Final Activity Report Summary - MTL (Multiple memory systems in the human temporal lobe: clues from remote memory and new learning in amnesia)
The present project investigated the neural bases of: (1) acquisition of new declarative information by the human brain, and (2) representation of old memories.
(1) Current theories of declarative memory propose two complementary learning systems in the mammalian brain: a system that supports rapid learning of arbitrary associations (mediated by a structure called the Hippocampus), and a system that only learns slowly by extracting generalities out of multiple similar experiences (mediated by regions in the Temporal Cortex). With the help of the Marie Curie IRG, we demonstrated rapid acquisition of novel arbitrary associations by adult amnesic patients with lesions to the Medial Temporal Lobe (MTL) including the Hippocampus. This was achieved through 'Fast Mapping' (FM): A process that supports vocabulary acquisition in toddlers, before the MTL is fully developed. FM is based on disjunctive syllogism, in which the child discovers the meaning of a novel word by inferring the speaker's intention and rejecting familiar items in the environment. In our task, amnesic patients and matched controls were simultaneously presented with familiar and unfamiliar objects (animals, fruit, and flowers) and had to answer a perceptual question about one of them.
% Following FM, patients were as good as controls on an associative recognition test both immediately and after a week's delay. They were, however, almost at chance on a matched associative encoding task without the FM component, which controls were very good at. Patients' data suggest that the tip of the temporal cortex may be crucial for associative learning through FM, and neuroimaging pointed to the basal ganglia which are sub-cortical structures, as possible contributors, but more experiments are needed to elucidate the neural mechanisms that support FM. Our data suggest a considerable revision of current memory theories is needed. They also offer hope for novel interventions with memory impaired individuals.
(2) In addition to its role in acquisition of new memories, the hippocampus is known to crucially mediate retrieval of new memories. Its role in retrieval of remote memories has puzzled researchers because remote memories appear intact following hippocampal distraction. Using behavioural methods and neuroimaging, the present study investigated the possibility the hippocampus in fact mediates a very specific memory function, namely the ability to re-experience events from the past (recollection), and that when tested appropriately, both remote and recent memories are affected in the same way following specific hippocampal lesions. We found that patients with lesions to the hippocampus were indeed impaired on their ability to retrieve associative information associated with remote events, even when memory for the events themselves was intact. We are currently analysing neuroimaging data to find out if this conclusion is supported by that method as well.
Together, the data from the two projects suggest that current ideas about the time-limited role of the hippocampus in memory acquisition and its exclusivity as a structure that support novel acquisition of arbitrary associations needs to be reconsidered.
(1) Current theories of declarative memory propose two complementary learning systems in the mammalian brain: a system that supports rapid learning of arbitrary associations (mediated by a structure called the Hippocampus), and a system that only learns slowly by extracting generalities out of multiple similar experiences (mediated by regions in the Temporal Cortex). With the help of the Marie Curie IRG, we demonstrated rapid acquisition of novel arbitrary associations by adult amnesic patients with lesions to the Medial Temporal Lobe (MTL) including the Hippocampus. This was achieved through 'Fast Mapping' (FM): A process that supports vocabulary acquisition in toddlers, before the MTL is fully developed. FM is based on disjunctive syllogism, in which the child discovers the meaning of a novel word by inferring the speaker's intention and rejecting familiar items in the environment. In our task, amnesic patients and matched controls were simultaneously presented with familiar and unfamiliar objects (animals, fruit, and flowers) and had to answer a perceptual question about one of them.
% Following FM, patients were as good as controls on an associative recognition test both immediately and after a week's delay. They were, however, almost at chance on a matched associative encoding task without the FM component, which controls were very good at. Patients' data suggest that the tip of the temporal cortex may be crucial for associative learning through FM, and neuroimaging pointed to the basal ganglia which are sub-cortical structures, as possible contributors, but more experiments are needed to elucidate the neural mechanisms that support FM. Our data suggest a considerable revision of current memory theories is needed. They also offer hope for novel interventions with memory impaired individuals.
(2) In addition to its role in acquisition of new memories, the hippocampus is known to crucially mediate retrieval of new memories. Its role in retrieval of remote memories has puzzled researchers because remote memories appear intact following hippocampal distraction. Using behavioural methods and neuroimaging, the present study investigated the possibility the hippocampus in fact mediates a very specific memory function, namely the ability to re-experience events from the past (recollection), and that when tested appropriately, both remote and recent memories are affected in the same way following specific hippocampal lesions. We found that patients with lesions to the hippocampus were indeed impaired on their ability to retrieve associative information associated with remote events, even when memory for the events themselves was intact. We are currently analysing neuroimaging data to find out if this conclusion is supported by that method as well.
Together, the data from the two projects suggest that current ideas about the time-limited role of the hippocampus in memory acquisition and its exclusivity as a structure that support novel acquisition of arbitrary associations needs to be reconsidered.