Editorial: En‘gaze’ment with text and talk

1 Some celebratory preliminaries

2016 has now eclipsed and I belatedly realise it has been 10 years since the journal changed its title from TEXT to TEXT & TALK in 2006. The title TEXT dates back to 1981 when Teun van Dijk and Janos S. Petöfi co-founded the journal and then van Dijk steered it expertly under his editorship. The transition occurred in 1998 when I assumed co-editorship with John Wilson and then became the sole editor from 2002.

In the spirit of performativity, the new title of TEXT & TALK was announced in 2006, accompanied by an editorial titled “Post-Text”, with the following remarks (Sarangi 2006):

The title of this editorial, “Post-Text”, is meant to capture a pervasive sense of transition. Curiously, the prefix “post” makes one mindful of the past, while signalling a determination to break away from this very past by looking for a change-charged future. It is perhaps in the nature of things to anticipate future events, and when that future engulfs the present, one begins to feel nostalgic about the familiar past.

[ … ]

I also envisage a future for the journal which will be a continuation of TEXT’s scholarly legacy. No doubt there will emerge a respecification of the journal’s remit, especially with the foregrounding of TALK in the new title. I very much hope that the journal under the new title will cross a range of disciplinary boundaries and generate a wider authorship and readership.

The crossing of disciplines brings with it a burden of ‘communicative responsibility’: How to make one’s writing (and talking) intelligible to others who do not necessarily sing from the same hymn book, although they may cohabit the same intellectual space. The new subtitle of TEXT & TALK already embodies three key disciplinary fields: ‘Language, Discourse and Communication Studies’. There are many more ‘ratified’ authors and addressees that TEXT & TALK would like to reach in disciplines such as anthropology, psychology, sociology, literature, history, education, organization studies, cultural studies, media studies, and so on. Increasingly, there will be the need for interdisciplinary dialogues and I will heartily afford journal space for such scholarly exchanges. The same is true about timely state-of-the-art position papers – covering both theoretical and methodological concerns – as a way of consolidating our cumulative knowledge of text/talk-based studies.

On the occasion of the 10^th anniversary of the new title, I wish to renew my vows and expand the currently dominant monolithic format of submissions to TEXT & TALK. In the future the following categories of submission will be considered:

While individual book reviews still fall outside the scope of the journal, review articles which bring together key book-length publications that are thematically organised and critically commented upon will be welcome. Special Issues on nominated topics will continue to be published.

The categories ‘Research Notes’ and ‘Systematic Literature Reviews’ deserve brief commentaries.

Research Notes are particularly suited for junior researchers who are currently undertaking doctoral studies or have assumed the status of post-doctoral researcher. It affords them the opportunity to share their preliminary research findings by writing up a shorter paper. This category may also appeal to teams of researchers working on funded projects and have preliminary findings to report. Research Notes are processed in a non-threatening way via a light-touch peer review system that is geared towards offering critical as well as constructive feedback. Generally speaking, Research Notes can serve as placeholders for ongoing research. Through sharing preliminary findings with the journal’s readership, there is the distinct possibility of follow-up fuller articles finding their way to the journal at a later date.

Systematic Literature Reviews have become a highly acclaimed genre in many natural and biomedical sciences. They are published in high profile journals on a regular basis as original contributions in their own right. The intrinsic value of systematic reviews of literature is less recognized in the disciplines of humanities and social sciences. There are many different formats within this genre as well as established protocols. An exemplar is as follows:

1. Designing a review protocol

2. Formulating a question

3. Identifying and selecting relevant studies

4. Systematic data extraction and collection

5. Synthesis and analysis of the data

6. Writing up and reporting systematic reviews

Systematic reviews, unlike opportunistic ones, help to identify genuine gaps in the field while providing a basis for benchmarking new research contributions in a given sub-field. The review protocols pay attention not only to findings and claims but also to methodological considerations and analytical apparatuses used by different researchers.

The current year, 2017, marks the onset of the 20^th year of my editorship of TEXT and TEXT & TALK – counting from 1998, the year of transition. Much more has been accomplished in the past 20 years than a change of title in 2006. In 2005 TEXT celebrated its silver jubilee anniversary. In 2006, coinciding with the title change, TEXT & TALK grew from 4 issues per year (600 pages) to 6 issues per year (800 pages), thus allowing more manuscripts to be published in a volume annually. The editorial board has changed regularly to accommodate new interests, while being responsive to shifts in trends in submissions. Over the years the TEXT & TALK editorial board has experienced the loss of key figures like John Gumperz, Dell Hymes, Janos Petöfi and, more recently, Chris Candlin, Geoff Leech and Geoff Thompson.

