The Marketplace of Ideas. A Corpus Study of Buy and Sell Metaphors in American Political Discourse. more

0 HELSINGIN YLIOPISTO The Marketplace of Ideas A Corpus Study of Buy and Sell Metaphors in American Political Discourse Sanna Franssila Master’s Thesis English Philology Department of Modern languages University of Helsinki October 2011 1 Contents 1 Introduction ...............................................................................................................................3 2 Studying metaphor in discourse ................................................................................................8 2.1 Metaphor in political discourse ........................................................................................12 2.1.1 Political language and metaphor studies .......................................................................12 2.1.2 The Conceptual Metaphor Theory and the political metaphor .....................................15 2.2 Metaphorical Vehicles in genres ......................................................................................17 2.3 Quantitative metaphor analysis ........................................................................................18 2.4 The metaphorization level: the proportion of literal and metaphorical expressions ........20 2.5 A summary: the research design ......................................................................................21 3 Negative metaphors in news discourse ...................................................................................23 3.1 Negativity and metaphorical expressions.........................................................................23 3.2 Partisan news discourse, political polarization and negativity in news in the United States ......................................................................................................................................25 4 The data and the method .........................................................................................................29 4.1 The election corpus ..........................................................................................................29 4.2 The comparison corpora: news magazines, Fox News, MSNBC, Limbaugh and National Public Radio...........................................................................................................................31 4.3 The identification of metaphorical expressions based on buy and sell ............................33 4.3.1 The identification of the most usual types of buy or sell metaphors .............................35 4.3.2 Borderline cases in identification ..................................................................................37 4.3.3 Similes and proverbs .....................................................................................................38 4.3.4 Discourse metaphors and comparisons .........................................................................40 4.4 The target of evaluation in the election news corpus .......................................................45 4.5 Negativity in the election news corpus ............................................................................47 5 Buy and sell metaphors in the corpora: the quantities and the proportional distribution ........51 6 The genre-specificity of the metaphorical buy and sell ..........................................................53 6.1 Buy it -type .......................................................................................................................56 6.2 Buy into and buy + [OBJECT] -type ................................................................................58 6.3 Sell + [OBJECT] -type .....................................................................................................61 6.4 Tough sell and does not sell -type ....................................................................................62 6.5 Sell out -type ....................................................................................................................64 6.6 A concluding note: buy and sell metaphors, genre and partisanship ..............................65 2 7 The metaphorization level of BUY and SELL ...........................................................................67 7.1 Buy it -type .......................................................................................................................68 7.2 Buy into and buy + [OBJECT] -type ...............................................................................69 7.3 Sell + [OBJECT] -type .....................................................................................................71 7.4 Tough sell and does not sell -type ....................................................................................72 7.5 Sell out -type ....................................................................................................................73 7.6 A concluding note: the proportion of the metaphorical occurrences ..............................74 8 Buy and sell metaphors in the election news...........................................................................77 8.1 Diachronic developments in the buy and sell distribution ...............................................77 8.2 Negative buy and sell metaphors in the election news .....................................................79 8.3 Non-negative buy and sell metaphors in the election news .............................................84 9 The targets of negative evaluation with buy or sell metaphors in the election news ..............86 9.1 The opponent party or the own party as a target ..............................................................86 9.2 The main targets of negative evaluation: the nominated Democratic candidates ............89 9.3 The targets of negative buy metaphors: Kerry, Clinton and Obama ................................90 9.4 Targets of negative sell metaphors: Gore and Kerry .......................................................91 10 Discussion .............................................................................................................................94 11 Conclusion ..........................................................................................................................104 References ................................................................................................................................108 Sources .................................................................................................................................108 Books and articles ................................................................................................................108 APPENDIX A. A list of abbreviations. APPENDIX B. A list of buy and sell metaphors (751 occurrences). APPENDIX C. Detailed information on corpora. 3 1 Introduction "Words are not actions," she said. "And as beautifully presented and passionately felt as they are, They are not action. You know, what we've got to do is translate talk into action and feeling into reality.” “Let’s have a reality check.” Presidential candidate Senator Hillary Clinton in a debate in New Hampshire 5th Jan, 2008 “And, you know, so, the truth is, actually, words do inspire, words do help people get involved, words do help members of Congress get into power so that they can be part of a coalition to deliver health-care reform, to deliver a bold energy policy.” “Don't discount that power. “ Presidential candidate Senator Barack Obama in a debate in New Hampshire 5th Jan, 2008 The Oxford English Dictionary (1989) dates the advent of buy in the meaning of ‘to believe’ as 1926 and considers it as late as 2007 an American, colloquial expression. Older figurative expressions with the stem word buy are, for example, ‘to be bought and sold’ (to bribe, 13th century) or ‘to buy a pig in a poke’ (to accept something without closer inspection, 15th century). Sell metaphors seem to be older and more productive: there are various phrasal expressions. The oldest occurrence is in the Lindisfarne Gospels ca. 950: Cuæð uutedlice iudam ‥ ðes forðon uæs sellend hine (John vi:71) 1 (Oxford English Dictionary 1989). Selling one’s soul is from the 16th century, sell someone down the river, sell short and sell out are from the 19th century and they are defined as American. Sell or sell on in the sense of promoting an idea, to be sold on in the sense becoming convinced, and sell oneself as promoting someone are all from the first half of the 20th century. The Oxford English Dictionary marks all these newer expressions as American and colloquial, sell out is specifically defined as “orig. U.S, political, slang”. In this thesis, I will study the occurrence of buy and sell metaphors in authentic political news discourse. Naturally-occurring language is represented in this study by four text corpora, which all are based on American news media. First, there is an election news corpus (1.6 million words) of news texts in print and online media covering the presidential election campaigns in the United States in 2000, 2004 and 2008. These texts represent either a conservative or liberal political affiliation, and there is one set of texts covering the primaries and another the general election phase in each election period. A 1 In modern English (International Standard Version): For this man, even though he was one of the twelve, was going to betray him. The original text in Vulgata, the expression underlined: dicebat autem Iudam Simonis Scariotis hic enim erat traditurus eum cum esset unus ex duodecim set unus ex duodecim 2 news article or an opinion article was included in the corpus if the theme was a presidential candidate. Thus, the focus is on the person-centered stories. To limit the data, I selected news coverage addressing only three or four candidates in each election, the nominated candidate of the Republican or Democratic party in each election and his main opponent in primaries: George W. Bush (Rep), John McCain (Rep), Al Gore (Dem) and Bill Bradley (Dem) in 2000, George W. Bush (Rep), John Kerry (Dem), Howard Dean (Dem) in 2004, and John McCain (Rep), Rudy Giuliani (Rep), Barack Obama (Dem) and Hillary Clinton (Dem) in 20082. Then there are three text corpora based on texts and transcripts in The Contemporary Corpus of American English (from hereon: COCA). I will compare the election data with the following data:    conservative and liberal news and opinion magazines ((3.8 million words), conservative and liberal political talk shows in cable TV (4.1 million words) conservative and liberal radio news and news commentaries (4.3 million words) The news magazines corpus is comprised of the following publications: Harper’s Magazine, The Mother Jones and The Washington Monthly for the liberal media, The American Spectator, National Review and U.S. News and World Report and USA Today Magazine3 for the conservative media. Fox News will represent conservative cable TV news and MSNBC liberal cable TV news. With regard to radio news data, I have chosen Rush Limbaugh data from the COCA to represent conservative radio news, even though the data is really from the short-lived Limbaugh cable TV show in 1992-1996. As there are no other suitable radio data available, and as can be expected that Limbaugh, the most popular political talk radio show host in the United States (Hall Jamieson & Cappella 2008, 46; Rush Limbaugh 2011), based his TV show on his radio show, it has to serve as “mock” radio data. National Public Radio will represent center-left radio news and news commentaries. In some cases I will compare data of this study with the entire COCA (410 mill. words) in order to illustrate how metaphorical expressions may be applied in American English in general in the 1990s-2000s. 2 With respect to the election 2008, I started to collect data in May 2007. So the selection of the candidates then was more a lucky guess of who might be nominated and who might be the main opponent. 3 USA Today Magazine has no connection with the USA Today newspaper. 3 The objective of this study is to investigate how certain pragmatic and grammatical factors affect the occurrence levels of buy and sell metaphors. Thus, the approach is principally quantitative, although some qualitative analyses will be carried out on the basis of quantitative findings. The variables explored in this study are: 1) genre/subgenre (political news, election news, written news, spoken news, TV, radio) 2) partisanship (liberal or conservative, liberal for Democratic and conservative for Republican) 3) multiword units based on buy or sell (for instance, sell out or buy it) 4) metaphorization level (the proportion of literal and metaphorical occurrences) 5) the evaluative value of the metaphors (in the election data only) 6) timeline (in the election data only: the 2000, 2004 and 2008 presidential elections) 7) the targets of evaluation (the election data only: presidential candidates and the Republican or the Democratic Party) There should be variation in different news genres, as opinion articles and TV talk shows differ, for example. Political cable TV and radio talk shows have emerged as significant political media in the United States after the 1980s. Opinion columns in periodicals are an old and respected form of political discussion, hence, they may be less open for colloquial expressions—such as buy or sell metaphors—than the new media. Election news represents in this research design a genre in which competition and partisanship together may induce more aggressive metaphor use. As conservative and liberal argumentation also differs, there should be variation in the partisan metaphor use. Partisanship could also be called more broadly a political ideology, and as the research corpora are comprised of liberal and conservative news media, they are strictly speaking not partisan, or at least not institutions of the Democratic or the Republican Party. The publications and the broadcasting channels in the conservative corpora, such as National Review, The Wall Street Journal, Fox News or Rush Limbaugh, represent each their respective conservative strand, but they do not adhere to the Republican Party. As for the liberal corpora, MSNBC cable channel and 4 publications such as The Nation or The New Republic follow their idiosyncratic liberal affiliation with no direct ties to the Democratic Party. National Public Radio is an independent non-profit membership news organization, which yet has sometimes been labeled liberal (McCauley 2002, 267). Hence, it represents the moderate, center-left voice in the political arena in this study, in juxtaposition to conservative talk radio (Limbaugh). I will pay special attention to multiword types of metaphor, since there is some evidence that collocations or phrasal expressions may be significant in metaphorization (Deignan 2005, 145, 160-161, 178-180, 210; Stubbs 1995, 382-386). Moreover, as my topic concerns metaphorical verbs, the objects and subjects of these verbs contribute to the metaphorical sense. In addition, I will explore how metaphorization levels vary: what is the proportion between metaphorical and non-metaphorical occurrence of buy and sell expressions? Criticism of the opponent party with negative metaphors is to be expected in the election data, but are there other differences in the evaluation? It is also possible that both parties (through partisan media) use the same metaphorical domains, as Charteris-Black (2004, 65-69) found in his study on British Labour and Conservative Party manifestos. Although metaphors arose from the same domains, there was variation in the individual realizations of these domains (ibid). As regards to terms conservative and liberal, I will follow a rather conventional definition of these political affiliations, as already was demonstrated by the choice of the comparison corpora. Conservative means a proRepublican ideological leaning with an emphasis on issues such as limited government intervention, strong national defense, individual responsibility or conservative cultural values, whereas liberal means a pro-Democrat leaning with an emphasis on issues such as progressive social or economic policies with government intervention, multilateral diplomacy or liberal cultural values. Metaphors play a role in the framing of political issues and politicians (see e.g. Mio 1997). The evaluative role of buy and sell metaphors in election data will be addressed from the aspect of their targets in the texts, with a focus on presidential candidates and political parties as targets. The timeline of the three elections in the 2000s is too short for comprehensive diachronic research, but as this period extends from pre-11/9 era to the 5 later phases of the Iraq war and the end of the George W. Bush presidency, the election corpus represents a period of crucial changes in American political life. Thus the diachronic analysis may yet reveal some directions in metaphor use. The immense and far-reaching changes in communication technology have greatly shaken media from the 1990s onwards, the growth and the significance of the Internet as the most important. Because of this, the years 2000, 2004 and 2008 may well represent a phase when online media emerged as an equal competitor of the traditional print media, radio and TV. There is some previous research on the rise of negativity in news (see e.g. Cohen 2008, Dagnes 2010). Negativity is often seen as an outcome of the hectic news cycles and tabloidization in the information age. Therefore this paper addresses specifically negative metaphors and their rise in news. The comparison corpora will not be used in the diachronic comparisons, as the COCA data chosen for this research design is not suitable for this. The COCA corpora in this study are all from the 2000s, except one, the conservative radio corpus, which is from 1992-1996. It should yet be beneficial to have data from a period when the incumbent president was a Democrat (Bill Clinton), as all other data represent mainly the George W. Bush’s era, the period in 2001-2008. Buy and sell were chosen as a research topic because in the exploratory phase of this study it seemed that their use varies in partisan election news and that buy metaphors seem to be a relatively new phenomenon at least in written genres. They also seemed to be clearly evaluative. Because the United States is the world’s most influential representative of free-market ideology, it may seem obvious that metaphors such as buy and sell which highlight free choice in transactions have emerged. They are, however, not the most frequent metaphors in political discourse. So far, studies have proved that conflict or contest metaphor occur very frequently in political discourse (see e.g. Howe 1988). There are, however, little studies on less frequently used metaphors in political discourse and their role. Furthermore, quantitative studies on variation or change in the language of American political news are scarce. 6 To sum it up, the purpose of this study is to find out answers to these research questions: 1) Do buy or sell metaphors occur in greater numbers in any genres of political discourse, are they genre-specific? Are there partisan differences in this respect? 2) Do buy or sell metaphors occur in multiword units and are there partisan or genre differences in the occurrence of these units? 3) Do buy and sell multiword unit metaphors occur more frequently than their literal counterparts? Are there partisan or genre differences in this respect? 4) Do buy or sell metaphors increase or decrease in the timeline of the three elections and does partisanship or a particular election (2000, 2004 or 2008) have any role in these possible changes? 5) How are buy and sell metaphors used in election news as evaluation? How does this evaluation, especially when it is negative, vary with respect to candidates, parties and the timeline of three elections (2000, 2004 and 2008)? Finally, there is the question of how these differences or changes can be accounted for, even though this may be at best tentative. Explanations of why these metaphors behave as they do can arise from various premises. I will focus in this paper on some wellknown phenomena linked with conservative and liberal political thought and practices and some prevailing views on mass media in the 21st century. My approach is to study metaphors in an empirical setting: how metaphors function in naturally-occurring language and how they work as parts of discourse the purpose of which is to persuade, to evaluate and to form in-group identities (such as partisan identity). To quote a researcher on Russian political discourse, the task is to analyze “discursive metaphors, i.e. the articulation of metaphors and their entailments at the level of the situated spoken or written text” (Macgilchrist 2011, 81). This analysis will be conducted by alternating quantitative and qualitative perspective since both are needed: quantitative analysis to discover the extent of use of different expressions, and qualitative analysis to identify metaphorical expressions and to assess evaluation in discourse, for instance. Charteris-Black (2004) calls this alternating procedure a dialogue: “each raising the questions that may be answered by the other” (32). 7 As abbreviations for metaphorical expressions based on the stem words buy or sell, I will use terms ‘buy metaphor’ or ‘sell metaphor’, (buy or sell will be written in italics). The lemmas of BUY or SELL will be written in small capitals. I will use terms ‘literal’ and ‘non-metaphorical’ to refer to non-metaphorical use of expressions (in Figures, abbreviation for literal is ‘lit.’ and for metaphorical ‘met.’). The term ‘non-negative’ refers to expressions which have a positive evaluation or neutral evaluation. For conservative and liberal, I will use ‘cons’ and ‘lib’ for short in the Figures and Tables. There is a list of abbreviations regarding the corpus data in Appendix A. 8 2 Studying metaphor in discourse In this section, I will present how metaphor in discourse has been studied and define some terms and definitions with regard to metaphor identification. I will introduce some ideas of political language and metaphors from the viewpoint of corpus studies, cognitive semantics and political communication research. I will also sum up how genre-specificity and metaphors have been examined in previous studies, with a special interest in quantitative metaphor research. Previous research on metaphorization is also discussed. The etymological origin of the word ‘metaphor’ is from the Greek meta = with/after and phereini = bear, carry, thus, it is transferring meanings (Charteris-Black 2004, 19). One of the simplest, yet a valid and apt metaphor definition, was made by Burke: “metaphor is a device for seeing something in terms of something else” (Burke 1989, 246). It has also been said that the function of metaphor is to hide and highlight at the same time. A frequently cited quotation in metaphor research literature compares metaphor with a solar eclipse: “it hides the object of study and at the same time reveals some of its most salient characteristics when viewed through the right telescope” (Paivio 1979; cited in Mio 1997, 113). Metaphor comprises two parts: the explainer and the explained, or the interpreting and the interpreted. Here is an example of a sell metaphor from the election corpus: (1) In this, Obama's [health care] plan is not dissimilar from Obama himself—filled with obvious talent and undeniable appeal, sold with stunning rhetoric and grand hopes, but never quite delivering on the promises and potential. And so he remains the candidate of almosts. (Klein 2007) The sense of this passage is that when candidate Obama offers his health care plan to voters, he sells it. Political proposal is interpreted or explained through commercial transaction. The simple lexical unit sold carries the interpretative or explainer function. Health care legislation, a considerably complex political process, is highlighted as a commodity. Naturally, this is not the only way of thinking of a political proposal. This is the hiding aspect of metaphor: the vast options of thinking of ‘something as something’ temporarily retreat (Charteris-Black 2004, 9; Kimmel 2004, 277; Lakoff 1980 10-13, 236-237), because of the interpretation provided by the explainer. Sometimes the resulting metaphor seems to be more than a sum of its parts, sometimes a metaphor 9 seems just a synonym or one term in a semantic field, for instance simply to offer in example (1). This sell metaphor as an offer is not a free offer but an offer with expectations of something in return (in election news, usually a vote or other form of support). In this study, I will mostly follow a metaphor definition suggested by a group of researchers in Metaphor Analysis Project4 (2006): “metaphor is a linguistic phenomenon found in discourse which has a potential to be identified as a metaphor”. As this definition is very general, it has to be amended with a notion that some incongruity or semantic tension has to be found in this discourse event (Cameron 2003, 9). Cameron also posits that metaphor may be signaled by an incongruous lexical item, but that metaphorical interpretation can also be created in the discourse context or in the larger co-text (ibid.). Low and Cameron (2002) underscore the complexity of figurative speech in discourse (85-87), and as this study is based on authentic discourse, it is reasonable to approach metaphors as discourse events potentially metaphorical at many levels. In example (1), the lexical item sold is incongruous with the context, a health care plan cannot really be ‘sold’ as a commodity. In the co-text, an additional contrast is created with comparing Obama’s health care plan with the candidate Obama himself: the phrases sold with stunning rhetoric and never quite delivering refer to the plan and to candidate Obama at the same time. The signal, not dissimilar, shows that there is a comparison at the utterance level. This example illustrates how multifaceted metaphors may be in authentic discourse. From the viewpoint of rhetoric, a linguistic metaphor is a trope: it is a figure of speech, like hyperbole or irony (Cameron 2003, 2). From the viewpoint of semantics, a linguistic metaphor may be seen as an anomaly. It has sometimes been called ‘loose speech’ inducing use of Grice’s co-operative principle: the listener/receiver must create or conclude the meaning (Partington 1998, 110). From the viewpoint of philosophy, a linguistic metaphor creates a breach of logic, it seems to “state an untruth that has some truth” (Cameron 2003, 2). The Conceptual Metaphor Theory suggests that linguistic 4 Metaphor Analysis Project and MetNet, Lynne Cameron led this research project. Other members included e.g. Graham Low, Elena Semino, Zazie Todd and Paul Rayson. 10 metaphors are instantiations of conceptual processing, whereby a source domain and a target domain form mappings in mind, such as POLITICS IS BUSINESS5 (Lakoff & Johnson 1980, 3-6 56-60). This would represent the viewpoint of metaphor as a cognitive function. Terms in the field of metaphor studies vary, and some terms have become linked to specific theoretical pre-suppositions. As I will not adhere directly to a particular theory in this study, I will use the terms as follows: Vehicle6 for the area of life a metaphor arises from, usually, but not always, a more basic element than the phenomenon it refers to, for example the commercial transaction in example (1) Topic7 for the area of life a metaphorical phenomenon addresses, the content of an utterance, for example the political proposal in example (1) metaphor for a linguistic phenomenon found in discourse, which has a potential to be identified as a metaphor (Metaphor Analysis Project and MetNet 2006) and may have a tension or incongruity between the Vehicle and the Topic (but not always), metaphor includes metonymy, simile and comparison linguistic in the connection of linguistic metaphor means that all kinds of linguistic phenomena, such as grammar, syntax, phrases or object relations can be a part of metaphor (Low & Cameron 2002, 88) multiword unit for a metaphorical phenomenon that extends to more than one lexical word (Low & Cameron 2002, 85) discourse unit for a lexical, grammatical or phrasal unit which operates at linguistic level and participates in creating a sense 5 Conceptual metaphors are normally written in capitals, to separate them from linguistic metaphorical expressions, as the separation of these two is crucial in this theory. 6 I will use Vehicle with a capital V to separate the use of this term from the lexeme, 7 I will use Topic with a capital T for the same reason. 11 utterance level for a larger stretch of discourse, the level at which a metaphorical phenomenon can sometimes be identified, for instance metonymy for a structure where a part of the Topic is used to refer to the whole Topic, or where the expression indicates a causal relationship with the Topic, as in: they know who us the better buy [sic] (Borosage 2000) simile for any expression or a chain of expressions which compares an item, entity, idea or process to another, often signaled with like, as, as if or similar to, as in: voters who stick with a party two or three times when they're young, like consumers buying a product, tend to develop an allegiance that stays with them (Roth 2007) comparison for a phenomenon in discourse, possibly stretching over several discourse units, in which an item, entity, idea or process is compared with something metaphoricity cline for a phenomenon when some metaphors are more metaphorical than others, or where it is difficult to separate literal and metaphorical usage (Low & Cameron 2002, 86) activation for the thoughts, ideas or a certain interpretation a metaphor brings to mind, the aspect of literal buy or sell it highlights (often called entailment in the Conceptual Metaphor Theory) resonance for the phenomenon when a discourse unit relates to metaphorical sense although it is expressed in a non-metaphorical sense (Low & Cameron 2002, 86) conventional metaphor for an expression which is so lexicalized that the incongruity seems no more apparent 12 novel metaphor for an expression which applies a Vehicle in a way not previously attested semantic field for a phenomenon when different expressions seem to share a common semantic nominator In this categorization of terms and in the procedure for metaphor identification, I follow some notions suggested by Low and Cameron (2002, 85, 86, 87-88), Cameron (2003, 36, 9-13, 22-23), Deignan (2005, 86-88, 195-197, 30-32) and Steen (2007, 321-322), but the definitions and the choice of terms are my own. These guidelines suit the quantitative, discourse-centered approach I wish to apply. It should also be noted that I will not address metaphors in spoken discourse in great detail. It has been claimed by Cameron (2003, 267-268; 2007, 109-112) that metaphorical occurrence in conversation has specific features often neglected in metaphor studies. Although I analyze spoken discourse (TV and radio talk shows), neither the interactional nature of metaphors nor metaphors repeated or elaborated in conversation will be specifically addressed. 