Introduction

Recent years have witnessed a surge in interest in preservation of endangered languages among language policy researchers, but insufficient attention has been paid to endangerment of regional dialects (Schilling-Estes and Wolfram 1999; Tulloch 2006). A range of case studies conducted in different geological contexts, such as Canada (Tulloch 2006) and China (Gao 2015) have revealed that dialect preservation is a dynamic and interactive process. Yet, little empirical research has been done to have an integrated holistic approach to language planning practices.

A process-oriented view towards language planning has become particularly important since the policy-making mechanism not only has to take into account the complex demographic and socio-cultural realities, where the local and micro language policies are to be initiated and implemented (Zhao and Baldauf 2012), but also needs to accommodate strong bottom-up voices, by involving individuals in making decisions about endangered languages. While the shift of focus adds substantial complications to language policy-making, it means roles of various stakeholders in the policy-making and implementation process should be emphasized in research on the process of formulation of policy on endangered languages and dialects. From this point of view, in the case of the somewhat oversimplified and marginalized language planning, the traditional top-down language planning process now needs to be complemented by bottom-up examinations of interactions between multiple stakeholders at a more contextualized level.

Integration of bottom-up voices in the decision-making process is of immense social significance particularly in a multilingual society such as mainland China, where Chinese, or Mandarin, is a linguistic system with ten major sub-branches (regional dialects), each having its own sub-branches (regional varieties of dialects). There has been an upsurge of social concerns about the vitality of Chinese dialects and such concerns have even triggered outcries or protests in different parts of mainland China. For example, Gao (2012) and Liang (2015) contend that in Guangzhou, one of the major Chinese metropolises, local citizens’ reactions, and struggles to defend Cantonese require language policy-makers and the public to engage in important dialogs related to the status of Cantonese as the regional lingua franca. Similar social tensions have also been observed in Shanghai, the largest city in China. Citywide and even nationwide debates over local children’s poor proficiency of ‘Hu Yu’ (沪语) or ‘Shanghai dialect,’ a regional variety of Wu dialect group spoken in much of the economically developed Yangtze River Delta on the east coast of China, have made headlines in print media coverage as well as social media spaces, leading to an ongoing movement to “save Shanghai dialect” across the academia and the general public since 2012. In light of these social realities, this study probes the dynamic interplay between language policy-makers and local stakeholders with regard to protection of Shanghai dialect in Shanghai.

The Process Model in Language Management Theory as an Analytical Framework

As players from multiple spheres are getting increasingly involved in the complex process of language planning, analytical, and theoretical approaches able to address the complexity for explanatory and interpretative purposes are needed. The Language Management Theory (LMT) (Jernudd and Neustupný 1987; Neustupný and Nekvapil 2003; Kimura 2014) has emerged as one such theoretical option to capture the dynamic process with intersecting perspectives on the interplay of bottom-up and top-down approaches. The LMT as an applied theory emphasizes the different processes in language planning by proposing a cyclical model as follows (see Fig. 1):

Fig. 1
figure 1

The LMT model (adapted from Kimura 2014, p. 267)

Full size image

As can be seen in Fig. 1, the model begins in a particular language use context or language situation with the noting of deviation from norms or expectations at multiple levels of different stakeholders. Following the noting stage, the deviation may be evaluated according to social conventions and preferences in the subsequent stage (Edwards 1994). Afterwards, an adjustment plan is selected and implemented. Kimura (2014) also incorporates the process of feedback as the final stage to the LMT process before new deviations are noted to trigger another cycle of language planning. This model of the process of language planning can integrate the bottom-up voices of multiple stakeholders during language planning because it has a thorough, dynamic, and interactive facet of micro language planning in dialect communities (Baldauf 2012).

