Case History

Case History


Large information-intensive organisations, such as insurance companies, banks, utilities and government agencies, conduct a great deal of routine correspondence with individual citizens, customers, or clients using standard letters notices and bills. These are highly complex computerised systems that draw down information from the organisations’ data bases and ‘assemble’ the output through high speed laser printers and digital output systems.

We all receive such correspondence in the form of bills, notices and letters.

CRI has been commissioned by many large organisations to develop such systems. One of the unexpected consequences of this work has been a massive productivity improvement in managing these systems, notorious for their complexity, and an expected but no less massive improvement in both sales and customer relations.

Here I report on that work, showing how such systems are being constructed and tested to ensure that they are comprehensible to their readers. We also draw out some of the implications of this type of ‘letter writing’, suggesting that it represents a relatively new genre which straddles mechanised mass-produced text and personalised correspondence.

I know that some people feel that marriage as an institution is dying out, but I disagree. And the point was driven home to me rather forcefully not long ago by a letter I received which said:
‘Darling, I love you and I cannot live without you. Marry me or I will kill myself.’!
Well, I was a little disturbed by this, until I took another look at the envelope and saw that it was addressed to ‘Occupant’.
Tom Lehrer, satirist and songwriter

As Tom Leherer’s acerbic observation suggests, mass produced letters sit uneasily at the intersection of public and private domains. One might even regard the term ‘letters’ as an inappropriate description of this type of distributed, yet individualised, communication.

This paper is about these misfitting missives; it is about the types of ‘letters’ that have emerged out of a research program undertaken on behalf of large Australian organisations to help them improve their letters, notices and bills for citizens, customers, and clients.

Mass produced letters in a consumer market

A brief history

There is nothing new about mass produced letters and notices. Even before the invention of the printing press, scribes faithfully copied existing letters to deal with routine correspondence. With the advent of printing in Europe in the fifteenth century, the mass production of all types of written material burgeoned.

Probably one of the first such ‘letters’ were Indulgences. They were printed by the church in large numbers, signed by church leaders, and sold on the promise of absolution for sins.

Letter of indulgence printed in 1454

Letter of indulgence printed in 1454

In this early stage of mass produced letters, one can clearly discern the uneasy and unequal relationship between, on the one hand, the institution and its false personalisation, and, on the other hand, the individual person regarding the communication as truly personal, an uneasy and unequal relationship that persists to this day.

The development of postal systems and the invention of the typewriter enabled institutions to use this false personalisation in a new, more customised way. Mechanised and industrialised copy typing made possible the large-scale production of individualised letters, written according to pre-constructed formulas. Even though each letter was individually typed, typists were taught to use standardised typewriter layout conventions, and the words, phrases, sentences, paragraphs, and in some case whole letters, were reproduced from established formulas. Thus, even during a period when letters were individually produced, the discourse that prevailed was highly conventionalised and controlled within the institutions.

More recently, with the introduction of corporate computerised databases, the advent of high-speed line printers, then the invention of high-speed laser printing followed by purely digital document production, there has been a radical transformation in the way that letters can be produced.

Three major changes can be discerned.
First the army of typists in the typing pools, who individually wrote letters, have disappeared. Thus any possibility of personalisation by individuals, small though it was, has disappeared.

Second, the computerisation of individual records, and the capacity to store and merge both fixed and variable data electronically, has transformed institutions’ capacity to automate business-letter writing.

Third, laser printing and digital output technology has opened up many more typographical, colour, and layout possibilities, extending the range of conventions that can be used in the construction of text.

Thus at the very time when business has become more anonymous, paradoxically it also has a much greater capacity than at any time in the past to customise and personalise communication with individuals.

Parallel with the recent technological changes, there have been political and economic shifts that have led to an increasingly powerful consumer lobby. Organised consumer interest groups have had a major impact on business regulation and practice.

Issues of comprehensibility, openness, and fairness for consumers have been brought to the fore through consumer protection legislation and regulation; businesses dependent on a consumer markets have responded to the rising tide of consumerism by adopting ‘customer focused’ policies and practices.

An opportunity for research

Taken together, the technological and political changes have created a climate in which it was possible to undertake research to rethink and redevelop ideas about business letters, their content, construction, and design.

