Study Guides (260,861)
CA (126,135)
York (10,271)
PSYC (1,204)
PSYC 3265 (12)
Final

PSYC 3265 Final: Memory Test 2 Lecture Notes

24 Pages
65 Views

Department
Psychology
Course Code
PSYC 3265
Professor
Norman Park

This preview shows pages 1-3. Sign up to view the full 24 pages of the document.
Retrieval 6:00
MTL Hippocampal System
- w/in hippocampus: sub-fields! Hippocampus and para-hippocampal cortex surrounding it
- hippocampus next to amygdala
- hippocampus well-suited for formation, storage and retrieval of memories! b/c it represents a closed
circuit. Info keeps flowing, processed, elaborated on over and over. Intricate sub-connections w/in
hippocampus to allow memories to be re-activated at retrieval
Hippocampal Formation
- Hippocampus proper
CA1, CA2, CA3, CA4 subfields *distinguished by cellular structure and connectivity
CA1 particularly vulnerable to damage from hypoxia (loss of O2 to the brain, can be from cardiac
arrest)
- Dentate gyrus:
narrow, concave, wraps around CA4 subfield.
- Subiculum
continuation of CA1 subfield
connects to entorhinal cortex hippocampus and back *closed circuit
- entorhinal neurons (major input to hippocampus) dendate gyrus CA3 CA1 subiculum AND back
around = entorhinal cortex *perferant pathways
- another pathway that exits the hippocampus via the fornix (another major output from the hippocampus
that leads to other brain areas/structures. Bundle of axon tracts that leave hippocampus and send info to
the thalamus, also info from mammillary bodies hypothalamus PFC)
- one pathway for remembering ANOTHER familiarity
- Schaeffer collaterals: auto-associative, allows for info to reverberate w/in the CA3 sub-fields pattern
completion. CA3 related to retrieval, complete memory of an event based on some cues AND info
continually processed in a loop w/in that 1 sub-field
Major input
- receiving info from all major associative neo-cortex areas = association cortex in frontal, parieto-occipital,
and temporal lobes (lateral regions)
- relayed from perirhinal and parahippocampal cortices to entorhinal cortex into the hippocampus
Major output
- Subiculum projects to entorhinal cortex and info flows right back out to association regions *info
elaborated on in the hippocampus, then sent out to the regions that were involved in the initial encoding of
an episode
- same regions involved in initial perceiving/encoding, re-activated when exposed to a stimulus on
subsequent trials (implicit or explicit)
- Fornix projects to diencephalon *important when distinguished b/w recollection and familiarity
* maybe memory not a set of systems WM, encoding, retrieval … MAYBE should be processes same
principles that apply to all types of memories)
PS/PC
find more resources at oneclass.com
find more resources at oneclass.com
- Pattern separation (PS, operate at encoding): process that parses similar events (share many features)
into more distinct, non-overlapping representations. Mechanism that allows us to code the details of similar
events as different.
- Pattern completion (PC, operate at time of retrieval): process that interacts with partial information to
reconstruct a complete event. A cue that allows you to retrieve the rest of the memory, based on partial info
re-construct the complete event.
- different processed map on to areas of the hippocampus!
Encoding versus Retrieval: PS versus PC
Current evidence
- initially predicted by Computational models
- Rodent studies = lesioned dendate gyrus, CA3 sub-field AND find the corresponding deficits
- High-resolution fMRI = look at activation at the level of sub-fields, see activation in dendate gyrus versus
CA. )ssues = cant separate CA and CA, see activation in sub-field versus dendate gyrus
- Aging and amnesia = patients w/ hippocampal damage show deficits. Hard to separate the 2 processes
- different processes believed to map onto different sub-fields w/in the hippocampus
- pattern separation occurs in the dendate gyrus. Info from entorhinal cortex, made more distinct here
- neurogenesis occurs in dendate gyrus. Dont know the function of those new neurons, we do know
there is new growth of cells
- pattern completion recurrent collaterals = takes a cue, and completes a pattern. Cue received w/in
region CA3, then can associate cues w/ other details of the event to complete the representation. Related to
retrieval
Patient BL
- 54 y.o. male, 13 years of education
- hypoxia-ischemia secondary to electrical injury/cardiac arrest loss of O2
- CA1 vulnerable to loss of O2, BUT his dendate gyrus specifically affected
- hard to form new memories, but not totally impaired
- average intelligence
- borderline-low average delayed recall
- intact performance on non-memory related cognitive tasks
