Tuesday, June 30, 2015

Detail #178: An Indefinite Pronoun (weaving a typological mess)

Consider utterances such as
"Ouch, I got something in my eye"
"Something made a noise, and it kept me up all night"
Now, in some circumstances, the identity of the referent is relevant to the further context (as per expectations that speakers learn while acquiring the language). In some circumstances, however, the identity might be entirely irrelevant - tiny particles you get in your eye, things that make noises, etc.

Such things could have a separate pronoun - one that is simply an indefinite, "antitopical" pronoun, which I will call the 'inconsequential'. Reusing the same pronoun in an utterance of roughly "paragraph"-length would indicate it's the same thing being referred to (but that still, it's identity is not all that interesting, though the continuity of its identity is maintained). Switching the referent to a more topical pronoun requires explicitly stating something like 'itinconsequential was [NP], itregular was' or some other construction along those lines.

Now, what other things could we weave around this concept? Maybe indefinite pronouns do not trigger third person congruence on verbs – but since only subjects (or only subjects and objects) have congruence on verbs, the pronoun has existed for other constituents, and been generalized to work as a subject as well. This could well lead to it having some morphological gaps - i.e. no case distinction between subject and object, and even possibly less. Of course, this might indicate that the language usually does have those cases, at least in the pronominal system.

(This of course is only an indication as to what kinds of questions we ask, rather than as to actually meaningful things about the language).

Further though, this might have some use for other purposes - Navajo has a possessive prefix that is somewhat similar in meaning, which is used to permit using inalienably possessed nouns even without specifying a possessor, i.e. shizé'é: my father, azhé'é: someone's father, a father.

So, this could easily extend into that kind of construction as well. 

As for getting no congruence on verbs, this could lead to a situation where a noun marked as object gets subject congruence (and the object congruence is omitted altogether), so the language essentially forms passives by 
subject: itinconsequential object: Noun, acc Verb: [subj: congruence with Noun]
Having both a case system AND subject and object congruence is not very common, as far as I can tell (some languages of Beringia and northernmost North America excepted), so it might just happen that this is somewhat typologically messed up. Let's instead go for the following solution: itinconsequential and other indefinites lack subject congruence, but not object congruence. In the case of no object congruence being present, the object has been promoted to syntactical subject status (while the indefinite has been demoted to unmarked oblique), whereas indefinite object congruence of course tells us that the subject is the regular noun.

A VP in which one indefinite object congruence is present, as well as an indefinite pronoun - and nothing else - the only interpretation is that the indefinite pronoun is the subject, and the indefinite morpheme on the verb is the object. Thus, "something acts on something" is distinguishable from "itinconsequential acts on something", but "something acts on something" is indistinguishable from "something acts on itinconsequential", and same goes for "itinconsequential acts on something/itinconsequential". The loss of these distinctions does not seem all that worrying, though.

Saturday, June 27, 2015

Grammaticalization Paths: Comparative Construction from Dough

I've been thinking a bit about neat grammaticalization paths, spurred on by hearing of a language where the development of the meaning of a particular word went something like
wood > stick > tool > [forgotten steps] > something like a perfect aspect marker
happened. Similar things probably have happened in loads of languages.

I propose the following path:
dough (noun) > swell (verb) > overflow, exceed, for instance wrt the size of a container
One of the common ways of comparisons to be formed is the exceed comparative, for which wals.info gives the following example:
Duala (Ittman 1939: 187)
nin ndabo e kolo buka nine
this house it big exceed that
'this house is bigger than that' 
So, seems rather possible that a noun for dough (or even more generally, some suspension of gunk in liquid, which then turns into 'dough') could become a verb for comparative constructions.

bibliography:

Leon Stassen. 2013. Comparative Constructions.
In: Dryer, Matthew S. & Haspelmath, Martin (eds.)
The World Atlas of Language Structures Online.
Leipzig: Max Planck Institute for Evolutionary Anthropology.
(Available online at http://wals.info/chapter/121, Accessed on 2015-06-27.)

