Sunday, January 31, 2016

Ŋʒädär: Some Exceptional Nominal Morphological Patterns

A few handfuls of Ŋʒädär nouns have exceptional morphological patterns going for them. There are three main classes of such patterns:
  • -st'i-/-st'ı- added for plural and obviative forms.
  • -rk'e-/-rk'o- added for all oblique cases and all plural and obviative forms.
  • -t'o-/-t'e- added for plural oblique forms and obviative oblique forms
These three morphemes are all nominalizers; -st'(i/ı)- most often signifies '-ness, -ity'; thus for instance k'oru (dry) → k'orust'ı (dryness), sar (weak) → sarst'i (weakness), iqe (man) → iqest'i (manliness) but also occasionally other meanings: ibik (sleep) → ibix(s)t'i (bed), rügvä (house) → rügvästi (inhabitant). However, it also appears in the plural and obviative  forms of nouns such as mother (yajo), king (kamma), edible mushrooms (sändö), male offspring (gumu), female offspring (t'äne).

The -rk'(e|o)- suffix is used both for verbal nouns and for 'concrete examples' of an adjective, i.e. 'a red thing', or 'a kindness' (this use of -ness in English is somewhat different from the meaning of -ness in the previous paragraph). Many tools follow this pattern, but so do certain vessels, i.e. reindeer or dog sleds.
The -t'(o|e)- suffix often serves the role of turning a verb into an instance of the verb, i.e. q'olku = die, q'olxt'o = a death. However, this also occurs with some animal names, i.e. leading dog (hark) having the plural/obviative oblique stem harxt'o-, reindeer (iseti) having the plural/obviative oblique stem ist'e-).

More nouns of these type will find their way into the dictionary.

Thursday, January 28, 2016

Ŋʒädär: Noun Morphology

Ŋʒädär nouns have case, proximative/obviativeness and number. In the obviative forms, number is omitted.

The cases of Ŋʒädär are the absolutive, the dative, the locative, the lative, the ablative, the genitive-comitative, the instrumental and the complement.

The absolutive takes no suffix. The dative singular takes -m/-im/-um/-üm/-ım. The locative takes -ŋa/-ŋä/-ŋe/-ŋo. The lative takes -lis/-lıs/-lus/-lüs. The ablative takes -line/-lınɤ/-lünä/-luno. The genitive-comitative takes -(h)Vs, where V is a reduplication of the last vowel of the stem. The instrumental takes -(r)Vk, with the same vowel reduplication as in the genitive. The complement is formed by the suffix -ɣuv/-jiy/-jüy/-ɣıv.

All cases except the dative (and absolutive) have very simple plurals - they are formed by inserting the morpheme -üv-/-uv-/-iy-/-ıɣ- after the stem. Some assimilation happens with the locative, giving -ümä/-uma/-inye/-uŋo.

The absolutive plural is formed by taking the full nominative, changing any word-final vowel to the most open vowel at the same position: u, o, a > a; i, e > e; ü, ö > ö; ı, ɤ > ɤ and affixing -r. If no word-final vowel exists, whichever of these four: {a, e, ö, ɤ} that fits the bill is inserted before -r.
The dative plural is exceptionally formed by a separate set of suffixes, viz. -ot'/-et'/-ät'/-ɤt'.

The obviative marker -qi(g)-/-qu(g)-/-qü(g)-/-qı(g)- takes the place of the number morpheme. The only case to have distinctive singular and plural forms in the oblique is the absolutive. The absolutive plural obviative turns the vowel into a widening diphthong with an -n for suffix, e.g. -qien, -quon, -qıɤn, -qüön. The dative also behaves exceptionally with the obviative marker: the suffixes are those of the dative plural and not of the dative singular.

The dative serves more roles than just that of a recipient; for several verbs, it marks the subject, and for some the object. It also marks purposes, causes, sources of information, the time at which something is to happen, by whose opinion something is or seems to be, or by whose perception something is known.

