This theses describes aspects of memory traces processing, its applications in simulation and development of modular memory subsystem simulator based on memory traces. The simulator also supports use for educational purposes thanks to built-in visualization which can be used to track address accesses in memory subsystem hierarchy.

Lexical access, the matching of auditory information onto lexical representations in the brain, is a crucial component of online language processing. To understand the nature of lexical access, it is important to identify the kind of acoustic information that is stored in the long-term memory and to study how the brain uses such information. This dissertation investigates the contribution of prosodic information to lexical access and examines language-specific processing mechanisms by studying three typologically distinct languages: English, Turkish, and Swedish. The main research objective is to demonstrate the activation of long-term memory traces for words on the sole basis of prosodic information and to test the accuracy of typological phonological descriptions suggested in the literature by studying electrophysiological measurements of brain activation. A secondary research objective is to evaluate three distinct electrophysiological recording systems. The dissertation is based on three papers, each examining neural responses to prosodic changes in one of the three languages with a different recording system. The first two papers deal directly with the interplay between prosody and the lexicon, and investigate whether prosodic changes activate memory traces associated with segmentally identical but prosodically different words; the third paper introduces morphology to this process and investigates whether prosodic changes activate memory traces associated with potential lexical derivations. Neural responses demonstrate that prosodic information indeed activates memory traces associated with words and their potential derivations without any given context. Strongly connected neural networks are argued to guarantee neural activation and implementation of long-term memory traces. Regardless of differences in prosodic typology, all languages exploit prosodic information for lexical processing, although to different extents. The amount of neural activation elicited by a particular piece of prosodic information is positively correlated with the strength of its lexical representation in the brain, which is called lexical specification. This dissertation could serve as a first step towards building an electrophysiological-perceptual taxonomy of prosodic processing based on lexical specification.

De nombreuses études ses sont intéressées à l’influence du sommeil sur la mémoire. Une période de sommeil comparée à une période d’éveil permet de réduire l’oubli, d’améliorer les performances mais également de réorganiser les traces mnésiques, favorisant ainsi la créativité et l’extraction de règles. Ces études se basent sur les modèles classiques de la mémoire et expliquent ces observations par une interaction entre les processus implicites et explicites au cours du sommeil. Cependant, les modèles classiques de la mémoire semblent limités pour décrire la totalité des processus de réorganisation des traces mnésiques observés après une période de sommeil. Dans ce travail de thèse, nous avons tenté de comprendre l’influence du sommeil sur la réorganisation des traces à travers le modèle Act-In, un modèle de mémoire à traces multiples se plaçant dans la lignée de la cognition incarnée.Nous avons développé notre recherche autour de quatre axes expérimentaux. Dans un premier temps nous souhaitions vérifier si une période de sommeil permet de transformer passivement une trace implicite en trace explicite. Nous souhaitions par la suite étudier l’effet du sommeil sur les processus d’intégration multi-composants et d’activation inter-traces. Selon nous, le sommeil permettrait de renforcer l’intégration des composants au sein d’une trace et il permettrait également de favoriser la mise en relation de plusieurs traces, conduisant à l’émergence d’éléments communs. Enfin, dans le dernier axe de ce travail de recherche, nous nous sommes intéressés aux conséquences des modifications nocturnes liées à l’âge sur la consolidation mnésique au cours du sommeil. / Numerous studies have been interested in the effect of sleep on memory. For instance, a period of sleep as compared to a similar period of wakefulness protects memories from interferences, improves performances, but also reorganizes memory traces favoring creativity and rules extraction. These studies are based on classical memory models and explain these observations by an interaction between implicit and explicit processes. However, these models seem to be limited to describe all the processes involved in memory reorganization. In this thesis, we tried to understand the influence of sleep on memory reorganization within Act-In memory model, a multiple traces memory model in the line of grounded cognition. We developed our research within four experimental axes. First of all, we wanted to verify if sleep could transform an implicit trace into an explicit one. We also wanted to study the effect of sleep on multi-components integration and inter-traces activation. We assumed that sleep could transform memory traces enhancing the integration of the trace components and could favor the link between different traces leading to the emergence of common elements. Finally, we were interested in the consequences of nocturnal modifications during aging on memory consolidation.

