Approaching religion through linguistics Methodological thoughts on a linguistic analysis of ‘ religion ’ in political communication

‘What if our language does not simply mirror or picture the world but instead profoundly shapes our views of it in the first place?’ Frank Fischer and John Forester posed this question at the opening of their book, The Argumentative Turn in Policy Analysis and Planning (1993). In the meantime almost 20 years have passed and it is unlikely that a single academic has not been confronted with this sentence in one way or the other. Still, applying a linguistic focus to policy research has not become a mainstream option. Policy fields, their rationalities, and structures are for the most part taken for granted and most effort is put into explaining and improving policies rather than asking critical questions about truth, power and language games.


Approaching religion through linguistics
Methodological thoughts on a linguistic analysis of 'religion' in political communication stephanie garling 'W hat if our language does not simply mirror or picture the world but instead profoundly shapes our views of it in the first place?' Frank Fischer and John Forester posed this question at the opening of their book, The Argumentative Turn in Policy Analysis and Planning (1993). In the meantime almost 20 years have passed and it is unlikely that a single academic has not been confronted with this sentence in one way or the other. Still, applying a linguistic focus to policy research has not become a mainstream option. Policy fields, their rationalities, and structures are for the most part taken for granted and most effort is put into explaining and improving policies rather than asking critical questions about truth, power and language games.

Keeping the categories unsettled
But what does it mean to be critical? Michel Foucault (1997) and Judith Butler (2002) differentiate between the notions of criticism and critique. Criticism is understood as making evaluative judgments about social goals and conditions. It aims to call into question the foundations of societies and to denaturalise social and political hierarchies. A policy analysis focussing on improvement and fault-finding, formulating proposals and deeper knowledge of the object can fulfil the preconditions of being critical. It may be based on moral objectives such as justice, empowerment and solidarity and may include an analysis of dominance. A critical approach applying this focus would give advice on how to influence the roles and positions of the stakeholders. It would name and categorise structural deficits. The main task of such an analysis is to find solutions to identified problems.
Critique, on the other hand, is understood as a practice in which questions are posed which deal with the limits of our most sure ways of knowing.
Critique is not generally definable, as this is dependent on its objects, but its objects will in turn define the very meaning of critique. Critique can therefore be understood as a problematisation of effects on truth. Questioning 'the exhaustive hold that such rules of ordering have upon certainty' (Butler 2002) is the task of the critique. What are the rationalities, programmes and logics that make the system acceptable? These limits can only be known if certainties are put at risk. Foucault speaks of a labyrinth into which one ventures and in which one can lose oneself while undertaking the endeavour (Foucault 2005: 19). Such a perspective has therefore no reassuring answers to give, as it cannot be known if the Other, the Unspeakable will produce a better world. But thinking otherwise and focusing on agenda-setting, inclusion and exclusion, relations of power and selective attention has the effect of maintaining unsettled categories and leaving space for antagonism -both of which are understood to be necessary aspects of the political (cf. Mouffe 2005). This article will show how the perspective of critique is chiefly left aside when focussing on 'religion' in empirical research. Therefore a method will be put forward that allows taking on the position of a critique, making it possible to focus on rationalities and logics that are inherent to the current usage of 'religion' .

Approaching religion through turning the term around
Even if all social scientists know that the world is socially constructed, as Steve Bruce stated in his evening lecture at the SOCREL Conference in Birmingham in 2011, only epistemological shifts allow us to open up to the reflexivity and problematisation that modes of critique, as described above, can offer in comparison to criticism.