CRI, with a broad agenda to help industry and government improve communication practices through research, was invited by a number of large corporate bodies in Australia to investigate and suggest ways of improving letters, notices, and bills (Sless 1992) .

The corporate bodies included government departments, utilities, and financial service organisations. While each body was largely concerned with its own communication problems, the Institute was able, through a series of case studies, to develop an overall approach to these types of problem (Sless 1996).

Here we look at the language and typography features that have emerged from this ongoing research program.

Assumptions and methodology

CRI’s research program on letters is broadly within the applied interdisciplinary context of information design (Sless 1992). It thus draws on the crafts of writing and typographic design, on the procedures of usability engineering (Nielsen 1993) and design methods (Jones 1980), and in particular on participative design methodologies (eg Schuler and Namioka 1993).
The theoretical underpinnings of the Institute’s research program are constructionist (Pearce 1995). Constructionism is concerned with how we mutually construct our social realities through conversation. This paradigm has its intellectual origins in the social philosophy of George Herbert Mead (1934) the aesthetic theory of John Dewey (1934), the moral philosophy of Martin Buber (1961), the linguistic and literary criticism of Bakhtin (Holquist 1981), and the language philosophy of the later Wittgenstein (1958).

Contemporary theorists within this broad church include social psychologists such as John Shotter (1993) and Rom Harr_ (Harr_ and Gillet 1994), communication researchers such as Kenneth Gergen (1994) and W Barnett Peirce (Pearce and Cronin 1980) and some designers. Recent interviews with Edwin Schlossberg and John Seely Brown in Mitchell’s recent book New Thinking in Design seem to be converging on a constructionist view of design similar to our own, though the interviews are too brief to provide a full view of their methodology(Mitchel 1997).

The central theoretical assumption of this paradigm, as it has been applied in this study, is that the basic unit of analysis in any communicative activity is the conversation. Thus in contrast to functional or structural linguistics, which treat text or discourse as legitimate units of analysis in their own right (Halliday and Hassan 1985, Saussure 1916) , the research reported here takes the relationship between text and user as irreducible. At the risk of oversimplifying, one can say that while traditional linguistics studies text, the methodology used here study the texts as they are used. thus instead of using formal linguistic categories to study examples of text, the methodology used here studies people using text. This leads the research to turn inside out the familiar linguistic domains of semantics syntactic and pragmatics: meanings are not immanent within texts either semantically or syntactically; rather they arise pragmatically, emergent properties of text usage. Semantic and syntactic features of language are treated as constructions, abstractions arising out of particular contexts of human action.

It is important to note that this work is not concerned with creating universal principles or generalisations about human behaviour. The detailed observation of human action is concerned with understanding the highly contextualised social interaction between people and documents. At most, it is possible to claim that this work leads to the creation and articulation of conventions, mutable rules of social conduct, not immutable laws of human behaviour.

No attempt has been made in this research to hypothesise underlying linguistic or cognitive processes. The research focuses on the actions of readers, and the textual features that enable or disable action. The observation of text usage is the primary data used in this type of research.

It is also important to note that the research reported here is interventionist: it is concerned with bringing about desirable change. From a constructionist point of view all research activities are interventionist; they, no less than the conversations they study, are generative activities giving rise to new possibilities which emerge out of the interaction between researcher and researched. As an aspect of information design, the research reported on here is by definition interventionist and constructive (Sless 1997). Moreover, from the point of view of the organisations commissioning the research, the objective has all along been interventionist, designed to improve an aspect of their communication.

Out of this research emerged principles for the construction of standard letters, of which the letter in this chapter (Fig 1) is an example.

In this chapter, I want to argue that this research has led to the creation of a hybrid set of conventions which may constitute an emergent new genre of business letters.

Some limitations of the data presented

This chapter discusses the manifest output from the research program undertaken by the Institute: a typical example of a unique letter that is produced for an individual customer by the combination of data base and high speed laser printing technology.

The full research program on standard business letters encompassed a multitude of organisational and technical issues to do with programming, writing, designing, and implementing standard letter systems.

The evidence from successful implementation of such systems suggest that well over 50% of the effort in a project can be dedicated to organisational and technical issues(Fisher and Sless 1991) . It would therefore be misleading to suggest that the material presented in this paper constitutes in itself the ‘solution’ to the standard letters problem. It is only one manifestation, the output of an intricate and complex process.