Over the years the journal has steadfastly maintained its interdisciplinary and international profile. Typically an annual volume carries 35 articles. In 2015 authors from as many as 17 countries were published and the readership spanned in excess of 20 countries. The aim is to reach further afield and as part of the current campaign, a selected number of articles from TEXT & TALK is offered as free downloads (see http://www.degruyter.com/view/j/text for details).

2 A brief tour across time: The many tribes and sub-tribes of text and talk analysis

TEXT & TALK continues to be characterised by its broad church orientation to language, discourse and communication studies with regard to methodology (theories and practices surrounding data) and analysis (theories and practices surrounding interpretation). The journal has been and will continue to remain inclusive in its overall methodological/analytical remit and publication strategy.

Here I cursorily document how text and talk analytic practices have evolved over the last six or so decades (see also Sarangi 2010). Since the 1950s when discourse analysis emerged as a disciplinary activity within Linguistics, it has branched out in many directions, even beyond language itself, by embracing other modalities as potential objects of study. Although Harris (1952) introduced the label ‘discourse analysis’ to emphasise the need for an upward extension of linguistics, it was not until Mitchell (1957) that we encountered discourse analysis grounded in real-life data. Mitchell started with Firth’s (1935) notion of ‘context of situation’ with the conviction that ‘meaning must be sought in use’. As early as 1935, Firth anticipated how conversation (i. e., parole in Saussure’s (1966 [1916]) terminology) would become central to linguistic inquiry:

Conversation is much more of a roughly prescribed ritual than most people think. Once someone speaks to you, you are in a relatively determined context and you are not free just to say what you please. We are born individuals. But to satisfy our needs we have to become social persons […] it is [in] the study of conversation […] that we shall find the key to a better understanding of what language really is and how it works.

(Firth 1935, cited in Stubbs 1993: 19)

The study of conversation, following Firth, is not only a means to appreciate ritualised conversational mechanisms with attendant constraints, but also to enhance our knowledge of language. Hymes’ (1964) model of ethnography of speaking was to extend this viewpoint. Later, in advancing the notion of ‘communicative competence’, Hymes (1972: 278) maintained that “there are rules of use without which the rules of grammar would be useless” and observed:

For that matter, a person who chooses occasions and sentences suitably, but is master only of fully grammatical sentences, is at best a bit odd. Some occasions call for being appropriately ungrammatical.

(Hymes 1972: 277).

According to Sherzer (1987: 296), ‘discourse is a level or component of language use, related to but distinct from grammar’.

Roman Jakobson’s (1960) model of communication with a nuanced notion of context, compared with the mathematical model of encoding and decoding of information as proposed by Shannon and Weaver (1949), has provided a useful frame of reference for language/communication studies. An analytical framework labelled ‘context analysis’ (Scheflen 1966), which is buried in the past, had the following basic assumptions:

1. That communication (for a given group or culture) has a regular, ultimately predictable structure consisting of definite units arranged in particular ways.

2. That the structural units of communication have a regular set of components and organisation which must be performed the same way (within a given range) or else communication will not occur.

3. That these structural units, consisting of communicational behaviours, are themselves merely components of larger units in a hierarchy of levels.

4. That institutionalised activities like psychotherapy use the basic communicational system of the culture modified in certain specialised but lawful ways.

5. That all participants have learned and know (mostly without consciousness) the system of communication for the groups in which they have membership.

6. That the communicative system of any type of group can be abstracted by context analysis.

Scheflen illustrated his analytical framework with the case of ‘cigarette lighting and looking up’ as the unit of analysis in a psychotherapy session, marking the transition of the heroic client to patienthood. He identified six distinct components of cigarette-lighting as follows:

1. Taking out a cigarette and bringing it and a pack of matches to her lap.

2. Waiting until the patient has finished a story.

3. Putting the cigarette in her mouth.

4. Waiting until the patient looks away.

5. Lighting up.

6. Discarding the match.

This kind of analysis illustrates how participants mutually orient to one another in context-specific and structured ways.

Among the many milestones in the development of discourse analysis, although some of them may not foreground the terms ‘discourse’ or ‘analysis’, are the following: Interaction process analysis (Bales 1950); symbolic interactionism (Blumer 1969); speech act pragmatics (Austin 1962; Searle 1969); systemic functional linguistics (Halliday 1978); grounded theory (Glaser and Strauss 1967); model of classroom discourse analysis (Sinclair and Coulthard 1975); comprehensive discourse analysis in the context of psychotherapy (Labov and Fanshel 1977); narrative analysis of personal experience (Labov and Waletzky 1967); metacommunication (Watzlawick et al. 1968); frames (Bateson 1972); frame analysis (Goffman 1974); cognitive sociology and interpretive procedures (Cicourel 1974); systematics of turn taking in conversation (Sacks et al. 1974); Lectures in Conversation (Sacks 1992); interactional pragmatics, activity types and language (Levinson 1979, 1983); discourse strategies and conversational inferences including contextual cues (Gumperz 1982); sociology of scientific knowledge and interpretive repertoires (Gilbert and Mulkay 1984; Potter and Wetherell 1987); genre analysis (Swales 1990); text linguistics and text analysis (De Beaugrande and Dressler 1981; Coulthard 1994) macro structures and ideology (van Dijk 1998); critical linguistics (Fowler et al. 1979); critical discourse analysis (Fairclough 1989, 1992); dialogicism (Linell 1998); mediated discourse and nexus analysis (Scollon 2001); multimodality (Kress and van Leeuwen 2001) and others (for overviews see van Dijk 1985, 1997; Schiffrin et al. 2001). Space constraints do not allow me even to gloss the salient contributions these analytic traditions have made, but the list above attests the variability of interpretation to be expected within discourse analysis under different labels. As Gilbert and Mulkay (1984) firmly attest:

Interpretation is therefore indeterminate, especially in relation to intentionality underpinning participants’ communicative behaviour. Such indeterminate accounts may be seen as unsatisfactory, but any ‘definitive versions’ are likely to be equally unsatisfactory ‘because they imply unjustifiably that the analyst can reconcile his version of events with all the multiple and divergent versions generated by the actors themselves’

(Gilbert and Mulkay 1984: 2).

The relationship between participants’ reasoning and analytic reasoning has always remained a thorny issue (Sarangi 2007). On several occasions, Cicourel has been cautious about equating participants’ reasoning with analytic reasoning:

Forms of reasoning are viewed as central to the researcher’s understanding of the way speakers and hearers presumably understand each other. The forms of reasoning we attribute to the participants of discourse parallel the reasoning we employ as researchers in making sense of the speech acts we record and listen to in arriving at some form of analysis. But as researchers we can, of course, specify formal aspects of discourse, produce systematic descriptions, and note emergent properties of the interaction. Yet we cannot attribute such properties unequivocally to the knowledge base of the participants.

(Cicourel 1980: 101)

The interpretive gap between the actor and the observer-researcher, according to Cicourel (1974), is part ideal and part academic:

In actual practice, however, the actor’s everyday theorizing is probably not much different from the observer-researcher. Both employ the same interpretive procedures and similar typification … The observer-researcher must rely upon interpretive procedures when subsuming “recognized” behavioural displays under concepts derived from his scientific vocabulary.

(Cicourel 1974: 39–40)

Over the last six decades, discourse analysis has taken so many shapes and forms that any genealogical mapping to match the branches to their roots may become a futile exercise. Consider the following analytic traditions/labels, which I have listed alphabetically although it would be possible to place them in a continuum of historicity to non-historicity or, alternatively, from etic to emic:

1. Activity Analysis

2. Account Analysis

3. Appraisal Analysis

4. Content Analysis

5. Context Analysis

6. Case Study (analysis)

7. Corpus Analysis

8. Conversation Analysis

9. Critical Discourse Analysis

10. Dialogue Analysis

11. Discourse Analysis (Theme-Oriented Discourse Analysis; Geneva Model of Discourse Analysis etc.)

12. Ethnography of Speaking/Communication

13. Frame Analysis/Framework Analysis

14. Genre Analysis

15. Grounded Theory

16. Interaction Process Analysis (Bales; Interaction Analysis Systems e. g. RIAS)

17. Interpretive Phenomenological Analysis

18. Multimodal Analysis

19. Narrative Analysis

20. Nexus Analysis

21. Positioning Analysis

22. Positive Discourse Analysis

23. Post-structural discourse analysis

24. Rhetorical Analysis (cross-cultural, organisational etc.)

25. Rhetorical Discourse Analysis

26. Stance Analysis

27. Text Analysis (Text Linguistics)

28. Video Analysis

29. Visual Analysis

There are other analytic traditions, which have disciplinary labels, e. g.

1. Critical Psychology

2. Discursive Psychology

3. Interactional Linguistics

4. Interactional Pragmatics

5. Interactional Sociolinguistics

6. Linguistic ethnography

7. Literary Stylistics

8. Microsociology

9. Rhetoric/Composition

10. Social Psychology

11. Systemic Functional Linguistics

The different strands of discourse analysis as we recognise them today do, however, share some family resemblance with one another, given their ontological and epistemological foundations in a broad sense, despite their differential orientation to what counts as data, categorisation, unit of analysis, evidence, claim, etc. The mention of discourse analysis now requires one to clarify what kind of discourse analysis is being undertaken, whereas labels such as conversation analysis and critical discourse analysis are considered more hegemonic (however, see Hammersley 1997 and the discussions about ethnomethodology and conversation analysis (Lynch 2000 and the other contributions in TEXT 20 (4)).

