Upgrade to remove ads
Only $2.99/month

PSY 211 Exam #3

Terms in this set (52)

What is a form of visual imagery?
When picturing someone's face, or the amount of chairs that are in your kitchen table, or even how you would get from point A to point B, these tasks they formed what's called a "mental image"
What are mental images for?
"Thought is impossible without images." -Aristotle
First psychological studies on mental imagery
Lab of Wilhelm Wundt (introspectionist)
-3 basic elements of consciousness (thinking)
1. sensation; 2. emotion; 3. imagery
Lab of Wilhelm Wundt
-So, for Wundt and the introspectionists images were an important component of thought
-thus studying imaging was a way of studying thought
-had people form mental images and then had them report (introspect) about the content and nature of those images
Historically what is known about behaviorists and mental imagery
-Behaviorists: not fans of the study of mental imagery
-unproductive because mental images are invisible to all but the "imager" thus not observable and not quantifiable
-"Epiphenomenon" -- a secondary effect or byproduct that arises from, but does not causally influence a process
What was an important development for mental imagery in the 1960s?
The cognitive revolution in the 1960s made it acceptable (again) to study mental imagery
Shepard and Meltzer (1971)
Measured reaction time to make judgments like: Whether the two images of blocks are the same
-found that it took longer to decide if the objects were the same if they were rotated further away from each other
-reasoned that participants were "mentally rotating" one object to see if it matched the other object
-menta rotation task
Alan Paivio
Concrete vs. abstract nouns
-concrete nouns can be easily imaged (truck, chair, cake, bird...)
-abstract nouns hard to form an image (truth, research, feminism, version...)
What did Alan Paivio find?
-memory for words that evoke mental images is better than for words that do not
-for example, in a free recall memory task where participants are given a list of words to study and 50% are concrete words and 50% are abstract: more concrete items were recalled
What does the Dual coding theory believe?
Concrete words have dual representations:
-imagistic
-linguistic
Abstract words have only linguistic representations
What does the Imagery debate express?
-linguistic representations are "Propositional": symbols, language
-Depictive representations: similar to real world objects (think pictures in your mind) importantly - these representations should reflect the spatial relationships of the features
-the question is whether the second (depictive) mechanism exists
The imagery debate: Can we represent everything we need to with propositions?
-semantic networks are propositional, could be both propositional & depictive
-if imagery is depictive it SHOULD rely on the same mechanisms as perception
-in other words, if what we think of as "imagery" is really propositional it should NOT rely on the same mechanisms as perception
Showing that imagery is "spatial" implies that it is depictive in nature - that it relies on visual representations. Is imagery spatial or propositional?
Propositional according to Pylyshyn (1973)
What did Pylyshyn propose?
-proposed that imagery is propositional
--can be represented by abstract symbols (think language here)
-spatial representation is an epiphenomenon
--accompanies real mechanism but is not actually a part of it
The imagery debate: What did Pylyshyn suggest?
Pylyshyn suggests that all mental representations are propositional and that we experience imagery as a secondary effect
What did Steven Kosslyn believe?
-imagery is visual-spatial
-involves use of same systems as perception
-it's not propositional
Kosslyn boat experiment
-had people look at a picture of a boat and then reimagine it but looking specifically at the anchor and were then asked if the boat had and a port hole/motor
-reaction time was found to be a function of how away two parts of the image are (faster for porthole than motor)
-suggests that mental images are depictive and spatial
-but wait...
Is imagery spatial or propositional?
-Pylyshyn:
--depictive imagery does not necessarily have to be involved in tasks where people are asked to describe mental images
Is imagery spatial or propositional?: Martha Farah
-cognitive neuroscientist Martha Farah used priming to observe imagery and its effects on perception
-Farah starts with the observation that propositional representations are a given for certain types of knowledge (dual coding)
-she reasoned that if she could show that imagery and perception are related/share mechanisms
-then this would be evidence for the existence of depictive representations (visual codes) in addition to propositional representations (linguistic codes)
Farah, 1985
-subjects asked to image a T or an H - (auditory)
-two successive squares flashed on the screen. Subjects did not identify the letter but determined what square it was in: first or second
-when the imaged letter was a T and the target was a T subjects were better at the task than when the imagined letter was an L and the largest was a T
What was found in Farah's experiment?
-subjects were at an advantage when the letter that they were asked to imagine was also the one in the square
-but are the imaged letter and the target using the same underlying neutral mechanism? (do imagery & perception share brain systems?)
Farah, 1985: Where is this underlying hardware and when is it active?
Used ERPs to ask this question
Evidence for a shared representational medium for mental images and percepts
-recall that participants were at an advantage when the letter that they were asked to imagine was also the one in the square (Farah et al 1988)
