A Survey of the Mixed Use of He/She for Chinese Freshmen in English Majors

The mixed use of he/she in oral English is a hot topic in the field of second language acquisition in the past decades and the related study is numerous. However, there are few studies on the misuse of the two pronouns in oral English for English majors in Chinese universities. This study aims to explore the current situation and characteristics of he/she misuse in oral English of Chinese freshmen majoring in English and analyze the factors that cause the misuse, so as to arouse the learners’ awareness of the error. This study focuses on two research questions: (1) What are the characteristics of he/she misuse in spoken English of Chinese freshmen majoring in English? (2) What are the reasons for the misuse? Based on the selected spoken language materials, this study explores the mixing types and error rate of he/she and analyzes the characteristics of the misuse through data. Through the questionnaire, this thesis study investigates the causes of the misuse. It is found that the mixed use of pronoun he/she is serious in the spoken language of English major freshmen, especially the misuse of “she” into “he”. The factors that cause the misuse are complex, involving mother tongue, attention distribution and working memory, pronunciation relationship, transfer of training, and social environment and personal factors, among which the analysis from articulatory phonetics deserves more attention.


Literature Review
, drawing up the "speaker blueprint", joins other researchers in slicing the production process into three main independent components: conceptualization, formulation, and articulation. Conceptualization refers to the preparation stage of message generation of language, including monitoring the process of forming the language; formulation refers to the construction of a fully specified linguistic message, involving grammatical encoding as well as phonological encoding; and articulation converts this abstract message into motor commands to the muscles. The three components are hierarchically ordered, and then the output from the upper stage can serve as the input toward the lower stage, which provides a theoretical basis for the study of the mixed use of the pronouns "he" and "she".  Levelt (1989) Antón-Méndez (2010a) holds the source of the error must be in the message generation stage of conceptualization. In a subsequent study by Antón-Méndez (2010b), it demonstrates that the gender error on his/her is due to the negative transfer of the mother tongue. As for Dong and Li (2011), the results show that the mixed use of pronouns he/she by Chinese EFL (English as a Foreign Language) learners is not a choice error in the form, but the lack of gender information of pronouns in the message generation stage. These researches are based on the message generation theory. Many other aspects on explaining the misuse phenomenon have been conducted like working memory and distance (Clark & Sengul, 1979), image-level theory (Tao, 2003), gender discrimination in language (Jiang, 2008) and basic level categories (Zhao, 2011).
However, very few scholars study it from the perspective of phonetics. According to Wei (2003), there are four initial consonants with retroflex sounds in Mandarin: Zh, Ch, Sh, R (among them, Zh and Ch are both affricates and aspirated). There is no initial retroflex sound in Zhuang language, so it is often read as other initiatives in Mandarin. Part of the Chinese show difficulty in articulating retroflex sound. Keating (1991) Vol. 11, No. 1;2021 somewhat different kind of retroflex fricative is found in Chinese and that is "sh". There exits similarity in articulation between /ʃ/ in English and "sh" in Chinese. It is noteworthy that this oral misuse of Chinese English majors can be analyzed through articulatory phonetics theory.

Methodology
This part will involve research questions, research methods, and data collection. The methods of research cover two ways-a corpus survey as well as a questionnaire survey so as to resolve the two research questions.

Research Questions
Although scholars at home and abroad have done some research on the confusion of "he" and "she", few people study it in terms of freshmen of English major. The main problems that this study aims to probe into are as follows: (1) What are the characteristics of he/she misuse in spoken English of Chinese freshmen majoring in English?
(2) What are the reasons for the misuse?

Research Methods
In order to answer the above two questions, two methods are adopted on the strength of the actual situation and specific acquirement of this study. They are:

1) Corpus Survey Method
This corpus comes from the audios of the final oral English test of the 2019 grade freshmen in the school of foreign languages of North China Electric Power University. The foreign teacher sets questions involving the daily life around students, personal values, and other contents. The students randomly select the questions and are given two minutes to prepare for the speech. The average duration of each audio is 3 minutes, with 61 recorded.

