Episode 89: ‘I Before E’ and All That

During the Middle English period, scribes developed a variety of spelling innovations to distinguish the sound of the various vowels. Some of those innovations were borrowed from French, and some were native to English.  In this episode, we explore those spelling techniques, many of which still survive in Modern English.


34 thoughts on “Episode 89: ‘I Before E’ and All That

    • Hi Gavriel,

      I tried to limit the discussion in this episode to developments that occurred during the Middle English period. I will have more to say about the vowels when we get to Modern English. I will also cover Noah Webster’s spelling reforms in some detail at that point.

  1. Something I would have liked to hear about is words ending in -ire, like “fire” and “ire”, since their pronunciation isn’t like other words with the i-consonant-e combination. Was their pronunciation or spelling different in Middle English?

    • Hi Gil,

      “Fire” was “fyr” in Old English. “Ire” comes from French and retains its original spelling. I think that both words follow the standard convention for ‘silent E’ in Modern English. Both words have a silent E on the end, and that E indicates that the letter I is pronounced as a long vowel. Of course, the pronunciation of ‘-ire’ can vary in Modern English. For example, “shire” has a long vowel, but “Hampshire” has a short vowel.

  2. Just wanted to say that History of English is a fantastic podcast. It has so much learning and to combine history and language just make for amazing storytelling and ties it down into something that is concrete and real to all of us. Thank you!

  3. Another great episode. Long African bus rides are helping me catch up!

    There are a couple of points I’d like to make. Firstly, you mentioned how a double i (without the dot) and a double u would be difficult to read in the scripts of the time. I’d like to add that these letters, as well as other letters with similar vertical lines with no ascenders or descenders that are curved at the top or the bottom (i.e. m, n and w) are called ‘minims’ by paleographers and are grouped as minims precisely because they are difficult to read. The word minim contains only minims and is thus an appropriate name.

    Secondly, you mentioned that the French accent aigu can often indicate that the é is pronounced like a long English ‘a’ vowel, as in the word café. This is correct in the final syllables of words. I think it’s worth adding that in earlier syllables (usually the first syllable), the accent aigu indicates a different vowel sound. Take the word décédé(e), the past participle of the verb décéder (to pass away or ‘become deceased’), which contains an accent aigu on all three ‘e’ letters. The é in the first two syllables is pronounced like the short English ‘e’ vowel sound to distinguish it from the ‘e’ with no accent, which is pronounced something like the English schwa (‘uh’). The final é is pronounced like the long English ‘a’ sound, thus making the pronunciation something like deh-seh-DAY. This first syllable accent aigu is very common in French (e.g. école, écouter, étudier etc), though English speakers may have confused the two. For example, some English speakers pronounce the French word élite as ‘AY-leet’, which seems to indicate an incorrect understanding of the French accent aigu (in French, it would be eh-LEET).

    I wonder if this is where monolingual English speakers’ propensity for incorrectly pronouncing the letter ‘e’ in Romance language words as an ‘a’ sound comes from. For example, many Americans pronounce Spanish words like queso, peso and Pedro as kay-so, pay-so and Pay-dro, which is not how they are pronounced in Spanish. Hopefully this is something you might address when it comes to modern English.

    • Thanks for the notes. I have been waiting for a good time to discuss the confusion surrounding the double ‘i’ and double ‘u’ in Middle English. I have already addressed the topic in a bonus episode at Patreon, but I intend to address it in the regular podcast as well.

    • I believe that the claim about décédé is simply incorrect. If you want deh-ceh-day, you would want to spell it dècèdé. My wife has been studying and speaking French for decades and agrees totally with this. E.g. élève is ay-lev. This accent is called the grave. You get the same effect from ê except it generally marks an omitted s.

      • “This accent is called the grave. You get the same effect from ê except it generally marks an omitted s.”
        I think, too, that if we mentally insert the omitted “s” the French “fête” turns into “”feste”” which seems to parallel the English “festival” where the “e” is short.
        Being a non-expert I cannot, of course, explain the long “e” sound of the English “feast” (grin)

  4. Another great episode! I love language change.

    Do you think the purpose of the double consonants of f’s and s’s (“puff”, “miss” I believe were mentioned) was to make clear there was no allophonic variation with the voiced versions of those sounds ([v] and [z]) if they were followed by a vowel or other voiced sound (e.g. “puff of air,” “Miss Emily”), like what happens with knife vs. knives, house vs. houses (verb) that have been mentioned?

    And would the purpose of the “ck” be to emphasize the hard “c” sound rather than the palatalized “ch” (like you mentioned “child” used to be spelled “cild” I think, and “wicca” is the older version of “witch”)? Or was that resolved in a different way (“ch” being implemented already?)?


    • The short answer to both questions is “I don’t know.” (That’s usually the best answer because speculation gets me into a lot of trouble.) A historical linguist could probably give you an answer, but my research didn’t offer any theories along those lines.

  5. Love the podcast! I discovered it recently and have been binge-listening my way through to the present.

    You mentioned in this episode that the technique of adding an “i” after another vowel to indicate that the vowel should be pronounced long never really caught on for the long “u” soun (except in Scotland). What about words like “suit,” “fruit,” “juice,” and “sluice”? Are they exceptions, or did they get their spellings through some other mechanism?

