When in March 2000 I described and published the stress rules of the Portuguese spelling system, as part of a larger project describing the pronunciation of European Portuguese, it met with quite some opposition. In Usenet group sci.lang, I insisted that the acute accent (´), the circumflex accent (^) AND the tilde (~) were all stress marks. This forced me to posit exceptions for words like sótão (attic), sotãozinho (small attic) and irmãmente (sisterly).
Opponents like António Marques and Harlan Messinger kept telling me that the tilde (~) is NOT a stress mark, but only indicates nasalisation.
They got a bit annoyed over my stubbornness. At the time I didn’t understand why, but I do now. Because they were right. This discussion, started by me on 3 March 2000 at 10:00, bore the subject “Portuguese : Where stress in words with double accents?” and Google Groups archived the thread here.
In late April 2008, the subject was discussed again, when I couldn't keep my mouth shut in response to this message by António Marques. In that thread, first I kept being stubbourn, but eventually on 1 May 2008 at 11:38 I declared myself convinced, writing:
In hindsight, I think you are right. I've learnt a lot in these past 8 years. That also means I still had a lot to learn, as I now know much more than I thought back then. I'll probably change my description on the web, and keep the old one only as a historic reference, to show how I had it wrong first.
After that eventual recognition of my mistake, it took me almost a year and a half, until September 16/17, until I finally corrected my description.
I am slow, but I learn!
Another nearly three months later, I also explained background and history of the change, in the article you are now reading.
Shortly before giving up my resistance, on 1 May 2008 at 10:46, I took the origin of the tilde into acount, as follows:
Perhaps we should look at history. The tilde ~ used to be an [sic] shorthand, for various things (as in q~ = que), but often for n or m. So ã comes from an of ana, and ão from ano or ion. From that perspective, you are right and I am not. But for today's Portuguese, my view also works.
By then I had forgotton that over 8 years earlier, António Marques mentioned that too: 6 March 2000 at 18:40
But to say that the tilde is a stress marker is useful in no conceivable way, other than holding than anything placed above a vowel is a stress marker. It just isn't. Want to know what it is? It's a superscript n!
In hindsight, what might have contributed to my unwillingness to admit that others were right, is my failure to fully understand what António Marques meant, due to what I think is a Portuguesism in his otherwise almost impeccable English (as far as I can judge that, being a non-native speaker myself).
I quote 4 March 2000 at 13:54:
It HAPPENS that words ending in a tilded syllable are nearly all acutes, so, no stress is marked - on the reverse, when they are not, a stress mark must be put on the stressed (always the next to last) syllable.
It happens that words ending in nasal sound, in liquid, or for which the last vowel is an i or a u are almost all stressed on the last syllable. SO, they are by default acute and need no stress marker. SO, words ending in tilded syllable are acute if they have no stress mark on the penultimate... but that doesn't make the tilde a stress mark no more than the i ou u are stress marks!
Here, António uses the English word acute to refer not to the diacritic that can be placed on a Latin script letter, the acute accent, but to a word that bears the stress on the last syllable. Indeed, that is one of the meanings of the Portuguese cognate of English acute, which is "agudo". The Porto Editora dictionary lists this as sense 7, as opposed to sense 8:
GRAMÁTICA diz-se da palavra que tem o acento tónico na última sílaba; oxítono
See also this table of Portuguese terms for stress pattern.