Harvard Slavic Dept Seminar, Dec 2, 2008 One, three, five… Resolving mismatches: number forms in native and One, three, five… heritage Russian Ming Xiang Boris Harizanov Maria Polinsky Main questions • Question 1: How do speakers represent Acknowledgments interface mismatches? Katia Kravtchenko Maša Dikanovic Tom Devine Olga Kagan Where can we look? Main questions Case study: Mismatch between • Question 2: Do native speakers and conceptual and formal features heritage speakers build similar representations when confronted with contrast between conceptual semantics Main data point: inflectional forms and surface form? combining number and case in Russian 1
Mismatches at the interface Russian genitive singular � Obligatory form with some negative existential When mismatch arises at the surface, predicates: what representations do speakers rely mal ´č ik-a/*mal ´č ik net ‘there is no boy’ on? no boy-gen.sg/nom.sg � Assigned by prepositions: do/u/bez/za mal ´č ik-a ‘to/by/without/for boy’ to/by/without/for boy-gen.sg – Surface morphological form: WYSIWYG � Possessive – Conceptual semantics sobaka mal ´č ik-a ‘boy’s dog’ dog boy-gen.sg – Some other abstract representation � Partitive nemnogo spirt-a ‘some alcohol’ some alcohol-gen.sg Russian genitive singular Counting in Russian • 1 mal’ č ik/apel ´ sin • The form of Russian genitive singular 1 boy.nom.sg/orange.nom.sg does not entail any special numerical value • 3 mal ´č ik-a/apel ´ sin-a • However, because of diachronic 3 boy-gen.sg/orange-gen.sg change, it is used with some plural • 6 mal ´č ik-ov/apel ´ sin-ov numerals as a count form: “ gen.sg .” 6 boy-gen.pl/orange-gen.pl ( we will consider masculine nouns only, because case � mismatch between form and concept distinctions are more transparent there) Explaining the mismatch Explaining the mismatch Relationship to number: Analysis 1: “Gen. sg.” is singular: • “Gen sg” is singular • Gen sg is the least marked form of Russian nouns (~possibly what is • “Gen sg” is not singular stored in the lexicon); the numeral Relationship to case: simply takes an unmarked N’, but for • “Gen sg” is part of the case paradigm higher numerals the plural is added on • “Gen sg” is independent of the case (Pesetsky 2007) system 2
Explaining the mismatch Explaining the mismatch Analysis 2: “Gen. sg.” is not singular: Analysis 3 (single categorial chunk): • It is a paucal form , which accidentally • “Gen sg” represents a separate looks like a singular (Rappaport 2002, category, whose number is Bailyn & Nevins 2007) unspecified (Zaliznjak 1968: 2.8) • as a special count form, it is not part of the case paradigm Comparing the analyses How is “gen.sg.” represented? A-1 A-2 A-3 In numerical expressions, if “gen.sg.” is underlyingly a true gen.sg., speakers � � � Is the form used with should represent it the same way as other 2-4 simply case-number feature combinations (such homophonous with as nominative or gen. pl.) the gen.sg. used outside numerals? Can we find processing evidence for this? � � � Is the form used with 2-4 part of Russian case system? Experimental paradigm Experimental conditions Expecting Receiving 12 conditions: X Y acceptable W Z unacceptable X Y Match Match Match Match NC N C zero unacceptable to a X/Y Z lesser degree? One (1) nom.sg. gen.sg. nom.pl. gen.pl. nom.sg. (nom.sg.) nom.sg. gen.sg./nom.pl gen.pl. Small N (2-4) gen.sg. nom.sg. gen.pl. nom.pl. gen.pl. gen.pl. (“ gen.sg.”) nom.pl./gen.sg. Big N (5+) gen.pl nom.pl gen.sg. nom.sg. nom.sg. gen.sg. (gen.pl.) “ gen.sg.” ? nom.sg./gen.pl. nom.pl. 3
Sample stimuli Sample stimuli Experiment 1: Acceptability-native speakers acceptability judgments 7 m atchNC • 1-7 scale (7 is the best) m atchN 6 m atchC • 60 items m atchzero 5 • 108 fillers 4 • 35 native speakers 3 2 1 num ber1 sm all N big N Experiment 2: Summary of the results self-paced reading • Native speakers show clear • 60 items grammaticality judgments • 108 fillers • No graded effect: what’s bad is just bad • 37 native speakers 4
