INTRODUCTION BACKGROUND & AIM

(1) School of Communication Sciences & Disorders, McGill University, Montreal, Canada; (2) Center for Research on Brain, Language and Music, Montreal, Canada; (3) Dipartimento di Scienze Cognitive, Università degli studi di Trento, Rovereto, Italia

Longstanding debate about “critical period” for L2: Maturational constraints on brain plasticity1? - Late-learners differ from native-speakers in performance and/or neurocognitive mechanisms2,3 - However, late age-of-acquisition confounded with low proficiency level and/or low exposure4,5 Aims: (1) Disentangle proficiency effects from AoA in “minority=language speakers” or “attriters” (= low exposure and declining proficiency in L1 due to immigration but high-proficiency/exposure to late-L2); (2) investigate neurocognitive correlates of L1 attrition which are, to date, largely unexplored.

- Supports anecdotal reports for confusing similar words - Effects of both condition and frequency suggests that lexical attrition is not only “degredation” or “forgetting”, but that L2-induced interference plays an important role.

References: [1] Lenneberg, E. (1967). Biological foundations of language. New York: Wiley. [2] Johnson, J. S., & Newport, E. L. (1989). Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language. Cognitive Psychology, 21, 60-99. [3] Ullman, M. T. (2001). The neural basis of lexicon and grammar in first and second language: The declarative/procedural model. Bilingualism: Language and Cognition, 4(1), 105-122. [4] Weber-Fox, C. M., & Neville, H. J. (1996). Maturational constraints on functional specialization for language processing: ERP and behavioural evidence in bilingual speakers. Journal of Cognitive Neuroscience, 8(3), 231-256. [5] Steinhauer, K., White, E. J., & Drury, J. E. (2009). Temporal dynamics of late second language acquisition: Evidence from event-related brain potentials. Second Language Research, 25(1), 13-41. [6] Paradis, M. (2007). L1 attrition features predicted by a neurolinguistic theory of bilingualism. In B. Kopke, M. Schmid, M. Keijzer and S. Dostert (Eds.), Language attrition: Theoretical perspectives (pp. 121-134). Amsterdam: John Benjamins. [7] Köpke, B. (2002). Activation thresholds and non-pathological L1 attrition”. In Advances in the Neurolinguistics of Bilingualism. Essays in Honor Of Michel Paradis, F. Fabbro (ed), 119-142. Undine: Forum. [8] Bates, E., & MacWhinney, B. (1989). Functionalism and the Competition Model. In B. MacWhinney & E. Bates (Eds.), The crosslinguistic study of sentence processing. New York: Cambridge University Press, 3-76.

(1) Do experimental data support anecdotal reports of word-finding difficulties and “confusing words”, especially when similar? (2) Is there evidence for cross-linguistic competition in attrition? (3) Are low-frequency items more vulnerable to attrition6?

PARTICIPANTS

EXPERIMENT 2: Lexical access in comprehension

EXPERIMENT 1: Lexical access in production RESEARCH QUESTIONS

DESIGN & STIMULI

EXPERIMENT 3: Online syntactic processing

GENERAL DISCUSSION

My Italian is not what it used to be: Investigating the neural correlates of L1 attrition and late L2 acquisition

Kristina Kasparian1,2, Francesco Vespignani3 & Karsten Steinhauer1,2

Graded N400

-Translation task (production) - Italian English (reported here) and English Italian - 29 “false-friend” items + 29 “control items”

- e.g. fabbrica = factory ≠ fabric (fabric = tessuto ≠ fabbrica) - Matched pairwise on frequency of item and of translation, length, number of syllables, word category, semantic field -Words displayed on screen and voice-response is registered (RT)

Small P600

(a) CONTROLS

● 4 groups: (1) Minority-speakers of Italian (L1) + English (late L2) = attriters

(2) L2 speakers of English (L1) + Italian (late L2) = bilingual controls

(3) Italian native-speakers in Italy = monolingual controls

(4) English native speakers = monolingual controls

Mismatch: To cover his head, the fisherman wears the mint of wool. Swap: To cover his head, the fisherman wears the chapel of wool. Correct: To cover his head, the fisherman wears the hat of wool.

Correct Per coprire la testa, il pescatore porta il cappello di lana. (To cover his head, the fisherman wears the hat of wool.)

Swap Per coprire la testa, il pescatore porta la cappella di lana. (To cover his head, the fisherman wears the chapel of wool.)

Mismatch Per coprire la testa, il pescatore porta la menta di lana. (To cover his head, the fisherman wears the mint of wool.)

N400 gradient not as clear

Large P600

Online acceptability ratings: (1) Mismatch and swap are rated lower than correct sentences in both groups; (2) Difference between swap and correct condition is greater in attriters than in controls.

