ISTANBUL PAST AND PRESENT: A LINGUISTIC LANDSCAPE PERSPECTIVE

This paper reports on a linguistic landscape study undertaken in Istanbul, Turkey from August 2016 to March 2017. A linguistics landscape study takes account of all the visible language typically found on signs in a bounded area such as a city, a commercial district or neighborhood. Combining historical, demographic and other data, a linguistic landscape analysis generates insights into the past and present-day interrelationships between languages in the context of their communities: in short, a sociolinguistic profile. Several such studies have been carried out in major urban centers such as Tokyo (Backhaus 2007) and Antwerp (Blommaert 2013). This is the first ever linguistic landscape study of Istanbul. This paper has two parts. Following a brief introduction to linguistic landscape studies, Part I reports on a survey of ten Istanbul mahalle (districts). The results of this quantitative analysis reveal, among other things, the overall dominance of the Turkish language throughout the greater share of Istanbul’s landscapes (69% monolingual Turkish signage across the landscapes surveyed); the pervasiveness of English (comprising 51% monolingual English of the total signage in one district); and the emerging visibility and importance of Arabic (12% monolingual Arabic in one district). Part II of the study examines Istanbul’s landscape as a whole from a more nuanced, qualitative perspective. This analysis is thematic in nature and will focus on the dynamic and volatile aspects of Istanbul’s linguistic landscapes as they transform and adapt to political and social upheavals and the changing realities in present-day Turkey. The results, on the one hand, show how recent commercial, social and political trends shape and configure distinct sociolinguistic regimes in each of Istanbul’s districts, and, on the other hand, reveal how the linguistic ‘deposits’ force a historicization of the linguistic landscape in the analysis. Among other things, these results demonstrate that linguistic landscapes cannot be ‘read’ or understood in any meaningful sense without, at the same time, bringing to bear on the interpretation past and present social, political, demographic events and trends.


INTRODUCTION
In the city-much like the air we breath, so too we are surrounded by visible language. Posters, billboards, huge electronic screens beaming adverts, an entreaty to locate a 'Lost Cat', a piece of paper taped to the shop door announcing 'Back in five minutes!', spray-painted expletives on store fronts and much more clutter the visual field: it would seem that cityscapes exist for the sole purpose of showcasing language objects. And rarely, if ever, does this clutter of language give us cause or pause for reflection. Like the air, it just is. But as Louis-Jean Calvet pointed out years ago in his comparison of the environnment graphique of Dakar and Paris (Calvet 1993), each sign has a story to tell. Not so much in what the sign 'says,' Calvet would be quick to point out, but, for starters, the story about the language or languages on the sign (Whose languages? Which language comes first?), the placement of the sign (Who owns this space? Why here and not somewhere else?), and the ownership of the sign (Who put it there? On whose authority?). It was the call to tell this side of the story, so to speak, that launched investigations into linguistic landscapes.
As Calvet noted, though signs are information bearing, they also include social meanings that point to or indicate relationships among languages and speakers in the wider community. Landry and Bourhis, who coined the expression "linguistic landscape," have said as much: "The predominance of one language on public signs relative to other languages can reflect the relative power and status of competing language groups" (Landry and Bourhis 1997:26). And because, as Jan Blommaert and others have put it, space is always "someone's space, not empty space", signs are placed in "social space" (Scollon and Scollon 2003;Blommaert 2013). From these perspectives, signs are implicated in matters of power, status, competition and inequality One purpose of these studies is to develop a thumb-nail sociolinguistic profile of a neighborhood or a district which uncovers power hierarchies and the sociolinguistic regimes (Blommaert and Maly 2014) that underpin them-that is, the assumptions, expectations, and behaviors that people have about the languages they may use (at home, for example) or must use (at the market place or court house, for example, in order to secure services). Ultimately a linguistic landscape study aims to explain the underlying dynamics and processes of change in language communities and their linguistic landscapes, using historical, demographic and other background information that contextualizes the unique conditions and circumstances of space over time.
