Douglas Maynard (2003) Good News, Bad News. Conversational Order in Everyday Talk and Clinical Settings

The conversation analytic perspective (CA) has delivered a great amount of rich empirical analysis of micro-level social systems. Being based entirely on conversational data (tape recordings usually), CA is a deliberately restrictive enterprise.  Sometimes CA studies are daring so far into the territory of (socio-)inguistics that sociologists are sceptical as to whether their results are still even relevant for the discipline. Douglas Maynard’s Bad News, Good News (2003) has the great benefit to relate a rigid conversation analytical perspective (CA) to broader concerns of sociological ethnography and political implications of this type of research. The topic of how bad and good news are delivered is a thankful one for this exercise. On the one hand it centres on a reasonably manageable sequence of conversation that in itself is dynamic and morally loaded, and on the other hand this sequence is central to the performance of actors in significant institutional contexts (hospitals, political communication). While I cannot say that I would find Maynard’s attempt fully convincing, it will undoubtedly remain one of the most important points of reference in this regard for some time to come.

Maynard begins the book with three chapters that deal with methodological questions surrounding conversation analysis. In the first chapter, he positions ethnomethodology (EM) in relation to the phenomenological tradition. Maynard understands the epoche (that phenomenologists aim to bring about artificially by “bracketing” our everyday understanding of the world) as a common occurrence in the life-world. Bad and good news, he claims, have the same “interruptive and sometimes utterly disruptive” character. As such, they result in the kind of “noetic crisis” (12) that Garfinkel told his students to bring about (equally artificially when compared to phenomenology) by acting out in front of their parents. There are a series of questions that might ensue from these comparisons and cross-references about the relation of EM to philosophy, about the status of “bracketing” a method or as a naturally occurring phenomenon and about the naturalness of data in EM. But Maynard does not go down this road and neither will I.

The second chapter mainly deals with a series of strategies and problems that are fairly well-known in relation to bad news delivery. Maynard only brings these up to argue for a more interaction-based approach that does not simply “pigeonhole” observations into abstract categories (64). Maynard also tries to argue throughout his book that bad and good news delivery should not be treated independently and – as is the case mostly – good news delivery should not be neglected entirely. Their similarity and even their apparent asymmetry, he claims, can teach us something about the phenomenon.

Chapter three deals with the relationship between CA and ethnography and is likely to be the most widely-read chapter of this book which can be recommended for any course on EM and CA. Ethnographic research on bad news is mostly “occupationally and substantively confined” (66) whereas Maynard says his research is based on an “activity focus” (65; cf. Drew/Heritage 1992; Levinson 1979). As such, Maynard needs to address the “contextual critique” leveraged against EM and especially CA. “(L)arger or broader structures, categories, or organizations”, Maynard explains, should only be considered “(1) (if) such categories are relevant to participants and, if so, (29 (if) they are procedurally consequential in the sense that participants display, in their talk and interaction, an orientation to them” (70). As a result, Maynard is sceptical of what he terms a “mutual affinity” between ethnography and CA (68). Ethnography and CA should, for his purposes, be characterized by a “limited affinity” in which ethnography has three uses: 1.) It helps to describe the setting and identities within it and to choose which features of them should be described in more analytical detail. 2.) It helps to understand and explicate unfamiliar terms, phrases and courses of action. 3.) It helps with “curious” patterns of action that a sequential analysis cannot fully explain. As can be seen from this list, while Maynard does not eschew ethnography fully as do some of his colleagues in CA, it can hardly be said that ethnography is allowed any substantial or systematic part in his research. Instead of expanding beyond the conversational interaction, Maynard advises to go “’deeper’ into the concreteness of a setting” (77).

Chapters 4 to 7 present the empirical research of the book. I want to present these parts only cursorily as CA is best read with its data next to it and is therefore often somewhat lengthy. Firstly, Maynard presents what he calls “The News Delivery Sequence” (NDS) which is best presented in this chart from the book: 

 The News Delivery Sequence (NDS) in Maynard (2003): 96.

The News Delivery Sequence (NDS) in Maynard (2003): 96.

The NDS proceeds from an announcement to a response to such announcement which will either encourage or discourage further elaboration. Depending on whether it is good or bad news, whether the news is unexpected – and depending on why it might be unexpected – the reaction will look or be expected to be different. An uncertainty on how to react to news (for example, is expecting another child good or bad news?) might lead to embarrassment and will have to be resolved interactionally. The announcement response as well as the elaboration will consequently tell observers a lot about the relationship of the interaction members and equally tell them something about that relationship as well.

