Separated by a Common Language

I'm British and I would pronounce Alicia as "aLISHya", but I've never known one so it's never come up.

Then there's the French-accented option. My Scottish aunt Teresa pronounces her name "Tə-RESS" (though some of her siblings go for Tə-REZZ), and spelling and pronouncing her name as "Tress" has become family shorthand in texts, Facebook, and the like.

@Kelvin, I'm American and I generally say the same, although it's also not uncommon for women of Latin American ancestry to prefer a pronunciation based on Spanish (Uh-lee-SEE-uh.)

In another wrinkle: I'm an Alisia ("aLEEsha") and while living in the US and the UK, pronunciation is all over the place, except when I lived in mid-Wales, where it was almost always correct. I developed the very unscientific theory that the "si" got pronounced as in Welsh (like Sian), and through good luck that was correct. Never found any evidence of that spelling being in any way related to Celtic languages, however.

A couple of points:

In the US, there's a fairly common name Marsha (eg Marsha Mason) which I suppose is a phonetic respelling of Marcia.

The UK vs US difference in pronunciation of Marcia, Alicia, et al reminds me of tissue -- tish-oo in the US, tiss-you in the UK. Except that as far as I remember from when I was a youth (1960s), tish-oo was the standard UK pronunciation too. I am always surprised when I hear BBC announcers saying tiss-you, iss-you, nego-see-ation, and so on. Has there been a recent change? (I apologize if you've covered this ishoo already, Lynne).

Leslie is a common male name in my wife's family, always with /s/, whereas they pronounce Leslie as a woman's name (which they do not use en famille) with /z/. I've picked this up over time.

American here (grew up in southern CA):

I've got only /s/ in Denise, Lisa and Joseph. I'd prefer /z/ for Leslie/Lesley, Louisa and especially Wesley, though /s/ seems OK for all of them.

And I definitely have louWEEza and never lewISSa, but maybe that's because my first, and only real-life, exposure to the name came from the Sound of Music film in which it's definitely louWEEza/sa (can't remember if it's /s/ or /z/ but the vowel is definitely /i/, not /I/).

And on the main name, a very close family member has the name Theresa and that definitely has an /s/ to me. It didn't even occur to me until I read the Language Log post that one could pronounce it with /z/!

Whereas I (BrE,Southern) didn't realise one could pronounce it with an /s/! And certainly not Wesley - John Wesley, founder of Methodism, is definitely with a /z/.

One of the things that does differ between the US and the UK is the stress on names (as, indeed, on other nouns and verbs, but this is a post on names). I remember watching a documentary about the Gunpowder Plot in which the American presenter constantly referred to "Father GarNETT" - but a British contributor pronounced it as "Father GARnet", which is the way that felt "right" to me. And I have several times been caught out when Americans pronounce their names differently to how I read them. The author Robin McKinley has written two novels set in the fictional country of "Damar", which I, in my British way, read as "DAYmar", and it wasn't until I heard Robin speak about the books that I learnt it should be called "Da-MARR"!

The worst is when it's Thérèse and the bearer insists on writing it with the accents, all the while pronouncing it Tuh-rayz.

I've never heard Marcia pronounced as anything but Marsha but I've met very few. I've just assumed that Marcia follows traditional Latin pronunciation. That rule sometimes also applies to Lucia but that varies, since it can also come from modern Italian, making it Luchia.

If Alicia is the Spanish for Alice, wouldn't you expect Americans to pronounce it Alissia and Britons to approximate it as Alithia?

Nick

I've just assumed that Marcia follows traditional Latin pronunciation.

There are competing traditions. Surely the English tradition would give MAH-si-uh. Compare Marcellus, Dacia, Perceval.

Classical Latin would have a k sound. Before whatever sound corresponded to letter-I, the 'c' became all sorts of sounds in different traditions of pronouncing Latin. In the traditions we inherited from French in became s.

How did we get 11 comments in without a reference to the canonical source for the "mar-sha" pronunciation?

My Mum is a Teresa - all her family call her 'tree-sa'.

I don't think Tezza is a nickname version of Teresa in Br. English. Tessa yes or even Terry. Bizarrely Tezza is a shortened form of Terry/Terrance (for a male) - maybe I'm wrong but I've never heard of Tezza for Teresa (& I've heard my Mum's name pronounced and shortened many ways!)

Very interesting subject.
When I travelled across the US I was surprised to hear so many versions of such a popular name as Laura (I live in Europe so I got used to Spanish and Southern BrE pronunciation only).
Besides, there's an American TV dance show with British host where the pronunciation of the contestants' names is a very funny bit :)

Another interesting one is (also how to spell it! - especially e vs. ë) Raphael
In Britain it's usually "raff-ay-ell" or "raff-a-ell".
In the US (at least in New Haven CT where I lived for several years) it was "ray-full"

Oops I guess it's not sibilant, but it's interesting all the same

I have to say, 'rayfull' is not something I've ever heard and it sounds rather unlikely.

Forvo doesn't have a US pronunciation for 'Raphael' but they do have one for 'Rafael', which is in line with how I know it to be pronounced: http://forvo.com/word/rafael/

Pronouncenames.com similarly has a three-syllable pronunciation, though it's less clear that this is by an American: https://www.youtube.com/watch?v=GaR8kXT2X3E

Elena and Louis are two more that spring to mind. Names should be pronounced the way the owner wishes, after all, it is their name. That assumes they know how they wish it to be pronounced; I rejoice in the middle name of Xavier and have still to make up my mind between X-Z, Z and with a French accent.

The rule about pronouncing names as the owner wishes should also apply surnames. Many Russians suffer in this regard - it should ShaRAPova, for example, not SharopOVa.

Names should be pronounced the way the owner wishes. I disagree. Western names, at least Christian names, have a long tradition, they do not come out of the blue.

Not strictly relevant, but I once knew someone whose French father and English mother agreed to call her Adele if a girl. They'd split up by the time their daughter was born, but the mother duly informed those enquiring of the girl's name. Unfortunately she had no idea how it should be spelt and made her best shot with E-R-D-E-L. Consequently schoolteachers doing rollcall pronounced it as one might expect and the poor girl had to suffer the other children's taunts of: "Erdel, Erdel, jumped over the hurdle!". Wouldn't happen today, of course.

