The importance of Input Processing in Second Language

The importance of Input Processing in Second Language Acquisition: the example of Greek grammatical gender for L 2 learners – implementation to other grammatical aspects Dr Patricia Koromvokis Ass. Lecturer Department of International Studies: Cultures & Languages, Macquarie University

Structure/ Aim of Presentation • PART A (Recent study): Input Processing: the example of Greek grammatical gender for adults L 2 learners: Theoretical Framework Rationale of the study – Methodological Framework – Findings – Discussion – Επεξεργασία γραμματικού γένους • PART B (Future study – AIM): Teaching applications : the importance of ‘less segmented’ linguistic input in teaching grammatical aspects to young students L 2 learners - Παρουσίαση/ Διδασκαλία γραμματικών φαινομένων με την χρήση ενός ‘λιγότερου διηρημένου’ γλωσσικού εισαγόμενου

PART A: Can L 2 learners acquire new grammatical features? - Grammatical Gender le table or la table? French der Tisch or die Tisch or das Tisch? German el messa or la mesa? Spanish il tavola or la tavola? Italian o τραπέζι or η τραπέζι or το τραπέζι? Greek the table English

What does processing of grammatical gender mean for a L 2 learner? Learning an L 2 involves the acquisition of both competence and performance in this language. For instance, learning grammatical gender in Greek involves acquiring both the knowledge of a word’s gender (i. e. , gender assignment, κήπος MSC) and how gender is expressed syntactically (i. e. , gender agreement, ο μεγάλος και τακτοποιημένος κήπος). The L 2 learner must then develop the capacity to systematically produce and process this knowledge. Ø The term PROCESS refers to the realization of gender agreement between the noun and other elements related to it.

Theoretical Framework: How is gender processed? Clothing Concept Dress Syntactic class node Gender Node Mas Fem Number node Syntactic Level Noun Neu Tense node φόρεμα MY RESEARCH Lemmas φόρεμ-α φ-ό-ρ-ε-μ-α Morphological Level Phonological Level

Theories of gender processing in L 2 Full Transfer / Full Access Model (Schwartz & Sprouse, 1996; White, 1989, 2003) Failed Functional Features hypothesis (FFFH) (Hawkins & Franceschina, 2004) Competition Model (Bates & Mac. Whinney, 1987; Mac. Whinney, 2002; Mac. Whinney, 2008) Missing Surface Inflection Hypothesis (Prévost and White, 2000)

Why L 2 learners have difficulty with grammatical gender? - Factors • Age-related insensitivity to grammatical features not instantiated in the L 1 (Franceschina, 2005; Hawkins, 2009; Sabourin and Stowe, 2008) • Performance limitations specifically affecting L 2 production (Alarcón, 2011; Montrul et al. , 2008) • More general limitations on L 2 processing due to increased demands on working memory (Gillon Dowens et al. , 2010; Keating, 2009, 2010) o • Different language learning environments (Lew-Williams and Fernald, 2010). • The nature and length of exposure to the L 2 and the resulting levels of proficiency (Gillon Dowens et al. , 2010; 2011) • The typological differences and similarities between the L 1 and L 2 and the possibility of feature-mapping or transfer of processing routines between these (Sabourin, Stowe, & de Haan, 2006; Tokowicz & Mac. Whinney, 2005).

When L 2 learners acquire better the grammatical gender? The ‘lexical gender learning hypothesis’ • Lexical gender learning hypothesis: Adults learn both nouns and their articles better when they first hear nouns used in context with their articles, prior to hearing them individually (Arnon & Ramascar, 2011; Gruter et al. , 2012). • L 2 learners used gender-marking on the determiner as predictive cue with newly learned novel nouns, but not with familiar nouns. (Gruter et al. , 2012) • Gender marking on dependent constituents, e. g. determiners and adjectives, may facilitate the predictive processing of the subsequent noun in grammatical sentences (Kamide, 2008).

