Text accessibility requires multiple formats

Language can be a stumbling block. How you write and present a text determines not only what you say, but also who is likely to understand you. If a text is too brief or uses too general terms, you may fail to communicate what you mean. But a flip side is that writing that is long or uses specialised words is likely to shut the door on some readers. In navigating this dilemma, the specific problems of accessibility that language poses should usually be at the front of your mind.

These accessibility problems are real. In practice, language often excludes. To put it simply, it divides between those who ‘get it’ and those who don’t. This is a serious issue that needs to be addressed within any community that values diversity and equality. Here I will focus on written language and reading needs. My argument is that because there is a plurality of reader experiences, communities should facilitate a plurality of text formats when there is a need for them.

A fallacy of generalisation

The aim of all writing is communication and writing fails if communication fails. Ernest Gowers wrote The Complete Plain Words (1954), one of the best known books on how to communicate well through writing. Gowers said that the golden rule is to pick the words that will convey your meaning, and only those words. Those words should be as plain as possible. No one is served by written documents that are needlessly complex or verbose.

But what is plain to the one, need not be plain to the other. It is all too easy to assume that, because a text is plain to you, it will be plain to everyone else as well. That is a fallacy. We can call it the fallacy of generalising reading experience. Such a generalisation is almost always a mistake, because there is great diversity among readers. An important source of this diversity is a difference in reading level.

A measure of reading level

Someone’s reading level is often expressed using the Common European Framework of Reference (CEFR). Their literacy scale ranges from A1 and A2 (beginner) to B1 and B2 (intermediate) to C1 and C2 (advanced). The scale was developed for people learning additional languages. However, over time it has come to be used to express first language ability as well. Someone who grew up in the Czech Republic and learnt to read texts in Czech can on this scale be scored as, say, a B1 reader if their ability to read Czech is intermediate, or as a C2 reader if their ability is advanced.

The B1 level is currently taken to be a norm for writing for a general audience. In the Netherlands, for instance, it is a guideline used by governments and public bodies for when they communicate with citizens across society.

Someone who has a B1 reading level in a language is an intermediate reader of that language. According to the CEFR classification, such a reader “can understand texts that consist mainly of high frequency everyday or job-related language”, and they “can understand the description of events, feelings and wishes in personal letters”. What is characteristic of texts intended for B1 readers, then, is that those texts will avoid specialised language, unless it is language that is common in the intended reader’s professional life.

Declining literacy and its consequences

Although B1 is a common norm, certainly not everyone is able to read at this level. An OECD survey from 2015 found that 1 in 6 of adults in England, and 1 in 5 in Northern Ireland, has literacy skills at or below the most basic level. According to the 2018 PISA report, in France just over 20% of 15-year-olds shows only the lowest level of reading proficiency, and among Dutch teenagers of that age this figure is closer to 25%. Furthermore, these figures are not stable, but in many countries literacy is declining.

This means that, compared to a few decades ago, increasingly many people struggle to read texts written for a general audience, let alone specialised texts. For this reason, successful communication will sometimes require finding alternatives to the B1-norm for general audience texts. (The Dutch government, for example, provides versions of some of its communication tailored to A2 readers.)

Language as an accessibility problem

All this proves that if you care about accessibility, you should take into account that language works differently for different people. It is all too common for written communication to become an exclusionary device. Just imagine how someone still learning Turkish may become sidelined in a society in which all important documents are sent to them in Turkish. Or think of a management meeting in which the circulated memo presupposes an understanding of the lingo specific to an organisation; this may make a newcomer feel left out and at a loss.

More generally, when members of a community or practice are expected to be able to read at a specific level, then this will put at a disadvantage any member who for whatever reason cannot meet those presumed standards. In these cases, it would be legitimate to wonder, ‘if I am really a member, then why am I not enabled to participate?’

Default reading levels

How should we respond to the specific accessibility challenges posed by written language? It may be tempting to suggest that to ensure accessibility, all texts must be written for a general audience. Because specialised language is not accessible to all, it must be avoided in documents that concern all—or so the thought goes. I think that this is a mistake. It is a mistake because the accessibility problem occurs regardless of which level of reading is adopted as the norm. No norm will be accessible to all.

