Hypertonal conlang
« previous post | next post »
Today's SMBC:
The mouse-over title: "Constructing the 30,000-tone language for Anne was more challenging."
The AfterComic:
The phrase "with every word consisting of two identical syllables" is misleading,as the tone-marking diacritics on the strip's transcriptions indicate, since the tone markings are different across the two syllables in e.g.
Chérŷl Chе̄rȳl
So the number of possible words in this language is not 175, but rather 175*175=30,625 — which is plenty for Steve's purposes, but probably a bit too low to be realistic.
Here's a graph suggesting that the effective (phonologically distinct) number of possible English words is between 19 bits (524,288) and 20 bits (1,048,576):
(Obviously we'd get a bigger number from a larger list of "words", especially one including lexicalized phrases. For a better explanation, see the presentation that it came from, or the brief discussion here…)
Update — See "Ambisyllabicity", 12/14/2024.
Jonathan Smith said,
December 8, 2024 @ 3:48 pm
Comic did the math right thus the mouseover title, but maybe should have written nerdviewese "the same two segmental syllables" in the first panel.
It would seem there is a kind of Poe's law at work in the world of constructed languages/scripts such that it's tough to create this kind of humor… more disturbingly and as displayed often enough on LL, similar fantasism pops up all the time in self-styled serious linguistics such that the con lang community winds up telling proto-Worlders or whomever "hang on now that's pretty implausible" :/ :/
Julian said,
December 8, 2024 @ 4:38 pm
The linked paper is well above my pay grade, but the image of the serious scientists standing beside the duck pond with a stop watch, a clipboard and a bag of bread is fun.
Lars said,
December 8, 2024 @ 7:48 pm
How do they reckon the syllables of "Cheryl" are identical (if indeed you count two syllables)?
Coby said,
December 8, 2024 @ 7:57 pm
I agree with Lars, it should be "identical two" not "two identical".
Yves Rehbein said,
December 9, 2024 @ 12:10 am
I recall that there were only two Proto-Afroasiatic roots in the Wiktionary, *x̣al and *dam . Now the category is up to 42 but still controversial. Not knowing what an epipharyngeal is I would frequently think about *XəR for a while as a Proto-World cover-all, conveniently shortened to *X. That's Cheryl modulo 175 tones, but it is also surreal.
Compare French sur, Spanish sobra, from L. super "over" with intervocalic voicing and deletion, but see Egyptian ḥr, compare Arabic ʕalā, “on, over”, from PAA **x̣al?
Now that's pretty implausible!
Chris Button said,
December 9, 2024 @ 9:35 am
It's interesting that E, R and Y are all getting diacritics. Nothing on L though.
Haha. I've never heard it pronounced as one. Where is the syllable break though:
– cher.yl
– che.ryl
I vote for the former in natural speech. Although I bet everyone uses the latter when sounding it out carefully to themselves.
Could you share some examples?
My gripe is more with reputable academic publishers that sometimes let shoddy linguistics go through since it is outside of their area of expertise
The real ability of many conlangers at linguistics should come as no surprise. David Peterson has even made a nice living out of it.
Philip Taylor said,
December 9, 2024 @ 9:58 am
"[…] where is the syllable break, though […]". The LPD votes for your first variant, but I find its initial consonant group very odd : /ˈtʃer əl/. For me, it is most definitely /ˈʃer əl/, although I would instinctively use your second syllabification.
Mark Metcalf said,
December 9, 2024 @ 12:34 pm
Reminiscent of Groot's language/vocabulary (i.e. "I am Groot") in the Guardians of the Galaxy movies.
Here are examples from the first film: https://www.youtube.com/watch?v=f_bJZGbSI1s
A bit more background: https://www.digitalspy.com/movies/a32301482/groot-language-james-gunn/
David Marjanović said,
December 9, 2024 @ 3:10 pm
A mistake for "epiglottal"?
Andrew Usher said,
December 9, 2024 @ 7:43 pm
Chris Button wrote:[blockquote]
Where is the syllable break though:
– cher.yl
– che.ryl
I vote for the former in natural speech. Although I bet everyone uses the latter when sounding it out carefully to themselves.[/blockquote]
If you're non-rhotic, you have no choice! But Americans definitely don't, and couldn't, as that vowel just doesn't exist without a following R. It's share – l, which could indeed be compressed to one syllable in fast speech exactly like girl, snarl, etc.\
Chris Button said,
December 9, 2024 @ 8:57 pm
@ Andrew Usher
Non-rhotic accents keep a rhotic intervocalically. And even add them sometimes–"drawing" as "drawring" is a classic example.
