Lost in Translation: International Relations as Inter-lingual Relations1 Einar Wigen Department of Culture Studies and Oriental Languages, University of Oslo, Norway Words: 11 026 Abstract To the extent that polities interact across linguistic boundaries, international relations are also inter-lingual relations. Given that relations and practices are given meaning in language, it follows that for inter-lingual relations to function smoothly, it has to be possible to give at least a minimum of shared meaning to mutual relations. Otherwise, the divergence of meaning and thus also of expectations will limit the possible extent and quality of those relations. This is one reason why some of the most powerful political blocs in the international system are centred on specific languages. Linguistic communities often also constitute political communities, for reasons of mutual intelligibility. Following the onset of internationalisation in the 19th century and globalisation in the 20th, linguistic divides have gradually narrowed, especially in terms of political vocabularies. This article proposes a theory of ‘conceptual entanglement’ as an approach to studying how compatibility of meaning comes about and is maintained between linguistic communities and hence also between polities. This is elaborated by the case of how the concept of ‘civilisation’ became embedded in the Ottoman language. Keywords semantics, Ottoman Empire, Turkey, international society, conceptual history, globalisation, Reinhart Koselleck

1

The author would like thank Morten Skumsrud Andersen, Kristin Haugevik, Helge Jordheim, Halvard Leira, Iver B. Neumann, Bjørn Erik Rasch, Ole Jacob Sending, and Bjørnar Sverdrup-Thygeson for commenting on early drafts. Tarak Barkawi also deserves grateful mention for sparking the initial idea for the paper.

1

[If a] sentence is faithfully translated into a foreign language: [is it] two distinct statements or one?2 Introduction With scholars like Michel Foucault (1970, 1972, 1977), Quentin Skinner (1978, 1981, 1996), Hans-Georg Gadamer (2004) and Jacques Derrida (1976, 1978) leading the charge, social theory has had its linguistic turn. Yet, none of these intellectual giants have paid much attention to interaction across languages. In their wake, the discipline of International Relations (IR) has gone through a linguistic turn of its own. While one might think that the study of such interactions would be of immediate relevance for the discipline specifically set up to study international relations – which are frequently also interlingual relations – nothing of the sort has emerged. IR, like much European and American social theory, is largely a monolingual affair. Consequently, it is blind to the transaction costs involved in crossing linguistic boundaries, despite some of the key names associated with the linguistic turn in IR being bi-, tri- or even quadrilingual (Hansen 2006; Jackson 2006; Neumann 1999; Rumelili 2007; Wæver 1998).3 Outside the linguistic turn, neo-liberal regime theorists like Robert Keohane (1984) have been highly attentive to transaction costs (see also Coase 1937, 1960; Williamson 1975, 1985). Keohane argues that the institutionalisation of cooperation reduces transaction costs, making future cooperation more likely (for a critique, see Fearon 1998). Other (mainstream) scholars like David Lake (2009) have argued that hierarchy reduces transaction costs. But language was never part of it. For instance, transcending the English/Kiswahili linguistic boundary involves higher transaction costs than interaction within English, both in terms of the meaning lost in translation and the increased resources needed in order to interact and govern efficiently. Such points seem to be lost on these scholars. As IR is steaming ahead into a ‘practice turn’ (Adler & Pouliot 2011; Andersen & Neumann 2012; Neumann 2002), I argue that it should take another look at language. More specifically, it should take a look at relations across languages: interlingual relations. The article is structured around the question what are the processes that have enabled meaningful social interaction across linguistic and hence also political 2

Foucault 1997: 101 Jackson 2006 does touch upon the fact that German and English are different languages, with different conceptual histories, but nevertheless treats interaction across the German/English boundary as if there is no semantic transaction cost involved. 3

2

borders? I start out by going through some of the extant theoretical approaches that are on offer, through which I seek to clarify the epistemological foundation and theoretical frame of reference for part two. The second part consists of an elaboration of a theoretical concept – conceptual entanglement – that may serve as a tool to make sense of what happens when social relations thicken between polities that previously have had either quantitatively few or qualitatively poor relationships. The third part consists of a case to illustrate this process, by tracing the entanglement of the Ottoman concept of medeniyet (settled habitation) with the French concept of civilisation. Language, Translation and Globalisation The UN has committed to six official languages4, institutionalising these languages’ superior status vis-à-vis the other territorially defined vernaculars. Although they are vernaculars (with the possible exception of Mandarin), they are, so to speak, less territorially defined than the non-UN languages. The American political scientist Robert D. Putnam’s (1988) theory of foreign policy being conducted in ‘two-level games,’ with one game taking place domestically and the other on the international arena, and each being subject to its own particular constraints and possibilities, may be a good starting point when thinking about this. The fact that foreign policy is communicated to other states in one language (such as English) and legitimised to the domestic audience in another (such as Turkish in the case of Turkey) means that statesmen are forced to draw upon two different sets of conceptual histories and two separate discursive structures enabling and constraining what is possible to formulate. This presents a host of problems for the actors involved, since what is legitimate on the domestic arena may not be as easily legitimised on the international arena and vice versa, problems that IR scholars seem to have ignored when conducting discourseoriented inquiries into international relations. The processes that set in motion the quantitative and qualitative increase in inter-lingual relations can be subsumed under the heading of ‘globalisation.’ Research on globalisation seems to focus on circulation and interaction quantitatively rather than a qualitative ‘thickening’ of social relations (see Holm & Sorensen 1995: 1, 279; Giddens 1990: 64, but see Steger’s definition (2009: 15), which includes the quality of 4

relations

by

using

the

Durkheimian

concept

‘consciousness’).

While

These are Arabic, Mandarin, English, French, Russian and Spanish.

3

communication technology makes it possible to communicate more quickly and more frequently across greater distances than before, studies of globalisation seldom tell us much about how the meaning content is transformed when translated and dislocated.5 However, geographic distances are not linguistic distances, nor are they cultural distances. A large proportion of this quantitative increase in social relations across the globe in fact entails intra-lingual relations (e.g. English-speakers in Britain interacting with English-speakers in Australia). Nevertheless, the increase in communication is part and parcel of the processes whereby European powers came to rule most of the surface of the Earth directly or indirectly by the early to mid 20th century (cf. McNeill & McNeill 2003). This imperial expansion made the English, French, Dutch, Russian, Portuguese and Spanish languages (each waxing and waning over time) languages of inter-polity relations (and inter-communal relations within imperial polities), supplanting regional linguae franca (comp. Windler 2001).6 This political conquest, along with its enabling technologies, brought with it an increase in the quantity of relations between linguistic communities (whether polities or not). This was due to semantically not very sophisticated processes (military conquest and domination is not immediately dependent upon understanding one another), and did not as such bring about either cultural homogeneity or semantic compatibility. To the extent that they were employed, more sophisticated forms of rule (i.e. those involving less brute force and more cooperation), following after conquest, were dependent upon communication across linguistic borders. Ruling foreign cultures efficiently depends on bringing down the semantic and the financial cost of inter-lingual transactions. Likewise, polities that were not directly subject to foreign imperial rule, such as the Ottoman Empire and Japan, were ‘expanded upon’ by international society (Neumann’s phrase (2011), see also Bull 1977; Bull & Watson 1984; Kayaoğlu 2010; Watson 1992), and increasingly had to legitimise their claims to rule by use of foreign 5

To the extent that it has dealt with this, IR has touched upon this largely by way of the ’history of transfers’ literature. One example is John Hobson (2004), who has conceptualised cultural contact through ‘resource packages’ that move from one culture to the next, being unpacked largely intact and unchanged upon arrival, affecting only the recipient culture. I take issue with such a view, since all inter-lingual relations are two-way streets (which is not to say that they involve two equal partners, but it always takes two to tango, however disinterested one of the partners may be). Instead, I come at this from a history of entanglement perspective (Werner & Zimmermann 2006). Inter-lingual relations entangle the different collectives involved, as well as changing what is ‘transferred’ – not to the same degree, of course, but nevertheless, both cultures become part of the relationship (something that is in line with Emirbayer 1997 and Jackson & Nexon 1999). 6 Even Russian-Chinese treaties of the 18th century were concluded in Latin, a language that neither used as their literary or bureaucratic language, but which was made possible by Jesuit missionaries at the Chinese court.

