Perspective differences at the Postal Giro

When we enter a large organization like the Postal Giro, we will be exposed to a large inhomogeneous mass of written and spoken language: we listen to the speech of workers engaged in work, interview them and talk to them, read manuals and look at the screen images they work with, and talk to management in offices removed from the shop floor. Although many of these verbal exchanges clearly refer to the same physical phenomena, the linguistic forms can be quite different. It soon appears that each linguistic form is adapted to the particular function the speaker fulfills in the total structure of work processes and power relations in the organization. Each sub-language expresses a particular perspective (on perspective, see Section 1.2.2.2) and reveals a partial truth about the organization.

In the development of the PGP system there was one major issue that in a nut-shell illustrates the intricate interplay between language, perspectives and position in the division of labor. In semiotic terms, the issue was simply the relation between signifier and signified, in this specific case taking the form of a disagreement between union and management about electronic scanning of paper forms.

Management wanted to get rid of all paper by optical scanning of the paying-in-forms and wanted all subsequent work involving the cards to be done by means of electronic card images on a video screen. The union demanded that the basis of data recording should still be the paper forms, so that a direct feeling of the amount of work remained intact. The union succeeded in preventing electronic scanning of paper forms.

In the previous part, I described the basic properties of computer-based signs. They have three main kinds of features (permanent, transient and handling features) and are organized in two main chains (concurrent and sequential). In this part, I shall treat composite computer-based signs. How are smaller computer-based signs combined to form larger signs?

It is very clear that complex signs corresponding to linguistic concepts like phrases, sentences, and periods exist. The player of video games will experience long nerve-racking combat episodes involving the hero and one or more enemies, and the user of a word processor will engage in major tasks like proof-reading a manuscript or reorganizing the composition of a text, involving extensive handling of tools and text. Since these larger activities are manifested in system processes, they must be considered signs also, and since they contain smaller recurrent elements with meaning (the hero and enemy, the tools and texts), they are composite signs.

Although the three main features of computer-based signs have been treated in detail, I have not yet said much of the analytical consequences of the fact that computer-based signs participate in two different chains.

I shall continue using the textbook on folk-dance from Section II. 1.2, since a dance instructor or a writer of textbooks on dance faces problems similar to those facing a designer of computer systems. Like the designer prescribing the sequence of computer-based actions and the stage of the screen, the dance instructor must teach the dancers how to move in time on the floor, how to coordinate their movements with their partner, how to hold each other, and how to establish collective patterns.

A psychological definition of language games

In the preceding section, I showed how to use semantic fields of lexical material for various purposes: when the system developer enters the organization she can use them to get an impression of conceptual differences of the organization, when she starts doing particular applications she can use them to design the general data structure, and when she has made a prototype, semantic fields are useful for evaluating its reception and for redesigning it. Finally, researchers studying the general effects of computerization on organizations can use semantic fields to record language changes in the organization, and use these data as a basis for setting up hypotheses about changes of culture and cognition.

However, language is more than a vehicle for interpretation and description. Besides the ideational functions, language also has an action aspect that is subsumed under Halliday's interpersonal functions.

Although it is not a bad idea to start with the ideational aspect, since professional languages are registers and tend to differ in vocabulary, the action aspect is equally important in systems development. Language in work situations is not only used to interpret and describe, but also to act verbally, and in a similar way, computer systems are used as media for distributing work, giving orders, and reporting. Designers not only need to know which kind of information is needed and how it should be structured semantically, they also need to know when it is needed and for which purposes.

In the preceding section, I showed how to use work and conversation patterns as a basis for design. In this last section I discuss principles for building these patterns into the system. What does it mean to “put work patterns into the system”? To what degree should it be done? How should it be done?