It is possible to limit the disciplinary paradigms which underpin these tribes and sub-tribes, but not in a one-to-one relationship, namely:

1. Dialogicism

2. Ethnography

3. Ethnomethodology

4. Functionalism

5. Hermeneutics

6. Literary/practical criticism

7. Phenomenology/Phenomenography

8. Post-structuralism

9. Structuralism

10. Symbolic Interactionism

Over the years, there have been debates across different analytic traditions, especially involving conversation analysis, ethnomethodology, critical discourse analysis and post-structuralism. Many of these dialogues have been largely defensive and at times acrimonious in challenging the legitimacy of the other, thus resulting in further compartmentalisation in a climate of intellectual apartheid (Widdowson 1995, 1996; Fairclough 1996; Schegloff 1997, 1998, 1999; Wetherell 1998; Billig 1999; Hammersley 1997, 2003a, 2003b; Potter 2003a, 2003b; Stubbs 1997; Lynch 2000).

It is evident from above that text and talk studies are methodologically and analytically pluralistic. Pluralisation of culture (as ‘cultures’) is something Raymond Williams celebrated. So, on the one hand, we can celebrate the proliferation of labels of analytical frameworks and regard such proliferation as necessary intellectual developments in our field. On the other hand, there may be cause for concern, with ambivalence and confusion underpinning various analytical labels and their operationalisation.

Much of discourse analysis is carried out within a broad qualitative paradigm, although some opt for quantitative measures on grounds of empirical rigour. Schiffrin (1987) points to two complementary kinds of analytical accountability: sequential and distributional:

When an analysis provides a comprehensive understanding of the coherence in a text, we may say that it has sequential accountability. When an analysis provides an explanation of why an element occurs in one discourse environment but not another, we may say that it has distributional accountability.

(Schiffrin 1987: 19)

Distributional accountability leans towards a quantitative approach as is the case in corpus linguistics. Other types of analytical accountability would include ‘categorical/rhetorical accountability’ and ‘thematic accountability’ – the latter comprising focal themes and analytic themes (Roberts and Sarangi 2005). In this sense content analysis and grounded theory – to different degrees – may be seen as precursors to discourse analysis.

In the context of healthcare discourse studies, Waitzkin (1991) summarises the drawbacks of both quantitative and qualitative paradigms: quantification studies, while being costly and tedious to carry out, cannot deal with the complex, deep structure of interaction as they systematically disregard underlying themes in context-sensitive ways; in qualitative studies the selection procedure and text presentation is not straightforward and interpretation is difficult to evaluate (cf. Gilbert and Mulkay 1984). It is worth noting that issues of reliability and validity are equally pertinent for qualitative research such as discourse analysis as they are for quantitative research (Peräkylä 1997). The fact that a team of researchers can attain inter-rater reliability during the coding procedure accounts for reliability, but it does not necessarily address issues of validity as far as participants’ meaning-making actions are concerned. Likewise, an increase in sample size can lead to decrease in validity.

Against this backdrop, some scholars, especially those working within institutional and professional settings, have moved away from tribalism in their attempt to marry their scholarly aspirations with practical relevance in an eclectic, ecological fashion. It is worth noting the rise of mixed methods, including eclecticism, in contemporary scholarship. There are risks and benefits of adopting a mixed and/or eclectic approach at various levels:

1. Theoretical mixing

2. Crossing disciplinary boundaries

3. Across methodological paradigms (quantitative and qualitative)

4. Within a research paradigm: primary and secondary data sources (e. g. combining DA and ethnography; interview data and naturally occurring data)

5. Combining analytical frameworks

When it comes to eclecticism, it may be useful to distinguish between opportunistic and systemic mixing of analytical paradigms – if they are indeed distinct paradigms in the first place. While systemic eclecticism is a reasoned way of integrating various frameworks in anticipation of one’s data analytic endeavour, also premised upon shortcomings in previous studies, opportunistic eclecticism is analogous to ‘butterfly collection’, i. e., picking and choosing different analytical trajectories as one goes along. Some relational pairs at the systemic level – e. g. corpus linguistics and critical discourse analysis; ethnomethodology and conversation analysis; ethnography and discourse analysis – are proving to be very productive in pushing the boundaries.

In juxtaposition to eclecticism, we can turn to Parker (1992) who prescribes 20 steps to analyse discourse dynamics:

Steps 1 & 2: selection of text

Steps 3–12: systematic identification of the subjects and objects in the text

Steps 13–20: examination of the ways in which the discourses that structure the text reproduce power relations

In a similar vein, Willig (2001, 2008) outlines 6 stages (using research interview data as exemplar):

No doubt, such step-wise procedures give discourse analysis an air of scientificity, but the steps/stages may be far more overlapping and recursive than meet the eye (e. g. practice vs. action; subjectivity vs. positioning). Additionally, there is the risk of over-proceduralising the discourse analytic practice which may drive out professional wisdom, artistry, intuition and judgement (Schön 1987).

