CHAPTER DISCUSSION AND CONCLUSION 7.1 Introduction: Triangulation (The key advantage of corpus informed research) Baker (2009) attributes considerable significance to the field of corpus linguistics because it successfully captures the “incremental effect of discourse” (p.13). This simply refers to the way underlying evaluations are effectively uncovered through an in-depth analysis of how language is widely employed across specific societal discourses. The proposed Integratedapproach to the Identification of Conceptual Metaphors (IICM) in this thesis aims to “triangulate” the entire process of metaphor research by going beyond the pure use of intuition as a springboard for subsequent metaphor analysis in a large collection of representative texts. The overarching aim of this thesis is to demonstrate the efficacy of the proposed IICM framework by examining a cross-section of the repeated patterns of metaphor use across all four British and American broadsheets in order to uncover the dominant metaphorical evaluations embodied by the respective broadsheet corpora. This is done in order to demonstrate a small snapshot of the range of discourse positions available to the broadsheet reader, contributing to the ongoing discussion in the way discourse prosody potentially influences and shapes shared cultural viewpoints. “Triangulation”, in the context of this thesis, refers to the use of the integrated approach advocated by the IICM, which essentially uses multiple frameworks and methods of analysis to validate the resultant findings. This   298   “integrated approach” embodies the combination and adaptation of various frameworks into an effective, unified system for metaphorical analysis. The strength of such an approach lies in three significant and inter-connected areas: • Firstly, the empirical and inductive nature of this proposed conceptual metaphor formulating process aims to provide an alternative framework for the triangulation of metaphorical findings via the current practice of intensive manual analysis of representative data. Thus, it aims to provide an inductive framework that potentially complements existing deductive methodologies. • Secondly, this proposed method carries the potential for an exhaustive analysis of metaphors in a large collection of representative texts if deemed necessary by the researcher due to the exhaustive lists of lexical items and USAS domains generated by both concordancing programmes. Thus, the IICM plays the role of a comprehensive integrated framework for large-scale metaphorical analysis. • Finally, the use of two leading online concordancing platforms (WMatrix and Sketch Engine) anchors the range of findings in potentially empirically robust interpretations and evaluations. This facilitates validity checks due to the detailed and repeatable procedures embodied in the IICM. At this juncture, it is important to acknowledge that informed researcher intuition is an integral part of all discourse-analytical corpus   299   research. This is due to the fact that while the USAS semantic domains and “word sketches” provide a springboard for the analysis, it is ultimately the researcher who determines the metaphoricity of the lexical units and formulates the relevant conceptual mappings. Hence, researcher bias cannot and should not be totally eliminated, as it is virtually impossible to deny one’s sociocultural and intuitive responses to language. However, the main aim of this thesis is the formulation of a repeatable method that strikes an adequate balance between semi-automated and manual analysis. Hence, it is hoped that the systematicity of the metaphor mining process proposed by the IICM would ensure repeatability and result in a more comprehensive analysis of the manifestations of metaphoricity. 7.2 Findings and contributions It is important to note that while the range of metaphors mined from these 250 articles per broadsheet might not be wholly encompassing, I believe that this preliminary methodological demonstration done via the IICMframework showcases its potential to surface the predominant evaluative trends pertaining to the issue under study. The utility of the IICM framework is demonstrated in the way subtle differences in the conceptual metaphors used within both national cultures are surfaced in the public debates. This is in spite of the fact that these metaphors are largely similar in range due to the close intertwining of the historical and political cultures between Britain and America. In the case of the BP oil spill, the metaphorical framing of this conflict between two traditionally strong allies has shown the potential for a   300   nationalistic and ideological bias. For example, the findings in the British broadsheets show the tendency of the relevant editors to adopt mitigating strategies in order to downplay the agency of the cause of the entire catastrophe (mainly attributed to BP, a company with distinctively British origins). These mitigating strategies include the strategy