A pioneer of the scientific study of memory Hermann Ebbinghaus

Document Sample
A pioneer of the scientific study of memory Hermann Ebbinghaus Powered By Docstoc
					         Subsequent modifications to the model:
       Collins & Loftus’ ‘spreading activation theory’
• to deal with critique of original Collins & Quillians model

       + no more strict hierarchical organization

       + introduction of different semantic distances
         between concepts (to account for relatedness effects)

         -> flow of activation between neighboring nodes
            varies according to distance

       + introduction of different types of links:

         ‘is a’ ‘is not a’ ‘has’   ‘can’   ‘cannot’
  Subsequent modifications to the model:
Collins & Loftus’ ‘spreading activation theory’

                                       types of links
                                       not shown)
   Accounting for semantic relatedness in Collins &
                    Loftus model
• systematic differences in RT for invalid (negative) sentences
       poodle is a bird vs poodle is a mineral
         -> faster RT for second than for first sentence
         -> why??
  ->   poodle and bird have several links to shared
       features (e.g. has eyes, has offspring, is animal)
  ->   spreading activation to shared features in favour of ‘yes’
       needs to be weighted against by distinguishing features;
       (e.g., can fly) in favour of ‘no’; weighing takes time
  ->   few shared features between poodle and mineral;
       many distinguishing features; not much weighing
What can studies in cognitive neuroscience tell us
  about how semantic memory is organized?

• category-specific semantic memory impairments:
  patients with deficits in semantic knowledge for living but
  not non-living things (resulting from herpes simplex
  encephalitis, stroke, or head injury)

  e.g. picture naming much worse for animals and plants
       than tools or vehicles

  other patients show opposite pattern (less frequent)
  Impairments in semantic memory other than
             semantic dementia
• Questions arising from observation of category-specific
  semantic memory impairments:

  Is living and non-living knowledge stored in different parts
  of brain?

  Is the brain’s knowledge base organized according to

   -> most researchers think ‘no’
 Category-specific impairments in semantic memory

• important differences between living and non-living objects:

       + living objects distinguished from each other primarily
         based on visual info
       + non-living objects distinguished primarily based on
         functional info

e.g.     camel: large animal with humped back, long neck etc.

         wrench: tool with jaws for gripping, turning, or twisting

-> category-specific impairments may reflect breakdown for
   specific types of object knowledge (visual vs functional)
   Neural organization of semantic knowledge
• sensory/motor model of semantic knowledge (A. Martin)

 core idea:
 meaning of concepts (words and objects) is not unitary but
 composed of different features stored in different brain

      - some features describe appearance of objects

      - other features describe functions; often functions
        are tied to specific actions (stored motor programs)

              e.g. fire extinguisher
 Evidence from functional neuroimaging to support
   sensory/motor model of semantic knowledge

• PET study by A. Martin (1995)

      presentation of b & w line-drawings of objects

      three different experimental conditions:

      (1) name object
      (2) generate real-world color for object     Note: not
      (3) generate real-world action for object    perceptual
Results from PET study by A. Martin (1995)

        colour > action       action > colour

-> suggests distinct neural basis for storage of
   different features
What can studies in cognitive neuroscience tell us
  about how semantic memory is organized?

• neuroimaging findings and findings from patients with
  category-specific semantic impairments suggest that
  different types of semantic features (i.e., knowledge)
  stored in different parts of human brain
   Implicit versus explicit memory: Definitions

• implicit memory:
  past experiences influence perceptions, thoughts and
  actions without awareness of person that any info
  from past is accessed

• explicit memory:
  conscious access to info from past (“I remember
  that..” )

  -> involves conscious recollection
  -> term generally used synonymously with episodic
                Implicit memory:
       What studies in amnesia have told us

• concept evolved out of neuropsychological research in
  patients suffering from amnesia; later studied in
  cognitive psychology, social psychology, developmental
  psychology etc.

• first systematic study in densely amnesic patient H.M.
  (Brenda Milner, 1960s, in Montreal)

• H.M.:
  in 1953, bilateral surgical removal of medial temporal-
  lobes to stop his epileptic seizures, which couldn’t be
  treated with medication
    Brief introduction to human neuroanatomy
Lateral view                         Medial view

Top view                             Bottom view
                   Brain lesion in H.M.
 • removal of amygdala, hippocampus, and parahippocampal
   gyrus in both hemispheres of patient H.M. for relief of

Medial view
of right
          Cognitive impairments of H.M.

• after surgery severe anterograde amnesia that has
  persisted with little improvement to this day

  -> no ability to keep track of daily events and
     episodes (no acquisition of new info in episodic

  -> no conscious recollection of episodes/events
     since surgery
       Anterograde amnesia in everyday life

Examples of problems:

• learn to recognize new people and remember their
• find way about in new environments
• remember where things were left
• remember what has happened during a day
• remember what one has done and what one hasn’t
• remember what was read in newspaper / seen on TV
               Amnesic patient H.M.
            Why is his case so important?
• first clear-cut demonstration of isolated memory
  impairment (not dementia!) due to damage in
  specific brain structure (medial temporal lobes)

-> established new
research field on
memory functions of
medial temporal lobes
in human and non-
human species
-> associated with
name of Canadian
psychologist Brenda
 Motor learning: a type of memory that is
 normal in H.M.
• mirror tracing task:
  builds on procedural
  memory for skills

• improvement in
  tracing performance
  with practice but no
  recollection of
  previous training
  (i.e. episodes)

-> suggests that H.M.
  retains some info over
  long-term and can
  access it implicitly
    Is normal implicit memory limited to motor
            learning in amnesia? No!
• Warrington & Weiskrantz (1970) demonstrate that
  amnesic patients show preserved implicit memory on list
  learning with word stem completion task

• Study: table, garden, telephone, lamp, umbrella etc.

  Test:      - Free Recall
             - Yes/No recognition
             - Word stem completion:     tab____?
             (Instructions: think of any word that fits stem)

• Finding: amnesic patients show memory deficits on first 2
  tests but normal performance on word stem completion
Interpretation of normal implicit memory in amnesia

• interpretation of normal memory effects on word-stem
  completion task by amnesic patients: repetition priming

  BUT: but how can we show that patients access info
  implicitly and do not explicitly remember words from the
  list? Maybe performance reflects residual weak explicit

  first hint:
  amnesic patients do not seem to be aware that any
  completed words were on study list

  -> subsequent systematic study by Graf et al. (1984)
     showed that task instructions are critical
           Repetition priming in amnesia:
     Lack of retrieval intentionality is important
• Graf et al. 1984
  Study:      intentional learning of word list
              (table, garden, telephone, umbrella etc.)
              - free recall
  intent.     - cued recall: complete word stem with specific
                word from study list
                             umb____ ??
  non         - word stem completion: complete word stem
  intent.       with first word that comes to mind

-> same cue in cued recall and word-stem completion
  but latter does not require conscious access to past
           Repetition priming in amnesia:
     Lack of retrieval intentionality is important

-> normal performance of patients on completion task but
   impaired cued recall and free recall
-> type of access (implicit vs explicit) to stored info critical

Shared By: