Trilingualism, Education and Ethnic Language Subjectivities

ABSTRACT

This study investigates the subjectivities of tertiary students in the Guangxi Zhuang Autonomous Region (GZAR) of the People's Republic of China. Specifically, we investigate self-reported language practices and attitudes in relation to identities in a language ecology with three prominent languages (Zhuang, Standard Chinese [Putonghua], and English). This mixed methods study shows that despite the population size and some educational support, Zhuang language vitality is challenged through subjectivities linked to promotion of Putonghua and Zhuang specific factors (e.g. script and dialectal variations). In addition, the data show multiple interpretations of what mother tongue might mean to people in this region, including those of Zhuang ethnicity. The study leads to insights on multilingualism in this region. It also suggests that a language ecology approach is useful for understanding subjectivities as multifaceted, with multiple influences, helping us better understand why implementation of national policies or community supports on their own are insufficient for language maintenance efforts.

KEYWORDS:

• Ethnic minority language
• language attitude
• multilingualism
• People’s Republic of China
• subjectivity

The People’s Republic of China (PRC) is linguistically and culturally diverse with over 80 languages and 55 recognized non-Han, ethnic minority groups collectively known as minzu (Grey, 2019). Current education and language policies in the PRC endorse a dominant status for Putonghua in the name of national identity and universal literacy, with an emphasis on English as an additional language for international communication and commerce. The languages of a small number of minzu (e.g., Tibetan, Mongolian, Zhuang) are included in the educational system in the name of ethnic autonomy and equality, creating trilingual education sub-systems. Among the minzu, Zhuangzu (Zhuang people) are the most numerous sub-group across the PRC and the dominant ethnic minority group in the Guangxi Zhuang Autonomous Region (GZAR). Prior research has reached contrasting conclusions about ethnic and linguistic identity among the Zhuangzu as well as questioning the long-term vitality of the language (e.g., Feng & Adamson, 2015; Feng & Sunuodula, 2009; Kaup, 2000). Few studies have examined the subjectivities of Zhuangzu in the GZAR, and none has specifically considered the subjectivities of tertiary students—those who have proceeded through the trilingual education system in the region—in terms of the three languages in the education system. This study examines self-reported practices and attitudes toward the three languages of the education system by university students in the GZAR as “the expression of self or the representation of the speaker’s perspective” (i.e., subjectivities) (Sonnenhauser, 2010, p. 278). We address the call for research on attitudes of key education stakeholders (Feng & Adamson, 2018) in relation to trilingual education and minority language vitality against the backdrop of China’s raising international stature (Lam, 2007), as well as nationalist and assimilationist ideology for ethnic minority groups (Beckett & Postiglione, 2012). We see these interconnections (stakeholder perspectives, minority language vitality, international stature, nationalism and assimilationist ideology) as being socio-psychological features of the language ecology (Haugen, 2001/2001) which influence and in turn are influenced by subjectivities.

Subjectivity is related to “our sense of ourselves as members of social categories” (McNamara, 2019, p. 1). Subjectivities are linked to identities which can be bound up in social and familial attitudes and practices (Lu & Guo, 2021; Zheng & Mei, 2021). Our focus is on multilingual subjectivities as “the attitudes, beliefs and ideologies of teachers (students in our study) and other stakeholders in education—towards linguistic diversity and multi/plurilingual practices” (Erling & Moore, 2021, p. 526−527). In this view, “subjectivities” can serve as a hypernym to attitudes, beliefs, and ideologies, recognizing their socio-psychological links with identities within a multilingual language ecology. Specifically, we investigate the subjectivities of tertiary students in a trilingual system in the GZAR including a recognized regional ethnic minority language (Zhuangyu—the spoken language of the Zhuangzu), a dominant national language (Putonghua), and an international language (English).

Subjectivities as attitudes and practices in a multilingual ecology

Attitudes are often considered as a measurable socio-psychological construct and as dispositions which can be ascertained via observable reactions (Sarnoff, 1970). They show connections between the way (some) people view (other) people and their language varieties (Edwards, 1982; Preston, 2014), and towards multi/plurilingual practices (Hall & Cunningham, 2020; Tsiplakides & Keramida, 2010). Thus, attitudes influence how people make decisions about language practices and learner experiences. In addition, attitudes and identities are social mediated (McNamara, 2019) and can be paradoxical not only in the extent to which they represent reality (or not), but also in being both valued and denigrated by the same individual or group for different uses in different contexts. For example, Kubota (2011), in a study of career goals, life experiences, and subjectivities of English learners in private language schools in a rural Japanese setting found that they might see English proficiency as crucial to job success even when their jobs did not require any use of English. Kubota found that English learning could have multiple meanings linked to the larger neoliberal discourse in society, perceptions of the value of English learning, and learner perceptions of competence. Subjectivities of language users are in close relation to the language ecological perspective as language ecology is sociopsychological (Haugen, 2001) and Mufwene and Vigouroux remind us that all language ecologies are settings for language contact and “such contact in the mental space is a consequence of interlocutor’s dispositions to use each other’s language” (2017, p. 77).

Studies looking at language attitudes (Allard & Landry, 1994; Baker, 1992) and specifically language attitudes in China (Shen et al., 2020; Zhang & Tsung, 2019) show the links to language use and vitality. For example, Wei et al. (2021) examined language attitudes of minority language (Chinese Mongolian) university students and the impact of socio-biographical variables. They found positive attitudes toward trilingualism and to each of the three languages (Mongolian, Putonghua, and English) as well as affective attachment to Mongolian, perceived high instrumental value for both Chinese and English, and belief in the high social influence of each language. Wei et al. (2021) also recognized that being marked as Mongolian on identity cards did not necessary mean the individuals were Mongolian speakers; thus, their study included only those who had learned Mongolian at home before starting school. However, they did not consider participant identity in relation to the three languages investigated. While previous research has focused on individual’s attitudes, this can impose limitations on the insights generated¹ if studies do not recognize the crucial links between attitudes, identities, and perception of self in relation to social categories.

Our study acknowledges the subjectivities of language users in a specific language ecology by examining self-reported language practices and attitudes toward the three dominant languages (Zhuangyu, Putonghua, and English) in relation to ethnic identity. Though other languages exist in this multilingual environment (e.g., other foreign languages are taught, and other Chinese varieties are used), these three languages have been particularly promoted through the school years in the current research context, as discussed in the next section. In this multilingual context, links between language attitudes, ethnicity, intra-group identification, and socioeconomics are not straightforward. Sociological and psychological factors are crucial to the subjectivities of language users; these subjectivities, in turn, have the potential to impact the language ecology, especially in terms of Zhuangyu vitality.

PRC as a multilingual nation: Zhuangyu, English and Putonghua

The GZAR, located in Southwest China, is an administrative region inhabited by Zhuangzu—the most populous minority group of China (19.57 million in China [National Bureau of Statistics of China, 2021]). In the GZAR, Zhuangzu account for 31.46% of the total population (GZAR Bureau of Statistics, 2015). According to the Office of Minority Language of Ethnic and Religious Committee in GZAR (2019), 84.95% of the total population of Zhuangzu in GZAR can speak Zhuangyu in their daily life. As this region provides tertiary education to support Zhuangyu (Grey, 2019), as well as Putonghua and English, it is a useful context for better understanding trilingual attitudes and practices as well as potential language vitality.

In terms of the language education system, Zhuangyu is reportedly seen as a “walking stick” (Grey, 2019, p. 22) to bridge Zhuangzu students in primary school until they can independently learn in Putonghua; thus, a transitional bilingual approach.