of passivising sentences in order to delete agency, the use of parallels in order to deflect attention (by alluding to the recent disasters in the US that highlight US Government incompetence) as well as the use of disaster and financial metaphors to describe abstract processes in terms of disaster and physical damage in order to further conceal the actions of those who have caused the disaster (see detailed discussion in Chapter 5). This is in contrast to the American broadsheets where the framing of the conflict generally focuses more on the pragmatic aspect of economic restitution by the guilty parties, emphasizing on the scale of the economic and ecological damage in an attempt to galvanise the citizenry against a common enemy. There is another strong unifying narrative strand that foregrounds a nationalistic stance, with the primary focus placed on the investigative aspect of the entire tragedy with the aim of castigating the CRIMINALS in the pursuit of justice as well as the parallel search for a deterrent measure to prevent the recurrence of a catastrophe on such a scale. This is done by voicing national outrage in culturally specific terms and analogies through the use of the 9/11 terror attacks or Hurricane Katrina as a parallel to the catastrophe. A detailed exposition of these metaphorical evaluations can be seen in Chapter 6. However, the dimension of intra-cultural bias is also evident in the way the liberal broadsheets in both the Britain and America (the NYT-corpus and   301   the G-corpus) frame the conflict in predominantly antagonistic and nationalistic terms. These liberal broadsheets tend to focus on the scale of the disaster and emphasise the range of negative emotions arising from the BP Oil Spill, favouring a narrative centering on the pursuit of justice and the relentless castigation of culpable parties. This can be seen in the higher percentage of negative evaluations for the key players in the oil spill in the NYT and G-corpus. This is in contrast to the conservative broadsheets (the WP-corpus and the TT-corpus) that tend to frame the conflict in pragmatic economic terms, focusing on the all-encompassing concern of economic and financial survival of the respective economies. This is   accomplished through the predominance of the WAR/CRIME/THREAT conceptual key in the WPcorpus (53.6%) and TT-corpus (68.1%), conceptualizing BUSINESS as a WAR/ GAME OF SURVIVAL in an effort to justify BP’s continued economic survival. Thus, it would appear that the conservative broadsheets adopt a less explicit, and a more subtle and mitigated approach to the metaphorical evaluation of the key actors in the oil spill – in an effort to foreground the more pragmatic, pro-business editorial ideology (language of ECONOMIC PRAGMATISM/ ACCIDENT SCENARIO). This is in diametric opposition to the liberal broadsheets which are more focused on the more explicit and intense metaphorical framings of the scale of the disaster and on the ideological adherence to notions of justice and ecological sustainability (MORALACCOUNTING METAPHOR/ CRIME AND PUNISHMENT SCENARIO). These findings clearly demonstrate that metaphor choice is motivated by both ideological outlook and the rhetorical aim of persuasion.         302   7.3 Limitations of the study and suggestions for further research 7.3.1. The need for a multimodal perspective There is an emerging need for discourse analysis to include the examination of the accompanying photos and positional strategies that appear in tandem with the corresponding printed pages or the webpages of the online broadsheets. The specific images that should be analysed include those accompanying the headlines taken from the relevant image banks. Two examples are embodied by Figures 7.1-7.2. Fig. 7.1. Image capturing the scale of the environmental catastrophe. Fig. 7.2 Image depicting the scenes of the initial blowout on the Deepwater Horizon Rig.   303   These images have a high meaning potential as they can be used not only to document particular moments in time and place but also to symbolise the generic. This increased use of images to represent disasters is part of a broader change in the use of photography in the mass media. Rather than merely documenting events, the photograph is increasingly central as part of layout and used symbolically. Furthermore, in many cases these online newspaper sites used their websites' interactive features to illuminate aspects of the story that would have been harder to digest in print or broadcast formats. This salience of this point is further illustrated in the way some of these online newspapers have done better than many traditional media outlets in making effective use of these interactive features on their websites to track key aspects of the BP disaster. For example, some widgets enable the direct monitoring of the amount of oil spilling into the Gulf through a visual video feed, whilst the NYT website offered a video animation that helped explain how a last ditch effort to prevent the spill failed. 