oturn home > Theory of presentation: overview > §1 Presentation

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

1    Presentation

This work puts forward a theory of presentation and discusses its implications for the design of technical documentation. This first chapter presents the theory in overview and spells out problems that became visible in the research process. The subsequent chapters will expand and relate the main concepts of the theory.

Presentation is users' display of resources in response to situated needs. Users display resources so that they become available for understanding. Display is taken to include not just the visual, but other sensory spaces such as the audible and the tactile. The display is directly available for the senses to afford noticing, inspection or comparison. Resources may emerge in presentation as articulated rememberings, gestures, hypotheses, sketches or messages, or exist already as navigable documents or document systems, such as photos, technical drawings, reports, databases, or technical manuals in print or on-line.

The basic modes of presentation, then, are the articulation and the navigation of resources. Articulation describes the generation and display of new resources. Navigation describes the display of already existing documents. In emergent presentation, both articulation and navigation respond to some trouble. Service engineers, for example, navigate technical diagrams or manuals to deal with engine troubles arising in their work context. Simultaneously, they also articulate new resources, for example, through verbal or written hypotheses, questions, or requests.

As resources become available as display, presentation begins to aggregate its problem in resonance to the patterns of the referent domain. For example, a particular set of engine symptoms resonates on established domain patterns that highlight appropriate resources, such as fault isolation manuals, line station logs, or accounts of service personnel. The momentum of problem aggregation creates situated needs which become questions and hypotheses aimed at saturating currently relevant problem dimensions. As emergent presentation aggregates the problem, its course accumulates a problem pattern which over time merges with domain patterns.

The ultimate aim of presentation is ending presentation: reaching close-out according to appropriate validation contexts. The validation context is a product of the domain: it defines what can count as understanding. In the service engineering domain, the validation context is grounded in the operation of equipment; in other domains, it may be negotiated as part of the emergent presentation setting.

One distinction that will reappear throughout this work is that between emergent presentation and schedules.

Emergent presentation is driven by situated needs and activities of users who draw on the repertoire of available resources. In contrast, the schedule is driven by an encoded accumulative order, as in plans, scripts, lists, curricula, video or animation. Here, one segment of users' overall resource is defined in advance: the document system display is encoded as an ordered, sometimes even paced, schedule.

The agency of the schedule competes with that of the user: while the former accumulates content, the latter aggregates a problem. In emergent presentation, users' aggregate the problem as it emerges in recursive contact with the referent domain. In contrast, the schedule pre-determines topics according to predicted relevance. While the schedule constrains a segment of the display, users' understanding hinges on their response presentation which attempts to understand the schedule's topics as solutions to implied problems. The success of such response presentation hinges on the interruptability, and possibly, re-ordering of the schedule [1] so that it becomes flexible enough to respond to emergent breakdowns or opportunities.

In presentation, navigation and articulation are often confluent. Confluence affords generation, comparison and qualification of resources. For example, while reading reports or inspecting diagrams, service engineers generate to-do lists or compare and qualify problems through talking to colleagues. Confluent articulation and gesturing qualify displayed resources in scheduled training sessions. Confluence extends to the concurrent presence of people and documents in meetings.

As presentation finds resources lacking, users articulate resource requirements iin critique, imagination and user design. Authorised design can start here, but usually takes in a range of other requirements too, which concern the business, the department, the customer, or the regulator. Design turns predicted requirements into a fixed resource architecture. Any spatial architecture favours some requirements over others by deciding which units are within the view boundary, which are directly adjacent, which are more remote, and which are not available at all. The fundamental problem for authorised design is that it has to predict presentation needs in a future setting which will be affected by the introduction of the resource. The problem of prediction is aggravated in the design of encoded schedules, for example, animation, since beyond spatial constraints, the schedule's pacing and accumulative order constrains the temporal dimension at the cost of emergent choice.

