Chapter 7: Representation and Manipulation of Knowledge in Memory: Images and Propositions

1. Chapter 7: Representation and Manipulation of Knowledge in Memory: Images and Propositions

2. Cognitive Psychologists Study Mental Representations • Methods • Self report • Rationalist approach • Empirical support • Experiments • Neuropsychological studies

3. Mental Imagery • Internal representation of items that are not currently being sensed • May be old, new, futuristic, imaginary • May involve any of the sensory modalities • Imagine a taste, a sight, a touch • Majority of research on Visual Imagery

4. Mental Imagery • Kosslyn proposes images are used to help solve certain types of problems • How many chairs are there in your house? • Do bunnies have whiskers?

5. Dual Code Theory • Paivio (1971) • We use two codes to represent information • Image codes- analogue codes, shares some perceptual features • Symbolic codes- arbitrary symbols to represent items (e.g., words) • Two codes are linked

6. Evidence for Dual Code Theory • Paivio compared concrete words (potato, horse) with abstract words (justice, love) • Found participants were better able to recall concrete words • Concluded that dual code was created for concrete words (analog & verbal label) but not for abstract words

7. Visual Codes are Processed Differently than Symbolic Codes • Visual information interferes with spatial information • Verbal labels interfere with spoken words • Sequence matters more for words, not so much for unrelated images • Thus, each type of code is affected by different manipulations

8. Evidence for Dual Code Theory • Brooks (1968) • One group saw a block diagram of a letter • Memorized it • Were asked to mentally travel the letter and indicate if the corner was on the extreme top or bottom Start

9. Evidence for Dual Code Theory • Brooks (1968) cont. • Second group saw a sentence • Memorized it • Were asked to classify each word as a noun by indicating "yes" or "no" • Verbal task A bird in the hand is not in the bush

10. Evidence for Dual Code Theory • Brooks (1968) cont. • Participants were then asked to respond in one of two ways • Say “Yes” or “No” • Point to the answer “Yes or No” • Why was this important? Yes No No Yes Yes No Yes No Yes No No Yes

11. Evidence for Dual Code Theory • Brooks (1968) Results For image task, RT was slower when pointing. For the symbolic task, RT was slower for the verbal response. Different pattern = different processing for different codes

12. Evidence for Dual Code Theory • te Linde (1982) • Participants answered questions about word or picture pairs

13. te Linde (1982) Results

14. Propositional Theory • Do not store in form of images • Instead have a “generic” code that is called “propositional” • Stores the meaning of the concept • Create a verbal or visual code by transforming the propositional code

15. Propositional Representations

16. Test Your Visual Imagery Ability! • Form a mental image of this picture • Which of the pictures on the next slide are part of this picture?

17. 1 2 3 4

18. Try Again with Another Design • Form a mental image of this picture • Which of the pictures on the next slide are part of this picture?

19. 1 2 3 4

20. Carmichael, Hogan, & Walters (1932) • Participants were shown simple figures with one of two verbal labels Hourglass Or Table Sun or ship’s wheel

21. Carmichael, Hogan, & Walters (1932) Results

22. Carmichael, Hogan, & Walters (1932) Results • Later participants were asked to draw items seen • Participants distorted the images to fit the labels • This pattern supports the idea that images may be stored propositionally not as original analog image

23. Analogical Limitations • Inability to see parts has led some to support a propositional code rather than an analogical code • Demonstrates mental images are not always precise

24. Finke, Pinker & Farah (1989) • Imagine a capital letter D. Rotate the figure 90 degrees to the left. Now place a capital letter J at the bottom • Imagine a capital letter N. Connect a diagonal line from the top right corner to the bottom left corner. Now rotate the figure 90 degrees to the right • Demonstrates that participants could manipulate images

25. Mental Imagery • Shepard & Metzler (1971) • Subjects had to decide whether displays had two similar shapes • Some pairs were similar, but rotated to various degrees

26. Shepard & Metzler (1971) Results

27. Mental Imagery Studies Demonstrate • Active process • Response times are proportional to degree of rotation • People can rotate images in three-dimensional space as easily as two-dimensional space • Images are “Mental Sculptures”

28. Functional-Equivalency Hypothesis • First proposed by Shepard and Kosslyn • Mental images are internal representations that operate in a way that is analogous to the functioning of the perception of physical objects

29. Neuropsychological Evidence & Mental Imagery • Cohen & Kosslyn • Same brain areas are involved in perception and mental rotation • Support for functional-equivalence hypothesis

