Hafta

Here's a sentence I read on this website (http://www-personal.umich.edu/~jlawler/aue/hafta.html)
Do you agree? I am nor sure, because I think I often hear Americans say "havta" with a V sound anyway.

<<...have to in its sense of must is canonically pronounced hafta /haeft@/, always with an /f/, never with a /v/, no matter how slowly or rapidly one is speaking. >>

I have heard both "hafta" and "havta".

I would go as far as to postulate that, in some cases, "havta" would seem to represent a perceptible increase in semantic emphasis.

I would say the opposite -- when I really mean something, I say HAFFta or even HAFF to. If it's not as emphasized, the F is not as pronounced, although I still don't think it's ever a full V for me.

I agree (as a native AmEng speaker from Massachusetts); I always pronounce "have to" with [f] (and "has to" with [s]). [v] doesn't sound natural to me there - "We halve to go." :)

This pronunciation scheme does provide the useful ability of distinguishing between "have to" and "have two". "I hafta fish" and "I have two fish" mean two completely different things. But I wouldn't go so far as two say it's the only "correct" pronunciation.

The word stresses would usually be entirely different in "I have to fish" and "I have two fish" -- they wouldn't be mistaken for each other anyway.

Another reason they wouldn't be confused is that people usually pronounce "to" as /t@/ in casual speech but they never pronounce "two" that way.

At least here in Milwaukee, there are three different normal pronunciations of "have to" outside careful speech, [ˈhɛ̯̆æ̆ftʲ(ʰ)ʉ̯u]/[ˈhɛ̞ftʲ(ʰ)ʉ̯u], [ˈhɛ̯̆æ̆ftʲəː]/[ˈhɛ̞ftʲəː], and [ˈhɛ̯̆æ̆fəː]/[ˈhɛ̞fəː], from most to least stressed. What is notable about this is that it is quite normal in everyday speech to not have any such reduction at all; the pronunciation [ˈhɛ̯̆æ̆ftʲ(ʰ)ʉ̯u]/[ˈhɛ̞ftʲ(ʰ)ʉ̯u] is not at all an anomaly introduced by the influence of higher registers but rather is the normal stressed pronunciation of "have to".

That said, pronunciations of "have to" with /v/ rather than /f/, that is [ˈhɛ̯æv̥tʲ(ʰ)ʉ̯u]/[ˈhɛ̞ːv̥tʲ(ʰ)ʉ̯u], feel just wrong to me outside the very highest registers. Even then, I would probably be inclined to just use "must" rather than "have to" in such registers, as the use of "have to" would just not really fit in to such registers IMHO.

Oh, I forgot to mention that in all the pronunciations listed above, it is actually the norm to pronounce them without the initial [h] in the dialect here, even when stressed. Yet at the same time, the initial [h] still shows up in all the cases listed above at times as a matter of free variation, even though it is more likely to show up when stressed and in higher registers.

<<Oh, I forgot to mention that in all the pronunciations listed above, it is actually the norm to pronounce them without the initial [h] in the dialect here>>

I see Travis, I remember talking about h-dropping in "have" forms some time ago. Generally speaking, I think h-dropping in all forms and usages of "have" is common in most dialects, in fast speech, when not stressed, but with the exception of negative contracted forms such as "hasn't" and "haven't". Those negative forms drop the H much more rarely, if ever.

What do you think? I know your accent has that too, but would you be able to consider this more generally, considering common dialects in the US in general?

>>What do you think? I know your accent has that too, but would you be able to consider this more generally, considering common dialects in the US in general?<<

I would say that it is by far norm in unstressed positions to have h-dropping in "have" and, for that matter, practically all "grammar words" in the vast majority of English dialects. I would be very much surprised if I encountered an English dialect which consistently lacked such h-dropping.

On the other hand, I think it is much less common to have h-dropping in "have" and other grammar words when stressed in English dialects that do not have general h-dropping; however, I myself am not really certain as to how widespread such stressed h-dropping of grammar words is in NAE dialects. Note that this is what my own dialect has, as it is generally (but not obligatorily) h-dropping for all grammar words, but non-h-dropping for all but the most common non-grammar words (as some very common non-grammar words such as "home" and "house" may have h-dropping here). And yes, I do normally h-drop even in cases like "hasn't" and "haven't".