Vincenzo Foppa, Young Cicero reading (Wallace Collection)

Behind this talk (by Jeffrey Johnson, author of several books on UX design) was one great idea: that reading is unnatural. What Johnson meant by this seeming paradox is simply: any infant (unless handicapped) learns to walk; any infant, placed at a young age among humans, will learn to speak the language of those humans. But reading is not something that everyone picks up automatically; like playing the piano, it has to be learned. Johnson stated, and it sounds reasonable, that over 50% of the world’s population either cannot read or has difficulty reading.

That was the great, and brilliant, idea;  but what came after it was not quite so convincing. The world divides, according to Johnson, into skilled readers and unskilled, and there didn’t seem to be a lot of hope for the unskilled. The speaker confidently stated that all the attendees were most likely skilled readers.

The BrightTALK format for presentations is not a slide show but a question and answer session, which means that the presenters (Scott Abel and Patrick Bosek) were able to dictate the event towards their own interests, and their questions set the theme. The questioners wanted to know about how people read in general. Johnson stated he was a web design expert, not a reading expert, but the questions by the presenters (and they had many questions) kept coming back to reading. “Do I read enough to my young child to make him a skilled reader?” was one of their questions. This wasn’t what I had expected, nor, I imagine, was it a question that Professor Johnson had anticipated.

I found the presentation tantalising, because clearly Johnson (who teaches interface design at the University of San Francisco) knew a lot, but didn’t get the chance to talk about it. When he talked about how we learn to read, he was on more shaky ground. He stated that young children can learn any language, which is correct, but he then claimed that having learnt a language, the child then “forgets” the sounds (he meant the phonemes) of other languages. We don’t forget phonemes; but we learn to recognize significant differences between sounds of one language, and we have difficulty identifying those differences later in life. OK, this was a small point, but it didn’t give me complete confidence for other points in the talk.

Text for websites

He was more confident when it came to suggestions about reading pages on websites. What were his suggestions for web design? Most writers think that more text is better, but Johnson’s advice was: minimize the text. Follow Steve Krug’s advice, and just before your site goes live, cut out half the text, then cut out half again. Don’t use technical vocabulary, unless you are sure your audience understands it (and you can’t be sure all of your audience does, so don’t use technical language).

All this sounds good, although it is facile simply to say “less text is better”. There may be occasions when the text is necessary, and you have to devise some way of making it more accessible – Johnson knows this perfectly well, and described it as “creating a visual hierarchy”, using bullet points, paragraph breaks, and so on.

Although Johnson is clearly aware of different types of reading for different purposes, more time was spent on vague questions such as “is reading top down or bottom up”, or “how can I read faster?” The answer should be, first define what the purpose of your reading is. We don’t read a recipe the way we read a guide to buying airline tickets. We don’t read a poem the way we read a detective novel. But the answer given was simply “practice reading a lot”.

How we read

Elsewhere in the talk, Johnson seemed to be aware of the famous paper by Kenneth Goodman on how we read (“Reading: A psycholinguistic guessing game”, published back in 1967). Goodman suggests, and it is commonly agreed today, that we don’t read every word; we read a couple of words, then guess what is coming next. Our eye moves forward in a series of leaps, and we only go back if we guessed wrong. The illustration shown (above) is an attempt to depict the reading process – it is actually a child reading what for them is a complex passage, but the idea is, I think, clear: that we guess our way forward with a text.

How a fourth-grader (9-10 year old) reads a text (From Goodman 1967)

He was more confident when it came to suggestions about reading pages on websites. What were his suggestions for web design? Most writers think that more text is better, but Johnson’s advice was: minimize the text. Follow Steve Krug’s advice, and just before your site goes live, cut out half the text, then cut out half again. Don’t use technical vocabulary, unless you are sure your audience understands it (and you can’t be sure all of your audience does, so don’t use technical language).

And of course, this method of reading will vary, depending on the goal. Reading a novel for study and reading for pleasure may even mean that the same book is read differently at different times. There is a major study to be done some day about the way people read. Software tools, such as Semantic Scholar and Scholarcy, are beginning to provide tools to aid the reading process for one purpose, doing a literature survey, but they are just the start.

In all, this was a fascinating, if slightly disappointing, session. It contained some great ideas, but this was not the definitive examination of reading that the title perhaps suggested. Unfortunately, many people will come away from the webinar thinking “if only I could read faster”, and sign up for those adverts you used to see in newspapers that promised speed reading skills, and hence bring you success in your career, your marriage and your leisure time.