If a possessum would be in the absolutive case were it not possessed, the genitive marks both the possessor and the possessum. In other cases, the possessum retains the case it should be having. Who owns whom is determined by the animacy hierarchy. Reflexive possession of objects is also marked by marking the subject and object of the verb with the genitive case. In fact, whenever the possessum is a subject or an object, there is no need for the possessor to be part of the same noun phrase - it can be flung almost anywhere in the clause. The interactions between possessum, possessor, the animacy hierarchy, and verb argument structure is somewhat complicated. Normally, having a lower-ranked noun own a higher-ranked one is not trivial except when they're subject and object of a regular, transitive verb. A number of strategies exist for this, though.

The instrumental also marks manners, entire time-spans, and is generally used to form adverbs. Peculiarly, it also marks the location of verbs like 'live' or 'reside'.

The ablative marks beginnings of time-spans. Origins, and when talking about distances to things – distances are from a place in Ŋʒädär, not to a place. Distances between two places are from both of them.

Besides the obvious direction and destination, the lative marks ends of time-spans, as well as recurring times in plural forms.

The locative marks place as well as most mental states, with perfective transitive verbs it can also mark destination.

Monday, January 25, 2016

Detail #251: A Register, Taboo and Exceptional Noun Class

In a culture with some rather strict notions of whom it is fit to marry - i.e. probably more strict than 'don't marry close relatives', imagine if the language has a register whereby if you speak to an unmarried, young person whose family relations you are expected to be knowledgeable about, you are forced to use a register whereby anyone forbidden to this young person is in a special noun class.
(I also imagine the culture could have other, non-genetical things that determine permissibleness, for instance, astrological or other similar superstitions.)

After the young person has married, this register is no longer used with him or her. However, the congruence markers can still be used for a few specific meanings: acts of fornication and general violations of the sexual mores of the culture will always have that noun class's congruence marker on the verb. 

Here, the marker also is unique in being the only such marker to permit reduplication; when doing so, it is intensive and judgmental, essentially conveying something like 'you fucking idiot really had to fucking go and fuck your cousin did you?'

Finally, in some minority communities with other taboos as well - including more formalized food taboos - the same morpheme has come to mark violation of any of the other taboos, regardless of noun class of the object, or even with objectless actions that are forbidden.

Ŋʒädär: A Mood Marked by Word Order

Much like in English, the imperative is marked in Ŋʒädär by fronting the verb (and omitting any 2nd person subject). However, this imperative-like mood also serves certain other functions. In a subclause, it serves to mark the purpose for which something was done, e.g.
vär ehi luqu-nta-jut törö-ntä-z ehi p'arŋuba
you (s)he trick-fut-inv get-fut-direct (s)he  inheritance
(s)he will trick you so that (s)he will get the inheritance
This fronting is the sole circumstance in which the negation goes after the (main) verb. Auxiliaries need not be fronted, but if they are, they too follow the main verb. Another use of this mood is 'until' with the future form if the main clause has a present form or past form in it. 
ehi gäbü-lö-: mat'a-nta-: k'ugoru
he scream-past-intr leave-fut-intr bear
he screamed until the bear went away

mat'a-nta-: k'ugoru a ehi gäbü-lö-:
leave-fut-intr bear * he scream-past-intr
'a' serves as a subclause delimiter.

Thursday, January 21, 2016

Challenge #1: My Own Solution?

In the very early days of this blog, I had this idea of posting challenges that people could come up with solutions for. I didn't get much response to them (although I had way fewer readers back then as well). However, in a moment of inspiration, I think I figured out a potential solution to Challenge #1.

Use some type of ingressive or imperfect aspect on the number; thus three-ingr-ptcpl = one that starts the count to three (where of course every 'counting' that this is the ingressive of is counting from the number one below, thus "the second").

Another way could of course be some causative-ingressive combination, three-caus-ingr-(ptcpl) or even three-ingr-caus-(ptcpl): the thing that starts to cause a count towards three, the thing that causes to start a count towards three.

It doesn't feel all that smooth, but it isn't all too stubbly either.

