ei <-> ie: English speciality?

[speising]

Very often on these and other fora, i come across spelling mistakes where instead of 'ei' people write 'ie' or vv.: theif, Murray Lienster, Albert Einstien, ... (and i could name a lot more examples if only i could think of them)
That's not a simple typing error, as those spellings will often be consistent throughout a posting.
For me as a native german speaker, this reads rather jarring, so i'm wondering whether for english speakers this is less obvious?

[Xenomortis]

It's a common source of misspellings in English.
So much so that, at least when I was at school, you're often taught the phrase "i before e except after c" (so you have field, fiend, ceiling, and receipt).
Of course, there are exceptions (neighbour, feint, weird, conscience, and society).
In fact, the exceptions outnumber the adherents, so it isn't a particularly good rule.

[jaap]

Well, English does have words with ei and words with ie where those vowels are pronounced the same, e.g. receive and reprieve.

For that vowel sound there is the general rule that after the letter c it's ei and otherwise it's ie, but there are some exceptions. I think that many English people are taught this rule badly at school when they are kids, and then are unsure which it is. And then there are all those words with ei or ie that are pronounced differently (weird, friend, height, eight, diet, and more) and they give up or misapply that rule to those words as well. [Edit: Such as the post above - those are not exceptions to the rule, those are words to which the rule does not apply. It only applies to the 'ee' sound.]

When learning English as a second language a little later in life, you tend to learn each word and its spelling individually, just accepting the spelling as it is - and when there are rules you have the maturity to understand how and when to apply the rules and what the exceptions are. Knowing a second language also makes it easier to recognise borrowed words which might not have standard spelling.

[speising]

makes sense. also, i probably pronounce the names in my examples above differently than you. (Einstein's clearly german anyway, but do you pronounce "Leinster" "Leenster"? i have no idea about the etymology)

[Xenomortis]

As a native English speaker and having read the name "Leinster" for the first time, I would guess a pronunciation of "len - ster", similar to Leicester (which is only two syllables before you get too confused).

[Diemo]

Yeah that is how it is pronounced.

Etymologically it comes from Laighin (pronounced line, is a name) and either tír or the Norse version which I can't write on the phone, which means land.

[Derek]

And then there are all those words with ei or ie that are pronounced differently (weird...

Do you not pronounce "weird" with an /i:/? As in /wi:rd/?

but do you pronounce "Leinster" "Leenster"? i have no idea about the etymology)

It looks like the name of an English town, so all pronunciation rules go out the window.

[speising]

but do you pronounce "Leinster" "Leenster"? i have no idea about the etymology)

It looks like the name of an English town, so all pronunciation rules go out the window.

Worse: Irish.

[gmalivuk]

And then there are all those words with ei or ie that are pronounced differently (weird...

Do you not pronounce "weird" with an /i:/? As in /wi:rd/?

For me at least, weird definitely rhymes with tiered.

[ThirdParty]

at least when I was at school, you're often taught the phrase "i before e except after c" (so you have field, fiend, ceiling, and receipt).
Of course, there are exceptions (neighbour, feint, weird, conscience, and society).
In fact, the exceptions outnumber the adherents, so it isn't a particularly good rule.

I was taught a longer version:

"I before E, except after C, or when sounded as A, as in 'Neighbor' and 'Weigh'; unless you're 'Weird'."

There are still some exceptions ("atheist" and "society" aren't really problems since they have syllable breaks in the middle of them, and "seize" at least has a C sound in it, but that still leaves us with words like "feisty" and "protein"), but it comes pretty close.

[Derek]

For me at least, weird definitely rhymes with tiered.

I agree, and stand by my original question?

There are still some exceptions ("atheist" and "society" aren't really problems since they have syllable breaks in the middle of them, and "seize" at least has a C sound in it, but that still leaves us with words like "feisty" and "protein"), but it comes pretty close.

"Feisty" has an /aɪ/, so it doesn't count. The rule only applies to /i:/. "Protein" is a counter-example though.

[jaap]

For me at least, weird definitely rhymes with tiered.

I agree, and stand by my original question?