4

vocabularies. At stake was their political survival. One of the key concepts of the time was the ‘standard of civilisation’ (see Gong 1984; Koskenniemi 2010; Bowden 2009; see also Salter 2002). According to the European-formulated international law of the time, any polity that did not attain this standard was a legitimate target for imperial subjugation. Successfully claiming to be ‘civilised’ would bring formal external legitimacy to their rule, thus being treated as an equal within international society rather than as an apt target for colonial expansion. Post-colonial studies have opened a field of inquiry by saying that the life worlds that emerge in the asymmetric relations between the cultures of the imperial metropoles and those of the former colonies need to be studied, and scholars working within this field are clearly aware of the existence of multiple languages (Spivak 1987, Chakrabarty 2000; Bhabha 1994). However, there is little in terms of specifying a methodological approach to this outside literary studies. The work of this article is to specify how these relationships can be studied by focusing on concepts and how concepts and what happens to meaning when interaction takes place across language boundaries. The reason one can treat inter-lingual relations between major Western languages as if they involved no significant transaction cost (moving between different languages as if they were discourses in a single language), is exactly because of the processes of conceptual entanglement that I develop theoretically in this article (cf. Pernau 2012; Werner & Zimmermann 2006). However, polities emerging in different traditions have political concepts that do not mean the same, mostly because their concepts have not become entangled (for a poignant example, see Everett 2008). Their emerging compatibility in translation is a social process that involves power, a certain creative tension and the practice of interpretation. Without a process of entanglements, concepts do not by themselves become compatible and translations will inevitably involve an excessive amount of interpretation. When languages are less entangled, there is a greater transaction cost involved in interacting across linguistic borders. Anyone communicating across such a linguistic border needs to include more context when translating a text and explain more to a conversation partner in order to get any kind of meaningful interaction. Even when one goes to such lengths, there is always a residual that cannot be translated, interpreted or explained (Tymoczko 1999: 23). When this residual is a relatively large proportion of

5

the text’s meaning, the transaction cost goes up, because there are great aspects of the relationship that cannot be given mutually intelligible meaning. International relations that are also inter-lingual relations require a smoothly operating system of translation. But, as the semiotician Yuri Lotman points out: […]

an

accurate

translation

presupposes

that

mutually

equivalent

relationship[s] have already been established between units of the two systems, as a result of which one system can be represented in the other. That is what makes it possible for the text of one language to be adequately expressed in another one (Lotman 2000: 37).7 I will call such mutually equivalent relationships between units of meaning ‘conceptual compatibility’. Since cultures and languages are always in flux, emerging through social historical processes, such relationships are inherently unstable. Compatibility at one point in time does not make for compatibility at another. In order to create and uphold conceptual compatibility there needs to be processes of mutual textual exchange on a regular basis (this may of course also be oral communication, which is almost impossible to trace historically). This leads me to agree with Robert Keohane (1984) that institutionalisation of cooperation reduces transaction costs, and thereby makes future cooperation more likely. Moreover, I would like to expand this view to interlingual relations, with its resulting conceptual entanglements. The more two linguistic collectives interact across the linguistic boundary, the more their concepts become entangled. The more their concepts become entangled, the more does the transaction cost decrease, thus inviting (but by no means guaranteeing) more interaction. Just like any relationship, interlingual relations involve power. As has been pointed out by key translation studies scholars: […] translation does not happen in a vacuum, but in a continuum; it is not an isolated act, it is part of an ongoing process of intercultural transfer. Moreover, translation is a highly manipulative activity that involves all kinds of stages in that process of transfer across linguistic and cultural boundaries. Translation is not an innocent, transparent activity, but is highly charged with significance at 7

But see Belloc 1931a and 1931b: ‘there are, properly speaking, no such thing as identical equivalents.’ See also Nida (2000: 30) for formal equivalence versus dynamic equivalence.

6

every stage; it rarely, if ever, involves a relationship of equality between texts, authors or systems (Bassnett and Trivedi 1999: 2). Conceptual entanglement is an effect produced by such power, and is thus seldom a meeting on equal grounds. Roxanne Doty (1997: 385) has touched upon ‘leaks’ and translations between discourses and fields, if not languages. She uses it as way to get at agency and argues that ‘A radical understanding of practice suggests that the possibility of agency results from a complex weaving together of the subject-positions and meanings that are available within multiple and overlapping discourses’ (Doty 1997:385). Halvard Leira (2003: 31-32) claims that this is a type of ‘inter-discursivity’; since subjects will have positions in different discourses, the boundaries between the discourses become unclear. This opens for the possibility that ‘re-writing the meaning of concepts within a discourse may be influenced by other discourses’ (Leira 2003: 32).8 True, all subjects will have positions in different discourses, but not all subjects will have competency in different languages. It is this scarcity that gives multi-lingualism into a potential privilege that makes it attractive for the state, and distinguishes a language from a discourse. There are numerous discourses within a language – and these discourses may transcend linguistic boundaries (which is another way of saying ‘conceptual entanglement’ – the phenomenon I investigate in this article) – but the degree of competence, skill and effort needed to master another language is much greater than for mastering a different discourse (especially given the differences of grammar and sound systems that come in addition to differences in meaning). However, as language is a continuum of overlapping elements and shared meanings, this is merely a working definition for distinguishing between ‘inter-discursive’ and ‘inter-lingual’, and deals with the distance between two systems of signs. The Russian literary theorist Roman Jakobson has proposed distinguishing between intralingual translation, which is ‘an interpretation of verbal signs by means of other signs in the same language’ and interlingual translation, which is ‘an interpretation of verbal signs by means of some other language’ (Jakobson 2000:

8

My translation from Norwegian original.

7

114).9 Note the affinity between Doty’s inter-discursive leaks and intra-lingual translation. The distinction has been criticised by Jacques Derrida (1985), who argues that differentiation is impossible because Jakobson ‘obviously presupposes that one can know in the final analysis how to determine rigorously the unity and identity of a language, the decidable form of its limits’ (Derrida 1985: 225). This is true ‘in the final analysis’, but I would follow his English translator, Gayatri Chakravorty Spivak, in arguing that in order to facilitate analysis of an ever-changing and fluid world one needs to keep something stable in order to narrate the analysis (Spivak 1987). Languages are not stable entities with agreed-upon boundaries. Rather, they are everchanging constellations of lexis that float into one another. But it makes sense for the sake of argument keep them separate and to treat them as if they were stable. With these caveats, the intra/inter-linguistic distinction may be a useful for IR. Hermeneutics and Conceptual History The linguistic turn in IR has largely drawn upon the theoretical developments associated with structuralism and post-structuralism, with Ferdinand de Saussure (1983) and Michel Foucault (inter alia 1997) as the respective key references, while ignoring German contributions such as Hans-Georg Gadamer (2004) and Reinhart Koselleck (inter alia 2004) (Der Derian and Michael Shapiro 1989; Bartelson 1995; Der Derian 1987; Neumann 1996, 1999; Wæver 1998; Hansen 2006). Nevertheless, there is no lack of ‘concepts’ in IR (see e.g. Griffiths & O’Callaghan 2002). But ‘concepts’ are generally tossed about without any systematic treatment, as a way for scholars to talk about meaning without saying ‘discourse’ (notable exceptions include Jordheim & Neumann 2011, Neumann 2002a and Jackson 2006).10 The godfather of German conceptual history, Reinhart Koselleck, was a student of Hans-Georg Gadamer’s. And, like Gadamer, he did not believe that meaning of a text was essentially tied to the intentions of the author (see Gadamer 2004: 158).11 Gadamer argued that understanding was linguistically mediated, through conversations with others in which reality is explored and an agreement 9

This was completed with a third type of translation, which is not relevant here, namely intersemiotic translation, which is ’an interpretation of verbal signs by means of nonverbal sign systems.’ Examples of this would be the filmatisation of a novel or the instrumental performance of a poem. 10 Although the reference is missing, parts of Friedrich Kratochwil’s (2011) analysis of the concept of practice can hardly be understood as anything but a Koselleckian analysis. 11 It is not commonly accepted that Schleiermacher really meant this, but that is a different matter. See Gjesdal 2009.

