明日とロボット このページをアンテナに追加 RSSフィード


認知心理学9. Visual Images: 視覚イメージ


Remember Baddeley?

  • verbal knowledge
    • Knowledge expressed in language (CP 156)
  • spatial knowledge
    • Knowledge of spatial relations that may be stored as images (CP 156)
    • are stored in different format.

A bit of historical context

  • Wundt (1880s):
    • one of the three basic elements of consciousness.
  • Watson (1913)- three main issues: (behaviorist)
    1. Can you use verbal reports to study images?
    2. Why is it difficult for some people to form mental images?
    3. What is "the mind's eye"?
  • Imagery research returned with the cognitive revolution (1950s-60s)

dual-coding theory

    • A theory that memory is improved when items can be represented by both verbal and visual memory codes (CP 161)
  • Paivio (1969):
    • memory can be improved if it can be represented by both visual and verbal codes.
  • have separate system for handling image and concrete, abstract.
  • Concrete abstract dimension:
    • Extent to which a concept can be represented by a picture (CP 159)

  • Concrete concepts have high imagery potential (Ease with which a cencept can be imaged (CP 159))
  • association value:
    • refers to the number of associations one can relate to a concept.
    • The number of verbal associations generated for a concept (CP 159)
  • Which is better for learning?
    • concrete is much better for learning.

  • Pavio, Smythe, & Yuille (1968)
  • Ss learned 16 paired-associates.
  • 1/2 were high in imagery, 1/2 low in imagery.
  • First term, image value of stimulus, second image value of response.
  • Conditions:
    • (H-H) juggler-dress
    • (H-L) letter-effort
    • (L-H) duty-hotel
    • (L-L) quality-necessity

  • Results


  • Being able to create images from both words facilitates learning and recall. Can imagine a juggler in a dress.
  • Learners were not told to use visual imagery.
  • Why are images effective?
    • provide an additional memory code that is independent of the verbal code. They forget one without forgetting the other.

More evidence for dual codes

  • Picture condition:
    • Congruent trials (which is larger in real life?):


    • Incongruent trials (which is larger in real life?):


  • Word condition:
    • Congruent trials (which is larger in real life?):


    • Incongruent trials (which is larger in real life?):



  • DV: time to decide which is larger in the world
  • Pictures <(is faster than) Words overall
  • Congruence effect for pictures but not for words
  • イメージの場合は、現実通りのサイズの比較だった場合が一番早く、現実とは違うサイズの場合がその次、
  • 文字はどちらともあまり代わりはなく、一番時間がかかる。

  • result


Semantic vs. Visual Elaboration: ''Bower & Winzenz (1970)''

  • Presented 30 concrete noun word pairs.
  • Conditions:
    1. repetition group
    2. sentence reading group
    3. sentence generation group
    4. Imagery
  • Recall and Recognition Test

  • Results
    • Recognition Test: no difference for any of the group
    • Recall Test: have differences
      • Repetition = 5.6
      • Sentence reading = 8.2
      • Sentence generation = 11.5
      • Imagery = 13.1

Using visual images


The Key Word System

  • keyword
    • A concrete word that sounds like an abstract word so that it can be substituted for the abstract word in an interactive image (CP 163)
  • keyword method:
    • Use keywords to improve paired-associates learning.
    • A mnemonic strategy using keywords to improve paired-associates learning (CP 163)
    • Learn the association between the name and the key word
    • Form an image of the key word.
  • Appropriate keywords important:
    • should sound as much like the keyword as possible.
    • Be different from other key words.
    • Concrete=easy to form images.

Atkinson & Raugh (1975)

  • Acquiring Russian vocabulary
    • Ss learn 120 Russian words
    • Two conditions: keyword group and control

  • Keywords not shown to control group.
  • Form an interactive image liking the keyword to the English translation.
  • Or generate a sentence if an image could not be formed.

  • ロシア語{覚えるためのキーワード}意味
  • Delo {Jello} Affair
  • Linkor {Lincoln} Battleship
  • Durak {Two Rocks} Fool
  • Krovat {Cravat} Bed

  • Results for recall of translation
    • Keyword Group: 75%
    • Control Group: 46%
    • 1 reason for fascination of these results: Russian is hard to learn because pronunciation is different form English.


  • Are imagery processes different from perceptual processes?

The imagery debate

  • Are mental images really images?

  • Pylyshyn vs. Kosslyn


  • if see this pic, we can remember and create mental images.

  • Propositional Networks


  • this is what we have in memory to generate mental image.


  • Imagine a map of Arkansas
  • Form an image of a black dot going from Clarksville to Fort Smith. Raise your hand when you get there.
  • Now form an image of a black dot going from Clarksville to Chicago. Raise your hand when you get there.
  • 長い距離を想像する方が時間がかかる。

Evidence for imagery: ''Kosslyn, Ball, & Reiser (1978)''


  • we scan mental image in the same way we scan real image.
  • visual buffer
    • A component of Kosslyn's model in which a generated visual image is maintained in STM (CP 174)

Evidence for imagery: Image Scanning Experiments


  • memorize and create mental images.
  • パーツからパーツまでの距離に比例して時間も増加

More evidence distinguishing verbal and spatial information

  • Sequential vs. parallel processing
    • sequential processing:
      • only one item at a time can be processed (verbal information)
      • Representation of knowledge in which only one item at a time can be processed (CP 167)
    • parallel processing:
    • more than one item at a time can be processed (images?)
    • Representation of knowledge in which more than one item at a time can be processed (CP 167)

Smith & Nielsen (1973)

  • Line drawings of a face or verbal descriptions of a face
  • Test face - is it the same one?
    • Varied number of relevant features (3, 4, 5)


  • show different faces. see line drawing or hear description.

