Semiotic Landscapes: Scaling Indonesian Multilingualism

This article presents a very preliminary description of a sample of photos of signage (e.g. posters, signs, billboards) drawn from around six hundred photos taken in Bandung in January 2019. Drawing upon scholarship on value and scale in general, and work on semiotic landscapes in particular, this paper seeks to extend earlier analysis of multilingual signage in Indonesia. I explore how an analysis of this signage can provide insights into multilingualism, inequality, and mobility in Indonesia, as well as how different social, political, and economic regimes effect the multilingual landscape.


INTRODUCTION
How can we characterize the shifting multilingual landscape of Indonesia? This paper starts to formulate an answer to this question via reference to images of street signage emplaced in Bandung's semiotic landscape. Among other things, research on semiotic landscapes has included the study of the emplacement of signage and the emplacement of texts and images on signage, the material make-up of signage, including the use of colors and images (Jaworski & Thurlow, 2010;Scollon & Scollon, 2003;Stroud & Mpendukana, 2009), who signage seeks to address, and what signage can tell us about the history of places, their populations, and inequality (Backhaus, 2005;Blommaert, 2013).
In synthesizing a wide range of sociolinguistic theory in general and work on semiotic landscapes in particular, I make four main points. The first is that political, economic, and social regimes need to be considered to understand the why and how of change in the constellation of linguistic features, colors, shapes, and so on that constitute signage. The second point comes from the first. Basically, to understand the meanings of signage in the semiotic landscape, we need to examine their history; in short, this requires a diachronic view of the emergence and decline of different regimes. Third, like social life in general, inequality can also be read from the semiotic landscape and such inequalities relate to different regimes. Fourth, attention to signage can tell us a lot about human mobility and processes of globalization. In my conclusion, I not only bring these four points together, but I do so through a reflection on the methodological strengths and weaknesses of this survey approach to multilingualism, while highlighting some potential methodological innovations to semiotic landscape studies that build on work that examines how semiotic features are valued via their circulation.

SEMIOTIC LANDSCAPING
This section briefly examines a relatively recent field, Linguistics Landscapes. In recent times, this field has increasingly been referred to as Semiotic Landscape studies as scholars have sought to provide more nuanced understanding of these landscapes. In particular, recent work has sought to understand connections between: 1) or private sponsor), value and authorization have received more direct attention as scholars have turned to reconceptualizing language. Within LA, language has been seen as an ideology that is a product of chains of discursive evaluations and of nation-building processes (Agha, 2007;Heller, Bell, Daveluy, McLaughlin, & Noel, 2015;Schieffelin, Woolard, & Kroskrity, 1998), while the other two fields (LE and SG) have reconceptualized competence and repertoires as fragmented and situation specific knowledge that can be strategically recontextualized, i.e. localized, to perform particular identities (Blommaert & Rampton, 2015;Garcia & Wei, 2014;Pennycook, 2010). Within studies of semiotic landscapes, language has thus increasingly been viewed as a mobile fragmented assemblage of signs rather than something that equates with labels such as language x or y (Blommaert, 2013;Jaworski & Thurlow, 2010;Pennycook, 2018).
Ideas of social class have been a constant feature of all three fields (LA, LE and SG) and have developed our understandings of different forms of literacy and its relationship with inequality (Blommaert, 2008(Blommaert, , 2010Heath, 1986;Rampton, 2006). Within the study of semiotic landscapes this has been conceptualized not just in terms of signage producers' ability to reproduce 'standard' language, but also their access to material resources in the production of signage (Blommaert, 2013;Jaworski & Thurlow, 2010;Stroud & Mpendukana, 2009). All of this has been analyzed through an analysis of the material make-up of signage, the use of colors and images, as well as the linguistic features of such signage. A focus on social class and inequality have also emerged as part of scholarship on centers and peripheries (Pietikäinen & Kelly-Holmes, 2013). The result of such thinking for linguistic landscapes studies was a move from studying signage that was found in large Western cities and urban centers, to smaller rural peripheries and to urban neighborhoods (Pietikäinen, 2014;Shohamy, 2017).