I remember seeing an online discussion a few years ago about the names "Ralph" and "Rafe"

Somebody had had a son and wanted to give him the name pronounced "Rafe". One side of the argument was that that name is spelt "Ralph" and pronounced "Rafe" and the spelling "Rafe was an abomination (despite the existence of the actor Rafe Spall). The other side was that these days, most people would pronounce "Ralph" as "Ralf" and therefore if the parents insisted on the "Rafe" pronunciation, that was the way they should spell it, which indeed they did.

There was an actress who was murdered last year named Sian Blake. I heard someone complain (possible one of the people complaining about the spelling "Rafe") that in news reports, her first name was pronounced Si-an, not Shan, but as she also spelt her name "Syan" I would guess the former was how she pronounced it.

I found an interesting discussion/poll about the pronunciation of "Tesla" on the Tesla site:

https://forums.tesla.com/forum/forums/how-do-you-pronounce-tesla-tezla-tessla

FWIW, I use /s/ for "Tesla", "Wesley", "Theresa", etc. Perhaps partly because my family pronounces our surname with /s/?

As for non-names, I grew up (in the USA) pronouncing "diesel" with /z/, but have switched to /s/, probably under the unconscious misapprehension that that is how it is pronounced in the UK.

Ron

I agree completely with Simonx that names (first names as well as last names) should be pronounced the way the owner wishes them to be pronounced. To insist otherwise seems incredibly rude.

@Ron Thanks ofr the link.
I work in motorsport, car parts etc. and I've never heard "diesel" pronounced with /s/ in the UK.
Laura

Hmm, car names. See also Jaguar. My father used to drive a Jaguar in the sixties, show-off that he was. I was surprised the first time I heard an American sports commentator pronounce the name jag-war. We always pronounced it jag-U-ah.

I work in motorsport, car parts etc. and I've never heard "diesel" pronounced with /s/ in the UK.

So here's the thing: Ordinarily I'd pronounce diesel with a /z/, but I can't rule out pronouncing it with an /s/. Similarly, I have a sister named Leslie and we routinely pronounce her name with an /s/. However, I'm reasonably sure I've heard her answer the telephone "This is Le/z/lie" and I can also imagine addressing her with the /z/ pronunciation under certain circumstances, e.g., when I'm extremely irritated with her. Moreover, I have no fixed pronunciation for the word either (EE-ther and EYE-ther) or vase (vahz and vayse).

I gather this isn't the case with people in the UK. Or perhaps it is and there's a degree of denial about it.

Dick, I (person in the UK, as you know) can neither wholly confirm nor wholly deny.

My pronunciation of either is not fixed (although EYE-thuh(r) is probably my usual). But my vase never rhymes with craze.

Ira Gershwin’s take on variance works better for American speech that British. Some of those are pronunciations that nobody uses here.

And does anybody anywhere say poTAHto?

@David L:

I've never noticed any other pronunciation for tissue besides TISH-you, and I'm American.

Regarding pronouncing someone's name as that would like it to be pronounced, it can be difficult to do that if they person has a different accent and/or dialect that you. You can't just assume they want to to try to mimic their pronunciation.

And I actually never knew till now that the Brady Bunch character's name is spelled Marcia, not Marsha. I have no association of the Marcia spelling with the "MAR-sha" pronunciation.

I (UK) had never heard the pronunciation "Treesa" until I heard someone refer to Theresa Green as an unintentionally funny name.
"Rafe" is certainly the traditional pronunciation of Ralph, as in "HMS Pinafore" where the rhymes obviously require Ralph Rackstraw's name to be pronounced that way.
I've never heard "vayse" in the UK, but our next-door neighbours when I was a child used to say "vawse".

@Dick, David
As far as "diesel" pronunciation is concerned, I just thought that my observation could help with possible research because I've been hearing this word more than 100 times per day all over the UK for 10 years. It's interesting, no /s/ version so far. Of course I'm in no position to rule anything out and as a non-native speaker I probably pay too much attention to pronunciation anyway.
Laura

Going off on a bit of a tangent, would anyone pronounce the "Th" sound in Theresa the same as in "three"? That would sound strange to me, but I've heard many Americans pronounce the name Anthony with a soft "th" sound, as in "anthem". As a native BrE (southern) speaker this would never occur to me. The usual pronunciation for me would be "Antony" (emphasis on the first syllable).

I know someone (here in SE England) who so hated being addressed as "Anthony" with the "th" pronounced, that he changed the spelling of his name, aged about 12 -- so it's "Anthony" on his birth certificate and "Antony" on every other official document.

When I was growing up in England I knew a Rafael -- his mother was Spanish, father English -- and his name was pronounced rayful.

@Ellen Kozisek: Perhaps I wasn't clear. I always hear 'tissue' in the US as tish-oo (possibly tish-you, although that seems less likely) and that seemed to be standard UK pronunciation when I was growing up.

But when I visit England now and hear people on the radio or TV saying 'tissue' or 'issue,' they seem most often to say something that rhymes with 'miss you,' without the 'sh' sound in the middle.

(US Midwest)
@Anonymous 11:43
Anthony, Bethany, Kathryn are all pronounced with a 'th' as in 'three'.
Likewise, the names Theodore, Thelma, Thaddeus are all pronounced with a 'th' as in 'three'.
Thomas and Theresa are both pronounced with a 't' as in 'tree'.

But my vase never rhymes with craze.

Actually, David, mine never does EYE-ther. As I mentioned, for me it's EE-ther vahz or vayse.

Re: pronouncing the name the way the owner of the name pronounces it, I gave up long ago. In fact, I anglicize it as much as possible myself (that is, BORis ZAKrin). I just wish people *spellt* it correctly, the first name at least. Seriously, Boris is not hard. Everybody's heard the name and has seen it written down. My brother's now-wife was excited to speak to a "real life Boris" the first time we spoke. So why is everyone spelling it with two R's, dangit!?

@Boris_Zakharin, I feel for you. Perhaps the influence of a certain orange tabby?

One of my ex-SIL's was named Louisa. She pronounced it Lou-EESS-a, but didn't seem to mind my calling her Lou-EEZ-a. Her sisters called her Weez or Weezy. (All in AmE, by the way.)
In Northern California, I never heard the city of San Rafael pronounced as anything other than "San Ra-fel".

I have seen "Theresa" spelled "Teresa" a number of times. And always pronounced with an /s/.

For comparison, here are a British and an American actor saying Wesley. The American /s/ is very pronounced, the British consonant less so:

If that's a /z/ Patrick Stewart is using, that's interesting, because it very clearly isn't a [z]: it's voiceless. It sounds like the short /s/ of my native southern German, where phonemic consonant length is alive and well but voiced obstruents were abolished some 1400 years ago...

Rudolf Diesel was German, specifically Bavarian; English /z/ is certainly a better approximation than English /s/.

Mrs. Redboots wrote: The author Robin McKinley has written two novels set in the fictional country of "Damar", which I, in my British way, read as "DAYmar", and it wasn't until I heard Robin speak about the books that I learnt it should be called "Da-MARR"!

Wow! I always pronounced it as you do. Learn something new...

That's going to take a lot of adjusting for me. Now I'm wondering what else in her worlds I've been mispronouncing. (McKinley is probably my all time favourite author, so it matters to me.)

– AiNJ

BrEng speaker in my sixties.

In my experience, Theresa is always Tǝreeza. Never an 's' or a 'Th'. Shortened forms Tess (as with 'of the D'urbervilles) and Tessa both pronounced with an 's'.
French version (rarer) is pronounced Tǝ-rayze'
Alicia usually 'Al-issya'. Aleesha likely to be regarded as a child's mispronunciation. It would not be perceived as a Spanish name - if it is such. So how it might be pronounced in Spanish would be irrelevant.
Marcia, invariably Mahssia. I've not encounter Marsha but if I had, would expect it to be spelt Marsha. I probably would not guess it as a variant of the same name.
Lisa variously 'Leiza' or 'Leeza'. 'Leessa' less widespread. I do know one, but she is half Italian.
Louise and Louisa always 'Loo-eeze' and 'Loo-eezǝ'. I've never heard any other pronunciation of either.
Likewise, Denise, Diesel and Joseph always a 'z'. An 's' would sound odd, foreign or affected. Dennis, though is an 's'.
Leslie or Leslie always a 'z' irrespective of sex. The male and female versions are pronounced identically.
Wesley, whether one of a family of great C18 evangelists, a current surname or a Christian name (less usual I suspect here) always a 'z'.

Siân is Shahn, whether it's got an accent on it or not.
In English, usually, Sean is pronounced Shawn, Siobhan is pronounced Shǝvawn with the emphasis on the second syllable, and Niamh, Neev.
Raphael is Raff-ayǝl
Vase is vahze. One hears vawze but it's generally regarded as wrong in the Alan Ross sense. I've never heard vayze, but I'm sure that would, in the Alan Ross sense, be even wronger.
Anthony always 't', with the stress on the first syllable. Garnett likewise the stress always on the first syllable however spelt.
Bethany, Kathryn, Theodore (if used at all) and Thelma always 'th'. Kathryn and Catherine normally pronounced identically. Katrine, however, is Kǝtreenǝ. Catriona is sometimes Kǝtreenǝ and sometimes Kǝtreeǝnǝ.
Kirsten varies between Kerstǝn and Kierstǝn, but Kirstie seems to be more likely to be Kersty than Kiersty.
Thaddeus if encountered is usually a foreign name.

I'm fairly sure I always say tishyou. One sometimes hears 'tissyou' but it sounds a bit precious, as would Patrissiya instead of Pǝtrisha. The shortened forms Trish and Trisha are always pronounced as written.

The car, or for that matter, the animal, always 'jag-U-ah'. Likewise, to BrEng ears not pronouncing the 'g' in Nicaragua sounds very affected.

Lucia is an odd one. I think the island is pronounced St Looshya but the opera is called in a sort of cod-Italian, Loocheeǝ di Lammamoor, even though the heroine knew herself as Lucy.

Dru - I am with you all the way (also sixty-something Brit) above.

I think some young women are called Lisa (LEE-sa), and the old-fashioned name Eliza would be E-LYE-za. I am not sure what to do about Liza Minelli!

The discussion about Adele/Urdel reminds me of overhearing a mother telling her daughter that the female character in the Harry Potter films was Armani ...

We understand words because we recognise them. We recognise them as different because we recognise little speech sounds as different.

But some speech sounds are more unequal than others.

A whole family of speech sound (technically ' phonemes ') with reduced inequality are those produced by turbulence somewhere in the mouth.

In English the turbulence is in four places — five if you're Scottish and if you can say it's a braw bricht moonlicht nicht. In Latin, the language which gave us our Roman alphabet, there were two
• at the exit of the mouth between the upper teeth and the lower lip
(There's a technical term I won't bore you with.)
• in front inside the mouth between the front of the tongue and the tooth ridge
(There's a technical term that's less obscure and was used by Lynne: ' sibilant')

However many turbulence places there's a theoretical possibility of double the number of speech sounds. They come in pairs
• one with simultaneous buzzing from the vibration of the vocal cords
(The technican term is voiced ).
• one without that vibration
(The technical term is voiceless .)

I said ' a theoretical possibility ' but in Latin there was no doubling; there were not four speech sounds but two. And when they decided to write the language down they borrowed two letters from their literate neighbours: F and S.

What does this all mean? Did Romans never make voiced sounds with turbulence at those two positions? The strange answer is that they didn't hear any voiced/voiceless difference. Or if a keen-eared Roman managed to hear the acoustic difference he or she ignored it as irrelevant. The acoustic difference wasn't meaningful; if you substituted the other member of the pair, you would still be saying the same word.
(In technical terms there could be two phonemes , divided into four allophones .)

Latin changed into Late Latin, which fragmented and changed into French, Italian. Spanish, Portuguese, Romanian and others. Among the changes
• In some languages the voiced allophones became meaningful phonemes
• The alphabet invented new letters to represent (among other things)
--- a voiced variant of F
--- a voiced variant of S
Actually, V and Z weren't new letters but they were give new values.