Ittmann, Johannes. 1939. Grammatik des Duala. Berlin: Reimer. (Not a direct source, but quoted as quoted by Leon Stassen - in this context it is not to be taken as an absolutely certain statement about the Duala language, but as an example of the kind of structure discussed. I find it highly likely it is an accurate description of Duala, but the relevance it has in this post is just as an illustration - even if it were a completely misanalyzed sentence, it would fulfill its function in this context. tl;dr - take it as an example, if you really want to know things about Duala, go to a more direct source than my post.)

Thursday, June 25, 2015

Possession in Dairwueh

Verbal possession in Dairwueh's already been described, but nominal possession has not been described this far. There's a number of small complications, but other than that, nominal possession in Dairwueh is fairly straightforward.

Normally, the genitive goes immediately after the noun,  although sometimes they're to the right of a demonstrative, or to the left of an indefinite pronoun - i.e. DET noun2.gen  noun1 = "this/that noun of gen's", "noun2.gen INDEF noun1" = "no/any/some noun1 of noun2's". The default interpretation of noun1 noun2.gen is that noun1 basically is definite.

Topicalizing possessors generally is done by fronting the possessor, inflected for genitive, and adding a possessive pronoun in front of the possessum. Intensifying the possessor can also be done with a similar

Personal possessive pronouns go before the noun, generally in the same syntactical position as determiners. The same reshuffling with determiners occur as with other genitives.

Dairwueh's interrogative pronouns have a tiny complication, viz. there are distinct feminine interrogative pronouns, although the masculine pronouns can be used for gender-neutral questions. This means if the person posing the question expects the answer to be a female, he will ask with the feminine interrogative pronoun. However, when the expected answer is an inanimate feminine or masculine noun, the inanimate interrogative pronoun is used.


Interrogative Pronouns in Dairwueh and Bryatesle

Bryatesle



animate inanimate
nominative tëm sëm
accusative tna sëm
dative tënk sënk
ablative tnam snam

Bryatesle, despite lacking a genitive, has a possessive interrogative pronoun as well,
erten
which is positioned phrase-initially in the NP.

Dairwueh

Dairwueh has an odd detail going on, viz. it has an actual gender distinction going on in its interrogative pronouns


animatefeminineinanimateplur anim.plur. inan.
nomtumtitisveterasva
acctutatnusvetetarsva
dattuvuttirsinteritsavit
gentudinturasiŋaterinsavit
loc-instrtudertuvatsiŋaterarsiŋa

The possessive pronoun stems are
masc: atov-
fem:- ativ-
neuter: siŋa (uninflected for congruence)
plur anim: atev-
plur inan: siŋa

We can see certain similarities between these two sets, and in fact they are among the most clear cognate sets between Dairwueh and Bryatesle.


Tuesday, June 23, 2015

Ćwarmin: Bryatesle and Dairwueh loans

Due to extensive contact between the three languages, Ćwar min has an extensive set of loans of different ages from Bryatesle and Dairwueh as well as their ancestral languages. Ćwar min, however, differs in certain significant ways from Dairwueh and Bryatesle:
  • vowel harmony
  • fixed initial stress (with loans, however, that is only slowly enforced)
We shall look at words of somewhat varying age.

pəktən < proto-dairwueh pak'təyn "hundred"
bičər "wheel" < proto-bryatesle 'bitars "wheel"
učuśan "plow" < proto-bryatesle ɨketr'sa- "plow"
gukula "viceroy" < proto-dairwueh  'gutkələ- "pay tribute"
sicə "vinegar" < proto-bryatesle sɨl'tse "wine"
Ćwarmin has been fairly conservative over time, with few consonant changes. However, we do find that words have generally changed to accomodate Ćwarmin vowel harmony.
More recent words have not necessarily changed in that way:

cixkan  "write" < dairwueh tsihkal "write numbers (in an accounting situation)"
dunvali < dairwueh dunvali, kingdom
re'sepaŋ < dairwueh re'sepank, criminal
te'buvu< bryatesle te'buxu, cake
In some recent loans, vowel harmony has begun spreading from the stressed syllable onwards (obtaining in some idiolects forms like resepəŋ pro resepaŋ). Stress tends to remain unmoved for a few generations, but since stress is fairly solidly word-initial in Ćwar min, each generation tends to increase the number of speakers who moves the stress to word-initial. Sometimes, they also fix the harmony, sometimes not, so you find idiolects anywhere along the line of te'buvu > 'tebuvu > 'tobuvu or 'tebivi.

