oturn home > Theory of presentation: overview > Appendix VI-Grounded theory applied

A theory of presentation and its implications for the design of online technical documentation
©1997 Detlev Fischer, Coventry University, VIDe (Visual and Information Design) centre

Appendix VI-Grounded Theory applied

The aim of this appendix is to illustrate the grounded theory development process outlined in chapter 2–Methodology. It provides samples of the material at various stages of the grounded theory process. The entire grounding process of the theory of presentation is too complex to be traced in detail. This appendix will therefore concentrate on one conceptual strand which resulted in the category of confluence (cf. chapter 7–Confluence). By reproducing excerpts from documents, I want to trace the emergence of this category from initial codes and their refinement and change in memos until to the final concept. The authors of the documents are indicated in the excerpts' headings.

The initial codes were chosen in the ‘open coding’ of reports written by the author and other witnesses (cf. subsection 2.2.2 Coding in §2–Methodology). Often, tentative codes were placed in the margins and further elaborated in memos. In other cases, parts of documents were subjected to line-by-line analyses. As the examples will show, such marginal coding and line-by -line analyses tend to produce codes across the whole spectrum of the emerging theory —homing in on a single code and relating it to other codes is done in memoing.

Sources and coding

The first sample shows a service engineer troubleshooting a problem. The example was caught through eavesdropping from a neighbouring desk and concurrent note-taking. The same evening I filled in missing bits based on abbreviations and my remembering of the event. I reproduce this excerpt here because it shows the type of confluent presentation drawing on multiple articulated and navigated resources that is characteristic for the domain of Service Engineering:

Excerpt from field note FN 17/5/95 (Detlev Fischer), 17/5/95

(P7) [Telephone] call about further problems with radial shaft O-Rings. J. takes the call and passes it on: ‘I hand you over to A., our new man.’

(P8) A. takes the call. ‘So you say 0.18?’ Makes calculation, says ‘Right, that's on the low end’, confirms: ‘Its not a happy one…right, we have to find some old seals. There is obviously very little you can do.’ Old standard parts are needed, ‘otherwise you have a problem…it looks like an ongoing bit of a glitch.’

(P9) A. recommends a ‘volume check’, explaining the relation of diameter and volume. ‘I'm pleased you see it as lucky, I wouldn't see it that way’…(listens)…‘as far as you reckon they are below the minimum targets’…(calculating)…‘For your information the tolerance on the drawing is from .015 to .0122…it seems to be that that's causing the swelling problem…’

(P10) He's on the phone, has a schematic, folders with cases?, an IPC?, a maintenance manual on his desk, makes notes, uses the calculator. ‘I left a message with that guy, if he is not coming across, I'm going to chase it up.’

(P12) A., right after the phone call, informally reports the problem to J.: ‘…a drawing alteration, problem is possibly a different mould of o-rings, different diameter of O-rings…W. has 19 cases which are all on the minimum, and two engines are waiting…’

(P13) A. then talks to someone else who stops by: ‘PTF - I spoke to him once, he promised to get back to me, but didn't ring back; I called again yesterday, but then the phone kept ringing …there must be something similar on some other engine. [Someone else is] probably the right guy to talk to…these things [O-rings] have a limited shelf life…if they send out the new stuff they must have sent out [all] the old stuff…none of these dimensions mean anything in isolation…’

(P15) 10 - 20 minutes later, A. reports to some one else he has consulted to sort out the O-ring problem: ‘A minor change of dimension 'escalates into a top priority problem.…Well, that's what I was told a minute ago, I believe it's true - we first have to look at the manufacturing process, [see if the parts are OK], make a volume check.’