This would be simple and straightforward but for the awkward facts:
• V and Z have different values in different languages
• The splitting of F and S into four phonemes is only partially complete

Fortunately, to understand the present states of letter Z and sound /z/ (and incidentally some strange spellings with letters F and V) we need look no further than English and French.

biochemist - many years ago Liza Minelli performed a song as part of her act all about "Liza with a zee, not Lisa with an ess".

we need look no further than English and French

Before 1066, written English and written French happily coped with letter-S for both [s] and [z] because they were simply allophones consistently used at different parts of a syllable. In both languages a sibilant was voiced between vowels and before a voiced consonant (with vocal cord buzzing) , voiceless at the end of a word and before a voiceless consonant.

But English had a voiceless sibilant in one of the words spelled seal so when Norman scribes had to write it they used the letter Z — which we still do with zeal. But there are still two words spelled seal and pronounced with a voiceless sibilant. So [s] and [z] qualify as distinct phonemes at the start of syllables. And that's why we use spellings with Z there

zoo (cf sue, Sue)
zip (cf sip)
Zack (cf sack)

Here we hear two different words, but the perception is less clear cut at the end of a syllable. We still understand if someone says
I came here by buzz
Bus like a bee

[In some accents of South West England, even zoo and Sue can sound the same.]

Between vowels and before voiced consonants such as /l/ many of us don't even have an acoustic distinction, let alone a perceptual distinction. Hence

1 Patrick Stewart's strange Shut up Wesley!. His sibilant isn't a typical /s/ or a typical /z/ because for him there's only one sibilant phoneme before /l/. In his accent (broadly shared by Dru, Biochemist, me and numerous other Be speakers) the typical allophone in this position sounds voiced like [z]. If somebody uses the voiceless allophone like [s] we don't hear the difference, or don't hear it as significant. The complication in the Star Trek clip is that Stewart's character is echoing somebody with a different accent.

2. Lisa, Louisa and Joseph still have the historical voiced allophone between vowels in this BrE accent. This actually extends to Louiseand Denise where the sibilant used to be between vowels in French, which is where we took the names from. Again, Dru, Biochemist and I among others don't hear any significant difference if another speaker uses a voiceless [s]. I (and, I think, the others) put it down to a minor idiosyncrasy of speech.

The converse can happen in an American accent.

• Kirk Poore's Louisa used a voiceless allophoneherself but didn't react when others close to her used the voiced allophone for her name of nickname.

• Dick Hartnell isn't always conscious of which allophone he uses in diesel, and his sister Leslie isn't totally consistent when naming herself. Interestingly, Dick associates the unusual allophone ([z] in this case) with a particular type of expressive utterance (exasperation in this case). This is often when we notice idiosyncrasies in speech that normally pass us by. Perhaps Leslie reserves the 'marked' (noticeably different) allophone for the unusual speech act of announcing herself.

To sum up.
• The split of 'S' in to two allophones is incomplete in English — except at the start of a syllable
• The voiced and voiceless allophones [z] and [s] are predictable according to position in conservative BrE speech (if not more generally)
• The choice of [s] or [z] is more open to many AmE speakers. Many prefer [s] where my accent prefers [z]. This may indicate a split into two phonemes /s/ and/z/ in those positions. But for Kirk and Lousia, Dick and Leslie (and no doubt other), the choice is still (at leat own part) one of allophones.

In my normal conversations using Northern Californian and Midwestern AmE, I have never hear anyone say Joseph with a /z/ sound.

Until last night, I'd never heard anyone say "Theresa" with a "th" sound. I was on my way home from a party late last night, and NPR was broadcasting the BBC World Service. One of the reporters talked about "Thereza" May. Without this thread, I would have missed the /z/ but the "Th" really stood out.

On the buzz/bus distinction, I have a memory of a Northern Irish friend at university who not only pronounced bus as buzz, but also maths as mazz. (But this was over forty years ago, so my memory might be faulty.)

As for pronouncing then th- in Theresa, I recall reading a comic book story when I was very young, not long after I learnt to read, which was set in London. I had heard the name of a river in London, but then I saw in the story the River Thames. I assumed there must be two rivers in London, on spelt something like Tems, and the other pronounced Thames.

Incidentally, I have noted two pairs of words in English such that the two members of the pair are spelled the same, but one has a voiced th, the other unvoiced.

Kirk

I meant to write that for you the choice is (at least in part) one of allophones. My spellchecker had other ideas.

In part you (and Louisa) are open to choice of allophones — but only (as far as we know) where the name in question is Louisa. In another part — where the name is Joseph — you see the [s] sound as the only norm

Paul Dormer

Incidentally, I have noted two pairs of words in English such that the two members of the pair are spelled the same, but one has a voiced th, the other unvoiced.

I'd be interested to know what they are, Paul.

As far as I know, the two sounds were allophones in Old English — perceived as the same sound or as near the same as made no difference. The rune-based letters þ and ð were used, but not with the two distinct values they now have in the International Phonetic Alphabet.

Apart from the demonstratives (this, that, these, those) and the, I can think of only two common environments for voiced [ð]: between vowels (e.g. rather, dither) and when a stem is used as a verb (e.g. teethe, bathe. To some extent the latter is true for [v] (e.g. thieve, calve.

LATER
I've just looked up a few things and found:

• Sure enough there was only one þ-ð phoneme in Old English.

• I overlooked some other ð-initial words, but they're all grammar words: there, then, though, thy, thine, they.

• There are two minimal pairs where one is a grammar word: ether~either, thigh~thy.

• And two minimal pairs where one is a noun and the other a verb: wreath~wreathe, mouth~mouth.

PS
three pairs if you count teeth~teethe

The two pairs I've found are:

thou, as in the second person singular pronoun, and thou as a thousandth of an inch.

mouth, the part of the body, and to mouth, as in to form words silently.

Three further thoughts on this.

1. 'Thy' is voiced, and 'thigh', spelt differently but the same vowel, is unvoiced.

2. The difference between 'teeth' and 'teethe', 'mouth (n)'and 'mouth (vb)' and 'calf', 'calve' is similar to the tendency where a two syllable word is used both as a noun and a verb, for the stress to shift down the word onto the second syllable when it is a verb, e.g. 'prospect', 'conduct', 'digest'.

