Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Global Chinese

Ed. by Wei, Li / Yanbin, Diao

2 Issues per year

See all formats and pricing
More options …

Visualizing language ideologies and verbalizing perceived linguistic boundaries: The case of Mandarin Chinese in contemporary Taiwan

Spencer C. Chen
Published Online: 2018-03-28 | DOI: https://doi.org/10.1515/glochi-2018-0002


As Mandarin gains prominence in the globalizing world, the use of Mandarin not only blurs geopolitical borders, but also the linguistic boundaries drawn between Taiwan and Mainland China. Employing the theoretical framework of language ideologies and the methodological approach of perceptual dialectology, this article investigates how the Taiwanese perform their identities by hand-drawing their perceived linguistic boundaries of Mandarin varieties. The data came from a multi-sited linguistic fieldwork project in Taiwan that engaged with forty-two Taiwanese (age 19–56) on tasks that include drawing dialectological maps of Mandarin and extended ethnographic interviews. The data show that rhoticity has been selectively rationalized and ideologized as boundary of linguistic differentiation between Mandarin varieties across the Taiwan Strait. This article further examines the ways in which the informants, by surpassing the geospatial indexicality of Taiwan Mandarin as a vernacular, rationalize and project the existing stylistic use of Taiwan Mandarin onto their sociopolitical construals, creating a space where Taiwan Mandarin is equivalent to Putonghua in symbolic power. This study demonstrates the ways in which perceived dialectological maps functions as visualized indexical fields that mirror the multiplicity and volatility of participants’ concurrent language ideologies about the complicated relationship between language and the micro- as well as macro-sociopolitical environments they inhabit.


随着汉语国际地位的提高,汉语的使用不仅日渐打破传统政治地理疆域,也逐渐淡化了海峡两岸之间的语言界线。本文基于语言意识(language ideologies)的理论架构,结合感知方言学(perceptual dialectology)的方法论,深入探究台湾人如何透过视像化感知方言界线来建构自己作为台湾国语话语者的身份认同。本文数据采自多点民族志,研究对象为 42 位观光旅行业从业人员。此外,分析亦加入深度访谈与受访者自行绘制的方言地图作为研究的辅助材料。研究结果显示,受访者普遍认为舌尖后音与儿化音是区别海峡两岸汉语语言变体(variety)的指标。与此同时,本文进一步探讨了这些旅行业受访者如何将台湾国语本有的地方区域特色转变为符号资本(symbolic capital),从而重新创构台湾国语的指示性(indexicality),促使台湾国语拥有同标准普通话一样的象征优势,进而将其转化为商用价值。本文旨在透过感知方言地图,突显语言意识是一特殊且具有指示性的意识形态,探究话语者如何透过语言投射出自身文化与社会政治认同。

Keywords: language ideologies; indexicality; dialectal maps; identity; Taiwan Mandarin

关键字: 语言意识; 指示性; 方言地图; 身份认同; 台湾国语

1 Introduction

The political division between Taiwan (officially, Republic of China) and Mainland China (People’s Republic of China, henceforward the PRC) since 1949 has protracted an ideological linguacultural contestation that led to the separate development of Mandarin in the former regime as Taiwan Guoyu, a.k.a. Taiwan Mandarin (Cheng 1985; Kubler 1981, 1985; Li 1985; Wei 2006; Wu 2009), and in the latter Putonghua, a.k.a. Modern Standard Mandarin (Li 2006). As the PRC gradually gains political economic prominence, Putonghua is steadily establishing itself as another transnational linguistic capital in the globalizing world. Across the Taiwan Strait, as Taiwan continues to depend on the Mainland Chinese market and yet strives to maintain its autonomy, the use of Mandarin hence blurs the geopolitical borders and sociolinguistic boundaries between Taiwan and the PRC. Thanks to multiple political economic and social transformations that occurred on both sides, the last three decades witnessed a surge of people’s interest in their counterparts’ Mandarin use as a result of increased non-governmental interactions (e.g. Zhang 2005). Taiwan’s opening of the Taiwan-PRC border to Mainland Chinese tourists in 2008 created new economic possibilities for its tourism industry. In the meanwhile, however, as the conflict between the two states’ respective political agenda continues, the notion of being “Mandarin-speaking people” is often loaded with nationalist ideologies, challenging not only people’s understanding of linguacultural boundaries (Friedrich 1989) but also the very geopolitical borders between Taiwan and the PRC. It is against this background that this article investigates the ways in which the Taiwanese differentiate Taiwan Mandarin from Modern Standard Mandarin, i.e. Putonghua to construct ideological boundaries that mark their distinctive identities as Taiwanese from their counterparts in the PRC. As such, this article, with its focus on the transnational relationship between Taiwan and the PRC in tourism section, offers a comparative perspective on the processes of the indexicalization (Silverstein 2003) of Taiwan Mandarin.

In his article “Indexical order and the dialectics of sociolinguistic life” (2003), linguistic anthropologist Michael Silverstein attempts to build on his theory of indexicality to show how any indexical sign (N-th order indexicality) – “a feature whose use can be correlated with a sociodemographic identity […] or a semantic function” (205) – could undergo a semiotic process which further enregisters the original sign with “an ethno-metapragmatically driven native interpretation” (212), thus creating an additional indexical relationship (N+1th order indexicality) between the original sign and the signed phenomenon. Following Silverstein’s theorization of indexical order, this article continues this investigation of the reproduction of indexical meaning through existing sociolinguistic elements with regard to the micro-social and macro-social implications of the signs, by drawing examples from multi-sited linguistic fieldwork in three tourist sites in Taipei, Taiwan. I outline the shift of Mandarin ideologies – from viewing Taiwan Mandarin as a regional vernacular to valorizing it as a distinct territorialized variety – to highlight how everyday interactions with the PRC tourists reflect the informants’ Taiwanese identities. I examine the ways in which the informants, by surpassing the geospatial indexicality of Taiwan Mandarin as a regional vernacular, rationalize and project the existing stylistic use of Taiwan Mandarin onto their sociopolitical construals, creating a space where Taiwan Mandarin is equivalent to Putonghua in symbolic power (Bourdieu 1991). I detail the ways in which Taiwanese informants, by manually drawing linguistic boundaries of Mandarin varieties on maps, visualize the perceived linguacultural borders and thus reinforce a sense of belonging by the very lines they created on their maps. I offer examples where the distinction between non-retroflex [ts, tsh, s] and retroflex [ʈʂ, ʈʂh, ʂ, ʐ] is mobilized to indicate the linguistic boundaries between Mandarin varieties across the Taiwan Strait. Subsequently, I discuss how articulatory differences between retroflex sibilants are further used to perceptually territorialize the Taiwanese and the Chinese separately. I argue that the reconfiguration of a standard-vernacular hierarchy between different Mandarin varieties involves semiotic processes (Irvine and Gal 2000; Silverstein 2003) that revalorize linguistic features of Taiwan Mandarin through foregrounding nationalist pride to compete with the global hegemony of Standard Putonghua. I supplement this analysis with transnational examples of possible alternatives for future studies of indexical order in understanding the effects of macro-social construals on speakers’ concurrent language ideologies about the complicated relationship between language and the micro-social interactions.