The excerpt shows the confluent use of resources in presentation. The user navigates on the levels of focusing (on schematics), switching (browsing a manual) and delegating within temporal tolerances (‘I left message with that guy, if he is not coming across, I'm going to chase it up’). The example also shows the difficulties of achieving a confluent display of needed resources (‘he promised to get back to me, but didn't ring back; I called again yesterday, but then the phone kept ringing …’) . While navigating, the user is engaged in confluent articulation, both as transformation (he produces new resources with calculator and note pad) and as emergent and transient articulation of hypotheses (‘it seems to be that that's causing the swelling problem’). The example shows confluence as necessary condition for successful presentation (‘none of these dimensions mean anything in isolation’).

Excerpt of field note FN 18/7/95 (Detlev Fischer), 18/7/95

(P11) One problem with DRUID is that it is all on one (or two) two disks, so anyone wanting to look at something ties up the entire ATA system. They would need more work stations to make it workable.


(P17) ‘Highlight communication [a Rolls Royce-internal publication] does trigger communication, but its not enough—often when people are called in serious cases to join a task force, you hear "wait a minute, we had this problem for two years but you never contacted us."’

(P18) In J.'s section they support the RB211-524 and 535 and the Trent, 'so day to day I would know [what happens]’. People talk when the desks are close. ‘I reckon if I were just sitting around the corner, I would already miss out 75% of what's going on…If I were sitting at another site, then immediately I would know nothing about what's going on.’

At the time of writing these field notes, neither the category of confluence not the core category of presentation existed as explicit codes. A lot of codes appear, change, and merge before the final terms emerge during memoing and sorting. The same excerpt from a witness report of a dialogical evaluation session which I coded once soon after the session and a second time half a year later may illustrate the development of concepts from initial codes and the re-evaluation of the source as the theory takes shape.

Excerpt of witness report LF 3.2 (Simon Turner), 3/5/95 with marginal coding (Detlev Fischer), 9/5/95

(P3) Q1 Why are there two filters?  

(P4) L. simply looked at the Overview  
and tried to answer the question  
on the basis of what he knew,  
and what he saw in the Overview.    
In fact he assumed that the filters were 
before the internal and stepaside  
gearboxes, which isn’t actually the case
He reasoned that there would be 
filters before each gearbox because  
these are most susceptible to damage  
from debris. It seems that his initial  
reaction was to respond to the question  
as a challenge - to guess from the 
general layout rather than interrogate 
the cinegram. 

surface effect 

background knowledge

viewed resource


witness knowledge


remembered resource


document independence

Here, open coding simply tacks labels to the source data. ‘Surface effect’ and ‘document independence’ will be explained as an effect of the evaluation setting: the initial lack of protocols for navigation. ‘Background knowledge’ and ‘viewed resource’ will lead to pattern and display. ‘Assumption’, ‘conjecture’ and ‘remembered resource’ will lead to hypothesis and projection based on resonance. ‘Reasoning/reproducing’ foreshadows articulation. ‘Witness knowledge’ describes the impact of the evaluator's presence on users' presentation. Here, the statement of the witness is actually wrong: there are filters before the gearboxes, and they are shown in the overview diagram.

The same report was re-coded in November '96:

(P3) Q1 Why are there two filters?

(P4) L. simply looked at the Overview  
and tried to answer the question  
on the basis of what he knew,  
and what he saw in the Overview.    
In fact he assumed that the filters were 
before the internal and stepaside  
gearboxes, which isn’t actually the case. 
He reasoned that there would be 
filters before each gearbox because  
these are most susceptible to damage  
from debris. It seems that his initial  
reaction was to respond to the question  
as a challenge - to guess from the 
general layout rather than interrogate 
the cinegram.

given resource







Re-coding now captures the confluence of resonant pattern and navigated display. Resonance generates ‘expectation ’and projects ‘noticing’ of the display so that both ‘match’. New aspects of the source take on meaning, for example, the fact that the user sees the small inlet filters that saturate his expectation, and ignores the much larger main filters which the question refers to. The term ‘given resource’ now emphasises the confluence of displayed and remembered resources in presentation. ‘Saturation’ and ‘offer’ will feed into the concept of close-out.