- hard to differentiate b/w the 2
- dendate gyrus shrunken, all other parts equal in V
Pattern Seperation
- related to encoding
- MST / BPS-O (Stark et al., 2013, 2015)
Study: 128 images of common objects, incidental encoding based on semantic cues of the object
(indoor/outdoor)
Test: 192 images, 64 studied (green), 64 similar lures (blue), 64 foils (red) need to decide if they
studied it: old judgement has to be identical, new hadnt studied it, similar item similar to old
- Participants = B.L. and 20 age-matched controls (mean age, 52; 10 women)
Pattern Separation Results
- hypothesis: if a PS issue, should do well at old, reject new but trouble w/ similar b/c PS means
distinguishing items that overlap from study to test
- results: identical to control for identifying old AND no issue saying the new items are new. Deficit: lures
(decifit specific to PS)
Pattern Completion
- take a cue, use it to fill in the rest of an event that they studied. Want to show dendate gyrus specific to PS,
need to show BL doesnt have issue w/ PL. PC appears to occur at retrieval
- Study: 5 line drawings of scenes (and their fragmentes), followed by brief forced-choice recognition
find more resources at oneclass.com
find more resources at oneclass.com
- Test: original 5 scenes learned items plus 5 new images new items. Need to describe the scene and
say how confident they are.
images presented at various levels of completeness
- Participants: B.L. and 17 age-matched controls (mean age, 51; 6 women
Pattern Completion Results
- when 100% complete scenes: can complete scenes for different levels of degradation at the same level of
controls, for the images he studies … PC intact
- when presented w/ new stimuli, hard to differentiate new and old items. Deficit specific to items that look
close to each other (reflects PS problem)
- PC seems intact dendate gyrus needed for PS (confusion b/w old and new iems) at encoding, not PC at
retrieval
Conclusion
- deficient pattern separation in B.L. provides compelling evidence that the dentate gyrus plays a critical
role in this process
- impaired pattern separation results in a bias towards pattern completion. Shows more PC than controls,
b/c PS doesnt happen
Encoding Versus Retrieval
Episodic encoding and binding in MTL
- MTL (contains hippocampus) believed to be responsible for binding features to an integrated memory
trace at encoding
receive highly processed input from many areas when an event is encoded
MTL then binds together input at encoding coherent memory trace
Retrieval and MTL
- at retrieval, a retrieval cue is encoded as part of the memory. Allows you to re-activate. w/ an obvious cue,
use MTL
- content addressable memory = retrieval cue often consists of some of input processed at encoding
- processed retrieval cue converges on MTL at the time of retrieval, triggers pattern completion (activates
other associations) within the hippocampus, which in turn reactivates information in neocortex (same
regions involved in initial perception at the time an event is experienced)
- MTL at the centre of encoding/retrieval, but a different role
- At encodingan event is processed by regions associated with different features of event; processed
information converges at hippocampus and features are bound
- At retrieval-- cue processed by regions associated with features of cue, and then converges to
hippocampus. Cue re-activates memory trace via hippocampus. if retrieval successful cue then connects
with memory trace projects to regions associated with memory.
- episodic retrieval: processes by which stored memory traces are retrieved
- assumed that retrieval produces subjective experience of consciously remembering the past. Feeling of
recollecting or re-experiencing all the details of the episode as if you experience it again
- episodic retrieval is assumed to depend on hippocampal regions that support pattern completion and
frontal lobes that support strategic retrieval cue that isnt effective or hard to access info, call on brain
regions in central executive i.e.. PFC, strategically retrieve info)
thus, there are multiple ways in which a retrieval cue can access memory trace, and partial
information may be enough to access memory trace
Episodic Retrieval
- evidence that hippocampus important for retrieval!
find more resources at oneclass.com
find more resources at oneclass.com

Loved by over 2.2 million students

Over 90% improved by at least one letter grade.

Leah — University of Toronto

OneClass has been such a huge help in my studies at UofT especially since I am a transfer student. OneClass is the study buddy I never had before and definitely gives me the extra push to get from a B to an A!