2.1 Metaphor in political discourse In the following sections, I will introduce some ideas of political language and metaphors from the viewpoint of political communication research and cognitive semantics. 2.1.1 Political language and metaphor studies It is generally agreed that political activity is by its nature linguistic activity: language does not just reflect the events, it is a part of the events (Geis 1987, 18; Heywood 2004, 3). Chilton (2004, 16-18) proposes that politics and language have a mutual evolution history: for the survival of species, humans developed systems of shared responsibilities, shared knowledge and shared procedures. In this process, cooperation was needed, and language serves cooperation. Language serves as a mark of in-group membership and it can create and maintain social order. But there is a contrasting force, too, the struggle for power. This struggle means competition for the assets and benefits in social life, a competition which sometimes amounts to coercion (Chilton 2004, 16-18, 45-47). To 13 avoid violence and to maintain civil society, the struggle for power has to be amended with negotiations which constitute the default position of politics (ibid.). An election, for instance, is one basic form of political negotiation. Even if this state of continuous negations creates complexity and unbalance in society, there seems to be no alternative. As Winston Churchill (1947) said: “[…] democracy is the worst form of government except all those other forms that have been tried.” Negotiations need a suitable language to be successful, as political goals can be achieved with persuasive language. Political persuasion or convincing can be direct or indirect. Due to the complex nature of metaphors and their assumed power to invoke assimilations and emotions, they are considered to be efficient indirect communication tools for social values or beliefs. Beer and De Landstheer (2004) state that "the power of metaphor is the power to understand and impose political order” (30). They emphasize that metaphors are important in the struggle to create collective meanings and to reflect the meaning of power (ibid. 7). There are also limits of using language as a political tool which can be seen in the sanctions society will impose on those who break the social code. In earlier eras, profanity, blasphemy or offensive language against a sovereign were considered a crime, for instance. In our age, political correctness is required for anyone who speaks publicly of minorities, for example, but sanctions do not extend to criminal punishments, they are more by nature social exclusion. As far as colloquial expressions are concerned, there have been different levels of acceptance for informal expressions in written genres or in public discourse. In sum, language use and its limits reflect the social structure of society. There is a noteworthy difference in the research of political language between Europe and The United States. In Europe, there is a strong tradition of Critical Discourse Analysis, but in the United States it is practically non-existent (Kaufer & Hariman 2008, 475-476). The American research tradition has focused on rhetoric and political communication in general, and metaphor studies of American political discourse have been based on political psychology or political persuasion in the fields of political science and communication (such as Howe 1988, Lakoff 1996, Mio & Katz 1997, Wilson 1990). The aim of Critical Discourse Analysis (CDA) is to reveal hegemonies created by discourses and to make power structures maintained by discourses visible 14 (Charteris-Black 2004, 29). Hence CDA has also an empowering, emancipating role. From the viewpoint of CDA, Chilton (2004) has proposed that the function of linguistic expressions in political discourse in general is to legitimize or to delegitimize, to represent or to misrepresent and to coerce (45-47). In this thesis, the analysis of power relations is limited to the analysis of some conservative and liberal positions expressed through metaphorical language in media, and I will not apply a CDA frame, which usually does not place great importance to frequency. Yet I concur with Chilton (1996) when he argues that metaphorical expressions may not be merely a stylistic choice since metaphors are often deeply embedded in discourses (37). In political discourse, they arise from the cultural and ideological context which interacts with language. Americans often hold that freedom of speech guaranteed in their Constitution allows a fairly equal struggle for power “in the marketplace of ideas”8 and do not consider their political system particularly repressive. Whether repression or freedom prevails in America is not an issue in this study. It may not be beneficial in this framework to argue that metaphors always reflect power or power deprivation (Charteris-Black 2004, 30). A general view on political news in the American context will be further elaborated in Section 3, where changes in news production and their effect on language use are addressed. Metaphors as political communication are one aspect of this study. My limited viewpoint, though, will be evaluation of some phenomena with metaphors. Many features of political framing and power relations expressed through language, as well as the multimodal nature of political communication (visual and symbolic, for instance) are excluded from this research design. As this study is mostly carried out with quantitative methods, it may yield some information on how the recurrence of metaphorical expressions is related to ideology and genre. These findings may later contribute to research on political communication at a more general level. 8 This phrase is said to be coined by Supreme Justice William Brennan in 1965, it refers to freedom of expression. 15 2.1.2 The Conceptual Metaphor Theory and the political metaphor I will now present some applications of the Conceptual Metaphor Theory in political discourse studies. As some of the previous research on conservative and liberal metaphor use in the United States is based on this theory, it is relevant to assess it. My summary is mostly based on Lakoff (1996), Chilton (2004, 2006) and Cienki (2005). In this context, I will use terms related with the Conceptual Metaphor Theory, which differ from the terms defined on pp. 10 - 11. In the Conceptual Metaphor Theory, linguistic metaphorical expressions are seen as instantiations of some cognitive metaphor, a thinking model (Lakoff & Johnson 1980, 36; 56-60). The theory suggests that abstract thinking is based on embodiment, our physical experience of the world. According to this theory, a source domain, usually more concrete, is mapped to a target domain in the mind. Conceptual metaphors are usually expressed as “A is B” structures, for example POLITICS IS BUSINESS. A linguistic instantiation of this conceptual metaphor would be, for instance: (2) Watching the two presidential debates, one can't help but notice the stark differences between each party's approach to these core issues. And it's hard to see how the general electorate is going to buy what the Democrats are selling. (Kudlow 2007) It is yet clear that not all buy or sell expressions can be identified as instantiations of this conceptual metaphor, which in itself is hypothetical. There is no clear evidence that this Conceptual Metaphor exists. Chilton has made some ground-breaking studies on political language and conceptual metaphor. Among his various studies, a study on Cold War discourse showed that security metaphors were often expressions that suit ‘container’, ‘path’ or ‘force’ schemas (Chilton 1996, 50-55). The container schema, for instance, emphasizes the importance of boundaries, the inside-outside contrast. Other frequent thinking models in political discourse seem to be ‘the front–back schema’ and ‘the up–down schema’ (Chilton 2006). Hence, politics could essentially be seen as going forward or backtracking and gaining or losing height. The basic thinking model of ‘above’ usually correlates with good or powerful, and ‘below’ with bad or being under control. The Latin-based expressions sovereign and subject reflect these senses. In a similar vein, only going forward is good, progression has a positive connotation. In modern Western politics, four other basic 16 thinking models have been found: ‘body politic’, ‘the ship of state’, ‘buildings’ and ‘machines’ (Peil 1983 as cited in Chilton 2006). Lakoff (1996) has made a suggestion for comprehensive, fundamental conceptual metaphors in American two-party politics: A STRICT FATHER for conservative ideology and A NURTURING PARENT for liberal ideology. Cienki (2005) has investigated this conceptual model using empirical data: he studied the language of the TV debates between Al Gore and George W. Bush in the 2000 United States presidential election. He found few expressions directly related to these concepts (Cienki 2005, 279). There are some other studies where Lakoff’s A STRICT FATHER and A NURTURING PARENT models have been investigated as possible explanations for American political discourse. In a study on the United States Congressional debates on the Iraq war, no clear evidence for partisan preference was found (Republican or Democratic), when the lemmas protect and punish were chosen as possible expressions of the Lakovian model (protect for Parent, punish for Father) (Miller & Johnson 2010, 47, 71-72). As the empirical research on Lakoff’s model has not been very productive, I will not apply it in this paper. While his ideas have received some positive response (Kövecses 2005, 174-176), they are often thought to be too simplistic, or not to correlate with the reality of political actions. For instance, national economy is not merely a linguistic concept or a thinking model: taxes cannot easily be replaced by membership fees, an alternative term for taxes suggested by Lakoff. If one fails to pay taxes, there are criminal sanctions, whereas if one fails to pay membership fees, consequences are not as drastic (Pinker 2007, 259-260). Moreover, NATION IS A FAMILY, a Conceptual Metaphor implied by this model, has been criticized for not having adequate explanatory power on capitalistic American society (Goatly 2008, 386-387). Pinker (2007) has argued that conceptual metaphors can only exist as competing predictions, like competing scientific theories, in need of empirical facts to be proved correct (260). Lakoff himself admits that his model is not fully developed and that there is no adequate methodology to empirically investigate these models (Lakoff 1996, 158). It has been said that “it is sometimes difficult when reading Lakoff to know where his political advocacy ends and his cognitive-linguistics scholarship begins” (Goldstein 2008). 17 2.2 Metaphorical Vehicles in genres There are two main questions with regard to buy and sell metaphors and genre: how genre-specific are these metaphors and is there any preference for certain types of metaphors in the political news genre? Since my intention is to study whether buy and sell metaphors are applied in a different way in different subgenres, it is relevant to examine previous studies on metaphors, their Vehicles and genre. Several scholars suggest that the most frequently found Vehicles in political discourse seem to represent a controversy of some kind: war, sport or competition (Deignan 2005, 27; Kövecses 2002, 62; Kövecses 2005, 174-176; Scheithauer 2007, 80, 84). Taking into account the struggle for power described in Section 2.1, this is not surprising. Journey, nature and family as Vehicles are frequent as well (Scheithauer 2007, 84). With respect to buy or sell metaphors, no studies addressing specifically them have been found. Although there are several case studies on how metaphors participate in framing political issues in the United States (see e.g. Ivie & Giner 2009, Kitis & Milapides 1997, Lee & Morin 2009, Scheithauer 2007), there are no studies on the genre-specificity of political metaphors. Vehicles in political news have not been studied exhaustively either. Howe (1988, 191) was one of the first to suggest that both war and sports metaphors dominate American political discourse. Mio (1997, 123-126) reported several Vehicles found in American studies, such as container, disease, machine, family or nature. Ivie and Ritter (1989) studied Republican George H. W. Bush’s representation of America as “heroic” in the 1988 presidential election as against his challenger’s (Democrat Michael Dukakis) more “managerial” framing of foreign politics. Kövecses (2005) has investigated cultural variation of metaphors and makes some observations on American politics. Vehicles such as war, personification (nations as persons, for instance) or sports are found in many cultures, but sports as a Vehicle seems to have various elaborated expressions in the American context, especially from football or baseball (Kövecses 2005, 120, 191). In China, the most usual domains for sports metaphors are table tennis, volleyball or soccer (ibid. 120). In addition, he points out that 18 show or entertainment are prominent Vehicles in American politics regarding especially elections (although he presents little evidence for this) (ibid. 187). 2.3 Quantitative metaphor analysis Quantitative methods have sometimes been applied in metaphor studies which focus on one genre or subgenre. Koller (2004) has studied metaphors in business news from the gender aspect and Charteris-Black (2004, 113-134) has studied metaphors in sports news, for example, but they have used a mixed-method approach combining corpus methods, the Conceptual Metaphor Theory and CDA. Hence, frequency as such has not been the main interest. Jäkel (2003) compared British and German financial news coverage, in which motion metaphors dominated. He comments on the question of frequency in this genre by saying that these metaphors probably comprise 90% of all linguistic metaphors in the research corpus, both in German and British business news (ibid. 313-315). So, metaphor frequency and genre seem to be interdependent according to his study. In a study on cross-cultural election night metaphors in media, Scheithauer (2007, 105) found that Vehicles are remarkably similar across cultures in this genre (journey, war, nature, for instance). Her data were the TV election night news in Britain (the 1997 election), in Germany (the 1998 election) and in the United States (the 2000 presidential election). Business as a metaphorical Vehicle ranks as the 5th most frequent in her study. The percentage for business metaphors, though, was low: 3% of all expressions are business metaphors (Scheithauer 2007, 92-93, 100). With respect to election news as a specific subgenre in political discourse, Vertessen and De Landtsheer (2008) make an interesting observation about the similarities of political language during elections and during crises (275, 282-283). A crisis in this context means an event that raises national anxiety level, such as a war. They apply a method called “collective evaluative measuring”, first developed by De Landtsheer as a “metaphor power model” (De Landtsheer 1998, 131-136; De Landtsheer & De Vrij 2004, 63-64, 172-177). In this model, metaphorical expressions are given a number for emotional value and intensity, and these numbers are then counted to present the affective density of a news report. Vertessen’s and De Landtsheer’s findings show the importance of metaphors in elections: a clear preference for metaphors during election 19 time was discovered in the Belgian news coverage they analyzed, whereas during the two other research periods outside election time, the affective, emphatic expressions seemingly decreased (Vertessen & De Landstheer 2008, 282-283). De Landtsheer (2009) comments on the lack of quantitative metaphor studies from her own theoretical viewpoint, political-semantic approach, in which metaphors are seen as rhetorical devices with no emphasis on any specific cognitive or linguistic theory. She says that her Metaphor Power Model is one of the few quantitative methods applied, although she concedes that some research has been conducted with mixed methods such as Zinken (2004). She also underlines the importance of studying metaphors as a political style (De Landtsheer 2009, 64). Following the research tradition in political communication, De Landstheer has found in her studies that metaphors increase in political discourse, when certain conditions are met. The conditions are as follows: Table 1. Context factors of metaphor use (De Landtsheer 2009, 73-74). context factors political crisis > military crisis > Economic crisis > Political party > Gender > Media technology > Media audience > Media organizer > Elections > who are stimulated to metaphor use Media, politicians Media, politicians Media, politicians Right wing, extremist Male Television Popular, broadsheet Commercial Media, politicians On the basis of De Landtsheer’s results, the research design of this study incorporates some of the variables attributed to metaphor frequency: extreme political affiliation, commercial TV (cable TV) and elections. 20 2.4 The metaphorization level: the proportion of literal and metaphorical expressions “No word is an island.” (Stewart 2009, 56) In this section I will address the proportion between metaphorical and non-metaphorical expressions. I will call this phenomenon metaphorization level. Literal and metaphorical proportion levels of lemmas have received little attention in metaphor corpus studies. Berber Sardinha (2008) says that when the term metaphoricity is searched for, or the term metaphorization, few results are found in metaphor research literature. Berber Sardinha studied differences of Language for Specific Purpose Portuguese (spoken business discourse) and general Portuguese. He found that there was a very high tendency in business Portuguese to use some words only in their metaphorical sense (Berber Sardinha 2008, 142-143). He also suggests models for probabilistic metaphor research (ibid. 143-144). In studies where phrasal structures have been investigated, some coincidental findings of metaphorization levels have been made. According to Stubbs (2003, 217), phrasal or multi-word expressions seem to be important features in naturally-occurring language, and thus, it is possible that these expressions carry more semantic value than single words or grammatical forms. Collocation is another phenomenon which has been studied in connection with metaphors, but also as a general issue in language acquisition or language change. Stubbs has investigated cultural connotation and collocation and has found some consistency with regard to some very general words, such as small, big or way. These words seem to be much more frequent in their metaphorical sense than in the literal sense (Stubbs 1995, 382-386; Stubbs 2003, 228). With regard to collocation and grammar, Deignan (2005) has made several contributions to metaphor research. Some of her findings indicate that differences of literal and metaphorical occurrence levels might be important, but this has not been her main objective. Considering the aims of this study, however, these findings are interesting and point out aspects of metaphorization that have received little attention. 21 Firstly, she suggests that based on her corpus research, some collocations, such as to pay a price or rise and fall, have a strong tendency to be used only in their metaphorical sense (Deignan 2005, 210). This observation may shed some light to polysemous words: patterns of language may offer clues to how they are acquired or learned (ibid). Their occurrence in fixed or semi-fixed patterns may help the disambiguation of senses (ibid.). Her observations could serve as a research model in this study: is the polysemous nature of BUY or SELL dependent on fixed expressions or linguistic patterns? Then there is the evidence of how some words do not occur at all as literal expressions in some grammatical forms. According to Deignan, budding, for instance, is 97% metaphorical and blossomed 98% (178-180, see also Berber Sardihna 2008, 128). All these observations confirm that the relation of the literal and metaphorical occurrences is worth investigating. Secondly, she argues that there seem to be several preferred forms or constraints in the grammar of metaphor (Deignan 2005, 145). Sometimes there are word class preferences: verbs, for example, have been found to count for 50% of metaphors in one study (Cameron 2003, 89). Deignan (2005, 160-161) posits that body part metaphors seem to fall into three categories: some relatively free (heart = center, hand = help), some relatively fixed (shoulders = responsibility) and some nearly always used with a word from the target domain (head of state). In an earlier study, Deignan and Potter (2004) found—almost as a by-product in a study on cross-cultural metaphors—that the metaphorization levels of body part words was high: 65% of head and heart occurrences and 50% of hand occurrences were non-literal in British English. Even though some of these expressions might be nearer to idioms than metaphors, Deignan (2007) argues that this could still be significant when examining metaphorization: metaphor and metonymy as formulaic phrases could prove to be a fruitful research area (104). 2.5 A summary: the research design To summarize, political communication research prefers to address metaphors as a style, a discourse strategy or as rhetoric, whereas the Conceptual Metaphor Theory addresses metaphor more as a fundamental thinking model. The style or discourse strategy approach emphasizes metaphors as a linguistic choice, while the cognitive view holds that thinking models define the linguistic expressions. 22 In this paper, the linguistic and pragmatic phenomena will be the major interest. My research interests in naturally occurring language—represented by corpora—are genrespecificity, metaphorical use of expressions based on buy or sell as compared with their literal use, multiword units of buy and sell metaphors and the evaluative aspect of these metaphors. In this study, buy and sell are not primarily seen as representations of some conceptual metaphors, although they may signify some similarity between political actions and commercial actions (such as an opportunity to choose between several options, at a cost). Principally, BUY and SELL are addressed in this study as polysemous words: they have several other senses in addition to their basic commercial sense. Metaphoricity is approached more as a cline than a clear division into literal and metaphorical, and in order to do that, the discourse context must be taken into account. Multiword unit metaphors will be specifically addressed. As Müller (2008, 58) says, language is the most elaborated vehicle for conceptual metaphors, thus, to look carefully at the linguistic structure is reasonable, this structure is not just “surface noise” (Deignan 2007, 101-102; Low & Cameron 2002, 84). Critical Discourse Analysis will not be applied in this study, as quantitative metaphor analysis, a method not typical for CDA will be carried out in this study. I will, however, consider Chilton’s notion of the culturally embedded nature of metaphors relevant, and this question will be addressed in this study to some degree, from the standpoint of buy and sell metaphors as intertwined with American business ideals. Mainly I will approach metaphors as a style and as a part of chosen ideological identity which may be reiterated by metaphor use. 23 3 Negative metaphors in news discourse In this section I will provide some ideas on how negative evaluation can be conveyed with buy or sell metaphors. I will also analyze how partisan politics and the structure of news media can affect the language in American political discourse. 3.1 Negativity and metaphorical expressions In a study on metaphor-based idioms and fixed expressions, Moon (1998) found out that 89% of these expressions have evaluative value (228). Even if the majority of metaphors are not idioms, evaluation may still be the raison d’etre of metaphors, the essence of them, possibly because metaphors enable indirect expression of evaluation. CharterisBlack (2004, 33) argues that one of the major contributions provided by Corpus Linguistics concerns evaluation: it has been found to be inherent at the discourse level even when lexemes are not particularly evaluative. However, negative evaluation in language is a large research topic, and in this study, I do not pursue to discover anything specific on negativity in language in general. As I will classify expressions into negative, neutral or positive, it is yet reasonable to define some textual and contextual factors of negativity. The partisan use of negative metaphors in news discourse is the principal interest, it will be further explored in Section 3.2. One objective point of departure to negativity is the definitions given in dictionaries for buy and sell expressions. In the Longman Dictionary (2005), for example, buy (as a figurative expression) is defined as to believe something that someone tells you, especially when it is not likely to be true. Many sell expressions, especially phrasal verbs, have a negative definition in dictionaries, such as sell out, tough sell, sell-by-date or sell short (The Oxford English Dictionary 1989). The etymology of proverbial sell expressions shows that the emerging expressions are often negative, such as sell one’s soul or sell someone down the river (ibid.). BUY or SELL as literal expressions seem to carry no negative weight. When they are used as metaphors negatively, the connotation usually arises from the object or subject of buying or selling, other lexical environment or syntax. The interpersonal nature of metaphors is also relevant: a writer or a speaker can express his or her own opinions with 24 metaphors, quote directly someone else’s opinion or refer to these possible opinions indirectly. For example in some cases in election news, the writer seems to communicate the feelings of the electorate which (s)he does not seem to share. The following extract (3) illustrates this phenomenon: (3) Find a plaid rock. Put a really good chameleon or a really good politician on a plaid rock, and they will turn plaid. Call it adaptability, call it retooling, call it selling out, it doesn't matter. It works. Al Gore's desire to become president overrode his own ego and made him willing to change his entire approach to campaigning. Bill Bradley, on the other hand, wasn't even willing to stop sucking on throat lozenges while giving his speeches. Gore became warm and fuzzy--many people vote based on which candidate they like most-- while Bradley remained cool, reserved, and remote. From first day to last he gave the impression, accurate or not, that he liked The People more than he liked people. (Simon 2000) Furthermore, there are different shades of negativity. Irony or asking rhetorical questions may provide a lighter negative frame than direct negativity. Example (4) shows how liberals may ironically frame Republican tax policies: (4) Michael Boskin of Stanford filled the same role in Bush Senior's White House and participated in the heretical act (for Republican true believers) of raising the top income tax rate to reduce the deficit. These two and some others are no longer pilloried as sellouts by the hardcore supply-siders (Newt Gingrich, who led the back-bench rebellion against George Senior's tax increase, has since retired to the stud farm). (Greider 2000) Another objective measure to gauge negativity is the use of no/not or other negative word in the context of metaphorical buy, which is more negative in general than sell. The particles no/not/never/none or other negative expressions (hardly, unlikely, doubt, refuse) are applied much more with metaphorical buy expressions than with literal buy expressions. In the election corpus of this study, for example, metaphorical buy expressions are used with negatives in 41% of the occurrences, whereas literal buy expressions are used with negatives in only 5% of the cases. It is however important to bear in mind that a mere no or not does not make an expression negative from the pragmatic viewpoint: what is bought, the object, is usually the evaluative element. Especially in conservative media, not buying is often used as a testimonial phrase, as praise to someone’s sensibility or judgment. An example: (5) There is nothing more dangerous to entrenched Washington power than a populist conservative who looks unlikely to buy into Washington's creature comforts. Take a close look at Governor Palin's record on ethics and energy in Alaska, and it becomes clear what this Beltway outburst is actually about. (“The Beltway Boys” 2008) In this study, negativity will be defined along these practical lines: is a presidential candidate assessed to be a good candidate for President? Is a party assessed as likely to win the election in the context where the metaphorical expression is used? Is the 25 expression in the context for or against a specific candidate or a party? The identification of the targets of evaluation and the evaluative value of expressions will be further elaborated in Sections 4.4 and 4.5. Because the emphasis is on the evaluation of the candidates, it may sometimes be necessary to interpret a fairly long passage of text to determine whether the negativity (or positivity or neutrality) expressed with buy or sell metaphors relates to a candidate, his/her party or other target. 