Experimental Psycholinguistics Flashcards
(308 cards)
1
Q
Linguistics
A
The scientific study of language and its structure.
2
Q
Psycholinguistics
A
The psychology of language. The study of the relationships between linguistic behaviour and psychological processes.
3
Q
Neuropsychology
A
The study of the relationship between behaviour, emotion and cognition on the one hand, and brain function on the other hand.
4
Q
Syntax
A
The rules of word order of a language.
5
Q
Lexicon
A
The mental dictionary.
6
Q
Inflection
A
A grammatical change to a verb or noun, changing tense or number.
7
Q
Broca’s area
A
A region of the brain concerned with the production of speech, located in the frontal lobe.
8
Q
Wernicke’s area
A
A region of the brain concerned with comprehension of speech, located in the temporal lobe.
9
Q
Model
A
An account that explains the data we’ve collected, but which goes beyond it.
10
Q
Falsification
A
Proving a theory wrong by observing a counterexample.
11
Q
Nativism
A
The idea that knowledge is innate.
12
Q
Empiricism
A
The idea that all knowledge is based on experience derived from the senses.
13
Q
Modularity
A
The idea that the mind is built up from discrete modules, sometimes said to be corresponding with identifiable neural structures in the brain.
14
Q
Interactivity
A
The idea that different modules involved in language can communicatie with and influence each other.
15
Q
Domain specificity
A
The idea that different aspects of cognition are built upon specialized learning devices.
16
Q
Pragmatics
A
The study of language in use and the contexts in which it is used.
17
Q
Segmentation
A
Splitting speech up into constituent phonemes.
18
Q
Specific Language Impairment (SLI)
A
A developmental disorder affecting just language.
19
Q
Poverty of the Stimulus
A
The argument that children are not exposed to rich enough data within their linguistic environment to acquire every feature of their language.
20
Q
Linguistic Universals
A
A pattern or feature that occurs across all natural languages.
21
Q
Parameter
A
A component of Chomsky’s theory that governs aspects of language and is set in childhood by exposure to a particular language.
22
Q
Universal Grammar
A
The core of grammar that is universal for all languages, and which specifies and restricts the form that individual languages can take.
23
Q
Critical period
A
The period after which language acquisition is much more difficult and less succesful.
24
Q
Aphasia
A
A disorder of language, including a defect or loss of expressive or receptive aspects of written or spoken language as a result of brain damage.
25
Cortex
The outer layer of the cerebrum, composed of folded grey matter and playing an important role in consciousness.
26
Lateralization
The tendency of neural functions to be specialized to one side of the brain.
27
Hemisphere
Half of the brain.
28
Lobes
Divisions of the cortex based on the locations of the gyri and sulci. The brain has six lobes.
29
Name the 6 lobes of the brain:
Frontal, temporal, parietal, occipital, insular and limbic.
30
Neuroimaging
The process of producing images of the structure or activity of the brain by different techniques.
31
Dichotic listening
A psychological test in which multiple auditory messages are presented at the same time.
32
Right ear advantage
The observation that when two different speech stimuli are presented simultaneously to both ears, listeners report stimuli from the right ear more correctly.
33
Paraphasia
Spoken word substitution, words are mixed up.
34
Speech errors
A deviation from the apparently intended form of the utterance.
35
Substitution error
The automatic replacement of one item in a sentence, word or phoneme when forgotten or unknown.
36
Speech planning
The process that speakers do not plan a whole utterance, but continue planning while they are articulating.
37
Intrusion error
A speech error that occurs because of thinking something whilst speaking.
38
Freudian slip
A speech error that occurs when the surpressed subconscious surfaces.
39
Lexicalization
The process of going from meaning to sound in speech production.
40
Picture word interference
A paradigm where participants have to name pictures while hearing words. The time of the auditory signal compared to the appearance of the picture varies.
41
Cascade processing
A model in which processes further down the line are already activated whilst the first are still completing.
42
Feedback
A model in which the later levels of processing can feed back and affect earlier levels.
43
Lexical bias
The tendency of speech errors to result in existing words at a higher rate than would be predicted by chance.