3. Ignoring the variations in how the 'u' is pronounced though it could be related, there tends to be a regional difference in BrEng between where 'us' is pronounced 'uss' and 'uz'. I don't know whether it's been mapped, but I suspect 'uz' is most frequent in a belt that starts somewhere round Birmingham, and possibly ends somewhere round Leeds. One also hears 'buz' for 'bus' but I don't think I've ever heard 'suz' for 'suss'

Surely if there are any minimal pairs between /θ/ and /ð/, regardless of the grammatical functions of the words in question, they cannot be merely allophones? That's kinda intrinsic to the definition of an allophone and a phoneme, after all. How common a sound is and how many minimal pairs you can find doesn't matter as far as I was taught -- if there is >1 minimal pair (the only phone different is the phone in question and the meaning is different, as in thy-thigh), the sounds are phonemes and not exclusively allophones of each other. Whether something's a "grammar word" or not is absolutely meaningless in the world of phonology afaik?

sparksbet

My point about θ and ð is that the split into two phonemes is a process that's not very advanced. My more general point is that English retains much of the old predictability of when s and z, f and v, θ and ð were allophones.

There was once a phonological rule that /θ-ð/ was voiceless in initial position. The rule has been destroyed by a number of words. But that number is very small and pretty much a closed set. As yet we don't invent new words beginning in /ð/. The ghost of the old rule still haunts.

Returning to the theme of the thread, the split of /s-z/ into two phonemes is a process that's more advanced than the /θ-ð/ split. But in accents like mine, there's still a vigorous haunting of the old allophonic rules, which explains why substituting [s] for [z] in those names makes no perceptual difference — or at most sounds slightly odd.

@David Crosbie
Ahhhh okay, that makes sense! That's actually a really satisfying explanation for why I have difficulty figuring out which I use in some names.

Looking beyond just names...

Frequency counts of English speech sounds and English spelling show a dramatic mismatch. The sound z is one of the more frequent phonemes, by at least one count in the top quartile. But letter Z has the lowest frequency of all. Hence its score of 10 in Scrabble.

Some of this can be explained by grammar. Three suffixes and two auxiliary verb forms have the same spelling

• noun plural
• noun possessive
• verb 3rd person singular
• reduced is
• reduced has

OK, some are differentiated by an apostrophe, but essentially they're all spelled s.

Yet there are three distinct pronunciations: s, z and iz. But the distribution is far from random:

• s
---after a voiceless consonant: cats, cat's, cats', The cat sits, The cat's here, The cat's gone

• z
---after a voiced consonant: dogs, dog's, dogs', The dog sits, The dog's here, The dog' gone
---after a vowel sound: cows, cow's, cows', The cow moos, The cow's here, The cow's gone

• iz
--- after sibilants and similar: faces etc, roses etc,wishes etc, garages etc, matches etc, bridges etc

The point is that these are all totally predictable. If we encounter a newly invented noun, we know exactly what its plural and possessive forms will be, and how immediately following reduced is or has will be pronounced. f we encounter a newly invented verb, we know exactly how the 3rd singular of the Present Simple will be pronounced. So we don't need to deploy the letters S and Z. One letter (with an added E if necessary) will do for all.

So why do we bother using letter Z at all?

In Shakespeare's time this was a rhetorical question — if not for him then at least for one of his characters, who utters the insult

Thou whoreson zed! Thou unnecessary letter!

.

Thou whoreson zed! Thou unnecessary letter!

Shakespeare was just having fun, but at around the same time the scholarly schoolteacher Richard Mulcaster was rather more serious. (Both writers are quoted in David Crystal's Spell it Out.) He likens letter S to a military leader acting in place of a monarch.

Z is a consonant much heard amongst us and seldom seen. I think by reason that it is not so ready to the pen as s is, which is become lieutenant general to z. ... It is not lightly (=frequently) expressed in English, saving in foreign enfranchisemnts (=borrowings).

And yet letter Z did make some progress — which Crystal puts down largely to the spelling of the verb-making suffix -ize. This was quite unnecessary; the spelling -se from French was recognisable as a z-sound. But classical scholars, particularly in the sixteenth century, were keen on -ize as a signal of Latin derivation — much as we still use ph as a signal of Greek. Ironically, z didn't really belong in the Roman alphabet; they used it to spell Greek words and their derivatives.

Subsequent to Shakespeare and Mulcaster, the New World Colonies plumped for the -ize spelling for most words. British writers continued to vacillate, until finally Oxford and its press decided (mostly) for -ize, driving rival publishers to -ise.

Inside the vacillation, Crystal sees a pattern:
• -se for words from French
• -ze for words from Latin/Greek
People still weren't consistent because they weren't necessarily savvy as to where the word actually came from.

With a familiar and frequent word like close the prejudice against Z was so strong that the same spelling was used for the adjective/noun (pronounced with /s/) and the verb (pronounced with /z/) Centuries later, Crystal points out, the invented term for a word in a test that involves closure was given the spelling cloze. Similar to the close pair are the noun house (/s/) and the verb house (/z/).

[This is a parallel with noun mouth and verbmouth. But the other split pair uses two spellings as in noun half and verb halve.]

Early borrowing from French like baptise ended up with an EYES pronunciation. The names Louise , Denise, Eloise etc were adopted (or re-adopted) later, so we kept the EE values of the vowel — but still with the French z-pronuncation. Given that we continue to pronounce the names the (almost) French way in Britain, I think it's safe to say that the EES-pronunciation is an American innovation.

The only words that come to mind with -sia spelling are ambrosia and Asia, both of which can be ZEEYUH in my accent, although Asia can go all funny with SH and ZH. Unfortunately, Shakespeare didn't use the word ambrosia, but David Crystal's Shakespearian Pronunciation Dictionary gives only a z-pronunciayion for Asia. Similarly, he gives only a z-pronunciation for Elysium. I'm reasonably confident that my z-pronumciations of aphasia, Malaysia, amnesia etc are conservative relics of original pronunciation, and that s-pronunciations, sh-ptronunciations and zh-pronunciations are innovations.

While -se represented Z in French, -ce represented Z. So I'm pretty sure that my s-pronunciation of Alice is the continuation of something old. Likewise, French -ci- and -cie were (our contained) S. It's easier to imagine an early SEEYUH changing to later SHUH than the reverse. Shakespeare (according to Crystal) pronounced patrician with an s-sound, so it's plausible that Patricia was also with SEEYUH. By extension, it's plausible that only Patricia has changed (to SH) in all dialects, while Lucia and Alicia have changed only sporadically. For most BrE speakers, I suggest that Marcia hasn't changed at all.