3.2 Partisan news discourse, political polarization and negativity in news in the United States In this section, some aspects in American partisan news, polarization and negativity, will be addressed as possible factors increasing buy and sell metaphors in news language. In general, both the content and presentation of news have been changing over the last 20 years. It has been said that news reporting today shows a tendency for personalization, dramatization and emotionalism (Scheithauer 2007, 76). This usually means a preference of human interest or conflict stories which are presented in simplified black-and-white, sentimental terms. It may also mean exaggeration of the controversial aspect. Sometimes this kind of news production has been called either “infotainment” (news presented as entertainment) or “confrontainment” (news presented as controversies) (Lauerbach 2007, 1336). These trends should be found in election news, where drama and controversy are usually the prevailing topics and where soft news9 stories have increased. Furthermore, Cook (2001, 186) argues, that media cannot be called a monolith institution anymore, and its fragmented audiences cannot be considered a one, fixed audience. Fragmentation may be seen in the use of language that signifies in-group identity: metaphors may be one linguistic device for it, especially in partisan media. Since the 1960s, visual media have taken over from the newspapers and magazines as primary sources for news (Cohen 2008, 58-66; Patterson 2002, 176), and television has been known to provide drama to elections, either real or produced (Gulati, Just & Crigler 2004, 241). In 1987, the Federal Communications Commission in the United States abolished the Fairness Doctrine: the companies having license on air waves were no 9 soft news = news on social or personal issues (family, health, food, travel or science, for instance) as against “hard news” (news on politics, crime or war). 26 longer required to broadcast balanced views on controversial public issues (Whittaker, Ramsey & Smith 2009, 10-11). The Telecommunications Act in 1996 further deregulated the restrictions in broadcasting (ibid. 11-12). Thus, in the 1990s, conservative talk radio and cable channels of every political affiliation joined the competition for audience and advertising (Gulati, Just & Crigler 2004, 246; Cohen 2008, 58-61). In the 2000s, Internet emerged as a new force in the media environment, and blogs, YouTube and social media (such as Facebook) gained foothold very rapidly (Cohen 2008, 58-61). Cable news, for instance Fox News and MSNBC, have introduced a new kind of media sphere: more partisan news and news produced as entertainment (Coe, Tewksbury, Bond, Drogos, Porter, Yahn & Zhang 2008, 201-202; Gulati, Just & Crigler 2004, 243). The arrival of soft news followed the deregulation of broadcasting rules, and big broadcasting companies were forced to compete with the “infotainment” and human interest stories cable TV preferred (Patterson 2002, 76-78). In a study on news coverage of national government, clear evidence was found that news reports on national politics and public issues have significantly declined from 1980s to 2001 (Sinclair 2006, 334335). In election news, reporting on polls, strategies, gaffes and controversies have become the prevailing topics, the “game-frame” (Gulati, Just & Crigler 2004, 240-241, PEW 2008). According to one study, 53% of news was so-called horse race10 news in the 2008 general election (PEJ 2008). All these changes in society may have influenced language in the news. American political climate has long been described, justly or not (Galston & Nivola 2006, 5-10), as polarized. On the one hand, this polarization means divided government (Pfiffner 2000, 34): the majority in Congress and the President have been from opposite parties for 30 years, excluding the years 1993-1995, 2003-2007 and 2009-2010. On the other hand, polarization may mean growing or sharpening differences between conservative and liberal parties with regard to central political issues: the role of the federal government, taxes, education, foreign policy and several social issues (racial and gender equality, abortion, homosexual rights, religion’s role in public life, stem cell research, sex education in schools or gun control). Many of these issues date from the 10 horse race = elections presented as a competition only 27 1960s, when American society moved to the liberal end of the ideological spectrum. The opposite turn to more conservative direction took place in the 1980s: the decade saw the Reagan revolution, the creation of conservative think tanks and the re-alignment of the conservatives of different interests (Sinclair 2006, 36-66). In the 1980s, social conservatives, neoconservatives and business-oriented conservatives joined their forces and the Republican Party won three presidential elections in a row. Substantial parts of the American electorate, though, have moved towards independent identity as a voter: registrations as an independent voter have been growing all the time since the 1950s (Pfiffner 2000, 33-34; Patterson 2002, 39-40, 42-44). In 2008, 37% of the voters in the United States identified themselves as independents (Bardes, Shelley, & Schmidt 2004, 280). Thus, in the elections the victorious presidential candidate has to appeal to this large center of voters (Pfiffner 2000, 30-31). Sometimes this political culture has been called more fragmented than ideology-driven: there are no central issues or parties that would gather voters into permanent coalitions (Pfiffner 2000, 35). In the historical perspective, there has always been a wide variety of ideological branches in the United States. Some researchers wish to remind that the present parties and the present media are still less divisive than, for example, parties and media at the beginning of the 20th century or in the 1960s (Easterbrook 2006, 258; Galston & Nivola 2006, 510). Others simply claim that polarization is the price Americans have to pay for the First Amendment of the Constitution, freedom of speech (Easterbrook 2006, 257-258). In general, though, a shift to more negative reporting has taken place in American news media. This trend has been confirmed especially in election news, even when the stories which address polls and campaign strategies have been excluded from studies (Patterson 1993, 66-77). It has sometimes been said that there is no conservative or liberal bias in American media, just a negative bias (Patterson 2002, 64-66). Even the tone of coverage of the presidents has become negative (Cohen 2008, 88-92), only national crises such as the pivotal event of 9/11 make media revert towards a more patriotic or elevated style (Patterson 2002, 65). Furthermore, there has been some debate on what is called a “structural bias”: losing candidates and candidates with poor poll numbers receive negative coverage regardless of the party (Gulati, Just & Crigler 2004, 239). 28 Some concerns have been raised over how media have been replacing the party process for example in the nomination of presidential candidates (Pfiffner 2000, 21; Patterson 1993, 267) and how media corporations have pushed their competitors from the market (Bagdikian 2004, 3). Thus, the original idea of free speech and free political participation has been said to be distorted. It has also been suggested that it is difficult to say which has been more influential in the rise of the negativity, partisan polarization or cable news, since these phenomena emerged in the same period, from the 1980s onwards (Cohen 2008, 100-101). With respect to language use, it is probable that conservative and liberal argumentation have not remained unchanged by these changes in media in general. The comparison of the election news with the comparison corpora, for instance Fox News corpus, should reveal something of this development. The election news data in this study is to a large extent based on opinion periodicals: nearly all articles in the liberal subcorpus are from high quality journals and ca. 72% of the conservative subcorpus articles as well. There is a long and worthy tradition of opinion magazines in the United States: the conservative National Review and the liberal The Nation, for instance, have been commenting politics for decades and they are respected as voices of ideological criticism (The Nation 2011, National Review 2011). Their circulation is not very wide, but they have a certain status in the political sphere. Yet a format change may have taken place in these partisan magazines in the 2000s: blogging has become a major format in written political discourse, and many longer, reflective articles are now published both in print and online (Cohen 2008, 58-61). With the advent of blogs, interaction between journalists and readers may have moved from Letters to the Editor section to the sphere of blogs, where responses are more immediate. As a consequence, there should be changes in the language of these up-market, even elitist political journals in the election corpus. It is not possible in the frame of this paper to study how the immediate response factor advanced by blogging (or other ‘ubiquity’ factors in the information age) has contributed to the language of the political magazines. However, it is reasonable to assume that polarization, negativity, increased colloquialism, the “game-frame” in elections and a more rapid news production process have enhanced the number of buy and sell metaphors, since they are informal and tend to be used negatively. 29 4 The data and the method In this section, I will describe the contents of the election news corpus in detail. The comparison corpora are also described (conservative and liberal news magazines, cable TV and radio). I will also outline how an expression is defined as metaphorical in this study. The metaphor identification procedure applied will be illustrated in Section 4.3. All concordance searches were carried out with AntConc3.2.1 concordance program. The results of comparisons will be shown as standardized frequencies of occurrences per one million words (for short, hereon: sf), if not otherwise noted. In the comparisons of occurrences, the differences are sometimes reported in percentages. These percentages are based on the standardized frequencies, not the absolute numbers (the Sections 5 and 7.6 are exceptions). Detailed information about the expressions defined as metaphorical and classified as having a positive, negative or neutral evaluation is in the Appendix B. 4.1 The election corpus There are 1 392 articles in the election corpus, of which 753 are conservative and 639 liberal. The corpus data of 2008 were mostly compiled from a website RealClearPolitics, which is a center-right oriented political website offering a daily collection of news from other media outlets. This compilation took place during two periods: from May 2007 to October 2007, and from August 2008 to October 2008, to cover both the primary period and the general election period. In addition, some material was compiled directly from major media websites, for example The Wall Street Journal’s website. With regard to partisan lines, I have included in the corpus only the evident, self-proclaimed conservative or liberal publications. As a result, the articles are mostly from well-known periodicals or newspapers and there may be more opinion articles than news reports in the corpus. The corpus data for 2000 and 2004 elections was compiled from electronic media archives in the summer of 2010. The major publications in the 2008 corpus were now chosen as sources, in order to make 2000, 2004 and 2008 corpora comparable with each other. The 2000 and 2004 corpora were created by doing searches with words such as Bush, Gore, McCain, Bradley, Gephardt, Dean, Kerry, Edwards, primary, primaries, caucus, Iowa, New Hampshire, presidential, election, electorate, electoral, Democratic, 30 Republican, candidate or candidates. The articles thus found were then chosen if the major theme was a candidate in that election (one of the nine). The publications (and their word count) in the election corpus are as follows: Table 2. The publications in the election corpus.11 conservative media The American Conservative The American Spectator American Thinker (online) The Arizona Republic Bloomberg (online) The Boston Herald Commentary Magazine The Christian Science Monitor Forbes Human Events National Review RealClearPolitics-blogs (online) Townhall (online) Union Leader U.S. News and World Report The Washington Times The Weekly Standard The Wall Street Journal all word count liberal media 6 208 81 436 15 895 703 9 213 1 825 7 026 1 084 4 400 20 607 160 517 45 430 1 461 527 97 328 12 827 116 058 204 272 786 017 all 834 733 The American Prospect The Huffington Post (online) In These Times (online) The Nation The New Republic The New Yorker Rolling Stone Salon (online) Slate (online) word count 114 476 2 477 2 285 172 097 186 776 147 272 21 382 153 607 34 361 The corpora of 2000 and 2004 do not have so great a number of publications as the corpus of 2008, because of the limited archives: the archive of The New Yorker for 2000 is not free, for example, and the archives for Slate or The Washington Times in 2000 do not exist. With regard to the 2000 subcorpus, it may not be as representative as the others, because in the years 1999 and 2000 electronic archiving may not have been 11 The decision to collect data from RealClearPolitics.com was made in order to apply at least some degree of randomized, semi-objective data selection. The greater number of conservative publications is due to this choice, since the site seemed to offer a larger selection of conservative media than liberal. See also section 6 on the article length. 31 regular or consistent. However, the compilation process showed that there was a saturation point at the level of ca. 190 000 words, after which the standardized frequency of buy or sell metaphors did not notably change. When divided into two parts following the ideological lines, the conservative subcorpus is 787 017 words and the liberal subcorpus 834 788 words. When divided into smaller subcorpora, the word counts are as follows: Table 3. Partisanship and the subcorpus word count in the election corpus. partisanship conservative liberal all election 2000 262 484 225 205 487 689 election 2004 224 181 283 419 507 824 election 2008 300 152 326 109 626 261 all 786 017 834 733 1 621 550 Since the aim was to include in the corpus the basic, most typical text types in the written subgenre of political news, news reports and opinion articles were preferred as data. I have generally attempted to avoid any conscious topic choice when copying the articles, so as to achieve a wide range of topics (although the candidates were the principal interest, not any general issues). The only topics or text types that I have consciously discarded are polling news, interviews, book reviews and articles about the candidates’ spouses—although the last principle is fairly difficult to follow regarding Bill Clinton, not only a spouse, but a former President. Interviews resemble more spoken discourse, polling news are not by nature interpretative and thus may not have many metaphors (at least, not buy or sell), and book reviews represent a different genre. 4.2 The comparison corpora: news magazines, Fox News, MSNBC, Limbaugh and National Public Radio The election corpus is compared with three different news corpora: conservative and liberal news magazines, conservative and liberal cable TV news and conservative and liberal/moderate radio news. The four subgenre corpora word counts are as follows: 32 Table 4. Word count in different corpora.12 word count election news news magazines cable TV radio all conservative 786 817 2 093 964 3 298 924 1 226 972 7 406 677 liberal 834 733 1 736 062 814 156 3 037 749 6 422 700 all 1 621 550 3 830 026 4 113 080 4 264 721 13 829 377 The news magazines corpus is comprised of these publications found in the COCA news and opinion magazines section over the years 2000-2008: Harper’s Magazine, The Mother Jones and The Washington Monthly for the liberal media, The American Spectator, National Review, U.S. News and World Report and USA Today Magazine for the conservative media. In Fox News, programs such as Brit Hume show, Greta Van Susteren show, Hannity and Colmes show, Bill O’Reilly show, John Gibson show and Fox Sunday show comprise the largest part of the conservative cable TV corpus. In the liberal cable TV corpus (MSNBC), Joe Scarborough show, Chris Matthews show and Keith Olbermann show account for 57% of the words, and in the National Public Radio corpus Talk Nation, All Things Considered, Daybreak and Morning Edition shows for 68% of the words. In the Appendix C there is information on the comparison corpora in detail. There is a difference in the time span between the Fox News and MSNBC corpora: MSNBC corpus is from 2004-2006, as none other period was available in the COCA, whereas from the FOX News subcorpus, I have selected the years 2001-2008. The Limbaugh data is from the 1992-1996 (none other available) and the NPR (National Public Radio) data is from the years 2005-2008. There are somewhat different program or article types in these corpora, and hence metaphor use may vary. The main difference between the election news corpus and the comparison corpora is that both conservative and liberal news magazines and the center- 12 The numbers on three comparison corpora from The Contemporary Corpus of American English are based on information given by Mark Davies on the www.americancorpus.org site. The latest information update was April 12th,,2011. 33 left National Public Radio publish or broadcast not only political news, but, for example, news on financial, educational or environmental issues. News on books, entertainment, life style or health are also frequent, and in Harper’s Magazine there is a tradition of literary essays. In the Fox News corpus, the data is mostly from political talk shows, but there are some business shows also, such as Neil Cavuto which accounts for ca. 10% of the Fox News corpus used in this study. Some other shows (John Gibson or Bill O’Reilly for instance) have a mixed program profile: a selection of topical issues, political or other. 4.3 The identification of metaphorical expressions based on buy and sell The lemmas of BUY and SELL in the four subgenre corpora were manually analyzed and expressions were classified into metaphorical and non-metaphorical. There are 3 137 BUY occurrences and 2 278 SELL occurrences in all the corpora together (Table 5). Table 5. The numbers of all the comparison corpora. BUY, all occurrences N = 3 137 BUY and SELL expressions in the election corpus and in conservative 165 465 687 305 1 622 liberal 198 505 125 659 1 487 all 363 970 812 964 3 137 election news news magazines cable TV news radio news all SELL, all occurrences N = 2 278 conservative 169 294 414 195 1 072 liberal 129 492 96 489 1 206 all 298 786 510 684 2 278 election news news magazines cable TV news radio news all After closely reading and analyzing the contexts of potential metaphors, the occurrences were then classified as metaphorical or non-metaphorical. When a discourse event potentially metaphorical is identified, there are several factors to be taken into account: 34   the discourse units, which may include marginally metaphorical parts and parts either larger or smaller than a lexical unit, the paradigmatic or syntactic aspects of the metaphorical phenomenon (such as the object of the verb buy or sell, unusual use of syntax or defining the paradigm of buy or sell), and  defining the Vehicle and the Topic: is there any incongruity or tension between them? As the objective of this study is to analyze also more complex structures of figurative speech, the focus was at the discourse level in the identification of possible metaphorical expressions. It should be noted, though, that there are some metaphorical phenomena in which Topic is difficult to locate, such as ellipsis or substitution (Low & Cameron 2002, 87; Cameron 2003, 10-11). These occurrences will be analyzed on the basis of the other aspects of metaphor: the utterance level, the discourse co-text, the grammatical function or the metaphorical resonance, for example. As I will compare metaphorical and literal occurrences in Section 7, this identification process affects also the number of literal occurrences. I have excluded the nouns buyer and seller from the paradigm of buy and sell, except for the expression buyer’s remorse, which occurs only in its metaphorical sense. The metonymic a buy and a sell are included but only the expression a tough/hard/difficult/soft/easy sell occurs frequently as a metaphor. New coinages of verbs such unsell out, oversell or resell are included, as well as an adjective sellable. Some idiomatic or proverbial expressions, such as buy time or sell a bill of goods are included. Proper names (e.g. The Best Buy, The Selling of The President) and compounds (e.g. best-selling or best-seller) are excluded. Because my American informants 13 think that to buy votes, to buy a Senator, buy influence, favors or legislation are synonyms with bribe and have no clear figurative sense, these expressions are not included as metaphorical. The same applies for buy off. 13 Eugene Holman, Joseph Flanagan. It is yet possible that there are metaphoricity clines in some expressions, such as buy votes, as Joseph Flanagan underscored (e-mail to author 27th Sept, 2011). 35 4.3.1 The identification of the most usual types of buy or sell metaphors I will give some examples of how the metaphor identification was conducted. Example (6) shows a typical way of using sell as a metaphor in political discourse: (6) MR RUSSERT Senator Chafee, one of your Republican colleagues we talked about earlier, Lindsey Graham, "[sic]Who has spent weeks attempting to recruit Democratic support for a plan to restructure Social Security, said that Republican [sic], made a strategic mistake' by initially focusing on a proposal to create individual investment accounts.... He said the accounts by themselves will not fix the solvency problem Social Security faces as baby boomers begin to retire. We've now got this huge fight over a sideshow. It's always been a sideshow, but we sold it as a main event.' ” [sic] do you agree? SEN CHAFEE Yes, he's making some good points, and as I said, there are other ways to address Social Security: raising the caps, indexing for inflation. These go a long way to addressing the solvency. (Meet The Press, 2005) Senator Chafee is asked about the measures to reform Social Security in the United States when president George W. Bush had proposed private saving accounts as a way to proceed (Social Security is a federal program called Old-Age, Survivors, and Disability Insurance, OASDI). Tim Russert formulates this effort in his question as It's always been a sideshow, but we sold it as a main event. Sold in this context means offering to be accepted. Alternatively sold can also be interpreted to mean the quite vigorous political process where the President and others give public speeches and interviews on the subject to create political momentum. The discourse unit is one word, sold, but the sense of sold is elaborated in the context with the object main event (and the sideshow). Main event and sideshow reflect the relative importance of the offered solutions. The Vehicle for this metaphor is commercial transaction and the Topic is political action or process. Perhaps surprisingly, the metaphor sold does not sound odd in a context where the Topic is related to finances. Here is another example (7) from the radio corpus, from a National Public Radio talk show on the same topic, Social Security: (7) But what's clearly the case here is that on Capitol Hill, which is a very important audience, the president had a very rough week. He wasn't about, Alex, trying to sell the American public [sic]. He was about trying to sell other Republicans [sic]. You had Senate Majority Leader Bill Frist saying that the pacing of the legislation is going to be up to the public, initially sounding as if he was putting it off and distancing himself from the plan. (Chadwick 2005) At first glance, it seems that there is something missing, because the President is perhaps not selling (in the sense of to betray) the American public or Republicans: there is no object of what or where to he is selling them nor a phrase sell out. In this extract from spoken data, the speaker leaves out the preposition to, but to identify this as metaphor, it 36 can be assumed that the discourse context requires that the sense is sell to, to offer for consideration or acceptance. The Vehicle is commercial transaction and the Topic is the President’s political proposals. As for metaphors based on buy, here is example (8) of a buy metaphor in the sense to believe. The Vehicle is commercial transaction and the Topic is tax cut proposal: (8) If so, the turf belongs to him. In this week's poll, on almost every major domestic issue, voters prefer Mr. Kerry; on some, such as health care, he has widened his advantage in recent months. Although Mr. Bush retains a small edge on taxes, voters buy the Democrat's argument on rolling back the tax cuts for wealthier Americans. There are a couple of sleeper issues, like stem-cell research, that Sen. Kerry hasn't capitalized on yet. There's a lot to work with. (Hunt 2004) Here is another, a more problematic buy metaphor (9): (9) PRICE-HEADLEY-BIG: I think so, Neil. I mean weve seen a lot of complacency among investors. You know, usually, when the summer lulls and everybody starts to not worry about the market anymore, it gets a little bit time to worry. And so were expecting that the next few months could be time to go to a safe haven portfolio, looking for stocks like Newmont Mining, for example, as a gold stock. I think, you know, Mets, 18 percent. The S&P is up 10 percent. That’s a great place to be over the next year. I’m saying watch out below. ! CAVUTO: What about you, Tony? ! TONY-DWYER-FTN-MI: Yes, I do nt -- I do nt buy the big sell-off. [sic] I mean to get a big selloff, youre going to have to have bad news and not just complacency. Youre also going to have to get bad news, and the news is not getting worse. The news is getting better. Now I think certain IT companies like Microsoft, for instance... ! CAVUTO: Information technology. ! DWYER:... yes -- have been sort of overvalued based on their growth. 14 (Cavuto 2003) The Vehicle is commercial action and the Topic is the expectations in the stock market which can lead to big sell-off. That a literal sell-off can be used with this buy metaphor indicates a very high level of conventionality of this metaphor type. Example (9) shows how identifying metaphors in spoken discourse may be difficult because of the fuzzy context in transcripts: spoken communication often relies on repetition and on the possibility of clarification if needed. Hence, a metaphor intentionally produced by a speaker and understood as a metaphor by a recipient is not easy to identify in transcripts. 