Tuesday, January 19, 2016

Detail #250: Partially Grammaticalized Subjects

Let's imagine a language that at some recent point has rearranged its noun classification. (Biological) Feminines have become distinct from other nouns, and now masculines and inanimates form the masculine gender.

The language has previously not had proper subjects, but is in the process of developing them. Specifically masculine animates are acquiring the status of being "proper subjects". Feminines, however, differ in having only some of the subject properties. (However, they acquire congruence marking on verbs to an even greater extent than animate masculines.) Inanimates trigger no verbal congruence anywhere.

As for congruence, this causes the following situation:
animate masculine agent present? → masculine congruence
feminine agent or patient or topic present? → feminine congruence (overrules masculine congruence)
inanimate masculine? → triggers no congruence
What syntactical things might we expect from these different levels of subjecthood? Let's look at some properties subjects often have. Let us start with reflexive binding. Masculines proper and feminines both have this subject property, but inanimates do not. Thus
She washes self/*her
Heanimate washes self/*him
Heinanimate wash *self/it
 Neuters behave differently with regards to control, so you can't say things analogous to
the house seems to be on fire
it would require
seems that the house is on fire
he/she seems to be doing fine
 would be permitted. Another thing could be that feminines and masculines can be implied, covert arguments. For feminines, this would probably apply both for subjects and objects. A thing that is difficult to make sense of with English examples is resistance to extraction. Apparently, subjects are cross-linguistically less resistant, but in English, perversely, they are more resistant:
Evert thinks that Robert gave Lisa a gift
*who does Evert think that gave Lisa a gift
what does Evert think that Robert gave Lisa?
who does Evert think that Robert gave a gift to?
In this language, "who does Evert think that gave Lisa a gift" would be permissible, but not
*what does Evert think that Robert gave Lisa
or even
*what does Evert think that [gap] is the tastiest dish?
who does Evert think gave Lisa a gift
who does Evert think received a gift from Robert
 would both be permissible, however not necessarily
?who does Evert think Lisa got a gift from
?who does Evert think Robert gave a gift to
More interesting things could be done with regards to subjecthood, but it's late in the evening and I think this post introduces enough ideas already.

Sunday, January 17, 2016

Bringing Your Child up in a Conlang: Long-Term Commitment

A really stupid thing many American school districts apparently do, is to require that people stop speaking their native language with their children and switch to English instead. The mistaken thinking that goes into this probably includes something like
learning a language is great effort, so learning two languages doubles all learning efforts in all subjects
i.e. a kid who is learning maths needs to learn maths twice to master it in both languages. We do know this to be mistaken, and we even know now that bilinguals potentially do better than non-bilinguals. Maybe talking about the same things a bit in more than one language by random chance will resolve potential ambiguities?

It turns out that this common policy not only is mistaken, in the sense that it does not solve the problem it sets out to solve, but even worse, in that it often creates new problems. One of these is that bonding between parents and children can be impacted negatively if the language they use is replaced by another language.

Thus, if you decide to exclusively speak a conlang to your child - which basically is what you need to do if you want the child to learn it - you have to be ready to go in there for the really long run. 

Are you sure your love for your own hobby is that strong? Are you sure you're in for such a commitment? A child is a big enough commitment on itself, adding this to the equation seems rather heavy.

Backwards with Idea #246: Some Ergativity

Post #246 originated upon reading a post that had been left to die in the drafts folder. Basically, the last clause was a cue to myself to come up with something. This was finally done about 120 posts later, and to some extent, this post in some sense only serves as a window into how ideas develop for me, if that interests anyone. So here goes, a peek into the creative process:

Getting back on the ergativity track, we can imagine a language that otherwise is very nominative-accusative, but where infinitives are ergative in alignment:
I see you = I see you
to see you =  you to see
seeing you = you seeing
to see (subj: you)= for you to see
seeing (subj: you) = your seeing
However, we could complicate this a slight bit.