That said, what better human social environment in which to test the explanatory power of analyses of linguistic landscapes than that of a globalized urban metropolis? The great metropolises today are cosmopolitan centers that combine the panoply of Silk Road communication traditions and strategies (lingua francas, improvised codes, language mixing) with emerging novel forms of interaction and discourse (spawned by present day mobility and digitalized information technologies) which have mutated social structures and challenged the concept of the 'community'-in times past, a more-orless bounded, stable, social formation redolent of predictability and order. To date, employing a wide range of methodological approaches, investigators have examined the linguistic landscapes of cities such as Tokyo (Backhaus 2007), Taipei (Curtain 2009), Bangkok (Huebner 2006) and Antwerp (Blommaert and Malay 2014). This paper will report on the first-ever study of the present-day linguistic landscapes of Istanbul.
For more than two thousand years, Istanbul has been one of the planet's great urban centers. Its unique bridging position between Europe and Asia made Istanbul's narrow Bosphorus strait the foremost cross-roads for East-West trade caravans, invading armies, and human migrations. Formerly the capital of the Byzantine Empire, Constantinople was captured by the Ottomans in 1453 and remained the capital of the Ottoman Empire until the founding of the Turkish Republic in 1923. Throughout the early years of the republic, in efforts to build an ethnically homogeneous nation state out of the cosmopolitan empire, not only did the government undertake a radical language reform project that resulted in the adoption of Latin alphabet and a large-scale replacement of vocabulary of Arabic and Persian origin for Turkish equivalents (see Lewis 1999), the government also undertook a program of Turkification for which large-scale demographic schemes were enacted that included deportations, depopulations, and repopulations. If not the political capital of the Turkish republic (which had been moved to Ankara), Istanbul remained the economic and cultural center of Turkey and was itself greatly transformed by these social and political upheavals of the early 1900s: consider that Istanbul in 1900 was 56% non-Muslim; by the late 1960s it was 99% Muslim (Mills 2010, King 2014. A hugely important city today of some 20 million residents, it is natural that we should want to examine the language situation of Istanbul and take stock of its rich linguistic heritage. As above, signs have a story to tell and therefore a history to document. Such was Pavlenko's message in her study on the linguistic landscapes of Kiev, Ukraine for which she took a "diachronic" perspective (Pavlenko 2010). Moreover the obligation of linguistic landscape researchers to consider in their analyses the historical dimension points to Jan Blommaert's criticism of much of the linguistic landscape work to date (Blommaert 2015). In the case of Istanbul, much of the landscape reaches deep into history (after all, it is the city that answers to three major historical eras and is known by names familiar to middle-school students everywhere: Byzantium, Constantinople, Istanbul); in order to understand the presence and relevance of these 'signs,' their place and function in the landscape, we must appeal to past events, trends, and ideologies that account for the transformations that we uncover. This paper will present the preliminary results of a survey of ten Istanbul linguistic landscapes. The study includes both a quantitative and a qualitative dimension, and a full exposition of the findings will await the full paper.
There are three research questions for this study: 1. How were the linguistic landscapes of Istanbul constituted during the closing decades of the Ottoman Empire? 2. What are the linguistic landscapes of present-day Istanbul? 3. What processes have shaped and are shaping Istanbul's linguistic landscapes?

PROCEDURES
In order to address research question #1 (and also to obtain a broader picture of the language situation of Byzantium/Constantinople/Istanbul across the centuries), I relied on research studies and historical sources, photographs, postcards, graffiti, fresco and mosaic inscriptions, and additionally, what I have called "alternative linguistic landscapes" (Wendel 2017, under review): calendars, postage stamps, grave stones, and visa stamps. For research question #3, I rely on historical accounts of Istanbul and Turkey, among other source materials, to explain the differences found between Istanbul's pre-republic and post-republic linguistic landscapes.
