Semantic Memory

Semantic Memory

1. An individual's store of knowledge about the world. Content is abstracted from experience and is therefore said to be conceptual, and generalised without reference to any specific experience

2. There is much overlap in the knowledge that each of us have stored in semantic memory (e.g. knowledge of the world) - but there are also individual differences (we generally have more knowledge for topics of interest)

   • Chassy & Gobet (2011): strong relationship between chess -playing skill and knowledge of opening moves (average players had memorised far fewer opening moves than masters)

3. Semantic (general knowledge about the world) vs Episodic (memory for events occurring at a specific time in a specific place) Memory

   • Separate Systems?
     • Spiers, Maguire & Burgess (2001): 147 cases of amnesia, episodic memory impaired in all cases - but many of cases had only modest problems with semantic memory. Suggests the two types of memory are different.
     • The extent of retrograde amnesia (prior events) for episodic memories often spans several years, but retrograde amnesia for semantic memories is generally reasonably intact (except for knowledge acquired shortly before onset of amnesia)
     • Semantic dementia: associated with damage to anterior frontal temporal lobe, but amnesia it is typically the medial temporal lobe that is most damaged. Ps find it very hard to access info about most concepts stored in semantic memory
       • Adlam, Patterson & Hodges (2009): ps with S.D. asked to remember what tasks they had performed 24 hours earlier, where and when during the session they occurred (episodic memory. Ps as a group performed at comparable level to healthy controls.
   • Interdependent Systems?
     • Burianova, McIntosh & Grady (2010): compared patterns of brain activation during episodic, semantic, and autobiographical memory retrieval. The same neural network including various frontal, temporal and parietal areas activated during these types of memory.
     • Kan, Alexander & Verfaellie (2009): A task apparently involving only episodic memory also involved semantic memory - amnesic patients and controls instructed to learn prices of grocery items. Prices of some items corresponded to p's prior knowledge (congruent) whereas others were incongruent. p's (who had fairly intact semantic memory) and controls had better memory performance for congruent pieces than incongruent. Ps with poor semantic memory unable to use their SM effectively on the memory tasks and so showed no congruency effect.
     • Irish et al (2011): there was a non-significant difference between patients with S.D. and healthy controls in their recall of recent autobiographical memories. But the S.D. ps performed much worse than controls when recalling remote autobiographical memories.
     • Irish & Piguet (2013): sensory and perceptual info can be used to recall recent autobiographical memories but not remote ones. Recall of remote memories requires semantic knowledge to provide a framework or scaffolding facilitating the retrieval of episodic memories.
   • Hierarchical network model (Collins & Quillan 1969)
     • S.M is organised into a series of hierarchical networks.
     • tested model by using task on which ps decided as quickly as possible whether sentences were true or false. According to the model it should be possible to decide very quickly that "a canary is yellow: is true as the two concepts are stored at the same level. "A canary can fly" should take longer as the concept and the property are separated by one level in the hierarchy etc) - this was found in the study: time taken to respond to true sentences became slower with increasing separation between subject of sentence and the property.
     • Limitations
       • Typicality effect: the time taken to decide a category member belongs to a category is less for typical than atypical members.
         • Rosch & Mevis (1975): oranges, apples, bananas and peaches were rated as much more typical fruits than olives, tomatoes, coconuts and dates.
       • Conrad (1972): when familiarity of sentences was controlled, hierarchical distance between subject and property had little effect on verification time.