Lecture 8 - Exceptional memory

Lecture 8 - Exceptional memory

1. John Dean: "the human tape recorder". Neisser 1982. Unknown to Dean (and most other people) at the time, recordings of all these meeting existed and were later compared to his sworn testimony so its accuracy could be assessed. Dean’s memory was not approaching “tape recorder” accurate. It is often claimed that gist was accurate, but “his memory for even the ‘gist’ of conversations was quite poor except where that gist had been rehearsed in advance or frequently repeated.” (from Neisser 1982). He recalled his role as more central than it really was (self referential memory). Nonetheless, while inaccurate on gist of specific conversations, his overall gist was accurate (i.e. who knew what, did what etc).

2. Luria 1968. S. (real name Shereshevskii) “there was no limit to the capacity of S’s memory or to the durability of the traces he retained”. Memory for word lists of 70 items Lasted 15-18 years. After only 3 min’s study, S learned a matrix of 50 digits perfectly. Formed mental images (incl. digit encoding). Extreme synesthesia (“what a crumbly yellow voice you have…”) Used method of loci (streets of Moscow)

3. Takahashi et al 2006. Hideaki Tomoyori (HT) set a world record reciting the first 40,000 digits of pi. Learned chunks of 10, associating the sound of a number with a word --> sentences --> visual image. Used cue words --> image --> sentence --> sequence. Uses retrieval practice effect (Landauer and Bjork, 1978) – retrieves information at successively longer intervals. Auditory digit span is 10 (control = 6.8), and visual digit span is 8 (control = 7.3) i.e. no general exceptional cognitive abilities.

4. The effect of strategies. Wilding and Valentine 1994

   • Good memorisers, 5 who claimed to use strategies and 3 who claimed to be naturals. Tasks amenable to existing strategies (e.g. tel. no.s, faces etc). Tasks less amenable to such strategies (snowflakes, position, order). Strategies resulted in better immediate memory but worse 1 week later. Those using strategies were better on tasks amenable to strategies than other tasks, but no task difference for natural memorisers.

5. The brains behind superior memory. Maguire, Valentine, Wilding and Kapur 2003

   • No difference in brain structure, e.g. grey matter volume. Differences seen in brain activation's when memorising/recalling consistent with claimed use of spatial strategies.
     • Ericsson 2003 - recent evidence from imaging brain activity during exceptional performance provides strong support for acquired nature of exceptional memory. 'the differential pattern of activation of brain regions during memorisation can be explained by strategy differences. Compelling evidence that ordinary people can dramatically improve their memory performance with appropriate strategies and practise. But, only some people persist with training and eventually reach these levels.

6. Can anyone develop exceptional memory? Ericsson, Chase and Faloon 1980

   • An undergraduate student (SF) with average intelligence and memory. Practiced memory span task for 1 hour/day, 3-5 days a week. Over 20 months (>230 hours) digit span increased from 7 to 79 digits. SF keen runner and coded groups of digits as running times e.g. “just under world record for the mile” etc. (using semantic memory) Recalled in groups of 5-6. Evidence that this is not increase in memory span (phonological loop) but use of strategies and association in LTM. SF asked for preceding or following group of digits: If part of same level 2 structure: RT= 4.4 secs, If part of different level 2 structure: RT = 10.1 secs. Conclusion: structure of memory is crucial to efficient retrieval
   • Ericsson (1988) proposed that there are three requirements to achieve very high memory skills: Meaningful encoding: the information should be processed meaningfully, relating it to the preexisting knowledge (e.g. SF). Retrieval structure: Cue should be stored with the information will aid later retrieval (encoding specificity), Practice: extensive practice so that he processes involved in encoding and retrieval function faster to produce automaticity. (i.e. unless you have naturally exceptional abilities, practice and effort will pay dividends!)
   • Higbee 1997 - investigated whether ordinary college students using phonetic mnemonic could duplicate two feats of S. Large n study where novices were provided with phonetic keywords along w matrices, nearly half of students successfully memorised each matrix with no errors. Experiment 2 was a small studyin which they were not given the key words with the matrices, 3/4 successfully memorised each matrix with perfect recall. Ordinary college students are capable of acquiring a level of skilled memory that rivals abilities that appear to be innate.

