Friday, October 31, 2014

Detail #117: Several Degrees of Definiteness vs. Lexical Distinctions

In a language with two levels of definiteness (definite vs. indefinite), a third level (specific) could be semi-present for some constituents by means of lexicalized distinction on verbs:

I look for a car (any car)
I search for a car (a specific car; I know which one it is, you probably don't)
I search for the car (that I mentioned earlier, and thus you know of it now too) 
Maybe this is done by some change in congruence on the verb - maybe all specific-or-definite verbs have a certain marking, while only definite nouns have a marking, or vice versa. Or maybe only certain verbs where the distinction is felt to be significant enough have pairs where the difference is meaningful. I would imagine the pairs would tend to look similar but not have any regular formation going on.

Detail #116: Some Unusual Things to Grammaticalize

Distinguish a few types of relative size of subject to object with regards to kinetic verbs! (May be distinguished by differential object marking? Or by some verb marking? Or by a weird auxiliary?)

I.e. 'John harvests(small-to-large) the field', 'David killed(medium-to-big) Goliath', 'The villain kicked(medium-to-small) the puppy',

One could also perhaps grammaticalize the amount of premeditation of first person future verbs by some way.

Detail #115: Local Cases

This is basically an idea that takes what is sort of present already (to some extent) in Finnish and puts it on steroids.

Imagine a case system of about six local cases. In Finnish, these are conventionally arranged as {on/by, in} * {towards, at, from}. However, let us not give them all that clear names yet.

Let's rather go by {A, B} * {1, 2, 3}. Now, we have this set: {A1, A2, A3, B1, B2, B3}. And now for the fun: let directionality of the different numbers vary with the verb and its aspect!

Type 1 verbs have the perfect aspect cause the cases to map thus: 1 → origin, 2 → location, 3 → direction (and A vs. B maintain the distinction 'in' vs. 'at'/'by'/'on'). The imperfect aspect, however, gets A1 → entirely unused, B1 →direction, A2 → direction, B2 → location , A3 → location, B3 → origin ... and the 'in' vs 'at'/'by'/'on' distinction is only partially preserved, where i.e. B3 conflates origins at/by/on or in the noun thus marked, and so on.

Of course, these mappings may be specific to classes of verbs, even unique to some verb (and most mappings may actually ignore most of the cases entirely), and the mappings may conflate things in various ways - sometimes, location and origin, sometimes location and direction, sometimes in and on/at/by, sometimes maybe even part of the on/at/by gets transferred to in and vice versa.

In part this is an exaggeration of case and adposition usage differences in reality - in some languages, a painting hangs on the wall, in some it hangs in the wall. In some, the mail slot in your door is on the door, in some it is in the door.

In English, the phrasal verbs likewise form a complex of preposition-verb pairs where the meaning is not trivially related to the preposition's usual use, and we can find languages where aspect does change the sense of location/direction - if you were heading somewhere, perfect aspect sort of implies you've gotten there somehow, and if you are coming from somewhere, you have been there.

Grammaticalizing and having a bunch of complications with regards to this is no big stretch, I imagine.

Tuesday, October 28, 2014

Detail #114: Aspect Shenanigans

Imagine a language that distinguishes punctual from non-punctual, but which exact type of punctuality is being discussed is not normally specified - so perfect and inchoative are marked the same in most circumstances - although there may be optional periphrastic ways of specifying further aspectual distinctions.

Of course, this might seem somewhat unrealistic, so how about restricting having only this particular distinction to only, say, irrealis moods?

Detail 113: Imperative as a Case

Imagine a case that marks out a specific noun as an answer to a question or as a suggestion or order.

Which one should I take? That.CASE.
 Would you pay your debts.CASE.
Should I bring the red or the white wine? White.CASE.
Now, this could easily develop into a situation where nominal verbs marked with such a case serve as imperatives.

On Bringing Kids up in Conlangs (Index, Introduction and statement of purpose)

Time to invite some controversy, I guess.

During my stints on the facebook conlang group, one topic that has regularly popped up has dealt with children. Specifically, whether there is anyone teaching a conlang to their children.

Few do this, but it is, I feel, my obligation to present an argument why conlangers should avoid doing it - there should be even fewer doing it out there than there is. There are several partial arguments to why, each argument contributing different reasons to avoid it. There are arguments from a number of fields - just ethics in general, sociolinguistics, psycholinguistics, language acquisition-related, etc.

