A Pronunciation Mystery: American ‘Father’

Letter A

Brooklyn Museum

Sometimes the most common words have the most illogical pronunciations. Such is the case with the American pronunciation of ‘father.’ It seems so self-evident this word is pronounced with a broad a (i.e. an ‘ah’ sound) that this fact barely seems worth examination. Yet on closer inspection, there is something quite mysterious about American ‘father.’

In many British accents, the broad a in ‘father’ makes sense: it might be thought of as part of the BATH set. That is to say, it’s part of that curious category of British broad-a words that also includes ‘bath,’ ‘can’t,’ ‘pass’ and hundreds of others. This set isn’t entirely logical in and of itself, but it at least makes sense that ‘father’ is a part of it.

But Americans (well, most Americans) have no BATH set. As commonplace as the ‘ah’ in ‘father’ might seem to be, it’s really rather strange that we say it this way in the US. Shouldn’t it logically be pronounced with the short-a in ‘cat?’

John C. Wells comments on this mystery in his Accents of English:

This lengthening is essentially the same as that in the BATH words, and it has not been satisfactorily explained how GenAm [i.e. General American English]comes to have /ɑ/ in father, palm etc, but not in calf, halve, and the other bath words.

Now, in terms of words like ‘calm’ and ‘palm,’ I might blame the letter ‘l‘ (even if it’s silent). I can’t say for sure, but there at least seems to be a type of rule where /a/ followed by/l/ results in a broad ‘ah’ sound in American English. (The ‘l’ in words like this is not silent for many Americans, including myself). But ‘father’ is truly perplexing.

The reason why ‘father’ has an ‘ah’ sound isn’t a mystery in and of itself. Several hundred years ago, English /a/ began a process of being lengthened and ‘broadened’ before fricatives (i.e. /s/, /f/, /th/, etc.) This was an inconsistent process: it impacted English around London the most; many Southern English dialects to some degree; Northern and Midlands English dialects only minimally (i.e. in ‘half’); and American dialects, which split from British English earlier in the process, very little.

The mystery deepens even further when one considers the word ‘rather.’ Some Americans (like myself) pronounce this word with a short-a (i.e. the vowel in ‘cat.’) Others pronounce it with a broad-a, so that it rhymes with ‘father.’ And for the life of me, I can’t figure out why. Is this a feature of certain regional dialects?

So why are ‘father’ and ‘rather’ exceptional? And while we’re at it, why does this not extend to ‘lather’ and ‘gather?’


About Ben

Ben T. Smith launched his dialect fascination while working in theatre. He has worked as an actor, playwright, director, critic and dialect coach. Other passions include linguistics, urban development, philosophy and film.
This entry was posted in American English and tagged . Bookmark the permalink.

77 Responses to A Pronunciation Mystery: American ‘Father’

  1. Ed says:

    Northern English accents do have the broad vowel [a: ~ ɑ:] in “father” for >95% of speakers. I’m unsure how long this has been the case. You can hear it with short [a] at 12:40 in this clip from Kes.

    As far as I know, all English accents have a broad vowel in “rather”, although Scottish accents may have a short vowel in both “father” and “rather”.

    “Can’t” is a similar case. All English accents have a broad vowel in “can’t”, but most American ones have a short vowel.

    • Ed says:

      By the way, I think that you meant to write “rather” instead of the first “gather” in your last paragraph.

  2. Cclinton says:

    Did you mean “father” and “rather” in that next to last line?

    I must also admit to never at all hearing the Broad-A “rather.” You live in the northeast right? If I remember right there is some A-Broadening features in some of those dialects.

    Also, I think I remember reading one of Well’s blog articles (or maybe the comments on it, I don’t really remember) where it was said that Irish (I think) actually had a THOUGHT vowel of all things for “Father.”

    • trawicks says:

      @Ed, @CClinton,

      Yup, thanks for pointing that out, guys!

      I’ll tell you two famous Northeasterners I can attest to using the Broad-A in ‘rather:’ Celebrity chef Anthony Bourdain and shock jock Howard Stern (an odd pairing, I know). Both grew up near New York, which is odd: there isn’t any kind of broad-a in the New York City accent.
      Given the infrequence of the word, it’s something you will only hear occasionally.

      • Jim Ancona says:

        “…there isn’t any kind of broad-a in the New York City accent.”