CORRECTION

While -se represented Z in French, -ce represented S.

David

Interesting stuff, and a great quote. (Of course it's thou whoreson ZEE in American productions. Isn't it?). Can you give the reference for it?

Btw, earlier you said:

"There was once a phonological rule that /θ-ð/ was voiceless in initial position"

But all those "grammar words" you mention have modern German cognates with a voiced initial consonant: der (the), dann (then), diese (this), etc. Surely if the ur-lingo had unvoiced initial th, Germans would now be saying "ter, tann, tiese" etc?

Zouk, the immediate quotes are from David Crystal's Spell It Out, The Singular Story of English Spelling.

The Shakespearian quote is King Lear Act II Scene 2 line 61. Kent has already addressed Gonerill's steward Oswald as

A knave, a rascal, an eater of broken meats, a base,
proud, shallow, beggarly, three-suited, hundred-pound,
filthy-worsted-stocking knave; a lily-livered, action-taking,
whoreson glass-gazing super-serviceable finical
rogue, one-trunk-inheriting slave; one that wouldst be a
bawd in way of good service, and art nothing but the
composition of a knave, beggar, coward, pander, and
the son and heir of a mongrel bitch; one whom I will
beat into clamorous whining if thou deniest the least
syllable of thy addition.

The Mulcaster quotation is from The Elementarie, which entreateth chieflie of the right writing of our English tung, published 1582, (pp 96, 123).

I'll need to check my facts before answering about 'the ur-lingo'.

David, on the -ize/-ise endings, I'm sure I was taught at school in the sixties that either ending was allowable in many words. Certainly didn't hear the French/Latin/Greek until much later. It wasn't until I started work in the seventies that I was told to use the -ise endings as "-ize is American."

But it's interesting you use "baptise" as one of your examples. Many years ago, my father showed me a family bible belonging to my great grandfather (1848-1916). My grandfather seems to have had a lot of sisters and my great grandfather listed their births on the fly-leaf. Also tucked in there was a baptism certificate for one of them. So I know that in the late nineteenth century in the Abingdon area, official printed certificates used "baptized" as the spelling.

I have never heard 'Ambrosia' pronounced any other ways than 'Ambrosia' or 'Ambrozha', assuming the latter is the usual sound 'tion' begins with, either with its classical reference or as a baby food etc. Is it really ever pronounced any other way?

On 'ise' v 'ize' I seem to remember being told that one should use 'ize' for formations from Greek words. That was Greek only and not Latin. One should not use it for any others. On that basis, 'baptize' would be preferable to 'baptise', but most others would be 'ise'. As it is, though, I have normally settled for 'ise' for all of them. It's consistent. It's not dependent on claiming to show off a knowledge of a language I haven't really got. And 'realize' is completely wrong. To me, it just looks uncouth.

Also, Asia pronounced any way other than Ayzha sounds affected to me, on a par with choirs that insist on singing 'sal-vay-ssee-on'. I know that with C16 and C17 music that is strictly correct, but it still sounds silly.

Thanks, David. So Kent isn't explicitly deprecating the letter Z, as I read it out of context (despite the lack of the comma that would call for), but implicitly so by calling the despised Oswald one -- "whoreson" serving adjectivally.

Paul

In Britain most of the words can take the Oxford spelling in -ize or the non-Oxford spelling in -ise. This doesn't mean you can choose randomly. Different publishers may impose different choices — or at least insist on consistency.

According to David Crystal it's widely agreed in Britain that these words are always spelled with S:

advertise, advise, apprise, chastise, circumcise, comprises, compromise, despise, devise, excise, exercise, franchise, improvise, incise, revise, supervise, surmise, surprise, televise.

He adds

QUOTE
This ins't a complete list, because there are derived forms (such as enfranchise, misadvise) as well as new usages (to merchandise, an enterprising proposal). Also, uncertainty over what the 'rule' is means that we sometimes see some of these verbs spelled with a z, especially in the USA.
UNQUOTE

The disadvantage of an ize-spelling policy is that you have to learn that list. With an ise-spelling policy, you don't.

Paul, Dru

baptise
Scribes in the Middle Ages would have been familiar with both the French bapiser and the Latin baptizare (but not with the Greek βαπτίζειν. This explains why early spellings included baptis, baptiz, baptyse, baptyze. Even in French, some scribes were so influenced by Latin as to write baptizer.

I read this in the OED entry for the word which uses baptize as its heading without acknowledging the other spelling. (Although some of the quotes use baptise.) I America, In believe, baptise is regarded as a British variant.

salvation
In his dictionary, David Crystal reckons that Shakespeare pronounced it salˈvɛːsɪən (sal VAIR si un). His printers spelt saluation.

Asia
Crystal gives the Shakespearian pronunciation as ˈɛːsiə with one stress or — when the metre called for two stresses —ˈɛːsɪˌa.

David said:

The names Louise , Denise, Eloise etc were adopted (or re-adopted) later, so we kept the EE values of the vowel — but still with the French z-pronunciation. Given that we continue to pronounce the names the (almost) French way in Britain, I think it's safe to say that the EES-pronunciation is an American innovation.

I'm afraid I'm too lazy to review the comments to see what's been said about the pronunciation of these names in American English, but Louise and Eloise are never pronounced with the /s/ sound on this side of the Atlantic. Denise, on the other hand, is virtually never pronounced with /z/.

Dick

Apple operating system allows you to search any open screen in any application. So I can tell you that Lynne said what you say about Denise and Louise , and that nobody mentioned Eloise before I did.

I'm surprised.

But I can still make the same guess about Denise.

Very tentatively, I suggest that Denise has been influenced by the pronunciation of Denis, whereas the LOO WIS pronunciation of Louis has not been so common as to influence Louise.

Last century's most famous Louis is pronounced LOO EE by most people I hear, although not by Armstrong himself.

I used to work with a Bosnian woman. As English was not her first language she was always trying to improve it and ask advice over idioms, etc. One day she was having an argument with another member of staff and I was called in to adjudicate. (I failed English Language O-level in 1968, so I don't see why I was considered an expert.)

