Cog 1 - Notes for test 1
Cog 1 - Notes for test 1 PSYC 3342
Texas A&M University-Corpus Christi
Popular in Cognitive Psychology
Popular in Psychology
This 33 page Class Notes was uploaded by Kimberly Notetaker on Sunday September 25, 2016. The Class Notes belongs to PSYC 3342 at Texas A&M University-Corpus Christi taught by Dr. Miguel Moreno in Summer 2016. Since its upload, it has received 4 views. For similar materials see Cognitive Psychology in Psychology at Texas A&M University-Corpus Christi.
Reviews for Cog 1 - Notes for test 1
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 09/25/16
How do Biological Systems Acquire Knowledge? (How do we gain knowledge?) We exist, and usually always in an environment o Energy, fluctuations of energy = information for us o EXAMPLES sound: we can feel it sometimes, we can hear it, and sometimes carries information like definitions, descriptions, etc Light, feelings, what we smell or taste o To take in that energy, we have to perceive that energy, which is where our basic senses come in (taste, smell, feeling, etc) o But, info only becomes useful when we can use that information to not only perceive what might be thrown at you, but to act on the perception by catching the ball or flinching from it What is knowledge? o It's about information o We have to understand the ball being thrown at us is expanding as it's coming closer and that we have to catch/dodge/etc How is knowledge organized? o Access quickly to not take too much time to catch that ball o Also how we don't access information (come exam, forget the answer to that ONE question) Learning implies change (triangle..), but a change in what? o Our fluctuations in environment are going to change o we're going to change by retaining and practicing new information o change how we interact with our environment/engage in the world o how we perceive the world The paradox of learning/knowledge o in order to learn, we already have to start with knowledge! o Read, write, know alphabet, how to see certain surfaces, recognize the sound of each sound for speech, know how to speak, o Requires that we build on something that we've already learned o Knowledge is necessary to acquire knowledge Plato Had an answer for the paradox of learning When we experience the world around us, things that are imperfect like nature (we'll never find a perfectly round, red object [maybe in our head! but 'memory' of that is not experienced from memory. in our mind, we can think about it]) cannot be found. o How can we know unless we think of what is perfectly round? o In our minds, we have concepts that are pure (pure forms, perfect ideas) which helps us understand the world around us! True knowledge can only be obtained through appreciation of pure forms, because only they have real, permanent existence o For Plato, the only world that really matters is the one in your mind where you can have perfect ideas Reminiscnece Theory of Knowledge o The immortal soul has complete knoeledge of the pure forms, this knowldge is only accessible through introspection Aristotle The actual work is real, and all knowledge can be extracted from it Knowledge is accessible by examining actual world and reasoning about it Knowledge is constructed by associating sensory information Descartes MindBody Dualism o Mind and body are distinct entities o Behavior of physical body can be explained mechanically Physical part of you is governed by the sciences (physics, biology, etc) and can move all on its own Eye twitches, muscle cramps, testing for joints, etc, those happen to us, it's our body doing it to us, but we're not doing its Involuntary actions Rudimentary notion of reflexes (automatic responses): reflection of our body, not us Nonphysical mind provides humans with consciousness, free will, and awareness Puppeteer that controls the mechanics of our body Voluntary actions Can modify, facilitate, or restrict involuntary actions o In a way, Plato and Descartes were about on the same page (Plato with his idea of the soul) Principle of Neural Representation (^^Descartes) We (our minds) have only direct awareness of states of the body The mind is not directly aware of the states of the environment BUT the mind can infer state of the environment from states /reactions of the body Energy of the world ["motions in light'] creates energy corresponding energies in the body ["motions in nerves"] Mind begins existence with knowledge: o God o Geometry/space/time/motion o Inference/reasoning capabilities Nativism/Rationalism (for Plato and Descartes) o N: Believe we are born with a certain amount of knowledge knowledge is inherited o R: Don't have to experience, can gain knowledge just by sitting down and thinking primary of reason o Person is a machine, mind is ghost "the ghost in the machine" o Can you build a formalism of Reason? o Can you build a mathematical system that captures the complexities of mental abilities? Formalism / Formal System A finite set of parts/pieces (limited) + A finite set of rules (limited) = Proper/logicalsequences (unlimited) o Examples o Genetics limited AT/CG, can only be used with one another, but endless diversity Chess pieces, rules of how/where to move, infinite