Zompist Bboard Again

Consider this a companion thread to the "One-syllable words with specific technical or rare meanings" thread from the old board.

For some reason, I find it intriguing that the Japanese word for "picture", 絵, consists of a single phoneme and phone, e. The shortest equivalents I know of in other languages are at least three phonemes (e.g. Welsh llun, Korean 화) and English picture consists of at least five.

So, with that in mind, I'm interested in seeing what some of the shortest words are for particular concepts. Yeah, we could argue endlessly about whether two lexical items from vastly different language are really equivalent or not, but let's don't. Same goes for phonology; I'll accept either phonemes or phones. For verbs, we could do stems as well as shortest attested finite forms.

A couple more to get started:
"water": French eau /o/, Acehnese i
"luck": Irish ádh /aː/
"tooth": Korean 이 /i/[*]
"hand": Guarani po
"forget": Korean 잊 /ic/ [SAFF: 잊어 /ica/], Thai ลืม /lʉʉm/

[*] 니 as a second element, so could be phonemicised as /ni/ with loss of initial, though this is tough to square with recent borrowings.

Lots of Mandarin one-phoneme examples, though I suppose tone should be counted as well. (In this context I suppose yì 譯 translate and yì 意 idea, meaning might be my favourite examples.)

Yes, counting gets awkward. Does Thai อา /ʔāː/ 'father's younger sibling' (e.g. 'aunt') count as one phoneme (a long vowel) or four - vowel quality, automatic glottal stop, tone and chroneme?

English eye /ai/ is one phoneme - but the same sound in Thai could be counted as five!

Richard W wrote: ↑Fri Jan 25, 2019 12:08 pmYes, counting gets awkward. Does Thai อา /ʔāː/ 'father's younger sibling' (e.g. 'aunt') count as one phoneme (a long vowel) or four - vowel quality, automatic glottal stop, tone and chroneme?

This is for shits and giggles. Police yourselves. (I do think you have to count tone at least unless there's an "unmarked" default.)

Richard W wrote: ↑Fri Jan 25, 2019 12:08 pmEnglish eye /ai/ is one phoneme - but the same sound in Thai could be counted as five!

Well, Luxembourgish has A so it's moot!

Japanese u "cormorant"
Irish aodh "fire".. used as a name element, not sure of pronunciation because wiktionary lists both /e:/ and /i:/
Swedish & Danish have å "river" and Swedish also has ö "island".

Loads of examples in Mandarin, as pointed out above....., but most are probably not used in all contexts . For example, I read that all of the names of the elements in Mandarin are monosyllabic. E.g. https://en.m.wiktionary.org/wiki/鍀#Chinese

Pabappa wrote: ↑Fri Jan 25, 2019 12:27 pmIrish aodh "fire".. used as a name element, not sure of pronunciation because wiktionary lists both /e:/ and /i:/

/e:/ is Munster; /i:/ is Connacht and Ulster. (The modern word is tine [ˈtʲɪnʲɪ].)

Pabappa wrote:Loads of examples in Mandarin, as pointed out above....., but most are probably not used in all contexts . For example, I read that all of the names of the elements in Mandarin are monosyllabic. E.g. https://en.m.wiktionary.org/wiki/鍀#Chinese

Because of lexical tone, I think it's difficult to find any truly monophonemic Mandarin examples. ("Toneless" words tend to be function particles.)

Trying to stick to languages I've studied.. Wiktionary is a too-easy treasure trove for this...

French /u/ 'where', /œ̃/ 'one', /ɑ̃/ 'year', also 'of that', /ɔ̃/ 'one, someone, we', /i/ 'to there', /y/ '1s/2s/3s/3p hoot'

Latin ī 'go!'

Sanskrit om ॐ is said to mean the whole of the Vedas, or the sound of creation... something grand, at least.
ṛc 'praise' isn't bad.

We could be here all day with particles and adpositions, but I think Russian deserves points for single-consonant prepositions: k, v, s.

Perhaps worth a mention: in the French sf comics series Le Cycle de Cyann, <O> is "the memory of the Blue Planet; the life symbolized by water; the sacred principle around which the deO organize their rites, sanctuaries, and establishments". It's also a sign of aristocracy: there are sumptuary laws governing how many O's you can have in your name.

Okay, thanks.

I remembered one more: Basque lo "sleep"

In spite of being a triconsonantal language, there's a few cases where masculine singular imperatives of basic ("form I") verbs in Standard Arabic end up being very short (the so-called "assimilated-defective" verbs) because of the morphophonological rules used.

The typical masculine singular imperative has the shape /ʔVCCVC/. If a basic CaCVCa verb has /w/ or /j/ as the first consonantal root, the first syllable (/ʔVC/-) is dropped, and if /j/ is the third consonantal root, the /-VC/ at the end surfaces as a short /i/.