2) Questionnaire Survey Method
This questionnaire adopts the setup of Chinese questions and is designed for all English majors. There are 14 questions in total. The first three questions are to record the subjects' school, major, and grade. Questions 4 to 9, 11, and 13 are in the form of multiple choices. The possible causes of misuse phenomenon are analyzed from different perspectives. Question 10 is to make statistics on the number of people whether there is mixed-use, question 12 is to make statistics on the frequency of mixed-use of the third person nominative pronouns, and the last question is to explore and supplement the causes of mixed-use in the form of short answer questions.

Data Collection
In this section, it contains corpus data collection and questionnaire data collection. Hence, collection methods, data size, and data requirements will be presented.

1) Corpus Data Collection
After excluding all these audio texts that cannot confirm the gender of antecedent (for example the doctor), the rest are composed into sub-corpus M1, and M1 has 52 copies. Subsequently, the requirement is to select those sentences with the wrong gender from M1 to form sub-corpus M2, which has 31 copies.
In line with the statistical method of Dong Yanping and Jia Tingting (2011), the error rate is calculated according to the percentage system, and the following three values are calculated: 1) the frequency of each singular third-person pronoun in M2 (such as heW and sheW); 2) the frequency of each pronoun in M1 (such as heT and sheT); 3) the error rate of each pronoun. This survey defines "he-error rate" as "he should be used but is misused into she". The formula for calculating the error rate of the pronoun "he": sheW／(heT-heW＋sheW), and the formula for calculating the error rate of "she": heW／(sheT-sheW＋heW).

2) Questionnaire Data Collection
There are 187 questionnaires and 185 are valid ones. Two of them are from students majoring in Japanese and Pedagogy, which is not suitable for the requirement of the study. According to the data, the majority of the subjects are from North China Electric Power University, a small part is from Dalian Foreign Studies University and Yan Shan University and other universities, and all majors are English or Translation. Students from grade one to grade four all participated in the survey, and the participation rate of each grade was relatively average.

Analysis and Discussion
Analysis and discussion are the main part of the whole thesis, so based on the two-research method, it will be ijel.ccsenet.org International Journal of English Linguistics Vol. 11, No. 1;2021 fallen into two sections-corpus-based analysis and questionnaire-based analysis.

Corpus-Based Analysis
In this section, corpus data are supposed to be analyzed and discussed in detail so that a clear situation of the mixed use "he" and "she" can be grasped.

Statistical Data from Corpus
Through the statistical method of Dong Yanping and Jia Tingting (2011), sub-corpus M1 and sub-corpus M2 are established, which helps count the numbers of correct pronouns and the misuse ones up, and then the error rate of one pronoun can be calculated. For instance, the formula for calculating the error rate of the pronoun "he": sheW／(heT-heW＋sheW). See Table 1 for statistical results: (1) Negative antecedents appear more often than positive antecedents. In 25.56% cases, the positive pronoun (he, him, his, himself) was mistakenly used to replace the negative antecedents, while in 5.26% cases, the negative pronoun (she, her, her, herself) was used to replace the positive ones.
(2) The nominative case and the possessive case are more likely to be mixed (18.28% vs 31.82%); the objective case and reflexive pronoun are less likely to be mixed (7.14% vs 0.00%). Among them, the error rate of possessive "her" is the highest (i.e., when people should have said "her", they express into "his"), which is 33.33%; the error rate of reflexive pronoun is the lowest, which is 0.00%.

Situations for Misuse
According to the survey results, the following misuse situation can be described: (1) The gender confusion of the third person singular pronouns is very serious in oral English. In the 52 spoken texts involved in this study, there are different degrees of gender misuse of pronouns.
(2) There are some rules for the gender confusion of the third person singular pronouns. Students are always more inclined to misuse negative pronouns, i.e., they use positive pronouns where they should have used negative pronouns (e.g., When students express "she", they speak out "he").
It should be noted that the actual error rate of third-person pronoun mixing should be higher than that reported here because this survey excludes all texts that are difficult to judge the gender of antecedents. For instance, the example (the doctor) mentioned at the beginning of the article is excluded. This antecedent here is a noun without gender mark in English, and the use of context pronouns also makes it impossible to judge whether the doctor is a man or a woman. Such texts tend to have a higher error rate.
Therefore, from Table 1, we can see that the mixed use of nominative he/she in oral English by freshmen in English major is severe, which is still worthy of attention by learners and researchers.

Questionnaire-Based Analysis
In this section, the questions in the questionnaire will be fully discussed and a mass of data will be displayed by graphs.

