Chapter 7 Flashcards
Describe Chassy and Gobet study of chess players proving that we have more info in semantic memory in those areas of special interest /importance to us
Analyzed over 70,000 games played by chess players of varying skill levels.
A very strong relationship between chess-playing skill + knowledge of opening moves
What are the differences between the subjective experiences associated with episodic + semantic memory?
Retrieval of info from episodic memory–>typically (but not always) accompanied by a sense of consciously recollecting the past.
No such sense of conscious recollection during retrieval from semantic memory.
Episodic memory involves “self-knowing” vs. semantic memory involves “knowing awareness”
Episodic memory–>a recently evolved, late-developing, + early-deteriorating past-oriented system, more vulnerable than other memory systems to neuronal dysfunction
What is Tulving’s view of semantic memory + episodic memory being distinct?
The adverse effects of brain damage on memory–>less for semantic memory than episodic memory.
147 cases of amnesia–>Episodic memory was impaired in all cases + only modest problems with semantic memory.
The extent of retrograde amnesia for episodic memories in amnesic patients–>often spans several years vs for semantic memories–>generally reasonably intact except for knowledge acquired shortly before the onset of amnesia
Define semantic dementia
Pattern of severe problems with semantic memory but relatively intact episodic memory.
Severe loss of concept knowledge from semantic memory even though their episodic memory + most cognitive functions are reasonably intact.
Associated with damage to the anterior frontal-temporal lobes as opposed to the medial temporal lobe.
Find it very hard to access info about most concepts stored in semantic memory
Describe a series of studies proving
Episodic + semantic memory often combine in an interdependent fashion in their functioning
Compared patterns of brain activation during episodic, semantic + autobiographical memory retrieval.
The same neural network including various frontal, temporal + parietal areas was activated during the retrieval of all these types of memory.
Amnesic patients + controls were instructed to learn the prices of grocery items.
The prices of some items corresponded to participants’ prior knowledge (congruent items), whereas others were incongruent.
Those with fairly intact semantic memory–>better memory performance for congruent grocery prices than incongruent ones.
Amnesiac patients with poor semantic memory–>showed no congruency effect.
There is a nonsignificant difference between semantic dementia patients + healthy controls in their recall of recent ABG memories but worse than controls patients when recalling remote ABG memories.
Sensory + 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 to facilitate the retrieval of episodic information.
Describe Loftus et al.’s studies on how categories vs. first letters as cues help recall
Elizabeth Loftus
The task of coming up with particular words, given a category + the first letter as cues.
Giving the category first and initial letter afterwards (e.g. fruit–p )–>faster responses than giving the initial letter before the category (e.g. p–fruit ).
Suggests it is easier to activate the category fruit in preparation for searching for the appropriate initial letter than all words starting with, say, p .
Because the category fruit is coherent/manageable, whereas words starting with p form too large a category to be useful.
Study in which the category was a type of psychologist + the initial letter that of the psychologist’s surname.
“Give me a developmental psychologist whose name begins with P” (Piaget) vs. “Initial letter P–a developmental psychologist.”
Novice students–>no difference between the 2 orders of presentation
Specialized students–>faster when the category was provided first b/c they’d developed categories such as “developmental psychologist,” while novices simply searched all “psychologists,” (underdeveloped categories)
Describe Collins and Quillian’s Hierarchical network model of semantic memory
- Semantic memory is organized into a series of hierarchical networks
- The major concepts (e.g. animal, bird, canary) are represented as nodes
- Properties or features (e.g. has wings, is yellow) are associated with each concept.
-It would waste space in semantic memory to have info about being able to fly stored with every bird name.
If those properties possessed by nearly all birds (e.g. can fly, has wings) are stored only at the bird node or concept–>cognitive economy.
- Property info is stored as high up the hierarchy as possible to minimize the amount of info needing to be stored in semantic memory.
- Tested this model using a task on which participants decided true/false sentences
- Possible to say true to “A canary is yellow,” b/c the concept (i.e. canary) + the property (i.e. is yellow) are stored together at the same level of the hierarchy
- In contrast, the sentence, “A canary can fly,” takes longer b/c the concept + property are separated by one level in the hierarchy.
-The sentence, “A canary has skin,” takes even longer b/c 2 levels separating the concept + property.
The more separation between the subject + the property –>longer response time to true sentences
We often use semantic memory successfully by inferring the right answer (e.g Davinci has knees is true b/c he is a human being, not b/c we have the info in our semantic memory)
Describe the limitations of the hierarchy model
A sentence such as, “A canary is yellow,” (common) differs from, “A canary has skin,” (uncommon) not only in the hierarchical distance between the concept + property but in familiarity.
When familiarity was controlled–>hierarchical distance between subject + property had little effect on verification time.
“A canary is a bird” and “A penguin is a bird.” should take the same verification time since both involve moving one level in the hierarchy.
However, takes longer for penguin than canary to decide a penguin is a bird than that a canary is a bird.
The members of most categories vary in terms of how typical/representative they are of the category–>faster verification time the more typical.