So, let me present to you the several partial arguments why you should not teach a conlang to your child.

2. You don't speak your conlang natively, reducing the ability to bond with your child

3. Introducing a child into a situation with an artificial lack of a speaker community to support its language learning seems rather peculiar (notice: this differs from teaching a child your native language if your native language's speaker community has dwindled away, due to reason #2)

4. Your child's language acquisition could be used for something more useful

5. Your child's language acquisition skills may be somewhat affected

6. Your child's hearing may be affected

7. Your child is not yours to do what you want with, a kid isn't an art project - unlike your conlang

8. Being an example to other conlangers, so as not to inspire linguistic ignoramuses from doing this in potentially harmful ways even if you could pull it off in a harmless manner

I will elaborate on each of these, probably giving each a post of its own, with literature to back most of the points up. However, it will be a slow going project, as I will have to fish up some sources - literature I have read years ago, etc. Other partial topics may appear as well.

However, I ask people who object to my reasoning to wait until the relevant post before responding with their objections.

Monday, October 27, 2014

Plans for the next twelve months

This blog is just about to turn two years. I figure presenting some plans for the incoming year may be a good idea.

So, here's the main things I will strive to develop:
  • More reviews of linguistic works; this far only one book has been reviewed, although I intend to write a review soon of Newmeyer's Possible and Probable Languages.
  • More pragmatics, and especially I'd like to get into information structure, a topic that I am not particularly knowledgeable about but want to get into. Information structure-related concerns seem to have potential for some great conlanging ideas.
  • More details from real-world languages, mostly presented in relation to original ideas. Basically as a way of introducing some real-world typology into the blog.
  • More Dairwueh, Barxaw, Tatediem, Bryatesle and Ćwarmin. (For the simple reason that they're getting to a point now where they are solid generators of ideas.)
  • Some observations and thoughts on the analysis of structures in conlangs (and structures in languages more generally), especially with regards to typologically uncommon (or even unattested) structures or absences of structures.
  • Some light controversy. I intend to write a detailed argument regarding bringing up children in conlangs.
  • Presenting a model for linguistics, and analyzing its benefits and drawbacks as well as some more general implications about linguistics. Not going to have any controversial stances there, so no attempt at a Chomskyan revolution from an amateur linguist, but rather an attempt at understanding what linguistics is about from another angle.
So there, some things to look forward to in 2015 (as well as the two months left of 2014). If I hadn't written this out, it's possible I might've forgotten some of these ideas or not just bothered to carry the ideas out. Now, I am at least somewhat committed to writing these things.

Saturday, October 25, 2014

Detail #112: A distinction in interrogative adverbial pronouns

Imagine a system with an adverb formed from some interrogative pronoun that signifies 'in what direction'. This is not all that uncommon ultimately, and Swedish, for instance, forms this by combining where and a preposition - but in Germanic style, puts the preposition last. (Strictly speaking, the adverb used is cognate to whither, rather than where.) Vartåt, so something like whither-to.

In a language where absolute directions are understood by the speakers, one could distinguish relative and absolute directions; maybe questions with an interrogative pronoun in the dual number are asked to distinguish left from right, whereas a plural number signifies desire to have an absolute direction.

Alternatively, of course, other means of distinguishing them could be used - imagine an English variety where what direction asks whether to go left or right, which direction asks for a compass bearing.

Of course, a thing that not all conlangers might now is that there are indeed languages where the speaker is assumed at all moments to know the cardinal directions' current, ehrm, direction.

Friday, October 24, 2014

bad conlanging ideas

Badconlangingideas will reach its hundredth bad idea soon. It is a great non-serious take on conlanging, and well worth checking out.

Detail #111: Noun Classifiers and Prepositions

It seems it could be quite possible for prepositions to develop into classifiers.

Consider 'I want five of those'. In some varieties of colloquial Swedish, certain mass nouns seem to sometimes take prepositions when preceded by words like 'mera' (more) or similar, although I currently fail at finding the examples I have seen in recent newspapers (of the Finland-Swedish variety).

One could easily imagine a situation then where different kinds of nouns prefer different prepositions, so you get a situation whereby e.g. more with time, more of salt, four of perch, any by friend, some on chieftain, etc.  Of course, what preposition is used is determined by the noun to the right, rather than by the quantifier (or determiner) to the left.

