Methodological ParadigMs
In order to consider different methodological paradigms that inform English language study, it is useful to step briefly beyond the disciplines of language and consider its place within social sciences and humanities more generally. Methodological trends can be witnessed as belonging to much broader patterns, and it is beneficial to view these trends in the light of the overarching academic arena where language studies belong.
In recent times, there has been an observable move from quantitative to qualitative methods in numerous disciplines across the social sciences and humanities. Sociologist Martyn Hammersley (1992) observes that during the 1940s and 1950s quantitative methodology was the dominant approach to social research, but since the 1960s qualitative methods have gained in popularity. Qualitative research has moved from the margins of many social science disciplines to occupy a far more central place. As qualitative methods have grown in popularity, a debate surrounding the relative values of quantitative and qualitative methods has emerged. This debate between the two paradigms is discussed thoroughly in D12. In this unit, we will focus on characterising and then illustrating the different types of paradigms that are followed in English language studies.
Historically, the basic principles of quantitative methods can be summarised as aiming for the following:
❑ objectivity
❑ neutrality
❑ replivability
❑ generalisation
❑ discovery of laws: in language study, instead of ‘laws’ quantitative researchers use the terms ‘rules’ and ‘norms’ – there are always exceptions to language patterns that can be found, so the term ‘laws’ is not a wholly accurate representation for language researchers
Classic methods that are typically employed by quantitative language researchers include standardised questionnaires, standardised interviews, where exactly the same set of questions are asked to numerous informants in exactly the same manner, and experimental settings where laboratory conditions are used to elicit spoken or written language data.
In contrast, qualitative research focuses on observing the social world as naturally as possible. From this perspective, the use of any artificial settings, be it through experiments, questionnaires or artificial settings such as interviews, is arguably flawed as researchers instead need to study ‘the social world in its natural state’ (Hammersley and Atkinson 1995: 6).
The main principles for qualitative methodology are often seen as being based upon a critique of quantitative methodology, in particular, upon survey and experimental methods. The validity of quantitative methods has been challenged by qualitative researchers. For example, any claims to objectivity and neutrality are fundamentally flawed, as quantitative researchers’ assumptions and perspectives on the social world will be imposed by the structured nature of the data collection, despite any claims to the contrary. This biases research, as it makes it very difficult to discover evidence that does not correspond with these assumptions and perspectives
To illustrate this point, Alvesson and Deetz (2000: 69) argue that questionnaires entail that informants have no choice but to ‘subordinate themselves to the expressions of the researcher’s subjectivity’, thus making it impossible for alternatives to the researcher’s position to be explored. Other criticisms include the following:
❑ Naturally occurring behaviour cannot be analysed by setting up artificial contexts such as experiments or standardised interviews. Therefore, evidence gained in these artificial settings cannot be used to make valid claims about what takes place in naturally occurring situations.
❑ You cannot rely on accounts of what people say they do without observing what they actually do in naturally occurring settings, as this fails to acknowledge that there can be a discrepancy between informants’ attitudes and their actual behaviour in specific social situations. Informants therefore need to be observed in their natural environments. If you as a researcher directly observe naturally occurring events, then you are no longer solely dependent on the respondent only.
❑ Quantitative research can imply that the aspects of language usage and social identities are distinctive and fixed. This may draw attention away from the fact that informants’ language usage will both change and develop depending upon context.
❑ Quantitative researchers tend to treat informants’ behaviour as something which is mechanically produced, thus arguably neglecting individual creativity and cognition. Therefore, the quantitative approach can appear to have a rather static perception of human interaction.
❑ As a social researcher, you are part of the social world you are investigating, and this factor cannot be ignored. Researchers should do everything in their power to reduce bias and the influence of idiosyncrasies in research, but appealing to objectivity or neutrality is problematic because they are both impossible to achieve. Researchers can never escape from their own ideologies as researchers in the social world or the interrelated fact that theoretical presuppositions are involved in all data (see A13–D13).
