In American English, the article “an” is used for a vowel sound to separate the words so they don’t blend together when speaking.
Normally, “a” always precedes a consonant, while “an” precedes a vowel. But “an” also precedes vowel sounds - i.e., the sound of the letter of the beginning of a word.
An apple
A banana
An hour
Hour starts with a consonant, but is pronounced with a vowel sound at the beginning. Thus, it is not “a hour” and rather “an hour”.
In the case of the example from the meme, id argue that either article works:
A “I’m…” - Typically when speaking, a person has a brief pause before they begin the quote. Since that pause would be enough to distinctly indicate two separate words, this sounds fine when being verbally spoken.
An “I’m…” - Looks great in text and would be the correct way to list it grammatically. However when speaking this aloud, since the person would have a brief pause when saying “an” and then the quote, it probably wouldn’t sound as great to some others.
My take - I like “an ‘I’m…’” best. Both in text and verbal form. Others may disagree as far as verbally said; however, grammatically in written form this is how it should be.
Edit: Fixed the inevitable autocorrects from typing this on mobile.
Because different dialects pronounce herb differently, sometimes the ‘h’ is pronounced, sometimes not.
I know you specified American English, but even within American English, you can find areas that differ on this, and I’m sure there are other words where this kind of thing crops up.
Also, I guess this may be worth mentioning as well:
Though this no longer seems to be as common as it was 10-20 years ago…
“An hero.”
Sometimes, either an unintentional misuse of this a/an rule will be ironically copied, and more widely used, essentially to either mock the original usage/user…
… or the a/an rule can be intentionally misused, as a way of infantilizing yourself, trying to come off as cutesy/ditzy, or maybe play up your own awkardness or inexperience, something like that.
Native English speaker here. This is incorrect, the “n” is added for phonetic help “a elephant” involves an awkward break between the two words, so enter “n” to help mouth muscles work around that.
This is the same reason for weird artifacts like: “a unicorn” because unicorn starts with. “Yoo” sound and so mouths don’t need the help of the “n” to break up the awkwardness.
Also seconds language but I’ve learned it depends on how the first letter of the following word is spoken. If the following word has a vocal sound (even if the first letter is technically a consonant), it’s ‘an’, which it thusly would also be here.
Edit: Instead of responding to individual comments, I’ll just put what’s going on with “a” and “an” here:
This alternation is a morphophonological process (specifically a sandhi alternation), whereby in native, fluent speech for most dialects of English, “a” is unconsciously placed before words beginning with a consonant, and “an” is unconsciously placed before words beginning with a vowel.
In contrast to what many people in this subthread seem to think, this is NOT to “ease pronunciation”. This is easily demonstrable since “a” and “the” have the same vowel sound in fluent speech (for most dialects of English), but while we get “a cat” but “an apple”, we don’t get “the cat”, but “then apple”. This alternation, therefore, is not a regular part of English speakers’ phonology (that is, part of the regular, unconscious processes that occur between sounds in all environments), but rather an idiosyncratic part of English’s morphophonology, in that it’s a phonological process that only happens in the presence of certain morphemes (simple words or word-pieces).
Why is this the case? Because “an” was originally just the word “one” that became reduced over time until it took on its own separate grammatical function, and later there was a regular sound change whereby “n” was deleted in certain specific unstressed environments before consonants, leaving an accidental alternation between “a” and “an” as a result of sound change.
This means that the “a”/“an” alternation in Modern English is not to “ease pronunciation” in any way - like with many phenomena in English (and all languages for that matter), it’s just a vestigial remnant of an accidental historical process.
We know this is the case because the exact same thing happened to “mine”, and in earlier dialects of English there was a similar alternation, “my cat”, but “mine uncle”. This alternation later collapsed in most dialects into our modern my/mine distinction, adding further evidence to the conclusion that this is not a phonological alternation, but a morphophonological one.
What all of this means, is that for a native English speaker that still has an “a/an” distinction (I don’t have one in my dialect, for example - I put “a” before everything when speaking fluently: “a cat”, “a apple”), if they don’t put a pause between “a” and “I’m” to signal the quoted speech, they would likely say “an I’m”, and if they do put a pause between “a” and “I’m” to signal the quoted speech, they would likely say “a I’m”.
Because “a” and “mindset” aren’t in a local configuration to each other, they will have no morphophonological influence on each other whatsoever (just like in “an able mindset”, for example).
So, while I won’t say that saying “it’s ‘a’ because of ‘mindset’” is wrong (because right/wrong aren’t really useful terms when describing language), I will say that it does not conform to the linguistic behavior of native English speakers when speaking fluently.