Linguistic Insecurity and Individual Agency

Any deviation in norms can be noted and evaluated as a symptom of linguistic insecurity. The notion of linguistic insecurity (LI) is theorized as a measure of speakers’ negative attitudes toward the prestige of certain linguistic forms, and a relation between a judgment of normativity and a self-evaluation for linguistic change in social stratification (Labov 2006). Calvet (2006) further proposes two types of linguistic insecurity, statutory insecurity as a result of speakers’ negative evaluation of the status of own varieties in comparison with another language or variety and identity insecurity that speakers may have when using a language different from that used by the community they identify themselves with (Escandell 2011). The notion of linguistic insecurity has become an increasingly important issue for language policy-makers as language planning is often motivated by feelings and attitudes of insecurity in the community (Ager 2001).

A closely connected concept that is different from the notion of language insecurity is agency, which refers to the individuals’ capacity to act linguistically as well as socially in a controlled and knowing way, usually juxtaposed with structure (O’Byrne 2011). The interplay in social life between the agency, our freedom to live as we choose, and the structure constraining how we live has proved to be a vital topic in socio-linguistics (Bell 2014). Language policy researchers have also begun to stress the significant role individual agency plays in the process of language planning (Liddicoat 2008). Zhao and Baldauf (2012) argue for the necessity of examining the role of the individual agency in different stages of language planning. Interactions between individual stakeholders when constructing emerging language use situations need to be systematically explored together with the power structure of relevant language policies. Therefore, this study addresses the following research questions to better understand the interplay between language policy-makers and multiple stakeholders in the discussion on ‘saving Shanghai dialect’:

  1. (1)

    How did local stakeholders inform the process of language planning to save the Shanghai dialect?

  2. (2)

    How did language policy-makers respond to the bottom-up voices of local stakeholders in the process?

The Study

This study is a part of a larger project that explores linguistic (in) security in multilingual/multidialectal Chinese society with a focus on the efforts to save Shanghai dialect in Shanghai.

The Research Context

In the past 60 years, China’s official language policy has in reality restricted the use and acquisition of regional dialects although it avows that the promotion of Putonghua as the national standard speech does not ‘wipe out dialects artificially, but reduces the scope of dialect use progressively’ (Guo 2004). In Shanghai, for example, 97 % of the municipal population can speak Putonghua (Zhang 2014). In contrast, Shanghai dialect has been undergoing rapid changes and is being assimilated into Putonghua. According to a report by Shanghai Social Academy, more than 40 % of kids in schools cannot speak Shanghai dialect well (Zhou 2012), leading to social tensions over dialect crisis in recent years. Calls for ‘saving Shanghai dialect’ have spread from the academia to the general public, from parents to educators and from communities to government agencies. Three discrete groups of stakeholders are involved in the process to taking actions; people with expertise, people with influence, and people with power in language planning (Zhao and Baldauf 2012). On January 5 2012, 82 Shanghai-based linguists and language professionals jointly signed a call for a public initiative to protect Shanghai dialect, which was covered in the local Oriental Morning Post (www.dfdaily.com) (Shi 2012). The proposal triggered successive interactions on dialect protection and policy adjustments between multiple stakeholders in the local community and language policy-makers. Such interactions feature a unique process of bottom-up language planning for analysis in this study.

Participants

The main study was conducted through face-to-face semi-structured interviews with ten individual stakeholders engaged or involved in the process of ‘saving Shanghai dialect.’ Among these stakeholders, there were two language policy officials of the Shanghai Municipal Language Committee (SMLC), an institution in charge of language management and affiliated to Shanghai Municipal Education Commission. Others included professionals involved in dialectological studies related to the Shanghai dialect, playing important roles in language planning. These individuals are two linguists working with departments of linguistics studies in Shanghai. Interviews were also conducted with two delegates to the local committee of the Chinese People’s Political and Consultative Conference (‘CPPCC’), which is a regional political advisory body made up of delegates from a range of political parties and organizations in Shanghai. Since pressure groups supporting dialect preservation have put in great efforts to promote Shanghai dialect on social media sites, two members of these groups were also interviewed. Both members were MA graduate students in the researcher’s affiliated university. Lastly, interviews were conducted with two parents whose children were included in trial programs to preserve the dialect in the kindergartens.

All participants indicated their willingness to be interviewed by signing informed consent forms. Their brief background information, with pseudonyms, is presented in Table 1.

