[Truncated abstract] Central to the study of general intelligence is Spearman?s g, one definition of which is general fluid reasoning ability. At a cognitive level of explanation, it has recently been hypothesized that individual differences in g is based in the executive functions of the frontal lobes. This theory can be contrasted with the theory that the fount of g is speed of information processing. The aim of this thesis was to test the idea that the two contrasting theories of g could be reconciled by invoking an alternative theory which suggests that there may be two g?s one related to individual differences in intelligence and attributable to differences in speed of information processing and one related to the development of intelligence and based in executive functions of the frontal lobes. This was done with a series of neuropsychological studies that tested groups of adults and children with and without putative central nervous system damage on tests of fluid intelligence, executive function, goal-neglect, and speed of information processing. In study 1, three adults with focal frontal lobe lesions and ten adults with frontotemporal dementia (FTD) were administered three common intelligence tests. In comparison to premorbid and crystallized measures on intelligence, the majority of frontal patients exhibited impaired fluid g. In study 2, 10 patients with FTD, 8 patients with Alzheimer?s disease (AD), 10 adult controls, and 15 adults with low fluid g, were tested on a fast and slow version of a goal-neglect task (thought to measure executive functioning) and a measure of speed of processing. A classical double dissociation was found. Frontal patients with impaired fluid g displayed goal-neglect but intact speed of information processing whereas the adults with low fluid g exhibited slowed speed of information 2 processing but not goal-neglect. It was concluded that the link between fluid g and goal-neglect in adults (demonstrated by previous research) was based on a speed of information processing confound in the goal-neglect task. In study 3, a series of hierarchical regressions were conducted to analyse the performances of 116 children aged 6- to 11-years on all tests. The statistical attempt to dissociate executive function and speed of information processing only provided tentative support for the hypothesis that executive functions are the basis of developmental changes in g

Notre but de comprendre les mécanismes constitutifs de la déficience intellectuelle et de la dyspraxie verbale, deux pathologies qui affectent les performances aux tests psychométriques. Nous avons sélectionné plusieurs processus pouvant contribuer à l’explication de ces troubles : la vitesse de traitement de l’information et la mémoire travail pour la déficience intellectuelle ; et l’apprentissage procédural pour la dyspraxie verbale. Dans une première étude, nous avons montré que dans la population « tout venant » d’enfants non déficients, la vitesse de traitement augmente avec l’âge. Notre seconde expérience a montré que des patients déficients intellectuels avaient la même vitesse de traitement que des enfants plus jeunes appariés sur l’âge mental. Par ailleurs, nous avons montré que ce ralentissement est similaire dans le syndrome de l’X-Fragile et dans la Trisomie 21. Inversement, la vitesse de traitement des enfants dyspraxiques verbaux est préservée. Ces résultats viennent discuter les deux modèles théoriques de l’intelligence que nous avons choisis comme référence, le modèle en cascade de Fry & Hale (1996) et le modèle d’Anderson (1992). Dans le second volet de nos recherches, nous avons mis en évidence un apprentissage procédural implicite comparable dans deux groupes d’enfants d’âges chronologiques différents malgré des temps de réaction plus importants chez les plus jeunes. La seconde étude montre une préservation de l’apprentissage procédural implicite dans le syndrome de l’X-Fragile et une altération spécifique dans la Trisomie 21. Cela démontre que ce processus est indépendant du QI et varie selon les syndromes. Notre dernière étude montre un trouble important de l’apprentissage procédural implicite dans la dyspraxie verbale, confortant notre hypothèse de départ. L’ensemble de ces résultats est discuté au regard des travaux antérieurs et des modèles théoriques afin de proposer des ouvertures tant théoriques que cliniques. / Our aim is to understand previous process in mental retardation and childhood apraxia of speech (CAS). There are both neurodevelopmental disorders which affect psychometric assessment. We have selected several processes which could explain these disorders such has speed of information processing, working memory for mental retardation and implicit procedural learning for childhood apraxia of speech. In our first study, we have demonstrated within two groups of typical children without disorders that speed of processing increased with chronological age. In the second study, our results have demonstrated the same speed of processing between a group with mental retardation and with mental age matched control group. Moreover, Down and Fragile X syndrome have the same speed of information processing. On the contrary, in our third study, children with CAS and chronological age matched control group have the same speed as process visual information. All results are important to discuss both theories of intelligence we have chosen: the cascade model (Fry and Hall, 1990) and the minimal cognitive architecture (Anderson, 1992). In the second experimental part, our first results have shown a different reaction time between the two groups of young children but both have learned the sequence in a serial reaction time task (implicit procedural learning test). The second study have demonstrated preserved implicit learning in Fragile X but altered learning in Down syndrome. The difference between to these two genetics syndromes is a proof of implicit sequence learning is independent of IQ. The latest results have shown severe implicit procedural learning impairment in CAS. We conclude that this is the cause of CAS.Our discussion is about all results and links between intelligence and disorders with regard to previous studies. Those results aim for both theorical and clinical openings.