2 A brief sociolinguistic review of Taiwan Mandarin

Every society has its own unique and complex geopolitical history that contributes to its contemporary belief system about itself and its language(s). Due to its geographical location on the Asia-Pacific trade route, since the seventeenth century Taiwan has received multiple waves of immigrants, each of whom imprinted the languages, sociocultural practices, and values of immigrants onto the existing social system. Their lingua-cultural ideologies in particular have contributed greatly to Taiwan’s own linguistic practices in relation to its multiethnic complexity (e.g. Brown 2004; Chen 2010; Sandel 2003; Wei 2005; Wu 2009). The political separation between the regimes in Taiwan and Mainland China throughout the post-World War II years has generated greater socio-political and lingua-cultural differences than the coarse distinction between totalitarian China (PRC) and democratic Taiwan captures. The regimes’ concepts of Chinese culture, social values, language, national identity, to name just very few, are now far from similar.

Mandarin Chinese, Guoyu (literally “national language”), was formally introduced in the aftermath of WWII to the Taiwanese. At the time, the majority of whom were monolinguals in Taiwanese with some practical knowledge in Japanese due to Japan’s prior imperial presence in the region. Subsequently, starting from 1946, Guoyu was enforced and promoted through various institutions as the only official language. Decades since its introduction to Taiwan in 1946, Guoyu has undergone sufficient linguistic changes that in the 1980s scholars started viewing it as a separate variety, Taiwan Mandarin, but not as a regional vernacular of Mandarin (see Cheng 1985). Cornelius Kubler (1981, 1985) and other linguists have compared and contrasted phonological, morphosyntactic, and lexical features of Modern Standard Mandarin (Putonghua) and Taiwan Mandarin (Cheng 1985; Hsu 2014, 1994; Jernudd 1985; Li 1985, 2006; Liao 2008; Sandel 2003). Citing such linguistic divergence as evidence, they call for a formal recognition of Taiwan Mandarin as a distinct Mandarin variety that has absorbed many features from Taiwanese local languages, particularly Taiwanese Hokkien. However, despite the linguistic distinctiveness of Taiwan Mandarin as documented by linguists, the ideology that privileges Taiwan Mandarin as a regional vernacular of Mandarin Chinese continues to exist. Coupled with the changing Cross-Strait relations between the two regimes, this mutual intelligibility of the languages further complicates the physical and psychological borderlines drawn between Taiwan and the PRC. At times, the Chinese nationalists utilize the “shared language” ideology as the lingua-cultural basis for the discourse of Chinese unification (Brown 2004). At other times, as I discuss later, the Taiwanese valorize Taiwan Mandarin as their apparatus for producing identity and nationalism at the micro-level. This is done in order to semiotically distinguish themselves from the mainland Chinese in response to the recent increased Taiwan-PRC cross-border interactions.

3 Language ideologies, identity, and Taiwan Mandarin

In its simplest, yet problematic definition (Rampton 2009), the term “speech community” refers to any group with a shared code of communication. Various attempts have been made to theorize speech communities since its inception. Linguists have devoted their attention to identifying a shared, and often homogenous, linguistic structure as the determining characteristic of group membership. Under this paradigm that prioritizes the denotational regularity of language, the complex relationship between a group of people and a specific linguistic code has been greatly reduced to a simplistic view. With the penetration of this assumption to the level of nationalist discourses, the ability to use this single linguistic code legitimates the speaker’s membership to a larger cultural collective (Anderson 1991). What it generates is the “one-language-one-culture assumption”, as Irvine elegantly puts it (1996: 123).

During the early years of the political separation between Taiwan and Mainland China, Taiwan had been institutionally positioned as a province of the Republic of China under the Chinese Nationalist Kuomintang (KMT) – the authoritarian regime that fled from Mainland China to Taiwan in 1949 seeking to establish itself as the sole rightful government of Greater China (Brown 2004) against the Chinese Communist Party (CCP). The Guoyu Policy (Guóyŭ zhèngcè, literally “National Language Policy”, implemented from 1946 to 1987) was hence imposed as part of the assimilation project whose goal was to Sinicize the Taiwanese population. Against this macro-social background, becoming speakers of Mandarin – practically a foreign language to then Taiwanese people – was highly politicized as the way in which Taiwanese people could truly obtain their membership in the Chinese nation-state (cf. Sugano 2012).

This idealized and neutral correlation between a group of people and a specific linguistic code has often led to the theoretical confusion between speech communities and language communities (Silverstein 1997: 127). Silverstein avers that “community must be…a degree term: membership of people in such communities radiantly depends on social-factional regularity” (1997: 128). According to him, speech communities should be carefully differentiated from language communities, the latter of which focuses primarily on denotational codes. Speech communities, instead, should be defined “based on patterns of indexical facts of linguistic usage-in-context…of who, normatively, communicates in which ways to whom on what occasions” (Silverstein 1997: 129). The perceptions of dis/similarities between linguistic practices and structures, theoretically known as language ideologies, support politico-economic and linguacultural projects of group distinction and language change (Irvine and Gal 2000; Kroskrity 2010; Silverstein 1979). Studies of language ideologies specifically deal with how ideologies mediate between linguistic practice and the sociocultural political structure speakers inhabit (Woolard and Schieffein 1994). The projects of looking at language ideologies have viewed communities as “political construct[s]” (Rampton 2009: 8) at the individual, communal, societal, and even to the nation-state levels regarding people’s actual language practice and beliefs, which in turn influence the development of these ideologies of language. On the one hand, speakers’ beliefs about linguistic structure and language use influence, project, and/or transform their language practice (Kroskrity 2000; Silverstein 1985). Conversely, individuals’ very use of language reflects, and potentially determines, their perception of the social reality they inhabit (Blommaert and Verschueren 1998; Silverstein 1985, 1996). Irvine and Gal (2000) further identify the three semiotic processes – namely, iconization, fractal recursivity, and erasure – that are central to the production of language ideologies. Iconization, as they understand it, includes participants recognizing, i.e. iconizing, certain indexical relationships between language practice and their immanent social reality. Fractal recursivity, furthermore, reuses these iconic associations and projects them to different levels of indexical meaning. In close conjunction with iconization and fractal recursivity, erasure involves a selective elimination of certain indexical associations in favor of the social actor’s own interests. In short, these three processes Irvine and Gal theorize provide intellectual means to describe and examine the ways in which social actors employ language ideologies to their identity construction.

Language ideological studies of issues of language and identity in Taiwan explain scenarios of language differentiation as the outcome of both language policies and the socio-political and economic dichotomies between Mandarin and non-Mandarin-speaking populations (e.g. Sandel 2003; Wei 2006; Liao 2008; Wu 2009; Chen 2010). This handling of these phenomena illustrates how the institutional enforcement of language assimilation by the KMT has fundamentally altered Taiwanese people’s attitude towards languages. For instance, referencing Bourdieu’s concepts of habitus (1977, 1991) and symbolic capital (1991: 50–57), Sandel (2003) discusses how the KMT’s enforcement of the National Language Policy has created a habitus where Mandarin dominates the practice of communication and where non-Mandarin speakers are forced to accept the pragmatic values of Mandarin as capital in the market, thus devaluing their native languages. As Taiwanese people gradually attained better socioeconomic status and political strength in the late 1980s, along with increased democratization, the discourses of language and national identity have undergone major transformations. Liao (2008) presents the ways in which iconic values of two subregional Taiwan Mandarin vernaculars shifted from indexing one’s ethnicity and place of origin to indicating the socioeconomic status, political affiliation, and to some extent, personality. Her perceptual dialectological study on Taiwanese college students maps out a folk dialect boundary (Preston 1999) that separates the perceived standard version of Mandarin from other perceptually accented-Mandarin varieties. Liao’s results point out that, while standard Mandarin is indexical of prestige, higher socio-political achievements, and northerner identity, accented-Mandarin varieties evoke sociolinguistic associations with the negative opposite values.

