Language Learning48:2,June1998,pp.245–279
The Acquisition of Ten-Aspect Marking in Japane as a Second Language
Yasuhiro Shirai
Cornell University
inspiredAtsuko Kurono
Japan International Cooperation Center
Rearchers have argued that the developmental quence of ten-aspect morphology in cond language
(L2)acquisition follows a universal pattern(the Aspectnice to meet you什么意思
Y asuhiro Shirai,Department of Modern Languages;Atsuko Kurono,Training Division,Tsukuba Office.
We would like to thank Kathleen Bardovi-Harlig,Kevin Gregg,and Foong Ha Yap for their helpful comments on an earlier version of the paper.We also thank Roger Andern,Yuko Koga,Minoru Tanaka,
and the participants of the minar on the crosslinguistic acquisition of ten-aspect taught by Y. Shirai(Cornell University,Spring1997)for their insightful discussions of theoretical and methodological issues related to this paper,and Jim Gair for his experti in South Asian languages.Needless to say,all the deficiencies are ours.We would also like to thank Yuichi Konishi for transcribing the interview data,the3Chine students for their cooperation in the interview, and the Japane Language Program for Overas Students at Daito Bunka University for recruiting them.We also thank the students and rearchers at Nagoya University who kindly agreed to participate in Experiment2.A preliminary report on Experiment1was prented at the19th Annual Boston University Conference on Language Development,and appeared in the conference proceedings.
This rearch was supported in part by grants from the Japane Ministry of Education,Science and Culture(No.05881078and08610530)and a special grant from Daito Bunka University Institute of Language Education Re-arch to Y.Shirai.
Correspondence concerning this article may be nt to Yasuhiro Shirai, Department of Modern Languages,Cornell University,Morrill Hall,Ithaca, New Y ork14853-4701,U.S.A.Internet:ys54@cornell.edu
245
yale university246Language Learning Vol.48,No.2 ,Robison,1995).Although such a tendency
fits the acquisition of typologically similar languages(Eng-
lish,French and Spanish),the universal claim requires
testing against non-Indo European languages.This study
tested the Aspect Hypothesis using L2Japane(JSL)
data.Experiment1showed that in oral production data,3
learners of JSL showed stronger association of achieve-
ment verbs and past marking(-ta),and of activity verbs and
progressive/durative marking(-te i-),than native speakers.
Experiment2showed that in an acceptability judgment
test of ten-aspect forms,17learners of JSL had more
difficulty correctly judging the acceptability of the progres-
sive/durative marker(-te i-)with achievement verbs than
with activity verbs.The results extend the applicability of
婴儿学英语the Aspect Hypothesis to a non-Indo-European language.
The acquisition of ten and aspect is an important compo-nent of cond language acquisition(SLA).In most languages,a learner uttering a ntence containing a finite verb has to linguis-tically encode either ten or aspect(or both).Also,from a func-tional viewpoint,ten-aspect is an important communicative device to establish time reference as well as to express how the speaker views the temporal contour of a situation.Second lan-guage(L2)rearchers have extensively investigated the develop-ment of temporality as a multi-faceted,complex phenomenon.One area of investigation concerns how L2learners express temporal information;rearch has found that learners u various devices, such as calendric reference or time adverbials,to establish tempo-ral reference(Bardovi-Harlig,1992;Dietrich,Klein,&Noyau, 1995;Klein,1986;Meil,1987;Schumann,1987).
The other intensively investigated area of ten-aspect de-velopment,the acquisition of ten-aspect morphology,has shown an interesting universal pattern in both first(L1)and cond language ,Andern&Shirai,1996).Several re-archers have claimed that the development of ten-aspect morphology in L1acquisition of various languages is strongly influenced by the inherent mantic aspect of the verbs to which
Shirai and Kurono247
the inflections are attached.This tendency,often referred to as the Aspect Before Ten Hypothesis or the Defective Ten Hypothe-sis,has been obrved in French(Bronckart&Sinclair,1973), Italian(Antinucci&Miller,1976),Greek(Stephany,1981),and English(Bloom,Lifter,&Hafitz,1980;Shirai&Andern,1995), among others(although Weist,Wysocka,Witkowska-Stadnik, Buczowska,&Konieczna,1984,prent a conflicting finding in the acquisition of Polish).The same tendency has been obrved in SLA.Robison(1990,1995)and Bardovi-Harlig and Reynolds (1995)in the acquisition of English,Andern(1991)and Ramsay (1990)in the acquisition of Spanish,and Bardovi-Harlig and Bergström(1996)in the comparative study of English and French have shown that the L2learner’s acquisition of ten-aspect morphology is guided by inherent aspect.The prent article reports a study on the acquisition of ten-aspect mor
phology in L2Japane.We argue that L2learners of Japane,despite the typological difference in the target language aspectual system, show the same tendency that has been obrved in previous studies on other languages.
