Wednesday, October 28, 2015

Dairwueh and Bryatesle Indefinite Pronouns

The Bryatesle indefinite pronouns, according to my old grammar for Bryatesle, followed the pattern given below:

(7) DN
(1) SK(2) SU(3) I/NS(4) QU(6) IN         A
(5) COND(8) COMP  A 

(9) FC
lermud, lertën, lerbat/lervind, lerden/lertar. White fields where never assigned a pronoun!
I like the ler- series of indefinites, especially as it is closely related to the relative pronouns of Bryatesle but not to the interrogative ones (which creates a nice unusual situation, typologically - although it probably could be explained by recourse to historical developments). Lerden and lertar differ by whether the free choice is 'free choice among some options' or 'free choice with infinite options'. Although I like this series, I am inclined to reject it now in favour of this somewhat simpler and improved series:

(7) DN
(1) SK(2) SU () (3) I/NS () (4) QU (6) IN
(5) COND(8) COMP

(9) FC
bale, dury, uka
An extra detail is that dury sometimes stretches in to I/NS, sometimes even as far as (2) SU, but I couldn't be arsed to make a gradient in the background of (3). Another extra detail is that mass nouns tend to take the definite secondary case marking when determined by an indefinite pronoun.

In Bryatesle, all of these can be used both independently and as determiners. They seldom take secondary cases, but when they do so they tend not to mark their primary case at all. Their inflectional stems are bal-, dir-, uk-/mys- (mys- appearing for non-nominative cases in a few dialects).

Dairwueh is somewhat similar:

(7) DN
(1) SK(2) SU(3) I/NS(4) QU(6) IN
(5) COND(8) COMP

(9) FC
vael-, nur-, mik-
Here, the blue field sometimes spills over into (6) IN, but more often, red spills into (2) SU, and at times it may reach (3) I/NS. Unlike Bryatesle, these cannot be used independently, and need some extra morpheme in order to be able to stand independently. In most dialects, this is the interrogative pronoun del, which is prefixed to the pronoun: delvael, delnur, delmik.

Tuesday, October 27, 2015

Detail #228: Conegative Verbs with a Lexical Twist

Consider the system present in many Finnic languages, where verbs have forms without (much) congruence whenever the verb is negated, i.e.

minä syön - minä en syö
sinä syöt - sinä et syö
hän syö - hän ei syö
me syömme - me emme syö
. . .
The examples above give I, you, he/she, we (don't) eat. Not all forms conflate the connegative and the third person:
hän tulee -  hän ei tule
Imagine some verbs whose meaning are somehow seen as 'negative' in some sense - lie, be missing, deceive, fail, forget, etc. Now, let's assume these lack the con-positive form entirely. So, in a normal positive utterance, they do not take congruence (or, if the language works differently from Baltic Finnic, take some extra marker or some other congruence or whathaveyou). In Finnish, for forget, you'd thus get this form:
*minä unohda (vs. minä unohdan)
*sinä unohda (vs. sinä unohdat)
*hän unohda (vs. hän unohtaa)
*me unohda (vs. me unohdamme)
. . .
Some amount of redundancy has been lost here, though. In a language that very much enforces some level of redundancy, we could imagine a positive marker becoming mandatory here, thus
*minä kyl unohda
*sinä kyl unohda
*hän kyl unohda
Where kyl is colloquial Finnish for kyllä, yes. One could go analogously to the negative auxiliary and do something like
*minä kyn unohda
*sinä kyt unohda

Mutatis mutandis for however your conlang deals with negation.

Monday, October 26, 2015

Indefinite Pronouns in Ćwarmin and Sargaĺk

Again I want to turn your attention to ye merrye conlangre's little typological treatise on indefinite pronouns. It is gold.

With Ćwarmin, we have to use several tables, because the different pronouns overlap each other too much otherwise:

(7) DN
(1) SK(2) SU(3) I/NS(4) QU(6) IN
(5) COND(8) COMP

(9) FC
bek' is barely ever used productively in the zone marked with gray. However, a number of fossilized expressions have bek' in that position.