Feel free to respond to this comment with any follow-up questions you have, and I’ll be happy to answer them.
First, I agree with most of what your saying, but:
This means that the “a”/“an” alternation in Modern English is not to “ease pronunciation” in any way - like with many phenomena in English (and all languages for that matter), it’s just a vestigial remnant of an accidental historical process.
Why do you frame that as a dichotomy? To ease pronunciation, we take the older form (containing the consonant at the end) when a vowel follows and the reduced form (without the consonant) when a consonant follows. We alternate between these forms to ease pronunciation. Same for “the”: Arguably, the “strong the” is not /þi:/ but /þıj/ ending in a constant (/j/) and is therefore favored when a consonant follows to ease pronunciation. Sometimes it’s used for emphasis which also happens with “an” so it’s basically the same phenomenon.
There are other factors at play, as you pointed out the break to indicate quotation and regional differences. Also the glotal stop might not be consciously perceived but still trigger the same result as any consonant.
I for one use the a/an distinction as I learned it at school while having a glottal stop heavy accent due to my native language so I will say stuff like /ʔən ʔɛpl/ and act surprised when people know where I’m from.
Good questions - hopefully the explanation here helps clarify my position.
To ease pronunciation, we take the older form (containing the consonant at the end) when a vowel follows and the reduced form (without the consonant) when a consonant follows.
We don’t, though. This is clear from the fact that “the” occurs in exactly the same phonetic environment (including the lack of stress), with exactly the same vowel, and it doesn’t show the same behavior. This data tells us that there’s no articulatory reason for this alternation. There is no phonotactic constraint active in English that speakers are getting around with this behavior - the process is specific to a single morpheme.
There are tons of other ways we could make this exact same sequence of unstressed schwa followed by another stressed vowel as well, and in exactly none of them do we ever see an “n” inserted to repair the hiatus the way we do with /r/ in many dialects (which one could analyze as an example of “easing pronunciation”, depending on one’s assumptions, though I probably wouldn’t with all of the deserved stigma around the ill-defined idea of “easing pronunciation”). This is telling us that this alternation has nothing to do with “ease of pronunciation”, since speakers clearly don’t need their pronunciation eased in this environment.
As for “strong the” specifically, we see a parallel form in “strong a”, which can also be argued to end with a yod, and which seems to alternate under the same conditions as “strong the” in most dialects, whatever those conditions are. For this reason, I don’t really think “strong the” is very relevant to the discussion.
When the sound change originally took place, of course, it could be argued that it was for “ease of articulation” purposes since the change was regular, but post facto explanations for sound change are always a bit dicey.
So, if you want to argue that the original source of the alternation was “ease of pronunciation”, well, sure, maybe, but it’s pretty clear from Modern English data that the “a/an” alternation has nothing to do with ease of articulation at all.
It’s a dichotomy because something either eases pronunciation, or it doesn’t, and in this case, the data makes it clear that it doesn’t. It may feel “easier” to speakers because it sounds wrong to them without it, but that’s due to morphophonology, not phonotactics, and it’s why we rely on tests like the above instead of speaker intuition whenever possible.
How about this: let’s take the f/v morphophonemic alternation in leaf/leaves, knife/knives, etc.
There’s a decent argument to be made that this medial voicing change in Old English was originally to “ease pronunciation”, but once this alternation became morphophonemic, the “ease of articulation” argument falls apart pretty quickly.
I don’t think any serious linguist would assert that it’s ‘life/lives’ in Modern English due to “ease of pronunciation” instead of “historical accident” when ‘fife/fifes’ and countless other later borrowings do not show the same alternation, and the ‘a/an’ alternation is this exact same sort of morphophonemic process.
You do not seem to be aware of that since your variety also lacks the a/an distinction, but ‘“strong the” before vowel’ is a rule at least promoted by my teachers. So it is the same phenomenon. This is true for other words, too like to (/tə/ vs /tuw/). “a/an” is just the only example visible in writing and your variety doesn’t seem to have these distinctions at all so your excused for not knowing about them. Lindsey has a video, I can look it up later.
since speakers clearly don’t need their pronunciation eased in this environment.
Saying it is possible to pronounce doesn’t mean it can’t be eased (is that wording right? You know what I mean). Language changes isn’t that regular. There are distances I sometimes walk or take the bus to ease the travel. This isn’t that strict. Language change often effects some words, or a single one but not others. For instance “listen” has a silent “t” in most varieties even tho it’s easy to pronounce and speakers didn’t need this shift, it just happened. Examples where this didn’t happen prove nothing.