Data Collection and Analysis

A variety of data were collected to arrive at answers to the research questions. Before the field research, local language policy texts, policy proposals, newspaper coverage, and social media posts (e.g., Weibo) on saving Shanghai dialect were studied. In total, the researcher first collected from the SMLC a total of 14 proposals from delegates to municipal CPPCC and People’s Congress and 44 petition letters through Xinfang, a kind of petitioning system in China, both related to the movement for preservation of the Shanghai dialect. These documents helped familiarize the researcher with the relevant issues.

The primary data were collected through three rounds of fieldwork during October 2012 to December 2012, August 2013 to October 2013 and November 2014 to February 2015, so that the dynamic features of the planning process could be captured. In each round of fieldwork, one-to-one interviews were conducted with ten participants, each one lasting between 50 and 80 min, in Chinese. During the interviews, participants were encouraged to talk about their viewpoints and experiences during the process of saving Shanghai dialect. Field notes were also taken during interactions with participants at different sites (e.g., offices, coffee shops, and restaurants).

Data analysis started during the fieldwork, when the researcher took down notes and reflected on the interviews. Three rounds of interviews were transcribed in Chinese, and then translated into English by an independent research assistant. The accuracy of the translation was checked by back translation. When the transcription of an interview was done, the researcher analyzed it qualitatively by identifying coding and categorizing the content within the framework of the process model of LMT inductively to reveal the interactions between multiple stakeholders in language planning.

The researcher started out with coding of processes in the planning phases; the codes reflect five stages in the LMT model: ‘Noting insecurity of Shanghai dialect,’ ‘Evaluation of the dialect in a crisis,’ ‘Adjustment plans for language planning,’ ‘Implementation of relevant language policy initiative’ and ‘Feedback on implementation.’ Then the researcher assigned concepts under each of the process codes through a range of themes coding and using the participants’ wordings to name the concepts. The researcher then identified the relationships between the concepts and classified them into topics through analytical coding (Richards and Morse 2007), with the topics subsumed under the process codes. A final version of the coding structure is presented in Table 2. Furthermore, data analyzed in the interviews were triangulated with multiple sources of supplementary data (e.g., policy proposals, petition letters, media coverage, and social media posts) during the process.

Findings

The results presented in this section are in accordance with the coding structure: the first layer headings of this section reflect the first-order process in the coding structure, adopted from the five stages in the LMT model; the second layer headings correspond to the second-order themes while the third-order concepts are embedded in the illustrations in each sub-section. Multiple sources of data are triangulated in this section. The first three stages in this structure correspond to the first research question whereas the other two stages respond to the second question.

Stage 1: Noting the Crisis of Shanghai Dialect

Perceptions of Linguistic Insecurity

Linguists and language professionals who first perceived formal linguistic insecurity undertook efforts to call for treating protection of Shanghai dialect as an important agenda for language policy-makers. The interview data show that long before the initiative by linguists and language professionals in 2012, feelings of anxiety and concern over the weakening proficiency in Shanghai dialect were widespread in academic circles and the local community (P1 and P2). P1 noted that many linguists began to express their worries about the status of Shanghai dialect as early as 2004. P2 remembered that the voices calling for preservation of Shanghai dialect have been heard outside the academic community since 2011 when Shanghai started a large-scale linguistic field survey. The following extracts are representative of participants’ views on the growing awareness of the sense of language insecurity over the use of Shanghai dialect. When asked about the decline of Shanghai dialect, P3 talked about his personal experience in the linguistic field survey:

The ideal speaker should be the one born and educated in the local community, with less than 4 years’ out-of-town living experience, and whose parents or spouse are all local people. We found few such qualified young speakers in some suburban areas… (P3, first fieldwork)

Similarly, P4 also noticed evidence of decline of the Shanghai dialect among local youngsters who have become less fluent in Shanghai dialect than in Putonghua. P3 and P4 were both linguists with professional knowledge of Shanghai dialect, representing people with expertise in language planning process. They noticed the deviation of Shanghai dialect from the ‘expected’ norm, which, as Neustupný and Nekvapil (2003) suggest, can be confirmed by grammarians on the basis of their linguistic awareness and practices. It is clear that P3 and P4 have a long held or ‘expected’ norm of the standard of Shanghai dialect (e.g., phonetics), but the findings regarding practices followed have intensified their worries about deviation from the norm. For this reason, they undertook enormous efforts to initiate and participate in the process of relevant language policy planning, using their technical expertise.