Saturday, January 16, 2016

Imperatives and Exceptional Imperatives in Dairwueh

I decided to include imperatives in Dairwueh for a variety of reasons. Here's a few features:
A number of verbs have slight suppletions, c.f.
ak! (sg), aksu! (pl)
- go (inf)
ral! (sg) rallu! (pl)
- hurry
ten! tessu!
vedeu -
help (inf)
vir! virru!
vesu - to listen

sim! smu!
gesem - tell, say
Some verbs drop part of the root:
bar! (sg), barru! (pl)
barinə - to be careful, to get out of the way, to be careful with something
dig! dixu!
endixu - stand

rud! ruxu!
berugu - leave a place

knu! knuvu!
aknuvu - keep, hold

Most roots lose any final vowel, and apply some morpheme out of the set {-u|-su, -ak|-su, -ar|-sur, -g|-xu}, where the left element is the singular and the right element the plural. The morpheme is lexically determined but tends to follow this schematic:
verbs of states: {-ar, -sur}
verbs of movement: {-ak, -su}
verbs of increase, and of stature: {-g, -xu}
other intransitive verbs: {-u, -su}
transitive verbs of applying a state to something: {-g, -xu}
transitive verbs of movement: {-u, -su}
other transitive verbs {-g, -xu}
Usually, the imperative is fronted, but this is not fully mandatory. There are also some syntactical complications. The main syntactical complication are 'imperatives with overt subjects', where some subject other than the addressee is syntactically in some sense the subject of the imperative. One of these is 'san|sannu', signifying 'look'. It can take a third person "subject".
kreruš san vorge
timberman look! strong
look how strong the timberman is
This might seem as an object, but it turns out that these quirky imperative-subjects can be coordinated as subjects of other verbs:
kreruš san vorge ke darav xogebars doŋbat-ta
timberman look! strong and works harder than smith
It also passes some other subjecthood tests, and therefore, peculiarly enough, qualifies as a subject despite the actual subject of the imperativey part of the verb is the second person.

Thursday, January 14, 2016

Detail #249: Nouns and 'Possessive Valency'

Sometimes, nouns can be imagined to have two distinct types of possessors. As examples of this, we could mention quite a few types of emotional states:
my fear of spiders
the police officer's love for donuts
These obviously correlate closely to the arguments of verbs - essentially these are zero-derived verbal nouns, and we find similar things going on for other types of verbal nouns as well. But there are also certain other nouns that can have a similar sort of transitive possession going:
their source of water
their day of rest
their museum of arts
We could of course imagine a language where such transitive possession was much more common, especially if also some kind of ownership for someone elses use was deeply rooted in the culture, giving us things like
her my house (the house she owns, and I use)
her his son (the son she has by him)
their my task (the task they have given me to perform)
your my debt (the debt I have to you)
your my letter (the letter that I sent you)

We can of course imagine ergative-like or nominative-like alignments for such possession, or even inverse or anything, basically. However, one idea that seems quite natural to me would be to have the marking be duplication of the genitive marker.

source of theirs of of water
their water's's source
Maybe only pronouns distinguish the two types of possession morphologically, and word order serves to distinguish the two?

One interesting thing here would also be how interrogative pronouns are dealt with - identical pronouns for both types of possession? Do certain nouns always require dual possession?

Wednesday, January 13, 2016

Detail #248: Congruence Failure for non-Agents and non-Patients

Really short idea: have verbal congruence be less mandatory for subjects and objects when they are not agents or patients. (In the case of passives, maybe patients as subjects should require congruence, but the other way around could be cool as well.) Subjects and objects can be other roles than agents and patients, e.g. stimuli (what is perceived), perceiver, recipient, instrument, etc, and many of these can appear as subjects in many languages, e.g.
this key (instrument) opens any door
I (perceiver) hear you
the music (stimuli) blared from the speakers

To make it a bit more interesting, have objects that are either recipients or patients require congruence.

Sunday, January 10, 2016

Ŋʒädär: Voices

Although one would think the inverse system would, in some sense, make passives superfluous, c.f.:
man bear kill-inv: the bear kills the man, the man was killed by the bear
man kill-dir bear: the man kills the bear, the bear was killed by the man
It turns out this is not the complete story, for a variety of reasons. In Ŋʒädär finite verbs, there are two possible voices: active and passive. Participles are a bit more complex: they have agentive, patientive and transitive.