Such criticisms of quantitative methods reveal the critical principle behind qualitative methods, namely, that the ‘nature of the social world must be discovered’ (Hammersley 1992: 12, emphasis in original). This can be achieved by the method of participant observation in order to produce detailed descriptions, sometimes referred to as ‘thick’ descriptions. Participant observation is generally defined as part of ethnography , where researchers physically join in and participate in the social world which they wish to study, so that they can observe it from an insider’s perspective.
However, despite historical differences and disagreements between quantitative and qualitative research, more recently social scientists have begun to question the purpose of the arguments between the two paradigms, suggesting that the debate itself is a fruitless exercise, which has resulted in detracting attention away from more important issues of theory and methodology.
We would agree that the dichotomies which exist between qualitative and quantitative approaches, and any dichotomies that exist in social science disciplines in general, are unhelpful and limiting, often resulting in researchers situating themselves in ‘armed camps’ (Silverman 2000: 11), unwilling to learn from each other. Researchers should instead be seeking common themes between different social science traditions in an effort to move disciplines forward by sharing ideas and expertise. If researchers stop viewing concepts as being in opposition to one another, then more integrated and arguably more sophisticated mixed-methodologies can be produced.
Furthermore, contrasting different approaches has led to the assumption that quantitative and qualitative methodologies are themselves harmonious and unified. T his is certainly not the case with either paradigm, both of which include many diverse approaches. On some occasions, qualitative methods may be deemed more appropriate by a researcher, but on other occasions, the very same researcher may deem quantitative methods to be more appropriate. In some circumstances, a mixture of the two methods may be used, depending upon the problem or the area of investigation which is at hand. However, this pragmatic, mixed-methods approach is not without its problems. We will explore this further in D12.
Researchers from all paradigms can improve the reliability and validity of their methodology by acknowledging clearly that they are actively involved in the social world in which they are studying. Hammersley and Atkinson (1995: 16) argue for a commitment to producing what is termed reflexive research, which is defined as the means by which researchers directly acknowledge that their orientations ‘will be shaped by their socio-historical locations, including the values and interests that these locations confer upon them’. This represents a rejection of the idea that social research can be carried out in isolation from the social world in which the researcher is studying and also an acknowledgement that the researcher as an individual in the social world will influence the research project.
Deciding upon which methodology or methodologies to use is a crucial process, which should be thought through carefully, even if you are only conducting a small project. To illustrate these different paradigms, we will now present an example of well-regarded English language studies for each of the different areas, including mixed-methodologies, to show the different approaches in practice.
an illustration: quantitative methods
Sociolinguist William Labov (1966) used an innovative and effective research method for his seminal quantitative study of New York City department stores, which is now commonly referred to as the rapid and anonymous method basically because it enabled him to compile sociolinguistic data from a number of different individuals very quickly, without needing to get any personal details from those who took part in his study.
Labov wished to analyse a particular phonological feature as articulated by shop assistants in three different department stores: Saks, Macy’s and S. Klein. Labov selected the three stores based upon perceived differences in the targeted socio-economic class groupings of shoppers. He was interested in investigating the popular and highly useful phonological variable of rhoticity, which, as we have already highlighted in A9, is a prestigious speech variant in New York.
He asked many different shop assistants a question about the location of goods in the store, where he already knew the answer to be ‘fourth floor’. He pretended not to hear the first time around and got the assistants to repeat their answers. In total, then, Labov ended up with four different examples of /r/ pronunciation per individual informant. He had examples where /r/ appears both in the middle of a word, known as word-medial positioning (‘fourth’), and also at the end of a word, known as wordf inal positioning (‘floor’). The phrase ‘fourth floor’ was ideal for presenting Labov with a number of different realisations of the rhotic pronunciation.
an illustration: qualitative methods
Qualitative or ‘thick’ descriptions can be basically defined as collecting anything and everything that you can when you are present as a participant-observer in a research setting, including audio or video recordings, written documents, field notes containing additional background details or any details that cannot be formally recorded for ethical reasons (see B12). The most productive way of gaining a thick description is commonly thought to be ethnography involving participant observation.