14 Neil Cavuto show seems to be a difficult show to transcribe based on the numerous exclamation marks and elliptic sentences. I did not mark the extract with [sic] because it would have made the text difficult to read. 37 4.3.2 Borderline cases in identification The main borderline type that is excluded as a metaphor is to buy in the sense of to bribe. As dictionaries state that buy off is a synonym for bribe, that phrasal expression is also excluded, even though in some cases it can mean trading of something else than money. There are 85 borderline expressions of this type in the corpora. Here are four examples of the excluded expressions: (10) It is the political convention of political conventions to pretend that political platforms matter. They don't. They are merely a way to buy off people who care deeply about issues, but can be bought off with the right language in political platforms. It's hard to imagine any President telling his advisers to follow a particular policy because of the language in his party's platform. This doesn't work for the poohbahs who are here to relive the good old days when conventions, like the press, really mattered. (Berman 2007) (11) What ever happened to John McCain's outrage at the Swift Boat ads and at Bush's failure to condemn them? McCain now says " I'd like to see him go furrier. But I'm not making this some kind of big issue. ”Was McCain bought off by the prominent speaking slot he was given at the Republican National Convention? Or was he promised a major cabinet post when Powell or Rumsfeld retires? The New Republic's Franklin Foer thinks it's because McCain still harbors presidential ambitions, for which he needs to burnish his regular Republican credentials. (Peters 2004) (12) Sanctioning Tyranny? joy [sic] Gordon's thesis that Iraqis suffer because of economic sanctions and not because of Saddam Hussein " Cool War, " November 20021 is sadly mistaken. Iraq's economy was already in shambles when I left the country in July 1980. Saddam was spending whatever money there was on weaponry. Rundown schools and hospitals were everywhere; water and electricity shortages were common. Yes, education was free, but only because Saddam used it to spread his propaganda. Yes, Saddam offered free medical care, but health care was used to buy ideological loyalty. To be treated in one of the better hospitals, you had to have a letter from a Baath Party official stating that you were a loyal member. (Biddle 2003) (13) From 2004 onward, however, McCain has been moving rightward again, emphasizing his support for the Iraq War and the War on Terror. So far, this move appears to have cost him support among independent voters and reporters without buying him many friends on the right. Conservatives still have the impression of him they formed when he was tacking left. Besides, even in the last two years he has taken some stands to which a lot of conservatives object. (Ponnuru 2007) As an example how a sell expression can have some metaphorical resonance, but is yet excluded, here is a passage describing how the John Edwards’s campaign is selling a briefing book: (14) Edwards's ads have been the inverse of Dean's. While Dean never talks about his upbringing and has yet to flesh out a comprehensive governing philosophy, Edwards's first ads were soft biographical spots highlighting his humble roots, and his latest ads emphasize his policy proposals. Like no other candidate, Edwards has a laser-like focus on specific policies that affect Americans in their everyday lives, from health insurance for children to help with tuition costs to reducing the burden of consumer debt. The campaign has so much faith in the power of specificity that its latest commercial mocks the whole idea of the political ad. "John Edwards believes America's challenges require more than can be said in thirty-second ads," says the voice-over--somewhat ironically, given how kind the 30-second ad has been to the senator's career. The spot consists mostly of a quote from, er, The New Republic, praising Edwards's domestic policy and offers viewers a phone number to order the senator's 64-page 38 plan. It's the only advertisement in the race that's selling a briefing book. There is no evidence that it's working. (Lizza 2003) This ‘selling’ could be ambiguous in the sense that a briefing book ‘sells’ (promotes) candidate Edwards as well. There is one expression which seems to refer to a purchase both literally and figuratively, with phrases people were too invested and bought into an illusion. I have included this expression as metaphorical: (15) He took prospective buyers on aerial tours, so they could scout property from the clouds. That many buyers would have no equity until the property was entirely paid off did not deter them. On opening day May 21, 1958, Phillips made $4.25 million. Two years and another $20 million in sales later, Phillips pulled out. He'd drawn back his curtain, but people were too invested to admit that they'd bought into an illusion. The Holly Corporation promptly pursued making Phillips's dream a reality. (Jeffrey 2002) There is a case (16) where the object resentment is unusual with buy and hence in the context it is ambiguous, but I have included this as metaphorical: (16) Steven Inskeep: How long do you think it took for America’s image to decline, and how long do you think it’ll take for it to get back? Undersecretary HUGHES: Well, it's been decades. In my own lifetime, I remember living overseas. I grew up, my dad was in the Army, and my mother always used to tell me that we were such a generous country, and yet we only bought resentment sometimes from the countries we tried to help. Yeah, one of my frustrations as a communicator is that a lot of times people don't even know what we've done. I was in Latin America, and on the way down, I asked our Latin America expert how much we had increased development assistance to Latin America during President Bush's administration. We've doubled development assistance to Latin America during President Bush's administration, and not one leader I met with knew that. (Inskeep 2006) Bought seems to refer to the result of helping foreign countries financially, and resentment is a side effect of this help. It could be said that this expression is a variation that belongs to the semantic field of bribe which buy shares. 4.3.3 Similes and proverbs One perhaps unexpected result is that in this data there are few similes based on stem word buy or sell. Some are more complex and can be interpreted only at the utterance level, such as this: (17) Even when they are aligned perfectly with the organization's strategic goals, many projects fail because there is an assumption that they somehow will be implemented, as if by magic. Like a shopper on a buying spree forgetting how the credit card bills will be paid or even how all the stuff will fit in the car the organization's leaders often fail to plan resource deployment so projects do not bump into each other or have teams squabble over common resources. (Graham 2006) 39 In this extract, the Topic is governing projects and the Vehicle is shopping behavior. Poor project management is directly compared with impulse shopping without required funds or transport for the things bought. The form of simile points out directly to weaknesses in managing several projects at the same time, although the content of this criticism depends on what one thinks of mounting credit card bills or delivering an overload of stuff home. In this case, the metaphorical discourse unit is the entire clause starting with the signal of comparison like. In one news article, a commercial simile is a starting point of the theme of presidential candidates as brands: (18) IT'S NO STARTLING insight to point out that presidential candidates sell themselves as if they were breakfast cereal or laundry detergent. So it was perhaps inevitable that this would be taken a step further by studying consumers . . . er, voters . . . along these same lines. Accordingly, a new survey asked respondents to define the candidates in terms of popular advertising brands. Bush supporters, for example, think Mr. Bush is Bud Light and Mr. Kerry is Heineken. Kerry backers see Mr. Bush as IBM and Mr. Kerry as Dell. (Steinberg 2004) Most of the comparisons are rather indirect, such as examples (19) and (20): (19) Candidates as cars: Would you rather buy a used car from Bush or Gore? The site asks people to select the kind of car each candidate reminds them of. Mr. Gore has slipped several percentage points lately, according to www.autotrader.com, as his car image has been changing among the company's target audience (blue-collar workers age 21 to 45).15 (Heinauer 2000) (20) In some respects, Kerry has been running a campaign more suitable for re-electing an incumbent senator: Look at my résumé, look at my record, look at my advisers. But as Ralph Whitehead, professor of public service at the University of Massachusetts, observes, "If you want to sell a vacuum cleaner, you don't talk about its virtue, you demonstrate it. And Kerry has not yet demonstrated his assets. He talks about them. On the other hand, with his strong position against the war, Dean has led." (Corn 2003) In example (19), candidates Bush and Gore are compared with used-car salesmen, an old adage of asking if someone is trustworthy. In the example (20), candidate Kerry’s campaign tactics is illustrated as poor by comparing him with a salesman who does not show the product but only talks of it. In all these cases, some aspect of political action is compared with commercial action. 15 In this example, two different, unrelated metaphorical expressions based on cars are applied: the used car salesman and candidates as cars. I will address only the first one which is based on sell verb and identify the Vehicle as commercial action. 40 There is one occurrence of sell in which a hypothetical structure and a hyperbole are applied (21): (21) The desperation of the supply-siders is most apparent in the columns of the Journal's editorial page, which harbors the most fervent apostles of the Laffer curve. This essential tenet of supply-side theory holds that tax hikes cannot raise government revenue because tax increases will slow down the economy, leaving the government larger chunks of a smaller pie. The Journal has frequently illustrated the point with a chart purporting to show that federal tax revenues have never risen above 20 percent of GDP and, therefore, never will. According to the Journal's logic, the upper-bracket rate hikes of 1990 and 1993 should have so devastated the economy that by now people would be selling off their tooth fillings and clubbing rats for food. It hasn't quite worked out that way, of course. (Chait 1999) In this case, selling off is literal, but the entire comparison can be understood as figurative. Selling off their tooth fillings and clubbing rats for food refer to the financial destitution which never took place. It can be asked if this expression is figurative and especially if the verb sell counts as metaphorical. I have included this expression as metaphorical. Then there are proverbs or other fixed expressions, such as sell one’s soul, sell your birthright or sell someone down the river. There are four expressions of this type in the corpora. Here is one example: (22) HAGERTY: Towey says if the court had gone the other way, the entire faith-based initiative could have been jeopardized. And he says the ruling comes at a critical time. After Hurricane Katrina, he says, religious groups rushed to the scene, and the government not only plans to reimburse some of them but wants them to play a larger role. Mr-TOWEY: I think it's going to bolster an awareness by governors and mayors that it's safe to partner with faith-based groups. I think it's going to tell faith-based groups, Oh, we can do this work without having to secularize and sell our soul in order to provide a public service.'[sic] It's a complete vindication of President Bush's Faith-Based and Community Initiative. (Norris 2005) There is a resonance in this expression, because the context is religion’s role in political sphere: the verb secularize creates ambiguity as it seems to be literal. 4.3.4 Discourse metaphors and comparisons Metaphor can occur in a longer stretch of discourse where the metaphorical resonance is created with metaphorical, marginally metaphorical and literal elements together. In the election corpus, there is this stretch of discourse in which the senses are related to each other and together create metaphorical sense: (23) "Politicians can't be overexposed," said Mr. Della Femina, the famous ad man. "Since they can have a life and death effect on us, no one gets tired of them. This is not, 'Am I going to get tired of Bruce Willis?' We're stuck with them. Given how much power they have, eight months of campaigning is a 41 drop. They are the only people selling a product you can't return. You're buying a president for four years."(Henninger 2004) In buying a president for four years, the sense is that a vote cannot be canceled and ousting the president is impossible. Thus, there is a direct comparison of a presidential election and a commercial action in the passage as a whole. The verbs sell or buy are not metaphorical as such. The words product, return, sell and buy together constitute the comparison, but none of them are clearly metaphorical by themselves. The reference to the movie star Bruce Willis also reinforces the ‘presidents as products’ comparison. Although this occurrence can be identified as metaphor in its entirety, it is not possible to say which discourse unit is metaphorical. So there is a specification problem. The Vehicle can be identified as commercial customer satisfaction, the Topic is the unsatisfactory level of information on presidential candidates. The cline of metaphoricity is also well represented by this example, and the identification must be made at the utterance level. There is incongruity in the utterances they are the only people selling a product you can't return and you're buying a president for four years, because a president is not a commodity on sale, and election follows different rules than a commercial action. Because the activated sense, satisfaction, has been created in the co-text before these expressions occur, these utterances are easily understood. Even a highly conventional metaphor, such as buy time, can be reformulated in discourse in a way that requires not only identification but also interpretation: (24) To put all this in context: Defense secretary Robert Gates admitted candidly in mid-March that without sectarian reconciliation among Iraqis the”strategy won't work. ”Indeed, the entire point of the surge is to bring such reconciliation about by, in Gates’ words, “buying the Iraqis time.”[sic] But that's the problem. The United States is ever more dearly buying time, and Iraq is ever more freely spending it. As this article goes to press, the parliament is set to embark on a two-month vacation, during which, if current trends hold, 200 more American troops will be killed. The Democratic Party, fresh from its wins in the midterm elections, understands this. (“The Bitter End” 2007). In this case, the direct quote by Gates, buying the Iraqis time, is somewhat unclear. The Vehicle is elusive as the usual buy time sense (to delay something in the hope of a better result) does not fit this context. It is possible that the Secretary of Defense Gates meant to use the phrase buy up, which literally means to buy quickly something as much as possible (Longman 2005). In the later reformulation the writer adopts Gates’s phrase, and a more complicated sense of time being not of essence in preventing violence becomes manifest with the phrase dearly buying time In addition, the commodity aspect of time is 42 activated with the new formulation spending time freely. The Topic is the war in Iraq in which the American actions are not deemed to be successful. The metaphorical sense is created at the utterance level, in which buy and time lexemes resonate as ambiguous expressions, as do dearly, freely and spending as well. They are fit for both literal and metaphorical interpretation: dearly (with cost), freely (without boundaries or without cost) and spending (lose or use). In example (25), there is no comparison, but the utterance is hypothetical and not literal in some sense: (25) The political press is abuzz with overblown stories of a Clinton-Obama rift. There are some hard feelings, but less than you'd think, given the closeness of the primaries. But I have a seven-point plan for uniting the Obama and Clinton wings of the party: Attack, attack, attack, attack, attack, attack. Attack. The way to unite an internally divided organization is to identify an external threat. The Obama delegates will be buying beers for the Clinton delegates once they're focused on how disastrous a third term for Bush-McCain would be. But no one is telling them. If the Democrats do not spend the remaining days of their convention -- hell, the remaining days of the campaign -- in an all-out assault on the ruinous Bush-McCain policies, they will lose. (Begala 2008) In this case, the hypothetical prediction of how the Clinton delegates she won in the primaries will become Obama supporters in the Democratic National Convention is described with the phrase The Obama delegates will be buying beers for the Clinton delegates. The buying act is a pragmatic reference either to persuasion, bonhomie or gratitude expressed as buying drinks. The ambivalence of the sense arises from the pronoun they in they’re focused: they can refer either to all delegates, to Obama delegates or to Clinton delegates. There is some part of the Topic missing: the conversion of the Clinton delegates to Obama supporters extremely rapidly, perhaps? If the basis for metaphor is that the verb buy should be metaphorical, this expression should be excluded. There is no incongruity or tension in the verb buy or in at the utterance level. Hence I have excluded it. Some of the more unusual objects of buying or selling are peanut butter, dog food, blue sky, sermon and sunshine: (26) Bauer manager Frank Cannon says Mr. Forbes is rejected merchandise. With Mr. Forbes losing in 1996 and spending an estimated $60 million on his current campaign, Mr. Cannon says, conservatives "know him, and they're not buying that brand of peanut butter”. (Calmes 1999) (27) There, flipping pork, was Lamar Alexander -- an eminently decent and qualified man who has visited 64 of Iowa's 99 counties, but just can't sell the dog food. (Tapper 1999) 43 (28) Obama merely needs to make the case that he's serious enough, so that when his ideas are matched with his charisma, he can build a national consensus for policies better than any other candidate. "We all bought Bill Clinton's blue sky," says Iowa Waterloo County Democratic Chairwoman Teresa Wolf, who is not backing a candidate. "Why not Obama's as long as it's a blue sky." (Dickerson 2007) (29) Bush and Gore Are Making Religion a Big Issue, and Praying Voters Buy The Sermon It’s been generations since so many politicians have talked so much about Jesus--and their personal relationship with him. Just back from a jog, George W. Bush is game to talk more. He slings a towel over his shoulder and plops into a seat on the patio of the Texas governor's mansion. "What are we talking about?" Faith. "Good. I like talking about faith." Five minutes later, he's discoursing on the Crucifixion and whether Jews can enter heaven. ("Governors don't decide that.") (Foer 1999) (30) “The fun... the sunshine, the relaxation. Our idea is to get the public to go see a ball game, win or lose, " said P.K., who then told a young Bill Veeck, who would later become one of the greatest impresarios in the history of baseball, to lant [sic] the ivy. It was his way of selling the fans the sunshine. (Cohen 2001) The sense of peanut butter and dog food is that a candidate cannot make himself accepted by the voters. Blue sky and sunshine are figurative expressions for some idealistic, too-good-to-be-true character of things. When voters buy the sermon, they accept the talk about religion the candidates make. Sermon is in the context a metonymy for religion. Then there are three rather extraordinary occurrences: I don’t want them buying all the chairs, sell chopsticks and swing voters are voters you can rent but not buy: (31) Both Obama and Clinton have talked about bringing all interested parties to the table to create universal health insurance. But Obama, who like Edwards distinguishes himself from Clinton by refusing contributions from political action committees and Washington lobbyists, also says, “I don’t mind insurance and drug companies having a seat at the table. I just don’t want them buying all the chairs.” (Moberg 2007) (32) "If Mr. Zhu Rongji had come to see me in the Oval Office as he saw Bill Clinton, I would have told him, 'Sir, you're going to stop persecuting Christians. You're going to stop bullying our friends on Taiwan. You're going to stop pointing missiles at us. Or you're going to have sold your last pair of chopsticks in any mall in the United States of America." (“On the Stump at Ames” 1999) (33) You can bet that between now and November, both candidates will be trying mightily to make that connection. "Swing voters are voters, you can rent, but you can't buy,” says Goeas. "They don't lock in permanently. Either you have to be there constantly or be the last one they're talking to." (Allen, Cohen, Smith & Tobin 2000) In example (31), candidate Obama takes a stance on the role of the insurance companies in the prospective negations over health care reform. An idiom a seat at the table which means having a say/power, participating in decision making, is used to describe the role of the insurance companies. A seat at the table is reformulated into I just don’t want them buying all the chairs. Buy seems to be used in the sense that comes close of the sense 44 bribe. Seat is converted into chairs. Chair is a metonymy for a table, and in this idiom it is a table at which negotiations take place. Hence, there seems to be a chain of metonymies: negotiations>a table>a seat/chair at the table. In example (32), there is a metonymic phrase Or you're going to have sold your last pair of chopsticks in any mall in the United States of America. Going to have sold is not metaphorical as such, because the larger context is China’s export to The United States, and selling is the essence of export. Chopsticks, in contrast, is metonymic, a part for the whole. In addition, it is a pejorative expression referring to Chinese products and how unessential or poor they are. There is also another metonymic structure, any mall for the domestic market of the United States. If the basis of a metaphor is defined in the largest possible context, it can be said that it is hardly sensible to assume that China’s premier Zhu Rongji would personally enter a mall in the United States to sell chopsticks. Therefore, the expression is metaphorical. The verb sell in sell chopsticks is a borderline case, it should be excluded if only the metaphorical value of words based on sell is a criterion for inclusion. Yet the utterance level is highly figurative, and it also represents a contrafactual structure: Pat Buchanan, from whose speech this citation is, was never elected President of the United States. When swing voters are referred to voters you can rent and not buy, as in example (33), the voters are compared to commodities. Yet it is interesting that in this case, one of the most usual senses of buy, to bribe, is not activated. Grammatically, voters you can buy, is a phrase similar to expressions often used in the sense of bribe. Rent could even strengthen the bribing scenario, but in fact it does the opposite, it disambiguates the sense from bribe to something else. The inconstancy of the support from swing voters is revealed as the sense of this expression (because renting is a temporary thing). The fickleness of swing voters becomes clear in the next sentence which describes the relationship with them. Different elements of buy become activated and this may explain the difference between the senses: when buy is used in the meaning of bribe, the (criminal) possession element of transaction is activated, but when buy and rent are used in the sense implied in this extract, the nature of possession as relative or conditional is activated. Buy has the sense of ‘becoming a constant supporter’. 45 Lastly, there is a phrase the better buy. In example (34), candidates Gore and Bush in the election of 2000 are compared by using a phrase: they know who is the better buy [sic]. (34) If elected, Bush will have a mandate and a majority to enact the reforms. Worse, he’ll get bipartisan cover from the New Democrats, the money wing of the Democratic Party. He’s also for negating the minimum wage, leaving it to the states; dismantling environmental regulation; repealing affirmative action; limiting a woman’s right to choose; and generally fronting for the corporate leveraged buyout of government. Think of the Gingrich Congress without Clinton’s veto. That’s why corporations are flooding the Republican Party with record contributions. They know who is the better buy [sic]. (Borosage 2000) A buy is here metonymic: the result for the process, a noun derived from the verb. In addition, the phrase is elliptic: who refers to Bush, who has been named, and to Gore, and Gore has not been named. This phrase resonates significantly with the sense of buy as bribe, because flooding with record contributions is in the same semantic field as bribe, but campaign donations are not criminal. Yet they can be considered acts of support with expectations, a return of favors in the future. In all these four examples (31, 32, 33 and 34), it can be doubted that the verb buy or sell as such is metaphorical, but I have included them as metaphors. In (31) and (33), the sense is very close to bribe or buy off. As insurance companies can have a legitimate role in negotiations and no reference to criminal or unfair activity is found, I have decided to include this multiword, phrasal figurative expression into metaphors. All the chairs phrase seems to refer to the too great influence of insurance companies in the health care in general. In (34), money, buyout and donations all refer to literal financial activity, but it does not seem to be criminal or dubious. The sense of the better buy is a better choice. Yet it resonates with the literal sense, if the aspect of supporting a candidate with record contributions is the activated sense. 4.4 The target of evaluation in the election news corpus The closest context of buy or sell metaphors, the subject or the object of it, for instance, was the measure in defining who or what is the evaluative target of an expression. If this practice did not reveal evaluative value, the target of evaluation was analyzed in the ca. 200 word context of the metaphorical expression. The expressions were divided into categories with regard to targets as follows: 46 1) presidential candidates, ten subclasses: Bush2000, Bush2004, McCain2000, McCain2008, Gore2000, Bradley2000, Dean2004, Clinton2008, Giuliani2008, Obama2008 2) the Republican Party, its policies or politicians or representatives (other than in 1, includes vice-presidential candidates and primary candidates other than 1): REP2000, REP2004, REP2008 3) the Democratic Party, its policies or politicians or representatives (other than in 1, includes vice-presidential candidates and primary candidates other than 1): DEM2000, DEM2004, DEM2008 4) other targets: OTHER200016, OTHER2004, OTHER2008 (when an expression addressed candidates of both parties, it was classified as OTHER) The target of evaluation was sometimes difficult to specify. The topic may be general in the stretch of text in which buy or sell metaphor occurs, but the larger context addresses one candidate’s tax reform proposal, for instance. There are also several expressions which address several candidates at the same time: these are generally included into OTHER category. The division into classes Republican, Democrat or OTHER was sometimes difficult, as many political issues are general, but can also be assessed to be core issues of either Republicans or Democrats. Labor unions and feminism, for instance, have traditionally been Democratic issues, and limited federal government and taxes Republican issues. In examples (35) and (36), I have interpreted that the expressions evaluate both McCain and his campaign tactics and Obama’s media coverage, but I have defined the target to be McCain (neutral) in the first and OTHER (negative) in the second: (35) The McCain campaign sees Palin as more than tough enough to withstand the feeding frenzy here this week -- and in fact, they think it might help her. Schmidt said the more the media digs into the story of Bristol Palin's pregnancy, the less the public will respect the press (and the more voters will buy the McCain team's assertion that Barack Obama is getting a free pass from reporters). The campaign ratcheted that argument up Tuesday, canceling a McCain interview with CNN after the network's Campbell Brown body slammed campaign spokesman Tucker Bounds Monday night. (Madden 2008) (36) Nonetheless, it's clear that Obama has lost control of this campaign. And he will not seize back the initiative with the sometimes halting, conversational and sadly reluctant sound bites he has been producing. The excitement Obama created at the beginning of the year has vanished, perhaps 16 I will use OTHER capitalized, when it is a term referring to a target of evaluation, in order to separate the term from the lexeme. 47 because his campaign (and, yes, many columnists) bought into the McCain campaign's demonization of the big rallies. Absurdly, McCain is now contesting the terrain of change -- and doing so at celebrity rallies of his own. (Dionne 2008) In the second example, the word demonization makes it related to McCain. Yet, as the theme in general is Obama’s campaigning style and its problems, I have concluded that the expression addresses candidates of the opposite parties together, and hence it is classified as OTHER (negative). In example (37), I have defined the target to be candidate John Edwards and not Hillary Clinton or Obama who are both referred to. Thus the target is the Democratic Party. The evaluative value is neutral, although if the target were Clinton, it would be negative. (37) If that's the case, then I think it has huge implications for both Obama and Edwards--namely, it suggests both men have been running the wrong campaign. In a nutshell, I think Edwards is the most obvious beneficiary of any buyer's remorse Hillary might provoke. Edwards is a white guy from North Carolina with a thick Southern accent and a moderate record in the Senate. Barack Obama is a black guy from Illinois with an exotic name and only three years in federal office. If you're an Iowa Democrat who wakes up in a cold sweat in January worried about nominating someone (Clinton) who may be a tough sell in Middle America, the person you probably run to is Edwards, not Obama. (For the record, I think there are all sorts of reasons to be skeptical of this snap judgment about who plays better in Middle America; I just think it's one the typical caucus-goer is likely to make.) (Scheiber 2007) Then there is a case where President George W. Bush and America are addressed together: (38) For governance, Bush has the task of leading a country that believes in American exceptionalism in a world in which that idea is, for many, off-putting if not repugnant. This is why Bush has taken pains to explain that the "nonnegotiable demands of human dignity" are not just American but universal, the gift of God or, if you will, imperatives imposed by secular ideas of liberty and equality. America's specialness has been its good fortune in asserting and trying to uphold those ideals earlier than others and having the strength, and therefore the obligation, to advance them around the world. Abu Ghraib makes that message harder to sell, but we must persevere. (Barone 2004) I have classified the target as Bush and the evaluative value negative, because the reference to Abu Ghraib war prison scandal is negative, and it is estimated to make selling a message harder for America, and the connection to America is via the presidency of Bush. The entire article, though, is very positive for Bush and emphasizes how his choices have been mainly right. 