Morphologically, words like dunvali often lose their final vowel in inflected forms, thus obtaining forms that are uniform as far as harmony goes for all other forms but nominative:
dunvali - dunvalutu (the kingdom), dunvaluc (kingdom.acc), dunvalututa (of the kingdom).
However, words like resepaŋ, the harmony either follows that of the closest stressed syllable (secondary stress is strong enough for its vowel harmony class to win out), or the rightmost remaining syllable:
re'sepaŋuc, resepaŋamca, ... but also
re'sepaŋ, resepaŋemce
depending on which of these approaches have won out in the particular idiolect spoken by our informant. Both approaches may coexist for any speaker, and may be lexically conditioned or even by sentence-level prosody.

Cancelled Draft: "Barxaw: Names"

I decided not to include the following bit in Barxaw, but will need to redesign it somewhat – quite a bit, indeed. I figured explaining what makes me unsatisfied with it is also an aspect of conlanging as an artform, so here goes. The draft:
Personal names in Barxaw tend to follow a few different patterns. Names often consist of two nouns in apposition  or an adjective and a noun. Names -with a few exceptions - are introduced by the article 'dә́(for men who are married), 'gù' (for unmarried, adult men), 'sé' for women regardless of marital status and 'ní' for children. The article is optional but not uncommon for topics and subjects, and mandatory in all other contexts.

Male names may relate to things associated with power, conquest, rule, strength, mastery, as well as symbols and instruments of such things. However, when names allude to symbols, they usually use synonyms for the relevant things, in order that one not accidentally attract the attention of authorities, and to make it clear whether a person of that name or the authorities themselves are referred to. Certain commonly held beliefs about the interaction of supernatural powers and secular powers are involved in this.

  • Wɛ̀n Érqə - 'right victory'
  • Ásɛ̀p Smó - 'ruling sword'
  • Mŋún Ráx - 'blue vestment' ('wáxé dòr' is the official designation)
Female names come in a few different patterns:
'sister [abstract noun]', or 'sister [symbolic noun]:
  • Evé Dìnaλ - sister of justice
  • Evé Érqə - sister of victory
  • Evé Smó - sister of sword
  • Evé Qiðzà - sister of fertility
'mother [abstract noun]' or 'mother [symbolic noun]':
[...]
As for names involving 'mother', these tend to be more common among aristocracy. It seems sister-names or [...]
I found it getting a bit too formulaic - a typical female name would come out as Sé Evé [actual distinctive morpheme]. That's somewhat too weak. As for symbolic nouns, this means I should probably establish a list of nouns that commonly symbolize things in Barxaw rhetorics - swords, roofs, specific-coloured vestments, etc. Thus, the naming customs in the Barxaw language will remain undecided on for a while longer.

Monday, June 22, 2015

Detail #177: A Highly Restricted Pair of Cases

Let's consider a language that lacks words for "left" and "right", but has a pair of case-like affixes that signify 'left' and 'right'. So
me-left : on my left, to my left, my left.
My left hand would be "me-left hand", not "hand-left", since "hand-left" signifies left of the hand, not the left hand.

This makes it possible to specify immediately left in terms of a thing. Of course, in this culture, you'd assume things also often have an orientation, that helps determining which part of it is its left, and which its right. Ships and wagons are oriented along their usual direction of movement, but houses are oriented depending on if you're outside or inside of the house while talking about it: when inside, it's as oriented while exiting the sleeping room in which the speaker sleeps, when outside, it's as per the orientation of someone entering through the main door.

Of course, not everything has two sides, and not everything has a reasonable orientation by which to deem its left and right, and thus most nouns just don't have this case.