        There used to be, when upper class New Yorkers had a distinct accent. Think of FDR or Hollywood actors in 1930’s movies.

        Should “water” be though of as another exception to the rule?

      • Katie says:

        Well, there sort of is a “Broad A” in the modern New York City accent too. Their lax-tense system is related to the Broad A system of RP. Their “tense a” occurs in some of the same environments as the RP “broad A”, e.g., before voiceless fricatives.

        • Katie says:

          (to continue from last comment)…It’s just that what was presumably [æː] at one time in the precursor to both accents changed in quality. In NYC it became closer [ɛə] and in RP it became more open and backer [ɑ̟ː]. So they went in opposite directions for some reason. Or maybe [ɛə] is the original quality. Who knows?

        • Dw says:


          I assume you’re referring to pairs such as can (be able)
          vs. can (container).

          This development is independent of, and considerably later than, the Southern England TRAP-BATH split.

          The historical sequence was something like this:

          Late Midle English: a

          Pre-fricative lengthening: a -> a: in BATH words in ancestor of RP

          a fronted to æ

          a: backed to A:

          æ tensed in appropriate environment in NY an environs

  3. IVV says:

    “Several hundred years ago, English /a/ began a process of being lengthened and ‘broadened’ before fricatives (i.e. /s/, /f/, /th/, etc.)”

    I’m glad you mention that it’s not at all complete in American accents. Otherwise, I’d have to zap that sad sack of sham fact.

    And here’s where I need to bring up “almond” again. Throughout most of the USA, almond uses the broad A. However, in the San Joaquin Valley, where most of the world’s almonds are grown, almond uses the short A (like the sound used in “salmon”). No idea why.

    “Goosey Goosey Gander,
    Where do you wander?
    Upstairs, downstairs,
    And in my lady’s chamber.”

    We all know this rhymes, even though it totally doesn’t.

  4. m.m. says:

    I was like, the only place i would expect someone to say ‘rather’ with any ‘broad-a’ like quality is in canadian/californian style accents. but indeed it happens elsewhere, as i remembered reading in the ANAE which notes “remnants” of the broad-a system could be found in eastern new england, which as you mention, doesn’t include new york, so it would be odd for stern and bourdain to do it. [although, both of them *were* born in the mid 1950’s. Plus, howard stern went to uni in boston, and anthony bourdain spend his early youth in northern new jersey before ending up in the hudson valley area]

    • Julie says:

      Umm…as a Californian, I’d have to say the only place I’ve heard a ‘broad’ a in “rather” is in mockery. Of snobbery and affectedness in general. I know I’ve never heard it naturally spoken.

      • m.m. says:

        I note that I say ‘broad-a like quality’, referencing the backing/lowering of the TRAP vowel found in the speech of some that may ‘sound’ like ‘broad-a’ to others. Indeed you may not have heard it, and it’s possible to never have come across a speaker who does right now because they’re not exactly a large group [especially in sacramento], but I’ve come across fairly backed TRAPs a *few* times with people in my age group [early 20s] in regular speech here in the la area.

      • Adam Cohen says:

        I’ve hear the mock broad A rather as well, but that is a comically exaggerated form, complete with non-rhoticity: rawwwwthuhh. When I say rather, the broad A is a much faster, more clipped sound.

    • trawicks says:

      Still trying to find a good example of broad-a ‘rather’ in an American speaker for you guys. I had a similar reaction when I heard it: uh, are you trying to sound British? Sadly, it’s very hard to find examples, as it’s a single word that you could speak for hours without saying.

      • Peter S. says:

        It’s not uncommon near Boston, but then we have broad a’s in a number of other words, most notably aunt.

      • Erica says:

        I would not be surprised to find a clip of John Kerry saying rather with a long, broad “ah” sound. You might have to go back to Vietnam-era John Kerry, (he changed the way he speaks sooooo dramatically), but that’s the kind of person (Boston Brahmin) that I’ve heard do that naturally. I live on the other end of Massachusetts. The end that speaks properly 🙂

        And I can’t remember hearing anyone say the joking/mocking version for ages — probably twenty years.