4.5 Negativity in the election news corpus Negativity was assessed according to the same principle as the target of evaluation: if the negative framing appeared in the close context of the metaphor, the expression was 48 classified as negative. When the context was not so close, expressions have been assessed case by case neutral, positive or negative. In many cases, the negative value of things bought is apparent (sell seems to be more neutral in this aspect). Examples (39), (40) and (41) represent unequivocal evaluation. The phrases phony optimism or claptrap are negative (39 and 41). Kerrys and Deans is more complex but the idea of not accepting Northern candidates in the South is clear (40). The context of buy as a metaphor indicates without doubt that the sense of buy is to believe (stupidly so). (39) Americans don't buy the phony optimism about Iraq that George Bush and Dick Cheney peddle -- their own intelligence agency doesn't believe it. If the question is framed, as the president prefers, whether it was good to have gotten rid of Saddam, most people say yes. Who wouldn't? (Hunt 2004) (40) North Carolina Sen. John Edwards has a powerful argument in his bid to be the Democratic nominee when he says, "What I give people is a candidate who can win everywhere in America." Translation: "We Southerners ain't gonna vote for no Yankee! You suckers up North will take our Clintons and Carters, but we just ain't buyin' Kerrys and Deans." (Maher 2004) (41) As Wright’s role model, James Hal Cone, put it: “If God is not for us and against white people, then he is a murderer, and we had better kill him. The task of black theology is to kill Gods who do not belong to the black community. . . . Black theology will accept only the love of God which participates in the destruction of the white enemy.” How much of this claptrap did Obama buy? Well, he stayed at Trinity for 20 year —until political expedience tore him reluctantly away. But has he really left the fold? (McCarthy 2008) It is not always the closest lexical environment that defines the negativity. In example (42), a reader has to assess if Bush-Rove politics are good or bad, and how does that relate to Obama’s campaign strategy (doesn’t mean he’s a wimp does not reveal evaluation without the context): (42) Obama won't do that, though, because his message is change, and because he has more integrity than McCain does. This is not "naïve" and "idiotic," as Terence Samuel at The Root calls Obama's response to the "lipstick affair" -- it is part of a larger strategy. Obama has shown as well as anyone that he is a rough-and-tumble politician who doesn't shy from a fight. But his campaign has made central his commitment to changing the way we do politics. That doesn't mean he's a wimp, but it does mean he can't buy into the Bush-Rove politics that McCain now espouses. Obama remembers, to recall the old adage, it's not worth wrestling a pig: You get dirty, and the pig likes it. (Fernholz 2008) I have identified example (42) as neutral: although the tone is generally positive for Obama, there is no direct stance on the tactics of the Obama campaign. There are some cases in which direct evaluation is avoided by using rhetorical questions, or other forms of indirect negativity (hypothetical structure, for instance). In example (43), I have assessed the evaluation to be neutral (the target is Obama). The occurrence in 49 example (44) is deemed as negative (the following prediction they won’t makes it negative). There is thematic repetition in both of these examples, and in cases like this, I have counted the occurrences as separate (two neutral buyer’s remorse metaphors in sample 43 and one neutral bought it metaphor and one negative voters will buy it in example 44. (43) Should Democrats Be Feeling Any Kind of Buyer's Remorse? As Democrats kick off their national convention to nominate Illinois Sen. Barack Obama as their nominee for president, there is little or no evidence that activists or insiders are having second thoughts about the party's standard-bearer. In other words, buyer's remorse has not settled in, and it probably won't unless Republican Sen. John McCain (Ariz.) nips Obama at the wire 10 weeks from now. Yet only the most uncritical party insider could avoid asking himself or herself the obvious question as delegates gather in Denver: Did Democrats, who two years ago placed no higher priority on selecting a candidate than on picking someone who could win back the White House in 2008, really pick the right person to carry the party's banner this year? (Rothenberg 2008) . (44) There's only one problem: there's a word for a spousal co-presidency in the English language or at least a system where one can ascend to higher office on the basis of marriage: It's called a "monarchy." So far, the press certainly has bought it. [….] The larger question, of course, is whether the voters will buy it, too. The guess here is that, ultimately, they won't. Royalism has never had that many fans on this side of the Atlantic. To the extent we've tried a similar idea recently -- with the current president, a/k/a "Junior" -- things haven't exactly worked out in spectacular fashion. 17(Stark 2007) Hypothetical structures are often neutral. In this example (45), the modal structure If you buy into this logic makes the assessment neutral: (45) For Gore, however, choosing a running mate may be more of a zero-sum game. [….] Of the 56 House Democrats from the swing states of New Jersey, Pennsylvania, Ohio, Michigan, Illinois, Wisconsin, and Missouri, 52 voted against PNTR-a ratio of 13 to one against the Clinton-Gore position.) And if you buy into this logic, as The Wall Street Journal's Al Hunt did in a recent column, you end up suggesting such candidates as Missouri's Gephardt or Durbin of Illinois. (Meyerson 2000) The following case (46) shows ambiguity: the interviewee considers Obama the best choice for Alabama veterans, but also admits that it’s a hard sell in Alabama. I have defined this occurrence as negative. (46) Eyes alight, Smith leaned forward and described how his battalion lost a paratrooper in the district of Sangin; he helped carry the flag-draped coffin onto a C-17. But, soon after his unit left Afghanistan, in April, “the Taliban were back in Sangin, because we didn’t have the troops to hold that city.” Smith now lives in Huntsville, Alabama, where he is Obama’s point man for veterans statewide. “I was told I’ve got the hardest job in the campaign. I’ve got to explain to veterans in Alabama that a liberal black attorney is better for them than a conservative white war hero. It’s true—but it’s a hard sell in Alabama.” (Packer 2008) 17 The topic is Hillary Clinton as a candidate and as a spouse of Bill Clinton. 50 Example (47) is an opposite case. The negative, ironical condemnation of the too easily satisfied press is one argument, but the Obama campaign’s prospects of winning the election look good in the view of this passage; thus, it is positive for Obama. If the target were identified as OTHER (press, mass media), the value would be negative: (47) This is what the journalists have been reduced to: the level of indentured field hands at a Russian monastery. With such a castrated press corps in tow, Obama doesn't have to work very hard to "sell" his message. The whole process has been streamlined, politically and culturally, to smooth the spread of the party's propaganda: The speech is already written, the press is already on board, and everybody's already working together to crank out those fish patties. (Taibbi 2008) Example (48) shows how one can avoid taking a clear stance immediately first by making a direct quote and then by asking a rhetorical question: (48) "His staff is making a good faith effort to reach out to pro-lifers," says Michael J. New, an assistant professor at the University of Alabama and author of a widely circulated National Review Online essay on how Giuliani could court pro-lifers. New has also discussed pro-life issues with Giuliani's staff. A focus on adoption is a "little more substantive than 'safe, legal, and rare,'" says New, referring to Bill and Hillary Clinton's abortion mantra. Still, "I'm really not sure that pro-lifers are buying what he has to sell here." What is Giuliani trying to sell? At a campaign stop in Fort Dodge, Iowa, on August 6, the mayor focused on adoption policy. (Continetti 2007) I have defined the first and the second metaphorical discourse units pro-lifers are buying and what he has to sell here as negative. The last discourse unit, trying to sell is defined as neutral for the candidate Giuliani. In total, then, this extract contains three buy or sell metaphors. 51 5 Buy and sell metaphors in the corpora: the quantities and the proportional distribution There are 396 buy metaphors in the corpora and 355 sell metaphors (in total, 751). The absolute numbers for identified buy and sell metaphors are as follows: Table 6. Metaphor distribution in the four subgenres. buy N = 396 election news news magazines cable TV news radio news all conservative 35 25 155 33 248 liberal 36 22 35 55 148 all 71 47 190 88 396 sell N = 355 election news news magazines cable TV news radio news all conservative 66 33 64 36 199 liberal 73 29 32 22 156 all 139 62 96 58 355 With respect to buy and sell, metaphors occurring in simple discourse units that express the basic figurative senses seem to prevail in this data: expressions based on the lemma BUY with the sense to believe something that someone tells you, especially when it is not SELL likely to be true (Longman 2005) and expressions based on the lemma with the sense to try to make someone accept a new idea or plan (ibid.) dominate. In the corpora, 90% of the 396 figurative buy expressions (N = 358) have the sense to believe, and if buyer’s remorse is included to the same sense, the number is 93% (N = 367). Within sell metaphors, 56% of the 355 metaphorical sell occurrences (N = 198) have the sense to try to make something accepted. I will call this sell + [OBJECT] -type, as the metaphorical sense depends on what is sold (sell oneself is included into this category). There are other types of sell metaphors that come very close to this sense:  tough/hard/difficult/easy/soft sell, 10% of the occurrences (N = 36) (difficulties in making something accepted) 52     sell well/does not sell, 6% of the occurrences (N = 21) (to succeed, to fail) selling point, 5% of the occurrences (N = 16) (an advantage in an offer) sell short, 3% of the occurrences (N = 9) (to undervalue or to belittle something) be sold, 3% of the occurrences (N = 9) (to accept something wholeheartedly) In total, these types account for 81% (N = 289) of the sell metaphors in this data. The other sell types that occur in these data are:   sell out, 13% of the occurrences (N = 44) (to betray) proverbs, such as sell down the river, sell one’s soul, and novel metaphors, 6% of the occurrences (N = 21) The metaphorical senses of these expressions were checked in three dictionaries: The Oxford English Dictionary (1989), Collins COBUILD English Dictionary (1995) and Longman Dictionary of Contemporary English (2005). Some information given in The Oxford Dictionary was presented in Section 1. There is somewhat different information on some expressions in the dictionaries. For buyer’s remorse, only The Oxford English Dictionary recognizes the expression and yet it is dated as 1966. The Oxford English Dictionary gives for hard sell the sense aggressive salesmanship or advertising, and enumerates examples for it from the 1950s in American usage. There seems to be no separate metaphorical sense for hard sell or tough sell in the OED, only the general sense of metaphorical sell, to convince (someone) of the worth of (something), is given. The Longman dictionary gives for hard sell two senses: a way of selling something in which there is a lot of pressure on you to buy and if an idea is a hard sell, it is difficult to get people to accept it). For tough sell, Longman gives only the metaphorical sense (something that is difficult to persuade someone about). The Oxford English Dictionary defines one use of to sell (to succeed, to fail) as finding purchasers and the Collins dictionary as when something sells it, is bought by the public in large quantities. No separate metaphorical definitions are given. 53 6 The genre-specificity of the metaphorical buy and sell In this section, a genre as a possible pragmatic factor and an explanation for occurrence levels of buy or sell metaphors is investigated. The question is whether buy or sell metaphors are genre-specific in the subgenres investigated in this paper (election news, news magazines, cable TV news and radio news). The partisan nature of these metaphors will also be addressed. A genre or a subgenre is, however, a problematic variable as it is no easy task to define the crucial differences between subgenres. Multiple criteria can be applied to define the differences. The spoken/written parameter is one factor which divides election news and news magazines in contrast to cable TV and radio shows. In the written news genre, there is a difference between news reports and opinion articles. There are practical factors in news production, such as the publication format, the length of an article or a segment, publication/broadcasting frequency and the planned audience. In the election corpus, for instance, there are daily newspapers and weekly or monthly published periodicals. One difference within the election news corpus is that in liberal media articles tend to be longer than in the conservative media (especially in The New Yorker). That is the reason for the greater number of conservative articles in the election news corpus (753 as compared with 639). Similar differences in editorial policies or favored formats may be found in other partisan media, and these practices may affect metaphor use. In spoken media, one genre parameter is that talk can be a monologue, a dialogue or a group conversation. Spoken news communication tends to be more casual and less reflective than written news reports. However, talk shows today are always somewhat pre-planned, the language use is more fabricated than natural conversation and participants or interviewees are usually expert commentators. Yet the elite opinion journals of the election news data (such as National Review or The New Yorker) represent a very different subgenre than cable TV, for instance. It is possible to compare spoken political discourse on commercial cable TV—which often aims at controversy—and the reflective opinion articles published weekly or monthly. They are certainly both political discourse, but they represent the opposite ends of the news commentary spectrum. The designated audience, the scope of the articles, the program format or the editorial choices may greatly vary. In this study, the production aspect of news business is not addressed. 54 However, it is important to underscore that genres and their similarities or differences depend on various factors excluded from this research design. First, the general occurrence level of all buy and sell metaphors in the four subgenres in partisan media is analyzed (Figures 1 and 2). The buy occurrence levels (Figure 1) follow the subgenre boundaries: buy metaphors are frequent in the election news and in the cable TV news but scarce in the news magazines and in the radio news. Hence it seems that buy metaphors are specific for one written news genre and one spoken news genre. There are slightly more buy metaphors in the conservative cable TV news and conservative radio news than in their liberal counterparts, but in general, there are no systematic partisan differences. Figure 1. Buy metaphors in the four subgenres by partisanship. N = 396, standardized frequency x/1 000 000 words Sell seems to be an election metaphor, as there are from 54% to 92% more sell metaphors in the election data than in the other subgenres (on average 72%) (Figure 2). There seems to be no great partisan variance in the election data. In the cable TV, sell seems to be popular in the liberal MSNBC, and in the radio data there is a conservative edge in the frequency. 55 Figure 2. Sell metaphors in the four subgenres by partisanship. N = 355, standardized frequency x/1 000 000 words In Sections 6.1-6.5, the link between subgenres and multiword unit metaphors will be examined, but in this analysis, pragmatic subgenre factors are limited to the written/spoken parameter and political news/election news parameter. The major factor that is addressed in this analysis is political partisanship. In this investigation, I will follow the notions on multiword unit metaphors discussed in Section 2.4. For instance, even if the distribution of buy metaphors suggests that they are characteristic for two subgenres, there may yet be differences in the buy multiword units inside one subgenre. I have chosen for genre-based analysis three categories of buy metaphors and four categories of sell metaphors. These multiword types are: 1. buy it/buy that/buy this (from hereon: buy it -type)18, -as in: do you buy that? 2. buy into -as in: not all of Edwards' campaign advisors initially bought into the idea for the spot 18 Buy it/that/this -type category is limited to those expressions in which it/that/this is not followed by a complementary clause or noun. Expressions such as “Asked by Byron York of National Review if he bought that theory, McAuliffe said he did” are excluded from this category, they belong to the buy + [OBJECT] category. 56 3. buy + [OBJECT] (other than it/this/that) - as in: voters bought this anti-Washington message 4. sell + [OBJECT], includes sell oneself - as in: in Texas, as elsewhere, the tort reformers exploited the rate hikes as part of a scare campaign to sell reform 5. tough/hard/difficult/soft/easy sell (from hereon: tough sell -type), - as in Giuliani was among those chosen to present the Administration’s idea of the issue to the public, and he did not use the soft-sell approach 6. sell well/ does not sell -type (from hereon: does not sell -type) -as in: for another, protectionism has not sold well on the presidential level since Herbert Hoover's campaign in 1928 7. sell out -as in: instead, Kerry looks like more and more of a sellout every time he shifts positions 6.1 Buy it -type There seems to be a preference for buy it -type metaphor only in the cable TV: Fox News and MSNBC use this phrase more when compared with other subgenres (Figure 3). It has a conservative edge in all subgenres but one, news magazines. Figure 3. Buy it -type metaphor occurrence. N = 129, standardized frequency x/1 000 000 words 57 Table 7. The absolute numbers of occurrence for buy it. buy it N = 129 election news news magazines cable TV news radio news all conservative 6 5 61 15 87 liberal 5 5 14 18 42 all 11 10 75 33 129 Here is an example of typical buy it occurrence from the conservative cable TV: (49) GIGOT:[…]We are back with our Dan Henninger and Mary O'Grady. Also joining the panel is Columnist John Fund. Dan, the narrative here on some -- part of some people who liked Eliot Spitzer what he did was that he was a great man who did great things as a prosecutor and just got tripped up here on this personal foibles. Do you buy that? HENNINGER: I don't buy it at all. I think Eliot Spitzer was a man who was simply unhinged from normal norms of personal and professional behavior. (“Fox News panel discusses”, 2008) This phrase may be frequent because it often works as a somewhat confrontational opening for discussion in TV talk shows: do you buy is a regular opening question, and it encourages an interviewee or a participant to respond, for instance, no, I don’t buy that. It may be typical for talk show genre in general. The entire COCA was consulted to find further evidence for the frequency of this phrase either on TV or on radio (Table 8). The subcorpora of the COCA in which buy phrases in general can be expected to be found are business sections of newspapers and business magazines (the COCA codes for these are NEWS: Money and MAG: Financial). These suborpora rank as second and ninth in this frequency list. All the spoken corpora in the COCA represent TV or radio channels. Spoken media have a high level of occurrences of this phrase in all subcorpora (independent TV channels rank as first, others from third to eighth, except National Public Radio, which is 17th). The trend is clear: this multiword unit is frequent in no other parts of the COCA than the business magazines, the business sections of newspapers and the TV or radio shows. 58 Table 8. The occurrence level of [buy] it|that|this -phrase19 in the COCA. [buy] it|that|this -phrase SECTION NAME PER MILLION TOKENS WORDS in the entire COCA, frequency ranking 1 2 3 4 5 6 7 8 9 [….] 17 SPOK:Indep MAG:Financial SPOK:FOX SPOK:NBC SPOK:CBS SPOK:CNN SPOK:MSNBC SPOK:ABC NEWS:Money [….] SPOK:NPR 43.51 42.23 41.58 41.43 40.01 32.98 31.93 29.31 28.48 […] 24.01 189 222 246 241 494 665 26 443 213 […] 409 4,343,343 5,256,801 5,916,135 5,817,458 12,347,287 20,161,550 814,156 15,114,475 7,477,721 […] 17,031,484 Even if it is taken into account that some occurrences are from business talk shows on TV, the numbers indicate that buy it metaphorical phrase may be typical for mediated spoken discourse. As this phrase is used on all major TV channels (NBC, CBS and ABC), on cable TV (CNN in addition to Fox News and MSNBC) and on independent channels (Oprah, Geraldo and Limbaugh), its use may not be genre-specific for political discourse. 6.2 Buy into and buy + [OBJECT] -type Buy into and buy + [OBJECT] types differ little with regard to semantics: both express believing in something. Yet the occurrence levels are different with regard to partisanship. Buy into is systematically applied more in the liberal media than in the conservative media, as can be seen from Figure 4: 19 The search was made with the following command: [buy] it|that|this 59 Figure 4. Buy into –type metaphor occurrence. N = 82, standardized frequency x/1 000 000 words Table 9. The absolute numbers of occurrence for buy into. buy into N = 82 election news news magazines cable TV news radio news all conservative 8 5 26 4 43 liberal 10 6 9 14 39 all 18 11 35 18 82 Here is an example of a typical buy into expression from the liberal election news: (50) But even if Obama's courtship of evangelicals and Catholics succeeds in smoothing over this sticking point, is it a wise strategy? By emphasizing their religious credentials, Democrats are implicitly buying into the right's phony charge that Democrats hate religion (see Ronald Aronson, "All Ye Unfaithful," page 52) without necessarily shifting the terms of what it means to be religious. (Posner 2008) Buy + [OBJECT] -type is slightly favored in conservative media, except in the conservative news magazines (Figure 5): 60 Figure 5. Buy + [OBJECT] -metaphor occurrence. N = 147, standardized frequency x/1 000 000 words Table 10. The absolute numbers of occurrence for buy + [OBJECT]. buy + [OBJECT] N = 147 election news news magazines cable TV news radio news all conservative 15 9 60 13 97 liberal 12 5 12 21 50 all 27 14 72 34 147 Here is a typical example of buy + [OBJECT] expression from the conservative election corpus: (51) Another Bubblehead blind spot. I’m bumping into a lot of critics who do not buy the legitimacy of small town mayorship (Palin had two terms in Wasilla, Alaska, population 9,000 or so) and executive as opposed to legislative experience. (Noonan 2008) Even if there is no great difference in the sense of these expressions, it is possible that the type buy + [OBJECT] is preferred in the conservative media, because it is more direct. Buy it multiword unit is also direct to the point of being blunt. They both have a quality of straight rejection when used with a negative no/not, and they are casual and colloquial. Maybe they thus better suit conservative ideology, which is generally considered to be a more practical and realistic approach than liberalism. The stereotypes of populist and 61 plain-spoken conservatism and elitist and complex liberalism can hence be illustrated in the use of these phrases. 6.3 Sell + [OBJECT] -type Sell + [OBJECT] -type is the most frequent type of the metaphors in this analysis. As was illustrated in Figure 2, sell is typical for election news genre, and the sell + OBJECT occurrences follow the subgenre boundaries, although in the liberal data this genrespecificity is less apparent, because there is no great variance between the liberal election news and the liberal cable TV news (Figure 6). Figure 6. Sell + [OBJECT] -type metaphor occurrence. N = 198, standardized frequency x/1 000 000 words Table 12. Absolute numbers of occurrences for sell + [OBJECT]. sell + [OBJECT] N = 198 election news news magazines cable TV news radio news all conservative 38 18 38 22 116 liberal 32 11 26 13 81 all 71 29 64 35 198 62 Here is a typical example of sell + [OBJECT] -type in the liberal election news: (52) But the assumption that the Democratic Party can make itself credible on defense through the personal heroism of its leaders trivializes its problem--much as the Republican Party does when it finds black and Hispanic spokesmen to sell its urban policies. (“Our Choice”, 2004) There may yet be differences in what is sold, though, but that is not in the frame of this study. I will suggest one possible difference: liberals seem to have referred to the Iraq war as something George W. Bush sold to Americans in a negative sense. This seems to concern the liberal MSNBC, and also the liberal election news (in 2004 and 2008) and the liberal news magazines. In contrast, the conservative media have used this type of expression neutrally about the positions of the Bush administration. The Iraq war became a topic in 2002, and the war was begun in 2003. Here is a typical example of sell the war phrase from the liberal media: (53) Because what matters is not what's on television but the ghostly afterimage that lingers in our minds and clouds our vision after we turn off the television. It’s all too easy to envision a time when, the White House will no longer feel compelled to sell a projected war to the American people but can merely pitch it to Jerry Bruckheimer, whose new series will show us why we need to spread our influence-preferably by force, since diplomacy is less apt to translate into compelling TV-throughout the Middle East. (Prose 2004) 6.4 Tough sell and does not sell -type There is a liberal edge in the tough sell type metaphors in every subgenre (Figure 7). Figure 7. Tough sell –type metaphor occurrence. N = 36, standardized frequency x/1 000 000 words 63 Table 13. The absolute number of occurrences for tough sell. tough sell N = 36 election news news magazines cable TV news radio news all conservative 7 3 5 0 15 liberal 8 6 3 4 21 all 15 9 8 4 36 Here is an example of a typical tough sell expression from the liberal election news: (54) If you're an Iowa Democrat who wakes up in a cold sweat in January worried about nominating someone (Clinton) who may be a tough sell in Middle America, the person you probably run to is Edwards, not Obama. (For the record, I think there are all sorts of reasons to be skeptical of this snap judgment about who plays better in Middle America; I just think it's one the typical caucus-goer is likely to make.) (Sheiber 2007) There is a conservative edge in the does not sell-type occurrences in all subgenres except in the news magazines (Figure 8). Figure 8. Does not sell -type metaphor occurrence. N = 21, standardized frequency x/1 000 000 words 64 Table 14. The absolute number of occurrences for does not sell. does not sell N = 21 election news news magazines cable TV news radio news all conservative 6 1 10 2 19 liberal 2 0 0 0 2 all 8 1 10 2 21 Here is an example of a typical does not sell expression from the conservative election news: (55) But no one likes naked calculation, and Clinton worries Democrats [sic] traumatized by the experience of their last few presidential candidates. Because liberalism typically doesn’t sell in American presidential politics, liberal candidates tend to run as culturally conservative centrists—i.e., phonies. It sank both John Kerry, who couldn’t even order a Philly cheesesteak properly, and Al Gore, who adopted three such utterly different personas in three national presidential debates that his performances could be a case study in abnormal-psychology classes. (Lowry 2007) Expressions tough sell and does not sell share a semantic field: they refer to difficulties in getting something accepted, in political discourse usually a proposal or a political candidate. For some reason, liberals prefer tough sell -type and conservatives prefer does not sell -type in the data. This kind of difference shows how ideological identity and metaphor use may be linked. The does not sell -type occurs rarely in the sense to sell well (to succeed) in the conservative data: it is used negatively in 81% of the cases (17 out of 21 occurrences). When tough sell and does not sell constructions are counted together, conservative and liberal media seem to use these expressions fairly evenly (28 occurrences in the conservative media, 27 occurrences in the liberal media, and the standardized frequency for both 4). 6.5 Sell out -type Sell out -type is favored only in the liberal election news (Figure 9). There may be some other stereotypes here, such as liberal prejudice of deeming the Republican Party as easily betraying the interests of the people, but one reason for the frequency of sell out is liberal self-criticism. This will be further explored in Section 9 and in Section 10. 