Taxonomy or list of English grammatical constructs?
April 26, 2011 12:33 PM
Taxonomy (or just a list) of English grammatical constructs suitable for use as a checklist for a second language learner?
It seems like it would be easier for an ESL student to get motivated if he or she just had an overall picture of what needed to be accomplished to achieve reasonable (where "reasonable" would, of course, vary) proficiency.
Does a list of _grammar_ points exist? Of course grammar is intimately tied to vocabulary and culture, but I'm thinking of a rough tool of some sort to give some kind of sense of scope.
I realize I could start from ESL textbooks' tables of contents, or Larsen-Freeman's "The Grammar Book", or a writing handbook, but none seems a perfect fit; most grammar textbooks don't promise to be complete unto themselves -- do they? Even across multiple volumes?
Is there some kind of reference list, somewhere? A detailed phrase-structure rule set might not be what's needed here - that may be more detail than I want. I'd like a several-page list that I can look at, or maybe hand a confident advanced student, and say, "you're 50% accurate on all these items, 95% on these, and haven't even looked at these - so if we estimate X amount of time for this, this, and this, you'll be at this level of proficiency by this date" (realizing, of course, that time estimates for learning various things are a whole other ballgame, and that no grammatical point is independent).
I tried Googling [English grammar taxonomy] to non-success.
It seems like it would be easier for an ESL student to get motivated if he or she just had an overall picture of what needed to be accomplished to achieve reasonable (where "reasonable" would, of course, vary) proficiency.
Does a list of _grammar_ points exist? Of course grammar is intimately tied to vocabulary and culture, but I'm thinking of a rough tool of some sort to give some kind of sense of scope.
I realize I could start from ESL textbooks' tables of contents, or Larsen-Freeman's "The Grammar Book", or a writing handbook, but none seems a perfect fit; most grammar textbooks don't promise to be complete unto themselves -- do they? Even across multiple volumes?
Is there some kind of reference list, somewhere? A detailed phrase-structure rule set might not be what's needed here - that may be more detail than I want. I'd like a several-page list that I can look at, or maybe hand a confident advanced student, and say, "you're 50% accurate on all these items, 95% on these, and haven't even looked at these - so if we estimate X amount of time for this, this, and this, you'll be at this level of proficiency by this date" (realizing, of course, that time estimates for learning various things are a whole other ballgame, and that no grammatical point is independent).
I tried Googling [English grammar taxonomy] to non-success.
You could try to work something up from TOEIC scores and/or reference books?
posted by No-sword at 1:17 PM on April 26, 2011
posted by No-sword at 1:17 PM on April 26, 2011
Nomyte - why didn't the projects get anywhere? Is the problem too complex?
The CGEL is interesting, but maybe not practical, as you say.
posted by amtho at 1:49 PM on April 26, 2011
The CGEL is interesting, but maybe not practical, as you say.
posted by amtho at 1:49 PM on April 26, 2011
Look to the EFL world. Michael Swan's Practical English Usage goes into detail on all kinds of forms and how they are used.
posted by mdonley at 2:35 PM on April 26, 2011
posted by mdonley at 2:35 PM on April 26, 2011
Nomyte - why didn't the projects get anywhere? Is the problem too complex?
A combination of client disinterest and problem complexity. It's really difficult to choose what to assess — there isn't really a literature on what language features predict overall proficiency. It's also obviously very dependent on the language (we weren't working with English). In the end we went with a big handful of tricky grammatical structures and phonological markers that (we hoped) would produce a big range of performance in our participants. Some did, some not so much, which was kind of interesting on its own, in a purely academic way.
Sadly, the problem with data-driven diagnostic assessment as a useful tool should be pretty apparent: the goal is diagnostic, not instructional. So even if you create a tool that classifies learners very accurately, it doesn't really open up a path forward for them. It's not as if learning to do well on the assessment is going to improve your overall proficiency. The learner needs to learn all of the language. The assessment is just a handy indicator of their learning state, it's not that it somehow contains the keys to the language.
posted by Nomyte at 5:52 PM on April 26, 2011
A combination of client disinterest and problem complexity. It's really difficult to choose what to assess — there isn't really a literature on what language features predict overall proficiency. It's also obviously very dependent on the language (we weren't working with English). In the end we went with a big handful of tricky grammatical structures and phonological markers that (we hoped) would produce a big range of performance in our participants. Some did, some not so much, which was kind of interesting on its own, in a purely academic way.