Constructivist and interpretative understandings of social phenomena lie at the heart of these shifts and call for a questioning of the 'great narratives' such as modernisation and development. Within the sociology of religion, 'secularisation' has been one of these grand themes. As stated in the editorial of this special issue, two main ways of looking at this phenomenon have developed. The one you might call the mainstream approach has turned towards religion, whereas the other is more concerned with turning the term around. Most research within the first field -the religious turn approach -focuses on questions of secularisation as a universal phenomenon: is secularisation on the increase or decrease and should secularisation theory be abandoned or not (cf. Berger 1999, Davie 1999, Berger and Davie 2008, Pollack and Olson 2008? Within this perspective the definition of religion is a substantial compon ent and it interrelates with three fields of academic controversy: firstly the correct position for such a definition within the research process (Weber 2005, Eliade 1958), secondly the width of such a definition (Durkheim 1915, Knoblauch 2000 and thirdly the question as to which essential features are the ones which outline the terms of religion appropriately and should therefore be included in such a definition (Smith 1991, Geertz 1966, Asad 1999. To some extent new approaches in this field, which is mostly based on measuring and comparing, are acknowledging secularisation as being multiple -like the modern ities (Casanova 1994, Eisenstadt 2000 -and trying to find shared characteristics as well as distinctions in-between and with different cultural backgrounds (Dobbelaere 2002, Wohlrab-Sahr andBurchardt 2012). Even if all these approaches take on so very many different perspectives and try to answer a wide range of research questions, what connects them is the fact that they are 'all dependent on a substantial concept of religion, in the sense that they assume -implicitly or explicitly -a substance of the category which can be traced and compared across time and space in order to measure the size and shape of the "thing" ' (Johansen 2011: 281).
Foucault characterises this kind of research as a 'virtual deductive edifice' (Foucault 2005: 62). He describes that within this edifice, terms are being arranged like stones which, step by step, constitute the edifice itself. They create disciplines and provide coherence and continuity. One of their great strengths is differentiation. That is how they call for clear inclusion and exclusion. Terms are being ordered through definitions and in being applied they likewise create order. They can be seen as descriptive and prescriptive at the same time.
Taking these aspects of terms for granted without focussing on their creativeness in constructing the world we live in is not the way research could be done when a position of critique is aimed for. But what would be an alternative to making further contributions to this edifice? Foucault sees this alternative in describing the organisation of the field of statements where the concepts appear and circulate: le jeu des concepts (Foucault 1994: 75). Focussing on this game of terms, as he calls it, allows for a broader perspective. Conceptual systems and their organisation and dispersion become central to the analysis. Adopting this perspective is what is meant by the title of this chapter; 'Turning religion' . It aims to offer analytical steps for a research that doesn't want to (directly) participate in the actual 'religious turn' but wants to observe its implications -turning the term 'religion' around and looking at the formation of its concepts. This will offer a research position which is a critique rather than a criticism. The researcher remains an observer rather than an adviser.
As it is described metaphorically in the editorial of this special issue, this perspective is taken on by a, so-to-speak 'smaller brother' . 'He' is not just turning towards 'religion' but turns the term around so that its very existence gets questioned, reduced to pure 'science fiction' and to something that 'merely distorts the field' and should be abandoned (Fitzgerald 1997: 93), or at least, he demands that its theoretical and practical implications be reflected upon (McCutcheon 2000: 132). These approaches focus on power, understood as something productive. Power relations are therefore seen as something that can be found everywhere where there is society. 'Religion' and its linguistic usage are seen as one aspect of these power games.
Still, most of these approaches get trapped as, with their extreme position of criticism rather than critique, they essentialise no less than the approaches they criticise (McKinnon 2002: 79). Besides, by solely concentrating on mere criticism makes people forget that 'religion' is being used by social actors every day and that 'non-academics make far more use of it, and they do not use it arbitrarily' (McKinnon 2002: 76). Avoiding this trap and taking the general usage of 'religion' into account makes it necessary to shift the focus towards the operative ideas (Bell 2006: 28) which are being built and which structure the 'edifice' of today's language. So the formation of the category or concept in itself should be the focus of empirical research , rather than just criticising its (mis)use within power relations.
Accordingly, it can be said that for such a wider understanding of 'how the idea of religion operates in the wider contemporary world' (Beyer 2003: 158), it is necessary to leave the described 'brothers' quarrel' behind and to develop new tools to turn the term around. Otherwise, research amounts 'to no more than an awareness that non-experts often use and have an understanding of the notion of religion' (Beyer 2003: 155). In the following this article suggests a term-based textual analysis as one possible tool with which to fill this gap and open up a discourse for critique rather than criticism.