Grammatical gender clues There are 3 possible sources of information that learners can rely on for processing the gender system of a language like Greek: 1. Semantic clues (Does the noun refer to something animate or inanimate? If it is animate, what sex is it? ) - ΦΥΛΟ 2. Morphophonological clues (What is the noun’s ending? )- ΚΑΤΑΛΗΞΗ 3. Syntactic clues (What forms do related articles, adjectives, pronouns and passive participles take? )- ΣΥΝΤΑΞΗ (Francheschina, 2005)

Determination of Greek grammatical genderthe role of Syntax Ambiguous Suffix –ος occurs in more than one gender values and the missing value is acquired at the phrasal level, that is through agreement in syntax with an item marked for a particular gender value (e. g. a determiner, an adjective). e. g. ο καλός δικηγόρος (o ka'los ðici'ɣoros MASC) (= the good lawyer) e. g. η καλή δικηγόρος (i ka'li ðici'ɣoros FEM) (=the good lawyer) (Ralli, 2002; Tsimpli, 2011)

My study: Interrelation between gender, sex and inflectional class in Greek SYNTAX gender sex SEMANTICS inflectional class MORPHOLOGY (Francheschina, 2005, p. 73)

Rationale of the study • Gap: There have been only 4 studies (Tsimpli, 2003; Tsimpli et al. , 2005; Agathopoulou et al. , 2008; Agathopoulou & Papadopoulou 2011; ) done upon the learning of the Greek grammatical gender of known words for L 2 ers. Which under detailed evaluation seem to contain specific and very important methodological problems. • Research Question: What is the relevant importance of morphology and syntax in the processing of the Greek grammatical gender by adult L 2 learners of Greek language? • Hypothesis: L 2 learners will perform better when syntactical information is also available, that is determiner and/ or adjective + noun sequences.

Method: Design of the Experiment 1 st Factor : 1. Morphological Information (noun ending) 2. Syntactical (article) + Morphological Information (noun ending) 3. Syntactical (article + adjective) + Morphological Information (noun ending) 2 nd Factor : 1. Masculine Gender 2. Feminine Gender 3. Neuter Gender 3 rd Factor : 1. Beginners 2. Advanced Dependent variables 1. Correctness of the response 2. Time reaction

Method: Participants • The participants were university students L 2 ers of Greek at the beginners and advanced level from Macquarie University, Sydney University and UNSW. • 151 visited the online survey. • Excluded: ✓ 9 participants did not provide consent form. ✓ 14 participants had Greek first language. ✓ 24 participants did not complete the survey. • 104 participants were met the selection criteria and their responses were analysed.

Method: Experiment - Part A

Method: Experiment - Part B The participants are asked to fill a gap filling exercise writing the appropriate English personal pronoun (he/ she/ it) according to the grammatical gender of the given Greek non words. Ένας μπλε πιφάρος ____ is blue Ένας κόκκινος κατάλης ____ is blue Ένας μπλε τανίκης ____ is blue

Method: Tests Reliability and Validity • Internal consistency reliabilities: Cronbach’s alphas were considered satisfactory (Anastasi & Urbina, 1997; Alexopulos & Kalaitzidis, 2004). Cronbach’s alphas 0. 72 | 0. 75 (quest. 1 and 2) on the base of correct answers. Cronbach’s alphas 0. 68 | 0. 72 (quest. 1 and 2) on the base of time reaction. • Construct validity: Principal Components Analysis revealed three factors for both questionnaires explaining for the first test 26. 84% and 28. 76% (correctness and reaction time) and for the second test 28. 75% and 25. 23%. • Confirmatory factor analysis confirmed the hypotheses of tasks’ construction. • Concurrent validity analysis confirmed the relation between the factors of both questionnaires. The results were considered satisfactory (Anastasi & Urbina, 1997; Alexopulos & Kalaitzidis, 2004). • Face validity of the experimentation confirmed the initial hypothesis.

Results for the first test ANOVA: Correctness and time reaction Significant: • type of information factor (F(1, 102)=16. 407 MSe=8. 008 p<. 001) |(F(1, 102)=15. 424 MSe=6. 005 p<. 001) • level of knowledge (F(1, 102)=6. 98, MSe=7. 606, p<. 05) | (F(1, 102)=7. 05, MSe=7. 554, p<. 001) Non-significant: • gender variation (F(1, 102)=3. 451) | (F(1, 102)=2. 422) • interaction between gender and level of knowledge (F(1, 102)=. 059) | (F(1, 102)= 1. 002) • interaction between information and level of knowledge (F(1, 102)=. 101) | (F(1, 102)=. 098) • interaction between the three factors (F(1, 02)=. 473) | (F(1, 02)=. 724)