Bryatesle: Postpositions

Not a particularly interesting list, but basically the most common postpositions in Bryatesle:

kajer - within, inside of, under supervision of, among,

peler - under, below, beneath, resulting from, performed by,

denër ­­- behind, after,

dedak - behind, after (obsolete)

pira - without

xera - with, by, according to, from (as far as family origins or such goes)

ribta - against, towards

dyra - from, out of, away from

gyner - along, by

furuk - across, through

fura - throughout, embedded in

furer - reaching, up to, all the way to

furta - in front of

Sunday, June 21, 2015

Detail #176: Quirky Case but for Inverse Alignment, and another idea

Although I've already presented a few interesting potential quirks for inverse alignment, let's try and insert the notion of quirky case into the fray.

In languages with case, quirky case is when a subject (or object) takes an unusual case in certain constructions or with certain verbs. This is different from differential subject/object marking, in that it doesn't really convey any extra information - it's a mandatory part of the construction. Examples of quirky case include German 'hilfen' (to help) which takes a dative object, not an accusative one, or Icelandic 'to lack' which takes an accusative subject.

Imagine that a bunch of verbs inverse the meaning of the inverse marker, for whatever reason. This is not all that far from a quirky case analogy for inverse alignment, but it strikes me as a bit uninterestingly obvious. Let's try something else.

How about ... certain verbs require the argument that is the highest in the hierarchy to have certain markers (i.e. case or adpositions or possibly some determiner of some kind). These verbs only mark the default alignment (so a null morpheme implies inverse alignment). Some other verbs require a separate marker on the argument lowest in the hierarchy, but likewise do mark the default alignment.

Adpositions or subordinated verbs affect the relation of arguments to the main verb, so that an argument that is syntactically bound to a subordinated verb (or adposition) automatically is lowered to the bottom of the hierarchy. Such "syntactical boundedness" is also marked morphophonologically by something like Celtic mutations or the like. 

Friday, June 19, 2015

Detail #175: Something with consonant gradation

Is it possible for a language to permit morphological processes to alter syllable weight by one "unit", but once the syllable weight is altered more than one unit, morphophonology sets in to stop further change?

This would require the language to have more than two syllable weights, which might seem slightly unlikely (although iirc some Finnic languages might qualify?) and there's a chance this violates some universal about how far languages can "count". Could make for interesting morphophonology, though.

Ćwarmin: Mood-dependent clitics

Although Ćwarmin has few morphologically explicit moods, it has a number of clitics that are restricted by "implicit" moods. This means that the mood may be implied by other means without these showing up - but these being there is a definite sign that that mood is what's rolling.

Indicative

-kot | -čet
Strengthens affirmation.

-sis | -sus

Marks surprisingness or relieved doubt - 'he arrived, after all'.

-jri | -vru
Somewhat like 'surely, certainly'.

Conditional

-ŋin | -ŋun
-ili | -ulu
As far as meaning goes these are pretty much identical. They emphasize a constituent, i.e. "if you sell at a loss" vs. "if you sell at a loss" vs. "if you sell at a loss".

Permissive mood

-nir | -nur
-nir/-nur might be seen as a kind of apprehensive permissive marker: go (then, if that's what you want), "okay, do it, I don't care".
-sur | -sir
-sur/-sir marks an almost jovial granting of permission - "sure, go ahead, eat as much as you like!". 

Both of these can go on any constituent.

Negative Permissive

-bət|-bat
 This serves to distinguish certain possible parsings of negative permission - if it goes on the negative particle, it means 'you are permitted not to', if it goes on any other constituent, it marks 'you are not permitted to'.

Alethic 

 -nti | -ntu
 Serves to emphasize an element.

Negative Alethic

-biśe|-buso
Often goes on the negative particle, but can go on the head of any constituent phrase to emphasize it. On nouns, it emphasizes that it is this particular noun or type of noun that cannot be so-and-so, on verbs that it is this particular verb that cannot be carried out with these particular arguments.

Imperative

-ara | -ərə
Intensifies the imperative. It is considered quite rude.

-vars | -vers
An encouraging marker, "yes, keep going on like that".

-ruč | -rič
An adversarial marker, "bring it on". 

A certain related marker

A separate marker that has an origin as a similar clitic is the -vin|-vun marker that can go on third person pronouns. With this suffix, the pronoun becomes a generic person, much like the German 'man' pronoun, or the Finnish passive verb or English 'one' or generic 'you'.