English has been the dominant foreign language and Chinese students have been the largest international group of English as a foreign language learner in formal educational settings (Qiufang, 2012). This continues to be true though the Belt and Road Initiative (BRI)—proposed by President Xi Jinping in 2013—supports other foreign languages for communication and collaboration (L. Huang, 2019). The BRI also supports Chinese as “the communicative language or another lingua franca” for Belt and Road countries (Gao, 2020, p. 8), in some ways mirroring the promotion of Putonghua as the lingua franca across ethnic groups in China. Within the PRC, education and language policies continue to endorse the dominant status and functions of Putonghua (Ng & Zhao, 2015). These forces might influence the subjectivities of tertiary students and how they see themselves “as members of social categories” (McNamara, 2019, p. 1) in a specific zone of language contact.

Region-based studies on language use and attitudes among different ethnic groups have been undertaken by Feng and Sunuodula (2009), Adamson and Feng (2015), Feng and Adamson (2018), and Feng (2021). These suggest various trilingual models are in place throughout the PRC with hierarchical language tiers placing Chinese varieties (Putonghua and other Han varieties) at the apex, English in the middle, and minority language(s) as stratified among the minority groups. Xu (2019) contended that Putonghua, at the top of the hierarchy, is seen as an “admission ticket” among ethnic minorities in China. According to Tsung (2014), ethnic minorities might worry that the reinforcement of Putonghua in the hierarchy of the Chinese language(s) squeezes the space for minority languages. Zhuangyu was explored based on case studies in Yunnan Province by Yuan et al. (2015) who concluded a depreciative model with weakening Zhuangyu vitality. However, Kaup (2002) found regional, sub-group disparities between Zhuangzu in Yunnan and in Guangxi. These disparities suggest the Zhuangzu/Zhuangyu in Guangxi is an underexplored case.

Given this backdrop, the purpose of this study is to investigate subjectivities of university students in the GZAR where Zhuangyu, Putonghua, and English form a type of trilingual sub-education system for ethnic minority students. The research questions are

1. What are the subjectivities of university students in GZAR towards Zhuangyu, Putonghua and English?

2. How are subjectivities shaped by and shaping of the language ecology?

Following Kubota (2011) and McNamara (2019), we take subjectivities to be the attitudes and identities expressed by participants in relation to the three languages. We note in particular McNamara’s explanation of the dual lens of subjectivity: the perception of self as member of social categories and the consciousness of how one is viewed by others (see also Garrett, 2007; Miller, 2012; Siegal, 1996).

Methodology

Eliasson (2015) in documenting the sociological dimension of ecology cited Hawley (1950) on the analytical unit, “The unit of observation … is not the individual but the aggregate which is either organized or in process of becoming organized … The subject of ecological enquiry is … the community.” To consider this community, we adopted a mixed-method design including a questionnaire distributed to students at eight universities in the region and interviews with 48 students from one minzu university.² Initially, parallel mixed data analyses (Riazi, 2016, p. 189) were used, followed by re-examining the results to consider complementarity. As Riazi explains, “The purpose is therefore not to cross validate results from one phase of the study with the results of the other phase, but rather to illustrate or to clarify the results from one method with the results of the other method” (p. 41). Prior to data collection, the project proposal was reviewed and approved by an accredited, university-based Institutional Review Board (NTU IRB2018–05-023). All participation was voluntary and, in all cases, individual consent was sought in Putonghua before data collection (individually and in writing for the interview, via anonymous written consent preceding the questionnaire).

Questionnaire

Huguet and Lasagabaster’s (2007) questionnaire of language attitudes—used in a transnational European study including majority, minority and foreign languages—was adapted for this investigation. The questionnaire items focused on attitudes and practices and analyses consistently linked attitudes and identities, supporting our view that the adapted questionnaire could surface subjectivities. For our study, items were translated to Putonghua by the first author, a trained translator. Some items which were not applicable to the PRC context were deleted. In addition, in order to identify language varieties, Huguet and Lasagabaster’s (2007) used the labels “L1,” “L2,” “L3.” This worked well in the international study as different participating countries had different language varieties. However, we were interested in languages in one specific context, so our questionnaire referred to Zhuangyu, Putonghua, and English. This influenced several items as in Figure 1.

Figure 1. Questionnaire: Adapted Language Labels.

In the original questionnaire, the assumption seemed to be that mother tongue (original questionnaire item 5) and L1 (original questionnaire item 8) were the same. In our adaptation, the underlying assumptions were that (a) a Chinese variety but not necessarily Putonghua was the mother tongue and initial language of learning, (b) Putonghua was learned at some point in life, and (c) English was learned as L3. These assumptions were based on the multilingual language ecology including the continued use of local Chinese varieties at home (Office of Minority Language of Ethnic and Religious Committee, 2019), the pervasive use of Putonghua in education and larger society, and the required study of English at school and for university admission.

Convenience sampling was employed to distribute the online questionnaire to university students in the GZAR. From September to October of 2018, 576 responses from undergraduate students in eight universities were received. The online questionnaire required all items to be completed so all collected questionnaires (n=576) were considered valid. Internal consistency was considered to be sufficient (Table 1).

Table 1. Internal consistency of questionnaire items (Cronbach’s alpha).

Proficiency			Language use			Attitudes
PTH	Z	E	Interlocutor	Activities	Importance of Z to do activities	PTH	Z	E
.947	α= .903	.910	.894	α= .937	α= .963	.948	α= .909	.905

Interviews

Forty-eight university students from one minzu university in the GZAR³ were interviewed (Table 2). The university offers programs in European languages (e.g., French and German), in Southeast Asian languages (e.g. Vietnamese, Indonesian Bahasa) and in Zhuang Studies. Interviewees were selected based on majoring in Zhuangyu, Putonghua or English language studies (as the language which were most promoted throughout their educational trajectory) and Zhuang affiliation (stated as Zhuangzu on their ID card or claims to be a Zhuang speaker). Participants were recruited through personal contacts and social media (WeChat). Interviews were conducted by the first author in Putonghua.

Table 2. Interviewees by academic major.

	Individual	Group
Zhuang	1	20
Putonghua	3	15
English	0	9
Total	4	44

The interviews were initially conceived as “group interviews.” However, some students could not join the groups and were interviewed individually. Interviews⁴ were semi-structured in three parts—personal information, language use/practices, and attitudes—to allow analysis of complementarity with the questionnaire and to allow participant voices to surface concerning subjectivities. The total interview time was 1,043 minutes (across all participants).

The questionnaire data were processed using SPSS (version 23). Demographic items (e.g., academic major, self-reported language proficiency) were coded using a numerical value as a label to allow for correlational analyses (Appendix A). Attitudinal questionnaire items were calculated by averaging the means (Baker, 1992). Descriptive statistics were calculated to reveal students self-reports on use of Zhuangyu, Putonghua and English on different occasions and to different people. Spearman’s rho correlations and one-way ANOVA tests were performed to examine factors influencing the subjectivities of the three languages.

Qualitative content analysis was employed to analyze the interview transcripts (Hsieh & Shannon, 2005). Analytical procedures included thorough reading of the full data set by the first author followed by collaborative coding of one individual and one group transcript with a second coder who was proficient in both English and Chinese, as well as a researcher in the field. After agreement on key words/phrases was reached, and description of the pattern of sample discourse was formed, the first author coded all transcripts. Subsequently, the first author used the keywords/phrases as well as re-reading of transcripts to analyse main themes related to language practices and attitudes. Coding was refined iteratively to gain an in-depth understanding of interviewee subjectivities. The authors then examined interrelationships across themes and transcripts with a view to understanding interactions of functions, within society, across languages, that is, from a language ecology perspective (Haugen, 2001). The first author took the additional step of revisiting the themes in relation to the specific research questions. Final peer checking was done by showing extracts of the derived themes and research notes as “polished or semi-polished product, such as the major findings” (Creswell & Creswell, 2018, p. 274) to the second author to comment on links to the findings, as well as for considering discrepant cases.