Health communication researcher Srikant Sarangi has conducted a number of studies in the medical profession where he has conducted ethnographic research and become a participant observer. He has then drawn upon his ‘thick’ descriptions to produce thorough, discourse-based analyses of medical interactions, including work on genetic counselling (see Roberts and Sarangi 2003). The participant observation method has the advantage of enabling researchers to pursue research questions which can be continually negotiated with those being researched, with the intention that the f indings can be of practical relevance to those being studied.
an illustration: mixed-methods
T he following is a quotation from Penny Eckert’s lengthy ethnographic participant observation in a Detroit high school called Belten High. In total, she spent over two years collecting data at this location. Here we have an example of linguistic data which is embedded within a rich ethnographic description of what is going on around the conversation. It gives rich detail to the positioning of the body as well as language – the advantages of which we discuss in D9 in reference to Eckert’s notion of communities of practice (see also Wenger 1998). The passage shows the importance of spatial features of context – how standing in specific locations is linked with the performance of identities alongside linguistic features:
At lunchtime in the spring of 1997, in an ethnically very heterogeneous junior high school in northern California, a crowd of Asian-American kids hangs out in a spot that is generally known in the school as ‘Asian Wall’. Girls stand around in their high platform shoes, skinny bell-bottoms, and very small T-shirts, with hips cocked. As they toss their heads, their long sleek black hair (in some cases tinted brown) swishes across their waists, the slimness of which is emphasized by shiny belts. Some of them talk to, some lean on, quiet-demeanored boys with baggy jeans and baggy shirts, with hair long on the top and shaved at the bottom. Linda turns away from her group of friends with a characteristic tilted head toss, bringing her hair around her shoulders; and with an exaggerated high-rise intonation on the pronoun, she calls to a boy who’s standing nearby. ‘What are you ?’ Another girl, Adrienne, who happens to be walking by, answers on his behalf. ‘He’s Japanese-Filipino.’ The boy smiles silently, and Linda turns back to her friends.
Eckert and McConnell-Ginet (1999: 185–6)
In addition to these ethnographic observations, Eckert also carries out very detailed quantitative analyses from recorded data of phonological and grammatical linguistic variables from a large number of realisations of phonetic and grammatical features over time. These quantitative studies are statistically tested and enable her to make valid and reliable arguments from her findings regarding language, regional identities, social identities and adolescence.
Eckert (2009) contrasts the ethnographic approach which she took at Belten High with what she describes as a ‘quick and dirty’ method of briefly going into various other schools in the surrounding area to examine the phonological manifestations of her adolescent identity categories. This method, which shares similarities with Labov’s ‘rapid and anonymous’ method described above, enabled Eckert to place her phonological analysis on a broad geographical continuum for the whole of the suburban region where her research took place.
In a reflexive evaluation of her own methodological practice, she notes that the ‘quick and dirty’ method lacked the textual detail and access to communities of practice which was so crucial in being able to characterise how adolescent groups use linguistic strategies to signal their identities. She concludes: ‘I was never as sure of the status of my data in the other schools as I was at Belten’ (2009: 150).
Once the decision regarding methodological paradigm(s) for conducting a particular English language study has been made, there are other decisions regarding data collection techniques and ethics that need to be taken. We will explore some of the most significant of these issues in B12.
A13 Language theories
Almost everything we have said about language in the first twelve strands in this book has been contentious at different points in history. The ways that people described language 300 years ago (see B8 and C8) were very different from more recent descriptions. Even observations from only 30 years ago which gradually became the general consensus (the paradigm) have gradually been questioned, rejected and replaced by new perspectives. No doubt our current thinking will also soon be superseded in due course.
This does not mean that linguists are generally incompetent: the fact that the f ield of study is constantly being revised is a consequence of the fact that language in general is immensely complicated and difficult to study. It ranges from material and measurable phenomena (writing, sounds) all the way to things that are very difficult to get at (meanings, implications). The close connections between language, thought and consciousness also take you very quickly to difficult philosophical considerations. Language covers the things that an individual says and writes, as well as the variations across large groups of people over geographical distances and throughout history. Language itself is a universal capability of all humans, and the particular languages that societies use have always changed over time.
