The Cot-Caught Merger

A Cot

A Cot! (Wikimedia)

One of the major distinctions in American English is something called the Cot-Caught Merger.  This is exactly what it sounds like: some dialects merge the sounds in words like cot, lot and Tom with the vowel in caught, paw, and thought.  Dialects in the Western United states almost always have this merger; most dialects in the Eastern half of the US do not (with the exception of Northeastern New England).

So, whereas somebody from New Jersey might pronounce cot and caught as IPA kɑt and kɔt (“caht” and “cawht”), somebody from Los Angeles might pronounce these words as IPA kɑt and kɑt(“caht” and “caht”). In other words, the same.

Seems like not a big deal, right? And yet you’d be surprised at how much passion it provokes in people. English-language geeks like myself will spend hours discussing the precise dimensions and specificities of the merger.  I was once part of a (now-defunct) language forum where there were three-hundred-page debates over this piece of linguistic minutiae.

Okay, maybe I’m exaggerating.  But let me get to the point here:  I think I’m cot-caught merging.

This seems strange.  I grew up in Connecticut, which is not a Cot-Caught merged state, and I lived in New York City, perhaps the least Cot-Caught merged place on earth, for 13 years (cot is IPA kat whereas caught is IPA kʊət — that is, caht and caw-uht). But I’ve recently noticed that my vowel for cot (or lot or cod) has started to take on a more rounded and backer quality, while the vowel in caught has begun to inch a little forward.

And I’m not the only one. Upon listening to a recording of my beloved significant other, a voiceover artist, I noticed that she tended to pronounce both sounds with a slightly rounded vowel (IPA ɒ). And she grew up in greater Philadelphia, which is another area of extreme difference between the two phonemes.

So what’s going on here?  Well, confession time: my girlfriend and I were both trained as actors, and let’s face it, actors aren’t great examples of “normal” dialect behavior. But I have read that the cot-caught merger is spreading like wildfire. Why?

As with all language change, there are bound to be numerous contributing factors, but one possibility I would posit is the rise of the technology sector over the past thirty years. Dialect changes tend to gravitate from where the money is, and until the housing boom, the money was in tech-driven areas like California and Washington State.

Then again, in most parts of the US, the Cot-Caught distinction wasn’t that great to begin with. In places where these sounds are kept distinct, Caught is often only a very slightly rounded version of Cot.

Anybody else feel like their “cots” aren’t all that different from being “caught” these days?


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.

28 Responses to The Cot-Caught Merger

  1. Laura says:

    Here in Chicago, I think dey are all cats.

  2. Aaron Bauman says:

    Related: the “Aaron-Erin” merger.
    Where I grew up in the midwest, they’re the same.

    So terribly annoying to have East Coasters tell me I’m pronouncing my own name wrong.

  3. Joe DeStreter says:

    @Aaron: well, sorry, but it *does* sound wrong to us [even if traveling and linguistics eventually teaches that there’s no ‘wrong’ or ‘right’ in dialects ;)]; we East Coasters hate when we hear Midwesterners or Californians say the name “Air-un,” b/c it’s not obvious without context whether they mean someone male or female [same thing with Don and Dawn, of course. Also “Ann” and “Ian” with some Great Lakes accents, eg Buffalo, Detroit, Cleveland…

    • trawicks says:

      To be fair, a lot of what you’re classifying as “East Coast” accents here are one particular stretch of the East Coast: The urban corridor from New York down to Baltimore or thereabouts. It’s a different story if you go up to New England, where “Aaron” is more commonly “air-un” outside of the major cities and “Ann” is often a bit like “Ian,” similar to the Great Lakes. “Don” and “dawn” are also famously merged Boston and much of northern New England.

    • Ellen K. says:

      I’ve only heard of the Don/Dawn merger as a (part of the) east coast thing, not a Midwest thing. And they are certain different in my experience, having lived all my life in the Midwest (not upper).

