TB6- Sentences and Meaning

Question 1

homopone=

homograph=

Answer 1

homopone= same sound, diff meaning
homograph= same spelling, diff meaning

Question 2

Piantodosi ambiguity and language

Answer 2

ambiguity makes language more effective by MINIMISING SPEAKERS EFFORT. ambiguous words are shorter, more common and fewer sound combos

Question 3

Swinney - where does activation spread?

Answer 3

similar words in semantics. If irrelevant meaning is activated, spreads to similar words to that too.
measured by SEMANTIC PRIMING- all lexical candidates activated at first then context comes in later (after a 3 syllable delay irrelevant words been suppressed)

Question 4

what does swinneys model not account for?

Answer 4

frequency of words

Question 5

ambiguous words take longer to read

Answer 5

tf multiple meanings are activated and compete

Question 6

if freq and context are in favour of one…

if mixed…

Answer 6

other word rep is suppressed

if one favour of one and one the other, competition due to equal activation eg context favours subordinate meanings

Question 7

what are noun phrases NPs, and constituents?

Answer 7

NP- moves as a unit and includes a noun (person/thing) and modifier
Constituent- words clumped into larger unit, reflected by pauses.

Question 8

what 2 things are structure, and what do these TF allow for?

Answer 8

structure is generative (can recognise and generate new egs) and hierarchical (reflects constituents who work with others to form larger ones)
allows for recursion.

Question 9

is recursion universal?

Answer 9

no, some languages have limits on sentence length.

Question 10

what do shadowing tasks show about incrementality and meaning processing?

Answer 10

show incrementality exists, and when you replace words with nonwords they repeat more slowly TF processing meaning

Question 11

what is a reduced relative clause?

Answer 11

grammatical structure with a relative clause where some function words removed, leading to ambiguity eg the horse (that was) raced past the barn fell
(another relative clause is can i have the pencil THAT I gave you?)

Question 12

how can you detect GPE in sentence reading?

Answer 12

time how long it takes to read the disambiguating part compared to the same sentence with function words included

Question 13

issue with reading time tasks where they press button for new part? self guided

Answer 13

unatural TF moving window paradigm introduced.

spillover effects- press next but didn’t fully understand tf take longer reading next part. can also use eyemovements

Question 14

garden path theory Frazier?

Answer 14

only structure considered and parser only computes one structure and meaning
incremental and modular.
SYNTAX THEN SEMANTICS

Question 15

constraint based approach

Answer 15

multiple interpretations of an ambiguous structure are simultaneously evaluated against constraints. PARALLEL.
SEMANTICS/CONTEXT INFLUENCES EARLY INTERPRETATION. can ramp up or suppress activation using these constraints.

Question 16

crits of GPT

Answer 16

relative clause sentences can be harder/easier to read with similar strutures.
reduced relative clauses dont always cause issues of interpretation- MacDonald

Question 17

has all the evidence been considered with garden path sentences according to GPT and constraint theories?

Answer 17

constraint- parser HAS considered all the evidence, just weighted in wrong direction
GPT- limits of understanding

Question 18

what predicts the difficulty of ambiguous sentences?

Answer 18

thematic relations associated with verbs eg how many Ps/roles- constraint says have access to this, reduced RT when thematic info strong
syntactic frames of verbs- constraint says parser accesses the semantic frames early on, but GPT that it just notices theyre verbs
lots of evidence parser takes into account statistical probability of frames

Question 19

evidence syntactic frames influence GPEs

Answer 19

trueswell and kim- self paced reading task with ambiguous sentences and priming verb subliminally to favour one type of frame. IDENTITY OF PRIME SHIFTED EXPECTATIONS AND AFFECTED RTs. semantic bias for one word can spread activation

Question 20

are syntactic and lexical ambituity discrete?

Answer 20

syntactic can come from lexical TF not as discrete as thought (lexical factors influence syntactic ambiguity interpretation)- mcdonald

Question 21

is there a frequency bias with GPS?

Answer 21

yes- maybe reduced relative clauses are harder because their structure is less frequent

Question 22

does context reduce GPEs?

Answer 22

yes, we can recruit it quickly to avert GPEs

Question 23

Tanenhaus apple and towel experiment

Answer 23

evidence for GP interpretation due to eye movements to the inappropriate towel. no gpe when context appropriate to the normally less preferred structure of garden path sentence.

Question 24

what do japanese case markers tell us about predictivity of sentences?

Answer 24

predictive language is sensitive to linguistic subtleties like casemarkers
brainwaves and eyetracking also show predictive processing

Question 25

ERP data N400 Delong

Answer 25

N400 seen for unpredictable words, evidence of it at the PREVIOUS WORD, so shows we predict words. surprisal- Hale (inversely correlated to predictability of an event/continuation) knowledge of language correlates with speed of predictions and vocab size (Borovsky)

Question 26

Grice- conversational implicature

Answer 26

extra part of meaning reflecting their intended meaning above the linguistic code hard meanings- linguistic code soft meanings-inferring they arise when we use Grices maxims to infer and implicatures are reasoned out, not automatic

Question 27

scalar implicature

Answer 27

a vague expression deliberately used rather than the stronger version, conveying the stronger version is untrue. children do from 6 perhaps. most 7-11s didnt but when given in a game many 3 year olds do (shulze)

Question 28

does conversational implicature take processing time/effort?