For the present day linguistic landscapes (research questions #2), data for this study were collected by the author during a sabbatical leave taken in Istanbul from August 2016 to March 2017. The neighborhoods included in this survey focused particularly on the historical quarters of the city, Fatih and Beyoğlu: four formerly Jewish, Greek, Armenian, and/or Latin communities (Fener, Kumkapı, and Harbiye on the European side; Kuzguncuk on the Asian side); two former principally European districts on or near the many-storied Grand Rue de Pera (İstiklal and Galata); three districts in Fatih that have been known as principally Muslim communities since the 1453 conquest (Haseki, Aksaray, and Sultanahmet); and finally, one outlaying district some five kilometers from the historic city center, Bağcilar. Whereas Bağcilar lies in a more recently settled region of Istanbul, Sultanahmet (with its Hagia Sophia church, Hippodrome of Constantinople, Basilica Cistern, Blue Mosque and Topkapi Palace) is the oldest district, the very heart of ancient Constantinople, and is today a major tourist destination.
I documented, taking photographs and notes (and whenever possible, interviewing shop-keepers and residents), all visible language along the streets surveyed including all signs (e.g., street names, tourist explanation panels, banners overhanging streets; signs in shop windows, on street-sellers carts, on the pavements), graffiti, and notices or handbills pasted or stapled on telephone posts, walls, or buildings. I did not include signs above the 2 nd level on buildings-the exceptions were huge usually political banners draped on buildings and very large commercial billboards. I lived in two of the research sites (Kuzguncuk and İstiklal) for several months; the other sites were visited multiple times before the formal survey was conducted. A total of 2,658 signs were documented for this study, an average of 266 signs for each site.
A data-driven language-based coding scheme was developed that most usefully captured generalizations and trends concerning the language on the signs. As part of the main analysis, each sign was put into one of four categories for Language: monolingual Turkish, bilingual Turkish and English, monolingual English, and Other. The Turkish and English bilingual group included signs on which both languages appear-'bilingual' is not intended to suggest any particular relationship between the two languages on the sign, i.e., the English on the sign may or may not be a translation of the Turkish. The Other group comprises all additional signs and includes other languages, monolingual and otherwise, and all language combinations. This means, however, that Turkish and/or English may also be found on signs within this category, appearing, for example, with Greek or Arabic. Signs were also categorized according to Agency, i.e., who or what entity placed them: top-down or bottom-up. A sign was top-down if it was placed by the municipal or national government, or a religious authority. A sign was classified as bottom-up if it was placed by a corporation, business, or an individual.
To summarize, in addition to taking photographs of the signs, I noted the following information.
• Language: Turkish, Turkish+English, English, or Other • Agency: whether a sign was top-down or bottom-up • The number of languages, the kinds of languages, and their order of appearance on the signs • Any historical background of the signs observable at the site • Any further notable characteristics of the signs: size, color, configuration.

RESULTS AND DISCUSSION
Many available sources are informative about the demographic composition of Istanbul during the Ottoman era. For the 17 th century, I have relied on the only other linguistic landscape study on Istanbul. Using contemporary accounts by travelers, dragomans, and transcription texts, Csató et. al. (2010) provide a portrait of the linguistic ecology of Istanbul in the 1600s. Their study also provides maps (see Fig. 1) showing the locations of the many religious and ethnic communities found at that time-points of contact they have aptly labeled, "areas of cross linguistic encounters" (Csató 2010: 421). A contemporary account of Istanbul for the early 1900s, written by Johnson (1922), also includes a "Nationality Map" showing identical locations for these same communities three hundred years on. For population estimates for Istanbul districts within the city, I have relied on the information provided in secondary sources (e.g., Mills 2010, King 2014. I have also accumulated abundant evidence from old photographs dating from 1870s through to the early republic years (e.g., Fig. 2), calendar leafs (e.g., Fig. 3); and postage stamps (which precisely document the transition from Ottoman Turkish to French-the then lingua franca), to modern Turkish written in the Latin alphabet). Altogether the secondary sources and 'alternative' linguistic landscapes demonstrate without any doubt that Istanbul's late Ottoman era linguistic landscapes reflected the cosmopolitan composition of the various resident communities (i.e., not transient or 'tourist' communities, but communities that had been founded hundreds of years prior). These findings suggest distinct sociolinguistic regimes were to be found in each district of the old city based in large measure on the resident populations of those districts.  The main results of the survey of ten Istanbul linguistic landscapes are found in Figure 4. These are arranged from top to bottom according to the percentage, in descending order, of Turkish monolingual signs for each district (see the 'Turkish' column in Figure 4). The districts with the highest percentage of Turkish monolingual signs are Bağcilar and Kuzcunguk, 94% and 92% respectively; the two districts with the least monolingual Turkish signage are Aksaray and Sultanahmet, 44% and 29% respectively. It is interesting that a Pearson correlation confirms that there are neither strong nor moderate correlations among the sign types: that is, for example, the proportion of monolingual English signs does not increase with a corresponding decline in monolingual Turkish signs. This suggests that each district has its own unique signature that is shaped by specific and local (and historically-based) contingencies. These results in Figure 4 suggest three configurations: Those districts whose signage is predominately monolingual Turkish (over 90%: Bağcilar and Kuzguncuk); those whose majority signage is monolingual Turkish (between 50%-75%: Istiklal -Haseki, 6 districts); and those whose signage less than 50% (Aksaray and Sultanahmet).