The next excerpt from a recorded dialogical evaluation session with a third year engineering student is followed by excerpts from its line-by-line analysis. Although confluence is still nowhere mentioned in the analysis, it becomes clear that presentation is driven by the momentum of aggregation and the continuous confluent and mutually corrective interaction between articulated and navigated problem aspect. The analysis also resonates - or rather dissonates—to the scientific context of mental representation (cf. footnote 6 on mental models in chapter 4–Problem).

Excerpt from a recording of the dialogical cinegram evaluation session MK 3.3, 18/7/95

D: evaluator (Detlev Fischer)
M:  user (Mark Kingdom)
[x]: pause of x seconds
[text]: partly audible, guessed from context
_text1_ , _text2_ : overlap

(P9) D: OK the first question, or the first task you could sort of take is—would be ahm: 'Why are there two filters in the oil system, not just one?'

(P10) M: Right. [2] So what do you want me to do, give you my opinion - or -- some [inaud.]?

(P11) D: Ahm - That's one of the interesting points anyway, to see how much you can actually fill in from your background knowledge_

(P12) M: _right_

(P13) D: _you know, you may not need to use it, and if you don't want to use it for answering a certain thing, if you are confident saying: 'That's why' , then_

(P14) M: _right_

(P15) D: _then that's fine_

(P16) M: _right_

(P17) D _then just tell me why.

(P18) M: Okay. I don't [really] want to do it that way, I want to do it using your system, because basically I don't know anything about the oil —Trent system anyway.

(P19) D _Yeah

(P20) M: [inaud,.] So…

(P21) D: But you would have a guess about why there are two filters, without_

(P22) M: _Ah, yeah! [3] ahm [4] Not really, actually, no! [Why don't I have a look at that] thing.

(P23) D: Filters

(P24) M: Yeah. It could be because of failure, if one clogs up

(P25) D: hhm

(P26) M: then at least you've still got a second one that will ahm —filter out the system …although that's only if they are working in parallel, if they are working in series it could be because one will take away the larger contaminants and one will take away the smaller contaminants afterwards.

(P27) D: Yeah

(P28) M: I don't know - they're my two guesses! (laughing)

Excerpts from line-by line analysis Micro MK 3.3 (Detlev Fischer),b30/6/95

(P21) D: But you would have a guess about why there are two filters, without_'

(P22) M: _Ah, yeah! [3] ahm [4] Not really, actually, no! [Why don't I have a look at that] thing.

(P23) D: Filters

The attempt to answer this question straight away demands a 'big step' - a 'spot-on' identification of the correct answer assumed to exist by virtue of the formal closure of the question. But the step is too big; the user feels he cannot take such step now, neither can he see, within the pace of the dialogical session, another spot where he could find a foothold to move closer. The pressure of the dialogical engagement implies that he must say something, so for the moment, he forsakes the answer, stepping out of the spotlight, commenting on himself in disarming honesty. (So this is a fight after all?) This instantly frees him to think about the question without the pressure to succeed, and to venture a general guess 'off record', which simply moves him into the right direction. It is a small step instead of the impossible big step.

What the analysis has captured here would later be conceptualised as a lack of confluence between articulation and navigation. ‘Foothold’ and ‘step‘ foreshadow the concepts of focus and navigation distance. The ‘pace’ of the dialogical session indicates the temporal tolerances of confluent resource use: the momentum of transient articulation of hypotheses under the pressure of the evaluation setting is in conflict with the time needed to navigate the display and aggregate the problem. (Later, as the session progresses, the user will take all the time he needs to navigate the display).