Leah — University of Toronto
Saarim — University of Michigan

Balancing social life With academics can be difficult, that is why I'm so glad that OneClass is out there where I can find the top notes for all of my classes. Now I can be the all-star student I want to be.

Saarim — University of Michigan
Jenna — University of Wisconsin

As a college student living on a college budget, I love how easy it is to earn gift cards just by submitting my notes.

Jenna — University of Wisconsin
Anne — University of California

OneClass has allowed me to catch up with my most difficult course! #lifesaver

Anne — University of California
Description
Retrieval 6:00 MTL Hippocampal System - w/in hippocampus: sub-fields! Hippocampus and para-hippocampal cortex surrounding it - hippocampus next to amygdala - hippocampus well-suited for formation, storage and retrieval of memories! b/c it represents a closed circuit. Info keeps flowing, processed, elaborated on over and over. Intricate sub-connections w/in hippocampus to allow memories to be re-activated at retrieval Hippocampal Formation - Hippocampus proper  CA1, CA2, CA3, CA4 subfields *distinguished by cellular structure and connectivity  CA1 particularly vulnerable to damage from hypoxia (loss of O2 to the brain, can be from cardiac arrest) - Dentate gyrus:  narrow, concave, wraps around CA4 subfield. - Subiculum  continuation of CA1 subfield  connects to entorhinal cortex  hippocampus  and back *closed circuit - entorhinal neurons (major input to hippocampus)  dendate gyrus  CA3  CA1  subiculum AND back around = entorhinal cortex *perferant pathways - another pathway that exits the hippocampus via the fornix (another major output from the hippocampus that leads to other brain areas/structures. Bundle of axon tracts that leave hippocampus and send info to the thalamus, also info from mammillary bodies  hypothalamus  PFC) - one pathway for remembering ANOTHER familiarity - Schaeffer collaterals: auto-associative, allows for info to reverberate w/in the CA3 sub-fields  pattern completion. CA3 related to retrieval, complete memory of an event based on some cues AND info continually processed in a loop w/in that 1 sub-field Major input - receiving info from all major associative neo-cortex areas = association cortex in frontal, parieto-occipital, and temporal lobes (lateral regions) - relayed from perirhinal and parahippocampal cortices to entorhinal cortex into the hippocampus Major output - Subiculum projects to entorhinal cortex and info flows right back out to association regions *info elaborated on in the hippocampus, then sent out to the regions that were involved in the initial encoding of an episode - same regions involved in initial perceiving/encoding, re-activated when exposed to a stimulus on subsequent trials (implicit or explicit) - Fornix projects to diencephalon *important when distinguished b/w recollection and familiarity * maybe memory not a set of systems (WM, encoding, retrieval …) MAYBE should be processes (same principles that apply to all types of memories) PS/PC - Pattern separation (PS, operate at encoding): process that parses similar events (share many features) into more distinct, non-overlapping representations. Mechanism that allows us to code the details of similar events as different. - Pattern completion (PC, operate at time of retrieval): process that interacts with partial information to reconstruct a complete event. A cue that allows you to retrieve the rest of the memory, based on partial info  re-construct the complete event. - different processed map on to areas of the hippocampus! Encoding versus Retrieval: PS versus PC Current evidence - initially predicted by Computational models - Rodent studies = lesioned dendate gyrus, CA3 sub-field AND find the corresponding deficits - High-resolution fMRI = look at activation at the level of sub-fields, see activation in dendate gyrus versus CA3. Issues = can’t separate CA1 and CA3, see activation in sub-field versus dendate gyrus - Aging and amnesia = patients w/ hippocampal damage show deficits. Hard to separate the 2 processes - different processes believed to map onto different sub-fields w/in the hippocampus - pattern separation occurs in the dendate gyrus. Info from entorhinal cortex, made more distinct here - neurogenesis  occurs in dendate gyrus. Don’t know the function of those new neurons, we do know there is new growth of cells - pattern completion  recurrent collaterals = takes a cue, and completes a pattern. Cue received w/in region CA3, then can associate cues w/ other details of the event to complete the representation. Related to retrieval Patient BL - 54 y.o. male, 13 years of education - hypoxia-ischemia secondary to electrical injury/cardiac arrest  loss of O2 - CA1 vulnerable to loss of O2, BUT his dendate gyrus specifically affected - hard to form new memories, but not totally impaired - average intelligence - borderline-low average delayed recall - intact performance on non-memory related cognitive tasks - hard to differentiate b/w the 2 - dendate gyrus shrunken, all other parts equal in V Pattern