The Awareness of Language Insecurity in Status and Identity

Unlike the language experts, the public’s involvement in the discussion to save Shanghai dialect stems from awareness of decline of the dialect in various social settings. For example, P7 and P8 found that many of local students could not speak standard local dialect on campus and prefer to speak Putonghua instead. The use of Shanghai dialect has been ‘prohibited in schools and administrative sectors since 1992’ (P1, P2 and P9). In business context, P10 has to switch to Putonghua for wider communication with other workmates.

The crisis of Shanghai dialect has attracted attention from local social elites also. P1 informed the researcher that delegates to the Municipal People’s Congress and CPPCC called for preservation of Shanghai dialect. P6, one of the active initiators of the proposal, a famous actor of Huajixi, traditional Shanghai style crosstalk, feels the dialect has declined as the young audience is losing interest in Huajixi partly because ‘they complained that they could not understand Shanghai dialect in our performances’ (P6, first fieldwork). P5 shared the same view, ‘Local young people seldom use their mother tongue… (Shanghai dialect) is definitely a name card of this city, laden with local culture and its history’ (P5, first fieldwork).

Language insecurity, perceived as an affective component of language attitude, is an emotion, not a rational construct (Ager 2001). As revealed in the present study, both the grassroots and social elites conceive the falling of Shanghai dialect more as a menace to the local identity than a linguistic threat, which poses as a feeling of insecurity. It can be observed that the general public was alerted to the ‘statutory insecurity’ arising from the dysfunction of the local speech in domains of their lives (Calvet 2006). However, as Réaumé and Green (1989) note in their analysis of the root value of linguistic security, ‘Apart from its instrumental value in communication, language is also an important marker of identity’ (p. 781). P5 and P6 are people with influence and exerted their agency by playing a key role in saving Shanghai dialect in response to their awareness of identity anxiety. They identified a threat significant enough to raise sufficient awareness of the decline of the dialect and its effect on the regional identity. P5’s use of the metaphor ‘name card’ well illustrates the social elites’ beliefs of language as a fundamental marker of regional identity and a highly visible component of group solidarity. Therefore, any deviation in norm of which has been closely related to the ‘identity insecurity’ of the regional dialect triggers a feeling of insecurity, turning the linguistic issue into a social concern.

Stage 2: Evaluation of the Dialect in a ‘Crisis’

Arguing for the Evaluation on ‘Crisis’

The interplay between stakeholders who initiated the moves for protecting the Shanghai dialect and language officials on evaluation of the crisis is quite apparent. On the one hand, both social elites and grassroots unanimously argue for the evaluation that Shanghai dialect is in a ‘real crisis.’ P5 found that ‘in talks with those senior citizens in local communities, they were really disappointed that their grandchildren could hardly speak Shanghai dialect’ (second fieldwork). P6 believed that Shanghai dialect was in crisis, because generational transmission almost stopped. He even quoted his son’s situation as an example, ‘I found it a pity that my 25 years-old son cannot speak good Shanghai dialect although our family was all born in Shanghai’ (second fieldwork).

On the other hand, language officials (P1 and P2) did not agree to call the situation a ‘dialect crisis.’ P2 acknowledged the decline of Shanghai dialect in the past 20 years but he explained it by saying that ‘(Shanghai) dialect loss, in the long run, is a natural process in the age of urbanization’ (second fieldwork).