      • Patrick Lane says:

        By an extraordinary coincidence, I had just read this post before listening to an old episode of NPR’s Pop Culture Happy Hour podcast (entitled “Scott Pilgrim And Our Great Big Blind Spots”), where I heard co-host Stephen Thompson (who appears to have grown up in Wisconsin) pronouncing “rather” with what sounds to me close to a broad-a, though I’m merely an English major and not a trained linguist. And furthermore, it definitely doesn’t come across as an affectation — I probably wouldn’t have really noticed it if I hadn’t just read this post.

        Anyway, you can hear his “rather” on the episode
        here at about the 33:00 mark.

      • Nathan says:

        How about this clip of Stewart Copeland, the drummer of the Police? He was born in Alexandria, Virginia, but he moved around a lot and he was obviously around English blokes a lot when he was in the Police, so I don’t know how much of an effect those things had on his speech.

  5. dw says:

    I’m going to humbly disagree with the suggestion that broadening of “father” should be classified alongside broadening of words like “bath”.

    BATH-broadening happened only in the environment of a following preconsonantal or morpheme-final voiceless fricative. Within that environment, it was pretty consistent for words in common use: exceptions are either words that didn’t yet exist in English (e.g. “gas”) or multisyllabic/learned words (e.g. “procrastinate”). Even apparent anomalies such as “castle” or “raspberry” can be explained by the fact that pronunciations at the time preserved consonants that were subsequently lost.

    “Father”, being before an intervocalic voiced fricative, obviously doesn’t belong to this category. Nevertheless, the evidence is clear that it rhymed with “gather” in late Middle English. Why, then, should it have been broadened in so many different contemporary accents?

    My own suggestion is that the word “father” was often uttered in a very respectful or religious context (the Christian notion of God the Father) which would be conducive to lengthening. The OED mentions that something similar happened to the word “God” (although this hasn’t survived in most of today’s varities of English). Dobson notes that there is some evidence of “father” having the FACE vowel in some dialects — this of course would be the result of lengthing the TRAP vowel of “father” before the Great Vowel Shift had operated.

    Another point of evidence is the word “master”, which has the “broad” value in many Northern English accents that otherwise don’t have BATH-broadening. This is also recorded with the “broad” pronunciation in Walker’s pronunciation dictionary, even though that work has little other evidence of BATH-broadening. Wells himself theorizes that the broad value of “master” in Northern English is reinforced by the context (e.g. school) in which the word is used.

    • Katie says:

      dw said…
      “The OED mentions that something similar happened to the word “God” (although this hasn’t survived in most of today’s varities of English).”

      That’s where “cor blimey” comes from, correct?

      This may be slightly off topic, but do you think the lack of a TRAP/PALM distinction in Scotland and Northern Ireland is a conservative feature?

      • dw says:


        The TRAP/PALM merger (“Sam”-“psalm”) is an innovation in Scottish/Northern Irish accents, just like the cot-caught merger which is its back-vowel equivalent

    • trawicks says:


      That’s an interesting theory about the connection between liturgy and pronunciation. There is, in fact, what sounds like a liturgical pronunciation of “God” in the American South (and African American churches as well), where the word is pronounced with a lengthened, rounded vowel for emphasis. (Come to think of it, “Jesus” seems to get a similar treatment, getting equal stress on each syllable). It does seem a common phenomenon.

      My slight rebuttal to your point about voiceless fricatives is that voiced fricatives in the context æ__C or æ___# are exceedingly rare, usually confined to basic function words (has, have), foreign loans (Mazda), or words that appeared late in the development of English (jazz). In my mind, it’s not so much that voiced fricatives are excluded from the lengthening itself so much as they are largely excluded from the environment where lengthening occurs.

      That doesn’t really explain why the ‘a’ in ‘father’ is broad, however. Although the word is unusual in that its ‘-er’ is not preceded by a morpheme boundary. I could almost see the word beginning as some kind of hypercorrection.

      • dw says:

        There are a fair number of short words with TRAP followed by /v/. Not so many followed by /ð /

        avid, cavern, gavel, gravel, ravage, (un)ravel, ravish, savage, travel, tavern

        blather, fathom, gather, slather
        (father, lather, rather)

        Dobson suggests that “rather” and “lather” were broadened late and on the model of “father”. However, this doesn’t explain why “gather” etc. weren’t similarly broadened.