number of different games Complicated, but deterministic sequences of outcomes (like our mind emotions, thoughts, memories, etc) Can a finite number of pieces and rules generate infinite outcomes? Russell/Hilbert (^^last question) Can you look at formal systems in the same way as scientists look at other things in nature? Through observation, classification, etc Mathematics as the object of inquiry, not a tool of inquiry Developed predicate calculus: a lanaguage suited to asking scientific questions about mathematics Questions about formal sustems o (1) Are they complete? o System has truths, all of whihc can be proben enturely within the rules of the system o (2) Are they consistent? o Can follow all of the system rules and never reach a contradiction o (3) Are they decidable/definite? o Can identify definiye sequence of steps form input to output Behavioral and Physiological Approaches Godel (Gurdel) Post problems with mathematics Provided two theorems o (1) For any consistent formal theory, it is possible to construct an arithmetic statement that is true but not included in the theory o Sometimes we have these puzzles that we can't explain o (2) If a formal system can be proven to be consistent AND complete from within itself, then it is inconsistent In a nutshell, formal systems CANNOT be both consistent and complete o Example: imagine that we can make a universal truthtelling machine that spits out only true statements, but it would print out all true statements (at the time) so, print all true statements, and print only true statements o Every machine has rules for how the parts of it work together. We give it an instruction, which is "the universal truth machine will never print out this statement". Flaw, because that statement is a true statement (paradox). If the UTM does print out that statement, it'll print out a false statement, which violates the only part. If the UTM never prints out the statement, then the statement is true, which violates ALL parts His theorems state that it is impossible for a system to generate all possible truths. We study something that is hard to measure, and we have to understand how the puzzles work in our minds. There always has to be a limit to what can be explained. Allen Turing **Lived before computers, so problems were worked out with a person Can you mechanize mathematics? It has parts, it has rules, so can make a machine with those parts and those rules (what he was really saying^^) Can you build a symbol manupulating machine that can represent a formal system? o The number 1 stands for 1 of something, so it's a variable. All the numbers we can create can be variables, as they can stand for a lot of things. Turing Machine the hypothetical machine that could compute numbers o Simple, had a reader that could determine the symbol written on a piece of paper. Also a writer that could write more symbols on a piece of paper. o Included an infinitely long strip of paper, which we can't do. Cells to hold each symbol, and would be only 1 or 0. o Need the rules! o Different for each process, of course For addition, depending on the state/settings (A, B, or C) and depending on what was on the piece of tape, would follow certain instructions. Print a 1, move to the right, and stay in the A position. Then, getting to the zero, change zero to 1, move to the right, and now in B state. So on until reads another 0 in the B state. Print 0, move left, and change to C state. So, depending on if it's a 1 or a 0, it'll either print a 0 and stop or just stop. Laborious, but maybe there's something in our mind that happens similarly, just a lot faster o Symbol manipulating system capable of any mathematical operation o Can we build a symbol manipulating machine that can represent thought? Can we mechanize thought? Turing Reductionism o Claims that all of our human states of mind/all our conscious awarenesses/products of our minds are the product of a symbol manipulating machine o "mental states are the computational states of a Turing machine" o Descartes we deal with representations of what is outside our mind o So maybe we compute the symbols of the outside o The mind is some type of Turing machine o Beginnings of artificial intelligence to get machines to do what humans can do Token Pshyicalism Any kind of matter that can support mental states can be a Turing Machine Carbon is not a privileged substrate Any kind of stuff may be able to have a human experience if that's true, then it doesn't matter what you're made of as long as you have the right rules and parts o Like those movies where you 'upload' your consciousness "It doesn't matter what the matter is" o What represents the mind is what it matters, which is what/who you really are o Anything can eventually be 'human' By way of contract... Type Physicalism o Can't replicate the human mind/experience into o Our perception of color is how we experience different wavelengths of light. But, even if a machine can read the color wavelengths, it cannot experience the color how each individual human does. o Mind and mental states require carbon based matter The Turing reductionism and Token physicalism allow us to model the mind and body as Descartes envisioned