1

Ok, we got a lot of words with suffix "-tion" in English like reflection or congestion.

But the way to pronounce "-tion" is different sometimes.

congestion /kənˈdʒes.tʃən/

reflection /rɪˈflek.ʃən/

SO, What are the rules to pronounce the suffix "-tion" in English, when to say "/-tʃən/" and when to say "/ʃən/"?

Tom
  • 4,737

2 Answers2

4

The most important thing to remember is to always use /ʃən/ after a vowel: for example, in words like condition or preparation – using /tʃ/ here would sound very unnatural. One word that is a weird exception is equation, where most speakers have a voiced consonant /ʒən/ for some unclear reason.

The pronunciation /tʃən/ is only required after an /s/ (in words spelled with -stion); /ʃən/ is generally used after any other consonant.

The only complication I can think of that applies to more than one word is the pronunciation after /n/, in the sequence -ntion. For many speakers, the contrast between /ntʃ/ and /nʃ/ is neutralized in some contexts, in particular after a stressed syllable (this is analogous to the "prince-prints merger"). So some speakers might pronounce a word like attention with [tʃən]; but these speakers might also have [tʃən] as a possibility in the word tension.

The difference between the pronunciation of the suffix in words like relation (or abstraction) on the one hand, and words like congestion on the other, is actually rather old (it predates the introduction of these words to English) and while it isn't reflected in the French and English spellings (which are based on the Latin etymology), it is reflected in the spelling of equivalent words in languages with more phonemically transparent writing systems:

Spanish: relación, abstracción, congestión
Italian: relazione, astrazione, congestione

Of course, the "rule" I outlined does not apply to all words that just end in the letters "tion" in English: the most notable counterexample is the word cation. However, I can't think of any counterexamples to my rule when dealing with the actual suffix -tion.

Note: I used the example words relation and abstraction rather than reflection because the latter word has alternated with reflexion in the past, giving it a more complicated history.

Community
  • 1
herisson
  • 81,803
  • +1 It's interesting to note that it would be very difficult to pronounce a /s ʃ/ sequence as we would normally assimilate the /s/ to the following /ʃ/. In fact any run of sibilants is unlikely. – Araucaria - Him Oct 15 '15 at 15:30
1

Interesting question. The examples you give, and a few others, seem to have /t/ lost between a voiceless stop (/p/ /t/ /k/) and following /ʃən/.

Greg Lee
  • 17,406
  • 1
    Well, the [t] was lost a long time ago in most positions, I think. After a vowel in words like "condition" or "preparation"... In French, these words have /s/, descended from earlier [ts]. The words were spelled with "cion" in some Middle English texts, which indicates that they were borrowed with [ts] (according to Wikipedia, [ts] later reduced to [s] in parallel changes in French and English). In contrast, words like "congestion" have [t] in French up to today. – herisson Oct 11 '15 at 03:55
  • @sumelic, But what happens to /t/ at the end of the stem in opt/option, act/action, etc.? Isn't that the point? – Greg Lee Oct 11 '15 at 04:08
  • oh, are you talking about synchronic analysis? – herisson Oct 11 '15 at 04:09
  • @sumelic, did you read the question? – Greg Lee Oct 11 '15 at 04:10
  • The question was about how to pronounce these words, but you introduced the idea that the /t/ is "lost" between voiceless stops and /ʃən/. I assumed you meant "[diachronically] lost", which doesn't seem to be the actual course of development (the loss of [t] predates yod-coalescence, and the [ʃ] was historically never a separate phoneme from the [t]). If you don't mean that, you might want to add some clarification to avoid other people like me getting confused by the wording. – herisson Oct 11 '15 at 04:13
  • I meant "lost" in the sense not pronounced. You might expect a /t/ to be pronounced, but you don't hear any /t/. Is it really this hard to understand? I don't understand what you want clarified. – Greg Lee Oct 11 '15 at 04:37
  • Well, I don't see why one would expect a /t/ to be pronounced, and I wouldn't say it was "lost" (when discussing these words in English alone) any more than I'd say the word initial has lost a /t/ between a preceding vowel /ɪ/ and following /ʃəl/ (or that presence has lost a /t/ between /n/ and /s/). Isn't the stem of act, ag (like in agitate or agent)? So it receives the suffix /t/ to form a verb, and the suffix /ʃən/ to form a noun. – herisson Oct 11 '15 at 04:43
  • @GregLee, Sumelic Well, we'd expect the /t/ to go there because it occurs in an environment in which we'd expect alveolar plosive elision. Namely it occurs at the end of a syllable and on the border or a morpheme boundary. The /t/ is surrounded by consonants and the preceding consonant matches for voicing. These are exactly the circumstances where we'd expect an elision of /t/ or /d/. Notice the same applies to the potentially expected /d/ in "attention". – Araucaria - Him Oct 15 '15 at 15:34
  • 1
    @Araucaria, yes, when you add a suffix to a stem, you expect that you'll wind up with the stem pronounced then the suffix pronounced. That is just how suffixation works in human languages. I can't fathom sumelic's difficulty with this. And as you say, it is perfectly ordinary for stem final /t/ to be lost between consonants in English -- it even happens optionally across word boundaries. – Greg Lee Oct 15 '15 at 16:10
  • My difficulty is with the idea that the word "reflection" consists synchronically in English of the verb /rɪˈflɛkt/ + a suffix /ʃən/. Is "division" morphophonologically /dɪvaɪd/ + /ʒən/? If so, why is the /d/ lost here? It seems simpler to me to postulate two distinct suffixes /t/ and /ʃən/ that alternate depending on the class of speech. – herisson Nov 14 '15 at 10:50
  • @sumelic, That's a straw man. No, "division" is not morphophonologically /dɪvaɪd/ + /ʒən/. And even if it were, how could that possibly imply that "reflect" has a suffix "-t"?? That is a very strange idea. – Greg Lee Nov 14 '15 at 14:02