  4. 'enry 'iggins says:

    “Also “Ann” and “Ian” with some Great Lakes accents, eg Buffalo, Detroit, Cleveland…”
    As long as you’re consistent in your hatred. If you hate it when people from the Great Lakes region say “Ann” then you also have to hate in when people from Eastern New England (“East Coasters” by anyone’s definition) say “Ann”, because they do the same thing.

  5. AC says:

    I don’t think the merger in the eastern US is really limited to northeast New England. I’m from southeastern WV, and most people I know the merger.

    Of course, the merger might be new to this area: a lot of people moved here from New England in the 1970’s. I’ve noticed southwestern WV usually makes the distinction: “cot” is kɒt whilst “caught” is like kɒʊt.

  6. Pingback: Phonics Friday #3 – Diphthongs and when to use two “L”s vs one “L” - Toshuō

  7. benw says:

    I wonder whether it depends on how old you are? I grew up in Tucson, AZ from age 4-17, which is supposedly a cot-caught merging state, but I don’t have the merger. I was in New Haven CT before that, and I’m 42, which means I was in Tucson from 1974 – 1987 and in CT from 1970 – 1974. Did Tucson not have the merger then? Did I somehow pick it up from CT? I dunno.

    Odd to say you are fronting “caught” and rounding “cot” — that’s the opposite of what it should be.

    BTW I’m a linguist but for whatever reason I never notice whether people merge or don’t merge the cot-caught vowels. Possibly this is because the two vowels for me are pretty close: caught has /ɒ/ (very much like British “not” but not nearly so short) and cot has a vowel that’s probably halfway between back /ɑ/ and central /ä/.

    One weirdness: For me, the vowel of “calm” is the same as “caught” (/l/ not pronounced). Did I back-form this from people saying /kɒlm/?

    Here’s another weirdness: my speech merges “Mary” and “marry” but “merry” is ever so slightly different (the merry vowel is low-mid, the Mary/marry is a bit higher, maybe mid, and they get even closer in fast speech). Where did this come from? I doubt this is normal in Tucson and the sounds are so close in my speech that often my own judgments can get a bit messed up, but in general they seem to agree closely with what non-mergers do (the merry vowel occurs only in open syllables and only when written “e” — but also in the word “bury”).

  8. Nicholas Chmielewski says:

    I’m from Cleveland, and I know that I am not cot-caught merged, /kät/ (750, 1250) and /kɔt/ (650, 850). Also, I know that I pronounce Ann as a diphthong, but Ian is two syllables.

  9. Pingback: Phonics Friday #8 – words that contain “all” or “alk” - Toshuō

  10. Daren says:

    This cot-caught merger you’re talking about is so weird because it seems to have sprung up out of nowhere and swept the continent. I have dictionaries from the 1950’s that include detailed prefaces, with precise explanations of very subtle dialectal differences in every English phoneme, yet not one of these dictionaries mentions this merger. Apparently, just a few decades the merger was still so rare (or stigmatized, perhaps) as to be unmentionable.

    Also, I’ve read that mainly younger people have the merger. Older people in most areas make the distinction.

    Yet, what’s strange is that even though the merger is so new, it can’t be traced down to any particular dialect. It has strongholds in widely separated parts of North America. That is, the merger has no obvious geographical origin.

    The only explanation I can think of is that this is a “delocalized” speech variation proliferated by the 20th Century mass media. Some rich westerners in Silicon Valley dropped the distinction, and then various people from many American dialects heard them on TV and decided to copy them. In other words, this merger is in some sense a socio-linguistic “fad.”

    This is how I interpret this article and the available evidence. I could be wrong, of course.

  11. Nicholas Chmielewski says:

    Lots of the comments mention the spread of the merger coming from Silicon Valley, however the Bay Area is the only part of the west that differentiates. The origin of the merger is from Scottish English in Canada (in Scotland it is all /kot/) It spread from canada to the west and north central. In western PA the origin is from the way the German Immigrants spoke English. I’m not sure what the origin is in NE, but they differentiate the vowels of father and bother. It is spreading to various parts of the midland from both the west and western pa.