Related to the above has been a focus on the concepts of timespace, or "scale" and its relationship to context (Blommaert, 2010). While it has long been accepted that the re-use of linguistic and other signs helps to create a context with tastes of the past (Bakhtin, 1981;Bauman & Briggs, 1990;Silverstein & Urban, 1996), more recently we have seen this idea applied to scale (Carr & Lempert, 2016). These ideas have been further developed into concepts such as "scale-making" (Goebel & Manns, 2020), whereby the use of a particular sign can rescale a communicative event from a face-to-face event involving just two participants, to a face-to-face event that seeks to involve a historical community. Within studies of signage and the signs that constitute them; 2) signage and inequality, 3) signage and political, economic, cultural and social regimes; and 4) what all of this can tell us about society and change more generally. As we will see in last part of this section, this intellectual work has also lead to methodological innovations.
The genesis of the field of linguistic landscapes is often traced to a paper by Landry and Bourhis (1997), although no actual street signage was analyzed in their work. As a paper that was concerned with student attitudes to signage found in the landscape, their main aim was to theorize the relationship between signage and the vitality of particular languages. Their observations were in line with sociolinguistic work of the day that conceptualized language as immobile and linguistic features as tied to a particular group residing in a particular territory. For example they noted that linguistic landscapes were "…. a marker of the geographical territory occupied by distinctive language communities within multilingual states." (Landry & Bourhis, 1997: 24). They defined linguistic landscapes as: The language of public road signs, advertising billboards, street names, place names, commercial shop signs, and public signs on government buildings combines to form the linguistic landscape of a given territory, region, or urban agglomeration. The linguistic landscape of a territory can serve two basic functions: an informational function and a symbolic function. (Landry & Bourhis, 1997: 25) As the field developed -including the launch of a journal, "Linguistic Landscape" in 2015 -the "linguistic" in the compound "linguistic landscape" has been put in tension with the idea of "semiotic landscape" (Gorter & Cenoz, 2020;Jaworski & Thurlow, 2010;Pennycook, 2018). Developments in scholarship on semiotic landscapes have also been tied to developments within the broader field of sociolinguistics, especially work in linguistic anthropology (LA), linguistic ethnography (LE), and the sociolinguistics of globalization (SG). SG requires a quick side note because it has drawn on both LA and LE to focus on globalization; essentially the flow of people, goods, services, and ideas between nationstates or countries, the complex connections between all of these, and the effects of these processes on individuals and on populations (Appadurai, 1996;Blommaert, 2010;Stiglitz, 2006;Vertovec, 2017).
While Landry and Bourhis (1997) implicitly pointed to the idea of social value (i.e. whether minority languages were valued or not), and authorization (i.e. whether signs were authorized by a government semiotic landscapes, these changes have been translated through acknowledgments that signage is constituted from recontextualized signs, i.e. signs from elsewhere, with the act of recontextualization also localizing the meaning of such signs (Oconnor & Zentz, 2016;Stroud & Mpendukana, 2009).
More recently, scale has been linked to social activity in the idea of "chronotope" a (Bauman & Briggs, 1990;Blommaert, 2016;Lempert & Perrino, 2007;Rutherford, 2015). Chronotopes have been seen as constantly under construction and part of chains of activities occurring in particular times and spaces (Blommaert, 2020). To exemplify these ideas, let's consider the case of Indonesia. We know that as nationbuilding infrastructures in Indonesia became increasingly funded under a "development" focused government (New Order) in the late 1960s, the Indonesian government constructed languages as spoken by a territorially bounded group of people (Errington, 1998;Heryanto, 1995). This happened within various chronotopic chains. The New Order government simultaneously supported the recirculation of an imitation of this language ideology in different chronotopic chains, such as in schools, universities, bureaucracies, and the media (Dardjowidjojo, 1998;Goebel, 2015). In terms of scale-making, classroom discourses helped to fashion the nation as Indonesian speaking, and the regions as ethnic language speaking.
We can contrast the scale of development era Indonesia, with another scale that emerged in the early 1990s. At this scale, regional languages became commodified as media professionals started to understand the economic value of these languages for reaching niche television audiences (Kitley, 2000;Loven, 2008;Sen & Hill, 2000). The multiple chronotopic chains that were part of the commodification of regional languages simultaneously occurred with another set of such chains; the then centralized government's long-term efforts to develop and promote Indonesian as the language of the state. Thus, we had multiple simultaneous chronotopes under construction not just at one scale but across different scales, with each scale influenced by a particular regime, whether political or economic.