Typicality effect–>Less time taken to decide a category member belongs to a category for typical members than the atypical ones
Oranges, apples, bananas, peaches–>much more typical than were olives, tomatoes, coconuts, dates.
Robins Eagles–>much more typical than Ostriches + penguins
Collins + Quillian was mistaken in assuming that concepts belong to rigidly defined categories. Most categories are loosely determined.
According to Ludwig–>members of a category = members of a family (certain characteristics are shared). Some members share more, some less + often not the same one or two.
Describe McCloskey and Glucksberg study that proves many concepts in semantic memory are fuzzy rather than neat and tidy
30 people tricky questions such as, “Is a stroke a disease?” and “Is a pumpkin a fruit?”
They found mixed results, not everyone agreed.
They tested the same participants a month later some people had changed their original answer.
Two reasons why there are individual differences in deciding which items belong to a given category:
1) There is ambiguity–> use different criteria for categorization
2) There is vagueness–>use different cut-offs to separate members from non-members.
Describe the Spreading activation model by Collins and Loftus
Designed to resolve problems with Collins and Quillian’s model.
logically organized hierarchies are too inflexible–>better to assume semantic memory is organized on the basis of semantic relatedness or semantic distance.
Semantic relatedness–>measured by asking people to decide how closely related pairs of words are or list as many members of a particular category (the ones produced most often= closely related to the category)
The length of the links between 2 concepts–>degree of semantic relatedness (e.g red closer to orange than sunsets)
Whenever a person thinks about a concept–>the appropriate node in semantic memory is activated–> activation spreads most strongly to concepts that are closely related semantically
(activation passed rapidly from robin to bird b/c of semantic relatedness but not for penguin)
Why is Spreading activation of central importance in Dell’s theory of speech production?
When we plan an utterance–> several of the sounds in the intended sentence become activated even before we start to speak.
Speech errors occur whenever an incorrect word is more activated than the correct one.
Describe the limitations of the spreading activation model
Generally proved more successful than the hierarchical network model b/c it is much more flexible.
The model typically does not make very precise predictions–>difficult to assess its overall adequacy.
1) The notion that each concept in semantic memory is represented by a single node is oversimplified (info about concepts is more spread out)
2) It implies that each concept has a single, fixed representation.
However, our processing of any given concept is flexible. (e.g difficult to lift the piano vs enjoyed playing the piano: our word processing focuses on heaviness in 2nd one, not the 1st)
Describe the concept of hierarchical categories with 3 levels
Concepts are organized into hierarchies + there are three levels within such hierarchies.
There are superordinate categories (e.g. item of furniture) at the top, basic-level categories (e.g. chair) at the intermediate level + subordinate categories (e.g. easy chair) at the bottom.
We do sometimes use superordinate categories (e.g. “That furniture is expensive”) or subordinate categories (e.g. “I love my new iPhone”).
However, there is generally a strong preference for using basic-level categories.
Participants named pictured objects.
-Basic-level categories were used 1595 times during the course of the experiment, subordinate names 14 times, and superordinate names only once.
-The preference for basic-level was b/c it provides the best balance between informativeness + distinctiveness.
Informativeness–>is lacking at the superordinate level (e.g. simply knowing an object is an item of furniture tells you little.
Distinctiveness –>is lacking at the lowest level (e.g. most types of chairs possess very similar attributes or features).
At basic level–>people use similar motor movements for interacting with category members. (Nearly all chairs can be sat on in the same way)
the basic level is the one usually acquired first by young children.
Who prefers basic-level categorization and who doesn’t?
Categorization is easiest at the basic level for 13 + 16-month infants and hardest at the superordinate level.
Study of birdwatchers + dog experts naming birds/dogs.
Both groups used subordinate names in their expert domain much more often than their novice domain.
Subordinate categories are much more informative for experts than for nonexperts.
Cultural factors are also important. s.
The Itza culture was more likely than the Americans to categorize plants, animals + birds at the subordinate level b/c of close contact with nature
Participants presented with a category label at the superordinate, basic, or subordinate level followed by the face of a familiar face (a celebrity) to decide whether the face matched the label.
Matching occurred faster at the subordinate level than basic.
Nearly everyone typically uses subordinate categories for faces.
Individual familiarity with objects at the subordinate level can produce very fast categorization (e.g Eifel tower rather than just tower)
Performance accuracy was greater for superordinate categorization than basic-level categorization when fast responding was required.
Participants were a few milliseconds faster to decide whether a scene contained an animal (superordinate level) than whether it contained a given species of animal (basic level)
Faster suordinate categorization b/c basic level is generally more informative–>requires more information processing.
Preference for categorizing at basic level =/= faster categorization speed
Patients with mild semantic dementia–>had accurate categorization at the basic + superordinate levels vs
severe semantic dementia–> better at the superordinate than the basic level because it required less information processing.
What can we conclude from concept hierarchy studies?
1) Categorization at all three levels within concept hierarchies is of value.
2) Categorization at the superordinate level is often the fastest because it requires less cognitive processing
3) Categorization at the basic level is often preferred because it combines informativeness + distinctiveness.
4) Categorization at the subordinate level is often preferred over basic level by those possessing relevant expertise (b/c former is more informative)