Weird rhymes with tiered/feared, but this is not the same sound as feel/receive/protein.
Just like peer and peel have different sounds.
(For me at least - I have a BrE accent.)

[WilliamLehnsherr]

Great thread, spiesing.

[chridd]

For me at least, weird definitely rhymes with tiered.

I agree, and stand by my original question?

Weird rhymes with tiered/feared, but this is not the same sound as feel/receive/protein.
Just like peer and peel have different sounds.
(For me at least - I have a BrE accent.)

I'm a rhotic American English speaker, and I perceive the eir/ier/ear/eer sound in those words to be a combination of the ee sound in peel + the r sound in red and car (which I perceive to be the same sound). However, transcriptions in IPA that I've seen of that sound in rhotic dialects tend to use /ɪɹ/ or /ɪɚ/, which would suggest that it's actually the vowel in pill + the r sound in car; it's possible that my perception of the sound is influenced by spelling. (When I try to say the sound slowly, it sounds like something between /i/ (peel) and /ɪ/ (pill).) My understanding is that British English tends to be non-rhotic, and that in non-rhotic accents (or at least non-rhotic British English), eer is a separate vowel/diphthong, rather than a combination of a vowel + consonant or the usual sound of ee.

In any case, the eir/ier/ear/eer sound tends to be spelled as if it were a combination of /i/ + r, so I would expect a spelling rule that applies to the /i/ sound to also apply to the eir/ier/ear/eer sound.

[ThirdParty]

"Feisty" has an /aɪ/, so it doesn't count. The rule only applies to /i:/

"Fiery" also has an /aɪ/, at least in the dialects that don't drawl it out into /aɪə/. So do lots of short words like "tie", "died", "lies", etc. So I'm inclined to think that the "i before e" rule does apply and the "feist"/"geist"/"heist" series represents a genuine exception.

I'm a rhotic American English speaker, and I perceive the eir/ier/ear/eer sound in those words to be a combination of the ee sound in peel + the r sound in red and car (which I perceive to be the same sound). However, transcriptions in IPA that I've seen of that sound in rhotic dialects tend to use /ɪɹ/ or /ɪɚ/, which would suggest that it's actually the vowel in pill + the r sound in car; it's possible that my perception of the sound is influenced by spelling.

I'm also a rhotic American English speaker.

Try singing the words "par" and "pore", and holding the final sound in both cases. Is it the same sound? If yes, they're genuinely /pɑɹ/ and /pɔɹ/. If no, they're /pɑ˞/ and /pɔ˞/. (For me they're the same sound.)

Regarding "car" vs. "red": I think I perceive them the same but pronounce them differently--"car" is /kɑɹ/ but "red" is more like /ɹʷɛd/. Feels like the same start-of-word rule that has "cap" be /kæp/ but "peck" be /pʰɛk/.

Incidentally, the /i/ in "peel" and "peer" feels slightly longer than the /i/ in "teen". I could imagine a Texan pronouncing "peel" and "peer" with two syllables each, but "teen" could only have one.

[Derek]

"Fiery" also has an /aɪ/, at least in the dialects that don't drawl it out into /aɪə/. So do lots of short words like "tie", "died", "lies", etc. So I'm inclined to think that the "i before e" rule does apply and the "feist"/"geist"/"heist" series represents a genuine exception.

I guess. I never thought of the rule as applying to those words, and I've never seen anyone have trouble spelling them correctly (I've never seen "tie" spelled "tei", or "heist" spelled "hiest", although any misspelling I've seen might have been passed over as typos). I think there may be a different rule (or rules), that doesn't need to be learned explicitly, at play here.

"Fiery" is a derivation of "fire", and I would argue that "fiery" has three syllable (and "fire" has two) and the i and e go in different syllables, so the i before e rule wouldn't apply, just like it doesn't apply in "society".

"Tie", "die", and "lie" all fit a similar pattern (a single open syllable). Are there any words that end in /aɪ/ and spelled as -ei? Are there any words not derived from this pattern that have an internal /aɪ/ that is spelled ie? Off the top of my head I can't think of any examples for either of these.

