Examples of truly unique conlang features?

[Tropylium]

A proximal/distal distinction in 2d person pronouns.

Or, according to whether the addressee is perpendicular to the speaker.

[ObsequiousNewt]

Nullar number (e.g. "no horses", but as a suffix.) Proto-Elmin had a five-way distinction in numbers—distinction of singular/plural and singulative/collective, and nullar.

[Aili Meilani]

I remember one my early conlang having negative number. Don't ask me what it was, I don't know.

[Rhetorica]

That's cool. Reminds me that I wrote down something about a "group of" suffix, which would turn "dog" into "dog pack" and "tree" into "forest".
And, now you come to mention it, if I suffix it onto "type of Noun/s" it could have a similar effect to yours.

Have you found anything similar in natlangs?

I've always thought a "complete set of" number made sense, but for some reason linguists only ever seem to observe paucal and dual numbers. Why wouldn't you use the same concept for "a pod's worth of peas" and "a pair of ducks"? A lot of things can have natural quantities attached to them.

[Rhetorica]

I remember one my early conlang having negative number. Don't ask me what it was, I don't know.

Maybe you use it to represent reversals, e.g. in motion? I could see "-3 people went home" as meaning "3 people came".

[R.Rusanov]

...for some reason linguists only ever seem to observe paucal and dual numbers...

Welsh? Arabic? Nilo-Saharan?

[Nortaneous]

inverse number innit

http://en.wikipedia.org/wiki/Kiowa_lang ... inflection

[Rhetorica]

Neat, but not negative number, so I doubt that was what was meant. That's just classes of nouns with collective forms instead of singulative. (Which, I suppose, is where all the languages with "complete set" numbers have gone...)

[finlay]

That's cool. Reminds me that I wrote down something about a "group of" suffix, which would turn "dog" into "dog pack" and "tree" into "forest".
And, now you come to mention it, if I suffix it onto "type of Noun/s" it could have a similar effect to yours.

Have you found anything similar in natlangs?

I've always thought a "complete set of" number made sense, but for some reason linguists only ever seem to observe paucal and dual numbers. Why wouldn't you use the same concept for "a pod's worth of peas" and "a pair of ducks"? A lot of things can have natural quantities attached to them.

This is called a collective plural.

[Benturi]

Nothing groundbreakingly original here, but a feature I have used in many of my conlangs is that all numerals are nouns, and the head of the nominal phrase is the numeral word and not the counted noun. Like "pair", "couple", "dozen", "score", etc. but applied to all numbers - including words indicating indefinite quantities, like "few", "several", "many", etc. Is there any natural language that does this?

Btw, the Kiowa / Tanoan inverse number thing is crazy.

[Imralu]

Nothing groundbreakingly original here, but a feature I have used in many of my conlangs is that all numerals are nouns, and the head of the nominal phrase is the numeral word and not the counted noun. Like "pair", "couple", "dozen", "score", etc. but applied to all numbers - including words indicating indefinite quantities, like "few", "several", "many", etc. Is there any natural language that does this?

I don't know of any natlangs, but all of my conlangs work this way.

In Ngolu, content words are all verbs (including numerals and quantifiers) and they are nominalised by using an article before them. Singular articles are used before number verbs.

izu daw
NOM.3p.SPEC.INAN.REL be.tree
(some specific) trees

zu egio daw
NOM.3s.SPEC.INAN.REL be.threesome be.tree
three trees / a trio of trees

If a plural article is used, it means multiple groups of the number specified.

izu egio daw
NOM.3p.SPEC.INAN.REL be.threesome be.tree
trios of trees / trees numbering some multiple of three

[----]

Dunno if it's completely unique but I've never heard of anything exactly like the way noun class/'gender' works in Taahu. Each noun is in one of two classes which I call animate and inanimate because actual animacy determines which class a noun goes in like 90% of the time. For clauses that include both an animate (na.) and inanimate noun (ni.), and the animate noun is the subject and the inanimate is the object/in the predicate, both are unmarked. But, if the ni. is the subject and the na. is the object, both participants get marked as if they are the subject of an intransitive clause and the verb itself is also marked as intransitive. For instance:

xú isée yíkpayi'in
1 bug/beetle AN>INAN-PL-eat.PROG
"I'm eating beetles"

iséew xwé kipayi'í náa
bug.INTR 1.INTR INTR.PL-eat.PROG=EMPH
"these bugs are really eating me up!"

