Conlang Random Thread

[Travis B.]

Another thing to take into account is stress differences - some words (e.g. content words) tend to be stressed more often than other words (e.g. grammar words) - which can in turn influence how sound change is applied to them. Also consider words that tend to be frequently collocated, which can interact with each other in their phonetic evolution and thus undergo "irregular" sound changes as a result.

[Kuchigakatai]

Yet another reason is simply analogy, either coming from words of the same type or subtype of word class, or from words from the same family (being derived from the same root and bearing related meanings), in which one word should undergo a sound change, but the other words influencing it have a different context and so keep the original sound, making the one word in question keep it too.

For example, let's imagine a conlang Alpha and its daughter lang Beta, in which the old /d/ merged with /ɾ/ as [ɾ] intervocalically. The present-tense verb paradigm of rebn 'to fish':

rebn reba reb rebl > rebn reba reb rebl

With a paradigm like this, you would expect a verb such as kod- 'to stay, remain' to show up as koɾa for the 2nd form:

kodn koda kod kodl > kodn koɾa kod kodl

But analogy could easily work from the other conjugations of the same tense here to keep the -d-, giving koda in the Beta language instead of koɾa.

Meanwhile, for an example of related words from the same word family influencing a word to conserve a sound, consider this family in Alpha:

kyad 'speech; word'
ukyadn 'to speak' (ukyadn ukyada ukyad ukyadl)
kyado 'aloud'
kyadnaw 'language'

You would expect kyado to become kyaɾo, but analogy with related words could potentially make the word conserve the -d-, giving kyado instead.

[Jonlang]

Thanks for the replies. I coined a word for one of my conlangs: gwors, and I had been playing around with it for a whole day when I remembered it should have lost the /w/ in the sequence /ɡwɔ/ and it ought to have been gors. And so the search began for a reason why this particular word retained its /w/.

[Jonlang]

I have an SOV conlang, but when I just came to attempt to translate the following:

"To know wisdom and instruction; to perceive the words of understanding." – The Book of Proverbs, 1:2.

I instinctively analysed it as:

Code: Select all

know.INF wisdom.NOM and.CONJ instruction.NOM; perceive.INF words.NOM understanding.GEN 

but I have no idea why I instinctively put the verb first, but placing it last seems wrong in this context... but I can't even explain what this context is!

[bradrn]

but I have no idea why I instinctively put the verb first, but placing it last seems wrong in this context... but I can't even explain what this context is!

Maybe it has something to do with the fact that you speak Welsh?

Or, less facetiously, it might have something to do with how this is a subordinate (specifically purposive) clause in the English, rather than an independent clause.

[jal]

but I have no idea why I instinctively put the verb first, but placing it last seems wrong in this context... but I can't even explain what this context is!

Unless I'm missing something, it seems to me you analyzed it in the same order as English has? Anyway, I'd think that SOV would have "wisdom and instruction know, words of understanding perceive", but that's probably what you yourself would have expected as well?


JAL

[Jonlang]

but I have no idea why I instinctively put the verb first, but placing it last seems wrong in this context... but I can't even explain what this context is!

Unless I'm missing something, it seems to me you analyzed it in the same order as English has? Anyway, I'd think that SOV would have "wisdom and instruction know, words of understanding perceive", but that's probably what you yourself would have expected as well?


JAL

Maybe. I think maybe the English to infinitive may have been influencing my thinking. I translated to sing is to laugh as:

Code: Select all

sing.INF laugh.INF be.GNO

Lúelanto hyarno set (all infinitives end in -o).

because it uses the gnomic is used for general truths and aphorisms.

[hwhatting]

I also would expect an SOV language to have *wisdom and instruction know-INF; the words of understanding perceive-INF" Plus, the SOV languages I know well (Turkic) normally have the order modifier-modified, so it would be *understanding-of the words peceive-INF.
Another thing is that Prov. 1:2 is an embedded clause to Prov. 1:1, so what a Turkic language would do here is:
English: 1:1 The proverbs of Solomon the son of David, king of Israel; 2 to know wisdom and instruction; to perceive the words of understanding;
3 to receive the instruction of wisdom, justice, and judgment, and equity; 4 to give subtilty to the simple, to the young man knowledge and discretion.
SOV Turkic: These (wisdom and instruction know-INF-NML understanding-of the words peceive-INF-NML wisdom-of, justice-of, and judgment-of, and equity-of the instruction receive-INF-NML subtilty to the simple, to the young man knowledge and discretion give-INF-MNL Israel-of king-of-NML David-of Son Solomon-of) proverbs (are).