Thursday, October 23, 2014

Dairwueh: Reciprocality

In Dairwueh, reciprocal subjects (as well as reciprocal objects of causative verbs) can be coordinated with a special conjunction, 'balə'. Thus "Ardil and Kreŋx see each other" turns into "Ardil balə Kreŋx tidir (ekin)", where ekin is an optional third person plural object pronoun, and tidir is the verb 'see' in third person singular. With a single noun in the plural, balə operates a bit like a preposition - in several dialects, it even triggers some non-nominative case when used with a single noun.

A few dialects repeat balə before each reciprocal subject - basically along the lines of "Balə Ardil balə Kreŋx tidir", but with sound changes and other grammar changes involved as well, so the actual dialectal form might be something like /bar aðil balkrəx:tiðr:/

For non-object reciprocals, explicit third-person plural pronouns resolve ambiguities.

The reciprocal conjunction interacts in complicated manners with non-nominative subjects. In a few dialects, the conjunction itself is marked by a case suffix, and the subjects have the 'default marked case', whichever that happens to be in the relevant dialect. This seems to have been a rather common strategy earlier, but now, most dialects affix balə (often in the form -valə) to the verb whenever this situation appears. Finally, another branch of dialects have generalized the non-object reciprocal approach - X.dat(subj) verb each other →balə X.nom verb they(gender congruence with X).dat. This seems to present a nominative subject, but congruence is seldom used, and it would rather seem as though the nominative constituent is just a hanging constituent that provides something for 'they' to refer to.

For reciprocal objects of a causative verb (i.e. 'X makes Y and Z hit each other'), matters are not particularly much more complicated. Causatives have a bit of a differential object marking - genitive vs. accusative, with the accusative implying a lack of volition on the part of the causee combined with a very active and intentional causer.  The conjunction, again, takes a case marker with the accusative, but no marker when corresponding to the genitive. With a plural noun as causee, balə/balar precedes as a preposition, and the noun often has the case it would normally have in that position.

For more complicated things like 'X and Y make each other hate tulips', extra pronouns are added, giving for instance 'X balə Y hate.cause eket tulips.acc', similar to the solution used for non-object reciprocals.

Wednesday, October 22, 2014

Detail #110: Verb Morphology Observations

One detail that seems quite common to conlangs that I guess is justifiedly common is a detail in verb morphology. It seems both nature and conlangers like verbs to mark person gives 83 languages without any person marking on their verbs in its sample, and about 300 with. Granted, of those 300, less than those without person marking exclusively have exclusively subject-marking. Conversely, about 200 mark both subject and object, 24 mark only the object, and six have a system where either the subject or the object is marked.

Anyways, we could maybe try and do something else. Maintain marking with roughly the same amount of different markings, but having some entirely different things instead of person.

Here's one suggestion, and I hope others can come up with other ideas that deviate in other ways:
present transitive, non-past intransitive, non-past existential, present intensive transitive, present intensive non-transitive (includes existential) 
past transitive, past intransitive (merges with past existential), past intensive (merges all)
future transitive, future intensive (merges all)

note: negative and intensive do not combine at all; an auxiliary can be turned intensive, though, and have a negative complement.

Tuesday, October 21, 2014

Dairwueh: A Short Description of the Case System and its Use

Dairwueh has a case system along these lines:
nominative | accusative | dative | genitive | locative-instrumental
The nominative serves a rather restricted set of uses - subjects, nominal and adjectival complements, arguments of prepositions that are semantically similar to complements of copulas, vocative uses, sometimes also topics that have semantic roles that are rather far from anything that could be the subject.

The accusative serves as direct objects, object of a few adjectives and prepositions, subjects of the verbs 'naxmal' (expect), 'rəntigi' (withstand), and 'drumal' ('repeat'), and complements of objects.

The genitive serves as possessors, definite transitive subjects, certain objects (of the verbs 'surdəl' 'to burden with, to be burdened with', 'gəš' 'to disapprove', 'əseaŋl', 'to marry'), pre-prepositional nouns and the objects of several prepositions.

