Or

[Ser]

I have a couple questions about "or".

Do you guys know if there are languages without a nice, short, one morpheme-long equivalent of non-interrogative "or" (as in "I am always [either] here or there")?

Also, is there a term for "or" questions, such as "Do you want X or Y?" These don't quite seem like polar questions to me, since in some languages like Spanish you can't answer (with the equivalent of) "yes". "Selection questions"? "Choice questions"?

[linguoboy]

Do you guys know if there are languages without a nice, short, one morpheme-long equivalent of non-interrogative "or" (as in "I am always [either] here or there")?

IIRC, Osage doesn't have such a word, but I'd need to double check my sources to be sure.

Hawai'ian equivalents are morphologically complex, i.e.: a i ʻole "and if not", ā VP paha (where ā is a variant of a and paha is a conditional marker].

ETA: Osage makes use of ímąðe "one or the other", which may be morphologically complex (cf. ímąhtaha "the other way").

E.g. ímąðe hǫbrį́ke watkxą́ hpáaɣe hta mįkšé
one-or-the-other beans squash make POT AUX.1S.sitting/lying
"I'll fix either beans or squash"

[Valdeut]

I have a couple questions about "or".

Do you guys know if there are languages without a nice, short, one morpheme-long equivalent of non-interrogative "or" (as in "I am always [either] here or there")?

Yes.

The following is from Martin Haspelmath (Ed.) "Coordinating Constructions", from the initial overview chapter written by the editor himself. Note that the term disjunction refers to or-constructions, while conjunction refers to and-constructions (coordination includes both):

Ohori° (§ 3.2) observes that ‘and’ words are much more frequent in discourse than ‘or’ words. Thus, we also expect that ‘or’ words are typically longer (and rarely shorter) than ‘and’ words.

[...]

Ohori° (§3.2) also notes that disjunctive coordinators tend to be more autonomous (i.e. they are more often free words rather than clitics or affixes), and that disjunction less often differentiates between NP coordinands and clause coordinands. All this can probably be explained as due to the lower frequency of disjunction markers.

In some languages, there does not seem to be any grammaticalized way of expressing disjunction at all. Ohori° (§3.1) describes cases of neutralization between ‘and’ and ‘or’. According to Kibrik° (§4), who describes coordination in Upper Kuskokwim Athabaskan, “there does not seem to exist any native way to express disjunction” in this language, only the English borrowing o is sometimes used. In general, disjunction markers are fairly easily borrowed (see Matras 1998 for the borrowing hierarchy ‘but’ > ‘or’ > ‘and’); another example is Dargi ya... ya..., borrowed ultimately from Persian.

In Lai, there is a native conventional way to express disjunction, but this is rather complicated and seems to be quite young. An example is (44).

(44) Hakha Lai (Peterson & VanBik°, ex. 12a)
làwthlawpaa=ni‘ vok ‘a-sií-làw=leè ‘aàr ‘a-tsook
farmer=erg pig 3sg.subj-be2-neg=cond chicken 3sg.subj-buy2
‘The farmer bought a pig or a chicken.’

Originally, the word for ‘or’ (‘asiílàwleè) was analyzable as ‘a-sií-làw=leè [3sg.subj-be2-neg=cond] and the literal translation of (44) is ‘The farmer, if it wasn’t a pig, bought a chicken.’ This is not a possible synchronic analysis because =leè is not used as a conditional marker anymore, but it is clear that the disjunctive marking strategy must be of very recent origin.

———

Also, is there a term for "or" questions, such as "Do you want X or Y?" These don't quite seem like polar questions to me, since in some languages like Spanish you can't answer (with the equivalent of) "yes". "Selection questions"? "Choice questions"?

I've seen them referred to as "alternative questions". Maybe there are other terms as well.

[zompist]

I'm not sure if Quechua counts for you or not. To express a disjunction "X or Y" you say X-chu Y-chu. Thing is, -chu isn't a dedicated word for 'or'; it's the question morpheme. It's like saying 'The farmer bought a pig? a chicken?".

[clawgrip]

Japanese is exactly the same, using the question particle ka, except you can drop the last one in the sequence, i.e. X ka Y (ka).

