BNIM:

TFA ‘Double-Experiencing’ Hypotheses (DEH)

+ Linked Following-Hypotheses (LFH)

The Biographic-Narrative Interpretive Method (BNIM) has been used in a variety of health research theses and post-doctoral research projects (as well as other areas). It involves a 2-subsession narrative interview and a protocol for interpreting the materials along two separate tracks and then bringing the results of the twin-track interpretations together afterwards.

One of the key components of track 2 – the interview transcript treated as a succession of successive micro-acts – is the TFA (/Teller Flow/Thematic Field/analysis).

Along this interpretive track, BNIM is concerned to try to get well-evidenced inferences about the ‘lived experience’ of the interviewee as they tell their told story in the interview. This involves a segment-by-segment consideration of the interview record, constantly inquiring into the ‘lived experience’ of the teller that might be thought to be ‘behind’ the actual recorded telling.

The danger is that hypothesising about the lived experience in the field of health – or any other field – can be relatively shallow and rather boring.

As a narrative researcher into the experience of ‘ill-health’ and of health workers, you might be interested -- whether you use BNIM or not – into the attempt below to ensure that researchers using BNIM nudge themselves towards a comparatively ‘deep’ inquiry into often complex and contradictory experience. It relates to the notion of a ‘BNIM panel’ in which three or four people including the researcher generate and eventual test and evaluate hypotheses about the ‘lived experiencing’ of an interviewee going through a biographic narrative, a told story of earlier experience.

All comments welcomed. If you like a free electronic copy of the current version of the BNIM Short Guide and Detailed Manual, just let me know

Tom

This is a revised (July 2012) version of a section in the BNIM Detailed Manual. It is in section (3.5.3.2). on finding the pattern of the telling of the told story

Note that the interviewee is having a double experiencing in their telling, and your (panel) hypotheses about “What are they experiencing at this point in the telling?” must implicitly have a double referent.

In the example below, the interview has recalled being badly treated in primary school by a teacher.

About-PIN chunk	Event being recalled : Being badly treated in primary school by a teacher
Chunk of Evaluation	“At age 35 this is not difficult at all: I suffered a bit, perhaps not even a bit”
The panel is then asked to put forward a hypothesis about a ’double experiencing’ at the point at which this Evaluation was given	What might be an EH about….	What might be an EH about…..
	Their then-experiencing	Plus Their now-in-the-interview experiencing
	Recalled At the time: repressed great anger and shame?	Now also: Sympathy with past self – sympathy with teacher – nostalgia for past reality?
The hypothesis of a double experiencing is =>	All of the above in complex joining and mutation, part-denying, part-asserting

Figure 20 Double experiencing when telling story: I suffered a bit

So the ‘hypothesis about the double experiencing’ at the point at which the interviewee says

“At age 35 this is not difficult at all: I suffered a bit, perhaps not even a bit”

might be (after the facilitator has done a lot of pushing (not for PINs but ) for more precise and fully-developed Experiential Hypothesising:

She is remembering that she felt great shame and anger when treated in that way, and that she had also felt that she had no choice but accept the treatment. Remembering it now, she wishes to appear in her own eyes and in the eyes of the interviewer as a mature adult and to do so, she distances herself from the child she was and her emotions then. Herself now also a teacher, she can also see that, without realising it then, she herself had been challenging the teacher in a way that she now knows teachers find it hard to deal with. She regrets a bit how far she has come from a childhood and a schooling which she retrospectively thinks of as being quite a happy and a privileged one, This ‘adult perspective’ makes her keep well away from identifying with the child she was, and does this by minimising what happened. That’s why she says

“At age 35 this is not difficult at all: I suffered a bit, perhaps not even a bit”

The concession to her child-self “I suffered a bit” is pretty much completely (though not completely) withdrawn when she says “perhaps not even a bit”.Her current self-image has to suggest that maybe she as a child “perhaps” didn’t suffer at all.

An earlier admission of being badly treated and by implication is suffering is in process of being withdrawn by her on the grounds, perhaps, “that for her now at 35 this is not difficult at all”.