Results for the first test Post-hoc test was carried out for the factor of Information (t-tests): There was a significant difference for the first dependent variable (correctness): i. Extensive syntactical Vs morphological ii. Extensive syntactical Vs syntactical iii. Morphological Vs syntactical (means 9. 5 and 5. 7 (d)=1. 99 p<. 001) (means 9. 5 and 8. 2 (d)=. 75 p<. 001) (means 5. 7 and 8. 2 (d)=1. 03 p<. 001) There was also a significant difference for the second dependent variable (time reaction): i. Extensive syntactical Vs morphological ii. Extensive syntactical Vs syntactical iii. Morphological Vs syntactical (means 5. 5 and 9. 1 (d)=1. 87 p<. 001) (means 5. 5 and 6. 2 (d)=. 85 p<. 05) (means 9. 1 and 6. 2 (d)=1. 24 p<. 05)

Graph for the first test

Results for the second test ANOVA: Significant: Correctness and time reaction • type of information factor (F(1, 102)=14. 231 MSe=7. 908 p<. 05) |(F(1, 102)=16. 004 MSe=5. 985 p<. 05) • level of knowledge (F(1, 102)=7. 051, MSe=7. 439, p<. 001) | (F(1, 102)=7. 85, MSe=8. 005, p<. 001) Non-significant: • gender variation (F(1, 102)=3. 221) | (F(1, 102)=3. 010) • interaction between gender and level of knowledge (F(1, 102)=. 061) | (F(1, 102)= 1. 102) • interaction between information and level of knowledge (F(1, 102)=. 125) | (F(1, 102)=. 101) • interaction between the three factors (F(1, 02)=. 523) | (F(1, 02)=. 689)

Results for the second test Post-hoc test was carried out for the factor of Information (t-tests). There was a significant difference for the first dependent variable: i. Extensive syntactical Vs morphological (means 10. 2 and 6. 4 (d)=2. 05 p<. 05) ii. Extensive syntactical Vs syntactical iii. Morphological Vs syntactical (means 10. 2 and 9. 3 (d)=. 85 p<. 001) (means 6. 4 and 9. 3 (d)=1. 03 p<. 001) There was also a significant difference for the second dependent variable: i. Extensive syntactical Vs morphological (means 2. 5 and 5. 1 (d)=1. 87 p<. 001) ii. Extensive syntactical Vs syntactical iii. Morphological Vs syntactical (means 2. 5 and 3. 1 (d)=. 85 p<. 05) (means 5. 1 and 3. 1 (d)=1. 24 p<. 05)

Graph for the second test

Summary of findings regarding processing Greek grammatical gender by adult L 2 learners of Greek process more accurately and quicker grammatical gender of new words when they are been presented with extensive syntactical information i. e. article–noun ending (ο κήπος) or article – adjective -noun ending (ο μεγάλος κήπος) than if they are first exposed to noun labels/ nouns combined only with the gender marked suffix e. g. δάσκαλος, φίλος. Ø This must be further investigated and been applied as a teaching method (future research).

PART B: L 1 vs. L 2: Different language learning environments Tight associations are formed between frequently co-occurring elements such as articles and nouns in early L 1 lexicons. But, the tight associations between determiners and nouns that emerge in the early L 1 lexicon are unlikely to arise in developing L 2 lexicons. Thus, different learning conditions affect learners’ ability to associate a noun with its gender class ( e. g. an artificial language learning experiment by Arnon and Ramscar (2009, 2012)

Future research goal: from grammatical gender to (teaching, learning, processing) Greek grammatical aspects to L 2 learners The importance of ‘less segmented’ input in teaching and learning: Øgrammatical gender Øverb endings Øvocabulary

‘Less segmented’ input in teaching vocabulary – Deep word knowledge/ Lexical Knowledge “Deep word knowledge” according to linguist scholars refers to the depth of word knowledge which it constitutes multiple levels of knowledge, i. e. grammatical category, pragmatic, etymology, and cognitive or/and the semantic value of word (Schoonen, R. , Verhallen M. , 2009, Corrigan, R. 2007). e. g. η καλοβαμμένη αίθουσα