Finally, findings from the quantitative and qualitative analyses were re-examined with a view to complementarity. The authors considered alignments, divergences, and ways in which the different data and analyses might illuminate the responses to the research questions (see Bazeley, 2018; Riazi, 2016) as they brought together the findings to address the research questions.

Findings

First, an overview of the questionnaire results is presented to give a broad perspective on attitudes toward the three languages and correlated uses/practices. Second, a summary of some of the key themes emerging from the interviews is presented, taking into account analytical complementarity. The research questions are addressed in the Discussion.

Questionnaire findings

Student profiles are summarized in Table 3. Items which are not crucial to our discussion have been removed (e.g., “Travel to an English-speaking country”).

Table 3. Student profiles: questionnaires.

Item	Category	Frequency	Percentage
University affiliation	GXU	242	42.01
	GXUN	114	19.79
	GXNU	77	13.37
	NNU	3	0.52
	GXUFL	35	6.08
	XJC	41	7.12
	Others	64	11.11
AY Intake	2015	117	20.31
	2016	111	19.27
	2017	157	27.26
	2018	191	33.16
Academic Major	Zhuang	45	7.81
	Chinese	27	4.69
	English	323	56.08
	Others	181	31.42
Gender	Male	124	21.53
	Female	452	78.47
Ethnicity	Zhuangzu	198	34.38
	Non-Zhuangzu	378	65.63
Hometown sociolinguistic context (i.e. use of the language in the community)	Zhuangyu	158	27.43
	Cantonese	214	37.15
	Guiliu	54	9.38
	Others	150	26.04
Mother tongue	Zhuangyu	97	16.84
	PTH	147	25.52
	Guiliu dialect	79	13.72
	Cantonese dialect	168	29.17
	Others	85	14.76
Hometown population size	>100,000	351	60.94
	≤100,000	225	39.06
Highest education level - Father	High school and below	431	74.83
	Junior college	72	12.5
	Bachelor’s degree	65	11.28
	Master’s degree	5	0.87
	PhD	3	0.52
Highest education level - Mother	High school and below	462	80.21
	Junior college	71	12.33
	Bachelor’s degree	36	6.25
	Master’s degree	4	0.69
	PhD	3	0.52

Results in Table 4 show hierarchical attitudes towards the three languages: neutral but more negative to Zhuangyu, positive to Putonghua, and neutral but more positive to English. As in Table 5, Zhuangyu seems to have little value as an instructional language (item 6) and is seen as not easy to learn (item 4). However, responses suggest university students’ hold a relatively positive affectional attachment to the language and that Zhuangyu has value for intergenerational transmission (items 7, 8, and 9). Though Putonghua is the most highly valued as well as showing relatively high affectional attachment and utility, the rather moderate mean for item 9 suggests that students do not consider Putonghua as the only language with utility. Taken together, the findings suggest that discourses of linguistic instrumentalism or utility (use as a lingua franca) play a role in developing the subjectivities of the students (Gao, 2020; Kubota, 2011). However, ethnic and family identities come into play as well.

Table 4. Attitudes towards the three languages.

	Attitudes towards Zhuangyu	Attitudes towards PTH	Attitudes towards English
Average	48.845	75.430	65.330
Category	Neutral	Favorable	Neutral

Table 5. Descriptive attitudinal statements.

		Zhuangyu		PTH		English
Items		Mean	SD	Mean	SD	Mean	SD
1	I like hearing _Language X_ spoken.	3.08	.883	3.95	.797	3.73	.840
2	_Language X_ should be taught to all pupils in Guangxi Zhuang Autonomous Region.	2.85	.980	4.04	.847	3.60	.903
3	I like speaking _Language X _.	2.76	1.012	3.99	.799	3.63	.929
4	_Language X_ is an easy language to learn.	2.60	.955	3.86	.915	3.05	.998
5	There are not more useful languages to learn than _Language X _.	2.40	.777	3.82	.905	3.32	.908
6	I prefer to be taught in _Language X _.	2.39	.870	4.02	.790	3.32	.900
7	Learning _Language X _ enriches my cultural knowledge.	3.51	.936	4.07	.781	3.98	.779
8	I would not mind marrying a _Language X _ speaker.	3.43	.935	4.12	.810	3.79	.900
9	_Language X_ is a language worth learning.	3.55	.853	4.18	.751	3.98	.808
10	If I have children, I would like them to be _Language X_ speakers regardless of other languages they may know.	2.97	1.007	4.11	.849	3.74	.930

Looking specifically at Zhuangyu and social uses, analyses of 15 items with a four-point scale of “unimportant—important,” showed the only item which rated a mean higher than 3 was “Live in Zhuang area/community.” “Be accepted in the community” had the next highest mean at 2.91. All other items had means between 1.84 and 2.36, indicating relatively low importance for use of Zhuangyu in daily life (Table 6).

Table 6. Students use of Zhuangyu for social activities.

Items	People	Unimportant	A little unimportant	A little important	Important	Mean
1	To make friends	189 (32.81%)	116 (20.14%)	182 (31.6%)	89 (15.45%)	2.3
2	Read	236 (40.97%)	151 (26.22%)	132 (22.92%)	57 (9.9%)	2.02
3	Write	243 (42.19%)	150 (26.04%)	125 (21.7%)	58 (10.07%)	2
4	Watch TV	227 (39.41%)	146 (25.35%)	147 (25.52%)	56 (9.72%)	2.06
5	Get a job	214 (37.15%)	139 (24.13%)	159 (27.6%)	64 (11.11%)	2.13
6	Be liked	194 (33.68%)	140 (24.31%)	177 (30.73%)	65 (11.28%)	2.2
7	Live in Zhuang area/community	66 (11.46%)	52 (9.03%)	180 (31.25%)	278 (48.26%)	3.16
8	Bring up children	168 (29.17%)	131 (22.74%)	177 (30.73%)	100 (17.36%)	2.36
9	Go shopping	251 (43.58%)	142 (24.65%)	134 (23.26%)	49 (8.51%)	1.97
10	Make phone calls	211 (36.63%)	132 (22.92%)	159 (27.6%)	74 (12.85%)	2.17
11	Pass exams	240 (41.67%)	138 (23.96%)	133 (23.09%)	65 (11.28%)	2.04
12	Be accepted in the community	101 (17.53%)	70 (12.15%)	185 (32.12%)	220 (38.19%)	2.91
13	Talk to friends at university	266 (46.18%)	160 (27.78%)	115 (19.97%)	35 (6.08%)	1.86
14	Talk to teachers at university	277 (48.09%)	153 (26.56%)	107 (18.58%)	39 (6.77%)	1.84
15	Talk to people out of university	256 (44.44%)	158 (27.43%)	118 (20.49%)	44 (7.64%)	1.91

In terms of language practices and media, Putonghua was overwhelmingly dominant. English TV watching was popular for those pursuing an English major and English TV watching correlated strongly with higher self-reported English proficiency (F 10.102*** df [4,571]). Table 7 shows the correlations for all variables. Subsequent tables show additional analyses to better understand these interrelationships.

Table 7. Correlations for language attitudes.