With changes in how we think about context more generally, the study of semiotic landscapes also moved to seeing space as being created and a result of economic, political and social regimes; hence the verbalization of "landscape" in the compound "semiotic landscaping" (Gorter, 2013;Jaworski & Thurlow, 2010;Pavlenko & Mullen, 2015). The object of study also moved from just public street and shop signage, to include private signage as well as graffiti, and signage that was mobile, such as inscriptions on the side of vehicles and litter (Blommaert, 2013;Gorter, 2013;Malinowski, 2018;Pennycook, 2018). At the same time, and as a way of accommodating the above conceptual changes and changes in the object of investigation, methodologies for investigating linguistic landscapes moved from a primarily quantitative approach where what mattered was how many languages could be found on a sign, to qualitative approaches (Moriarty, 2014;Shohamy, 2017). These qualitative approaches typically were influenced by ethnographic epistemologies that focused on grounded theory and local theories of meaning.
Ethnographic methodologies used includes the historicization of signs by relating them to the social, economic, and political conditions of their production and display, their connection via imitation to other signage and signs (Blommaert, 2013;Stroud & Mpendukana, 2009;Taylor-Leech, 2020), interviews with people about how they interpret signage (Curtin, 2015;Ferguson & Sidorova, 2018). Some of the scholarship looked at the interpretation or reception of signs on signage by examining their imitation on other signage (Stroud & Mpendukana, 2009), while some examined government documents or talked with those who produced signage to get insights into the intent and imagined audience of such signage (Backhaus, 2005;Edmond, 2017). Some scholars, such as (Spolsky, 2020: 6) consider this latter perspective an important, but rarely studied aspect of semiotic landscapes.
In summing up this section, we can highlight some of the insights that we can gain by studying semiotic landscapes. While the limitations of quick quantitative surveys is clear (e.g. gathering hundreds of images of signage and counting the languages on them), when such data is considered in relation to other data sets, the reconceptualizations noted above, and an ethnographic epistemology, such studies can provide a range of insights. For example, we can gain insights into who are imagined to be contemporary consumers or readers of signage, and what this can tell us about the changing make-up of a population. In addition, such surveys can provide insights into political, economic, social, and cultural actors who are currently active, and insights into changes in the value of particular labels (language and otherwise). In the next section, I will focus on these five areas as I analyze photos of signage that I took in Bandung in 2019.

RESEARCHING SEMIOTIC LANDSCAPES IN BANDUNG
The city of Bandung is the capital of the province of West Java. Bandung is often characterized as the center of exemplary Sundanese (Moriyama, 2005;Uhlenbeck, 1964). Since the beginning of political and fiscal decentralization and the democratization processes in 2001, Sundanese has increasingly become valued with successive governors inviting civil servants to wear emblems of Sundanese-ness and to use Sundanese on particular work days (Goebel, 2017;Moriyama, 2012). Like other major provincial Indonesian cities, it has a long history of attracting migrants from around Indonesia (Bruner, 1974), as well as being a city that has long experienced the effects of and or lead globalization processes. For example, Indonesia's first president, Soekarno, hosted the non-aligned nations movement in the 1950s (Ricklefs, 2001). As we will see in sections 4 and 5, signage in Bandung points to the continued effects of globalization on the semiotic landscape. This global scale can be contrasted with another, that of the local government and private commercial scales. As we will see in section 6, both have produced informational signage in both Indonesian and Sundanese, with some mixed language signage as well.
The images that I examine in the following sections are drawn from a broader study that I carried out in January 2019. In this broader study, I was interested in the language of political signage that was being inserted into the landscape in the lead-up to a major election. While this broader study gathered over six thousand images from the regencies of Indramayu, Majalengka, Kuningan, Cirebon, and the cities of Semarang, Serang, Cirebon, and Bandung, in this chapter I focus on just a handful of the six hundred or so images I gathered in Bandung. These images were taken over a two-and-a-half day stay. Like many of the studies of semiotic landscapes noted above, I took these photos while walking the main streets of Bandung's city center as well as some of the nearby neighborhoods. I did this with my Indonesian spouse, who had little to say about the signage that I was interested in photographing.
My initial interpretations of signage was complimented with insights about intention which I gained through my access of company, government, and mass media websites that commented on this signage. In essence, providing insights into both intent (in the case of government and company websites) and reception (in the case of comments about signs in the mass media). These interpretations were further supplemented by reading signage in relation to work on regimes of language in Indonesia (Goebel, 2010(Goebel, , 2015(Goebel, , 2020Goebel, Cole, & Manns, 2020), especially the political, social, and economic circumstances surrounding the emplacement of signage.