8

reached. This agreement represents a new understanding (Ramberg & Gjesdal 2005). What is achieved in understanding is a fusion of horizons between the two conversation partners. When reading a text, this means that the horizon of the reader fuses with that of the text, thus making him ‘see’ what the text ‘sees.’ Much in the same way as Gadamer takes interpretation to be productive, so does Yuri Lotman take disturbances in transmission to be productive of meaning. The main difference between the two is that where Gadamer had as a methodological goal to ‘understand’ a text by engaging with it over and over and to understand this text by reference to a tradition that was assumed to be the same for both author and reader, Lotman saw disturbance in itself as productive and was interested in what happens to meaning when a text is displaced from the semantic system of its production (Andrews 2004: xxiii; Lotman 2004: 4-6, see also Luhmann 1995). Gadamer treated improved understanding as a methodological goal of engaging with a work. Lotman on the other hand, was more interested in what happens when meanings collide. The emphasis here is on the individual reader, and Gadamer never takes this to the collective level. This rests on the key distinction between meaning as mediated by language and as constituted in language. Because he ignores both the social, collective aspect of text interpretation and the power aspect of text translation and interpretation, Gadamer’s approach, and especially his concept ‘fusion of horizons,’ may serve as a sensitising device for IR, but not as a methodological tradition. Koselleck, however, worked at the level of collective meaning.12 While he did not break with hermeneutics as such, he placed less emphasis on individual acts of interpretation. Collective meaning took the place of individual understanding as the objective of scholarly effort. However, there was no principled opposition to Gadamer’s transfactualism, and Koselleck did not shy away from writing about understanding as such (cf. Jackson 2010). In Begriffsgeschichte, meaning is embedded in concepts and words. Concepts, however, stand ‘above’ words in that they embody more meaning, and are somehow qualitatively different from ‘mere words.’ Every concept is associated with a word, but not every word can be considered a concept. Social and political concepts have a certain generic quality that is applicable to a number of situations and are always multi-faceted, something that makes them objects of competition (Koselleck 1982: 12

For the Koselleck reception in IR, see Jordheim & Neumann 2011; Jackson 2006.

9

418). ‘The meaning of a word can be determined exactly through definition, concepts can only be interpreted’ (Koselleck 1972: XXIII).13 Any attempt to fix such relationships is a political act, as this tries to force a particular interpretation of concepts on society, and to constitute society in a particular way.14 While words can be defined, concepts always have to be interpreted by reference to what Koselleck calls their ‘semantic field’ (Andersen 2003: 34). Koselleck did not consider the distinction between concept and word to be an analytical prioritisation, but something that belonged to the concepts in and of themselves. A word may have several possible meanings, but a concept combines in itself an abundance of meanings. Thus a concept may be clear, but it must be ambiguous. It bundles together the richness of historical experience and the sum of theoretical and practical lessons drawn from it in such a way that their relationship can be established and properly understood only through a concept (Koselleck 2011: 20). I would take a more pragmatic view and say that the distinction between word and concept is an analytical one, made on the basis of the extent to which the word/concept carries multiple meanings and the role it has in the language under study. The main consideration of whether it can be useful to treat something as a concept is the extent to which an inquiry into the history of a word and its synonyms yields interesting results for an analysis of social relations. Koselleck, however, takes concepts, or rather key concepts,15 to be those concepts without which the historian cannot write history (Koselleck 2011: 19). Key concepts in IR would either be those concepts without which our actors cannot conduct their affairs or those without which we cannot conduct ours as IR scholars. ‘Key concepts’ therefore, is not a stable entity. And the key political concepts of the Ottoman language were different not only from English concepts today, they were also different from Turkish concepts of today and

13

Translation by Niels Åkerstrøm Andersen, 2003: 37. This is linked with Friedrich Nietzsche ([1887] 1996) Genealogy of Morality, where he argues that ‘only that which is without history can be defined’ (60), as history will always imbue words with more than one meaning. 15 Grundbegriffe – literally ‘basic concepts’, but ‘key concepts’ is an established translation, see Ifversen 2011: 87. 14

10

of French, English and German concepts of the time. There is a semantic gap both along the diachronic and the synchronic dimension. Language Games An important departure point for the literature on how social relations are constituted in language is the theoretical concept of ‘language games’ as developed by the Austrian philosopher Ludwig Wittgenstein (1975; 2009). ‘Game’ is used as a metaphor for a set of norms and rules that outline what is socially acceptable to claim. These norms and rules are not objectively given, but socially accepted. They need not be there, but since they are, all players have to abide by them in order to be taken seriously. In short ‘a language-game is something that consists in the recurrent procedures of the game in time’ (Wittgenstein 1975 §519), where the appropriateness of a statement is determined not by the correspondence between the statement and an empirical fact, but between a statement and a response. Wittgenstein’s language games are intrinsically tied to practice (Wittgenstein 1975 §229). Systems of propositions and ‘the rest of our proceedings’ (which I would call practices) come as totalities (Wittgenstein 1975 §140; see also Foucault 1997: 97; Neumann 2002b), which have to be learnt (Wittgenstein 1975 §476, §535). So, languages are systems of truth claims, and the truth of new propositions may only be evaluated by reference to the pre-existing system. This is an interesting way of exploring language, but it makes language into an ever-expanding structure of truth claims. Moreover, it does not tell us anything about what happens when a truth claim is brought in from a different system by a person familiar with both systems. I may then sum up by saying that while concepts are focal-points of intersubjective knowledge, these concepts are imbricated in practice, much as Wittgenstein would claims that ‘Our talk gets its meaning from the rest of our proceedings’ (Wittgenstein 1975 §229). Practices are competent performances. More precisely, practices are socially meaningful patterns of action which, in being performed more or less competently, simultaneously embody, act out, and possibly reify background knowledge and discourse in and on the material worlds (Adler & Pouliot 2011: 6).

11

A way to bridge practice theory and conceptual history is to say that the boundaries for acceptable conceptual interpretation set boundary posts for acceptable practice. The acceptability of an interpretation is an inter-subjective affair, liable to struggles, but shared within a collective (for a similar line of reasoning, see Krebs & Jackson 2007). Discursive struggles – struggles over what practices are legitimate, how and why – are usually focused on the interpretation of one or more key concepts. Successfully intervening with a new interpretation, thus adding that interpretation to the multiplicity of interpretations that are there already, moves the boundary posts and widens the scope for what practices can be legitimised by use of the concept. Making a particular interpretation illegitimate – or simply ‘wrong’, thus detaching it from the concept – narrows the scope, and makes the concept unusable to legitimise practices for which it had previously been used (this is something Koselleckian conceptual history does not allow for). Since concepts are polyvalent, they can be used to legitimise a practice to more audiences than one. By using a particular concept, such as in the Ottoman case of ıslah – which at once meant ’correction of bida’,’ (innovation/malpractice) and ’reform’ in the sense of adopting societal models from Europe, the Islahat Fermanı (Rescript of Reform 1856) meant both a rescript to correct religious malpractice and a rescript to reform the state in the European meaning of the concept of ’reform’, with which it was in the process of being entangled. One could of course argue that using a certain concept means invoking references to certain texts, but this would be too neat, because neither speaker nor listener is likely to think or make mention of the specific texts, or even know of their existence. Users of a concepts do not make explicit reference to other texts, they draw upon a intersubjectively held body of knowledge that is largely implicit. Only when an interpretation or conceptual usage is not intersubjectively held does a user have to ‘activate’ intertextual linkages and make explicit reference to texts to invoke authority. Since there will inevitably be groups whose position is legitimised by the current interpretation of particular political concepts, putting forward an interpretation that is not already intersubjectively shared will often, though not necessarily always, involve contestation. Putting forward a new interpretation or trying to delegitimise an established one is an act to move the boundary posts of what practices are possible to legitimise by reference to the concept (for a Weberian take on the problematique, see Krebs & Jackson 2007). 12

I would contend that not only are concepts discursive resources to draw upon when giving meaning to practice, the very act of interpreting them is a practice in and of itself. Routine interpretation, that is, interpretation within the established and accepted boundariesfor legitimate interpretation, draws on, embodies and reifies background knowledge. What is remarkable about the Ottoman Empire and the Ottoman language in the period between the eighteenth and the early twentieth century, is that this corpus of legitimate texts to be invoked when making a new interpretation or delegitimise and thereby ‘erase’ an old interpretation of a concept changed. While the Ottomans had had plenty of interaction with Europeans, and there was no end to the of technological and military entanglements between the Ottoman Empire and what may anachronistically be called European polities, the body of texts that made new concepts and new conceptual interpretations legitimate was almost entirely separate. Political meaning was only to a very limited extent shared across the linguistic divide. The practice of interpreting a text usually draws on a particular body of intersubjectively shared but implicit knowledge, but when the interpreter needs to convince someone, or assumes that he or she will need to back up their interpretation, they will inevitably have to invoke an authority, which in a text society is a textual authority. And certain people are better positioned than others to make their interpretations become accepted (cf. Doty 1997: 385) – they speak from a subject position vested with the authority to make interpretations, such as the Şeyhülislam, the highest rank among the ulema of the Ottoman Empire. Choosing a text or a body of texts to draw upon when making a contentious interpretation, and using that text, is a practice. What changed in practice terms was that when faced with political contention, certain Ottoman hommes des lettres started drawing upon French texts rather than Arabic texts when interpreting concepts. But here it gets complicated. For a period Ottoman hommes des lettres such as Namık Kemal implicitly interpreted the very Arabic texts which it was customary to draw upon when making a contentious interpretation by reference to French texts. There were parts of the audience who would not have accepted the French texts as legitimate for re-interpreting Ottoman concepts, but other parts which accepted the French interpretation and recognised it even when it comes via Arabic. By going via re-interpretation of the Hadith it was possible to speak to multiple audiences. Part II: Conceptual entanglement 13

