Usable words are words that make sense to the people who have to read them. When we at CRI started our work, we assumed that the advice given to us about writing clearly and plainly, including using ‘the right word’, would work. It turned out not to be the case. As soon as we started doing baseline measurement studies—diagnostic testing of existing public documents—we discovered an unexpected variability in usages across subject domains, idioms, populations, and demographics. Was the craft tradition of writing wrong?


Advice on clear writing has arisen out of formal linguistics and persuasive communication with roots in grammar and classical rhetoric. This advice is confident and to a great extent authoritarian in many senses of that term: it has been built on the authorities of the past, most notably in the practical art of rhetoric. One of the clearest common indicators of this tradition is that readers are frequently referred to as an audience.

In contrast, contemporary practice in information design is based on a firm background of practical knowhow built up over centuries of trial and error. This has culminated in rules not unlike the prescriptions of rhetoric and classical grammar, but applied to the visual rather than verbal nature of texts. At CRI, with our backgrounds in visual information design crafts such as typography and graphic design and in empirical and critical communication research, 1Inside Communication Research we were at first hesitant to critique the cognate craft of clear writing. Who wants to throw the baby out with the bathwater?

But the hesitation did not last long. To start with, describing readers as an audience struck us as odd. Audience is clearly a hangover from the rhetorical tradition, when well practised orators persuaded and manipulated their listeners to a certain point of view. (Sometimes they succeeded, of course, just as the occasional clever television advertisement might create a temporary rise in sales.) As well, for some time we had been sceptical of the underlying assumption about the nature of communication as a linear process from author to reader (often sender to receiver). ‘Getting the message across’, in other words. From our point of view, informed by critical communication research and our own changing notions about the nature of communication, we saw the linear model as too limited for the task in front of us.

We see readers not as passive receivers of messages already formed by active senders but as active creators of meanings that flow from their social and individual positions—readers can be citizens, consumers, workers, parents, students, and many more. Reading is as much an act of creation as writing—readers construct meanings out of reading what others have written.

When we first invited people to participate in our testing of documents at the baseline measurement stage, we discovered a degree of unanticipated variability in the way people read and constructed meanings out of the texts in front of them.

baseline measurement in information design

This is not a trivial matter. People misunderstand their entitlements and rights, their obligations, and the agreements they have entered into, either because the words used by writers aiming at clarity have different meanings to those used by the readers, or because the words used by writers are not in the readers’ vocabularies.

There are other problems associated with specific subject domains. Readers use ways of talking to each other about a subject that differ from those of the writers, in word choice, emphasis, and structure. Sometimes a writer’s perspective is quite different to the readers’—we found that readers sometimes reject documents, saying for example:

They’ve only written it to cover themselves. They’re not interested in helping us.

In a broader sense readers see some public documents as socially inappropriate to their circumstances. The consequences have been, on occasion, quite serious: parliamentary inquiries, loss of institutional and corporate credibility, disadvantaged citizens, sickness and death. There have also been many instances where organisations’ productivity and profits suffer.

We quickly realised that we had to take language variability and usage into account in our Scoping and Baseline Measurement stages, before we started developing new prototypes.

We therefore augmented some of the questions we ask at these early stages to provide us with insights into this variability, so that we can take it into account at the Prototype stage.

These Guidelines are to help our Members and Fellows take advantage of what we have learnt from this experience.

As we said in a recent newsletter:

We can help you get it right before you write. Easier than getting it wrong and fixing it later.

At the heart of these Guidelines are special ways of listening and observing.

Sorry to interrupt your reading of this interesting paper. The publishing of this paper is funded by our Members and Fellows. If you are a Member, please log in to read the rest. If not, and you value what we do, please become a Member.