<p>Diss. (sammanfattning) Umeå : Umeå universitet, 1993, härtill 4 uppsatser.</p> / digitalisering@umu

If remembering is discriminative behaviour along the dimension of time and if, as Sargisson and White (2001) argued, generalisation around a peak can occur in this behaviour, then the peak shift which has been shown in discrimination along so many other stimulus dimensions, might also occur in remembering. To examine this hypothesis, 6 hens were trained in a delayed matching-to-sample procedure at delays of 2 and 4 s. The probability of reinforcement for correct responses was initially 0.9 at both delays until performance stabilised. A generalisation probe was then carried out by inserting unreinforced trials at delays of 0, 1, 1.5, 2.5, 3, 3.5, 4.5, 5 and 6 s in a session amongst normal training delay trials. The generalisation functions had a slight peak around 2 s. After further training, a second generalisation probe showed a slightly declining function. The probability of reinforcement at the 2 s delay was then dropped to 0.1, so that in the terms of the classical generalisation/peak shift paradigm, 2-s delay trials became S¯ and 4-s delay trials became S+. A third generalisation probe then was conducted. This resulted in a flat function from 0 s to 3 s, and a large, clear peak in discriminative performance at 4.5 s over all hens. After more of the same differential reinforcement training, a fourth generalisation probe showed a broad curve peaking at 3 s, with minima at 1 s and 6 s and a global maximum at 0 s. Another training condition was then run, with the probability of reinforcement at the 2-s delay dropped to 0, to see if increasing the aversiveness of S¯ would again result in a peak shift. A fifth generalisation probe was then conducted. This showed a sharp decline in discriminability at shorter delays, a dip around 2 s, and a very small area shift beyond 4 s, but no clear peak shift. This was interpreted as being due to overlearning, with the consequences of remembering at S¯ no longer significantly affecting performance at S+. A final training condition was then run, with S¯ moved from 2 s to 3 s with zero probability of reinforcement, and for only a short period, to prevent overlearning. It was predicted that this would cause peak shift to re-occur. A sixth generalisation probe was then conducted. This found a further decline in discriminability at shorter delays, a shift in the dip from 2 s to 3 s, and a large, clear peak at 4.5 s. This demonstration of peak shift in a remembering process would not have been predicted by any traditional theory of memory, but strongly supports the conception of remembering as discriminative behaviour along the stimulus dimension of time.

Notre but était de mieux comprendre les mécanismes associés à la récupération en mémoire à court terme acoustique à l’aide de mesures électrophysiologiques. La tâche utilisée était une version modifiée de la tâche de Sternberg dans laquelle les participants devaient retenir deux, quatre ou six sons purs hors de la gamme musical bien tempérée, ne pouvant pas être facilement recodés verbalement. Après un intervalle de rétention silencieux, ils entendaient un son et devaient indiquer si celui-ci était présent ou absent dans l’ensemble mémoire. En utilisant plusieurs conditions de charge et en contrôlant pour les durées de stimulation, nous avons comparé les effets de position sérielle, ainsi que les effets de charge, sur les données comportementales et sur les données électrophysiologiques mesurées lors du balayage mnésique. Notamment, nous avons trouvé un effet de récence important peu importe le nombre d’items mémorisés, se traduisant par des temps de réponses courts et des taux de bonne réponse presque parfaits, ainsi que par l’augmentation de l’amplitude de la P3, une composante de potentiels reliés aux évènements (PRE). Les données liées aux autres positions sérielles étaient similaires, indépendamment de la charge mnésique, et montraient des performances moins bonnes et une P3 moins ample. Une méthode de décomposition du signal, l’analyse en composantes indépendantes (ACI) nous a permis d’observer et de décrire les différences électrophysiologiques, dans le temps, entre la récupération d’un son parfaitement retenu (le dernier) et celle d’un son peu retenu. Nos résultats suggèrent l’existence de deux sous-types de mémoire sensorielle. / We aimed to understand better the processes involved in acoustic short-term memory retrieval using electrophysiology. We used a modified Sternberg task in which participants had to encode two, four, or six pure non-musical tones (out of the well-tempered musical scale) that cannot be readily recoded verbally. After a silent retention interval, we presented them with a probe tone and they had to report its presence or absence in the memory set. By using multiple load conditions, and by controlling for stimulation duration, we compared the serial position effects, as well as the load effects, on behavioral and electrophysiological data during memory scanning. In particular, we found a recency effect, similar across loads, where shorter response times, near perfect accuracy, and an increased P3 amplitude in the event-related potential (ERP) data were observed. Serial position data for all other positions were similar regardless of the memory load and showed lower performances (response times and accuracy), as well as smaller P3 components. We also performed a signal decomposition analysis, the independent component analysis (ICA), which allowed us to observe and describe better the time courses of the electrophysiological data for the retrieval of a perfectly memorised tone (the last one), and a lesser memorised one (any other serial position). Our results suggest the existence of two subtypes of sensory memory.