Wednesday, June 17, 2015

A general consideration of language complexity, pt 2

I wrote this post a couple of years ago. I figure I should complete the trains of thought that started in it.

I listed a bunch of 'models' for what grammar is – granted, in rather exaggerated forms – and presented some of their drawbacks. Finally, I hinted at my own favourite model. So, let's look a bit closer at this particular model.

A model of grammar needs to account for the following facts about language:
  • grammar changes over time
  • grammar is learned by new speakers (kids, immigrants)
  • every speaker has grammar
  • even a person who does not know the formal definitions of grammatical terms (and who may not be able to consciously identify something as an object or an indirect object) may very well have a great mastery of the grammar of his or her language.
  • grammar needs no central authority to exist, nor any platonic ideal form.
Now, this can be accounted for, in rought terms, by the following:
  • a new speaker learns by observing the language and creating hypotheses about how the language works. The speaker uses these hypotheses to parse and create utterances. Hypotheses are sometimes updated, but depending on the individual and on what evidence he is exposed to, some mistaken hypotheses can persist throughout life. Oftentimes, one hypothesis is not enough - there may be several different hypotheses about the same thing (for one example, consider a word that means different things in different registers or regiolects). The hypotheses are really 'patterns', and we could basically state a hypothesis as 'this pattern is used when these particular preconditions hold'
  • the speaker's grammar is the set of hypotheses that are in active use in his or her speech and parsing
  • the grammar of the language is some kind of 'weighted average' of the grammars of the speakers - speakers who speak much might be more relevant than speakers who speak little. A person who's taken a vow of silence is a member of the speech community, but will have little effect on the grammar. Calculating an actual coefficient for each speaker is impossible – but even worse, the notion of an 'average' of a set of similar rules is also somewhat ill-defined. 
We observe certain things about this:
  • human brains are fairly comparable. Certainly there are geniuses and idiots, and certainly any population will have a variation between those extremes. However, human linguistic skill is remarkably similar – we don't find populations that simply cannot learn language.
Since most of language learning is not "directed" by a teacher, we can assume that any two brains in any two situations will find similar amounts of patterns – not the same amount, but comparable amounts, at least as far as order of magnitude goes. Of course, there will be individual variation – but as we look at an increasingly large population, and discard such patterns that only few speakers have, we will probably get even closer to the same average amount of grammar for pretty much all speaker communities of sufficient size.

Monday, June 15, 2015

A Diversion: A New Song


Yet another song in an unusual musical scale.

Detail #174: Congruence as Precision-marker with Numerals

In Russian, precision is marked by postposing the number:
ten cars = ten cars
cars ten = about ten cars

Other strategies are possible for the same kind of thing: having approximate numerals derived from the regular ones, having a particle ('about ten'), etc. However, in a language with lots of congruence going on, some exceptional congruence thingy could be a reasonable approach.
ten.PL.CASE car.PL.CASE = ten cars
ten.SG.CASE = about ten cars
However, another possibility could be to actually have no number marking in the usual construction:
ten.SG.CASE car.SG.CASE = ten cars
ten.PL.CASE car.SG.CASE = about ten cars
The latter is not all that far from how Finnish constructs 'tens of', 'hundreds of', etc - although with those, you get
ten.PL.CASE car.PL.CASE
whereas Finnish usually - with a few exceptions - does
ten.SG.CASE car.SG.CASE
(except with nom / acc where ten.SG.CASE car.SG.PART happens). Finnish also has its plurale tantum words - words without singular forms - take plural congruence on its numbers, so '(exactly or approximately) ten parties' is constructed the same way 'tens of cars' is.

Another could of course be to have an indefinite (~singular) article precede the number:
a ten cars
This actually happens in some colloquial varieties of Swedish.
 