1.0.1    Other meanings of presentation

A note may be in order to relate the use of the term presentation in this work to other common understandings. Like information, presentation is often used to refer to some material display, screen or page, static or scheduled. This view of presentation appears in conveyor-belt metaphors that pervade the literature on multimedia design [2]. According to this view, presentation transports ‘information’ from source to target, from resource to user or audience [3].  This view of presentation, which makes it something outside the user, implies that ‘presentation’ can be designed as an external resource.

In contrast to this view, this work suggests defining ‘presentation’ as ‘a process generating the presence of objects’. In this rendering it is clear that this presence cannot hinge on the display alone—its locus must be someone's presence of mind acknowledging the displayed object as present [4]. Any implication of meaning implies ‘presence-for-someone’. However, it should be noted that this definition of presentation does not grant ontological preference to the perceiving subject or the mechanisms of perception. Presence is socially transmitted and depends on social negotiation even where an isolated individual confronts the resource.

1.1    Aims, sources and outcomes

The aim of this work is to give designers of on-line technical document systems a workable theory which articulates the relation between designed systems and other resources that have a part in users' presentation.

Usually, diverse theories cover partial aspects of presentation, for example, theories of perception and cognition, problem solving, document and display design, learning and instruction, mental modelling, information retrieval, organisational behaviour, and so on. Here, the attempt is made to see these aspects as confluent within the overall activity of presentation, and to explore their interdependencies. The theory has implications for document design, particularly for the design of scheduled animation sequences (cf. section 8.4 Schedule design in chapter 8–Design).

The aim of the prototype that accompanies this work is to demonstrate how animated images and diagrams can function as rich views and interfaces at the same time. Perhaps most importantly, the prototype demonstrates a way of including animated sequences within the framework of a node-based on-line document.

The theory of presentation has two different sources. The first source is the ‘natural’ domain of a large engineering company, Rolls Royce; the second is the ‘artificial’ domain of evaluations of the cinegram prototype [5] with a range of users. While the field work captured data by observing the ‘naturally occurring’ activities in the Service Engineering department at Rolls Royce, the design and evaluation of the cinegram generated data in a new and artificial evaluation context with its own rules and emergent behaviour.

The novel outcomes of the research project are (1) a grounded [6] theory of presentation and its import for the design of animated node-based document systems; (2) the cinegram, a prototype of a diagrammatic and animated on-line document; and (3) experiences and recommendations concerning the application of grounded theory method for documentation research and design.

(1) The theory of presentation situates documents as one resource among many in users' settings, and relates existing navigable resources to emergent resources that users articulate during presentation. This link between navigation and articulation points at the importance of mediating—technically linking—between users' occasioned design and authorised document system design. The application of the theory to the design of authored schedules (in this case, animation sequences within node-based systems) has highlighted the importance of segmentation (for example, determining a useful access granularity, predicting possible effects of projection between confluently displayed events, or predicting adequate points of entry into the event context on the level of the schedule as a whole and that of the stand-alone segment), and the problem of integrating schedule interfaces within the spatial interface of a node-based architecture, i.e., designing the boundary and interactions between timed and spatial context.

(2) The cinegram shows the oil system of the Trent 700, one of Rolls Royce's latest large commercial turbine engines. My own design activity recursively embodied my changing understanding of users' problems and my changing idea of documentation design. The material result, the cinegram prototype, allowed a recursive evaluation and analysis of use situations, and through the medium of use, critical evaluation and change of design decisions.

The novel aspect of the cinegram is its use of animation both as a local dimension of individual ‘looping’ nodes which afford rich views of individual system components, and in the context of schedules which use several animated nodes as segments in order to illustrate functional or procedural dimensions of the referent system. A CD-ROM documenting the latest version of the cinegram prototype is provided in the back sleeve of this dissertation.

(3) To my knowledge, the grounded theory method has not yet been applied to the domain of documentation research and design. Grounded theory is an import from social research. Some authors have recommended its application to digital library design and in information studies, and some recent practical applications exist in related areas (cf. section 2.2 Grounded theory in chapter 2–Methodology). Experiences and recommendations concerning the practical application of grounded theory method can be found in subsection 9.3 Comments on grounded theory method in chapter 9–Conclusion.

