In my last blog I showed how Teachers of English as a Foreign Language had abrogated to their profession  the title of Applied Linguistics.

The idea of applying these insights to the language of native speakers of English was squashed for a generation by well-meaning but dogmatic linguists in the sixties, who excoriated Basil Bernstein’s early work which had suggested that the difference between educated and less well- educated pupils in English secondary schools might be a linguistic deficit in the language of working-class children.

Unlike those linguists, I had substantial experience in teaching a mixture of middle and working-class pupils in English secondary schools, and I always thought that it was a shame that Bernstein’s early work was not followed up. A decade or so later I was struck by a talk by Thelma Henderson (University of Warwick), a fellow student on the M.A. course in Applied Linguistics at Essex University in 1975. She described her problems in teaching the appropriate language to her A-level pupils in East Africa. I remember thinking that her problems in teaching academic English to non-native speakers sounded very similar to mine in teaching A-level English to native speakers in England.

Once you see the problem as a problem of language, rather than a deficit in intelligence, something can be done about it. At a Systemics Conference in 1976 I heard Eugene Winters describing his analysis of the discourse structures of academic English (using articles from The New Scientist). I thought then that an understanding of those structures would be a quick way of improving the written English of native-speaking pupils.

It was not until the eighties that I was free to put this idea into practice: Book 3 of Breakthrough to Learning takes pupils systematically through Winters’ discourse structures (Problem/Solution, General/Particular, Compare/Contrast), using Michael Hoey’s early book On the Surface of Discourse.

It was my experience of workshops writing materials for SWAPO refugees in Zambia that led to my own contribution to the particular features of Academic English. Our task was to improve on some earlier materials written to teach English to the elite of a future independent Namibia. These very bright radicalised young adults had no teachers and were semi-literate in any language. Although we wrote materials for tape recorders, it was optimistic to think they would work in the desert camps where SWAPO was concealed from the South African army. This brought the members of the workshops up against the very basics of communication.

Before I left for Zambia, I had drafted a paper comparing two pieces of literary criticism, both about George Eliot’s Daniel Deronda, one by the traditional critic, F.R. Leavis, and the other by the feminist critic, Catherine Belsey. I showed how the different grammatical structures embodied the different philosophical positions of the two writers.  When I returned from Zambia, I came back to my paper and realised that what both pieces had in common was more interesting than what made them different. They were both extremely esoteric pieces of academic English that few people would understand (or indeed wish to understand).

I set about analysing the two pieces at the levels of the word, the sentence and discourse. The word proved to be the most crucial one. Various measures of reading difficulty had singled out long words and long sentences as indicators of reading difficulty. Not surprisingly, I found the same feature in my analysis of the two texts. However, I went beyond the mere measurement of the length of the words to examine why the words of academic discourse are so long. It is because they are abstract words, created out of short concrete words with or without the addition of a suffix. Also this structure is concealed in English by the fact that the concrete word itself is usually derived from Latin (or Greek) not the basic Anglo-Saxon of English.

This preponderance of long abstract words is not there for decoration: it makes possible the abstract frameworks (variables) which the scholar imposes upon the data of their subject.

This is all explained with examples in the article (Mason, Mason and Quayle 1992), which can be downloaded from this website and the background material on abstract language on my companion website www.languageofideas.co.uk