[Astraios]

Lakota uses naíŋš, which is the conjunction na “and”, plus the adversative pronoun íŋš, so that together naíŋš means “and on the other hand”, used in both interrogatives and non-.

Lé naíŋš hé opȟétȟuŋ.
“He bought [either] this or that.”


Russian uses the word i “and” plus the interrogative particle li “whether, if”: On kupíl íli to, íli drugóje.


Hungarian uses vagy, which is shortened from az vagyon “that is”: Vagy ezt vette meg, vagy azt.


And although Turkish, Kurdish, and Persian do have a dedicated monomorphemic “or”-word, they also tend (to varying degrees) to compound it with conjunctions:

Turk: Ya bunu veya şunu aldı., where veya is Persian “and-or” (an alternative is ya da, from Persian “or” followed by Turkish “and”).
Kurd: Wî an ev an jî ew kirriye., where an is “or”, and jî is an optional (but commonly added) adversative or additive enclitic.
Pers: Yâ in vayâ ân râ xaride ast., where yâ is “or”, and va “and” is optional (but less common).

[vokzhen]

Naxi uses A mɔ33 ni31 B, which appears to be derived from mə33-ni33 NEG-COP, which is available for nouns only. Nouns can also use A lɑ33 B (the conjunctive coordinator), making conjunction and disjunction identical. Verb disjunctive coordination uses A nɔ33 B, and clausal uses A B nɔ33. In addition, two negative clauses juxtaposed can have either a conjunctive or disjunctive meaning.

Nuosu has the complex ap nge ox go NEG COP PERF COMP, which the author translates as "(if) it is not the case that." It appears to only be available for coordinating clauses. There is also a word ddap that the author glosses "or," but it only exists in A ddap NOT-A questions, e.g. ne la ddap ap-la [2S come or NEG-come] "will you come?", and as a sentence-final interrogative particle. The only example of a sentence translated with "or" and linking nouns is a negative verb "not eat x, y, or z," which doesn't take *any* kind of coordinator as lists don't take coordinators.

Cho'l has the relational noun -ik'oty- that is an and/with/instrument morpheme (and takes agreement even when linking clauses!), but no disjunctive. It borrows a bunch of coordinators from Spanish. The other Mayan languages I've looked at appear to be similar, no native disjunctives; the "learn Yucatec"-type grammar I looked at gave an example of a "disjunctive" that appears to just be two question sentences said in sequence, "Is there X? Is there Y?", to get the disjunctive meaning.

[hwhatting]

Russian uses the word i “and” plus the interrogative particle li “whether, if”: On kupíl íli to, íli drugóje.

But that is really only etymology - on a synchronous level, it's a one-morpheme conjunction.

[finlay]

Japanese is exactly the same, using the question particle ka, except you can drop the last one in the sequence, i.e. X ka Y (ka).

That, or like German, it uses the word "or" for questions (how do we know its basic form is the question particle?)

[Vijay]

I have a couple questions about "or".

Do you guys know if there are languages without a nice, short, one morpheme-long equivalent of non-interrogative "or" (as in "I am always [either] here or there")?

Malayalam doesn't have one. You say things like "I am always here if not there."

[Ser]

<Whimemsz> Ojibwe has a number of different words that mostly are some combination of maa (a discourse particle) and ge~gaye (a general conjunction), with the basic meaning "maybe/perhaps". Gemaa or maage being the most common, but then there's also just gaye on its own, or gemaa gaye. Also possible are other terms meaning "maybe" like goni and combos like gonimaa

[Imralu]

Someone who speaks Korean may give better imput but I was taught the phrase "Don't speak Korean. Speak English ... or I'll kill you" (Don't ask!) and the word or phrase for "or" comes out to something like geu-roh-ji-an-eu-myeon which I have been told by several Koreans is how you would say "or" in that context and it means basically "and if not" ...

[linguoboy]

Someone who speaks Korean may give better imput but I was taught the phrase "Don't speak Korean. Speak English ... or I'll kill you" (Don't ask!) and the word or phrase for "or" comes out to something like geu-roh-ji-an-eu-myeon which I have been told by several Koreans is how you would say "or" in that context and it means basically "and if not" ...