As a threshold – albeit sounding reductionist – I have been using for my own research and teaching purposes the mnemonic of discourse as follows (Sarangi 2010) (see Figure 1):

Figure 1: DISCOURSE mnemonic.

In any given discourse analytic task, a choice has to be made rather than giving attention to everything that the notion of discourse encompasses. When it comes to operationalising or practising discourse analysis, the mnemonic above can be translated in to the following list:

1. Three Cs: context; contextualisation cues; categorisation/category-work

2. Three Fs: frame; footing; face-work

3. Three Is: indexicality; implicature/inferencing; intertextuality/interdiscursivity

4. Three Ms: modalisation; metalanguage; mapping

5. Three Rs: rules (norms); roles (positioning); rhetorical devices

6. Three Ts: turn-taking; topic management; target audience/addressee

Accompanying the above list are an identifiable cluster of rhetorical/discourse devices which function as foot soldiers and can be used in any configuration in text and talk studies:

1. character work (identity; role-set)

2. event work (including hypothetical scenarios)

3. facework

4. framing, footing, formulation

5. Contrast (also called parallels capturing both similarities and differences)

6. Constructed dialogue (including reported speech, reported thought)

7. Extreme case formulation

8. Typical case formulation

9. Concession

10. Repetition

11. Questions (including question-answer and initiation-response-feedback sequences)

12. Pronominal reference

13. Particles

14. Metaphor

15. Laughter

16. Overlaps and interruptions

17. Contextualisation cues

18. Etc.

The three-part sets and the list of rhetorical/discourse devices, which are by no means exhaustive, cross the divide between text analysis and talk analysis, although one needs to remain sensitive to the similarities and differences across the two modes (Leech 2000).

3 An editorial reminiscence

Editorship of a journal for 20 years cannot be without its rewards. In processing hundreds of manuscripts and in making decisions about rejections and acceptances I seem to have acquired a stock of experiential knowledge that I wish to share here as I have done on numerous occasions over the past 15 years in many different parts of the world in the form of international publishing workshops.

I believe four distinct but inter-related forms of knowledge contribute to successful authorships. I would label these as follows:

1. Subject (disciplinary) knowledge (the what-dimension)

2. Linguistic/communicative knowledge (the how-dimension)

3. Genre knowledge (i. e., journal article genre vs. textbook genre; genre of cover letters responding to reviewers)

4. Contextual/procedural knowledge (about publishing in specific journals, the journal publishing marketplace)

Over the years the following issues have been commonly identified by various reviewers as shortcomings of individual submissions, leading to rejections.

1. Missing focus (in title, abstract, research question)

2. Absence of theoretical orientation, if applicable

3. Lacking analytic sophistication (paraphrase vs analysis)

4. Mismatch between theory/literature review and data analysis

5. Poor contextualisation of relevant research in the field; unsystematic review of literature (in terms of themes, study settings, theoretical, methodological and analytical frameworks)

6. Weak evidence for making generalised claims

7. Lacking clarity of expression and organisataional structure

8. Absence of originality (covering familiar territory)

9. Written in a textbook style; reads like a PhD thesis chapter or coursework/assignment

10. Not worthy of publication in the journal concerned

Of the many weaknesses, I want to focus on the analytical aspects as many submissions are turned down because of a lack of fit between data and analysis, i. e. the poverty of interpretation (dumbed as glossing or paraphrasing of data).

4 Analytic ‘gaze’ vs ‘glance’

My creative spelling of en‘gaze’ment in the title of this Editorial is designed to draw attention to the notion of ‘gaze’ as a powerful metaphor in categorisation of phenomena – with corollaries elsewhere, e. g. clinical gaze, ethnographic gaze. Categorisation of observable phenomena is central to all human activity, but within language/communication studies, categorisation is not only the object of study but also the process through which analysts make sense of others’ categorisation practices. Description is the basis of categorisation, which is central to both the natural sciences and the human/social sciences. According to Dunn (1978), while within the natural sciences descriptions take the form of statements and assertions which can withstand inter-observer reliability, within the human and social sciences ‘reporting’ is the preferred form of description which may have no corresponding inter-observer reliability.

An attendant concern in our categorisation practice is what is widely known as ‘category mistake’. Following Gilbert Ryle (1949):

When two terms belong to the same category, it is proper to construct conjunctive propositions embodying them. Thus a purchaser may say that he bought a left-hand glove and a right-hand glove, but not that he bought a left-hand glove, a right-hand glove and a pair of gloves. ‘She came home in a flood of tears and a sedan chair’ is a well-known joke based on the absurdity of conjoining terms of different types. It would have been equally ridiculous to construct the disjunction ‘She came home either in a flood of tears or else in a sedan chair’.

