PresentationPDF Available

The Influence of Script Knowledge on Language Processing: Evidence from ERPs

Authors:

Abstract

Previous research has shown that the semantic expectedness of a word – as established by the linguistic context – is negatively correlated with N400 amplitude. While such evidence has been used to argue that the N400 indexes semantic integration processes, findings can often be explained in terms of facilitated lexical retrieval, which, among other factors, is influenced by lexical/semantic priming. In the present study we examine this issue by manipulating script event knowledge – a person’s knowledge about structured event sequences – which has been previously shown to modulate the N400. An ERP-study (German) investigated whether N400 modulation by a mentioned script event is due to priming alone, or is further sensitive to linguistic cues which would be expected to modulate script influence.
The Influence of Script Knowledge
on Language Processing:
Evidence from ERPs
Elisabeth Rabs, Heiner Drenhaus, Francesca
Delogu, Matthew Crocker
AMLaP 2017 Sep 9, 2017
One first example
Before the movie starts, I like to get a snack.
There’s nothing like eating a big box of
popcorn < soda < car
(measurement: N400 amplitude)
(paraphrased from
Metusalem et al., 2012)
One first example
Before the movie starts, I like to get a snack.
There’s nothing like eating a big box of
popcorn < soda < car
One first example
Before the movie starts, I like to get a snack.
There’s nothing like eating a big box of
popcorn < soda < car
One first example
Before the movie starts, I like to get a snack.
There’s nothing like eating a big box of
popcorn < soda < car
Script Knowledge vs. Priming?
Lexical
Priming?
Script
Knowledge?
Simple/
automatic
complex/
compositional
Script Knowledge vs. Priming?
Lexical
Priming?
Script
Knowledge?
Simple/
automatic
complex/
compositional
Use of script knowledge influences
processing difficulty of words in
context beyond priming
Script Knowledge vs. Priming?
Lexical
Priming?
Script
Knowledge?
Simple/
automatic
complex/
compositional
Disentangle/quantify influences of script knowledge and priming
Use of script knowledge influences
processing difficulty of words in
context beyond priming
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Make
sentence
more
plausible
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
inactive active
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
active
inactive/
unmentioned
Experiment 1:
Stimuli (2x2 Design, German)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: Instead of going to the post office, she went to the pharmacy.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Context size:
1 script
Vs.
2 scripts
Target:
Fits active script
Vs.
Fits inactive script/
unmentioned
Experiment 1:
Predictions
Active-script fitting objects
easier to process than
inactive-script fitting objects:
Reduced N400 for active-
script fitting conditions
Lexical Priming
Script
Knowledge
Processing of unprimed
objects difficult:
Attenuated N400 for all
conditions compared to
unmentioned one
Experiment 1:
Behavioural Data
Mean ratings:
ØActive-script fit better
rated than inactive-script
fit/unmentioned
ØNo difference between
2-script vs. 1-script
conditions
2 Scripts 1 Script
Fit 1.4 (SD 0.7) 1.4 (SD 0.6)
No Fit 3.4 (SD 0.9) 3.4 (SD 0.8)
1234
Rating: 1=good, 4=bad
Condition
2scripts/
activefit
2scripts/
inactivefit
1script/
activefit
1script/
unmentionedfit
Instead of post officepharmacy / prescription
Instead of post officepharmacy / parcel
Pharmacy / prescription
Pharmacy / parcel
Experiment 1:
ERP Results
300-500 ms:
ØMain effect for active script fit: N400 for Inactive-fit
(2/4) vs. Active-fit (1/3)
ØInteraction with context size: larger N400 for 1-
Script/unmentioned vs. 2-Script/Inactive-fit
200 0 200 400 600 800 1000
6 4 2 0 24
Cz
Instead of post officepharmacy / prescription
Instead of post officepharmacy / parcel
Pharmacy / prescription
Pharmacy / parcel
1 Script,
active-fit vs. unmentioned
2 Scripts,
active-fit vs. inactive-fit
Experiment 1:
Interpretation
200 0 200 400 600 800 1000
6 4 2 0 24
Cz
Lexical
priming
compositional
script knowledge
Instead of post officepharmacy / prescription
Instead of post officepharmacy / parcel
Pharmacy / prescription
Pharmacy / parcel
Script Knowledge
Lexical Priming
Active-script fitting objects
easier to process than
inactive-script fitting
objects:
Reduced N400 for
prescription vs. parcel
Facilitation for inactive-fit
vs. unmentioned:
Inactive scripts are
accessible, can still prime
objects
A possible confound?
Post office pharmacy prescription
Post office pharmacy parcel