44
Similarity bias
The finding that speech errors are more likely to occur the more similar the target and intrusion sound are.
45
Semantic priming
The idea that words are easier to process when the preceding words are related in meaning.
46
Syntactic priming
The finding that the syntactic structures of speakers can be influenced by exposing them to sentences with a particular structure.
47
Mediated priming
When a target word is activated that relates indirectly to the given word, through an intermediate word.
48
Tip of the tongue
When you know that you know a word, but you can not immediately retrieve it.
49
Incremental processing
A model of processing where information is directly integrated with earlier information as soon as possible.
50
Parsing
Analysing the grammatical structure of a sentence.
51
Global grammatical ambiguity
A sentence that has at least two interpretations.
52
Local grammatical ambiguity
A sentence that has a phrase that is ambiguous, but the whole sentence is not ambiguous.
53
Garden path sentence
A grammatical correct sentence that starts in a way that the perceptor's first interpretation will be incorrect.
54
Reanalysis
A change in a word or phrase resulting from replacing an unfamiliar form by a more familiar form.
[when something is heard wrong]
55
Minimal attachment
The strategy where the syntactic structure with the least nodes is built, the simplest.
56
Late closure
The principle that new words tend to be associated with the phrase currently being processed instead of structures further back in the sentence.
57
ERP
Event Related Potential
58
Event Related Potential
Electrical activity in the brain after a particular event. Measured by EEG.
59
Constraint based model
Theory that all possible interpretations of a sentence are activated and the most probable is then selected.
60
phoneme
A sound of the language: changing a phoneme changes the meaning of a word.
61
allophone
A realization of a phoneme: does not contribute to differences in meaning.
62
speech gesture
Non-verbal communication in which visible bodily actions communicate particular messages.
63
vowel
A speech sound with very little construction in the airstream.
64
consonant
A sound produced with some obstruction of the airstream.
65
source
The thematic role where the theme is moving from.
66
filter
The learner's attitudes that affect the relative success of second language acquisition.
67
voicing
consonants produced with vibration of the vocal cords.
68
pitch
The relative low/highness of voice. Determined by the frequency of the vibration of the vocal cords.
69
fundamental frequency
The lowest frequency that is produced by the oscillation of the whole of an object
70
harmonic
A component frequency of an oscillation.
71
prosody
The melody of speech, to do with duration, pitch and loudness.
72
spectrum (of sound)
The amount of vibration at each individual frequency
73
Formant
Each band of frequency that determine the phonetic quality of a vowel.
74
bottom-up processing
Purely data-driven.
75
top-down processing
Processing that involves knowledge coming from higher levels.
76
Lexical access
Accessing a word's entry in the lexicon.
77
Lexical decision task
Press a key when you identify a word in the string of letters passing on the screen, and another key if it is a non-word.
78
Lexical decision
Deciding whether a word is familiar or not without accessing the meaning.
79
Naming task
Name the word that you see on the screen. The time it takes to start speaking, is measured.
80
cross-modal priming
Participants have to listen and watch.
81
Uniqueness point
A point, measured from the beginning of the word, from which the word is phonetically different from all other words in the lexicon.
82
cohort
All the lexical items that share an initial sequence of phonemes (thus activated by that sequence)
83
Gating
A task where the sound of a word is progressively more revealed.
84
Pseudo-words
A string of letters that forms a pronounceable non-word.
85
Non-word
A string of letters that does not form a word.
86
Dyslexia
Disorder of reading.
87
Homophone
Two words that sound the same.
88
Lexical neighbors
Words that can be formed with the target word through deletion, addition or substitution of a single phoneme.
89
Eye-mind assumption
Says that the eye keeps fixating on a word until it is completely processed.
90
statistical model
A mathematical model which is modified/trained by the input of data points.
91
Mathematical model
Specifies a relation among variables in either input to output or (x,y) relation.
92
Probabilistic model
Specifies a probability distribution over possible values of random variables.
93
Trained model
Takes a collection of possible models as input and a collection of data points and selects the best model.
94
on-line technique
Measure variables that tap into language processing as it happens.