I guess this is almost like a direct-inverse system or something?

[Trebor]

A great idea for a thread.

Here are some interesting, though perhaps not unique, features already present in or being considered for a new conlang of mine, which is still in its earliest stages.

1) There are three types of articles--generic, indefinite/partitive, and definite.

2) As in Igbo, words can consist only of either tense or lax vowels but not both. However, unlike that Nigerian language, my project is supposed to be as isolating as humanly possible.* So, preceding articles must conform to the quality of the vowels in the noun modified.

3) Forming an inverse-number system, nouns will be categorized into three groups--(a) singular, (b) plural, and (c) mass/collective. But, unlike the Tanoan languages where this phenomenon is attested, the same marker will not be recycled to indicate non-default number in (a) and (b). Again unlike those North American aboriginal oddballs, but as in Welsh, there may be a way of deriving a singulative form of (c).

4) There may be different sets of personal pronouns for independent and dependent clauses.

5) There will be some unusual morphosyntactic alignment.

a) As in Dyirbal, the presence of first- and second-person referents makes the clause act nominative-accusative, while third-person arguments make it behave ergative-absolutive. But, unlike that Australian aboriginal language, so does the appearance of third-person referents which belong to the speaker or hearer.

b) Or, verbs are subdivided into two classes--one which takes nominative-accusative and the other which calls for ergative-absolutive.

c) Or, some other option not to be found in Indo-European languages will be employed.

Any comments or suggestions are welcome, as well as recommendations of linguistics papers.

*How to make derivational morphology interesting is an issue I will really have to investigate. What to do about Indo-European participles, I wonder?

[jal]

I'm currently developing July 2014, a yet unnamed conlang, that has syntactic agreement on the verb (subject, direct object, oblique object) but has semantic case marking on its constituents, and indicates focus/topic as part of the subj/obj/obl marking on the verb (fused particles). Word order is typically VFT (verb-focus-topic) but the constituents may be moved around, which is indicated by the order of the verbal agreement proclitics. Also, the order of the markers is the exact opposite of the order of the constituents, so that both the clitic and the noun of the focus in VFT order are closest to the verb. The agreement marker can also indicate a "null" constituent, which is akin to a pronoun. I have no idea whether something even remotely like this exists in a natlang.

In my old conlang Kotanian (which is in desperate need of being dusted of), I have a rather insane amount of verbal suffixes, for tense, aspect and mood, the latter subdivided in directive, conditional, certainty, concord and opinion/attitude, and voice, all in all 48 of them (and additionally there's two adjectival and one adpositional verbalizer). It works out quite nicely though, and enables one to say things like pèrtòynawibra, "making full to my dissatisfaction", or loprèmilôa, "already had been lost", or, my personal favorite nyatyalòynèmiètòmabra, "unfortunately will be made detrimentally fine in that case". All examples from here.


JAL

[Ser]

WeepingElf commented in my thread on Kgáweq' that he had no knowledge of any natlang analogue to the success affix. A few of the success affixes may be present to some degree (indicating conativity and evidentiality) but others indicate success, failure, or qualifications thereof.

Mandarin has a resultative affix (-成 -chéng) which sometimes indicates the successful completion of something. 翻译 fānyì 'to translate sth' > 翻译成 fānyì chéng 'to finish translating sth'.

I guess you could say this rather represents completing something rather than successfully doing something, it's hard to distinguish what's to successfully complete something, and to completely or thoroughly do something...

Failing to do something is very similar to do something in a wrong way, too, about which we could have several English mis- verbs like "misquote", "misidentify", "mismatch", "misjudge"...

Though I guess the thing about your conlang is that these aren't derived verbs or constructions but inflectional affixes?

[jal]

Mandarin has a resultative affix (-成 -chéng) which sometimes indicates the successful completion of something. 翻译 fānyì 'to translate sth' > 翻译成 fānyì chéng 'to finish translating sth'.

Which would be a normal telic particle? Perhaps the "success" thing is just a variant of a telicity marker.

it's hard to distinguish what's to successfully complete something, and to completely or thoroughly do something...

It would be purely pragmatic when to use it, I guess.