ØNo variation in script order: active script always last
ØDistance between script and object small for active script, large for
inactive script
ØRecency Effect: Priming stronger for more recent words, weaker for
more distant words
Experiment 2:
Design & Predictions
Replicate exp. 1, but change order of scripts in context:
“Instead of going to the post office, she went to the pharmacy.”
“She went to the pharmacy instead of going to the post office.”
Experiment 2:
Design & Predictions
Replicate exp. 1, but change order of scripts in context:
“Instead of going to the post office, she went to the pharmacy.”
“She went to the pharmacy instead of going to the post office.”
Script
Knowledge
(+ Priming)
Difference between
2-script conditions
should vanish
no recency effect No change in results
between experiments
Lexical
Priming
recency effect
Experiment 2:
Stimuli (2x2 Design)
Context Sentence:
Introduction: Roberta’s cold had gotten worse.
1 Script: She went to the pharmacy.
2 Scripts: She went to the pharmacy, instead of going to the post office.
Target Sentence:
Active Script: She entered and handed over the prescription with a smile.
Inactive Script: She entered and handed over the parcel with a smile.
Experiment 1
200 0 200 400 600 800 1000
8 6 4 2 0 24
Cz
200 0 200 400 600 800 1000
8 6 4 2 0 24
Cz
Experiment 2
Active-script Fit
vs.
Inactive-script Fit
Experiment 1
200 0 200 400 600 800 1000
8 6 4 2 0 24
Cz
200 0 200 400 600 800 1000
8 6 4 2 0 24
Cz
Unmentioned vs.
2-Script/Inactive-Fit
2-Script/Inactive-Fit vs.
2-Script/Active-Fit
Unmentioned vs.
2-Script/Inactive-Fit
2-Script/Inactive-Fit vs.
2-Script/Active-Fit
Experiment 2
Summary
N400 attenuated by events mentioned in the prior context
Summary
N400 attenuated by events mentioned in the prior context
Script Knowledge Lexical Priming
Summary
N400 attenuated by events mentioned in the prior context
Script Knowledge Lexical Priming
Influence of active
and inactive
events
Summary
N400 attenuated by events mentioned in the prior context
Script Knowledge Lexical Priming
Influence of active
and inactive
events
Attenuated N400 for
words associated
with a mentioned
(active or inactive)
event
Summary
N400 attenuated by events mentioned in the prior context
Script Knowledge Lexical Priming
Influence of active
and inactive
events
Greater attenuation
in N400 for words
associated with an
active vs. inactive
event
Attenuated N400 for
words associated
with a mentioned
(active or inactive)
event
Summary
N400 attenuated by events mentioned in the prior context
Script Knowledge Lexical Priming
Influence of active
and inactive
events
Greater attenuation
in N400 for words
associated with an
active vs. inactive
event
Attenuated N400 for
words associated
with a mentioned
(active or inactive)
event
Findings robust to the order of events mentioned in the context
Thank you!
N400 attenuated by events mentioned in the prior context
Script Knowledge Lexical Priming
Influence of active
and inactive
events
Greater attenuation
in N400 for words
associated with an
active vs. inactive
event
Attenuated N400 for
words associated
with a mentioned
(active or inactive)
event
Findings robust to the order of events mentioned in the context
Experiment 1:
Example Sentence German
Context Sentence:
Introduction: Robertas Erkältung war schlimmer geworden.
(Roberta’s cold had gotten worse.)
1 Script: Sie ging in die Apotheke.
(She went to the pharmacy.)
2 Scripts: Anstatt zum Postamt zu gehen, ging sie in die Apotheke.
(Instead of going to the post office, she went to the pharmacy.)
Target Sentence:
Active Script: Sie trat ein und übergab das Rezept mit einem Lächeln.
(She entered and handed over the prescription with a smile.)
Inactive Script: Sie trat ein und übergab das Paket mit einem Lächeln.
(She entered and handed over the parcel with a smile.)
Experiment 1:
Example sentences
Gabriela had the night off. Instead of walking her dog, she went bowling.
She tied her shoes and took the bowling ball/leash from the shelf.
Erwin had a lot to do. Instead of taking a driving lesson, he began working in
the office.
He sat down and started the computer/car without hesitation.
Lisa enjoyed loitering. Instead of taking care of her budgie, she began surfing
the internet.
She sighed quietly and opened the browser/cage while yawning.
Jan was late. Instead of stopping at the coffee dispenser, he went to the ticket
machine.
He inserted coins and chose the ticket/beverage by pressing a button.
Janina didn’t feel like doing chores. Instead of sweeping the floor, she started
painting.
She tightened her grip and swung the brush
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.