As my example I shall use the data entry parts of the data entry (registrering) process classified as control style in Section II.2.2.2. Here I noted that the PGP-system is in control in this part of the work process and guides the worker through a fixed standardized sequence of operations:

In addition, it automatically provides the next card of the same batch, and the first card of the next batch when the last of the previous batch is finished and the batch tallies. If it does not tally, it enters the completing subview. The system supervises the user's work, and directs the worker's attention to and from the screen by means of error messages like

a type of control which was consciously introduced by the developers: they wanted to “stop a phase where you work with the card, and initiate a phase where you work with the screen”.

I have already classified this standardization of work as a “frozen” variant structure.

I start with a reassessment of selected linguistic traditions. Since they were developed to meet needs other than those I want them to serve, a careful reexamination is not out of place.

Empirical characteristics of two work languages

The first issue I shall take up is simply whether they are able to handle the language variety in which most computer systems must function, namely the languages of people at work, and in order to do so we need to know what a “work language” is.

The car repair shop

The typical Danish car repair shop is a small scale enterprise, consisting of the owner and a couple of skilled workers, possibly an apprentice and a bookkeeper, as well.

Most of the work can be performed by one worker alone, so cooperation is seldom a physical necessity. Planning and execution of the work procedure is integrated. The mechanic performs both tasks. Personal relationships between employer and employee survive, such as the master-journeyman relationship. Normally, a small stock of tools is used: hammers, screwdrivers, nippers, etc.

In the early eighties, I investigated communication in a small repair shop, and the following pages are based on this project. The research was done in collaboration with one of the mechanics employed in the garage. A taperecorder placed on a bench recorded the conversation that took place during work. (Everybody in the garage knew about the recording.) The mechanic also produced written descriptions of whole working days.

The analysis of the data took 1 1/2 years. The mechanic was employed as an assistant teacher in an advanced course at the university.

In Part I of this book I presented a general framework for understanding language and computers, and in Part II showed how to apply this framework to analysis and design of computer systems. What remains to be done now is to broaden the scope of our analysis, and show how the same framework can be used for describing and designing computer systems in a practical context of work. In order to do that I need a concrete example, and I shall use data from the 1986 project at the Postal Giro Office in Stockholm, “Professional language in change”, mentioned in Section 1.1.1.

Before I start on this last stage of the journey, I think it would not be out of place to make a short summary of what has been accomplished so far.

Although the theoretical framework of the book is not natural science but a structuralist tradition of semiotics and linguistics, the project of the book is a part of an ongoing paradigm change in computer science itself: from seeing the computer system as a self-sufficient mathematical object, the focus is gradually being shifted to the relations between system and work context. An important practical motivation for this shift is simply that it hopefully will enable us to construct computer systems that meet the needs of the users in a better way. However, the shift is also theoretically motivated by the structuralist framework, since structuralism is characterized by focusing on relations, not on objects, as the real existing phenomena.

What is computer semiotics?

Semiotics is the science of signs and their life in society. A sign is anything that stands for something else for someone.

Semiotics treats all kinds of signs: verbal language, pictures, literature, motion pictures, theater, body language. Computer semiotics is a semiotic discipline that studies the nature and use of computer-based signs.

The motivation for the discipline comes from the nature of computer systems: although in many respects computer systems can be conceived as tools in analogy with typewriters, pencils, brushes and filing cabinets, they differ from these tools by not primarily existing or being used as physical objects, but as signs. The pencil of the drawing program is no real pencil that can be used to chew on, it is merely stands for a pencil, represented by a collection of pixels on the screen. Computer systems resemble other media by primarily acting as carriers of meaning.

Even if computer systems basically are symbolic tools, this symbolic nature has only attracted serious attention in recent years.

The reason for this is threefold: from being a tool only for specialists, computers have now been integrated into many occupations, and therefore the demands for interpretability have risen. The mode of operation and the meaning of the data must be easy to grasp for the secretary or the manager whose time should not be spent deciphering cryptic codes but writing agendas or making decisions. A good interface has become a necessary prerequisite for a good system.

We have now looked at work language, the language variety that must be central to our concerns, and made a preliminary list of the descriptive shortcomings of classical structuralism with respect to this language type.