    • dw says:

      I live in the Bay Area, and most people here have the merger. Perhaps the best evidence is my four-year-old daughter, who has lived all her life here and has it in spades.

  12. Pingback: "Orphan Black's" House of Dialect Mirrors | Dialect Blog

  13. Hcat says:

    I’m a native Southern Californian, but being 63 I have somewhat of a cot-caught distinction. Most of my younger neighbors don’t. But The Lord-lard distinction is very marked for Californians – to merge them sounds Texan. To me “lord” and stressed “or” are “loard” and “oar” very different from any vowel of “caught” or “law” and it confuses me that dictionaries use the same symbol for both.

    • Andrea says:

      They use the same symbol for the vowels in “or” and “caught,” even though they don’t sound the same to you, because the (unmerged) vowel in “caught” /ɔ/, is there, but possibly rhotacized in /ɔr/; that is, the vowel is being affected by the following sound, which would account for its sounding different (it IS different, but very slightly, and the underlying sound is the same). Or it could just be that it isn’t as different as it seems to be, because it’s much more difficult to distinguish between vowels and the American /r/ than sounds like the /t/ in “caught”.
      Also, the environment of “before /r/” is the exception, even for the dialects that merge; /ɔ/ does not merge to /ɑ/ before /r/. Maybe the rhoticization saves it? I really don’t know why. But words like “par” (/pɑr/) and “pour” (/pɔr/) will still be different, even for someone who solidly merges “caught” and “cot” (and “pa” and “paw”).
      Perhaps it really has more to do with the following consonant than is usually thought to be the case. The voiceless stop /t/ had already shortened the vowels, at least slightly, before any merging effect came around. Does this merger happen as strongly before voiced stops which have not been shortened? Maybe there’s a reason this particular word pair has been nominalized for the entire merger; maybe it is the most exemplary case of it.

      • benw says:

        No, the vowel in most Americans’ pronunciation of “port” and “core” is more like the vowel of “coat” than “caught” — it’s basically the same vowel but minus the glide at the end, plus an /r/. That’s what the poster means by the spellings “loard” and “oar” (= lord, or). This doesn’t apply to New York non-rhotic speech, where “lord” and “laud” merge, but does to Black non-rhotic speech, where “floor” sounds like “flow” not “flaw”.

        • Andrea says:

          Well, the vowel in my “port” and “core” is definitely more ɔ than o ( see, and so is most of America’s if Wikipedia is to be trusted (and I’d say it is in most cases), and that’s why the same symbol is used in dictionaries. If you say it differently, then your dialect simply isn’t being represented.

          But if “oar” has the same vowel as “coat” for you, (for me this sounds like “o’er” as in “over” without the [v] and is not how I would say “or” even when stressed), then it isn’t relevant to the “caught”/”cot” merger anyway, because that’s a totally different vowel.

          For instance, for “horrible” posh Britians say [ˈhɒɹɪbɫ̩], some New Yorkers, New Jersians, and Carolinans say [ˈhɑɹəbɫ̩] (what is meant by the vowel in “cot” and kind of like how I say “car”), and most Americans and Canadians say [ˈhɔɹəbɫ̩] (what is meant by the vowel in the unmerged “caught” and like how I say “core”). I don’t use [ɒ] unless I’m trying to adopt a British accent, and is simply [ɑ] + rounding.

        • Tom says:

          He didn’t say (i.e. write) “oar” had the same vowel as “coat”, he said the vowel in that word was more like the vowel of “coat” than that of “caught”. He would still have a distinction between “oar” and “o’er” presumably. I think I do use phone [ɒ]*, but it’s in words like “hawk”, etc. not in words like “box” as in English accents.

          * I don’t think this is uncommon in American English as a phone. As a phoneme, distinct from both that of dawn and that of father, it’s almost completely non-existent AFAIK. But that doesn’t mean that Americans aren’t allowed to use the phone [ɒ]. Not that the vast majority of us would be aware of it if we did use it anyway.