95
off-line technique
Measure variables related to the subsequent outcomes of processing.
96
incrementality
The phenomenon that each word in the sentence is interpreted immediately when it is encountered.
97
encapsulation
The extent to which different knowledge sources are formally separated. Critical property of modularity.
98
predicate proximity
A preference to attach as close to the head of a predicate phrase as possible.
99
Click location task
Participants listen to an utterance and have to point at the place of a click that occurs somewhere in the utterance.
100
Dichotic switch monitoring
Participants have to indicate when the speech moves from one earphone to another.
101
What does fMRI measure?
Blood flow in the brain.
102
What do PET-scans measure?
Emissions from a radioactive substance that is injected in the bloodstream.
103
What is a CAT-scan?
An X-ray taken from different angels.
104
What methods measure electrical activity?
MEG and EEG.
105
SAT
Speed-Accuracy Trade-off
106
Speed accuracy trade-off task
Participants have to response to a linguistic stimulus as soon as they hear a tone after the stimulus. Longer interval between stimulus and tone -> less errors.
107
Priming techniques
Research the effect of processing a prime item on the processing of a target item.
108
Cross-modal priming
Priming where the prime item is (for example) spoken and the target item is (for example) written.
109
Gating technique
Establishes when listeners interpret words in relation to information in the speech stream, and whether the uniqueness point predicts recognition time.
110
What does eyetracking tend to show?
That comprehension is incremental and immediate.
111
RSVP
Rapid Serial Visual Presentation
112
Rapid serial visual presentation
Presents sequences of words at a fixed fast rate. Participants have to do a second task while reading the words.
113
Saccades (in RSVP)
Fast eye movements from one letter or word to another.
114
Regressions (in RSVP)
Saccades moving right to left instead of left to right.
115
Moving window technique
When the window of text moves depending on where the reader is fixating.
116
What does the moving window technique show?
That readers only take in information from a small region of text and try to integrate everything they read with the information processed.
117
Implicit priming task
Participants have to learn pairs of two words. In one condition, all 2nd words start with the same syllable and this speeds up articulation.
118
Referential communication task
A task to study dialogue.
P1 describes visual patterns and P2 has to identify the right pattern from a set.
119
EEG
Electro-encephalography
120
What does an EEG measure?
Electrical activity at the scalp.
121
MEG
Magneto-encephalography
122
What does an MEG measure?
Changes in magnetic fields associated with electrical activity in the brain.
123
+ and - of fMRI
Precise about the location, but not about the time course.
124
+ and - of ERP (EEG)
Precise about the time course, but difficult to find the source of the activity.
125
+ and - of MEG
Good localization and temporal resolution, but only sensitive in certain neural parts.
126
N400
A peak in the negative change around +- 400ms after stimulus. Associated with integration of words into the sentential context.
127
P600
The positive change around +- 600 ms after stimulus, associated with syntactic integration processes.
128
Why is ERP mainly used for immediate processes?
It gets noisy after sometime.
129
What fields are the five questions of psycholinguistics in?
Development, structure, production, comprehension and attrition.
130
What are the two main approaches in cognitive sciences?
Generativism and connectionism.
131
Serial account
Says that people initially adopt one analysis and reanalyze when inconsistency is realized.
132
Ranked parallel account
Says that people adopt more than one analysis, but rank one higher than others. Difficulty occurs when they have to rerank.
133
Garden-Path theory
Assumes serial processing.
134
Active-filler strategy
Minimal chain principle, deals with unbounded dependencies and assumes that analysis with gap filling is preffered over analysis without.
135
Interactive accounts
Assume all potentially relevant sources of info can be used immediately and affect earlier processing decisions. = constraint-based
136
Construal theory
Claims that structural parsing principles operate for primary syntactic relations (with arguments) and that the processor uses non-syntactic information for non-primary relations.
137
Referential theory
The principle of parsimony.
138
The principal of parsimoiny
The processor accepts the analysis that requires the fewest unsupported presuppositions.
139
Revision-as-last-resort principle
Says that reanalysis only happens when no other option is possible. This is incorrect for correct grammatical sentences.