Ideologies of language are not static snapshots of societal conceptions about language; instead, they are both reflections and instruments of the concurrent socio-political reality. Since the late 1990s, the discourse that foregrounds Taiwan as part of the language community of Mandarin has been constantly challenged and rebutted. Despite the KMT government and the PRC’s assertion of Taiwan-China unification based on the mutual intelligibility of language and shared cultural practices, the Taiwanese do not unanimously agree with such assertions of language isomorphism. Focusing on intra-national relationships, findings from the previously mentioned sociolinguistic studies show their limits in capturing the conceptions of language, identity, and nationalism of Taiwanese society with respect to the current extra-national situation – namely, the rise of the PRC as a global powerhouse. As Meek (2009) highlights, the study of language ideologies is difficult in that “[the] sentiments of community members are never coherently solidified […] [community members] expressed a range of concerns not easily compartmentalized into a singular language ideology” (170). Such fluidity, or “indexical mutability” as Penelope Eckert calls it (2012: 93–97), is a constant process of (re-)defining indexical signs through transformable linguistic practices, for example, enregisterment (see also Agha 2007; Silverstein 2003).

Thus in the sections that follow, I engage with Taiwanese people that have intensive interactions with the mainland Chinese and seek to understand the change in language ideologies about Taiwan Mandarin and other Mandarin varieties regarding this volatile Taiwan-PRC relationship. Specifically, I am concerned with the following research questions:

  1. How Taiwanese people redefine their Taiwan Mandarin as opposed to the perceived Mainland Mandarin 1?

  2. How Taiwanese people use understanding in conceptualizing an imagined boundary (Avineri and Kroskrity 2014) set against the PRC?

  3. How do they utilize linguistic resources as the semiotic basis for constructing that boundary and their speaker identities?

4 Methodology and data

This study implemented the mixed methods approach that combines participant-observations with ethnographic interviews (Briggs 1986), and dialectal mental maps (Preston 1999; Preston 2002). Through this ethnographic approach, this article critically investigates the implications of observed sociolinguistic phenomena and informants’ self-reported ideologies and attitudes about Mandarin varieties and identities.

The primary set of data comes from several short-term fieldwork projects conducted during 2014 and 2015 in Taipei, Taiwan. I followed closely the sales persons working at the souvenir stores in three tourist sites, namely, Beitou Hot Spring, National Palace Museum, and Taipei 101. I recruited 42 informants ranging in age from 19 to 66, of which 15 were male. All informants are native to Taipei City and currently live in the Taipei Metropolitan Area. All but four of them (40 individuals) considered themselves bilingual in Taiwan Mandarin and Taiwanese; the remaining four informants considered themselves as bilinguals in Taiwan Mandarin and Hakka (2 individuals) or Mandarin-speaking monolinguals (2 individuals). All informants use Taiwan Mandarin as their home language, and over 50 percent of them (23 individuals) reported that they code switch between Taiwan Mandarin and Taiwanese at home.

The semi-structured interviews with these informants were carried out and audio-recorded in informal settings near the named sites. The interviews centered on their perception of language differences/similarities between Taiwanese people and mainland Chinese, their experience with mainland Chinese tourists, and their self-reflection on their Taiwanese identity. 2 Audio-recorded interviews, amounting to a total of approximately 44 hours, were later transcribed and analyzed in the original language. Excerpts presented in this article are translated in English along with the original language transliterated in Pinyin. In cases where the analytic focus is on the content of the narrative, I present the relevant excerpts solely in translated English.

In addition, in order to pin down how they, not just discursively, but also visually, comprehend dialectal boundaries of Mandarin varieties, the informants were requested to draw on paper their version of Mandarin maps. Specifically, in each interview session, they were first provided with blank papers to draw, afterwards papers with preprinted geopolitical maps that show territorial or prefectural borders. As they drew, they were asked to discuss, and to enact their opinions if they chose, the variational differences of Mandarin varieties that they named. Together with field notes, these dialectal mental maps serve as the complementary data to the interviews.

5 N-th order of indexicality: Taiwan Mandarin as a vernacular of Mandarin

Under the Guoyu Policy (Guóyǔ Zhèngcè), educational institutions required non-Mandarin-speaking population to acquire Guoyu and denounced the inability to learn the language properly as a moral failure (Sandel 2003; Sugano 2012). Much emphasis and effort was placed on “correcting” or “improving” Taiwanese people’s phonology of Mandarin. It thus came as no surprise to me, both as a researcher and speaker of the language, the phonological peculiarities were unanimously stressed throughout the interviews of all the perceived differences between the two named Mandarin varieties. The informants would start the discussions of phonological differences focusing on the accurateness of pronunciation. For instance, even though Wei, 3 a salesperson in his late-twenties, touched upon intonation, he linked the accentual distinctiveness of Taiwan Mandarin solely to pronunciation.

Excerpt (1):

“Nonstandard” Pronunciation 4

01WeiDuì ā. Ránhòu, fāyīn yě bù xiàng Běijīng nàme biāozhǔn
Yeah. And then, [our] pronunciation is not as standard as Beijing [Mandarin]
02Jiùshì wǒmen qiāngdiào de wèntí, yǔdiào de wèntí.
This is a matter of our accent. A matter of our intonation.
0… ((3 lines omitted))
06SCCÈn-èn. Zěnme shuō ne?
Um-hm, how would you say [about this differentiation]?
07WeiSuǒyǐ wǒ jiù shuō, yǎozì méiyǒu nàme zhǔnquè
That’s what I said, pronouncing less accurately
08jiùshì Táiwān-rén jiǎng de Guóyǔ
is exactly Taiwanese’ Guoyu.
09Xiàng zh- ch- sh- bùshì zhèyàng fāyīn 5.
So such sounds are not pronounced [precisely] zh-, ch-, sh- [i.e. retroflexes].
10Shì yǒu qīnqiēgǎn.
[Such pronunciation] conveys the feeling of being at home [i.e. familiar].

As we see here, both fāyīn ‘to emit sounds’ (lines 1 and 9) and yǎozì ‘to bite words’ (line 7) can be understood in English as “to pronounce.” The word yǎozì stems from a practice in Peking Opera that strictly requires the performers to articulate the lines cleanly to ensure the beauty of the songs. The simultaneous use of the terms fāyīn and yǎozì reflects an ideology that there appears to exist the “standard” to follow in pronouncing words.

This was especially the case when informants brought up the issues of the lack of retroflexes in Taiwan Mandarin as an evidence of the “deviations” from the “standard [form]”. The informants, like Wei, understand the Taiwanese pronunciation as “less standard” and “less accurate” in comparison to the Beijing Mandarin (i.e. Putonghua), since Taiwanese do not produce the retroflex consonants zh-, ch-, sh-, and r- exactly as [ʈʂ], [ʈʂʰ], [ʂ], and [ʐ] respectively. This is because in the textbooks and primary school Guoyu classes, rhotacization has been overtly emphasized (e.g. Kubler 1981, 1985; Cheng 1985; Li 1985; Sandel 2003; Chung 2006b), and retroflex initials have been categorically and indexically referred to as the “standard” form. Nonetheless, in the actual speech, it has been documented that since as late as the late 1970s Taiwan Mandarin has witnessed “the gradual loss of … retroflexivization which in turn characterizes the language” (Li 1985: 123; see also Cheng 1985; Kubler 1981, 1985). The informants were well aware of the discrepancy between the actual production and the prescribed articulation of these sounds. Consider the following group discussion I had with two salesclerks, Yuu and Haul.