Che.ryl is like ha.ppy
Cher.yl is like happ.y
The second examples better reflect how we actually speak. But many analyses break them to give CV.CV according to the first examples instead, which is indeed how people often break syllables when sounding them out individually in isolation.
Jonathan Smith said,
December 9, 2024 @ 9:52 pm
In USofA it seems that standardly (?), preceding vowel quality is linked to syllabic l- in such frames, e.g., for me there is ≈/er.ɫ̩/ as in carol and ≈/ɑrl/ as in Carl but no such thing as */erl/ or */ɑr.ɫ̩/. Meaning no contrast would be lost by conceiving all such arrangements (to include Cheryl) as monosyllabic, which I suspect must be the situation in at least some southern/Appalachian Englishes. Example here, though of course this is a phonemic issue, not a function of speech speed or straightforwardly determinable from phonetic signal.
Re: LL fantasies, commenters tend to raise the alarm on individual threads, so no reason to recap here probably :D
Jonathan Smith said,
December 9, 2024 @ 9:56 pm
*oops, no need to mark "dark l" above, sloppy copy/paste work by me
Lars said,
December 9, 2024 @ 11:37 pm
@Jonathan Smith: Now you've reminded me of that Neil Sedaka song, "Oh Cheryl". Clearly, Cheryl is a descendant of the name "Carol" which has undergone the same changes that happened in early French (ca -> che)
Philip Taylor said,
December 10, 2024 @ 6:41 am
I seem to recall that the Neil Sedaka song was "Oh Carol", Lars. Perhaps someone else recorded "Oh Cheryl".
Chris Button said,
December 10, 2024 @ 7:59 am
@ Andrew Usher
Pondering a little more on this, I'm now wondering how you would instinctively break Cyril?
Jonathan Smith said,
December 10, 2024 @ 8:33 am
As a point of comparison FWIW I see people from "the south" broadly speaking giving eye-dialect "durl" for Darryl…
Jerry Packard said,
December 10, 2024 @ 9:39 am
The Neil Sedaka song was "Oh Carol”, to which Carole King responded by penning “Oh Neil”.
J.W. Brewer said,
December 10, 2024 @ 7:01 pm
"Cheryl" was one of the 50 most popular given names for US-born baby girls for years-of-birth 1944 through 1971, so you'd think it certainly ought to pop up in the lyrics of some songs from the era when women of those generational cohorts (real or fictional) were likely to be mentioned in songs. I can think of a few songs mentioning "Sherry," which can be a nickname for Cheryl as well as a freestanding name, but nothing explicitly mentioning Cheryl is coming to mind, which is annoying me.
As will happen with once-in-vogue given names, especially female ones, when it fell from favor, it fell fairly quickly and inexorably. Out of the top 50 after '71, out of the top 100 after '79, out of the top 250 after '85, out of the top 500 after '91, and out of the top 1000 after '97. How old does Steve in the cartoon look to be? The cartoonist himself was apparently born in '82, when Cheryl was #159 on the charts, so just playing the percentages he may not implausibly have known or even broken up with a Cheryl or two growing up. But Steve looks younger than that.
Andrew Usher said,
December 10, 2024 @ 7:50 pm
Chris Button:
Of course I know non-rhotic accents pronounce the R in 'Cheryl'. That's the point. You said 'when sounding it out carefully' and if you did so as Cher-yl you'd be putting an R sound before a pause, which I imagine you'd want to avoid. I'm not judging which syllabification is correct – both can be, as syllable boundaries don't really exist in speech.
Jonathan Smith:
All those names – Carl, Carol , Cheryl, Cyril – follow the same pattern for me – two syllables, the second syllabic L. My kind of speech just doesn't allow for /rl/ in one phonemic syllable, though I recognise that compression by loss of syllabicity is possible, and normally is confined to higher speeds. The same applies to vowel breaking before R or L as in 'mile' or 'fire', but also 'trial' and 'higher' which have been merged for a long time despite morphology (there's a possible vowel difference, but that's a different point).