The key weakness of linguistic turn social theorists, with the exception of Yuri Lotman (see e.g. 2000), is that they are, at least in their approach to theory, monolingual (but see Bassnett & Trivedi 1999 for an example of how translation studies deals with this).16 They are attentive to the importance of language for social relations, but have little grasp of the extent to which languages may diverge. To improve this situation, I propose ‘conceptual entanglement’ between languages as an analytical concept to study the process of how concepts in different languages, and thus also languages, become increasingly compatible. Conceptual entanglement is about the movement of meaning ‘in’ and ‘out’ of a language, and the entanglements between concepts in languages. One reason why this has been left largely untheorised is that the main European languages, in which social theorists (and IR scholars) are sometimes bilingual, are so closely synchronised that it is easy to become blind to the fact that this situation emerged historically and that there is an on-going social process that maintains conceptual compatibility. While I will illustrate this in the case study that makes up the third part of the article, a European example contrasting the familiar with what is unfamiliar to most readers may be in place at this point. Because the meaning of a concept depends on historical and cultural context, meaning is always specific to the language in which it is used (‘belong’ would have been misleading here, as concepts move across languages). Thus, the French concept citoyen is not exactly the same as the German Burger, but they are entangled concepts, concepts that take their meaning from a great number of translations of political texts between French and German, and from semantic adjustments and interpretations made by a large number of French-German bilingual authors. Moreover, they have been established as equivalent concepts in translations of political treatises over a long period of time. Each translation of a French text into German which uses the concept Burger to translate citoyen is an act of conceptual entanglement which reinforces the conceptual compatibility between citoyen and Burger. But it does not make them the same (comp. Tymoczko 1999: 23), because although one may say that it brings in one more part of the French textual canon (to use Gadamer) or the French archive (to use Foucault), it does not replace what is already there, it merely adds to the repertoire of possible interpretations that 16

Michel Foucault knew classical Greek, Latin, German and English in addition to his native French, but he seems not to have theorised the relationships between these languages, nor drawn extensively on sources outside French except as one monolithic whole in which translation involved little semantic transaction cost and was not productive of meaning.

14

may be made of a particular concept, or reinforces particular French interpretations in German (for culture as repertoire, see Swidler 2003). All the while, both Germans and Frenchmen (at least the intellectuals and the political class) are aware that there is a difference between citoyen and Burger. Even when texts have not actually been translated, many European intellectuals are and have been bilingual. This means that many of those who have written on the concept Burger in German may have read French texts in which the concept citoyen is used. This influence is nigh impossible to trace, but also difficult to rule out. Moreover, such an influence is unlikely to result in an out-and-out direct usage of the French interpretation. Rather, it is most likely that the German author relates to this French concept by adjusting his own narrative accordingly, either by implicitly or explicitly contrasting his interpretation with the French or integrating some of the French interpretations in his own usage. This is often done without explicitly acknowledging the source, and as anyone who has every written anything of a certain length can attest to, it is not given that the author remembers where he has it from even if he wants to acknowledge the source (as I have already mentioned, references tend to become more explicit when the chips are down). This process of relating to interpretations in other languages is part and parcel of the process of conceptual entanglement. Adjusting one’s interpretations based on what one has read in ‘foreign’ languages one is familiar with alters one’s ‘own’ language ever so slightly in the direction of conceptual compatibility with the other language. As a contrast to the French-German concepts of citoyen and Burger, the Ottoman concept teba’, which is now usually translated into English as ‘subjects’ (in the plural) was not, at least not prior to 1856, extensively entangled with the English concept subject. Therefore, it meant something else.17 The historical traditions within which they emerged were politically separate (which is not to say that they were completely unentangled). The meaning of the Ottoman concept was instead more entangled with its Arabic and Persian equivalents, and thus expressed meaning equivalent to concepts in these languages. This made possible seamless translation

17

In 1902 the Ottoman encyclopaedist Şemseddin Sami wrote the following for the lemma tābiʿ [singular of teba’] – […] 1. Someone who follows another. Who follows [tabiʿet] and imitates: those who have come to follow the great [religious] scholars and imams of [one of the four orthodox] schools of Islam [meẕāhib]. 2. One who submits and complies. Someone who is obedient/compliant. Who is being led. The Kingdom of Bavaria is the subject [tābiʿ] of the Empire of Germany [Ālmāniya imparaṭorluğu] (Şemseddin Sami 1317 [1901/02]: 370).

15

between Arabic and Persian on the one hand, and Ottoman on the other. For Ottoman texts to be made available and given meaning in a European language, and vice versa, a great deal of interpretation, contextualisation and explanation was needed. Texts could not flow easily between these two clusters of entangled languages, because the key concepts had different histories and thus different meaning. Moreover, the syntax, style and writing conventions (such as Ottoman texts not having punctuation and very few finite clauses), made it extremely difficult to make texts written in one such entangled cluster legible in a language in the other. In short, there was an enormous transaction cost involved in interacting across the Ottoman/French linguistic divide. The key point about conceptual entanglement is that concepts do not only take their meaning from relationships within the language they are used, but also in their relationship with concepts that are treated as equivalents in translations. Where De Saussure (1983: 15) and Foucault (1997: 91) emphasise the relationships between signs (within a language) and Koselleck (1972: XXIII; see also Andersen 2003: 34) argues that a concept has to be interpreted by reference to its semantic field (which is also monolingually defined), I would argue that translation ‘equivalents,’ and compatible concepts are highly important in the way that concepts are meaningful. Since there is an interpretational cost involved in translation. It happens slightly less frequently than references are made among the ‘internal’ relations of the discursive structure. Burger, citoyen and citizen may be conceptually compatible in that it is possible to use them as equivalents in a translation. But their ‘primary’ relations are to other political concepts within German, French and English respectively. Only to the extent that translated texts have become part of the native political tradition can they say to constitute part of the archive that gives these concepts their meaning. Relations to other languages can be said to be secondary, constituting an ‘outer’ tier of the archive (or rather a more secluded part of the archive accessed less frequently?). I should probably take pains to point out that the distinction between a ‘primary’ and a ‘secondary’ archive is analytical, that it is not explicit in discursive practice and that it goes against the grain of post-structuralist sensitivities. But it may be useful to think of the archive not as one monolithic entity, accessible to all, but rather a repertoire where not everyone has competency to master every single aspect, nor knowledge of its existence (cf. Swidler 2003). Depending on the social context, one may nevertheless sometimes use such interpretations if one has the proper subject position to speak from. A nineteenth 16

century Frenchman knowledgeable in Ottoman had an good subject position from which to speak on any topic pertaining to French society and interpret Europeanderived concepts authoritatively vis-à-vis Ottoman conversation partners, even though the audience did not have the same repertoire of meaning, nor knowledge of an archive that would give the statements meaning without asking for further clarifications. It was up to our hypothetical Frenchman to, in the first instance, define concepts and introduce the repertoire that may make his statements meaningful (before they took on a life of their own). This is not to say with Marx that the ruling ideas are the ideas of the ruling classes, or that the ruling interpretations were the interpretations of the European imperial states, but that the discourse was set up in such a way that interpretations coming from certain subject positions were more likely to be accepted than those from others. Language Hierarchies The main European languages display a high degree of conceptual compatibility at least as pertains to political and religious vocabulary. This is the result not only of intense translation activity (largely centred on translation to and from English, French or German), but also the remnants Latin being established by the Church hierarchy as the language of religion as well as knowledge production. This dominance was all but total until it was challenged first by the use of vernacular languages for poetry, exemplified by Dante Alighieri’s Divine Comedy, written sometime between 1308 and 1321, and Geoffrey Chaucer, writing half a century later. A more serious challenge was the emergence of vernacular as a religious languages with the disintegration of Western Christendom under one ecclesiastical hierarchy following Luther’s obstinacy in 1517, made possible among other things by the appearance of moveable print in Europe (Hobson 2004). Furthermore it was challenged as a language for political treatises in Europe starting some time around the publication of Nicolò Machiavelli’s The Prince (1532)18. To speak of the universality of the French language in eighteenth-century European diplomacy is a received idea. Yet, if it is true that French was used more often than any other language, the reality is more complex. 18

Niccolò Machiavelli’s The Prince, first circulated in Latin manuscript under the title De Principatibus (About Principalities) in 1513, and was then published in Italian in 1532.

