Les troubles praxiques font partie des critères d’inclusion decertaines pathologies neurodégénératives et sont de nature à réduirel'autonomie de ces patients. Pourtant, les rares travaux dans cedomaine prennent rarement en compte les spécificités liées auxpathologies dégénératives (e.g., évolution, lésions diffuses,phénotypes différents), sans compter qu’il n’existe pas de consensusthéorique sur l’apraxie. Les troubles d'utilisation des outils ontlongtemps été expliqués par la perte de représentations stockées (e.g.,connaissances sur les objets ou les gestes) mais des modèles récentssuggèrent de distinguer l’utilisation proprement dite de l’usagecommunément admis. De même, l’apraxie est fréquemment mise enrapport avec la perte d’une représentation du corps mais une mise enperspective historique de la littérature permet de constater qu'il s'agitd'un postulat tout à fait récent. La présente thèse émet une propositionalternative en opposant à la notion de représentation corporelle desprocessus plus dynamiques et neurologiquement ancrés tels quel’intégration sensorielle. Au plan théorique, la présente thèse proposeainsi une mise en perspective historique et une revue critique de lalittérature, ainsi qu’une réflexion épistémologique opposant savoir etraisonnement, permettant de distinguer des troubles spécifiques dugeste et des répercussions "dans le geste" d'autres niveaux deperturbations. Au plan clinique, ce travail de neuropsychologieclinique et expérimentale examine ces différents niveaux dans uneperspective différentielle et montre que les dissociations envisagéessont cliniquement opérantes. / Apraxia is a diagnostic criterion for neurodegenerativediseases and can reduce autonomy in these patients. Yet, the fewstudies that have been done in this domain rarely take into accountspecificities of dementia (e.g., evolution, diffuse lesions, differentphenotypes) and there is no theoretical consensus on apraxia. Tooluse disorders are mainly explained by the loss of stored knowledge(e.g., about tools or gestures) but recent models suggest to distinguishthe actual use of tools from common usage. Likewise, apraxiai sfrequently linked with the loss of body representations but a historicalperspective of the literature demonstrates that it is a relatively recentassumption. The present doctoral thesis submits an alternativeproposal by opposing body representation and more dynamic andneurological processes, that is, sensory integration. Thus, at thetheoretical level this work proposes a historical perspective and acritical review of existing literature, as well as an epistemologicalreflection that contrasts knowledge and reasoning, and that makes adistinction between specific gesture disorders and consequences ofother cognitive impairments on gesture production. At a clinical level,this neuropsychological research examines these different levels ofimpairments in patients through a differential approach. Results showthat the aforementioned theoretical considerations are clinicallyvalidated.

Este trabalho teve como objetivo estudar o desenvolvimento do comportamento manipulativo em macacos-prego (Cebus apella) em semiliberdade, com ênfase na ontogênese do comportamento de quebra de cocos. A coleta de dados foi realizada ao longo de dois anos e meio, o que permitiu um acompanhamento longitudinal dos filhotes com idade inferior a dois anos. As dinâmicas sociais relacionadas à aprendizagem de quebra também foram estudadas. Os resultados mostram que os macacos começaram a manipular objetos com cerca de um mês de idade e conseguiram quebrar cocos com sucesso a partir dos dois anos. Eles observaram seus co-específicos quebrando cocos, sendo que os maiores observadores foram os imaturos, especialmente os juvenis. Os alvos de observação foram preferencialmente aqueles macacos que apresentaram uma maior taxa de quebra. Durante a observação de co-específicos, houve raros registros de agonismo, ou seja, há grande tolerância social. Houve intercalação entre brincadeira e quebra de cocos por macacos imaturos. Grande parte da aquisição do comportamento de quebra de cocos pode ser atribuída às experiências individuais, mas também há oportunidades para que ocorra a aprendizagem social por meio de observação direta do co-específico proficiente, já que os observadores são bem tolerados. / The objective of this work was to study the manipulative behavior of semifree-ranging tufted-capuchin monkeys (Cebus apella), focusing on the ontogenesis of nutcracking behavior. Data collection was done during a period of two years and a half, and so, a longitudinal study with monkeys under two-years old could be performed. Social dynamics related to nutcracking learning were also studied. The results show that the monkeys started manipulating objects when they were around a month of age, but only subjects over two years of age were able to sucessfully crack nuts. They observed conspecific nutcracking: immature monkeys were the main observers, especially juveniles. The main targets of observation were the monkeys who had the highest nutcracking rate. Agonism was rare during conspecific observation, what means that social tolerance was high. There were events in which immature monkeys alternated play and nutcracking behavior. An important part of nutcracking acquisition can be attributed to individual experiences, but there are also opportunities for social learning through direct observation of proficient conspecific, once observers are well tolerated.