140
DNN
Deep Neural Networks
141
Artificial neural networks
Mathematical objects that compute functions from one sequence of real numbers to another sequence, using different layers of neurons.
142
Word embeddings
Distributed representations of words in vectors.
143
RNN
recurrent neural networks
144
Recurrent neural network
Processes the sentence from left to right, maintaining a single vector h.t (the first t words of a sentence).
145
Gated RNNs
Possess a mechanism that allows them to better control the extent to which information in their hidden state is updated after each word.
146
LSTM
Long short-term memory networks
147
Attention (in DNN)
Relaxes the hidden-state bottleneck and allows the network to take into account all the previous hidden states when computing the next state.
148
Transformer
A processing architecture that relies on attention only to carry information across time. Can be trained effectively on very large corpora.
149
In what three common settings are DNNs used?
1) Classifier
2) Language model
3) Sequence-to-sequence setting
150
How is a DNN used as a classifier?
The network outputs a discrete label for a sequence. A supervised setup trained on a corpus of sentences.
151
How is a DNN used as a language model?
The network receives the first n words of a sequence and assigns a probability to each word that could come up next. Unsupervised, only a text corpus needed.
152
How is DNN used in the sequence-to-sequence setting?
The network generates an output sequence for each input sequence, configured by chaining two RNNs. Requires pairs of input and output sequences, but no further annotation.
153
Number prediction task (DNN)
The DNN is exposed to a sentence prefix in which the word following the prefix is expected to be a verb and is asked to predict whether the verb that comes next should be singular or plural.
154
Attractor
Nouns that intervene between the head of the subject and the verb.
Example: 'the length of books'. Books = attractor.
155
Large Language Models (LLMs)
Large computational models that try to guess words in their context, using probabilities. They are non-modular.
156
Neural substrate
The part of the central nervous system that underlies a specific behavior, cognitive process or psychological state.
157
ICE
Intercranal electrophysiology
158
LFP
local field potentials
159
What has been the method to test the claim of lexical selection by competition?
The picture-word interference paradigm.
160
REH
Response exclusion hypothesis
161
Response Exclusion Hypothesis (REH)
States that lexical selection does not depend on the number and activation of potential competitors
And that spreading of activation only has a facilitative effect.
162
What does it mean when a model is called discrete?
They make the claim that the stages of word retrieval or lexical access are operating in strict temporal succession: first lexical-semantic and then phonological word forming.
163
AoA
Age of Acquisition
164
The age-of-acquisition effect
Words and pictures with earlier learned labels are processed faster than words and pictures with later learned labels.
165
Word frequency effect
When a word that occurs highly frequently in a language is processed faster or more accurately than those that are less frequent.
166
Which one is stronger: the AoA effect or frequency effect?
AoA
167
Why do some researchers dismiss the effect of AoA in lexical processing?
They say it is just another measure of word frequency.
168
What two types of multiple regression have been used in studying AoA?
Simultaneous multiple regression and stepwise multiple regression.
169
Simultaneous multiple regression
Multiple predictors are entered into the regression equation at the same time.
Each predictor's unstandardized regression coefficient represents the change in dependent variable when the predictor is changed by 1.
170
Multicollinearity
When multple independent variables in a model are correlated.
171
Stepwise multiple regression
Variables are entered into the regression equation in a specific order
172
Factor analysis
Intercorrelations between variables are examined as a measure of association.
173
Name 5 theories that have been proposed to account for AoA effects.
1) the bilateral representation hypothesis
2) the phonological completeness hypothesis
3) the cumulative frequency hypothesis
4) the semantic locus hypothesis
5) the lexical-semantic competition hypothesis.
174
The bilateral representation hypothesis
Says that AoA effects exist because early acquired words are represented in both hemispheres and later acquired words only in the left hemisphere.
Not supported.
175
The phonological completeness hypothesis
Earlier learned words are easier to pronounce than later acquired words. The later ones have to be generated every time whilst the early words are stored as an entity.
176
The cumulative frequency hypothesis
Early words have been encountered more times than later words. Predicts whether the number of encounters predicts processing time.