Failing to do something is very similar to do something in a wrong way

I wouldn't say that. At least failing to finish doing something doesn't mean you did it wrong (you could be interrupted, for example).


JAL

[Jūnzǐ]

In one of my filler languages for Cytàm (previously known as Ceutàm; I've referred to it on the CBB), I have started working on a language with consonant labialization harmony, known provisionally as Tlwěkwīdwù ([t͡ɬʷěkʷȳdʷù]).

[Pogostick Man]

WeepingElf commented in my thread on Kgáweq' that he had no knowledge of any natlang analogue to the success affix. A few of the success affixes may be present to some degree (indicating conativity and evidentiality) but others indicate success, failure, or qualifications thereof.

Mandarin has a resultative affix (-成 -chéng) which sometimes indicates the successful completion of something. 翻译 fānyì 'to translate sth' > 翻译成 fānyì chéng 'to finish translating sth'.

I guess you could say this rather represents completing something rather than successfully doing something, it's hard to distinguish what's to successfully complete something, and to completely or thoroughly do something...

Failing to do something is very similar to do something in a wrong way, too, about which we could have several English mis- verbs like "misquote", "misidentify", "mismatch", "misjudge"...

Though I guess the thing about your conlang is that these aren't derived verbs or constructions but inflectional affixes?

Partially. Here's the vanilla list of success affixes:

-woʼ- success
-g- accidental success
-kʼ- barely
-t- partial success
-se- almost success
-ą- failure
-(ʼ)s- catastrophic failure
-ƛ- conative, unknown outcome

[clawgrip]

WeepingElf commented in my thread on Kgáweq' that he had no knowledge of any natlang analogue to the success affix. A few of the success affixes may be present to some degree (indicating conativity and evidentiality) but others indicate success, failure, or qualifications thereof.

Actually, English basically has this: "manage to" is used to indicate successful completion of an action when there is some potential for failure. It's not a suffix, but it's as much a modal construction as "have to", "want to", "used to", "be going to", etc.

[Kou Daoguang]

Mandarin has a resultative affix (-成 -chéng) which sometimes indicates the successful completion of something. 翻译 fānyì 'to translate sth' > 翻译成 fānyì chéng 'to finish translating sth'.

I guess you could say this rather represents completing something rather than successfully doing something, it's hard to distinguish what's to successfully complete something, and to completely or thoroughly do something...

Indeed. With the bar set that low, any completed affirmative resultative verb in Chinese could be hailed a success. Too, I think the choice of (-成 -chéng) with 翻译 fānyì also has to do with translating from one language INTO another, 成 chéng being intimately tied to 'becoming' as well as 'completion'.

[Jūnzǐ]

...for some reason linguists only ever seem to observe paucal and dual numbers...

Welsh? Arabic? Nilo-Saharan?

Yeah, I actually prefer the singulative-plurative-collective system of the Nilo-Saharan langs to the singular-plural system of English.

inverse number innit

http://en.wikipedia.org/wiki/Kiowa_lang ... inflection

I've been trying to decide which filler lang on Cytàm will get inverse number: Probably Ɂunetiɂiho ([ʔu̟neˈtiʔɪ̟ho]) will get it, being a highly synthetic, highly fusional, dependent-marking, marked nominative-neutral language.

[Pole, the]

inverse number innit

http://en.wikipedia.org/wiki/Kiowa_lang ... inflection

Oh, this one has been on IOL this year!

[Haplogy]

What about something for a primitive language that has only counting words for, say, one, two, a few, lots, enough, and not enough, or something, but which also has all of these as grammatical number. It's not an obscenely large amount of grammatical numbers, and it's all the counting words, too. Are there any natlangs that do this?

[jal]

What about something for a primitive language that has only counting words for, say, one, two, a few, lots, enough, and not enough, or something, but which also has all of these as grammatical number. It's not an obscenely large amount of grammatical numbers, and it's all the counting words, too. Are there any natlangs that do this?

I would say you wouldn't need any counting words in that case.

Q: Omani man de?
A: Manon / mantu / masom / malos / manof / males de!

Q: "How many men are there?"
A: "One man / two men / a few men / lots of men / enough men / too little men are there!"


JAL

[Trebor]

Are there any natlangs that conjugate verbs for aspect, mood, and voice, but not at all for person, number, gender, and tense?

And do any natlangs possess any of the features mentioned in my previous post in this thread?