(7) DN
(1) SK(2) SU(3) I/NS(4) QU(6) IN
(5) COND(8) COMP

(9) FC
suv-:, ink'-

(7) DN
(1) SK(2) SU(3) I/NS(4) QU(6) IN
(5) COND(8) COMP

(9) FC
Sargaĺk has a slightly simpler system:

(7) DN
(1) SK(2) SU(3) I/NS(4) QU(6) IN
(5) COND(8) COMP

(9) FC
mar-, vur-, jok-. Vur- sometimes reaches into COMP.
In Ćwarmin, the pronouns can be used both independently and as determiners. Sargaĺk does not permit their independent use, but forms compounds with the adjectives 'good', ete, thus getting forms like marete, vurete, jokete. With human referents, it often uses kime instead, obtaining markime, vurkime, jokime. In certain situations, suma, 'saint' can be used, obtaining marsuma, vursuma, joksuma; this can either be a very profound religious expression or very rude.

I am working out a similar set of graphs for Dairwueh and Bryatesle, with the intention of comparing and contrasting these systems through some example sentences.

Sunday, October 25, 2015

Detail #227: A Syllabary with a Quirk

Consider a language where there is a bunch of very uncommon syllables, i.e. these syllables appear in about a handful of words. Now, syllabaries easily get a fair share of symbols, and having symbols for syllables that basically don't form "minimal pairs" ever, due to their being very uncommon in the language might seem like a waste of effort.

So, have a symbol for "other syllable", that appears in a handful of words, but is realized in very different ways in them.

Detail #226: A Twist on Verb Markings

Let us consider a language where verbs usually have a fair share of meaningful morphemes on them whenever they're used in whatever form, i.e. a verb is nearly never used without semantically meaningful affixes. By this rather vague terminology I mean to say none of the morphemes used are exclusively used to mark the verb as belonging to this or that morphosyntactic category - i.e. nothing is exclusively a marker of a verb being an infinitive or participle or third person plural or whatever - generally it also marks some other central thing, such as location, direction, intensity, ...

Now, almost all verb phrases will adhere to the above. However, some verbs permit simpler markings, especially a few imperatives - help (in a situation of dire need), run, and a few others. However, another situation is extreme rudeness. Omitting the markers entirely is nearly always parsed as if the most rude possible affixes were included. Thus hold means wank, run means running away like the coward he is/you are, etc. Not just the meaning of the verb, but the tone in which it is used is turned as negative as possible.

Friday, October 23, 2015

A Noun Formation Pattern common to Ćwarmin and Sargaĺk

Sargaĺk and Ćwarmin share a few patterns in word formation, patterns that have spread from one to the other, and generally also tend to be present in the area at large. These are, however, not present in Dairwueh and Bryatesle, which do not form part of the same convergence zone.

A good example of a very common pattern is the noun pair question - answer. The verb to ask is ŕebat in Sargaĺk, and kinil and kinən in Ćwarmin. The nouns for 'question' are related - ŕeber in Śargalk, kinij in Ćwarmin.

However, answer also derives from the same root, using the suffix -ada in Sargaĺk and -elə/-ola in Ćwarmin, thus ŕebada, kinelə. The suffix goes back to an early ĆŊ participle, the passive causative participle. It has later come to be used on nouns that are results of other nouns, as well as nouns that are results of verbs.

Thus, you also get nouns like Ćwarola, "those caused by the Ćwar people", which essentially means 'the next generation of Ćwar", or Sargada, with a similar meaning with regards to the Sarg people.

Wednesday, October 21, 2015

Detail #225: Along the Lines of Gnomic Aspect

I got my first contribution ever, and I am quite happy about that! Since I am a thorough kind of guy, I discussed the idea a bit with the contributor, and together we came up with a bunch of ideas.

The starting point, really, was articles, or rather degrees of definiteness. Observe how for general statements, languages may differ in their use of articles and even numbers:
the car is a vehicle | a car is a vehicle | cars are vehicles
Analogous sentences to the ones given above are attested in different languages for analogous statements. Some languages even permit more than one of them.
However, what other things could we do with this? Well, we could imagine treating definitions (and other universal statements) as dealing with mass nouns instead of with count nouns, regardless if the noun normally is a mass noun or not:
car is (a) vehicle
Or we could break the verbal congruence:
(a|the) car is* (a) vehicle
* without overt third person marking, so 'be' would be better in English here.

cars is vehicle(s)
One could also imagine using some type of partitive case for the complement (or the subject) with regards to statements of this type. Or, finally, let's come up with a new article altogether!
nu car is (a|nu) vehicle

Some languages apparently have a gnomic aspect for statements such as these. I have no idea whether this also can mark truths about an object instead of truths about a subject, such as:
engineering turns nu idea into reality
engineering turn.gnomic idea into reality
This is a truth about engineering as well, but we may want to focus on the idea, and tell us something about how ideas relate to engineering and reality. With just a gnomic aspect, we can't really tell what we're focusing on in this (well, word order readjustments and whatnot could fix that, of course.)

