Approaches to Discourse in Dementia

Jacqueline A. Guendouzi, Nicole Muller

• Publisher: Psychology Press
• Year: 2006
• ISBN:
  9781135623432
• Edition: 1
• Topic: Didattica
• Subtopic: Didattica inclusiva

The qualitative analysis of naturally occurring discourse in neurogenic communication disorders, specifically in dementia studies, has experienced recent burgeoning interest from wide-ranging disciplines. This multidisciplinarity has been exciting, but has added contextual confusion. Thisbook advances the study of discourse in dementia by systematically exploring and applying different approaches to the same free conversational data sets, collected and transcribed by the authors. The applied methodologies and theories comprise a useful sourcebook for students, researchers, and practitioners alike.

1 Dementia and Its Discourses

THIS BOOK AND ITS CONTEXT

In recent years, there have been steady advances in the neurology, neutrophils, and neuropathology of dementing diseases and in experimental studies probing their communicative and cognitive consequences. There is also a growing body of work, in dementias and other neurogenic communicative disorders (such as aphasia following focal brain injury or traumatic brain injury), that uses qualitative methods to analyze naturally occurring interactions between persons with a communicative disorder and persons without. The focus of our own work, and of this book, is this comparatively young, although by now well-established, tradition of qualitative research in communicative disorders. Our approach to the analysis of interaction or discourse brings with it certain assumptions and practices, which we revisit repeatedly and in some detail as we progress through the book:

1. Our focus on communication disorder could be described as contextualized and emergent. The object of analysis is the interaction between two or more human beings as it unfolds. Order or disorder, success or breakdown in communication, arise out of the unfolding interaction in its context. Communicative interaction is a joint, collabo-rative, distributed, and contextually situated activity. Disorder, in this perspective, is not primarily concerned with the limitations of an individual person’s communicative faculties but with the mutually perceived success or lack thereof of a communicative interaction. This also implies that communicative disorder, for our purposes, needs to be described, analyzed, and understood within the realm of real communicative events: the prototypical such event being conversation.¹
2. Our attempts to analyze naturally occurring interaction are interdisciplinary in nature. Researchers and teachers in clinical communication studies have some scavenger-like characteristics: Promising-looking methods from linguistics, pragmatics, anthropology, sociology, various branches of psychology, and more recently, complexity theory and network theory are ruthlessly adopted, adapted, and applied. As we discuss further later on in this chapter and throughout this book, this interdisciplinary approach holds both a great promise and some danger. On one hand, having multiple descriptive frameworks and analytic methods available makes for a rich tool kit, and thus, it should in theory become easier to capture at least some of the complexity of human interaction. On the other hand, if frameworks and methods are divorced from their philosophical backgrounds, understood imperfectly, and applied inappropriately, the resulting picture will be less than illuminating (see Perkins, 2000, 2002). There is of course some degree of overlap between the various approaches, methods, and theories that make up the umbrella category of discourse analysis (DA). However, one must not assume that the differences are superficial and consist only of different sets of labels for the same phenomena. Rather, terminologies and categories bring with them the legacies of different methodological preferences, theories, and philosophical orientations. These legacies determine how different schools of discourse analysis approach (and indeed construct) their data, which in turn will determine which questions can usefully be answered by any given approach and which conclusions can sensibly be drawn.

In writing this book, we have pursued three main goals. First of all, we wished to provide readers with an interest in the analysis of spoken discourse with an overview of the main approaches and theories in DA. Second, we aimed to show how these approaches and theories can be applied to data from clinical contexts by offering detailed analyses of data extracts from conversations with persons with dementia. Third, in presenting these analyses, we wished to illustrate some of the characteristics and challenges inherent in interaction in the context of dementia. The title, and to a lesser extent the structure of this book, are a deliberate echo of Schiffrin’s (1994) admirable work, Approaches to Discourse, which over the years has been an invaluable resource to many researchers, students, and teachers in discourse studies. In the remainder of this chapter, we discuss notions of discourse, discourses, and interaction before moving on to a brief introduction to dementia and communication in and with dementia.