Kuzguncuk
Kuzguncuk is a small residential community of several thousand people on the Asian side, tucked in a narrow valley leading down to the Bosphorus, just two kilometers north of Üskudar-the latter being the former Greek city of Scutari which in the 1880s had a Muslim majority population, but also half as many Armenians. In the present day, there are two populations living in Kuzguncuk: a tightly knit, traditional elderly Turkish community, and a community of younger newcomers including art gallery and boutique shop owners, artists, and the like (Mills 2010). There are also a few longtime foreign residents from Europe, North America and Japan. However, Kuzguncuk had been, at least from the 1600s up until the early 1900s (Csato et.al. 2010) a chiefly Greek and Jewish neighborhood. In its day, the neighborhood also included a sizable community of Armenians. Population statistics (cited in Mills 2010, p. 43) for 1914 are indicative: 1,600 Armenians, 400 Jews, 250 Greeks, 70 Muslims, and 4 foreigners (i.e., Europeans). These facts explain the neighborhood's two Jewish synagogues, three Greek Orthodox churches, one Armenian church, the largest Jewish cemeteries in Istanbul with gravestones dating back over 200 years, and a large Greek cemetery, among other non-Muslim landmarks in the valley. These facts also suggest that Kuzguncuk's sociolinguistic regime in the prerepublic years was multilingual, including Greek, Ladino, and Armenian-and that what little signage were to be found along the main shopping street during these early years would include these languages. Note that these languages were used in everyday communication: the signs would have been intended for residents and local shoppers, not for tourists and transients.
Today, however, there are only a handful of Greeks and Jews resident in the community (Mills 2010). The Turkification policies of the early 1900s and events through the 1950s (e.g., the Istanbul pogrom of 6-7 September 1955) and 1960s (e.g., the deportation of thousands of Istanbul Greeks in the wake of the 1963 'Bloody Christmas' massacre in Cyprus) are the causes for the departure of the non-Muslim Kuzguncuk population.
Kuzguncuk has in the past 20 years undergone significant gentrification. Recently, there have been efforts to restore many of the old row houses that had been abandoned or fallen into disrepair. The main street leading up from the Bosphorus, İcadiye Sokak, is lined with small businesses, pastry shops and bakeries, restaurants, and art galleries. The restaurants and tea houses always have customers, most of them local on weekdays, usually seated outside on small tables and chairs, spilling out at times into the street. A popular tourist destination for locals (particularly for businesses who specialize in wedding photos for the soon-to-be married couples who yearn for the nostalgic Istanbul scenes of an imagined 'yesteryear') its small food establishments do a fast business on the weekends.
As above, Figure 4 shows Kuzguncuk a strongly Turkish sociolinguistic landscape profile (e.g., 92% of its signs are monolingual Turkish). We can deduce that the target audience is Turkish, for both the residents and the weekend visitors. A few details of Kuzguncuk's signage will bear this out. We note that 244 of the 266 of signs are monolingual Turkish, so what is going on with the 22 signs? Of the 22 such signs: 5 are bilingual Turkish and English, ten are monolingual English, and 7 are categorized as Other. All of the five bilingual and ten English signs are commercial. In all cases, English and the other foreign languages (but see below) are used for their cachet of modernity and foreign cultural references (e.g., the "Betty Blue" restaurant), and minimally for the information they bear. Several examples follow: Turkish + English -"Harmony Sanat Gallery" [Harmony Art Gallery]-One of the five 'Turkish+English' signs, the use of the English expression "harmony" gives the art gallery a warm, calm and perhaps spiritual air. The use of the English word harmony is used as much for its cachet of modernity as it is for its informative content.