        • trawicks says:

          There are a number of ‘av’ words, but few before morpheme boundaries or consonants. Words that end in ‘-ave’ are usually part of the FACE set (‘have’ and maybe a few other exceptions notwithstanding), and /vd/ is obviously a much rarer consonant cluster than /ft/! Come to think of it, it’s interesting that we have, orthographically-speaking, words that end in /ff/, but no corresponding /vv/. Much of this imbalance, I’m assuming, dates back to Old English’s treatment of voiced fricatives as mere allophones of their voiceless counterparts.

        • dw says:


          There are a number of ‘av’ words, but few before morpheme boundaries or consonants.

          Yes, but the “th” in “father” isn’t before a morpheme boundary or a consonant either. Or do you consider “father” to consist of two morphemes: “fath” “-er”?

  6. Adam Cohen says:

    RE: Almond pronunciation in San Joaquin Valley,

    I’ve never noticed anyone pronouncing it with a short A, and I live right next door in the Sacramento Valley. Everyone that I know, myself included, pronounces it with a broad A.

    RE: Rather,

    I pronounce it with a broad A, and I’m a native Californian. I also pronounce waft with a broad A, although that one seldom comes up in conversation.

    • IVV says:

      I’m only reporting what I grew up with. My parents came from Oakland and Orange County, and they hadn’t encountered (s)almond until they arrived in Stanislaus County (for which there is absolutely no consensus on whether the ending “s” is silent). Many people often said that when you shake the tree to drop the almonds, you “knock the ‘L’ out of it” and the (s)almond becomes an all-mond. Or, you know, into it, if you were an all-mond or ah-mond speaker.

      Now, I could easily believe that this has changed over the past 30 years. But that’s what I heard, and that’s what I grew up with, and it’s an oddity that very restricted geographically.

      • Ellen K. says:

        So, then, how does having a L or not connect with how the vowel is pronounced?

        • IVV says:

          Check the links below. The second speaker has a full “knock the L out” story, but reverses the pronunciations I mention here. I don’t remember there being much consensus on what’s right.

    • Julie says:

      I think the “ammond” pronunciation is getting rarer these days. My husband says his Sacramento-born mother used it (although his equally native father did not) . As a Coast native, I’ve always pronounced it with a broad a and dark l. I’ve never heard it pronounced otherwise in my 30+ years in Sacramento, but I’ve certainly heard it many times over the years.

    • Fritinancy says:

      I have an acquaintance, now in his late 60s, who uses broad-a “rather” consistently and non-ironically. He’s a native of Pasadena (Southern California) who attended boarding school on the Central California coast; he’s never lived outside the US.

      From my own California upbringing I remember older folks from both the LA area and the Bay Area who used broad-a “rather” and “ammond.” (An aside: San Francisco natives used to have a distinct accent, and as I recall broad-a “rather” was one of its features.) Pronouncing “rather” that way may have been part of an effort to sound cultured when you’d grown up in what was still the provinces: an aunt of mine, born in SoCal in 1913, used to say “thee-AY-ter” for “theater.” I think “ammond” was considered a little fancy, too–compare French amandine, which used to be common on upscale restaurant menus.

      • Victor Hugenay says:

        In my mind the broad A pronunciation in rather is not restricted to California it can also be found in the Pacific Northwest region! To give an example Washington State native Kurt Cobain also used to use a broad A in word like rather. Do you guys think that there is a correlation between pronouncing broad A’s and the phenomenon of lowering which is native to the region?

  7. m.m. says:

    Refering to this,, ‘almond’ has ‘broad-a’ with nearly 70% of californians who took it.

    +1 for “waft” with a broad-a [though technically my ‘broad-a’ is more of a [ɒ] xD]

    • Julie says:

      70% is probably conservative. I don’t think it’s used at all in any major urban areas anymore. I would expect to hear it in the smaller towns. On the other hand, that same survey does not report more than a handful of Californians who call a small stream a “crick,” but I grew up without having heard any other pronunciation (until I moved to Sacramento to attend college). The maps, in fact, show surprisingly little regional variation.

      • m.m. says:

        “Crick” is definitely more a minority in the west.
        As IVV mentions, like the central valley, which tends to sound more southern than the rest of the area, it seems the west has a trend where the farther inland you go from the coastal metropolitan areas [not necessarily rural], the more midlands the dialect becomes.
        I myself grew up never hearing a poor-pour merged person, and my default perception was that california was more fir-cure merged than anything, until I started going inland more often, and all of a sudden I was surrounded by poor-pour merged people.