(P24) M: Yeah. It could be because of failure, if one clogs up

Modelling has started with naming the focusing point and likely failure event suggested by the question - imagination now begins to sketch in the connections, events, direct consequences of events, and propagation of effects outwards from the initial focusing point. The sketching thought moves along its own construction while modelling, rather then holding fast 'the model as accumulated ideational construct'. If it hits upon something which has the scent of relevance (in the light of the emerging answer space?), imagination turns into verbal account, which again brings the modelling process into relief, and as external production, becomes amenable to qualification, correction, and refinement

The ‘imagination’ that begins to ‘sketch in’—project and aggregate—‘ connections…outwards from initial focusing point’—the current focus—will lead to the concept of resonance of the focus on the emerging problem pattern—here still called the ‘answer space’—and domain patterns (here, the generic engineering knowledge of the user). The momentum of this ‘sketching’—the user's problem aggregation—has temporal characteristics which will later be described as temporal tolerances of confluent presentation. ‘If imagination hits on something with the scent of relevance’ —the result of resonance —it flips over to ‘verbal account’—articulation of a new transient and confluent resource which becomes addressable for future articulation.

(P25) D: hhm

the communicative encouragement [week confirmation] given by the evaluator encourages Mark to carry on along the lines of redundancy design. So the speed or pace of the surface production may be dependent on the user's reading of the evaluators reaction (down to body language). What would have happened if the evaluator had stayed silent?

Here, the evaluator is recognised as a confluent resource with an impact on the user's presentation. The nature of the ‘pace’ which here seems dominated by the evaluator's actions and reactions will be conceptualised as the pace of the schedule in training sessions or evaluations which often lacks confluence with the pace of users' emergent response presentation. ‘Surface production’ will become the articulated display.

(P26) M: then at least you've still got a second one…

the use of the article ‘one’ indicates that the pair of filters is established as main focus around which the rest of the system is sketched in. A structural account would describe this as the concept of reserve filter; but it is not clear if the concept emerges on the surface, as a described activity, or as an elaboration of a [pre-existing] structural concept…But the user starts with ‘then at least’ which indicates [that] he first sees the blockage happening and switches from here to the ‘at least’ remedy of the second one- which could potentially clog up as well. One and the second one may indicate that at this point, the user thinks of a simple model of both filters one after the other in one line. This would be emergent replacement modelling, in other words, there is no clear concept of what it would imply if both filters were fitted either in line or parallel (with some branching apparatus). The emergent account breaks down as soon as the consequences of the blocked filter in the same line become apparent:

(P26) …that will ahm - filter out the system

the hesitation shows that the user discovers a contradiction in his produced surface level argument, and that he is now fishing for a better explanation. The foothold may be the latest unit of surface production, with older statements fading away quickly when not constructively used as footholds in the ongoing argument. Focus [on] the field of view, or head lamp of thought where emergent sentence production comes to light when speaking and thinking become slightly asynchronous.

The contradiction—dissonance—appears because the thought is articulated and thereby enters the confluent display so as to allow comparison. The ‘fading’ of ‘statements…not constructively used as footholds in the ongoing argument‘ directly leads to the notion of transient articulation, a transience determining the temporal tolerances for confluent presentation. The dissonance relates to the asynchronicity noticed in the analysis which points at a difference of mode: while the articulated resource is paced by the rules of language, resonance in comparison and on memory is immediate.

(P26) although that's only if they are working in parallel,

This is a simple improvised remedy to the impossibility of using the second unblocked filter in the same line as the previously blocked filter unless they are bypassed - a concept that would solve the users logical problem, but one which so far (although certainly known) has not been considered. What happens here is the discovery of combinatory variability afforded by the theme of two components in a flow system which could theoretically be in line or in parallel - but this does not necessarily imply that the user is actually making a conscious choice between these two options. He tries the simplest arrangement, that applied as explanation produces a contradiction, which then triggers an alternative combination. There is no reason to believe that this range would exist in a structured model-like fashion. At the moment when the world parallel appears, there is no sign of an underlying appreciation or hypothesis of the routing mechanism which would be needed to guide the oil stream through a parallel filter if the first filter gets blocked. Language seems like sketching with a rapidly disappearing ink.