Furthermore, because there are so many different dimensions to language, it is impossible to hold them all in equal and perfectly objective balance at the moment of study – every researcher comes to the study of language with ideas, ideologies, commitments and perspectives that influence which part of language they choose to study, which questions they decide to ask, which methods they decide to employ, and how their findings can best be interpreted from their own angle.
l inguistics as a social science
T he objective of language study – like any properly progressive discipline – is to gain a better understanding than the state of existing knowledge. The best way of ensuring this development is to be healthily sceptical of all ideas and ask questions to determine the basis of those ideas. Being critical does not simply mean finding fault and rejecting others’ work: there is an ethical imperative on you also to be creative and suggest solutions where you find problems. The general scientific principle – even for a complex social science like linguistics – that underlies all this is the notion of falsifiability . Any statement you can make about an aspect of language must be able to be disproven. Consider, for example, the following two statements:
Language is the blood of the soul into which thoughts run and out of which they grow
(poet Oliver Wendell Holmes
Language is always changing
(linguist David Crystal)
Only the second of these can be proven false – for example, you could try to find a living language somewhere in the world that has remained exactly the same for at least three generations. You would only need to find one contrary example to falsify Crystal’s general statement – or at least force him to amend the phrase ‘always changing’ to something weaker, like ‘tends to change in general’. The first statement is certainly poetic and certainly meaningful, but it is not possible to disprove it, because as a metaphor it is already literally false, and the range of meanings it might evoke in different interpretations means that it does not say anything precisely. In short, the first sentence is a poetic and expressive statement of an opinion and viewpoint that is untestable, and the second is a scientific statement that is testable and disprovable.
Falsifiability must apply to all assertions in language study, from simple statements like the ones above to entire theoretical frameworks such as the way in which we carved up the constituents of clauses in A4 and B4. All statements about language can then be regarded as theories offered as descriptions of language available to be falsified. They cannot be proven true, since you can never say what new evidence, new techniques or new approaches might emerge in the future to disprove the theory, so the best we can do is to hold a set of theories about language which are as good as we can manage for the moment. This means that in language study, as in any analytical study, everything in this (or any other) book is best thought of as being only provisionally not wrong.
different approaches to language study
Of course, when a theory receives criticism, or a new perspective is suggested, not everyone always agrees that the original theory is disproven. Some adherents to the original theory will defend it, arguing against the criticism. Others will adapt the original theory to meet the challenge offered by the new proposal. And, of course, some researchers will have their positions changed by the new debate and will work to develop other theories too. This process means that the historical development of modern linguistics has not been a neat march from one descriptive position to another, with everyone in perfect step. If you take a snapshot at virtually any historical moment in linguistics, you will find adherents of several contradictory and complementary theories existing at once and debating with each other. On occasion, this has become so passionate that one spat in the 1970s was even termed ‘the linguistics wars’.
For example, the study of language through most of recorded history has been through rhetoric (learning the art of speaking and persuading) and through prescriptive treatises on correct grammar and good usage. Through the later nineteenth century, philology was the name for language study, with a strong focus on the historical development of languages, the etymology of words, sound-change rules in pronunciation, and literary history.
Around the turn of the twentieth century, linguists began to be interested not so much in these diachronic developments but in more synchronic patterns, looking at a snapshot of the language of the present as a system of structures and symbols. This was structuralism, which dominated language study for the first half of the twentieth century. Structuralism shifted the emphasis of study away from historical processes and onto social ones, which meant that links between linguistics and anthropology were easier to make.
In the 1950s, a revolution in language study occurred in the name of
generativism , which created the discipline of modern linguistics. In some respects, generativism continued the structuralist approach of aiming to discern underlying systematic patterns and structures in language, rather than being interested in language history or social variation. Indeed, generativists explicitly claimed to be interested only in a speaker’s underlying
competence rather than their observed
performance , though of course their intuitions about well-formedness constituted important information that shed light on deeper structures. In other respects, generativism marked a break with structuralism in the disregard of social, performative aspects of language. Generativists believed that the capacity for language – and even many of its deepest structural patterns – was innate in humans.
Comments
Post a Comment