English -"Olive and Beyond"-One of ten monolingual 'English' signs to be found along the 400m of Icadiye Cad. surveyed. The sign appears above the store entrance, inscribed onto a fine piece of wood-there is no other signage on this store. The store features cold-press olive oil and olive oil products and has all the trimmings of being up-scale and exclusively for olive oil gastronomes. The sign "Olive and beyond" is informative about the store's main product, and at the same time, suggests in a marginally poetic fashion, a further universe of olive products. Although monolingual English, this sign is nonetheless intended for the neighborhood's Turkish shoppers, residents and local tourists alike. There are very few international tourists visiting this community, the use of English in this case is meant to appeal to an educated and discriminating Turkish clientele.
Turkish + French -"Sarmaşık // Café de // Balık // Keyfi" [ivy // café of // fish // mood] or approximately: "Ivy In-the-mood-for-Fish Café]"]-This is one of the seven 'Other' signs found in Kuzguncuk. The use of both Turkish and French (café de= 'café of') here is amusing, and the French expression gives the culinary experience a frisson in this modest Kuzguncuk eatery. There is no doubt that all Turkish speakers understand café de, not the least because the word "cafe" with and without the accent aigu is found everywhere in Turkey (as is the rare Turkish transcription of the French word, "kafe", though now we see examples of the English "coffee" everywhere).
In the examples above, the foreign languages on the signs are not chiefly informational in their purpose or only minimally so. The exceptions to the above are found in the signage for the two Greek churches and the Jewish temple, all designated as top-down. Altogether, there are five such signs along Icadiye Cad.-all of them monolingual Greek, Hebrew, or bilingual Greek and Turkish. The signs give the name of the church or synagogue (as is typically found in Istanbul, these Christian and Jewish edifices are today found behind high walls of concrete or metal, sometimes barbed wire topping the wall) and provide a spiritual message. They are intended to be informative, but there are only a handful of people today in Kuzguncuk who may be able to read them. Few Greeks and Jews live in the neighborhood-in fact, there is no sense in which we can say that, today, Kuzguncuk is a Jewish and Greek community.
One must ask, therefore, to what extent these signs or inscriptions mounted on the gates or doors of the churches and temples along İcadiye Sok. (or wherever they are found in Istanbul) constitute a part of the 'living' linguistic landscape? Should these signs be given their own classification, one that acknowledges their past relationship to the landscape, but also notes a presumed present-day irrelevance, except perhaps as archeological artifacts? The thing is, as above, there are pockets of these once vibrant communities still living in Istanbul. The Greek Patriarchate, the holy see of the Orthodox church, is located in the Fener district; that of the Armenian Patriarchate is found in the Kumkapı neighborhood. Small but dedicated groups are active in keeping their religious institutions alive. These signs are relevant to a small historical minority of Istanbul's population.
In fact, these and other religious or historical inscriptions found throughout the older quarters of Istanbul play a critical role in the overall linguistic landscape. Not only are they reminders of the sweeping transformations that these communities experienced through the first half of the 20 th century, but these also force us to adopt a historical perspective on the whole project-they historicize the landscape as few other signs have the power to do. Most importantly for this study, they also, as mentioned above, function as evidence for a sociolinguistic regime vastly different from that of the present day.