Sunday, June 14, 2015

Ćwarmin: Conditionals

Ćwarmin has three different ways of forming conditional sentences. The first way is fairly similar to the English way, and are used whenever there is a difference in tense between the protasis and apodosis.
uś [...] ćen xaukunc (bec) keləmce
it  [...]   if  read-2sg(past) you know-recentpast
if you had read it, you would know/have known
Ćen is the conditional particle, pretty similar to English 'if' in function. It can also be used with a verb phrase as such to indicate irrealis mood - with second person verbs, it tends to be a mild imperative, with first and third person it conveys an optative sense. It usually goes before the verb, or as the final word in the conditional subclause. It can go in the wackernagel position from the other end of the subclause as well:
uś ćen [...] xaukunc bec keləmce
There is no word that would serve as a "then ...", so the apodosis cannot be introduced by any particle. Often, the apodosis will be introduced by its subject or the verb, which is somewhat unusual as the verb usually goes last.

The two other constructions are each other's opposites. In the one, the verb of the protasis is an infinitive marked with the genitive, and the apodosis has any tense whatsoever. The subject of the conditional verb too is in the genitive.
bacak xaukutoś (bec) keləmce
you-gen reading-gen you knew

In the other, the conditional is a regular finite clause, and the apodosis is subordinate - and formed likewise, but with the verb in the singular reflexive accusative - again, the subject of the infinitive in the genitive (if it's the same as the subject of the main clause, it can be omitted):
bac xaukunc (becək) kelsin
you read(past) your (own) knowing
This construction doesn't really have a good English analogue - the closest would perhaps be
you have read it, so you know (it)
But unlike the English construction, it's not indicative - it is parsed just as "irrealisly" as it would have been, had it been the previous construction instead.

The tense of both of the verbs are parsed as the same as that of the finite one in both of these constructions.

Saturday, June 13, 2015

Detail #173: Implicit 'other'

Certain nouns might so often appear in collocations with 'other', that a grammaticalization process might create some derived nouns. In this situation, for noun X, X.affix simply signifies 'a/the other X', the referent of which of course depends strongly on context. Most nouns in the language do not have this form, but rather form 'other' with a separate particle, not with an affixed morpheme.

However, meanwhile maybe some former synonyms also come to work as such 'other'-nouns, and you get pairs of words that historically basically meant the same, but now one of them simply signifies "the other X", in a context where a primary X has been established. Basically, some nouns get a suppletive other-form, now that other-forms as a thing exist.

Of course, this seems most likely with nouns whose referents often appear in small groups and where distinguishing one from another is important.

 

Friday, June 12, 2015

Detail #172: Subjective Adjectives

In a language with quite rich verbal morphology - including markers for subject, object and indirect object - certain adjectives require indirect object markers whenever they're used as attributes. When they're used as complements, the copula takes the indirect object (unless the complement use basically consists of turning them into verbs, in which case they of course retain the complement).

The indirect object serves to mark who perceives the thing as having such a quality. So, for instance 'beloved' would always mark whoever it is that loves the thing or person thus described.

Of course, this makes it very easy for us to sneak in some modern rationalist views into a language, and make it certain that its speakers would mark subjective things as subjective (i.e. delicious-to-me food) and thus somehow make the language more 'relativistic' or whatever. This is not, however, necessarily the case - it might just as well serve to make the language carry more coding for social hierarchies.

The set of adjectives having this would include a number of usual suspects - obviously subjective ones - beloved, favourite, beautiful, ugly, sweet, sour, tasty, bitter, stinky, etc. Maybe some adjectives of political or affiliative kind: loyal, radical, various clan/party-based adjectives. Perhaps certain occupations need a subjective marker. Obviously, covert occupations and such may have adjectives that come with this: thieves, spies. Perhaps religious heretics and various infidels that hide among the right believers? Maybe people with unspeakable vices. 

The social hierarchy coding I mentioned in the paragraph before the last one would obviously appear in how someone is loyal-to-me (I perceive him as loyal), or loyal-to-them (they perceive him as loyal), etc; clearly someone I perceive as loyal is also likely to be loyal to me, although I might also be talking of someone who's loyal to my enemy (and my impression is the important thing). Same goes with the various clan/party-adjectives: there might be specific adjectives for loyal-to-hostile-clan and loyal-to-friendly-clan as well as loyal-to-our-specific-clan. 

However, one could also include some adjectives that are not clearly subjective to us: pregnant, adult, permitted, forbidden, broken, whole.