o He argued that we have two existences our physical body and our mind o Both say it doesn't matter the physical, as long as you have the 'hardware' or the mind, the true source of who you are o Mind as a symbol manipulating machine independent of the particular physical characteristics of the body o Mind/body dualism Throughout the whole course, we'll talk about "reduce thought to computation" relying heavily on metaphor o Approach cognitive psychology using a computer metaphor How can we model the mind as Descartes envisioned? Godel numbering express formal systems numerically Turing Machines manipulate numerical states mechanically Turing Reductionism represent mental states mechanically Token Physicalism Mental states in any medium ALL OF THOSE ABOVE = mind as a symbol manipulating machine Independent of the particular physical characteristics of the body Reduce thought to computation using computer metaphor Behavioral Approach (NOT Behaviorism) Information Processing (IP): abstract description of how information is manipulated flow diagrams o Input and output o IP is the sequence from input all the way to the final output o Infer components of the mind based on behavioral measures Pshyiological Approach (neuroscience) Neurological correlates of behavior Some link between brain and behavior links between brain structure and cognitive function Infer component of mind based on physiological changes In both approaches, we have to rely on an inference about the parts and understandings of the mind Behavioral Approach Donders How do the relations between changes in the stimulues and behavior inform us about the mind? Simple vs. Choice Reaction Time (RT) o State Theory: thought = train of processes or stages RT is the sum of all different processing times, like an assembly line (all working together, speed of assembly depends on how quickly each person does their job) o Task 1: Stimulus >...> A > C >... > Response o Show light, if there, press button Process visual information and move to make the reaction o Task 2: two lights, determine which one o See image, make choice, make muscle move (A > B > C) In difference of times, we can see subtraction method (RT2 RT1 = Stage B processing time) o Subtraction Method: RT2 RT1 Stage B processing time o Task2 inserts a new stage without altering others Sternberg Search Experiment Experiment done in class replicated from our hw assignment Example of behavioral approach o Stimulus presented, behave somehow to react Attempt to understand how memory works o and also how to use that information, like making comparisons Avoids pure insertion assumption (Donder's) o bc it was always the same task Instead, it manipulated the memory set size o memory set = "3 9 6" is stored in memory o Probe = 9, z = "not present", / = "present" No new stages inserted when you change the trials; memory content manipulated o 9 > perceive probe > search memory [maybe what happens is you do a comparison one at a time, like the first or second or even third in the set. Is 9 in the set of 123? No. Example of serial search] > done with search, so we make a decision (sometimes we betray ourselves by pressing the wrong button for example) > generate a behavioral response > / o Search process could be exhaustive Exhaustive vs. SelfTerminating Exhaustive o Uses all options/possibilities o Every item in STM is always examined/compared to the probe o More items, more time spent o absent / present same slope o reaction time is faster when memory set size is smaller, slower when bigger Both go at same rate because it's not any different whether it's absent or present Slope is linear, meaning amount of time to search for that one item is about the same for each SelfTerminating o Reflects more what's typical of what we do in our daily lives o Stop the search as soon as the information/itemwe're looking for has been found o Reaction time will be different each time o Present is half as steep as absent (chart) Degraded Stimulus o The second condition in the experiment done in class (the unfair one _) o Influences perception stage only o Longer RT, slope is the same Sternberg Search: Class Data o Slower RT when memory set size is bigger o Using a serial kind of search o present and absent overlap o using exhaustive type of search o Linear relation between RT and set size o RT for present and absent are approx. the same o Slope for present and absent are approx. the same o Which type of search process is supporyed? Most likely selfterminating' o Comparative system in our minds that help us compare things in our memory in an exhaustive way, comparing many things in our memory, not just one thing Sternberg's Model as Prototype of Information Processing Theory No attempting to attritbute information processing to any brain location or brain process o Don't care about region, care about how it works Highly symbolic o No consideration of neural representation Computer metaphor high speed scanning o Using search feature when looking for a file Assumption of additive finishing types o Assumed each stage took a certain amount of time (look at probe, compare, make decision, mind to control body to create response) o All of those times each add up