Monday, December 30, 2013

A paper worth reading

There is a fascinating paper by David Tuggy on a verb in Orizaba Nawatl, detailing a verb that has no stem - it consists, entirely, of grammatical affixes. Well worth reading.

In addition, having it linked here I will not displace it by cleaning out my bookmarks when they're getting too many, or by reinstalling the OS and not keeping the bookmarks backed up. Which I suspect may be appreciated by the friend who brought this to my attention, as I have already asked him to link it a full four or five times over the last several years.

Sunday, December 29, 2013

Detail #72: A diachronic origin for present participles

In some language with a moderate case system, whatever case can be used for complements goes on the infinite to form an adverbial/complemental participle. Over time, the participle is reanalyzed and can be used as a full adjectival participle, both in attributal, complemental and adverbial ways.

An overview of this process, with a Slavic influence, would posit that instrumentals are used on some complements. This is especially useful, as it lends itself to two developments:
1) "by [infinitive]", i.e. marking the infinitive to show the manner in which something was carried out.
2) "as (implicitly: doing) [infinitive]", i.e. marking the infinitive to show that the infinitive is a complement telling us something about the subject (or possibly object).

The second option easily could include elliptical uses of subclauses (X who is verb-instr. -> X verb-instr) and soon, the participles may be very much like classical participles for those.

I am thinking of doing something similar for that case system with regard to how participles came about. In that, the case markers that form present participles will be closely related to the complements and the instrumental-comitative.

Friday, December 27, 2013

Detail #71: A weirdo place to put auxiliary-like things

Make some auxiliaries be genitive attributes of the subject, and add a copula between the subject and a somewhat less inflected (possibly infinitive) form of the verb:
[word indicating obligation]'s men are eat → men have to eat
sometimes [word indicating occasional occurence or possibility]'s me am doubt the wisdom of this plan → sometimes I might doubt the wisdom of this plan
These nouns can also appear as dummy subjects for verbs that usually do not take subjects, or for modal passive-like constructions (passive in that the subject is omitted, but an object is retained).

Detail #70: A participle-like form

In the participle-rich language I described way back, one of the participles could perchance work a bit like this:

Applies to a verb related to

  • having a social status (to rule, to serve, to minister, to teach, to lead an army, to be an outcast, to be exiled, to be ostracized, to be a man of religious significance, to be a woman of religious significance, to be a monk, to be a soldier)
  • obtaining a social status (generally the previous verbs in transitional forms, to marry, to be baptized (or analogous religious verbs), to be granted membership in certain kinds of fraternities and organizations, to join things, to enlist in the army, to be entrusted with a responsibility, to be granted some privilege or title, to be granted a higher rank, to ascend a throne, to be found innocent, to be found guilty, to be imprisoned, to be released)
  • loosing a social status (to abdicate, to be discharged, etc)
The resulting form is used with people associated with the person to whom the verb refers:

abdicate-prtcpl mother: the mother of the abdicated (king)
defrocked-prtcpl sister  : the sister of the defrocked (priest)
army-lead-prtcpl friends : the friends of the general

Often, due to the social importance of the parent-child relationship, sons, daughters and parents are the most often used nouns when these participles are attributes, but they may appear with any kind of social relationship. At times, the participle is used as a head of a noun phrase itself, and then may refer to some relevant relation to the person that has the referred to status.

At times, it is used as a complement, and may have different connotations:
he was being-man-of-religious-significance.prtcpl : he made (his contextually relevant relation) turn into a man of religious significance (generally this indicates something comparable to bishop)
she wanted heal.prtcpl = she wanted her son/daughter to be a doctor
he resented being-granted-priviliges.prtcpl = he resented that his brother(?) was granted privileges (which by implication he wanted for himself)

Thursday, December 26, 2013

Detail #69: A detail regarding comparative and superlative adjectives

I am generally somewhat cautious about including comparatives and superlatives in non-European languages (granted, there's no one-to-one correspondence between languages that have them and Indo-European languages, neither one way or the other), and I generally advise against their inclusion in conlangs.

However, let's assume a conlang having at least two, possibly all three of these. Let's add a few features:

There's significantly fewer comparatives and superlatives than positives.
Now, that's not particularly weird by itself - some adjectives' comparative and superlative forms nearly never get used in a literal sense, and some probably lack both reasonable literal and figurative senses. Examples of those lacking a reasonable literal sense are deader, deadest, and more alive, most alive. 

