Group 12 Gender in Third-person Singular Pronouns English Mandarin - - PowerPoint PPT Presentation

Gendered and Genderless: The mental organization of Chinese personal pronouns

Group 12

Gender in Third-person Singular Pronouns

English Mandarin Cantonese Masculine He

他 [tha1]

佢 [khøy5]

Feminine She

她 [tha1]

Neuter/ Genderless He/They

他 [tha1]

The Bilingual Gender Error

Mandarin/Cantonese-English bilinguals sometimes use gender mismatched pronouns, i.e., using “he” to refer to females and “she” for males. E.g. Mark visits Starbucks every morning. She thinks their coffee is the best. Ruby visits Starbucks every morning. He thinks their coffee is the best.

?

Self-paced English reading experiment by Dong et al. (2014)

Native English speakers: longer reading times on pronouns that mismatched their antecedents’ gender than those that matched i.e. “mismatch effect” e.g.1 matched pronoun - Mark (male name) is a student. He (masculine) goes to school by bus. e.g.2 mismatched pronoun - Mark (male name) is a student. She (feminine) goes to school by bus.

Self-paced English reading experiment by Dong et al. (2014)

Native Chinese speakers: initial absence of mismatch effect; but appeared when participants were prompted by human pictures that explicitly showed gender information (visual cue) Conclusion: → Native Chinese speakers do not process gender information →Biological gender processing deficiency in the conceptualizer

Cheung (1994)

Participants: 60 children Experiment: 1. Read aloud sentences with two antecedents

• 2. followed by a gender matched or mismatched pronoun

(他 ’He’ or 她 ’She’)

• 3. Offline questions: test the ability to make anaphoric reference

Results: Aware of the gender distinction between 他 and 她 But no significant effect of gender cues on accuracy Conclusion: 他(He) and 她(She) share the same semantic and phonological access codes

ERP Studies

ERP = Event-related potential evidence for gender processing in pronoun resolution in Chinese speakers N400 effect semantic violations P600 effect syntactic violations E.g. 這位女患者情緒低落，醫生鼓勵他振作起來。 This female patient was in low spirits, the doctor encouraged him to cheer up.

Qiu et al. (2012) & Xu et al. (2013)

• When the gender mismatched pronoun occurs in the clause or sentence

immediately after the antecedent e.g.這位女患者情緒低落，醫生鼓勵他振作起來。 This female patient was in low spirits, the doctor encouraged him to cheer up.

• ERP studies have shown both N400 and P600 effects
• Chinese speakers can detect both semantic and syntactic violations in

gender mismatch situations

• shows that both semantic and syntactic processing occur in pronoun

resolution

• Dong et al.’s and Cheung’s results are not consistent with ERP results

Two aims

• 1. Call into question…

Dong et al. (2014) and Cheung (1994): “There is a deficiency in biological gender processing for linguistic purposes in Chinese native speakers.” If Chinese speakers show gender mismatch effect, the idea of “shared semantic entries and a gender processing deficiency in the conceptualizer

• f

Chinese speakers” is rejected → Not yet clear if this is the case in Chinese

• 2. alternative account by examining the

lemma level (gender specification) of the mental lexicon Apply the “Multi-level Spreading Activation Model” (Bock & Levelt, 1994) → 3 levels of representation:

• I. conceptual level (semantics)
• II. lemma level (syntax)
• III. lexeme level (morphology and

phonology)

Multi-level Spreading Activation Model

ask /a:/ /s/ /k/ /ɪ/ /ŋ/ ask ask Legal Phonological Combinations Conceptual Stratum Lemma Stratum Form Stratum ing related concepts syntactic information Morphology Phonology

?

VS

Masculine Feminine Genderless Conceptual Level

(Concepts of gender)

Lemma Level

(Syntax)

Lexeme Level

(morphology + phonology)

?

/th/ /a/ /kh/ /ø/ /y/ 他 她 佢

/th/ /a/ /kh/ /ø/ /y/

他 她 佢 Legal Phonological Combinations

Multi-level Spreading Activation Model

Experimental Design

• 1. Preliminary Study for sample selection
• 2. Self-paced reading
• 3. Offline questions

Preliminary Study

• 1. Self-report by target participants
• Rate their own competence in Standard Chinese and Cantonese

respectively on a scale of 1-10

• Report the number of hours spent on using the written form of each

variety per day → Target participants: senior high school & university students

• 2. Test
• Translation task and reading comprehension

*Helps eliminate potential confounding variables of frequency effect and imbalanced competence

