The importance of readability

‘Translation involves the transfer of implicit and explicit lexical meaning from one complex linguistic and social system into another linguistic apparatus with the objective of facilitating streamlined cross-cultural comprehension, bringing continents, people and businesses closer together, assisting commercial enterprises to expand into new international markets and business environments.’

Crikey, that was quite a mouthful!

Today’s blog post will look at readability, also referred to as plain language, plain English or clear language, depending on who you’re asking. Readability, as the word suggests, is all about how easy it is to read something and understand what is being said. It’s about providing information clearly and concisely without using unnecessarily verbose turns of phrase or complex syntax – the opposite of the gobbledygook you read above.
Writing with readability in mind is not about ‘dumbing down’ content, oversimplifying a complicated subject or swapping out important technical terms for unsuitable ones. Rather, readability helps make everyone’s life easier by providing information in a way that removes ambiguous phrasing and is easily digestible. Reading a sentence just once should be enough if it’s well written; cutting out unnecessary words helps us all.

How can readability be measured?
The most common way of measuring readability in English is by using the Flesch Reading Ease score. This metric focusses on word length, word complexity and sentence length. The formula yields a score ranging from 0 to 100, with higher scores meaning higher readability.
New, cutting-edge ways of determining a text’s readability include NeoSCA, a tool that analyses the syntactical complexity of a document, rather than just word and sentence length. It counts how frequently nine different structures appear in a text – including words, sentences and dependent clauses – as well as 14 syntax indicators, such as dependent clauses and mean sentence length.

Who helps promote readability?
A range of actors have shown their commitment to readability and plain language, including non-profit organisations, public sector institutions and well-known writers.
One of the main UK-based organisations promoting readability is the Plain English Campaign. An independent group, it helps organisations to communicate clearly with their audiences by cutting out unnecessary jargon and misleading turns of phrase. They are passionate about public information being clear and understandable.
The European Union has also worked hard to promote clear language. The European Commission, for instance, organises clear language information events for the public and training for its staff. What’s more, the translation department at the European Parliament is called the Directorate-General for Translation and Clear Language, which shows the institution’s commitment to clarity and readability.
In Politics and the English Language (1946), George Orwell explains the importance of clear language. He criticises the writers of his time for their pretentious word choices and is fed up of seeing the same set phrases lazily tacked together. He goes as far as to say that ‘the whole tendency of modern prose is away from concreteness’. Orwell suggests his famous six rules, one of which is: ‘[n]ever use a long word where a short one will do’ – an important principle of readability and clear language. For Orwell, a lack of clarity makes the text less original and unreliable. In another rule he urges writers to ‘[n]ever use a foreign phrase, a scientific word or a jargon word if you can think of an everyday English equivalent’.

What is the relationship between AI and readability?
Depending on the prompt or input AI-generated texts can contain quite complicated syntax, ambiguous wordings or both. This presents two major concerns for readability: (i) clusters of vague adjective-noun pairs, and (ii) a general lack of substance and meaning.
To provide a concrete example, I asked a well-known AI tool to write a 200-word text about England’s chances of winning the World Cup.
Here is an extract from that text:
‘Recent deep runs in major tournaments have instilled a resilient elite-mentality within the squad, shedding the historical psychological burdens of past generations. Under pressure, this group has proven they can control tempo and navigate complex knockout fixtures.
However, the path to glory remains fraught. Success will hinge on defensive solidity against elite counter-attacking sides and the manager’s ability to maximize tactical flexibility during critical in-game shifts.’

Issue 1: Too many noun phrases
The extract above is incredibly noun heavy. Complicated adjective-noun pairs chop up the sentences, making them harder to understand. These include: ‘deep runs’, ‘elite-mentality’, ‘defensive solidity’, and ‘tactical flexibility’. This is a perfect example of how abstract nouns can often impede meaning and flow. We find ourselves nodding along to the text without fully understanding it.

Issue 2: A lack of substance and meaning
The text doesn’t tell us about whether England will win the World Cup, nor does it compare England to other teams, or mention a single player. The text dances around the topic with vague words that sound appropriate, even technical, in the context, such as: ‘in-game shifts’, ‘maximize tactical flexibility,’ and ‘navigate complex knockout fixtures’. What exactly they mean, however, is unclear.
Of course, this is only one example of an AI text, generated after one prompt. Therefore, it is not representative of all AI-generated text. However, this is a good example of how AI forms strings based on how likely words are to appear together. This yields a text that, based on data, sounds plausible, but does not communicate much substance.
For a text to be readable, a human is required somewhere in the process. AI alone cannot evaluate the clarity and quality of written text and emulate human writing expertise. AI can certainly be useful, but humans are still best placed to tell whether a text makes sense to them.
Final thoughts
Readability has become increasingly important for all kinds of organisations. As Orwell notes, readability is a responsibility that we all share.
This is also true for businesses. When posting marketing campaigns and website content, for example, companies need to have their potential clients in mind. Visitors to the site need to be able to understand a business’ products and unique selling point (USP) quickly, without having to decode jargon or read through paragraphs of long-winded text.
In the age of AI, drafting a marketing email or producing copy has never been easier. You can type a prompt, press enter and a text appears. However, it’s important to think about the message you want to convey, the words you need and don’t need to do so, and above all how your copy will land with the reader.
Can you think of any other important aspects of readability? Do you have an opinion that you’d like to share?
Please get in touch at: office@astls.co.uk

Links:
https://pypi.org/project/neosca/0.0.19/
https://translation.ec.europa.eu/get-involved-european-language-activities-and-initiatives/clear-language-events_en
https://standards.education.gov.uk/standard/plain-language
https://www.orwellfoundation.com/the-orwell-foundation/orwell/essays-and-other-works/politics-and-the-english-language/
Hemingway editor -> gives writing a readability grade: http://www.hemingwayapp.com/