She was using -ize endings and her colleague was complaining those were American spellings. She in turn was using a copy of the pocket OED as proof that she was right.

I tried to explain all that about how in many words it was optional, it all depended on where the word was derived from, -ize wasn't exclusively American, and that Oxford University Press preferred the -ize endings, whereas the company house style was the -ise endings.

I don't think either person was happy with that answer.

Paul, did she say BOZnia or BOSSnia?

David said:

Last century's most famous Louis is pronounced LOO EE by most people I hear, although not by Armstrong himself.

I would certainly say LOO EE and not LOO IS, though as someone who was born in St. Louis and spent the first 10 years of my life in one of its leafy suburbs I can also attest that no one there (the lyrics to the song "Meet Me in St. Louis" notwithstanding) said St. LOO EE -- it was almost always St. LOO IS.

Also, though it isn't as common as Louise, there are also people here named Louisa whose name is typically pronounced with the /s/ and not the /z/ sound.

I can't remember. :-)

Zouk Delors

But all those "grammar words" you mention have modern German cognates with a voiced initial consonant: der (the), dann (then), diese (this), etc. Surely if the ur-lingo had unvoiced initial th, Germans would now be saying "ter, tann, tiese" etc?

There are different layers of ur-lingo, Zouk.

At the base is Proto Indo-European *t.

The next reconstruction up is Proto Germanic by Grimm's Law usually written *þ with a possible sound value something like .θ.

In the next layer up, *þ is still in East Germanic (as seen in Gothic) and in North Germanic (as seen in Old Norse etc). This voiceless spirant is also presumed in Common West Germanic.

In the next layer,
• Old English kept the voiceless þ sound — with a voiced allophone between two vowels, of which the first was stressed
• Continental West Germanic retained some sort of spirant — usually spelled th in the earliest written texts. The books I've consulted tend to think that this started as a voiceless sound but became voiced.

In the top layer
• Middle English allowed the voiced sound to spread to the words we're talking about, thus creating a separate phoneme /ð/
• Scandinavian languages lost the þ sound in favour of a stop — with the exception of Icelandic
• In the early forms of German, Dutch, Frisian, the voiced spirant ð changed to the stop d. I've seen a suggestion that this started in the south with those grammar words — probably because they were unstressed — then spread northward and to all words with ð.

BTW: it's worth mentioning that you British pronounce the "s" in resource with a /z/, while we Americans never do. My copy of the compact (that is, first) edition of the OED indicates that whenever the R volume was compiled early in the 20th century resource was also pronounced in the UK with the /s/ and not the /z/ sound.

And since the French word is ressource -- sans /z/, you might say -- we'll have to look elsewhere for an explanation of the shift.

To be slightly pedantic, I think in BrEng 'resource' is 'z' where the stress is on the second syllable and 's' where it is on the first. The different stress patterns have slightly different meanings. With the stress on the second syllable, whether a noun or less frequently a verb, it means 'supply'. That is its usual meaning. With the stress on the first syllable it is emphasising the 're' bit and means something more like 'to find a fresh supply' or 'to replace with fresh supplies'. That usage is less frequent.

In the first 're' is 'rǝ' but in the second, it is 'ree'.

Incidentally, writing that, I've just noticed that in BrEng in 'emphasising' the first 's' is an 's' but the second is a 'z'. But in 'emphasis', they are both 's'.

John Wells in hid Longman Pronunciation Dictionary reports a survey of BrE speakers with preferences

50% -ˈzɔɪs
45% -ˈsɔːs
6% ˈriː- sɔːs/zɔːs

A famous Louis of the 19th century, R.L. Stevenson, was called by his middle name because he had a cousin Robert. He was originally Robert Lewis and kept that pronunciation after he changed the spelling, though modern readers often say LOO EE when referring to him.

As someone who lives in a St Louis suburb, I can testify that anyone saying Saint LOO EE here would be look at as if they told a joke in extremely poor taste. In fact, I would say that in most cases, current usage for spelling LOU EE is "Louie", as in comedian Louie Anderson or the song "Louie Louie". "Louis" and "Lewis" are pronounced exactly the same.

I've just heard an American voice on the radio, saying something like 'the relic is now housed in the museum' , using the /s/ in 'housed' rather than BrE /z/.
BrE would also have the /z/ in 'housing'.

I can go either way on Theresa, whereas for each of the names in the bulletted list I have only one pronunciation (/s/ for Denise, Lisa, Joseph; /z/ for Leslie, Wesley, Louisa).

And then there's the final consonant in "Jesus", with which I have only now made the connection to this post. Here in Australia, it's /z/ — Jeezuz. You do occasionally hear people hissing it (Jeezusss), but that's so strongly correlated with a certain type of preacher that it comes across as an affectation. In fact, it never occured to me that anyone might have phonemic /s/ in "Jesus" until a certain conversation in an online linguistics forum (many years ago now), where an American was surprised that I have /z/. I was at least as surprised that he didn't.

On the topic of name pronunciation differences more broadly, there is probably a blog post's worth of examples in which American and British English place the stress on a different syllable. I remember being surprised by John Wells once, when he remarked that the feminine "Gabrielle" is stressed on the final syllable — I was perfectly well aware that this is true in America, but John is British! So what gives? Either (a) all of the Gabrielles that John knows happen to be American, or (b) The first-syllable pronunciation is an Australianism. (Or something else more complicated and nuanced.)

In my British accent there's final stress on Gabrielle and Michelle. In another British accent the Beatles let the music put stress on -chelle (although there's also stress on Mi-). They feel like French names, and the Beatles song reflects this with ma belle and even a verse translated into French.

There's a counter-example to the ear in the actual name of Mrs Redboots, which is usually stress on the first syllable But we spell it Annabel. I'd assume something different from the spelling Annabelle.

On further thoughts, the same is true for Christobel. Oh yes, and Rachel.

So it's a matter of spelling.

Lisa: Italian style with z, Spanish with s. Is this an influence on AmE, perhaps more Spanish in eg California, New Mexico etc? Used to be more z than s in BrE, I think. My niece was usuallly Lisa with z in UK but often pronounced with s where she lives in California.

David Crosbie wrote: There's a counter-example to the ear in the actual name of Mrs Redboots, which is usually stress on the first syllable But we spell it Annabel. I'd assume something different from the spelling Annabelle.