									Proficiency			Initial Age									Attitude toward
Variables	G	Intake	School	Major	Zhuang Ethnicity	MT	EF	EM	Z	PTH	EL	MT	PTH	EL	HP	ESC	ETV	SC	PP	HS	Z	PTH	EL
1. Gender	-
2. AY intake	−.128**	-
3. School	.026	−.006	-
4. Major	−.271**	.243**	−.185**	-
5.Ethnicity- Zhuangzu	.012	−.049	.052	−.096*	-
6. MT	.010	−.060	−.043	−.046	.529**	-
7. Education (Father)	−.070	.044	−.052	.049	−.025	−.089*	-
8. Education (Mother)	−.031	.079	−.013	−.030	−.040	−.084*	.679**	-
9. Zhuangyu proficiency^1	−.013**	.026	.066	.069	−.444**	−.368**	.014	.000	-
10. PTH proficiency^1	−.048	−.027	.070	.042	.002	−.038	.206**	.192**	−.035	-
11. English proficiency^1	.135**	−.059	−.042	−.162**	.040	.026	.062	.061	−.110**	.161**	-
12. Initial age, mother tongue	−.135**	.156**	.111**	.055	−.075	−.063	−.078	−.091*	.094*	−.094*	−.034	-
13. Initial age, PTH	−.009	−.042	.020	−.032	.078	.155**	−.314**	−.318**	.040	−.226**	.001	.264**	-
14. Initial age, English	.006	−.074	.009	−.077	−.021	.066	−.255**	−.277**	.122**	−.261**	−.090*	.063	.395**	-
15. Hometown Population	−.005	−.024	.015	−.042	−.087*	−.086*	−.185**	−.192**	.146**	−.127**	−.021	.137**	.209**	.167**	-
16. English-speaking countries	.036	.025	.001	−.022	.021	.069	−.313**	−.333**	.063	−.164**	−.074	.121**	.199**	.208**	.114**	-
17. How often watch English TV	.148**	−.110*	−.011	−.302**	.032	.007	.031	.068	−.049	.104*	.251**	.012	−.080	−.097*	−.503	−.132**	-
18. Sociolinguistic context	.059	−.072	−.023	−.001	.483**	.425**	.003	.012	−.348**	.048	.058	−.086*	−.046	−.115**	−.173**	−.078	.059	-
19. Private or public university	.041	.107*	−.597**	.171**	−.122**	−.037	.038	.111**	−.023	−.084*	−.026	−.098*	−.038	.000	−.085*	−.023	−.017	−.081	-
20. Type of high schools	.158**	−.120**	.112**	.104*	.009	.017	.052	.094*	−.004	−.014	.039	.024	−.049	−.002	−.047	.014	.085*	−.023	.086*	-
21. Attitude -Zhuangyu	−.057	.056	.029	.093*	−.406**	−.278**	−.099*	−.116**	.424**	−.108**	−.040	.136**	.067	.142**	.137**	.099*	−.022	−.383**	.068	−.010	-
22. Attitude - PTH	.023	−.033	−.033	.014	.131**	.018	.112**	.097*	−.130**	.202**	.012	−.037	−.175**	−.149**	−.143**	−.091*	.042	.113**	.002	−.052	−.139**	-
23. Attitude - English	.104*	−.020	.000	−.144**	.052	.023	.042	.039	−.123**	.128**	.195**	−.013	−.122**	−.112**	−.073	−.128**	.252**	.088*	−.025	−.059	−.058	.532**	-

1. Language proficiency is self- reported.

Bold shows statistically significant correlations: *p < .05. **p < .01.

One-way ANOVAs (Table 8) show that self-reported proficiency and initial age of language learning can be linked to university, academic major, mother tongue and sociolinguistic context with some differences by language. Post hoc analyses, as in Table 9, show higher Zhuangyu proficiency was related to attendance at a minzu university (GXUN), Zhuang as an academic major, mother tongue and sociolinguistic context. Somewhat surprisingly, initial age of mother tongue learning was not significantly related to mother tongue or sociolinguistic context. We return to this point later in our discussion.

Table 8. Self-reported language proficiency and initial age of language learning: one way ANOVAs.

	Self-reported Proficiency			Initial Age of Language Learning
	Zhuangyu	PTH	English	Mother Tongue	PTH	English
University	F 6.269*** df (6,569)	F 4.498*** df (6,569)	F 2.139* df (6,569)	F 1.608 df (6,569)	F 1.569 df (6,569)	F 2.981** df (6,569)
Academic major	F 43.613*** df (4,571)	F 6.103*** df (3,572)	F 8.088*** df (3,572)	F 2.652* df (3,572)	F 1.437 df (3,572)	F 1.875 df (3,572)
Mother tongue	F 30.016*** df (4,571)	F 2.736* df (4,571)	F 1.838 df (4,571)	F 1.395 df (4,571)	F 23.752*** df (4,571)	F 20.805*** df (4,571)
Socio-linguistic context	F 49.597*** df (3,572)	F 1.430 df (3,572)	F 1.060 df (3,572)	F 2.568 df (3,572)	F 1.499 df (3,572)	F 4.046** df (3,572)

Table 9. Self-reported proficiency and initial age of language learning: Post hoc Bonferroni.

	Self-reported Proficiency			Initial Age of Language Learning
	Zhuangyu	PTH	English	Mother Tongue	PTH	English
University	GXUN > GXU=GXNU=NNU=GXUFL=XJC=Others	GXUN < GXU=GXNU=NNU=GXUFL=XJC=Others	GXU > GXUN	n.s.	n.s.	GXUFL > GXNU
Academic major	Zhuang > PTH=English= Others	English=others > Zhuangyu	English= others > Zhuang	PTH > others= English= Zhuangyu	n.s.	n.s.
Mother tongue	Zhuangyu > PTH=Guiliu = Cantonese= others	Zhuangyu < PTH	n.s.	n.s.	PTH < Zhuangyu	Guiliu dialect=PTH < Zhuangyu=Others=Cantonese
Socio- linguistic context	Zhuangyu > Cantonese =Guiliu- =others	n.s.	n.s.	n.s.	n.s.	Zhuangyu =Cantonese- =Guiliu- > others

Table 9 also shows that Zhuang majors self-reported as less proficient in English, while English majors self-reported as less competent in Zhuangyu. There were differences for initial age to learn Putonghua and English (Table 8 and Table 9) and for attitudes to Zhuangyu and English (Table 10 and Table 11). These results could suggest that students with Zhuangyu as mother tongue learned Putonghua and English later. Mother tongue was positively correlated with self-reported proficiency in the same language, for example, Zhuangyu and Putonghua (Table 8 and Table 9). Attitudes toward the three languages were related to mother tongue, academic major, and in some cases to proficiency or sociolinguistic context (Table 10). Specifically, Zhuang major students showed more unfavorable attitudes to English while English majors showed more unfavorable attitudes to Zhuangyu (Table 11).

Table 10. Attitudes toward Zhuangyu, PTH and English: One way ANOVAs.

	Attitudes to Zhuangyu	Attitudes to PTH	Attitudes to English
Mother tongue	F 30.016*** df (4,571)	F 4.228*** df (4,571)	F 4.496*** df (4,571)
Academic major	F 30.811*** df (3,572)	n.s.	F 7.538*** df (3,572)
Zhuangyu proficiency	F 45.370*** df (3,572)	F 3.944*** df (3,572)	n.s.
PTH proficiency	n.s.	F 9.293*** df (3,572)	n.s.
English proficiency	n.s.	n.s.	F 12.502*** df (3,572)
Sociolinguistic context	F 49.642*** df (3,572)	n.s.	n.s.

*p < .05. **p < .01. ***p < .001.

Table 11. Attitudes toward Zhuangyu, PTH and English: Post hoc Bonferroni.