Seperation - related to encoding - MST / BPS-O (Stark et al., 2013, 2015)  Study: 128 images of common objects, incidental encoding based on semantic cues of the object (indoor/outdoor)  Test: 192 images, 64 studied (green), 64 similar lures (blue), 64 foils (red)  need to decide if they studied it: old judgement (has to be identical), new (hadn’t studied it), similar (item similar to old) - Participants = B.L. and 20 age-matched controls (mean age, 52; 10 women) Pattern Separation Results - hypothesis: if a PS issue, should do well at old, reject new but trouble w/ similar b/c PS means distinguishing items that overlap from study to test - results: identical to control for identifying old AND no issue saying the new items are new. Deficit: lures (decifit specific to PS) Pattern Completion - take a cue, use it to fill in the rest of an event that they studied. Want to show dendate gyrus specific to PS, need to show BL doesn’t have issue w/ PL. PC appears to occur at retrieval - Study: 5 line drawings of scenes (and their fragmentes), followed by brief forced-choice recognition - Test: original 5 scenes (“learned items”) plus 5 new images (“new items”). Need to describe the scene and say how confident they are.  images presented at various levels of completeness - Participants: B.L. and 17 age-matched controls (mean age, 51; 6 women Pattern Completion Results - when 100% complete scenes: can complete scenes for different levels of degradation at the same level of controls, for the images he studies … PC intact - when presented w/ new stimuli, hard to differentiate new and old items. Deficit specific to items that look close to each other (reflects PS problem) - PC seems intact  dendate gyrus needed for PS (confusion b/w old and new iems) at encoding, not PC at retrieval Conclusion - deficient pattern separation in B.L. provides compelling evidence that the dentate gyrus plays a critical role in this process - impaired pattern separation results in a bias towards pattern completion. Shows more PC than controls, b/c PS doesn’t happen Encoding Versus Retrieval Episodic encoding and binding in MTL - MTL (contains hippocampus) believed to be responsible for binding features to an integrated memory trace at encoding  receive highly processed input from many areas when an event is encoded  MTL then binds together input at encoding  coherent memory trace Retrieval and MTL - at retrieval, a retrieval cue is encoded as part of the memory. Allows you to re-activate. w/ an obvious cue, use MTL - content addressable memory = retrieval cue often consists of some of input processed at encoding - processed retrieval cue converges on MTL at the time of retrieval, triggers pattern completion (activates other associations) within the hippocampus, which in turn reactivates information in neocortex (same regions involved in initial perception at the time an event is experienced) - MTL at the centre of encoding/retrieval, but a different role - At encoding—an event is processed by regions associated with different features of event; processed information converges at hippocampus and features are bound - At retrieval-- cue processed by regions associated with features of cue, and then converges to hippocampus. Cue re-activates memory trace via hippocampus. if retrieval successful cue then connects with memory trace projects to regions associated with memory. - episodic retrieval: processes by which stored memory traces are retrieved - assumed that retrieval produces subjective experience of consciously remembering the past. Feeling of recollecting or re-experiencing all the details of the episode as if you experience it again - episodic retrieval is assumed to depend on hippocampal regions that support pattern completion and frontal lobes that support strategic retrieval (cue that isn’t effective or hard to access info, call on brain regions in central executive i.e.. PFC, strategically retrieve info)  thus, there are multiple ways in which a retrieval cue can access memory trace, and partial information may be enough to access memory trace Episodic Retrieval - evidence that hippocampus important for retrieval! - Eldridge et al. (2000): importance of hippocampus during retrieval of episodes supported by = hippocampus activated during successful retrieval but not during unsuccessful retrieval Brewer et al., 1998 - Indoor/outdoor judgments to complex scenes during event-related fMRI - Subsequent memory test 30 min. after scanning (new vs. old)  Remember (distinct recollection of seeing  more MTL activity: parahippocampus and PFC at encoding/retrieval) vs. know (feel familiar) vs. forgot - close correspondence b/w areas active at encoding/retrieval - regions not activated when material forgotten - cues that lead to successful retrieval: strength of encoding memory. If info encoded deeply (strong memory trace), more likely info will be retrieved - activity in parahippocampal and frontal cortex at encoding of complex pictures predicted how well material later recalled (not activated when material forgotten) - strength of encoded memory predicts subsequent retrieval (one of the factors that determine retrieval success) - patient BL: encoding depends on dendate gyrus BUT intact retrieval Retrieval Hippocampal patients (e.g., HM) - impaired memory for info acquire long ago, for some memories - temporally graded retrograde amnesia where remote memories (long before lesion) could be retrieved but recent memories could be accessed, suggesting that MTL involved in retrieval BUT not always necessary for retrieval of old memories Frontal lobe patients - damage to PFC regions: hard time w/ retrieval - test: PFC patients given famous faces from different decades (only in that decade). Hard time retrieving names of faces from all time periods (before and after lesion) - flat gradient in retrograde amnesia; remote memories can’t be retrieved under recall (remembering famous names in response to photos) *worse than controls for all time periods  Improved with recognition (structure) - lesions lateral - patients worse that controls for all time periods, better for more remote faces. Hard time naming faces in ALL time periods - if it’s a recognition test (given cues: correct and incorrect names): improved to normal overall, compared to recall - in recognition give people cue and distraction  this is retrieval support - issue not of storage/encoding deficit (info acquired prior to lesion), or consolidation (shouldn’t be able to recognize correct at test above chance) - evidence that PF have a deficit in memory retrieval, unlike MTL patients Externally vs. internally generated cues - external cues easier to answer (info available w/in the cue, relates to the memory you try to retrieve) versus re-constructing the memory to get to the same answer (engage in strategic retrieval, effortful) - MTL/hippocampal module (mediates encoding, storage, & retrieval of explicit, associative/cue-dependent memory). External cue: MTL takes cue and completes the pattern at retrival - w/o accessible cues, need additional neural region needed for:  devising strategies to retrieve info when cues unavailable (to re-construct the memory). Same processes to monitor and verify outcome of search  ex/ confabulation: affect ability to monitor outcome of search b/c they cant distinguish b/w true and false memories Retrieval: Divided Attention (DA) - successful at retrieval b/c of proper attention to cue. - DA: try to retrieve info while doing another task. deplete attn at the time of encoding/retrieval, affect retrieval success. Tests w/ dual task: try to learn/retrieve words while doing a distractor task - Q: Is interference effect from DA at retrieval material- or process-specific? *perhaps the 2 test was just another memory task (do they need to compete) - Experiment 1: Is mnemonic processing in a distracting task crucial to produce interference with retrieval? - Experiment 2: Does the semantic or phonemic aspect of the distracting task influences size of interference effect on memory retrieval? Experiment 1 - participants intentionally encoded auditorally presented list of words (common nouns), followed by free recall task - prior to retrieval, engage in distracting task (DA = demanding, didn’t add memory load): animacy/syllable decisions about words presented visually on a screen  participants continued to perform one of the distracting tasks while simultaneously trying to recall aloud the studied word list  participants also performed a full-attention (FA) condition, in which the distracting task ended prior to free recall Results - distracting tasks that required animacy or syllable decisions to visually presented words, w/o memory load, produced large interference on free recall performance - when study list of words auditorally, engage in phonological processing. IS is this or the meaning that’s most important? - 2 experiment w/ DT requiring phonemic decisions about visually presented nonsense words OR pics and make a semantic judgement. Process words at study semantically, then 2 pic test should interfere w/ memory more. If process phonemically, then non-sense word task should interfere more w/ retrieval of study words  larger interference effect than 1 that required semantic decisions about pics  more effect for non-sense words, DA affected free recall more for non-sense words (phonemic DT) BUT not for semantic condition - loss of info when the task required them to make an animacy or syllable decision during the secondary task. Ability to remembered studied words that were auditorally presented declined w/ DA task (even if task not a memory task) - suggests when DA, deplete resources needed for retrieval, retrieval can be effortful (if all resource not available, will suffer Conclusion - free recall disrupted by competition for phonological or word-form representations during retrieval - equally resource-demanding picture- based distracting task produced significant interference w/ memory retrieval, effect was significantly smaller in magnitude - although smaller, decline in memory from picture task suggests that competition for semantic representations or for general resources can also disrupt retrieval  possible that any