7.3.2. The cross-classification of metaphors Grady, Taub & Morgan’s (1996) delineation of all metaphors into primitives (“simpler sub-mappings”) and compounds (“metaphors that are compositional in nature”) attempts to address two core issues plaguing the metaphorical analysis in large representative texts (p.179-181):   304   1. The inefficiency of analysis i.e. the issue of organising the sprawling metaphorical complexes that arise in order to account for all the relevant correspondences from linguistic evidence. An ambitious preliminary attempt to account for the range of resultant conceptual metaphors from large representative corpora using Grady, Taub & Morgan’s concept can be seen in Figures 6.9-6.12 in Chapter 6. 2. The issue of accounting for duality (i.e. the cross-classification of metaphors). A simple hypothetical example is embodied in Figure 7.3 below. The cross-classification or metaphorical “duality” is accounted for in Figure 7.3 where the structural metaphors of POLITICS IS WAR and BUSINESS IS WAR are diagrammatically reconciled by depicting the Gulf Economy being a victim of poor business strategy and bad politics. These metaphors are classified under the overarching WAR/CRIME scenario with the conceptual key OIL IS A THREAT).   Fig  7.3   Pictorial   Representation   of   Metaphorical   Primitives   and   Compounds   (adapted   from  Grady,  Taub  &  Morgan,  1996)               BUSINESS  I  S  WAR        THE  GULF   ECONOMY  IS  A   VICTIM             CONCEPTUAL  KEY  (SCENARIO):   THE  OIL  IS  A  THREAT   (WAR/CRIME  SCENARIO)     305     POLITICS  IS  WAR   Hence, Grady, Taub & Morgan’s framework claims to result in the efficiency of analysis by adequately accounting for the problematic issue of metaphorical cross-classification. This is achieved by accounting for the conceptual gaps arising from the independent motivations of each metaphorical primitive through the simple use of the overlaps to represent the resultant metaphorical compounds. However, the element of subjectivity in all experience of metaphor is inevitable as there are too many eclectic, overlapping factors that affect the individual analyst or text consumer’s emotional response to language. Hence, even though the analytical method proposed in the IICM is clear and the reader is free to challenge metaphor classifications, the subjectivity inherent in the scenarios result in interpretative ambiguity on the part of both the textproducer and text consumer in the attempt to establish a common scenario. This can be clearly seen in the presuppositions inherent in the WAR scenario that enables it to be used as an embodiment of two opposing target domains – POLITICS and BUSINESS. This clearly shows how the text-producer’s intentions behind the use of a metaphor are largely dependent on the text-consumer’s socio-political outlook and experiences. Thus, a framework that automatically includes the cognitive-semantic elements (social and metaphorical contexts) should be included in subsequent permutations of the IICM. Since metaphor draws on the unconscious emotional associations of words, which are often rooted in contextual and cultural understandings, the eventual development of a holistic approach to conceptual metaphor formulation that incorporates cultural and contextual knowledge will serve to further refine the current IICM framework.   306   7.3.3. The evolution of metaphor (A longitudinal perspective) Another possible area of future research would lie in the tracking and analysis of the diachronic development of metaphorical concepts. This is especially so as both politicians and the media are mainly focused on the “fastest and widest possible dissemination and acceptance of their representations”, accounting for the “density of micro-changes in usage patterns that we observed in variations and reinterpretations of certain prominent metaphor formulations” (Musolff, 2004, p.121). The potential implications of these findings can be significant in accounting for the eventual survival or demise of the relevant actors in the BP Oil Spill e.g. the current British and American metaphorical representations of BP is of significance in accounting for BP’s prospects and long-term survival in the US and beyond (as lawsuits are being formally filed from 2013). However, in order to substantiate and formulate these metaphors to track the evolution of public discourse, separate corpora from across all four broadsheets have to be collected. These corpora also have to be collected from different and clearly delineated timeframes in order to track the “evolutionary transformation” between the successive and predominant metaphorical representations of the key entities under study. Hence, a longitudinal approach to the study of the metaphors from the same issue across a range of different domains (scientific journals, literary references, organization and management manuals and brochures, law texts, political speeches and religious references) would effectively showcase how a set of particularly interesting set of metaphors is “formulated, developed,   307   shared and passed on, exploited, transformed between individuals and groups of participants within and across a discourse event” (Steen, 2010, p.254; Gentner and Bowdle, 2008). 