[Copper Bezel]

Gradeschool grammar is that "the E makes the I long" in those cases, that is, makes it [aj], like "slime" and "dine." It's definitely a separate rule.

I stare at "weird" every time I write it and have variously second-guessed myself in the past. It trips up the circuits for that rule to me.

[hooked]

Does anybody know the statistical frequency of the occurrence of "ei" vs "ie"? Ideally this should be weighted by frequency in the corpus.

[ThirdParty]

Does anybody know the statistical frequency of the occurrence of "ei" vs "ie"? Ideally this should be weighted by frequency in the corpus.

After a bit of googling and creative dictionary searches, my impression is that there are about twice as many words containing "ie" as words containing "ei", but the "ei" words are about twice as frequent on average, so it works out to about even.

[gmalivuk]

COCA tells me the top 100 words with "ie" account for about 1/3 more tokens than the top 100 words with "ei"

[KarenRei]

It's a common source of misspellings in English.
So much so that, at least when I was at school, you're often taught the phrase "i before e except after c" (so you have field, fiend, ceiling, and receipt).
Of course, there are exceptions (neighbour, feint, weird, conscience, and society).
In fact, the exceptions outnumber the adherents, so it isn't a particularly good rule.

Indeed.

[a@b tmp]$ egrep -i -c ".*[^c]ie.*" /usr/share/dict/words
15848
[a@b tmp]$ egrep -i -c ".*[^c]ei.*" /usr/share/dict/words
5255
[a@b tmp]$ egrep -i -c ".*[c]ie.*" /usr/share/dict/words
869
[a@b tmp]$ egrep -i -c ".*[c]ei.*" /usr/share/dict/words
303

I before E is just more common in general - including after C. In fact C doesn't have any significant effect on the ratio. And "c(ie|ei)" is such a small portion of the total that it really doesn't justify having a special part of a "rule" about it anyway, there are far more "[^c](ie|ei)" cases.

Now to be fair, the way I did it is not the best. For one, it confuses compound words - for example, "ageism", that's not really an "ei". Also one should really bias the list towards common words. But no matter how you do it, the fact remains that it's a bad "rule".

[KarenRei]

I haven't attempted to make an optimization function, or to use a statistical breakdown of usage of words, but just from looking at it, it seems one's best bet would be to have a rule that incorporates common "stems" to words that are "ei" - for example:

"I before E, except in words like weigh, neigh, reign, vein, veil, ceiling, either, seize, or height"

or

"I before E, except after v where it's not followed by a w or the end of a word; before a g that's not 'ge'; and words like ceiling, either and seize"

It's easier to make stronger rules for i before e that hits tons of positives with few negatives - for example "ties", "tied", "lier", "cies", "ries" etc are almost always ie. But IE really is the default, it's notably more common, so rules that define it aren't as useful. Regardless, the above seems to get lots of the common words, and while grep shows lots of false positives, they're generally in things like compound words or suffixes formed by y-dropping and the like, things someone wouldn't be likely to mix up in the real world.

I wonder how true the common "except when pronounced 'ay'" correlary to the traditional rule is...

Hmm, now that I think about it, probably the best way to teach this one would be to find as many, varied "ei" words as possible and make up a sentence that uses them all (and no ie words). That way kids would just need to memorize one sentence and could get most words right

[Hominid]

I always assumed that the "i before e except after c" thing was only about the /i/ sound, so the "neighbor/weigh" addendum seems kind of pointless.

The "rule" has enough exceptions that it's useless, though.

[Eebster the Great]

I always assumed that the "i before e except after c" thing was only about the /i/ sound, so the "neighbor/weigh" addendum seems kind of pointless.

The "rule" has enough exceptions that it's useless, though.

Is that really a reason to necro the thread, though?

[Hominid]

I always assumed that the "i before e except after c" thing was only about the /i/ sound, so the "neighbor/weigh" addendum seems kind of pointless.

The "rule" has enough exceptions that it's useless, though.

Is that really a reason to necro the thread, though?

Whoops, sorry, didn't notice the date of the thread.