The active voice is obviously the way the verb usually works. The passive is formed by introducing the morpheme -sta-/-stä- after the direct/inverse marker. Direct and inverse marking still appear, and there are complications about. In Ŋʒädär, the subject of an active verb can be demoted to an oblique with a passive verb, marked by the oblique instrumental case. The 'alignment' of the verb will depend on where the new subject on the hierarchy with regards to dative subjects. Left of dative subjects, you have direct marking, to the right of it, you have inverse marking.

With participles, the agentive voice expresses an agent, the patientive a patient, and the transitive has both an embedded subject or object, and is an attribute of an external subject or object, thus:

The patientive an agentive voices have, however, the same morpheme marking both of them, viz. -ran/-rän, and whether the head is the subject or object of the participle is marked by the inverse/direct morphemes as compared to the rank of dative subjects. One notable dividing line is that between proximative and obviative third person animate, with the obviative ranking lower than the cutoff point.
galʒa-z-ran iqek
help-dir-ptcpl man
helping man
galʒa-jut-ran iqek
help-inv-ptcpl man
helped man

galʒa-z-ran iqe-qi
help-dir-ptcpl man-prox
helped man(obviative)
galʒa-jut-ran iqe-qi a helping man(obviative)
help-inv-ptcpl man-obv
sikiä-z-rän hark - a leading dog
sikiä-jÿt-rän hark - a led dog
sikiä-z-rän hark-oqu - a led dog(obviative)
sikiä-jyt-rän hark-oqu - a leading dog(obviative)
Oftentimes, an obviative noun will be placed before the participle to decrease the risk of mistaken parsing, but the order given in the above examples is not infrequent either.

The transitive participle has a direct object or a subject, which generally is marked with the dative-genitive. 
imb-im galʒa-z-up igek
child-dat help-act-ptcpl man
the man who helped the child, the child-helping man

imb-im galʒa-jut-up igek
child-dat help-inv-ptcpl man
the man who was helped by a child, the child-helped man
The above example also shows that within steps of the hierarchy, there can be sub-hierarchies. These sub-hierarchies, however, are somewhat verb-specific; in other words, generally speaking, two verbs will not have the major categories of the hierarchy cross any lines, (with some very few exceptions), but within the third person, for instance, there may be differences with regards to noun phrases depending on the verb that is used. Galʒa, for instances, assumes that the more experienced, strong, powerful, influential or closer to 'ideal adult male' the subject is, the more likely it is to be the subject.

Another detail is that a participle might move a noun up or down within its rank - i.e. an animate noun that is also the agent of a participle, is more likely to be the default subject than any other animate noun (even over the obviative/proximative divide), whereas the patient of a participle sinks a bit in rank.

Saturday, January 9, 2016

A Ćwarmin Anecdote

Ćwarmin humour, much like ours, is often in a sense codified as 'jokes'. However, jokes are not told in isolation nor are they presented like 'hey, did you hear the one about the ...'. Instead, the better the anecdote is sneaked into just any discussion, the more the smoothness of the joke-teller will be appreciated. Here is one example.

olbanco  murutuka Taśum karn-utoś mərə-in-i, 
that-compl. Muri--acc-in Taśum long-gen travel.past.3sg
that's like when Taśum from Muri village travelled so far

[that] (on) his return

un seben ragan taxugar-un-uv
him.dat right left [swap places].past.3_pauc
his right and left [had] swapped sides
This illustrates one aspect of Ćwarmin humour: short absurdist statements. It also illustrates another use of the nominative complement, viz. particularly with the demonstratives olba and arna, it can serve as a sort of full statement in itself, basically "it is thus/it is so/it is this". In this example, it's a bit more abstract, viz. "it is like with [...]", the following utterance being the referent to which the pronoun refers.

