Some notes on Laqar

[Travis B.]

consider this...

OW!!!

(Consider that a conlanging compliment.)

And that is a typical monosyllabic CVC verb stems. That is ignoring that there are also CV, CV:, CV:C, CVCV, CV:CV, CVCCV, CV:CCV, CVCV:, CVCCV:, CV:CV:, CV:CCV:, CVCVC, CV:CVC, CVCCVC, CV:CCVC, CVCV:C, CVCCV:C, CV:CV:C, and CV:CCV:C verb stems. There are also subclasses based on whether they undergo palatalization or not or whether they contain uvular consonants (Note that these are their forms in Proto-Laqar, not Early Middle Laqar.)

[Vardelm]

*hurtful things*

You sir, are ruthless.

[Travis B.]

*hurtful things*

You sir, are ruthless.

In future stages of Laqar I am going to have to do some significant leveling, as, as it is right now, Laqar's verbal system is a threat to the sanity of its speakers.

[Travis B.]

Here is another paradigm, of a verb with a CVCV Proto-Laqar stem containing a uvular consonant.

[Travis B.]

Laqar developed a wide range of subordinating conjunctions by fusing nouns with case markers while dropping preceding complementizers. These are:

	form
if	yòq'e
because	yòq'o
so that	yòča
than	bẽne
before	kéča
after	c'íča
when	dák'ə
while	dáko
where	mòk'ə
how	rók'ə
whether	kók'ə
as	rólo
but	sák'ə
despite	sálo
lest	sétə

These follow the subordinate clause and precede the main clause.

[Travis B.]

Here are some general relational nouns:

	base	3rd sg. m.	3rd sg. f.	3rd pl. m.	3rd pl. f.	gloss
before, in front	ke	kéwə	kéri	kèyo	kéxô	chest
after, in back	c'i	c'íwə	c'íri	c'íwo	c'éxô	back
over, above	buk	búkə	búkri	búko	búkô	hair
on top of	loś	lóśə	lóśri	lóśo	lóśô	head
on the bottom of	mírrə	mírrəwə	mírrəri	mírrəwo	mírrôxô	buttocks
under, below	k'íčə	k'íčə	k'ečrí	k'íčo	k'ečò	feet
in, into, out of	ną	nãnə	nãri	náno	nãnô	stomach
on, onto, off of	dô	dòwə	dòri	dòwo	dòxô	skin
outside	síbə	síbəwə	síbəri	síbəwo	síbôxô	wind
around	dékə	dékə	dékri	déko	dékô	circle

The following are some relational nouns that must be used with specific cases, rather than with any relevant case:

	base	3rd sg. m.	3rd sg. f.	3rd pl. m.	3rd pl. f.	gloss
instead of	máweča	mòwəča	môríča	mòwoča	moxòča	place=ALL
as	rólo	rówəlo	rórilo	ráwolo	róxôlo	way=INST
because of	yòq'lo	yòq'əlo	yòq'rilo	yòq'olo	yòq'ôlo	reason=INST

[Travis B.]

The relational nouns directly follow their objects, which syntactically are their possessors, and are marked with a case, which usually except for certain relational nouns is a locative case. They inflect for the gender and number of their objects, like all possessors. Also, they may be used as normal nouns; whether they are relational or not is a matter of context.

[Travis B.]

One construction in Laqar that I should explore more is compound verbs, formed by a gerund marked with -cə (-tsʰæ in Proto-Laqar) which is marked for the patient of the main verb, if there is one, combined with a finite verb marked with both the patient and the gerund.

[Vardelm]

I dig the relational nouns. I've thought about using them before, but seeing them here makes me realize they will fit in really well with a conlang I've only just started on. I'll be creatively borrowing some inspiration from here!

Are those essentially inflected prepositions per Celtic? More insanity, which is good.

[Travis B.]

I dig the relational nouns. I've thought about using them before, but seeing them here makes me realize they will fit in really well with a conlang I've only just started on. I'll be creatively borrowing some inspiration from here!

Are those essentially inflected prepositions per Celtic? More insanity, which is good.

The relational nouns aren't as exotic as they sound when you consider English - "I put the can of beer on top of the car" - "The barnacle was stuck on the bottom of the boat" - "The rabbit sat in front of the lawnmower" - and so on.

[Vardelm]

The relational nouns aren't as exotic as they sound when you consider English - "I put the can of beer on top of the car" - "The barnacle was stuck on the bottom of the boat" - "The rabbit sat in front of the lawnmower" - and so on.

Oh, definitely. It's just that the use of body parts for that will play well with the themes in that lang. It's a small thing, but it adds flavor.

[Travis B.]

The relational nouns aren't as exotic as they sound when you consider English - "I put the can of beer on top of the car" - "The barnacle was stuck on the bottom of the boat" - "The rabbit sat in front of the lawnmower" - and so on.

Oh, definitely. It's just that the use of body parts for that will play well with the themes in that lang. It's a small thing, but it adds flavor.

Actually, relational nouns are very commonly derived from body parts.

[Travis B.]