Answer 28

yes eg bott and noveck- pragmatic preferred to semantic but RTs slower and more errors are made when you instruct people to make pragmatic

Question 29

DeNays & Schaeken dot pattern harder > choose

Answer 29

FEWER pragmatic interpretations (choose semantic/literal)

Question 30

ferreira- do speakers try to avoid ambiguity?

Answer 30

when all unrelated gave single noun 65% time, when 2 bats much better at giving info. lexical ambiguity (2 types of bat) > bare nouns 40% time tf different ambiguities tap into different stages of speech production with nonlinguistic earlier than linguistic ambiguity

Question 31

arnold- do speakers try to avoid ambiguity when attaching prepositional phrases?

Answer 31

ambiguity did NOT affect syntactic choices but relative weight did-speakers choose whatever is easiest for speaking. ferreira and dell- same result, when recalling from memory decision to insert "what" was not affected by ambiguity but was by accessability of pronoun

Question 32

BROWN AND DELL feedback from hearers

Answer 32

speakers mentioned instrument 2x as much when unpreictable, but in followup speakers didnt take listeners perspective into account (issue-backchannel feedback in body language)

Question 33

keysar

Answer 33

all automatic language processes are WITHOUT AN EGOCENTRIC FRAME OF REFERENCE and fast general egocentric processes separated from slower reasoning/mindreading

Question 34

swinney multiple access model

Answer 34

meanings accessed first, THEN context- autonomous/modular support- Swinney priming found both meanings activated in lexical decision at first Tanenhaus- priming for verb and nonverb based on interpretation and later only the contextually appropriate one

Question 35

Schvanenvelt- context guided single reading lexical access

Answer 35

interactive model with EARLY CONTEXT effects support-Tabossi and Zardon- dont always activate all meanings (dominant meaning for subordinate) Lucas metaanalysis- weaker interactive models say sentences BIAS ACTIVATIONS so some interaction

Question 36

Ferreria- are speakers better at avoiding linguistic or nonlinguistic ambiguity?

Answer 36

better at avoiding nonlinguistic which is evidence of common ground

Question 37

2 theories for indirect speech

Answer 37

1- 2 stage mechanism with literal THEN pragmatic- searle | 2- one stage parallel- ONLY LITERAL MEANING CONSIDERED (?)

Question 38

incremental parsing- tyler and marslen-wilson

Answer 38

FASTER responses for APPROPRIATE context, supports incrementality

Question 39

garden path theory assumptions

Answer 39

``` Serial Autonomous Incremental Innate (only incremental is supported still) ```

Question 40

cuetos and mitchell

Answer 40

not innate like GPT says as spanish prefer early closure. | gibson- memory can be a constraint

Question 41

does animacy influence GPEs?

Answer 41

eye tracking shows GPE bigger for animate. a semantic effect so against GPT

Question 42

Tanenhaus

Answer 42

visual content of sentences influenced likelihood of GPE (language and visual world interact) TF supports referential context and not GPE

Question 43

Referential context

Answer 43

altmann and steedman- 1 safe vs 2 safe context. MINIMAL ATTACHMENT SLOWER WITH 2 SAFES AGAINST GPT. referential theory needs to specify. -incremental -context affects selection -context can override GPE

Question 44

simon effect

Answer 44

longer RT for incongruent patch. | about response congruency

Question 45

did bilinguals or monolinguals do better in simon effect incongruent condiiton? Bialystok & Craik,

Answer 45

BILINGUALS DID BETTER and even bigger gap with age Better “executive control” – more able to ignore/inhibit/suppress irrelevant information – often related to frontal lobe function

Question 46

advantages of being bilingual?

Answer 46

executive control eg stroop ambiguous interpretation cognitive reserve flanker task-response inhibition

Question 47

disadvantages of being bilingual?

Answer 47

vocabulary slower lexical access in picture naming worse at category fluency (but better letter)

Question 48

Emmorey et al- unimodals or bimodals faster at flanker/ go no go?

Answer 48

unimodals faster but no difference in accuracy | (bimodals and monos same) suggests advantage in cognitive control comes from swapping languages in the same modality

Question 49

weaknesses to Emmorey study?

Answer 49

AoA not controlled well-5y diff attention couldve played a part due to position of chevron perception- bimodals do 2 at once modularity seen but not as much as fodor suggests

Question 50

what does emmorey study tell us about cognitive abilities of bilinguals?

Answer 50

``` It’s not enough to know two languages – They need to be either: • particularly similar • or you need to be practised at switching between them – code switching vs. code-blending ```

Question 51

what does studying bilinguals tell us about plasticity/modularity?

Answer 51

Language is not a modular, encapsulated system as Fodor as argued – it makes use of general cognitive resources • Long-term repeated practice with a language can have a significant impact on our general cognitive abilities – Suggests a great deal of plasticity