There's no morphological correlation between regular adjectives and comparatives.
Now, to get this weirder, let us just make the "lexical resolution" of comparatives and superlatives less than that of regular adjectives: handsome, stylish, good-looking, pretty, beautiful, all correlate to two comparatives with a slight meaning-difference between them; difficult, heavy, wearisome, tragic, taxing, all are represented by one particular comparative. There need not be any direct similarity between any of the positive adjectives and the comparative adjective they're associated with.

The comparative form is a comparative form mainly on account of its use - it can be used in comparative constructions in ways that a positive adjective cannot, and of its meaning - the meaning actually is 'has more of the quality of this class of qualities than the other referent in the construction'.

For the superlatives, the system might be somewhat different. I am inclined to entirely drop those, but having most adjectives have distinct (and morphologically predictable) superlatives - even the comparative lexemes would have them - might be somewhat interesting. Finally, I think the superlatives would be more likely to be used as intensifiers, whereas the comparative is more clearly restricted to comparative constructions.

Sunday, December 22, 2013

Detail #68: Participle-morphology on non-verbs

In some conlang with a system of four participles - active perfective, passive perfective, active imperfective, passive imperfective - and where copulas usually are not dropped, negated adjectival complements for some reason take no copula, but instead take participle markings. Which particular participle goes on the adjective depends on lexical as well as contextual factors. Since the copula is dropped, this usually leads to the aspect of the chosen participle marking to serve tense-aspect-mood functions, while the voice sometimes marks details that usually would be visible in which particular copula was picked. However, the copula system of the language has at least three different copulas, so some distinctions are lost in the negative.

Thursday, December 19, 2013

Detail #67: Nouny 'and' and a noun class

I recently read of a language where the conjunction and is inflected like a verb. Suddenly the notion of having a very nouny and appeared to me.

Morphological template:
[noun class/number][root, -əl- or -əj-][dual/plural/optional possessive suffixes][case]
 Now, as may be gleaned by careful observation, number congruence appears in two places - in part with the noun class morpheme, in part in the slot on the other side of the root. And belongs to a pseudoclass of nouns, including several functional words and just a few regular nouns (if, I, anything, something, nothing, a few dummy objects, what when not being a determiner of a noun, ...)

The number is either dual or plural, for which the noun class does have markers (shared with another noun class though). To refer to the noun phrase later, the pronoun will be marked for this noun class.

The slot after the root is kind of interesting, though: if both nouns are possessed by the same owner, this slot will be filled with a corresponding affix, and no number congruence whatsoever. If no owner is present (or there are different owners for the nouns), the number congruence will be either dual or plural. Plural congruence permits stacking several ands, in which case only the first has to take case. If a pronoun is included, this also is marked as a possessive marker, so 'my friend and I' is 'friend-my dual-əl-mine-[case]', my friend and you is friend-my dual-əl-yours-[case].

Often, the conjunction goes after the two nouns, reverse polish notation style, but it may be shifted around due to the relatively free word order of the language.

Case follows the language's usual case marking paradigm, and marks the case of the whole phrase. The subordinate nouns may take the same case. Another option would be to use some kind of wastebasket case or not to distinguish certain cases when coordinated, e.g. object and subject cases default to nominative on the nouns (but not on the conjunction), whereas local cases are maintained on the noun, so e.g.
I bought a newspaper and an icecream -> I bought class_of_trivial_things-newspaper-nom foods_and_plants-icecream-nom pseudoclass/dual-əl-acc
I visited the museum and the concert hall -> I visit-1sg-past class:big_things-museum-to class:big_things-concerthall-to pseudoclass/dual-əl-to
If the coordinated nouns have different case - a permitted thing in this language - the case of the conjunction is generally the genitive (for core case coordination) or the somewhat wastebaskety locative.

There is a tendency for heavy constituents to go first among the two nouns, and in the case of trivial constituents, there is a discernible hierarchy where the greater the animacy, social importance and contextual relevance, the more likely the noun is to be placed first.