The ability to use tools is one of the hallmarks of intelligence. Tool use is fundamental to human life and has been for at least the last two million years. We use tools to extend our reach, to amplify our physical strength, and to achieve many other tasks. A large number of animals have also been observed to use tools. Despite the widespread use of tools in the animal world, however, studies of autonomous robotic tool use are still rare. This dissertation examines the problem of autonomous tool use in robots from the point of view of developmental robotics. Therefore, the main focus is not on optimizing robotic solutions for specific tool tasks but on designing algorithms and representations that a robot can use to develop tool-using abilities. The dissertation describes a developmental sequence/trajectory that a robot can take in order to learn how to use tools autonomously. The developmental sequence begins with learning a model of the robot's body since the body is the most consistent and predictable part of the environment. Specifically, the robot learns which perceptual features are associated with its own body and which with the environment. Next, the robot can begin to identify certain patterns exhibited by the body itself and to learn a robot body schema model which can also be used to encode goal-oriented behaviors. The robot can also use its body as a well defined reference frame from which the properties of environmental objects can be explored by relating them to the body. Finally, the robot can begin to relate two environmental objects to one another and to learn that certain actions with the first object can affect the second object, i.e., the first object can be used as a tool. The main contributions of the dissertation can be broadly summarized as follows: it demonstrates a method for autonomous self-detection in robots; it demonstrates a model for extendable robot body schema which can be used to achieve goal-oriented behaviors, including video-guided behaviors; it demonstrates a behavior-grounded method for learning the affordances of tools which can also be used to solve tool-using tasks.

A quebra de frutos encapsulados por macacos-prego com o auxílio de ferramentas é tipicamente objeto de observação e scrounging por coespecíficos, bastante tolerados e em geral mais jovens e menos proficientes. O presente estudo teve como objetivo examinar o processo de escolha, pelos observadores, dos alvos de observação, e se esta escolha pode otimizar as oportunidades de scrounging. A pesquisa foi realizada com um grupo de macacos-prego selvagens (Cebus libidinosus) na Fazenda Boa Vista (Piauí - Brasil), numa área de ecótono cerrado/caatinga. A partilha de alimento fora dos episódios de quebra de cocos e o uso de ferramentas para a quebra de outros itens alimentares encapsulados também foram abordados. Indivíduos de todas as classes de idade e sexo se envolveram na observação da quebra de cocos, havendo uma grande variação individual. Nossas análises mostram que os macacos preferencialmente escolhidos como alvos de observação foram aqueles que apresentaram maior Freqüência, Proficiência e Produtividade na quebra de cocos. Apesar de haver interações agonísticas durante os eventos de observação, os alvos se mostram muito tolerantes à observação e ao scrounging - 25% dos episódios de quebra são observados e mais da metade destes eventos de observação resultam em scrounging. Os observadores tiveram a oportunidade de comer os restos dos cocos e de manipular os itens do sítio de quebra. Estas observações reforçam a noção de que as condições e possibilidades vivenciadas pelo observador-scrounger otimizam as oportunidades para que ocorram processos de aprendizagem socialmente mediada, ao menos por realce de estímulo. Tal exposição próxima ao comportamento do