(Ryle 1949: 23–24)

A classic illustration of category mistake would be to characterise folktales on the basis of ‘tales with fantastic content’; ‘tales of everyday life’; and ‘animal tales’ (Propp 1968). There is no particular reason to believe, for instance, that ‘animal tales’ could not contain ‘fantastic content’ or that ‘tales with fantastic content’ could not be about ‘everyday life’. Here we have potentially overlapping categories, thus each individual category while being meaningful, becomes less sustainable in its own right. A grounded theory approach, which privileges a ‘constant comparison method’, may come up with a more robust categorisation system concerning Propp’s folktales.

Behind all categorisation is the perspective/stance of the observer. One may say that category lies in the eyes of the observer, from the vantage point of observation – what I will refer to as analytical gaze and glance.

There are two dominant metaphors to characterise the act of categorisation/interpretation: the tool box metaphor and the lens metaphor. Under the tool box metaphor, the classic slogan is the ‘Law of the hammer’: “It is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail” (Maslow 1966: 15). The ‘law of the hammer’ amounts to narrow-minded instrumentalism, or over-reliance on a familiar tool, which is characteristic of scientific inquiry. In discourse analysis, it means making analytic tools our focal interests, as can be seen in many methodological/analytical traditions which work with coding systems and impose them on emergent data.

The coding of data is motivated by a desire to quantify the interaction process in a systematic, rigorous way so that it can influence outcome measures. This analytic activity overlooks the fact that interaction is a dynamic, cumulative activity which defies any one-to-one correspondence between linguistic/semiotic form and function and that meaning is negotiated sequentially and simultaneously in an ongoing basis. Many of the codes may be quasi social action labels, similar to static speech act taxonomies, and therefore reductionist (Sarangi 2010). From the viewpoint of discourse analysis, early critics of coding are Aaron Cicourel and Elliot Mishler. Cicourel observes:

Coding the transcript in terms of the meanings of singular utterances is difficult even when looking at the remarks immediately preceding and following the utterance being coded, yet they can lead to interesting and important frequency counts of different kinds of remarks. This kind of coding can also lead to misinterpretations of the extended dialogue because of the fact that both patient and doctor remember earlier remarks and use them to initiate and comment on later remarks.

(Cicourel 1981: 421)

Mishler (1984) points to the fault lines when coding is the primary tool of data analysis. According to him:

[Coding systems] violate the two most essential features of natural discourse, its organisation and structure, and the contextual grounding of meaning. The use of standard coding systems further compounds this problem since the categories in such codes do not refer to the special and particular content of medical interviews. Finally, there is a marked tendency in these studies to interpret findings on the basis of implicit assumptions that reflect a medical bias.

(Mishler 1984: 57)

The lens metaphor is more nuanced. It stands for microscopic attention to the observable phenomena. Goffman (1974) uses the lens metaphor alongside the metaphor of tool box as follows:

The question ‘What is it that’s going on here?’ is considerably suspect. Any event can be described in terms of a focus that includes a wide swathe or a narrow one – and as a related but not identical matter – in terms of a focus that is close up or distant. And no one has a theory as to what particular span or level will come to be the ones employed. To begin with, I must be allowed to proceed by picking my span and level arbitrarily, without special justification.

(Goffman 1974: 8)

While Goffman promotes a loose sense of eclecticism in his approach, he was able to come up with microscopic observations and categorisations of everyday life. Many analytic traditions, such as conversation analysis and corpus linguistics, would insist on systematic and scientific rigour than what Goffman practised.

Let me dwell further on the notion of ‘gaze’ as it is something a discourse analyst of any persuasion is expected to embody in their scientific enterprise. In The Birth of the Clinic, Foucault (1973) coined the term ‘medical gaze’ to characterise the stance of the doctor in making a diagnosis and identifying the disease causation. According to him:

The medical gaze embraces more than is said by the word ‘gaze’ alone. It contains within a single structure different sensorial fields. The sight/touch/hearing trinity defines a perceptual configuration in which the inaccessible illness is tracked down by markers, gauged in depth, drawn to the surface, and projected virtually on the dispersed organs of the corpse.

(Foucault 1973: 164)

Foucault continues:

The observing gaze refrains from intervening: it is silent and gestureless. Observation leaves things as they are; there is nothing hidden to it in what is given. The correlative of observation is never the invisible, but always the immediately visible, once one has removed the obstacles erected to reason by theories and to the senses by the imagination.