I would suppose the compression is more likely when a vowel follows in the next word, as 'higher and higher', which I can easily imagine saying high-r'n-high-er. This is entirely different from British smoothing as a syllable can be lost, but the diphthong (PRICE, MOUTH, or CHOICE) always remains.
You were right to say it was unnecessary to mark dark L, as syllabic L is always dark in English; indeed, I can't imagine a syllabic clear L.
Chris Button said,
December 10, 2024 @ 9:02 pm
@ Andrew Usher
Ah, I see what you mean now. Yes, that makes sense.
So, back to my other question… how would you break Cyril then? I would follow Cheryl to give Cyr.il but I'm curious because of how the pronunciation of "near" in "nearby" is not the same as "near" by itself for many speakers. So, I could see a potential cher.yl vs cy.ril clash for some.
@ Philip Taylor
Yes, that's a new one for me too! But, you're right, it's there as an option in the ever-reliable LPD bible.
Andrew Usher said,
December 11, 2024 @ 9:01 am
As I stated, 'Cyril' would be just like 'Cheryl' for me. I'm aware of what you are saying about the British 'near' vowel; I don't know whether that would cause variation but I'd guess not because neither vowel is diphthongal in this environment.
On the other hand, if the vowel preceding R were a diphthong, I'd have to go the other way, and I think everyone would agree with By-ron : 'by' is a legal syllable on its own, while 'byr' (as just mentioned) is not for me. This is really the same phenomenon, the syllable break falling between two approximants, but here they are the glide of the diphthong then /r/, in the other they are /rl/.
Chris Button said,
December 11, 2024 @ 10:19 am
Ah yes, Byron works great!
In natural speech, I would break it following LPD principles as Byr.on to be consistent with Cher.yl, but that is where the confusion comes in since deliberate pronunciation would give by.ron.
That is what I was trying to get at with Cyr.il for rhotic speakers by analogy with nearby [nɪr.baɪ], where the first syllable rhymes with Cyr, but near on its own doesn't perfectly rhyme as [nɪᵊr] instead.
Philip Taylor said,
December 11, 2024 @ 10:42 am
Andrew — "'by' is a legal syllable on its own, while 'byr' (as just mentioned) is not for me" — is "byre" (a cow-shed) not in your idiolect ?
J.W. Brewer said,
December 11, 2024 @ 10:53 am
First syllable of Cyril is (for rhotic AmEng me) homophonous w/ "sear,"* and standalone "sear" is a perfect rhyme (for me) with standalone "near."
First syllable of Byr.on would be homophonous with "byre" ("A barn, especially one used for keeping cattle in"), which rhymes with "fire" and like "fire" isn't perfectly monosyllabic in my idiolect because of something about the diphthong.
*I note that wiktionary gives "sear" as /sɪɚ/ in AmEng but /sɪə(ɹ)/ in BrEng.
J.W. Brewer said,
December 11, 2024 @ 11:09 am
Contrast cyr.il with su.rreal, i.e. /səˈɹiː(.ə)l/. I wonder if the stress placement has some effect on which syllable the "r" seems attached to? (If you think it somehow sticks to both adjoining syllables in "surreal" then try my best guess as to the local pronunciation of Mount Sarrail in Alberta instead, which I'm assuming is completely Anglicized and does not attempt to mimic the way the French WW1 general it is named for would have pronounced his own surname in French.)
Benjamin E. Orsatti said,
December 11, 2024 @ 12:47 pm
In Pittsburgh, we say /ʃɹ̠w/ (bcs w ht vwls), making the /ɹ̠/ do the work of a vowel, which is why it has never occurred to me that the word could possibly exceed one syllable in length.
Philip Taylor said,
December 11, 2024 @ 1:34 pm
Andrew has "Carl" as bisyllabic, Benjamin has "Cheryl" as monosyllabic — I can only assume that they live as far apart as it is possible to be in the United States of America !
J.W. Brewer said,
December 11, 2024 @ 1:55 pm
@Philip T.: For partial insight into Benjamin O., you may want to go to this article and use control-F to find the discussions of monophthongization: https://en.wikipedia.org/wiki/Western_Pennsylvania_English. Pittsburgh is the place where "iron" is stereotypically a monosyllable that rhymes with "barn."
