Ronny Scholz

Towards a post-material prosperity?
An analysis of legitimising narratives in German crisis discourses from 1973 and 2008.1

Résumé | Plan | Texte | Bibliographie | Notes | Citation | Auteur(s)

Résumé

L’article compare des récits de la première crise pétrolière (1973) et de la crise financière récente (2008) dans la presse allemande. Avec un intérêt particulier pour la formule « la croissance engendre de la prospérité », l’étude combine des méthodes quantitatives et qualitatives : les narrations repérées à l’aide d’outils textométriques sont ensuite interprétées au niveau textuel à l’aide de l’analyse des topos et de la théorie post-structurale de Laclau et Mouffe.

Abstract

In this paper I compare narratives representing the first Petrol Crises (1973) and the recent Financial Crisis (2008) in the German press. With a specific focus on the mantra-like phrase ‘growth generates prosperity’ I propose a mixed methods approach to large text corpora. The narratives unveiled with quantitative methods are interpreted on the text level applying topos-analysis and Laclau and Mouffe’s post-structuralist theory of hegemony.

Table des matières

Introduction

1The public debate on the ‘Financial Crises 2008/2009’ has shown how the economic system of capitalism or its German variant, the social market economy, has been put under legitimising pressure throughout such periods. This paper compares two crisis discourses in the German press in this regard with corpus-linguistic methods. In particular I shall be focussing on the first Petrol Crisis (1973/1974) and the ‘global’ Financial Crisis (2008/2009), both of which were accompanied by more or less drastic political measures. While the public awareness for the former increased dramatically after the German Federal government opted for a Sunday ban on driving, the latter first made headlines with several major bank failures such as Lehman Brothers - and subsequently with a number of unprecedented measures such as a 500 billion Euro bail-out package for banks, and a car-scrap bonus scheme to stimulate the German car industry. Whereas the Petrol Crisis is often referred to as turning point in the political economy of Western states in the post-World War II era and accelerating influence on the break with Keynesian principles, the Financial Crisis triggered a discussion on limits of globalised financial capitalism and on enhancing the regulation of financial markets on a global scale. During both crisis periods discussions arose which either re-emphasised and/or questioned the legitimacy of capitalist economic order. The dominant climate in crisis periods in which established orders and systems are put into question has been recognised by political counsellors, coining the term ‘crisis honeymoon’ due to their being very fertile for the enforcement of political measures and substantial structural reforms which are otherwise unthinkable in non-crisis situations (Williamson, 1994). The major shifts in the monetary policy of the European Central Bank from targeting the control of inflation to a direct political involvement with crisis management across states might serve as an example here (Maesse, 2015a). In this sense, crises are interesting objects for discourse researchers as they are periods of enhanced discussions during which a more or less established discursive order undergoes a struggle. In this discursive struggle, meanings of positions, oppositions and arguments are attacked, defended, redefined and re-stabilised.

2This paper studies the communication of these discussions, focussing on the discursive struggle which the current economic system in Germany underwent during both of the above mentioned crises as it was represented in the German press. With a particular interest in narratives, the paper studies the question of how far crisis discourses fosters the emergence of alternative discourses on an economic order. Therefore it analyses how arguments of protagonists in politics and industry are tied to narratives in order to justify political measures in the past, present or future. The study uses corpus linguistic tools to trace the narrative elements of each crisis which are then interpreted with regards to the discursive struggles around a hegemonic meta-narrative legitimising the current economic system.

3The study is part of a large-scale project on ‘Linguistic Constructions of Social and Economic Crises in Germany from 1973 until 2009’ which was financed by the German Research Foundation until 2012. In the project we explored the hypothesis that ‘crises’ are to a large extent ‘effects’ of the relevant discourse in which the political decision-making process is embedded. The fact that a certain period is perceived as an economic crisis or as an economic upswing is considered as a result of ‘knowledge by description’ (Warnke, 2009). This knowledge is transferred and organised through media and discourses and is less the result of the personal experiences of the people. The paper argues that these descriptions very often take the form of narratives which blend events, problems, political and economic actors and institutions with arguments favouring and therefore legitimising certain political solutions and delegitimising others. Therefore the paper emphasises the analysis of narrative elements and structures. The ‘narrative’ as an analytical category has been conceptualised in the Social Sciences and Humanities for several decades (Lyotard, 1979; Ricœur, 1980; White, 1973), and has recently been developed more explicitly for discourse studies (De Fina and Georgakopoulou, 2012; Viehöver, 2014). In contrast to approaches starting their analysis on the textual level using qualitative methods this paper argues for a corpus driven approach to narratives with text-statistical methods focussing on the example of economic growth and prosperity as part of a legitimising meta-narrative.

4In the first section of this paper I am going develop how we can use narratives as analytical categories in discourse studies. The second section presents the methodology focussing on the French corpus linguistic approach named ‘lexicometrics’ which I present as a quantitative heuristic to discourses and narratives. The third and fourth part contains the empirical analysis. Using quantitative methods, section three outlines the differences and similarities between the Petrol Crisis and the Financial Crisis corpus on a lexical level. For the analyses presented in section four I focus on one particular case of a narrative that I name the ‘growth and prosperity narrative’. All text sequences containing or referring to this narrative in both corpora are presented and interpreted. Section five discusses the findings of section four on a broader scale, applying topos-analysis and Laclau and Mouffe’s hegemony theory.

