LIN 260 Final

what is cognitive science?

the study of the mind. how does the mind process and represent information?

components of cognitive science

linguistics, anthropology, psychology, philosophy, computer science, neuroscience

operant conditioning

reward and punishment

classical conditioning

two stimuli are repeatedly paired to elicit a conditioned response

B.F. Skinner

proposed operant conditioning as the main driver of language acquisition. verbal behavior

are brains like computers: hardware level

no. computer processors are primarily serial. brains operate in parallel

are brains like computers: abstract level

yes. they perform computations over representations

computational theory

what is the goal of the computation, why is it appropriate, and what is the logic of the strategy by which it can be carried out?

representation and algorithm

how can this computational theory be implemented? in particular, what is the representation for the input and output, and what is the algorithm for the transformation?

hardware implementation

how can this representation and algorithm be realized physically?

computational level example

input: two numbers. output: their product

algorithmic level example

whatever step-by-step process you learned

hardware level example

your brain, pen, and paper

domain specificity

specialized for certain kinds of info

innate and developmentally regular

instinctive, low variability

fast and automatic

you don’t have to think about it

information encapsulation

modules don’t rely on one another

farah et al

is there a specific module for facial recognition? results found that there was

the faculty of language

the combination of cognitive modules that allow us to produce/process language

faculty of language in the narrow sense (FLN)

just the language-specific parts

faculty of language in the broad sense (FLB)

FLN plus supporting modules

damage to these areas

aphasia

face blindness

prosopagnosia

broca’s aphasia

mostly structure

wernicke’s aphasia

mostly meaning

lateralization

the brain is divided down the medial plane into left and right

split-brain patients

severed corpus callosum. behave unusualy in experimental settings

syntactic structures (chomsky)

how to precisely characterize what makes some sentences grammatical, but not others

grammaticality

is an utterance well-formed in language x?

felicity

is an utterance sensical in language x?

grammatical and felicitous

alice cut the onions and peppers with a knife

ungrammatical but felicitous

what did alice cut the onions and BLANK with a knife?

grammatical but infelicitous

colorless green ideas sleep furiously

ungrammatical and infelicitous

furiously sleep ideas green colorless

finite-state machine

an exact formal model for the representation of language

long-distance dependencies

THE APPLES that alice picked ARE red

probability

the laws governing the likelihood of events

bayes rule

the probability of b given a is equal to the probability of a given b times the probability of b divided by the probability of a

black box

something we can’t see inside of

the information processing approach

generate testable predictions, measure behavior, analyze results, evaluate with respect to theory

donders

how long does it take to make a decision?

decision time equation

RT2 - RT1 = (time to perceive light + time to decide + time to press button) - (time to perceive light + time to press button)

parallel search

look at memory all at once then answer yes or no

serial self-terminating search

look at each individual memory, end when answer is found

serial exhaustive search

look at each individual memory, answer cannot be concluded until end

sternberg

applied the subtraction method for memory search. found support for serial exhaustive search

deduction

by logic, infallible

induction

by generalization, fallible but often works

shepard test conditions

class 1 - only need to track one attribute; class 2 - need to track two attributes; class 3 - each has one exception; class 4 - essentially arbitrary

category

a collection of things regarded as having shared characteristics

concept

hypothesis or description about the shared characteristics of a category

definitional accounts

each feature is individually necessary and jointly sufficient to assign an observation to a category

superordinate

more general

subordinate

more specific

individually necessary

each feature must be present

jointly sufficient

it is enough for each feature to be simultaneously true

the family resemblance problem

the smith brothers. the one with the most family features is most typical

exemplar theory

newly encountered example brings to mind the most similar stored exemplar. individual instances

prototype theory

newly encountered example brings to mind the prototype of that class. idealized example

what is linguistics?

the science of language

phonetics

concrete speech sounds

phonology

abstract organization of sounds

morphology

organization of words

syntax

organization of utterances

semantics

literal meaning of utterances

pragmatics

meaning as part of the discourse

voiceless

vocal folds don’t vibrate: p

voiced

vocal folds vibrate: z

stops

full constriction: b

fricatives

partial constriction: f

phonemes

sound categories

minimal pairs

words that differ by only one phoneme

productive

can apply to new instances. allows language to be infinite (plural -s)

unproductive

only applied to a fixed list of words, do not generalize (counterclockwise)

over-regularization

a child applies a grammatical rule too broadly to an irregular word

over-irregularization

DEFINE THIS

u-shaped learning trajectory

children memorize past forms, children learn a +ed and overapply it, children learn restrictions on the rule

the wug-test

what pattern will a participant extend to a word they haven’t heard before?

dual route models

regulars and irregulars are represented/processed differently. regulars are productive rules, exceptions are minor rules or memorized

single route models

regulars and irregulars are represented/processed the same way. no fundamental difference, it’s just a matter of frequency

symbolic approach (pinker)

minds work like computer programs, math, and logic with symbols and rules. an application of serial search

if-then-else (pinker)

if verb is used in the past tense then: 1. if it’s irregular, then look up the irregular form; 2. else add -ed, the default rule

connectionism

a recurring trend in computation cognitive science, based on ANNs. what if we could model behaviors by modeling neurons directly?

continuous

an acoustic or visual signal

discrete

exists at multiple levels and is governed by combinatory rules

the production pipeline

step 1: prepare a message; step 2: select words and frames; step 3: positional processing; step 4: articulation

anticipation slip

using a sound/word you will need later too early (leading list, reading list)

exchange slip

swapping two sounds/words (spear bill, spill beer)

perseveration slip

keeping a sound/word around too long (beed needle, beef noodle)

syntactic accessibility (bock)

a syntactic priming experiment. the sentence read before being shown a picture impacts the next phrases structure

word recognition

going from sound to meaning

homophone

same sound but different meaning

cohort theory of spoken word recognition

hearing the first phonemes of a word triggers all the words that start that way

uniqueness point

the point at which a word can be distinguished from all others

ambiguous sentences

sentences that have two different meanings (i saw the man with a telescope)

low attachment

the inaccurate interpretation of the sentence

high attachment

the accurate interpretation of the sentence

garden path sentences

sentences that lead you down the wrong path, you need to backtrack and correct (the horse raced past the barn fell)

modular prediction

garden path should happen regardless of semantic/pragmatic context

interactive predicton

garden path should be avoidable with semantic/pragmatic context