alvo poderia influenciar aspectos mais finos do comportamento dos observadores, ao longo de sua história de aprendizagem. O presente estudo aborda estas interações entre manipuladores de ferramentas e observadores/scroungers pela primeira vez em uma população selvagem e discute as semelhanças e discrepâncias em relação aos resultados anteriormente obtidos com uma população em semil-liberdade. / The tool-aided cracking of encapsulated fruit by capuchin monkeys is a frequent target of observation and scrounging by conspecifics, well-tolerated and usually younger and less proficient. The present study aimed to examine the process of observational targets choice by the observers, and whether this choice can optimize scrounging opportunities. The research was conducted with a group of wild bearded capuchin monkeys (Cebus libidinosus) in Fazenda Boa Vista (Piauí, Brazil), in an ecotone area between cerrado and caatinga savanna-like environments. Food sharing in contexts other than nut cracking, as well as the use of tools to crack open other food items were also analyzed. Individuals from all age and sex classes were involved in nut cracking observation. Our analyses show that the monkeys preferentially chosen as observational targets were the ones exhibiting the greater Frequency, Proficiency, and Productivity in nut cracking. In spite of a few agonistic interactions during observation events, the targets tend to be very tolerant to observation and to scrounging (25% of the nut cracking events are observed, and more than half of these resulted in scrounging). The observers had the opportunity of eating nuts leftovers and of manipulating items from the cracking sites, both in the presence and in the absence of the targets. These findings strengthen the idea that the conditions and possibilities experienced by the observers-scroungers optimize the opportunities for socially biased learning processes, at least by stimulus enhancement. Such close exposure to the targets behavior could also influence finer details of the observers behavior in the long run. The present study is the first one to deal with these interactions between tool manipulators and observers/scroungers in a wild population and discusses the similarities and differences from previous observations from a semi-free population.

This thesis is a literature study on how existing research on embodied tool use may support the use of the use of the computer mouse within three dimensional environments, followed by an analysis of a typical scenario in the use of three dimensional environment. Problems with interaction in this domain are well known to designers of 3d programs but not well understood, which results in programs in which mouse controllers are used to control three dimensional objects being more difficult to learn and less efficient to use than would be possible if the interaction was better understood. The problems are often identified by their symptoms, such as the drag-threshold problem, picking problem, and the object rotation/viewpoint management problem, but this thesis will explore what the cause of those problems is, and identifies them all as a single cognitive problem which is found to be caused by a rift between the functioning of the two dimensional tool in use (the mouse and cursor) and the simulated three dimensional environment with which the cursor is interacting. Analyses are performed on a scenario, and result in a pinpointing of the problem and possible solutions to the interaction part of the problem (with design guidelines emerging), as well as finding the possibility that the cognitive roots of the problem result from an incompatibility between body-schema frames of reference for movement between the two dimensional parts of the action and the three dimensional part of the action.

