Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
95 Cards in this Set
- Front
- Back
What are the two basic tasks of language?
|
We have to comprehend and produce speech/language. In essence, language is an arbitrary intermediate that simply conveys abstract ideas.
|
|
Why does Daniel say that words are arbitrary? Are sounds arbitrary?
|
The physical sounds do not give words meanings. We assign meanings to sounds based on our culture and/or the language we speak.
Probably not, there are physical properties to sound that we can make associations with. |
|
What are the levels of language? (There are 5).
|
Sentence --> Phrase --> Word ---> Morpheme --> Phoneme
|
|
Do phonemes always correspond to letters?
|
No.
|
|
What are phonemes? Are phonemes universal across different languages?
|
The smallest unit of SOUND. There are 40 in english. There are different phonemes in different languages.
|
|
What are morphemes?
|
The smallest unit that signals meaning! This can be entire words.
|
|
What is a content morpheme? What is a function morpheme?
|
A functional morpheme (as opposed to a content morpheme) is a morpheme which simply modifies the meaning of the word, rather than supplying the root meaning of the word.
|
|
Are words good enough to unambiguously convey meaning?
|
Nope. Phrase structure is really important too. Different phrase structures have different meanings, and the inflection cues that distinguish these meanings may be subtle (think: micropauses)
|
|
What is syntax?
|
The rules that determine word order
|
|
What is meant by phrase structure descriptions?
|
a
|
|
What are some problems with phrase structure approach to meaning?
|
Ambiguities remain: one phrase can mean different things, and different phrases can mean the same thing.
|
|
Surface Structure
|
The (valid) phrase structure of the current utterance
|
|
Deep Structure
|
A (valid) canonical phrase structure that has the same meaning as the surface structure.
Canonical phrase structure is stored in memory. This is like... an exemplar/cognitive economy approach to recognize phrases. |
|
Transformational grammar
|
Rules that transform a deep phrase structure into surface phrase structures with the same meaning.
|
|
Lexical ambiguity
|
Word has two different meanings
|
|
Syntactic Ambiguity
|
Words can be grouped together into more than one phrase structure
|
|
Swinney Experiment
|
Primes people using a sentence, then asks people to judge whether a word is real or not (lexical access point) and words that were related to the lexical access point.
Immediately after primer word, people get both meanings of word. After a little bit of time, though, people only get contextually appropriate meaning. |
|
What is the subject-verb-object order experiment related to?
|
People respond to active voice faster than passive voice when the action (verb) is reversible. But they respond/comprehend active and passive sentences equally when the verb is non-reversible.
|
|
Broca's Aphasia:
|
Difficulty with production: Slow, halting speech.
Simple grammar: no function words (be, of, the). Comprehension largely intact. |
|
Wernicke's Aphasia:
|
Fluent speech, but makes little sense (word salad).
Made-up words, word substitutions. Difficulty with comprehension |
|
What happens in language acquisition?
|
We have to learn how to distinguish language sounds from other sounds.
We need to be able to parse sounds into phonemes -- atomic units of language sound -- and then parse phonemes into words. Then, once we have the words, we need to assign meanings to words. |
|
What are some problems in assigning meanings to words?
|
The same sound can refer to different things, and different sounds can refer to the same thing.
|
|
What are rules we need to learn for language?
|
We must learn how words are combined
We must generalize to novel sentences (so we can't just memorize wordings) We need to acquire rules that can be applied to new sentences. |
|
Do children receive feedback on their grammar?
|
Nope! Adults only correct meaning, not grammar. Furthermore, correcting grammar/pronunciations doesn't even help. Kids correct themselves despite no negative feedback.
|
|
So, if kids aren't getting feedback on their grammar, what does this suggest?
|
This suggests that there are linguistic universals. Aka, this means that there are general language principals that are embodied in every language.
|
|
Development: Phonemes
|
In first year, infants can discriminate all phonemes from all languages. Gradually lose discriminations that are not important in their own language.
|
|
What's "Motherese?"
|
The way moms talk to their kids. They speak with higher pitch, slow rate, exaggerated intonations, pausing signals, and phrase boundaries. This aids with parsing.
|
|
Holophrastic
|
One word stage. No syntax, need context to disambiguate. Undergeneralization AND overgeneralization for first 75 words. Do UNDERSTAND some phrases
|
|
Telegraphic Stage
|
Two word stage. Correct use of word order. Can convey a lot of information succinctly.
|
|
Learning Syntax/Rules/Generalization
|
Start learning past tense, and nonsense words.
