Lecture 9 - the remembering brain P2 Flashcards
short term memory
memory for active information
long term memory
stored information that isn’t presently/consciously accessible
working memory
a system for the temporary storage and manipulation of information
- the active manipulation of info within the STM store to serve higher cognitive functions
Baddeleys (2000) model of working memory
seperate STM stores and an executive system for manipulating and controlling info within the stores
- visuospatial sketchpad
- episodic buffer
- phonological loop
where is the phonological store in the brain
- lies posteriorly
- in parietal lobes
Smith et al 1996 - PET study on Model of Working Memory
- tested Ps on short-term retnetion fo either letters (verbal STM) or location of markers (visual STM)
- found distinct brain regions are active in two tasks
- verbal STM - left hemisphere
- visuospatial STM - right hemisphere
what did Miller (1956) suggest the capacity of the STM was:
7 +/- 2 - meaningful ‘chunks’ of information
what does chunking rely on
- relies on LTM and on LTM and STM interactions
what did Cowan 2001 suggest the capacity of the STM is
lower - around 4
why is the span of the STM capacity not related to the meaningfulness of chunks? (BADDELEY 1996)
- span length lower for polysyllabic words (eg. skeleton)
- span length lower for phonologically similar words (map, cat, cap etc.)
what does the capacity of the STM depend on according to Baddeley
- opportunity to rehearse the material - ARTICULATORY SUPRESSION
Span tasks rely on what two components according to Baddeley?
- phonological store component (verbal STM)
2. and a rehearsal mechanism
Paulesu et al 1993 - PET study on the separation of the phonological loop
PS performed task either:
a. STM for letters (both store and rehearsal components)
b. rhyming judgements (rehearsal only)
- phonological store > LEFT SUPRAMARGINAL GYRUS
- rehearsal system > BROCAS
two models of Visuospatial STM
SLOT MODEL: a small number of memory ‘slots’ each capable of storing a single visual object
RESOURCE MODEL: no upper limit on no. of items stored - BUT the more items held in memory - the less precisely they can be recalled.
Luck and Vogel (1997) on the slot model
- brief displays of arrays (incl. coloured squares and oriented coloured lines)
- PS could retain information about only FOUR colours or orientations in visual WM at one time.
BUT they could retain info about colour AND orientation of four objects
so - visual STM capacity understood in terms of integrated objects NOT individual features.
Bays, Catalao and Hussain (2009) and the resource model
- Ps exposed to arrays which contained 1-6 coloured squares
- they were presented with test array with a prop to recall one of the squares colours.
- precision of recall decreased with increase of item on the array
Ranganath et al 2004 - neural correlates of visuospatial STM (procedure)
- delayed matching sample: PS shown Face/place and asked to keep it in mind - then test stimulus shown
- in delayed paired associates - PS had previously learned to pair particular face and place images together
ignore
okay
what does the delayed-response task measure
this measures working memory
delayed response task in monkey
monkey must continue t retain the location of the unseen food during delay period (WM)
- lesions to prefrontal cortex impacts this ability
- but is this due to deficit in forming associations or working memory?
working memory vs associative memory - delayed response task
- food is paired with a visual cue
- task measures animals ability to retain long-term rules
- no need for animal to retain visuospatial info during delay
- PFC lesions do not damage this ability (only damage WM task)
response of PFC neurones in delay response task
- prefrontal neurons differentially respond to stages of experiment (cue-delay-response)
- neurones active during delay period - these neurons provide neural correlate for retaining visuo-spatial info
how long do PFC neurons stay active for in delay-response-task
until they give response (varied with each task)
PFC in working memory for faces (Druzgal and D’Esposito, 2003) - PROCEDURE
- four stimuli presneted successively in encoding section - either intact or scrambled faces
- Ps had to remember only inatct faces
- number of intact faces varied with each trial (more faces.= more demands on WM)
- probe presented was either a match or not
PFC in working memory for faces (Druzgal and D’Esposito, 2003) - FINDINGS
- activation in lateral PFC was sustained during delay - needed for encoding of probe
- activity greater when more faces had to be remembered
- FFA showed greatest activation when face was shown, but LPFC showed greater sustained activation in delay period
- BUT FFA activity not at baseline
- support idea that PFC critical for WM
WM: interaction between PFC and posterior cortex
PFC activation reflects a representation of task goal
WM relies on interaction between PFC and other regions of brain e.g. FFA
Petrides’ theory of working memory
suggested division of function sin PFC
- two separate processes: maintenance and manipulation
Petrides’s neuropsychological findings
patients with PFC damage impaired at self-ordered pointing task
- whenever Ps needed to retain spatial info - activation in ventrolateral PFC
- whenever Ps needed to retain and update new locations - activation seen in dorsolateral PFC
function of the ventrolateral PFC
LONG-TERM MEMORY ENCODING
- selecting and maintaining info within WM
function of dorsolateral PFC
- manipulating and monitoring information in WM
What is the Modal Model (Atkinson and Shiffrin)
- they propose there are different stores for STM and LTM
Patient HM and different STM and LTM stores
- inability to make new memories (no transfer to LTM)
- but intact STM
Patient KF and different STM and LTM stores
- left parieto-occipital damage
- normal LTM, good long-term learning
- smal STM span
unitary models of WM/LTM
- working memory is just temporary activation of LTM memories
- Central executive proposed to be responsible for selecting and activating LTM representations to bring them into ‘focus of attention’
- BUT capacity?
WM/LTM final conc.
- working memory and LTM are not completely distinct
PFC and memory encoding (Kellley et al 1998)
- encoding of words or semantic materials involve the LEFT PFC
- encoding of spatial info or faces involve the RIGHT PFC
PFC and retrieval
- activity in DLPFC increases with increased demands
e. g. free recall, recall (vs recognition) low confidence judgements = INCREASED DLPFC activity - PFC damage results in more severe impairment during free recall as compared to recognition.
source monitoring
the process by which retrieved memories are attributed to their original context