NML: a nominalizer that allows complements to stand before the noun, e.g. Kazakh -ghy.

So a Turkic-style SOV language would move the introduction from I:1 to 1:4. You could keep "Israel-of king-of-NML David-of Son Solomon-of" in I:1, but while that would keep the text more in line with the original structure, it would make it even harder to parse.

[Torco]

despite my many years conlanging, I'm embarassingly low in knowledge of linguistics beyond an intuitive, tinkerer's understanding. so, quick question for those with more knowledge: what do I call these two particles?

for context, they're from a highly isolating conlang, probably as isolating as eng. one is so, which functions as "such that" or "of" in the sense that you just put it after a phrase and it becomes a relative clause. like japanese 'no' (as in kimi no uso, you of lie, your lie (in april is a very good anime btw)) or the way we like to use "de" or "con" in spanish (el gato de mi prima, el gato con botas), we call em 'preposiciones' but there they're doing relativizing work, no? The odd thing about it is that it fulfills all of the relativizing roles, as in every relative clause is whatever so whatever (or )This is a... relative pronoun? a posessive pronoun? a relativizing posessive pronoun? an example.

fet

3P.SING.INF

a

PLUR

kanr

feather

kai

to.have

ay

and

kiro

stable

lumam

temperature

so

of

yi

to.be

They are endothermic, with feathers.

oka

1PS

so

of

to

small

haedda

cat

kosta

away

hem

walk

i

PAST

My little kitten walked away.

Also, how do I call another particle if it's only, or dominantly (in any case, paradigmatically) used to answer questions?: like, if this cat is se haedda but if you ask me which cat, I'll answer "this cat" by saying sem haedda instead of *se haedda ? this is a... responsitive demonstrative? answeric demonstrative?

[bradrn]

These things never have single ‘best’ answers, but I’d call them a ‘relativiser’ and ’question-answering demonstrative’ respectively.

[Imralu]

fet

3P.SING.INF

a

PLUR

kanr

feather

kai

to.have

ay

and

kiro

stable

lumam

temperature

so

of

yi

to.be

They are endothermic, with feathers.

oka

1PS

so

of

to

small

haedda

cat

kosta

away

hem

walk

i

PAST

My little kitten walked away.

I'd gloss that as GEN just because using "of" or "de", if you're glossing in Spanish, might lead to the expectation that its use is more similar to that of a particular English or Spanish preposition. I feel like "genitive" is sufficiently broad, although someone who's really into Latin might be like "Hmm, this usage is not like the genitive at all!", but that would be a them problem.

I don't quite follow as to how possessive relationships are a form of relativisation. Both are a form of modification, subordination et. al., but as far as I'm concerned, relativisation is a way of subordinating a clause and "my" is not a clause (although some languages could easily express it as such: "that I own" for example).

Also, how do I call another particle if it's only, or dominantly (in any case, paradigmatically) used to answer questions?: like, if this cat is se haedda but if you ask me which cat, I'll answer "this cat" by saying sem haedda instead of *se haedda ? this is a... responsitive demonstrative? answeric demonstrative?

Responsive is already a word and would work fine here, I think. Maybe glossable as RESP.DEM or RDEM.

[Imralu]

Any interesting divisions of days/nights etc. in your conlangs? Is there a separate word for nychthemeron, or does it coincide with the word for "daytime" (as in English) or "nighttime". Does a new day begin at midnight? At dawn? At dusk?

In Hibuese, "days" and "nights" are considered separately.

Code: Select all

          daytime  nighttime
next:     gyaga    dwoga
current:  gyanye   dwonye
previous: gyazi    dwozi

It might be tempting to translate them as ...

gyaga = tomorrow
dwoga = tomorrow night
gyanye = today
dwonye = tonight
gyazi = yesterday
dwozi = last night
... but that is often wrong.

Before dawn, the coming day, "today", is gyaga. Gyanye makes no sense as there is no current day. After dusk, the day that has just ended, "today", is now gyazi.. The word tonight would be translated as dwoga during the early morning or the daytime, but once the sunsets and "tonight" is underway, it is dwonye.

So the true translations are more like:

gyaga = tomorrow / today
dwoga = tomorrow night / tonight
gyanye = today
dwonye = tonight (or "last night" if you're one of those weird people who calls the pre-midnight part of the current night "last night" just because it"s now after midnight)
gyazi = yesterday / today
dwozi = last night / this morning (before dawn)

But then also, these can be used in a narrative with a different time context*, meaning that gyaga can also mean "the next day"/"that day", gyanye can also mean "that day" and gyazi can also mean "the day before", "that day" etc.

