The best thing you can get from doing an MA is a hightened awareness of sloppy thinking. Appeals to science are often used to get you to drop your guard, and can even make you doubt your own common-sense reaction to obvious nonsense. Here’s the story of how a sloppy, poorly-written article in the limited domain of word recognition led to the claim that people who read texts never skip words.
On Jan 13th, Susan Godsland tweeted:
Eye Movements and Reading ”Reading is accomplished with letter-by-letter processing of the word” http://www.readingrockets.org/article/eye-movements-and-reading
I clicked on the link, and read the article which included this:
(W)e do not skip over words, read print selectively, or recognize words by sampling a few letters of the print, as whole language theorists proposed in the 1970s. Reading is accomplished with letter-by-letter processing of the word (Rayner, Foorman, Perfetti, Pesetsky, & Seidenberg, 2001, 2002). Fluent readers do perceive each and every letter of print. Thus, we can distinguish casual from causal, grill from girl, and primeval from prime evil. Better readers process the internal details of printed words and match them to the individual speech sounds that make up the spoken word. Even when “chunks” are recognized, they can be analyzed into their individual phoneme-grapheme correspondences on demand.
I replied to Godsland saying:
The article says “we do not skip over words, read print selectively”. We do.
And Pamela Snow, who knows nothing about me, chimed in with the daft comment
I’ll defer to @KevinWheldall, @maxcoltheart and/or @annecastles – all better qualified to comment!
at which point, I thought I’d better have my say elsewhere.
The article in the Reading Rockets website gives no evidence for the claims made, fails to put research on word identification into the wider context of reading comprehension, and fails to make it clear that its aim is to support the view of the Reading Rockets team that phonics is the best way to teach children how to read. There is no hint to the unwary reader who clicks on the link provided by Godsland that the Reading Rockets website is devoted to the promotion of phonics – one of the two schools of thought on teaching reading to young children, and one of the protagonists in a long-standing, acrimonious feud among educationists in the USA.
On one side are proponents of a phonics approach in reading, and on the other, those who advocate a “whole language” approach. Proponents of phonics say that the perceived general decline in reading in the USA is the result of the adoption of whole language instruction, and they support their case by saying that scientific studies “prove” that phonics instruction produces better reading scores than other methods. Those who take the phonics approach say that reading is a process of phonological coding – we assign speech-based codes to the printed text we read – and in support of their case they emphasise the research they’ve done on eye movements. Using instruments which measure eye fixations and saccades while subjects read, the studies try to give supporting evidence for the claim that readers use phonological code, rather than accessing meaning directly from the visual word form. This explains the importance of the claim trumpeted by Godsland in her tweet that ”reading is accomplished with letter-by-letter processing of the word”, the implication being that we should start by teaching young children letters and then teach them how to “sound out” words.
The “whole language” approach starts from the premise that we can access meaning directly from the visual stimulus of the printed words, and adopts the constructivist theory of education whereby children learn by connecting new knowledge to previously learned knowledge. Many will recognise the influence of Vygotsky’s “zone of proximal” development (children learn things a bit beyond their current understanding through help from more knowledgeable peers or adults) here. Phonics is said by some to represent a bottom-up approach (readers “decode” the meaning of a text), while whole language is said to represent a top-down approach (readers construct their own personal meaning for a text by using their prior knowledge to interpret meaning). It seems to me to be prtetty obvious that a hybrid approach, which combines the phonic and the “whole language” approaches is needed, but the mention of bottom-up and top-down approaches leads to the wider question of reading comprehension and the absurdity of Godsland’s myopic comments.
When Godsland endorsed the Reading Rocket claim that reading is accomplished with letter-by-letter processing of the word and that we don’t skip words or read print selectively, she, like the authors of the article, failed to properly define the domain of the studies referred to. Word recognition is a small domain, important in understanding the initial development of the reading skill, but not so relevant to more macro-orientated considerations of reading comprehension. In one tweet Godsland cites the excellent article by Asby and Rayner (2007) as evidence supporting her assertion, completely failing to appreciate that Asby and Rayner are not saying (how could they!) that readers never skip words or read selectively, but rather that there is evidence to support the claim that skilled readers, especially in their first encounter with a text, tend to process texts thoroughly “from the letter level on up” rather than engage in a “psycholinguistic guessing game.” This reasonable claim is made in support of the argument that young children should be taught how to be good readers, not just encouraged to read a lot, and that consequently children should be taught letter-to-sound mapping. What it doesn’t do is settle the argument about how much weight should be given to phonic and “whole language” approaches, or give any support whatsover to Godsland’s ostrich-like misinterpretation of the short, very poorly-written article published in Reading Rocket.
When those of us interested in SLA look at how people read, we have a different point of view to that of primary teachers helping young children to read well. We’re interested more in reading comprehension than in word recognition, and we look at how readers use bottom-up and top-down processing, use strategies such as skimmming and scanning, and approach the text from a Text as a Vehicle for Information (TAVI) and Text as a Linguistic Object (TALO) perspective. In all such work, we rightly assume that readers often skip words and read parts of a text selectively.
So, keep on your toes everybody, and enjoy your reading .-)
Asby, J. and Rayner, K. (2007) “Literacy Development: Insights from Research on Skilled Reading” In Dickinson and Neuman (eds) Handbook of Early Literacy Research, Volume 2. N.Y. The Guilford Press.