Narratives as an analytical category for discourse studies

5This paper addresses the concept of narrative as it was developed by Ricœur (1980; 1983; 1991), White (1973; 1989) and Schiffrin (2000) as an analytical category enabling the analyst to understand how the collective memory of an event or a series of events in a society is compiled and organised. Bringing this into accordance with a Foucauldian perspective on discourse, narratives can be considered as important structuring entities taking the form of stories as one kind of discursive practice which “systematically forms the objects of which they speak” (Foucault, 1969, p. 66-67). Discourses are structured substantially through narrative elements – such as a series of events, locations, a time frame, a structure of protagonists which are organised within a narrative plot. Narratives can be considered as meta-code that helps to interpret the world by forming and (re-) constructing experienced realities and horizons of expectations in a society. Therefore each narrative sets up a certain order of things and hence a certain order of knowledge (Viehöver, 2014, p. 213-214). As different narratives can refer to the same elements by telling a different story they prepare the ground for conflictual discussions in a society and therefore become crucial objects for discourse research, especially when taking into account how power is exerted by telling a certain story (for an introduction, see De Fina & Georgakopoulou, 2012, p. 125-154).

6In order to conceptualise the narrative within Foucault’s approach to discourse I propose to distinguish the term narrative from the term narration. Whereas the latter should be reserved for the very act of telling a story or an account told in a particular text by a particular narrator, a narrative is to be understood as a mode of a discursive practice beyond the delimitations of a single concrete text. It is a recurrent pattern moulding occurring events with the help of linguistic signs and utterances into a plot which accords to the cultural habits in a given society. Thus, a similar understanding can be found in Greimas’s semantic and narration theory in which he describes narratives as mechanisms underlying the production of meaning in society (Greimas, 1966; Maesse, 2014). This was developed further by Jameson (1976; 1981; 1991) as a theory of cultural symbolic structure.

7Taking a micro-perspective on narratives, we can distinguish at least four typical plot patterns in Western societies which to a certain extent can be recognised on the macro-level of narratives: tragedy, romance, comedy and satire (e.g. Frye, 1957, 1963; White, 1973). A narrative normally has an episodic structure. It has a beginning, a body and an end which take place within a certain period of the narrated timeframe. When analysing the course of a story we can distinguish different functions: the cardinal function which refers to decisive moments changing the direction of the narrative; and the catalysis which is a consecutive moment in the narration without impact on the course of events. The plot organises numerous events, incidents and episodes; it also provides the timeframe and the location of the narrative; and it opens a matrix of possible actors and actions (Viehöver, 2014, p. 235-236).

8When narratives blend events and social or political problems with arguments and political solutions they produce a meaningful story legitimising a certain range of political actions. At the same time, such a story has the potential to delegitimise political actions that don’t appear logical in the order of the narrative presented. In this sense we can speak of legitimising and delegitimising narratives (Viehöver, 2014, p. 213). Meta-narratives (Lyotard, 1979) are a special type of legitimising narrative. If narratives are relatively free in terms of the plot and the story they tell, meta-narratives are rather fixed with this regard. Thus, in general the plot of a meta-narrative describes a horizon of expectation concerning the development of the world within a certain cultural era. It is an overarching narration delivering important presuppositions from which other narratives draw, often implicitly, as sources for their argumentation. As a cultural era, the modern age is strongly shaped by the meta-narrative of Enlightenment. It conveys the conviction that all phenomena of the world can be explained by scientific arguments and that man can become emancipated from ignorance and subjugation by means of understanding and the urge for egalitarian endeavour. Besides Marx’s historical materialism, the meta-narrative of the modern age contains also the narration of capitalism whose technical and industrial development will help mankind to become emancipated from poverty. In general meta-narratives implicitly convey a certain set of social values. In a given society these are important for the legitimisation of institutions, social and political practices, ways of thinking and legislations. All meta-narratives have in common that they are only legitimised through an idea to be realised in the future (Lyotard, 1987, p. 32-33).

Corpus-linguistics: a quantitative heuristic to analyse narratives

9As pointed out in the previous chapter, narratives are multi-layered. However, when analysing them we normally approach narratives via the textual surface with different methods. During the last few decades, corpus-linguistic methods have become an important tool in the analysis of language and discourse. Numerous large digitalised text collections have been made available to numerous communities of researchers. And, in order to analyse these corpora, a large variety of corpus linguistic methods have been integrated in software with more or less intuitive graphical interfaces usable even with a low level of computer language knowledge. Notwithstanding the opportunities these new infrastructures offer for the analysis of discourses on the lexical surface, it has to be asked how far they can account for the analysis of narratives as outlined above. What can we draw from quantitative text data in order trace narratives and better understand the construction of society?

