psychology essay
10/4/2019
1
Goals of this course… • Outline the basics of our visuospatial perceptual systems
• Establish and evaluate the two main competing theories of representation and cognition – symbolic and non- symbolic / embodied.
• Investigate cognitive processes and how they relate to the embodied experiences of visual and spatial perception
• Explore the symbol grounding problem and the nature of “understanding” something
• Discuss the current and future state of cognitive science as a discipline
• Expand your mind dude
How to succeed in this course 1. Come to class engaged and prepared to think
2. Ask questions when things are unclear • Do NOT hesitate! If no questions are asked, I’ll assume it’s
understood.
3. Don’t get caught up in writing down everything I say – take notes strategically • I will audio record and post my lectures, unless this is abused • A good strategy – one sentence per slide summing up takeaway point • Focus on ENGAGING with the material, not mindlessly copying me
4. Keep up with all readings • Readings are fair game on RPPQs • I’ll make sure to give you a discussion section in between when the
reading is assigned and when it is quizzed on
5. Take advantage of discussion sections • You can go over readings and discuss your projects and homework
here
6. Get an early start on your project proposal
7. Review material every weekend
Introduction to Visual and Spatial Cognition
10/4/2019
2
Close your eyes and relax…
How did you answer these questions? • Did any visual imagery come to mind while
answering these questions? How much did you “will” this?
• Did you imagine what it would be like to actually be there, or do the action? How real did this seem?
• Can you imagine things you’ve never done?
• Why is blue “cold”?
• When adding apples, did you visualize the apples? Did you intend to do this, or did you just do it?
• When solving the second math problem, how did you keep track of the numbers?
10/4/2019
3
Visual and Spatial Cognition • The idea that our visual perception, bodily
perception, spatial awareness, and sensorimotor abilities are involved with (necessary for?) cognition.
• How “thinking” utilizes perceptual processes, especially visual and spatial ones
• How “thinking” utilizes motor processes, specifically moving through and interacting with an environment
What is mind?
Mind-Body Problem • Dualism – Mind and body are
separate entities.
• “Although I possess a body with which I am very intimately conjoined. . . . it is certain that this (that is to say, my soul by which I am what I am) is entirely and absolutely distinct from my body, and can exist without it.”
-Descartes
• “Since our inner experiences consist of reproductions, and combinations of sensory impressions, the concept of a soul without a body seem to me to be empty and devoid of meaning. ”
– Einstein
10/4/2019
4
Penfield Experiment
Desmurget et al., 2009
Artificial Intelligence
• If dualism is wrong, and mind is not some ethereal quality…
• Then we should be able to build an artificial brain, and therefore an artificial mind.
10/4/2019
5
Turing and Computers
• Turing machine – theoretical computer that can calculate anything calculable.
• Given an input and a current state, provides proper output.
Brains as Turing Machines? • Input comes in • Input represented as
symbols (in this case, binary values)
• Symbols operated on according to rules (and based on state of the system)
• Rules transform input symbols into output symbols
• Information is then output • How your laptop works,
essentially. But is this how the mind works?
http://www.elsevier.com/connect/wp-content/uploads/2013/06/Turing-Machine.jpg
Side note: What is a “symbol”? • A symbol is anything that stands
for / represents something else
• This is not a pipe (but it is a symbol of a pipe).
• The digit “2” is a symbol. The quantity two is not.
• A map is a symbol. The actual location is not.
• The word “dog” is a symbol. My dog itself is not a symbol. But this picture of my dog is also a symbol. • Are the symbols “dog” and the
picture of my dog equivalent? Is one less symbolic than the other?
10/4/2019
6
Artificial Intelligence • What sort of things are symbolic processors and
Turing machines good at? What aren’t they good at?
• Original AI using “Turing style” symbolic processing not successful at many “human” tasks. Modern AI works a bit differently, and has had more success.
• Two forms of artificial intelligence • Weak AI – what we currently have – AI that can do
human-like tasks, such as object recognition, driving, or speech processing. Built for one specific task.