        • Julie says:

          I think the rural north coast has more in common, generally, with the Pacific Northwest than with the inland counties or with any of the metropolitan areas. I think that I usually merge poor/pour, but I’m sure my husband does not.

          This video is close to my heart, but I’d like to know how you perceive the accents: Fishermen’s Fate

        • IVV says:

          I merge poor and pour (and pore). In fact, it was a common answer in school when discussing homophones (like to/too/two). Nobody thought twice about it.

        • IVV says:

          After a first listening to the Fishermen’s Fate video, I thought it was interesting to hear how accents change with the generations–but also how they stay the same. Only one of the older speakers raised the short-A sound slightly (the boat broker) but the others had very open short-A’s.

          Their sound is very Northern Californian to me. It’s not the Portland or Seattle sound. It might be Northwest-colored in the same way the Modesto speaker is Midland-colored, but that’s about it.

        • m.m. says:

          I only perceive a bit of something ‘western’ in the video.

          IVV’s “It might be Northwest-colored” reminds me of the minor term ‘or-cal’, seemingly the area north of chico up to the border. Though fort bragg would seem to be the south extent coast wise.

  8. IVV says:

    That sounds about right. In fact, some people will refer to the nut as an all-mond, but the tree as (s)almond. It’s a really interesting distinction, and I’m not entirely sure where it came from, and how it’s changing over the years.

  9. Adam Cohen says:

    29. almond
    a. all-mond (first syllable sounds like “all”) (61.96%)
    b. ah-mond (no l) (17.82%)
    c. aw-mond (if different from “ah-mond”) (3.55%)
    d. I say something in between l and nothing (13.76%)
    e. other (2.90%)

    I pronounce it ah-mond (no L). Until I saw this statistic, I was unaware that I was out-numbered by people who pronounce the L.

    • Nick says:

      Wow! The OED needs to check this out, they don’t even bother listing an L pronunciation for Americans.

      • Ellen K. says:

        Wiktionary has

        IPA: /ˈɑmənd/, IPA: /ˈɑlmənd/, IPA: /æmənd/

        For the U.S., and the first of those only for the U.K.

        • Ed says:

          I’ve never understood how the RP pronounciation of words such as “almond” is determined, since you don’t hear the word on the BBC very often.

          Everyone I know says [ɒlmənd].

        • dw says:


          I’ve never understood how the RP pronounciation of words such as “almond” is determined, since you don’t hear the word on the BBC very often.

          The same way as your [ɒlmənd], or indeed any other “pronounciation” of any other word, is “determined”: by the normal processes of linguistic change.

          RP is not defined as “what is heard on the BBC”!

        • Ellen K. says:

          I imagine Ed meant how a dictionary determines what the RP pronunciation. It’s worth noting that in Wiktionary, RP = British pronunciation. Capturing a particular accent isn’t the point.

        • dw says:


          Well, in the case of “almond”, I never heard anything other than /ˈɑːmənd/ in England: indeed I was unaware of any other pronunciation until I moved to the US. It may be a North/South thing in England; in which case RP usually follows the Southern pattern.

        • Ed says:

          @ dw: I mean what is used by BBC newsreaders – not by Radio 1 DJs. Some people (Peter Trudgill, Peter Roach) use the phrase “BBC English” instead of RP, even though it’s clear that they’re referring to exactly the same thing.

          Words such as “almond” are not part of everyday discussion for most people. I don’t understand how you can isolate the RP pronunciation from the non-RP pronunciations in such cases, since you can’t do a survey for every word. There is inevitably some subjective feeling that goes into pronunciation dictionaries. In John Wells’s excellent LPD, I struggle to understand why he gives [as a first form] twɒt for “twat” and mæntʃɪstə for “Manchester”. I have literally never heard the former.

          I have heard people say [ɑmənd] for “almond”, but I’m under the impression that it’s confined to areas further south than this. Where I live in Yorkshire, it is always [ɒlmənd]. I don’t rule out the possibility that there are a small number of words in the UK for which there is no non-regional form, and I think that this may be one of them.

        • m.m. says:

          >>I struggle to understand why he gives [as a first form] twɒt for “twat” and mæntʃɪstə for “Manchester”. I have literally never heard the former.