-also, there was also more brain activation over occipital sites
Imagery & Perception
-the interpretation is that imagery and perception share mechanisms because they activate (rely on) the same neural hardware
-also, this activation occured early in processing (~150 ms)
Imagery & Perception: Farah
-for Farah the interpretation is that imagery and perception share mechanisms
-but if this is true, shouldn't we be able to find other brain imagining evidence for this interpretation?
Imagery and the Brain
-fMRI: (striate cortex - visual area)
-BOLD fMRI responds to both perception and imagery - though less to imagery
Kosslyn and coworkers (1999)
-TMS to visual area of brain during perception, imagery, and control task
-response time slower for both perception and imagery but not control
-suggests that brain activity in visual areas of the brain play a casual role for both perception and imagery
Neuropsychological case studies
Unilateral neglect
-patient ignores objects in one half of visual field in perception
-also in imagery
Where does this leave us?
-there appears to be overlap in the brain areas during perception and imagery
-which many cognitive psychologists have interpreted as evidence for dual coding theories
Artificial intelligence has advanced to the point where self-driving is coming...
-Waymo (Google) says that by 2020 - its fleet of self-driving Jaguars will be doing as many as a million trips per day
Why is Waymo's statement about his self-driving Jaguars so important?
-some argue that it is simply companies like Uber trying to cut their drivers out of a job...
-on the other hand, there are 40,000 traffic deaths a year in the US
--the sooner we get safe self-driving cars the sooner we can reduce that number
--get more cars on the road at once reducing traffic delays without having to build more highways or adding more lanes
What is AI?
-not just self-driving cars
-popular notion such as, Star Wars, The Terminator & ex machina, are one way to think about AI
What is real world AI?
Intelligence demonstrated by machines, in contrast to natural intelligence displayed by humans and animals
What does general AI do?
-attempt to build computational models of cognitive, emotional and/or social intelligence processes meant to simulate human intelligence
-or some aspect of human intelligence
What are robotics?
-human like/mobile intelligent agents
-robots use general AI to guide their behavior
T/F: At it's core all AI is software - that is, a computer program
true
Alan Turing
Turing machine - 1936
-proved that digital computers are possible
Turing test - 1950 (called the Church-Turing thesis)
-a test of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human
Turing test
if a human observer cannot determine by question if a given answerer is a human or a computer then the computer can be said to have the ability of the human
History of AI
John McCarthy - 1955
-AI name is coined
Dartmouth Conference of 1956
-very optimistic outlook for what AI would accomplish based on how fast digital computers were advancing
-within 20 to 30 years there will be intelligent systems and robots!
-this prediction proved to be widely overoptimistic
What are some goals AI has taken on?
Expert systems
-medical decision making
-game intelligence
Natural language processing
-getting computers to understand human interaction and respond to commands
The original goal of AI was to figure out, Can human intelligence be simulated?
-General intelligence (GA)
-in the human intelligence field this is called "G" and is what intelligence tests like the Standard Binet are suppose to measure as IQ
--things like reasoning, decision making and problem solving
-that is, can human intelligence be duplicated or even exceeded? there are two schools of thought on this question
What are the two forms of AI?
-Strong AI
-Weak AI
Strong AI
-the mind is a program
--based on the notion that programs are independent of their hardware
-any program can, in theory, run on any computer
--if we knew the right "code" we could run your mind on my laptop...
T/F: Strong AI says a "mind" or an "intelligence" is a program
true
John Searle's "Chinese Room" experiment
-computer programs manipulate formal symbols
-but they dont understand they have syntax but no semantics
What does Searle propose?
-Searle proposes a room where he (a non-Chinese speaker) sits
-outside the room a Chinese speaking person slops written Chinese questions through a slot into the door
-they are trying to determine if the room contains an "intelligent" Chinese speaker
What did Searle find?
-in Chinese room
--Searle is the equivalent of the computer
--a rule book is the equivalent of a program
-symbols are handed in and he looks up rules and hands back output
-however, he (the CPU) doesn't understand Chinese
--although observer outside the room might think so
What conclusion did Searle come to?
According to Searle, this analogy applies to other forms of cognition as well and therefore it proves (correct)
Weak AI
-the mind is software (of a sort)
-however, the hardware it runs on might place particular constraints on the nature of the program it runs (minds are more than software...)
According to weak AI...
the best we can do is simulate certain aspects of human intelligence
What is language?
A system of communication
-all communication involves transmission of information
-however, human language is arguably a unique form of communication
--because it is limited to humans (or is it?)
How does human language differ?
-it is the only (known) system of communication that includes both a finite set of abstract symbols (words)
-AND a
YOU MIGHT ALSO LIKE...