DISCOURSE IN DEMENTIA STUDIES

The terms discourse and discourses mean many different things to different people. It is not our intention here to survey the many existing definitions; for such a survey, the reader is directed to, for example, the opening chapter of Jaworski and Coupland (1999).² Schiffrin, Tannen, and Hamilton (2001) distil three major definitional strands from the literature, namely, “(1) anything beyond the sentence, (2) language use, and (3) a broader range of social practice that includes nonlinguistic and nonspecific instances of language” (p. 1). The definition of discourse used in any analysis of language or language use, whether spoken or written, depends of course on the theoretical and analytical framework adopted. Schiffrin (1994) draws a basic distinction between formalist and functionalist approaches to the analysis of discourse. Formalist, or structural, frameworks focus on the relations between constituent elements of discourse “(1)” as defined by Schiffrin et al. (2001) (see Grimes, 1975; Stubbs, 1983; van Dijk, 1985a, 1985b, 1997), whereas functional approaches analyze language in use including the purposes and functions of linguistic form in human interaction: in other words, Schiffrin et al’s discourse “(2)” and/or “(3)” (see Brown & Yule, 1983; Fasold, 1990). Discourse “(3)” underlies critical discourse studies, which presuppose language and its use as a part of society and reciprocally, society as constructed through linguistic and nonlinguistic (discursive) practices (see Fairclough, 1989).

The field of clinical discourse studies is truly multidisciplinary in nature, but typically the frameworks used are adapted from nonclinical contexts to clinical ones. Each theoretical bias, or approach, brings with it its own framework for description, categorization, and analysis and crucially, its own understanding of the nature of its data. Thus, what makes up discourse differs as widely in clinical as in nonclinical discourse studies. The explicit or implicit definitions of discourse in published work on dementia are on a continuum that spans elicited spoken or written discourse in experimental settings on one end and free conversation or narrative on the other end. This continuum partly overlaps with an analysis continuum in which quantification of predetermined discourse characteristics occupies one end and qualitative methods attempting to describe, rather than count patterns, the other. It comes as no surprise that elicited discourse in experimental set-tings tends to team up with quantitative analyses, whereas qualitative research tends to insist on naturally occurring language data.

At the experimental end of the continuum, there are studies on discourse data that fall squarely under Schiffrin et al.’s (2001) definition “(1),” and the classic studies carried out and reviewed by Ulatowska and colleagues (see, e.g., Ulatowska, 1985; Ulatowska &Chapman, 1995) may serve as useful illustrations and recommended reading here. Ulatowska and Chapman’s (1995) definition of discourse and their view on the use of discourse data in dementia studies, is worth quoting here:

Although discourse may be composed of a single word, a phrase, a sentence, or a combination of all the above, discourse typically consists of a sequence of connected sentences. The coherence of discourse is determined by how well this sequence of sentences is related. Discourse provides a promising avenue for investigating behavioral changes in dementia, because discourse can be manipulated to explore how cognitive impairment impinges on linguistic functioning and communicative competence. This approach is possible because of the inherent nature of discourse, which entails a complex interaction among linguistic, communicative, and cognitive processes. Precisely, discourse is expressed linguistically (i.e. via words and sentences) and is defined communicatively (i.e., a unit of language that conveys a message). Moreover, complex cognitive processes (e.g. memory, attention, perception, and retrieval) underlie discourse comprehension and production. (p. 115)

Studies in this tradition typically use elicitation procedures for connected spoken and written language in controlled experimental settings. In their review of discourse studies in dementia, Ulatowska and Chapman (1995, p. 116ff) list the use of oral and written picture description tasks using various stimuli, orally presented narratives for retelling, delayed story retelling, self-generated narrative, the elicitation of summaries and morals for stories, and interview dialogue. Students and researchers working in, for example, the assessment of aphasia or brain injury will be familiar with various assessment materials and techniques that use these tools, and there is no need to go into detail here discussing them.

Among Bayles and Kaszniak’s (1987) “recommendations for evaluating the communication deficits of dementia patients” (p. 174), we find the following important passage:

Analyze Communication Beyond the Sentence Level. The third basic principle in testing the dementia patient is to include an analysis of a unit of language larger than the sentence, namely, discourse. Discourse is the most natural and common type of communication and requires the integration of all types of linguistic knowledge. It is a medium through which the social, psychologic, and linguistic aspects of communication can be studied. With discourse analysis, the clinician can quantify the emptiness of language associated with deme the fragmentation of thought, and anomia. (pp. 174–175)

Again, discourse is essentially understood as language structure beyond the sentence level, discourse is elicited in a testing situation, and the analysis of discourse is primarily seen as a means of quantifying deficits associated with disorder. One may of course question the naturalness of language produced through a picture elicitation task or in a story retelling, but the research questions we pursue here are explicitly not questions of naturalistic communicative functioning but questions of underlying competences and knowledge and their impairment through the effects of dementing conditions.