And your f/v example fails because it’s fossilized. The “e” that softens the f to v went silent but the v stayed voiced and even voiced the s (to z). The a/an distinction on the other hand is productive. It’s “an honor to join a union” since the h in honor is silent and “union” starts with [j]. People even say “an historical event” because the unstressed h is too weak. Even the glottal stop, while not consciously perceived as a consonant, can trigger this. Lindsey also shows this in a video.
When the sound change originally took place, of course, it could be argued that it was for “ease of articulation” purposes since the change was regular, but post facto explanations for sound change are always a bit dicey.
So when did it stop ease the articulation? When it fossilized? When it stops being productive? Because, as shown above, it didn’t. It’s still regular. Silent letters don’t effect it, it’s still all about pronunciation, about easing the articulation and only implemented where it does this job. And it always only effected this word so it was never as regular as [f]>[v] between vowels. It always, and still does, effect this one word in a very regular way.
And it’s not an inserted “n”. I think of it as an “a/an alternation” but you can also think of it “an losing its n” just like to (/təw/) loses its w. And this framework also explains why “my/mine” didn’t stay: when /i:/ shifted to /aj/, there was a consonant at the end anymore and the /n/ no longer needed to ease the articulation.
I hope you see why I don’t think your position is very convincing. How can you ignore than simplifying pronunciation is a key factor in language change?
I’m not disagreeing with your larger point but I don’t necessarily buy the part of your explanation saying
This is easily demonstrable since “a” and “the” have the same vowel sound in fluent speech (for most dialects of English), but while we get “a cat” but “an apple”, we don’t get “the cat”, but “then apple”
because in most dialects (at least of American English) “the” before a consonant uses ə while before a vowel sound it’s ē.
I don’t think that’s accurate, but I’d be happy to see a source proving me wrong. I looked briefly, but wasn’t able to find a paper dealing with that alternation specifically (though I didn’t look very long, and there may very well be one).
Also, I’m pretty sure that for the dialects that do use “strong the”, they also use “strong a” in exactly the same environments, which to my mind makes it a non-issue.
Either way, there are plenty of other ways to get a word-final unstressed schwa followed by a word-initial stressed vowel, and we never see an “n” repair in any of those other situations either - the important point is that this is a process centered entirely around a single lexical item, and it doesn’t make sense for a process affecting a single lexical item in a common environment to be “easing pronunciation”.
We expect our management to know when to use ‘an’ instead of ‘a’.
Would it be “an”? Does the a/an rule apply to whatever the next word is or does it apply to the word it is targeting? “An mindset” would be incorrect.
In American English, the article “an” is used for a vowel sound to separate the words so they don’t blend together when speaking.
Normally, “a” always precedes a consonant, while “an” precedes a vowel. But “an” also precedes vowel sounds - i.e., the sound of the letter of the beginning of a word.
Hour starts with a consonant, but is pronounced with a vowel sound at the beginning. Thus, it is not “a hour” and rather “an hour”.
In the case of the example from the meme, id argue that either article works:
My take - I like “an ‘I’m…’” best. Both in text and verbal form. Others may disagree as far as verbally said; however, grammatically in written form this is how it should be.
Edit: Fixed the inevitable autocorrects from typing this on mobile.
This does get confusing with say…
“A/An herb.”
Because different dialects pronounce herb differently, sometimes the ‘h’ is pronounced, sometimes not.
I know you specified American English, but even within American English, you can find areas that differ on this, and I’m sure there are other words where this kind of thing crops up.
Also, I guess this may be worth mentioning as well:
Though this no longer seems to be as common as it was 10-20 years ago…
“An hero.”
Sometimes, either an unintentional misuse of this a/an rule will be ironically copied, and more widely used, essentially to either mock the original usage/user…
… or the a/an rule can be intentionally misused, as a way of infantilizing yourself, trying to come off as cutesy/ditzy, or maybe play up your own awkardness or inexperience, something like that.
English being my second language, from why I’ve learnt, “a […] mindset” is correct.Edit: I stand corrected
The rule I’ve always used is that if the first letter of the word immediately following it is a vowel, it’s “an” and if it isn’t use “a”.
For example, “an apple” or “a potato”. If there is an adjective, go by that first letter, for example “a large apple” or “an average potato”.
For anyone scrolling, I’ve followed a similar rule. Except an is used anytime the following word makes a phonetic vowel sound. E.g ah, eh, ee, oh, ooh
Tell that to British midlanders: “Can I have a glass of wo’a?”
Native English speaker here. This is incorrect, the “n” is added for phonetic help “a elephant” involves an awkward break between the two words, so enter “n” to help mouth muscles work around that.
