Wellsean Syllabification and Recapitulation Symbols in the LPD

Question

Those of you who deal with phonetics and phonology of English, and perhaps other languages as well, will surely have read John C. Wells’s article “Syllabification and allophony”, which you can find here:

http://www.phon.ucl.ac.uk/home/wells/syllabif.htm

There are some things that he discusses in the article that I do not understand. Perhaps you can help me sort them out.

For example, the syllabification of words district and extra. He states that district is /ˈdɪs.trɪkt/ and extra /ˈeks.trə/.

I understand that district cannot be */ˈdɪst.rɪkt/, because a syllable-final /t/ would be tappable and “glottallable”. However, it doesn’t brake any phonotactic constraints, does it? There are monosyllabic words that can both start and end the way both of those two syllables in this incorrectly syllabified district do. So why does he state that rule (4) is the reason such a syllabification is wrong?

On the other hand, if /tr/ is indivisible*, why is district not */ˈdɪstr.ɪkt/? That would be the result of rule (2), correct? Is it because /t/ of /tr/ is aspirated, thus it must begin a syllable, and not end it?

Since district comes from Latin dis- and stringō, stringere, from which of these two ⟨s⟩ is /s/ a remnant of and should that have any implications on Wellsean syllabification, i.e. could one posit a possible */ˈdɪ.strɪkt/?

If /dr, tr/ are indivisible, why, then, do we have Woodrow as /ˈwʊd.rəʊ/, woodruff as /ˈwʊd.rʌf/, or midriff as /ˈmɪd.rɪf/, but he allows for bedroom to be /ˈbedr.ʊm/ (it can also be /ˈbedr.uːm/, /ˈbed.ruːm/, /ˈbed.rʊm/, according to his Longman Pronunciation Dictionary)? Why hasn’t he allowed for the possibility of /ˈwʊdr.əʊ/, /ˈwʊdr.ʌf/, and /ˈmɪdr.ɪf/ in those entries respectively?

I was also wondering why *extra is /ˈeks.trə/ (and where exactly has the syllabification of this word been widely discussed)? If /tr/ is indivisible and a stronger syllables glues onto itself as many phonemes as it can without breaking phonotactic constraints, and petrol is /ˈpetr.əl/, why is extra /ˈeks.trə/ and not */ˈekstr.ə/? Aspiration? Etymology? Both? Something else?

Another example is the word monstrous /ˈmɒns.trəs/. If this was ever **/ˈmɒnst‿ər.əs/, and if a stronger syllables attracts onto it as many phonemes as it can without breaking the phonotactic constraints, why does he syllabify it as /ˈmɒns.trəs/?

In the end, regarding /dr, tr/, what are those two? Are the phonemes, which I think they’re not, or something else? What is the succinct way that defines them? Are they consonantal clusters that sometimes coalesce into an affricate, making them an allophone not of a single phoneme, but of two when the two meet?

Of course, Wellsean syllabification has exceptions, which, I believe, are often a result of how Wells himself feels a word is syllabified. The only one that comes to mind right now is the name Sade, as in – the singer Sade, which is /ˈʃɑː.deɪ/, but there are many more.

---

---

Second matter regards Wells’s aforementioned Longman Pronunciation Dictionary? Those of you who have it will known that when recapitulating syllables, he will use the symbols bullet ⟨•⟩ and ⟨-⟩ hyphen.

Does any one of you who regularly rifles through the dictionary know when does he use which?

I presume that the bullet, obviously, stands for one syllable, and the hyphen for two or more, but why does he sometimes omit the stress marks and sometimes includes them or why does he sometimes recapitulates the whole word with bullet symbols and stress marks when he can easily bundle several syllables into a hyphen without creating confusion? On the other hand, although one can say it’s glaringly obvious to which syllable the alternative pronunciation refers to, the entry reexamine just uses hyphens and no bullet symbols?

Just when I start thinking I might have got it, another entry appears which demolishes my theory.

Then he also sometimes splits pronunciation with a semicolon. Why?

Minor matters, I know, but it itches me to solve it.

I’d also like to point out that I think he wrote somewhere that he was given limited space to explain certain matters and choices he made (probably on his blog somewhere), but I wish he expanded on them while he was writing the blog regularly.

*It seems that /tr/ is sometimes /tr/ and sometime /t/ + /r/, but it’s not really explained when. I guess one has to use one’s own pronunciation as a guide.

