In some Latin alphabet Polybius squares, the I and J are merged into one letter. Normally, this doesn't really create any problems, as I and J are a vowel and a consonant, so there wouldn't really be a case where they could easily be swapped. The only one I can think of is ion and jon. Are there any more?

  • Comments are not for extended discussion; this conversation has been moved to chat. – tchrist Dec 3 '16 at 19:16

Oh please. We've already been there, before j was invented and English worked just fine then too:

However, there were some (now obsolete) rules to distinguish between vowel and consonant forms. Bullokar explains:


I has two sounds, one agreeing to its old and continued name, and is then a vowel, the other sound agreeing to the old name of g and my g', and then is a consonant: and is always used for a consonant, when it begins a syllable and a vowel next after it in the same syllable.

Even without knowing that rule, I can read "pre-j" books easily, simply because I am a native speaker. Most of the time, it's obvious it's really a j, for example subiect.

Even so, there are plenty of words that are spelled the same way in English already (some of which are also homophones). For example, it's always quite obvious if Turkey is the country, or a bird (and its meat). It's all about context.

On the other hand, English has grown used to having a separate j and i. These newer parts of the language will be the root of any problems.

For example, this change would drastically affect shorthand writing.

  • In SMS, ik and jk are two different things ("I know" and "just kidding").
  • This would affect some acronyms, potentially changing the pronunciation. (Example: SJC becomes SIC.)
  • Replacing all the j's with i's would also invalidate a very large percentage of programming code, given their use as iterator varible names, among other purposes. (Interesting history behind that, BTW.)

Also, some words with double (or more) i/j will be practically impossible to parse:

  • bijou would become biiou
  • Ijo would become Iio
  • Fiji would become Fiii
  • Beijing would become Beiiing (although we could use the old romanizarion Peking instead)

This type of change would only cause a few words to have the same spelling, and typically this would not be a problem (again, context). Actually, sometimes this would consolidate different, synonymous spellings (alleluja vs. alleluia, Maia vs. Maja).

I found some words that would now overlap.

(Method: I got a list of words from /usr/share/dict/words containing j. I replaced the j's with i's and removed all the words that had the red spellcheck line. I ignored pairs of alternative spellings, and pairs where one or more word was always a name, not in English, or obscure slang.)

I found these words:

  • field/fjeld
  • hajj/haji
    • These words are derived from the same Arabic word. There are a whole bunch of alternate spellings for hajj that cropped up, including alfaie/alfaje.
  • iamb/jamb
  • ione/jone
  • iota/jota
  • kai/kaj
    • One is NZ slang, the other is Indian; probably a non-issue with overlap.
  • maiolica/majolica
    • Majolica is earthenware made to imitate maiolica
  • rai/raj
  • raia/raja
  • sai/saj
  • tai/taj

Of these words, the only ones that are likely to be confused are field/fjeld, hajj/haji, and maiolica/majolica. Which is nothing.

How did they make it work before?

You could use 'g' for the 'j' sound in most of those words - mgb

Yes, g was often used where i would have been ambiguous. You can still see this in words: for example, edge instead of edie.

Additionally, particualarly with words of Spanish origin, it would be appropriate to use h, as in the a variant spelling of marijuana: marihuana.

*A note about names

Everyone with a name from before a change in orthography would keep the original spelling. (The DMV is already too crowded; now imagine all the Johns need to get their driver's licence changed all at once.)

You can see this in other languages, too. On German.SE, äüö explains:

No, personal names are always spelled as they are officially registered on birth. They do not fall under the changes of rules of New German Orthography.

What you would find is that people will start naming their babies differently, adapting to the new English orthography.

  • Comments are not for extended discussion; this conversation has been moved to chat. – tchrist Dec 3 '16 at 19:15

Some names could be ambiguous. For example:

  • Jo (short for Josephine) vs. Io (the Jovian moon)
  • Jan vs. Ian

Proper name ambiguity could have a real potential for causing problems that context clues as mentioned by @Laurel might not be so good against.

For example,

Io the astronomer was looking at the Iovian moons when Ian, Ian, and Iohn came in. Iohn asked Io for a picture of Io and Ian.

Has our brave heroine Io (Jo) been asked to provide a picture of:

  • Jan and the moon Io
  • Ian and the moon Io
  • Jan and Jo
  • Ian and Jo

In reality, of course, if this actually happened, people would soon come up with some convention to prevent misunderstandings, even if it wasn't splitting the "i" and the "j". Perhaps they could agree that the name Ian should be respelled "Iiann" and Jan would stay "Ian".

Christmas just won't be Christmas without massive tidal forces and volcanism!

