If the word-final letter sequence <tion> is preceded by a vowel sound, the most common pronunciation of the sequence is ʃn. There are, as you would expect, exceptions. The word equation is pronounced ɪˈkweɪʒn by most GBE speakers. Some words ending <ition> are variable, some speakers using ɪʃn, and other using ɪʒn. The word transition is an example.

If the sound preceding the sequence is s, then the pronunciation is tʃn. as in combustion, digestion, question and the like. For other preceding consonants the pronunciation is ʃn, as in deception, invention, protection etc.

However, I have been noticing over the past few years younger speakers who use tʃn after consonants other than s, especially after n. So, for example, invention is pronounced ɪnˈventʃn by these speakers. I don’t know how widespread this is in the UK, or if it also occurs in accents other than UK English. I wonder if this is the beginning of a sound change which will result in a categorical difference between preceding consonant and preceding vowel environments.

11 thoughts on “<-tion>

  1. Invenchuns
    These /ɪn`ventʃnz/ you’ve been hearing promp me to ponder var·ous c·nsid·rations. It remien·s me for one thing how alməs totally unaware I uʒely am of the preznce or apsnse of epenthetic tees. An· I tend to wunder w·ether historicly-speaking the· was a tee oridg·nally to be he·rd th·t then dropt out. B·t more-to-the-point I wund·r if these /ɪn`ventʃnz/ are emfatic raather thən ordn·ry forms for their uzers ie are they /ɪn`ventʃnz/ or [ɪɱ`ventʔʃnz] or [ɪɱ`venʔʃnz]? Or are such notations point`less.☺

  2. It’s my impression, and just that because I have no objective evidence at present, that the fricative part of pronunciations like ɪnˈventʃn is shorter than one would expect if the were a sequence and more like that in a “true” affricate.

    The epenthesis argument doesn’t hold for pronunciations like dɪˈseptʃn and kəˈrektʃn, both of which I have heard.

  3. @John Maidment:

    We may need to distinguish between two phenomena:

    In words like “invention”, I would suggest that epenthesis is the most likely explanation for [ntʃ]. This epenthesis is well-known, easily explained, and found in other words such as “financial”.

    A prounciation like [dɪˈseptʃn] requires, as you say, a different explanation. I admit I’m surprised to hear that any native speakers actually use this pronunciation — I’ve never noticed it myself, and it doesn’t exactly roll off the tongue! I would expect pressure of articulatory simplicity to lead fairly rapidly to loss of the labial element: from [dɪ’sɛp̚tʃn] to [dɪ’sɛtʃːn] and then possibly to [dɪ’sɛʔtʃn] or [dɪ’sɛtʃn].

  4. I think that whether or not a prounciation such as [dɪˈseptʃn] “rolls off the tongue” of a native speaker of British English may well depend on that speaker’s exposure to other languages. It took me quite some time to hear the “tstv” consonantal cluster in the middle of Polish “wydawnictwo”, but once I could hear it I found I could also pronounce it without difficulty. Exposure to the Slavic family of languages certainly accustoms one to far more consonantal clusters than one encounters in English, and I suspect that those who have had the benefit of such exposure would find that [dɪˈseptʃn] does indeed roll off the tongue without difficulty.

    Philip Taylor

  5. @Philip Taylor:

    Apologies to you and any other anglophone Slavicists I’ve offended; I was making implicit reference to the phonotactics of English, which I don’t think normally permit a hetero-organic stop + affricate cluster in English, other than across a morpheme boundary.

  6. Dammit: of course they do — “capture”, “rupture”, etc. I hereby withdraw my “doesn’t roll off the tongue” comment, and shall listen out for these pronunciations in future 🙂

  7. dw: “We may need to distinguish between two phenomena”

    Indeed, but they don’t have be unrelated.
    Things may have started off with epenthesis in invention, attention, mention, essential, etc. (quite well-documented: e.g. LPD), and then transferred to other pre-consonantal environments by analogical exten(t)ʃ(ə)n.

    So John Maidment may be quite right, and this could indeed be “the beginning of a sound change which will result in a categorical difference between preceding consonant and preceding vowel environments.”

    But it’s hard to make dɪˈzɪʃnz with so little (few) data – I quite agree.

  8. Some feedback for Phillip Minden — Phillip, I forwarded your modified version of my [tstv] cluster to a Polish friend who speaks perfect accentless English and asked his opinion as to which was more accurate, [tstv] or [tstf]. He replied “Well, 50/50? I’d prefer your version (i.e., [tstv]), though sloppiness (sort-of) prevailing in the modern world would lead me to believe your opponent’s approximation might be what you’d hear on the streets… A different interpretation of the 50/50: neither version is totally adequate, the consonant in question as spoken where I am around bangs my ears as something in between [tstf] and [tstv]. The initial consonant is definitively to be transcribed as [v].” Needless to say, I did not describe you as “my opponent” in the original message.

  9. Philip,
    thanks. I think that more or less confirms it – just go and ask the average German whether Rad is pronounced like Rat. My guess is the answer will be “Well, 50/50? I’d prefer d, though sloppiness (sort-of) prevailing in the modern world would lead me to believe an approximation of t might be what you’d hear on the streets.”

Leave a Reply

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