(P26) if they are working in series it could be because one will take away the larger contaminants and one will take away the smaller contaminants afterwards.

this is an alternative correction of the hypothesis of consecutive blocking, perhaps because there is something unsatisfactory about the parallel arrangement—it could be that from the foothold of the statement if they are working in parallel the user can already briefly sense on the horizon problems of having to elaborate the still unknown apparatus which would make that assumption work, (i.e. render it free of conscious contradictions). So there is no model of apparatus, but a foreshadowing of problem [and of] still unspecified complexity? Also, it increases the offer to the evaluator by rescuing/repairing an earlier suggestion. The user introduces an order of first, filter out large particles, then, small particles. It would make little sense [the] other way round. This is probably conventional.

Another example of line-by-line analysis refers to the following excerpt from a witness report by Sue Flynn covering a session with two users in stage 2 of the cinegram evaluation:

Excerpt from witness report DB/TR 2.2 C, (Sue Flynn), 12/3/95

(P2) Q1 What is the purpose of the oilsystem?

They systematically investigated the ‘overview’ diagram in order to decide the purpose of the oil system. They scrolledthrough the text field and read the text. (overview —> text, scrolled.)

I reproduce the analysis of just one line of the report in order to show how many dimensions can be unearthed when investigating the source carefully. First, the analysis turns to users' mode of investigation:

Excerpt from line-by-line analysis Micro DB/TR [2.2], 2/4/95

(P2) They systematically investigated


[investigated:] possibly a method, a heuristics, probably dialogical 1) between document and users, and 2) between [two] users;

How could that dialogical process evolve?

# seeing/picking an icon, either alone or supported through communicative #gesture like pointing; #looking at it before or after # reading its label; # inferring meaning from situated label connotation back to icon; # discussing the meaning (communicative validation), or: one of the users feeling or being more confident and/or assertive than the other decides the matter provisionally and on the fly - meaning '‘until further notice’; then, # proceeding to next icon, possibly adjacent, and doing the same thing, but possibly helped by the patch of meaning already created, and so on? Has this description anything to do with what 'actually' happens? Introspection confirms this.

The analysis develops the confluence of actions (each marked by an asterisk) that take part in investigating the overview diagram; this will lead to the concept of display confluence bridging navigation gaps between the target tokens of noticing and switching, and users' confluent articulation and gesturing. The ‘patch of meaning already created’ will later be called the emergent problem pattern to which the focus of presentation resonates.

(P2) They investigated … in order to decide

the order is implicit in the setting and 'administered' set of questions/tasks; they have to decide; there is a feel of: 'they couldn't care less, but this is what is asked, so get on with it'. Which points at Simon's ATA (Awareness of task artificiality). This part of the analysis relates to the presentation context: the setting, the situational protocol, and the power differential.

(P2) They investigated the…diagram…they scrolled through the text and read the text

They first focus on the diagram, then read the text; they have to manipulate (scroll) the text in order to read it, the diagram is just there: they investigate the diagram, but simply read the text. This sound as if the diagram affords a more 'interactive' mode of use; they seemed to be discussing the overview diagram. How did that work? Did they focus on one component which poses a question in its isolation, 'begging the question', which is then answered by further looking at adjacent and other components, which generate further questions? The difficulty of establishing what has happened in the session can be contributed to the granularity of observation ( recording through note-taking) and the reference distance between the source document (which was written by another person) and the activities it refers to. The analyst has no confluent memory of the situation which would help clarify ambiguities or increase the granularity of the description—drawbacks which have led to the choice of a dialogical evaluation method (cf. section Stage 3–dialogical evaluation in appendix V–Evaluation).

(P2) they investigated the 'overview' diagram

- Is there a correlation between the overview-nature of the view and the method of investigating? Could one investigate a component view in the same manner?


The concept of confluence appeared in memos as dimension of articulation and navigation before it became a category (and chapter) in its own right during the sorting and writing stages of theory generation. To conclude this appendix I will reproduce excerpts of several memos developing some aspects of confluence.

Excerpt from memo Articulation 7/12/95

One aspect is the confluence of the different presented media in the process of articulation, which includes references over time (deictic language, and physical pointing & gesturing), comparisons (document —document', account—document) and negotiations (account-account, plan-plan—that is, potential action).