Data presentation

The letter shown below has been modified to ensure confidentiality of both the company and the individual, but in all other respects it is the output from an actual system.
Standard Letter ExampleThe letter is the end result of a complex methodology involving many rounds of diagnostic testing and refinement with customers to ensure that all variants of letters in the system are comprehensible and acceptable in tone to the company’s customers, as well as adhering to the company’s business rules and system constraints (Sless, forthcoming). Diagnostic testing, as the name implies, is concerned with identifying faults in the document that lead to inappropriate readings by users. Each round of diagnostic testing—usually involving between ten and twenty participants—is followed by refinement or changes to the design to remove any faults and enhancing the design so that it performs optimally with users. The testing and refinement cycle is repeated until the letter is performing optimally. As with the letter, the exact results from the testing are confidential to the company. However, the principles applied and the results as manifest in the letter, do not involve any breach of commercial confidentiality

This letter replaced a traditional letter and accompanying forms. The letter it replaced, though produced on a high speed laser printer, followed the normal typewriter conventions for letter design, using a single non-proportional font and multiple pages. The new letter uses the same customer data and business rules as the original. However, this letter differs in three ways from its predecessors: in its production, performance, and content.

First, no single person actually wrote this letter. It was produced by ‘assembling’ a variety of prefabricated components according to a set of algorithms. Indeed, if a letter in this system needs to be modified in production then it is the algorithms and components that are changed, not the letter itself.

Second, the letter’s performance—the extent to which it is accessible and usable by customers—is not only better than its predecessor, its usability is largely known ahead of its implementation. Typically, a letter of this type results in far fewer transaction errors or queries than the traditional business letter.

Third, as this chapter discusses, this letter does not look like a conventional business letter, nor does it follow the normal conventions of business letter writing.

Four important characteristics define the conventions used in these letters.

1. Three structures in one

The most noticeable initial characteristic of this letter is its division into three zones, made visually distinct by the background preprint stationery and the arrangement of elements on the page: the top left is the nearest to a traditional letter, the top right is a summary of the customer’s policy information, and the bottom section is a payment slip. Each zone is designed to be used differently by readers: the letters zone is to be read, the right hand column is to be used for reference only, and the bottom slip is to be completed for payment. These correspond to the three main groups of tasks that readers need to perform in their dealings with the insurance company.

These structures were, like all aspects of the letter, derived by a combination of design decisions and subsequent refinement about the structure and content of the letter, as a result of diagnostic testing of the letters with users.

This ‘mixed’ genre—letter, reference list, and payment slip—is a good example of what can be achieved with contemporary technology. The fact that the genre makes sense to users, is indicative of a widespread ability to deal with such mixed genres on a single document.

2. Typography

The second characteristic is that the document uses a range of fonts and spacing to guide the reader, unlike traditional business letters which follow the conventions of typewritten text, using a single mono-spaced font for all text. In traditional business letters, the burden of providing readers with visual queues to the structure and content of the text depended almost entirely on the vertical spacing of elements: addressee, date, reference, salutation, paragraph breaks, closing salutation, signature space and name of sender. White space, in the form of blank lines differentiated between different parts of the letter.

Contemporary high speed laser printers give a wider range of choice over fonts and spacing that did not exist in a typewriter, giving a much wider range of visual queues for guiding readers. However, it is important to realise that this enhanced choice falls short of the choices available in today’s word processing software, and a long way short of what is possible using high end desk top publishing software. This means that the range of ways in which typography and spacing can be used is restricted, even though it is a significant enhancement on the typewriter.
In the example shown, the software used to create the letters imposed a number of limitations. These included: the number of fonts that could be used, the font family they could be chosen from, the interline spacing within each zone, and the letter spacings both between and within words.

3. Unity of form and function

Most linguistic analyses of text ignore typographic considerations. Moreover texts on business writing tend to give typographic issues a low priority, and mention it only briefly (eg Murphy 1989, Petelin and Durham 1992, Putnis and Petelin 1996). However, as Waller persuasively argue, typographic features are an integral part of the way in which we construct meaning out of text (Waller 1987).