7.3.4. The lexical unit as a starting point: Limitations Finally, it has been established that the basis of CMT is not lexical but rather, conceptual. This suggests the need to move away from lexical analysis as the key point of departure in metaphor analysis across large datasets. Thus, the key issue with the current metaphor identification procedure in the IICM is the fact that the springboard for all its analysis stems from the individual lexical or multi-word unit. It is however, clear that metaphorical concepts can be found even in the absence of metaphorical language e.g. allegory (Pragglejaz Group, 2007, p.24). This simply means that socio-historical, contextual and specific linguistic factors as well as the co-text have to be taken into account for a thorough metaphorical analysis. This point is aptly represented by Oster (2010, p.730) who states that “(current) metaphorical pattern analysis consists of choosing a lexical item from the target domain, extracting a random sample of its occurrences in the corpus, identifying all metaphorical expressions that the search word is a part of, and grouping them according to general mappings.” This statement highlights the key shortcoming that when the starting point of the analysis is lexical, the list of co-occurrences under consideration only occurs in the vicinity of the node word. Expressions that are used figuratively to describe the targetted concept without naming it cannot be detected with this method. Hence, this might be   308   the impetus that pushes the analyst towards an exhaustive approach to the IICM framework by analyzing virtually the full range of empirically suggested items from the concordancing software. This simply means that the suggested limitation of the lexical approach does not seem to be a substantial drawback to the IICM even though it results in unnecessary tedium. At this juncture, it would be appropriate to list the methodological distinctions between Oster’s study of FEAR as an emotion concept and the methodological considerations in this thesis. Oster’s study can be seen as a landmark methodological demonstration that combines Kovecses’ (1990, 2008) Lexical Approach and Stefanowitsch’s (2006) Metaphorical Pattern Analysis with additional “promising conceptual tools for semantic or pragmatic analysis (including collocations, co-occurrences, semantic prosody and semantic preference)” (p.755). However, the study conducted in this thesis differs in the following methodological parameters: 1. Where Oster’s point of entry into the analysis was based on going directly to the pre-determined target lemma of “fear” (p.734), the IICM adopts a more inductive point-of-entry to the study of the selected target domains by selecting the most (comparably) frequent lexical item from empirically-salient USAS semantic domains (as indicated by the LL-value) across the four broadsheets. This provides a bottom-up approach to the data that is empirically-based. 2. Oster’s study on the linguistic and metaphorical manifestations of “fear” as an emotion concept was based on the Corpus of Contemporary American English, a large general corpus of 300   309   million words. However, this thesis extends his methodology by embarking on a comparative study across four separate corpora on the metaphorical representations of core concepts in the BP Oil Spill. Thus, the pre-selection of comparable sub-corpora (as outlined in the data-collection process in Chapter 1) is needed rather than a general corpus for the purpose of establishing a comparative study. 3. Finally, the typological classifications of the metaphorical expressions also differ in this thesis. Oster’s in-depth study of the metaphorical realisation of the emotion concept of FEAR essentially divides FEAR into several sub-types: SOMETHING INSIDE THE BODY; a FORCE (e.g. an ANTAGONIST, AUTONOMOUS FORCE); an AUTONOMOUS BEING; an ILLNESS/INSANITY; an OBJECT or a PLACE/CONTAINER (p.739). However, the study in this thesis focuses on the range of metaphorical sub-types that emerge from the metaphorical study of the selected entities in the discourse (such as BP itself, the British and American Governments, the spilt oil as an entity, the notions of POLITICS and INTERNATIONAL RELATIONS in general etc.). These metaphorical subtypes are divided into Structural Metaphors, Ontological Metaphors, Orientational Metaphors and Image/Analogy-Based Metaphors respectively.             310   7.4 Concluding remarks In this thesis, I have attempted to introduce a unique approach to Conceptual Metaphor analysis that uses an existing semantic annotation tool (USAS) within WMatrix in combination with “word sketches” in Sketch Engine. This is done through the formulation of a proposed new metaphorical analytical framework: the Integrated-approach to the Identification of Conceptual Metaphors (IICM). The key methodological parameters lie in the following areas: 1. The combined utilisation of two leading online concordancing platforms, WMatrix and Sketch Engine as an empirical point of entry for metaphorical analysis in a large collection of representative texts: • WMatrix provides an indication of the semantic categories that stand out from the reference text. The most (comparably) frequent lexical unit from these salient semantic domains will then be selected as the Target Domain for further metaphorical analysis. The conceptual metaphors will only be formulated from a detailed analysis of the co-text and the context (socio-linguistic, political and situational) of the empirically suggested lexical units. This is further systematised through the reliance on corpus-based dictionaries in the actual formulation of the conceptual metaphors. •   The Word Sketch function in Sketch Engine further refines the 311   metaphorical mining procedure. This is because the actual formulation of the conceptual metaphors from linguistic evidence will be based on the node word’s grammatical and collocational behaviours. The additional Part-of-Speech annotation afforded by word sketch refines the metaphorical analysis because “collocates may differ considerably between nouns and verbs of the same lemma” (Oster, 2010, p.735). This simply means that the concordances examined will be further parsed according to the grammatical relation of the node word, resulting in a more refined approach to the formulation of conceptual metaphors from linguistic evidence. This is a significant departure from the previous practice of merely examining the ‘arbitrary window of text around the headword’ for the first pre-designated number of lines (Kilgarriff et al., 2004, p.3). 2. The formulation of the Integrated-approach to the Identification of Conceptual Metaphors (IICM), a methodology that proposes an inductive and empirical point of entry to the analysis of large swaths of data (see section 3.4.1. and Appendix A). • The three broad phases and the accompanying steps embodied by the IICM provide a common springboard for academic disagreements due to its systematicity. This is because the explicit commitment to empirical evidence from these online concordancing programmes as well as the detailed series of   312   accompanying steps will enable other researchers to base their disagreements on the same platform with respect to the same set of evidence, thus potentially facilitating subsequent academic discussions on metaphor in discourse. • Finally, the adaptation of Grady et al.’s (1996) useful distinction between primitive and compound metaphors results in the efficiency of analysis by organising sprawling metaphorical complexes that often arise from large-scale discourse studies into a coherent relational framework. The resultant diagrammatic depiction organises the range of conceptual metaphors into a coherent diagrammatic depiction. The hierarchical aspect of the representation in terms of frequency is reflected by the corresponding size of the circles. This adaptation of Grady et al.’s idea also facilitates a comparative cross-cultural and intra-cultural analysis, as the specific experiential embodiments in each context will be foregrounded through the specific comparison of the choice of primitive metaphors. This is due to the fact that a direct intertextual comparison of the analogy-based metaphors (which tend to be more experientially embodied) enables the analyst to see the direct contextual and cultural influences that are bound by context and culture (Johansson Falck & Gibbs Jr, 2012, p.254).   313   In other words, this proposed framework minimises the use of “predetermined search strings” that potentially incorporate a degree of researcher bias as the first point of entry into the metaphorical analysis of the corpus. This is because the selection of target domain concepts for analysis in this framework is based on the empirical suggestion of candidate semantic fields proposed by USAS and the empirical salience of the collocations proposed by word sketch in Sketch Engine respectively. In the process, this proposed framework also minimises the chances of missing out on large chunks of relevant and possibly significant lexical items as the researcher can theoretically adopt an exhaustive approach to the analysis to the full range of semantic domains and lexical items generated by WMatrix and Sketch Engine. As a result, the IICM is potentially a lot more inductive, generates more comprehensive results than the use of pre-determined search strings and most importantly, is repeatable and empirical. The inductive approach embodied by the IICM can also be used as an alternative framework to triangulate the findings of the intensive manual analysis for a more robust claim to the study. With subsequent iterations, the IICM can potentially evolve into a framework that attempts to embody a combination of automatic semantic analysis, key word identification and a well-documented conceptual metaphor-identification procedure into a repeatable method that strikes an adequate balance between semi-automated and manual analysis.   