그렇지 않으면
/ku.leh-ci anh-umyen/
be.so-ASS NEG-HYP[*]
"if not so..."

You could also translate this as "otherwise".

Etymological note: The verbs 그렇다 /kuleh.ta/ and 않다 /anh.ta/ are both derived by means of the common light verb 하다 /hata/--그렇다 from the medial deictic 그 /ku/ and 않다 from the negative prefix 안 /an/.


[*] The mood glosses are ad-hoc. Martin defines -지 as "the uncertain fact". Declaratively and interrogatively, it's used to assert something the speaker believes to be true, so I've glossed it "assumptive", but its use before the verb 않다 is syntacticly required. -면 essentially translates to "if".

[Imralu]

Ah, thanks for that.

geu-reoh-ji | anh-eu-myeon

그렇지 않으면

I'm not sure what transliteration you use, but it's cool to see that I nearly got it right after all these years. I don't always hear the difference between ㅓand ㅗ, and I forgot one of the syllable final ㅎs.

[linguoboy]

Ah, thanks for that.

그렇지 않으면

I'm not sure what transliteration you use, but it's cool to see that I nearly got it right after all these years. I don't always hear the difference between ㅓand ㅗ, and I forgot one of the syllable final ㅎs.

Well, medial /h/ is silent in voiced environments anyway. But that deletion applies after metathesis with stops, so geu-reoh-ji would actually be pronounced geu-reo-chi (and is sometimes written this way, particularly as a standalone word meaning "yes").

[Imralu]

Ah, thanks for that.

그렇지 않으면

I'm not sure what transliteration you use, but it's cool to see that I nearly got it right after all these years. I don't always hear the difference between ㅓand ㅗ, and I forgot one of the syllable final ㅎs.

Well, medial /h/ is silent in voiced environments anyway. But that deletion applies after metathesis with stops, so geu-reoh-ji would actually be pronounced geu-reo-chi (and is sometimes written this way, particularly as a standalone word meaning "yes").

Yeah, I was proud of myself for remembering that the ch sound there is h-j, and not too disappointed I forgot the silent h. My pronunciation was good enough to fool Koreans into thinking I was fluent. I'm not at all.

[clawgrip]

There are a number of other "or" phrases in Japanese that are longer: mata wa, arui wa, moshiku wa, and soretomo. They mostly all have overlapping meanings that are variously used to contrast opposites or phrases with similar syntactic or semantic phrasing, or in questions.

Identical phrasing:
...higaeri ni iku ka, mata wa nihaku mikka no shūmatsu kōsu de iku ka...
go on a day trip, or go on a 3-day 2-night weekend course

Similar thematic content:
Sanka suru hito wa denwa moshiku wa mēru shite kudasai.
Participants, please call or email us.

Opposites:
kyōryoku, arui wa, kyōsō
cooperation or competition

Dono bubun ga tadashii mata wa machigatte iru ka tte iu koto ga wakareba...
If we can determine which parts are correct and/or incorrect...

Questions:
Sore wa chō desu ka, soretomo ga desu ka?
Is that a butterfly or is it a moth?

[Imralu]

And although Turkish, Kurdish, and Persian do have a dedicated monomorphemic “or”-word, they also tend (to varying degrees) to compound it with conjunctions:

Turk: Ya bunu veya şunu aldı., where veya is Persian “and-or” (an alternative is ya da, from Persian “or” followed by Turkish “and”

In Turkish, ya X ya Y is also common as a kind of "either or". I don't really know how common either is or what the usual usage is but my impression seems to be that veya is used on its own more, without a preceding ya.

There's another word for "or" which is yoksa, which is from yok "absent" (used to form negative existential sentences, and also sometimes informally as a general negator) plus the conditional suffix -s2, so basically it means "if there isn't" or "if not".

Edit:

Oh, and in questions with alternatives, you can simply use two instances of the question particle m4.

Kırmızı mı, mavi mi?
^{red Q | blue Q}
Is it red or blue?