10Besides the numerous studies by discourse linguists using qualitative analytical categories such as topos-analysis (e.g. Wengeler, 2003), metaphors (e.g. Böke, 2005; Musolff, 2005) or cognitive frames (e.g. Ziem, 2014) different quantitive approaches to text corpora have been developed to analyse discourses. For example, in the UK, corpus linguistic tools used in discourse studies have been developed in the tradition of lexicography (McEnery & Hardie, 2013) for instance ascertaining key word by contrasting a smaller research corpus with a large reference corpus (Baker, 2006; Baker, 2014; Baker & McEnery, 2005). In contrast, in France, corpus linguistic tools have been developed more closely in the context of discourse analysis and multivariate statistics. Under the label of lexicometrics, computer assisted discourse analysis has been developed since the 1970s (Tournier, 1975; Tournier, 1993). Lexicometrics is a corpus-driven (Tognini-Bonelli, 2001) approach emphasising exhaustive, contrastive methods (Lebart, Salem & Berry, 1998) by analysing partitions based on Benzécri’s (1980; Salem, 1982) multifactor analysis or Reinert’s (1983) descending hierarchical classification. These methods are used as heuristic tools to explore the data and in order to reveal the underlying lexical structure of the text material concerning the chronological development and/or the lexical distance between different discourse participants.

11It is only since recently that the number of studies combining quantitative and qualitative methods of discourse analysis has increased (eg. Felder, Müller & Vogel, 2012). While some scholars try to retrace language use patterns in a certain part of society (Bubenhofer, 2013; Bubenhofer & Scharloth, 2011; Bubenhofer & Scharloth, 2013) or in the representation of a certain topic (Müller, Freitag & Köder, 2010), others emphasise rather the heuristic strength of lexicometric methods allowing the researcher to explore a corpus with a non-hermeneutic eye in order to find parts relevant for the research question to be analysed in-depth with qualitative methods (Angermüller & Scholz, 2013; Kuck & Scholz, 2013; Scholz & Ziem, 2013; Scholz & Ziem, 2015).

12Lexicometrics is a quantitative approach to corpora which was developed in French discourse analysis in order to unveil lexical characteristics in the discourse of different speakers and periods without using interpreting methods. It is based on statistical methods such as multifactor analysis or hyper-geometrical distributions which are barely used in corpus linguistic studies of German or Anglo-Saxon provenience (for an introduction see Scholz & Mattissek, 2014). Of course, one can find a substantial methodological overlap when looking at methods such as collocation or n-gram analysis. However, in contrast to the Anglo-Saxon tradition of corpus-linguistics which developed in the context of lexicography and language-teaching, the lexicometric approach was born at the heights of structuralism in France within a debate on how to unveil ideology in the analysis of political discourse (Scholz & Fiala, forthcoming). In order to achieve this aim lexicometric scholars have emphasised the advantages of measuring all lexical items of a given corpus at once exhaustively, systematically and automatically. These measurements draw on a complete table of the vocabulary containing the frequencies of each word-token in the different parts of the corpus to be created by the researcher according to metadata such as author or date of publication(Lebart, Salem & Berry, 1998, p. 45-79). Furthermore, the lexicon of a corpus is compared, classified and analysed from both a synchronic and diachronic perspective. A corpus can be analysed with methods of multivariate data analysis (Benzécri, 1976; Cibois, 1983) in order get an idea of the corpus structure. It can be measured if certain words are over- or under-represented in the texts of certain authors or certain research periods (specificity of vocabulary). Hence, this method helps to explore which words are preferred by certain authors or in certain periods in comparison to the texts of the other authors or periods (keywords). Furthermore, one might be interested in which immediate textual environment a word is used. One method that can help to answer this question is collocation analysis, which measures which words are over- or underrepresented in the immediate context (cotext) of other words. At the same time, one can always access the actual text in order to analyse it with qualitative methods by looking for text sections with the keywords in context (concordances) which have raised the researcher’s interest in the quantitative part of the research.

13A corpus for lexicometric analysis is to be specifically designed according to the research question concerning a certain discourse analytical problem. Normally the corpus is not usually analysed by contrasting it with a large reference corpus representing a certain written standard language or a variety of it; rather it is analysed by contrasting the quantity of lexical items in the different parts of a corpus. These so called corpus partitions are introduced by the analyst according to the metadata of texts integrated into the corpus following a certain discourse research related hypothesis. For instance, by implementing the partition author, the researcher follows the hypothesis that the discourse positions of discourse participants (authors) take effect in the language practices which manifest themselves among others in the lexicon of each participant, and therefore can be traced with text statistical methods. This way of analysing the corpus demands certain preconditions for the construction of a corpus. The texts included in a corpus should belong to the same text register (Lee, 2001) in order to make sure that the results exclusively represent the characteristics of a certain discourse without being biased by the characteristics of the register (Lebart, Salem & Berry, 1998, p. 153-159).

14These methodological implications comply very well with the corpus requirements of the discourse- historical approach in which a discourse is defined as a virtual text corpus containing all texts, in the broadest sense, concerning a certain topic (Busse & Teubert, 1994; Hermanns, 1995; Jung, 2000). The actual research than has to be conducted in a concrete text corpus which ought to be representative for the discourse analysed, in terms of participants, arguments, time span and so forth. In this regard, we conducted the five case studies from which this paper draws, two of these being the Petrol Crisis and the Financial Crisis. In order to gain access to the crises discourses we built a large text corpus of press articles deemed to represent the public debate concerning these crises.