Benjamin E. Orsatti said,
December 11, 2024 @ 2:49 pm
J.W. Brewer,
Ah, who needs all those extra sounds? You can tell the difference between the Greek letter, "tau", a bathroom "towel", and a "tile" floor from context, right?
But it gets worse: the further down the Appalachians you go, the more diphthongs you lose. In fact, somewhere just south of Morgantown, W.Va., "oil" turns into the Swedish word for creek (crick?), /å/.
Philip Taylor said,
December 11, 2024 @ 4:56 pm
« You can tell the difference between the Greek letter, "tau", a bathroom "towel", and a "tile" floor from context, right ? » — That really threw me. Is the Greek letter "tau" not pronounced (approximately) /tæf/ ? That is certainly the only way that I have ever heard my Greek colleagues pronounce it.
Andrew Usher said,
December 11, 2024 @ 9:06 pm
Any pronunciation of 'tau' with /f/ is from Modern Greek, and not historically authentic. Americans all rhyme it with 'vow' I think, which is the closest to the ancient Greek pronunciation. Traditional English would be 'taw', as spelled.
As for the difference between Benjamin's speech and mine, I don't think it's a matter of geographical distance so much as that his pronunciation is definitely regional, as he admits, while mine is non-regional.
Philip Taylor said,
December 12, 2024 @ 6:33 am
OK, so if your pronunciation is non-regional, Andrew (and therefore presumably representative of, or at least similar to, the pronunciation of the majority of Americans), is it therefore the case that "Carl" is bisyllabic for the majority of Americans ? If so, I am surprised. What about "furl", "marl", "snarl" and "whorl" ?
J.W. Brewer said,
December 12, 2024 @ 8:44 am
@Philip Taylor: Those are all words that, at a minimum, reveal the semi-incoherence of the notion of "syllable" in edge cases, such as those presented by /ɹ/ and /l/ sequenced as if a cluster. The word-final /l/ tends to force at least a little bit of syllabic color onto the immediately-preceding approximant, presumably for physiological reasons, because a word-final stop does not have the same effect. One might say that words ending with that consecutive-approximant sequence come out as a syllable-and-a-half. Maybe we need to think in terms of morae rather than syllables? The easiest way, and perhaps the only way, to make such words unambiguously monosyllabic is to be non-rhotic. But that seems like cheating.
J.W. Brewer said,
December 12, 2024 @ 8:49 am
For "Carl" in particular, one might note that female given names derived from Carl like Carleen or (via some intermediate steps) Charlotte allocate the consecutive approximants to different syllables: Car.leen and Char.lotte. This imho tends to confirm the notion that trying to get them into the "same syllable" is inherently awkward and we thus subconsciously sidestep the awkwardness when it is possible to do so.
Philip Taylor said,
December 12, 2024 @ 8:50 am
"The easiest way, and perhaps the only way, to make such words unambiguously monosyllabic is to be non-rhotic." — Ah, I think you have just identified the reason for my surprise. My topolect is non-rhotic, and I did not consider (a major error on my part) the possibility that Andrew (and other Americans) might sound the "r" in those words …
Benjamin E. Orsatti said,
December 12, 2024 @ 9:58 am
Philip,
Aye, there's the rub. We have a superpowered "r" that behaves extremely aggressively towards its neighboring consonants, for example, often bludgeoning poor /l/ into /ɫ/ before finally pummeling it into /w/.
Chris Button said,
December 12, 2024 @ 10:15 am
@ J.W. Brewer
I think so. It has far more explanatory force and can handle concepts of fortis/lenis and tense/lax.
Jonathan Smith said,
December 12, 2024 @ 10:36 am
I mean… no, -rl (two liquids [ɹ̠ɫ] or however you care to represent it) is a legit coda cluster in Gen. Am. and is basically a survival from Old English. Commenters above may be (1) actually innovative in some way I'm not familiar with or (2) overthinking. Re: the latter, phonetics / "sonority hierarchy" don't (have to) matter — stuff can just be one syllable (cf. obstruent-liquid codas in Icelandic, French…) So to most (?) Americans and in all (?) dictionaries, just as tin and splint are equally one syllable, nor and snarl are equally one syllable.