Excerpt (2):

Pronouncing “Hěn-Rè” and “Hěn-Lè” ‘very hot’

01SCCNà nǐmen zìjǐ jiǎnghuà huì fèn zh-, ch-, sh-? Huì ma?
Then when you guys talk, do you distinguish zh-, ch-, and sh- sounds? Do you?
02YuuTā jiù jiū-zhèng guò wǒ!
He did correct me!
04YuuTā hěn xǐhuān jiūzhèng bié-rén.
[tsɤn⇃] [lən↿]
He likes to correct people[’s pronunciaton]
06Yuur- gēn l- bù fèn.
[I don’t] differentiate the r- sound from the l- sound
07HaulTā jiù shuō “Hěn lè” ér bùshì “hěn rè”
Yeah he would say “HĚN-” [‘very hot’] but not “HĚN-” [‘very hot’]
          [ɾə⇃]         [lə˥˩]
08SCCNi zìjǐ gānggāng bù jiù jiǎng hěn lè?
But you [i.e. Haul] yourself just said , didn’t you?
09HaulHā hā, méi yǒu ba!?
Haha, no way, did I?
10SCCXiàng xiànzài xiǎo péngyǒu shì-bù-shì bǐjiào bùhuì fèn le?
So is it true that nowadays children don’t distinguish the sound differences?
11YuuBù tài huì zhèyàng
Not really so [i.e. distinguishing the sounds]

As this interview took place in the afternoon of a hot summer’s day, the complaints about the heat and humidity constantly entered our conversation. The word ‘hot’ came naturally into the discussion about the Taiwanese pronunciation. While in Guoyu the designated pronunciation of is [ʐɤ⇃], with a retroflex sibilant [ʐ], mid-back unrounded vowel [ɤ], and a high-falling tone [⇃], what Yuu uttered instead was much closer to non-retroflex (lateral), mid-central unrounded mid-falling [lə⇃]. Both informants here were aware of the fact that Yuu failed to produce the retroflex sibilant [ʐ] (which is also evidenced, though unnoticed by them, in Yuu’s other retroflex sounds, e.g. rén in bié-rén ‘other people’ in line 4, which he produced [lən↿] instead of the prescribed version [ʐən↿]). Haul metapragmatically commented on Yuu’s “mistake” in uttering [lə⇃] instead of the “correct” [ʐɤ⇃]. What passed his attention, however, is that both of his versions of (in line 7), a flap [ɾ] in the former and a lateral [l] in the latter, were in fact the Taiwanese-accented variants of the “correct” [ʐɤ⇃]. When attention was directed to the “non-standard” pronunciation of this type, informants would attempt an immediate self-repair of the questioned linguistic token. They often used “jiū zhèng” (cf. line 4) or “jiǎo zhèng”, literally meaning ‘to correct, to amend, to set straight’, to refer to the action they take or being forced upon them in dealing with these accented sounds.

Underlying such discourses is a standard language ideology (Lippi-Green 2011; Milroy 2001; Silverstein 1996) at work that sees superiority in the standard form and disfavors non-standard variations. Guoyu, or Mandarin Chinese in general, including Putonghua spoken in mainland China, uses Beijing-based Mandarin vernacular as the phonological basis (Chao 1968; Li 2006). Amongst all Mandarin sounds that do not exist in Taiwan Mandarin (e.g. the postvocalic Erhua [ɻ], diphthongs [uo] and [ou], etc.; see Kubler 1981, 1985; for further discussion), retroflex consonants have received attention the most. As such, gradually, the articulation of retroflex initials has become the iconic index (Irvine and Gal 2000) of prestige, educational and socio-economic achievements in (semi-)formal settings (Chung 2006b). The inability to accurately utter retroflex sounds indexes lack of socio-economic achievements and evokes associations with working-class locations where Taiwanese is widely spoken (Liao 2008; Sandel 2003). This iconic indexicality creates an indexical layer where the distinction between the standard and vernacular is assigned socio-spatial values. Speaking the “non-standard” Taiwan Mandarin thus subjects Taiwanese-speaking populations, as users of a vernacular Mandarin, to the broader Mandarin-speaking population.

Recall that in Taiwan, Mandarin Chinese was an enforced effort under the Guoyu Policy (Guóyǔ Zhèngcè), and that the teaching materials emphasized on instructing Beijing dialect-based Mandarin phonology. Thus, in its earlier stages, Mandarin learning has been associated with the linguistic coercion imposed upon Taiwanese people. Thus, it is not surprising to hear informants associating Guoyu with the manifestation of linguistic coercion and Taiwan Mandarin with a lack of official status. The informants were well aware of the disjuncture between their actual speech and the “standard” form. Figure 1 shows the parallel of the sociolinguistic context of Taiwan and the PRC that the informants mapped out.

The hierarchical order of standard Mandarin in Taiwan in comparison to the PRC (China).
Figure 1:

The hierarchical order of standard Mandarin in Taiwan in comparison to the PRC (China).

The blue- and red-colored squares represent the speech communities of Taiwan and the PRC, respectively. The larger grey square diagram illustrates the ideology that essentially submits Taiwan and the PRC to a larger, Beijing-centered, Mandarin language community. In either context, under the standard language ideology, the Beijing dialect-based standard variety is regarded as prestigious and hierarchically superior in comparison to other non-standard forms.

With that said, many of my informants’ responses seem to suggest a similar and yet different indexical relationship. This can be more clearly exemplified using the dialectal maps drawn by the informants themselves. Yuan (see Figure 2 below for her map) made use of lexical variations, intonation, and 捲舌音 juǎnshé yīn ‘tongue-curling sounds’ (retroflexes), to metapragmatically label the perceived dialectal boundaries between the Mandarin-speaking areas in Greater China. From her labeling of the Taiwanese use of Mandarin as 台湾腔 6 Taiwan qiang ‘Taiwan accent’ (as indicated in the circled word on the right) we can see the standard ideology at work here. The use of 腔 qiang ‘accent’ invokes a socio-spatial indexical order (Silverstein 2013) that strongly indicates her view of Taiwan Mandarin as a vernacular of the standard Mandarin. Rather than treating it as a variety with its own specific linguistic features, Taiwan Mandarin is reduced from its transnational social meaning and practically conforms to the intranational social significance, by whose standard, Taiwan Mandarin is a linguistic deviation of Putonghua.

Yuan’s map of Mandarin language community.
Figure 2:

Yuan’s map of Mandarin language community.

Furthermore, what is also noticeable is the fact that the accents many informants described are abstract and unspecific, and arguably imagined. As shown in Figure 2 here, there is only the dichotomy of 台湾腔 Taiwan qiang ‘Taiwan accent’ vis-à-vis 大陸腔 Dalu qiang ‘Mainland accent’. Two important features of these “accents” are the projected homogeneity and attitudinal evaluations. Whereas of Taiwan qiang is considered more 平 píng ‘flat’ and 溫和 wēnhé ‘gentle’, Dalu qiang is characterized by its 高低起伏 gāodī qǐfú ‘punctuative intonation’ and 衝 chōng ‘bellicosity’. This seems to suggest a changing indexicalization of Taiwan Mandarin. Rather than complying with the N-th order of indexicality (Silverstein 2003) that socio-spatially degrades Taiwan Mandarin as a vernacular, the informants validated their own natural speech to reconceptualize the indexicality prescribed by r-coloring sounds. As Karen Chung also points out, “textbook [Guoyu] exists mainly as an idealized language that is studied and exists in one’s consciousness, but is seldom consistently practiced” (2006b: 198). This tells us that in actual practice, Taiwan Mandarin occupies a very different but more practical speech domain than textbook Guoyu does.