For this purpose, a pre-prepositional noun is a noun that kind of forms its own hanging utterance in a sentence in combination with a prepositional phrase:
Kadŋi ori i sembern, tauhat ma derilŋa
sit-3sgI he at table.dat, beer.gen in glass.loc-instr
Often, pre-prepositionals are body parts, clothing articles or tools: coat on him, hat on head,  knife in hand, hand on oar, foot in air. One adposition only ever appears with both preprepositionals and a regular complement - vas - which basically works a bit like 'being', the complement always basically being a descriptor of the preprepositional. If the subject is a pronoun or otherwise markedly definite the preprepositional is assumed - especially if a concrete noun - to be the possession of the subject or otherwise under it control.

The dative serves as indirect object, object of several prepositions, subject of several verbs, object of several verbs, possessors of certain nouns (generally highly animate ones), direct object of verbs that have an intensifying dummy object.

An intensifying dummy object is a semantically bleached object whose only function is to intensify the verb. These often are grammaticalized expletives:
Šamgito si xaugna mešdar
rule she INTENSE.acc husband.dat - she rules her husband strongly, xaugna originally being the name of a really minor demon in Dairwueh myths.
The locative-instrumental serves to mark place and instrument, but also manner with abstract nouns. It appears with some prepositions and as object of a few verbs. In many dialects it also marks possessums, overruling most of the case rules given above - in most dialects that have this, there's still the exception that subjects still tend to be nominative or genitive when possessums.

Sunday, October 19, 2014

Detail #109: A Preposition Trick

Mark focus or topic marking for preposition phrases by reduplication of the preposition in sentence-initial position:

in I told you he's in London!
with he does his job with pride. 
Obviously this benefits from a language with case marking, so that nouns that happen to follow the reduplicated preposition are not mistakenly parsed as complements - or at least that the likelihood for such misparsing is reduced.

Detail #108: More on Adjectives as a non-class

Let us say we conflate adjectives and nouns. Let us further say we make a rule whereby intensified nouns rank higher than less intensified nouns, and higher ranks go to the left or the right of lower ranks. Let's further use English comparatives and superlatives as were they intensifiers instead of comparison markers for the sake of this post. So with a right-dislocation of high intensity nouns we get.

the old captain
the captain oldest

Now, let's imagine more 'stereotypical' nouns too get intensivity marking, so e.g. boss = boss, bosser = middle management boss, bossest = head honcho.

the old boss, the boss older, the older bosser, the bosser oldest, the boss oldest, the old bossest, etc.

I am not sure if this is realistic, but somehow it feels like the most overtly intensive noun sort of could be attracted to head-of-phrase or phrase-initial or whatever prominent position. 

Saturday, October 18, 2014

Detail #107: Fun with Verb-like Adjectives

Many conlangers get rid of adjectives in one way or another - basically, it is an easy way of adding some level of exoticness, since adjectives as a word class are omnipresent throughout Europe. Thus, conflating them with verbs or with nouns is an easy way of avoiding Standard Average Europeanness. A few languages that fall within SAE range conflate nouns and adjectives, which might be why this option is less often talked about.

There are, of course, two ways of going about this: one is the hardcore way that genuinely utterly conflates verbs and adjectives. At this point, the only difference may be a slight statistical difference - some verbs may be used as attributes more often than others, whereas some might even be almost exclusively used as predicates. The other way is to maintain some kind of slight distinction - adjectives are a subclass of verbs that might have some syntactical and morphological peculiarities to them.

However, most constructed languages fall short there - the grammar asserts adjectives behave like verbs, and that is what we get. We get to know how to use verbs as attributes and as predicates, maybe a few bits about comparison and how to intensify verbs, etc.

However, from the descriptions we get, I am generally left with the feeling that this in a way is window dressing - we still end up using the verb that corresponds to 'red' in a way that closely corresponds to how we use the adjective 'red' in English. What if we could do some violence to the structural similarities, what if we could force some innovative phrasings and ways of structuring an utterance.

An example could be one of the early posts in this blog - 'X has adj Y'  'X adj.pres Y', although you might want to introduce some voice morphology or somesuch. 'X has a red house' → 'X red.voice house'. What happens if there's more than one adjective? "X reds and bigs his house", with the same markings on the verbs, naturally. Normally adjectives will be intransitive (although this needn't be the case, necessarily - some adjectives, like 'fearful (of)', 'bored (with)', might very well turn out to be transitive in some languages, even if that language has a distinct class of adjectives!), so a transitive use could, for some verbs (or even some 'verbs with adjective tendencies') indicate something like 'having an X of the quality the verb expresses'.

