Syllabification of "windy".

Stephen W.   Sat Feb 24, 2007 12:14 am GMT
Why do so many dictionaries get the syllabification of "windy" wrong? It's /wInd.i/, not /wIn.di/. Compare "brandish" /br{nd.IS/ "wield, wave" and "bran-dish" /br{n.dIS/ "dish for bran". Clearly the two words sound different. They sound different because the syllabification is different. The first syllable in "brandish" ends at the /d/ and the first syllable of "bran-dish" ends at the /n/.
Lazar   Sat Feb 24, 2007 12:37 am GMT
I completely disagree with you.

<<They sound different because the syllabification is different.>>

No, they sound different because "brandish" has only one stress, whereas "bran-dish" has a secondary stress: ["br{n.dIS] versus ["br{n%dIS]. Your syllabification scheme is the one proposed by JC Wells, and I think the problem is that British dictionary-makers like Wells have a weird aversion to marking post-tonic secondary stress.

I can think of a hypothetical word "brandish" ("brand-ish", relating to a brand or brands), which I would intuitively syllabify as ["br{nd.IS]. I could also think of a hypothetical contrast between "baby" ("babe-y", relating to babes) and "baby" (infant): the infant is definitely ["beI.bi], whereas the contrived hypothetical adjective would be ["beIb.i].
Lazar   Sat Feb 24, 2007 12:43 am GMT
And just to be clear about the first example you gave, I do think the syllabification of "windy" is ["wIn.di].
Stephen W.   Sat Feb 24, 2007 8:12 pm GMT
What about the syllabification of "handle"? Would you say it's /h{nd.=l/ or /h{n.d=l/? A dictionary I have at home tends to list syllables as always ending immediately before a syllabic consonant, and thus lists /h{nd.=l/.
Lazar   Sat Feb 24, 2007 8:27 pm GMT
I think I'd syllabify it as ["h{n.d5=].
Stephen W.   Sat Feb 24, 2007 8:27 pm GMT
For example, while it lists "candy" as being /k{n.di/ it lists "candle" as being /k{n.dl=/ due to the syllabic l.
Stephen W.   Sat Feb 24, 2007 8:29 pm GMT
<<For example, while it lists "candy" as being /k{n.di/ it lists "candle" as being /k{n.dl=/ due to the syllabic l.>>

Typo. It lists it as /k{nd.l=/. I find it odd that it claims "candy" and "candle" actually have syllable breaks at different places.
Buddhaheart   Thu Mar 01, 2007 6:21 am GMT
We must admit syllabification is a controversial subject. Not many phonologists can agree on any principles or methods. However, we shouldn’t have any qualms on well-established English phonotactic
constraints or rules.

JC Wells of LPD adopts a principle to facilitate the prediction of distribution of allophones. LPD wants to see syllabification parallels the morphology of the word. Hence “windy” is syllabified as /``wInd,i/ as windy = wind + y.

D Jones of EPD follows the widely known MOP (Maximal Onsets Principle) according to which we would place as many consonants as possible with the onset of a syllable without violating any English phoneme sequencing constraints. “Windy” could be syllabified as /``wI.ndi/; but ‘nd’ is not a permissible cluster in the onset position, it should be divided as /``wIn.di/.

The word is syllabified similar to EPD by Lexicologists like Gage Canadian & Webster’s. I must stress one or the other approach does go against one’s intuition. Either way, the word is pronounced the same to me. Perhaps think of the /d/ being ‘ambisyllabically’ shared by both syllables
/``wIn[d]i/ in the analysis of LPD. This again is another controversial area of phonology I won’t dare venture into much!

The word “brandish” is also syllabified by the above lexicologists according to their principle or policy. “Bran-dish” is not exactly an accepted word like “brand-new”. Or perhaps it is? If we must treat it as a
compound noun, a stress generalization would put the primary accent on the first word “bran”. I would syllabify it as /``br{n dIS/. A secondary stress may also considered in the second part of the word /%dIS/. A pause between “bran” & “dish” in speech serves also to distinguish this word from “brandish”.