distracting task that is resource demanding disrupts retrieval somewhat, due to added complexity of coordinating two tasks - DA at retrieval less detrimental to memory performance than DA at encoding - conclude:  DA affects ability to retrieve, even if secondary task doesn’t demand memory. Especially when task is phonemic in nature  suggests its material-specific  Even thought this is the best condition: DA where secondary task is a semantic judgement, performance is still worse compared to full attention. Maybe we need some general resource (not material-specific) for intact successful retrieval  Possible any distracting task that places demands on PFC, disrupts retrieval due to added complexity of co-ordinating 2 tasks - need more processing at encoding than at retrieval - internally generated cues: info not readily available to help you retrieve, need to generate your own cues/strategies - need a set of neural strategies to do this - need to attempt retrieval, monitor and verify Retrieval Attempt - lateral Posterior PFC (BA 44/46) involved in goal-directed attempt to reconstruct past (same region as involved in WM). No useful cues, need more effort to remember the past (re-construct: problem solve, hold info online).  involvement when effort needed in retrieval attempt (e.g., shallow encoding)  independent of retrieval success, more concerned w/ the strategies you are attempting  material-dependent: L = verbal, R = visuo-spatial  elaborate processing of material beyond retrieval (ex/ reminds you of another memory … it’s still active!) - anterior frontal pole (BA 10 – especially right)  selective to remembering  role in ongoing monitoring (responds for long duration)  material-independent (verbal vs. non-verbal)  involved in retrieval success (alerting system indicating that something is from the past – old vs. new). Activated when successful in retrieving the info; subjectively feel recollection! - NEURAL evidence that there are separate process: during retrieval involved in attempting to retrieve/build up strategies and others involved in monitoring content of retrieval/indicating success Retrieval Content - content stored in regions during encoding  reactivation of regions involved in initial perception of info (modality- specific) - overlap between regions activated during perception and retrieval of same info not complete (i.e., cortical reinstatement = re-activate regions involved at the initial experience of the event) - at study presented w/ word dog, association w/ a picture of the word dog. At test, person sees the word dog, and have to retrieve the pic associated w/ it (the one they saw at study). Regions activated: one’s involved in initial encoding of the dog picture - see re-activation of regions involved in initial encoding of dog word and picture, less extent but same area activation at retrieval - different pathways involved in remembering/re-collect (re-experience event and all the details), familiar (feeling they know, can’t re-create context and all details) Recollection versus Familiarity Recollection versus Familiarity - specific, detailed (re-experience an episode include context) versus vague feeling of “pastness” (ex/ know you know the name) - controlled (recollection, processing demands) versus automatic - episodic memory (past personal events) gives rise to recollection (process by which we remember and re- experience an event) whereas semantic gives rise to familiarity (feeling of knowing w/ specific spatial/temporal context) Remember/Know Paradigm - “remember” response if specific instance of study can be called to mind - “know” response = item known to have been studied, no specific memory formed Remember/Know Procedure (Eldridge et al.) - Blocked design fMRI = trials presented in close succession and averaged together - Event-related fMRI = able to separate individual trials/event (takes advantage of fast signal acquisition properties of MRI). Look at activation of each trial! SO can look at regions involved/activated predicted successful retrieval (at retrieval) - study list of words on a screen  at test (in scanner): decide if they studied the word of not. If they said old (studied) then had to decide if they remembered studying the word or just know the word OR know (familiar, not sure they studied) - during remember/know scanned. Results: MTL (hippocampus) activated during remember responses more than know or familiarity. Activity at encoding predicted future recollection. Recollection versus Familiarity - Familiarity judgement is faster - Operate in parallel; Independent, can operate separately - Familiarity is strength-based (can rate confidence level); Recollection is a threshold (Y or N, context) - Familiarity = more automatic; Recollection = controlled and effortful MTL and recognition Memory Controversy - Unitary declarative memory hypothesis  Medial Temporal Lobe memory system = all the same regions involved in familiarity (weaker traces of recollection, simpler) and re-collection. - ** Functional dissociation (dual process) hypothesis  Extended hippocampal system = system devoted to recollection (MTL and fornix/thalamus etc…) separate from system of regions in familiarity Recognition: “Extended Hippocampal System” - Two processes: recollection & familiarity = (1) Recall: recollection (context of study) (2) Recognition: recollection or familiarity - Extended Hippocampal System for recollection (damage  impaired recollection). Extra-hippocampal MTL structures support familiarity, bypass hippocampus to PFC Decomposing Recognition - given a study list, then a recognition test: Indicate old/new. If old  remember or familiar. Some words recognize and recollect studying, other words may be familiar. - false alarms versus hits. Familiarity related : curved line  degree of familiarity (gradient for words) versus recollection: all or nothing  slope of line - recall is mostly recollections. Hippocampal damaged: if recollection impaired, so will recall ALSO likely recognition memory (recollection and familiarity) will be slightly better (less remember, more know) Functional Dissociation within MTL: Patient Evidence - Better recognition > recall = Patient YR, Hypoxia. Suggests recollection impaired - More severe deficits for R than K = Hypoxia vs. extensive MTL - Delayed Non-Match-to-Sample research in monkey and rat = (1) H Lesions mild deficit for recognition (2) Perirhinal lesions less severe than hippocampal damage Confabulation - Most often drawn from actual experiences - can include non-personal info historical facts, other aspects of semantic memory - Accounts/memories aren’t usually internally or externally consistent - Patients are not concerned or aware of the distortions - no purpose; secondarily may arise to explain internal inconsistency Explanations - Compensation for lack of episodic details available - Temporal disorder (don’t know temporal context of events), lose source memory - Deficit in retrieval, specific strategies (vs. associative or automatic cue-driven retrieval) . Combined w/ impaired ability to stop responses and/or monitor the results of those searched! Confabulation: Temporality Account - increased tendency to confuse the time at which info was last encountered; can usually be traced back to fragments of actual experiences - possible explanations = (1) failure to rep new info w/ normal saliency in memory, so established info intrudes into ongoing/recent thinking (2) fail to suppress activated memory traces and mental associations in the face of current reality Continuous recognition task - similar to PS task: failure to PC is fail to PS? Interference … similar memories and can’t be made distinct (details lead into each other) - run 1: series of items w/ some items repeated, pax asked to indicate recurrences - run 2 (~1 h.): same pics, randomized order w/ diff items selected as target items (recurring items). before each run, subjects asked to forget that they had seen pictures before and indicate recurrences only within current run - result: patients produced increasingly more false +ve responses from run 1 to run 2 - what result would be associated w/ explanation 1 (fail to strongly rep new info)? Semantic Memory Semantic Organization - categorization of world isn’t an arbitrary historical accident, reflects psychological makeup, is subject to investigation - categorize world and knowledge of world/ourselves, organized specifically, subjective (depends on how we/others experienced the world, how we share info). There are common elements that form basis of certain cognitive theories - linguistic based on computational model - Berlin and Kay (1969) investigated colour names across 100 different languages - order/frequency of colours used is consistent across cultures: B/W, R, G/Y, B … - e.g., 2 words for colours they tend to be B/W; 3 B/W/R etc… - Rosch-Heider (1972) experiment w/ American and Dani (only had words for B/W) - based on study: Zuni and English remember focal colours (rep of all colours w/in that spectrum, exemplar e.g., pure green) better than non-focal (more variability, not as great of an exemplar e.g., purple) - Experiment 1 = showed a single coloured chip, required to recognize it from a set of 160 chips learnt; both US and Dani subjects perf better w/ focal colours - Experiment 2 = Dani associate colours w/ clan names, better w/ focal colours - Results = same colours were focal for Dani as for US subjects; concluded = not language that makes certain colours easier to remember, rather perceptual salience Semantic Memory: Definitions - general world knowledge; language, conceptual knowledge (relate concepts/ideas)  shared among people; interface b/w language, perception, & cognition - concept: mental rep that determines how things are related/categorized (specifies category membership). All concepts have an underlying word, but there aren’t labels for every concept (ex/ concept brow dog, but no label) - lexical representation: word label for a concept, ex/ TOT: can’t retrieve lexical representation even if you have the concept - cognitive economy: semantic memory organized to avoid excessive duplication (cognitive representations shouldn’t be overly redundant) - basic/default level of