Local netizens exhibited great agency by launching successive online discussions appealing for the protection of Shanghai dialect ever since 2012 (P7 and P8). For example,

‘More than 83 % of respondents among 30,000 voters over one night strongly agreed or basically agreed that Shanghai dialect was in a crisis.’ (P7, second fieldwork)

“Shanghai dialect should be ranked on the agenda of endangered languages list.” (P8, second fieldwork)

In order to verify the above perceptions, the researcher collected data on Sina Weibo, one of the most popular micro blogging sites in China and counted 1189 posts and reposts with a total of 9800 comments concerning the topics of ‘saving Shanghai dialect’ or ‘dialect crisis in Shanghai’ during the period from 2012 to 2014. It is observed from the posts and successive comments in the virtual world that there is an evident desire for preventing Shanghai dialect from extinction and that has had an overwhelming impact on language policy-makers.

Framing the Cultural ‘Crisis’ of Haipai (海派)

In response to the official discourse that ‘dialect loss is a natural process,’ regional stakeholders framed in the interactive process a particular focus on cultural identity of the city. P5 proposed to the municipal government in 2013 that ‘(Dialect crisis) is more of a sign of the decline of Haipai culture with old memories and local knowledge of this city’ (P5). P6 expressed the same worry about the Haipai culture:

‘Just imagine what will be left after 20 more years in Shanghai concerning local culture…I do hope Shanghai dialect will remain as the only symbol (italics added) of the regional identity’ (P6, second fieldwork).

Social elites have demonstrated their pride in the Shanghai dialect as it is associated with its historical cultural heritage labeled as Haipai (literally Shanghai style), which refers to the mixture of classical Chinese culture along the Yangtze river and western cultural elements based in Shanghai. It is observed that P5 and P6 framed the ‘cultural crisis’ to call for collective actions to protect the regional dialect as a voice of group identity (Edwards 2011). The researcher’s Internet search also confirmed that Online discussions on saving Shanghai dialect echoed the framing of cultural crisis by numerous posts connecting cultural values with the regional dialect and some posts became ‘popular’ by claiming ‘the gradual fall of Shanghai culture with the dialect crisis’ (P8, second fieldwork).

It is clear that the social discursive constructions on the dialect crisis arising from feelings of insecurity are by no means a series of incidental events, but accumulated collective actions. The observed agency in evaluating the dialect crisis and framing the cultural crisis reflects a symptom of social resistance to the current language policies and signposts desires of the local community to interact with the relevant decision-makers.

Stage 3: Adjustment Plans for Dialect Planning

Plans and Strategies Proposed for Adjustment

Plans and strategies for saving the Shanghai dialect and to retain dialect vitality were proposed by various stakeholders between 2012 and 2014 in the form of complaints and petition letters, which were later passed on to the SMLC. It was found that the number of letters of this kind had risen rapidly in the past three years, accounting for 90 % of the total number of complaints and letters concerning language issues in Shanghai (P1). P2 classified them into ‘four tenets,’ i.e., (1) to increase dialect usage in the public and the media sectors; (2) to teach Shanghai dialect in schools and kindergartens; (3) to develop a writing system and standardization for Shanghai dialect; and (4) to support creative and literary writing in Shanghai dialect for aesthetic purposes.

The strategies proposed by the general public resonated with the interview data from proposal makers (e.g., P5 and P6) and drafters (P3, P4, and P7). They suggested that Shanghai dialect should be taught as a part of formal education (P3, P5, P6, and P7), and language education planning in terms of teaching materials, course books, and language proficiency benchmarks for teachers should be taken care of (P4). P5 urged language planners to expand the scope of the use of Shanghai dialect by adding dialect-medium programs in the mass media, adding public interest broadcast services in Shanghai dialect and establishing local TV channels and Radio stations to broadcast in Shanghai dialect. On the front of literacy practices, the government should encourage publishing a range of novels and other literary works (P6). These plans and strategies form a uniting ‘upstream’ agency for preserving the dialect vitality by motives to adjust (1) the uses, in contexts of public service and the mass media; (2) the users, by way of language teaching and language spread for cultural prestige; and (3) the usage, through standardization efforts. All the three motives were articulated in successive interactions between regional stakeholders and the SMLC.