So maybe combine a gnomic article with a gnomic aspect to make it possible for general truths to be about non-subjects – but then again, the gnomic article per se seems sufficient. But then again², redundancy is useful in languages.

Thanks for the contribution, Daniel!

Sunday, October 18, 2015

Detail #224: Inverse Things with Possession

Possession is a thing where inverse markings could make much sense - consider, for instance, a hierarchy of nouns where apposition is the way to mark possession:
cow farmer
It's even likely that there are languages where both orders mark the same possessive relationship:
cow farmer = farmer cow → a|the farmer's cow
Imagine now a particle or affix that goes somewhere, and inverses the possession:
cow e farmer | cow-e farmer | cow farmer-e | farmer-e cow | farmer cow-e | farmer e cow| e cow farmer| e farmer cow
The rule for locating this particle might be anything along the lines of "before the two nouns", "after the first noun", "on the higher-ranking noun", "on the lower-ranking noun", "after the two nouns" or whatever you prefer, really - heck, anywhere in the vicinity of the two nouns could make good sense as well.

However, let's go on with some other possibilities. Let's introduce inverse number on possessive pronouns! This language has distinct singular and plural personal pronouns, i.e. "I" vs. "we", but the possessive pronoun is distinct from both. Let's give it as "oune". Depending on the noun that is possessed, this is either parsed as "our X" or "my X". The number of the noun may influence the parsing, so "oune books" means "our books", but "oune book" means "my book". However, this does not happen with all nouns: "oune shoes" and "oune shoe" both are parsed as having "oune" refer to the first person. Here I would go for an affix that fuses with the number marking of the noun in order to mark inverse number on the possessive pronoun:
-ev is plural, inverse possessor number, -in is singular, inverse possessor number:
oune bookev: my books
oune bookin: our book

oune mother : our mother (c.f. oune mom : my mother)
oune mothers: our mothers
oune methrev: my mothers (including previous generations of mothers)
oune methrin: my mother
We could do other weird things, though: maybe inverse marking for reflexive possession. This would only go for third person possessors.

he sees his sister: sister is assumed reflexive, so "he sees his own sister" is how it is parsed.
he sees hin sister: hin is inverse, so this is parsed as another referent's sister

he sees his face: face is assumed non-reflexive, so "he sees his (the other guy's) face" is the required parsing
he sees hin face: he sees his own face

Saturday, October 17, 2015

Detail #223: A Small Evidentiality System

Consider a language where the verb has congruence for both subject and object. These use the same marker - the subject goes first, the object goes second. However, an evidentiality prefix exists that always goes first. This prefix signifies that the knowledge is as stated by the subject or object. The noun according to which it did happen retains its congruence marker, and the other marker is lost. The noun according to whom it happens is also always topicalized, and therefore fronted.

Friday, October 16, 2015

Sargaĺk Invariant Nominal Suffixes

One of the conlinguistic historical mysteries of Sargaĺk is the origin of the invariant suffixes that some nouns take. Most of them are very short, and in fact Sargaĺk itself carries an example of such a short suffix. Below is a table of Sargaĺk as inflected in different case forms to demonstrate the effect of the invariant suffix (which does have morphophonological variation, however):
nominative: sarga-ĺk
accusative: sarga-ta-ĺk
pegative: sarga-tat-ĺk, sarga-tat-ok
comitative: sarga-mai-ĺk, sarga-mai-jk
locative: sarga-ru-k, sarga-ru-ĺk
ablative: sarga-tsa-ĺk
lative: sarga-rne-ĺk, sarga-rne-jk
A few other examples are: widow, tam-ra, forest, dun-sus, cause, reason, pin-hac, life, sin-sus, time, baus-hac, limp foot, nic-hac, clan, family, sask-ra, breast, vun-tak, breath, kime-hac. Counting all dialectal forms, the full list runs at about 95 entries, and the set of suffixes contain -ĺk, -sur, -hac, -tak, -ra, -ver, -savn, -ark - in order of how common they are. Finally, there's four suffixes that each appear in one word, and they are somewhat peculiar. More on those later.

There does not seem to be any strong connection between these suffixes and morphological gender, nor does there seem to be any strong semantic reason to consider them derivational affixes - although -ra often appears with terminology that is related to clans and families and procreation, and -hac often seems to be some way of turning abstractions or qualities into things - nic-hac seems to be related to nikar, limp, and kime-hac to kimede, life, spirit.

