Schacter Ch 3,2

Dec 3
Development
Development sets up the brain to receive and interprete relevant information.  It does this through trophic and activity dependent processes.

  1. During the earliest points in development trophic factors dictate the basic architecture of the system.
  2. Once the basic architecture is set up then activity dependent processes are involved in organizing the specific connectivity patterns and intrinsic connections.  These patterns will be influenced by the input that the system receives.
Implications of this developmental scenario for learning and memory theories are :
  1. Perceptual aspects will be stored first.  For instance learning that the points on a line are generally representivie of a single object.
  2. Storage and retrieval systems are in flux throughout the first few years of life.
  3. Memories from childhood are stored in a very different fashion, if at all, and therefore are not retrievable by adult retrieval systems.
  4. Source misattribution can be explained by this theory as can omission, false beliefs etc. In ch 7 poor source memory per se is e attributed to a slowly developing frontal lobe system.  This would fit in with what the developmental scenario proposed above because the frontal lobe system is the last to develop.
PDP
 Given that the system is set up properly and the basic perceptual aspects of the world are encoded how is new information stored?  One possiblility is that it is stored as a database with all the characteristics of a given experience stored and linked together and each new experience is separately cataloged.  Under the database scenario information should either be retrievable or not.  While omission may be support this view most observations of memory distortion (blending, etc) point to another mode of storage. This other mode is referred to as a parallel distributed network and is the result of a activity dependent changes in connections between many cells in different systems.
Under this hypothesis characteristics of a given experience are stored relative to prior knowledge and linked together with the prior experiences.  Hence if the prior knowledge is not very extensive new information will be stored ina very minimalistic manner.  Memories  are then reconstructed when they are retrieved.  This mechanism for encoding and storing information can be simulated on a computer.  Output from networks such as these demonstrate many of the phenomenon discussed in the Memory Distortion book such as omission, blending, false beliefs, cryptomania.

These models employ biologically plausible networks and synaptic modification rules that mimic those seen in studies of activity dependent synaptic plasticity.