Challenging the Push for Putonghua

Regional stakeholders were found to have challenged the mandatory promotion of Putonghua in line with the national language policy. For instance, P5 proposed to abolish the prohibition on the use of Shanghai dialect in schools and kindergartens since The Law of the National Commonly Used Language and Script of the People’s Republic of China (The Language Law) never stipulated this restraint but in reality, Shanghai dialect was excluded from the schools so as to promote Putonghua. P5’s argument was echoed by P9, who felt puzzled that ever since 1992, school teachers and students were not allowed to speak Shanghai dialect in schools. P6 further problematized the policy to push Putonghua in the past 20 years:

‘Since almost 97 % of the residents in Shanghai can speak Putonghua while the local dialect is on the verge of extinction, why not stop the push for Putonghua now and switch to promote Shanghai dialect with the same efforts?’ (P6, second field work)

Similarly, P7 said, ‘it is time to slow down the pace of pushing for Putonghua as a standard language which has ousted the local Shanghai dialect.’ P8 appealed that the Language Law, which neglects hundreds of local dialects is in a dire need of revision. In other words, these voices attribute the dialect crisis to the centralized policy to push Putonghua and argue against the inadequate consideration of dialects in the Language Law. These quotes imply that individual stakeholders asked for policy clarification and adjustments to expand the survival space for regional dialects so that their rights and status could be redefined by the Language Law.

Stage 4: Implementation of the Dialect Planning

Policy Shifts in Response to Culture Crisis

Influenced by the above-mentioned regional stakeholders’ requests, the SMLC responded first by commissioning ad hoc projects to investigate the language situation in Shanghai and then decided to improve the cultural transmission and preservation of Shanghai dialect (P2, second fieldwork). This policy shift is also encapsulated in P1’s comments as presented below:

‘The SMLC supports dialect transmission and preservation. We have coordinated other sectors to provide Shanghai dialect broadcast services in public transportation, a number of dialect-medium TV and radio programs in the media were on air, and 20 pilot kindergartens and schools were under pilot programs of cultural experience in Shanghai dialect.’ (P1)

In reply to P5’s proposal in 2013, SMLC committed to take measures to preserve the local culture and the Shanghai dialect. The participants pointed out the changes noted in the media and public service sectors, where Shanghai dialect programs and broadcasting could be found (P3, P4, P7, and P10). It suggests that the policy shift emerged, to a large extent, in response to the ideologically framed ‘cultural crisis’ arising from tensions about the linguistic insecurity. The policy window on dialect transmission was kept open in the interactive process between SMLC and regional stakeholders, which revealed the backdrop of the National Strategic Blueprint for Medium- and Long-term Language Reform and Development (2012–2020), which made the first ever reference to linguistic harmony and framed language in terms of a cultural marker.

Policy Constraints in Regard to Dialect Protection

It must be noted that in the process of dialect protection individual stakeholders’ agency is vulnerable to constraints imposed by the national policy. When asked about the plans challenging the push for Putonghua, P2 expressed his hesitation: ‘It is unrealistic to abolish or even undermine the policy to promote Putonghua.’ P6 interpreted the hesitation as policy constraints the local language planners were faced with, but he was disappointed that SMLC was reluctant to take specific measures to protect Shanghai dialect. P4 also noticed that the mention of specific ways for dialect protection, such as dialect teaching and use in educational contexts was sparse; P2 explained that they were cautious about the feasibility of implementation in terms of curricula, teaching materials and teachers.

As implementation is the process of putting the planning options into real policy actions, under the general policy umbrella with relevant constraints, individual stakeholders did affect the local language policy-making in certain ways. However, it needs to be noted that language planning at the bottom may not change the power structure of national language policy, which has been traditionally centralized and controlled at the top in China. The responses of the SMLC, the local language policy-maker in fact mirror the impact of the power asymmetries reflected in policy outcomes regarding dialect protection, which demonstrates limited space for change under the language regimes of policy structure.

