Hmm, I'm not sure where you guys are getting your rules from. I mean, about the only snippet I could find on anything like this (in grammars published in English-speaking countries and aimed at native-speaker teachers), and that made any sense at all to me, was this:
Leech, on page 79 & 95 in his [i]Meaning and the English Verb[/i], Third edition, wrote:MUST: 117 C. LOGICAL NECESSITY/reasonable assumption (common) This use of must normally has no negative or question from; but see section 137b.
137b. [Regarding auxiliary negation - FH] Must not (='logical necessity') has a dubious status in BrE, but seems to be gaining ground, particularly in AmE. Notice the logical equivalence of must not in this sense and can't (='impossibility'): She must not be on campus today is virtually equivalent to She can't be on campus today. Less problematic in BrE is the use of the contracted form mustn't in tag questions following must in the 'logical necessity' sense: They must have hundreds of people looking for jobs, mustn't they?
Anyway, perhaps the answer to "your" question José is that those two people over on Englishforums.com (ooh and maybe now Maciek too LOL
) are maybe a bit confused and/or too rule-bound (they're non-native speakers, right?), because for a supposed pair of grammatically closely-related clauses (in a genuine tag question structure at any rate, and bear in mind that some grammars call tags
clause tags!) to be the same propositionally, they surely have to contain matching auxiliaries. And I don't see (contrary to what Rafaelinrio is suggesting) what would be wrong epistemically-speaking with
You must be a teacher, mustn't you/You have to be a teacher, don't you (tag has falling intonation in each case) - imagine a game show where people are guessing and forming suppositions about the professions of mystery guests. (That's not to say that using 'must' and certainly not 'have to' in this way will make the speaker's reasoning always sound "considered" or "open" enough - these modals are a bit "strong" compared to other ways of expressing the modality e.g.
You're probably...). Certainly as a native speaker I'd definitely prefer to say either of those sentences than Rafael's "alternative" (IMHO*)
He must be a teacher, isn't he? And did you spot the
'Question tag for "It must be a cat, _____...' related discussions link on the page you gave, José? (Here it is just in case:
http://www.englishforums.com/English/Qu ... r/post.htm ). That seems to have a lot more native-speaker input on it and contain far sounder advice and examples, IMHO.
That's not to say that questions containing auxiliaries differing to ones earlier in the sentence don't occur, but they are probably different enough from tags that we should really call them something else to avoid the confusion (how about non-tags, or NTs for short, in the meantime?).
Here are a few examples off the top of my head (italic is the example itself; underline shows timing and punctuation different to tag questions; bold is what I personally would prefer to say/stress):
1)
You must be tired - oh, aren't you?/you aren't?
2) (in reply to your "Hey Fluffy...even when it's logical conclusion/deduction": The 'must' in 'You must be tired' can really ONLY be logical deduction/epistemic modality; that is, this is obviously NOT deontic/"obligational" modality we're talking about here).
I mean, I've never seen anybody commanded to be tired - have you?, versus
- but were you/have you been'? (Note in this particular example how the auxiliary in the NT can differ from that in the statement, DEPENDING ON THE MEANING ONE WANTS TO CONVEY, but that the subject of the NT is
completely different in both cases).
Then there are Chineselindsay's and Rafaelinrio's remaining examples, which we could adapt along lines similar to the above:
*
He must be a teacher, isn't he? > He must be a teacher - what/eh, isn't he?!
*
He must fly a lot, doesn't he? > (
I'll leave you to do this one!).
Ultimately, I have never read (or certainly don't recall ever reading) rules like the ones Rafael states (and remember that I own a LOT of grammar books!), which simply do not chime at all with my native intuitions:
when we use must with the idea of obligation we use mustn't to make the tag question
when we use must with the idea of assumption/deduction we use another auxiliary verb do make the tag qusetion such as to be or do/does.
examples:
You have worked a lot. You must be tired, aren't you?
She must be the new teacher, isn't she?
Finally, another factor that might be motivating certainly Chineselindsay's thinking is that Chinese speakers (well, at least Singaporeans speaking English/Singlish at any rate! Though I've noted that CL is apparently from Suzhou, Jiangsu province, the PRC!) seem to have quite a liking for the invariable tag
is(n't) it?.