Methodology -method -data
Consequently, this article is a text on method. It suggests one way in which -when asking the research questions in the epistemological way described above -the empirical data chosen could be approached. This method of a term-based textual analysis could be applied to multiple forms of other empirical backgrounds but as every method needs examples to be illustrative, comprehensible, and open to scrutiny, the following characterisation of such an analysis is based on examples from the policy field of German and Swiss development aid. It contains segments of a whole project focussing on the understanding of 'religion' in foreign aid and is based on programmatic texts from this policy field, such as speeches, training handbooks, conference invitations, and presentations. All materials are produced by political actors working in this policy field and are published on the internet and/or via, for example, leaflets, brochures and handouts. As stated before, the aim of these examples is not to shed light on the usage of 'religion' in foreign aid, but to assign the method of a termbased textual analysis. Therefore, the given examples are not interrelated, but are chosen because they indicate the benefit that a term-based textual analysis has for focussing on the formation of the concept of 'religion' . They facilitate an exposure of the inherent knowledge circulating within this process and reveal certainties and rationalities that keep the system running and open them up to a critique.

Term-based textual analysis
[T]he way that we use words is a significant index of how we think. Also, more actively, it is a significant factor in determining how we think. To understand the world, and ourselves, it is helpful if we become critical of the terms and concepts that we are using. (Cantwell Smith 1996: 16.) Two analytical methods For the analysis of 'terms' it is possible to differentiate between two schools. These schools come from different academic disciplines, ask different questions and form different ways of approach. Still, both can help to illustrate the so-called 'language games' and their intertwining trails. The first school -arising from Wittgenstein -can be spoken of as actionoriented approaches. They take linguistic utterance as communicative acts and therefore as actions. Subjects are intentionally speaking according to their mentality and consciousness. Therefore, contexts play an important role for understanding the language game. The second school arises from the so-called structureoriented approaches, which focus on formation rules. Coming from Saussure they are taking language as a system which can be analysed according to its structure. For this school the meaning of linguistic signs exists only in relation to other signs. Identities and differences make the mechanisms of language (Saussure 2006: 107). The subject is not seen as autonomous and sovereign. Therefore, linguistic practices are analysed in detachment from the activity of the subject: recipient, author, audience are not the focus of analysis but merely semantic structures within the discourse. Expressed through the metaphor of a game, we could say that the first school focuses on the player, the venue and the season, whereas the second puts the focus on the rules of the game.
But what does this mean for a text analysis which focuses on the use of a specific term -such as, in our case, 'religion' -and aims at a critique rather than criticism?
Avoiding causal relations through structure Structure-oriented approaches offer a way to avoid singular, context-based causal relations. No matter what kind of context (situational or historical), only presumptions of the author's intention, recipient's reactions or text functions enacted by assumed causal ities lead to plausible statements. But causal relations are often artificial and superficial (Werner and Zimmermann 2006: 47) and therefore constitute a gateway for normative and essential ising conclusions. Therefore, an analysis which aims at critique rather than criticism needs to be based on a structure-oriented approach. It needs to focus on the semantic structure rather than the text's context and function. For this endeavour, texts are not seen as homoge neous, unmediated and particular (Foucault 2005: 27). Instead, it becomes necessary to question their unity. Its unity needs to be suspended with the overall aim of restructuring it and retrieving a unity that turns the previous one around. A useful means of effecting this restructuring lies in the text itself, as it consists of three different layers: the thematic, argumentative and grammatical levels (cf. Marxhausen 2010: 229-34). At the thematic level overall topics of the text and its outlines are set and the themes unfold. At the grammatical level, different linguistic means are used to make these themes understandable and plaus ible to the audience. And finally at the argumentative level convincing and evaluative strategies as well as presumptions are to be found.
The analytical differentiation into these three layers of a text helps to fulfil the described operation of questioning and restructuring texts in the way of a critique. They will therefore frame the following sections with the purpose of showing the usefulness of such an analysis, by visualising some cases of inherent knowledge that lie within the semantics surrounding the term 'religion' in political communication. The following passages in the first instance illustrate different linguistic means (topoi, metaphors and verbal strategies) which can be helpful for such an analysis. And secondly, examples are integrated that derive from the research project mentioned above which deals with programmatic texts on foreign aid and development. 1 Text semantics The analysis of topoi gives a guideline by which to explore the first layer; the thematic level of a single text or a whole corpus, as it offers the chance to disclose rule-complying enunciations. They can reveal frames and patterns by which meaning is structured. Topoi are to be understood as plausible argumentation schemes which belong to the premises. They connect the argument(s) with the conclusion as they justify the transition from the argument to the conclusion. Topoi can be seen as content-related warrants or conclusion rules (cf. Reisigl and Wodak 2009: 110-11). They can be obtained through paraphrasing reduction or through deducing the topic of the text(s).