	Attitudes to Zhuangyu	Attitudes to PTH	Attitudes to English
Mother tongue	Zhuangyu>Guiliu>others	PTH>others	PTH>Guiliu>others
Academic major	Zhuang>Chinese>English=others	n.s.	English=others=Chinese>Zhuang
Zhuangyu proficiency	Very fluent=fluent>unfluent>very unfluent	Very unfluent>very fluent	n.s.
PTH proficiency	n.s.	Very fluent=fluent=unfluent>very unfluent	n.s.
English proficiency	n.s.	n.s.	Fluent=very fluent=unfluent>very unfluent
Sociolinguistic context	Zhuangyu-speaking community>others	n.s.	n.s.

Table 11 shows that students with Putonghua as their mother tongue had parents of higher education level, a proxy for socioeconomic status. This might suggest the importance which parents of higher education level attach to Putonghua. Considering the positive correlation between Putonghua as mother tongue and attitudes to English (Table 12), it seems likely that students from higher SES families perceive both languages as having high social capital. Zheng and Mei (2021), for example, investigated beliefs of lower/higher SES parents in urban PRC. They found all parents shared a belief in the importance of English but parents from the higher SES community had stronger beliefs about the necessity of investing in Putonghua learning.

Table 12. Parents’ education level, one-way ANOVAs with post hoc Bonferroni.

	Father’s Education Level	Mother’s Education Level
Mother tongue	F 9.230*** df (4,571)	F 9.792*** df (4,571)
Post hoc test (Bonferroni)	PTH >Guiliu =others=Cantonese=Zhuangyu	PTH >Guiliu =others=Cantonese=Zhuangyu

*p < .05. **p < .01. ***p < .001.

While Zhuangyu is an officially recognised minority language and close to 85% of the population are reported to use the language in daily life, questionnaire respondents identified numerous “mother tongues” when given the chance to self-nominate, many of which are not officially recognised (Table 13). Respondents also stated various ages for initial learning of mother tongue, including ages 3−5 or 6−10 which suggests they learned their mother tongue in an educational, rather than home, setting (Table 14). These data imply that respondents had different interpretations of “mother tongue” which may not align with understandings of mother tongue as the home language nor with ethnic identity.

Table 13. Mother tongue identification.

Items	Officially Recognized Varieties	Frequency
Mother tongue	壮语 (Zhuangyu)	97
	普通话 (PTH)	147
	桂柳话 (Guiliu)	83
	粤语/白话 (Cantonese/Cantonese subdialect)	168
	客家话 (Hakka)	31
	瑶话 (Yao minority language)	3
	闽南话 (Hokkien)	1
	廉州话 (a Cantonese subdialect)	2
	潮汕话 (Teochew)	1
	平话 (Ping language)	3
	大冶话 (Daye dialect)	1
	Other varieties (based on individual designations)
	地方/地区/家乡方言 (Local/regional/hometown dialect)	12
	土白话 (Local Cantonese subdialect)	1
	土话 (local indigenous dialect)	4
	宾阳话 (Binyang dialect)	1
	玉林话 (Yulin dialect)	6
	艾话 (Ai dialect)	2
	乐庭话 (Leting dialect)	1
	文桥话 (Wenqiao dialect)	1
	海南话 (Hainan dialect)	2
	蒙山话 (Mengshan dialect)	1
	全州话 (Quanzhou dialect)	1
	地老/佬话 (Dilao dialect)	3
	钦州话 (Qinzhou dialect)	2
	无 (Nil)	1
	其它 (others)	1
TOTAL		576

Table 14. Initial age of language learning.

	Mother Tongue	Putonghua	English
Initial age of learning a language
0–2	414 (71.9%)	166 (28.8%)	2 (0.3%)
3–5	133 (23.1%)	207 (35.9%)	32 (5.6%)
6–10	29 (5%)	197 (34.2%)	345 (59.9%)
11–13	0 (0%)	3 (0.5%)	179 (31.1%)
14+	0 (0%)	3 (0.5%)	18 (3.1%)

Summary of questionnaire findings

In summary, the questionnaire findings show respondent subjectivities through reported attitudes and attachment toward and uses of the three languages. Language attitudes maintain a hierarchical positioning for the three languages: Putonghua > English > Zhuangyu. In terms of use, Zhuangyu was seen as somewhat important as cultural enrichment and for those living in Zhuang communities but, overall, Zhuangyu was not seen as having high social functionality/importance. Putonghua was largely preferred for media, but respondents also commented on English TV watching as a common practice, especially those pursuing an English major or concerned with higher English proficiency. Family SES (using parental education as an indicator) positively correlated with Putonghua as the mother tongue. Those studying Zhuang major and those from sociolinguistic contexts with Zhuangyu use report higher proficiency and more positive attitudes toward Zhuangyu. However, ethnic identity did not have a strong correspondence with linguistic identity; for example, those who identified as Zhuangzu did not necessarily list Zhuangyu as “mother tongue.” Below we present the overview of findings from the interview analyses in relation to the main themes emerging from the questionnaire.

Interview findings

Table 15 shows the profiles of interviewed students, followed by presentation of key themes and illustrative examples. Key themes that emerged in the questionnaire and interviews included (a) language hierarchy and perceived utility, (b) self-reported proficiency, and (c) ethnic identity and mother tongue.

Table 15. Student profiles: Interview.