I suggest that this selectivity of awareness (Kroskrity 2000, 2010; Silverstein 1996) of pronunciation brings up a different but more crucial aspect of Taiwanese people’s conceptualization of Taiwan Mandarin. Such awareness does not, however, stem from the standard language ideology that projects negative value judgments on “the non-standard” Taiwan Mandarin. Instead, informants expressed feelings of being home, closeness, familiarity, and Táiwān de wèidào ‘Taiwan’s flavor’ towards Taiwan Mandarin. In the next section, I offer my own interpretation of this revalorization and the reconfigured ideology towards “the non-standard” deviation of Guoyu now known as Taiwan Mandarin.

6 N + 1th order of indexicality: national identities and territorial boundaries

Placing Taiwan Mandarin at the opposite end in comparison to Standard Guoyu is not to characterize it merely as a regional vernacular of Mandarin Chinese. Rather, the informants’ selective attention and disattention to specific linguistic features substantially reveal their ideologies towards both Mandarin varieties in question. Such selectivity of awareness is an attempt to restore the status of Taiwan Mandarin and more crucially to disassociate this variety from Beijing dialect-based Mandarin, which is the basis of Guoyu and Putonghua. The informants focused primarily on phonological and lexical aspects in their metalinguistic commentary when comparing the two Mandarin varieties. In most cases, when asked to demonstrate such differences, the informants would intentionally showcase articulatory differences between retroflex initials and postvocalic Erhua [ɻ] using mutually intelligible lexical items. They often followed up such demonstrations with half-joking self-ridicules. They narrated awkward moments when they would “automatically”, but “unnaturally”, “curl their tongue [for retroflex sounds]” when talking to mainland Chinese tourists. Cher, a counter clerk in her fifties, revealed this very “weird” moment here:

“[I] don’t know why but whenever I talk to [mainland Chinese tourists], automatically my tongue gets tied. And like [so does] my husband too. [He said,] “Why do you curl you tongue whenever you talk to them?” [And I said,] “You do the same thing! [You change your] Taiwan Mandarin, whenever you see them you keep curling your tongue!” [Then he said,] “Oh did I really do that?” That’s it, I found that many people would unconsciously [curl their tongue].”

This metapragmatic awareness and the stylistic use of language suggest a clear perceived boundary between retroflex and non-retroflex sounds. While the retroflex sounds are indexical of the (perceived) “Mainland Mandarin,” non-retroflex sounds index Taiwan Mandarin. By underlining and producing the “non-standard”, non-Beijing elements in Taiwan Mandarin, they symbolically position themselves as equal in status to mainland Chinese, rather than subordinate to them. To understand how this comes to be, consider the following example of Rye’s enactment of his version of “Mainland Mandarin”:

Excerpt (3):

“No wénhuà

1RyeYě yǒu shuō jiùshì, “Nǐ zhēnde hěn méiyǒu wénhuà yē”
“Also there's like, [you say] “you have completely no wénhuà [i.e. uncivilized]!”
2Tāmen jiù hui shuō, “Méi wénhuà-ér zhēn kěpà!”
And they would say, “[having no] wénhuà-er is terrible!”

In this short string of speech, Rye at first uttered his Taiwan Mandarin version of “Nǐ zhēnde hěn méiyǒu wénhuà yē” ‘you are completely uncivilized’ (line 1). He immediately followed up enacting his version of Mainland Mandarin (i.e. the phrase in red in line 2). The differentiation was done through a change of phonological details, in this case, (1) the retroflex initial [ʈʂ] for zhēn ‘really’, (2) postvocalic Erhua [ɻ] after wénhuà ‘cultured, civilized’, and (3) tonal and prosodic contrasts (see Figure 3 below for illustration).

Prosody of the two utterances of “You’re completely uncivilized”.
Figure 3:

Prosody of the two utterances of “You’re completely uncivilized”.

First of all, note the phonetic difference of zhēn “really” between Rye’s first alveolar affricate [tʃən˥] and his second retroflex affricate [ˈʈʂən˥]. 7 He specifically enforced a stressed [ˈʈʂ] to differentiate from his usual relaxed [tʃ]. In addition to the non-retroflex vis-à-vis retroflex distinction, note also the rhotacization, the postvocalic Erhua [ɻ], of the common noun wénhuà [uən↿ huaɻ⇃] in his Mainland Mandarin version. Moreover, although never overtly commented upon by my informants themselves during this and other interviews, in practice many informants, while they enacted their stereotypically “Mainland Mandarin”, would change the prosody as well as the syntactic structure of the utterance. In his enactment, Rye displayed a drastic prosodic difference and a syntactic distinction between the two versions (see underlined tokens in Figure 3).

In his normal speech, instead of pronouncing the words de [tə] and méiyǒu [mɛɪ↿ joʊ⇃˦] in their fully developed form (i.e. as designated in Standard Guoyu and PTM), he exhibited phonetic reduction in the particle de [ʔ] and phonetic contraction in méiyǒu [mɛɪ↿‿oʊ⇃] that are common in Taiwan Mandarin (Chung 2006a). In addition, observe that the third tone (the low-dipping tone) in the second syllable of méiyǒu [mɛɪ↿‿oʊ⇃] has no rise, a feature of Taiwan Mandarin influenced by Taiwanese (Kubler 1981: 105–106). Perhaps these features that diminish the distinctiveness of each syllable made the speaker believe that Taiwan Mandarin is more “flat” and “soft”, a point often made by my informants. In comparison, in Rye’s “Mainland Chinese voice”, every syllable was fully developed and clearly articulated. Additionally, here, on top of tonal differences, we can also observe alternations in other suprasegmental features. Unlike the relatively “flattened” Taiwan Mandarin utterances, the conceived stereotypical Mainland Mandarin utterance is distinctive in its high-low punctuative pitch contour (see Figure 3 above). It is characterized as a string of consecutive and audibly distinguishable individual beats. Such prosodic distinction contributes to the perceived differences between Taiwan Mandarin and Mainland Mandarin.

Lastly, let us briefly look at the syntax of negation for the phrase “not have” in both pieces of this utterance. According to Kubler (1981, 1985), due to the influence 8 from Taiwanese-speaking learners of Guoyu in the earlier stages, Taiwan Mandarin demonstrates a higher frequency of as well as grammatical tolerance for the addition of yǒu and méiyǒu as aspectual auxiliaries before main verbs (Kubler 1985: 161–165), a feature that Beijing Mandarin [i.e. Putonghua] does not permit. Kubler observes that, “in Taiwan Mandarin, méiyǒu is usually used in all positions where Beijing Mandarin would have only méi” (1985: 162; italics added for formatting consistency). Here, we see the exact pattern of such syntactic construction. Rye reduced yǒu from méiyǒu in his Mainland Mandarin utterance. Despite the fact that he regularly uses méiyǒu in his normal speech, he did not metalinguistically comment on, or show any awareness of his syntactic adjustment in his Mainland Mandarin utterance.