177
The semantic locus hypothesis
Early acquired words have more connections in the semantic systems than later acquired words. Search for lexical items might be biased towards more/better connections.
178
The network plasticity hypothesis
AoA effects should be larger when the mapping from input to output is arbitrary (mapping hypothesis)
179
The lexical-semantic competition hypothesis
Size of frequency effect & AoA effect are equivalent and correlated. Says they stem from the same learning process.
180
Immediate naming
A word is presented on a computer screen and the participant has to name the word aloud as quickly and accurately as possible.
181
Speeded naming
Like immediate naming, but with a deadline.
182
Delayed naming
Immediate naming but with a delay.
183
What is found to be the best predictor for naming latency for noun?
word length
184
Have AoA effects been found in delayed naming?
no
185
LDT
lexical decision time
186
visual lexical decision task
Letter strings are presented on a computer screen and participants need to decide whether they make up a word or not.
187
Self-monitoring
The processes that speakers use to monitor their own speech and intervene when necessary.
188
Theory of mind
The idea/conscience that other people's ideas and stuff are different than your own.
189
Why do only humans have language? (g vs c)
G: language module
C: bigger/better brain
190
How do we learn a language? (g vs c)
G: innate rules
C; statistical learning
191
How do we produce words? (g vs c)
G: stepwise (concept, meaning, form)
C: parallel (all levels activated together)
192
How do we parse sentences? (g vs c)
G: Incrementally (word by word)
C; Competing structures
193
What is lost in aphasia? (g vs c)
G; Specific knowledge
C: general ability
194
Self-paced reading
Participants read a sentence one word at a time, at their own pace. The reading times per word are recorded.
195
What words are seldom fixated on during reading?
Function words & short words (2-3 letters)
196
Foveal
A region in the visual field, representing the center 2 degrees.
197
Parafoveal
A region in the visual field, representing the 5 degrees next to the foveal area.
198
Peripheral
The outer region of the visual field.
199
Visual world paradigm
Participants look at a visual scene constructed by the experimenter while they listen to audio. Eye movements are tracked.
200
Surprisal theory
Says that reading time (RT) is proportional to the surprisal of the word in the context (the P that the upcoming word is not predicted in the context).
201
Agreement attraction (error)
when a verb gets a wrong agreement form because of an attractor
202
Referential theory of meaning
The meaning of a word is the object in the world to which it refers. (critique: 'empty references' are still meaningful)
203
Image theory of meaning
Meanings are 'mental pictures' associated with the word. (critique: more detail in pictures, abstract words_
204
Definitional theory of meaning
Meaning of a word is its definition
205
Family resemblence (theory of meaning)
A word is defined by its relations to other members of a set.
206
Prototype theory of meaning
The set is defined by its memvers and prototypical elements. (critique: not every set has a prototype)
207
Globules
Theory of ML where a word is a bundle of features. Related words have atoms/fields in common.
cancelled.
208
Wordwebs
Theory of ML where a word is defined by the words that are connected to it.
209
Free association
A concept where the participant has to name everything that comes to mind, without censorship.
210
coordination relation (words)
Words that cluster together as a group (ex. colors)
211
Collocation relation (words)
Words that can be found together
212
Super-ordination relation (words)
Words that define the group of the stimulus word. Ex. red-color.
213
Graph
Underlies every semantic network, consists of a set of nodes/vertices and a set of edges/arcs.
214
Arcs
Joins a pair of nodes, directed.
215
Edge
An undirected link between two nodes.
216
Undirected graph
A graph consisting of only edges.
217
directed graph
A graph containing only arcs.
218
neighborhood (in semantic networks)
Á subset of nodes consisting of some node and all of its neighbords.
219
in-degree (of a node)
the number of arcs going in that node.
220
k.i (in directed networks)
The degree of node i.
221
Path (in undirected graph)
A sequence of edges that connect one node to another.
222
Path (in directed graph)
A set of arcs that can be followed along the direction of the arcs from one node to another.
223
Path length (semantic networks)
The number of edges or nodes along that path.