15The research periods, 01/10/1973-31/01/1974 and 01/09/2008-30/04/2009, were established after having read a small selection of press articles across the supposed crisis period and after having taken into account the corresponding research literature of in the economic, social and historical sciences (e.g. Butterwegge, 2006; Nützenadel, 2007; Plumpe, 2010). During this process we determined the search terms which helped in finding the relevant texts for the corpus. For the Petrol Crisis we used the search terms energy [Energie], economic activity [Konjunktur], recession [Rezession], growth [Wachstum], unemployment [Arbeitslosigkeit] and crisis [Krise]. For the Financial Crisis we used only the search term financial crisis [Finanzkrise], although this brought an almost unmanageable number of texts, which had to be part-read manually in order to decide on their relevance for the project. We also took into account the relative frequencies of keywords in newspaper databases in order to determine the beginning and the end of a certain crisis narrative in the public discourse.

16The texts retained for the corpus have been taken from five national, quality newspapers deemed to have an impact on public discourse and the formation of public opinion. In particular the texts were collected from three German national daily periodicals: the Frankfurter Allgemeine Zeitung (FAZ), ahigh profile daily representing mainly liberal conservative ideas; the Süddeutsche Zeitung (SZ), ahigh profile daily representing liberal centre-left positions; and Bild, a daily tabloid often criticised for an underlying right wing populism; and the weekly journal DieZeit, representing left and centrist ideas;as well as the weekly magazineDerSpiegel. The articles were found in different databases accessible over the Internet or from different archives and libraries2. The corpus contains 5,841 press texts with a total of 3,249,459 word-tokens; 2,027 articles with 924,545 word-tokens were retained for the Petrol Crisis and 3,814 articles with 2,324,914 word-tokens were collected for the Financial Crisis.

Differences and similarities in the representation of the Petrol Crisis and the Financial Crisis

17In order to trace common narrative patterns between different crises we first try to uncover differences and similarities in the representation of the Petrol Crisis and Financial Crisis in the German press. This can be done with methods measuring the specific vocabulary (keyword-analysis) and the common vocabulary (proportionally distributed vocabulary). The study remains limited to nouns as they convey the largest part of the semantic value we are interested in at this point of analysis. The specific vocabulary consists of word-tokens that are proportionally overrepresented in one part of the corpus (e.g. Petrol Crisis) when compared to the other (e.g. Financial Crisis) (Lebart & Salem, 1994, p. 171-196; Lebart, Salem & Berry, 1998, p. 129-145). In contrast, the common vocabulary (vocabulaire commun in Labbé & Monière, 2003) consist of word-tokens that are proportionally distributed in both parts of the corpus. Therefore this vocabulary can be considered as the lexical base of both crisis corpora which might give some insight into the characteristics of a vocabulary that is typically used in times of economic crises in the German press in general.

The crisis specific vocabulary

18Looking at the specific vocabulary3 of both crises represented in Tables 1 and 2, one can discern that certain narrative elements occur in both crises. Table 4 represents categories which were developed as a first heuristic step, trying to identify narrative elements in the data. In this sense we find in both corpora specific vocabulary that refers to locations, a time frame, involved actors and a description of each crisis event which contains different ways of naming the crisis event, symptoms and objects of the crisis which are generally describing a problem in terms of numbers and measures to be tackled with different political measurements.

Table 1: The most frequent keywords of the Petrol Crisis corpus

Examining words denoting locations, for the Petrol Crisis the geographical localisation seems to be more specific in terms of country names than for the Financial Crisis. As we know, it was also claimed that the latter was a global phenomenon. However, it seems that its global impact is less reflected by the names of countries mentioned in the media. Both crises seem to take place at very different levels of experience. In texts relating to the Petrol Crisis we find rather nouns referring to the experience of everyday life, such as petrol stations and refineries, or names like Federal Republic, [Saudi]Arabia, [European] Community, Middle East and Persian Gulf which refer to political entities and countries somehow involved in the crisis.

Table 2: The most frequent keywords of the Financial Crisis corpus

19In contrast the Financial Crisis seems to take place in the rather virtual space of the financial markets [Finanzmärkte, Finanzmärkten, Märkte] which manifest themselves only indirectly in the everyday experience of ordinary people’s lives. This impression can be confirmed by looking at the political measures such as ban on driving or rationing in the Petrol Crisis which interferes with everyday life. In contrast, the Financial Crisis is constructed in much more abstract terms such as debts, recession, bailout package, financial markets, and world economy.

20There is not much information in the tables represented concerning a specific time frame of each crisis. However, if we examined a more extended version of the table we would be able to see that texts relating to the Petrol Crisis refer to shorter periods (winter, days, time, month) than texts regarding the Financial Crisis, which refer rather to longer periods (times, year(s)). Looking at the concordances of these words, we see that the Financial Crisis is narrated rather by referring to longer periods of past events (e.g. since … years, in the 1920s years) whereas the Petrol Crisis is narrated rather by referring to short-term past and future periods (e.g. every month, in the past or nextmonth, in this month). This suggests that the narrative on the Financial Crisis seems to use more historical references than the Petrol Crisis, which might indicate in turn that it is the Financial Crisis rather than the Petrol Crisis that is constructed as a ‘historic crisis’.