Les perturbations des activités de vie quotidienne constituent l’un des critères diagnostiques du Trouble de l’Acquisition de la Coordination (TAC). Malgré leur impact sur l’autonomie de l’enfant, les troubles d’utilisation impliqués dans ces activités quotidiennes demeurent peu explorés (e.g., mécanismes sous jacents, profils inter-individuels, recul évolutif). A notre connaissance, la notion d’utilisation d’outil n’a pas fait l’objet d’une modélisation développementale spécifique.Elle se situe aujourd’hui au carrefour des modèles sur le développement perceptivo-moteur et des théories sur l’émergence des capacités cognitives de résolution de problème chez l’enfant. Des modèles récents développés chez l’adulte apraxique suggèrent que l’utilisation d’outil est le fruit d’un processus dialectique entre un raisonnement technique (i.e., analyse abstraite des moyens techniques pertinents pour un but donné) et des habiletés sensori-motrices (i.e., guidées par les représentations sur les transformations mécaniques à opérer). A l’appui de ces modèles, la présente thèse propose d’une part d’analyser le développement typique de l’utilisation d’outil et des processus sous-jacents.Nous explorons d’autre part l’hypothèse d’un développement atypique de l’utilisation d’outil chez des enfants diagnostiqués TAC. Nos résultats apportent des arguments en faveur de la contribution du raisonnement technique au développement typique de l’utilisation d’outil. Sur le plan clinique, l’analyse de l’étendue et des causes sous-jacentes des difficultés d’utilisation d’outil confirme l’importance d’un tel examen pour la validation des critères diagnostiques du TAC. / Deficits in daily living activities are one of the diagnostic criteria of the Developmental Coordination Disorder(DCD). Despite their impact on child’s autonomy, tooluse disorders involved in these daily activities remainunder-assessed (e.g., underlying processes,interpersonal profiles, persistence disease). To ourknowledge, there is no theoretical framework of tool useformulated in a specifically developmental perspective.This concept actually remains at the crossroads between models of the perceptual-motor development and theories about cognitive processes of problem resolution in children. Recent models developed inadults with apraxia suggest that tool use is a dialectical process between a technical reasoning (i.e., abstractanalysis of technical means and ends) and sensory motorskills (i.e., managed by the representations on the mechanical transformations to be operated). According to these models, we first aim to analyze the typical development of tool use and the underlying processes.We then explore the assumption of an atypical development of tool use in children with DCD. Ourfindings provide evidence for the involvement of technical reasoning in typical development of tool use.From a clinical perspective, the analysis of tool use impairment and underlying deficits confirm the relevance of such an assessment for the validation of the diagnostic criteria of the DCD.

L'étendue et les causes des déficits de coordination visuo-manuelle des enfants nés prématurés à bas risque ne sont pas vraiment connus. L'objectif principal de la thèse présentée ici était d'étudier le développement de la coordination visuo-manuelle chez douze prématurés à bas risque, nés après 33-36 semaines de gestation et examinés entre 5 et 23 mois d'âge chronologique. Au cours de la thèse, nous avons mené une série d'études avec des objectifs spécifiques. Dans l'étude 1, nous avons observé le développement de la coordination visuo-manuelle, la prise d'objets, le contrôle de la motricité bimanuelle et la latéralité chez ces enfants prématurés vus entre 6 et 12 mois et comparés à des enfants nés à terme à âge gestationnel équivalent. La perception visuelle des enfants a également été testée à 5 mois et mise en relation avec la coordination visuo-manuelle. Les résultats montrent une perception visuelle globalement normale chez les prématurés qui ne différent des enfants à terme que par une fixation visuelle plus faible. Par contre les enfants prématurés présentent un développement plus lent pour l'intégration visuo-manuelle, la prise d'objet, la coordination bimanuelle et la latéralité. L'hypotonicité et l'âge gestationnel sont les principaux facteurs responsables de ce retard. Aucun retard n'est observé pour le développement postural et locomoteur. Dans l'étude 2 nous avons recherché l'effet de la prématurité sur le développement de la compréhension qu'un objet dont toutes les parties ne sont pas à portée de main peut être saisi par n'importe quelle partie (principe de cohésion). Nous avons présenté aux mêmes enfants prématurés une condition « objet composite » dans laquelle une balle colorée est attachée au bout d'un manche blanc, seul le manche étant à portée de main. Dans une deuxième condition (« déconnection invisible »), la balle paraissait connectée au manche mais ne l'était pas. Les résultats montrent que, en tant que groupe, les bébés prématurés ne présentent pas de retard dans leur prise d'un objet composite. Cependant les enfants prématurés hypotoniques ou nés après 33-34 semaines étaient significativement