AUTHORIZING AND VALUING SIGNS OF CHINESE-NESS
Indonesia has had a history of devaluing and erasing some languages from the semiotic landscape. The erasure of Chinese characters, other emblems of Chinese-ness, and Indonesian's of Chinese ancestry themselves in the 1960s is, of course, the most well-know (Coppel, 1983). Since the Gus Dur presidency of the early 2000s, Indonesians of Chinese heritage and signs of Chinese-ness have increasingly been valued (Hoon, 2012), including the use of Chinese characters in the semiotic landscape. The analysis below suggests that in public commercial and religious domains this positive valuation seems to be continuing, although in other social domains this is not the case (Aspinall & Meitzner, 2019;Hatherell & Welsh, 2017).
Of the six hundred or so photos taken in the central district of Bandung, little signage contained Chinese characters. Those that did have them were either commercial or religious, including two global banks, two mosques, and two food and drink establishments. To varying degrees all were examples of three types of globalization, global flows of capital, global flows of people, and the localization of global signs, including texts and architecture. In the case of figure 1, the China Construction Bank Indonesia (CBC), it was the product of the 2016 merger of two Indonesian financial groups that now is controlled by the China Construction Bank (CBC) who has a controlling share in the company (CCB Indonesia, 2020). This signage used the traditional symbol for money (following the CCB), and then simplified Chinese characters that are translated in the smaller English text located below these characters 1 . In the case of figure 2, the Industrial and Commercial Bank of China Indonesia, its website notes that it is the product of the The websites of each of these banks provides some insights into the intent of these banks and potentially its signage. For example, CBC's website states that it seeks to work with all levels of Indonesian society, while facilitating increased commerce with China. Similarly, according to its website, ICBC seeks customers from all walks of Indonesian life and to encourage joint ventures between Indonesians and Chinese. Despite the stated intent of each of these banks on their respective Indonesian language websites -that is, to work with Indonesians from all walks of life -the signage in figures 1 and 2 do not evidence such an intent given the lack of Indonesian. In both cases, the signage excludes anyone who cannot read the Chinese characters or who recognizes the corporate emblems. In the case of CCB, the signage also excludes anyone who cannot read English. Note too that the CCB signage separates Chinese characters and English by placing English underneath the Chinese characters. We can refer to this layout of texts as a type of codeswitching but not codemixing. This contrasts with the languages on the other signage that we will consider in section 5, and through these comparisons we also get to see a picture of codemixing and the use of standardlike foreign languages as addressing an elite, while the texts that are typically examples of codemixing seem to address people from the less economically able parts of Indonesia.
In considering the imagined audience of the signage in figures 1 and 2, we can say that while Indonesia has a minority population who are of Chinese heritage, it is less clear whether they are able to read these characters. The use of the English translation underneath the Chinese characters in figure 1 also points to another audience; those able to read English. Both audiences would thus appear to be limited. Even so, the presence of such localized global banks and the Chinese characters that label them index Indonesia's political climate. In particular, this signage tell us a story of the authorization of global flows of capital and language from non-English speaking countries. Similar flows of languages, and people can be evidenced if we examine places of worship, such as the Al-Imtizaj mosque built in 2010 (figures 3-5).
Figures 3-5 represent an example of what Pennycook (2018) describes as a semiotic assemblage. Like another mosque in downtown Bandung (Lautze-2), in addition to multiple language scripts -including Indonesian, romanized Qur'anic Arabic, Chinese characters 2 , and old Sundanese script (aksara Sunda) in figures 3-4, Qur'anic Arabic and Indonesian (figure 5) -the architecture, colors, and materials used inside and outside this mosque are emblematic of the mixing of Chinese religious architecture in religious sites known as klenteng with Islamic architecture (Mulia, 2017;Pradewo, 2018;Wibowo, 2018). The mosque was designed and built to not only tell the story of Indonesians of Chinese heritage who converted to Islam, but to also make an  explicit statement about unity in diversity (Indonesia's national motto), especially the intermingling of those of different ethnic backgrounds. The building of this mosque was made possible through an alliance between Indonesians of Chinese heritage, who were Muslims, and the local government in the late 2000s. In a real sense, the mosque was authorized by the local government in a more complex manner than the building of the typical mosque one finds in Indonesian neighborhoods. Each of these emblems index different eras (i.e. scales) and features of globalization including the chains of activities and people involved (i.e. chronotopic chains). The Sundanese script is based on Sanskrit (figures 1 & 2 bottom right script located beside the Chinese characters on the left below the text "Masjid Al-Imtizaj). This script indexes the mobility of people and other religious traditions from a distant past and far off land. Similarly, the Chinese characters and the architecture, colors, and building materials all point to a more recent past of mobile people with other religious practices, among other things. In this case, those who emigrated to the Indonesian archipelago a few hundred years ago. The Qur'anic Arabic script in figure 5 does similar indexical work pointing to a distant past, while the Indonesian informational texts about how to conduct oneself within the mosque point to a much more recent era, that of Indonesian nation-building. Note too, that in contrast to figure 1, the Chinese characters and the halal sign are underneath and smaller than the Indonesian text. This suggests that the Indonesian text has more importance than the Chinese script.