Thursday, June 11, 2015

Detail #171: Ordinals with Orientation

Consider English words such as topmost, foremost, leftmost, etc. It is easily imaginable for a language to have ordinals with mandatory marking for orientation, i.e. each ordinal has a morpheme that encodes the orientation of the set, and which end of the set the element is at. A few examples:

topmostfirst-[vertical, positive]
bottommostfirst-[vertical, negative]
hindmostfirst-[horizontal, away from observer, negative]
foremostfirst-[horizontal, away from observer, positive]

Of course, it's not unimaginable that this marking would be optional on most ordinals - maybe only first, second and third mandatorily mark for this, or maybe only the first ordinal in some kind of pragmatic unit need to specify orientation, and it's free for the rest.

As for the markers of the orientation, mayhap a combination of cases and possessives may work, i.e. "hindmost" would be formed as 'first-out-yours', whereas "foremost" is 'first-to-yours'.

Tuesday, June 9, 2015

Bryatesle Personal Adverbs

There are a number of adverbs in Bryatesle that historically go back to verb forms. These basically go anywhere in the sentence, and almost all of them only come in one form - the largest chunk are former first person verbs, followed by a chunk of former second person verbs - and mostly singular, although not entirely. The translations attempt to catch the idiomatic senses of the usages-

ktat - I hear(d), "allegedly"
gzunet - I hope, "hopefully"
zd̪edmat - I surmise (telic), "probably"
varbet - I long_for, "may it happen soon"
xulnat - I notice(d) (telic), "apparently"
vixret - I say, "for sure"
tkasrat - I disapprove (telic), "alas, sadly, unfortunately"
A few plural 1st person verbs are these: 
galenim - "we praise", as an adverb often used in sentences with religious statements as an intensifier
kulknim - "we remember", often used in recollections of public events and statements, even if the witness restating it is a singular person. Might be best translated as "as made public".
gilsam - we serve (telic), "as commanded", "as you wanted", "as per instruction"

All of these are used as predicates as well, except varbet, which has been replaced by the verb surbret.

For second person verbs commonly acting as adverbs we have
kter - you hear, "for sure", but can also be part of a command structure, an imperative auxiliary adverb, essentially
stumer - you are tired (atelic), "unfortunately, sadly"
xuster - you reach (atelic), "soon"
vildar - you guard (telic), "already"
Finally, a couple of third person verbs: 
bn̪eder - it has sufficient strength (atelic), "it is possible that" 
krima - it creaked (telic), "suspiciously enough"
rupurez - it smells (atelic), "suspiciously enough"

A great variation of similar usages for any number of verbs appears throughout subsets of the speaker community.

Monday, June 8, 2015

Detail #170: Possessive Comitative Constructions

To me, it is intuitively quite clear that the comitative is a context where a distinction between reflexive possession and other third person possession is somewhat more likely than in other oblique positions. In the spirit of that intuition, let's go for a peculiar construction for reflexive possession!

Let the comitative be marked by an adposition. When there's a reflexive possession going on between the subject and the comitative, i.e.
he1 went there with his1 wife
the language marks this by having a passive marker on the comitative adposition, thus
he1 went there with.PASS wife.
The alternative construction
he went there with his wife
would parse as
he1 went there with his2 wife .

A Discovery in Personal Data-Archaeology

I thought it had disappeared entirely, but ... finally I have rediscovered it! About ten years ago, sano made a script and font for Bryatesle.

Now, it just so happens that embedding fonts on blogger is not an entirely reliable technology. Still, I've had it working once (on firefox), and I hope chrome's gotten up to the task as well.

Sunday, June 7, 2015

Detail #169: Present Participle Morphology on Nouns

A language with zero copulas could easily take present participle morphology and put it on nouns used as complements whenever they denote a temporary or non-essential quality:
man captain-PCPL = the man is a captain
man cousin me.dat = the man is my cousin
The latter is a non-temporary quality: being cousins is mostly a fixed quality. Let's extend this idea a bit further: certain nouns that are rather unchangeable can be intensified:

she mother = she is a mother
she mothering = she has recently become a mother, and thus has more mothery obligations than usual

We further extend this so such nouns turn into regular nouns.