65 Figure 9. Sell out -type metaphor occurrence. N = 45, standardized frequency x/1 000 000 words Table 15. Absolute numbers of occurrences for sell out. sell out N = 45 election news news magazines cable TV news radio news all conservative 4 6 8 6 24 liberal 13 4 2 2 21 all 17 10 10 8 45 Here is a typical example of sell out expression in the liberal election news: (56) Neither Obama, who would like to argue that he could transform politics in the old Confederacy, nor Giuliani, who has been selling out to the likes of Ralph Reed since the 2002 midterms, should care so much about his chances in the Deep South. (Keating 2007) 6.6 A concluding note: buy and sell metaphors, genre and partisanship As was suggested at the beginning of these comparisons, some metaphor types may account for the frequency of buy metaphors in these subgenres. The Figures 3, 4 and 5 in Sections 6.1 and 6.2 show that there is some variation in the occurrence of the buy multiword types. Figure 10 shows the level of buy metaphors when buy it –type and buy + [OBJECT] -type expressions are excluded. The buy metaphors in the cable TV news 66 are mostly of these types. If they are excluded, buy expressions are genre-specific only in the liberal election news, and moreover, there is a liberal edge in the occurrences in all subgenres.20. Figure 10. Metaphorical buy expressions without buy it -type and buy + [OBJECT] type. N = 120, standardized frequency x/1 000 000 words, absolute numbers Table 16. Absolute numbers for metaphorical expressions without buy it and buy + OBJECT] -type. buy into + others, N = 120 election news news magazines cable TV news radio news all cons 14 11 34 5 64 lib 19 12 9 16 56 all 33 23 43 21 120 Second, the comparisons of sell multiword unit metaphors shed some light on the genrespecificity of sell in the election news subgenre: it is attributed to sell + [OBJECT] expressions especially in the conservative election news and sell out expressions in the liberal election news. 20 In the figure 10, all buy metaphor types are taken into account, such as buyer’s remorse and other metaphors, not only buy into -type. 67 7 The metaphorization level of BUY and SELL In this section, I will compare the occurrence levels of metaphorical expressions and literal expressions. There may be differences in subgenres when it comes to the proportion of metaphorical and literal buy or sell phrases: buy or sell metaphors may be favored more in some subgenres than others, and this can affect literal use. Yet it is not very straightforward to examine the literal versus metaphorical proportions, as there may be various reasons for the literal BUY or SELL to occur, some more accidental, some more systematic than others. The quantity of news on economy or on business in the political news genre is an obvious factor, as are the campaign finances (advertisement purchases, fundraising or donations) in election news. The financial backgrounds of the candidates are usually examined in campaign news, such as Bush’s career as a businessman, Hillary Clinton’s investments in cattle futures and real estate speculation, Kerry’s and McCain’s wealth and property management with their respective wives, both millionaires, or Barack Obama’s real estate deals. In the 2000 election, Gore campaigned on the economic achievements of the Clinton era, in 2004 Bush wished to enhance the ownership society, and in September 2008, worsening economy and worldwide fears of deep depression became the main issue in the presidential election. Health care insurance (buying it, making it accessible to all Americans) was one of the contested issues in the 2008 election. What could actually be expected is that in news genres where coverage of economy, business or finances is common, there would be less buy or sell metaphors, as one of the common sense expectations is that metaphors are effective only if they are from other domains than the theme or the Topics of a text. If there still is a relatively high frequency of buy and sell metaphors in election news, it suggests that this metaphor use is attributed to the situation, election, which thus overrides the contrasting domain principle. In what follows, I will analyze how the use of BUY or SELL in the literal sense varies in political news subgenres. The proportions of literal and metaphorical occurrences will be compared. The comparisons will be carried out with regard to different expression types. 68 These types are the same as in the subgenre comparison of metaphorical expressions (see Section 6 for detailed classification): 1. buy it 2. buy into 3. buy + [OBJECT] 4. sell + [OBJECT] 5. tough sell 6. does not sell 7. sell out 7.1 Buy it -type It seems that when buy it expressions are applied as metaphors, it sometimes has a reverse effect on literal use: when metaphors increase, literal expressions decrease and vice versa (Figure 11). This phenomenon happens mostly in the election news and in the news magazines, and to a lesser extent in the cable TV. Figure 11. Buy it: metaphorical vs. literal occurrence. N = 91 for non-metaphorical expressions, standardized frequency x/1 000 000 words 69 Table 17. Absolute numbers of occurrences for literal “buy it.” buy it N = 91 election news news magazines cable TV news radio news all conservative 1 9 29 2 41 liberal 0 12 6 32 50 all 1 21 35 34 91 7.2 Buy into and buy + [OBJECT] -type Literal “buy into” expressions are not frequent in any subgenre, and there are no reverse patterns (Figure 12). The literal “buy into” is naturally typical only for business news. Figure 12. Buy into: metaphorical vs. literal occurrence. N = 19 for non-metaphorical expressions, standardized frequency x/1 000 000 words Table 18. Absolute numbers of occurrences for literal “buy into”. buy into N = 19 election news news magazines cable TV news radio news all conservative 1 2 6 1 10 liberal 4 4 0 1 9 all 5 6 6 2 19 70 To some extent, there is a reverse pattern in buy + [OBJECT] occurrences: when metaphors increase, literal expressions decrease and vice versa, especially in the liberal media (Figure 13). Figure 13. Buy + [OBJECT]: metaphorical vs. literal occurrence. N = 2 518 for non-metaphorical expressions, standardized frequency x/1 000 000 words Table 19. Absolute numbers of occurrences for literal “buy + [OBJECT]”. buy + [OBJECT] N = 2 518 election news news magazines cable TV news radio news all conservative 96 429 484 269 1 278 liberal 121 465 84 570 1 240 all 217 894 568 839 2 518 71 7.3 Sell + [OBJECT] -type It is significant how little non-metaphorical sell + [OBJECT] -type expressions there are in the election news: there are slightly more literal expressions than metaphorical in the liberal media, but there are no great differences (Figure 14). This further strengthens the case of sell being a genre-specific metaphor. In other subgenres, there are considerably more literal sell + [OBJECT] expressions. Figure 14. Sell + [OBJECT]: metaphorical vs. literal occurrence. N = 1 766 for non-metaphorical expressions, standardized frequency x/1 000 000 words Table 20. Absolute numbers of occurrences for literal “sell + [OBJECT]”. sell + [OBJECT] N = 1 766 election news news magazines cable TV news radio news all conservative 44 254 326 140 764 liberal 55 453 60 434 1 002 all 99 707 386 574 1 766 There is also a slight reverse pattern: when sell + [OBJECT] -type metaphors grow, nonmetaphorical expressions decline, especially in the liberal media. The greatest gap between literal and metaphorical occurrence levels is in the newsmagazines. This may be 72 due to their regular coverage on business news. Yet the occurrence of literal SELL expressions may be circumstantial, the topics or issues that require the use of this phrase are unknown. 7.4 Tough sell and does not sell -type Literal tough sell expressions are scarce in these subgenres (Figure 15). Literal tough sell expressions probably are typical only for business or marketing news. Figure 15. Tough sell: metaphorical vs. literal occurrence. N = 8 for non-metaphorical expressions, standardized frequency x/1 000 000 words Table 21. Absolute numbers of occurrences for literal “tough sell”. tough sell N = 8 election news news magazines cable TV news radio news all conservative 0 0 2 0 2 liberal 0 2 0 4 6 all 0 2 2 4 8 “Does not sell” -expressions occur as literal only in the conservative cable TV and in the radio news (Figure 16). 73 Figure 16. Does not sell: metaphorical vs. literal occurrence. N = 33 for non-metaphorical expressions, standardized frequency x/1 000 000 words Table 22. Absolute numbers of occurrences for literal “does not sell”. does not sell N = 33 election news news magazines cable TV news radio news all conservative 1 2 12 6 21 liberal 1 0 1 10 12 all 2 2 13 16 33 7.5 Sell out -type There are extremely little non-metaphorical “sell out” expressions in these data. In the radio news, literal “sell out” expressions occur at higher level in both partisan media than the respective metaphorical expressions (Figure 17). 74 Figure 17. Sell out: metaphorical vs. literal occurrence. N = 51 for non-metaphorical expressions, standardized frequency x/1 000 000 words Table 23. Absolute numbers of occurrences for literal “sell out”. sell out N = 51 election news news magazines cable TV news radio news all conservative 2 5 10 10 27 liberal 0 2 3 19 24 all 2 7 13 29 51 7.6 A concluding note: the proportion of metaphorical occurrences To sum it up, the proportion of metaphorical occurrences out of the all occurrences is now examined with regard to expression types. The percentages of the metaphorical expressions in each type are presented in Figures 18 and 19. The basis for this calculation is the absolute number of sell expressions, not the standardized frequencies. As can be expected, buy + [OBJECT] and sell + [OBJECT] have the lowest proportions, since this kind of phrase is frequent in literal use. The more fixed multiword types score the highest percentages. All the sell metaphor types score high numbers in the election data. Only one multiword type, buy into, has a slightly higher number in the cable TV and in the radio data than in the election data. 75 Figure 18. The proportion of metaphorical occurrences: buy (three types) Figure 19. The proportion of metaphorical occurrences: sell (four types) 76 Table 24. The absolute numbers of occurrence: metaphorical buy (three types). buy metaphors, N = 358 buy + [OBJECT] buy into buy it all election news news magazines cable TV news radio news all 27 14 72 34 147 18 11 35 18 82 11 10 75 33 129 56 35 182 85 358 Table 25. The absolute numbers of occurrence: literal BUY (three types). BUY literal, N = 2 631 buy + [OBJECT] buy into buy it all 217 894 568 839 2 518 5 6 6 2 19 1 21 35 34 91 222 924 609 876 2 631 election news news magazines cable TV news radio news all Table 26. The absolute numbers of occurrence: metaphorical sell (four types). sell metaphors, N = 300 sell + [OBJECT] tough sell does not sell sell out election news news magazines cable TV news radio news all 70 29 64 35 198 15 9 8 4 36 8 1 10 2 21 17 10 10 8 45 all 110 49 92 49 300 Table 27. The absolute numbers of occurrence: literal SELL (four types). SELL literal, N = 1 858 sell + [OBJECT tough sell does not sell sell out all 99 707 386 574 1 766 0 2 2 4 8 2 2 13 16 33 2 7 13 29 51 103 718 414 623 1 858 election news news magazines cable TV news radio news all 77 8 Buy and sell metaphors in the election news In this section, only the election news data will be investigated. The occurrence levels of buy and sell metaphors in the election news corpus will be addressed from two different standpoints: the diachronic distribution of buy and sell metaphors in the 2000, 2004 and 2008 elections and the evaluative use of these metaphors (the numbers of negative and non-negative metaphors). 8.1 Diachronic developments in the buy and sell distribution There are 210 buy and sell metaphors in the election news corpus. There are more sell metaphors than buy metaphors in the election news, as was seen in Figures 1 and 2 in Section 6: there are 71 buy metaphors and 139 sell metaphors. There are some differences in the occurrence levels in the timeline (Figure 20). Both buy and sell occurrences have increased from 2000 to 2008, but the increase has not been continuous, there is fluctuation. It is difficult to say in a timeline of only eight years if this increase is permanent or if fluctuation will continue. Still, the increase is 17% in the conservative corpus and 24% in the liberal corpus, they are relatively high numbers in such a short time. Figure 20. Buy and sell occurrence in the election news in 2000, 2004 and 2008. standardized frequency x/1 000 000 words 78 Table 28. Absolute numbers for buy and sell occurrence in timeline. N = 210 conservative liberal all 2000 2004 2008 29 26 55 32 36 68 40 47 87 all 101 109 210 When the trends of buy and sell occurrences are analyzed separately, there is considerable variance (Figure 21). The greatest increase of occurrences concerns buy metaphors in the conservative media: the occurrence level has grown from 27 (sf) to 69 (sf), which is 63% over three elections. In the liberal media, there is a slightly different pattern, it concerns sell metaphors: the occurrence level of sell has increased from 67 (sf) in 2000 to 104 (sf) in 2008. That is a 36% increase, and 104 is actually the highest partisan occurrence level in 2008 (and second only to conservative sell in 2004, sf 107). Figure 21. Buy and sell occurrence and partisanship in 2000, 2004 and 2008. standardized frequency x/1 000 000 words 79 Table 29. Absolute numbers for buy and sell occurrence and partisanship. N = 210 buy sell all conservative 35 66 101 liberal 36 73 109 all 71 139 210 Based on the evidence in Figures 20 and 21, it can be concluded that even if there is no difference between conservative and liberal media with regard to buy and sell metaphors in total, they yet have a different partisan occurrence profile in the timeline. The increase of buy in the conservative media and sell in both partisan media may result from the increase of negative buy or sell expressions, which will be investigated in Sections 8.2 and 9. 8.2 Negative buy and sell metaphors in the election news In this section, I will show how the conservative and the liberal media use buy and sell metaphors as evaluation in general. The focus is on the quantitative analysis of metaphors with regard to timeline, partisanship and the negativity/neutrality aspect. The occurrences of buy and sell are investigated both together (both buy and sell, the general negativity level) and separately (how they differ with respect to negativity). There are 132 negative buy or sell metaphors in the election corpus (63% of all, 210, thus, there are 78 non-negative expressions). In the conservative media, there are 26 negative buy metaphors and 40 negative sell metaphors, in the liberal media there are 22 negative buy metaphors and 43 negative sell metaphors. The absolute numbers are as follows: 80 Table 30. Negatively evaluative buy and sell metaphors in the corpus: the absolute numbers. negative buy metaphors N = 46 conservative buy liberal buy all 2000 N 4 6 10 2004 N 6 9 15 2008 N 15 7 22 all N 25 22 47 negative sell metaphors N = 83 conservative sell liberal sell all 2000 N 9 12 21 2004 N 18 13 31 2008 N 13 20 33 all N 40 45 85 In the election corpus, there are slightly more negative expressions in the conservative news and non-negative in the liberal news (buy and sell are counted together, Figure 22): Figure 22. Negative and non-negative buy and sell expressions and partisanship. standardized frequency x/1 000 000 words Table 31. The absolute numbers for negative and non-negative metaphors. N = 210 conservative liberal all negative 65 67 132 non-negative 36 42 78 all 101 109 210 81 In general, negative metaphors have increased, whereas non-negative metaphors have stayed at the nearly same occurrence level (in this calculation, buy and sell are counted together) (Figure 23). The growth rate of negative buy and sell metaphors is 36% from 2000 to 2004, and then there is a 3% decline in 2008. Figure 23. Negative and non-negative metaphors, timeline of 2000, 2004, 2008. negative N = 132, non-negative N = 78, standardized frequency x/1 000 000 words This indicates that the negative value inherent in these metaphors may be one reason for their growth. Increased negativity in news in general may have influenced the increase of these metaphors. There is variance in the partisan distribution of these expressions, when they are analyzed with respect to the timeline of the 2000, 2004 and 2008 elections (Figure 24). The liberal negative buys and sell metaphors stay at a fairly even level from 2000 to 2008, on average at 82 (sf). In the conservative media, negative metaphors grow 60% from 2000 to 2004 to the level of 116 (sf). Then the occurrence level drops 20%, to 93 (sf). The average increase from 2000 to 2008 is hence 40%. 82 Figure 24. Negative, partisan metaphors (both buy and sell). conservative N = 65, liberal N = 67, standardized frequency x/1 000 000 words There are two phenomena here worth noting. First, it seems that negative metaphors have increased in the conservative media, but not in the liberal media. It is possible that the partisan polarization and the increase of negativity depicted in Section 3.2 concerns more conservative media than liberal media. The national trauma of the 9/11 attack and the wars in Iraq and Afghanistan can also have a role in this, since conservative and liberal ideology adhere to different beliefs and agendas as regards to these events and the presidency. On the other hand, negative buy or sell metaphors were in 2000 more characteristic of liberals, although not at the level of the conservative media in 2004 and 2008. Negative evaluation—at least with buy and sell metaphors—seems to have moved from the liberal media to the conservative media from 2000 onwards. Second, the 2004 and 2008 elections were framed by George W. Bush’s agenda as President. Republicans were the party of the incumbent president in 2004 and 2008, and from 2003 to 2007 they also had a majority in Congress. In 2000, when liberal media used more negative metaphors, the Democratic Party had an incumbent President, and the incumbent Vice President Al Gore was the Democratic candidate. This hold on power may be reflected in metaphor use, although it is difficult to say why the party in power (or its media) had a need to use negative metaphors. Why opposition rhetoric should be less negative than the governing party’s remains unclear. It is more plausible that the partisan polarization accounts for the differences. As this finding concerns only buy and sell metaphors, it cannot be generalized. In Section 9 and 10 the targets of negative criticism are addressed in detail. 83 In the conservative media, sell metaphors have been in negative use more than buy metaphors in 2000 and 2004 (Figure 25). In the election of 2004, sell metaphors reached the level of 80 (sf), which is a 66% rise from the level of 2000. Then in 2008, there is a 34% decline, but yet a higher number than in 2000. Negative buy metaphors have been growing sharply, the increase is 78% from 2000 to 2008: so, they seem to be the reason for increased negativity in the conservative media. Sell metaphors have grown on average 50% from 2000 to 2008. Figure 25. The timeline of negative metaphors in the conservative media. conservative buy N = 25, conservative sell N = 40, standardized frequency x/1 000 000 words In the liberal media there is neither growth nor permanent decline of negative buy, but there are less negative buy metaphors than sell metaphors (Figure 26). Negative sell metaphors have grown a little (6%). 84 Figure 26. The timeline of negative metaphors in the liberal media. liberal buy N = 22, liberal sell N = 45, standardized frequency x/1 000 000 words 8.3 Non-negative buy and sell metaphors in the election news It is now fairly well established in this study that the increase of negative metaphors is related to the negative use of buy in the conservative media, and that negative sell metaphors have grown in the conservative media as well. Now the non-negative buy or sell metaphors are further investigated. The absolute numbers in the election corpus are as follows: Table 32. Non-negative buy and sell metaphors non-negative buy metaphors N = 24 conservative buy liberal buy all 2000 4 4 8 2004 2 3 5 2008 4 7 11 all 10 14 24 non-negative sell metaphors N = 54 conservative sell liberal sell all 2000 12 4 16 2004 6 11 17 2008 8 13 21 all 26 28 54 Non-negative buy and sell metaphors grew 49% in the liberal media from 2000 to 2008 (Figure 27). In the conservative media, non-negative buy and sell metaphors have decreased, but the decline has not been continuous, on average 40%. 85 Figure 27. Non-negative buy and sell metaphors 2000, 2004 and 2008. conservative N = 36, liberal N = 32, standardized frequency x/1 000 000 words So, non-negative metaphors have increased in the liberal media and decreased in the conservative media. It is, however, important to bear in mind that this does not mean that there is less negativity in the liberal media: there may be other devices to express negative evaluation in the liberal media. 86 9 The targets of negative evaluation with buy or sell metaphors in the election news In this section, I will show how conservative and liberal media evaluate negatively the nine presidential candidates with buy and sell metaphors, and how metaphors are used to criticize either the opponent party or the own party (the Republican Party for conservatives, the Democratic Party for liberals). The timeline of the occurrences will be analyzed as well. I will also suggest some reasons for the differences found. 9.1 The opponent party or the own party as a target When it comes to targets of evaluation, there is an interesting difference (Figure 28): liberals employ negative buy and sell metaphors considerably more of their own candidates or their own party than conservatives of their own. There are 43% more of self-critical expressions in the liberal media than in the conservative media and 34% more metaphors critical of the opponent party in the conservative media than in the liberal media. Figure 28. The target of negative evaluation: the opponent party or the own party. standardized frequency x/1 000 000 words 87 Table 33. The absolute numbers for the target of evaluation. buy + sell, N = 132 opponent party (criticism) own party (self-criticism) other all conservative 40 19 6 65 liberal 29 33 5 67 all 69 52 11 132 This observation lends support to the common stereotype of Democrats in American politics: they are often seen as having more internal party wars, and these intramural differences are also made public more easily than similar differences inside the Republican party (Dagnes 2010, 9-10). There may also be simpler explanations to this difference: Democratic candidates Al Gore and John Kerry in the 2000 and 2004 elections have often been called uninspiring. This may be hindsight, because both Gore and Kerry lost the election, and Gore an exceptionally close election which was solved only by the historical Supreme Court decision which ended the vote recount in Florida. Bill Clinton has been ascribed to say that the difference between Democratic and Republican behavior in elections is that “Democrats fall in love, Republicans fall in line” (Pelosi 2010). Thus he underlines the need of the Democrats to get inspired. The rise of non-negative metaphors in the 2008 election data also suggests that there may be some truth in the Democratic need of inspiration, as Barack Obama was hailed as a charismatic candidate in 2008. In contrast, it is possible that conservative media truly support the nominated candidate of the Republican party, or at least restrain from colorful public criticism over elections. In the 2008 election, for example, Republicans had a wide range of presidential candidates in the primaries21, and a controversial vice-presidential candidate, Sarah Palin. Yet none of the tension this may have caused is seen in the number of negative metaphorical expressions addressed in this study, except perhaps in the small rise of negative buy metaphors used of the 2008 candidates, McCain and Giuliani (see Section 9.3. Figure 31). Ronald Reagan famously said that he follows the 11th commandment: 21 The colorful populist Mike Huckabee, the feisty ex-New York mayor Rudy Giuliani, the libertarian Ron Paul, a senator turned into TV-star (Law and Order), Fred Thompson, Mormon Mitt Romney, for instance. 88 “Thou shall not speak ill of a fellow Republican” (Fastenberg 2011). Perhaps this adage still counts. This difference becomes even more pronounced when the timeline of conservative negative assessment of the opponent party is analyzed: it seems that after the 2004 election the frequency fluctuates little. Buy and sell expressions that assess negatively the opponent party grew sharply (78%) in the conservative media from 2000 to 2004, and the level has stayed high in 2008. In the liberal media, the level of negative buy and sell metaphors which target the opponent party has decreased a little from 2000 to 2008 (Figure 29). Figure 29. Criticism: the opponent party as target. conservative N = 40, liberal N = 29, standardized frequency x/1 000 000 words The highest number of negative buy and sell metaphors which target the own party (46 sf) is in the liberal corpus in 2004. Conservatives are significantly less self-critical, at least with buy or sell metaphors (Figure 30). 89 Figure 30. Self-criticism, the own party as a target. conservative N = 19, liberal N = 33, standardized frequency x/1 000 000 words 9.2 The main targets of negative evaluation22: the nominated Democratic candidates With respect to presidential candidates, John Kerry, Barack Obama and Al Gore lead as targets of negative buy and sell metaphors (Figure 31), they are all the nominated candidates of the Democratic Party. George W. Bush and Hillary Clinton come even at the fourth place. 22 Two Republican candidates participated in two elections: George W. Bush in 2000 and 2004, and John McCain in the primaries of 2000 and as the nominated candidate in 2008. The frequency is based on the word count of 2000 and 2004 subcorpora for Bush and 2000 and 2008 corpora for McCain. This slightly skews the numbers for McCain, because he was not anymore written about in the general election period in 2000. If only the primary phase of the 2000 subcorpus were taken into account, his numbers would be higher. 90 Figure 31. Candidates as targets of negative buy and sell metaphors. N = 85, standardized frequency x/1 000 000 words 9.3 The targets of negative buy metaphors: Kerry, Clinton and Obama Conservatives employ negative buy metaphors of Hillary Clinton and Barack Obama, liberals employ them of John Kerry (Figure 32). Figure 32. Candidates as targets of negative buy metaphors. conservative N = 25, liberal N = 22, standardized frequency x/1 000 000 words John Kerry is the target of most liberal buy metaphors, and the liberal media seem to use them of the candidates of the opponent party very little. It is important to recall here that 91 the absolute number of buy metaphors is low, 47, and that in the conservative media—as illustrated in Section 8.2—negative buy metaphors were scarce in 2000 (N = 4) and 2004 (N = 6). When conservatives target Clinton and Obama with buy metaphors and very little others, it is because conservative media have just started to use more buy metaphors in 2008 (N = 15). In 2008 even their own candidates, Giuliani and McCain are targets of buy metaphors in the conservative media. The absolute number for negative buy occurrences in the case of Kerry is six in total, four in the liberal media and two in the conservative media. The negativity about Kerry may partly be explained by the following example (57) from the liberal media, in which buy metaphors are applied negatively twice. It is also interesting that this comment on the weaknesses of the Kerry campaign is in the middle of a positive assessment of Kerry, his capability of responding to criticism constructively. This shows how the Democratic selfcriticism works in discourse. (57) AFTER NINE TIMES AROUND THE TRACK, I'M CONVINCED that the presidency is something that requires more fate than ambition. Sometime Jupiter aligns with Mars, sometimes it doesn't. And when Kerry started campaigning in earnest in early 2003, he-not for the first time in his career-came out of the blocks miserably. Since 1982, every one of his fights has required a second wind. What I think is most relevant to a possible Kerry presidency is that he has, up until now, always listened to criticism when he has been screwing up, and he has responded forcefully. The initial year of his presidential campaign was almost fatal because of two rookie mistakes influenced by hubris: Kerry bought into front-runner-ism via fund-raising yardsticks, and, worse, he bought into a presentation based mostly on himself, his war record, and his résumé. What was missing from the calculus was a Democratic electorate in Iowa and New Hampshire (and nationally) that was more interested in how national policy might improve its members lives, not just in Iraq or even in the much-celebrated "anger." (Oliphant 2004) Here is an example of a buy metaphor critical of Hillary Clinton in the conservative media: (58) There is a sense that Sen. Obama is rising, a sense for the first time in this election cycle that Mrs. Clinton just may be in a fight, a real one, one she could actually lose. It's all kind of wonderful, isn't it? Someone indulged in special pleading and America didn't buy it. It's as if the country this week made it official: We now formally declare that the woman who uses the fact of her sex to manipulate circumstances is a jerk. (Noonan 2008) 9.4 Targets of negative sell metaphors: Gore and Kerry With regard to sell metaphors, John Kerry is the candidate of whom most conservative negative sell metaphors are used, but Al Gore is negatively framed by conservatives and liberals alike (Figure 33). Obama is addressed negatively with sell metaphors in the 92 conservative media at a slightly higher level than Gore. Hence it is usually the nominated candidates of the Democrats who are criticized with these metaphors. Figure 33. Candidates as targets of negative sell metaphors. conservative N = 40, liberal N = 45, standardized frequency x/1 000 000 words It is noteworthy that in an election where the incumbent vice-president of their preferred party is a candidate, liberals turn their back on him (at least with sell metaphors). The Clinton era legacy with its scandals may have been a mixed blessing for Democrats, and it can have a role in the dissatisfaction with Gore. The absolute number of occurrences is eight in the case of Kerry and eight in the case of Gore as well. Here is an example of a sell metaphor critical of Gore from the conservative media: (59) New Democrats would disagree with this assessment. They would probably note that Gore also came out in his speech for targeted tax cuts and demanded further reductions in crime; that he boasted of his support for the Gulf War; that he talked about reducing the national debt rather than spending the surplus; that he continues to support the death penalty and missile defense. All true. But these positions highlight the perversity of Gore's political strategy: He is selling mostly Clintonian policies with the rhetoric of an unreconstructed liberal. (Ponnuru 2000) Yet another example, a sell metaphor critical of Gore from the liberal media: (60) On environmental issues, for all of Gore’s well-documented failings, the two candidates speak and act as if they come from different planets. Again, Gore is both an environmentalist and a political pragmatist. Judged by the demanding standards that Gore himself laid out in his book Earth in the Balance, he is a sham and a sellout. To take just one example, the Clinton/Gore Administration opened up Alaska’s precious National Petroleum Reserve, selling the first oil-drilling leases in May 1999. Compared with George Bush, however, Gore is Mother Nature herself. If elected, he will 93 arguably be the most environmentally sensitive and sophisticated politician ever to occupy the Oval Office. (Alterman 2000) Yet again this example (60) shows that liberal self-criticism can be complex: the negative evaluation of the candidate is in the middle of a long comparison which in fact judges Gore to be a superbly better candidate than George W. Bush. There is a corresponding occurrence to this, example (38) in Section 4.4, in which President Bush is generally well appreciated in the article (in the conservative media), but a single reference to the Abu Ghraib prison scandal makes the expression harder to sell negative, and hence the expression is classified as negative. George W. Bush is the second in frequency as a target of negative sell metaphors by liberals, but the third target is again a Democrat, Hillary Clinton. George W. Bush’s relatively low numbers with both buy and sell metaphors show the limitation of this study: he must have been framed negatively by Democrats in both elections in which he was a candidate (in 2000 and in 2004), but it must have taken place with other devices than these metaphors. The so-called structural bias, the incumbent President as a candidate remaining more shielded from campaign criticism (Zeldes, Fico, Carpenter & Diddi, 2008, 397) may have contributed to the fact that Bush received less scrutiny in the liberal media. It is also good to bear in mind that the United States was at war during the election 2004 (Afghanistan 2001, Iraq 2003) (Benenson & Moore 2008, 420-421). Another structural factor in American politics may explain the scarceness of negative assessment: at a time of external threat, criticism against the leader of the country will be more easily regulated or self-censored in media. Moreover, George W. Bush’s troubles began in earnest only in 2005, when the intelligence information about the WMD threat (weapons of mass destruction) in Iraq was revealed as dubious and the hurricane Katrina and its aftermath made Bush highly unpopular (Benenson & Moore 2008, 421). 