21When looking for words referring to actors there are few surprises. As expected, one finds actors of the corresponding economic area that is in crisis – such as petrol companies in the Petrol Crisis corpus and banks, banker, investors, manager, central bank, hedge funds, minister of finance in the Financial Crisis. However in the Petrol Crisis texts, the field of the actors involved seems to be broader as we find also trade unions, car drivers and employees (not shown in table). Moreover, while the specific vocabulary of the Petrol Crisis contains a greater variety of political actors at a national level, in the Financial Crisis texts the only actors at a national level we find are the state and the chancellor as well as some supranational actors like the European Central Bank and the International Monetary Fund. Therefore the Financial Crisis seems to be presented much more as a phenomenon for which global actors are politically responsible than in the case of the Petrol Crisis.

22With regards to the naming of the crisis, we find energy crisis [Energiekrise] and petrol crisis [Ölkrise] in the specific vocabulary of the Petrol Crisis corpus. It seems that the Financial Crisis is more often referred to in terms of an economic crisis [Wirtschaftskrise] and in terms of a ‘general’ crisis [Krise]4. It’s also worth mentioning that we can find terms that might be used to explain an energy crisis (e.g. resources, petrol, and fuel oil) and a financial crisis (financial market, hedge funds, credits, value, and risk). However, we cannot find terms that might help conceptualise the economic crisis within the Financial Crisis corpus. This, and also the overrepresentation of the simple term crisis might indicate that the Financial Crisis is narrated in terms of a universal crisis that touches all economic areas and many other areas of society.

The common crisis vocabulary

23After having briefly summarised some observations in the specific vocabulary of each of the two crises, we now want to look at those nouns and proper nouns that are proportionally distributed in both crises, and therefore might be (with precautions) considered as a common crisis vocabulary. As one might expect, the number of content words that are used to a similar extent in both crisis corpora is not very high. The words that are most common to both corpora are represented in table 3. Highlighted according to the same categorisation used in table 1 and 2 one can find among the most frequent words future, growth and investments. Common actors in both crises are the Americans, the political party CSU and the heads of governments. Most common locations are Brussels, foreign countries and Europe. Interestingly nationalisation is a political measure that is discussed in both crises to the same extent. Of course, at this level of research without looking at the co-text we cannot say with what political implications this term has been discussed. However, as I am aiming to trace narratives relating to the political economy that are common to the discourse of both crises, the term growth seems to be the most promising. In order to investigate the co-text of this word we first studied its collocations separately in both crisis corpora. In both corpora the word co-occurs with prosperity (Wohlstand), employment (Beschäftigung), stability (Stabilität), percent (Prozent), gross national product (Bruttosozialprodukt), and economists (Wirtschaftswissenschaftler, Ökonomen). In order to deepen the analysis of a narrative relating to the political economy in both crises I decided to have a closer look at the sections of text containing the words growth and prosperity.

Table 3: Common vocabulary in the Petrol Crisis and Financial Crisis corpora
(proportional measurement)

Table 4: Explanation of highlighted text in Tables 1 to 3 based on semantic categorisation

Growth and prosperity in crisis situations

24The exploration of the common vocabulary has shown that growth is a key term in both crises and one of the main collocates of growth is the word prosperity; as such, we will now go on to analyse text sequences containing both words with a qualitative approach which allows us to interpret the meaning of words in their co- and contexts. For the first step, we will focus on the question how each of the two crisis events is related to the question of economic growth and prosperity.

Table 5: Growth and prosperity sequences in the press discourse on the Petrol Crisis (For German original text see annex)

25All analysed text sequences seem to refer to growth and prosperity as a presupposed and unquestioned set of social values. I understand this set of social values in terms of a meta-narrative of capitalist society which is then discussed in the subsequent section of this paper. For the sake of the argument it is important to remind the reader that the German word Wohlstand which I translate with the English word prosperity refers not only to economic affluence and wealth but also to immaterial welfare, i.e. wellbeing and subjective judgements about the quality of life in general. As I am going to argue later, it is exactly this ambiguous meaning which lends the term its popularity across expert discourses of political economists and debates relating to social progress and wellbeing in the public and private sphere.

26Table 5 displays all text sequences containing the words growth and prosperity from our corpus on the Petrol Crisis. All examples acknowledge in their own right, implicitly or explicitly, the importance of petrol for economic growth and prosperity. With the exception of sequence 2, which represents a historical perspective, and sequence 5 which seems to unveil a new truth about the European Economic Community, they recognise that the important role of petrol has come to an end or might have to be changed in order to gain or maintain prosperity. This need for a new source of growth and prosperity seems to establish the very object of the crisis.