together is our reaction time o Independent/serially ordered processing stages o Don't all happen at the same time, we have to create the representation of the probe first, and then go on to the next stage of above Physiological Approach How do the changes in the brain inform us about the Mind? o Patient HM (Henry Molaison) HM Henry Molaison Bicycle accident at age 9 o Lived in a time where helmets weren't really a thing Increasingly severe epileptic seizure since age 10, unresponsive to medication o overactivation of neurons across the cortex Seizures getting worse and worse At age of 29, went in for neural surgery (in its infancy, not good) bilateral hippocampus removal (both of the removed) The surgery worked for its purpose, removing the seizures, but no LTM Short term memory was normal as long as he wasn't bothered/distracted Most memories prior to operation (LTM) accessible Surgically removed brain areas are important for transferring information from short term memory to long term memory Told us that there were at least two types of memories First Contact with the World Energy > Sensation, Energy > Brain E > S: Physical( ) energy converted to Psychologica( response In some sense, S copies E, however imperfectly. It is the 1st stage in getting the world inside the mind How does a signal get passed along from sense organs to the brain? Coding by neurons Coding by neurons o Light hits the rod (or cones) o If energy is less than threshold, then nothing happens If energy is more than threshold, cell fires Sends an impulse to the next cell o Impulses are the currency of the nervous system 2 types of problems: o Get signal down one cell o Get it across to the next cell o Different problems lead to different solutions Structures of Neurons Receiving end o Dendrites around the cell body Pathway o Axon Sending end o Terminal ending Sensory neurons take signals from receptors and send them to the brain Motor neurons take signals from brain and send them to muscles Interneurons relay signals from one part of the nervous system to another Problem 1: Withincell signals Traveling down the axon Begin with resting state with no activity o Check out chemicals inside and outside o Fluids contains ions (electrically charged) o Characterize axon with respect to electrical voltage inside vs. outside o Resting potential is negative Stimulate cell enough to exceed threshold o Voltage difference changes o Membrane destabilizes o Positive charge travels down axon o Signal is propagated from one end to another without losing strength o Action potential = electrical positive signal when the neuron The Code There's baseline stimulation, moderate stimulation, and intense stimulation Amplitude and speed of action potentials do not change o Cell fires or it doesn't: longer/stronger stimulation does not influence individual action potentials o All or None Law E > Intensity coded by: o frequency of firing (Rate Law) o # of neurons firing The positive charge makes its way down the axon o Sacs of chemicals sit in terminal endings o When the action potential gets to the bottom... chemicals (neurotransmitters) are released into the synaptic gap or synapse Problem 2: Betweencells From axon to dendrite Use behavior to infer physiology On a person, one zap has no reflex, so it's below threshold With two zaps, it's enough to create a reflex o Zap happened at same time, but in different places (spatial summation) o Even if zap happened on same spot but at different times (temporal summation) Chemical neurotransmitter signals (excitatory or inhibitory) add up o Eventually trigger next action potential o Where does it go? o What happens when it gets there? BottomUp Processing Stage 1: Transduction Transforming one form of energy into another Our body has to translate what we detect by variations of nerve impulses Source of light from the outside world, it'll hit objects and either be reflected by a shiny object, absorbed, or scattered (depending on the surface). Our mind detects that property and gets to our eye. o How is light energy picked up by the eye and turned into a neural impulse? We go to stage 2 Stage 2: Neural Processing We have structures at the back of our eye that help us identify what we're looking at o Rods and cones o Cones right behind our pupil, and responsible for sharp images and where dark/light (shadows) o Rods sense change of light Since we're looking at light, mostly using sensory neurons Lots of cells in use, so greater sensitivity in detection of light BUT, fewer cells in the center of the eye for greater acuity Some neurons that are stimulated make sure that the next one doesn't get stimulated. Others, though, will try to get the next one ready to cause an action potential. So, either an excitatory or inhibitory action. Separate receptors converge onto a ganglion cell o Pool information to ganglion cells o Some will be excitatory (inside region), others inhibitory (outside region), to the ganglion cell o Collection receptors is the ganglion's receptive field Look at powerpoint slide for illustration Lateral Inhibition o Neural circuitry/mechanism for detecting stimulus features by the convergence inhibitory and excitatory signals o To do with