Perhaps, without realising it, in the interview she is re-enacting the repression of suffering (of great anger and shame) that she felt as a child that she had to do in the classroom. Does the interviewer now represent her class peers at the time, her then-teacher and her then-parents?

This is an ‘expanded version’ of the sort of deliberation about experiencing that might go on in or after a BNIM TFA panel attempting to spell out a ‘double experiencing hypothesis’.

The facilitator would have to find a way of ‘condensing’ such a deliberation into a few phrases on the flip-chart sheet.

He or she might then ask the following question to the panel:

Do any Structural Hypotheses about her situated subjectivity (in that then- child-phase of subjectivity, in her current phase of subjectivity within health care) suggest themselves?

There might or might not be one, or none, or more than one……

A ‘Double-Experiencing Hypothesis’ about the (often contradictory, ambivalent) experiencing during the doing and the closing of the interview segment is put forward, and it is part of the BNIM method that such hypotheses should be preferably linked to an eventual ‘guess’ (one or more ‘following hypotheses’) as to what the shape of the next or succeeding text segments would be like if the ‘Experiential Hypothesis’ were true

(Linked Following Hypothesis = hypothesis about what might follow in the data-series, defined more narrowly or more broadly).[1]

E.g. If the ‘double-experiencing hypothesis’ (DEH) is that the interviewee is embarrassed about showing emotion when giving a detailed story about topic H (Health), then one ‘linked following hypothesis’ (LFH) about what will happen next or soon in the text might be that he or she will shift either to a less emotion-inducing topic (e.g. Work) or to a different way of talking about the same topic H (rather than re-living a narrated moment very concretely, s/he might keep the emotion better at bay by either keeping the topic H but shifting to an abstract argument, or by launching into ‘a declaration of values’). S/he might possibly do both (changing topic and shifting out of narrative).

The last paragraph summarises a range of Following Hypotheses that might arise if the Experiencing Hypothesis (embarrassment about/showing emotion when giving detailed story – PIN – about topic H) were true.

Once the panel has registered its alternative models of the situated experiencing subject at that point (Experiential-Hs + some Following-Hs for each Experiential-H), then the actual next segment is revealed.[2]

This new chunk of data might confirm some previous (following) hypotheses, force a shift in others, lead us to reject still others. Some earlier hypotheses might – and for a long time they do -- remain untouched. Having considered the impact of the new event chunk on earlier hypothesising, the panel then turns to generate new hypotheses and counter-hypotheses in the light of the new event-chunk.

Taking the example in the text, if the speaker did shift topic to an emotionally-neutral one and also started talking in an abstract position-taking way, then the experiential hypothesis of ‘Embarrassment about showing emotion’ would appear confirmed.

But, if the shift was to an even more delicate topic and a continuation of narrating key moments in detail (PINs), then the previous (‘Embarrassment about showing emotion’) hypothesis about why the previous segment came to an end might have to be revised or even rejected.

For a magisterial account (and a magisterially difficult one !) about the revealing difficulties of narrating, see Schütze (2005).

Then the exercise is repeated.

Experiencing hypothesising, and linking them to ‘Following Hypotheses’:

In the “I suffered a bit, perhaps not at all” example cited above (p.552), the ‘Experiencing/ Experential’ hypothesis was developed in a way which showed the potential of such hypothesising to explore quite deep into subjectivity.

In the next example, we can see how a panel not pushed towards deeper understanding (deep-EHs) can actually stay excessively shallow (and short) in this crucial work of ‘imagining situated subjectivity’ in the TFA panel.

START OF CRUCIAL DIGRESSION ON ELABORATION OF ‘EXPERIENCING HYPOTHESES’ IN TIGHT CONNECTION TO ANY ‘FOLLOWING HYPOTHESES’ DERIVED FROM THEM TO ‘TEST’ THEM

1. The hypotheses (especially the ‘Experiencing Hypotheses’ EHs) need to be detailed

2. The following hypotheses (FHs) are only of interest if linked to specific EHs (Linked Following Hypotheses)

Notice that you need to move yourself towards quite specific hypotheses about experiencing. Just putting a one-word EH will not prove to be sufficient.