(Foucault 1973: 107)

Here we have an invitation towards a-theoretical, non-imaginative observation which is emergent in character. Foucault then goes on to suggest a distinction between gaze and glance which is relevant. He writes:

In fact, the gaze implies an open field, and its essential activity is of the successive order of reading; it records and totalizes; it gradually reconstitutes immanent organizations; [ … ] it forms, as it were, the privileged articulation of two fundamental aspects of saying (what is said and what one says). The glance, on the other hand, does not scan a field: it strikes at one point, which is central or decisive; the gaze is endlessly modulated, the glance goes straight to its object. The glance chooses a line that instantly distinguishes the essential; it therefore goes beyond what it sees; it is not misled by the immediate forms of the sensible, for it knows how to traverse them; it is essentially demystifying.

(Foucault 1973: 121)

The idea of going beyond what is seen resembles the Chomskyan metaphors of deep structure and surface structure phenomena. Elsewhere Foucault characterises discourse analysis as an activity constituted in ‘doing more’. Consider Foucault’s (1972: 49) conceptualisation of discourse:

Of course, discourses are composed of signs; but what they do is more than use these signs to designate things. It is this more that renders them irreducible to the language (langue) and to speech. It is this ‘more’ that we must reveal and describe.

[Italics in original]

The notion of gaze sits comfortably with ethnography, as practised by Clifford Geertz (1973). His notion of ‘thick description’ can be equated with en‘gaze’ment with the observed phenomena at a deep-structural level.

If gaze is about deep-level analysis then glance concerns selectivity or even intuitive bias/judgement. It is worth reminding ourselves of Malinowski’s (1935) confessional account of ethnographic practice:

The main achievement in field-work consists, not in a passive registering of facts, but in the constructive drafting of what might be called the charter of native institutions. The observer should not function as a mere automaton; a sort of combined camera and phonographic or shorthand recorder of native statements. While making his observations the field-worker must constantly construct: he must place isolated data in relation to one another and study the manner in which they integrate. To put it paradoxically, one could say that ‘facts’ do not exist in sociological any more than in physical reality; that is, they do not dwell in the spatial and temporal continuum open to the untutored eye. The principles of social organisation, of legal constitution, of economics and religion have to be constructed by the observer out of a multitude of manifestations of varying significance and relevance. It is these invisible realities, only to be discovered by inductive computation, by selection and construction, which are scientifically important in the study of culture.

(Malinowski 1935: 317)

Following from this, how would one characterise a discourse analytic gaze or glance? The circumference of analytic activity is a matter of limiting one’s observation/interpretation in light of what is taken for granted within a discipline – the doxa. The topic/resource distinction was aptly captured by Moerman (1988) with the following observation:

All natives take their native knowledge for granted, take it to be nothing other than the nature of the world (Geertz 1973). But how could the conversation analyst recognise an utterance as a pre-invitation, for example, without trading on covert native knowledge of dating practices and the special significance for them of Saturday night?

(Moerman 1988: 4)

Here neither gaze nor glance will help an outsider analyst to reveal what underlies local native practice. Thus the limits of description have to be contained with. Consider the following descriptive account about the rain-making ritual associated with the Toka tribe:

One morning, most of the men and women from three neighbouring villages assembled in the village in which I stayed. After a while, they set out on a march, together with the people from this village, out of the settlement. The march, in which altogether about eight people took part, was led by a middle-aged man. The marchers sang and occasionally stopped and danced for a while to the accompaniment of more songs. After having marched for about three miles, they stopped in the bush. The women cleared a small patch of ground by removing the grass. Then all except the leader of the march sat down on the ground and slowly clapped their hands in rhythm. The rhythmic clapping of hands was occasionally accompanied by the beating of a drum. While this went on, the man who had led the procession poured some water, beer and milk on the cleared patch of ground and simultaneously delivered a short speech, obviously addressed to no one in particular.

(Holy and Stuchlik 1983: 35)

Holy and Stuchlik make the point that a description of actions and practices without appealing to the emic beliefs and values underpinning the rain making ritual will only remain a description devoid of understanding and meaning making. The sequence of ritual actions and the participation framework described have deeper meanings, including invocation of ancestors to bring about rain.

The disciplinary gaze and the tools accompanying the gaze extend to other areas of scientific activity. A sense of taken-for-grantedness becomes second nature to analytic practice. Agnes Arber (1964) provides the example of an average biologist who may take his/her microscope for granted and simply use it as a tool, whereas a maker of the optical instrument would critically appraise the microscope itself as an instrument. Likewise, a philosopher may question the theoretical assumptions underlying the biologist’s observations without necessarily appreciating the findings. Once again, the topic/resource dilemma is played out cross-disciplinarily. There is thus an appeal to division of expert labour, while acknowledging the pitfalls of infinite reflexivity.

