I gave the keynote address at the AWELL Conference a few months ago, and one of the practical suggestions I included in my talk for writing teachers working with English language learners was to try to draw on corpus-based grammars and style guides based on real language use vs. those that try to pass off personal style preferences as hard and fast “rules”.
Here is an excerpt from that talk, with the references I included. The friendly folks of #tleap on Twitter pointed me toward some of them. I haven’t personally used all of these in the classroom, so welcome any comments and feedback on how well any of them hold up to real-life use.
I suggest working from a grammar of English that draws on data from language corpora to determine what the constitutive rules of English are. For those of you that don’t know, a language corpus is a “body” of hundreds of millions of lines of text of the language in use. Some corpora compile thousands of newspaper and magazine articles, others academic texts, others fiction writing. So you have a huge repository of the language as its actually used by real people in real contexts. And you can analyze the language in a corpus using technological tools to determine the rules of a language based on how the language is actually used. So for example, if you’re wondering “Can I really never end a sentence in a preposition in an academic text?” corpus data will tell you that in fact, you can; lots of great writers do it all the time. They can’t all be wrong. Use determines the rules in the descriptive approach, remember? So that tells you that not ending a sentence in a preposition must simply be a style choice. (Though it’s debatable if it’s something everyone does literally all the time, if you’d really want to give much time to it. )
So here are some English Grammars that are corpus-based:
Some of them have “learner’s editions” which are aimed at learners of English at lower levels of proficiency. They all endeavor to present the rules of English as it is actually used. Several will contrast frequency of use of a particular item in different contexts, for example in fiction writing versus academic writing.
The Garner texts (2016 editions), for example, highlight particular language points where there is disputed use. For example, maybe the traditional style guides say one thing, but lots of people do another. It gives corpus data showing the real patterns of use of that particular item, how common each variant is, and if there has been change over time.
So for me, these are a more accurate choice. Why would we want to enforce grammar rules with multilingual writers that people don’t actually adhere to in real-life written academic English? And If you really want to nerd out, you can delve into corpora yourself to check out certain grammar points. The Google N-Gram viewer or the MICUSP corpus are good places to start. You may also be interested in reading linguist Steven Pinker’s “The Sense of Style”, which is written by a well-known author who grounds his advice for good writing in linguistic theory.