As stated above, past studies have accounted how retroflex sounds in the Taiwanese context are indexical of socio-spatial values – prestige, higher socio-political achievement, and post-WWII Chinese-descendent identity. Such an indexical field, however, is created based on the intra-national analogy between Taiwan Mandarin (and/or Taiwanese-accented Mandarin) and Standard Guoyu. With respect to the Taiwan-PRC comparison, my informants have shown what I see as an example of fractal recursivity (Irvine and Gal 2000) that reconfigures the indexicality of retroflex sounds. These informants are, relatively speaking, in constant and intensive contact with the mainland Chinese compared to the majority of the Taiwanese. Also, given that many of informants as salesclerks rely on the business with the mainland Chinese tourists, they have to consciously accommodate their own speech form with their mainland Chinese guests. These factors help to erase (Irvine and Gal 2000) the indexical link between the sounds and other socio-political realities within Taiwan that the past literature has pointed out. Take as another example the map made by Amy, a nineteen-year-old counter clerk at a souvenir store that specializes in Taiwanese traditional snacks (Figure 4).

Amy’s map of Mandarin language community.
Figure 4:

Amy’s map of Mandarin language community.

As seen in her own analysis (see the scale she made at the far-left side of her map), if one must compare and scale different Mandarin varieties based on accent, Taiwan Mandarin (labeled as “TW”) is least accented, while Putonghua, which she labelled as 北京 ‘Beijing’, i.e. Beijing Mandarin, is most heavily accented – 捲舌腔 juǎnshé qiāng ‘tongue-curling accent’, as she put it. Instead of viewing Taiwan Mandarin as a vernacular, she labelled it as the 最標準 zuì biāozhǔn “most standard” Mandarin variety.

Yet, with regard to such metapragmatic labeling, she followed up arguing firmly that such analogy is inappropriate, as Taiwan Mandarin should not be considered a part of Mainland Mandarin communities. Also, similar to Yuan (cf. Figure 2), Amy provided attitudinal evaluations of Taiwan Mandarin and Mainland Mandarin varieties. In comparison to the Taiwanese 軽声 qīngshēng ‘softness’ and 細語 xìyǔ ‘gentleness’ in their speech, the mainland Chinese are evaluated as 不舒服 bùshūfu ‘unpleasant’, 快 kuài ‘fast’, and 剁剁逼人 9 duōduō bīrén “aggressive”. Such dichotomized rationalizations further help to reinforce the semiotic boundary drawn between Taiwan Mandarin and Mainland Mandarin, the very semiotic process “whereby diverse behavioral signs (whether linguistic, non-linguistic, or both) are functionally reanalyzed as cultural models of action” (Agha 2007: 55).

As seen in these examples, the indexical connection between retroflex sounds and the post-WWII Chinese-descendent identity and higher socioeconomic status has instead been projected onto a different analogical layer that contrasts the Taiwanese speech with the mainland Chinese speech. While the retroflex/non-retroflex distinction formerly indexed class or status differences, it now indexes national identity. These semiotic processes produced an additional indexical association, the N+1th order of indexicality to signify territorial boundaries and national identities.

As shown in Figure 5, this reconceptualization reconfigures the prescribed hierarchical order under the standard language ideology as stated in the previous section (see also Figure 1). At one level, it valorizes Taiwan Mandarin and frees it from being presented as a regional vernacular of Standard Guoyu and/or Putonghua. The sociolinguistic hierarchy is perceptually abated to the degree where Taiwan Mandarin is considered as equally prestigious as Standard Guoyu. At another level, curling the tongue for the r-coloring sounds is perceived as a non-Taiwanese linguistic practice.

The N-th and N+1th order of indexicality of Taiwan Mandarin.
Figure 5:

The N-th and N+1th order of indexicality of Taiwan Mandarin.

This N+1th layer of indexicality reverses the Chinese ideology about Mandarin Chinese, which takes Taiwan Mandarin as merely one of its many regional vernaculars. Rather than indexing sociolinguistic differences within Taiwan, the retroflex/non-retroflex distinction now becomes a well-defined boundary line that clearly distinguishes Taiwan Mandarin from Mainland Mandarin as the language of a different speech community (as marked by blue-colored square diagram and the dotted double line). Figure 6 further demonstrates how the informants constructed a semiotic boundary between Taiwan Mandarin and Mainland Mandarin.

The reconfiguration of the order of Mandarin in Taiwan in comparison to the PRC (China).
Figure 6:

The reconfiguration of the order of Mandarin in Taiwan in comparison to the PRC (China).

Rhotacization and retroflex sounds have thus been rationalized to be iconic of Mainland Mandarin, blurring the line between Guoyu and Putonghua. I see this highlighting of traditionally non-standard linguistic elements and metapragmatic attitudinal judgments as a semiotic endeavor to establish a clearer boundary between Taiwan Mandarin and Mainland Mandarin, as well as their speaker’s identities. Reflecting on the earlier discussion of Rye’s use of non-retroflex sibilants, low-dipping [⇃˦] third tone, phonetic contraction (cf. Figure 3), we see how these phonological features play in my informants’ ideologization of Taiwan Mandarin, as demonstrated in their metapragmatic comments, maps, and enactments. These result in the perception of Taiwan Mandarin being relatively ‘flattened’, i.e. having a much smoother and fluent prosodic contour. Thus from the perspective of Taiwan Mandarin speakers, Mainland Mandarin differs in that the tonal distinctions still prevail, in particular the full-fledged third tone, which make syllable-initials and the ‘up-and-down’ prosody very salient. Each word is thus a distinguishable and forced beat to the ears of many Taiwanese people. This explains why they perceive Mainland Mandarin as very “heavy”, “strong” and even “boisterous”.

Avineri and Kroskrity (2014) contend that what defines a speech community is based on its members’ phenomenological conceptualization of social boundaries and temporal borders regarding their perceptions of the other and temporalization of shared memories. More specifically, it is not the language per se that differentiates one group from another, but rather, it is the semiotic perception of a system, language being only a part of it, that constitutes them as a group unique from others. It is by appealing to the sociality of speech and the perceptual differentiation of Mandarin varieties that Taiwanese people are able to semiotically present themselves (Irvine and Gal 2000) as not belonging to the larger Mandarin-speaking language community. The perceived particularities of Taiwan Mandarin enable these Taiwanese salespersons to claim agency and membership in a speech community that is different from other PRC Mandarin speech communities, while at the same time utilizing linguistic resources for doing business with the mainland Chinese tourists. The perceived linguistic distinctiveness between these two varieties of Mandarin across the Strait are projected and semiotically mobilized as the justification for the political and socio-cultural diversion between Taiwan and the PRC.

7 Conclusion

As Mandarin Chinese gradually attains the global status and sees a significant increase in its speaker populations, a careful handling of the sociolinguistic complexity of Mandarin communities within and without the Greater China becomes crucial to studies of global Mandarin Chinese. This article dealt with such complexity by focusing on the sociolinguistic and political economic implications of the stylistic uses of Taiwan Mandarin in tourism sector in Taiwan that targets on Chinese tourists.

The history of Taiwan as a receiver of immigrants and their distinct ideologies towards language, ethnicity, and nation is reflected in the multiplicity of Taiwanese identities. The increased interactions with the PRC (Mainland China) in recent years at the state and individual level have inspired Taiwanese society to search for its identity and autonomy. Particularly with respect to the opening of the Taiwan-PRC border to mainland Chinese tourists in 2008, Mandarin becomes an exceptionally profitable asset for the promotion of “exotic” Taiwanese-ness to “other” Mandarin-speaking people. In the meanwhile, with regard to the recent Cross-Strait relation, the PRC’s global-level politico-economic influence and its firm stance against Taiwanese nationalism have greatly increased Taiwanese people’s anxiety about Taiwan involuntarily becoming more regionalized and dependent on the PRC in the globalized world. Speaking Taiwan Mandarin enables its speakers to assert and maintain their Taiwanese identity in the face of “other” Mandarin-speaking people.