In the letter shown in figure 1, the typographical variety available through the new technology is used as an integral part of the letter; there is a direct correspondence between form and function; the appearance and position of the text matches the function of the text for the reader.

As part of the research and development of these letters, we created and tested a set of rules governing the functions of the text from the users point of view and the use of the typographical and spatial specifications to match those functions. Particular functional features of the text always appear in the same font and in the same positions relative to other elements, and the whole document follows a consistent typographic hierarchy.

The illustration in figure 1 clearly shows how function and form come together in a strict hierarchy. Within the letter zone, there are always topic headings followed by statements or instructions, which are sometimes followed by an explanation or elaboration. Each of these three functional elements—topic headings, instructions, and explanations—have their own distinctive typographical manifestation; section headings are in bold and preceded by a space, the instructions or statements are in roman, and the explanations or elaborations are in italics.

The diagnostic testing used in this project showed—as in many others—that a consistently applied set of rules that are visually distinct to readers, help those readers use the documents appropriately. (Mackenzie 1992)

4. Temporal sequencing

The letter follows the temporal order of events from the readers point of view, with the headings making this temporal order visibly part of the structure by using the following section headings:
1. What has happened
2. What we would like you to do
3.What will happen if you do not do this

This type of temporal structuring is a routine part of such letter systems and is also used in other types of documents (Penman 1993). Its use is derived from widely used narrative structures in which the telling of a story follows the temporal sequence of events. Thus it is a widely understood cultural convention. It has also been found to be highly effective in other types of documents. For example, in consumer medicines information, such a structure has been found to be highly effective, taking consumers through the sequence of actions which begin with ‘before you take the medicine’ and ending with ‘how to store and dispose of the medicine’ (Sless and Wiseman 1994).

From the user’s point of view, such a structure has two functions: first, it provides a coherent way of understanding the document as a continuous interconnected series of actions; second, it provides a highly usable reference structure, so that the reader can easily find the particular section they are interested in, at any time in the future.

For example, it would not be unusual for a customer to read the letter when it arrives and put it to one side to deal with later. When the customer returns to the letter at a later date, their primary question is: what do I have to do and when? At that moment their reading is quite different from their first reading. They are more likely to scan the letter, looking specifically for the answer to their question, rather than reading the whole letter, starting from the beginning. The letter is structured to help them also with this second type of reading activity.

4. Absence of false personalisation
One of the most enduring features of the business letter enduring long after the demise of personalised correspondence—is the use of the opening salutation ‘Dear, and the closing salutation ‘Yours Faithfully etc,.

In discussions with many businesses and government agencies, it is clear that many managers feel extremely uneasy about abandoning this formal convention. The reasons are complex—resistance to change; concern over losing the ‘authority’ of authorship either personally or legally; and a concern that customers will not read the letter because it does not have the appropriate letter conventions.

There are, however, a number of practical considerations that make the retention of the convention difficult. The person ‘signing’ the letter is not necessarily the person who will deal with the customer if they ring up or write, particularly with contemporary business call centres. In an attempt to maintain a level of personalisation, some organisations do direct customers to specific individuals who know about their ‘case’.

Unfortunately, because of the rapid turnover of call centre staff, this too becomes impractical. Further, even in situations where it is practical to give customers a specific person to contact, adding a facsimile of their signature to the electronic laser printing system, increases the memory requirements of the system, which can slow down the mass production of the letters.

In the event, the testing of these letters with customers revealed no problems with the absence of these formalities. Indeed, most customers did not even notice the abandoned convention and did not see anything odd about it when it was drawn to their attention. Whether this would be the case in cultures outside Australia, we don’t know, but the fact that these letters work effectively within their own context and milieu suggests that, at least in Australia, we have seen the emergence of a new genre, a new set of workable conventions for business letters.


The emergence of a new genre in business letter writing is the result of a combination of changes in technological, social and methodological factors. In the future we may expect to see this genre evolve as these factors themselves change. Indeed the work reported here has already evolved, from the large scale production environment of a mainframe computer and high speed printers accessible only to large organisations, into the more modest and generally accessible desk top system that can be used by medium to small organisations.