en retard par rapport aux enfants prématurés hypertoniques. Ces résultats des deux premières études montrent que les nourrissons prématurés, même à bas risque, présentent un retard de développement visuo-manuel dans leur première année de vie et on peut se demander si ce retard se prolonge au-delà de la première année. La réponse à cette question est présentée dans l'étude 3. Nous avons comparé le développement de l'utilisation d'un outil chez les enfants prématurés, âgés alors de 15 à 23 mois, avec celui d'enfants nés à terme d'âge gestationnel équivalent. Comprendre l'affordance d'un outil est une étape importante dans la période sensorimotrice. L'utilisation d'un outil pour prendre un objet hors de portée, par exemple, se développe normalement au cours de la deuxième année. Nous avons présenté aux enfants prématurés un jouet coloré hors de portée et un râteau blanc à portée de main dans cinq conditions de relations spatiales entre le jouet et le râteau. En tant que groupe, les enfants prématurés ont présenté un développement normal de l'utilisation de l'outil. Cependant, ce développement normal n'apparaît pas chez tous les enfants prématurés. Les enfants prématurés hypotoniques ou nés après 33-34 semaines de gestation ont montré un retard significatif dans leur acquisition de l'utilisation de l'outil. Ces résultats montrent que les enfants nés prématurés à faible risque peuvent être retardés dans leur développement de nouvelles compétences manuelles au-delà de la première année. En suivant ces enfants au-delà de deux ans, il devrait être possible de déterminer quels sont les tests les plus prédictifs de déficits neuropsychologiques persistants parmi ceux que nous avons utilisés. Ces résultats devraient néanmoins être confirmés dans une population plus large. / The extent of and reasons for visual-manual coordination deficits in moderate and late low risk preterm born infants are not well known. The aim of the thesis presented here was to investigate the development of visual-manual coordination in twelve low risk preterm infants born after 33-36 weeks of gestation and examined from 5 to 23 months of age. We conducted a series of studies with specific goals. The goal of study 1 was to investigate the development of visual-manual and bimanual coordination as well as handedness in the preterm infants between the postnatal ages of 6 and 12 months, after having evaluated their visual functions at 5 months. Gross motor development was also evaluated every month. Preterm infants were compared to a control group of ten full-term infants according to corrected age. Results show that preterm infants had a visual perception close to normal, with only visual fixation being lower than in full-terms. Preterm infants had delayed development of visual-manual integration, grasping, bimanual coordination and handedness. Tonicity and gestational age at birth were the main variables associated to the delays. There was no delay for gross motor development. The goal of study 2 was to investigate at what age preterm infants integrate the notion of object cohesiveness into their action plan. The same preterm and full-term infants as in first study were presented with a composite object task at 6, 8, and 10 months in two conditions, "composite object" condition, in which a bright ball was attached to the end of a neutral handle, and "invisible disconnection" condition in which the ball seemed to be connected to the handle but was not. Only the handle was within reach. Corrected age comparisons between preterm infants and full-terms showed no significant difference between the two groups. However, the hypotonic preterm infants had a significantly lower performance than the hypertonic infants. Based on the results from studies 1 and 2 it seems that preterm infants, in particular the hypotonic infants, have a delayed development in their first year of life, and one can wonder if this delay remains at later stages. To answer this question, in study 3 we compared the development of tool use in the preterm infants, aged 15 to 23 months, with that of full-terms according to corrected age. Understanding the affordance of a tool is an important milestone in early sensorimotor period. Using a tool to bring within reach an out-of-reach object, for instance, has been shown to develop during the second year in full-term infants. Here we presented preterm infants with an attractive toy out of reach and with a rake-like tool within reach in five conditions of spatial relationships between the toy and the tool. As a group, preterm infants showed a normal development of tool use. However not all preterm infants developped tool use without delay. Hypotonic and moderate preterm infants were significantly delayed in their acquisition of tool use. These results, show that some low risk preterm infants can still be impaired for the development of new manual skills beyond the first year. By following these low risk preterm children, it should be possible to evaluate which of the delays observed in our studies are the best predictors of later neuropsychological deficits in childhood. These results should also be confirmed with a larger sample of preterm infants.