• Strong AI – what we do not currently have – AI that is indistinguishable from a human mind and can do all that humans can do to (at least) the same ability as a human.
• What differences exist between modern weak AI and human intelligence?
How can we know that an artificial intelligence has “mind” in the same way
that we do?
• If there were machines bearing the image of our bodies, and capable of imitating our actions as far as it is [practically] possible, there would still remain two most certain tests whereby to know that they were not therefore, really men. Of these the first is that they could never use words or other signs arranged in such a manner as is competent in us in order to declare our thoughts to others . . . so as appositely to reply to what is said in its presence. The second test is . . . to act in all the contingencies of life in the way in which our reason makes us act.
– Descartes
10/4/2019
7
The Turing Test
The Turing Test
• Is passing the Turing test enough to consider a computer intelligent?
• Is it enough to consider the computer conscious?
• Can a computer have the same mental experience as a human?
http://ichef.bbci.co.uk/news/660/media/images/75395000/jpg/_75395401_eugene-goostman-600×338.jpg
Standard Cognitive Science
10/4/2019
8
Computational Theory of the Mind • Implied from Turing – the brain is a symbolic
processor
• Perception cues symbolic representation, which can be operated on (processed) mentally to provide the desired output.
• Sensation -> Perception -> Symbolic Representation -> Action Planning -> Action
Standard Cognitive Research
• Most standard cog research follows these basic principles: • Design an experiment that
isolates one cognitive process • Manipulate something
between groups • Measure something and see
if there is a group difference using frequentist statistics
• Use the results to support / refute theories
Standard Cognitive Research
• Assumptions: • Cognition in the lab is roughly equivalent to cognition in
the world (it’s ecologically valid) • The way people do the task is equivalent
• The environment and state of their body (sitting in front of a computer) don’t impact cognition to a meaningful degree
• Cognitive processes can be isolated (e.g., visual search does not require any linguistic abilities) • This is also called modularity, as in two processes are
“modular” and are distinct in the brain
• Differences in mean DV values across conditions (as determined by significant p-values) means the two underlying processes are, in some way, different • And this difference is not due to something like task demands
or subject bias
10/4/2019
9
Standard Cognitive Research
• As an example, let’s look at some word priming research. • Participants read words off of a
computer screen
• IV – either the first letter is primed or not (using some sort of priming method, whatever kind)
• DV – reaction time (how long it takes to start reading the word)
• Results – participants are faster at reading words when the first letter has been primed
_____ b____
beach beach
Longer RT Shorter RT
Standard Cognitive Research • How do we explain this data?
• Levelt (1999) • The process of speaking a word has multiple “stages”
• One stage is phonetic encoding, a serial process
• When you prime the first letter, it can be phonetically encoded before the rest • This means that the primed letter cues a representation of that
letter to be slotted in to the representational frame of the to be spoken word
• Essentially, how a computer would do this. Symbol manipulation.
• When the rest of the word is presented, you don’t have to phonetically encode the first letter (it’s already there) so…
• Therefore, you see a reduction in RT for primed words!
Standard Cognitive Research • Is this the only explanation?
• From a computationalism point of view, this feels like it tells the whole story. The data is explained in a well- defined theory • This model is also explicitly
computational and can be programmed using algorithmic programming – WEAVER++ model
10/4/2019
10
Standard Cognitive Research • Thus, aside from providing the
input and producing the output, perception and action do not play a meaningful role in cognition. • Standard cognitive science typically
follows this pattern • “Envatted” brain
• But would it be possible to explain this data WITHOUT the symbolic representation?
• Could we appeal to sensorimotor processes to explain the RT difference instead of appealing to symbolic representation?
• We will come back to this later.
What would cognition entwined with perception
and action look like?
Imagery and Thought • Does thought occur
completely separately from perception?
• Let’s take imagination (a form of thought)
• Imagine the following situation…
• Imagination relies on the visual system and mental imagery. • Eye movements during
imagery • Much more to come here
10/4/2019
11
Memory in the World
• Suppose our question at hand is “How do people remember lists of items?” like a shopping list.
• Standard cognitive science • Give people lists of…