While there is some reason for celebration in the development of a genre that makes the mass produced business letter easily accessible to customers, there are also good reasons for expressing caution about this development. There is evidence from studies in plain English that customers can be lulled into a false sense of confidence about their capacity to understand a business document because it is in a ‘plain English’ style (Sless 1996). It is only the repeated testing and retesting applying the most rigorous standards of good conduct that ensures the letters are intelligible rather than just seeming to be so.
Even though this genre eliminates the false personal relationship of earlier letters, it does not alter the basic uneasy power relationship between large corporate bodies and individuals—a relationship that is always open to abuse. It would take more than a new letter design to change that relationship.


Buber, M. (1961) Dialogue. In Between Man and Man, 17_59. Translated by Ronald Gregor Smith. (First published in 1929 as Zwiesprache.) London: Fontana Library Edition.

Dewey, J. (1934) Art as experience. New York: Milton, Balch.

Fisher, P. and Sless, D. (1990) Information design methods and productivity in the insurance industry. Information Design Journal, 6(2), pp 103-129.

Gergen Kenneth J 1994. Realities and Relationships: Soundings in Social Construction. Cambridge, Mass: Harvard University Press.

Halliday, M. A. K. & Hasan, R. 1985. Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective. Victoria, Australia: Deakin University Press.

Harr_, R. and Gillett, G. 1994. The Discursive Mind. Thousand Oaks, Calif: Sage.

Holquist, M. (ed.) (1981) The Dialogic Imagination: Four Essays by M.M. Bakhtin, Trans. Caryl Emerson and Michael Holquist. Austing: Austin University Press.

Jones, C. J. (1980). Design Methods : seeds of human futures. London: Wiley.

Mackenzie, M. (1992) New Standards for ABS publications; a case study. In Documents in context. Australian Journal of Communication 19(3), 43_70.

Mead, G. H. 1934. Mind, Self and Society. In Strauss Anselm (ed) 1964, George Herbert Mead on Social Psychology: Selected Papers, 115_282. Chicago: University Press.

Mitchell, C. T. (1997) New Thinking in Design : Conversations on Theory and Practice. New York: John Wiley & Sons.

Murphy, E. M. (1989) Effective writing: plain English at work. Melbourne: Pitman Publishing.

Nielsen, J. (1993) Usability Engineering. AP Professional, Cambridge MA.

Pearce W B (1995) A sailing guide for social constructionists. In Leeds-Hurwitz Wendy (ed), Social Approaches to Communication, 88_113. New York: Guildford Press.

Pearce W B. and Cronen, V. E. (1980) Communication, Action, and Meaning: The Creation of Social Realities. New York: Praeger.

Penman, R. (1993) Conversation is the common theme: Understanding talk and text. Australian Journal of Communication, 20(3), 30_43.

Petelin, R. and Durham M. (1992) The Professional Writing Guide: writing well and knowing why. Melbourne: Longman Professional.

Putnis, P. and Petelin, R. (1996) Professional Communication: Principles and Applications. Sydney: Prentice Hall.

Saussure, F. de (1916) Course in General Linguistics. Edited by Charles Bally and Albert Sechehaye 1966. Translated from the French by Wade Baskin. New York: McGraw-Hill.

Schuler, D. and Namioka, A. Eds. (1993) Particpatory Design: principles and practices. London: Laurence Erlbaum.

Shotter, J. (1993) Cultural Politics of Everyday Life: Social Constructionism, Rhetoric and Knowing of the Third Kind. Buckingham: Open University Press.

Sless, D. (1992) Designing documents that work. Xploration, III (1), 14_16.

Sless, D. (1992) What is information design? In Sless, D. & Penman R. (Ed) Designing information for people. Canberra: Communication Research Press, 1_16.

Sless, D. (1996) Better information presentation: satisfying consumers?, Visible Language, 30(3) 246_267.

Sless, D. (1997) Theory for practice. Communication News 10(4) 1997

Sless, D. (1997)

Sless, D. and Wiseman, R. (1997) Writing about medicines for people: useability

guidelines and glossary for consumer product information. Canberra: Australian Government Publishing Service.

Waller, R. (1987) The typographic contribution to language. Unpublished PhD Thesis. Reading University: Department of Typography and Graphic Communication.

Wittgenstein, L. (1958) Philosophical Investigations. Translated from the German by G E M Anscombe. 2nd edition (revised). Oxford: Blackwell.