Learning for irregular past tense is U-shaped. Initially use right form, but then overgeneralize to -ed. Children are able to apply general rules to nonsense cases, which implies that language learning is generative, not just imitation. |
|
Learning Word Meanings: Parts/Wholes
|
Input gives clues about parts/wholes
“This is a rabbit; these are his ears” “his” cues the fact that “ears” is a part |
|
Learning Word Meaning: Bias toward Shape
|
Inherent bias that new words refer to shape
If something is a “biff”, children will claim other objects with same shape also “biffs |
|
Critical Period Effects: Social Isolation
|
Remember Jeanie -- After critical period (age 10-12) she could not learn english.
|
|
Critical Period Effects: 2nd Language
|
You are way better at learning languages when you are younger than the critical period.
|
|
Critical Period Effects: Sign Language
|
See the same critical period effects.
|
|
What are the two basic problems in motor skill acquisition?
|
Skill and Serial order acquisition
|
|
Fitts Power Law of Learning
|
A Mathematical idea that the same amount of practice doesn't necessarily buy you the same amount of skill every time.
|
|
What are the stages of skill learning?
|
Cognitive, Associative, Autonomous
|
|
Cognitive Stage of Skill Learning
|
Declarative knowledge.
Commit facts to memory. Rehearse as you try to perform. Requires attention--can’t do second task. May be independent of skill: the best teacher may not be the most skillful, but rather someone who knows how to describe the cognitive stage well. |
|
Associative Stage of Skill Learning
|
Strengthen connections that lead to desired result
Feedback is important: See which actions lead to desired result Get rid of actions that lead to errors |
|
Autonomous Stage of Skill Learning
|
Fast
Executed with less attention/consciousness Less verbalization Less dependent on verbalization Declarative knowledge less available Feedback Less important Lower level (proprioceptive not visual) Need for consistent practice where a stimulus always gets the same response |
|
ACT-R (anderson)
|
Computational model -- adaptive control of thought
Explains how we go from declarative knowledge to implicit procedural knowledge. Proceduralization: Take declarative knowledge and turn into productions Composition: Take several productions and join them together into one |
|
Response Chaining
|
Feedback from one movement triggers the next one
It's wrong. It takes too long. Movements occur too quickly. |
|
Motor Program
|
Representation of the plan for movement and movement sequences
Fast, doesn’t require feedback Abstract Hierarchical - Abstract high level, specific low level Composed of subprograms -- Less abstract representations of movement sub-parts |
|
Rosenbaum experiment
|
Evidence for Hierarchical motor programs.
Brought people into lab, taught them a sequence of drumming their fingers. You can break down motor program into subprograms based on reaction time. Jagged reaction time curve. |
|
Evidence for Subprograms: speaking
|
Reaction time to say first word increases when the number of words to be said increases
|
|
Fitts Law (#2)
|
Speed/accuracy tradeoff.
|
|
What are 2 types of problems?
|
Well-structured and Ill-structured
|
|
Well-structured problems:
|
Completely specified starting conditions, goal state, and methods for achieving the goal (proofs)
|
|
Ill-structured problems:
|
Some aspects are not completely specified -- finding the perfect mate, etc.
|
|
What are the stages of problem solving?
|
You have to
1. form a representation 2. construct a plan 3. execute plan 4. check/evaluate yourself |
|
What is a problem space?
|
It's the WHOLE range of possible states and operators, ONLY some of which will lead to the goal state.
Operators: actions that move in between states. So what we do in a problem space is: we try to navigate the states and operators in the best way possible. |
|
What do we mean when we talk about problem representation?
|
For many problems, the representation may make it easier or harder to solve
For example, geometry problems are easier graphically, but algebra problems are easier to represent as equations. |
|
What is an isomorphic problem?
|
An isomorphic problem is a problem where we can rephrase/re-explain/represent the problem in a different way, to make it easier. E.g. the monk problem.
|
|
How do we use analogies to solve problems?
|
We retrive a representation of a problem from meory that is similar to the problem we currently face.
|
|
How do we mess up analogies?
|
We tend to miss deep similarities between problems, because we focus on surface similarities.
We can use analogies, but we FAIL to notice them -- military ray experiment. |
|
Top-down preconceptions
|
When we look at a new problem, we tend to encode it in a way consistent with long term memory.
|
|
Functional fixedness
|
We see an object as having only a fixed, familiar function.
|
|
How do we get trapped by a familiar perspective?
|
We like to take the point of view of the familiar process.
|
|
Stuck in Set:
|
Improper applications of analogous situations.
|
|
Algorithm
|
Completely specified sequence of steps that is guaranteed to produce AN answer (usually guaranteed to produced correct answer)
Can be slow and laborious |
|
Heuristic
|
Short cut/ Rule of thumb. Never guaranteed to produce correct answer, but usually quick and easy.
|
|
Think aloud protocol
|
When we're solving a problem, the experimenter just asks us to try it aloud.