In the fifth chapter, Maynard distinguishes three forms of news delivery according to who is the “consequential figure” (121), i.e. which person is affected by the news. He accordingly differentiates between first party news (news about me), second party news (news about you) and third party news (news about another). Part of the results of this chapter are, again, summarised in a table replicated below.  The reader might wonder what the purpose of these taxonomic exercises might be that are reminiscent of the very “pigeonholing” Maynard had initially criticised. Especially the section on second-party news (e.g. on stoic reactions to news) may be helpful for practitioners, such as nurses – and, in fact, Maynard’s data is mainly derived from such settings. Beyond that, it is less clear to see the scope of the analysis.

The next chapter asks why bad and good news tend to be delivered differently. Specifically, Maynard argues bad news creates more trouble and makes more interactional work necessary. The reason for this, he argues, can be found in Sacks’s elaborations on “being ordinary” (Sacks, Lectures on Conversation, vol. II, part IV, lecture 1, pp. 215-222). Maynard argues, “(inasmuch) as the everyday social world is achieved as externally real and ordinary, it has an accomplished benign structure”  (183). What might interest us here is the deductive move Maynard (and Sacks) employs here: from small and detailed interactional data we draw conclusions about something as general as “the everyday social world”. Of course, institutional settings might influence how difficult it is to convey negative news and how unexpected bad or good news appear in context. The degree of difficulty and the interactional work necessary to overcome it might vary accordingly. Maynard’s discussion of doctor-patient or nurse-patient interaction goes to some length in this direction as well. Nurses with news about HIV test results, for example, tend to favour distanced behaviour.

Chapter seven addresses the often-noted phenomenon that the messenger and the message tend to become confused in bad news (but very rarely in good news). In Antigone, Sophocles famously had a guard engage in a lengthy pre-announcement (followed by an even lengthier elaboration) when he delivered the news that someone buried the corpse king Creon had forbidden to be touched:

Creon
And what is it that so disheartens you?

Guard
I want to tell you first about myself—I did not do the deed, nor did I see the doer, so it would be wrong that I should come to any harm.

Creon
Like a bowman you aim well at your target from a distance, and all around you hedge yourself off well from the deed. It is clear that you have some unheard-of thing to tell.

Guard
That I do, for terrible news imposes great hesitation.

Creon
Then tell it, will you, and so unburdened go away?

Guard
[…] So here I stand, as unwelcome to you as I am unwilling, I well know. For no man delights in the bearer of bad news.”

 The guard is reporting to Creon in a modern adaptation of Sophocles'  Antigone . (c) Magda Molin

The guard is reporting to Creon in a modern adaptation of Sophocles' Antigone. (c) Magda Molin

The Persians- the always present enemy of the ancient Greeks – were known to execute messengers of bad news. It can be assumed that Sophocles wanted to morally educate the spectators of his play by teaching the importance of the boundary between the message and the messenger. Over 2400 years later, the blurriness of this distinction is still able to burden interaction in a way that needs to be countervailed by practices Maynard seeks to describe.

The last chapter of the book finally tries to connect conversation analysis to broader social implications. Many authors before have doubted whether this is possible at all and Maynard’s attempt also leaves much to be desired. The “benign order of everyday life”, Maynard argues, prevents politicians from admitting mistakes and leads to such disastrous results such as the systematic denial of epidemics and their consequences. This diagnosis seems pretty disingenuous and is also empirically out of sync with the rest of the book. To transpose results from “naturally occurring” data onto highly institutional and strongly charged settings ultimately sacrifices the specificity of the issue at hands for the sake of supposed clarity.

Overall, Maynard’s book offers a lot of detailed analysis of interactional data from which EM-scholars can draw a lot of inspiration. The relation to political questions, however, still needs to be clarified and the connection of CA to ethnographic data also ends up being much less rigidly specified than one might wish for. “Limited affinity” might be a good pragmatic solution, but it does not really touch the problems CA is faced with when dealing with institutional and culturally (or morally) sensitive data. The taxonomies of interaction sequences observed by Maynard can help especially practitioners from the fields that were observed for this book. But we should ask: To what extent is the NDS a “natural event” and to what extent do institutional contexts influence it? How could we operationalise the observation of such contexts more effectively? What effects have unwanted or institutionalised repetitions of the observed sequences influence on the efficiency and the way news deliveries operate? For example, considering the amount of medical knowledge available online and the de-legitimation of the medical profession to some extent, how do challenges to bad news affect the NDS? Does the delivery of bad news in the first instance (e.g. when there is just a suspicion) affect proceeding news deliveries (e.g. impede revisions). Such chains of NDS and their institutionalisations might need more investigations and it is yet to be seen to what extent CA could be helpful for such projects.