View the edtech Discussion Logs by month
View the Prior Message in edtech's December 1992 logs by: [date] [author] [thread]
View the Next Message in edtech's December 1992 logs by: [date] [author] [thread]
Visit the edtech home page.
I don't often contribute on this list, though I find the informa- tion that passes through here very useful. The recent discussion of criteria and programs to assess readability and writing skills has mainly been pretty practical, but I think people should be aware of the severe limitations of these approaches. Most of them are based on either (1) grammati- cal, and even spelling, punctuation judgments that are DECONTEX- TUALIZED, or (2) superficial criteria with no basis in an under- standing of how we make meaning with text (linguistic text semantics). On (1): Unlike a human reader, these programs do not take into account the meaning, or even a guess at the meaning, of a form. A correctly spelled word, or a possible grammatical construction, will not be flagged even if it is wrong in context. These pro- grams are useful to a writer to flag POSSIBLE errors, but thene require writer judgment. And they miss a lot of actual errors. On (2): Criteria dependent on frequency of use of words, types of grammatical constructions, sentence length, etc. are extremely superficial. Apart from familiarity with the subject matter and genre conventions of a particular specialist discourse (and all discourses are specialist in one sense or another), readability of academic material depends in large part on none of these vari- ables very critically (at least within the range of normally enountered styles, not including artificial examples). What it does depend on, very often, is the stylistic shift in academic writing away from the grammar patterns of spoken, conversational language: the use of nonfinite constructions, verbal process nominalizations, use of empty abstract relational verbs in place of explicity clausal conjunctions, etc. A very good account of this, and a practical quantitative index, can be obtained from the work of the linguist Michael Halliday as described in his book *Spoken and Written Language*, Oxford University Press, 1989. ---------------- JAY LEMKE. City University of New York. BITNET: JLLBC@CUNYVM INTERNET: JLLBC@CUNYVM.CUNY.EDU