224
Strongly connected graph
A directed graph where a path exists along the arcs between any pair of nodes.
225
L (semantic networks)
The average distance
226
D (semantic networks)
Diameter
227
C (semantic networks)
clustering coefficient: the probability that two neighbors of a random node are themselves neighbors.
228
P(k) in semantic networks
degree distribution
229
formula for C
C.i = T.i / (k.i / 2).
Met T.i = the number of connections between the neighbors of node i.
230
Small-world structure
A combination of short average path lengths L and relatively high cluster coefficients C.
231
bipartite graph
A graph with two different kinds of nodes: word and semantic category, with connections allowed only between nodes of different kinds.
232
preferential attachment (network models)
A kind of network growth process in which nodes are added to the network successively by connecting them to a small sample of existing nodes selected with probabilities proportional to their degrees.
233
On the basis of what three principles are new nodes added probabilistically to existing nodes?
1) Differentiation
2) P(differentiating a particular node) is proportional to its current complexity (how many connections it has)
3) Utility variation
234
Differentiation (principle of semantic network growth)
The idea that the meaning of a new word or node typically consists of some kind of variation on the existing word
235
Utility (semantic network growth)
Modifies the probabilitiy that a node will be the target of new connections.
236
LSA
A machine learning model that makes a high-dimensional semantic space based on how words are used in passages. Static and distributed representation.
237
Localist representation
Each concept is associated with a single distinct node in the semantic network.
238
Lemma
Semantic and syntactic info
239
Lexeme
phonological form
240
SOA
Stimulus onset asynchrony
241
stimulus onset asynchrony
The time between the onset of the picture and the auditory-presented word.
242
SA
speech sensorimotor adaptation
243
Speech sensorimotor adaptation
An alteration of the performance of a motor task that results from the modification of sensory feedback.
244
Compensation
A response to a feedback perturbation that is in the direction opposite to the perturbation.
245
Adaptation (Villacorta)
Compensatory responses that persist when feedback is blocked or when the perturbation is removed.
246
proprioceptive feedback
feedback pertaining to limb orientation and position and visual feedback.
247
RSA model
Rational Speech Act model
248
Conversational maxims
A set of principles described by Grice as a theory about how people think about speaker's intended meaning and so arrive at pragmatic implications.
249
Implicature
An inference about the meaning of an utterance in context that goes beyond its literal semantics.
250
Rational speech act model
A class of probabilistic model that assumes that language comprehension in context arises via a process of recursive reasoning about what speakers would have said, given a set of communicative goals.
251
Scalar implicature
arises when a speaker did not use a stronger alternative term, leading to narrowing of the interpretation.
252
Social recursion
Reasoning that involves two people thinking about each other
253
uncertain RSA models
Extension of RSA models that allows for joint inferences about both the speaker's intended meaning and other aspects of the interaction.
254
Speech-shadowing task
Subjects wear headphones with a stream of speech. They need to repeat as they hear it.
255
the average degree
256
[gamma] in network models
The power-law exponent for the degree distribution.
257
random graph (network models)
A network where each pair of nodes is joined by an edge with probability p
258
Scale-free network
A network with a degree-distribution that is power-law distributed.
259
What structure do almost all natural networks have?
Scale-free / small-world
260
What kind of network is the mental lexicon?
scale-free
261
Articulation stage in Levelt's model
From phonetic plan to overt speech
262
Motor planning (stage)
The connection between speech sounds and required articulatory movements, on the level of individual articulators.
263
Motor programming (stage)
Specifies timing and amount of muscle movements necessary for production
264
Motor execution (stage)
Sends the neural signals to peripheral systems & transforms into coordinated muscle activity.
265
Internal self-monitoring
Used during motor planning to avoid wrong planned speech movements being executed.
266
External self-monitoring
Comprises fast somatosensory feedback & slow auditory feedback, is used for adaptation of articulation and error correction.
267
How does auditory feedback play an important role in the production of speech sounds?
It is a teaching signal for the acquisition and adaptation of speech motor programs.
And it is a guiding signal for the online control and correction of speech movements.