No, you can't intuit the spelling of my name from its pronunciation - in the UK, at any rate, the stress is always on the first syllable, no matter how it is spelt. Alas, when you consider the many and varied spellings of it that I have to put up with!

In Britain the stress in Annette normally now goes on the second syllable, as in Jeanette. But there is a suspicion from old spellings and verse that in the past, Annette might have been stressed on the first syllable, as in Janet.

Another name where I feel /z/ is more British and /s/ more American is "Jesus".

@David Crosbie @Paul Dormer @dru

The most detailed historical inquiry into the ize/ise question that I've yet found is at the excellent Random Idea English blog.

There's more detail (lots more) and more considered argument in David Crystal's 'Spell It Out'.

vp

John Wells’ Longman Pronunciation Dictionary gives /s/ for BrE and /z/ for AmE.

David

"There are different layers of ur-lingo, Zouk"

Thanks for the reply. It confused me slightly, but it's difficult to say more without knowing how to produce the special characters. Thanks for taking the trouble anyway. The relevant "ur-lingo" in this case is probably Proto-Germanic, I think.

Thanks also for the reference to Grimm's Law. On looking into it a bit I see shifting consonants do sometimes gain or lose voicing so the underlying assumption of my query was probably ill-founded anyway. I had in mind the way foreigners or dialect-speakers deal with the two English 'th's: usually with either 'd' and 't' (eg Irish) or 'z' and 's' (eg Germans).

Incidentally, Arabic has four or five 'th' consonants, which are variously rendered in the different dialects

I do still wonder how we can know how things were pronounced by people long dead before recording technology arrived.

Someone up there mentioned Liza Minnelli and her name song, which was the first thing I thought of when Lisa was mentioned. If you don't know it, here (start at 7:07) for Say Liza (Liza with a "Z").

Anthony is definitely said with the soft TH of "thin" and not a T or the hard TH of "there" here on the northern high plains of western South Dakota. One of our neighbors' sons who played with my son was named Anthony and the TH was always pronounced by him, his mother, and everyone else in the neighborhood. Although he didn't use it, Tony however is the shortened form and I've occasionally wondered why the TH becomes a T before this.

My mother also spelled her name Annabel (her father named her after Poe's poem Annabel Lee, but she always pronounced it with the stress on the last syllable, the BEL.

Dark Star

The odd thing is not that the h is missing in the short form, but that it's been added to the long form. See https://en.m.wikipedia.org/wiki/Anthony_(given_name)

Going back to Theresa; the most geographically widely-spread use of the name I know is the Maria Theresa thaler (https://en.wikipedia.org/wiki/Maria_Theresa_thaler) named after the empress, and definitely pronounced with a z.

Could the Americans among you confirm how you would pronounce the coin?

Could the Americans among you confirm how you would pronounce the coin?

No. Before you posted the link I'd never heard of this coin. (And since it's Austrian, I wouldn't begin to know the preferred pronunciation.) Am I to assume that Britons know all about it? If so, good on you.

I can say, however, that Americans pronounce Mother Teresa's name with an /s/. To be specific, as tuh-RAY-sah. On the other hand, I've always pronounced the first name of Teresa Wright, the popular American movie actress of the 1940s, the same way most Americans would pronounce Theresa: tuh-REE-sah. But I think that's already been covered.

Just for the record (and a long time after all the other comments) it has just emerged that Barbra Streisand is annoyed that Siri mis-pronounces her surname -apparently it should have /s/ in the middle. I know I would definitely use a /z/.
And - Liza Minelli - my query way upstream concerned the vowel in the middle - is it EYE or EE?

And - Liza Minelli - my query way upstream concerned the vowel in the middle - is it EYE or EE?

It's EYE.

The American electric car manufacturer, Tesla, is invariably called Tessla in the USA and Tezla in the UK. It's founder, Elon Musk, who is South African by birth, also refers to his company as Tezla, although he now has an American accent that sometimes exhibits undertones of British and South African English.

TECHNICAL

Test to see whether the new subscription widget works (for me).

I watched an American-produced YouTube video the other day in which two animated figures were used to present the two sides of an argument. Their names were Theresa and Anthony, pronounced thereeza and antheny (sorry but I don't know how to produce a schwa or any other special character). Amazingly it was interesting enough to hold my attention to the end, although the "h"s never ceased to jar. And I'm one who has to turn the radio off if an interviewee starts a second sentence with verbal-tic initial "So".

Wow, this is blowing my mind! I have been American for 40 years and I have never heard or heard of Louisa pronounced LouISSa. (I also use the Z for Joseph and Lesley/Leslie and have not been aware of the option of the S pronunciation, but that doesn't shock me as much.) For reference, I have lived in the northeast, Ohio, and most of my family is from the west coast.

I regret that I am so late to this thread.

Thank you for this post. I'd been googling in vain to find out why I say "Joseph" with an /s/ and many neighbors (mostly older) use "Jozeph" with a /z/ ("JOH-zf" is how it sounds to me). I'm an American and was born in Philadelphia: Many of the "Jozeph" users were first-generation Italian- and Irish-Americans, but that could be a coincidence.

I'd heard the /z/ version on the animated comedy "King of the Hill" for years (the Bobby Hilly character is particularly strong in this pronunciation), and no one else around me found the pronunciation gap remarkable.

Lynne: just recently found your site, and I am absolutely hooked (much to my wife’s amusement). Sorry for the very late reply. I am working my way through old posts.

I agree completely for the BrE use of z in the names you give ( except for Joseph, Which to me is always an s). What I really find amusing is the way non-Scots pronounce the girl’s name Ailsa.

It is meant to sound like Ayl -za, with the first syllable rhyming with ale, ail. English friends who would always say “Deneez” seem to find it really, really difficult to use z in Alisa. Even the first syllable is often problematic. The name often comes out sounding like Elsa.

Names should be pronounced the way their “owners” intend. Please, please note that McPherson is not pronounced mac-feer-son. Similarly, I always assumed that the island Nevis was named after the Scottish mountain, and should not be pronounced Nee-vis. I have realised that I have absolutely no idea how those who live there pronounce their island’s name. I was really surprised to find in a later post that the guy who named the tallest mountain pronounced his name Eev-wrist.