27If we look at the text sections containing the lexical items growth and prosperity in the corpus of the Financial Crisis in table 6 we can certainly observe one striking similarity with the text sections gathered in table 5. First of all, the mantra-like phrase ‘growth generates prosperity’5 seems to be even more stabilised and affirmed in the discourse than this was the case previously in the Petrol Crisis. This can be inferred from the content of the text sections. With the exception of examples 10 and 14, if we could observe a certain refrain from common economic practices concerning the generation of growth and prosperity with the help of petrol in 1973/1974, this willingness to change habits seems to be missing completely in the text sections displayed here for the Financial Crisis in 2008/2009. Similar to table 5 the very object that is discussed as the originator of the crisis, for instance the financial system, the financial markets, hedge funds or even greed, are presented in the same way as sources for growth and prosperity. Especially in the examples 7, 8, 9 and 11 such a line of argument is defended rather than criticized. Particularly remarkable, is example 8 as it presents state intervention as being equal to uncertainty for investments and hence as a risk for growth and prosperity which certainly reflects most clearly an underlying ideology advocating free markets and containing all state activities on the economy. Such a statement presented in public discourse might have been unthinkable in the 1970s when Keynesian ideas of a state controlled political economy were still in vogue, and were considered important guidelines in order to ensure economic growth and the subsequent material prosperity of the population.

28The sequences in table 5 and 6 show that in each utterance the dependence of prosperity on growth is presupposed. This presupposed character becomes evident because the composition of both words is never an object of the discussion in any of the sequences. Only in a few sequences is the validity of interrelation questioned, especially when growth has not resulted in sufficient prosperity (1) or when the semantic meaning of prosperity is defined in more detail (4, 14). As a consequence the dependence on growth seems less, or not at all obvious. Nevertheless, these sequences also reproduce the presupposition that prosperity depends on economic growth.

Table 6: Growth and prosperity sequences in the press discourse to the Financial Crisis (For German original text see annex)

29I suggest that this composition is to be understood as a part of a meta-narrative which, as outlined above, describes a horizon of expectations concerning world development. In this particular case the semantic elements growth and prosperity are part of a narrative of capitalism in which its technical and industrial development will help mankind to emancipate from poverty.

Economic growth engenders prosperity – a continuing story legitimising the capitalist economic order

30The preceding section of the study has shown that the “growth and prosperity narrative” is reproduced but also criticised in both of the crisis periods analysed. In both cases the narrative is adapted to the predominant content of each crisis discourse. We can understand better what happens on the discursive level if we consider it through the lens of Laclau/Mouffe’s discourse theory. In these terms the relation between petrol or the financial system, growth and prosperity can be understood as a chain of equivalent (floating) signifiers which are elements articulated with temporarily fixed meanings as moments in a particular narrative within a (crisis) discourse. The elements become equivalent moments of a chain when their difference is subverted by an identical reference (Laclau & Mouffe, 1985, p. 127). In the present example the identical reference is provided by the signifier prosperity. We agree that the term prosperity remains rather undefined in most of the cases where we encounter it in the corpus. This is exactly why it can be understood by readers in all sorts of contexts, which refer it to different meanings. However, in all contexts it remains a positive social value. As prosperity is a social value recognised at least by large parts of the population in Western societies as a desirable social or individual state to be acquired and maintained, it has a dominant value influencing all other elements of this chain of signifiers by subverting their meaning relating it to the social value of prosperity. Without the prospect of the contribution to prosperity, the meaning of growth, petrol and the financial system (and its equivalents such as hedge funds, research and innovation) would have a very different value within the semantic structure of the crisis discourse. Possibly these elements would not be able to acquire such a prominent and positive value within the crisis discourse. In this respect prosperity has a dominant position in this chain of equivalent discursive moments. Further to Lacan, Laclau/Mouffe call such points of partial fixation of meaning nodal points(Laclau & Mouffe, 1985, p. 112). If we try to understand prosperity as an empty signifier which represents an absent totality mobilising all different kinds of actors in the social sphere, the articulation of the above mentioned equivalent elements establish a hegemonic relationship (Laclau, 1996, p. 42-44). It remains hegemonic as long as prosperity (with all its different meanings) is recognised as a key element to be maintained or acquired in the future development of society. We could also argue that growth establishes a nodal point in the discursive structure because it occurs much more often as a term in both corpora than prosperity. However, I would like to point out that if growth does not refer implicitly or explicitly to prosperity each time it is mentioned, it would not have the same value in the structure of the discursive elements in which it is articulated.

31As a nodal point, prosperity becomes an important argument in public discourses which, whenever it is articulated in a chain of equivalent moments, increases the persuasive force within a certain argumentation. In crisis discourses a common pattern of argumentation is the topos of threat. A topos is a place in discourse where argumentation takes place according to a certain pattern. One such pattern concerns opposites: “If action Y is desirable in relation to object X, the contrary action Y‘ should be disapproved of in relation to the same object X”(Žagar, 2010, p. 17). Related to our case this topos could be translated as follows: If unregulated financial markets; hedge funds; greed etc. are good for prosperity, the abolition of these elements should be avoided. Furthermore, topoi can be defined according to their topic (e.g. Wengeler, 2003). An example is the topos of threat which has it’s a high rate of occurrence in public discourses during periods of crises. For instance, the Petrol Crisis and the Financial Crisis, are represented as respectively endangering the social value of prosperity which, according to the analysed discourse, is caused by growth.