This particular joke has also made the rounds in Sargaĺk villages, in this form:
ot'a sin miv-tsa Muri-tsa Tasum t'obas t'obas,
yea*, how** village-abl Muri-abl Tasum far far travel.past_prfct.3sg
that's like Tasum from Muri travelled very far
t'oba-tsa t'orme-an-u
far-from return-past_prftct.3sg
having returned from far away

limi-ta is kanra merenr-an-u
left he right replace.past_prfct.3sg
left replaced right for him
* cheap translation; ot'o serves to introduce statements of comparison or contrast or even surprise
** 'demonstrative how'

Friday, January 8, 2016

Ŋʒädär: Inverse Alignment with Instances of Quirky Case

Ŋʒädär has a typical direct-inverse alignment with some small twists. We shall first look at the basic situation, however. The hierarchy is as follows.
1st singular > 1st plural > 2nd singular > 2nd plural > 3rd animate proximative > 3rd animate obviative > 3rd inanimate obv > 3rd inan prox
Thus, a second person subject with a first person object needs an inverse marker. 
vär ehi luq'u-nta-z | ehi var luq'u-nta-z
you (s)he trick-future-direct
you will trick him/her

vär ehi luq'u-nta-jut
you (s)he trick-future-inverse
(s)he will trick you
However, for non-absolutive* subjects and objects, which occur with a number of verbs, interesting stuff happens. The hierarchy gets some quirks
1st singular > 1st plural > 2nd singular > 2nd plural > 1st, 2nd, and 3rd person animate dative subjects > 3rd animate proximate > dative objects > 3rd animate obviative > 3rd inanimate prox > other non-nominatives subjects and objects > 3rd inanimate obv
Thus, for instance, "I have it"
saɤ-ŋa is ot'u-t
me-at it is-(short inverse marker)
A tuplet where we can see the peculiarity of the dative subject location is one of the verbs for "like", jÿŋärä:
saɤ-m is jÿŋärä-z
I-dat it like-direct
I like it

saɤ-m ehi jÿŋärä-jyt
I-dat he/she like-inverse
I like him/her
Similar things happen with quirky objects, c.f. the verb sadɤk'a 'feed', which takes a dative object:
saɤ is-im sadɤk'a-z
I it-dat feed-direct
I feed it

saɤ-m ehi sadɤk'a-z
I-dat he/she feed-direct
he/she feeds me

saɤ-m ehi-qi sadɤk'a-jut
me-dat he/she-obv feed-indirect
he/she feeds me
Thus, for quirky case verbs, the direct/inverse operates as though the obliquely marked nouns were regular subjects or objects that just happen to be in quite different spots in the hierarchy than normally expected.

* I've opted to call the unmarked noun form by the term absolutive in Ŋʒädär.

Tuesday, January 5, 2016

Detail #247: Definite Articles as Derivative Affixes

Consider abstract nouns. For certain of them, definiteness* might not make much semantic sense. Of course, the language might have syntactic rules regarding definiteness that sometimes force a noun to be definite, but let's imagine that this language doesn't have such rules.

Why not have definiteness on these nouns instead be a way of deriving 'proper' or 'good' or 'ideal' X? Friendship vs. the friendship: the first is friendship in general, the latter is ideal friendship. Skill vs. the skill, the first is just ability in general, the latter the ideal ability.

* depending on how definiteness works in your language, a topic to which I have intentions of returning

Detail #246: A Subtle Spot for Sneaking Ergative Alignment In

Let's consider nouns with verbal complements, i.e.
a bit to eat
some way to go
Now, let's imagine that the language has intransitive subjects and objects of the embedded verb marked the same way:
a bit to eat
a man to sing
whether these mean 'singing man/eaten bit' or 'a man with the potential of singing' etc is neither here nor there - heck, these need not even really be part of the noun phrase, they could be part of a structure such as:
we need a man to sing
we need a bit to eat
even with a verb where 'to sing|eat' more clearly is an argument of the main verb rather than a part of the noun phrase than in this case. However, let's construct the transitive subject (even with implicit object) differently:
we need a man so he can hunt?
we need a man that hunts?
we need a man for his hunting?
we need a man he to hunt?
This would be a rather restricted place for ergativity to pop up, but it seems fairly realistic to me at least.