* Is there a word for the contextually understood temporal reference point, such as in a narrative? I feel like I knew one, but can't think of it.

[jal]

Interesting! Since I've only created a-posteriori or at least Earth-bound conlangs lately, nothing special to mention about them with regards to days/nights etc. (And I'm pretty sure I've never before encountered the word "nychthemeron" :D)

Sajiwan has "de" (day) and "nay" (night), where, like English, "de" is used both for daytime and the 24h cycle, and "nay" can both refer to the entire nightime period as well as late (dark) evening ("ifning" otherwise). It uses adjectives "pwaya" (previous) and "nes" (next) to refer to past/future occurences, so "pwaya de" is yesterday, "pwaya nay" last night (or, "the night before" like "Kwismas pwaya nay" - "Christmas eve"), "nes de" is tomorrow, etc.


JAL

[Jonlang]

Today I've done a bit of easy Sunday conlanging - which means mostly pondering and trying stuff out rather than getting anything done.

One of my conlangs is supposed to be Latin-ish with SOV order but it is pretty free to move stuff about in its literary register and a more rigid SVO order in its colloquial register. I thought about maybe having subordinate clauses be V1, so it would work like:

the king saw that she cried

Code: Select all

king.NOM.SG see.IND.PST.SG cry.SUBJ.PERF.SG pron.NOM.3RD.SG

(the) king saw (that) cried she

but would this be naturalistic? It uses the subjunctive for subordinates so there's no word corresponding directly to English that.

[Imralu]

Today I've done a bit of easy Sunday conlanging - which means mostly pondering and trying stuff out rather than getting anything done.

One of my conlangs is supposed to be Latin-ish with SOV order but it is pretty free to move stuff about in its literary register and a more rigid SVO order in its colloquial register. I thought about maybe having subordinate clauses be V1, so it would work like:

the king saw that she cried

Code: Select all

king.NOM.SG see.IND.PST.SG cry.SUBJ.PERF.SG pron.NOM.3RD.SG

(the) king saw (that) cried she

but would this be naturalistic? It uses the subjunctive for subordinates so there's no word corresponding directly to English that.

Hmm, yes? If the subjunctive is taking over the role of an overt conjunction that would normally be at the beginning of the clause, then ... sure. In Swahili, relativised verbs tend to appear at the beginning of their clause before the subject, although this tendency seems to be stronger in older texts than it is now, with SVO being used more now. This doesn't happen in complementiser clauses, as in your example, but the same principle could apply because, what I believe attracts the verb to the beginning of the clause is that it contains something that relates directly to what appears before the clause (the relative affix relating to the the antecedent of the relative clause).

E.g. "This is the house the house where my friends live":

Code: Select all

(1)
Hii  ni nyumba wa-na-po-ishi         marafiki zangu.
This is house  they-PRES-REL.16-live friends  my
               [relative_clause___________________]

(2)
Hii  ni nyumba marafiki zangu wa-na-po-ishi
This is house  friends  my    they-PRES-REL.16-live 
               [relative_clause___________________]

(3)
Hii  ni nyumba amba-po    marafiki zangu wa-na-ishi.
This is house  REL-REL.16 friends  my    they-PRES-live
               [relative_clause________________________]

Sentence one shows the relativised verb wanapoishi fronted to the beginning of the relative clause. The relative affix -po- relates to the antecedent nyumba and class 16 is a locative class, meaning that -po- generally means "where" or also "when".

Sentence two has the subject of the relative clause, marafiki zangu, before the relativised verb, leaving the beginning of the relative clause unmarked. It's pretty clear in this case, but in any case where you get two noun phrases together that could potentially be understood as one, there could be a garden path.

Sentence three has the relative affix -po- hosted on the word amba-, which is are much more flexible way to introduce relative clauses, as only certain verb forms allow direct relativisation.

Personally, I like sentence (2) and it seems to be perfectly grammatical these days, although some grammars and textbooks still proscribe it, and I can understand why. A relative clause that begins with a relativised verb or amba- is immediately clear and does not present as much opportunity for garden paths as one where the relative clause simply begins with the subject.

In German, which basically does the opposite of your language, some analyses explain the subordinating conjunction as occuping the space at the beginning of the clause that the verb would otherwise inhabit and that this sends the finite verb to the back end of the clause(or prevents it from being fronted under other analyses that see German as inherently SOV with fronting of the finite verb in independent clauses).