94 10 Discussion On the basis of data in this study, it seems that sell metaphors are specific for election news subgenre. Business as a Vehicle in political discourse, even in election discourse, is not the first and foremost with respect to frequency. War, sports, family, journey and building Vehicles usually dominate political discourse (Deignan 2005, 27; Kövecses 2002, 62; Kövecses 2005, 174-176; Scheithauer 2007, 80, 84). Still, it is significant that figurative expressions based on sell are somewhat specialized terms in the American political news genre: there are from 55% to 92% more sell metaphors in the election data than in the other subgenres (on average 72%). Moreover, the metaphorization level of sell is high in the election news: on average 47% of all SELL expressions are metaphorical in this data. Therefore, elections as events seem to enhance use of sell metaphors. However, it is unknown if sell metaphors occur at any notable level in other genres than political news discourse: further studies would be needed to confirm the genre-specificity of metaphorical sell. If some metaphors become genre-specific, they may, in the first place, affect the way issues are interpreted in this genre, and secondly, they may seem unusual in other genres. As was discussed in Section 2, politics in a democracy is a state of ongoing negotiation. Since buying or selling is a kind of a negotiation between a buyer and a seller which aims to mutual benefit, satisfaction or compromise, sell and buy share a semantic field with to negotiate, to make a deal, to bargain, to offer a choice, to accept or to suggest—or even to take a risk, to gamble or to betray. Some buy or sell metaphors tend to occur in multiword units, and the occurrence level of these multi-word types varies with regard to subgenre and partisanship. The most frequent metaphor type in this data, though, is sell + [OBJECT] –type, which strictly speaking is not a fixed phrase since the object varies. Yet the object of selling is often the element that makes an expression a metaphor, and often it contributes to negative evaluation. In the election discourse, it is natural to feature campaigns or candidates as selling themselves, selling ideas, tax cuts or a health care reform. 95 In a similar vein, the president’s political operations are often called selling in the cable TV, news magazines or radio news. In the entire COCA, president or his administration is the subject of selling something in a metaphorical sense 156 times in the period of 1990-2010 23 . Hence it can be assumed that the political actions or operations of a President are regularly portrayed with sell metaphors. Election period usage may have moved to the news coverage on the presidents. Dagnes (2010, 11) argues that campaigning for office and campaigning for public policy can no more be seen as separate processes. To some degree, the data especially in the comparison corpora indicate that sell + [OBJECT] -type metaphors were applied frequently with respect to the Iraq war. The war was framed in the liberal media as sold to the American public on false premises by President Bush and his administration (14 occurrences), whereas in the conservative media, this sell the war phrase was rarely used, except in one discussion on Fox News about the right moment or way for President Bush to sell the war. Thus, it could be argued that ideology may affect the frequency of sell metaphors, such as issues which partisan media wish to frame as selling. There might be other similar differences in the objects of sell, but this study will not address this question comprehensively. Sell out phrase in the sense of political betraying or not keeping promises is American usage which emerged in the 1890s according to The Oxford English Dictionary. “Sell out” as a literal expression is still in use in the sense of closing a business permanently. In its literal sense it may be neutral or negative. The data in this study show that the metaphorical phrase was used especially in the liberal election news: liberals usually wish to insinuate a betrayal of higher ideals, and they use this expression equally of their own politicians and of conservative politicians (46% of sell out expressions, six occurrences, were used as negative evaluation of the Democratic politicians in the liberal election news, whereas no negative sell out expressions were applied of Republican politicians in the conservative election news). 23 The search terms were president, administration, Clinton, Bush or Obama in 1 to 6 word span before and after the lemma of sell. 96 As such, sell is not a significantly partisan metaphorical expression in itself. Some types of sell metaphors, however, seem to constitute the genre-specificity of sell metaphors more than others: sell + [OBJECT] -type in the conservative election news and sell out type in the liberal election news. So, there is this minor difference in the partisan use of sell: conservatives and liberals prefer different types of sell expressions in the election data. Partisanship correlates sometimes with usage patterns in a very subtle manner. Liberals seem to prefer tough sell -type and conservatives does not sell -type, even though there is only a marginal difference in their metaphorical senses (in the literal, commercial sense, tough sell means an energetic effort to get someone to buy something, does not sell means not finding purchasers). This suggests that there are phrases which seem to belong to either conservative or liberal identity. With respect to buy metaphors, they have a different occurrence profile than sell metaphors. First, they seem to be favored in the election news and in the cable TV news. In the news magazines and in the liberal radio news buy metaphors are scarce. Second, the metaphor type preferred in the conservative and liberal cable TV news and to some degree in the conservative radio news is the buy it -type. This indicates that there may be a register difference in usage: the buy it -phrase may be typical for spoken news. There are some partisan features in the use of buy metaphor types: while the buy it-type is slightly favored in the conservative data, there is a clear, consistent liberal edge in the occurrence of buy into expressions. Moreover, the buy + [OBJECT] -type shows a slight conservative edge in the election news and in the radio news. These types also share a semantic field; they all have the sense to believe. Hence it seems that stylistic choices may explain this phenomenon: I would suggest that the buy it and buy + [OBJECT] types are more direct and brusque than the somewhat elaborated buy into-type. This difference may reflect the stereotypes of practical and direct conservatism as against idealistic and complex liberalism. In addition, buy into -expressions resonate sometimes with other senses than to believe: a parallel sense to apply, to put into use can be identified in some cases. Here is an example (61) of this: (61) BROWN: Fox News caught up with Romney in Des Moines. He was forced to window shop today because of the weather. It may be time for Romney to buy into a much more 97 aggressive Iowa strategy. The latest Rasmussen poll shows Huckabee well out in front of the previously long-time Iowa leader Romney. But when asked repeatedly about going negative on Huckabee, Romney ignored the question. (Hume 2007) From the empirical data in this study, it can be seen that there is a metaphoricity cline in the use of buy and sell metaphors, they do not always seem to fall neatly into metaphorical or non-metaphorical categories, as was demonstrated in Section 4.3. Some extended metaphors in authentic discourse resonate with literal or marginally metaphorical senses, which is exemplified in examples (24), (31), (32) and (33) in Section 4.3.4. This finding supports the results by Deignan (2007): she has addressed semi-fixed metaphorical expressions and points to the metaphoricity cline found in them (104). Yet most of the buy or sell metaphors are very conventional and easy to identify, these conventional expressions account for 93% (N = 701) of all buy and sell metaphors. This is in accordance with Steen, Dorst, Herrman, Kaal & Krennmayr (2010), a study that showed that 99% of metaphorical expressions are conventional in some genres (785, 789). However, it is worth noting that despite the high level of conventionalized buy and sell metaphors, both buy and sell can be taken into novel use, as illustrated by the two expressions buying the Iraqis time or sell the dog food (repeated here from Section 4.3.4 examples 24 and 27): (62) Indeed, the entire point of the surge is to bring such reconciliation about by, in Gates’ words, “buying the Iraqis time. "[sic] But that's the problem. The United States is ever more dearly buying time, and Iraq is ever more freely spending it. (“The Bitter End”2007) (63) There, flipping pork, was Lamar Alexander -- an eminently decent and qualified man who has visited 64 of Iowa's 99 counties, but just can't sell the dog food. (Tapper 1999) Some probably new uses of metonymic sell can be seen in the following examples: (64) ROBERTS: You know, I asked the Democratic pollster that over the weekend, and his answer was absolutely not. The Democrats are nowhere. No one, according to the polls, knows who they are or what they stand for, and perhaps if the public is sufficiently angry with the [Republica]ns, if oil prices go up or they feel strongly enough that the country is off on the wrong track, it could boost the Democrats in the next election regardless of their own plans. But right now it's very hard to see. As this pollster put it to me, No equivalent of a Contract with America that the Republicans had to take the House in the midyear election in 1994,' the Democrats hope to do something with the Republican ethics problems, that -- Bill Frist selling his HCA stock has raised some questions. And, of course, ongoing investigations of Tom DeLay -- they're trying to say majority leaders in both Houses are ethically tainted. They're going to have a little trouble making that a big sell to the American public. MONTAGNE: Well, Cokie, does the fact that the Democrats seem to be, as you describe it, drifting mean that the president doesn't have to worry about his next appointment to the Supreme Court? (Montagne 2005) 98 (65) During an interview at the Sioux Falls headquarters of Vote Yes for Life, the organization she heads that's campaigning to uphold the state ban, Unruh spoke about her own abortion years ago and her campaign to focus not on confrontation but on the personal stories of women like her who have had or considered abortions. But activists on both sides are riveted by the bottom-line question voters here will answer next week: Can Unruh's new sell persuade South Dakotans to endorse a state law that ignores a U.S. Supreme Court precedent and makes no specific abortion exceptions in cases of incest or rape, or when a mother's health is threatened? (Halloran 2006) Even though metaphors in spoken discourse were not specifically addressed in this paper, there is a significant finding in this respect: seven out of twelve discourse metaphors in the section 4.3.4 are direct quotes (examples 23, 24, 26, 28, 31, 32 and 33). This shows that it might be useful to examine novel metaphors and complex discourse metaphors especially in spoken discourse. When the occurrence level of some multiword metaphor types is high, the literal occurrences of the same multiword phrase sometimes seem to decrease, and vice versa. This phenomenon concerns sell + [OBJECT] -type, buy + [OBJECT] -type and to some degree buy it -type, but less other multiword types. So a reverse occurrence pattern is occasionally found. On the basis of this study, Deignan’s (2005, 2007) results regarding multiword unit metaphors as one driving force in metaphoricity are to some degree corroborated, as far as buy and sell metaphorical phrases are concerned. One objective of this study was to find out how the evaluative value of buy and sell metaphors varies in partisan election news. Negative buy and sell metaphors seem to be more frequent than non-negative expressions in the election data, 63% (N = 132) of the occurrences are negative and 37% (N = 78) non-negative. Negative metaphors have grown only in the conservative media, the growth rate has been on average 56% from 2000 to 2008 (with some fluctuation). Negativity in the conservative media has risen, because non-negative buy and sell metaphors have decreased in the conservative corpus, and because of the considerable and steady increase (78%) of negative buy metaphors from 2000 to 2008. Sell metaphors have also grown in the conservative media at the average of 58%. In the liberal media, the level of negative expressions has stayed fairly even from 2000 to 2008, but there is an interesting difference between the conservative and the liberal 99 media with regard to the target of negative evaluation: liberals criticize their own candidates or party much more than conservatives their own. There are 43% more selfcritical expressions in the liberal media than in the conservative media, and 34% more metaphors critical of the opponent party in the conservative media than in the liberal media. Negative buy or sell metaphors are mostly used of Democrats John Kerry, Al Gore, and Barack Obama. George W. Bush comes only as the fourth in this order of negative criticism. Conservative echo-chamber negativity 24 , strengthened with the liberal selfcriticism especially with regard to Kerry and Gore, may explain the high numbers of the Democratic candidates. Yet it seems that some buy or sell metaphors which criticize Kerry or Gore could also be classified as neutral: the context of these metaphors is complex. George W. Bush’s low numbers show the limitations of this study: he must have been negatively framed by liberal media, but not with buy or sell metaphors. It is also possible that Bush in 2004 had an advantage of a structural bias in media, both as an incumbent President and as a wartime President (Gore as an incumbent Vice-President in 2000 did not receive such privilege). The election of 2004 seems to have been a significantly polarized period when negative buy or sell metaphors were used by both parties. The bitter, contested election in 2000 which Democrats finally lost, the 9/11 attack in 2001 and the Afghanistan and Iraq wars may have contributed to this. It was discussed in Section 3 that news production process often seems to lead to simplification or dramatization of issues. Metaphors serve this aim by providing easy access to abstract concepts and by conveying direct or indirect evaluation. Changes in news language towards more informal style have been noted in several diachronic studies (see for instance Hundt & Mair 1999). As buy and sell metaphors are colloquial, their increase may be related to increased informality. New media and the broadcasting legislation changes reported in Section 3.2 may have contributed to the increase of metaphors as well. Yet it can be questioned if metaphors simplify too much or offer a biased version of politics. The hiding aspect of metaphors may not be beneficial for 24 ‘Echo-chamber effect’ means ratcheting up issues dear to conservatives by repeating stories other media have already published, each media outlet legitimizing the other. (see in detail Hall Jamieson & Cappella 2008). 100 understanding power relations and operations in politics. A frequent complaint of some, especially the left wing of the liberals, is that a “dumbing down” process happens too often in media with regard to complex social issues (Lilleker 2006, 69-72; Dagnes 2010, 61-62, 76, 88). A variant term for this phenomenon is “tabloidization” (Baker & Ellege (2011, 147): it refers more broadly to the ‘infotainment’ or ‘confrontainment’ phenomena. In this study, however, no great difference was found in buy and sell metaphor use between conservatives and liberals: both partisan media apply these expressions. It is possible that in the ongoing competition of making an appealing representation of daily politics, media needs to use all the tools they can, and progressive political media have to revert to simplifications, following the ‘plain-spoken’ conservatism. The self-critical assessment with buy or sell metaphors is cherished only by liberals, although this liberal negative evaluation is a rather complex phenomenon in discourse, as was shown in examples (55) and (58) in Section 9. Further studies on how conservative or liberal media reify and reiterate their own ideological identity and frame the opponent ideological identity would be needed, and this study shows that it could be fertile to study evaluative metaphors applied by conservatives and liberals. Is the emergence of metaphorical expressions dependent on some pragmatic factor, and if it is, how can this emergence be detected and explained? In the view of this thesis, a need to express negative evaluation indirectly or a need to question credibility seem to be possible reasons for the increase of buy or sell metaphors. There may have been a need to create an expression which principally means ‘not believing’, and buy metaphors serve that end. Sell seems to have a specialized role: political offers or proposals seem to be expressed in this way for some reason, maybe because the political campaigns have evolved into huge operations with forceful outreach to citizens. Both buy and sell metaphors may have filled a semantic gap. The negativity inherent in these expressions has probably enhanced their use. Channell (2002, 55) raises the question of how negative polarities of evaluative lexis seem to have significantly increased: this observation was made during the compilation of the Collins COBUILD Dictionary of English (1995). This paper supports the finding that negative expressions seem to have increased. Political polarization and cable TV may have attributed to the negativity as well. 101 There could be other driving forces in the evolution of language that can be discussed here only fleetingly because of the limitations of this study. The language of business pervades American society, and as freedom of trade is an essential American belief, it is easy to see how elections could be seen as a series of transactions whereby the best offer finds a buyer after a fair competition. Writing about elections as consumer actions and commercial comparisons may have enhanced the use of business metaphors. Referring to presidential candidates as brands (such as cars or beer) has not been an entirely unknown practice in media. As early as 1972, the columnist Art Buchwald was quoted in the TIME magazine when he was commenting on the ambiguous situation with one candidate in the Democratic primaries, Ed Muskie: (66) To Humorist Art Buchwald last week, Muskie's breakdown seemed more like an automobile recall. Citing "engineering difficulties and lack of consumer acceptance" for withdrawal of "the Muskie" from the market, Buchwald noted the faint hope for final victory as a compromise candidate: "While the 'Muskie' will not be sold in Ohio, Michigan, New York or California this year, it will be on display in the showroom at the Miami Convention Center in case anyone still wants to buy it." (“The press: the hairline fracture”, 1972) It can be assumed that historical reasons have influenced the evolution of buy and sell as metaphors. As Kövesces (2005) has said, our selection of metaphors depends on who we are, how we construct our identity and what our ultimate concerns or interests are (180). As a nation, the United States is without doubt the business leader of the world, and it may well be that this American identity is reflected in metaphor use. The limitations of this study are based on the research design. First, it is obvious that a study of only two lemmas (BUY and SELL) does not allow an analysis of metaphors in political discourse in great detail, or a more sophisticated analysis of negative evaluation. Second, the most difficult problems are the complexity of metaphor identification, the definition of the evaluative value and the target of metaphorical expressions. This was discussed in Sections 3.1, 4.3 and 4.4: there is always some subjectivity in the classification of expressions. The negative self-evaluation by Democrats might prove to be more multi-layered, if the classification procedure had not been based on the closest discourse context of a buy or a sell metaphor. Even though the negative value of a buy or sell metaphor was in many cases apparent, the larger discourse context seemed to be equally important in evaluation This shows that it is difficult to study evaluation on the basis of lexis alone. Third, the period 2000-2008 with only three elections is too short for 102 conclusive results about the diachronic change in the occurrence of buy or sell metaphors. Other limitations concern the research corpora. The most important weakness is the mock radio corpus which really is comprised of the Rush Limbaugh cable TV show. It can be assumed that the show was mostly based on the format of the Rush Limbaugh show on radio and that Limbaugh himself did not change his radio persona to do that TV show. The fact that it is already problematic to have a subcorpus based on a single news commentator’s show makes it even more important to state that the conservative radio data may not be as representative as other subgenres. With regard to the representativeness of other subcorpora, the 2000 election news corpus may not be as representative as other election subcorpora, because of the lack of the electronic archives or because archiving may not have been regular or consistent. There may be other unbalances in the corpora with regard to partisan identity in media. It was not always possible to easily define a news medium as conservative or liberal. There are a few publications in the conservative election news corpus that would possibly be called more centrist than conservative, such as U.S.News and World Report, Forbes or Bloomberg. Secondly, editorial policies may be the reason for the uneven distribution of some metaphors in the corpora. Buy metaphors occur rarely in The Mother Jones and Harper’s Magazine, which comprise 35% and 32% respectively of the liberal news magazine corpus, so most of the buy metaphors found in the liberal news magazine data are from The Washington Monthly (33% of the corpus). There are no buy metaphors in The New Yorker data, which comprises 18% of the liberal election news corpus. These numbers show that there might be differences in metaphor use with respect to publication type. Further studies could be suggested: are colloquial expressions or negatively evaluative metaphors typical for some news magazines, for instance? In spoken discourse, metaphorical expressions may increase because of repetition, characteristic for conversation, or because participants adopt expressions from each other. Moreover, lack of clarity in transcripts makes it sometimes difficult to interpret the meaning. With regard to adopting expressions from other participants in conversation, there is one incident in the Fox News corpus when the right time for the President to ‘sell 103 the Iraq war’ was discussed, and there are five sell metaphors in a period of ca. two minutes’ conversation. So, if they all were excluded as repetition, the claim that only liberals discussed the war in Iraq in terms of selling would be even stronger. 