New Caledonian crows (NC crows; Corvus moneduloides) are the most prolific avian tool-users and crafters, using up to three unique tool types derived from numerous plant materials. Since the discovery that wild populations of NC crows use and manufacture different tools in different locations with no measured environmental correlates to these distributions, the process by which NC crows acquire their tool-oriented behavior has been investigated. Two major findings were discovered in 2005: NC crows have a genetic predisposition to manipulate stick like objects, and they increase their rate of manipulation when exposed to social influences. Since then, much of the research into the sociality of wild NC crows has focused on direct social influences, especially the parent-juvenile relationship, yet no social network of wild NC crows has been described. In my thesis, I characterized a new proximity-logging device, Encounternet, and outline a four-step plan to assess error in animal borne devices; uncovered drivers, such as relatedness, space-use, and environmental factors, of wild NC crow sociality, and experimentally manipulated the social network, revealing immediate changes to the number of day-time and roosting partners, the breakdown of first-order relatedness driving sociality, and an increase in the amount of time NC crows associate; and revealed an indirect pathway via tools left behind by conspecifics allowing for the transmission of tool-properties between unrelated NC crows. Altogether, I furthered our understanding of wild NC crow sociality through the use of an animal-borne device, experimental manipulation in the wild measuring the response of the NC crow social network, and demonstrated the utility of animal-borne devices in mapping the network of a population of wild birds.

Partindo das evidencias encontradas na literatura sobre o desenvolvimento do uso de ferramentas de percussão para quebra de frutos encapsulados, propomos um projeto focalizando não apenas quando as alterações motoras aparecem, mas também na relação dessa trajetoria de mudança com duas possíveis vias de influencia do nicho ecológico sobre a trajetoria de mudança comportamental que da origem ao uso de ferramentas de percussão para quebra de frutos encapsulados em macacos-prego (Sapajus libidinosus). Em primatas, os aspectos físicos e sociais (isto é, artefatos: martelos, bigornas e restos de cocos, e o comportamento dos coespecíficos) canalizam as oportunidades individuais de aprendizagem, permitindo maiores níveis de manipulação de objetos apropriados em contextos apropriados para o surgimento do uso de ferramentas e facilitando a exploração de várias fontes de forrageamento. Por tanto, procuramos determinar o papel de duas variáveis específicas de influência de nicho construídas: reuso dos artefatos e scrounging, no processo de desenvolvimento do repertório manipulativo em um grupo silvestre de macacos-prego (Sapajus libidinosus). Especificamente: 1) Como é a trajetória de mudança na atração e reutilização de artefatos relacionados ao craqueamento de noz (martelos, bigornas e cascas de nozes) e como isso se relaciona com o surgimento de ferramentas de percussão; E, 2) Como relaciona-se o scrounging, no contexto da quebra de frutos, com a emergencia de habilidades manipulativas adequadas para a quebra de frutos encapsulados. Propomos uma pesquisa de campo longitudinal de uma população de macacosprego livres no Brasil, através da observação naturalista. O resultados indicam que não só a idade está relacionada tanto com as mudanças na atração e reuso dos artefatos e com o scrounging, mas também que ambas variaveis estão relacionadas e podem estar em função das mudanças no repertorio manipulative exibido pelos individuos, de tal forma que não só os individuos mais velhos exibem maiores indices de reuso e menores indices de scrounging, como também os indices de variabilidade comportamental e proficiencia na quebra estão intimamente relacionados com os indices de reuso dos artefatos e scrounging / Based on the evidence found in the literature on the development of the use of percussion tools for encapsulated fruit breakage, we propose a project focusing not only \"when\" motor alterations appear, but also on the relation of this trajectory of change with two possible ways of influencing the ecological niche on the trajectory of behavioral change that gives origin to the use of percussion tools to