17

Mediterranean diplomacy escaped the predominance of the French language. From Constantinople to Maghreb, Italian remained, until the first decades of the nineteenth century, the main language of oral communications (Windler 2001: 85).19 So, (New) Latin only gave way to French as a diplomatic language in the 18th century, but was not universally used. And although Latin was challenged, it did not disappear entirely. Even in a Protestant country like Sweden, the ‘father of modern taxonomy’ Carl von Linné published his main scientific treatises, Systema Naturæ and Species Planetarum in Latin as late as 1735 and 1753 respectively. Among Muslim polities, Arabic has played a similar role, though in the absence of a unified ecclesiastical hierarchy (and the widespread use of Persian as a literary language outside those areas where Arabic was used as a mother tongue), its influence was less monolithic but more persistent. The presence of such linguae franca contributes to conceptual entanglement between the vernacular languages that share a common lingua franca. For matters of state, Ottoman Turkish was an extremely important language, largely because of the formal reach of the Sultan’s writ. In such a position, it was even used as the language of inter-polity correspondence for some political entitities that were subordinate to the Ottoman sultan. […] in the eighteenth century, the documents of [Tunisian] beys to European powers were almost always written in Turkish [i.e. Ottoman]. The predominance of the Turkish language in documents that the bey authorized was a manifestation of the formal recognition of the suzerainty of the Ottoman sultan. The Porte only considered itself bound by the Turkish version of the capitulations. The parallel writing of treaties with the regencies in the respective languages of the parties, in which each only signed the version written in his own language, indicated that each of the interlocutors bound himself only by his own legal order. Certainly, the European powers had versions written in their own language, but in the case of divergence it is the Turkish text that was followed. At the end of the eighteenth century, French, 19

"Diplomatic History as a Field for Cultural Analysis: Muslim-Christian Relations in Tunis, 17001840," 85; Ferdinand Brunot, Histoire de la langue française des origines à nos jours. Tome VIII, Le Français hors de France au XVIIIe siècle (Paris: A. Colin, 1967), pt. 2. L'universalité en Europe.

18

English, and Spanish diplomacy was still resigned to this situation (Windler 2001: 86-87).20 If one is to take ‘in the case of divergence it is the Turkish text that was followed’ literally, French, English, and Spanish diplomacy did, in certain situations and for certain texts, engage in a practice of textual interpretation that functioned according to a language hierarchy in which the Ottoman language was an authoritative source of reference. This would make for a certain degree of conceptual entanglement across the Ottoman/European linguistic divide, but within power relationships that were completely different from the late 19th century. Relationships of conceptual entanglement are seldom (if ever) mutual, as there are usually hierarchies of languages at play, even among vernaculars. In Turkish it is nowadays quite often legitimate to call upon a conceptual interpretation drawn from English.21 Reverse the situation, and the Englishman will probably not see the relevance of a Turkish concept or concept interpretation unless in a very particular context. The practice of re-interpretation and the corpus of authoritative texts available to reference are unevenly stacked, so as to put the Turkish speaker at a disadvantage. To the extent that linguistic communities are bilingual, it always involves some sort of language hierarchy within the community, and those language hierarchies are context dependent. Among Turks English conceptual interpretations are usually more authoritative than say Russian when it comes to most secular affairs, but for religious affairs Arabic would trump every other language as a legitimate source of concept interpretation. Even within languages, there are potential hierarchies between interpretations, and the expectations of the universal validity of some but not others. As Winston Churchill once famously remarked, Britain and the United States are ‘countries divided by a common language’. This is underpinned by such linguistic facts that all Brits (or at least British intellectuals) know that Americans mean something slightly different by the concept of democracy, but not all Americans (or even American intellectuals) know the same fact. Quite simply, they don’t need to know, because the relationship is not symmetric, even between communities within the same language.

20 21

My emphasis and inserts. Though whether it is a mark of social distinction depends on the context.

19

But the potential for diverging interpretations, and thus also potential asymmetries, are not as great as between languages. What is it that travels? The question is, what is it that travels? Texts travel and people travel. But since texts are dependent upon interpretation to give meaning at all, and since the practice of interpretation is an activation of background knowledge, the result of the interpretation of a text is highly dependent on the background knowledge of the interpreter. Where the culture of the interpreter is highly different from that of the writer and his immediate society and audience, the interpretation will almost inevitably be at variance with that of the culture within which it was written. As the missionary-cum-anthropologist Daniel Everett (2008: 243) argues, Language is a by-product of human cognition, rather than a special universal grammar, in conjunction with the constraints on communication that are common to evolved primates (such as the need for words to appear out the mouth in a certain order, the need for units like words for things and events, and so on), the overarching constraints of specific human cultures on the languages that evolve from them. Without someone to interpret the texts travelling with them, it is unlikely that texts will have more than a mere resemblance when they are culturally and linguistically displaced. However, with a dense network of texts and people crossing back and forth, a new frame of reference for the interpretation of texts can emerge. And it did in the Ottoman Empire of the late 19th century, just as it did in most of the internationally interconnected world of High Colonialism and after. Since discourses are relational, what Michel Foucault calls discursive structures do not travel whole, except where two or more people with an ability to use these structures travel from one place to another. Settler societies, such as those in South Africa, Australia and North America, are examples where this has happened. English political discourse travelled more or less intact with the settlers, taking whole discursive structures with them. This is qualitatively different for places such as Turkey, Iran and China, where European political meaning arrived little by little. When studying semantic change as entanglement with European politics, conceptual 20

history has its advantages over Foucauldian discourse analysis, as it is better to trace the semantic shift in concepts than to look at entire discursive structures. If the structures themselves do not travel (at least not alone), then looking at the entire structures makes less sense than focusing on the concepts. Just as discursive structures do not travel whole, so do concepts not travel as concepts. When a concept is introduced into a new linguistic context it is no longer a concept in the sense that Koselleck defines it, namely as a word that is so semantically multi-faceted that it cannot be defined and without which history cannot be written or politics be carried out. I would contend that what ‘travels’ from one language to another when a text is translated or when it is simply moved into a new historical context is individual interpretations. If all translations involve interpretations, what they do in terms of concept formation is to add one concept interpretation at a time. Concepts travel as single interpretations or definitions, at least in the first instance. That instance is empirically very difficult to find and ascertain, but it is nevertheless there as a starting point from which the word then becomes a concept by the addition of more interpretations and the integration of that concept into the political vocabulary of the recipient language. The word thickens into a concept as new interpretations are added. These interpretations may come from both within or without the recipient language, but the source language is often an authoritative source of new interpretations. Only after it has become a concept can it be developed into an institution with its attendant distribution of roles and established practices. Let us see how this played out in the case of the entanglement of the Ottoman concept medeniyet and the French concept civilisation. Part III: The Case Few concepts were as central to both international society and the domestic affairs of 19th century polities as civilisation. This was a concept that all polities that were in contact with Europe had to relate to because of its status in international law, and was a concept that was central to structuring international relations (Bull 1977; Bull & Watson 1984; Elias 2000; Salter 2002). The Ottoman Empire was no exception, and civilisation became an important concept in Ottoman political language from the mid19th century, and initially played a role in the relationship between the Ottoman Empire and European states. In the late 19th and early 20th century it was also put use in structuring relations within the Ottoman Empire (see Makdisi 2002; Deringil 1998). 21

The most pressing concern for non-European polities was that of the ‘standard of civilisation’; polities that were not deemed civilised were according to international law considered land that had no state and which were therefore a free-for-all for colonial expansion (see Gong 1984; Koskenniemi 2010). That is, if the political elite of a polity did not act in a certain prescribed manner and display certain Eurocentrically formulated qualities, they were not considered to have a state.22 Since these qualities were defined by Europeans, it became imperative to deal with how and to what extent one should and could borrow from the Europeans. Umran Following the translation of the Maghrebi scholar Ibn Khaldun’s famous work on history, Muqaddima, from Arabic in the middle of the eighteenth century the Ottoman language had a concept with the interpretation a ‘totality of human habitation’ – namely umran (al-Azmeh 1990: 10). This concept, although in one sense similar to civilisation (and in Arabic later becoming its established translation), lacked much of the semantic polyvalence that civilisation carried. It had no civic connotation and nothing to do with politeness, education or refinement. Umran – the totality of human habitation – was made up of medeniyet (settled/city human habitation) and bedeviyet (nomadism/beduinity) (Ibn Khaldun 1967). As I will go on to show, it was medeniyet that became the established Ottoman and later Turkish translation of civilisation. The etymological origin of medeniyet is the Arabic root m-d-n, from which is derived, among other words, medina (city). Medeniyet can be translated as ‘that which pertains to the city,’ ‘urbanity’ or ‘city culture.’23 Ibn Khaldun argues that the asabiya ‘community spirit’ of the bedeviyet is superior to that of the medeniyet, and thus the nomads are bound to conquer medeniyet, where they will become decadent and lose their asabiya. Nevertheless the luxuries of medeniyet are necessary for increasing the prestige of the ruler and the realm vis-à-vis neighbouring polities. Medeniyet, in Ibn Khaldun’s Muqaddimah, is a necessary evil; the source of both decadence and prestige. Sivilisazyon 22

See Deringil 1998 for an extensive study of how Ottoman statesmen European practices in displaying power on the international arena in the late 19th and early 20th centuries. 23 Ibn Khaldûn (1967) The Muqaddimah: An Introduction to History. Franz Rosenthal (trans.), N.J. Dawood (ed.) Bollinger Series. Princeton, NJ: Princeton University Press, page?