ERP results: (a) Controls show a significant N400 in both violation conditions, followed by small P600; The N400 is numerically graded, but difference between mismatch & swap not statistically significant. (b) Attriters show a different pattern, namely an N400 only for the mismatch condition, and a large P600, especially for the swap condition.

Results suggest: (1) A different pattern of processing in attriters, where “swap” not readily detected;(2) Less automatic and more elaborated processing in attriters (“double-take” further downstream); (3) ERPs reveal a group difference that behavioral judgments are not sensitive to.

Online acceptability ratings

(b) ATTRITERS

RESEARCH QUESTIONS RESULTS

DESIGN & STIMULI - Nouns differing only in final vowel, but different meaning e.g. cappello (hat) vs. cappella (chapel); busto (chest) vs. busta (envelope) - Sentence contexts created for each word (48 pairs = 96 contexts) - 3 conditions: (1) Correct (each word in pair occurs in its proper context) (2) Swap (words in pair are switched, e.g. cappella in cappello context) (3) Mismatch (word from another pair, e.g. menta in cappello context) - Balanced design: Each target words occurs in each condition - Task: Acceptability judgment (rating 1-5)

(1) Is there evidence of lexical attrition in online comprehension, (aside from behavioral tasks or production)? (2) Do attriters differ from native-speakers in how the brain processes Italian in real-time? (3) Can we show effect of “confusing words” within L1 with ERPs?

RESEARCH QUESTIONS

DESIGN & STIMULI

RESULTS (1) Is there evidence of attrition in morphosyntax (although studies suggest that grammatical knowledge is less vulnerable than lexicon7)? (2) Do attriters process and interpret Italian differently than native controls due to interference from the dominant-L2 grammar?

1. V-NP subject

Il gatto che caccia i topi corre nel giardino. (The cat that chases the mice runs in the garden)

2. V-NP object

I topi che caccia il gatto tremano di paura. (The mice that chases the cat tremble with fear.)

3. NP-V subject

Il gatto che i topi caccia corre nel giardino. (The cat that the mice chases runs in the garden.)

4. NP-V object

I topi che il gatto caccia tremano di paura. (The mice that the cat chases tremble with fear.)

-Relative clauses: Cross-linguistic differences in grammaticality and in cue strength7 (semantic cues in Italian; word-order in English) - Conditions 1-4 are grammatical in Italian (although 2 & 3 are less preferred because they are less frequent) - Conditions 2 & 3 are impossible in English (ungrammatical) - Strong agent-patient relationships ( semantic cues ) - Noun1, Noun2 and verbs matched on length + frequency across conditions, and number counterbalanced per condition - Task: Acceptability judgment (rating 1-5)

V-NP contrast (condition 2 vs. 1)

Online acceptability ratings

(b) ATTRITERS

V-NP subject: The cat that chases the mice runs in the garden. V-NP object: The mice that chases the cat tremble with fear.

Controls show an N400 for object condition, suggesting that they process semantic cues first, then a P600.

Attriters show a tendancy toward left-lateralized negativity + significant P600 (morphosyntactic violation effect, no reliance on semantic cues)

Attriters are similar to controls in ratings for conditions 1&4 (grammatical in English + preferred in Italian, but rate 2&3 (ungrammatical in English) as less acceptable than controls do

- Controls show an early frontal positivity (P3A?), with no evidence of an early negativity, followed by a posterior P600. This pattern is consistent with a mild garden-path effect, rather than an outright morphosyntactic violation. - Attriters show a different pattern, characterized by a tendency toward an early negativity (not clear yet whether LAN, N400 or left-temporal negativity). - Behavioral and ERP results suggest that English-dominance affects how attriters accept and process relative clause word orders that clash between the L1 and L2.

N400

P600

NP-V contrast (condition 3 vs. 4)

Left-lateralized negativity

P600

(a) CONTROLS

NP-V subject: The cat that the mice chases runs in the garden. NP-V object: The mice that the cat chases tremble with fear.

P300a

P300a

P600

P600

(a) CONTROLS

(b) ATTRITERS Negativity (N400? LAN? LTN?)

Negativity

?

? P600?

- Online measures such as ERPs provide crucial information about attrition, as some effects are not revealed in behavioral tasks and offline linguistic tests. - Preliminary evidence of ongoing brain plasticity for L1 and L2, even in adulthood.

- Three experiments provide the first evidence of attrition effects in (1) lexical access during translation, (2) online lexical processing, and (3) online syntactic processing. - The more dominant and highly-proficient L2 influences the L1 lexicon and grammar.

RESULTS (b) Number of errors by condition and frequency (c) Types of errors by condition (a) Accuracy (correct items out of 29)