Kristin Lemhöfer & Julia Egger INTRODUCTION

0 downloads 0 Views 3MB Size Report
ERP correlates of picking up new foreign-language words in dialogue ... …then, you take a whisk ... 40 low-frequent non-cognate target words, 120 well-known filler words. Dialog game (alternating with 'virtual partner') with price comparisons [2] ... Block 1: listening to unknown vs. known words (targets vs. fillers; N=35). Cz.
ERP correlates of picking up new foreign-language words in dialogue Kristin Lemhöfer & Julia Egger Donders Institute, Radboud University, Nijmegen, The Netherlands, University

INTRODUCTION

DESIGN

• New words in a second language (L2) are often learned incidentally in every-day situations

…then, you take a whisk…

• What are the ERP correlates of this incidental learning from spoken input? • Attempt to simulate this learning in the EEG-lab

R E S U LT S ( C n t d . )

• ‘Virtual partner’: pre-recorded female native speaker of British English

EEG: SUBSEQUENT MEMORY EFFECT

• Target (unknown) words always used by ‘virtual partner’ first

Block 1: unknown words later produced vs. not produced; N=14

• Training and testing embedded in the same game

• Mainly frontal Late Positive Component (LPC) from about 500 ms for learned words

• One exposure per target word in first block, one in second block • Retention interval (between input and testing): 3 trials

• ERPs during encoding are predictive of subsequent incidental learning

R E S U LT S BEHAVIORAL

METHOD

EEG AFTER LEARNING

P A R T I C I P A N T S & M A T E RF3I A L S

Fz

F4

-4

-4

-4

-2

-2

-2

0

0

2

2

E E G4

4

Block 2: known vs. just learnt vs. not yet learnt words; N=14

• 37 right-handed native speakers of Dutch (students) (more to be analyzed) • Regular use of L2 English (mean LexTALE score [1]: 72%) • Unaware of taking part in a learning study (cover story: make consistent price comparisons; in English because experimenter 0 does not speak Dutch) 2 • 40 low-frequent non-cognate target words, 120 well-known filler words 4

PRETEST

Block 1: listening to unknown vs. known words (targets vs. fillers; N=35)

Task: estimate the price of each object FILLER FILLER CRITICAL

0

500

0

500

0

C3

Cz

C4

-4

-4

-4

-2

-2

0

0

• Not yet learnt words remain to show an LPC in block 2 just as 500 at first exposure

“A lamp costs 15 Euros.” à KNOWN “An apple costs 50 Cent.” à KNOWN

-2

“A… the cooking thing costs 3 Euros.” à UNKNOWN

0

MAIN EXPERIMENT

(if known: exclude From analysis)

2

2

Dialog game (alternating with ‘virtual partner’) with price comparisons [2] 4

Listening to ‘virtual partner’

Own price comparison “A …6whisk is cheaper than a lamp”.

“An apple is cheaper than a whisk”.

0 3 intervening trials



à Agree or not?

[email protected]

4

6

6

500

P3

CONCLUSIONS • First study on on-line ‘picking up new words’ during dialog in L2

2

4 0

500

Pz

• LPC for novel vs. already known words (no N400) • Subsequent learning is predicted by LPC during encoding 0

REFERENCES

-4

• No N400, other than in L1 pseudoword ‘learning’ during reading -4 [3]

-2

• Broadly-2distributed Late Positive Component (LPC) from about -2 700 ms

[1] Lemhöfer & Broersma (2012). Introducing LexTALE: A quick and valid Lexical Test for Advanced Learners of

-4

0

0

2

2

2

4

4

4

6

6

6

0

500

• 500 Word-like ERP signature of newly learnt words after only one exposure

P4

0

0

500

• Just learnt words start to become indistinguishable from words known long before after only one exposure

English. Behavior Research Methods, 44. [2] de Vos, Schriefers, ten Bosch & Lemhöfer (under revision). High learning rates for interactive L2 vocabulary acquisition in a lab-based immersion setting. [3] Batterink & Neville (2011). Implicit and explicit mechanisms of word learning in a narrative context: An event-related potential study. Journal of Cognitive Neuroscience, 23.

0

500