22

When European interpretations of the concept of civilisation were first introduced into Ottoman, they were linked neither to umran nor to medeniyet, but to the French word civilisation, which transcribed into the Arabic script became a neologism in Ottoman (see Baykara 2007: 15). The first known instance of trying to explain the concept in Ottoman was in 1834. Probably not coincidentally, this came after a decade or two of European writings, in English and in French, on the topic of the presence or absence (mostly the latter) of civilisation in the Ottoman Empire. European observers were particularly concerned with the ‘liberty of Greece,’ which had recently been fighting for independence from the Ottoman Empire, and which in European languages was linked to civilisation (see Baykara 2007: 24, 26, 34-35). Writing a letter from Paris, where he was ambassador, the Ottoman statesman Reşid Paşa first introduced the concept ‘civilisation’ into Ottoman and explained it as ‘the education of man and a practicing of orderliness’ (Baykara 2007: 29-30).24 From the very beginning, it was linked to an interpretation having to do with orderliness and the education and refinement of the individual. One key quote used by conceptual historians is Friedrich Nietzsche’s ([1887] 2007: 59) ‘all concepts in which an entire process is semiotically concentrated defy definition; only something which has no history can be defined.’ At this moment in time, sivilisazyon (as the Ottoman transcription of civilisation is written in Turkish today) not only could be, but was in fact defined. This explanation that Reşid Paşa here gave was quite simply a definition, and absent knowledge of the French tradition, there was no other way to define it until it became a more established part of Ottoman vocabulary and came to be interpreted differently. It did not, however, take long before the concept became polyvalent in the Ottoman context as well, although it did by no means come to mean exactly the same as the French civilisation. Three years later (in 1837) another Ottoman statesman, Sadık Rıfat Paşa, argued that a new system had been enforced in Europe since the end of the Napoleonic Wars. This system Sadık Rifat called ‘civilisation’ (sivilisazyon), and he claimed that it was based on the determination to maintain peaceful and friendly relations between states (Mardin 2000: 180).25 This was soon after Sadık Rıfat Paşa had been to Vienna, where he met and spoke extensively with the Austrian 24

‘Civilisation usûlüne, ya’nî terbiyye-i nâs ve icrâ-yı nizâmât husûslarına.’ See Neumann 1999a: 176 fn53. 25 Sadık Rıfat Paşa Muntahabat-ı Âsar, II, 3-4; VII, 67; VIII, 35.

23

Prince Metternich. Both Sadık Rıfat Paşa and Mustafa Reşid Paşa were central to the proclamation of the Tanzimat, and were, along with their mentor, Pertev Paşa, strongly influenced by Metternich and the French orientalist Silvestre de Sacy (Mardin 2000: 177-179; Lewis 1968: 106). Sadık Rıfat linked this concept to a representation that a state flourished whenever its subjects could reap as much as possible of their own labours. This meant that they would have to be protected from arbitrary rule. It was the insecurity that resulted from arbitrary rule that ultimately created wars. Peace and prosperity were mutually dependent, and started with the protection of the subjects. (Mardin 2000:180-181). No individual subject and likewise, no country were created for the sake of states. On the contrary, since states are simply graces of God granted to worldly rulers to protect and safeguard the welfare and well-being of their domains, in the administration of state affairs they have to act in conformity with the rights of their people and the laws of the state; and therefore no arbitrary or despotic act can be observed… Since these matters are reckoned as the fundamentals of the major policies for the civilised nations of Europe, it is an indispensable urgency to endeavour and make utmost efforts so as to prepare the necessary conditions for the Ottoman State (Devlet-i Aliyye) to attain all these (Sadık Rifat Paşa 1837 quoted in Lewis 2000: 132, see also Berkes 1998: 130-131).26 Civilisation, here, was linked primarily to the conduct of states and statesmen. It was not the honing of an individual’s character that ensured civilisation, but the institutionalisation of the relationship between ruler and ruled, as well as between rulers. There may be other, unknown, interpretations introduced between these two, but here we have a second interpretation of the concept. One may say that from this time onwards, the Ottoman concept of civilisation had an internal aspect (that of domestic orderliness and the education of man) and an external aspect (namely relations between states). This latter aspect was implicitly linked with what was then a particular system for the maintenance of (if not peaceful, then at least routinised) 26

The translation is Lewis’s.

24

relations between European Great Powers — the Concert of Europe. Civilisation may have been interpreted in other ways as well, but the Concert of Europe was certainly interpreted as a civilising project; the ‘civilised nations in Europe.’ While it became more complex after this (and probably also was a bit more complex than my presentation here suggests), one particular discursive event was formative of the concept’s continued usage. The incident that seems to have been a key event in institutionalising the concept as part of Ottoman statecraft happened in 1844, when an Ottoman Armenian converted to Islam, and then later converted back. While the state had nothing against conversion to Islam (and indeed encouraged it), converting from Islam to Christianity was an act of apostasy. Since there was a death penalty for apostasy in the Ottoman Empire at the time, the man was sentenced to death. News of this reached the diplomatic community in Istanbul, and the British ambassador (among others) intervened on the man’s behalf. The main argument was that death sentences administered on the basis of conversion was ‘un-civilised,’ something that brought into play the ‘standard of civilisation,’ in that civilised nations do not practice death penalty for conversion (at least not to Christianity). Due to the diplomatic pressure (if not necessarily because they were swayed by the weight of the civilisatory argument), the Porte bowed out and nullified the death sentence (for the events, see Deringil forthcoming). To simplify, the concept civilisation here gained a third politically forceful interpretation. It also became a more established part of Ottoman vocabulary and got linked to Ottoman statecraft, restricting the scope of what was considered legitimate state practice. One could no longer continue certain established practices without coming into conflict with this concept, and ultimately endangering the state’s survival. Medeniyet Some time between 1840 and 1860, the word sivilisazyon was replaced by medeniyet (most commonly in the variant temeddün) as the main translation of civilisation. The two had been brought together by Mustafa Sami Efendi, who was the first to have used medeniyet in this was in 1840 (Palabıyık 2010: 170). In 1860/61 (1277) the influential reformist statesman Ahmet Cevdet Paşa (182227-1895)28 translated the last

27

Richard L. Chambers (1973: 440) claims he was born in 1822 and Christoph K. Neumann (1999: 18) claims 1823.

25

third of Ibn Khaldun’s Muqaddimah. Ahmet Cevdet soon started using medeniyet as a key analytical concept in his very influential twelve volume Tarih-i Cevdet (Cevdet’s History) (1854-1877) (Neumann 1999a: 168), and moreover, used it in a meaning that was similar not only to Ibn Khaldun’s interpretation, but also to the previous sivilisazyon. This increased the concept’s importance in the Ottoman vocabulary, and seems to have been key in replacing sivilisazyon with medeniyet, thus merging the two concepts and creating a polyvalent concept that called upon two traditions, and thus came to mean more than the French civilisation while still lacking some of the semantic connotations that it held in French. Although the Tarih-i Cevdet was commissioned by the Ottoman Encümen-i Daniş (a committee modelled on the Academie Française) (Neumann 1999: 16-18), whose members included some of the most important European orientalists of the time29, it is unclear to what extent this provided Ahmet Cevdet Paşa with opportunities to exchange views on civilisation and further entangle his analysis with European ways of writing history of the time. However, given Ahmet Cevdet Paşa’s curiosity, there is no reason to believe that he did not avail himself of their advice. The Tarih-i Cevdet is greatly influenced by Ibn Khaldun’s Muqaddime, and employs much of the same analytical vocabulary (see Neumann 1999a: 154-156; 168183). It also deals extensively with the French Revolution and the deeds of Napoleon Bonaparte, in ways that cannot be meaningfully analysed outside a frame of reference that includes contemporary European philosophy of history, including the German philosopher Georg Wilhelm Friedrich Hegel. But despite Ahmet Cevdet’s extensive referencing, these are notable in their absence. Nevertheless, from such works as Namık Kemal’s article ‘İbret’ in 1870 and ‘Mukaddime-i Celâl’ in 1875 onwards, civilisation was linked to progress (in Ottoman terakki or terakkiyat).30 Modern civilisation was something that was historically ’ahead,’ and thus organised historical development temporally. None of these authors reference Hegel, but it is clear that his

28

It may be worth mentioning here that Ahmet Cevdet Paşa was a protégé of Mehmed Reşid Paşa, the statesman mentioned above who in 1834 wrote that civilisation was ‘the education of man and the practicing of orderliness.’ 29 The British orientalist Sir James Redhouse, the Italian orientalist Thomas Xavier Bianchi (17831864) and the Austrian diplomat and orientalist historian Joseph von Hammer (later von HammerPurgstall) were members of the Encümen-i Daniş (Berk 1999: 38). 30 Namık Kemal ([1875] 1972) ‘Mukaddime-i Celâl’ in Celâleddin Harzemşah, Istanbul: n.p. ‘İbret’ in İbret, No. 3. Quoted in Şerif Mardin 2000: 405-406.