N. R. Hanson (1958), in the tradition of philosophy of science, and in relation to experimental psychology, argues that perception is intertwined with concepts – hence any observation is theory-laden as opposed to ‘phenomenal seeing’. He offers a distinction between ‘seeing as’ and ‘seeing that’: “it is a matter of logic, not merely a matter of fact, that seeing as and seeing that are indispensable to what is called in science, seeing or observing” (Hanson 1958: 86). He explains:

Seeing that threads knowledge into our seeing; it saves us from re-identifying everything that meets our eye; it allows physicists to observe new data as physicists, and not as cameras […] Observation in physics is not an encounter with unfamiliar and unconnected flashes, sounds, bumps, but rather a calculated meeting with these as flashes, sounds and bumps of a particular kind.

(Hanson 1958: 20–24)

‘Seeing that’ approximates ‘glance’ as I have been discussing so far – to align with intuition, judgement, bias etc. To an extent it also aligns with Garfinkel’s (1967) proposal about the ‘documentary method of interpretation’, which extends Mannheim’s original conceptualisation:

The method consists of treating an actual appearance as “the document of,” as “pointing to,” as “standing on behalf of” a presupposed underlying pattern. Not only is the underlying pattern derived from its individual documentary evidences, but the individual documentary evidences, in their turn, are interpreted on the basis of “what is known” about the underlying pattern. Each is used to elaborate the other.

(Garfinkel 1967: 78)

The notion of ‘glance’ also intersects with Donald Schön’s (1983: viii) characterisation of competent practice:

Competent practitioners … exhibit a kind of knowing-in-practice, most of which is tacit … Indeed, practitioners themselves often reveal a capacity for reflection on their intuitive knowing in the midst of action and sometimes use this capacity to cope with the unique, uncertain, and conflicted situations of practice.

This is echoed in many discussions about the novice-expert continuum in a given professional context. With reference to nursing, Patricia Benner (1984) identifies five distinct stages to becoming an expert, although it is difficult to map the taxonomy discoursally on to different stages of professional development, e. g., discourse analytic expertise. But Benner’s definition of expert performance has a ring of familiarity:

The expert performer [is one] who no longer relies on analytic principle (rules, guidelines, maxims). The expert nurse, with an enormous background of experience, now has an intuitive grasp of each situation and zeroes in on the accurate region of the problem without wasteful consideration of a large range of alternative diagnoses and solutions.

(Benner 1984: 31–32)

The significance of intuition in expertise has been receiving attention over the years; however there is prevailing scepticism about how to define and identify occasioning of intuition (English 1993).

5 Conclusion: Rigour or relevance

Donald Schön (1983) draws attention to the dilemma of ‘rigour or relevance’ in professional practice more generally – which can be extended to discourse analytic practice. He stipulates the dilemma as follows:

This dilemma of ‘rigour or relevance’ arises more acutely in some areas of practice than in others. In the varied topography of professional practice, there is a high, hard ground where practitioners can make effective use of research-based theory and technique, and there is a swampy lowland where situations are confusing ‘messes’ incapable of technical solution. The difficulty is that the problems of the high ground, however great their technical interest, are often relatively unimportant to clients or to the larger society, while in the swamp are the problems of greatest human concern. Shall the practitioner stay on the high, hard ground where he [she] can practice rigorously, as he [she] understands rigour, but where he [she] is constrained to deal with problems of relatively little social importance? Or should he [she] descend to the swamp where he [she] can engage the most important and challenging problems if he [she] is willing to forsake technical rigour?

(Schön 1983: 42)

On many occasions, as a practising discourse analyst in the healthcare setting, I have confronted this dilemma and have survived to tell the tale. My interpretation of data has occasionally been challenged by healthcare practitioners (for a fuller account, see Sarangi 2015) as well as by discourse analysts from cognate traditions. I have taken solace in the notion of ‘descriptive discretions’ as formulated by John Dunn (1978):

The prevalence of descriptive discretion is not in itself epistemologically alarming, though it does weaken the prospects of any very crisply incremental development of the social sciences. In itself it comes no closer to imperilling the reality of human performance than variations in the taste of landscape artists come to altering the physical properties of mountains. To make it epistemologically alarming and not merely methodologically troublesome, it would be necessary for it to extend to the assertion and negation of the same description (identically interpreted) of the same phenomena by two different competent, sincere and attentive observers.

(Dunn 1978: 155–156)

Inter-observer or inter-analytic convergence is more likely to happen within a specific tradition of discourse analysis and is most likely to be a remote possibility across different traditions. Not only the stances of the analysts but also their language of analysis (i. e., metalanguage) are bound to be variable, with unintended epistemological consequences. If one were to invent a ‘discourscope’ modelled after the microscope, even though the microscope has been outlived by new technologies of observation in the biosciences, we would still have to adjust the lens of ‘gaze’ or ‘glance’ depending on our analytical motivations.