J.W. Brewer said,
December 12, 2024 @ 10:55 am
@Jonathan Smith: You can certainly make up (and justify/rationalize on various methodological grounds) a technical definition of "syllable" which unambiguously covers things ending with that two-liquid coda. My perspective is that when you actually say those words out loud they feel notably different in the mouth (longer and more internally segmented) than more prototypical monosyllables ending with a coda of r+stop. Compare "Carl" to "cart," or "snarl" to, um, "snart" (rhymes with start). There's obviously a continuum; many AmEng speakers would say that "girl" and "squirrel" are not quite perfect rhymes, which fits with my one-and-a-half-syllable explanation of words like "girl."
Philip Taylor said,
December 12, 2024 @ 11:11 am
Just as a point of interest, how many syllables would a rhotic American speaker identify in the Czech word Jiří (when correctly pronounced, of course) ? For me (non-rhotic Briton) it is just two [ˈjɪr̝·iː], but I wonder whether a rhotic American might find three [ˈjɪ·r̝·iː].
Jonathan Smith said,
December 12, 2024 @ 12:57 pm
@ J.W. Brewer: in general syllable count judgements (by reference to say verse or 4-year-olds clapping) are extremely consistent across speakers, esp. when you remove dialect variation. But OK there seems to something to your intuition: this paper finds that a minority of Am. Eng. speakers, when such a choice is offered, will give "1.5 syllable" judgments for certain liquid-final words, so e.g. for feel male, and pool, 2 of 18 (always the same 2?) choose 1.5 syllables rather than 1 (see Table 4).
Re: theoretical implications / authors' interpretations, IDK. But it would make sense if similar were true for -rl. Not sure about -rt. And in general "feels longer" isn't informative wrt consistency of syllable count judgments across speakers, status of syllable as some kind of computational primitive, etc.
Chris Button said,
December 12, 2024 @ 1:18 pm
So let's set the world on fire
We can burn brighter
Than the suuuuuun…
Benjamin E. Orsatti said,
December 12, 2024 @ 3:36 pm
Philip Taylor said:
2 syllables: (1) Yih (short "i"); (2) ree (long "i"). Source: Jiří Hrdina, superstar center for the back-to-back Stanley Cup Champion 1990-90 & 1991-92 Pittsburgh Penguins.
EVERY Pittsburgher my age (46) and older knows how to pronounce Jiří; it flows off the tongue almost as smoothly as "Lemieux".
Philip Taylor said,
December 13, 2024 @ 5:42 am
But "Yih-ree" is nothing like the correct pronunciation of "the Czech word Jiří", Benjamin. It may well be how Jiří Hrdina anglicised it for the benefit of his fellow team players (and match impires !), but the first syllable fails to reflect the "ř" in any way. If I wanted to "eye-spell" Jiří, I would go for something along the lines of (rhotic) "Yearzhee".
Confession : long before I came to learn Czech, Polish and Slovak, I had a colleague called Jiří Studnicka (this was at Westfield College, University of London). And to my later enormous embarrassment, we all referred to him as (eye-spelling) Jiree Studknicker (for non-rhotic "knicker"). When (many years later) I finally learned how we should have been pronouncing his name, I tracked him down and telephoned him, addressing him correctly in Czech and then apoologising for my earlier mis-pronunciation of his name. To say he was surprised would be quite an understatement.
Andrew Usher said,
December 13, 2024 @ 9:45 pm
I must disagree. English doesn't have the Czach /ř/, and it could be argued (and seems to my ears) that an ordinary English R is the closest approximation in English phonology. As for the first syllable being 'Yih' or 'Year', for those that make a difference between them, the former is more appropriate as the Czech vowel is short. Could you transcribe how you now say "Jiří Studnicka"?
I don't really trust syllable count judgements of that sort. It's well-known that on linguistic matters, people tend to be influenced by what they've consciously learned over what they should be able to hear – here, the spelling and the conventional syllabification that goes along with it. Most (American and British) speakers don't distinguish 'real' and 'reel', and all rhyme 'foul' and 'towel', but many will deny that they do so on examination, and if asked to pronounce them may do so in an abnormal way to support their belief (and I assume that is normally not conscious).
Jonathan Smith said,
December 13, 2024 @ 10:58 pm
re: fire / brighter, fire is a two-syllable word for many Am. Eng. speakers (like me and this genius lyricist it seems.) Rhymes with higher. This is reflected at Table 4 in that paper too (meaning words like this don't work so great in the study.)
re: foul / towel, nope, towel and owl have two syllables period full-stop point-blank for some (like me). So no they most assuredly do not rhyme with monosyllabic foul and howl. The issue is, ironically just as Andrew Usher notes, projecting one's own modelings onto the speech of others (e.g., me thinking my 2nd grade teacher was simply an idiot for insisting that owl only had one syllable.)