In this article, through analyzing metalinguistic comments, self-drawn dialectal maps, and language enactments, I have attempted to demonstrate how Taiwanese people, particularly those in tourism section, rationalize the language structure and use that are indexical of Taiwan Mandarin to support ideological differentiation and boundary maintenance between Taiwan and the PRC. I began with an analysis of the multiplicity of speaker’s awareness of the linguistic differences between the two Mandarin varieties. Moreover, I detailed the ways in which Taiwanese informants, by manually drawing linguistic boundaries of Mandarin varieties on maps, visualize the perceived linguacultural borders and thus reinforce a sense of belonging by the very lines they created on their maps. I offered examples where the distinction between non-retroflex [ts, tsh, s] and retroflex [ʈʂ, ʈʂh, ʂ, ʐ] within Taiwan is enregistered (Agha 2007; Silverstein 2003) to indicate the linguistic boundaries between Mandarin varieties across the Taiwan Strait. This reconfiguration of standard-vernacular hierarchy between different Mandarin varieties involves semiotic processes (Irvine and Gal 2000; Silverstein 2003) that revalorize linguistic features of Taiwan Mandarin through mobilizing a nationalist pride to compete with the hegemony of Standard Guoyu and Putonghua. Through this article, I supplemented the existing literature on Taiwan Mandarin that is preoccupied with international comparisons with transnational examples. I hope that I have offered possible alternatives for future studies of global Chinese in understanding the effects of macro-social construals on Chinese speakers’ concurrent language ideologies about the complicated relationship between language and the micro-social interactions.


Earlier versions of this article were originally prepared for and presented at the annual meetings of the Language, Indexicality and Belonging, Linguistic Anthropology Conference in the University of Oxford, April 7–8, 2015, and in American Association for Applied Linguistics 2016 Conference in Orlando, Florida, April 9–12 in the colloquium “Chinese in the Globalizing Era: Negotiating Identities and Ideologies”. I would like to especially thank Paul Kroskrity, Erin Debenport, Bal Sharma, Nancy Hawker, Kinga Kozminska, and the anonymous reviewer for their critical comments, editorial advice, and encouragement during various phases of the work involved in producing this article. Thanks also to Michael Silverstein, Wenhao Diao, Norma Mendoza-Denton, Elinor Ochs, Marjorie Goodwin, Olga Yokoyama, and conference participants for insightful and encouraging discussion.


  • Agha, Asif. 2007. Language and social relations. Cambridge: Cambridge University Press. Google Scholar

  • Anderson, Benedict. 1991. Imagined communities: Reflections on the origin and spread of nationalism. London: Verso. Google Scholar

  • Avineri, Netta & Paul V. Kroskrity. 2014. On the (re-)production and representation of endangered language communities: Social boundaries and temporal borders. Language & Communication 38. 1–7. Web of ScienceCrossrefGoogle Scholar

  • Blommaert, Jan & Jef Verschueren. 1998. The role of language in European nationalist ideologies. In Bambi Schieffelin, Kathryn Woolard & Paul Kroskrity (eds.), Language ideologies: Practice and theory, 189–210. New York, NY: Oxford University Press. Google Scholar

  • Bourdieu, Pierre. 1977. The economics of linguistic exchanges. Social Science Information 16(6). 645–668. CrossrefGoogle Scholar

  • Bourdieu, Pierre. 1990. The logic of practice. Stanford, CA: Stanford University Press. Google Scholar

  • Bourdieu, Pierre. 1991. Language and symbolic power. Cambridge, MA: Polity Press. Google Scholar

  • Briggs, Charles L. 1986. Learning how to ask: A sociolinguistic appraisal of the role of the interview in social science research. New York, NY: Cambridge University Press. Google Scholar

  • Brown, Melissa J. 2004. Is Taiwan Chinese?. Los Angeles, CA: University of California Press. Google Scholar

  • Chao, Yuen Ren. 1968. A grammar of spoken Chinese. Berkeley, CA: University of California Press. Google Scholar

  • Chen, Su-Chiao. 2010. Multilingualism in Taiwan. International Journal of the Sociology of Language 2010(205). 79–104. Google Scholar

  • Cheng, Robert L. 1985. A comparison of Taiwanese, Taiwan Mandarin, and Peking Mandarin. Language 61(2). 352–377. CrossrefGoogle Scholar

  • Chun, Allen. 1996. Fuck Chineseness: On the ambiguities of ethnicity as culture as identity. Boundary 2 23(2). 111–138. CrossrefGoogle Scholar

  • Chung, Karen S. 2006a. Contraction and backgrounding in Taiwan Mandarin. Concentric: Studies in Linguistics 31(1). 69–88. Google Scholar

  • Chung, Karen S. 2006b. Hypercorrection in Taiwan Mandarin. Journal of Asian Pacific Communication 16(2). 197–214. CrossrefGoogle Scholar

  • Eckert, Penelope. 2012. Three waves of variation study: The emergence of meaning in the study of sociolinguistic variation. Annual Review of Anthropology 41. 87–100. Web of ScienceCrossrefGoogle Scholar

  • Friedrich, Paul. 1989. Language, ideology, and political economy. American Anthropologist 91(2). 295–312. CrossrefGoogle Scholar

  • Hsu, Hui-Ju. 2014. Zài zúqún yǔ yǔyán jiēchù xià xíngchéng de Táiwān Huáyǔ – Cóng shēngxué fēnxī de jiēguǒ kànqǐ [Taiwan Mandarin, a Mandarin variety formed under the social and language contact between various Chinese dialects and their speakers]. Language and Linguistics 15(5). 63–662. Google Scholar

  • Hsu, Jia-Ling. 1994. Englishization and language change in modern Chinese in Taiwan. World Englishes 13(2). 167–184. CrossrefGoogle Scholar

  • Irvine, Judith T. 1996. Introduction. Journal of Linguistic Anthropology 6(2). 123–125. CrossrefGoogle Scholar

  • Irvine, Judith T. & Susan Gal. 2000. Language Ideology and linguistic differentiation. In Paul Kroskrity (ed.), Regime of language: Ideologies, polities, and identities, 35–83. Santa Fe, NM: School of American Research. Google Scholar

  • Jernudd, Björn H. 1985. Chinese language contact: An introduction. Anthropological Linguistics 27(2). 119–121. Google Scholar

  • Kroskrity, Paul V. 2000. Regimenting languages: Language ideological perspectives. In Paul Kroskrity (ed.), Regime of language: Ideologies, polities, and identities, 1–34. Santa Fe, NM: School of American Research. Google Scholar

  • Kroskrity, Paul V. 2010. Language ideologies: Evolving perspectives. In Jürgen Jaspers, Jef Verschueren & Jan-Ola ÖStman (eds.), Society and language use, 192–211. Amsterdam: John Benjamins Publishing Company. Google Scholar

  • Kubler, Cornelius C. 1981. The development of Mandarin in Taiwan: A case study of language contact. Ithaca, NY: Cornell University dissertation. Google Scholar

  • Kubler, Cornelius C. 1985. The influence of Southern Min on the Mandarin of Taiwan. Anthropological Linguistics 27(2). 156–176. Google Scholar