Statistical Data from Questionnaire
As mentioned above, the first four questions are about the statistics of background information in the survey. The rest is mainly about the investigation of the reasons for the mixed use in he/she oral English and the discussion.    Figure 4 ndents, they h egories, namel t into "never", r" mixed-use the probability r" mixed-use i s of study, the .90%; in the g mes" 32%, and ysis, we can se e versa. ttended a onen 6 is to expl ically learning ng before enter have studied E ly 5−10 years "seldom", "so is 1.82%, the y of "always" is 2.86%, "seld probability of group of 15−20 d "always" 0% ee that the long -to-one/one-tolore whether t English, espe ring the univer tion 6-The Rat he students ha of the students ved profession   ⑬ Too little practice, if you speak too much, you will remember more clearly and form a memory ⑭ Among languages, there are nouns of both genders (such as various occupations). People are accustomed to using "he" to refer to those kinds of nouns.
⑮ Influenced by talkers' environment, their habitual address choices will be affected by their friends' gender.
⑯ Influenced by their own gender, they prefer to use the same-sex third person.
⑰ It may be influenced by the pronunciation relationship between "he" and "she".
⑱ Because of too many contents needing express, people forget the subject gender, which affects the use of pronoun.

Reasons for Misuse
The explanations given by the students in the questionnaire are simple, but they provide the most authentic first-hand information for us to accurately explain the causes of he/she mixed-use in oral English of students in English majors. On a basis of relevant linguistic theories, this section analyzes the reasons of the mixture of Chinese EFL learners' third person nominative pronouns.
(1) Affected by Mother Tongue When it comes to the reasons for the mixed use of the third person nominative pronouns, it is greatly influenced by mother tongue, like answer ①. First of all, message generation theory has to be mentioned. According to the speech production model of Levelt (1989), we know that language message generation occurs in the conceptualizer. There are differences in information needs between the two languages. "他" and "她" in Chinese do not differ in pronunciation; The third person nominative pronouns in English have different pronunciations and meanings. Bilinguals may be able to prepare language information according to L1 when speaking L2 (for example, no pronoun gender information is required). Speaking L2 may have no grammatical problems, but its content is abnormal (for example, pronoun gender error) which affects the wrong use of pronoun he/she.
Besides, when people gain new knowledge, they will certainly involve the transfer of old knowledge. If the old knowledge promotes the learning of new knowledge, it is called positive transfer; otherwise, if the old knowledge hinders the learning of new knowledge, it becomes negative transfer. Antón-Méndez (2010b) holds even high-level bilinguals are influenced by mother tongue transfer in second language speaking. For Chinese English learners, the negative transfer of their mother tongue often affects their second language learning. English and Chinese belong to different language families. They are quite various in phonetics, vocabulary, grammatical structure and so on. Therefore, in the process of constructing learners' interlanguage, there are many components of "negative transfer" of their mother tongue. As for the third person single pronoun, in Chinese, "他" and "她" are different in writing, but there is no difference in pronunciation; in English, there are obvious varieties in writing and pronunciation between "he" and "she". Hence, in oral communication, the learners are easily influenced by Chinese and make mistakes frequently.
(2) Affected by Attention Distribution and Working Memory Through answers ⑤⑥⑦⑨⑫, we can conclude that attention distribution performs a certain function in he/she mixing. Under Tao (2003), he emphasizes the mixed use is due to meaning transmission, focusing on the overall content of transmission and the prominence of subject and predicate. In terms of this questionnaire survey, most of the English learners pay much attention to the expression of the overall meaning of sentences in the process of communication, thus ignoring the use of sentence structure and form. At the same time, some students point out that the lack of prominence of the antecedents leads to the wrong choice of nominative pronouns.
Of course, answer ⑱ also indicates that working memory has a related effect on the mixing of "he" and "she". Clark and Sengul (1979) finds that the greater the distance between anaphora and antecedent, the longer time it takes to understand. This also suggests that the longer the distance between anaphora and antecedent, the more requirements for working memory. Therefore, if the distance between personal pronouns and reference words is too far, for learners with poor working memory ability, they are intensely easy to forget the reference words when expressing, thus affecting the use of personal pronouns.
(3) Affected by Pronunciation Relationship In English, the pronunciation of "he" is /hi:/, and "she" is /ʃi:/. Although the consonants of the two words are different, they both have the same vowel pronunciation-the long vowel /i:/. Therefore, when the speaker is faced with the choice of "he" and "she", it is very easy to misuse because of the close relationship between pronunciation.
However, although the two words have similarities in pronunciation, their consonant pronunciation is different, which also causes the misuse of pronoun. For many Chinese people, when they speak Mandarin, there is no distinction between blade-alveolar and retroflex, that is, the initial "s" and the initial "sh". Most people tend to pronounce "sh" as "s", so it is difficult for Chinese students to pronounce "she" well. According to IPA (International Phonetic Alphabet), when consonant /ʃ/ is pronounced, the tongue end is close to the back of the gingiva, the tongue body is raised close to the upper jaw, the lips are slightly rounded and slightly protruded, and friction sounds are formed when air flow passes through. It's a voiceless consonant and the vocal cords don't vibrate. Pan (2012) holds that the Chinese initial "sh" is a tip to gingival affricate (abbreviated to as post tip affricate), which is similar to pronunciation of /ʃ/. Hence, we can deduce that Chinese students also have difficulty in pronunciation of /ʃ/. According to Wei (2003), though people are easy to pronounce "H" as "F" when the initial "H" and vowel "U" is spelled together in some parts of southern China, there is no misreading when "H" is spelled with other vowels. So, there is no big problem for Chinese second language learners to pronounce consonant /h/.
From what have been mentioned above, students are more inclined to "he" in the choice of "he" and "she".
(4) Affected by Transfer of Training Zheng (2011) points out that, on the one hand, transfer of training refers to the inappropriate teaching in foreign language, which causes learners to generate incorrect understanding of some language points and grammar rules of the target language. These misunderstandings are often deep-rooted due to the initial stage of learners' learning, that is, the period when language habits are formed; on the other hand, language improper use of speech materials can also lead to training transfer. In the questionnaire survey, answer ④⑧⑪⑬ argues that people usually take this problem less seriously when choosing words and making sentences, oral practice is too little, and "he" has more opportunities to be used in the teaching materials and teachers' expressions. All of these belong to transfer of training that contributes to the misuse of "he" and "she".