The Aspect Hypothesis
First,we describe the four categories of inherent aspect of verbs,defined in terms of the temporal properties of the situation the verb(phra)describes.1This classification is derived from the work of scholars such as Vendler(1967)and Mourelatos (1981):
1.Achievement—that which takes place instantaneously,and
is reducible to a single point in ,recognize,die,reach the summit,etc.).
2.Accomplishment—that which has some duration,but has a
necessary ,run a mile,make a chair,build a hou,etc.).
3.Activity—that which has duration,but without a necessary
,run,walk,play,sing,etc.).
248Language Learning Vol.48,No.2
4.State—that which has no dynamics,and continues without
additional effort or energy being ,e,love,hate, want,etc.).xiao
criaIn terms of mantic features,achievements are punctual whereas all other categories are durative.Achievements and accomplish-ments are both ,having an inherent endpoint)whereas activities and states are atelic(non-telic).States are non-dynamic by definition whereas activities,accomplishments,and achieve-ments are dynamic.
As mentioned above,studies on the acquisition of verb mor-phology that systematically investigated the relationship between verb morphology and inherent aspect have consistently indicated that children acquiring an L1and adults acquiring an L2are strongly influenced by the inherent aspect of the verb to which the morphology is attached.To take an example from the acquisition of English,learners initially give the past ten inflection predomi-nantly to achievement verbs,and rarely to state verbs.Both L1 children and L2adults initially restrict the progressive inflection to activity verbs;L1children rarely overextend it to state verbs (Brown,1973;Kuczaj,1978),whereas adult L2learners some-times do(Robison,1990).
adverbialRearch has obrved this tendency,with some variation due to language-specific differences,in the acquisition of French, Italian,Polish,Spanish,Portugue,Greek,Chine,and other languages in L1acquisition,and in Spanish,English,and French in SLA.The generalization,which we call the Aspect Hypothesis (following Andern&Shirai,1994;Robison,1995),can be sum-marized as follows(Andern&Shirai,1996;Bardovi-Harlig& Bergström,1996;Shirai,1991):
1.Learners u(perfective)past making on achievement/ac-
complishment verbs,eventually extending u to activity and state verbs.
2.In languages that encode the perfective/imperfective dis-
tinction morphologically,imperfective past appears later than
Shirai and Kurono249
perfective past,and imperfect past marking begins with sta-tive and ,atelic)verbs,then extends to accomplish-ment and ,telic) verbs.
3.In languages that have progressive aspect,progressive
reali是什么意思marking begins with activity verbs,then extends to accom-plishment/achievement verbs.
aeis4.Progressive marking is rarely incorrectly overextended to
stative verbs (in L1acquisition).
Although there appear to be some disagreements(Klein, Dietrich,&Noyau,1995;Meil,1987)2,recent studies have es-tablished a solid empirical basis for the Aspect Hypothesis in both naturalistic and classroom SLA in English,Spanish and French (Andern,1994;Bardovi-Harlig&Reynolds,1995;Robison,1990, 1995;e Andern&Shirai,1996;Bardovi-Harlig,1995a for review).The tendency occurs not only in frequency counts of oral and written production data but also in paper-and-pencil-tests (e.g.,Bardovi-Harlig&Reynolds,1995),in which learners have more difficulty in supplying appropriate forms in obligatory con-texts and where the pattern of their respon is consistent with the Aspect Hypothesis (e.g.,more accurate u of past ten form with achievement verbs than activity or state verbs). The gener-alization is so robust that Bardovi-Harlig and Bergström(1996) stated that“we can conclude that untutored cond language learners,foreign language learners,and instructed cond lan-guage learners all show remarkably similar quences”(p.323). This developmental ,past ten form starts with achievement verbs,and progressive starts with activity verbs) arguably has the status of a universal in SLA.
日语等级考试网
However,the studies concerning the Aspect Hypothesis have investigated only typologically similar languages (English,Span-ish,and French);to claim universal status for the Aspect Hypothe-sis,rearch must cover the acquisition of non-Indo-European languages,too.Bardovi-Harlig and Bergström(1996) stated:
