262 EXTENSION: LINGUISTIC READINGS of this clause would have to specify some sort of agency, either explicitly in the form of active or passive non-middle clauses: X shattered the rear windows. The rear windows were shattered by X. or implicitly, in the form of a passive with the agent deleted: The rear windows were shattered. If we consider the three narrative clauses of the original text together, there is the impres- sion that the processes expressed by them are simply self-engendered, uninduced by any external cause. This is odd, given that they depict the violent effects brought about precisely by an external cause. The impact of the bomb explosion is further dimin- ished and, for that matter, trivialized, through the syntactic frame which incorporates these three clauses. For instance, each clause is presented in an asyndetic ‘listing’ fash- ion without the use of formal connectors. Furthermore, the consequences of the explo- sion – especially the French girl’s loss of an eye – are clearly undermined by the following adversative clause (‘but essentially the children escaped scot-free’), the content of which is manifestly at odds with what has gone before. [. . . O]ther descriptions of violence in The Little Drummer Girl exhibit similar transitivity patterns. [. . .] What this brief analysis has sought to demonstrate is how a certain type of transitivity pattern, especially when developed in conjunction with a positively- shaded modality, can function as an ironizing technique. In the le Carré example, this convergence of modality and transitivity served to highlight the persona of the speaking voice whilst distancing the purported central event of the narrative. So, where the physical horror of certain events is suppressed, the opinionated subjectivity of the narrator is foregrounded. The use of the ironic narrative technique has been com- mented upon by critics of le Carré, one of whom remarks specifically of The Little Drummer Girl that Where explanations seem unnecessary they are given, where a particularly horrifying incident seems to demand the narrator’s acknowledgement of that horror, none is forthcoming. Often there is a throwaway matter-of-factness to descriptions or, again, a disturbing faux-naiveté. Such techniques give a cool pathos to parts of the narrative. (Barley 1986: 162) While endorsing these observations in general terms, I would want to add that a stylistic analysis will go some way towards explaining just why and how such ‘throw- away matter-of-factness’ and ‘cool pathos’ is created. Indeed, the rationale behind much modern stylistics is that not only does the use of linguistic models offer greater ‘purchase’ on texts but that it also provides the basis for comparative analyses of other texts using those same linguistic models. This comparative principle will further underpin the discussion of the final extract in the stylistic part of our transitivity programme. In order to provide a complete picture of the point of view spectrum, a text which exhibits negative modal shading will be used to round off this section. This will be a short sample of ‘Gothic’ horror fiction which represents a genre of writing not covered so far and which should be worth exploring in terms of transitivity. In the Paul Simpson SWIN|KCrEIB1Qqc8svpQueSEh0w==|1282036319 TRANSITIVITY AS POINT OF VIEW 263 extract below [. . .], a first-person narrator discovers that he is in the company of a less than personable companion: Was It – the dark form with the chain – a creature of this world, or a spectre? And again – more dreadful still – could it be that the corpses of wicked men were forced to rise, and haunt in the body the places where they had wrought their evil deeds? And was such as these my grisly neighbour? The chain faintly rattled. My hair bristled; my eyeballs seemed starting from their sockets; the damps of a great anguish were on my brow. My heart laboured as if I were crushed beneath some vast weight. Sometimes it appeared to stop its frenzied beatings, sometimes its pulsations were fierce and hurried; my breath came short and with extreme difficulty, and I shivered as if with cold; yet I feared to stir. It moved, it moaned, its fetters clanked dismally, the couch creaked and shook. ‘Horror: a true tale’, Blackwoods 89 (1861) The markers of negative modal shading take the form of epistemic modal verbs (‘could it be’), modal lexical verbs of perception (‘my eyeballs seemed starting’; ‘it appeared to stop’) and comparators based on reference to physical stimuli (‘My heart laboured as if I were crushed’; ‘I shivered as if with cold’). The ways in which transitivity patterns intersect with this modal shading are interesting. Despite the control that the narrator appears to have over his mental faculties, all suggestions of physical self-control disappear. Material processes of supervention signal the lack of command that the narrator has over, so to speak, his body parts. Here are a few illustrations: actor process My hair bristled actor process circumstances My eyeballs seemed starting from their sockets actor process My heart laboured The abject fear which this linguistic strategy is presumably designed to convey is one feature of the extract, but the attempt to convey suspense, I would suggest, relies on another textual feature. This second pattern relates to the transitivity patterns associ- ated with the protagonist’s ‘grisly neighbour’. If we look closely at the type of process in which the apparition is directly involved, all it really does is ‘move’ (material action) and ‘moan’ (verbalization?). Of course, there is the suggestion that it is responsible for more ‘happenings’ than are attributed to it directly. Consider, for instance, the ominous sequence The chain faintly rattled. The process of ‘rattling’ is one which allows an ergative interpretation. Although it may therefore permit agency, it can still be represented, as it is here, in the form of an agentless, middle clause. In other words, the chain just rattles, and any responsi- bility for the rattling is left unspecified and must be inferred from the context. Similarly, cause and effect relations are also suppressed in the final three material Paul Simpson 264 EXTENSION: LINGUISTIC READINGS processes of the extract, where agentless, middle clauses help develop a picture where inanimate objects appear to have a will of their own: its fetters clanked dismally, the couch creaked and shook. This pattern of transitivity squares neatly with the viewing position of the narrator. The apparition’s involvement in the movement of inanimate objects, although not perceived directly, is none the less imputed. The spatial point of view established here is therefore very much akin to a cinematographic technique which is employed almost to the point of cliché in horror films. This is the technique where the action is shot from within a darkly lit room. The camera pans towards the door and then brings in the door-handle in extreme close-up. The door-handle turns. Then the door creaks as it begins to open . . . Although certainly not consciously contrived to do so, the sequence of clauses just used to describe the movement of the door in my example exhibits the same type of transitivity pattern as that used in the passage of horror fiction examined above! The three short analyses undertaken in this section were designed to illustrate the potential of the transitivity model in stylistics. Although this type of analysis will not provide an exhaustive account of a text’s meaning, it should at least offer some insights into one important feature of message construction. Furthermore, some of the ways in which transitivity and modality interact have been assessed, and this has, I hope, enriched further our understanding of point of view in narrative fiction. Issues to consider q In principle, the framework Simpson sets out here can be used to explore the actional and reflective sequences of any literary work, though of course it will be of most interest in those texts in which the transitivity pattern is particularly thematically significant. In the extract above, Simpson uses examples from a literary classic, a spy thriller and a creepy tale; you could use the framework to explore other types of literary works. Can you find any characteristic patterns – in terms of transi- tivity – associated with particular genres? q Elsewhere in his book, Simpson shows how ‘free indirect discourse’ – a sort of blend of speech and thought – is used to create a ‘dual voice’ between character and narrator. He suggests free indirect discourse is the stylistic key to irony. Irony is a tricky problem in stylistics, since it is at least as much a readerly matter as a product inherent in the text. Irony tends to be studied within pragmatics (see B3) in order to capture this contextual dimension. However, there is no question that there are also accompanying stylistic features that encourage an ironic reading. Find a text that is usually considered to be ironic, and try to account for the stylis- tic features that point a reader to an ironic interpretation. Is free indirect discourse involved, or is the irony carried mainly by other means? q Simpson finds a few examples where his stylistic analysis offers a more system- atic and evidential basis for an impression that is articulated by a literary critic (here, Levin and Barley). You can easily find similar literary critical impressions, opinions, assertions and interpretations in discussions of literary works. Focus on one and see if the literary critic was actually onto something by trying to find the proper stylistic evidence to support their position. Paul Simpson QUALITATIVE, QUANTITATIVE, AND MIXED METHODS RESEARCH 265 QUALITATIVE, QUANTITATIVE, AND MIXED METHODS RESEARCH The extracts from Zoltán Dörnyei’s work here present us with a number of impor- tant arguments through which we can assess the viability of quantitative, qualitative and mixed-method approaches to conducting English language studies. One point which Dörnyei makes early on when talking in general terms about the social sciences is abso- lutely crucial to your own study of the English language: you are researching people, and therefore any study you conduct will always have to be selective. In any English language project that you devise you will need to decide upon the most reliable and valid way to be selective whilst ensuring that you conduct your pro- ject in a manner which enables you to gain access to the data you require. Once you have collected your data, you need to ensure that it will enable you to produce a mean- ingful analysis which answers your research questions/examines your chosen area of study. In any study of the English language which you encounter as a reader/listener, whether it be quantitative, qualitative or mixed-method, you need to develop the skills that enable you to consider critically the methodological choices that researchers have made so that you can judge the reliability and validity of their research findings to make an informed assessment of the overall viability of their arguments. Zoltán Dörnyei (reprinted from Research Methods in Applied Linguistics (2007), Oxford: Oxford University Press, pp. 25–46) The qualitative–quantitative distinction Although at first sight the difference between qualitative and quantitative data/research appears to be relatively straightforward, the distinction has been the source of a great deal of discussion in the past at every conceivable level of abstraction. Without dwelling on this issue too long, let me offer a taste of how things can get very com- plicated when we start discussing the qual–quan contrast. To start with, is there really such a contrast? And if so, where exactly does it lie? Richards (2005), for example, points out that the numerical versus non-numerical dis- tinction does not give us clear enough guidelines because qualitative researchers would almost always collect some information in numbers (for example, the age of the participants), and similarly, quantitative researchers usually also collect some non- numerical information (for example, the gender or nationality of the participants). So, as she concludes, ‘qualitative and quantitative data do not inhabit different worlds. They are different ways of recording observations of the same world’ (p. 36). Arguing in a similar vein, Miles and Huberman (1994) assert that in some sense, all data are qualitative because they refer to ‘essences of people, objects and situations’ (p. 9); sometimes we convert our raw experiences of the social world into words (i.e. qual), at other times into numbers (i.e. quan). Therefore, Sandelowski (2003) actually concludes that qualitative research is not clearly distinguishable from quan- titative research because there is no consistent manner in which such a comparison can be made. Even though I agree that qual and quan are not extremes but rather form a con- tinuum, we still tend to compare them all the time. Why is that? I would suggest that D12 Zoltán Dörnyei 266 EXTENSION: LINGUISTIC READINGS the almost irresistible urge to contrast qualitative and quantitative research goes back to three basic sources of division between the two approaches: (a) an ideological contrast, (b) a contrast in categorization, and (c) a contrast in the perception of individual diversity. Let us look at these contrasts one by one. Ideological differences Although scholars in the social sciences (for example, in sociology) have been using both qualitative-like and quantitative-like data since the beginning of the twentieth century, the qual–quan distinction only emerged after number-based statistical research became dominant in the middle of the twentieth century and some scholars started to challenge this hegemony flying the ‘qualitative’ banner. [. . .] Thus, the terms ‘qualitative’ and ‘quantitative’ were originally introduced as part of, or rather for the purpose of, an ideological confrontation. In a thoughtful analysis, Schwandt (2000) describes qualitative inquiry in general as a ‘reformist movement’, uniting a wide vari- ety of scholars who appear to share very little in common except their general distaste for the mainstream quantitative paradigm. As he writes, qualitative inquiry is a ‘home’ for a wide variety of scholars who often are seriously at odds with one another but who share a general rejection of the blend of scientism, foundationalist epistemology, instrumental reasoning, and the philosophical anthropology of disengagement that has marked ‘mainstream’ social science. (p. 190) Having been created in the spirit of antagonism, we should not be surprised that the two terms are still often used to represent contrasting views about the world around us. Contrasting categorizing/coding practices One thing that is common to every research approach is that the almost limitless infor- mation obtainable from the social world around us needs to be reduced to make it manageable. Researchers typically use ‘categories’ or ‘codes’ to structure and shape this information, but this is where the similarities between qual and quan end. We find that the nature of the categories and the categorization process in qual and quan are very different. In fact, Bazeley (2003: 414) argues that ‘Codes – the way they are generated, what they stand for, and the way they are used – lie at the heart of differ- ences between quantitative and qualitative data and analysis tools’. Quantitative researchers define the variables they work with well in advance and assign a logical scale of values to them, which can be expressed in numbers. Thus, quantitative research can start a research project with precise coding tables for pro- cessing the data (for example, within the ‘gender’ variable, ‘male’ is to be assigned 1 and ‘female’ 2). Qualitative researchers also use coding extensively, but the qual cat- egories are different in two important ways. First, they are not numerical but verbal, amounting to short textual labels. Second, they are usually not determined a priori but are left open and flexible as long as possible to be able to account for the subtle nuances of meaning uncovered during the process of investigation. For example, if we wanted to draw the boundary between two countries in an unknown terrain, the quan approach would be to take the map and after defining the size distribution of the two countries, draw straight lines using a ruler. In contrast, the qual approach Zoltán Dörnyei QUALITATIVE, QUANTITATIVE, AND MIXED METHODS RESEARCH 267 would resist this top-down decision making but would expect the boundaries to nat- urally emerge using the inherent geographical properties of the terrain (for example, rivers and mountain ridges). Different approaches to individual diversity Most data collected in the social sciences, regardless of whether it is qual or quan, is related to people – what they do, what they are like, what they think or believe in, what they plan to do, etc. Because people differ from each other in the way they perceive, interpret, and remember things, their accounts will show considerable variation across individuals. The problem is that no matter how well-funded our research is, we can never examine all the people whose answers would be relevant to our research question, and therefore we have to face the fact that the final picture unfold- ing in our research will always be a function of whom we have selected to obtain our data from. Both qual and quan researchers acknowledge this link between the specific sample of participants examined and the results obtained by the research, but the two camps consider the issue in a very different light. Quantitative researchers regard the sample-related variation as a problem which needs to be fixed. The quan solution is to take a large enough sample in which the idiosyncratic differences associated with the particular individuals are ironed out by the sample size and therefore the pooled results largely reflect the commonalities that exist in the data. Qualitative researchers, on the other hand, question the value of preparing an overall, average description of a larger group of people because in this way we lose the individual stories. They see this as an undesirable reduction process because in qual terms the real meaning lies with individual cases who make up our world. Of course, qualitative researchers are not oblivious to the fact that individuals are different, but rather than believing in a higher-level meaning that can be arrived at by summing up individual cases, they hold that there are multiple meanings to discover. Thus, quantitative researchers follow a ‘meaning in the general’ strategy, whereas qualitative researchers concentrate on an in-depth understanding of the ‘meaning in the particular’. However, the story does not end here because the ‘big number’ approach of quantitative researchers has offered an additional bonus for quan data analysis, statistics. [. . .] Statistics versus researcher sensitivity Once quantitative researchers had gone down the ‘meaning in numbers’ path, a welcome bonus emerged. Mathematicians have found that if we have a sufficiently big sample size, the characteristics of the people in this group will approach a very special pattern termed ‘normal distribution’. This means that within the sample a few people will display very high values, a few others very low ones, with the bulk of the sample centred around the middle or average range. This is the all-important ‘bell-shaped curve’ [. . .], and it has been found that the greater the sample, the more ‘normal’ the distribution and the more regular the curve becomes. [. . .] What makes this bell-shaped curve so important is that it has unique properties upon which it is possible to build a whole range of mathematical procedures that have led to the development of ‘statistics’. Zoltán Dörnyei . conducting English language studies. One point which Dörnyei makes early on when talking in general terms about the social sciences is abso- lutely crucial to your own study of the English language: . researching people, and therefore any study you conduct will always have to be selective. In any English language project that you devise you will need to decide upon the most reliable and valid way. which answers your research questions/examines your chosen area of study. In any study of the English language which you encounter as a reader/listener, whether it be quantitative, qualitative or