This means that:
- the masc. sg. imperative of /waʕa:/ (root /w-ʕ-j/) is simply /ʕi/ 'pay attention!'
- the masc. sg. imperative of /waqa:/ (root /w-q-j/) is simply /qi/ 'protect!'
- the masc. sg. imperative of /wafa:/ (root /w-f-j/) is simply /fi/ 'carry out [the promise]!'

These are not particularly surprising for their length (French has a larger number of CV imperatives for one), but because they're found in a triconsonantal language.

zompist wrote: ↑Fri Jan 25, 2019 1:25 pm/ɔ̃/ 'one, someone, we', /i/ 'to there', /y/ '1s/2s/3s/3p hoot'

Oh, French...

/õ/ is also ont '3P have'.

/i/ is also a common colloquial pronunciation of il 'he'.

/y/ is also the past participle of "have", avoir eu. (And the historic pasts eus/eus/eut '1S/2S/3S had'.)

/a/ is as/a '1S/2S have'.

/ɛ/ is hais/hais/hait '1S/2S/3S hate', the imperative aie 'have [sth]!' (and the rare preposition ès 'in').

And es/est '2S/3S be' can be either /ɛ/ (more formal) or /e/ (more colloquial).

Makassarese: u' (hair)

Indonesia ironically has shorter word for more advanced concepts than basic concepts, barring grammatical words:

am : Not limited to a certain person or race; universal. Usually used as "gereja am" which means a church that is universal regardless of denomination and time.

Indonesia is usually disyllabic at minimum

Danish ble "diaper"
Possibly Hebrew peg "premature baby".....I remember reading this long ago but its hard to confirm
English bat "bat (animal)"... it seems plain to us, but in much of Europe this animal is often named with a compound such as french chauvesouris or Spanish murciélago...I remember working at a toy store where a Bat Lord toy was translated as "El Señor de los Murciélagos".
What about bee? Can anyone beat 2 phonemes for that, or at least tie?


edit: confirmation of /peg/:
https://translate.google.com/#view=home ... ure%20baby , although it might be /pag/. it wont ler me link directly to iy so i have to press the translation on the other side.

- Æ e i A.
- Å, æ e i A æ å!

http://www.bbc.co.uk/languages/yoursay/ ... gian.shtml

French always comes to mind for me too:

/u/ - “August”, “holly”

Although I guess “août” is not really a true example because it is pronounced /ut/ in some environments. But as far as I know, “houx” is just /u/.

Finnish is not going to be faring well in this race most of the time, but an arguably two-phoneme pii for 'silicon' could be tough to beat.

Tropylium wrote: ↑Sat Jan 26, 2019 12:02 pm Finnish is not going to be faring well in this race most of the time, but an arguably two-phoneme pii for 'silicon' could be tough to beat.

You count it wrong. There is 3 phoneme there. /p/, /i/, and /:/ (the chromoneme is counted as separate phoneme, not combined with the vowel as in English)

Chroneme as a distinct phoneme is not a universal analysis; chroneme as a segment even less (similar to the Thai issue above). Even if you do subscribe to that, monosyllabic content words do not have a contrast between [V] and [Vː], and they could be analyzed as /CV/ with length being predictable…

Complications aside, even three phonemes for a "non-classical" element still seems like a good score. (Maybe less so for long-known metals & the like; French & Catalan or 'gold'; English tin, lead; Hungarian vas 'iron'; Swedish kol 'carbon'…)

Linguoboy wrote: ↑Fri Jan 25, 2019 12:18 pm

Richard W wrote: ↑Fri Jan 25, 2019 12:08 pmYes, counting gets awkward. Does Thai อา /ʔāː/ 'father's younger sibling' (e.g. 'aunt') count as one phoneme (a long vowel) or four - vowel quality, automatic glottal stop, tone and chroneme?

This is for shits and giggles. Police yourselves. (I do think you have to count tone at least unless there's an "unmarked" default.)

Aha! This word has the common tone (so unmarked?), default initial, default length (*) and the default vowel quality. Does this then count as 0 phonemes?

*Several transcription schemes take the default length as long and mark short vowels as short. I did consider mentioning this on the 'natlangs as conlangs' thread. When the Thai orthography needs an extra mark to indicate vowel length, it is the short vowel that gets the extra mark. When there are signs that incorporate the distinction, the sign for the long vowel is more complex, but these pairs hark back to Indic languages, where it makes sense to treat shortness as the unmarked length.

What does “common tone” mean?

While it’s often the case that distinctions which are unmarked phonologically are also unmarked in the orthography, this can’t simply be taken for granted (even when dealing with a writing system that *isn’t* heavily influenced by being adapted for a completely unrelated language); it has to be demonstrated.

Interestingly the plural of "egg" in Cajun French, /ø/, is shorter than the singular, /œf/. In Munster Irish, they're the same length: /i:/ vs /ov/.

Vietnamese has ô [ʔo˧˧] for umbrella. Bonus points for it looking like an umbrella over someone's head, in some fonts anyway.