1.2    Problems

The theory of presentation deals with two clusters of problems: those related to emergent presentation and those related to schedules. The focus on schedules is a result of my interest in animation within node-based systems at the outset of research.

The problems which are here summarised were not conceived as hypotheses at the outset of the research programme. They crystallised as a result of the research process and grounded theory generation. They are not treated separately but appear in many facets throughout the theory. Chapter 8–Design will also indicate design solutions.

1.2.1    Emergent presentation

Problems of emergent presentation can be summed up under four points: identifying appropriate resources, identifying descriptors for appropriate points of entry, difficulties in achieving confluence between different resources, and uncertainties in defining the appropriate validation context.

(1) Resource identification. This concerns the linking between emergent presentation triggers, such as sets of symptoms, and appropriate resources. Given an emergent need, the first problem may be knowing that an appropriate resource is provided, then, where it is located, and then, how it can be accessed. This problem is more acute in emergent settings without strong domain patterns.

(2) Descriptor and point of entry. Once a resource is identified, the difficulty is turning the trouble into a descriptor which matches terms of the document system and leads to the right point of entry. What does the document system need to know about users' need to know? Often there is no single clear-cut symptom which would suggest the point of entry into the document, but rather a multiplicity of symptoms and possible causes ‘which makes straightforward troubleshooting by the book impossible.’ An example is an airline calling with a fault: ‘We've had LOP and HOC, now we have found a dodgy seal - can that have caused it?’ [FN 4/7/95 P48] If symptoms appear to be correlated, which one should users treat as primary, opening a segment of the search space, and which ones as secondary, qualifying the search within the chosen segment?

(3) Confluence. The use of material resources for emergent needs is a secondary process that suspends dealing with the primary trouble. The navigation of documents often suspends the articulation of the problem. How can presentation engender confluence of appropriate material and transient resources within situated temporal tolerances?

(4) Validation context. This relates to the conditions for presentation close-out. How can users tell that they have solved a problem? How can they tell that there is not more to be found which might reveal important problem dimensions? How can they tell that what the resource offers is indeed valid and appropriate, i.e., saturates the emergent need?

1.2.2    Schedules

The design of a schedule embodies a prediction of presentation in the acumulative order and pacing of its content. This creates an agency conflict which must constantly be negotiated between schedule and active user. The agency of the schedule rests in the selection and order of content, and, in the case of timed schedules such as training sessions or animation sequences, in its pace. The schedule embodies the instructive activity of designers who supposedly know what is needed and what should come next. As a process is cast as schedule, it treats users emergent understanding as the noise of reception.

On the other hand, the agency of the user follows situated needs which arise in the emergent process of presentation and negotiate processes in the setting and the referent domain. The tension between the momentum of emergent presentation and the predictive nature of the schedule surfaces in breakdown—both in the breakdown of schedules and in the breakdown of users' emergent understanding. The context provided by plans and schedules often becomes irrelevant or even misleading once a disturbance and its knock-on effects takes the plan off course.

The problems for schedule design all relate to prediction and can be divided into three different aspects: prediction of problem relevance, prediction of accumulation, and prediction of response presentation.

(1) Prediction of problem relevance. The schedule design for complex domains may fail to predict a meaningful fit between preconceived schedule and emergent need since it cannot account for contextual and situated dimensions of problems.

(2) Prediction of accumulation. Schedule design attempts complete coverage of a selected process, scenario, or problem type. It often relies on sequential accumulation of meaning so that a particular segment in the middle of a schedule can be correctly understood only if the user has seen and understood the preceding views or segments. However, the schedule cannot predict users' situated navigation strategies, e.g., their points of entry and suspension and their skimming and skipping decisions. The problem is aggravated in paced schedules since their agency carries on regardless of users' ‘window of attention’, i.e. the time during which users actually pay attention to the schedule.