While it is not hard to imagine who these signs seek to address (Muslim Indonesians of Chinese and other ethnicity), as with the bank signage, those who can read the Chinese characters will be few. There are even less who can read the old Sundanese script, aksara Sunda, despite its increasing presence in the landscape since 2015, including the text underneath some Indonesian language street name signs in Bandung (Nurwansah, 2015). In contrast to the Al-Imtizaj mosque, which was also built to service the nearby market, the two commercial signs in figures 6 and 7 are respectively located in an area close to the cultural center in the Braga district of Bandung (figure 6) 3 , and nearby to a number of major international hotels and banks ( figure 7).
Pempek Ny. Kampto is a fishcake restaurant franchise that has its origin in the cooking of a Palembang Chinese fisherman (figure 6). The shopfront in this has Indonesian, such as Pempek Ny. Kampto "Mrs. Kampt's Fishcakes", Majelis Ulama Indonesia "Islamic Council of Indonesia" as part of the green circular symbol that has halal "allowed [to be eaten]" inscribed in Qur'anic Arabic inside this circle. On other co-located signs there is also Indonesian sejak "since" and Indonesian for names of the food on the menu billboard. There is also red Chinese characters underneath the shopfront business name. While the Indonesian indexes an audience and customers who can read Indonesian, the Islamic and Chinese features of this signage seem to navigate the tensions between indexing authentic Chinese food (often associated with non-halal food) and food that is certified as halal (i.e. allowed to be eaten by Muslims) by the Islamic Council of Indonesia. Note too, that in contrast to figures 1 and 4 the location and size of both the Chinese characters and the Qur'anic Arabic suggest they are potentially less important than the Indonesian.
In the case of figure 7, the top bit of text is Chinese and relates to "being drunk" and "Panda bears" respectively, while the below English text is a gloss of sorts with the literal translation of Panda, with Tipsy ("on the way to getting drunk") preceding it. It  is interesting because these texts limit the audience to those who can understand each set of texts in figure  7. The wider landscape along this street which has a number of hotels and banks, some of which also have signage that is exclusively English language points to an internationally mobile audience and expatriates from English and Chinese language areas, and potentially Indonesians who have the economic resources to invest in English, Chinese or other foreign language education. As an establishment that sells alcohol and thus would not attract pious Muslim patrons, Tipsy Panda contrasts with Pempek Ny. Kampto (figure 6). Note too that while the simplified Chinese characters are located at the top of the sign, the English is in the center and much larger, suggest the importance of the English text.
Thus far the analysis has focused on signs that seek to address those who can be labelled as wealthy (figures 1-2), Muslims of Chinese heritage (figures 3-5), and Indonesians seeking authentically Chinese eating experiences (figure 6), and tourists and expats (figure 7). In terms of mobility, most of this audience are either internationally mobile (figures 1-2), immobile locals (e.g. figures 3-6) or foreigners (figure 7). By taking a look at the materials that this signage is made from, we can also get insights into the economic resources of the sponsors of these signs. In the case of figures 1-2, it is clear that as banks they have significant resources, and the embedding of the signage into the architecture of the building also points to this. This is also the case for the mosque and the fishcake restaurant, while the signage and the architecture of the Tipsy Panda point to much more modest resources. Note too, that all signs had clear distinctions between different languages, typically achieved via their emplacement within a particular street sign. In the following section, I will focus more squarely on what signage can tell us about mobility and the economic resources of the imagined audience of this signage.

SIGNS OF PHYSICAL AND ECONOMIC MOBILITY
Indonesian universities have long invested in the teaching of foreign languages such as English (Zentz, 2017), although investment in Korean, Japanese, and Chinese, is much more recent. In some cases, this growth is related to Indonesian academics' experience doing post-graduate work in countries such as Korea, Japan, Taiwan, and Hong Kong. This has also co-occurred with a massive increase in Indonesians working abroad as domestic helpers, agedcare providers, and factory workers in places such as Taiwan, Hong Kong, and Korea, and to a lesser extent Japan and Canada (Lorente, 2018). Some of the images we will examine in this section relate to advertisements for the study of foreign languages, study abroad, and translation and interpreting services (figures 8 and 9). The audience that this signage appears to seek are those who aspire to be internationally mobile or those who are internationally mobile. This audience, thus might include those already studying these languages at university or those whose jobs require contact with people from Northeast Asia. Analysis of these images also tells us that the audiences being sought have different economic means. Thus, attention to this type of signage can not only tell us something about multilingualism in Indonesia, but also about the mobility of the audience, their economic means, and the relationship of this to language.