This is the same reason for weird artifacts like: “a unicorn” because unicorn starts with. “Yoo” sound and so mouths don’t need the help of the “n” to break up the awkwardness.
Also seconds language but I’ve learned it depends on how the first letter of the following word is spoken. If the following word has a vocal sound (even if the first letter is technically a consonant), it’s ‘an’, which it thusly would also be here.
So much badlinguistics in this subthread.
Edit: Instead of responding to individual comments, I’ll just put what’s going on with “a” and “an” here:
This alternation is a morphophonological process (specifically a sandhi alternation), whereby in native, fluent speech for most dialects of English, “a” is unconsciously placed before words beginning with a consonant, and “an” is unconsciously placed before words beginning with a vowel.
In contrast to what many people in this subthread seem to think, this is NOT to “ease pronunciation”. This is easily demonstrable since “a” and “the” have the same vowel sound in fluent speech (for most dialects of English), but while we get “a cat” but “an apple”, we don’t get “the cat”, but “then apple”. This alternation, therefore, is not a regular part of English speakers’ phonology (that is, part of the regular, unconscious processes that occur between sounds in all environments), but rather an idiosyncratic part of English’s morphophonology, in that it’s a phonological process that only happens in the presence of certain morphemes (simple words or word-pieces).
Why is this the case? Because “an” was originally just the word “one” that became reduced over time until it took on its own separate grammatical function, and later there was a regular sound change whereby “n” was deleted in certain specific unstressed environments before consonants, leaving an accidental alternation between “a” and “an” as a result of sound change.
This means that the “a”/“an” alternation in Modern English is not to “ease pronunciation” in any way - like with many phenomena in English (and all languages for that matter), it’s just a vestigial remnant of an accidental historical process.
We know this is the case because the exact same thing happened to “mine”, and in earlier dialects of English there was a similar alternation, “my cat”, but “mine uncle”. This alternation later collapsed in most dialects into our modern my/mine distinction, adding further evidence to the conclusion that this is not a phonological alternation, but a morphophonological one.
What all of this means, is that for a native English speaker that still has an “a/an” distinction (I don’t have one in my dialect, for example - I put “a” before everything when speaking fluently: “a cat”, “a apple”), if they don’t put a pause between “a” and “I’m” to signal the quoted speech, they would likely say “an I’m”, and if they do put a pause between “a” and “I’m” to signal the quoted speech, they would likely say “a I’m”.
Because “a” and “mindset” aren’t in a local configuration to each other, they will have no morphophonological influence on each other whatsoever (just like in “an able mindset”, for example).
So, while I won’t say that saying “it’s ‘a’ because of ‘mindset’” is wrong (because right/wrong aren’t really useful terms when describing language), I will say that it does not conform to the linguistic behavior of native English speakers when speaking fluently.
Feel free to respond to this comment with any follow-up questions you have, and I’ll be happy to answer them.
First, I agree with most of what your saying, but:
Why do you frame that as a dichotomy? To ease pronunciation, we take the older form (containing the consonant at the end) when a vowel follows and the reduced form (without the consonant) when a consonant follows. We alternate between these forms to ease pronunciation. Same for “the”: Arguably, the “strong the” is not /þi:/ but /þıj/ ending in a constant (/j/) and is therefore favored when a consonant follows to ease pronunciation. Sometimes it’s used for emphasis which also happens with “an” so it’s basically the same phenomenon.
There are other factors at play, as you pointed out the break to indicate quotation and regional differences. Also the glotal stop might not be consciously perceived but still trigger the same result as any consonant.
I for one use the a/an distinction as I learned it at school while having a glottal stop heavy accent due to my native language so I will say stuff like /ʔən ʔɛpl/ and act surprised when people know where I’m from.
Good questions - hopefully the explanation here helps clarify my position.
We don’t, though. This is clear from the fact that “the” occurs in exactly the same phonetic environment (including the lack of stress), with exactly the same vowel, and it doesn’t show the same behavior. This data tells us that there’s no articulatory reason for this alternation. There is no phonotactic constraint active in English that speakers are getting around with this behavior - the process is specific to a single morpheme.
There are tons of other ways we could make this exact same sequence of unstressed schwa followed by another stressed vowel as well, and in exactly none of them do we ever see an “n” inserted to repair the hiatus the way we do with /r/ in many dialects (which one could analyze as an example of “easing pronunciation”, depending on one’s assumptions, though I probably wouldn’t with all of the deserved stigma around the ill-defined idea of “easing pronunciation”). This is telling us that this alternation has nothing to do with “ease of pronunciation”, since speakers clearly don’t need their pronunciation eased in this environment.
