1

The interaction of lexical and grammatical

aspect on language processing

Foong Ha YAP, Stella KWAN, Patrick CHU, Emily YIU,

Stella WONG, Seongha RHEE*, Yasuhiro SHIRAI#

Chinese University of Hong Kong Hankuk University of Foreign Studies*

University of Pittsburgh#

2

Acknowledgements

Research Assistants

Huisung Kang Mikyong Shin

Computer Technician

Hyung-Jun Kim

Artists Calvin Chan Kimmee Lo Grankie Wong

3

Funding

We gratefully acknowledge support from:

Direct Grant 2004-05 (#2010255) from the

Chinese University of Hong Kong

Competitive Earmarked Research Grant 2005-07 (#2110122) from the Research Grants Council of Hong Kong

4

Overview

Introduction to aspectual system Inherent lexical aspect vs. grammatical aspect

Background on aspectual asymmetry The notion of ‘perfective advantage’

Mental Representation of events Previous studies in English, Cantonese, Mandarin and Japanese

Aim of the present study

Korean reaction time study

Methodology Results and discussion Conclusion

5

Introduction to Aspectual System

Aspectual System

Lexical aspect

(situation aspect)

Grammatical aspect

(viewpoint aspect)

state activity accomplishment achievement

Perfective Imperfective

semelfactive

6

Inherent Lexical Aspect

Accomplishment: [+telic], [+durative]

E.g. sakwa-lul mekta

eat ACC apple ‘eat an apple’

Activity: [-telic], [+durative]

E.g. swuyenghata

swim ‘swim’

7

Introduction to Aspectual System

Aspectual System

Lexical aspect

(situation aspect)

Grammatical aspect

(viewpoint aspect)

state activity accomplishment achievement

Perfective Imperfective

semelfactive

8

Grammatical Aspect

Imperfective marker –ko iss

depicts an ongoing action e.g. ku namca-nun swuyengha-ko iss-ta.

the-man-TOP swim - PROG-DEC

‘The man is swimming.’

Perfective marker -ess

depicts a completed action

e.g. ku namca-nun swuyenghay-ss-ta.

the-man-TOP swim - PERF/PAST-DEC

‘The man has swum.’

9

Processing Aspectual Asymmetries

English Magliano & Schleich (2000)

Madden & Zwaan (2003)

Cantonese, Mandarin Chan et al. (2004)

Yap et al. (2004)

Japanese Yap et al. (in press)

10

time space

protagonist

intentionality

causation

Reference: Zwaan & Radvansky (1998)

Our mental representation of events

11

Temporal information & event representation

Magliano & Schleich (2000)

Different aspectual markers impose different interpretations on the temporal duration of activities.

12

Madden & Zwaan (2003)

Are perfective constructions processed faster than imperfective ones?

Focus: accomplishment verbs

Temporal information & event representation

13

The subject was shown a sentence …

Either

The boy was drawing a picture. (imperfective)

Or

The boy drew a picture. (perfective)

Previous methodology

14

The subject was then shown two pictures …

ongoing vs. completed

56% 76% Accuracy rate:

15

Aspectual asymmetries (processing speed) across different languages

English (Medden &

Zwaan, 2003)

Mandarin (Yap et al., 2004)

Cantonese (Chan et al., 2004)

Japanese (Yap et al., in press)

Perfective NA 1315 ms 1118 ms 1512 ms (-ta)

Imperfective NA 1397 ms 1253 ms 1809 ms (-teiru)

Speed

Advantage

NA perfective perfective perfective

Accomplishment verbs only

16

Aspectual asymmetries in Cantonese (Yap et. al.,

2006)

Activity verbs Accomplishment verbs

Perfective (zo2) 1211 ms 941 ms

Imperfective (gan2) 1125 ms 1032 ms

Speed Advantage imperfective perfective

Processing speed for activity and accomplishment verbs

Perfective Advantage may not be observed across all verb types

17

Aim of the present study

Examines 2 Korean aspect markers:

Perfective –ess vs. Imperfective –ko iss

Inclusion of sentences without any aspect markers (‘null’ sentences) for comparison

e.g. ku namca-nun swuyengha-ta.

the-man-TOP swim -DEC

‘The man swims.’

Focus on accomplishment verbs and activity verbs.