The example below is from a recent BNIM trainee’s record of a TFA panel, and it refers to the interviewee’s experiencing of the delivery of the SQUIN at the start of the interview. The ‘comment’ is by me. If panel members just started with one-word accounts of the ‘experiencing’ interviewee (.g. ‘nervous’), the experienced panel-facilitator would not have not just left the matter there but he or she would, in the panel, have pushed the one-worders to elaborate each of their initial ‘one-word experiencing’ hypotheses.

Let us see what did happen, as provided in the TFA record.(The comment was my comment to the trainee on the draft she sent me). .

Line

Textsort

Gist

EH: Hypotheses about the experiencing of the interviewee (EH)

FH: Hypotheses about what might follow (FHs) if a particular EH was true

001

I: Initial Question

D1.Delivery of SQUIN

Interested in young peoples’ experiences of care and leaving care

Uncomfortable

Reluctant

Nervous[RA1]

FH1. Will Pause

Unproven by D2

FH2. Will Ask what I wanted to know

Unproven by D2

FH3. Will give chronological account of care

Proven

Slightly revised –

Figure 21 TFA panel record - undeveloped hypothesising, unlinking FHs to EH

In the example above, each experiencing hypothesis is only given ‘one word’ (although interestingly, the FHs in the final column are given more and better elaboration). In addition, it is not clear to which of the three ‘one-word EH’, each of the FHs is attached. The structure needs to be elaborated.

1. If, in the final column, FH3 comes to be confirmed by later chunks, we don’t know how this relates to any of the three EHs in column 3. Therefore, our guess at an FH was right, but this doesn’t link to confirming a particular EH. And it is ‘good hypotheses about lived experiencing’ with which we are concerned.

What was the interviewee experiencing at this moment of the interview? is the question, and, since the FHs are not related to particular EHs, neither the refutation of some FHs nor the confirmation of other ones brings us any closer to more knowledge about the EHs.

If there is no linkage back from a given FH to a particular EH, there is no point in putting forward a single FH and, even if one does put them forward, there is no point in checking back from later chunks as to whether an FH is confirmed or disconfirmed….. because it does not link back to strengthening or weakening any particular (EH) hypothesis about the (normally double) experiencing of the interviewee at particular moments of telling their story.

No linkage from FHs to a particular EH; no point (or very little) in collecting FHs at all. It is in the complex ‘lived experiencing in the telling of the story’ which reveals that complex situated subjectivity and FHs are only worth doing if they link up precisely to elaborated hypotheses about ongoing lived experience (the double experiencing of the past in the present) of the teller as they are telling. My argument: if there is No linkage from FHs to a particular EH; then there is no point in collecting FHs at all.

See textbook QRI (Wengraf 2001 p.278-9 see how more elaborated you can be about experiencing Hs, how more elaborated you need to be if you are going to generate for your audiences accounts of lived experiencing which are worth the elaborate and sustained effort of following the BNIM interpretive procedure.[3]

Compare Figure 20 above, with the ‘revision’ on the next page:

Line	Textsort	*Gist*	EH: Hypotheses about the experiencing of the interviewee (EH)	FH: Hypotheses about what might follow (FHs) if a particular EH was true
001	I: Initial Question	D1.Delivery of SQUIN Interested in young peoples’ experiences of care and leaving care	E.H.1.1. Uncomfortable about such an open question	FH 1.1.a Will Pause Unproven by D2 FH 1.1.b Will Ask what I wanted to know in particular Unproven by D2
			E.H.1. 2. Comfortable about being given such freedom	FH2.1.2a Will give chronological account of care Proven FH 1.2b Will start with an overall evaluation of the experiences and not a story
			E.H.1.3. Suspicious of motivations and interests of researcher	FH 1.3a Will ask “Why are you doing this research?” FH 1.3b. will ask “Who is going to read what you write? FH 1.3c. Will say “I don’t know if I’ve got anything to say”
			EH 1.4. Anxiety about capacity to give a good story	FH 1.4a Will say “I don’t know if I’ve got anything to say”

Figure 22 TFA panel record - more elaborate, more linked EH, FH

Please note that

(1) the EH column is not just marked by abstract unspecific subjective states with one-word descriptors like ‘uncomfortable’, ‘suspicious’, ‘anxious’, but more elaborate hypotheses about experiencing. ‘Uncomfortable’ might be uncomfortable because torn between wanting to talk about matter M1 but wanting not to talk about matter M2 which is very linked, and wanting to not think at all about how he changed his mind about M3….. etc. A complex ‘experiencing of uncomfortability’ which needs to be -- rather than alluded to with a one-word unfocused descriptor -- -- spelled out in the equivalent of a pararagraph a little more like the “suffered a bit, perhaps not suffered at all” example on p. 552.