In the case of figure 8, Northern Lights Education Centre (NLEC), the use of the Japanese characters on the right (reading "Let's learn Japanese and go to Japan"), and the almost exclusive use of English suggests the audience to be someone who already has a good understanding of either Japanese or English. Such forms of competence can only be obtained with significant investment in quality education and/or study abroad. Contrast this with the language used on figures 9 and 10, where the speech bubble "Hey you join us" is one of the only fragments of English on the signage. The remaining and very significant amount of text is in Indonesian, or in a mix of Indonesian with some English lexicon (e.g. yang fun "that which is fun"). This mixing points to a localization or syncretization of English forms. This formatting choice points to the primary target being those who have little access to English resources, and potentially those who imagine becoming mobile, but who actually have little economic means to do so.  (Goebel, 2010), the audience of figure 8 would be from kelas atas "upper class" or kelas menengah ke atas "middle to upper class", while the audience of figures 9-10 would be referred to as kelas-menengah ke bawah "middle to lower class". We can see such contrasts in other types of signage too. For example, figure 11 addresses an upper class audience, while figure 12 addresses a middle to upper class audience.
In the case of figure 11, we have a banner that is exclusively in English with just Indonesian street and place names. The use of English points to an audience of a privileged few in Indonesia and potentially to expatriates. This is reinforced via the 1,500 million rupiah starting price, which is at least five times the amount the majority of Indonesians could pay for a home. Access to high levels of economic resources is also a precondition for air travel in Indonesia. We get insights into the imagined repertoires of this segment of Indonesian society via advertisements that link economic ability (credit cards and thus the banked population) and mobility (air travel), as in figure 12.
The credit card advertisement which links getting a credit card with discounted Garuda Airlines flights ( figure  12) is sponsored by the Indonesia-based chapter of United Overseas Bank. While the global website for this bank has no information on its base of operations, it does note that it aspires to be leading bank in Asia. Of importance for our analysis, however, is the fragments of English "cashback" in an otherwise Indonesia sentence, the exclusive use of English "The airline of Indonesia" underneath the Garuda label and symbol, and the "apply now". Unlike figures 9-10 that were examples of a localized English, here all instances of its use are actually highlighted using italics. This and the fact that the signage advertises a credit card and air travel helps to focus on a specific audience; those who have the resources to be mobile and those who have more than grassroots literacy in English.
The imagined Indonesian audience for figures 11 and 12 can be contrasted with those for figures 13 and 14 below. Figure 13 is another sign advertising homes for sale, while figure 14 is advertising the local department of transports bus services. Both signs are almost exclusively in Indonesian. Figure 13 advertises Rumah baru minimalis "new modest homes" located in a complex designed for low to middle income workers of the state electricity company (PLN). Rather than advertising a luxury two story home, this modest sign with no images advertises very modest size homes starting at just 72 m2 on a 95m2 block (type   figure 11, and is only present at the top right hand corner as part of the brand of the real estate company Balindo Realty. This brand mixes the Indonesia name "Balindo" with the English "Realty, while underneath this is its territorial remit, lingkar selatan "southern area" and underneath this is the English "we grow as you grow". With the rest of the text in Indonesian, it is not hard to guess that the imagined audience are those who are literate in Indonesian, and who most likely are low ranking civil servants and others who receive a modest income (Goebel, 2010: chapter 3). Note that being able to read English has little to do with being able to understand the sign, and thus these fragments appear to be more about indexing some form of prestige or modernity of the real estate company. Note too, that in contrast to figure 11, the material that this sign is made from and the lack of graphics suggests that the real estate company, like those it seeks to house, has modest resources. Figure 14 also tells a story that links Indonesian with those with modest means. In this case, we have a banner sponsored by the traffic control wing of the local government of Bandung, which reads Dishub kota Bandung. Beside this logo, which is located on the right corner of figure 14, there is also another logo which is the symbol of the Bandung local government. Underneath it is text that originates in what local bureaucrats refer to as Kawi "old Javanese" (Pemerintak kota Bandung, 2020). The text says gemah ripah wibawa mukti "Fertile soil [means] a wealthy people". The symbol itself was first adopted in 1953, and its colors and shapes index a host of meanings too (Pemerintak kota Bandung, 2020), although the emplacement of this symbol on the signage and its size suggest that it is a minor part of the overall message, acting mainly as an emblem that indexes the sign's sponsor and their authority.