314   [...]... through the formulation of a proposed new metaphorical analytical framework: the Integrated-approach to the Identification of Conceptual Metaphors (IICM) The key methodological parameters lie in the following areas: 1 The combined utilisation of two leading online concordancing platforms, WMatrix and Sketch Engine as an empirical point of entry for metaphorical analysis in a large collection of representative... BEING; an ILLNESS/INSANITY; an OBJECT or a PLACE/CONTAINER (p .73 9) However, the study in this thesis focuses on the range of metaphorical sub-types that emerge from the metaphorical study of the selected entities in the discourse (such as BP itself, the British and American Governments, the spilt oil as an entity, the notions of POLITICS and INTERNATIONAL RELATIONS in general etc.) These metaphorical... grammatical and collocational behaviours The additional Part -of- Speech annotation afforded by word sketch refines the metaphorical analysis because “collocates may differ considerably between nouns and verbs of the same lemma” (Oster, 2010, p .73 5) This simply means that the concordances examined will be further parsed according to the grammatical relation of the node word, resulting in a more refined... needed rather than a general corpus for the purpose of establishing a comparative study 3 Finally, the typological classifications of the metaphorical expressions also differ in this thesis Oster’s in- depth study of the metaphorical realisation of the emotion concept of FEAR essentially divides FEAR into several sub-types: SOMETHING INSIDE THE BODY; a FORCE (e.g an ANTAGONIST, AUTONOMOUS FORCE); an AUTONOMOUS... adopts a more inductive point -of- entry to the study of the selected target domains by selecting the most (comparably) frequent lexical item from empirically-salient USAS semantic domains (as indicated by the LL-value) across the four broadsheets This provides a bottom-up approach to the data that is empirically -based 2 Oster’s study on the linguistic and metaphorical manifestations of “fear” as an emotion... sprawling metaphorical complexes that often arise from large-scale discourse studies into a coherent relational framework The resultant diagrammatic depiction organises the range of conceptual metaphors into a coherent diagrammatic depiction The hierarchical aspect of the representation in terms of frequency is reflected by the corresponding size of the circles This adaptation of Grady et al.’s idea also... WMatrix provides an indication of the semantic categories that stand out from the reference text The most (comparably) frequent lexical unit from these salient semantic domains will then be selected as the Target Domain for further metaphorical analysis The conceptual metaphors will only be formulated from a detailed analysis of the co-text and the context (socio-linguistic, political and situational)... exhaustive approach to the analysis to the full range of semantic domains and lexical items generated by WMatrix and Sketch Engine As a result, the IICM is potentially a lot more inductive, generates more comprehensive results than the use of pre-determined search strings and most importantly, is repeatable and empirical The inductive approach embodied by the IICM can also be used as an alternative framework...shared and passed on, exploited, transformed between individuals and groups of participants within and across a discourse event” (Steen, 2010, p.254; Gentner and Bowdle, 2008) 7. 3.4 The lexical unit as a starting point: Limitations Finally, it has been established that the basis of CMT is not lexical but rather, conceptual This suggests the need to move away from lexical analysis as the key point of. .. framework to triangulate the findings of the intensive manual analysis for a more robust claim to the study With subsequent iterations, the IICM can potentially evolve into a framework that attempts to embody a combination of automatic semantic analysis, key word identification and a well-documented conceptual metaphor-identification procedure into a repeatable method that strikes an adequate balance between . unifying narrative strand that foregrounds a nationalistic stance, with the primary focus placed on the investigative aspect of the entire tragedy with the aim of castigating the CRIMINALS in the. terms and analogies through the use of the 9/11 terror attacks or Hurricane Katrina as a parallel to the catastrophe. A detailed exposition of these metaphorical evaluations can be seen in Chapter. explicit and intense metaphorical framings of the scale of the disaster and on the ideological adherence to notions of justice and ecological sustainability (MORAL- ACCOUNTING METAPHOR/ CRIME AND