7. Flashbulb memories, Brown and Kulik 1977

   • Flashbulb memories concern unexpected but personally relevant events. “primary live quality” “Indiscriminately preserves the scene” “permanent store”
   • The "Now Print" theory. Livingston 1967
     • Theorized that flashbulb memory formation results from: Recognition of a high level of novelty or unexpectedness. A test for biological significance to the individual. If true, a “Now Print!” order is executed by the memory, recording not only the novel event itself but all recent cognitive activity. This results in an unusually detailed memory for the contents of consciousness at that moment. n.b. There was no empirical evidence for this view at the time it was proposed..
   • Operational definition of FM currently uses 4 “Canonical Questions”: Where were you when you first heard? Who else was present? How did you first hear the news? What were you doing when you first heard? A “flashbulb memory” is said to occur when a subject accurately recalls at least 3 of the 4 answers to the canonical questions. Proportion reporting FMs depends not only on event but personal relevance.
   • Accuracy
     • Theory predicts that FMs should be: accurate (“indiscriminately preserves the scene”) complete (“permanent store”)
       • Neisser and Harsh 1992: FM tested was the explosion of the shuttle Challenger, Jan. 25th,1986. Next morning, 106 participants completed a questionnaire based on Brown & Kulik’s canonical questions. Accuracy measured as test-retest consistency. 14% inconsistent with previous report. 9% now not remembered. Even material participants were very confident about were not always accurate.
       • Greenberg 2004 - on at least 3 occasions, george bush was asked how he heard the news of the 9/11 attacks. His answers contain substantial inconsistencies.
   • Strong version (accurate and complete) not supported by evidence. Weak version of FM hypothesis: Flashbulb memories are more accurate and complete than would be expected if they were produced by ordinary memory mechanisms.
   • Comparison of flashbulb and non-flashbulb memories Weaver 1993
     • On January 16, 1991, [American] students encoded an "ordinary event" as part of a demonstration on flashbulb memories and completed a questionnaire modeled after S. A. Christenson (1989). By coincidence, the bombing of Iraq began the same day. A similar questionnaire asking about memory for the bombing was given to students 2 days later. In April 1991 and January 1992, both questionnaires were again completed, allowing comparisons of the students' memories for the 2 events. Although few differences appeared in the accuracy of the 2 memories over the intervals, confidence levels for the memories surrounding the bombing were significantly higher. Furthermore, retention and confidence did not differ from April 1991 to January 1992. These findings argue against any special-mechanisms hypothesis in the genesis of flashbulb memories but suggest that 1 thing that makes flashbulb memories unique is the undue confidence with which these memories are held.
       • Although few differences appeared in the accuracy of the 2 memories over the intervals, confidence levels for the memories surrounding the bombing were significantly higher. Furthermore, retention and confidence did not differ from April 1991 to January 1992. These findings argue against any special-mechanisms hypothesis in the genesis of flashbulb memories but suggest that 1 thing that makes flashbulb memories unique is the undue confidence with which these memories are held.
   • What is special? events are distinctive, suprising and consequential and rehearsed.

8. Finkenauer et al 1997 - old idea of memory as individual possession - or is it social faculty? To explain FBM superiority in recall, it is assumed that certain characteristics of the discovery situation lead to strong association between the actual news and surrounding details of context. On one hand, FBM are individual as they consist of each persons memory for their personal discovery context. But also social - they involve collective shared memory for actual event and interpersonal rehearsal plays and important role in maintaining and consolidating FBM. Opens possibility that individual memory contents become social through interpersonal communication and collective remembering.