The imagined audience of the other text on this banner, Bandung Baik, bersama transportasi publik "[create a] Good Bandung with public transport" are literate Indonesians. Although the text implies an effort to hail environmentally conscious passengers, more generally buses are for those who do not typically own cars or motor bikes; that is, those who might be able to afford the smallest of homes such as the type 72/95 in figure 13. While the audience are mobile Indonesians, this is local mobility only, with the only other sign of mobility being the old Javanese motto that points back to travelers bringing Sanskrit texts from India in the distant past.
In contrast to the permanent signage discussed in section 4, the emplacement and the quality of the material used in the signage discussed in this section all suggest varying degrees of ephemerality. While inequality in Indonesia is nothing new, the signage that we have examined shows how the use of different languages and their formatting as mixed or separate give us insights into how these languages can be mapped onto the economic circumstances of their respective imagined audiences as well as the different imaginings about audience mobility. In the last section, I will turn to the use of a regional language, Sundanese. Rather than being further down the ladder in economic terms, as we find with many indigenous languages around the world, we will see how Sundanese has increasing cultural value.

REVALUING SUNDANESE
In this section we will see how fragments of a regional language, Sundanese, is valued at a number of scales. At the provincial level of West Java most government sponsored signage has the provincial symbol which, like the one for Bandung city ( figure 14), not only has a specific text but different colors and shapes to signify a host of meanings. In figure 15 we see this symbol on the signage for the provincial office of the ocean and fisheries department (middle left). Together with the other shapes and colors, the text, Gemah Ripah Repeh Rapih, roughly glossed means "West Java is a rich area where the land is fertile and inhabited by wealth people who are spiritually calm, peaceful and who work together for the common good" (Pemerintah propinsi Jawa Barat, 2020).
If the explanation provided by the Bandung city government noted earlier (figure 14) is taken into account, especially in relation to the first two words of this text, then the provenance and classification of this text could be old Javanese. However, other sources, such as the Indonesian Wikipedia labels the whole text as Sundanese (Wikipedia, 2020). To the right of this text is Indonesian. Less ambiguous in provenance and category is the language found on local government signage, as in figure 16. The text in the signage above the bins reads nyampah teh di dieu "put your rubbish here". The signage also has the Bandung city council emblem on the right side, with BAGUM "general department" underneath, which indicates who sponsors this sign. Note too, that the bins below are labelled in Indonesian. While the words nyampah "to rubbish" and di "in" can also be classified as Indonesian and hence are syncretic in nature, this signage attracted commentary from a local blogger at the end of January 2018 along with on Facebook, and Picburn, and Instagram sites. In the blog this language was described as Sundanese (Krisnomo, 2018).
The above type of directional signage with Sundanese text could be found in other places serving other functions, such as "don't litter here" signage beside waterways in the central city parklands. At another scale, this time three administrative levels down from the city government to the level of RW "neighborhood" (kota "City" > kecematan > "shire" > kelurahan "suburb" > rukun warga (RW) "neighborhood) we find territorial markers such as figure 17.
The text in figure 17 is almost exclusively in Sundanese. The few exceptions are the name of the neighborhood at the top (notably the largest text), the name of the Bandung city government and the program that sponsored this monument Program Inovasi Pembangunan dan Pemberdayaan Kewilayahan (PIPPK) "Program of Regional Innovative Development and Energization", and the words maju "to go forward", waktu "time", masih "in process/still", diri "self", harga diri "self-worth". According to the head of the public relations section of the city council, this type of infrastructure work was done as part of a much larger beautification project conducted in Bandung in 2018 (Prasatya, 2020). While the imagined audience is limited to those who can read Sundanese and those who have an understanding of the seven proverbs being cited, nevertheless the use of Sundanese and the marking of a specific territory (RW07) discursively projects the idea that those in this territory understand Sundanese. In doing so, it represents an instance of scale-making.
Signage with fragments of Sundanese could also be found in private advertisements for services, as in the Grab advertisement in figure 18. Grab is the Indonesian answer to Uber and this image is part of a larger mural advertising all the different Grab services.
