These exercises also are not meant solely for the purpose of improving a particular document either. They are meant as exercises to get the students to think about (by typing in and searching–a form of writer response) the errors that they make.

When people spellcheck (and again I use myself as an example) they do not often learn the correction so much as “accept” the correction the way they might just make all the changes a teacher marks on a draft without learning why the correction needs to be made.

This exercise tries to encourage students to make note of their common errors (specifically grammatical and maybe broadly stylistic), to think of how to “find” them in a search, and to replace when appropriate.

Take the imaginary person who mistakes compliment for complement. Spellchecker won’t catch this. Grammar check probably won’t catch this, but if the student is aware of a tendency to make this mistake, searching for instances of compliment and complement might be an efficient way of weeding out the mistake while testing his or her knowledge of the rule.

Btw, my grammar checker caught nun of the errors in this post.

Mark, maybe I’m misunderstanding the sense in which you mean GREP. Spelling and Grammar checkers are already amazingly optimized for situations in which commonly occuring errors need to be considered, often on a case-by-case basis, then dealt with. The results of asking a writer to Find/Replace or GREP common errors will usually look like this: 1. Writer hits “Spellcheck” 2. Writer skips each checks, noting common ones on a scrap of paper. 3. Writer uses Find/Replace as instructed to manually enter an edit that could have been done using the spellchecker “Replace All” button. 4. Writer wonders what the point was.

I use GREP constantly as a programmer and when editing large documents - but not to correct common errors. Instead, I generally use it to rename, paraphrase, and restructure ( http://writerresponsetheory.org -> http://writerresponsetheory.org ). There are good examples of when you should eschew the specialized frontends to Find/Replace that are the grammar / spellchecker - but I’m not sure that grammar and spelling mistakes are among them. But perhaps we are saying the same thing in different terms….

From Bradley Morgan’s 1983 introduction to the first issue:

“For students and professors in a wide variety of disciplines, writing is???or should be???a primary tool for learning. The computer can help with all phases of the writing process, from the heuristic mustering of an idea-base to oft-neglected revision. It can provide a quantitative measure of a writer’s style???or allow a professor with a standalone system to offer detailed, student-specific comment sheets. The word processor not only saves time, conserves labor, and solves problems, but it also reinforces the traditional mission of writing programs.”

You say:

I think it is a mistake to ask students to fix their own most commonly occuring problems, especially because they often don???t know what those problems are or don???t know to what extent they can be pattern-matched.

But this is exactly what I hope to teach students to do. I imagine this exercise as coming after the first few papers in which the teacher has already given them feedback on their most common mistakes. I encourage students to keep their own lists.

If done in a lab, the teacher could help them hands-on to use the exercise, though I like your ideas for editing the text as well.

The goal eventually would be for this to be a process that they could apply before turning in final drafts but that they could use in adapt as they become more aware of their process.

I enjoy this lesson concept a lot. Two reactions:

GREP: first, it seems to me that the GREP excercise needs more thinking through, in particular claryfing to students how/why the use of advanced Find/Replace is different from using a spellchecker. For clarity, we might want to refer to the history but call call the excercise itself a Find/Replace excercise if it will be occuring inside Word / OpenOffice / etc. and not using GREP syntax - although ideally the excercise would still use wildcards. I think it is a mistake to ask students to fix their own most commonly occuring problems, especially because they often don’t know what those problems are or don’t know to what extent they can be pattern-matched. Perhaps better to ask them to clarify a key term, phrase, or conceptual label used often in their paper, e.g. “Mr. Arthur Doyle -> Sir Arthur Conan Doyle,” “people -> citizens” etc. (”America -> the United States” is a common problem, for example).

Tracking and versioning: ideally, this would round up a defined sequence of previous excercises. That is, you write something (1.0), clarify a key term (2.0), get feedback (3.0)… the trick is to decide where the “perverse versions” that break the spellchecker, grammar checker, and thesaurus should fall in the sequence.