previous example with different light coming in and the ganglion firing rate o Receptive fields care about size and shape, BUT NOT orientation o Specifically about size (mostly) and shape does not distinguish between horizontal or vertical But since orientation influences what objects mean, we have more convergence in order to help with identifying orientation o Modest firing rate: Stimulus smaller than excitatory center o Maximum firing rate: stimulus "fits" excitatory center o Reduced firing rate: stimulus hit both excitatory center and inhibitory surround To help with orientation... o Receptive fields overlap o We can now access a collection of receptive fields, as orientation is important o To cells in the cortex (brain) o They have receptive fields too, but very different from ganglion cells (look like a bullseye, concentric circles). In the brain, look like layers of a cake o Some cortical cells prefer vertical rays of light, others prefer horizontal or oblique (diagonal) simple cells help understand the orientation o Lower visual processing includes ganglion cells, and simple cells which detect edges and lines with particular orientations. Maximal response to stimuli of a particular orientation at less than or greater than 15 degrees o Although they have a preference for orientation, doesn't mean they'll act on that ONE preference Like when looking for your favorite chocolate bar, but can't find it, still want chocolate, will pick something else based on first option. If wanted Snickers, try to get something similar, not gum for example. o Some designs in our nervous system are features detectors, which prefer movement of those features in a particular direction o also movement these cells are called complex cells Response (to orientation & size & motion direction!) Maximal response to stimuli of a particular orientation, AND motion. if it's not acting how it's supposed to (detecting one accurately but maybe off on the other), won't fire at it's maximum The mind has to have access to a lot of simple and complex cells in order to really know what's going on. Cortical cells as feature detectors? Provide some where, what, and what's it doing information Ambiguity: firing rate is reduced if orientation or size or motion is just right > which is it? Firing rate is the only vocabuary that our mind understands, the nervous system MUST use this Primary Visual Processing Receptors copy properties as best they can o Omissions, distortions, errors Some characteristics of distal stimulus (Sd) are recovered by feature detectors o Voice distal stimulus o wavelengths of sound from voice proximal stimulus (bc closest to our mind) Signals travel to specialized brain mechanisms Where does the signal/information go? o To the brain to be processed o Looking at a bird, the bird is the distal stimulus, and the process of seeing the bird is the prox stimulus o Once hitting the back of the brain, creates a visual representation What happens when it gets there? o More specific copies or representations are needed Regions of left and right eyes correspond in cortex o Cortical cells form retinotopic or topographic maps of LVF and RVF o Ex Primates experiment o to look at the semi circle on the presentation once they knew they were looking at the image, they harvested the brain found the pores open (stimulated) and stained the brain to see where exactly in the brain the activity that takes place image was upside down at the back of the brain, so there's somewhat of a neural image o Spatially distorted, reflecting importance of receptor region o Point eyes to something, it's something we want to look at, so takes a lot of brain space to process that information o Amount of brain space for each function represents the importance of that sense Receptor signals do not remain separate o Many to 1 (or few to 1) convergence (receptors to next cells) o Few to 1: more acuity, more specific to see what's going on Many to 1: More like seeing what's brushing on your leg Pooling, editing information > How are features organized? o Slide with black lines o See a bird WHEN it moves o There musr be a rule of thumb for what features go together? o Gestalt heuristics h what is mostly true ex: in texas, bigger is better. in a way, it can be true. Form Perception Feature detectors provide inventory of l, , / , \ Gestalt laws (principles/heuristic) of perceptual organization one curvy, one straight line o other options of seeing 12 different line segments o Pragnanz: of the alternatives allowed, choose the more likely set of nine triangles o can say 3 rows or 3 columns of triangles o now we see that circles are in the middle to replace the triangles o now we want to put it in columns to group together o Similarity: look alike things go together o part of one group Set of 6 vertical lines evenly spaced out o Now see that they are grouped in two's o Proximity: nearer things go together like a hospital system/buildings or a campus with multiple buildings, people who sit next to each other, assume they know each Black line coiled around white pole two objects o could be five different objects, but we don't tend to assume that o Good continuation: smooth edges belong together o