R-lessness

Something was said tonight that piqued my linguistic interest and, as I’ve been pretty light on linguistic content of late, I thought I’d share it.

We (my housemate, a few friends and I) were at the pub tonight, when we met an Irishman, from Clare county. My housemate is named after a capital city of a country just north of Australia, which is a non-typical Anglo-Australian name. When he introduced himself to said Irishman, he said his name was ['mo:sbi]. Upon hearing this, the Irishman repeated it back, to make sure he understood correctly, and said ['moɹsbi], with a clear rhotic segment. I am sure he didn’t know the name beforehand, just to mitigate against any possible priming effects.

What interested me was the fact that the Irishman, who speaks with a rhotic accent (that is, with post-vocalic ‘r’) managed to extrapolate the correct phonemic form of my housemate’s name, including the ‘r’, even though the way it was presented to him was entirely r-less.

Is it the case then, that some long vowels in r-less Englishes are assumed by r-full speakers as being a short vowel followed by an /r/?

This is an area of linguistics that I know very little about; how speakers of different dialects and accents of English manage to overcome the accentual differences between the ways in which they speak and deduce the right form. The same goes for the North-American pronunciation of my nickname; I pronounce it [hɔs], yet North-Americans have no trouble at all converting that directly to [ha:s].

On the other had, it may be a neutralisation effect; if I were to hear the name [ha:s] in a typical rhotic North-American accent, I may permissibly take it to translate into my accent as either Hoss, my nickname, or Haas, as in Mary Haas, for instance.

</stream of consciousness>

5 Responses to “ R-lessness ”

Comments:

Piers Kelly says:
August 14th, 2008 at 10:02 am

This is my take on what’s going on:
When I was little I used to mentally transliterate everything that I heard so that I would ‘see’ a written representation of speech in my head. I’m not so aware of this any more but I have a feeling that many people do this on a subconscious level. So my suspicion is that your Irishman has heard ['mo:sbi], ‘transliterated’ it as ‘Moresby’ then reproduced it again as ['moɹsbi] via the mentally reconstructed written representation. He would have to have already understood something about Australian English pronunciation and its relationship to the spelling system to do this. I reckon that the English spelling system in all its irregular glory has more influence on our pronunciation than we imagine. Personally I don’t think in IPA! I’ve written something along these lines here: http://perezsez.blogspot.com/2008/08/say-it-dont-spray-it.html

jangari says:
August 14th, 2008 at 10:16 am

Cheers, Piers¹.

Yeah, that’s what I think is going on, but it’s interesting that long vowels in ɹ-less dialects are perceived phonemically by speakers of ɹ-full dialects as being a normal length vowel followed by an /ɹ/, right?

As if the ɹ-dropping rule was: Whenever you get the sequence /Vɹ/, assimilate the /ɹ/ to the vowel, thereby lengthening it.

So, to reverse this, which is what ɹ-full speakers will have to do to accurately understand people: Whenever you get a long vowel, consider (at least) the possibility that it was underlyingly a /Vɹ/ sequence.

I guess the question then becomes ‘does this get overgeneralised’? I think it probably does. as I said in a conversation with Michael Walsh the other day, the Bostonian English question [wɛ:z ma:li] has three possible interpretations, and therefore three answers: “I don’t know where she is”, “He’s probably buried in Jamaica” and “Next to Mauritania”.

¹Sorry Piers, I couldn’t resist.

Claire says:
August 15th, 2008 at 3:05 am

But the underlying long vowels are a little different acoustically from the ones that map to Vr. Not all of them, and not for all speakers, but enough of them for enough speakers. Also, I bet there’s frequency involved. I bet that oː maps onto oɹ (really ɒɹ, maybe) more often than onto oː.

Incidentally, weːz isn’t a diphthong for Boston English speakers – it’s wiəz.

David Marjanović says:
December 28th, 2008 at 12:25 am

When I was little I used to mentally transliterate everything that I heard so that I would ’see’ a written representation of speech in my head. I’m not so aware of this any more but I have a feeling that many people do this on a subconscious level.

I still do that — not for everything maybe, but for much. And for Chinese, I do it in Pinyin.

I know of a native English speaker who does it too.

But the underlying long vowels are a little different acoustically from the ones that map to Vr. Not all of them, and not for all speakers, but enough of them for enough speakers.

In my native non-rhotic German, postvocalic /ʀ/ manifests as the second half of a clear diphthong, [ɐ̯]. This also often happens in non-rhotic English, though apparently not for all speakers, for [ɜ] less commonly than for the others, and if it still happens, the [ɐ] (or [ə] or whatever) is considerably shorter than the first part of the diphthong, which is outright long (if stressed), while in… at least southeastern German both parts have the same length unless the first part is phonemically long.

David Marjanović says:
December 28th, 2008 at 12:28 am

Listen to Schwarzenegger if you don’t know what I mean. “/foːr moːr jiːrs/”!!!