crack nuts in capuchin-monkeys (Sapajus libidinosus). In primates, physical and social aspects (ie, artifacts: hammers, anvils and remains of coconuts, and co-specific behavior) channel individual learning opportunities, allowing greater levels of manipulation of appropriate objects in appropriate contexts for the emergence of use of tools and facilitating the exploration of various foraging sources. Therefore, we sought to determine the role of two niche-specific variables constructed: artifact reuse and scrounging, in the process of developing the manipulative repertoire in a wild group of monkey-prey (Sapajus libidinosus). Specifically: 1) How is the trajectory of change in the attraction and reuse of artifacts related to nut cracking (hammers, anvils and walnut shells) and how does this relate to the emergence of percussion behavior. And, 2) How scrounging is related, in the context of the fruit break, with the emergence of adequate manipulative skills for the breaking of encapsulated fruits. We propose a longitudinal field survey of a population of free monkey-natives in Brazil, through naturalistic observation. The results indicate that not only age is related both to changes in attraction and reuse of artifacts and to scrounging, but also that both variables are related and may be due to changes in the manipulative repertoire exhibited by individuals, such that not only do older individuals exhibit higher reuse rates and lower scrounging rates, but also the indices of behavioral variability and proficiency in the break are closely related to the artifact reuse indices and scrounging

L’utilisation d’outils est un trait définitoire du genre Homo. Il est donc fondamental de mieux connaître les bases cognitives et cérébrales nous permettant d’utiliser des outils. Les modèles cognitivistes actuels expliquent l’utilisation d’outils à travers l’hypothèse de l’activation d’une mémoire gestuelle (i.e., engrammes gestuels ou visuo-kinétiques, ou connaissances sensorimotrices sur la manipulation ; voir Rothi, Ochipa, & Heilman, 1991 ; Buxbaum, 2001) ? Cette hypothèse ne permet toutefois pas de comprendre l’utilisation d’outils nouveaux. Une hypothèse alternative a été établie, suggérant que toute situation d’utilisation d’outils (familiers et nouveaux) requière un raisonnement technique (e.g., Osiurak & Badets, 2016). Ce type de raisonnement, qui impliquerait le lobe pariétal inférieur gauche, nous permettrait de formuler l’action mécanique et d’évaluer les propriétés physiques des outils et des objets. Dans le cadre de cette hypothèse, l’une des finalités de cette thèse était de mieux comprendre les troubles d’utilisation d’outils chez des patients cérébro-lésés. Le présent travail s’est également porté sur l’investigation de la cognition numérique. Par ce terme, nous ne faisons pas uniquement référence au calcul mental ou à l’arithmétique. Nous englobons également ce que Dehaene et Cohen (1995) ont nommé code analogique dans leur Modèle du Triple Code. Ce code stockerait les représentations des quantités numériques au sein des lobes pariétaux. Autrement dit, il contiendrait le sens du nombre (Dehaene, 1997) permettant d’associer une étiquette symbolique (e.g., chiffre arabe) à la quantité correspondante. Au quotidien, ce serait grâce à ces représentations que nous pourrions comparer ou estimer la numérosité des ensembles d’objets. L’objectif principal de cette thèse était de rapprocher, tant au niveau cognitif que cérébral, ces deux domaines d’intérêt que sont l’utilisation d’outils et la cognition numérique. En effet, nous avons remarqué que ces deux capacités nécessitaient toutes deux un processus commun d’estimation de la magnitude (i.e., magnitude des propriétés physiques et magnitude des quantités numériques). En outre, au niveau cérébral, elles nécessitent l’activation de régions communes dans le lobe pariétal. Pour penser ce lien, nous nous sommes appuyés sur la théorie de la magnitude (ATOM) formulée par Walsh (2003). Celui-ci postule que toutes les magnitudes, c’est-à-dire toutes les dimensions qui peuvent être décrites par des relations « plus que/moins que », soient traitées au sein d’un système commun et unique dans le lobe pariétal droit (Bueti & Walsh, 2009). Nous avons supposé que la magnitude des propriétés physiques pourrait être traitée dans ce système au même titre que les magnitudes discrètes (e.g., numérosité) et continues (e.g., temps, espace). Nos