Could English prosodic stress be considered phonemic?

[salem]

...in negative statements, at least?

Take the sentence, "She didn't steal my purse":
if stressed as "She didn't steal my purse", the implied meaning is that it was not her but someone else who stole the purse;
as "She didn't steal my purse", there was no purse-stealing whatsoever;
as "She didn't steal my purse", she did something other than stealing it;
as "She didn't steal my purse", she stole someone else's purse but not the speaker's;
and as "She didn't steal my purse", she stole something other than the purse.

Anyways, stress seems to be used for designating which word gets the negativity. Are there any other languages that do this?

[KathTheDragon]

This works for positive statements too. Consider the positive form of the sentence you gave.

She stole my purse - 'Twas she and none other who did it.
She stole my purse - Emphasis on the stealing as opposed to another form of taking the purse, like borrowing.
She stole my purse - Mine and none other's.
She stole my purse - Drawing attention to the purse, as opposed to anything else being stolen.

[CaesarVincens]

This really can't be called phonemic in a proper sense because it has to do with pragmatics rather than distinctions of sounds.

A more phonemic distinction of stress in English is the verb-noun pairs like increase and increase. Though again, this might not be phonemic per se but rather different stress rules for different parts of speech.

I believe using stress for pragmatic emphasis or topicalization is fairly common among the methods of such things, but I don't know the statistics on it.

[Imralu]

as "She didn't steal my purse", there was no purse-stealing whatsoever;

I disagree strongly that this sentence implies there was no stealing.

[salem]

as "She didn't steal my purse", there was no purse-stealing whatsoever;

I disagree strongly that this sentence implies there was no stealing.

Really? What would you consider it to imply, then?

[Porphyrogenitos]

as "She didn't steal my purse", there was no purse-stealing whatsoever;

I disagree strongly that this sentence implies there was no stealing.

Really? What would you consider it to imply, then?

It could easily be used in a context in which there had been stealing.

"Just face the truth - she's the reason your purse is gone."
"She didn't steal my purse!"

[salem]

as "She didn't steal my purse", there was no purse-stealing whatsoever;

I disagree strongly that this sentence implies there was no stealing.

Really? What would you consider it to imply, then?

It could easily be used in a context in which there had been stealing.

"Just face the truth - she's the reason your purse is gone."
"She didn't steal my purse!"

Ah, but in that context I'd expect stress to be both on "didn't" and "purse", as opposed to only on "didn't". Unless you have "steal my purse" all at the same pitch? Which seems kind of unnatural, so I'd hazard a guess not.

[Imralu]

as "She didn't steal my purse", there was no purse-stealing whatsoever;

I disagree strongly that this sentence implies there was no stealing.

Really? What would you consider it to imply, then?

Giving a negative auxiliary greater stress generally seems to indicate contradiction (ie. you claimed that she stole my purse - that's incorrect). To me, it lacks any further connotations about who or what did anything, it's simply a sentence about what someone didn't do.

[zompist]

Phonemic certainly isn't the right word here. You're not talking about whether a sound is or isn't part of the phonemic inventory. And it's not exactly as if you're asking if stress is phonemic in English-- it clearly is, along the lines suggested by CaesarVincens.

I confirmed with my wife that Spanish works exactly the same way. You can say "La chica no me robo la bolsa" to indicate that it was someone besides la chica, and so on.

[Salmoneus]

Phonemic certainly isn't the right word here. You're not talking about whether a sound is or isn't part of the phonemic inventory. And it's not exactly as if you're asking if stress is phonemic in English-- it clearly is, along the lines suggested by CaesarVincens.

I confirmed with my wife that Spanish works exactly the same way. You can say "La chica no me robo la bolsa" to indicate that it was someone besides la chica, and so on.

Is stress really phonemic in English? There are plenty of verb/noun stress differences, but can these ever actually be the sole differentiation between two utterances? Isn't that precluded by verbs and nouns not being able to go in the same places anyway? So the stress would be an accidental property, not a distinguishing one (you could even describe it as there being one word and a prosodic rule - certain verbnouns are stressed this way when in this part of a sentence and that way in another).

But maybe somebody can think of a genuine minimal pair?