(5) Affected by Social Environment and Personal Factors
From the perspective of history, there has been the idea that men are superior to women since ancient times. On this basis, there will be gender discrimination in language. According to Yu (2015), all specification is designed for men. The male language is the standard while the female language is only a subsidiary or a variant. "he" stands for men, "she" stands for women, and people are more inclined to use "he". Therefore, sexism can also lead to language choice.
The family and interpersonal environment also affect the choice of vocabulary to a certain extent. According to Sun (2010), sexuality is the ratio of male and female features in discourse. There are a lot of male characteristics in the discourse, so the masculinity of the discourse is high, and vice versa. When people live in a more masculine environment, the degree of discourse tends to be masculine; on the contrary, it tends to be feminine.
Personal emotion is another vital factor, like answer ②. Emotional factors are mainly related to the speaker's emotional state of mind. When a person is excited, anxious or nervous, the expression of language will be greatly affected, giving rise to repetition or other verbal errors. In teaching, we also find that as learners answer questions or make speeches in class, the frequency of confusion between "he" and "she" is relatively high, which may be the result of anxiety and tension.

Conclusion
Through the research of the corpus, we know that the gender confusion of the third person singular pronouns by freshmen in English major is far more serious and regular in oral English. Both he/she and him/her are misused, and people are prone to use the third person positive pronoun and ignore the third person negative pronoun. Additionally, the actual error rate of the third person pronoun mixing should be higher than that reported here, because the texts which do not clearly reflect the antecedents is removed from the corpus in advance. In the research of the questionnaire, there are five factors that may contribute to the misuse of the pronoun "he" and "she", including mother tongue, attention distribution and working memory, pronunciation relationship, transfer of training, and social environment and personal factors. Among them, pronunciation relationship is the perspective that many scholars have not got involved with. Some of the responders point out that it is easy for them to confuse the pronunciation of he/she because the two hold some similarities in articulation to a certain degree, and also show differences in pronunciation. In the future, it is hopeful that we can check our findings from more levels of EFL and make a significant breakthrough in he/she pronoun research.