  14. Jimmy Carl Black says:

    What’s interesting (and a bit frustrating) to me about almost everything written on the cot-caught merger is that it doesn’t mention that the merger that takes place in the accents of Boston and some other areas of Eastern and Northern New England is the reverse of what it is elsewhere. To me, this is a completely separate merger which should be classified differently. In the typical American cot-caught merger (e.g., someone from the Midwest), the two words are pronounced as kɑt. BUT… in places such as Boston, they are both pronounced as kɔt. Why isn’t this explained and seen as significant? That’s not the same merger at all!

    Think of it this way. Say there are two subspecies of the same species of bird that once lived side-by-side in the same geographic range: one red and one yellow. Over time, they interbred and one sub-species took on the color of the other. BUT, in one region, the birds all became red, and in other region, they all became yellow. This “merger” wouldn’t be described as being the same in both regions. We would consider that the red subspecies assimilated into the yellow in one region; that the yellow subspecies assimilated into the red in the other.

    Likewise, given the differences, this isn’t really one linguistic merger at all, but two opposite ones. In most regions of America (where this has occurred), “caught” began to pronounced as “cot.” In other regions (e.g., part of New England), “cot” began to be pronounced as “caught.”

    • dw says:

      I think you have a point here. The fact that Bostonians merged COT into CAUGHT may explain another distinctive feature of the Boston accent — its resistance to the FATHER/BOTHER merger found everywhere else in North America.

      The idea is that the Boston LOT vowel no longer existed (having been merged into THOUGHT) at the time the FATHER-BOTHER merger operated everywhere else.

    • Nathan Brown says:

      Agreed. I have sometimes wondered about the origins of the caught/cot merger in New England — it’s not a feature of English speech, or of the Irish who followed them. The only place nearby that has the merger is Canada. I would be surprised if the merger came from there, since most of the Canadians who settled New England were French speakers. But then again, Canadian raising is also prevalent in much of New England (it’s the norm in the Boston area nowadays), so maybe there’s more Canadian influence on New England speech than we know.

      It seems to be spreading in some areas of New England — younger people in the Boston area seem to have the merger more strongly and consistently than older ones, I’d say. And it’s become usual in Vermont, too; I think the early linguistic studies showed Vermont to keep the two sounds separate. Even people who are rhotic and otherwise show little trace of a New England accent typically have the merger.

  15. Nathan Brown says:

    I think you’re right that the merger is spreading. I sometimes hear it from people from areas where you wouldn’t expect it, like Upstate New York (the merger is pretty prevalent in the Adirondacks, among younger speakers, and I’ve heard it from some younger people in the suburbs of Albany).

    As you said, the two vowels aren’t much different for many non-merged Americans — they’re about the same length, the only difference is the rounding. It’s not like in England, where “caught” is very audibly longer and more closed.

    Some accents with other shifts seem to resist it — in areas where you have the Northern Cities Shift, New York City-style “caught” tensing, or a Southern drawl that can turn “caught” into a diphthong, you don’t have the merger. I’d be willing to bet that the merger is going to become the norm, if it hasn’t already, in most accents that don’t have any of these resistant features.

  16. smiley faces says:

    My brother recommended I might like this web site.
    He was entirely right. This post actually made my day.
    You can not imagine simply how much time I had spent for this information! Thanks!

  17. Karen says:

    I think I may have a merger of sorts: I am originally from NW Pennsylvania but have lived in Minnesota for the past 35 years. I merge the words “vowel,” “towel,” and “owl” with “val,” “tal,” and “al” so they rhyme with “pal.” When I moved to MN in 3rd grade my teacher sent me to speech classes to “lose my accent” (once they realized it was not a speech impediment) but apparently they didn’t take. I’ve been troubled by that ever since. Would someone please tell me this is common somewhere, (even if it isn’t)?

  18. Pingback: Accent of the Week: Wisconsin (with some Minnesota thrown in) | Language and other musings