26

influence is there somewhere, and that this brought intertextually in when interpreting medeniyet. Having mapped out some of the key developments of the entanglement of the French concept civilisation and the Ottoman concept medeniyet in the early to mid19th century, it becomes clear that as the process of conceptual entanglement went on, the two concepts became ever more compatible. They did so by mainly by changes in the interpretation of the Ottoman concept. Establishing the concept of civilisation as an integral part of Ottoman political vocabulary was also an aspect of ‘entering’ international society. By the employing and relating to this concept the Ottomans increasingly used the established standards of legitimising policies on the international arena. A history of the Ottoman concept medeniyet and its semantic field could of course be made much more detailed, but the details provided should be sufficient to show the analytical utility of conceptual entanglement, conceptual compatibility and entangled concepts. Conclusion Conceptual entanglement takes place by the very process of trans-polity and translinguistic intertextual linkages that for their ‘success’ rests upon the very same mutual intelligibility for that it is productive of. My argument here is that this conceptual entanglement between Europe and the Ottoman Empire that started in the early 19th century and which is still ongoing, is a matter of Horizontverschmelzung between Ottoman and European languages, and thus also between their users. So, rather than hermeneutics being merely a matter of method, I am arguing that it is a social process that also takes place outside the scholarly community, and that it takes place all the time (something that does not run contrary to Gadamer’s approach, but has a slightly different emphasis). Another important point is that it was the Ottomans and later the Turks who took turns in European hermeneutic circles, and not the other way around. This is a key point on the relationships of power between Ottomans/Turks and Europeans. Horizonverschmelzung between Ottomans and Europeans rested upon Ottomans reading European texts. Only to an extremely limited extent did Europeans read Ottoman texts. Ottoman understanding of Europe was limited, but from the late 19th century onwards, European understanding of the Ottoman Empire on its own terms was a lot worse. Moreover, it is a matter of the horizon of expectation of the Ottomans and Turks no longer resting upon their ‘own’ historical experience, but 27

rather upon the historical experience of Europeans (as this was represented in text and translated/transmitted into Ottoman). I would propose that this is by no means a unique feature of Ottoman history, but a structural point that has relevance to other polities. I am not saying that it was the same everywhere, but that today, politics is formulated through concepts that have some kind of European past, either as native concepts ‘absorbed’ interpretations from certain key European concepts, or as entire loan words from European languages. It is well known that nothing is more difficult than a dialogue in two different languages in which one person speaks one and the other person the other, each understanding the other’s language but not speaking it. As if impelled by a higher force, one of the languages always tries to establish itself over the other as the medium of understanding (Gadamer 2004: 386). This is not only the case on the individual level and only under these particular circumstances. It seems also to be the case in inter-lingual relationships between polities that enjoy a difference in prestige on the international arena. Meaningful claims to legitimacy are seldom formulated in languages such as Norwegian, Turkish or Polish, but rather in English and French. The external aspect of sovereignty, then, hinges, not on claims formulated in the language of the polity itself, but the language that is established as the acceptable medium in the inter-polity relationship. This need not be French or English universally. Russian is used in Russia’s relations with its ‘near abroad.’ French has lost out except in relations involving its former colonies. Now, domestic political vocabularies are likely to, over time, become semantically synchronised with those of the hegemonic language of inter-polity relations. Prior to the 18th century that was Latin in Europe, but other languages played similar roles in other places, with Persian and Mandarin being among the most important. In the 19th century, this was to a large extent French and in the 20th and early 21st, it has for the most part been English. But this need not be so. Just because most languages IR scholars (and other social scientists, anthropologists excepted) are likely to come into contact with have already been largely synchronised with English, at least in terms of political vocabulary, it does not mean that this situation is given by nature. Conceptual entanglement with English as the privileged language cannot be taken for granted. It came about in particular international power relations and it is on-going. 28

Even though it seems to be operating perfectly at the moment (at least for Englishspeaking polities formulating their claims to legitimacy and conducting their international affairs in English), it may not last forever. Formulating those claims in a Chinese conceptual vocabulary may offer some challenges. References Adler, Emmanuel & Vincent Pouliot (2011) ‘International Practices: Introduction and Framework’ in Emmanuel Adler & Vincent Pouliot (eds.) International Practices. Cambridge & New York, NY: Cambridge University Press. Andersen, Niels Åkerstrøm (2003) Discursive Analytical Strategies: Understanding Foucault, Koselleck, Laclau, Luhmann. Bristol: Policy Press. Andersen, Morten Skumsrud & Iver B. Neumann (2012) ‘Practices as Models: A Methodology with an Illustration Concerning Wampum Diplomacy’ Millennium – Journal of International Studies 40(3): 457–481. Appadurai, Arjun (1996) Modernity at Large: Cultural Dimensions of Globalization. Minneapolis and London: University of Minnesota Press. Ataç, Nurullah (1940) ‘Tercümeye Dair Notlar’ Ulus, 4th Oct., 2. Ataç, Nurullah (1941) ‘Tercümeye Dair,’ Tercüme, 1(6): 505-507. Al-Azmeh, Aziz (1990) Ibn Khaldūn. London & New York, NY: Routledge. Bartelson, Jens (1995) A Genealogy of Sovereignty, Cambridge: Cambridge University Press. Baykara, Tuncer (2007) Osmanlılarda Medeniyet Kavramı. Istanbul: IQ Kültür Sanatı Yayıncılık. Bhabha, Homi (1994) The Location of Culture. London and New York, NY: Routledge. Belloc, H. (1931a) On Translation. Oxford: Oxford University Press. Belloc, H. (1931b) “On Translation,” Bookman 74:32-9, 179-85. Bowden, Brett (2009) The Empire of Civilization. Chicago, IL: University of Chicago Press. Bull, Hedley (1977) The Anarchical Society: A Study of Order in World Politics. London: Macmillan. Bull, Hedley & Adam Watson (eds.) (1984) The Expansion of International Society Oxford: Clarendon.

29

de Carvalho, Benjamin (2006). ‘War hurts: Vietnam movies and the memory of a lost war’, Millennium, 34(3): 951–962. Chomsky, Noam (1975) Reflections on Language, New York: Pantheon Books. Coase RJ. (1937) ‘The nature of the firm’. Economica 4:386–405 Coase RJ. (1960) ‘The problem of social cos’t. Journal Law Econ. 3:1–44 Der Derian, James (1987) On Diplomacy: A Genealogy of Western Estrangement. Oxford: Basil Blackwell Press. Der Derian, James & Michael Shapiro (eds.) (1989) International/Intertextual Relations: Postmodern Readings of World Politics. Lexington, MA: Lexington Press. Derrida, Jacques (1976) Of Grammatology. Translated by Gayatri Chakravorty Spivak Baltimore & London: Johns Hopkins University Press. Derrida, Jacques (1978) Writing and Difference. Translated by Alan Bass. Chicago, IL: University of Chicago Press. Derrida, Jacques (1985) ‘Des Tours de Babel,’ Joseph F. Graham, (ed.) Difference in Translation. Ithaca, NY: Cornell University Press. Doty, Roxanne Lynn (1997) ‘Aporia: A Critical Exploration of the Agent-Structure Problematique in International Relations Theory’, European Journal of International Relations 3(3): 365-392. Dunne, Tim (1998) Inventing International Society: A History of the English School. St. Antony’s Series. Basingstoke: Macmillan Press. Eco, Umberto (2000) ‘Introduction’ in Lotman, Yuri (2000) Universe of the Mind: A Semiotic Theory of Culture. Bloomington & Indianapolis, IN: Indiana University Press. Emirbayer, Mustafa (1997) ‘Manifesto for a Relational Sociology’ The American Journal of Sociology, 103(2): 281-317. Everett, Daniel (2008) Don’t Sleep, There Are Snakes. London: Profile Books. Fearon, James D. (1998) ‘Bargaining, Enforcement, and International Cooperation’ International Organization 52(2): 269 - 305. Foucault, Michel (1970) The Order of Things: An Archaeology of the Human Sciences. New York, NY: Pantheon books. Foucault, Michel (1972) The Archaeology of Knowledge. New York, NY: Pantheon Books.