Sadly, the problem with data-driven diagnostic assessment as a useful tool should be pretty apparent: the goal is diagnostic, not instructional. So even if you create a tool that classifies learners very accurately, it doesn't really open up a path forward for them. It's not as if learning to do well on the assessment is going to improve your overall proficiency. The learner needs to learn all of the language. The assessment is just a handy indicator of their learning state, it's not that it somehow contains the keys to the language.
posted by Nomyte at 5:52 PM on April 26, 2011
Wouldn't that depend on the data that drives the assessment? I guess I assumed you were using real-world output (speech or written text), analyzing it for errors, and identifying which errors were most common and were clustered together -- then creating an assessment from that. Identifying common errors, especially if the groupings proved illuminating, at different stages of learning might provide focus for instruction, mightn't it?
posted by amtho at 6:00 PM on April 26, 2011
posted by amtho at 6:00 PM on April 26, 2011
Here's the taxonomy and order I've used for nine years with superb results:
1. S-V agreement
2. Pronoun agreement
3. Pronoun case
4. Parallelism
5. Comparison
6. Modifiers (misplaced/dangling)
7. Verb agreement/tense up to and including subjunctive mood. Special focus on past perfect and participles
8. Clause coordination/subordination
9. Collocation/diction/phrasal verbs (no way to make this complete, but there are some great lists out there that serve as solid starting points).
VERY important that feedback at each step be focused--that is, in the S-V phase, don't comment on/edit/correct all the unlearned errors; just focus on the s-v stuff. THEN it should get cumulative, with special emphasis on the skill at hand.
posted by Joseph Gurl at 9:23 PM on April 26, 2011
1. S-V agreement
2. Pronoun agreement
3. Pronoun case
4. Parallelism
5. Comparison
6. Modifiers (misplaced/dangling)
7. Verb agreement/tense up to and including subjunctive mood. Special focus on past perfect and participles
8. Clause coordination/subordination
9. Collocation/diction/phrasal verbs (no way to make this complete, but there are some great lists out there that serve as solid starting points).
VERY important that feedback at each step be focused--that is, in the S-V phase, don't comment on/edit/correct all the unlearned errors; just focus on the s-v stuff. THEN it should get cumulative, with special emphasis on the skill at hand.
posted by Joseph Gurl at 9:23 PM on April 26, 2011
While that's much less detailed than I'd originally envisioned, Joseph, I actually think it will be very helpful - the smaller parts implied by the more general items should be easy to discover. Thank you!
posted by amtho at 4:44 AM on April 27, 2011
posted by amtho at 4:44 AM on April 27, 2011
A really good resource (for teacher and student alike) is Rules for Writers by Diana Hacker.
posted by Joseph Gurl at 7:15 AM on April 27, 2011
posted by Joseph Gurl at 7:15 AM on April 27, 2011
The British Council / EAQUALS Core Inventory is a taxonomy of the English language based around key language points at various CEF levels.
(I try to stay fairly unidentifiable on MeFi, but I should state that I have some connection to this document.)
posted by Busy Old Fool at 12:19 PM on April 27, 2011
(I try to stay fairly unidentifiable on MeFi, but I should state that I have some connection to this document.)
posted by Busy Old Fool at 12:19 PM on April 27, 2011
Yes! That's extremely, extremely helpful, BOF. Thank you so much!
posted by amtho at 6:41 PM on April 27, 2011
posted by amtho at 6:41 PM on April 27, 2011
This thread is closed to new comments.
I've worked on projects that tried to develop data-driven language proficiency assessments ("You got xx% correct on a test of construction X, your overall proficiency level is Y."). Sadly, those projects never really got anywhere.
I think a lack of comprehensive, quantitative approaches to measuring proficiency is a huge blank spot in the field of second language acquisition. Even the federal government, which has a huge number of employees who get bonuses on the basis of their language proficiency, relies heavily on impressionistic measures like the oral proficiency interview.
posted by Nomyte at 12:53 PM on April 26, 2011