No.	Name^1	Gender	AY Intake	Academic Major	Ethnicity	Language proficiency			Hometown	Family / Community Language
						Z^2	C^3	E^4
1	Yu	F	2018	Chinese	Zhuangzu	4	3	2	Laibin, Liuzhou	Zhuangyu, Guiliu dialect
2	HJing	F	2017	Chinese	Zhuangzu	2	4	3	Jingxi, Baise	Zhuangyu,
3	Chua	F	2017	Chinese	Zhuangzu	3	4	2	Yongning, Nanning	“Tu Hua”
4	Li	F	2016	Chinese	Zhuangzu	4	3	2	Shanglin, Nanning	Zhuangyu, Yaoyu, Hakka
5	Lin	F	2016	Chinese	Zhuangzu	3	4	2	Donglan, Hechi	Zhuangyu, Guiliu dialect
6	Qian	F	2016	Chinese	Zhuangzu	4	4	2	Longan, Nanning	Zhuangyu, PTH, Cantonese
7	Pei	F	2016	Chinese	Zhuangzu	2	4	2	Yongning, Nanning	PTH, Cantonese
8	Yu	F	2016	Chinese	Zhuangzu	2	4	2	Laibin, Liuzhou	Zhuangyu, Guiliu dialect
9	Rui	F	2016	Chinese	Hanzu	1	4	2	Jingxi, Baise	Guiliu dialect, Zhuangyu
10	Su	F	2016	Chinese	Zhuangzu	1	4	3	Yongning, Nanning	“Cun Hua”
11	Xing	F	2016	Chinese	Zhuangzu	1	4	2	Fangchenggang	Hakka, Cantonese
12	Li	F	2016	Chinese	Zhuangzu	4	4	2	Shanglin, Nanning	Zhuangyu, PTH
13	Lin	F	2016	Chinese	Zhuangzu	3	4	2	Donglan, Hechi	Zhuangyu, PTH
14	Hua	M	2015	Chinese	Zhuangzu	4	3	3	Bama, Hechi	Zhuangyu
15	Fei	F	2017	Chinese	Zhuangzu	3	4	2	Hengxian, Nanning	Cantonese, Hakka, Zhuangyu
16	Dan	F	2015	Chinese	Zhuangzu	3	4	2	Bama, Hechi	Zhuangyu
17	Ning	F	2015	Chinese	Zhuangzu	1	4	2	Yongning, Nanning	Ping language
18	Yyi	F	2018	English	Zhuangzu	3	3	3	Longzhou, Chongzuo	Zhuangyu, PTH
19	Fang	F	2018	English	Zhuangzu	4	3	2	Debao, Baise	Southern Zhuang dialect
20	Huan	F	2017	Chinese	Zhuangzu	2	4	3	Ningming, Chongzuo	Cantonese, PTH
21	Bei	F	2017	Chinese	Zhuangzu	3	4	3	Pingguo, Nanning	Zhuangyu
22	YFang	F	2017	Chinese	Zhuangzu	3	4	3	Liangqing, Nanning	Cantonese, Zhuangyu
23	Hai	F	2018	English	Zhuangzu	4	4	3	Hechi city	Zhuangyu
24	Wei	F	2018	English	Zhuangzu	3	3	3	Longan, Nanning	Zhuangyu
25	Yan	M	2018	English	Zhuangzu	4	4	3	Maonan, Hechi	Zhuangyu
26	Hao	M	2018	English	Zhuangzu	1	4	3	Xiangzhou, Liuzhou	Guiliu Dialect
27	Ye	F	2018	English	Zhuangzu	3	3	3	Zhongshan, Hezhou	Zhuangyu
28	Wan	F	2018	English	Zhuangzu	3	4	3	Hechi city	Zhuangyu, Guiliu Dialect, PTH
29	Yi	F	2018	English	Zhuangzu	3	4	3	Laibin, Liuzhou	PTH
30	Linn	F	2015	Zhuang	Zhuangzu	4	4	2	Wuming, Nanning	Zhuangyu
31	Lun	F	2015	Zhuang	Zhuangzu	2	4	3	Laibin, Liuzhou	Guiliu Dialect, PTH
32	Chun	F	2018	Zhuang	Zhuangzu	2	4	3	Jingxi, Baise	Guiliu Dialect, Zhuangyu
33	Die	F	2018	Zhuang	Zhuangzu	4	4	3	Tiandeng, Chongzuo	Zhuangyu
34	Cyu	F	2018	Zhuang	Zhuangzu	4	3	3	Rongshui, Liuzhou	Zhuangyu, Miao language
35	Lan	F	2018	Zhuang	Zhuangzu	4	4	3	Laibin, Liuzhou	Zhuangyu, Guiliu Dialect
36	XJing	F	2018	Zhuang	Zhuangzu	1	4	3	Hengxian, Nanning	Ping language, Cantonese
37	Wen	F	2018	Zhuang	Yaozu	4	4	3	Dahua, Hechi	PTH, Zhuangyu
38	Zhou	M	2018	Zhuang	Hanzu	2	4	3	Beihai city	Cantonese, PTH, Zhuangyu
39	Zi	M	2018	Zhuang	Zhuangzu	3	4	2	Liujiang, Liuzhou	Zhuangyu
40	Guo	M	2018	Zhuang	Zhuangzu	1	3	2	Liuzhou city	Guiliu Dialect
41	Xia	F	2017	Zhuang	Zhuangzu	3	4	2	Daxin, Chongzuo	Zhuangyu, Cantonese, PTH
42	Mei	F	2017	Zhuang	Zhuangzu	3	4	3	Fengshan, Hechi	Zhuangyu
43	Yi	F	2017	Zhuang	Zhuangzu	4	4	3	Baise	Zhuangyu
44	Luo	F	2017	Yao	Zhuangzu	4	4	3	Youjiang, Baise	Zhuangyu
45	Qi	F	2017	Zhuang	Zhuangzu	3	4	2	Tianyang, Baise	Zhuangyu
46	Jing	F	2015	Zhuang	Zhuangzu	4	3	2	Shanglin, Nanning	Zhuangyu
47	Bi	F	2018	Zhuang	Zhuangzu	3	3	2	Ningming, Chongzuo	Zhuangyu
48	Zhi	F	2018	Zhuang	Zhuangzu	1	4	2	Laibin	Guiliu dialect, PTH

¹All names are pseudonyms.

²Z=Zhuangyu.

³C=Chinese.

⁴E=English.

Of the 48 interviewees, all but three identified as ethnically Zhuang. Almost half (22) were Chinese majors. Nine were English majors, 18 were Zhuang majors and one listed “Yao” as major. In terms of Zhuangyu proficiency, 15 rated themselves as low (“very unfluent” or “unfluent”); all others rated themselves as “fluent” or “very fluent.” For English proficiency, 23 rated themselves as “unfluent” (though none rated themselves as “very unfluent”) and 25 rated themselves as “fluent” or “very fluent.” All rated themselves as “fluent” or “very fluent” in Putonghua. Unlike the questionnaire results, there were no noticeable relationships between academic major and English or Zhuangyu proficiency (i.e., some English majors listed higher fluency in Zhuangyu than English; some Zhuang majors listed lower Zhuangyu proficiency than English).

Language hierarchy and functionality

Corroborating the questionnaire findings of language hierarchy, student interviews showed a prioritisation for Putonghua. This included not only utility (Extract 1, Yu) but also affective attachment and Chinese identity (Extract 1, Pei). This agrees with prior research which has shown the unwavering status of Putonghua in multilingual Chinese society (e.g., Jiang & Dewaele, 2019; Y. Huang & Fang, 2021). The comments also reflected the broader social discourse of Putonghua as lingua franca and showed how the students positioned themselves as members of Chinese society through language.

Extract 1

Yu: It’s not whether you like to learn or not, but a lot of people use it, and we should learn. Otherwise, we can’t integrate into the mainstream. Anyway, it’s very important.

Pei: I think Chinese people should speak Putonghua, use it to talk, to communicate, to work, to study and to live a life. Our current life should be inseparable from Putonghua.

English was seen primarily through an instrumental lens—as a tool or resource for work (Extract 2, YFang). A few students showed positive affective attachment to English feeling it was “beautiful” or “cool” (Extract 3, Wan).

Extract 2

YFang: In my opinion, language, including English, is just a tool. It is also a resource. If you are good at English, you can have more advantages than others. It is competitive advantages.

Extract 3

Wan: I like it… I think English is also beautiful to hear, and I usually watched that kind of American series, and then I feel that it is very cool to watch TV without subtitles.

However, some held negative attitudes toward English, seeing a lack of utility and identifying Putonghua and Zhuangyu as more important (Extract 4, Qi); they distanced themselves from membership in countries such as the United States and Britain and affiliated themselves with China through Putonghua/Chinese.

Extract 4

Qi: I started learning English from kindergarten, my kindergarten was a bilingual class. But now I cannot speak English fluently in dialogue. My English was fine when I was in junior high school, but in senior high school I lost my interest in learning English. I was not interested in it and did not want to learn it … because in this language environment, I feel that we need no English. If I learn Chinese, I can read novels, can read other kinds of Chinese related books. If I learn Zhuangyu I can communicate with families and learn some minority culture, but learning English I really am not interested in it, and I am not interested in British or American culture.

Attitudes toward Zhuang were decidedly mixed with some expressing a “responsibility” regardless of personal interest (Extract 5, Chun) while others viewed the value of Zhuangyu as declining in terms of utility (Extract 5, Ye). As in Chun’s comments, the responsibility to learn and maintain Zhuang was a community responsibility, voicing an ecological perspective to language, ethnicity and identity. In contrast, when Ye saw Zhuang use as only “for special moment,” this minimizes rather than prioritizes Zhuang because the community of use for Putonghua is larger, where it can be used as a “common language.”

Extract 5

Chun: Although Zhuangyu is not the language I am most interested in, it is right to learn Zhuangyu. I think there’s a sense of responsibility … Because this work always needs a group of people to do, and it may be us … I think if I can really learn Zhuangyu very well, I think this responsibility can be placed on my shoulders. Because no one did it, but I do, I feel really proud of it.