          I used to be the same, thinking “twat” only came in [æ] form, but I’ve since been exposed to many instances that range somewhere from [ɑ] to [ɒ].

  10. boynamedsue says:

    Gather is short a in all English dialects, including RP, probably something to do with it being a verb and so frequently being used with suffixes. Lather I’m not sure about, it not being a high enough frequency word for me to have a broad knowledge. I think RP ‘lather’ was long a in the ’40s, but not now.

  11. IVV says:

    Some (s)almond speakers:
    About almond farmers
    Shaking the trees

    An all-mond speaker:
    Nuts need bees

    An ah-mond speaker:
    Fresh almonds (It’s kind of between ah-mond and all-mond)

    If your family has been raising almonds for generations, you’re more likely to be a (s)almond speaker. So it’s a dying breed.

    • Adam Cohen says:

      That was really weird.I suppose that those guys are part of the 2.90% who are in the other category. When I played it for my Dad (he’s from southern California), he said that he recalled hearing the short A form for almond every now and then in his youth (1950s).

    • IVV says:

      Another quick note: The first two speakers are from Ripon and Denair respectively, about 30 miles away from each other. The third speaker (the all-mond speaker) is from Modesto, which is directly between the first two speakers. Yes, the Modestan has more Midland influences (that kind of Southern-inflected Western/GenAm sound).

      The last speaker is from Valencia, closer to Los Angeles. Note the complete cot-caught merger with “almond” and “father-in-law” and “apricot”–but not “all”.

      • m.m. says:

        I’m not hearing the ‘all’ in the valencia video, but I’d’ve attributed that to dark-L influences, although the ‘fall’ didn’t sound affected.

        • IVV says:

          At 0:18: “they fall on the ground and the kids all pick them up”
          at 0:23: “they’re all shriveled”

          But yeah, I’d say that “fall” matches “father-in-law”, “apricot”, and “pop”. Maybe you’re right, the difference feels subtle the more I listen to it.

        • m.m. says:

          Having listened to it knowing what to hear now, I can’t tell what I’m perceiving being a fully backed merged person. They seem to all fall somewhere between ‘broad-a’ and and back of central /a/ for me, so they all are merged to my ears.

        • IVV says:

          I think that’s the same thing I hear, too, m.m.

          I always figured Californians merged to the open backed unrounded vowel for cot, caught, and father, but when I hear pure examples of the vowel, it always sounds too backed to me. I think now that Californians, and Southern Californians especially, merge the vowels to a central open unrounded vowel.

          I think you’re right.

        • m.m. says:

          Some papers i’ve read about the subject would agree with that, as they seem to find the final merged vowel in californians leaning towards low center, which I do hear and to my ears sounds bad, because I have the other result found where the vowel is more back and possibly rounded, though I cant be sure how often I hear it because my ears don’t hear it as ‘different’ sounding compared to the central one.

  12. IVV says:

    I have a comment that needs moderation. It has links to various Californian almond farmers to get that sound down.

  13. IVV says:

    Wow, and I just realized we’ve got a speaker that goes from all-monds to (s)almonds. So, yeah, it goes in all directions–that’s a “shake the L out” speaker!

  14. Jonathan Daniels says:

    OK – I love reading these comments and was sent here by the great Lynnguist, but I confess I have no formal training in these matters. It’s just something I’m interested in. Having said that, I was thinking about father and something struck me as perhaps cultural and not geographic – the import that the word “father” has in a religious sense (Holy Father) and in a patriarchal sense. I also know that Americans tend to have high respect for British accents, again, as a cultural quirk. Perhaps the father pronunciation caught on in the US for reasons that “gather”, “rather”, etc. did not – IE, the cultural importance/significance of the word “father” combined with the American conception of British English as high-culture. I’m just thinking out loud here, and I’m sure that’s not the reason, but it struck me as a possibility.

    • Jonathan Daniels says:

      Woops – that’s LYNNEGUIST. How can I be taken seriously here if I can’t even spell more source’s name correctly? Apologies!

  15. dw says:

    And while we’re at it, why does this not extend to ‘lather’ and ‘gather?’

    I (RP unaffected in this respect by living in the US) have broad-A in lather and rather.