A middle ground between experimental-quantitative and naturalistic- qualitative studies is occupied by work that uses more naturally occurring language but analyzes the data thus gathered by means of predetermined categories in an attempt to quantify the characteristics of discourse produced by persons with dementia (see, e.g., Garcia & Joanette, 1994; Hamilton, 1994b).

At the end of the clinical discourse studies continuum opposite the experimental- quantitative end, we find work that is explicit in using naturally occurring, interactive language, often in the form of conversations or self-generated narratives. In this tradition falls the groundbreaking work by Hamilton, Ramanathan, and Sabat (e.g., Hamilton, 1994a; Ramanathan, 1997; Sabat, 2001) in Alzheimer studies. Hamilton (1994a, p. 18) discussed an important point of test situations in which participants are asked to produce “spontaneous” language on a specific, very narrow topic: The participants may consider the task trivial (or maybe even demeaning?) and may not see the need to communicate about, for example, a button, an envelope, a nail, or amarble placed in front of them (as in, e.g., Bayles, 1982). As Hamilton (1994a) puts it, “There must be a need for the patient to communicate before we can be sure that our statements regarding the patient’s ability to communicate are valid” (p. 19). In fact, we could say that in testing situations, we have at aminimum two levels of discourse happening at the same time: The actual “experimental” utterances, that is, instructions given by the tester, and responses by the test participant, which are embedded into the surrounding test situation, a discourse including all nonverbal, situational, and so forth factors as well as all utterances not directly pertaining to the experiment. That is, we have an instance of discourse “(1),” as defined by Schiffrin et al. (2001), embedded into an instance of discourse “(2)”/“(3).”

Hamilton’s (1994a, 1994b) work is strongly data driven; her data consist of naturally occurring conversations with a single participant. Hamilton (1994a) herself adopted the role of the participant observer, using a “personal research approach which is meant to supplement the technical approach normally taken in studies of Alzheimer’s disease” (p. 31). The term “personal research approach” is from Kitwood (1988); the approach contrasts with that of the objective, depersonalized observer of traditional ex-perimental science in that “at its core, it works interpretively and empathetically, going far beyond the measurement of indices or the codification of behaviour. In all of this the researcher takes a personal risk ” (Kitwood 1988, p. 176, as cited in Hamilton, 1994a, p. 31.).Acrucial point in Hamilton’s (1994) work is that the analytical tools and categories emerge out of the data rather than being given a priori, in the form of an analytical framework superimposed on the data gathering and analysis processes. Hamilton describes her approach thus:

Sense-making difficulties and unusual moments in the conversations would pique my linguistic curiosity along the way and lead me to wonder about possible interrelationships in the data. But it was not until after the final conversation examined in this study that I began to use these observations and considerations to form the analytical tools and frameworks which would allow me to carry out quantitative analyses which seemed to be true to the data. (pp. 30–31)

Ramanathan (1997), in discussing Hamilton (1994) and Sabat (1991), draws attention to the jointly distributed character of face-to-face interaction and the changing role of the interlocutor over time as dementia progresses. Thus, here, the interviewer or conversational partner is seen as a facilitator, accommodating to the partner’s interactive needs (see also Sabat, 1991). Ramanathan uses life narratives rather than interviews or conversations as primary data, although the narratives were collected in face-to-face interactions, thus incorporating an interactive element. Indeed, the interactional processes or schemas are a major focus of analysis in a work that “examines how particular interactional features inhibit extensive and meaningful talk from the patient, whereas others facilitate it.… A close examination of language as a social product and process can be another fruitful approach to understanding the Alzheimer experience” (Ramanathan, 1997, p. 6).

