Is there something like Syllable Dictionary?

Paul M   Wednesday, June 09, 2004, 06:56 GMT
I was thinking.
Do you know how many possible syllables there are in English?
If there is a some kind of list of every syllables there are in English, I wouldn't mind having a look at it..

Of course if I caculate with 20 vowels and 24 consonents, the possibilities are almost limitless. say.. there is one phoneme syllable.. that's like 21? 22?
Two phonemes syllables will be like.. umm..20*24 + 23*20 = 940
Three phonems syllable can be.. 20*24*24 + 23*20*24 = ???
That can't be right..

Wait I just searched the web and ther are roughly 12000 syllables in English.
I hope there are some kind of syllables dictionary.. :(
Someone   Wednesday, June 09, 2004, 07:02 GMT
This isn't helping, but why would you need a syllable dictionary?
Paul M   Wednesday, June 09, 2004, 08:26 GMT
I guess I want to see some pattern in them..
And wouldn't it be fun to mix and match them to make a nursery rhyme! :)
Willy   Friday, June 11, 2004, 03:56 GMT
"compromise" does not rhyme with "promise"

/'kom-pruh-''mIz/ /'prom-is/

[Ety. com- or pro- with -ise from Latin]


Where's the verb standardization in conjugation for English? Oh, it doesn't have!

cantare to sing

podere can (an auxiliary verb)

vivire to drink
Jim   Friday, June 11, 2004, 06:27 GMT
Twelve thousand sounds about right. It's big but not limitless. There are patterns there, I'm sure, but they're not all necessarily going to be easy to pick out when you're dealing with twelve million syllables.

I read that all the syllables of English can be fit into a certain pattern. This pattern is initial consonant cluster, vowel then final consonant cluster. Final consonant clusters have from zero to four consonants. Initial consonant clusters have from zero to three consonants.

Of those intial consonant clusters with three consonants, I read, the first is always /s/ and the second is always /p/, /t/ or /k/. It also seems that the third is always /w/, /j/, /l/ or /r/. Some of these are really rare, e.g. /skl/ as in "sclerosis".

Hey, talking about syllables, you're from Corea, aren't you Paul? I'm from Sydney. In Japanese katakana "Sydney" comes out as three syllables: "Shidoni". In English it's only two syllables. It can't be helped: that's the way katakana is. However, in hangul (Is my spelling right?) they could have written it as two syllables. The thing I can't figure out is why they didn't. Do you have any idea?
Paul M   Friday, June 11, 2004, 15:08 GMT
Thanks Jim for your reply.
I was hoping that the number is less than that frankly.
I wonder what the most frequently used syllables are.

Yes, I'm from Korea but I'm living in Sydney too. :)

Yes, you are absolutely right about that.
Sydney can be written it in 2 syllables and we just don't..
It could've written as

sI nI

instead of..

sI d. nI

And it would've made much more sense.

Ok, guess is.... this
When consonants become a part of final clusters in Hangul, they tend to lose a lot of their characters if it's one of the fricatives or plosives. So, in this instance, /d/ would very much sound like /n/ because it's heavily affected by the next consonant. If you somehow force the /d/ sound, then /d/ would've sounded (to a Korean) as another syllable and since there can't be a syllable without a vowel in Korean, we say /d/ as /d./

Also, the fact that Koreans are used to read one glyph as one syllable where as in English, one glyph is one phoneme (usually) doesn't help either as it's not very clear exactly where we should divide the syllables in English words.

Hope that helps..
Paul   Friday, June 11, 2004, 15:24 GMT
If there are 12,000 possible syllable in English, probably less than 3,000 are used. I would like to see a frequency distribution.

Also if you were using a Syllabic Alphabet. You would probably split it into 4 forms.
1. Stand-alone Vowel
2. Vowel plus Consonant Cluster
3. Consonant Cluster plus vowel
4. Optional Closing Consonant cluster for type 3.

Syllables would be written by symbols for 1, 2, 3, 3+4

Any opinion on how many symbols such a English syllabic alphabet would require?
Jim   Monday, June 14, 2004, 00:21 GMT
In my post above I wrote "... dealing with twelve million syllables." I meant "thousand" not "million".

Paul M, your expanation of why it's written as three syllables makes sense. I suppose I have much to learn about Hangul.

I think you're right, Paul, that not all possible syllables are used. For example, consider all the possible syllables starting with /skl/, there must be hundreds but, as far as I know, the only one used is /skl../ ("sclerosis" and "sclorotic").

Paul, if I understand you correctly you're splitting the set of all English syllables into:

1 V i.e. vowel
2 VC i.e. vowel then consonant cluster
3 CV i.e. consonant cluster then vowel
3+4 CVC i.e. consonant cluster then vowel then consonant cluster

If you're using 4 for final consonant clusters, then you could replace 2 with 1+4.