True enough, if only specialised texts are offered to a diverse community of readers, it is likely that such documents will not be accessible to the entire community. Not everyone may have the ability to read and understand such texts. But the same holds when written communication in an organisation is aimed at people who read a language at a B1 level (which many governments and public bodies use as a norm for general audience communication). As mentioned earlier, a significant and growing proportion of many communities is not able to read that community’s dominant language at B1 level. Such readers will be left out if there are no forms of communication available that meet their specific needs.

In other words, it won’t do to to single out a specific reading level for which your text needs to be suited, because whatever reading level you pick, there will be people who cannot read at that level. And what should you do if you want to reach those people as well? I think that the answer is that texts should be tailored to meet people’s needs. But tailoring texts to specific needs of one group of people, may well produce results that are inaccessible to others. For example, a translation into Arabic or Finnish can help some, but will again not be accessible to all.

There are no B1 texts, only B1 readers

Moreover, the assumption made in all this is that we can meaningfully classify texts in terms of reading levels. That we can say of a specific letter or report that this text is suitable for B1 readers. But the assumption that there are ‘B1 texts’ is problematic, because the group of B1 readers is itself diverse.

Between those who would score as B1 readers there will be differences as to what counts for them as ’everyday or job-related language’. Everyday life is not homogenous. Differences in people’s upbringing, employment, and life experience will make for differences in which vocabulary these readers will experience as ‘plain’. For example, someone who scores ‘B1’ on a reading test may well be able to read the professional manuals or instructions pertaining to their day job, documents that contain terms that would mean little to many other people. This illustrates that even if a document is plain for some B1 readers, it is highly likely that there will be B1 readers for whom it will not be plain.

The upshot is that there are no B1 texts, only B1 readers. There is no form of writing that works for all. Instead, a community is best served by a variety of writing forms and ways of delivering them, tailored to the specific needs of the people you are communicating with. This will always be imperfect, but it is the best you can do.

The use of specialised vocabulary

There are other problems with expecting of organisations and communities that shared texts are all written tailored to a B1 reader, at least if by that is meant that these texts only use language that avoids specialised vocabulary.

For one, in legal contexts specific terminology may be required for a text to function at all. More generally, avoiding all specialised vocabulary will often make texts unhelpfully long and complex. This is because specific language can improve readability just as it can undermine it. The richer a vocabulary, the wider the range of words you have at your disposal, the more efficiently a thought can be expressed. (This is illustrated by how the number eleven can be expressed using different numeric systems with vocabularies of different sizes: eleven in binary, which has a vocabulary of two terms, must be written as ‘1011’; in decimal, a vocabulary of ten, it is shortened to ‘11’; in hexadecimal, a vocabulary of sixteen, it is reduced to ‘B’.)

Removing all specialised vocabulary from a community’s communicative registers would mean that certain things could only be communicated in a form that replaces specific terminology with a general description using only more basic words. This would inevitably lengthen these texts. But length can itself be an obstacle for people. In that case, the accessibility problem would not have been solved, but only displaced. Or the community would simply have to stop writing about that topic.

Once more, the conclusion to draw is that accessibility asks for a diversity of approaches. Different readers may require different formats.

Not one size fits all

I hope this makes clear that because there is a plurality of reading needs, communities should facilitate a plurality of text formats when there is a need for them. The irony is not lost on me that my elaboration of this simple point has taken a considerable number of words, and led me to take you through various lines of argument. But I have tried to be as plain as I could without losing sight of the things I meant to say.

When it comes to accessibility, one size fits all is rarely the solution. And also when you consider the specific problems posed by written language, there simply is no form of writing that is accessible to all. This includes my own writing, I am well aware.¹ If you want to ensure accessibility of communication, you should expect that whatever form you choose, there will be some people to whom it is not accessible. No text is ideal.

In the end, none of this absolves you of the task of writing as plainly as possible. If you want to say something, there is no excuse to abandon the golden rule.² Pick the words that will convey your meaning, and only those words.