  • 2
    It would certainly make these kinds of things worse, but they already exist in English without any problems and without any misunderstanding-preventing conventions. Names are bad in particular. Take for example: "Read to Jean? No, I read to Jean!" Each "read" could be past or present, and while one Jean is probably not pronounced like the other, we can't tell which---8 different possibilities. We make do using context. – SamM Dec 1 '16 at 19:13

A few minor ambiguities that haven't been mentioned yet (found using a Python script):

  • hajj (the journey) vs. haji (the traveler). The ambiguity could be avoided by using hajji for the traveler, which is a more natural transliteration. But people like @Laurei don't like parsing haiii. :-)
  • raja (monarch) vs. raia (a fish, etc.)
  • saj (utensil or tree) vs. sai (music)
  • Taj (place or person name) vs. tai/Tai (fish or ethnic group)
  • fjeld vs. field (HT: @Airymouse)

Arguably the less familiar ones are pretty unclear without context anyway. Also, some of them could be disambiguated with capital/lowercase, depending on which sense is meant; but that wouldn't help on a Polybius square.

All of these are "foreign" words (except field, but fjeld is), i.e. relatively recent loan words that still feel like they belong to the source language. So the fact that English didn't use to distinguish between I and J isn't very relevant to these cases.

But even with distinct I and J, we have plenty of ambiguity in our English borrowings of foreign words. And we manage that pretty well using explanations and context when needed.

Update: In case anyone wants to verify my method or improve on it, here is the Python3 script.

  • These are definitely not English words. They are foreign loan-words at best, and as such, even if English did not have distinct letters I and J, these words could still be spelled with those letters because they were derived from languages that did. Or, at the very least, they would have some disambiguating spelling, as happens with German names that have umlauts (König => Koenig). – Cody Gray Dec 1 '16 at 15:42
  • 1
    @Cody: It's a mistake to say that these are (all) not English words. Yes, they are loan words from other languages, but raja (also spelled rajah) has definitely entered English, as have hajj and hajji, and some of the others (e.g. raia) are in English dictionaries. "These words could still be spelled with [distinct I and J]" -- sure, but the question asked what words would become ambiguous without distinct I and J, as on a Polybius square. – LarsH Dec 1 '16 at 16:03
  • P.S. I agree that some of these, like sai and saj, are arguably not English words. However they are used sometimes in English communication, and I think for that reason they fit the question. But I'd be willing to concede those words. – LarsH Dec 1 '16 at 16:27

I don't think it would make sense.

Even if there are only a handful of times where the substitution of one for the other will make a different intelligible word, there are going to be plenty of times that an unfamiliar word will have to be "tasted" both ways to figure out which sound is going to make the most sense, or people mispronouncing the word because of what it looks like (depending on which got folded into which).

There are letters that have different pronunciation in different contexts - c being pronounced like an s or a k is one example I thought of - and this sort of confusion is present when someone isn't familiar with the word, or has been pronouncing it by sight. Yes, people can figure it out, and they will if they must... but it is one more point of confusion, one more place for people to stumble, and it isn't necessary.

What might make more sense is if you were to fold the J into G, since their pronunciation is similar, and leave I alone. Functionally, you still have one fewer letter, but it will be a more intuitive substitution.

As for points of confusion, as per your original question - iamb/jamb is from deadrat's comment. I thought of Ian/Jan (and I think names are particularly vulnerable to this sort of miscommunication, since we don't know what they were intended to be just from spelling). Or how someone might spell Fiji, or Reykjavík. Or anglicized words from other languages (random example - jyotishi, which is easier to find means astrologer if you know the first one and last two are different characters). Or if you're using initials, or acronyms. Again, the problem is when you don't know what the word is supposed to be, or you don't already know how it is pronounced.

Besides, what would we do with sayings like "jot and tittle" or "iota", when their meanings are no longer distinct? After all, the sayings come from pointing out little difference to which we should pay attention... why else would be differentiating between the jot over a j and the iota over an i (and the tittle across a t) matter? is it the iot over a i or the jota over a j?

  • 3
    i18nguy.com/twain.html – Jim Dec 1 '16 at 2:13
  • 2
    As Laurel mentions, people actually did this, and it worked out fine. – sumelic Dec 1 '16 at 3:40
  • 1
    @sumelic - yeess... but to be picky I should point out the "it worked out" eventually included someone inventing the j and separating out the letters, I assume because it was easier that way. As I said, people certainly can do it, and would adapt if they must, that doesn't make it not tricky and annoying and likely not necessary, especially since I don't know why they must. – Megha Dec 1 '16 at 4:08
  • 2
    See the Wikipedia entry for J; "j" wasn't created to make English easier. Early English simply didn't need to distinguish between consonant and vowel uses of "i" like Latin did. Indeed, early use of "j" in English was for Roman numerals ("xiij" for 13) and French loan words, where Latin /j/ was turning into /ʒ/. That sound is similar to the /dʒ/ found in native English, which is not word-initially; that helps explain why we write "edge", not "ej", but "jam", not "dgam". – chepner Dec 1 '16 at 5:41
  • 2
    @PierreArlaud - "by now only a memory in the minds of old dodders" - it's almost certainly some kind of slang for elders, as doddering old people. It is more familiar as a verb, though, so it takes a bit of thinking to place :) – Megha Dec 1 '16 at 9:33

Interesting that you should say "latin alphabet", as the original alphabet used by the Romans for writing Latin had no J, and they managed iust fine, using I in its place.

If you don't believe me, ask Iulus Caesar :-)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.