Sultanahmet
Let us now turn to Sultanahmet, the district with the lowest percentage of monolingual Turkish signs. Sultanahmet, in contrast to Kuzguncuk above, has been a Muslim district since the 1453 conquest, albeit a unique one. Photographic and other evidence suggests that in the late Ottoman years, this district (particularly the area that is surveyed for this study) was not the bustling tourist or commercial district it is today-it was largely devoid of shops or residential buildings. The sociolinguistic regime at the time must have particularly been dominated by Ottoman Turkish and Turkish vernaculars. In modern times, however, Sultanahmet has been a major tourist destination bringing millions of international visitors every year to Istanbul. Divan Yolu Cad., the street surveyed for this study, is the main street fronting many of the historical sites and is lined with kebab and fastfood restaurants, sweets and pastry shops. souvenir and money change shops, pharmacies, and travel agencies. These business cater to the tourist trade, both domestic and international. Figure 4 shows that Sultanahmet has a linguistic landscape profile very different from that of Kuzguncuk or indeed any of the other districts surveyed for this study. Monolingual English signs account for 51% of the total while monolingual Turkish account for only 29%. Thus the majority language here (English) is not the mainstream language (Turkish) in this district's landscape. Altogether, English is found on 66% of the signs (i.e., including all the English found on the bilingual Turkish and English signs, the Other signs, and the English monolingual signs) of the study site along Divan Yolu Cad.
A breakdown of these results in terms of agency is revealing (see Table 1). The signs are mostly bottom-up commercial signs for all groups (above 79%), but in the case of English monolingual signs, the percentage is 100%. These signs (all sizes, but many small ones with information about domestic tours) were particularly to be found in the front windows of the travel agencies (e.g., "Marco Polo Travel Agency") advertising domestic tours, souvenir shop (e.g., "Jasmine Spice Shop and Turkish Delight") and restaurants ("Cosy Pub Restaurant") along Divam Yolu Cad. About half of these signs were informational, the other half functioned symbolically for their cachet of modernity. For the Turkish and English bilingual signs, while 26 signs were commercial, 7 are top-down. Many of these latter signs had been placed by the official bodies such as the cultural ministry (for historic building renovation) or by the municipality (information panels on the Sultanahmet tramway station) and are obviously intended for international and domestic tourist populations. The commercial bilingual signs were found among souvenir shops, pharmacies, and beauty and tattoo businesses. There are 28 signs categorized as Other, one of which was top-down, but the balance of 27 were commercial. The languages on these signs included French, Russian, German and two monolingual Persian-mainly to be found on money change shop windows and restaurant menus on display (in four languages: Turkish, English, Arabic and Russian). Interestingly, 16 of these signs included the Arabic language. These statistics attest to the strength of the international and domestic tourist market in this district, particularly the reliance on English for its appeal both as a language of commercial tourism and its symbolic value.

Further observations on Istanbul's linguistic landscape
Arabic -The year 2016 was a very poor one for tourism in Turkey, particularly for Istanbul. There were few tourists from Russia or from European and North American countries, but larger numbers coming from the Arab states, Pakistan, Indonesia, and Malaysia. For this reason, we find a good number of Arabic signs in Istiklal, Aksaray, Kumkapi, and Sultanahmet. So for example, of the 28 signs in the 'Other' category in Sultanahmet, nearly half of them were monolingual Arabic or included Arabic. Of the 34 'Other' signs in İstiklal, 19 included the Arabic language.
French-as a lingua franca -French was the international language of culture and diplomacy in Europe from the 18 th century until the end of World War II; it was also widely used among the different speech communities in Istanbul, and also as a means for the Ottoman administration to communicate with its subjects. Over time, Turkish incorporated several thousand French words, much as English absorbed Norman French after the 1066 Battle of Hastings. The first non-Ottoman language used on Turkish postage stamps was French in 1876 (e.g., "Emp. Ottoman", "Postes Ottomanes", "Vingt Kurus"). Even in the 1960s, the Turkish government issued visas for foreign passports that were bilingual Turkish and French (e.g., for a diplomatic visa issued in 1961: "Délivré le ____ ", "Nombre de Visites Permises ____" ; the blank spaces on the visa stamp filled in by hand with French words in ink: "22 Mars 1961" and "Plusieurs"). For this reason, one can always count on finding French words in Turkish orthography inscribed in on buildings and the like around the city. The Greek consulate, for example, along the study site in Istiklal has inscribed above it: "Consulat General de Grece." In fact, words of French origin are pervasive on signage throughout Turkey: "noter"=notaire, "avukat"=avocat, "kuaför"=coiffure, "bank"=banque.