Methodology

• 1. Tutorials & trials: ensure participants are familiar with the gender of the

four proper nouns used in the experiment

• 2. Self-paced reading
• Participants are presented with chunks of sentences
• Proceed chunk-by-chunk by pressing spacebar on the keyboard
• 3. Offline questions
• Identify the referent of the pronoun in self-paced reading
• Answer by pressing “A” or “L” on keyboard, corresponding to the

referents’ positions

Example of the self-paced reading and offline question

大文 很 怕 麗麗， 因為 他 的 脾氣 很 壞 啊！ Dai Man very afraid Lai Lai because 3.M.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because he has a very bad temper!” Question: 誰的脾氣很壞？(Who has a bad temper?) Answer: 大文 (Dai Man) 麗麗 (Lai Lai)

+

大文

很怕

麗麗，

因為

他的

脾氣很壞

啊！

誰的脾氣很壞? 大文 麗麗

Example of fillers

Manipulation on animacy

小明 很 怕 那 蝴蝶， 因為 牠 非常 膽小 啊！ Siu Ming very afraid that butterfly because 3.SG.animal very easily frightened PART “Siu Ming is very afraid of that butterfly because it is very easily frightened!” Question: 誰很膽小？(Who is easily frightened?) 小明 蝴蝶

Manipulation on plurality

小明 很 怕 那 一隻 蝴蝶， 因為 牠們 非常 醜陋 啊！ Siu Ming very afraid that CLF butterfly because 3.PL.animal very ugly PART “Siu Ming is very afraid of that butterfly because they are very ugly!” Question: 句中有多少隻蝴蝶？(How many butterflies are there in the sentence?) 1隻 多於1隻

Independent Variables & Dependent Variables

Independent variables: manipulation: Gender specification Mandarin: Gendered third-person singular pronouns (他/她) Cantonese: Genderless third-person singular pronoun (佢) Dependent variables: Self-paced reading tasks: Reading times Off-line questions: Accuracy *Accuracy for ensuring that participants actually read the sentences *Data from participants with low accuracy are rejected

Conditions

Matched condition

大文 很 怕 麗麗， 因為 她 的 脾氣 很 壞 啊! Dai Man very afraid Lai Lai because 3.F.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because she has a very bad temper!”

Mismatched condition

大文 很 怕 麗麗， 因為 他 的 脾氣 很 壞 啊! Dai Man very afraid Lai Lai because 3.M.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because he has a very bad temper!”

Cantonese, Genderless

大文 好 怕 麗麗， 因為 佢 嘅 脾氣 好 壞 啊! Dai Man very afraid Lai Lai because 3.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because he/she has a very bad temper!” Masculine: 小明, 大文 Feminine: 芬芬, 麗麗

Predictions

Prediction: Gender mismatch effect exists in Chinese

• 1. Longer reading time will be found in the mismatched condition than the

matched condition

• 2. The matched condition will have shorter reading time than the genderless

condition Reading Times: Matched < Genderless < Mismatched

Mismatched: 大文 很 怕 麗麗， 因為 他 的 脾氣 很 壞 啊！ Dai Man very afraid Lai Lai because 3.M.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because he has a very bad temper!” Matched: 大文 很 怕 麗麗， 因為 她 的 脾氣 很 壞 啊！ Dai Man very afraid Lai Lai because 3.F.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because he has a very bad temper!”

Matched: 大文 很 怕 麗麗， 因為 她 的 脾氣 很 壞 啊！ Dai Man very afraid Lai Lai because 3.F.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because he has a very bad temper!” Genderless: 大文 很 怕 麗麗， 因為 佢 嘅 脾氣 很 壞 啊！ Dai Man very afraid Lai Lai because 3.SG POSS temper very bad PART “Dai Man is very afraid of Lai Lai because he/she has a very bad temper!”

Implications

• 1. Matched < Mismatched
• 他 (he) and 她 (she) have different lemma entries, they contain different

syntactic information

• Gender is processed semantically and syntactically
• 2. Matched < Genderless
• Genderless has a longer reading time because there is less semantic and

syntactic information

Reference

Bock, K., & Levelt, W. J. M. (1994). Language production: Grammatical encoding. In M.A. Gernsbacher (Ed.), Handbook of Psycholinguistics (pp. 945-984). London: Academic Press. Cheung, K.Y. (1994). Pronoun resolution in school-age children’s reading. The HKU Scholars Hub. The University of Hong Kong. Dong, Y., Wen,Y., Zeng, X.(2014). Exploring the Cause of English Pronoun Gender Errors by Chinese Learners of English: Evidence from the Self-paced Reading Paradigm. Journal of Psycholinguist Research. 1-15. Qiu, L., Swaab, T. Y., Chen, H. C., & Wang, S. (2012). The role of gender information in pronoun resolution: Evidence from Chinese. Plos ONE, 7(5). Xu, X.D., Jiang, X.M., ＆Zhou, X.L. (2013). Processing Biological Gender and Number Information during Chinese Pronoun Resolution: ERP Evidence for Functional Differentiation. Brain and Cognition. Vol.81(2), p.223-236