(2) the FHs are precisely linked to the EHs from which they are derived

(3) the same FH might be put forward which links back to two or more EHs. That is, FH 1.3c and FH 1.4a might get ‘confirmed’ by a future chunk if the interviewer did go on to say in the next chunk ““I don’t know if I’ve got anything to say”. If the interviewee did go on to say that, we might say that EH 1.2. was not strengthened by this new chunk, but that we could not say that which of EH1.3 or EH1.4 were more confirmed than the other by this new chunk. It would have to be left open.

(4) The same predicted FH might relate to more than one EH, and then, were the FH confirmed at some future point, then this cannot on its own be taken to support any and only one particular EH.

(5) The confirmation or disconfirmation of an FH on its own is a loose not a tight indicator about the possible strength of weakness of any particular EH. But it is not irrelevant either.

END OF DIGRESSION ON ELABORATION OF ‘EXPERIENCING HYPOTHESES’ AND OF THEIR BEING TIGHTLY CONNECTED TO ANY ‘FOLLOWING HYPOTHESES’ PUT FORWARD.

Why is this an important point?

The point is not so much to put FHs on the sheets of paper and then tick them or put a cross against them as new chunks of data confirm or disconfirm them. That is a means to an end.

The purpose or end is for the slow development in the minds of the panel or panellists (namely the researcher and allies) of a guided complex sense of the interviewee’s likely general modes of experiencing their experience. We are searching to empathise more closely with better-imagined lived experiencing.[4]

The capacity to imagine the interviewee’s state of mind in the interview oriented simultaneously to the ‘talking in the interview’ event and to the ‘remembering past situated events and states of mind’ requires a capacity for empathy and identification rather like that of a historian, in fact, exactly like that of a (certain type of) historian. See discussion of ‘imaginative identification’ on p. 162, and Kate Grenville’s Search for the Secret River.

If the hypothesising in the Teller Flow (TFA) panel is not both ‘double’ and ‘deep’, then the structural hypotheses about the interviewee will themselves have to remain ‘simple and shallow’. This is an over-simplification itself!

[1] A ‘linked following hypothesis’ is one which hypothesises about the shape of next or future segments that might follow in the interview were the preceding speculative ‘experiential hypothesis’ true. It is a draft operationalisation of the ‘experiencing subjectivity-in-context hypothesis’ (this latter can only be tested indirectly by a risky prediction about as yet un-known data, i.e. by a ‘following hypothesis’, a hypothesis about what might ‘follow’ logically and historically) were the experiential hypothesis to be true.

[2] The important mechanics of such ‘registration’ are dealt with in Wengraf 2001, ch.12. The flow of chunks and of generated and/or supported and/or disconfirmed hypotheses needs to be registered on A2 flip-charts which get pinned up on the wall for all to keep the flow of panel thinking in collective sight and in collective mind and for ‘further noting’ (and memo-ising) as the panel proceeds.

[3] Froggett and Wengraf (2004) has an example of this sort of work in a discussion of research team (panel) dynamics. What is going on below the surface in the panel so as to produce the work of experiential hypothesising?

[4] The extended ‘expansion’ of one particular ‘Experiencing Hypothesis’ (about repressed anger and shame in the past leading to further self-repression in the interview: “I suffered perhaps not at all”) on p. 552 above represents a discussion in the pane;. On the sheet of flipchart paper, the facilitator can only write a few words to remind panel members during the further discussion (and herself afterwards) of a discussion and of their experience of that discussion. For example: “Anger and shame repressed at time; self-distancing in the interview” might be one (not particularly good) condensation of that discussion onto the flip-chart. A few words just to remind panel members.