By way of reminder: Conditions Self-paced reading demo nom.sg. nom.pl … one boy- in glasses …. gen.sg. gen.pl. nom.sg. три маленькие мальчики в очках nom.pl … three boy- in glasses …. gen.sg. gen.pl. nom.sg. nom.pl … six boy- in glasses …. gen.sg. gen.pl. Privileged status of ‘one N’? Critical N-native speakers 540 • Way of expressing indefinites, 520 which are less marked/unmarked 500 m atchNC • Nominative sg is the citation form 480 m atchN m atchC • ‘one N’ may be more frequent 460 m atchzero 440 • Anything else? 420 400 number1 small N big N Summary of results Spill over - native speakers m atchNC 5 4 0 m atchN 5 2 0 online performance, native speakers: m atchC 5 0 0 m atchzero • sensitivity to grammaticality in number 1 4 8 0 4 6 0 and big number environments, but not in 4 4 0 4 2 0 small number context 4 0 0 3 8 0 • no graded effect 3 6 0 3 4 0 3 2 0 3 0 0 num ber1 sm all N big N 5
Linguistic implications Main questions • “Gen.sg.” does not receive the • Question 2: Do native speakers and representation as a sg genitive— heritage speakers build similar otherwise subjects would have shown representations when confronted with sensitivity to the wrong representations contrast between conceptual semantics and surface form? • The exact representation of “gen.sg” is yet to be determined Experiment 3: Biographical data acceptability judgments, heritage speakers • 6 males, 17 females • age ranges between 19-26 (average 22.5) • age moving to the States • Offline 1-7 scale (7 is the best) • between 0 to 13; average 7 years old • 60 items 6 • 108 fillers 5 • 23 heritage speakers 4 3 2 1 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 More age m oving to the states Summary Acceptability-heritage speakers 7 m atch NC • Although still noisy, heritage speakers m atch N 6 m atch C show the basic grammaticality pattern, m atch zero 5 except for the big number context 4 • The fine detailed patterns on individual 3 features are yet to be determined when 2 more data are collected 1 num ber 1 sm all num ber big num ber 6
Experiment 4: self-paced reading, heritage speakers Critical N -heritage speakers • 60 items 950 900 • 108 fillers match NC 850 800 match N • 25 heritage speakers, advanced (can 750 match C read Cyrillic) 700 match zero 650 600 550 500 number 1 small number big number One step forward, Some tendencies… two steps back… • Number trumps case in the small N and • Does the partially matched number feature have a facilitation effect for big N environments (these are semantically plural context (any number semantically plural) above 1)? • In the big N category, the use of nom. sg is tolerated better—possibly an • Is the big number context really special, indication of the #+citation form as the or it is just noisy at this stage? unmarked version Heritage speakers: Critical N Spill over - heritage speakers • Heritage speakers show the basic 6 0 0 grammaticality pattern, except for the big 5 5 0 number context 5 0 0 m atch NC m atch N 4 5 0 • However, the sensitivity to grammaticality in m atch C the small number context is strikingly different m atch zero 4 0 0 from what we found with native speakers 3 5 0 3 0 0 • As with other grammatical patterns, this may num ber 1 sm all big num ber num ber be evidence of divergent grammar (different from the baseline) (cf. Polinsky 2008) 7
Heritage speakers: Spill over Native and heritage speakers • The spill over region confirms that • It is the heritage speakers, not the heritage speakers show the basic native speakers, who showed the grammaticality pattern in all three sensitivity to grammaticality in the small numerical contexts number context! • This suggests that heritage speakers build their representation more directly • Finer details on individual features are based on the input form they see yet to be determined Tentative conclusions “Gen.sg.” is different for the two • For native speakers, there is no clear subject groups evidence of graded grammaticality • Native speakers: “Gen.sg.” is a special effects based on partial feature match form and does not map to the deep representation of gen. sg • Confronted with the surface mismatch at the morphological-conceptual • Heritage speakers: what you see is interface, native speakers pay attention what you get, “Gen.sg.” is gen. sg. to deeper representations, while heritage speakers rely more on surface representations Tentative conclusions The nature of “gen.sg.” • Heritage speakers do not recognize surface homophony Recall the three analyses: • Gen. sg. is gen. sg. is gen. sg. • Heritage speakers are more tolerant to • “Gen.sg.” is non-singular (paucal) morphological mismatches • “Gen.sg.” is not a case form • For heritage speakers, conceptual number might potentially play some role (stay tuned) 8
Recommend
More recommend