For coordinated non-nouns, a naked əl/əj suffices, and then often goes both between and after the verb phrases or adjectives. For subordinate clauses to which the speaker may want to refer, [singular]əl[3sg grammatical class possessor][case] may serve to form a subordinating conjunction:
he said singular-əl-3sg grm.class.acc he returns tomorrow
ta mege nəleta miǵris neltle
 he said that he returns tomorrow

Tuesday, December 17, 2013

Detail #66: A tiny noun morphology detail

In this sketch, a language with question particles (or affixes) and a vocative case, the combination of these two is parsed as 'are you X?'.

King-voc-question : are you the king?
John-voc-question: You are John, right?
Sick-voc-question: are you sick? 

Monday, December 9, 2013

Conlanging by Fiat: Unjustified Claims for Typological Quirks

There is a thing that bothers me about the approach to linguistics that some conlang descriptions rest on.

Not because it relies on bad linguistics (which it does), but because it relies on bad epistemology and a generally bad approach to understanding how things works and what makes them what they are.

The example I will give is phonology-based, but similar kinds of mistakes appear in grammatical descriptions.

I guess I have to explain this in greater detail . When we discuss phonology, we are making several related kinds of statements about the sound system of a language - we are making a claim about what actual phones appear in the language, the distribution of these phones, and the parsing of these phones in terms of what phonemes they belong to - basically how the brain will analyze the sounds it hear, and finally there's a more meta thing that no one ever thinks about. This final thing is how would a brain that doesn't know a single language assign the phones they hear to phonemes in a way that is natural for the human brain. Most conlangers skip that step, and sometimes make rather unjustified analyses that not only are unlikely and unnatural, but basically unlearnable except by being explicitly taught the way they are designed - which 'natural' speakers generally wouldn't be, most people acquire their phoneme system by unconscious analysis of the language they hear around them.

Sunday, December 1, 2013

Ćwarmin: The adjective and a bit about the noun phrase

Despite my almost obsession with congruence between adjectives and nouns in NPs, I have decided not to do that in Ćwarmin. As previously noted, Ćwarmin has about 20 cases. The case suffix normally goes on the head noun, but sometimes a determiner will carry the main case marking of the clause, with the head noun either taking the genitive, the general ablative or the possessed objective (the last of which never goes on the determiner).

On occasion, an adjective may be inflected in some case other than the case of the noun. The only cases that do not appear on adjectives standing as attributes of nouns are the reflexively possessed accusative and the accusative.


There are a number of determiners - demonstratives, indefinite determiners (including the negative determiner), amount-related determiners (numbers, 'many', 'a few', 'all', etc). As mentioned, these sometimes carry the whole noun phrase's case, in which case the noun either agrees with it in case (if the whole phrase is nominative, accusative or genitive), or takes either the general ablative or genitive.

Those determiners which can have a singular as well as a plural meaning usually have number congruence with the head noun. 

Some special lexemes

'One' - er - in combination with the nominative complement case or genitive, eramće or erća, roughly works like specific or particular. The genitive is most often used with timespans or with mass nouns, but does occur with other nouns as well.
erəmće seltimgə - a specific fisherman
erćə toŋugul - one particular winter
erćə mehwi - a particular loaf of bread

Ćul - 'few' - combines with the complement cases for a meaning along the lines of 'just a few'
ćulamćo nedim, ćulamćan nedim - just a few bits.
In the genitive it signifies 'too few'.

Cases with Adjectives

The cases can also alter how the adjective interacts with the head noun, both for adjectives in attribute position and in complement position.

Negative - the negative simply negates an adjective, essentially like sticking in- or un- or a- before an adjective in English.

Complement cases - the complement cases tend to mark temporary or clearly subjective qualities. The nominative complement is used with rather objective qualities or qualities an animate noun intentionally acquired or maintains, whereas the accusative complement more often implies that the quality has been caused to the noun without its participation or regardless of his or her wishes.

General ablative - roughly like affixing -ish to an adjective in English when used as an attribute. Also for complements that mark what something appears to be. In situations where this and genitive or instrumental could work, the ablative has priority.

Genitive - generally used to intensify the adjective, but also marks flavors when they're complements.

Instrumental - generally used to intensify the adjective. Also marks various sound qualities when as complements.

Locative cases - sometimes nouns with these cases are used as adjectives.

Dative - for some peculiar reason, this marks first-hand knowledge evidentiality.

Comitative-with and comitative-to mark different levels of evidentiality - basically hearsay and 'anything further off than that'.