  • Li, David C.-S. 1985. Problems and trends of standardization of Mandarin Chinese in Taiwan. Anthropological Linguistics 27(2). 122–140. Google Scholar

  • Li, David C.-S. 2006. Chinese as A Lingua Franca in greater China. Annual Review of Applied Linguistics 26. 149–176. Google Scholar

  • Liao, Silvie. S.-W. 2008. A perceptual dialect study of Taiwan Mandarin: Language attitudes in the era of political battle. Paper presented at the Proceedings of the 20th North American Conference on Chinese Linguistics (NACCL-20), Columbus, OH, 25–27 April. Google Scholar

  • Lippi-Green, Rosina. 2011. English with an Accent: Language, ideology, and discrimination in the United States. London: Routledge. Google Scholar

  • Meek, Barbra A. 2009. “Language ideology and aboriginal language revitalization in the Yukon, Canada.” In Native American Language Ideologies: Beliefs, Practices, and Struggles in Indian Country, edited by Paul V. Kroskrity and Margaret C. Field. Tucson, AZ: The University of Arizona Press. Google Scholar

  • Milroy, James. 2001. Language ideologies and the consequences of standardization. Journal of Sociolinguistics 5/4. 530–555. Google Scholar

  • Preston, Dennis. 1999. Handbook of perceptual dialectology, vol. I. Amsterdam: Benjamins. Google Scholar

  • Preston, Dennis. 2002. Handbook of perceptual dialectology, vol. II. Amsterdam: Benjamins. Google Scholar

  • Rampton, Ben. 2009. Speech community and beyond. In C. Nikolas & J. Adam (eds.), The new sociolinguistics reader. New York, NY: Palgrave Macmillan. Google Scholar

  • Sandel, Todd L. 2003. Linguistic capital in Taiwan: The KMT’s Mandarin language policy and its perceived impact on language practices of bilingual Mandarin and Tai-gi speakers. Language in Society 32. 523–551. Google Scholar

  • Shìjiè Huáyǔ-wén Jiāoyù Huì [World Chinese language association]. 2012. Guóyǔ yùndòng bǎinián shǐluè [A brief history of a century of national language Guoyu movement]. Taipei: Mandarin Daily News. Google Scholar

  • Silverstein, Michael. 1979. Language structure and linguistic ideology. In Paul Clyne, William Hanks & Carol Hofbauer (eds.), The elements: A parasession on units and levels, 193–247. Chicago, IL: Chicago Linguistics Society. Google Scholar

  • Silverstein, Michael. 1985. Language and the culture of gender: At the intersection of structure, usage, and ideology. In Elizabeth Mertz & Richard Parmentier (eds.), Semiotic mediation, 219–259. Orlando, FL: Academic Press. Google Scholar

  • Silverstein, Michael. 1996. Monoglot a standard in America: Standardization and metaphors of linguistic hegemony. In Donald Brenneis & Ronald Macaulay (eds.), The matrix of language: Contemporary linguistic anthropology, 284–306. Boulder, CO: Westview Press. Google Scholar

  • Silverstein, Michael. 1997. Encountering language and languages of encounter in North American ethnohistory. Journal of Linguistic Anthropology 6(2). 126–144. Google Scholar

  • Silverstein, Michael. 2003. Indexical order and the dialectics of sociolinguistic life. Language and Communication 23. 193–229. CrossrefGoogle Scholar

  • Silverstein, Michael. 2013. Discourse and the no-thing-ness of Culture. Signs and Society 1(2). 327–366. CrossrefGoogle Scholar

  • Sugano, Atsushi. 2012. Taiwan no Gengo to Moji: Kokugo, Hōgen, and Moji Kakaku [The language and characters of Taiwan: The national language, dialects, and character renovation]. Tokyo: Keisfōshobō. Google Scholar

  • Wei, Jennifer M. 2006. Language choice and ideology in multicultural Taiwan. Language and Linguistics 7(1). 87–107. Google Scholar

  • Woolard, Kathryn & Bambi Schieffelin. 1994. Language ideology. Annual Review of Anthropology 23. 55–82. CrossrefGoogle Scholar

  • Wu, Ming-Hsuan. 2009. Language planning and policy in Taiwan: Past, present, and future. Working Papers in Educational Linguistics 24(2). 99–118. Google Scholar

  • Zhang, Qing. 2005. A Chinese yuppie in Beijing: Phonological variation and the construction of a new professional identity. Language in Society 34. 431–466. Google Scholar


  • 1

    In this article, unless otherwise specified, the term Mainland Mandarin does not refer to any specific Mandarin varieties. Rather, the term is used to refer to the perceived Mandarin variety, something that my informants believe all mainland Chinese speak regardless of their home provinces. 

  • 2

    In order to not impose any preconceived categories, I used open-ended questions such as “What are the languages you consider your native language?”, “What are the languages you use most often at home?” and “How would you describe your identity?” This approach enabled me to understand how my informants labeled the linguistic varieties they use/encounter on everyday basis and identity categories they self-identify with in their own term. This also avoids the risk of eliciting biased results from using overtly and covertly politicized local terms for Chineseness (see Chun 1996; for the discussion on the ambiguities of ethnic terms for Chinese; see also Sugano 2012). In this article, I used only the terms they self-identified using, such as Taiwan Mandarin, Taiwanese, etc. 

  • 3

    Throughout this article, all informants are referred to by pseudonyms to ensure their anonymity. Personally identifiable information is also removed. 

  • 4

    In all transcription excerpts, Zhuyin symbols (a.k.a. Bopomofo) are italicized and transcribed in their corresponding Pinyin symbols. Zhuyin is the phonetic notation system used in Taiwanese educational institutes. It was first developed in 1912–13 to enhance literacy by the Republic government (Shìjiè Huáyǔ-wén Jiāoyù Huì 2012Shìjiè Huáyǔ-wén Jiāoyù Huì 2012) and implemented in Taiwan since WWII by the KMT. It consists of 37 symbols, of which 21 are used to designate consonants and 16 vowels. As demonstrated in Excerpt (1), most Taiwanese would make use of Zhuyin symbols to discuss phonological aspects of Taiwan Mandarin. 

  • 6

    All Chinese characters used in this article are exact duplications of what my informants wrote on their maps. As such, inevitably, there is an inconsistency in the use of traditional Chinese characters and simplified characters. 

  • 7

    The particle de in zhēnde in line 1 is a particle that links the attributive word zhēn with its head. The meaning of zhēnde, however, is completely identical to zhēn in line 2. 

  • 8

    Words ū ‘have’ and bóu ‘not have’ in Taiwanese are more extensively used in the syntactic structure than the Mandarin counterparts yǒu and méiyǒu (Kubler 1981, 1985; Cheng 1985). 

  • 9

    Instead of writing Chinese characters 剁剁, she used Zhuyin symbols ㄉㄨㄛˋwith the mathematical symbol square foot “ 2” to present her categorization of 攻擊性gōngjī xìng ‘aggresiveness’. For clarity, here I used the corresponding Chinese characters. 

About the article

Published Online: 2018-03-28

Published in Print: 2018-03-26

Citation Information: Global Chinese, Volume 4, Issue 1, Pages 11–36, ISSN (Online) 2199-4382, ISSN (Print) 2199-4374, DOI: https://doi.org/10.1515/glochi-2018-0002.

Export Citation

© 2018 Walter de Gruyter GmbH, Berlin/Boston. Copyright Clearance Center

Comments (0)

Please log in or register to comment.
Log in