Note: protocols do not reflect all of the reasons why a person tried a specific configuration |
|
What is difference reduction?
|
SAME THING AS HILL CLIMBING
At any point in a problem space, we always try to select the operator that moves us closer to the goal state. This doesn't really work (orc/hobbit problem) |
|
Means-end analysis:
|
We identify the largest difference between current state and goal state. We set a subgoal to reduce that difference. We find and apply an operator to reduce the difference. etc etc.
|
|
Problem Solving: Working backwards
|
Transform goal state so it is more similar to it is the initial state! (reverse of means-end). Useful if too many paths leading from initial state.
|
|
How can expertise harm your ability to problem solve?
|
Can sometimes harm:
Functional Fixedness Water jug ‘mental set’ |
|
Chess Study
|
In chess study, experts memory was no better than beginners.
Memory for meaningful configurations much better. Memory for random configurations slightly worse (probably hindered by schemas) Chess masters know 50,000 chess patterns. Chess masters intentionally study these patterns. |
|
Do experts spend more time on representation or less?
|
More time. They take less time to complete it.
|
|
Do experts use more or less means-end analysis?
|
Less! They have more pre-stored solutions in LTM
|
|
What is deterministic/Deductive reasoning?
|
Predetermined that if the premises are true, the conclusions must be true.
|
|
What is probabilistic/inductive reasoning?
|
This is what's PROBABLY true. I can REASON that this is true.
|
|
What's normative reasoning?
|
How one ought to reasons -- rules of logic
|
|
What's descriptive reasoning?
|
How people actually reason -- heuristics
|
|
What two things you need to draw probabilistic conclusions?
|
Current evidence and base rates
|
|
Bayes Theorem
|
A normative theory of probabilistic reasoning
|
|
Availability Heuristic
|
A heuristic we employ when we estimate frequencies or probabilities by the ease with which instances or associations are brought to mind
|
|
Simulation Heuristic
|
Base judgements on how easily you can imagine: how things will turn out in the future, and how things would have turned out in different circumstances
|
|
Conjunction Fallacy
|
Occurs when people mistakenly believe that a conjunction of events is more probable than a single event.
|
|
What are 3 causes of conjunction fallacy?
|
Availability
Simulation Representitiveness Heuristic |
|
Framing Effect
|
The same information presented in different forms can lead to different decisions
|
|
Non-Normative Heuristics
|
a
|
|
Risk Behavior
|
In a positive frame, people are going to be risk averse
In a negative frame, people are going to be risk seeking |
|
Deductive Reasoning: Quantifier
|
Categorical Syllogisms
A = B. B=C. A=C. If premises are true, conclusion is true. Quantitatively define a category |
|
Deductive Reasoning: Comparative
|
A<B. B<C. A<C.
This time we are comparing things. |
|
Deductive Reasoning: Conditional
|
If you do A, then you will get B.
You do A. Therefore you should expect B. |
|
Sound Argument
|
Both deductively valid AND the premises are true.
|
|
Content Effects
|
The content affects whether we endorse the syllogism or not, despite the "correctness" of the syllogism.
|
|
Atmosphere Effect
|
If either premise is particular, prefer particular conclusion
If either premise is negative, prefer negative conclusions |
|
Figural Effect
|
We like our terms in linear order.
|
|
Conversion Effect
|
We infer bad conclusions from premises. We convert the premise into a deductively invalid conclusion.
|
|
Pragmatic Reasoning Schema
|
Humans are not built to follow abstract logic problems -- but we are rational. We've evolved to reason in PRAGMATIC situations.
|
|
Formal Rules
|
People are logical, but make mistakes. Assume that formal rules of logic are built in.
Errors occur because we misinterpret premises, some rules are unavailable, and we can't find proof for the conclusion. Problem: why does content matter, then? Why would formal rules be built in, if they only apply to infrequent tasks? |
|
Mental Models
|
People construct mental models (representations) that correspond to the premises, describe them, and then try to falsify conclusion by constructing alternative models.
Errors arise from WM limitations Problem: Constructing counterexamples is a very complex process that is specific to deduction. Would subjects with no training use it? |
|
Verbal Reasoning
|
Repeatedly use linguistic processes to try to extract more information from problem statement
Rather than applying formal rules or searching for counterexamples Errors arise because linguistic processes are adapted to demands of communication, not deduction Includes inferred information that may not be deductively valid May exclude deductively valid information that wasn’t emphasized |