268
Feedback control system
1) Auditory & somatosensory feedback is compared to the goal.
2) Error = difference between feedback & goal.
3) Error is used to generate motor command to change state of feedforward control system.
269
Feedforward control system
Motor commands are instantiated as preplanned trajectories and executed by the articulatory system in a time-locked manner.
270
DIVA
Directions into velocities of articulators
271
What do neurocomputational models for speech contain?
Feedback controllers and and feedforward controller.
272
SimpleDIVA
A 3-free-parameter computational model that estimates contributions of feedback and feedforward control mechanisms.
273
What 3 subsystems does SimpleDIVA model?
1) Auditory feedback control (alpha.A)
2) Somatosensory feedback control (alpha.s)
3) Feedforward control/learning rate (lambda.FF)
274
Articulation disorder
Inability to produce certain phones, typically s- and r-sounds.
275
Phonological delay
Speech disorder where error patterns are developmental but delayed.
276
Childhood apraxia of speech
Multi-deficit motor speech disorder.
277
Verbal communication loss can be caused by...
Injuries and neurodegenerative diseases that affect gross and fine motor production, speech articulation and language understanding.
278
ALS
amyotrophic lateral scelerosis
279
LIS
locked-in syndrome
280
Locked-in syndrome (LIS)
Caused by neurodegenerative conditions.
Complete loss of motor control, preventing communication.
281
BCI
Brain-computer interfaces
282
Brain-computer interfaces
An assistive technology that provides a new communication channel for people with LIS
283
How do BCI technologies work?
They create a (bi)directional communication interface which reads the signals generated by the human brain and converts them into the desired cognitive task. It also feeds digital signals back into the brain.
284
What are the 4 steps to use a BCI?
1) Signal production
2) Pre-processing
3) Feature extraction
4) Classification
285
Signal production in BCI use
What are the properties of the signals to be recorded? How should they be captured?
286
Pre-processing step in BCI use
Record, unmask and enhance the information within the signal.
287
Feature extraction step in BCI use
Extraction of the main characteristics of the signal
288
Classification step in BCI use
Signals are classified depending on their features.
289
Communication principle:
Use an expression that is true and more informative than other true alternatives.
290
Rational speech act theory
Communication is a rational update on beliefs. Listener & speaker update probabilities rationally.
291
Why do children fail at pragmatic reasoning?
They lack the theory of mind. BUT: can reason pragmatically tho not aware of the right alternatives.
292
competence
Knowledge of language
293
Neurophysiological methods
Used to identify a state of consciousness and unconsciousness. Techniques measuring activity in the brain.
294
Ecological validity
The ability to generalize study findings to the real world. Subcategory of external validity
295
Controllability
The ability to say that the effect observed is because of the causes you manipulated and not some other effect.
296
Family resemblance
A way to define semantic category membership. Words that are related, are close together in the network.
297
Semantic primitives
Features that desribe how to decompose languages.
298
Meaning atoms
The smallest units in which a language can be defined.
299
Spreading activation
A method for searching associative networks, computes activation valeus for nodes.
300
Blends (speech errors)
Speech errors in which two suitable words fall together and are fused into one word, so both words are partially chosen.
301
Blocking
A morphological phenomenon where a word cannot surface because it is blocked by another form whose features are more appropriate to the surface form's environment.
302
Meta-analysis
Examination from data from a number of independent studies of the same subject, in order to determine overall trends.
303
What is the difference between random, small-world and scale-free networks?
Scale-free is distributed with a power law, so most nodes have few connections and a few nodes have many connections (=hub).
Small-world have local clustering and short pathl lengths between clusters.
Random networks do not have hubs. Low clustering coefficient.
304
Power-law distribution
Relationship between two quantities. Loopt naar een limiet toe. Exponentiele afname of toenamen.
305
Syntax-first
Modular theory of processing in which semantic info is processed at a later stage than syntactic info.
306
Redundancy
When information is expressed more than once, when lesser words can be used.
307
Somatosensory
pressure, pain or warmth in contrast to feelings localized to one organ/place like sight, balance.
308