résultats ont mis en évidence un trouble de l’utilisation d’outils nouveaux chez les patients LBD, sans difficultés apparentes pour estimer les propriétés physiques. Les patients RBD étaient déficitaires dans toutes les conditions évaluant la cognition numérique, contredisant les prédictions issues du TCM. Ces patients étaient également en difficulté pour estimer la longueur mais pas le poids. Comme des associations entre estimation de la longueur et du poids, et entre estimation de la longueur et cognition numérique ont été observées dans les différents groupes, nous suggérons que le système de magnitude soit divisé en sous-systèmes. Fait étonnant, nous avons trouvé une association entre utilisation d’outils et calcul approximatif chez les patients LBD supposant une tentative de compensation de l’utilisation par le calcul. Finalement, il semble que l’utilisation d’outils et la cognition numérique reposent sur des mécanismes neurocognitifs distincts, puisque les différents types de magnitudes ne paraissent pas être traités au sein d’un système commun et unique. / Tool use is a defining feature of the genus Homo. It is therefore fundamental to better understand the cognitive and cerebral bases that allow us to use tools. The current cognitivist models explain tool use through the hypothesis of an activation of gestural memories (i.e., gestural or visuo-kinetic engrams, or sensorimotor knowledge of manipulation; see Rothi, Ochipa, & Heilman, 1991; Buxbaum, 2001). This theory is unable to explain the use of novel tools. An alternative hypothesis suggests that any situation of tool use (familiar and new) requires technical reasoning (e.g., Osiurak & Badets, 2016). This reasoning, involving the left inferior parietal lobe, would enable to formulate the mechanical action and to evaluate the physical properties of tools and objects. One of the aims of this thesis was to better understand the tool use disorders in brain-damaged patients, within the framework of the technical reasoning hypothesis. This work has also focused on the investigation of numerical cognition. By this term we refer to mental arithmetic and math, but also to analogical code (see the Triple Code Model, Dehaene & Cohen, 1995). It corresponds to the representation of numerical quantities, stored in the parietal lobes. In other words, this code would contain the sense of number (Dehaene, 1997) to associate a symbolic label (e.g., Arabic digits) with the corresponding quantity. In everyday life, this representation would be critical to compare or estimate the numerosity of object sets.The main objective of this thesis was to explore, at cognitive and cerebral levels, whether links exist between both fields of interest that are tool use and numerical cognition. Indeed, we noticed that both capacities need a common process of magnitude estimation (i.e., physical properties and numerical quantity). In addition, at the cerebral level, they require the activation of common regions in the parietal lobe. We relied on the Theory Of Magnitude (ATOM) formulated by Walsh (2003). It postulates that all magnitudes, namely the dimensions described by “more than/less than” relationships (e.g., Is this stick long enough to reach a given place?), are processed within a common and unique system, in the right parietal lobe (Bueti & Walsh, 2009). We assumed that the magnitude of physical properties could be processed in this system as well as the discrete (e.g., numbers) and continuous (e.g., time, space) magnitudes. Our results highlighted a disorder of novel tool-use in LBD patients, who nevertheless had no difficulty in estimating physical properties. The RBD patients were impaired in all conditions assessing the numerical cognition, refuting the predictions derived from TCM. They were also impaired in the estimation of the length but not of the weight. As associations between estimation of length and of weight, and between estimation of length and numerical cognition have been observed in the different groups, we suggest that the magnitude system be divided into subsystems. Surprisingly, we found an association between tool use and approximate calculation in LBD patients assuming an attempt to compensate tool use by calculation. Finally, it seems that tool use and numerical cognition rely on distinct neurocognitive mechanisms since the different types of magnitudes might not be processed within a common and unique system of magnitude