As for “strong the” specifically, we see a parallel form in “strong a”, which can also be argued to end with a yod, and which seems to alternate under the same conditions as “strong the” in most dialects, whatever those conditions are. For this reason, I don’t really think “strong the” is very relevant to the discussion.
When the sound change originally took place, of course, it could be argued that it was for “ease of articulation” purposes since the change was regular, but post facto explanations for sound change are always a bit dicey.
So, if you want to argue that the original source of the alternation was “ease of pronunciation”, well, sure, maybe, but it’s pretty clear from Modern English data that the “a/an” alternation has nothing to do with ease of articulation at all.
It’s a dichotomy because something either eases pronunciation, or it doesn’t, and in this case, the data makes it clear that it doesn’t. It may feel “easier” to speakers because it sounds wrong to them without it, but that’s due to morphophonology, not phonotactics, and it’s why we rely on tests like the above instead of speaker intuition whenever possible.
How about this: let’s take the f/v morphophonemic alternation in leaf/leaves, knife/knives, etc.
There’s a decent argument to be made that this medial voicing change in Old English was originally to “ease pronunciation”, but once this alternation became morphophonemic, the “ease of articulation” argument falls apart pretty quickly.
I don’t think any serious linguist would assert that it’s ‘life/lives’ in Modern English due to “ease of pronunciation” instead of “historical accident” when ‘fife/fifes’ and countless other later borrowings do not show the same alternation, and the ‘a/an’ alternation is this exact same sort of morphophonemic process.
You do not seem to be aware of that since your variety also lacks the a/an distinction, but ‘“strong the” before vowel’ is a rule at least promoted by my teachers. So it is the same phenomenon. This is true for other words, too like to (/tə/ vs /tuw/). “a/an” is just the only example visible in writing and your variety doesn’t seem to have these distinctions at all so your excused for not knowing about them. Lindsey has a video, I can look it up later.
Saying it is possible to pronounce doesn’t mean it can’t be eased (is that wording right? You know what I mean). Language changes isn’t that regular. There are distances I sometimes walk or take the bus to ease the travel. This isn’t that strict. Language change often effects some words, or a single one but not others. For instance “listen” has a silent “t” in most varieties even tho it’s easy to pronounce and speakers didn’t need this shift, it just happened. Examples where this didn’t happen prove nothing.
And your f/v example fails because it’s fossilized. The “e” that softens the f to v went silent but the v stayed voiced and even voiced the s (to z). The a/an distinction on the other hand is productive. It’s “an honor to join a union” since the h in honor is silent and “union” starts with [j]. People even say “an historical event” because the unstressed h is too weak. Even the glottal stop, while not consciously perceived as a consonant, can trigger this. Lindsey also shows this in a video.
So when did it stop ease the articulation? When it fossilized? When it stops being productive? Because, as shown above, it didn’t. It’s still regular. Silent letters don’t effect it, it’s still all about pronunciation, about easing the articulation and only implemented where it does this job. And it always only effected this word so it was never as regular as [f]>[v] between vowels. It always, and still does, effect this one word in a very regular way.
And it’s not an inserted “n”. I think of it as an “a/an alternation” but you can also think of it “an losing its n” just like to (/təw/) loses its w. And this framework also explains why “my/mine” didn’t stay: when /i:/ shifted to /aj/, there was a consonant at the end anymore and the /n/ no longer needed to ease the articulation.
I hope you see why I don’t think your position is very convincing. How can you ignore than simplifying pronunciation is a key factor in language change?
And all the prescriptivists just collapsed onto their fainting couches.
(I kid, nicely done. Also fuck prescriptivists.)
I’m not disagreeing with your larger point but I don’t necessarily buy the part of your explanation saying
because in most dialects (at least of American English) “the” before a consonant uses ə while before a vowel sound it’s ē.
I don’t think that’s accurate, but I’d be happy to see a source proving me wrong. I looked briefly, but wasn’t able to find a paper dealing with that alternation specifically (though I didn’t look very long, and there may very well be one).
Also, I’m pretty sure that for the dialects that do use “strong the”, they also use “strong a” in exactly the same environments, which to my mind makes it a non-issue.
Either way, there are plenty of other ways to get a word-final unstressed schwa followed by a word-initial stressed vowel, and we never see an “n” repair in any of those other situations either - the important point is that this is a process centered entirely around a single lexical item, and it doesn’t make sense for a process affecting a single lexical item in a common environment to be “easing pronunciation”.