Saturday, June 6, 2015

Detail #168: Split Quirky Case / Split Differential Object Marking

Quirky case and differential object marking can of course be "split" in the sense that split ergativity or some other similar system can be split.

As for split DOM, Baltic Finnic sort of has that already: only non-negative verbs distinguish telic from atelic objects. However, seeing the negative verb as the 'typical', non-split context seems weird, and this might be why it's not reported as a split DOM all that much. 

As for split quirky case, that would be all that more interesting really - some verb that can take quirky case subjects/object only does so under certain circumstances - and picking the non-quirky subject/object indicates something as well in that circumstance. This makes it slightly more lexically restricted - given that quirky case generally signifies a lexically restricted set of (usually) verbs that require an odd case for their subjects or objects.


This can of course also go on for participles and the like. As an example I'll use a 'fake' Finnish where 'to have to', viz. 'täytyy' has Split Quirky Subjects (in real Finnish, the subject is invariably in the genitive). Let's also imagine that this split only happens in the transitive:
minu-ntäytyysyödä
1sg-genmust(-3sg)eat
I
musteat
Note that the "3sg" is a sort of 'placeholder' - impersonal and third person verbs are marked the same in Finnish.
Let us now imagine that in past tense, there's a quirky subject thing going on, where the subject can be nominative as well - and this affects verb morphology too:
minä
täyty-i-nsyödäruoka/ruuan*
I
must-past(-1sg)eat(telic)food
I
had tofinish eatingthe food
BEWARE: the above sentence is malformed Finnish; Anyone using google to find attestations of nominative subjects with this construction beware!
minun
täyty-i
syödäruoka
1sg gen
must-past(-3sg)eat(telic)food
I
had tofinish eatingthe food
What kind of a distinction could we have by this very restricted quirky subject split? Maybe something about how the obligation came about - or perhaps even more interesting, "minun täytyi" might imply that this was part of a causal chain with relevance in the presence, i.e. "I had to finish eating the food - and that's why I am late", whereas "minä täytyi" lacks that implication.

* The Finnish object is funny - whenever there's no nominative subject, a singular telic, non-negative object is in the nominative, otherwise in the genitive. (In the plural, it works slightly differently: the nominative and the accusative are identical, and thus "ruoat"/"ruuat" would cover both options.)

Thursday, June 4, 2015

Detail #167: A musing on noun morphology

Sometimes, languages have different combinations of things in different parts of the verbal paradigm. For instance, the Russian past tense verb has congruence for gender, but does not mark person in any way. Meanwhile, the Russian present tense verb marks person - but does not mark gender in any way.

Is there perchance some way of conflating marking in some clever way in the nominal morphology? Obviously we could go for a very simple solution - have core cases distinguish gender in their morphology while non-core cases don't. However, this is only maybe the most obvious such conflation.

Let's go a bit further afield and try to come up with something a bit weirder. How about ... case, definiteness and number marking gets lost on nouns whenever they are possessed and whenever they are possessors. (Note: I assume this conlang uses case a bit like German or Ancient Greek, so not too excessively over-reliant on it.) However, when possessed, there's an explicit gender prefix as well as a prefix that codes for whether the possessor is the noun higher or lower in the animacy hierarchy (possession is done by simple apposition, and both orders go). For arguments that are not preceded by a preposition, this also triggers the presence of inverse marking on the verb.

Conclusion to the post: not a very good idea, and not as interesting really as the Russian verbal split. It seems things along the line of the Russian split are easy to come up with for verbs (just replace categories that trigger some agreement or under which some other category is permitted - i.e. have negative verbs not distinguish realis and irrealis, or have the future tense merge some aspects or whatever, etc) than it is for nouns (I've been trying to come up with something more clever for days!)

So this is a bit of a challenge as well – please, come up with something better!

Tuesday, June 2, 2015

Detail #166: Slightly onwards with #165

Merging persons for some class of person throughout a language may be fairly unlikely to develop in a real language. However, in a language where possessives and subject/object pronouns form entirely different systems, it could be reasonable for such a difference to exist. Such a system would have even more obvious implications regarding social status and such.