32Within this theoretical frame we can interpret the above cited text sequences as follows: The object of each crisis (Oil, price of resources or financial products and financial markets etc.) are set equivalent to growth which is set equivalent to prosperity. They are articulated as a chain of equivalent moments of a hegemonic discursive formation arguing for prosperity in society. Within this hegemonic formation we find a narrative which articulates growth and prosperity closer to each other than other moments of this formation. In this narrative, prosperity heavily depends on growth. If prosperity can be considered as a social value itself, growth cannot. However, by articulating growth as an absolute precondition for prosperity, growth gains the importance of a social value itself. Growth can now function as a quasi-nodal-point having almost the same value within the discursive structure as prosperity.

33There is no question that such a discursive movement cannot simply be manufactured by the participants represented in the German press, or whomsoever. It might help to distinguish between a macro-level and a micro-level of discourse in order to understand how this movement is possible. On the macro-level there are at least two major forces preparing the ground for movement. There is a set of social values amongst, for instance, the prosperity and welfare of the population on which the members of a society implicitly or explicitly agree, and which are normally not a matter of discussion anymore. Furthermore, there are discourse participants who are legitimated to define how these values can be reached and maintained. In Western societies this position was traditionally taken by politicians. However, in recent years we can observe a rise in the use of expert knowledge in order legitimise political measures. Economic experts seem to have gained a dominant legitimate position throughout the last few years (Maesse, 2015b). It is the discourse of economic experts – in particular the discourse of neoclassical growth theory (e.g. Solow, 1956) and its descendants – which has gained a dominant position in the public discourse on prosperity. It is this expert discourse that coins the mantra-like combination “growth engenders prosperity”. The experts translate prosperity into growth as the economic variable that can be measured and manipulated through the knowledge of economic experts, which in turn automatically becomes the source of legitimate political decisions - and hence political power. The set mantra is then reproduced in the public discourse on a micro-level in order to serve a particular discourse participant as argument for or against a particular position. As a result of this constellation the need for a discussion of what kind of prosperity can be generated by economic growth is less obvious. Hence, questioning the narrative that “growth engenders prosperity” becomes more difficult.

34Furthermore, in periods of crisis there is a dramatic element which further obscures the value of prosperity within the discursive structure. This is because within a topos of threat the focus within the production of meaning shifts from the endangered object to the action that can tackle the danger. The more imminent the danger of a crisis for the basis of growth - and subsequently of prosperity - is represented, the more urgent the action against the danger is needed. In consequence, growth and prosperity become only the targets of political measures against the crisis and are not a matter of debate. At the same time (almost all sorts of) political measures against the crisis become vital for society and are therefore highly legitimised. Moreover, the argumentative force of the narrative affects those concepts that are semantically linked to it within the discourse, for example: the resource of oil, the economic system of social market economy, research and innovations as well as hedge funds become vital for society. In contrast, such a meaning is withdrawn from the European Economic Community in example 5, or a strong state in example 8. In Laclau’s terms the argumentative force of the empty signifier prosperity is triggered by the absence of totality in the meaning of the articulated chain of equivalent moments which is partly fixed through the social value of prosperity. With such a key role the term also becomes crucial in alternative debates which do not rearticulate the existing discursive structure, but articulate its nodal-point differently. Thereby the meaning of prosperity becomes reconstructed differently - for instance independent from economic growth and based on the quality of life, time for the family, leisure, friends and culture (14).

Summary and conclusion

35With regard to methodology, the paper has shown how narratives can be analysed with quantitative methods of discourse studies. As a heuristic approach used in a corpus driven logic these methods can lead to results that could have not been anticipated prior to the study. This is possibly because the statistical measurements allow the discovery of results from a perspective that is temporarily alienated from the textual material, and does not follow the hermeneutic reflex which is triggered whenever we encounter a text. These rather rigid methods make it possible to map the discourse according to its dominating narrative elements. The aim of using quantitative methods was to demonstrate their heuristic strength and their potential to narrow down the vast amount of corpus data in a meaningful way to one particular discourse phenomenon. Other studies in the same corpus or in a bigger reference corpus focusing on different co-occurring words of growth could help, in future, to draw a bigger picture of how narratives are used in crises discourses. Instead of using the abundance of corpus linguistic methods that could be applied to analyse narratives, such as the analysis of concordances, poly-co-occurrences (Martinez, 2012), or n-grams I have preferred to present in this paper how different methods can be interlocked in order to reduce a vast corpus to a manageable amount of text sequences that are relevant to the research question and can be treated with qualitative methods of discourse studies. Of course, to analyse narratives in depth we have to use our context knowledge and qualitative hermeneutic methods. The study has shown that Laclau/Mouffe’s theory of hegemony can provide exciting insight into the semantic relations between lexical elements which have been identified with text statistics. Therefore the mixed methods approach presented for the analysis of narrative seems to be promising.

36With regard to the narrative elements analysed, we have seen from the quantitative analysis that the keywords in the Petrol Crisis indicating locations and political measures refer to an everyday experience of the crisis which involves protagonists at a national political level; whereas the Financial Crisis, which refers rather to abstract locations such as the financial markets and seems to be more strongly constructed as a historical crisis similar to the Great Depression of the 1920s, involves protagonists at a supranational level such as the European Central Bank or the International Monetary Fund. There is evidence that the Financial Crisis is rather narrated in terms of a universal crisis that touches all economic areas and all other areas of society. This higher dramatic tension in the construction of the Financial Crisis might explain why the political measures against this crisis could involve the infusion of billions of Euros and Dollars without raising any public opposition in a society that by that time had been suffering a constant decline in net wages for 20 years (Brenke, 2009).