An example from the research project on the understanding of 'religion' in foreign aid is the topos 'ambivalence' , which is closely interlinked with the current use of the term religion in political communication. Thus a sense of the reality of religion being ambivalent is created and it seems obvious that religion has ambivalent features: It can be good or bad. It can serve stability just as it can be misused. It can be helpful in creating understanding and at the same time it can be a trigger for wars. The topos 'ambivalence' , which is explicitly applied in some of the texts (e.g. Ma1: 18; Ma2: 10-13; Ma3: 9) connects these arguments with the conclusion that 'we' -the actors in foreign aid -need to be attentive about (Ma4: 13), or carefully monitoring (Ma5: 29) 'religion' .
Looking for different motives and features that arise and are repeatedly articulated, topoi still don't need to be expressed as explicitly as is done with religion being ambivalent. But the interrelation can also be reinforced through denying it or describing it as 1 The materials used to illustrate the theoretical thoughts derive from a research project on understandings of 'religion' in German, Swiss and Austrian foreign aid. All of them are programmatic texts from this policy field, including speeches, training handbooks, conference invitations, and presentations. They were produced by political actors engaged in foreign aid and were published between the late 1990s and 2009 on the internet and/or via leaflets, brochures, handouts etc. The original language of the texts is German. The English translations used in this article were done by the author. something opposite. An example of how a topoi, in this case 'Islam and terrorism' , can be supported, even while denying the connection of the two terms, is instanced in the following sentence: 'Terrorism shouldn't be named in the same breath as Islamism' (Ma4: 48). This statement does precisely what is supposed not to be done. Even by challenging the topos of Islam and terrorism their interrelation is reinforced. Named after a book by George Lakoff (2004), this effect could be called the 'Don't think of an elephant' effect. The reader or listener does exactly what he/she is asked not to do, merely by being asked not to do it and the picture of an elephant -or the inter-linkage of Islam with terrorism -is strengthened.
As we have already seen on the basis of these few examples, topoi can occur implicitly or explicitly, but they always set the main terms of argumentation and frame the limits of what is assumed to be true. They give an overall structure of what is to be said and what stays aside and they are therefore an important aspect of a structure-oriented analysis which aims to retrieve knowledge and make it approachable for critique.
At the grammatical level metaphors are a very interesting and often-used device as they have a strong impact on our perception of the world. Like topoi, they stabilise lines of argumentation and provide justifications. They are not just purely aesthetic ornaments but are often used to arouse emotions (Kienpointner 2005: 229). For example, in the metaphor of 'religion' having 'a dark side' (Ma6: 38), a feeling of uncertainty and discomfort is conveyed. There has to be a light side too, but caution is necessary as things can change suddenly and the negative and destructive parts could manifest. Metaphors create connections between issues, reduce the complexity of complicated topics and conceal that of others. Ordinary conceptual systems are fundamentally metaphorical (cf. Lakoff and Johnson 1980). In the case of the 'dark side' metaphor a picture arises which suggests that 'the side we see is only one aspect of this' (Ma2: 27) and that there is an element of the unknown, something that causes uncertainty.
Metaphors play 'a central role in defining our everyday real ities' (Lakoff and Johnson 1980: 3). They are able to transfer something abstract and unknown into a more familiar, concrete and mater ial idea. They may visualise something which doesn't actually exist. These pictures drawn with language are very effective in building up identification and help to evaluate and prescribe behaviour. Another example from the material of the communication in foreign aid consists of two opposing metaphors which are both used to describe religion and its features. One describes religion as 'deeply rooted' (Ma6: 16) and 'deeply anchored' (Ma7: 22). The second uses the metaphor of religion being 'not cemented' (Ma6: 12). Thus, on the one hand, religion is not solid, but rather flexible and mobile. It is not a kind of binding material with which to build a house, like cement, but something that is constructed, that can be influenced and interpreted. On the other hand, the use of the word religion in combination with terms like anchor and root creates a picture of something that lies in depth in the groundwater or earth and which sustains, supports, and persists. A deeply intrinsic understanding of religion is supported and further associations are stimulated. For example in the case of the root, understood as the part of a plant that provides for other plant components above, the question arises: What might these parts of the 'plant' be? What is it religion provides for? Instantly more ideas come up and are interconnected with the issue without being directly mentioned within the text itself. Consequently, metaphors have the potential to highlight specific aspects of a matter and set others aside. Regardless of whether a religion is described as fluid or solid, as constructed or intrinsic, both kinds of metaphor are ontological, giving entity and substance to the notion of 'religion' . It becomes possible to refer to it, categorise, group and quantify it.