Ye: I feel that I won’t use Zhuangyu for my future work. Maybe if he wants me to promote some Zhuangzu culture, maybe Zhuangyu will be used at this time. Anyway, Zhuangyu is not a very common language, its use must be used for special moment, in family or in community.

Though students tended to have a hierarchical attitude toward the three languages, with some disparaging English or Zhuang utility, overall interviewees felt positively about multilingualism and saw GZAR as a multilingual context. Linn, for example, felt that students who were not multilingual (including Zhuangyu and other local varieties) were disadvantaged (Extract 6).

Extract 6

Linn: I feel there is a language environment. Like Lulu in our pragmatic class, her parents don’t let her learn any Zhuangyu, just to let her speak pure Putonghua. I think her parents’ behaviour was wrong. You clearly have this environment, there is condition to learn Zhuangyu. I think you can still let your children be both good at Putonghua and speak good Zhuangyu. They are from Hechi. People over there could speak Guiliu dialect and Zhuangyu, and speak English, this multilingual environment is very good.

Self-reported proficiency

As above, students felt Putonghua proficiency was important for utility in the broad community and as part of their Chinese identity; they also self-reported high Putonghua proficiency. Self-reports on English proficiency were more mixed with most students seeing English as valuable instrument but not necessarily easy to learn and they did not view themselves as highly proficient. The greatest ambivalence was shown toward Zhuangyu proficiency with some interviewees feeling it was not useful (Extract 5), and others feeling it was an important part of the multilingual ecology (Extract 6). As “many children do not learn this language anymore” (Extract 7), there was concern that Zhuangyu “will be forgotten” with intergenerational shift to Putonghua—a community loss.

Extract 7

Yan: But the proficiency in Zhuangyu may not grow with age. Some will be forgotten. Because now many children, those in our hometown, could not speak Zhuangyu. I think in the future Zhuangyu may be forgotten, because many children do not learn this language anymore.

Ethnic identity and mother tongue

Zhuangzu tend to perceive themselves as Zhuangyu-speaking Han based on a long history of co-residence with Han and use of a standardised Chinese writing system (e.g., Grey, 2019; Kaup, 2000; Lu & Guo, 2021). Even Zhuang major students showed a laissez faire attitude towards the relationship between ethnic and linguistic identities (Extract 8, Chun, Die, Cyu, Jing, & Wen). As above, some did see learning Zhuangyu as a responsibility, but many saw it as lacking social functionality (Extract 5) and unnecessary for Zhuang ethnic identity (Extract 8, Lin). Others had mixed feelings. Wan, for example (Extract 8) stated that Zhuangyu was not “equivalent to the ethnic identity” and parents might feel use of Zhuangyu would negatively impact proficiency in Putonghua, but at the same time Wan claimed that lack of Zhuangyu weakened community identity.⁵

Extract 8

Chun, Die, Cyu, Lan, Jing, and Wen: We don’t think Zhuangzu people have to speak Zhuangyu, but it is better if he/she could speak.

Lin: I think everyone looks really similar today, very few people care if you are Zhuangzu people, or why couldn’t you speak Zhuangyu, in fact, a lot of Han people can’t speak Chinese … Many people think that Han people speak Putonghua. I don’t think you can say that if you are a Zhuangzu people you should speak Zhuangyu.

Wan: I don’t think speaking Zhuangyu is equivalent to the ethnic identity of Zhuangzu, because many Zhuangzu from our hometown can’t speak Zhuangyu. I think parents now want their children to speak fluent Putonghua. They think that Zhuangyu will switch their Putonghua into Jia Zhuang,⁵ so they won’t be allowed to learn Zhuangyu. This is not good, because it is a bit of weakening our Zhuang people’s identity, the weakening of national culture, because Zhuangyu also belongs to our Zhuang culture, I think so.

A final point worth noting arose only in the interviews: Zhuangzu in our study did not always recognise Standard Zhuang and instead considered regional variations of Zhuangyu as more authentic (Extract 9).

Extract 9

Fei: I don’t know Standard Zhuangyu, I don’t know how to define, because Zhuangyu is a dialect, and I think Zhuang dialects in each area are different, I do not know how to define the standard.

This is in contrast to the view that standardisation can support minority language vitality. Groves (2010), for example, examined perceptions of Cantonese as a “dialect” or a “language” among Hong Kongers, “Mainland Chinese” and “Mainland speaking Chinese” and found significant correlations in relation to perceived standardisation, autonomy and reduction (e.g. lack of a writing system, limited functionality).

Discussion

Subjectivities and language ecology

This study examined the subjectivities of those who have proceeded through the trilingual education system in the GZAR. In response to the first research question in which we ask the descriptive question “what are the subjectivities of these students toward the three languages,” the findings highlight the multifaceted nature of subjectivities. Overall, there was a clear hierarchy among the three languages which prioritised Putonghua based not only on utility but also on identity: Being Chinese meant being Putonghua speaking.

Based also on perceptions of utility and identity, subjectivities toward English were mixed. Questionnaire results showed that respondents thought English could enrich cultural knowledge and was worth learning but was not as important as Putonghua. More positive views tended to be more instrumentalist and correlated with higher family SES (Extract 2). The finding indicates a fissure between these tertiary students and recent language planning efforts to “avoid utilitarianism” (Gao, 2020, p. 13). Despite the positive views, self-reports on practices showed that most students did not make use of English outside of classes. Only English majors, for example, engaged much with English media. Interviewees presented a more nuanced view with some stating that English could offer a competitive advantage or be useful for international communication; others viewing English as just a school requirement and not relevant in their lives. There was little mention of the students’ subjectivities as learners of English (cf. Siegal, 1996). While English might be “cool” to some, it was not part of their identity; instead, English-speaking was linked to Western identities for countries such as the United States or Britain.

Zhuangyu was seen as the lowest in the hierarchy of three languages despite some moderately positive attitudes toward use in different social situations (Table 5) and some negative views of declining use among families (Extract 7). Comments in the interviews strengthened the overall finding that Zhuangyu was not a priority. Very few felt Zhuangyu was important for social functionality and even those few saw it as being limited in scope, for example, “not a very common language, its use must be for special moment, in family or in community” (Ye, Extract 5). Among some Zhuang majors (Extract 8), Zhuangyu was aligned Zhuang ethnic identity but more as a matter of responsibility than personal affiliation.

When considering this language ecology and the potential for language vitality, the findings present a pessimistic view of declining utility and interest in Zhuangyu despite national and regional policies to support the language. To the extent that participants do not see Zhuangyu as functionally useful or linked to personal/familiar/ethnic identity, there is unlikely to be strong support for maintenance. This agrees with prior research that sees Zhuangyu vitality as declining (e.g., Adamson & Feng, 2015; Feng & Sunuodula, 2009; Kaup, 2000).

In terms of the second research question, on how subjectivities are shaped by and shaping of language ecology, we note the interrelations of multiple factors (Table 7). These interrelationships were also evident in the interviews. This is in keeping with an ecological perspective which expects interdependence of factors (Fill & Mülhäusler, 2001; Mackey, 2001). In analysing across instruments, key themes emerged as “shaped and shaping” aspects of subjectivities within the language ecology which linked attitudes (toward the three languages) and sense of self (as social member and subject to social discourses). These included (a) pervasive instrumentalist views of language; (b) diverse perspectives on ethnic language, mother tongue and ethnic identity; and (c) a less than promising outlook for Zhuangzu vitality.