In the next, constructive sections I will describe those parts of the tradition I wish to keep, and propose changes in other parts that make the theory better suited for the purposes at hand. Although the next few sections are concerned with language proper, I shall append short indications of how the concepts described can be used to understand computer systems, so that readers primarily interested in computers may judge if reading on is still worth the effort.

After presenting the key concepts (Section I.2.1), I show in Section I.2.2 how they can be applied to analysis of work processes and organizations, in particular the car repair and the Postal Giro examples from I.1.

Section I.2.3 extends the framework further to deal with computer systems. Computer systems are interpreted as media for human communication, and as a consequence of this, the theoretical focus is shifted from the system itself to the relations between system and user (Section I.2.4).

Basic concepts for describing symbolic acts

One of the basic points of structuralism is that units of language are defined relationally.

Computer-based signs

I start my investigation of computer-based signs from the viewpoint of the expression plane, partly because the computer's means of expression are different from other media, partly because the history of linguistics tells us that expression analysis is easier than content analysis. What are the characteristic features of computer-based signs, and in which respects do they differ from other kinds of signs we know?

Handling, transient, and permanent features

The prototypical computer-based sign is composed of three classes of features:

• A handling feature of a computer-based sign is produced by the user and includes key-press, mouse and joystick movements that cause electrical signals to be sent to the processor. Handling features articulate user actions.

• A permanent feature of a computer-based sign is generated by the computer. It is a property of the sign that remains constant throughout the lifetime of a sign token, serving to identify the sign by contrasting it to other signs. Permanent features articulate system states into parts.

• A transient feature of a computer-based sign is also generated by the computer, but unlike permanent features, it changes as the sign token is used. It does not contrast primarily to other signs, but only internally in the same sign, symbolizing the different states in which the sign referent can be. Transient features articulate system transformations.

As our first example, let us look at a version of the game Breakout. The system displays a paddle, a ball and a brick wall. The ball bounces back and forth, and the player must hit it with the paddle.

The first part of the book ended with a sketch of the computer system as a medium, and the purpose of this second part is to turn this sketch into a more detailed portrait. What are important properties of computer systems from this perspective, and how should systems be structured?

It should come as no surprise that the interface, defined in Section 1.2.3.3 as a collection of computer-based signs, and its functions relative to the linguistic environment and working context reign supreme in this approach. The basic function of a system is to generate processes in which an interface can be expressed, just as the only reason for having the costumes and wings in a theater is the experience they may contribute to giving the audience. Although this view on design is relatively new in computer science, it exists under the name of user centered design.

This view makes it difficult to maintain the tradition of separating functionality from the interface. Traditionally, functionality is what can be done with the system, while the interface is the manner in which it is done. Thus, two word processors have different functionality if one allows the user to make an automatic table of contents while the other does not. They have the same interface if operations are done in the same way, for example if text is selected by dragging the mouse over the desired piece of text, whereas their interface is different if one uses the mouse while the other requires the user to type line and character numbers in to select text.

In this chapter and in Chapter 5 we shall present many mathematical applications of computer graphics. In order to draw the line somewhere (pardon the pun) we shall restrict ourselves to the mathematics associated with the plane and in particular with curves in the plane. There is a whole other realm, just as fascinating, connected with objects, such as curves, polyhedra and surfaces, in three-dimensional space. Nevertheless what the computer actually draws is, in these cases too, a curve or system of curves in the plane – that of the screen. The extra complications come from taking a three-dimensional object and associating with it a curve or system of curves – for example its outline when seen from a distance, or a sequence of such outlines or a sequence of plane sections of the object. We touch on this in a discussion of the swallowtail surface in Section 4.14.

The computer can (with our help) draw curves and collections of curves which are just too complicated to attempt by hand. For some purposes a rough sketch of a curve does very well – you will probably have drawn many such sketches by hand, and we are certain that the art of curve-sketching by hand is still an art well worth acquiring. However, for some other purposes, such as the illustration or discovery of facts connected with the differential geometry of curves and families of curves (not to mention surfaces), accurate drawings are essential.