Turkish, the role of English, language regimes -With the exception of Sultanahmet and Aksaray and perhaps a few other districts not included in this survey (e.g., upscale neighborhoods such as Nişantaş), Turkish dominates the linguistic landscapes of Istanbul. And yet, inasmuch as English is used more for its symbolic value than for true informational purposes for English-reading audiences, Turkish is truly the dominant language of Istanbul's signage. Based on my visits to many districts outside of the old city centers, I believe that 90%-plus monolingual Turkish signage will be the rule and not the exception-as we found in Bağcilar. So while the data show that English has penetrated the landscapes of Istanbul, a fact attributable both to its status as an international language and globalization processes (Crystal 2003, Garcia 2001, these findings suggest that English, though pervasive in one or two sites, plays a limited role in Istanbul's overall language scheme, and that it functions as a cachet of modernity and in its role in the international tourist trade. Significantly, English plays no role in the formation of Turkish identity or citizenship. And yet, the several districts in the old city today enjoy a cosmopolitan air in the sense that international visitors will always be accommodated in one language or another by the Turkish shop keepers and restauranteurs. This is especially true for Sultanahmet, but also for Aksaray, Istiklal, Galata, Haseki which are regularly visited by different groups of language speakers-not only tourists, but visa seekers and people in commercial trade. That said, Sultanahmet is indeed a small island in the sea of Turkish language dominance in Istanbul. The language regime on the street in Sultanahmet (and incidentally the Grand Bazaar) is that of a polyglot's paradise. Shopkeepers and touts are often functional in four or five languages and take pride in their linguistic virtuosity. They will greet a foreign visitor in English or Japanese or Arabic, and segue into German, Chinese or French if they've missed the mark. Behaviors and expectations and attitudes about speaking languages other than Turkish are enthusiastically positive. In contrast, the language regimes on the streets of Bağcilar and Kuzcunguk are decidedly monolingual-although you will find goodly numbers of educated, fluent English speakers among the newcomer population in Kuzguncuk.
Kurdish -Finally, in all my observations of public signage taken across Istanbul, including graffiti, I have not come across one instance of the use of the Kurdish language, notwithstanding that there is a significantly large population of three to four million Kurds living in Istanbul today. This is ultimately owing to the government's treatment of the ethnic Kurds (comprising 15%-20% of Turkey's total population), and the government's implicit and explicit language policies (Zeydanlioğlu 2012) that have either "invisibilized" the Kurds or proscribed all use of Kurdish across the Turkish Republic (Haig 2003) in the post-independence years. In such a politically charged atmosphere, Kurds will not risk publically revealing their identity by writing Kurdish on signs. Literacy in Kurdish is also reported to be very low, estimated to be around 27% -likely far lower for women than men (see Ethnologue 2017 under the listing for Turkey)-one outcome of decades of top-down assimilationist policies. There had been a relaxation of the government's language policy on Kurdish in 2008 (ostensibly to fulfill requirements for EU candidacy), but any hope for an enlightened and humane approach to this issue seems to have been lost in light of both the terror attacks of the past several years and the attempted coup of 15 July 2016.

CONCLUSION
This survey of Istanbul's districts has shown how linguistic landscapes and their corresponding sociolinguistic regimes are transformed configured by historical, social, and commercial contingencies. It is clear that the what we observe today cannot be interpreted except in the light these contingencies. As we have seen in some detail, the sociolinguistic profiles of Kuzguncuk and Sultanahmet manifest influences from many sources. In the present day, it is English, with its accents of cosmopolitanism and modern chic and its use as the lingua franca of the international tourist trade, that contributes the greatest share to languages other than Turkish in Istanbul's public spaces. However, it is the Turkish language that dominates the linguistic landscapes of greater Istanbul.

ACKNOWLEDGEMENT
I thank my colleagues at Boğaziçi University's Department of Foreign Language Education for their kind support for this project during my 2016-2017 academic sabbatical in Istanbul, Turkey.