Whatever the psychological reality, people deeply know and regularly manipulate syllables. Like… Ubbi dubbi for instance. ANMD HANG ON… at 1:11 this person says "MUBaiuble" for "mile"!! This would not occur to me in a million years; of course "mile" has one syllable and it should be "MUBail". In fact if not for this thread I could easily have not even understand WTF this word was… TIL
Andrew Usher said,
December 13, 2024 @ 11:58 pm
If you really have two syllables for 'owl' but one for 'howl', this is an idiosyncrasy. But I wouldn't believe it until hearing them from you in running speech. The thing is not projecting one's mental model onto the speech of others, but first projecting it onto one's own (perceived) speech, and secondarily to others believed to have 'the same' dialect. This is what your 2nd-grade teacher undoubtedly did, and what you might well be doing as well.
And I'm rather sure 'mile' is also not monosyllabic, being in the same category. It is necessary to be able to hear your own speech – with the help of recording if needed – as it really is, to make any useful contribution to discussions involving phonology. Though the study did, I don't use the category '1 1/2 syllables' – that can only mean '1 or 2, I'm not sure' or '2 with one very short for a syllable' (the latter actually being true here).
Note that I don't always rely on my own judgement in these matters of determining what is normal English. Sometimes I use Youglish to get a spectrum of real-speech examples; yes, they are biased toward educated speakers and formal settings, but that's exactly what you want to hear non-regional, standard English. For 'mile' I just listened to the first 50 examples (American and British) and only 3 were possibly monosyllabic: one case of markedly rapid speech and two of liaison onto a following vowel – both phenomena I mentioned, and occurring in my own speech also. For 'Carl' in rhotic accents the ratio was not as strong but still definitely in favor of two syllables, with monosyllabic examples largely in fast speech.
Philip Taylor said,
December 14, 2024 @ 12:05 pm
Andrew — "English doesn't have the Czach /ř/, and it could be argued (and seems to my ears) that an ordinary English R is the closest approximation in English phonology" — perhaps, but my question was not about how a native English (or American) speaker would/should pronounce "Jiří" : it was about "how many syllables would a rhotic American speaker identify in the Czech word Jiří </when correctly pronounced". The correct pronunciation of Jiří is the pronunciation that most native Czechs would afford it (there are some native speakers of Czech who cannot pronounce the "ř" but they are very much in the minority, although a former Prime Minister was one). As to « how [I would] now say "Jiří Studnička" », my familiarity with the IPA conventions for Czech is not especially great (far weaker than my familiarity with the IPA conventions for British English, for example) but as a best approximation I would hazard [ˈjir̝iː ˈstudɲɪtʃka] ("'yearzhee 'stoodneatshkuh") — I disagree with your assertion that "the [first] Czech vowel is short".
Jonathan Smith said,
December 14, 2024 @ 1:06 pm
Andrew Usher — you remain confused. One cannot reliably "hear" this sort of difference in one's own speech or that of others. It is phonemic. Some (most? IDK) Am. Eng. speakers model owl as having one syllable, whereas others have two — see e.g. at Wiktionary here. This and many other cases are well-known. Turns out same is true (has become true?) for mile and similar, as shown above — this is not reflected in dictionaries (yet) AFAIK.
Andrew Usher said,
December 14, 2024 @ 10:51 pm
I think it is you that is confused. I assume you aren't talking about the physical ability to hear phonetic differences but the difference between phonetic and phonemic representations. While this is a valid concept, phones and phonemes should have a predictable correspondence (by definition) and postulating that individual speakers can have different phonemic representations that happen to lead to the same phonetic output, while not actually wrong, is generally not useful. We do not have access to a speaker's mental phonemic representation, and asking them for it is not reliable as speech is a subconscious process. Actual linguistic research avoids this method when possible.
Philip Taylor:
I see, but 'Jiří' has two syllables and I can't imagine any reasonable pronunciation that would elicit a different count. I suppose some botched attempts at the ř might sound like another syllable, but can't imagine any right now.