30

Foucault, Michel (1977) Discipline and Punish: The Birth of the Prison. New York, NY: Pantheon Books. Gadamer, Hans-Georg ([1960] 2004) Truth and Method. London & New York, NY: Continuum Books. Giddens, Anthony (1990) The Consequences of Modernity. Cambridge: Polity Press. Gjesdal, Kristin (2009) Gadamer and the Legacy of German Idealism. Cambridge & New York, NY: Cambridge University Press. Gong, Gerrit W. G. (1984) "The Standard of "Civilization" in International Society. Oxford: Oxford University Press. Harris, Richard L. (1995) “The Global Context of Contemporary Latin American Affairs”, in S. Halebsky and R. L. Harris (eds.), Capital, Power, and Inequality in Latin America. Boulder: Westview Press. Holm, Hans-Henrik and Georg Sorensen (1995) (eds.), Whose World Order? Uneven Globalization and the End of the Cold War. Boulder: Westview Press. von Humboldt, Wilhelm ([1836] 1971) Linguistic Variability and Intellectual Development. George C. Buck and Frithjof A. Raven (tr.) Coral Gables: University of Miami Press. Ifversen, Jan (2011) ‘About Key Concepts and How to Study Them’ Contributions to the History of Concepts. 6(1): 65-88. Jackson, Patrick T. (2006) Civilizing the Enemy: German Reconstruction and the Invention of the West. Ann Arbour, MI: University of Michigan Press. Jackson, Patrick Thaddeus & Daniel Nexon (1999) ‘Relations before States: Substance, Process and the Study of World Politics’ European Journal of International Relations 5(3): 291-332. Kayaoğlu, Turan (2010) Legal Imperialism: Sovereignty and Extraterritoriality in Japan, the Ottoman Empire, and China. Cambridge: Cambridge University Press. Keohane, Robert (1984) After Hegemony. Princeton, N.J.: Princeton University Press. Koselleck, Reinhart (1982) ‘Begreiffsgeschichte and social history’ in Economy and Society 11(4): 409-427. Koselleck, Reinhart (1985) Futures Past: On the Semantics of Historical Time. Cambridge, MA: MIT Press.

31

Koselleck, Reinhart (1987) ‘Historik und Hermeneutik’ in Reinhart Koselleck and Hans-Georg Gadamer (eds) Historik und Hermeneutik. Heidelberg: Carl Winter Universtitätsverlag. Koselleck, Reinhart (1988) Critique and Crisis. Oxford: Berg. Koselleck, Reinhart (1989) ‘Linguistic change and the history of events’ in Journal of Modern History 61: 649-666. Koselleck, Reinhart (1996) ‘Response to Comments’ Harmut Lehmann & Melvin Richter (eds.) The Meaning of Historical Terms and Concepts: New Studies on Begriffsgeschichte. Washington D.C.: German Historical Institute. Koselleck, Reinhart (2011) ‘Introduction (Einleitung) to the Geschichtliche Grundbegriffe’ Contributions to the History of Concepts 6(1): 7-25. Koskenniemi, Martti (2010) The Gentle Civilizer of Nations: The Rise and Fall of International Law 1870-1960. Cambridge: Cambridge University Press. Krebs, Ronald R. & Patrick Thaddeus Jackson (2007) ‘Twisting Tongues and Twisting Arms: The Power of Political Rhetoric’ European Journal of International Relations 13(1): 35-66 Kristeva, Julia (1980) Desire in Language: A Semiotic Approach to Literature and Art. New York: Columbia University Press. Lake, David (2009) ‘Hobbesian Hierarchy: The Political Economy of Political Organization’ Annual Review Political Science 12( ):263–83 Leonhard, Jörn & Ulrike von Hirschhausen (eds.) (2011) Comparing empires: encounters and transfers in the long nineteenth century. Schriftenreihe der FRIAS School of History / edited by Ulrich Herbert & Jörn Leonhard. Lotman, Yuri (2000) Universe of the Mind: A Semiotic Theory of Culture. Bloomington & Indianapolis, IN: Indiana University Press. Lotman, Yuri (2004) Culture and Explosion. Marina Grishakova (ed.) & Wilma Clark (trans.). Berlin/New York: De Gruyter Mouton. Luhmann, Niklas (1995) Social Systems. Stanford: Stanford University Press. Mitchell, Timothy (1991) Colonising Egypt. Berkeley & Los Angeles, CA: University of California Press. Neumann, Iver B. (1996) ‘Self and Other in International Relations’ in European Journal of International Relations 2(2): 139-174. Neumann, Iver B. (1999) Uses of the Other: The East in European Identity Formation, Minneapolis: University of Minnesota Press. 32

Neumann, Iver B. & Jennifer Welsh (1991) ‘The “Other” in European delf-definition: An addendum to the literature on international society’ Review of International Studies 17: 327-348. Neumann, Iver B. (2002a) ‘This Little Piggy Stayed at Home: Why Norway is not a Member of the EU’ Lene Hansen & Ole Wæver (eds.) European Integration and National Identity: The Challenge of the Nordic States. London & New York, NY: Routledge. Neumann, Iver B. (2002b) ‘Returning Practice to the Linguistic Turn: The Case of Diplomacy’ Millennium 31(3): 627-651. Nietzsche, Friedrich ([1887] 1996) On the Genealogy of Morals, trans. by Douglas Smith. Oxford: Oxford University Press. Palabıyık, Mustafa Serdar (201) "Travel, Civilization and the East: Ottoman Travellers' Perception of 'The East' in the Late Ottoman Empire." PhD diss., Middle East Technical University. Parker, Walter C. (2005). Teaching Against Idiocy, Bloomington: Phi Delta Kappan. Pernau, Margrit (2012) ‘Whither Conceptual History? From National to Entangled Histories,’ Contributions to the History of Concepts 7(1): 1–11. Putnam, Robert D. (1988) ‘Diplomacy and Domestic Politics: The Logic of TwoLevel Games,’ International Organization 42(3): 427-460. Ramberg, Bjørn & Kristin Gjesdal (2005) ‘Hermeneutics”, Stanford Encyclopedia of Philosophy. Salter, Mark B. (2002) Barbarians and Civilization in International Relations. London: Pluto Press. Saussure, Ferdinand de (1983) Course in General Linguistics. translated and annotated by Roy Harris. London: Open Court Publishing. Şemseddin Sami (1317 [1901/1902]) Qāmūs-i Türkī. Dersaʿādet [Istanbul]: Aqdām Maṭbaʿası. Skinner, Quentin (1978) The Foundations of Modern Political Thought. Cambridge & New York, NY: Cambridge University Press. Skinner, Quentin (1981) Machiavelli. New York, NY: Hill and Wang. Skinner, Quentin (1996) Reason and Rhetoric in the Philosophy of Hobbes. Cambridge: Cambridge University Press. Swidler, Ann (1986) ‘Culture in Action: Symbols and Strategies’ American Sociological Review 51(2): 273-286. 33

Swidler, Ann (2003) Talk of Love: How Culture Matters. London & Chicago, IL: Chicago University Press. Watson, Adam (1992) The Evolution of International Society. London & New York, NY: Routledge. Werner, Michael and Bénédicte Zimmermann (2006) ‘Beyond Comparison: Histoire Croisée and the Challenge of Reflexivity,’ History and Theory 45: 30-50. Williamson OE. (1975) Markets and Hierarchies: Analysis and Antitrust Implications. New York, NY: Free Press Williamson OE. (1985) The Economic Institutions of Capitalism: Firms, Markets, and Relational Contracting. New York, NY: Free Press Windler, Christian (2001) ‘Diplomatic History as a Field for Cultural Analysis: Muslim-Christian Relations in Tunis, 1700-1840’ The Historical Journal 44(1): 79-106. Wittgenstein, Ludwig (1975) On Certainty. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by Denis Paul and G. E. M. Anscombe. Oxford: Blackwell Publishing. Wittgenstein, Ludwig (2009) Philosophical Investigations. Revised 4th edition (P.M.S. Hacker & Joachim Schulte, eds.). Chichester: Wiley-Blackwell. END

34