Finally, we get to the exceptional suffixes. The suffixes -apar and -sindas are bisyllabic and have an asyllabic stem to which they are suffixed: ĺ-apar, sand bank,t-sindas, traditions of the village, also attested as k-sindas in a few dialects.

The final exceptionalal suffixes are -kartuk, birch sap harvest and -bersan, border between clans' claims which share the prefix ε-, viz. the empty string. Thus, one could just as well say that for these two nouns, case affixes are exceptionally prefixes. Thus, the case table for kartuk is:
nominative: kartuk
pegative: tatkartuk
comitative: maikartuk
locative: rukartuk
ablative: tsakartuk
lative: ŕnekartuk
Finally the question of historical origin of such a system remains. There are some tantalizing phonological similarities to the secondary cases in Bryatesle, and there are semantic clues that these may be cognates - i.e. the words that use them seem to be such that they could often have occurred with the relevant suffix. Another clue is in some old folk songs, where these suffixes do appear on other words, and these words may have different suffixes in different verses. The origins of -apar, -sindas and -bersan are probably a secondary development after the split from Proto-Dairwueh-Bryatesle. This analysis would require that the Proto-language from which they all originated had a secondary case system in place, and that it was lost in Dairwueh, as well as in Sargaĺk, with these few suffixes being a fossilized remain of it.

Wednesday, October 14, 2015

Detail #222: An Extension of Switch Reference

Let us consider two different situations in a switch-reference situation: the referent may be a noun in the same speaker's utterance, or it may be a noun in another speaker's utterance.

This is an interesting difference and we could possibly do something with it. 

For reference to the same subject as the previous speaker's most recent verb referred to, you'd use the same-subject marker and some kind of second-person possessive or whatever.

Tuesday, October 13, 2015

Detail #221: Infinitives and their Article

Some languages, such as English, Swedish and German (to some extent) have markers that sometimes appear before infinitives. Some languages might even have them appear before all their infinitives.

This seems to be a device that seldom is used by conlangers to add any redundancy - how about having separate markers for transitive infinitives, yet like in English not always having the marker appear?

Or how about having it show congruence with the inherent aspect of the verb - e.g. distinguishing punctual verbs such as 'to hit' from more static ones like 'to sit' by different markers, maybe 'to hit', but 'for to sit'?

This doesn't even seem particularly unlikely to come about!

Sunday, October 11, 2015

Detail #220: Some Fun with Negation

Consider pairs of meaning such as "undo" vs. "not do", "be uninterested" vs. "be disinterested", "untie" vs. "not tie", etc – and other negations along other distinctions as well.

Now, we could go for lexically determined negation - i.e. "not do" is more likely than "undo", but "untie" is more likely than "not tie". Therefore "not do" and "untie" both get the same, "simple" negation marker. "Not tie" and "undo" get an inverse negation marker instead, which is slighly more complicated (i.e. maybe requires an auxiliary or something).

We could also have a third and maybe a fourth negation marker that go for negating even more unusual parts.

Another venue where inverse negation could happen is in contexts like this:
I don't like all flavours of icecream(, but I do like some) vs.
I dislike all flavours of icecream
This gets into scope and such, and how the negation interacts with subjects and objects could also be interesting for a language with this property. Maybe the "regular" negation has different syntactical properties from the inverse negation, giving the same semantic sense of negation different syntactical properties depending on the verb, forcing distinct constructions into existence in order to express the same ideas with regards to scope and all for verbs of different classes.

Saturday, October 10, 2015

Detail #219: Ergative-like Topic-Comment Structure

Consider a language where the comment part is introduced by a subordinating conjunction, here presented by 'that':
man that lives in a cave
the man, he lives in a cave
However, if there is no comment, and we're just attracting the attention to a topic, the topic is also marked by this:
that man
a man, the man
However, this seems like a rather restricted thing without much real "interest" or even any way of extending it into having more of an impact on the language. Are topics and comments just too far off from anything we could turn into a system analogous to erg-abs?

I suspect they are, but I always keep second-guessing myself on it.

Friday, October 9, 2015

Detail #218: Inverse Evidentiality

Certain verbs could imaginably have default evidentialities, and it's imaginable that these would not mark the evidentiality for the default one at all, but use some marker for the other evidentiality.

Further, one could easily imagine that different person markings imply different evidentialities for certain verbs! For example, "I snore" might behave differently as far as default evidentiality goes than "you snore" does.

(Thanks badconlangingideas' editor for the idea.)

