Dreaming: a process of self-organizationKahn and Hobson (1993) proposed that dreams are a product of self-organization of brain during sleep. As a complex system far from equilibrium state, the dreaming brain may form a new pattern by the interaction between components within this system. At REM sleep stage, signals from neuronal clusters self-organize and form image fragments, then the image fragments interact and produce images, and finally these materials are associated into a relatively continuous narrative (i.e., dreams). This process above (...) happens under a weak control of brain, and the conditions of this state mainly include: (a) the inhibition of external stimuli or feedback; (b) the changes of neuromodulatory systems (e.g., the governance of aminergic neurochemicals and the weakness of cholinergic neurochemicals); (c) the bombardment of PGO wave to the visual cortex; and (d) the changes of neural activity in brain (e.g., the activation of limbic system and the reduction of prefrontal regions). In this situation, the brain could focus on the internal world and integrate various psychological and physiological changes into dream content (Kahn et al., 2002, 2000). However, self-organization mechanism could not determine which memories will be activated and incorporated into dream content (Kahn, 2013), although it provides an approach how an ordered structure, behavior or pattern spontaneously emerges from the interaction between components or subsystems without an e... (shrink)

In the philosophy of memory, singularism is the view that episodic memories are singular mental states about unique personally experienced past events. In this paper, I present an empirical challenge to singularism. I examine three distinct lines of evidence from the psychology of memory, concerning general event memories, the transformation of memory traces and the minimized role temporal information plays in major psychological theories of episodic memory. I argue that singularist views will have a hard time accommodating this evidence, facing (...) a problem of transitional gradation. I then look at some potential consequences for the larger debate, highlighting the way in which singularism has featured in three important recent arguments in the philosophy of memory. (shrink)

The Hard Question of memory is the following: how are memory representations stored and organized so as to be made available for retrieval in the appropriate circumstances and format? In this essay, I argue that philosophical theories of memory should engage with the Hard Question directly and seriously. I propose that declarative memory is a faculty performing a kind of cognitive triage: management of information for a variety of uses under significant computational constraints. In such triage, memory representations are preferentially (...) selected and stabilized, but also systematically modified and integrated into generalized, model-like representational structures. Further, I propose a hybrid theory of remembering, which takes into account both the nature of the cognitive processes underlying remembering and the norms that govern representational success in relevant cognitive/epistemic contexts. (shrink)

High demands on cognitive functions in daily life call for an efficient memory system that reduces interference between memories and enables generalizations across similar events. By means of pattern separation, similar memories are stored as distinct, non-overlapping representations, whereas during pattern completion, previously stored memories are reactivated by partial environmental cues. These two functions are critically reliant on the hippocampus. Evidence from computational models, studies in rodents, as well as human data support the idea that pattern separation and completion are (...) mediated by the hippocampal dentate gyrus and CA3 regions. However, studies in humans lack information regarding mechanistic aspects of causality regarding the anatomical structures of the hippocampus and pattern separation and completion processes. The aim of this thesis was to elucidate the role of the human hippocampus and its subfield-specific contributions to pattern separation. We examined natural lesion models, by means of selective CA1 lesions during a transient global amnesia, and in a rare form of limbic encephalitis, where neurodegeneration preferentially shows in the dentate gyrus and CA3. The results showed that pattern separation measured by a mnemonic similarity task was best predicted by the volume of the dentate gyrus, whereas recognition memory was stronger associated with the volume of CA1. We also found a strong deficit in pattern separation associated with selective CA1 lesions. We then examined pattern separation performance after post-encoding sleep in healthy humans to clarify the neurobiological processes of memory consolidation. We demonstrated the relevance of hippocampal information processing during sleep in the stabilization of separated information. This might also suggest a link between pattern separation and completion processes during sleep-dependent memory consolidation. (shrink)