(3) Prediction of response presentation. Users' response presentation aggregates meaning at its own pace dependent on emergent understanding. The form of the schedule implies that its designer implements predictions as to users' probable needs by providing selected content at selected times. But the schedule cannot predict the contingencies of use which determine the point at which users will want to override the schedule and take over agency. To the extent that schedules substitute emergent action with set procedures, they can even weaken users' ability to cope with the contingencies and unexpected contexts created by breakdowns [7].

These points explain why scheduled resources are particularly prone to situated misunderstandings. Applied to the design of animation within node-based on-line documents, they indicate techniques that retain the richness and internal differentiation of time-based displays while avoiding the disadvantages of schedules by maintaining independence and integrity for every independently addressable document node.

1.3    Outline of this work

Chapter 2–Methodology accounts for the choice of grounded theory method and describes its basic operations with examples of its application in this work. It outlines the field work carried out at Rolls Royce and describes the relation between theory development and design practice.

Chapter 3–Context covers the patterns and protocols that trigger and constrain presentation. By comparing the service engineering and the evaluation domain, it describes the role of validation contexts, business processes, diverging presentation interests of domain and users, the make-up of stable and emergent settings, the power differential between participants, and the factors that contribute to the weight of presentation.

Chapter 4–Problem describes the aggregation of the problem from the initial trigger until presentation close-out. It introduces the concept of resonance that turns resource units into tokens for aggregation. It looks at the problem pattern that accumulates through presentation. It shows how resonance on problem and domain pattern produces projection into the display that explains situated misunderstandings.

Chapter 5–Articulation shows how aggregation produces new transient and material resources. After discussing transient articulation and its hypotheses, the chapter deals with transformations between technical and human settings that generate material resources. This leads to the concept of reference distance between root and derivative resource and the effect of the form of the resource type on the selection of content features in transformation.

Chapter 6–Navigation describes three different levels of navigation: delegating presentation between settings, switching between resources or views , and noticing the display. It develops the concepts of navigation distance and navigation pleasure which both affect users' choice of resources.

Chapter 7–Confluence discusses the confluence of articulated and navigated resources in the display within the temporal tolerances of the setting or the referent domain.

Chapter 8–Design shows how conflicting requirements turn into specifications and a resource architecture which forces boundary decisions on all levels of resource. It then discusses the agency conflict between user and schedule. After a description of the solution to schedule design chosen in the cinegram, the chapter turns to design problems relating to the schedule interface and schedule segmentation.

Chapter 9–Conclusion extends the scope of presentation by looking at the conflict between emergent presentation and the world view implied in protocols and schedules. It then identifies worthwhile areas of research into on-line document systems and schedule design, and ends with comments on the grounded theory method.


Footnotes to chapter 1Presentation

[1] John Lindsay pointed out that the possibility of re-ordering a stack of tasks (shopping, picking up the kids, etc.) is a decisive advantage of cars compared to the fixed schedule of public transport (personal communication).

[2] A recent example is Bergan (1995) who lists ‘increased bandwidth of communication’ as the first benefit of the multimedia approach to human computer interaction. He continues: ‘Presenting information through multiple media channels in many cases enables more information to be conveyed than is the case with a single media channel.’

[3] A contrasting view can be found in Lindsay (1994), who argues that ‘information’ is not an object, but a process between user and system.

[4] I ignore at this point the difference between individual and shared awareness, a point to which I will return in section users and settings in chapter 3Context.

[5] Cf. Glaser & Strauss (1967), Glaser (1978), or Strauss (1987). The grounded theory method is described in chapter 2Methodology.

[6] The prototype evaluations that were carried out in the field can be seen as qualitative experiments (cf. Kleining 1994 pp148; Moser 1995 pp130).

[7] A related problem is the often-voiced concern about automation of complex control tasks which turn the operator into a mere monitor. The passivity of monitoring replaces the experience of feedback that is necessary should the system break down and control be transferred back to the operator. For an impressive list of related concerns, see Funk et al (1996), ’Perceived human factors problems of flightdeck automation’.

Last update: 08 November 2007 | Impressum—Imprint