Today I was curious about the rarity of the consonant cluster sr in the English language. I found a WordReference forum from 2006 that asked about the matter.

The general response is that because English historically lacks the sr cluster, native speakers are unfamiliar with it, and unlikely to produce it. (The most familiar word that features sr is Sri in Sri Lanka, which I hear pronounced most often as Siri and, occasionally, as Shri.

This answer makes intuitive sense, yet surely many sounds have made their way into the English language over its history.

Can we trace any common sounds today back to a first appearance (probably in a loan word)?

  • I am less interested in sound change by way of drift—unless this is the only way new sounds have entered the language.

  • While I am curious about the full course of English history, for the sake of answerability I will restrict my question to Modern English.

  • I am especially interested in sounds that reappear in words that native speakers will not consider foreign. (From @Cascabel's comment, the /ñ/ in canyon.)

  • 1
    @Cascabel Go ahead. I think this is an interesting question.
    – Lawrence
    Commented Jul 18, 2018 at 23:32
  • 1
    @Unrelated That's an important constraint that narrows the question. I've taken the liberty of adding it to the question text. Please feel free to edit further or to roll back the change.
    – Lawrence
    Commented Jul 18, 2018 at 23:41
  • 1
    @Cascabel I am interested in both, but if the question was too broad, that was my attempt at narrowing
    – Unrelated
    Commented Jul 18, 2018 at 23:44
  • 1
    @Cascabel, canyon is a fine example: though of Spanish origin it is not considered foreign by English speakers. (Orthography is irrelevant to me.)
    – Unrelated
    Commented Jul 18, 2018 at 23:46
  • 1
    Union has the /ñ/ sound, and comes from Latin. (12th century, according to etymonline, but I don't know how it was pronounced back then). Commented Jul 19, 2018 at 1:05

2 Answers 2


I don't think there is any example of the exact thing you seem to be asking for: a "common phoneme" that has an identifiable "first appearance" in Modern English. There are some relatively uncommon phonemes that do not occur in native vocabulary, but that do occur in words that entered English around the Middle English period.

The consonant /ʒ/

The sound /ʒ/, as in the word “vision”, is commonly described as a phoneme that was introduced to English through foreign influence, even though the most common English words with /ʒ/ actually didn’t come directly from words that had /ʒ/ in a foreign language. Rather, /ʒ/ most often shows up as a result of coalescence of earlier /zj/. But /zj/ didn't occur as far as I know in native English words, so /ʒ/ only shows up in words from French and Latin (at least partially: glazier is built on the native stem glass, but has a French suffix -ier).

There are also word with /ʒ/ where it came directly from French /ʒ/, as in mirage, prestige or genre, but the sound is a bit more marginal and unstable in word-initial or word-final position: for example, some people replace /ʒ/ with /dʒ/ in "genre". (Actually, even in word-medial position, /ʒ/ is a rather "marginal" sound in that it has low contrast with /ʃ/ and there are a number of words that show variation between the voiced and voiceless sound in different accents or idolects: e.g. Asia, version, fission. But the "marginal" status of the contrasts between /ʒ/ and /dʒ/ and /ʒ/ and /ʃ/ isn't something that we only see with non-native sounds: it's well known that the contrast between /θ/ and /ð/ in English has a similarly low "functional load".)

I don't know what the "first" English word with /zj/ or /ʒ/ was.

The "choice" diphthong /ɔɪ/

Like /ʒ/, the diphthong /ɔɪ/ (as in the word "choice") is an infrequent sound that characteristically shows up in non-native vocabulary (mostly words from French).

Some common consonant clusters do not occur in native English vocabulary, and many common sounds don't occur (or barely occur) in certain contexts in native English vocabulary

If we relax the requirement that the examples be "phonemes", we can look at consonant clusters like /sk/, which doesn't occur word-initially in vocabulary inherited from Old English because of a sound change that turned it into /ʃ/. (I'm not sure if there are any exceptions. I specified "word-initially" because in word-final contexts, there are words like "ask" where I think the /sk/ > /ʃ/ change did not apply in Old English because of the existence of metathetic variation between /ks/ and /sk/.)