The questionnaire indicated that all three languages potentially have utility. However, questionnaires and interviews stressed instrumentalist views of languages which in turn supported a higher status for Putonghua and English as well as the notion of multilingualism. Some interviewees expressed a “resource view’” toward Zhuangyu, with support for minority languages (Extract 6), but this support was anchored more in discourses of community responsibility than utility or personal identity. The latter might be linked to the diverse perceptions of “mother tongue” evidenced in our data. As show in the questionnaire, participants were raised in sociolinguistic contexts including Zhuangyu, Cantonese, Guiliu and others. Their responses also indicated multiple interpretations of what mother tongue might mean to people in this region (Table 13). Interview respondents were relatively unaware of Standard Zhuang as a school language and potential lingua franca in the ethnic community (Extract 9), disaffirming standardisation as a path to enhanced vitality. Voices affirming a link between Zhuang identity and Zhuangyu tended to be few and faint (Extract 8).

Khubchandani (2003) recognised that the initial language learned at home (characterised as “native language”) could be distinct from the language of affiliation (characterised as “mother tongue”). Capturing these nuances across languages and contexts can be tricky. Li (2015) considered the meaning of “mother tongue” in the PRC in light of regional varieties and representation in Chinese script. He contended that “yu” [语] (the Chinese character used in our questionnaire) was the constituent morpheme of mother tongue, and claimed this was different to “yan” [言] which is closer to “regional dialect” or “mother dialect” (p. 24). Our questionnaire respondents did not seem to make this distinction as they provided a large variety of responses, some of which might be considered as regional dialects, “local indigenous dialect” [土话]. These findings show that for many, Zhuang ethnic identity does not require Zhuangyu proficiency. The findings also suggest that stakeholder understandings of “mother tongue” merits further investigation in relation to local subjectivities.

McNamara highlights that languages are “subject to a variety of discourses” (McNamara, 2019, p. 100) which involve positioning the self in relation to language varieties and language choices, and that these are closely linked to the subjectivities of learners/users in a specific environment. We argue that unpacking the interrelationships between subjects (language users) and languages requires examination of the language ecology, including intersections of family, education, community while recognising the “subjective experience of the learner as an individual with a history, both personal and cultural” (McNamara, 2019, p. 116). This has implications for language vitality as “the ecology of language shift is the study of interrelated sequences of causes and effects producing changes in the traditional language behaviour of one group under the influence of another” (Mackey, 2001, p. 68). The results show that these shifts are not based only on influences of one group to another. We suggest the concept of subjectivities helps us to evaluate the statuses, uses and affiliations of languages as expressed by community members. By examining subjectivities, the analysis maximises the opportunity to capture the dynamics in this multilingual ecology—the living, hierarchical, and complementary nature of multiple codes in a given social structure.

Conclusion

Our study investigated tertiary students’ subjectivities in relation to multilingualism in the language ecology of the GZAR. The analyses lead to better understandings of multilingualism in this region where influences of family, community language use, media use, and education are evident. Overall, the students view multilingualism—including official minority languages and regional dialects—as positive. The students clearly show hierarchical attitudes toward the three languages, prioritising Putonghua, with strong views linking Putonghua with Chineseness and weaker views linking ethnic identity to Zhuangyu.

Our study has two limitations. First, language ecologies are complex and dynamic with many interrelated features. Our analysis focused on tertiary students as a population subject to trilingual policies and practices. However, other languages could be considered in a more detailed analysis (e.g., the other foreign languages taught at the universities and the other Chinese languages used in the region). In addition, including other community members and social groups from this population of 19.57 million Zhuangzu might complete or better capture the dynamics of the language ecology.

Second, these data and the work of some others (e.g., Groves, 2010; Lu & Guo, 2021) suggest that perceptions of “mother tongue” do not align with ethnic identity as in local policy statements. In addition, ethnic identity as understood and experienced in this context may be different to language-ethnicity affiliations elsewhere, for example, South Asia (Khubchandani, 2003) and Europe (Huguet & Lasagabaster, 2007). Subjectivities of “mother tongue” and what this means in terms of ethnic identity in the PRC deserves more exploration.

In conclusion, we suggest that a language ecology approach helps us better understand why implementation of national policies or community supports on their own are insufficient for language maintenance efforts. Instead, we must consider multifaceted user subjectivities. We see that a language ecology approach allows ambivalence to surface, highlights attitude/use (dis)connections and reveals where subjectivities are more symbolic than actual. This is neatly captured in one interview in which Wan reported being actively discouraged from Zhuangyu learning in favor of Putonghua at home, but continuing to use Zhuangyu with her grandmother. She stated that students in high school could not speak Zhuangyu and this was unimportant, but went on to say that some parents didn’t allow their children to learn Zhuangyu and “this is not good, because it is a bit of weakening our Zhuang people’s identity, the weakening of national culture, because Zhuang dialect also belongs to our Zhuang culture, I think so” (Extract 8). Understanding these positive/negative/mixed attitudes and practices is essential for understanding identities and informing language vitality efforts.

We also suggest that subjectivities are shaped by and shaping of language ecologies. As McNamara explains, subjectivity is “performatively constituted” (p. 89) through competing discourses, especially in multilingual language ecologies. He reminds us that language learning, including choice of which language to learn and use, is not purely functional but is linked to the “subjective experience of the learner as an individual with a history, both personal and cultural” (McNamara, 2019, p. 116).

Acknowledgements

We would like to express our sincere gratitude to the study participants whose invaluable contributions made the research possible. We would also like to thank the second coder, Mr. Hui (Hugh) Zhang, for his committed support to this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Notes on contributors

Ying Wu

Ying Wu is a lecturer from Shenzhen University with a PhD from National Institute of Education, Nanyang Technological University, Singapore. Her research interests are in the dynamicity and multidimensionality of language ecology, including understanding ethnic minority language and education in China.

Rita Elaine Silver

Rita Elaine Silver is an Associate Professor at National Institute of Education, Nanyang Technological University, Singapore. Her research interests are in classroom discourse, literacy, multilingualism, and in research methodology and ethics.

Notes

1. We thank an anonymous reviewer for raising this point and encouraging us to make it more explicit.

2. Universities with specialties in ethnic affairs, authorized by State Ethnic Affairs Commission.

3. The name of the university is confidential.

4. The interviewer did not function as a facilitator, as in a focus group, but as a traditional interviewer. (See Gubrium et al. [2012, p. 32] for discussion.).

5. Jia Zhuang is a variety recognised by the local Guangxi people, indicating a phonological shift from Zhuangyu to Putonghua. See more of this issue in Grey (2021) and Wu et al. (2022).

Appendix A.

Coding of questionnaire item responses

Table

	1	2	3	4	5	6
1. Gender	Male	Female
2. AY intake	2015	2016	2017	2018
3. School	GXU	GXUN	GXNU	NNU	GXUFL	Others
4. Academic major	English	Zhuang	Chinese	Science & Engineering	Other
5. Whether a Zhuangzu (Ethnicity)	Yes	No
6. Mother tongue	Zhuang	Putonghua	Guiliu	Other
7–8. Parent educational level	High school and below	Junior college	Bachelor’s degree	Master’s degree	PhD
9–11. Language proficiency	Very unfluent	Unfluent	Fluent	Very fluent
12–14. Initial age of learning a language	0–2	3–5	6–10	11–13	14+
15. Hometown population	More than 100,000 citizens	100, 000 or less
16. Whether been to a native English- speaking country	Yes	No
17. How often watch English TV	Never	Very few	1–2 times per week	3–5 time per week	Everyday
18. Sociolinguistic context	Zhuang-speaking community	Cantonese-speaking community	Guiliu-speaking community	Others
19. Private or public university	Private	Public
20. Type of high schools	Comprehensive natural sciences	Comprehensive social sciences	Arts	Physical education
21–23. Language attitudes	Strongly disagree	Disagree	Neither agree nor disagree	Agree	Strongly agree