What other things could this conflation lead to? I guess one thing could be rephrasings when several of the nouns in an NP share adjectives: the old woman helped the old man → old.3pl, the woman helped the man.

I sort of feel like there could be greater difference in how stuff is pragmatically structured - like a genuine, well-thought out lack of adjectives should have a greater impact on information structure than what conlangs actually deliver.

I suggest this post be read as a pointer rather than a criticism - here are some things to think about, where little work has been done. If you want to break ground, this is one place you can do so. I realize this is a tricky thing to think about - I wanted to come up with way better examples than I managed to, but I am pretty sure I am not the most imaginative conlanger out there, and conlanging should be a way of practicing your imagination anyway!

Friday, October 17, 2014

Detail #106: Weird Modal Shenanigans in Verbs

Imagine a language where modal marking and voices have become somewhat conflated, i.e. verb-passive basically is parsed as some kind of irrealis mood. However, let us consider some possible complications!

Different verbs could have different types of modal meaning in different moods - passive could be jussive for some verbs, permissive for some, a potential mood for some, and a precative imperative for some verbs.

Basically, I am thinking there could be a language where there is one primary modal distinction, and

  • that this modal distinction is different for different verbs - but probably only three or four different classes of verbs or somesuch
  • that this modal distinction primarily is expressed by what previously was a passive voice, and that this leads to the situation where the main modal distinction also leads to syntactical shenanigans.
Of course, other less prominent modes may be marked by more traditional ways that don't interfere with what historically has been voice marking. Possibly, voice marking retains its original meaning when other explicit voice marking is present. A new indicative passive may be forming in some manner. 

One big question that should be obvious now, of course is: how do we get a verb for which the usual meaning of the passive is 'jussive' to mark potential or permissive or precative? Maybe the association with different moods is just a statistic thing - some verbs are likely to appear in contexts where precative is more likely, some are more likely to appear in collocations where jussive is more likely, and thus saying that 'go.pass' is permissive just says a thing about its most common use, and precative or potential or jussive use is actually inferred from the context, but we observe that it most often just happens to be permissive in the language and therefore pretend go.pass really is permissive. Alternatively, there may be subtle differences in use - have a statitical tendency to omit the agent with jussive and precative for verbs that don't default to jussive or precative meanings, have some weakening adverbs or such that make non-potentials potential, have some case marking that can go on a participant acquiring permission to turn non-permissive verbs into permissive ones.

Another idea that could fit with this is using negative passives as a kind of intensive negation.

A Challenge!

Pull off grammaticalization of some expletive so it becomes a regular part of the language. However, do this in an elegant fashion.

Thursday, October 16, 2014

Looking for a thing

I have a recollection that some while ago, I read about a language that had a fairly rich system of tenses - hodiernal, hesternal, and a few pre-hesternal ones, even. However, this language had a complication!

Several of the tenses covered non-contiguous bits of time, some even being used both when referring to some particular bit of the future and some particular bits of the past. 

If anyone happens to know a source that mentions this language, tell me! - or if it's a conlang and my recollection is very flawed, or if I have dreamt it all up some night. Or if someone's been pulling my leg somewhere. If comments don't work, I can be reached at moc.liamg@okkeim (obviously reversed).

Detail #105: Titles

I am almost entirely certain some language somewhere does this already.

A language where military titles, ecclesiastical titles, professional titles, etc all are mass nouns.

Wednesday, October 15, 2014

Bryatesle: Inflectional Noun Morphology (in tabular form)

A previous post on Bryatesle's phonology and a basic summary of its typological features can be found here.

