I once had a student who translated Chaucer's famous line (thanks to Dr. Virago, we all know it), ""Te-he,' she seyde, and clapte the wyndow to," as,
"'Teach,' she said, 'and close the window, too.'"
Insert joke here about closing the barn window after the horse is stolen.
I expect upper-division English majors to read Chaucer (and many other Middle English texts) in Middle English. I would not expect this of lower-division students in a gen ed course, but it seems to me that if you're a major, you should read Chaucer, not someone else's version of Chaucer. A few years ago, the Chaucer Review published an essay analyzing summaries of the Wife of Bath's prologue (Crags' Notes, Regal Notes, FireStarter notes), and the ways minor misstatements added up to serious distortion; I think the same thing goes for translation. Most of Chaucer's writings are poetry, so the words matter even more than in prose, and you should read for more than just plot. Told by someone else, the Miller's Tale would be just a fart joke.
Some students take to ME quite easily; others take awhile; some never really catch on. My experience suggests that it helps a lot to have (a) a large vocabulary in Present-Day English, so that you recognize archaic and dialect words more easily, and (b) a flexible approach to spelling. Further, being able/willing to sound words out helps; I suspect the student who gave the translation cited above had learned to read by recognizing the shapes of words ("whilom" came out as "William," and a host of similar errors).
I encourage students to use a translation, preferably the Harvard Chaucer Page's interlinear translation, for the General Prologue, and start weaning themselves off translations while reading the Knight's Tale. This used to be self-enforcing, when the HCP gave only the GP . . . the GP and a little more . . . but now the HCP has the whole CT, and various other reputable online sites have most of the Chaucer corpus available in translation.
So, are we to give up and assume that Chaucer, like Beowulf, will now be read only in translation at the undergraduate level? Last time I taught Chaucer, I kept to what has become my usual sequence of texts, and continued to encourage students to do their own translations and check them against the HCP, to wean themselves from the interlinear translation, to bring their assigned books to class and quote and discuss ME in their papers. But I noticed a lot of printouts being used, and there was a notable drop in participation the first day (mid-semester) we discussed Sir Orfeo, which I assigned from the TEAMS site. (Students didn't seem to get as far as acquiring the Tolkien translation, so maybe if I didn't mention the online translations of Chaucer, they wouldn't find them? I doubt it, though.) I think that when faced with ME sans interlinear translation, a lot of them boggled.
Two solutions immediately spring to mind. One, give in to the use of translations and cover a lot more ground than I currently do. (I used to cover more, but back in the day, students seemed to read better . . . please join me in a chorus of my-how-the-world-has-changed-for-the-worse-since-I-was-young.) Two, go the other direction entirely, and do fewer texts but in much more depth, so that students will become intimately familiar with the ME, and thus, I hope, will stop feeling intimidated by it and develop a real appreciation for the language and its strange shades of meaning that no longer exist.
My preference is for the second; and that leads to the problem of which texts to choose. If you were going to spend a whole semester on 2-3 Chaucerian texts (with some supporting critical and source material), which would you choose, and why?