categorization: used in communication to describe concepts Investigation Methods - (1) Lexical Decision: concept label is a word/non-word (2) Category Decision: item fit it a category? Underlying concept (3) Property Decision: item has a certain property? (4) Similarity Judgment: similarity on a scale - Models attempt to account for speed (RT) and accuracy of decisions  may give a sense of how concepts are associated w/ one another (Faster decision = likely info represented in a closer semantic space) Semantic Networks - Semantic network: collection of interrelated concepts and their pathways in semantic memory - Node: Point/location in semantic space corresponds to a concept or simple fact about a concept (connected to one another). Semantic space = distance b/w nodes - Activation: node activity, ready to fire, mental activity to access info from network. Activation exceeds threshold, concept fires  activation spreads to related concepts What should models of semantic memory explain? - Basis for similarity judgments (how/why of similarity) - Hierarchical category structure - “Basic Level” in hierarchy of categories (pref for describing objects at 1 level: not too vague or specific) - Typicality effects within categories (robins vs. penguins) Collins & Quillian’s Hierarchical Model - implements cognitive economy (something stored, not repeated at a lower level) - nodes connected by links, specify rel b/w nodes. “ISA” link (lower level node: type of higher level node). Each node/concept can be described w/ defining properties; lower levels can have more specific properties (properties not repeated) - pax presented w/ sentence verification task (T/F): RT slow as you move down hierarchy. Difference b/w RT’s equivalent (i.e., linear relationship b/w nodes) - applies to properties of an item too! - pax begin at node in network that rep subject of sentence, travel through network until find necessary info. Travelling takes fixed time for each link; RT’s slower for info further away. Times tells us how closely concepts are associated w/ each other Difficulties with model (1) Conjoint Frequency of 2 nodes confounded w/ semantic distance. Ex/ 2 words may be closely associated AND tend to co-occur often (2) Prototypicality effect: an item considered a better exemplar of a concept than others (despite the same semantic distance) (3) Relatedness Effect: more related 2 things are, harder to disentangle Smith’s Feature Overlap Model - Knowledge stored as semantic features of concepts (some concepts have more overlapping features than others). Meaning of word encoded by decomposition to smaller units of meaning (i.e., features) rather than position in a network of meaning - Individual concepts are sets of features, and concepts similar if they share features (share more features = more similar) - unlike hierarchical model, boundary between concepts is more graded - two possible routes for making category decisions = (1) Fast Route: global comparison of features; is there overlap? (2) Slow Route: looking up necessary and defining features for a category. More feature overlap  decision takes longer Difficulties with model: (1) Some categories don’t have obvious defining features common to all members (2) Cognitive Economy: only non-redundant facts should be stored (do we have “bird” represented a separate time in each list even loosely associated with “bird”?) Schemas - remember new material in terms of existing knowledge structures called schemas (represent some aspect of the environment, our experience, or beliefs) - learning conceptualized: active process by which people attempted to make sense of what they had experienced based on previous learning (“effort after meaning”). Add to semantic store, but skew details of those memories - study effects of schemas on memory: investigate memory for a NA folk tale (structured but unfamiliar material). showed that pax tended omit material that was strange to them or to distort it in ways that fit their expectations (view, belief …) - this type of knowledge structure enables people to make sense of partially observed/described situations. Adaptive purpose, we may not experience all the details, so this is a form of filling in. - schemas have variables that are inferred, all encoded together Scripts - developed scripts that represent commonly experienced social events (shared knowledge). E.g., going to a restaurant, going to a bank, taking a bus - ex/ Restaurant script = (1) Props (2) Agents (3) Entry conditions (4) Results - scripts can have scenes, each scene has several steps Category-Specific Deficits - Can dissociate syntax from semantics, but are there dissociations w/in semantics? How is it organized? Unitary (1 system for all categories) or multiple conceptual systems? Hierarchical? Distributed? Localiz
More Less
Unlock Document


Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

Unlock Document
You're Reading a Preview

Unlock to view full version

Unlock Document

Log In


OR

Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


OR

By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.


Submit