[Whimemsz]

<Whimemsz> Ojibwe has a number of different words that mostly are some combination of maa (a discourse particle) and ge~gaye (a general conjunction), with the basic meaning "maybe/perhaps". Gemaa or maage being the most common, but then there's also just gaye on its own, or gemaa gaye. Also possible are other terms meaning "maybe" like goni and combos like gonimaa

Right, so, some examples:

Ziinzibaakwad ina, gemaa gaye doodooshaaboo giwii-tagonaan gigaapiiming? "Do you put sugar or milk in your coffee?"
ziinzibaakwad ina, gemaa gaye doodooshaaboo gi-wii=dagon-am-n gi-gaapii-m-ing
sugar Y/N, maybe and milk 2-DESID=mix.in-TI-INAN:OBJ 2-coffee-POSSD-LOC (Southwestern Ojibwe - from The Ojibwe People's Dictionary)

Wiindamawishin wegone dowa waa-minozaman maa gaye wiiyaas maa gaye giigoonh "Tell me what you want to cook, meat or fish."
wiindam-aw-ishi-n wegone dowa <IC>wii=minw-z-am-an maa gaye wiiyaas maa gaye giigoonh
tell-TA-1sOBJ-IMPER what type <REL>DESID=well-cook-TI-2s:CONJ maybe meat maybe fish (older Eastern Ojibwe - from Nishnaabemwin Dictionary)

Nga-nsaa na, gnimaa gegoo iw gaawiin geyaabi gegoo ji-zhi-gshki'ewzisig? "Am I to kill him, or perhaps something so that he no longer has any power?"
ni-ga=niS-aa na, gonimaa gegoo iw gaawiin geyaabi gegoo ji=izhi=gashki'ew-izi-si-g
1-FUT=kill-DIR Y/N, perhaps something that:INAN NEG still anything IRREAL=thus=powerful-be:AI-NEG-3:CONJ (Odawa - from Nishnaabemwin Dictionary)

Awiya omaa nizhiinishkaag. Na, odaabin, gemaa niga-zaagijiwebinaa "Someone [a doll] here is crowding me [in the car]. Here, take her, or else I will throw her out."
awiya omaa ni-zhiinishk-aw-igw. na, odaabi-n, gemaa ni-ga=zaagiji-webi-N-aa
someone here 1-crowd-TA-INV. look, take-IMPER, or 1-FUT=out-by.hand-TA-DIR (Saulteaux - from "A Collection of Saulteaux Texts" by Harold Logan)

[Ser]

Today I was thinking Classical Chinese might be another language without a nice equivalent of "or". My handbooks of Classical Chinese grammar actually don't say anything about disjunction, but after a bit of searching I found the following in Joseph Needham and Christoph Harbsmeier's Science and Civilisation in China, vol.7, page 120:

The interesting point is that Classical Chinese has no current equivalent for the declarative vel ‘or’, and the exclusive aut ‘(either) or’. [...]

In pre-nominal position, we very occasionally find the particle jo 若 used to mean ‘or’:

• Please let the ruler or the heir apparent come (chhing chün jo ta tzu lai 請君若大子來). [Tso Chuan, Duke Ai 17, fu 2]

During Han times, moreover, we begin to find the particle huo 或 ‘some’ used to mean ‘or’. The earliest example that I have found, from the -2nd century, comes from the Huai Nan Tzu 淮南子:

• Generally, when people think or plan, they all first consider something as right and only then proceed to doing it. That they are right or wrong (chhi shih huo fei 其是或非) is what distinguishes the stupid from the clever. [Huai Nan Tzu, ch. 9]

How, then, does one ordinarily say ‘This is fish or fowl’ in pre-Han times (and after)? [...] In formal logic ‘p or q’ is understood to be exactly synonymous with ‘if not-p then q’. The ancient Chinese availed themselves of just this logical equivalence. Instead of ‘This is fish or fowl’ they said something like ‘If this is not fish then it is fowl.’ The pattern for a declarative disjunction in Classical Chinese is ‘fei 非 p tsê 則 q’ ‘if not-p then q’. This is indeed a current pattern in ancient Chinese literature. It is current exactly where we would expect a disjunction.

Unfortunately he doesn't give an example.

As for interrogative "or", Needham says that the standard way to render it is to ask two questions: "Is it fish? Is it fowl?" "Quite often", he says, chhih 其 is used (which I guess is the pre-classical 'surely'), and i 抑 'or rather' may also be used.