Sunday, October 4, 2015

Detail #217: Verbs from Numbers

Form verbs, with transitivity markers, from numbers, with the meaning 'take on a certain number'. Greater numbers of course go more approximate:
he fived to fight the men: he took on the men, five in number, in a fight
he thousands to reject objections: he rejects objections by the thousand

Saturday, October 3, 2015

Detail #216: Really Abstract Outline

We're quite used to the notion of "converting" a verb into a noun so we can talk of it as an object. This is useful with regards to describing an action, or talking about its properties. "Navigare necesse est" and all that. However, what if the information in a language was structured in such a way that it made more sense to go the other way?

Certainly lots of languages permit forming verbs from nouns - especially verbs of turning into, making something into, or even being or acting like something. Also, many languages permit similar constructions for time spans and various types of spans (verbs derived from 'holiday' marking 'to spend ones holidays', etc). But that's not what I'm talking about.

I am thinking like turning nouns into verbs for the sake of getting them into the noun-aspect-mood spectrum, or even other similar spectra. Turning nouns into verbs for the sake of marking transitivity or voice things, habituality.

Something like "she father-habitual-passive" : she has daddy-issues.

That example doesn't quite do my idea justice, though. It needs a very much wider range of things wherein actual nominal referents are turned into verbs for a variety of reasons. This, again, is one of these ideas I've been having for years without ever coming up with exactly what the idea is, just a vague notion of in what direction it lies.

Detail #215: Quirky Case Participles

It would be pretty cool to have verbs that take quirky case to have the quirky case carry over to the participles as well. So, e.g. to lack takes an object in the ablative; thus:
I lack common sense-abl
Now, we go on to talk about the common sense which is lacking:
the lacked common sense-abl prevented him from a good career
In most languages, sense would be whatever case prevent takes for its subject in this context. Notice, however, that this may let us drop the idea of active and passive participles for those participles that have quirky case arguments. (If we go and have nominative and accusative covered by this too, we've kind of broken nominative and accusative with regards to normal sentence structure, so let us not go there – feels like if the case system entirely breaks down at that level, it won't survive at the participle-level either.)

Of course, we could have some limited suffixaufnahme going as well - i.e. stacked case markers, one for agreement with the participle, one for agreement with the main verb.

Detail #214: A Complication with regards to Predicative Possession

One of the ways of forming predicative possession (i.e. "I have an apple") is by a kind of locative circumlocution: "by me is an apple". In some languages, the apple is the actual grammatical subject, but in some it's not - quirky case can permit for non-nominatives to be subjects. We'll assume case marking is done by affixes in the language we're making up.

Now, we might want to have a way of forming participles and gerunds and whatnots from the predicative possessive usage, yet we probably don't have any particular voice for it. Normally, a noun that is an argument of a clause will take the case marker for the clause in which it is an argument, i.e. you won't say
"by the something-being man"
when you mean to say
"the man by whom something is"
Of course, if we have adpositions, we could do something like
"the something by-being man"
but since we're doing this by cases instead, we would need something else to carry the case. (Or, we'd need to break the assumption that case pertains to the grammatical environment outside the noun phrase.)

So, we go down the route of having being and been and so on pertain to possession, and we come up with separate solutions for the copula sense of the verb. Here we introduce some nice possibilities: we could have specific lexemes meaning being (participle), being (gerund) and been and so on. We could even derive them in somewhat odd ways: maybe if we have a noun meaning 'capacity (as in in the capacity of ...), role, office (as in a position within some organization), title, stature' we could derive the gerund from that: "roleness", "officeness". The present participle might be derived from the verb in some way - thus-being?

So, at this point we have a situation where:
be.GERUND illicit substances can lead to problems with the law →
having illicit substances ...
And to express, say 'being a nurse is no cakewalk' you'd go for 'roleness of nurse is no cakewalk'

I haven't really dealt with the past tense participles here, but dealing with that is left as an exercise for the diligent reader.

A further idea though: extend this thing to other verbs where you let the non-nominative subjects lead to slightly other meanings. Have various periphrastic constructions form the equivalent of gerunds and participles.

Friday, October 2, 2015

Detail #213: A Cornucopia of Intensifiers

Make adjective intensification a very complicated affair - basically, each adjective has a small set of permissible intensifiers. 

One thing might be to restrict the semantic fields you permit combinations of. For instance, you maybe don't want a thing to be shiningly big or hotly loud or something, i.e. go for rather coherent semantic thing. Also, you may want to avoid seeming contradictions: hotly cold, strongly sick, ...

Now, you let meanings drift, yet keep the associations fixed.