Semantic memory

Semantic memory

1. Semantic memory

   • General factual info - permanent memory store for our world knowledge (no representation of when or where we learned the info) e.g what is the capital of Greece, what colour is a canary
   • Episodic vs semantic - epsisodic = remembering an event at a given time in a given place, memory for events occurring at a specific time - semantic = remembering what coffee tastes like, general knowledge
     • Patients with amnesia exhibit impaired episodic memory but modest problems with semantic memory (amnesia is mostly associated with damage to the temporal lobe) Patients with semantic dementia exhibit severe problems with semantic memory but relatively intact episodic memory - SD is mostly associated with damage to the anterior temporal lobes - these two forms of long term memory operate relatively independent of each other
   • Importance - we know how to speak a language because the sounds and meanings of words are stored in semantic memory - retrieve specific info from concepts we have previously learned - recognise objects
     • Definitions - semantic memory = an individuals store of knowledge about the world - concepts = mutual representations and the fundamental units of thought - propositions = relationships between concepts
       • 'A circle is on top of a triangle = verbal representation - ON (CIRCLE, TRIANGLE) = propositional representation - image itself = mental image representation - amodal vs embodied and situated accounts

2. Defining or core features approach

   • Word meaning is a list of defining, core features e.g cat = animal, mammal, pet - but where do you draw the line? is a monk a bachelor? some words do not have consistent features e.g game - some meanings are not equally good across contexts e.g red (as in hair) and red (as in fire engine red)
     • Hierarchical netwrok model (Collins and Qullian) - the meaning of the word is represented by a hierarchy of nodes and links and is given by how it is embeded within a network of other meanings
       • Support = sentence verification task - decide as quickly as possible whether sentences are true or false - a canary is yellow, a canary can fly, a canary has skin = true - a canary is a fish = false - reaction time increases - but - relatedness effect - the more related 2 things are the harder it is to disentangle them - e.g a pine is a church, a pine is a flower - typicality effect - not all true statements with the same semantic distance repsonded to equally as fast - a canary is a bird, a penguin is a bird -
         • Spreading activation model (collins and Loftus) - activity at one node causes activation as other nodes via links - spreading activation is automatic and decreases the further it travels in the networks (e.g the total amount of activation that can be spread is limited)
           • Sentence verification task - priming effect - semantic and mediated priming -
             • The priming effect occurs when presenting one stimulus at time 1 helps people respond to another stimulus at time 2 - people respond faster to target words preceded by related, compared to unrelated, primes - semantic priming effect - duck primes goose, horse does not - mediated priming effect - lion primes stripes via activation of 'tiger'
               • McNamara - same thing but some primes closer semantically than others, e.g red primed for rose, and red primed for flower - greater semantic priming when the prime target words were very closely related semantically
                 • IMPORTANT - does priming result from sharing nodes in the network or does priming occur simply because words co-occur>
                   • Associative vs semantic priming - word pairs - some associated, some associated and semantic, some unrelated and some just semantic e.g traffic and jam, lemon and orange, bread and tower and bread and cereal - task is to read and remember - associated word pairs decreased the magnitude of N400 but semantically related word pairs did not - semantic association through the semantic networks is not the same as simply relating words because they co-occur
                     • But - the notion that each concept is represented by a single node is over simplified - what about abstract concepts such as justice? does each concept have a fixed mental representation?

3. Embodied and situated approach

   • Indexical hypothesis - verbal knowledge is tied to our perceptual experience - e.g a bell activates a strong auditory response, rainbow a strong visual response and coffee a strong olfactory response - affordances - are determined by the interaction of our perceptual abilities and the physical characteristics of our bodies and the physical properties of objects in the real world
     • Support - gestures while speaking are universal - behavioural demonstration of the censoring motor influences on language processing tasks (inter relation meaning action and emotion) - neautoscientific demos of the overlap between areas in the brain contributing to language and to action
       • Gestures as visible embodiment - gestures reflect and stem from speakers mental images and facilitate speech production and meaning representation

Images and files in the map