18

Methodology

Task

• Sentence – Picture Matching Decision Task

Subjects • 30 university students from Hankuk University of Foreign Studies, Seoul

Materials • 96 sentences (48 target sentences and 48 distractors) • 8 trial practices

19

Experimental Procedures

B. Presentation of a sentence in visual form (in Hangul)

e.g. ku namca-nun swuyengha-ko iss-ta. the-man-TOP swim- PROG-DEC ‘The man is swimming.’

C. Presentation of a mask ‘XXXXXXXXXX’

D. Presentation of a picture:

A. Presentation of a mask ‘XXXXXXXXXX’

20

Demonstration

21

XXXXXXXXXXXXXXXXXXXXXXX

22

그 남자는 수영하고 있다.

23

XXXXXXXXXXXXXXXXXXXXXXX

24

25

26

L R (match) (mismatched)

Keyboard

27

Experimental Design Two within-subject factors for the visual prime sentences: 1. Lexical aspect activity verbs vs. accomplishment verbs 2. Grammatical aspect perfective marker vs. imperfective marker

vs. null sentences (without grammatical aspect markers) e.g. ku namca-nun swuyengha-ko iss-ta. the-man-TOP swim- PROG -DEC ‘The man is swimming.’

Lexical Aspect: Activity Grammatical Aspect: Imperfective

28

Results (Reaction time)

Imperfective –ko iss

Perfective ­-ess

Null Speed Advantage

Activity Mean = 993ms SD = 399

Mean = 1201ms SD = 425

Mean = 916ms SD = 371

Accomplishment Mean = 1050ms SD = 402

Mean = 1134ms SD = 400

Mean =1133ms SD = 429

Imperfective p<.000

No difference p=.014

1. Comparison between Imperfective and Perfective 2. Comparison between Null sentence and Imperfective/perfective

Null p<0.000 No difference p=0.075

No difference p=.009 No difference p=.645

3. Comparison between Null sentence in Activity and Accomplishment

Activity p<.000

29

Prototype Theory

Activity verbs e.g. swuyengha-ta.

swim -DEC ‘swim’

[+durative] [-telic] Ongoing phase (swimming) endpoint (completion of the event)

ku namca-nun swuyenghay-ss-ta. the-man -TOP swim - Perf –DEC ‘The man has swum.’

ku namca-nun swuyengha-ko iss-ta. the-man-TOP swim - Prog -DEC ‘The man is swimming.’

Prototype Non - Prototype

Imperfective marker -ko iss

Perfective marker -ess

30

Prototype Theory

Accomplishment verbs e.g. cip-ul cis-ta house-ACC build-DEC ‘build a house’

[+durative] [+telic] Ongoing phase (building a house) endpoint (completion of the house)

ku namca-nun cip-ul ci-ess-ta. the-man-TOP house-ACC build-PERF–DEC ‘The man has built a house.’

ku namca-nun cip-ul cis-ko iss-ta. the-man-TOP house-ACC build-PROG–DEC ‘The man is building a house.’

Prototype Prototype

Imperfective marker -ko iss

Perfective marker -ess

31

Conclusion

Aspectual asymmetry in Korean Activity verbs: imperfective facilitation

Accomplishment verbs: no perfective advantage

The perfective advantage in processing speed is not robust across all verb types.

Prototype theory: More prototypical combinations of grammatical and lexical aspect are processed significantly faster than less prototypical combinations.

32

References

Madden, C.J. & R.A. Zwaam. 2003 How does verb aspect constrain event representation?

Memory & Cognition, 32, 663-672. Chan, Y.H., F.H. Yap, Y. Shirai, & S. Matthews. 2004. A perfective-

imperfective asymmetry in language processing: Evidence from Cantonese. Proceedings of the 9th International Symposium on Chinese Languages and Linguistics, pp. 383-391. Academia Sinica and the Graduate Institute of Linguistics, National Taiwan University, Taipei.

F.H. Yap, Y. Inoue, Y. Shirai, S. Matthews, Y.W. Wong, &Y.H. Chan. (in press). Aspectual asymmetries in Japanese: Evidence from a reaction time study. Japanese/Korean Linguistics, vol. 14. Stanford, CSLI (Center for the Study of Language and Information).

F.H. Yap, Chan, Y.H., Y. Shirai, L.H. Tan, S. Matthews & P. Li. 2004. A perfective-imperfective asymmetry in the human mind: Evidence from Cantonese and Mandarin. Paper presented at 28th International Congress of Psychology, Beijing, August 8-13.