THAT IS: HOW DO WE GO FROM AN INPUT, SUCH AS THE VISUAL CONFIGURATION "A" TO CATEGORIZING IT AS THE LETTER IT IS (AND THUS ACCESSING RELATED INFORMATION, SUCH AS THAT IT IS A VOWEL, THE FIRST LETTER OF THE ALPHABET AND SO ON)

SOME HISTORY:
FROM DISTAL (THE STIMULUS IN THE "REAL WORLD") TO THE PROXIMAL STIMULUS (THE STIMULUS AS IT IS REFLECTED ON THE SENSORY RECEPTOR) TO PERCEPTION (ACCESS TO MEANING)

NOT A 1:1 MAPPING FROM THE INFORMATION AVAILABLE PROXIMALLY TO OUR PERCEPTION (OR INTERPRETATION) OF IT

SOME EXAMPLES:

AMBIGUOUS FIGURES

GESTALT PSYCHOLOGY:
-THE PHI PHENOMENON (NOTE A REACTION TO THE "COPY" THEORIES OF THE "EMPIRICIST-ASSOCIATIONISTS")

GESTALT LAWS OF PRAGNANZ: PROXIMAL AMBIGUITY IS NATURALLY RESOLVED BY LAWS OF ORGANIZATION-RULES THAT SUGGEST WHICH PARTS OF THE PROXIMAL STIMULUS BELONG TOGETHER
EG. PROXIMITY, SIMILARITY, COMMON FATE, GOOD CONTINUATION

BASED ON THE "POP-OUT" PHENOMENON

PROCEDURE: PARTICIPANT IS PRESENTED WITH A VISUAL DISPLAY. THE TASK IS TO RESPOND AS RAPIDLY AS POSSIBLE WHETHER OR NOT A GIVEN "OBJECT" IS PRESENT IN THE DISPLAY (MEASURE RT FROM ONSET OF DISPLAY TO RESPONSE)

IN ONE SUCH CASE THE "OBJECT" WOULD BE A GIVEN SIMPLE FEATURE (SUCH AS: A VERTICAL LINE OR A SPECIFIC COLOUR OR A GAP)

THE MANIPULATION, ON POSITIVE TRIALS (TRIALS IN WHICH THE TARGET IS ACTUALLY PRESENT IN THE DISPLAY) IS THE NUMBER OF DISTRACTERS PRESENT AT THE SAME TIME

EG., WHEN LOOKING FOR HORIZONTAL LINE, POSITIVE DISPLAYS WOULD INCLUDE
 

A.	¯	(NO DISTRACTER CONDITION
B.	\ ¯       \	(THREE DISTRACTERS)
C.	\       \ \    \  \    \    ¯	(SIX DISTRACTERS)

LOGIC. IF A FEATURE IS EXTRACTED AUTOMATICALLY (WITH A MINIMAL AMOUNT OF ATTENTION), THEN THE RT TO IDENTIFY THE TARGET SHOULD PROCEED IN PARALLEL; IF IT DEMANDS AN ATTENTION-DEMANDING PROCESS, THEN SEARCH SHOULD BE SERIAL IN NATURE

FIND: EVIDENCE FOR A PARALLEL SEARCH WHEN THE FEATURE IS BASIC (Question: what would the graph look like...remember the independent variable is the number of distracters and the dependent variable is the RT to correctly respond that YES the target was present)

ALSO NOTE ASYMMETRIES: THAT IS, A FEATURE MIGHT DEMONSTRATE THE PARALLEL FUNCTION (E.G, SEARCHING FOR A TILTED LINE IN A DISPLAY WITH VERTICAL DISTRACTERS) BUT NOT THE REVERSE (EG., SEARCHING FOR A VERTICAL LINE AGAINST A BACKGROUND OF TILTED LINES)

-THE CONJUNCTION OF SIMPLE FEATURES

-ILLUSIONARY FEATURES

-CAN ALSO STUDY USING THE SEARCH TASK THAT WE DISCUSSED FOR SIMPLE FEATURES

SAME TASK, BUT NOW ONE CAN JOIN SIMPLE FEATURES, SUCH AS ¯AND I|
EXAMPLE SEARCHING FOR THE LETTER T IN A BACKGROUND OF LETTER L DISTRACTERS
EG,
 

T

L T

L    L L   T   L     L  L   L

NOTE: NOW FIND EVIDENCE OF A SERIAL SEARCH; TREISMAN CONCLUDES THAT WHEREAS EXTRACTION OF A SIMPLE FEATURE MIGHT BE AUTOMATIC, THE "GLUEING" TOGETHER OF FEATURES TO PRODUCE COMPLEX OBJECTS IS ATTENTION DEMANDING

EXAMPLE: THE PHONEME RESTORATION EFFECT

TAKE A WORD ; ERASE A PHONEME (PHONEMES ARE THE BASIC SOUNDS OF A WORD, THEY ARE THE BUILDING BLOCKD OR THE MOST PRIMITIVE AUDITORY FEATURES OF A WORD)

TASK: PLAY THE MODIFIED WORD AND ASK THE PERSON TO IDENTIFY WHICH SOUND IS MISSING (OR IDENTIFY WHICH WORDS HAVE BEEN MODIFIED) AUDITORY FEATURE

FIND: PEOPLE CAN'T DO THIS TASK.. .IE. SOMEHOW THE PHONEME IS MENTALLY RESTORED (Task for you: how could a feature net model of the sort described on pp. 50-52 explain this)

*EEL (*=DELETED PHONEME)

*EEL ON THE CAR
*EEL ON THE ORANGE
*EEL ON THE SHOE