Metaphors create dimensions in which it is pos-Byzantine Metaphor for the Soul and Death. Photo by Ken Fairfax.

Wikimedia Commons
sible to speak about the immaterial and also make it possible to reason about 'religion' in one way or the other (cf. Lakoff and Johnson 1980: 25). The analysis of metaphors surrounding the term 'religion' therefore constitutes a very interesting element of a textual analysis which aims at retrieving inherent knowledge in the communication regarding it. At the argumentative level verbal strategies are a crucial means. Verbal strategies can be understood as procedures for persuasion and argumentation (cf. Salamun 1988: 11). Verbal strategies are all part of the metaphor 'argument is war' (Lakoff and Johnson 1980: 4). They are often referred to when speaking of the means for winning a semantic fight or conquer terms (cf. Grinth 2002: 68). Verbal strategies follow the idea that achieving victory is a question of how to say things. They often work with presumptions and isotopes and can be emotive, evaluative or interrogative and they can appear in the forms of a statement, claim, criticism, appeal or argument.
The statement, 'Religion is a generally-acknowledged moral attitude' (Ma8: 9), is one example from the project on understandings of 'religion' within foreign aid using this means. The sentence indicates very directly what is supposed to be true -in this case that religion is a moral attitude. As this is a generally acknowledged fact, it would be odd to believe something different. The phrase therefore bestows more plausibility and prominence upon the statement. It marginalises other opinions and confers a high status on religion in combination with morality.
Another example of how verbal strategies are applied is: 'In many respects religious belief forms the behaviour of human beings' (Ma9: 25). The nominalisation of religious belief leads to the plausibility that it can form something. It can be active. The perspective that only human beings are able to form religious beliefs is dismissed and religious beliefs are being essentialised with the same consequences as are mentioned in the context of the ontological metaphors above; religion becomes an entity which can be referred to, which can be categorised and therefore in the logical next step, quantified.
The following example particularly presupposes a great deal of inherent knowledge: 'The coming into contact of different religions leads to head-shaking' (Ma10: 28). It is firstly a prerequisite that there are different religions and that they are able to get into contact with each other. Again, an ontological understanding of 'religion' is reinforced. At the same time, it is assumed that readers associate the same idea and feeling with the shaking of heads. This becomes dif-ficult if -as in Romania and Bangla desh -shaking your head means that you agree with something. But what this sentence refers to is, in fact, disagreement. The statement that disagreement arises when different religions meet is therefore being naturalised and reinforced with this sentence. It seems to be a normality.
Another example of a verbal strategy which is quite common in the context of religion is in-group referencing. Religion as a western construction (Ma2: 8) or described in 'eurocentric' (ibid.) terms are two examples of this. In-group references secure not only the interpretational sovereignty, but also deprive the out-group -in this context non-western -of the capacity of a sovereignty of use (cf. Grinth 2002: 68). Even when acknowledging or criticising the dominance of the in-group, its power is still reinforced and its members are left at the centre of knowledge and its construction.
A similar means are new word creations -neologisms -which are also a very popular verbal strategy in the semantic field of 'religion' . One of the most accepted examples of this is the term world religion. But also terms connected to each other via hyphens, such as 'religious-fundamentalist' (Ma11: 31), 'political-religious' (Ma12: 47) or 'Maya-Spirituality' (Ma13: 52) are examples of neologisms which are well-known. New creations of this kind make it easy to combine or connect two issues that were previously standing on their own. Additionally they offer the chance not only to fill these new terms, but even to coin them. For example, in the case of the term 'world religions' , Tomoko Masuzawa shows how through the invention of the term in a time of transformation, control and order in the international system have been re-established (Masuzawa 2005: 41). As we have seen, the ways in which arguments are delivered also carry a lot of inherent knowledge about the term 'religion' . With a focus on the argumentative level -besides the focus on the grammatical and thematic level -it becomes possible to retrieve this knowledge, make its structures visible in discourse and open up discussions.