References • Alarco n IV (2011) Spanish gender agreement under complete and incomplete acquisition: Early and late bilinguals’ linguistic behavior within the noun phrase. Bilingualism: Language and Cognition 14: 332– 50. • Agathopoulou, E. , & Papadopoulou, D. (2011, April 3 -5). Probing the adult initial state: of non-native Greek: A case study. 19 th International Symposium on Theoretical and Applied Linguistics. Thessaloniki. • Agathopoulou, E. , Papadopoulou, D. , & Zmijanjac, K. (2008). Noun-Adjective agreement in L 2 Greek and the effect of input-based instruction. 1 -19. • Arnon, I. & Ramascar, M. (2012). Granularity and the acquisition of grammatical gender: How order-of -acquisition affects what gets learned. Cognition, 122, 292 -305. • Bates, E. , & Mac. Whinney, B. (1987). Competition, variation, and language learning. In B. Mac. Whinney (Ed. ), Mechanisms of language acquisition. Hillsdale, NJ: Lawrence Erlbaum Associates. • Clahsen, H. , & Felser, C. (2006). How native-like is non-native language processing? Trends in Cognitive Sciences, 10(12), 564– 570. • Corbett, G. G. (1991). Gender. Cambridge: Cambridge University Press. • Dimitrakopoulou, M. , Fotiadou, G. , Roussou, A. , & Tsimpli, I. -M. (2006). Features and agree relations in L 2 Greek. In A. Belletti, E. Bennati, C. Chesi, E. Di Domenico, & I. Ferrari (Ed. ), Language Acquisition and Development: Proceedings of GALA 2005 (pp. 161 -166). Newcastle: UK: Cambridge Scholars Press. • Franceschina, F. (2005). Fossilized second language grammars: The acquisition of grammatical gender. Amsterdam: John Benjamins. • Hawkins, R. , & Chan, C. (1997). The partial availability of Universal Grammar in second language acquisition: The ‘failed functional features hypothesis’. Second Language Research, 13, 187– 226.

References • Gillon Dowens M, Vergara M, Barber HA, and Carreiras M (2010) Morphosyntactic processing in late second-language learners. Journal of Cognitive Neuroscience 22: 1870– 87. • Gillon Dowens M, Barber H, Guo Dowens M, et al. (2011) Gender and number processing in Chinese learners of Spanish: Evidence from event related potentials. Neuropsychologia 49: 1651– 59. • Grutter, T. , Lew-Williams, C. & Fernald, A. (2012). Grammatical gender in L 2: A production or a real time processing problem? Second Language Research, 28, 191 -215. Harley, B. (1979). French gender “rules” in the speech of English-Dominant, French-Dominant and monolingual Frenchspeaking children. Working Papers in Bilingualism, 19, 129 -156. • Ralli, A. (2002). The role of morphology in gender determination: evidence from Modern Greek. Linguistics, 519 -551. • Keating GD (2010) The effects of linear distance and working memory on the processing of gender agreement in Spanish. In: Van. Patten B and Jegerski J (eds) Research in second language processing and parsing. Amsterdam: Benjamins, pp. 113– 34. • Lew-Williams C and Fernald A (2010) Real-time processing of gender-marked articles by native and non-native Spanish speakers. Journal of Memory and Language 63: 447– 64. • Mac. Whinney. (2002). Extending the competition model. In R. H. Roberto, & A. Jeanette (Eds. ), Advances in psychology (pp. 31– 57). North-Holland. • Mac. Whinney, B. (2008). A unified model. In N. Ellis, & P. Robinson (Eds. ), Handbook of cognitive linguistics and second language acquisition. New York: Routledge. • Sabourin, L. , & Stowe, L. A. (2008). Second language processing: When are L 1 and L 2 processed similarly. Second Language Research, 24, 397– 430. • Schwartz, B. D. , & Sprouse, R. A. (1996). L 2 cognitive states and the full transfer/full access model. Second Language Research, 12, 40– 72. • Tokowicz N and Mac. Whinney B (2005) Implicit and explicit measures of sensitivity to violations in second language grammar: An event-related potential investigation. Studies in Second Language Acquisition 27: 173– 204.

Ευχαριστώ πολύ patricia. koromvokis@mq. edu. au 31