Actually I think I'd try a system similar to Hangul where each syllable-symbol is constructed of smaller symbols representing the individual sounds. Of course, it would have to be tailored to the different peculiarities of English.
Paul   Monday, June 14, 2004, 07:11 GMT
Ok, that makes sense.
I would consider splitting the set of all English syllables into:

1, V i.e. set of all standalone vowel syllables
1+4 VC i.e. set of all syllables made up of a vowel then consonant cluster
3 CV i.e. all consonant clusters followed by a vowel
3+4 CVC i.e. consonant cluster then vowel then consonant cluster

So if we created such a system how many symbols would be needed?
Any estimates?
I'd say 22 for Categorey 1.
Jim   Monday, June 14, 2004, 08:01 GMT
Maybe 22 would do for category 1. Give or take a few ... probably give more than take. Which 22 do you have in mind?

I think you could also use a symbol (or two) to indicate voice. This way you could reduce the number of symbols needed. It would also have spin-off when it comes to words like "cats", "dogs", "sacked" and "sagged".
Paul   Monday, June 14, 2004, 13:50 GMT
My understanding is that consonants in a Consonant Clusters in English with the exception of Liquids must be entirely voiced or unvoiced.
That is to say you can have str and zdr, and sp and zb, and ks and gz, and ts and dz (all consistent), but never sdr, sb, gs,tz or ds.

This greatly reduces the number of valid combinations in a cluster
Jim   Tuesday, June 15, 2004, 02:34 GMT
Also with the exception of semivowels and nasals, e.g. "twig", "snuff", "smut", "few".
Paul   Tuesday, June 15, 2004, 05:52 GMT
This increases the number of valid combinations in a cluster, but I would guess that there are still less than 300 syllabic components, that are even used.

Regards, Paul V.
Jim   Thursday, June 17, 2004, 06:29 GMT
I get the feeling that there are probably more than 300 syllables.


Sri Lanka
Jim   Thursday, June 17, 2004, 07:50 GMT
I tried counting them yesterday and it seems that there are something between fifty and one hundred valid initial clusters. I think that it would be difficult to put an exact number on it though because it would depend on your accent.

My above list includes place names, words from other languages and a technical term with some initial clusters that you don't tend to find in home-grown English words. Still, if you're aiming at designing a syllabic orthography for English, you'd want to include such things.

An interesting thing that I found when counting initial consonant clusters yesterday is that most of them can be constructed (at least in my accent) like this:

1st consonant: /s/ or nothing
2nd consonant: stop, fricative, affricative, /n/, /m/ or nothing
3rd consonant: semivowel, liquid or nothing

Note: I'm considering /tS/ and /dZ/ as a single consonant rather than a stop followed by a fricative. I say "most of them": note how /ts/, /Sn/ and /Sm/ are missing. Of course, you could include these like this:

1st consonant: /s/, /S/, /t/ or nothing
2nd consonant: stop, fricative, affricative, /n/, /m/ or nothing
3rd consonant: semivowel, liquid or nothing

But this might give you the impression that there are twice as many valid initial clusters. I think I'd want to go the other way ... at least for the purposes of counting them. Specifically, I'd want to divide the set of all initial clusters into triple consonants and other clusters. All triple consonants fit the pattern:

1st consonant: /s/
2nd consonant: unvoiced stop
3rd consonant: semivowel or liquid (excluding /W/, i.e. "wh")

The others fit one of these patterns:

1st consonant: /s/
2nd consonant: unvoiced stop, /f/, /tS/, /n/, /m/, semivowel, liquid

1st consonant: /S/
2nd consonant: /n/, /m/, /w/, liquid

1st consonant: stop, fricative, affricative, /n/, /m/ or nothing
2nd consonant: semivowel, liquid or nothing

1st consonant: /t/
2nd consonant: /s/

Of course I could take it further but not right now. Note how I include /sf/ as in "sphere". I get the feeling that it's probably a /sf/ rather than a /sv/ in "Svalbard" ... I don't know Norwegian but if it's like German this would be a good guess.

Note that I've included /stS/ (which would be a triple cluster if you count /tS/ as two consonants). I've also included /stj/. Now something worth pointing out is that, as far as I can make out, they're just two different ways of pronouncing the same thing.

That is, for example, /stSu:p..d/ and /stju:p..d/ are just two different ways of pronouncing "stupid" (some also say /stu:p..d/). So perhaps, from an orthographic point of view, we could ignore /stS/.

In fact we could ignore both (and about a dozen more) if we consider the /j/ as part of the vowel. It seems that for every consonant cluster containing /j/ except where /j/ is the only consonant the following vowel is either /u:/ or /u../ (/o:r/ in some accents but it's equivalent to /u../: they're both spelt "ure"). All it seems we need to do is count /ju:/ and /ju../ as vowels (traditional orthography does this anyway) and we can simplify things quite a bit.

P.S. "Tsar" is also spelt "czar" but it's still pronounced the same.