Acquisition of L2 morpho VS NS processing

Acquisition of L2 morpho VS NS processing

1. Factors (from Rogers et al 2015) + Gor 2010

   • Type and token frequency in input
     • Ellis 2002
     • Ellis & Schmidt 1997,1998
   • L1 background: properties of infl morpho L1
     • Ellis and Sagarra 2010, 2011
     • Rast 2010
   • Individual differences (verbal WM)
     • Brooks et al 2006
     • Kempe Brooks 2008
     • Kempe et al 2010
   • Metaling awareness of construction
     • Brooks & Kempe 2013
     • Grey et al 2014
     • Marsden et al 2013
     • Robison 2005
   • Morphological richness of L2
   • L2 proficiency
   • Properties of input
     • Type of instructional condition
       • Alanen 1995
       • DeKeyser 1995
     • AoA
     • amount of exposure and practice

2. Theories Differences L1/L2

   • Processing constraints
     • Competence = includes internalized linguistic representations only, regardless of accuracy in natural communication
     • low WM capacity
     • poor decoding
     • inadequate processing speed
     • MacDonald 2006
     • Missing Surface Inflection Hypothesis
       • Prevost & White
       • production-specific mapping pbs
         • mapping pbs between abstract features and morphological form (Lardiere 1998)
       • < L1 prosodic constraints?
     • = performance deficiency
       • difficulties in :
         • accessing
         • retrieving
         • controlling
         • but K internalized
       • Ellis & Larsen-Freedman 2006?
     • Sharwood Smith (competence/control model)
     • Sorace
       • interlanized ling representations vs procedural K necessary to access it
   • More than processing constraints
     • Shallow Structure Hyp
       • reliance on lexical cues > syntactic
       • storage > decomp
       • Clahsen & Felser
     • Declarative/Procedural Model
       • declarative 1st = storage
       • then procedural = decomp
       • Pinker
       • Ullmann
       • L2 rely more on declarative system
     • Selective integration
       • L2 ignore morphosyn cues when not functionnaly loaded
       • Jiang 2007
       • Factors
         • = factors contributing to integration
         • L1 transfer
         • perceptual salience
         • input frequency
         • redundancy
     • = competence problem
       • competence = ability to use a structure correctly in spontaneous language production without paying attention to it
       • incomplete acquisition of morpho knowledge
       • knowledge not integrated, not used in spontaneous speech
       • correct use comes from EK
       • Johnson & Newport = CPH
       • Krashen
       • Jiang (transfer-based model)
   • Prosodic Transfer Hypothesis
     • properties of L1 prosody transferred
     • Goad & White 2003
       • English tense &amp; agreement: require adjunction to PWd, not possible in Mandarin (in French either, voir Goad et al 2011)
     • Goad et al 2011

3. Evidence

   • MEG
     • + dual-route
       • syntactic MMN to ungram past: generated productively
         • Bakker 2013
   • Behavioral
     • No evidence of transfer for tense marking Aux acquired before infl morpho
       • Paradis 2008
   • EEG
     • Nb/Gender agreement
       • Implicit training > more like NS
         • Morgan-Short 2012
           • IMP: AN + P600 vs EXP Pos + P600
         • Morgan-Short 2010
       • L2 can show NS-like processing
         • Aleman 2014
           • same L2/NS: P600
     • Tense marking
       • White 2012
         • L2 proficiency det P600, not L1
         • L1 infl latency
         • P600
       • Newman 2007 (L1)
         • Regular: LAN + P600
         • Irregular: P600
   • fMRI
     • late learners low / moderate = different mechanisms
     • high proficiency = early/late bilinguals = similar areas
     • late learners = greater activation left IFG
     • Roncaglia Denissen 2015 (review)