Context -the field of concomitants As stated before, a term-based textual analysis focuses mostly on the text structure and less on its context. Still, context is, of course, not completely left aside during interpretation. But to avoid causal constructions on the basis of the research's background knowledge, context factors are only integrated if they are explicitly named in the texts themselves. Michael Werner and Bénédicte Zimmermann (2006) refer to this procedure as 'pragmatic induction' . They argue that as elements can be repositioned in relation to one another and are not taken as generic, and having a pre-established nature, it helps to 'get away from the external often artificial nature of the context' and 'the trap of essentialism and its overly static categories' (47). Pragmatic induction would therefore offer a perspective to integrate a reflection on the principles governing the definition of terms and its contexts (ibid.).
Foucault refers to these statements 'which are active among the statements studied' (Foucault 2005: 64) as the 'field of concomitants' and describes them as serving four different functions: analogical confirmation, general principle or accepted premises, transferable models and higher authority (ibid.). Thus, statements in the field of concomitants are all more general then the actual studied statements. They can also concern and belong to completely different domains of objects.
This description of Foucault indicates one way in which to incorporate context factors in the analysis through pragmatic induction: The analysis of topoi, as described before, can be one possible way of in-tegrating context into a structure-focussed approach and avoiding its often artificial nature.
To make this argument more illustrative, one example which can be taken from the empirical material of the research project on foreign aid is that of the frequently deployed references to the attacks on the World Trade Centre on 11 September 2001. This event and its assigned consequences can be seen as a topos that structures the argumentation and therefore serves as a principle with higher authority in the way Foucault has described it. These references therefore indicate one field of concomitants surrounding the term 'religion' . Many of the publications analysed in the project are part of large special programmes set up as investigations of the attacks and are described as 'considerations of security and the preservation of democratic values having priority' (Ma11: 5), marked by suspicion and uncertainty. ' At least since then it has become clear how little we know about Islam' (Ma4: 13). It is argued thet new necessities have developed dealing with other religions especially Islam.
Through being part of this field of concomitants, 'religion' becomes closely interlinked with the war against terrorism, fragmentation and fundamentalism. As it is directly pointed out in one publication: '9/11 gave rise anew to a need to deal with the issue of "religion" ' (Ma4: 13) or as expressed elsewhere: 'With 9/11 an awareness of the necessity to deal with "religion" has evolved ' (Ma14: 24).

Entanglements
With these and other more or less vague references, 'religion' and '9/11' are tied more closely together and it seems to have become a reality that both issues are combined causally. It seems to prove the argument right that the ambivalent nature of religion has once more shown its dark side. And it is only logical to be cautious about it.
A discursive knot -understood as a 'statement where several discourses are entangled' (Jäger and Maier 2009: 47) -has thus developed. The theme under consideration can be linked to one or several others at the same time. Thus, in one text several fragments can be contained or 'entangled' (Jäger 2001: 47). The tools of a term-based textual analysis described here and focussing on the three different levels of text (thematic, grammatical and argumentative) offer the chance to extrapolate the themes which are active in communication, as well as their formation rules, connections, and inter-linkages. They make it possible to question the text's unity and offer a way to deconstruct and later restructure texts. The description of a knot such as that used by Siegfried Jäger and Florentine Maier therefore makes a fitting metaphor for this process of discourse analysis, as it visualises a line which gets tangled (through the differentiation on various text level with a wide range of analytic al means), without abandoning its inner structure. It symbolises internal consistency without denying conflicting, fractured and uneven parts. Knots can be unravelled, their unity can be questioned, and they can be re-fastened in different ways. Every knot is different depending on the material(s) used and the person knotting it together. The knot can symbolise antagonism as well as harmonic parallelism. And last but not least, a knot needs patience. One must begin with its small (linguistic) strands and with their interpretation in order to make shifts, fractions and transformations as well as transmissions and continu ity visible. This article aims to encourage this kind of analysis and indicates how much inherent knowledge there is, when a closer look is taken at the discursive practices that surround the term 'religion' in political communication. It has indicated tools for analysing such complex forms of communicative action and gives room for its analytical potentiality to open up to further critique in terms of a problematisation of the certainties and rationalities that keep the system running.  Stephanie Garling is currently working at the GIGA German Institute of Global and Area Studies in Hamburg, Germany. In 2000-7 she studied political science, law and economics at the University of Leipzig. Her PhD thesis dealt with the construction of religion in foreign aid. In addition to this, her main research interests are questions of identity, critical development theory and discourse analysis. Her last publication was a volume edited together with Simon W.