Stage 5: Feedback on Implementation

Nuanced Changes in Dialect Planning

Several participants in the third round of fieldwork reported nuanced changes that took place in the dialect planning process. For example, dialect preservation steps such as documentation and records were greatly improved (P4), communal dialect clubs and dialect activities were sponsored (P7), and children in some pilot dialect-teaching programs were glad to pick up Shanghai dialect (P10). However, as P3 argued, ‘preservation and protection are different… I didn’t see any substantial changes in protecting Shanghai dialect.’ (third fieldwork) P5 pointed out that there was still much space to promote Shanghai dialect. The participants also expressed their concerns about dialect teaching. ‘The pilot programs of awareness raising activities in schools are not enough to help the kids acquire the dialect,’ P6 commented. P9 reflected that ‘kids need more exposure to Shanghai dialect in and out of class.’ Apparently, local stakeholders problematized the implemented policies and attempted to push for greater promotion of Shanghai dialect through interactive negotiations.

Calling for a Paradigm Shift in Dialect Planning

The participants who were actively engaged in discussions on Shanghai dialect planning process reflected on the paradigm of policy-making in Shanghai. P2 recalled that,

‘It was rare to tackle language planning issues coming from grassroots because local language policy-making and implementation was traditionally solely subject to the central government.’ (P2, third fieldwork)

Although some adjustments were made largely in response to shifting societal and contextual realities, the participants’ demands for dialect vitality mostly remained unaddressed (P4), which created new challenges for language policy-makers. One participant even challenged the role of the SMLC in implementing policy in a top-down manner and neglecting voices at the bottom (P8). Similarly, P7 believed that ‘voices from the local community unanimously of the view that teaching Shanghai dialect in schools and kindergartens should be carefully considered before it triggers new tensions’ (third fieldwork). P6 requested the SMLC to engage with multiple stakeholders in policy-related discussions and to take account of societal realities in language policy-making. It seems that with top-down influences, language planning in Shanghai will continue to be a dynamic and cyclical process that copes with shifting socio-linguistic realities and that interactively drives policy development within the constraints imposed by the national policy.

Conclusion

This paper has examined how local stakeholders exerted efforts to inform and impact language planning in response to social tensions emanating from a sense of language insecurity within the LMT framework. It presents the case of Shanghai dialect with an aim to reveal and interpret the interplay between top-down and bottom-up dimensions underlying the language planning process. The process of saving Shanghai dialect is unique as it illustrates a typical bottom-up case by demonstrating individual stakeholders’ agency struggling and striving for regional language rights, which is indicative of a democratic micro-level planning process in the formation of a macro language policy (Shohamy 2006). The findings may have relevance for language policy-making processes in other multilingual/multidialectal contexts with similar contextual constraints where linguistic diversity at the local contexts are experiencing and encountering shifts in language use. Besides, this study may also enrich the applicability of the LMT framework which reveals the interactive and dynamic process by unbundling individuals’ responses and influences on language planning.

The findings also reveal the social political duality of structure and agency that can be observed in two aspects. On the one hand, as Haarmann (1990) argued, ‘as practical language policy planning can only attempt to approach certain ideal objectives as closely as possible’ (p. 124), the power to push the national standard speech remains unchanged with institutionalization of planning activities. It is evident that the power of the grass roots is relatively meager and the organizational impact of measures is likely to be weak. Therefore, the plans or strategies proposed from the bottom up may not be fully implemented or achieved. However, agency exerted by individual stakeholders with expertise and influence is an important force that needs to be addressed in language planning discourse as in the case of Shanghai dialect in this study. On the other hand, the process of saving Shanghai dialect may invite the question of ‘structuration’ in language planning. One may wonder whether individuals make their language choices and perform actions through agency within the power structure of language policy. In other words, the structure of language policy in China may provide ‘rules and resources’ which give meaning to and shape the language situations local stakeholders would like to find themselves in (O’Byrne 2011). It would be ideal if the present overwhelming impact of structural forces upon the individual stakeholders’ agency could be negotiated in a more interactive and democratic dialog without predefined policy constraints. Nevertheless, language policy-making in a specific context is still an interactive process at the top, and at the bottom it takes societal realities into account with language researchers assuming the mediating role so as to appropriate realistic rules and substantial resources into the efforts to protect Shanghai dialect.