In a certain Sherlock Holmes story (The Adventure of the Norwood Builder), there's a cry of 'Fire', repeated three times.

The relevant segment of the story:

"Might I ask you all to join in the cry of ‘Fire!’? Now then; one, two, three——"

"Fire!" we all yelled.

"Thank you. I will trouble you once again."


"Just once more, gentlemen, and all together."


The audiobook version is read by someone who (to my non-native ears, at least) speaks with a cultured, pleasant-sounding British English.

In his reading, the shouts seem to come out more like 'fɑː' than 'faɪə(r)'. You can listen to it in this mp3 file, from 16:24 to 16:45.

Is that how shouting may distort pronunciation, or is it a perfectly ordinary pronuncation that sounds a bit strange only to me?

  • I'm not prepared to listen to 20 minutes of sound files, but shouting (and exhilaration, panic etc) will obviously alter the sound characteristics. Wind instruments behave differently at higher volumes, their timbre changing. – Edwin Ashworth Mar 21 '17 at 10:44
  • 2
    You can actually pull the slider right to 16:24 without having to listen to the preceding 16 minutes and 23 seconds. – Bepe Mar 21 '17 at 10:48
  • But you need to listen to context to assess a 'base level'. Note how the pronunciation changes between the 'Would you all shout "Fire", please' and the actual shout. Dictionaries give the normally voiced version (assuming they get it right) and shouting almost always distorts pronunciation, and not consistently. Shouted words, like song lyrics, are probably best considered off-topic. – Edwin Ashworth Mar 21 '17 at 11:05
  • When someone actually shouts, there is often an added emotional or some other factor affecting the distinctness of their pronuncation. But in this case, it's only someone reproducing a shout for his listeners, which shouldn't prevent him from pronouncing it loudly but clearly. I was thinking that perhaps for a native English-speaker a distinctly pronounced shout (like a long, loud faɪə(r)) sounds contrived, unnatural? – Bepe Mar 21 '17 at 11:15
  • It depends how good the actors are. – Edwin Ashworth Mar 21 '17 at 14:02

The [fɑː] that you heard is an ordinary pronunciation variant of "fire". This kind of variant doesn't only occur in shouted words. In fact, I'd think a "triphthongal" pronunciation [aɪə] (what you probably think of as the "usual" pronunciation of "fire") would be more, not less common in an emphatic setting like that.

What you are hearing is "smoothing," a feature of pronunciation that optionally applies to the sequences /aɪə/ or /aʊə/ in many non-rhotic accents of English. The high glide or semivowel in the middle is elided, and in modern accents the schwa is usually assimilated to the first vowel, resulting in a long monophthong. In modern British accents with smoothing, generally /aɪə/ is simplified to [ɑː] (more or less the same vowel used in the word "father") and /aʊə/ is simplified to [aː] (a vowel that doesn't really occur in other contexts; it's similar to a long version of the vowel in "trap"). Some other slightly different outcomes are possible in more old-fashioned accents.

You can read more detailed descriptions of smoothing on Geoff Lindsey's Speech Talk blog ("Smoothing, then and now"; this has some nice audio files) and John Wells's phonetic blog ("triphthongs, anyone?", "the Cowell jumped over the moon," "our cake").

A slightly different, but related phenomenon: even accents that don't usually have smoothing in content words like "fire" may have it in certain frequent function words. I have a rhotic accent, but I usually have "triphthong" smoothing in our [ɑ˞], and I think I have it at least sometimes in I'll /ɑl/ and while /wɑl/ (but I think when this word is unstressed, I often reduce it even further to [wəl]). These words may also behave differently in accents that do have a generally-applicable process of smoothing; for example, it seems some British English speakers might use [ɑː] for our despite using [aː] for the smoothed pronunciation of power.

  • Very useful answer, thank you. Still, in the recording, the same narrator uses two different pronunciations within seconds of each other. If it's not because of the shouting, then could this be to show a difference between the cultured Holmes and the more uncouth county constables (who provide most of the shout)? Might there be some such notion attached to a smoothed vs unsmoothed pronunciation? – Bepe Mar 22 '17 at 10:54

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.