[CaesarVincens]

I doubt English has phonemic stress in the way that say languages with noun-noun or verb-verb distinctions regularly occur.

On wiki, it mentions compounds in English with different meanings but the same components as its example of paper bag (a bag made of paper) vs.paper bag (a bag for [news]paper[s]).

[KathTheDragon]

Does that count as a minimal pair?

[CaesarVincens]

It could. I'm not sure it does per se. Minimal pairs should be ideally the same part of speech (so they are used the same way) with one pair pf close phonemes and the rest the same such as /tun/<tune> and /dun/<dune>.

This is to say, if the distinction is only of the stress and they are used the same way then it could be a minimal pair, but...

With paper bag, the stress seems less a phonemic and more a pragmatic distinction.

For me, I think:
It's a paper bag not a plastic bag. (the bag is made of paper)

A bag for paper(s) is probably stressed for me more like paper bag

The wiki doesn't cite any sources in regards to the compounds, so...

The difference is probably that for me, paper bag is a lexical compound and a paper bag is an ad hoc (syntactic) compound.

Compare "He lives in the White House" and "He lives in the white house"

[salem]

as "She didn't steal my purse", there was no purse-stealing whatsoever;

I disagree strongly that this sentence implies there was no stealing.

Really? What would you consider it to imply, then?

Giving a negative auxiliary greater stress generally seems to indicate contradiction (ie. you claimed that she stole my purse - that's incorrect). To me, it lacks any further connotations about who or what did anything, it's simply a sentence about what someone didn't do.

Fair enough. "No purse-stealing whatsoever" was a bit strong, and I apologize for that.

Phonemic certainly isn't the right word here. You're not talking about whether a sound is or isn't part of the phonemic inventory. And it's not exactly as if you're asking if stress is phonemic in English-- it clearly is, along the lines suggested by CaesarVincens.

I confirmed with my wife that Spanish works exactly the same way. You can say "La chica no me robo la bolsa" to indicate that it was someone besides la chica, and so on.

Would "meaningful" work, then?

[zompist]

Is stress really phonemic in English? There are plenty of verb/noun stress differences, but can these ever actually be the sole differentiation between two utterances? Isn't that precluded by verbs and nouns not being able to go in the same places anyway?

I'm not sure you can appeal to syntax in a phonological rule... wouldn't Occam, at least, suggest that it's preferable just to allow that stress is phonemic? Besides, it's not consistent— words like "regret" and "defeat" don't have the shift.

But anyway, how about rocket/Rockette? Or Barack/barrack, or barette/Barrett, or natal/Natal. (Edit: poutine/Putin.)

(We have to look at somewhat unusual words, as English stress is generally predictable. Nouns with final stress aren't rare, but they're certainly a minority.)

[CaesarVincens]

I'm not sure you can appeal to syntax in a phonological rule... wouldn't Occam, at least, suggest that it's preferable just to allow that stress is phonemic? Besides, it's not consistent— words like "regret" and "defeat" don't have the shift.

But anyway, how about rocket/Rockette? Or Barack/barrack, or barette/Barrett, or natal/Natal.

(We have to look at somewhat unusual words, as English stress is generally predictable. Nouns with final stress aren't rare, but they're certainly a minority.)

I think those are some good cases for marginal phonemic differences in stress. Foreign names and words are somewhat suspect when talking about this sort of thing, but some become naturalized enough to be a good candidate.

I wouldn't look at the syntax either but morphology can play a role for some cases.

[Lyhoko Leaci]

Is stress really phonemic in English? There are plenty of verb/noun stress differences, but can these ever actually be the sole differentiation between two utterances? Isn't that precluded by verbs and nouns not being able to go in the same places anyway?

I'm not sure you can appeal to syntax in a phonological rule... wouldn't Occam, at least, suggest that it's preferable just to allow that stress is phonemic? Besides, it's not consistent— words like "regret" and "defeat" don't have the shift.

But anyway, how about rocket/Rockette? Or Barack/barrack, or barette/Barrett, or natal/Natal. (Edit: poutine/Putin.)

(We have to look at somewhat unusual words, as English stress is generally predictable. Nouns with final stress aren't rare, but they're certainly a minority.)

Don't forget desert (n)/desert (v)/dessert, which aren't unusual at all.