Word-initial /v/ or /p/ typically indicates that a word is non-Germanic, although there are a few exceptions like vat (which developed from a dialectal variant pronunciation with voicing of word-initial /f/) and path (whose Proto-Germanic etymon has an unclear etymology).

  • Thank you, this is very interesting and what I’m looking for
    – Unrelated
    Commented Jul 19, 2018 at 5:59
  • There's scrape, for which the OED's etymology mentions both Old English scrapian (to scratch) and Old Norse skrapa (to erase). So chances are any exceptions to the rule /sk/ > /ʃ/ would be blamed on Old Norse. From the OED citations, it looks to me like in Middle English, the word ended up being pronounced /skr/ and inheriting both the Old English and the Old Norse meanings. Commented Jul 19, 2018 at 10:40
  • Just to add to this, the sound /ʒ/ is not as foreign as implied by the common examples 'azure' or 'leisure'. There's lots like 'decision', 'conclusion', 'measure' that seem to have a generic sound change of palatalization from /zj/, from Latinate neologisms in ME (as pointed out by sumelic). Also, as to /ɔɪ/, there's common enough non-Latinate 'boy', 'toy', ... hm...checking, others sure do seem to mostly come from Anglo-Norman. But is that really what the OP is looking for?
    – Mitch
    Commented Jul 19, 2018 at 13:09
  • @Mitch: The etymologies of "boy" and "toy" have been, and as far as I know still are, considered unclear.
    – herisson
    Commented Jul 19, 2018 at 15:49
  • 1
    So probably the fact that /ʒ/ was an English phoneme by 1800 is the thing that enabled us to pronounce mirage and garage with the "right" consonant /ʒ/. Wikipedia estimates that /ʒ/ became a phoneme sometime in the 17th century. Commented Jul 19, 2018 at 23:35

The initial consonant cluster /gw/ doesn't seem foreign to English speakers, at least to American ones, but all the words that use it are either Spanish (guava, guano, guacamole) or Welsh (Gwendolyn, Gwen).

Maybe we don't find it foreign-sounding or difficult to pronounce because the similar consonant clusters /dw/ (dwarf) and /kw/ (queen) are native to English, being inherited from Old English.

But maybe the assimilation of initial /gw/ into American English is fairly recent. The name of the geoduck clam, in the Pacific Northwest, comes from an American Indian word pronounced with /gw/, and in 1937 it was pronounced both as /ˈgwi.dʌk/ and /ˈɡu.i.dʌk/, and was occasionally spelled gweduck. The standard pronunciation found in all the dictionaries is now /ˈɡu.i.dʌk/; I don't know whether the pronunciation with /gw/ still exists. Presumably the pronunciation /ˈɡu.i.dʌk/ arose because people were reluctant to start a word with /gw/.

A similar change seems to have happened with buoy. See this answer. Apparently one pronunciation of this word in England several centuries ago was /bwɔɪ/. This pronunciation appears to have disappeared today, probably because /bw/ is an uncommon consonant cluster at the beginning of words. It has been replaced by either /bɔɪ/ or /bu.i/ in the U.S. (and only /bɔɪ/ in the U.K.).

  • 1
    Also jaguar from Tupi, Guam from Chamorro, exceptions proving the rule.
    – Mitch
    Commented Jul 19, 2018 at 13:16
  • 1
    I've noticed a lot of people say jag-wire or, less drastically, jag-war, indicating that perhaps still struggle with the sound. Thanks @Mitch and Peter
    – Unrelated
    Commented Jul 19, 2018 at 17:35
  • 1
    I thought 'geoduck' was one of them newfangled Pokemon.
    – Mitch
    Commented Jul 19, 2018 at 18:41
  • 1
    @tautophile: I assume that the first people to write down geoduck spelled it goeduck, and the o and e were transposed by other people through confusion with the prefix geo-. Commented Jul 22, 2018 at 18:22
  • 2
    @Peter Shor: you're probably right. Of course, as many people would say about this whole discussion, "Frankly, Scallop, I don't give a clam."
    – tautophile
    Commented Jul 22, 2018 at 20:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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