While it is quite common to find fragments of regional languages in television advertising (Goebel, 2015), it is less common to find them in the semiotic landscape of Bandung, something that applies to other Indonesia towns (Rohmah & Zuliati, 2018). In figure 18 we have two speech bubbles that use two terms of address, kang "older brother" and teh "older sister" in an otherwise Indonesian text. In terms of imagined audience, the use of Indonesian addresses anyone literate in Indonesian. The use of Sundanese kin terms would not create much of a problem for locals, because this in one of the first things that mobile Indonesians do is learn to use local kin terms as a way of performing local-ness and belonging (Ewing, 2020;Goebel, 2010). The type of mobility that is advertised, being doubled on the back of a motorcycle, addresses an audience who do not have their own car, or chauffeured car, and/or live in an area that four wheel vehicle access is difficult. In short, it addresses those with modest economic resources.
In linking the semiotic landscapes discussed thus far to the historical circumstances surrounding their production, we can say that in the five years prior to the appearance of the government sponsored texts, civil service bosses in the Bandung city council invited their charges to wear emblems of Sundanese-ness and use Sundanese in meetings on specific days (Goebel, 2017;Moriyama, 2012). It would be interesting to engage with the designers of the slogans on the bins to understand whether and to what extent the usage of Sundanese on them was a reflex of this ongoing revaluation of Sundanese. The neighborhood signs appears to be a special case because of how its emplacement, and linkage with a specific territory (RW07) help to presuppose an audience who can read Sundanese. In doing so, this instance of scale-making also erases the other inhabitants of this neighborhood were some will almost surely be migrants from other areas of Indonesia, as is the case in Indonesia's other large provincial capitals (Goebel, 2010;Grijns, 1983)

CONCLUSION
In this paper I have drawn upon the broad field of sociolinguistics, but especially work on semiotic landscapes to provide a grounded theory of what public signage can tell us about multilingualism in contemporary Indonesia across time and space and how this signage relates to human mobility, social class, and different political regimes. What we found was that Chinese characters along with English had a strong presence in the landscape, although each example of signage typically had a unique audience that could be distinguished along economic grounds, especially access to resources. Similarly, we noted that the economic resources of the sponsors of signage could also be ascertained from the materials that they were made from along with their emplacement. The emplacement of text within signage generally followed what other research tells us about emplacement in multilingual signage, namely that the most important text is either largest, placed above subordinate text, or central, while signs of authorization are more peripheral in signage. While Indonesian was by far the most common language used in signage, there were regular instances of the use of bits of Sundanese, sometimes used to address particular audiences and at other times also used in scale-making exercises that imagined a community of Sundanese speakers. As pointed out in the critique of semiotic landscape research, short data gathering fieldtrips like the one reported here certainly have their limitations, especially if we are concerned with the multiple processes that help shape signs and the intent of such signage (Spolsky, 2020: 6). Although as I also pointed out, some of this type of data can be gleaned from internet sources, as can insights into how Indonesians actually perceive these signs (i.e. the reception component of this process of localization). Despite such limitations, however, such short survey type work can also highlight some interesting areas for more in-depth longer term fieldwork. The case of Sundanese usage on Bandung city council bins and other signage being one interesting case.
Focusing on the linguistic forms found in this signage can also be deceiving however. As Shohamy (2017) has noted, it is important to ask what the signage doesn't tell us about multilingualism. Indeed, the use of Sundanese to address a whole neighborhood seems to erase the fact that as a major city, Bandung has attracted many migrants from other parts of Indonesia. This would mean that as in the past (Bruner, 1974), many of Bandung's inhabitants would speak other regional languages, and there is some evidence of this if we consider signage for restaurants, sidewalk eateries, and mobile food sellers who have place names in their establishments'/business' title indexing a place of origin outside of Bandung. There would also be those from the West-Java countryside who would speak other varieties of Sundanese (Uhlenbeck, 1964).

1) According to my Chinese language consultant the
English is essentially character to word translation. She also pointed to the significance of the large characters as part of the brand of this bank and of ICBC because other branches used exactly the same simplified Chinese characters. 2) According to my Chinese language consultant, these characters can be seen as either traditional and simplified. Their literal meaning character for characters are "harmony/co-existence/integration" for the first two, then the third is "Islam" and the fifth "temple". 3) According to my consultant the script is simplified and there are two interpretation of the first two characters. The first is "sweet" and "a lot" respectively, while the second is that it phonetically represents "Kamto". The latter two characters are "fish" and "pancake".