This content may be corrected if I find errors in the tables. These tables have been somewhat updated since my previous Bryatesle grammar was written back in 2007 or somesuch.
First, the regular cases without any secondary cases on top:
mascsgvaried  -ak/-ik/-yk  -e/-ë  -ity/-ïty  -eny/-ëny-im/-ïm
plurvaried  -uku/-veku-(u)mex-(u)rsi-uny/-vunyvarious
neutsg -u, variedsame as nom-yn-ity/-ïty-iny/-ïnyN/A
 With possessive:
nomaccdatablvoc and excl are absent
mascsg-unë  -an  -ar  -ent/-ënt  

plur-uvu -(v)ekux-(v)emxi-(v)emxi

femsg-ela/-ëla-ei/-ëi-ir/-ïr-ing, -ïn


neutsg -unë-unë-ent/-ënt-ent/-ënt


With partitive: 

mascsg-u  -eze/-ëze-er/-ër    

plur-(v)utë -(v)ux-(v)ux



neutsg -ur-ur-yr


 With definite:
nomaccdatablvoc and excl are absent
mascsg-unë  -an  -ar  -ent/-ënt  

plur-ven -ver-(v)emxi-(v)emxi

femsg-në-nyk-ir/-ïr-am, -em


neutsg -es-ek-ent/-ënt-ent/-ënt

 The secondary subject case is formed by attaching fem: -(n)isr, masc: -(n)isr, neut -(n)isn, plur -(n)yx to the regular case marking, the reciprocal object attaches -sus to the regular case marking, and suggestion marking attaches -ki to regular case suffixes.

Sunday, October 12, 2014

Comments issue

Apparently, comments may not work properly right now, I will look into it.

Update: Thanks for test-comments you people have been posting, I have obviously deleted some of them - but I know that at least about half of the comments people have left have not shown up, so something is definitely not as it should be.

Some Ćwarmin Adverbs of Time and other considerations

Here are some adverbs of time used in Ćwarmin, a language intended to be somewhat Uralic-like (although nothing prevents Indo-European, Turkic and various Caucasian influences).

twiǧerćel - formed from the verb twiǧel, participate, with the suffix for 'mandatory action'. Has come to signify time spanning until the regular full village meetings, which serve a ritual function as well as a civic function. When referred to as a noun, the full time span is 'cawxur', which basically serves the role 'week' does in English. A cawxur is usually nine days, with about one in four being ten-day cawxurs, and one annual eleven-day cawxur. It is clearly related to the number nine, 'cawke'. With the determiner 'bax', it signifies something taking place in the previous twihel cycle, and with the determiner 'saŋ' ('up'), it signifies something between the incoming twiherćel and its successor. Jaźe twiǧerćel ('') signifies something spanning a full twiǧerćel cycle. A number before twiǧerćel sets the number of cycles that something will span. A number of weeks spanned in the past is also preceded by awwun - "was". The pseudoparticiple used for the span of an action -enit is not used for this because the action itself only covers a few hours at most (or even shorter where village clergy and leaders are less meticulously religious, and fewer communal issues are needed to be discussed at the civic part).

lentapritaś - literally milk.obligatee.gen-sing-specific "of him who is obligated to milk (cattle)", a poetic name for 'in the early morning', appeared in the Ćwarmin analogue to the Romantic era. The morphemes involved are lentek (a verb for harvesting animal products except meat - eggs, wool, milk, blood, but also the meat of snails, frogs and small fish), -apir (noun marking person obligated to do an action), -itaś (specific genitive singular).

śpanit - in the nights, in night-time, at night derived from śpal (sleep) + -enit. A generalized adverb of time, so not when speaking about a specific night, which uses the noun kojom, night - often in the accusative form kojuc. (The root is irregularly shortened, or rather the nominative has an unusual -om suffix).

jirune - soon (time.dative)
merg - often
ćimij - oftentimes, ćiwuru - sometimes. These are verbs, ćimid has third person plural inflection, ćiwuru is paucal third person. Ćimil basically means 'do, happen, act, perform, make, make happen' - happen for more abstract subjects, do for more concrete subjects. As a regular predicate verb it most often appears directly after a subject or object, but as an adverby verb it either is sentence initial or goes after the main verb. When used as a main predicate it almost never drops its subject pronoun.
ćimamce - just, a moment ago, recently. Not commonly used, as the verb itself can be inflected for recent past. However, with verbless constructions and as an answer to when something happened, this is a common lexeme.

Further, due to its rather wide span of meanings, it's really a rather ambivoiced verb - "I did" and "it happened/got done by me" are both quite similar in practical use. Since the immediate past makes no person distinction, it's quite possible for there to be no subject at all, and even just oblique patients and oblique agents, whereas the present and the past tend to have an actual subject - either a promoted patient or an agent - that triggers congruence.

Friday, October 10, 2014

Detail #104: Onwards with detail #101!