30. Functional-Equivalence Evidence • Kosslyn (1975) • Examine how participants scan and use images • Some participants imagine an elephant next to a rabbit • Others imagine a rabbit next to a fly • Then answer questions about the rabbit • Does the rabbit have whiskers? • Does the rabbit have ears? • Does the rabbit have a beak? • Reaction time to answer is measured

31. Kosslyn (1976) • Asked college students and fourth graders simple questions about animals • Does a cat have claws? • Does a cat have a head? • Varied the type of instructions used to answer questions • Imagery instructions • No imagery

32. Kosslyn (1976) Results • In imagery condition, questions were answered faster if the attribute was larger • In no imagery condition, questions were answered faster based on distinctiveness of characteristic for the animal, no impact of size

33. Size Judgments (Moyer, 1973) • Which is larger, moose or roach? • Which is larger, wolf or lion? • When objects are similar in size, participants imagine both objects and then compare the size of the objects in their image • Similar results when making comparisons of actual physical objects • The closer in size, the longer the reaction time

34. Functional-Equivalence Evidence • Kosslyn (1983) • Memorize map • Later ask to scan image • Manipulate distance between items in scan • Hut to grasses • Lake to Hut • Measure reaction time

35. Kosslyn (1983) Results • Linear relationship between the distance to scan and actual reaction time of participants • Further support for functional-equivalence hypothesis • Mental images are internal representations that operate in a way that is analogous to the functioning of the perception of physical objects

36. Finke (1989) on Functional Equivalence • We use similar transformation on objects and mental images • Spatial arrangements of a mental image are similar to spatial arrangements of actual object • Images can be used to generate information not explicitly stored during encoding • Processes of visual system are used on both mental images and visual objects

37. Demand Characteristics • Major criticisms of Kosslyn’s Research • Pylyshyn • There is only one code, propositional • The results due to task demands • The instructions imply some necessary relationship between the physical distance and time required

38. Demand Characteristics & Mental Scanning • Participants give the experimenters the pattern they expect • Intons-Peterson replicated research but mislead experimenters • If experimenter expectations are part of demand characteristics, then leading them to believe that longer distances would lead to faster responding should alter the results • Evidence was found to support demand characteristics idea

39. Demand Characteristics & Mental Scanning • Jolicoeur & Kosslyn (1985) • Created a false demand characteristic for a U shaped function for participants • Proposed that Gestalt principle of proximity makes close points “hard”, and distant points would also take longer • No experimental expectancy effect found • Supports idea that image is being used

40. Johnson-Laird (1983) • Proposed there are three types of mental representations • Propositional representations which are pieces of information resembling natural language • Mental models which are structural analogies of the world • Mental imagery which are perceptual models from a particular point of view

41. Characteristics of a Mental Model • A representation of a described situation rather than a representation of a text itself or the propositions conveyed by a text • The structure corresponds to the functional relations among entities as they would exist in the world • A simulation of events in the world, either real or imaginary

42. Evidence for Mental Models • Kerr (1983) • Studied participants who were blind • Created a tactile Kosslyn Map study equivalent • Participants had to study the island, given a physical map to touch • Asked the same scanning questions • Found the same pattern of results—longer distances, longer reaction times

43. Visual Imagery & Spatial Imagery • Visual Imagery (images are visual) • Seeing colors • Comparing shapes • Spatial Imagery (analog spatial format) • Rotating objects • Aiming and shooting at a target

44. Neuroscience Evidence • Farah (1988) • Brain Injury Case Study (L.H.) • Gave some visual tasks • Color identification, object naming • Gave some imagery tasks • Mental rotation, mental scanning

45. Farah (1988) Results • L.H. • Poor visual image skill • Normal spatial image skill • Thus, both types of imagery must exist

46. In Sum, Researchers Have Proposed • Evidence for analog codes • Evidence for propositional codes • Evidence for mental models • Evidence for a mental imagery that is spatial • Evidence for mental imagery that is visual

47. Cognitive Maps: Historically • Tolman – Rats • von Frisch – Bees • Thorndyke – Humans

48. Creating Cognitive Maps • Gain increased spatial knowledge • Using three types of knowledge • Landmark (special buildings) • Route-road (procedures to get to one place from another) • Survey (global map-like view)

49. Tversky (1993) • Cognitive maps more like cognitive collages • Constructionist view of creating cognitive maps • Distortions can occur when using heuristics

50. Heuristics Affecting Cognitive Maps • Rotation heuristic • Tend to ‘regularize’ tilted landmarks in maps to appropriate E-W or N-S axis • Alignment heuristic • Students view two maps of the Americas • One a correct map, and a second map which was altered (South America was moved westward with respect to North America) • A majority of students thought the altered map was the correct one