37In the common vocabulary of both crisis corpora I identified content words of which growth was to be analysed in more detail. In combination with the co-occurring word prosperity I identified an important meta-narrative which legitimised the current political economy. We have seen that the ‘growth and prosperity narrative’ occurs in both crisis corpora – related to the length of both corpora much more often in the Financial Crisis corpus. With regards to the structure of semantic values within the discourse we have seen that in both corpora the crisis objects (Oil, Petrol price, financial products) are set equivalent to growth and prosperity. This constellation of discursive elements prepared the ground allowing the justification of most extraordinary political measures in both crises discourses.

38Laclau/Mouffe helped to understand this narrative in terms of a hegemonic formation in which prosperity takes the dominant position of a nodal point which represents the absent totality in the chain of equivalent articulated moments. In this chain growth has a privileged relation to prosperity because in this hegemonic formation prosperity appears to be impossible without growth. Growth seems to become a social value to the same extent as prosperity already is. Thereby, the explicit articulation of the signifier ‘prosperity’ within this hegemonic formation becomes less important. Nonetheless, it is always implicitly or explicitly articulated as a common goal to be reached through growth. Hence, prosperity functions as an empty signifier – the centre of an absent totality which organises the structure of the discursive elements.

39Furthermore, as these objects are articulated in a topos of threat, both crises are represented as a danger for growth and prosperity. As such, a threat to growth becomes a threat for prosperity and hence a threat for society. Political measures which address this threat can therefore be represented and understood as measures to save society. In such a discursive constellation most exceptional political actions as such as a Sunday ban on driving, or a 500 billion Euro bailout package, consequently appear to be legitimate measures.

40If we try to zoom out of these discursive micro structures towards a more generalised perspective on the mechanisms within the public political discourse, we can conclude that the discourse follows a hierarchical order based on the question of accountability. In this order the meta-narrative sets the frame in which legitimate positions can be acquired. In order to articulate a legitimate position, discourse participants have to account for the social value ‘prosperity’. Hence, this meta-narrative with its set social value must be situated at a higher level of the discursive order. Not all discourse participants that articulate a legitimate position become automatically relevant (Sperber & Wilson, 1986) for the discourse. There is a variety of forces that influences the relevance of a position and hence its discursive power, not all of which can be described here – one important force is of course the institutional power of participants with which for instance a politician can acquire high relevance in a discourse on social questions. However, institutional power in itself is not sufficient; rather the discourse participant in one form or another also needs to engage with the expert knowledge which is considered to be helpful in gaining the social value strived for. As the narrative refers to economic growth, of course, economists and their various argumentations become sources to stabilise a legitimate position in the public discourse. For instance, politicians refer to economic expert knowledge to make their arguments (see also Maesse, 2015b). As the public discourse serves a public debate and tries to translate more complex phenomena, it will not reproduce the whole expert discourse but only an eclectic choice of positions and argumentations of it. Even if the public discourse could represent all expert discourses, it would only reproduce versions of a much more complex reality reduced with the help of reliable and valid scientific methods.

41Such imaginaries tell a consistent story about the much more complex, messy and incommensurable economic processes it refers to (Jessop, 2004; Sum & Jessop, 2014). In terms of a cultural political economy we might want to look for imaginaries that tell a different story in order to help build a different form of capitalism. This comparison of the media discourse representing two major crises of capitalist economies during the last 50 years has shown how the flexibility of meta-narratives which were applied in different crises with very different objects helps to maintain, re-establish and restabilise belief in the legitimacy of the current political economy. Furthermore, as we have seen, the re-articulation of meta-narratives can help to legitimise the oddest excrescences of the latest developments in financial capitalism – for example, when articulating greed, which elsewhere is considered as one cause of the Financial Crises, as source for growth and prosperity. In this discursive constellation the formula ‘growth engenders prosperity’ is reproduced as a presupposed, fixed mantra that seems impossible to be unveiled and deconstructed in order to instigate a debate not only on alternative ways to gain prosperity but also on what we mean by prosperity in the first place. In this sense the discourse of the press seems mainly to reproduce positions for which, following a certain positivist reductionism (Holcombe, 2013, p. 20-31), neoclassical growth theory has once argued (e.g. Solow, 1956). These positions might be valid when the largest part of the population is deprived of basic material goods, and mass production is the main mode of production in the capitalist industry as was the case for long periods during the first three quarters of the 20^th century in Western societies. However, such a limitation of material wealth no longer represents an understanding of prosperity in late capitalism, where for instance physical and psychological health and quality of life play a much more important role than they previously used to play in public discourse. New indices like the ‘Legatum Prosperity Index’ try to account for such a post-material understanding of prosperity and a recent, quite popular but elaborate, account of post-material prosperity has been provided by Jackson (2009). Further discourse research might investigate under what discursive conditions such a concept might gain momentum in the public debate.

Bibliographie

Annexes

Pour citer ce document

 

 

Ce(tte) œuvre est mise à disposition selon les termes de la Licence Creative Commons Attribution 4.0 International.