104 11 Conclusion Metaphor in discourse and metaphors as discourse events which join together grammar, context and political objectives were the principal interest in this study. The following research questions guided this study: how genre-specific are buy and sell metaphors, how are they distributed in partisan political news and what kind of metaphorization levels of buy and sell can be found in authentic discourse? Moreover, negative evaluation conveyed with these metaphors in the partisan news coverage of three presidential elections in the United States (2000, 2004, and 2008) was one aspect in this study. The data comprised four text corpora which represent American political news genre: election news (1.6 million words), news or opinion magazines (3.8 million words), cable TV news (4.1 million words) and radio news (4.3 million words). All corpora had a conservative and a liberal subcorpus. On the question of genre-specificity, sell was found to be a genre-specific election news metaphor: there were on average 72% more sell metaphors in the election data than in the other subgenres. Moreover, the occurrence levels of metaphorical and literal sell + [OBJECT] -type expressions were nearly even in the election news. All other sell metaphor types were more frequent than their literal counterparts in the election data. With respect to political subgenres in this study, there were some indications that different metaphor types might be preferred in some subgenres more than others, such as the buy it -type in spoken news. Although neither buy nor sell were partisan as such, some types of theme were more partisan than others. The buy into -type seemed to be characteristic for liberal political discourse. Some metaphor types were favored in partisan subgenres, such as the sell + [OBJECT] -type in the conservative election news and the sell out -type in the liberal election news. Sometimes there were reverse patterns in partisan use: conservatives preferred the does not sell –type metaphors and liberals the tough sell -type, and to some extent this applied also to the conservative use of the buy + [OBJECT] -type and the liberal use of the buy into -type. With regard to negative evaluation, negative metaphors were found in the election news data more than non-negative metaphors: there were 36% more negatively evaluative buy and sell metaphors than non-negative in the election news. The increase of negative buy 105 metaphors (78%) in the conservative election news was especially great from 2000 to 2008. Liberals used negative buy and sell metaphors of their own candidates and party much more than conservatives of theirs. Some of the familiar stereotypes of liberals and conservatives are confirmed by these numbers. Democrats (liberals) are generally seen as having more internal disagreements and their intramural differences are also made public more easily. In contrast, the Republicans (conservatives), even though they are often described as having “a large tent” of a party (social conservatives, business-oriented conservatives and neo-conservatives), seem to have more discipline: they support the nominated candidate of their party in public, or at least restrain from colorful criticism during the elections. There is one caveat to this notion: the liberal negative selfevaluation was in some cases complex and dependent on the large discourse context. Metaphors seemed to be used more in order to create a stylistic effect than to actually criticize. The future research topics suggested by this study concern at least five domains. First, what kind of further evidence can be found for the findings that sell metaphors are genrespecific in election news or that certain buy metaphor types, for instance buy it -type, may be typical for spoken political subgenre? Genre-specificity of metaphors has not been studied exhaustively from the empirical point of view: what kind of genre differences are there in metaphor use in authentic discourse and why? This study shows that even within closely related political subgenres there is variance. The relationship of metaphors both with genre and register should be more fully explored, as Steen et al. (2010), Krennmayr (2011) or Pasma (2011) have already done. Second, the multiword type of metaphor as an explanation for metaphorization level is in accordance with earlier studies (Deignan 2005, Deignan 2007). This raises the question whether phrasal metaphor in authentic discourse leads the evolution of metaphors. The proportion of the literal and metaphorical occurrence levels in general could also be further examined, in order to understand how metaphor occurrence is dependent on Topics and contexts. Third, the evaluative aspect of buy and sell metaphors and the partisan differences found in this study show something of the relationship between metaphor use and ideological 106 identity: constructing ‘self’ and ‘other’ or ‘us’ and ‘them’ identities in political discourse could be investigated with regard to other metaphors than buy or sell. As Critical Discourse Analysis has increasingly been applied together with cognitive or corpus methods (see for instance Hart 2011, Salama 2011), this could prove to be a productive new research area. Fourth, a study on the diachronic development of buy and sell metaphors could reveal pragmatic factors that have induced the growth of, for instance, buy metaphors in the last 40 years. Finally, a note on the quantitative research design applied in this study. There is the pivotal question: does frequency matter, and if it does, how does it matter? The departure point of this study was that ideological identity may be revealed in metaphor use by comparing quantities, and especially the quantities of negatively evaluative metaphorical expressions. This expectation was at least partially confirmed and shows that shifting foci between quantitative and qualitative analysis is useful. The ongoing debate over different levels of metaphor occurrence and identification shows the significance of the topic (see e.g. Kövesces 2008, Steen et al. 2010). This paper suggests that the relationship between frequent metaphorical Vehicles and some less frequent but genrespecific metaphorical expressions should be explored further. To conclude, this study shows that a quantitative study in which many comparison corpora are consulted can reveal characteristics of metaphors previously unnoticed. The research design of this study can fairly easily be applied in other genre studies. However, the way into consistent view on metaphors both as a discursive and conceptual phenomenon seems long, because no central forces or general principles have been found in metaphor use in authentic discourse studies so far. Taking small steps in this field, one by one, will some day lead to better metaphor identification procedures, better discourse analysis and better understanding of cognition and political identity. As a highly unlikely contender for the highest office in the United States, candidate Obama referred to the power of words in 2008: “don’t discount that power” (see the quotation in the beginning of Section 1). Words carry a sense, words carry social identities and words carry potential political clout. As I started with Obama, I will end with Obama, with a possible discourse metaphor from authentic spoken discourse: it 107 shows all the complexities of elliptic, context-situated figurative speech which requires cultural and political knowledge to be understood. In 2009, when the health care reform proposal received poor poll numbers in the United States and the trust to the new President was fading rapidly, Obama was asked in the Oval Office by his advisers, if he was still feeling lucky. He replied: “My name is Barack Hussein Obama and I am sitting here. So yeah, I’m feeling lucky.” (Alter 2010, 399) 108 References Sources The Election Corpus. In the possession of the author. The Contemporary Corpus of American English (COCA). www.americancorpus.org Books and articles Allen, J., Cohen, W.S., Smith, W. & Tobin, M. (2000, May, 22). The interstate 80 election. U.S. News & World Report. Retrieved Jul 20th, 2010, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/detail?vid=8&hid=122 &sid=5f59907f-518b-429a-bf1ee81a63482661%40sessionmgr112&bdata=JnNpdGU9ZWhvc3QtbGl2ZSZ zY29wZT1zaXRl#db=a9h&AN=3094938 Alter, J. (2010). The Promise: President Obama, Year One. New York: Simon & Schuster. Alterman, E. (2000, October, 16). Bush or Gore: does it matter? The power of the presidency. The Nation. Retrieved Jul 20th, 2010, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/pdfviewer/pdfviewer?si d=400281be-068f-471f-98dfb14f813b9d34%40sessionmgr114&vid=7&hid=122 Bagdikian, B. (2004). The New Media Monopoly. Boston: Beacon Press. Baker, P. & Sibonile, E. (2011). Key Terms in Discourse Analysis. London: Continuum International Publishing Group. Bardes, B. A., Shelley, M. C., & Schmidt, S. W. (2004). American Government and Politics Today; the Essentials. Belmont: Thomson/Wadsworth. Barone, M. (2004, May, 17). No, it's not the American way. U.S. News & World Report. Retrieved 20th Jul, 2010,. from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/detail?sid=5acf42fe- 109 9c73-4b26-8e981b130eeac679%40sessionmgr12&vid=2&hid=8&bdata=JnNpdGU9ZWhvc 3QtbGl2ZSZzY29wZT1zaXRl#db=a9h&AN=13043121 Beer, F. & De Landstheer, C. (eds.) (2004). Metaphor, politics, and world politics. In F. Beer & C. De Landtsheer (eds.), Metaphorical World Politics. East Lansing: Michigan State. Pp. 5-52. Begala, P. (2008, August, 26). Please, Democrats, attack. RealClearPolitics.com. Retrieved Sept 2nd, 2008, from http://www.realclearpolitics.com/articles/2008/08/please_democrats_attack. html Benenson, B. & Moore, J.L. (2008). Elections A to Z. Washington, DC: CQ Press. Berber Sardinha, T. (2008). Metaphor probabilities in corpora. In M. S. Zanotto, L. Cameron & M. de C. Cavalcanti (eds.), Confronting Metaphor in Use. Amsterdam: John Benjamins. Pp. 127-47. Berman, A. (2007, September, 17). Democrats sizzle in Iowa. The Nation. Retrieved Sept 19th, 2007, from http://www.thenation.com/doc/20071001/berman Biddle, W. (2003, February). Anxiety, anonymity, amnesia. Harper’s Magazine. Retrieved Aug 2nd, 2011, from the COCA. Borosage, R. L. (2000, August, 21). Vote for Gore. The Nation. Retrieved Jul 20th, 2010, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/pdfviewer/pdfviewer?vi d=7&hid=122&sid=400281be-068f-471f-98dfb14f813b9d34%40sessionmgr114 Burke, K. (1989). On Symbols and Society. Chicago: The University of Chicago Press. Calmes, J. (1999, November, 17). Campaign 2000: Forbes and Bauer fight for mantle of GOP Right, but many conservatives plan to pull Bush Lever. The Wall 110 Street Journal. Retrieved Jul 20th, 2010, from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=46403557&sid=1 &Fmt=3&clientId=23404&RQT=309&VName=PQD Cameron, L. (2003). Metaphor in Educational Discourse. London, New York: Continuum. Cameron, L. (2007). Confrontation or complementarity? Metaphor in language use and cognitive metaphor theory. Annual Review of Cognitive Linguistics, 5, 107 -135. Cavuto, N. (2003). Interview with Ben Stein, Price Headley, Tony Dwyer, Tobin Smith. Retrieved Aug 2nd, 2011 from Fox News, COCA. Chadwick, A. (2005). Analysis: President Bush's proposal for Social Security; rifts in the CIA over new post of intelligence czar; Syrian occupation of Lebanon. Daybreak. Retrieved Aug 2nd, 2011, from National Public Radio, COCA. Chait, J. (1999, December, 27). Don't worry, cut taxes: Bush fakes left, goes right. The New Republic. Retrieved Aug 29th, 2011, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/detail?vid=4&hid=25& sid=908bb5fb-d95f-48ec-8e78540c3176f5d0%40sessionmgr11&bdata=JnNpdGU9ZWhvc3QtbGl2ZSZz Y29wZT1zaXRl#db=a9h&AN=2587364 Channell, J. (2001).Corpus-based analysis of evaluative lexis. In S. Hunston & G. Thompson (eds.), Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: Oxford University Press. Pp. 38-55. Charteris-Black, J. (2004). Corpus Approaches to Critical Metaphor Analysis. Basingstoke: Palgrave Macmillan. Chilton, P. (1996). Security Metaphors: Cold War Discourse from Containment to Common House. New York: Peter Lang. 111 Chilton, P. (2004). Analysing Political Discourse: Theory and Practice. New York: Routledge. Chilton, P. (2006). Metaphors in political discourse. In Encyclopedia of Language & Linguistics. Pp. 63-65. Retrieved Jun 3rd, 2011, from http://www.sciencedirect.com.libproxy.helsinki.fi/science/referenceworks/9 780080448541 Churchill, W. (1947). Speech in the House of Commons Nov 11, 1947. Retrieved May 12th, 2011, from http://www.quotationspage.com/quote/24926.html Cienki, A. (2005). Metaphor in the ‘‘Strict Father’’ and ‘‘Nurturing Parent’’ cognitive models: Theoretical issues raised in an empirical study. Cognitive Linguistics, 16, 279-312. Coe, K., Tewksbury, D., Bond, B. J., Drogos, K. L., Porter, R. W., Yahn, A. & Zhang, Y. (2008). Hostile news: Partisan use and perceptions of cable news programming. Journal of Communication, 58, 201-219. Cohen, J. E. (2008). The Presidency in the Era of 24-Hour News. Princeton: Princeton University Press. Cohen, R. (2001, August). Down and out at Wrigley Field. Harper’s Magazine. Retrieved Aug 2nd, 2011, from COCA. Collins COBUILD English Language Dictionary (1995). s.v. “buy”, s.v. “sell”. London: HarperCollins. Cook, T. (2001). The future of the institutionalized media. In W. Bennett & R. M. Entman (eds.), Mediated Politics: Communication in the Future of Democracy. Cambridge: Cambridge University Press. Pp. 182-200. Corn, D. (2003, December, 22). Can Kerry get real? The Nation. Retrieved Jul 20th, 2010 from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/pdfviewer/pdfviewer?vi 112 d=9&hid=122&sid=400281be-068f-471f-98dfb14f813b9d34%40sessionmgr114 Deignan, A. (2005). Metaphor and Corpus Linguistics. Amsterdam: John Benjamins. Deignan, A. (2007). Linguistic and figurative patterns in natural language data. Ilha do Desterro, 53, 95-107. Deignan, A. & Potter, L. (2004) A corpus study of metaphors and metonyms in English and Italian. Journal of Pragmatics.36, 1231-1252. Dagnes, A. (2010). Politics on Demand: The Effects of 24-Hour News on American Politics. Santa Barbara: Praeger. De Landtsheer, C. (1998). The political rhetoric of a unified Europe. In O. Feldman & C. De Landtsheer (eds), Politically Speaking: a Worldwide Examination of Language Used in the Public Sphere. Westport: Praeger. Pp. 129-145. De Landstheer, C. (2009). Collecting political meaning from the count of metaphor. In A. Musolff & J. Zinken (eds), Metaphor and Discourse. New York: Palgrave MacMillan. Pp. 59-78. De Landtsheer, C.& De Vrij, I. (2004). Talking about Srebrenica: Dutch elites and Dutchbat. How metaphors change during crisis. In F. Beer & C. De Landtsheer (eds), Metaphorical World Politics. East Lansing: Michigan State University Press. Pp. 163-189. Dickerson, J. (2007, July, 30). Obama's Achilles' heel: What's he ever done? Slate.com. Retrieved Aug 2nd, 2011, from http://www.slate.com/id/2171292 Dionne Jr., E. J. (2008, September, 12). What Obama should do to defeat McCain's lowly tactics. The New Republic. Retrieved Aug 2nd, 2011, from http://www.tnr.com/politics/story.html?id=47d43e3b-197b-4674-b82bc5953737f2a9 113 Easterbrook, G. (2006). Comments: Two alternative perspectives. In P. S. Nivola & D. W. Brady (eds), Red and Blue Nation? Vol. 1, Characteristics and Causes of America's Polarized Politics. Stanford: Hoover Institution on War, Revolution, and Peace, Stanford University. Pp. 255-260. Fastenberg, D. (2011, February, 4). Top 25 Political Icons. Ronald Reagan. TIME. Retrieved March 25th, 2011, from http://www.time.com/time/specials/packages/article/0,28804,2046285_204 5996_2046088,00.html Fernholz, T. (2008, September, 11). Everybody calm down! Obama is hitting back. www.prospect.org. Retrieved Aug 2nd, 2011, from http://www.prospect.org/cs/articles?article=everybody_calm_down_obama _is_hitting_back Foer, F. (1999, December, 6). Running on their faith. U.S. News & World Report. Retrieved Aug 2nd, 2010 from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=47625834&sid=9 &Fmt=2&clientId=23404&RQT=309&VName=PQD Fox News Panel discusses news of the week (2008). Journal editorial report. Retrieved Aug 2nd, 2011, from Fox News, COCA. Galston, D. W.& Nivola, P.S. (2006). Delineating the problem In P. S. Nivola & D. W. Brady (eds), Red and Blue Nation? Vol. 1, Characteristics and Causes of America's Polarized Politics. Stanford: Hoover Institution on War, Revolution, and Peace, Stanford University. Pp. 1-47. Geis, M. (1987). The Language of Politics. New York: Springer-Verlag. Goatly, A. (2007). Washing the Brain: Metaphor and Hidden Ideology. Amsterdam: John Benjamins. 114 Goldstein, E. R. (2008). Who framed George Lakoff? Chronicle of Higher Education, 54, B6-B9. Retrieved Oct 20th, 2010, from http://search.ebscohost.com/login.aspx?direct=true&db=ehh&AN=3401195 9&site=ehost-live&scope=site Graham, A. (2006, September). Projects that work. USA Today Magazine. Retrieved Aug, 2nd, 2011 from COCA. Greider, W. (2000, July, 8). Bush's touchy-feely economics. The Nation. Retrieved Aug 2nd , 2010, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/pdfviewer/pdfviewer?vi d=4&hid=12&sid=6e2bb941-6bba-4737-a14da65a8d2678bc%40sessionmgr10 Gulati, G.J., Just, M.R. & Crigler, A. N. (2004). News Coverage of Political Campaigns. In L.L. Kaid (ed.), Handbook of Political Communication Research. Mahwah: Lawrence Erlbaum Associates. Pp. 237-256. Hall Jamieson, K. & Cappella, J.N. (2008). Echo Chamber: Rush Limbaugh and The Conservative Media Establishment. New York: Oxford University Press. Hart, C. (2011). Force-interactive patterns in immigration discourse: A cognitive linguistic approach to CDA. Discourse & Society, 22, 269-286. Heinauer, L. (2000, October, 25). Capitalizing on democracy - with election approaching, marketers ride to bank on candidates' coattails. The Wall Street Journal. Retrieved Jul 20th, 2010 from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=62883647&sid=2 &Fmt=3&clientId=23404&RQT=309&VName=PQD Henninger, D. (2004, March, 26). Wonder Land - skiing as news: Is John Kerry overexposed? The Wall Street Journal. Retrieved Jul 20th, 2010 from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=589682321&sid= 3&Fmt=3&clientId=23404&RQT=309&VName=PQD 115 Heywood, A. (2004). Political Theory: An Introduction. Basingstoke: Palgrave Macmillan. Howe, N. 1988. Metaphor in contemporary American political discourse. Metaphor and Symbolic Activity, 3, 87-104. Hume, B. (2007). Political headlines. Retrieved Aug 2nd, 2011, from Fox News, COCA. Hundt, M. & Mair, C. (1999). 'Agile' and 'uptight' genres: The corpus-based approach to language change in progress. International Journal of Corpus Linguistics, 4, 221-242. Hunt, A.R. (2004, September, 23). Neither guy has made the sale yet. The Wall Street Journal. Retrieved Aug 2nd,2010, from. http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=697760591&sid= 4&Fmt=3&clientId=23404&RQT=309&VName=PQD Inskeep, S. (2006). Hughes: No short-term fix for U.S. image abroad. Morning Edition. Retrieved Aug 2nd, 2011, from National Public Radio, COCA, Ivie, R. L. & Giner, O. (2009). American exceptionalism in a democratic idiom: Transacting the mythos of change in the 2008 presidential campaign. Communication Studies, 60, 359-375. Ivie, R. L. & Ritter, K. (1989). Whither the "Evil empire": Reagan and the presidential candidates debating foreign policy in the 1988 campaign. The American Behavioral Scientist, 32, 436-450. Jäkel, O. (2003). Motion metaphorized in the economic domain. In H. Cuyckens, T. Berg, R. Dirven & K. Panther (eds), Motivation in language Amsterdam: John Benjamins. Pp. 297-318. Jeffrey, C. (2002, November). Go west, old man. Harper’s Magazine. Retrieved 2nd Aug, 2011, from news and opinion magazines section in the COCA. 116 Kaufer, D. & Hariman, R. (2008). Discriminating political styles as genres: A corpus study exploring Hariman's theory of political style. Text & Talk, 28, 475500. Keating, P. (2007, June, 6). Go west. Rudy and Barack are both wrong about redrawing the electoral map. www.tnr.com. Retrieved Aug 2nd, 2011, from http://www.tnr.com/article/go-west Kimmel, M. (2004). Metaphor variation in cultural context: Perspectives from anthropology. European Journal of English Studies, 8, 275-294. Klein, E. (2007, May, 30). A lack of audacity. How Obama's health care plan resembles the candidate himself. www.prospect.com. Retrieved Jun 3rd, 2007, from http://www.prospect.org/cs/articles?article=a_lack_of_audacity. Koller, V. (2004). Metaphor and Gender in Business Media Discourse: A Critical Cognitive Study. New York: Palgrave MacMillan. Kövecses, Z. (2002). Metaphor: A Practical Introduction. New York: Oxford University Press. Kövecses, Z. (2005). Metaphor in Culture: Universality and Variation. Cambridge: Cambridge University Press. Kövecses, Z. (2008). Conceptual metaphor theory: Some criticisms and alternative proposals. Annual Review of Cognitive Linguistics, 6, 168-184. Krennmayr, T. (2011). Metaphor in Newspapers. Dissertation. Retrieved Aug 28th, 2011, from http://dare.ubvu.vu.nl/bitstream/1871/19774/1/dissertation.pdf Kudlow, L. (2007, May, 10), Democrats dooming themselves to defeat. RealClearPolitics.com. Retrieved May 16th, 2007, from http://www.realclearpolitics.com/articles/2007/05/democrats_dooming_the mselves_t.html 117 Lakoff, G. (1996). Moral Politics: What Conservatives Know That Liberals Don’t. Chicago: London: University of Chicago Press. Lakoff, G. & Johnson, M. (1980). Metaphors We Live By. Chicago: University of Chicago. Lauerbach, G., & Aijmer, K. (2007). Argumentation in dialogic media genres—Talk shows and interviews. Journal of Pragmatics, 39, 1333-1341. Lee, R. & Morin, A. (2009). Using the 2008 presidential election to think about “Playing the race card”. Communication Studies, 60, 376-391. Lizza, R. (2003, December, 1). Air war. The New Republic. Retrieved Aug 2nd, 2010, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/pdfviewer/pdfviewer?vi d=4&hid=122&sid=3bd2d350-29fb-40fc-b1503c1a9da88149%40sessionmgr111 Longman Dictionary of Contemporary English. (2005). s.v. “buy”, s.v. “sell”. Harlow: Pearson. Low, G. & Cameron, L. (2002). Applied-linguistic comments on metaphor identification. Language and Literature, 11, 84-90. Lowry, R. (2007, January, 25). Hillary the Calculator. National Review. Retrieved Sept 2nd, 2008, from http://article.nationalreview.com/?q=ZWQ5NTY1ZGU2MWEzOTgwMD QwMTE0MzRmMjgzN2ZiN2E Macgilchrist, F. (2011). Journalism and the Political: Discursive Tensions in News Coverage of Russia. Amsterdam: John Benjamins. Madden, M (2008, September 3). McCain: No, really, we vetted her! Salon.com. Retrieved Aug 2nd,2011, from 118 http://www.salon.com/env/feature/2008/09/05/sarah_palin_down_syndrom e/print.html Maher, B. (2004, February, 7). Lose the twang, y'all. Enough with the Civil War complex: It's time for Southern Democrats to get enlightened about voting Bush out of office. Salon.com. Retrieved Aug 2nd, 2011, from http://www.salon.com/news/opinion/feature/2004/02/07/maher McCarthy, A.C. (2008, October, 20). Obama’s assault on Our Unum: What “fundamental change” is all about. National Review.Retrieved Oct 23th, 2008, from http://article.nationalreview.com/?q=MmI4NmQ2NWI2NWZlY2JhZjE3N TRhNDFmMjc1YzhhOGY= McCauley, M. (2005). NPR: The Trials and Triumphs of National Public Radio. New York: Columbia University Press. McChesney, R. W. (2004). The Problem of the Media: U.S. Communication Politics in the Twenty-First Century. New York: Monthly Review Press. Meet The Press. (2005, May, 13th). Retrieved Aug 2nd, 2011, from MSNBC, COCA. Metaphor Analysis Project and MetNet. Retrieved Aug 2nd, 2010, from http://creet.open.ac.uk/projects/metaphor-analysis/index.cfm Meyerson, H. (2000, July, 31). Gore's mating ritual. The American Prospect. Retrieved Aug 2nd, 2010, from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=56599799&sid=1 0&Fmt=4&clientId=23404&RQT=309&VName=PQD Miller D. & Johnson J. H (2009). Strict Fathers or Nurturing Parents? A corpus-assisted study of Congressional positioning on the war in Iraq. In J. Morley and P. Bayley (eds), Corpus-Assisted Discourse Studies on the Iraq Conflict: Wording the War. New York: Routledge. Pp. 34-73. Mio, J. (1997). Metaphor and politics. Metaphor and Symbol, 12, 113-133. 119 Mio, J. & Katz A. (1996). Metaphor: Implication and Application. Mahwah: Erlbaum. Moberg, D. (2007, September, 19). Obama’s in the eye of the beholder. InTheseTimes.com. Retrieved Aug 2nd, 2011, from http://www.inthesetimes.com/article/3329/obamas_in_the_eye_of_the_beh older/ Montagne, R. (2005). Analysis: Political aftermath of Hurricane Rita. Morning edition. Retrieved Aug 2nd, 2011, from National Public Radio, the COCA. Moon, R. (1998). Fixed Expressions and Idioms in English: A Corpus-Based Approach. Oxford: Oxford University Press. Müller, C. (2008). Metaphors Dead and Alive, Sleeping and Waking: A Dynamic View. Chicago: University of Chicago Press. Nation, The (2011). In Encyclopædia Britannica. Retrieved March 20th, 2011, from http://www.britannica.com.libproxy.helsinki.fi/EBchecked/topic/40422 3/The-Nation National Review (2011). In Encyclopædia Britannica. Retrieved March 20th, 2011, from http://www.britannica.com.libproxy.helsinki.fi/EBchecked/topic/405341/N ational-Review Noonan, P. (2008. September, 3). Open mic night at MSNBC. The Wall Street Journal. Retrieved Sept 5th, 2008, from http://online.wsj.com/article/SB122044753790594947.html Norris, M. (2005). Analysis: Judge rules use of religion in hiring decisions is OK. All Things Considered. Retrieved Aug 2nd, 2011, from National Public Radio, the COCA. Oliphant, K. (2004, August). The Kerry I know. The American Prospect. Retrieved Aug 2nd, 2011, from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=670965091&sid= 11&Fmt=4&clientId=23404&RQT=309&VName=PQD 120 On the Stump at Ames (1999, August, 27). Human Events, Retrieved March 12th, 2011, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/detail?sid=80313c47c4b1-4b7b-b318 f3f041b67df8%40sessionmgr12&vid=2&hid=8&bdata=JnNpdGU9ZWhvc 3QtbGl2ZSZzY29wZT1zaXRl#db=a9h&AN=2212803 Our Choice (2004, January, 19). The New Republic. Retrieved Aug 2nd, 2010, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/resultsadvanced?sid=11 6e6708-8b9f-4abe-b0913c5c381b0d29%40sessionmgr114&vid=3&hid=122&bquery=(JN+%26quo t%3bNew+Republic%26quot%3b)+AND+(Our+Choice)&bdata=JmRiPW E5aCZ0eXBlPTEmc2l0ZT1laG9zdC1saXZlJnNjb3BlPXNpdGU%3d Oxford English Dictionary, The (1989). s.v. ”buy”, s.v. “sell”. Retrieved 1st Nov, 2009, from http://dictionary.oed.com.libproxy.helsinki.fi/entrance.dtl. Packer, G. (2008, September, 8). Vets for Democrats. www.newyorker.com. Retrieved Aug 2nd, 2011, from http://www.newyorker.com/talk/2008/09/08/080908ta_talk_packer?printabl e=true Partington, A. 1998. Patterns and Meanings. Amsterdam: John Benjamins. Pasma T. (2011). Metaphor and Register Variation: The Personalization of Dutch News Discourse. Dissertation. Retrieved Jun 18th, 2011, from http://dare.ubvu.vu.nl/bitstream/1871/19630/1/dissertation.pdf Patterson, T. E. (1993). Out of Order. New York: Knopf. Patterson, T. E. (2002). The Vanishing Voter: Public Involvement in an Age of Uncertainty. New York: Alfred A. Knopf. 121 Peil, D. (1983). Untersuchungen zur Staats- und Herrschaftsmetaphorik in literarischen Zeugnissen von der Antike bis zur Gegenwart. Munich: Wilhelm Fink. Pelosi, C. (2010). Democrats' dilemma: Fall in love or fall in line? www.huffingtonpost.com. Retrieved Feb 28th, , 2011, from http://www.huffingtonpost.com/christine-pelosi/democrats-dilemma-fallin_b_795389.html. Peters, C. (2004, October). Tilting at Windmills. The Washington Monthly. Retrieved Aug 2nd, 2011, from COCA. PEW 2008. Winning the Media Campaign. Project for Excellence in Journalism. PEW Research Center. Retrieved Nov 18th, 2010, from http://www.journalism.org/node/13308 Pinker, S. (2007). The Stuff of Thought: Language as a Window Into Human Nature. New York: Viking. Pfiffner, J. P. (2000). The Modern Presidency. Boston: Bedford/St. Martin's. Ponnuru, R. (2007, March, 19). Coming McCain Moment: Taking a second look. National Review. Retrieved Aug 2nd , 2011, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/pdfviewer/pdfviewer?vi d=4&hid=122&sid=48e762b8-7fcd-4786-bd08178480f36918%40sessionmgr114 Ponnuru, R. (2000, September, 11). The Gore Straddle. National Review. Retrieved Aug 2nd, 2011, from http://web.ebscohost.com.libproxy.helsinki.fi/ehost/pdfviewer/pdfviewer?vi d=4&hid=122&sid=48e762b8-7fcd-4786-bd08178480f36918%40sessionmgr114 Posner, S. (2008. October, 27). Democrats chase evangelical votes. The Nation. Retrieved Oct 20th, 2008, from http://www.thenation.com/article/democrats-chase-evangelical-votes 122 Prose, F. (2004, March). Voting democracy off the island. Harper’s Magazine. Retrieved Aug 2nd, 2011, from the COCA. Roth, Z. (2007, May). The Upstart. The Washington Monthly. Retrieved Sept 2nd, 2011, from COCA. Rothenberg, S. (2008, August, 20). Should Democrats be feeling any kind of buyer's remorse? www.realclearpolitics.com. Retrieved Aug 26th, 2008, from http://www.realclearpolitics.com/articles/2008/08/should_democrats_be_fe eling_an.html Rush Limbaugh (2011). In Encyclopædia Britannica. Retrieved Jul 20th, 2011. from http://www.britannica.com.libproxy.helsinki.fi/EBchecked/topic/341204/R ush-Limbaugh Salama, A. H. Y. (2011). Ideological collocation and the recontexualization of WahhabiSaudi Islam post-9/11: A synergy of corpus linguistics and critical discourse analysis. Discourse & Society, 22, 315-342. Scheiber, N. (2007, September, 24). A role reversal for Obama and Edwards. www.tnr.com. Retrieved Sept 25th, 2007, from http://www.tnr.com/blog/the-plank/role-reversal-obama-and-edwards Scheithauer, R. (2007). Metaphors in television election night coverage in Britain, the United States and Germany. In A. Fetzer & G. Lauerbach (eds), Political Discourse in the Media: Cross-cultural Perspective. Amsterdam: John Benjamins. Pp. 75-108. Simon, R. (2000, March, 20). The primary lessons: The campaigns aren't pausing, but a look back shows where they need to go. U.S. News & World Report. Retrieved Aug 2nd, 2010, from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=51049783&sid=1 2&Fmt=2&clientId=23404&RQT=309&VName=PQD 123 Sinclair, B. (2006). Party Wars: Polarization and the Politics of National Policy Making. Norman: University of Oklahoma Press. Stark, S. (2007, September, 15). Does America want a Clinton monarchy? RealClearPolitics.com, Retrieved Sept, 23rd, 2007, from http://www.realclearpolitics.com/articles/2007/09/will_voters_embrace_a_c linton.html Steen, G. (2007). Finding Metaphor in Grammar and Usage: A Methodological Analysis of Theory and Research. Amsterdam: Benjamins. Steen, G. J., Dorst, A. G., Herrmann, J. B., Kaal, A. A., & Krennmayr, T. (2010). Metaphor in usage. Cognitive Linguistics, 21, 765-796. Steinberg, B. (2004, August, 30). Is Bush a Ford and Kerry a BMW? Poll voters define senator, president as rival brands: Challenger vs. Old Reliable. The Wall Street Journal. Retrieved Aug 2nd, 2010, from http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=685229281&sid= 13&Fmt=3&clientId=23404&RQT=309&VName=PQD Stewart, D. (2009). Semantic Prosody: A Critical Evaluation. New York: Routledge, Stubbs, M. (1995). Collocations and cultural connotations of common words. Linguistics and Education, 7, 379-390. Stubbs, M. (2003). Two quantitative methods of studying phraseology in English. International Journal of Corpus Linguistics, 7, 215-244. Taibbi M. (2008, September, 18). Obama on the trail. www. rollingstone.com. Retrieved Aug 2nd, 2011, from http://smirkingchimp.com/thread/17003. Tapper, J. (1999, August, 16). They feed horses, don't they? Salon.com. Retrieved Jul 24th, 2010, from http://www.salon.com/news/feature/1999/08/16/straw The Beltway Boys (2008, September, 3). The Wall Street Journal. Retrieved Aug 2nd, 2010, from 124 http://proquest.umi.com.libproxy.helsinki.fi/pqdlink?did=1546662531&sid =14&Fmt=4&clientId=23404&RQT=309&VName=PQD The bitter end (2007, June). The Washington Monthly. Retrieved Aug 2nd, 2011, from COCA. The press: the hairline fracture (1972, May, 15). TIME. Retrieved Aug 8th, 2011, from http://www.time.com/time/magazine/article/0,9171,903469,00.html#ixzz1 UPyt2vuY Vertessen, D. & De Landstheer, C. (2008). A metaphorical election style: Use of metaphor at election time. In T. Carver & J. Pikalo (eds), Political Language and Metaphor. New York: Routledge. Pp. 271-285. Whitaker, W. R., Ramsey, J. E. & Smith, R. D. (2009). Mediawriting: Print, Broadcast, and Public Relations. New York: Routledge. Wilson, J. (1990). Power to the people: Political metaphors. In J. Wilson (ed.), Politically Speaking. Oxford: Blackwell. Pp. 104-130. Zeldes, G. A., Fico, F., Carpenter, S. & Diddi, A. (2008). Partisan balance and bias in network coverage of the 2000 and 2004 presidential elections. Journal of Broadcasting & Electronic Media, 52, 563-580. Zinken, J. (2004). Metaphor practices in the German Wende discourse. Journal of Multilingual & Multicultural Development, 25, 424-436.