  16. dylan says:

    I’ll add that this curiosity also extends to Northern England. In my accent I have no broad a like southerners in words like bath, grass, answer etc, but I do in father and rather, and I think the same would apply for many other speakers of Northern English accents. A very strange phenomenon

  17. AW says:

    In northern English traditional dialect ‘father’ was pronounced [fe:ðə], so I would guess that [fɑ:ðə] is a loan from RP. I think this applies to most other words in the PALM set as well.

  18. 456 says:

    I’m a bit late to this post, and I’m not sure if this is entirely on topic, but anyway: I was wondering about words that have a broad-a in AmE but a short one in BrE. Like the first vowel of ‘pasta’, and the first of ‘aqua’ (at least when said as part of the name Aquaman when he’s referred to on The Big Bang Theory…). How did that arise?

    • AW says:

      I think they’re mostly words from latinate languages. The ‘a’ in these is fairly similar to the British short ‘a’, but somewhat different from the ‘raised’ American one. In American the broad-a is a better phonetic match.

      That’s my theory anyway.

      • Dw says:

        Exactly: and this phenomenon, which I like to think of as the MAFIA set, is by no means confined to Romance languages: it can be found in virtually any loan words. Other examples would include Rachmaninoff (Russian), Nissan (Japanese), Gandhi (Gujarati), and Gadaffi (Arabic).

  19. Andy says:

    I should imagine that ‘almond’ is a word that many English people until comparatively recently could go a whole lifetime and never use – possibly why there no regionally-linked way of pronouncing it ever developed.
    I’ve usually heard it pronounced with the short ‘a’ in both areas I’ve spent my life in (ie the South East and West Yorkshire) but also occasionally with long ‘a’ in both.
    I’d guess the long ‘a’ variety being almost completely confined to those who have been to university and had the rougher edges of their accent knocked off/or never had them in the first place (perhaps been to a fee-paying school).

    Incidentally the word ‘master’ was mentioned as an anomalous example of the broad ‘a’ in northern England – to add to that the word ‘plaster’ seems to follow that principle in certain parts of the north too.

  20. Pingback: This Week’s Language Blog Roundup | Wordnik ~ all the words

  21. Not all British, or even English, accents have a “long a” in “can’t”. How do they pronounce it in Bath, for instance? And how do they say the name of their own city?

  22. Alexander M. says:

    I believe that this “mysterious” pronunciation is due to the fact that “father” is a very fundamental, essential, even primal word; thus, it may have not gone through the vowel transformation that affected most other similarly structured words. If not in and of itself, then by the virtue of its connection to Papa, Dada (initially pronounced with a broad “a” by babies and toddlers, before they master the short “a”). Speakers of other Indo-European languages may have affected the phonology here in addition to other factors. And in most IE languages the broad “a” is used in their cognates. I would liken “father” to “Mama” to an extent. I know, it is frequently spelled “momma”, but come on! It is “mama” with a broad “a” in both syllables… And my last theory, although not very scientific, brings up a comparison of the words “mother” and “father” (self-explanatory)… Regardless, my first “hypothesis” regarding the word “father” being fundamental, essential and primal, and thus, not being a subject to the phonological shifts and transformations, is the one I believe to be most likely…

  23. m.m. says:

    …“Mama” to an extent. I know, it is frequently spelled “momma”…

    …something just snapped in my mind
    fully low back merged speakers write momma, but non merged speakers write mama

    but come on! It is “mama” with a broad “a” in both syllables

    maybe for standard mergers. in southwest PA and people with the canadian or californian shift, itd be with [ɒ] or “short o” sound as the layman say. at least the first syllable. the second is almost always a schwa.

  24. dw says:

    Rereading this thread, in particular my own earlier suggestion that the word “father” was often uttered in a very respectful or religious context (the Christian notion of God the Father) which would be conducive to lengthening, I realized that I’d missed something really obvious.

    The Lord’s Prayer, which would likely have been recited daily or at least weekly by a large proportion of Anglophone North Americans from the colonial era up to fairly recently, begins: Our Father — the metrical context naturally leading to an emphasis on the first syllable of “Father”.

    I had to recite the Lord’s Prayer daily in state-funded elementary school in England as recently as the 1980s. I know that the Supreme Court rulings in the early 1960s would have ended any tradition of such recitation at public schools in the US, but I wonder how widespread they were before that? And, of course, many Christian denominations recite the prayer as part of their own liturgy.