    • Without thoroughly researching the history of those words, I am not entirely sure if they constitute an exception. All of those words were borrowed from French, and “suit” and “fruit” were both sometimes spelled with an ‘i’ in Old French. However, “juice” and “sluice’ apparently acquired their respective ‘i’s after they were borrowed into English. I am not sure if the spelling was changed in those words to mirror pre-existing words like “suit” and “fruit,” or if they got an ‘i’ as part of this spelling convention. (Though I suppose it would be difficult to call it a ‘spelling convention’ if it was restricted to a couple of words.)

  6. I’ll start by echoing how much I enjoy this podcast. I’m binge-listening, as you might say, and am in the 90s now.

    I took great interest in these episodes. Between listening to these last two podcasts and helping my grade-schoolers with spelling, it really drove home that the spelling rules were developed for readers, not writers. By which i meant they seem to have been designed so you knew whether the vowel was long or short when reading the word, rather than from the perspective of someone who hears a word and wants to know how to write it.

    It is interesting that, after taking into account subsequent vowel shifts, the rules do still work for a lot of words, at leas the basic ones. So many of the basic words my kids have to learn in the early grades. Lots of silent “e”s to indicate the vowel is long, lots “ai”s to indicate a long-A sound, etc. Presumably many of these words had a more classic A sound rather than our current long A sound, but still.

    Again, thank you.

  7. Kevin,

    In relistening to this episode, I latched onto your mention that double-A’s used to be common. Is this the origin of the double-A spelling in the biblical name Aaron (also the name of my oldest son)? I’ve always wondered where the two A’s came from.

    Thank you.

    • I think the spelling of ‘Aaron’ is derived from Latin which used the same spelling. The name is ultimately from Hebrew ‘Aharon,’ so I assume that the current English spelling reflects an older pronunciation with two distinct ‘a’ sounds in the first part of the name. The name was rendered in early English as both ‘Aaron’ and ‘Aron.’

    • Indeed it does. To my knowledge, there are only a handful of words that violate the rule as long as it is restricted to the ‘ee’ sound: weird, either, neither, seize, leisure (US pronunciation), sheik, caffeine, codeine, protein, and holstein. Most of those are loanwords that retain a spelling used in the original language.

      • Of course, many of those also have pronunciations that don’t use the ee sound – eye-ther, leisure rhyming with pleasure, sheik pronounced shake. All of those are how I’d pronounce them, and maybe the spelling reflects those as the original versions.

      • Thoroughly enjoying your podcast. Thank you.
        When briefly learning German I was taught that for the ie or ei pairs the second letter provided the sound pronunciation, for example ‘niece’ (neece) and ‘stein’ (stine). Please could you advise whether this convention was ever adopted in English?

  8. “ceiling” around the 25m45s mark.
    A long time ago I met “Oh Ciel!” (“heavens!”?) in The Gondoliers by Gilbert and Sullivan and in my innocence supposed that Ceiling and Ciel were related through a sense of “above”. Back in those days I had not been introduced to “cognate”.

    Given that you mention only two ei-after-c instances (ceiling and variations on perceive), I wonder why the dictionary makers did not stick with the presumably Latin origins and use “cieling”.
    This is a matter of curiousity only.

    Your podcasts are 10 times richer the 2nd time through, and I am looking forward to my third pass.

  9. Your description of I before he actually clarified some things to me. Some of those exceptions that you listed off no longer sound like exceptions. HEIR is not an exception because it is pronounced as long A. So it’s the same as neighbor and way. And it’s an EI because back in those days along E sounded like “Ehh.” So instead of a long E sound as in I before E; it’s making a long A sound. One of your other examples of an unexplained exception also struck me that way. Ancient is a dipthong, both i and e are pronounced, and they are both short. Sufficient and proficient are the same way. They aren’t indicating a long I; they are indicating a dipthong.

    • In British English, “heir” is pronounced as “air”. So I suppose that whether words are exceptions or not, depends on the speaker’s own accent.

  10. Oh you appears to be an example of where the spelling actually change the pronunciation. Because it was being spelled OU, we began pronouncing it as a dipthong. “ H aw oo s”, “H aw oo nd”.

  11. I feel like your explanation of biter versus bitter is a little more complicated than it actually is. By adding the second T, you’re breaking up the pattern of vowel-consonant-silent E. So the only reason why it is a short vowel when there’s two T’s is because you no longer have vowel-consonant-silent E. At least that’s what I was taught in school.

  12. I am enjoying this podcast immensely. Canadians who travel in America are sometime told that their pronunciation of ‘about’ sounds like ‘a boot’ Your description of the great vowel shift and subsequent modern American pronunciation shifts for words like house and mouse seems to explain the different pronunciations of the ‘ou’ letter combination in ‘about’, with the Canadian pronunciation being intermediate. But as a Canadian, I am intrigued whenever an American hears the long u (a boot) as in the Old English pronunciation, (abut?)

  13. In British English, “heir” is pronounced as “air”. So I suppose that whether words are exceptions or not, depends on the speaker’s own accent.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.