It seems perfectly reasonable to me that your subjunctive verb form could be pulled to the beginning of the clause to occupy the space that the absent complementiser would take, seeing as it fulfils that role too.

[Imralu]

Interesting! Since I've only created a-posteriori or at least Earth-bound conlangs lately, nothing special to mention about them with regards to days/nights etc. (And I'm pretty sure I've never before encountered the word "nychthemeron" )

Sajiwan has "de" (day) and "nay" (night), where, like English, "de" is used both for daytime and the 24h cycle, and "nay" can both refer to the entire nightime period as well as late (dark) evening ("ifning" otherwise). It uses adjectives "pwaya" (previous) and "nes" (next) to refer to past/future occurences, so "pwaya de" is yesterday, "pwaya nay" last night (or, "the night before" like "Kwismas pwaya nay" - "Christmas eve"), "nes de" is tomorrow, etc.

Yeah, Hibuese, the current incarnation of this conlang that's been kicking around for ages, is also Earth-bound, but with a bit of hand-waving, they've been able to escape a lot of colonial influences and are a bit more a priori in terms of language and culture.

Check out Swahili time though, which is very much Earth-bound but still managed to hold onto their own way of defining days in spite of globalisation. In Swahili, you'll look at a clock saying 7:00AM and pronounce it as saa moja asubuhi (hour one morning). In spite of using clocks the same way as everyone else, they start counting the hours of the day from 6:00AM and, essentially, it means, if you're looking at an analogue clock, you imagine the hour hand projecting backwards to offset it by six hours.

[foxcatdog]

Where do negatives usually come from? I'm struggling to find the form of a negative auxillary since *-na the old negative also appears in the negative. copula. I have *san "to lack" as a source but it already became the standard way of negating a noun.

Edit i have 3 current sources the second of which would be attested at and earlier point as "to need".and the third of which would be attested at an earlier point as "to leave"

*manja and *maniima both negative auxiliaries either from *ma+nea “do/go + bad/wrong” or *man+ia “remove from need + anticausative”

*mori and *morima “from “mauria die/destroy + anticausative”

*lauja and *lawa from *laweja and *lawejala from *lamba+ia “stay + anticausative”

[hwhatting]

One pathway is intensifiers or complements - e.g. French intensified the inherited negation ne with pas "step", so je ne sais pas "I don't know" literally means "I don't know a step". In colloquial French, the old negation is dropped, so you say je sais pas, and pas now carries the sense of negation. You can do that with other intensifiers, e.g., in 18th century French point "point" competed with pas. English not is originally a compound of something like ne and aught "something", and IIRC the Scandinavian negation ikke is the result of the old negation being dropped from an equivalent combination "not something".
For another pathway, modern Standard Arabic has a way of negation where "other than X" became to mean "not X" (negation with ghair).

[foxcatdog]

So...
Proto Amaric
pai na nawo tuni
3.sg.m NEG talk often
"He doesn't talk much"

Old Amarin
mai na naao mauria tuni
3.sg.m NEG talk need often
"He doesn't talk much"

Current Amarin
mai lauja naau runi
3.sg.m NEG talk often
"He doesn't talk much"

*lauja is a contraction of *laweia "to leave (its got the same ending and *mauria) and *lawa is its imperfective form a contraction of *lawejala. The *w retention is irregular.

[foxcatdog]

In terms of phonological conservatism Old Amarin or Najansiipaa Amarin is quite conservative retaining 12 out of 20 of its numerals in their original form. Current Amarin or Kilsimaa Amarin only retains four.

Numerals
*nari “one” from *nati > *nati
*sina “two” from *śina > *sina
*isju “three” from *eisku > *eisku
*saa “four” from *santa > *saa
*wikan “five” from *wikan > *wikan
*nura “six” from *noura > *noura
*sanis “seven” from *saniś > *sanis
*palka “eight” from *palka > *palka
*luni “nine” from *luni > *luni
*iski “ten” from *iske > *iske
*ana “eleven” from *jana > *ana
*tauri “twelve” from > *tauti > *tauti
*ilai “thirteen” from *pilai > *pilai
*issaa “fourteen” from *iski+saa
*kaja “fifteen” from *kanka > *kaja
*sawi “sixteen” from *sambi > *sawi
*torra “seventeen” from *torta > *torta
*naskii “eighteen” from *naskii > *naskii
*ińas “nineteen” from *eińat́ > *eińas
*mini “twenty” from *meine > *meine