the spiral makes us assume it's just two objects example: something hanging on the wall, we know wall continues behind whatever is hanging on it The bird example from earlier o Common fate: pieces that move together go together o traffic light: green light group goes together while red stay together o wheels on a car working together to move together Seeing forms precedes recogniziing patterns o Gestalt principles of organization segment the visual elements into objects o Transduce, detect, realize, and then what? Stage 4: Pattern Recognition Biederman's recognitionbycomponents theory Familiar object may be recognized as a configuration of familiar component parts o Like how we read: words made up of familiar shapes (letters), and would be able to understand new words by seeing the letters and knowing how they sound o Similar to how we use the 26 letters in our alph to create different words Object Recognition involves breaking down complex objects into simple component parts but, what are the parts? o He created GEONS o From a small number (36) of basic geons its is possible to creae a large number of unique combinations Look at slide o View variant properties: geon's unique features visible from nearly any perspective o Each geon can be distinguished easily from nearly any perspective: Discriminability o Like letters, each one is easy to discern o Noise resistance: geons can be distinguished despite visual noise o Accidental perspective o Geon's unique features are no longer visible all due perspective seeing a pregnant person from behind and not seeing the bump so not thinking they're pregnant Biederman's recognitionbycomponents theory: 4 Stages of processing Edge extraction o Provided by early neural proessing like lateral inhibition mechanisms o using ganglion cells to fire off and send info about the object Then, subobject segmentation o Edges and lines are grouped into objects o Gestalt principles of grouping (like the similarity, prox, continuation, etc) THEN, geon categorization o Subobjects are identified as particular geons o Like recognizing letters in a word Lastly, object recognition o Object is recognized as a configuration of geon pieces o Like recognizing a word Object recognization should fail IF the geons cannot be recognized (third step) Example/experiment on slide with looking at pictures to idenitfy objects, look at it!!! Basically about recognizing geons even under noisy conditions middle column more reconizable because more than enough geons abstract art as an example Same length of Line in Middle and Right column figures Center figures: Intersections intact o Geons still recognizable o Correctly identified 90% of the time when viewed for less than a second Right figures: intersections removed or changed o Geons unrecognizable o Correctly identified only 30% of the time even when viewed for 5 seconds The Modern Helmholtz like program: Information Processing Successive stages of storage and transformation Of what? o "information" o Images, copies, features, representations o Rules developed from experience help o Unconscious inference Energy > Sp: o Sp o Retinal image pitch pattern cutaneous image o Sp then leads to pooling, editing in pathways o Visual, auditory, somatosensory Energy: patterns of excitation and inhibition in Sp, then patterns become more specific when being pooled, edited in pathways, patterned are organized in the brain, and THE PATTERN IS RECOGNIZED YAY BottomUp Processing is not sufficient Encoding/transformations driven by proximal stimulus o May see something we recognize but turns out to be something else we don't recognize The pattern of light on the retina is exactly the same o The representations in our mind are different What makes our percepts different even when the Sp are exactly the same? o TOPDOWN PROCESSING: encoding/transformations driven by expectations and knowledge o topdown more important in realworld situations Misreading Effect: Perceiving what is expected to be in the sentence Auditory only: hear /ba/ Visual only: looks like /ga/ Auditory and visual = /da/ The visual information creates an expectation (topdown) The auditory signal (bottomup) mismatches the visual expectation Perception is an interaction of bottomup and topdown information Helmholtz Perception comes from proximal stimulus (Sp) + rules (knowledge) o Perceived size is derived from perceived distance o Like the MullerLyer experiment we did this past week o The pattern of light on our retina from the further line (Sp) o The pattern of light on our retina from the closer line (B) o SpA = SpB Depth cue: linear perspective says "A is farther than B" Size distance scaling: Sp size decreases as distance increase So... o If A & B project same size Sp o And is A is farther than B o Unconscious inference > A must be bigger than B What about the MullerLyer Illusion? (the experiment we did!) Helmholtz: Perception comes from proximal stimulus (Sp) + rules (knowledge) o > perceived size is derived from perceived distance o Rule: RI size decreases with distance, so... o If A and B project the same size Sp And if A is farther than B Inference > A must be bigger than B
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'