Combine detail #101 with a system where different nouns (and ~adjectives, obviously, as detail #101 presupposes a vague or nonexistent distinction between these two classes) belong to different declensions, mainly determined by morphophonological things.

Thus, case distinctions will come and go rather randomly for phrases - "that old man" might distinguish dative-genitive, accusative, nominative and instrumental, but "that young man" might distinguish dative, genitive-instrumental, accusative-nominative, and "this learned man" might distinguish dative, genitive, instrumental and accusative-nominative.

Thus, the amount of cases distinguished in the language would get somewhat perplexing.

We add to this that with deictic determiners involved, some adpositions might be left out - yet the case goes on whichever noun in the NP attracts case most strongly, so "at this X house" ≃ "here X house" -> that house/that X, depending on what X is. For some nouns, the deictic determiner may even be the highest-ranking attractor, in which case "≃there) X Y" would be the outcome.

Thursday, October 9, 2014

Detail #103: A very limited split ergativity

Imagine a language with a fairly regular nominative-accusative alignment, which takes an odd turn with verbs of movement. Verbs of movement have an ergative case, and for them, the nominative serves more as an absolutive case.

Jeten.nom see.3sg Igar.acc = Jeten sees Igar
Jeten.nom sing.3sg = Jeten sings
Kuwan.nom walk.3sg = Kuwan walks 
Kuwan.erg  walk.3sg bag.nom = Kuwan carries the bag
Diraj.nom return.3sg = Diraj returns
Diraj.erg return.3sg car.nom = Diraj returns the car
Accusatives of course appear somewhat randomly instead of nominatives on occasion, and this kind of performance mistake is slowly leading the language into a more classical accusative alignment, although with a specific subject case for verbs of movement - losing the distinction between transitive and intransitive verbs of movement. Meanwhile, there's a slight tendency for the ergative marker - "-gri" - to be prefixed to the verb instead and parsed as a causative morpheme for verbs of movement.

Sunday, October 5, 2014

An Observation on Conlanging

Conlanging has a particular problem as far as hobbies go. I don't think this problem is unique to it, but it is fairly strongly present.

Obviously, conlanging is somewhat knowledge-intense. For newcomers, this is in part alleviated by some guides - Zompist's Language Construction Kit, various threads for newcomers at various fora, people answering newcomers' questions for help with a bunch of advice. (The same bunch of advice, however, has now been repeated for about 15 years without any updating, without anyone even looking at what results the advice leads to.)

Since conlanging has become more popular in recent years - a result of Avatar, Game of Thrones, Defiance and possibly other movies and series that I may be forgetting right now - more newcomers that don't have all that much knowledge in linguistics enter conlanger communities than ever before, and are presented with the same advice.

Pretty much everyone who offers advice are in an almost heart-warming agreement as to what the first thing a conlang needs is. That thing is a almost unvariably a phonology. And there is a nice way of illustrating a phonology - a table with IPA symbols in it.

Of course, in reality, there's much more to a phonology than that. There is all the allophony (which some conlangers do get into), there is archiphonemes and mergers, there's phonotax, there's syllable structure, there's suprasegmentals, there's feature spreading, there's prosody, ...

However, this leads to a situation where every new conlanger posts a tabular representation of a phonology, or even a few of them - each for a new language once the previous one has been abandoned. And we get drowned in relatively boring phoneme inventories. Phoneme inventories are useful for conlangers, but they are about as exciting as any other excel table.

My intention when starting this blog was to show that compelling conlanging can be done without even having a table of phonemes in mind. That one can describe and develop a language quite far without phonology. And once you have an idea about the kinds of constructions - the morphosyntax, the morphology, even the more scary stuff in pragmatics and semantics, - you can start thinking about what kind of a phonology you want to add as skin on your Frankenstein's monster.

Heck, you can even come up with a loose sketch of morphophonological features of the language without having a clear idea of the phonology - and I think that might be a good idea, even, as you then can design the phonology to make realistic morphophonemic alterations that pattern the way you want them to.

I hope this blog this far has inspired people to skip ahead past the phonology-section of the Language Construction Kit, and to start describing wildly different aspects of their languages without fleshing the phonology out. Maybe, in the future, we will see more conlangs introduced not by posts describing their phonologies, but posts describing their most fascinating, well-developed and quirky parts.