Answer 1

Partial answer:

As commenters and other answers have mentioned, morpheme boundaries may be important to the syllabification of a word. Wells's rules are not meant to apply to all words: the existence of exceptions is made clear in the Introduction (which refers to "competing pronunciations that differ only in syllabification" and "morpheme boundaries ... that upset the implicitly expected syllabification").

If /dr, tr/ are indivisible, why, then, do we have Woodrow as /ˈwʊd.rəʊ/, woodruff as /ˈwʊd.rʌf/, or midriff as /ˈmɪd.rɪf/, but he allows for bedroom to be /ˈbedr.ʊm/ (it can also be /ˈbedr.uːm/, /ˈbed.ruːm/, /ˈbed.rʊm/, according to his Longman Pronunciation Dictionary)? Why hasn’t he allowed for the possibility of /ˈwʊdr.əʊ/, /ˈwʊdr.ʌf/, and /ˈmɪdr.ɪf/ in those entries respectively?

Wells does not argue that the sequences /dr/ and /tr/ are indivisible. What he argues is that the presence of affrication in /dr/ or /tr/ implies that there is not a syllable boundary between /d/ or /t/ and a following /r/. Whether /dr/ and /tr/ are pronounced with affrication cannot be deduced just from knowing the sequence of phonemes: it must be determined empirically by listening to how people pronounce them. Presumably, Wells did not include transcriptions like /ˈwʊdr.əʊ/, /ˈwʊdr.ʌf/, and /ˈmɪdr.ɪf/ for one or more of the following reasons:

• he felt that he had not heard people pronounce these words these ways

• he felt that these pronunciations were not common enough to be worth mentioning

• he felt that these pronunciations were not standard enough to be worth mentioning

could one posit a possible */ˈdɪ.strɪkt/?

Wells would say that this violates, for no clear reason, a constraint against ending a stressed syllable in a short vowel. Compare the example mentioned in the section "The phonotactic condition":

Better is /ˈbet.ə/ rather than */ˈbe.tə/ not only because the /t/ triggers pre-fortis clipping and is tappable, but also because a syllable /ˈbe/ would not be phonotactically well-formed.

I'm not sure how to answer your questions about words like extra, district, monstrous because I've also wondered how to fit these in with Wells's idea that consonants are preferentially syllabified with the stronger adjacent syllable, and that /tr/ is a valid coda in words like "petrol". Perhaps Wells would say that /tr/ and /dr/ can only be analyzed as codas when preceded by stressed vowels.

Answer 2

(Disclaimer: ..surely have read John C. Wells’s article “Syllabification and allophony... no to both of the things, i am a high school student interested in linguistics)

1) since you labelled this question as phonotactics, each language has its own phonotactics, thus making some syllables feasible and pronouncable and "seamless" in that particular language thus:

a) /ˈdɪst.rɪkt/ -> you would have to literally pronounce it like two words ~seamless~

b) /ˈekstr.ə/ -> "kstr" ~feasible~ in English , also American English generally allows a vowel-only-last-syllable when it is aspirated so even if "ekstr" was possible, it would have to end with "əh" (with that h flying above)

2) A language should have syllables that exist or fit phonology coszily

a)/'dɪstr.ɪkt/ -> both syllables don't already exist in major dialects of English (by syllables I mean separate syllables independently, not like "distract" also has the same string of characters so it should be right to separate distr)

b) wood , mid, row and riff already exist as common words / prefixes (which I know is not the case here) and that's the point, it's (idk psychologically?) easier to break words into known parts rather than maintaining some rules. Take a non-existent word - Bigred , did you pronounce this like big-red in your mind? Why not bigr-ed (maybe that's right because of some rule which says gr is indivisible as it exists in grass , Groot and so on) but no big and red are "easier" to pronounce.

3) you can (mostly) only pronounce what the spelling says (it's stating the obvious)

a) since monstrous is not monsterous it is pronounced as ˈmɒns.trəs/ and not the other one.

/ˈdɪ.strɪkt/ (I am not really sure of, but I think two consonant clusters make it a really heavy syllable in English, but in other languages like Hindi you can actually pronounce it easily and it is a spoken variant around here in India)

In the end, regarding /dr, tr/... (Those are too many linguistic terms hope the info above and people around here help)

(Curly waves mean a cut through the word)