It is not our intention here to make a case that discourse elicited under experimental conditions represents “worse” (or indeed “better”) data than conversations or narratives collected under circumstances that are as “natural” and spontaneous as possible. Rather,we wish to point out that we are dealing with different types of data, and approaches to analysis, in each case, and that ideally, experimental-quantitative and naturalistic-qualitative studies should complement each other. A brief mention of Sabat’s (2001, p. 3ff) discussion of classical and romantic science, following Luria (1987a), may be useful at this point. Researchers working within the tradition of reductionist classical science in experimental settings “look upon events in terms of their constituent parts. Step by step they single out important units and elements until they can formulate abstract, general laws.” (Luria 1987a, as cited in Sabat, 2001, p. 3).

As Sabat (2001, p. 3ff) points out, the focus on generalization is crucial here. Classical science aims to distil the general, the average, out of typi-cally large amounts of data gained in experimental settings and subjected to various processes of quantification. Equally important in classical science is the focus on the reduction of complex organisms, functioning in a complex environment, to their constituent parts, which in turn are examined separately and the results expressed in terms of test scores. Hence, the use of standardized tests and controlled settings, elicitation procedures for discourse that are kept as constant as humanly possible across participants, and the quantification of discourse characteristics as in the work discussed in, for example, Ulatowska and Chapman (1995), or Bayles and Kaszniak (1987). Aggregates from these separate examinations are combined into averages and compared with average scores from other groups (in dementia research, typically normally aging participants) who have undergone the same tests. Thus, an individual’s functioning is not only compartmentalized but grouped with and matched against the compartmentalized functioning of others. The aim is to identify the general, typical bundle of symptoms of dementing conditions, such as Alzheimer’s disease (AD) as evidenced in certain types of language use, in comparison to persons without these conditions. This is of course a necessity in the search for the abstract, general laws of classical science, and in research efforts aimed at the improvement of clinical diagnosis of AD and other dementing diseases, which require the establishment of a reliable, recognizable, and at least to an extent generalizable symptom complex, this approach is of great value and indeed necessary (Sabat, 2001, p. 4).

Our concern in this book is not, however, the clinical diagnosis of dementing conditions, nor, in the first instance, their pathological symptom complexes. The diagnosis of, for example, AD as experienced by persons withADand their carers, and certainly the presence of AD and its effects on human interaction, are of course part of the universe in which we collected our data and carried out our analysis. However, our starting points and primary concerns in this book are not with the quantification of symptoms or typical traits. In terms of data, we begin with naturally occurring conversations or as natural as conversations can be while they are being recorded on video or audio tapes. In terms of analysis, our aim is to examine various approaches to language use or discourse. One of our concerns here is a meta-analysis, in that we aim to investigate what questions can sensibly be asked from which approach, and what we can learn about our assumptions concerning language, language use, and linguistic interaction by applying methods developed on the implicit basis of “normal” discourse practices to “disordered” data. In terms of dementia, persons with dementia, and their carers, our focus of investigation lies with individuals functioning and interacting more or less successfully within the complexity of their environments. Thus, our ideological home, if one may use the term, is closer to what Luria (1987b) calls “romantic science,” than to classical, reductionist science. In Luria’s words (1987b, as cited in Sabat, 2001), “Romantics in science want neither to split living reality into its elementary components nor to represent the wealth of life’s concrete events in abstract models that lose the properties of the phenomena themselves. It is of utmost importance to romantics to preserve the wealth of living reality, and they aspire to a science that retains this richness” (p. 14).

Principles of Qualitative Research and Discourse as a Metacategory

At this stage, we need to ask ourselves whether it is really possible to achieve Luria’s romantic (1987b) science and to genuinely preserve the “wealth of living reality.”

Rempusheski (1999) draws a distinction between realists and idealists in scientific inquiry, defining the realist as conducting research in the tradition of logical positivism, following a “deductive approach and quantitative research methods wherein phenomena are viewed objectively and reduced to numerical configurations” (p. S45). The idealist on the other hand “explores a world known through perceptions and the subjectivity of humans,” in the tradition of “postpositivist philosophy of science,” which “follows an inductive approach and qualitative research methods wherein the context is maintained and the subjectivity is acknowledged” (Rempusheski, 1999, p. S45).

Qualitative approaches to inquiry have found increasing numbers of followers in neurogenic disorders of communication in recent years. The philosophy and objectives of qualitative research make it a paradigm (or maybe rather a bundle of paradigms) that is attractive in areas in which the researcher is not or cannot be certain from the very beginning exactly wha...