  • result


  • picture: number of the feature does not influence to the processing time
  • description: number of the feature does influence the processing time.

Manipulating mental images

  • Are mental images really like real world percepts?
  • Cooper & Sheppard (1973)


    • IV: degree of rotation.
    • Task: is image normal or mirror.
    • DV: Time to make decision.


  • たくさん回転している方が、理解するのに時間がかかる。

Interference studies- Brooks, 1968


  • I.V. - type of response required (verbal vs. spatial)
  • D.V. - time to respond
  • Visual task- letter tracing
  • 角をたどっていって、凸の角の時はYes、凹の角の時はNo

Spacial taskだからyes/noを言うためのverbal を処理する部分は影響を受けない。

Interference studies

  • Verbal task- repeat a sentence and indicate ‘yes’ if the word is a noun
    • A bird in the hand is worth two in the bush.
  • Verbal response = saying the word (yes or no)
  • Spatial response = point to Y or N


  • Results
    • Spatial responses interfered with letter tracing, verbal response interfered with noun identification, but not vice-versa.
  • Thus, verbal responses interfere with verbal material.
  • Spatial Responses interfere with visual material.

Evidence from neuroscience

  • event-related potential (ERPs)
    • A diagnostic technique that uses electrodes placed on the scalp to measure the duration of brain waves during mental tasks (CP 173)
  • Cerebral blood flow
    • Measurement of blood flow to localize where cognitive operations occur in the brain (CP 173)
  • Both find activation of visual processing areas (occipital lobe) during imaging.

Brain damaged patients


  • どれだけ近くにいるかわからない。

More relations between perception and images


Dissociation between imagery and perception

  • R.M.(患者の名前)- could recognize objects and copy drawings, but could not imagine images or draw from memory.
  • C.K.- couldn’t recognize objects but could draw from memory (and later didn’t recognize his drawings**).
  • Therefore, imagery doesn’t exclusively rely on perceptual system.

In every case I have looked at, hypothesizing pictures or depictions does not provide any explanatory advantage over what I will call the “null hypothesis” that image content is represented as symbolic expressions, even though it may feel more comfortable because it comports with one’s subjective impression of what is going on. I, for one, get very nervous when I find a theory in psychology making claims that are consonant with how it looks to me from the inside - I know of too many examples where how it feels on the inside is exactly the wrong kind of theory to have.
(Zenon Pylyshyn)

Pylyshyn’s ideas

  • Imagery is an epiphenomenon
    • Mental phenomena can only be caused by physical phenomena, not by other mental phenomena.
    • like the hard drive light on a computer

  • Tacit-knowledge explanation
    • Performance on imagery tasks depend on Ss beliefs or knowledge of the task

Pylyshyn’s interpretation of the imagery data

  • Mental scanning
    • task demands
  • Interference
    • only shows common representational components, not -necessarily pictures
  • Rotation studies
    • unusual (yet memorized) shapes are hard to rotate

Pylyshyn’s arguments against images

  • Also- imagine a familiar word.
  • Mentally scan that image and spell it backwards
  • Blind individuals show similar results in mental scanning studies.

More shortcomings of mental imagery

  • If something is seen often, we should have a good mental image of that object.

Which one is like a real penny?


  • maybe mental image is not the same to the visual image

More shortcomings of mental imagery

  • Is Reno east or west of Los Angeles?
  • people tend to think its on the east, but
  • WEST**


Why it’s good imagery ≠ perception: ''Johnson, Raye, Wang, & Taylor (1979)''

  • Reality monitoring
    • Discriminating between actual and imagined events (CP 178)
  • Shown pictures and names of concrete nouns 2, 5, or 8 times (Ps imagined objects when a word appeared)
  • At end, asked how many times they saw the picture

Reality monitoring cues

  • Sensory information
    • perceptual events have more detail than imagined events
  • Context
    • Perceptual events occur in a external context that contain more information
  • Cognitive operations
    • If images were created without conscious awareness, we would have poor memory for the cognitive operations use to generate them.

Well so what.._

  • Does visual imagery have an application outside of education

part 3

Theoretical Function of Visual Images: ''Schnotz & Bannert (2003)''

  • The integration of text and Graphics.
  • Descriptive system:
    1. Includes the external text.
    2. An internal representation of the texts surface structure.
    3. And a propositional representation of the text content.

  • Depictive:
    1. The external graphic representation.
    2. A perceptual representation of the graphic image.
    3. And a mental representation of the image’s semantic content.

textとimageのそれぞれが、text surface structureとvisual imageを作り、verbal/visual organizationを作る。
そして、propositional representation/mental modelができる。そして、propositional representationとvisual modelはお互いにinfluenceしあう。

Type of Graphics: ''Carney & Levin (2002)''

  • what types of graphics are actually involved.
  • Representational


  1. Organizational: help organize information with some types of meaning.


  1. Interpretational: help make abstract, hard concept to easier to understand.


  1. Transformational: act as neumonic


  1. Decorational: seem to make up most part of text book.