows is to show that these forms of dissatisfaction stem from the fact that brain evolution and development were either entirely ignored, or insufficiently taken into account in inquiries about the structure of mental contents. I will discuss how evolutionary and developmental approaches to human cognition are now in a position to substantially alter the central paradigms currently used in cognitive science.