So there is (1) transient (oral/gestural) and (2) permanent (writing, sketching, re-locating, etc.) articulation: the latter falls into production and conversion/transposition (copying, print-off, submission, clipping, summarising, re-filing)

Another aspect is where this articulation is materialised concurrently or in consequence by the production of permanent resources, for example:

An articulation of resources is resource conversion, e.g., re-filing, filtering and aggregation of data to produce other resource types or aggregates - this does not increase factual knowledge, but changes the level of accessibility. Articulating resources = turning to usable form, an available resource stored in unusable format (microfilm); copying/clipping what is hidden in journals; e.g. data-mining

Excerpt from memo Focus < resources (Detlev Fischer),13/12/95

Focus and potential resources

In general resources are liked to:

……? hhm…

(The counterexample is that Service Engineers talk to the Reps rather than to the airlines apart from critical cases—is the issue sensitivity and adaptation to customer's [organisational] climate? Surface changes to documentation in order to avoid hurting the customer's feelings point in that direction)

Presentation trajectory (history) (Detlev Fischer ), 5/1/96

Presentation theme

Once the process [of presentation] has started, users have a theme, often with a descriptive template e.g., ‘cold start’ [MK 3.3. P126] which is like an emitted pattern, creating visibility making them able or unable to notice a matching token in the surroundings.


The theme that the instance [of trouble] suggests is put into relief by a consonance between instance and pattern (a fit, leading to structural coupling): what is lacking is highlighted by a level-out of instance and pattern and leads to factual requests; what is problematic about the instance is brought out by interferences between instance and pattern, i.e. by those patches where instance and pattern fail to meet or fit.

Is ‘presentation trajectory’ and ‘extended event space’ same as ‘answer space’? ‘Space’ has no temporal connotation. ‘Extended event space’ includes many events which might overlap and interact and have their respective expected future and remembered history. It allows for the notion of confluence between some events/activities, or alternatively, shifts of attention between simultaneous but separate activities.

Confluence (Detlev Fischer ), hand-written 3/7/96 and sorted within Navigation

Confluence of document and documented object:

there is oscillation between two processes, e.g., installation and documentation of installation. Resuming-suspending: with each resuming, a reorientation finding the paragraph where I left off, finding the next part to assemble.

Three parts (in Trent installation): Document, equipment, label on equipment. (Finding problem with large set size)

[Confluence on] screen [vs.] desk

The screen forces within-document replacement. The desk affords extension of real estate. [The user in] book navigation can have fingers between pages; a limited [number of] bookmarks, but near-confluent.

The system has the logic of trail instead, provided the label still makes sense.

The screen has between-document obliteration - less confluence than desk comparison.

Articulation - Navigation confluence brings in the difference between emergent theme trajectory and given system architecture. The theme produces the needs for reference/proof) and thus occasions navigation; at which point time scales become important. When the theme is a co-operative production; its trajectory and tolerances depend on two or more users. Context contributes to theme window and pressures/urgency which filter resources according to availability. Better a bad resource right now than a perfect one in a minute, or in a week. Particularly when navigation forces suspending articulation (fetch a file) or put the situation in a wait loop ('just a sec, just starting up the system') and the new display forces a re-tuning which is disruptive, particularly when it is not available for both (all) participants. and I will quote Idea of display encompassing the current document (system) vie but also many more resources / navigation as within display and switch-delegate / distance as technical and noticed / resonance on patterns affording projection that can bridge the navigational gap.

An important tool for theory development was the drawing of conceptual diagrams, a kind of visual memoing. The diagram in figure VI.1 sketches the confluent activities of course schedule, trainer, and users (trainees) during a lesson at Rolls Royce's Customer Training School.

Conceptual diagram showing confluence during a training lesson

Figure VI.1. Conceptual diagram showing confluence during a training lesson.

Last update: 08 November 2007 | Impressum—Imprint