I figured I should work on Laqar some more, but I am not sure what to do with it next - I really like just how fusional it has gotten, even though that is just asking for it to go under massive amounts of reanalysis in later stages (e.g. like going from Old Irish to modern Irish). For some things, little needs to be done, e.g. adjectival declension. However, nominal declension is significantly more complex, and verbal conjugation is even more complex. I do not really know even where to start to rework its morphology, and to be honest, I like its complexity, even though it is only realistic that it would be reanalyzed in later forms.

[Travis B.]

One problem is that even if I did not massively reanalyze Laqar's morphology but only turned it into discrete declensions and conjugations, I would still need to keep tables of all the possible forms for each declension or conjugation, which is easier said than done (note how large the tables for two different verbs are; tables like those would have to be created for each individual declension or conjugationn).

[Rounin Ryuuji]

. I do not really know even where to start to rework its morphology, and to be honest, I like its complexity, even though it is only realistic that it would be reanalyzed in later forms.

Complexity can be resilient in some cases (I understand Welsh to be extremely fond of it); Spanish comes to mind, with keeping the past tense where French and Italian seem to have switched to using what was originally their perfect tenses, and Japanese to some extent — the nidan verbs of Classical Japanese (except one little straggler) have all fallen into the more regular ichidan paradigm, but then the yodan verbs got an extra stem and became godan verbs instead (they also absorbed some of the nidan verbs sporadically).

Otherwise, I enjoyed reading about this, but don't feel I have much to add (and I'm probably not saying anything you don't already know anyway).

[willm]

One problem is that even if I did not massively reanalyze Laqar's morphology but only turned it into discrete declensions and conjugations, I would still need to keep tables of all the possible forms for each declension or conjugation, which is easier said than done (note how large the tables for two different verbs are; tables like those would have to be created for each individual declension or conjugationn).

One alternative to making that many declension and conjugation tables would be to decide that, perhaps barring some truly irregular words, the forms are predictable from principal parts (and maybe conjugation/declension class). You could work out tables for a few examples and then just list the class and necessary forms for the rest. This probably requires quite a bit of leveling by analogy (given how complex the morphology is, it seems unlikely, though not impossible, that a very large number of different declension and conjugation classes could exist without significant simplification), but it would make the problem somewhat more manageable. You may already have thought about all of this, though.

[Travis B.]

One alternative to making that many declension and conjugation tables would be to decide that, perhaps barring some truly irregular words, the forms are predictable from principal parts (and maybe conjugation/declension class). You could work out tables for a few examples and then just list the class and necessary forms for the rest. This probably requires quite a bit of leveling by analogy (given how complex the morphology is, it seems unlikely, though not impossible, that a very large number of different declension and conjugation classes could exist without significant simplification), but it would make the problem somewhat more manageable. You may already have thought about all of this, though.

Even if I drastically reduced the number of classes, even then each class still would have many forms, as demonstrated by my conjugation tables above.

[Travis B.]

The obvious places where I can level are where voiceless laterals and velars have different fates depending on what is adjacent to them. But beyond that there are a good number of natural classes stems can fall in, specifically CV, CV:, CVC, CV:C, CVCV, CV:CV, CVCV:, CV:CV:, CVCCV, CV:CCV, CVCCV:, CV:CCV:, CVCVC, CV:CVC, CVCV:C, CV:CV:C, CVCCVC, CV:CCVC, CVCCV:C, and CV:CCV:C.

[Travis B.]

I have been trying to wrap my brain around how to synchronically analyze Early Middle Laqar and am getting nowhere. I first tried to analyze it in terms of vowel alternations based on word length and stress. But this does not take into account the fact that there is very frequent unstressed short vowel elision, and in cases this is followed by vowel-glide sequence reduction and consonant elision or assimilation, which may hide the original vowels.

[Ares Land]

A few ideas, in case it helps:

• Conjugation tables aren't very useful when you mark both patient and agent. I've played with it, but ultimately, it's best to keep tables of person suffixes, other inflections, and stem change separates. It looks like you could keep a table of indicative person markers, and a table for the subjunctive, for instance.
• I'd suggest a few gaps on those polypersonal agreement tables. For instance, is 1S>1S reflexive? If so, how about marking this differently? (With intranstive marking + a separate prefix, for instance). How about 3S > 3S? Is it reflexive or not?
  How about 2S>2P, or 2SM>2PM? You can of course keep them, but in that case, those would be used very likely - they'd be formed by analogy.
• What I generally do is form the regular paradigms on the basis of a dozen or so common verbs. I run the sound changes on these, than try to see which patterns happen.
• Generally: unpredictable forms are replaced by regular ones, by analogy with the most common ones. Keep a few unpredictable bits for flavor. For instance, in Old French, the 1st person singular was irregular in many cases: j'ains 'I love', il aime 'he loves' > j'aime. But je vais, il va was so common it was kept (even though the regularized version je vas existed historically and hangs on in some dialect.)
  Maybe most of your verbs have a regular, say, 2SM intransitive past imperfective q'altóne but a few keep the less predictable q'alętóne
• For your own sanity, you need to figure out the principal parts: that is, the minimum number of forms you need to work out a verb's complete paradigm.
• Work out how many irregular verbs you want, and derive the principal parts for these.

But from what I see... the verbs could be reworked by analogy but it's still perfectly realistic to keep it that way. The Greeks did just fine for millennia with much worse!