Gaaziketti

Basic idea behind the grammar

In Gaaziketti, just about any kind of word can modify (qualify) just about any other kind of word. Verbs, nouns and adjectives fall into essentially a single part of speech.

For example, kossakaidu means 'jeweller' -- if thought of as a noun.

But consider this one-word sentence: Kossakaidu.

Here kossakaidu acts as a verb, meaning: 'There is a jeweller.'

Kossakaidu le raazu. means 'The jeweller criticises (something).'

But raazu, thought of as a noun, means 'criticising' (n.) or 'criticism'.

Thus " Kossakaidu le raazu. " can be read as 'There is criticising by the jeweller.'

Adjectives are generally treated as verbs. E.g. cooqi: to be silent. For such an "adjective" (which is really a verb), if you want to treat it as an attributive adjective, e.g. as in 'the silent movie', then you put it like this: 'the movie which is silent'. This isn't particularly unwieldy in Gaaziketti, because there is a concise relative pronoun (d). Thus cooqi de haiga le means 'the movie which is silent'/ 'the silent movie'.

The syntax tends to be agglutinative.

Sounds

The vowels are read approximately as in Spanish. But double vowels are pronounced long. Double consonants are pronounced double. Like Finnish.

'q' is pronounced like the 'ni' in 'onion'.

'c' is pronounced like the 'ch' is 'chair'.

'n' when at the end of a word, is pronounced like the 'ng' is 'song', or else is assimilated to the following consonant, as in Japanese. Elsewhere, it is pronounced like the 'n' in 'sonnet'.

'ai' is pronounced like 'eye' in 'eye'.

'au' is pronounced like the 'ow' in 'cow'.

Other sounds are pretty much pronounced as one would expect. But unvoiced consonants are not aspirated. E.g. 'p' is pronounced always as in 'spot', never as in 'pack'.

Lexicon sample

I have arbitrarily picked the following words from the lexicon for the purposes of illustrating grammar. See Gaaziketti: Lexicon for more (randomly selected) words.

kossakaidu jeweller

nauzu to be original

mennu name

raazu to criticise

cooqi to be silent

haiga movie

kuubu slave

deela enemy

cecca soldier

paa to be (only for linking noun phrases)

Word order

SOV (usual)

or

OSV

In general, a modifier precedes what it modifies.

Articles

These follow the noun and are, in general, compulsory, for any phrase intended to function as a noun phrase. But pronouns, of course, do not need articles.

Articles receive compulsory case endings. (See below.)

Sing. def. l

Pl. def. k

Sing. indef. s

Pl. indef. m

Case endings

nominative -e

accusative/ genitive -o

dative -ai

Note that these are added to articles and pronouns, but not to nouns.

Personal pronouns

These are listed below in order of nom., acc., dat. forms. Possessive (genitive) forms, as attributive, precede the noun phrase, and the noun must still take the article, e.g.: qo haiga le, my movie; gotos nauzun mennu ke, their original names.

1st person singular: qe, qo, qai; plural: goqe, goqo, goqai.

2nd person singular: ve, vo, vai; plural: gove, govo, govai.

3rd person animate singular: te, to, tai; plural: gote, goto, gotai.

3rd person inanim. singular: ze, zo, zai; plural: goze, gozo, gozai.

Parsing markers

These markers indicate exactly what qualifies what. They can be added to any morpheme ending in a vowel (or diphthong).

Gather one element, modify next element (nil)

Gather one element, do not modify next element -s

Gather two elements, modify next element -n

Gather two elements, do not modify next element -k

It's kind of hard to explain how these work. If you've got a noun phrase, say, and it's modifying the next phrase to appear in the sentence, then you don't need to tack on any parsing marker at all. But say you want noun phrase X not to modify the very next phrase in the sentence. Then you tack on -s. Say the order of phrases is XYZ. Say you want X to modify Z and Y to modify Z. This generally occurs when Z is appearing as a verb, and X and Y are the subject and object respectively. Then X gets -s and Y gets -n. But say you've got WXYZ. You want W, X and Y each to modify Z. Then W gets -s, X gets -k (which serves to "gather" W and X into a unit containing two phrases, both of which will end up modifying the same unit, without modifying the very next unit), and finally -n. Note that Z gets no marker, because it appears as a verb.

Verbal endings

Indicative “there is” (nil)

Imperative (informal) -p

Imperative (formal) -vaa

Infinitive -ce

Relative pronoun

The relative clause precedes the modified noun.

There are two versions of the relative pronoun. There is a concise version which cannot handle more complicated types of construction; and there is a less concise version, which can handle complex constructions better.

The concise option: d-

Declined according to the modified noun's case with respect to the verb in the relative clause.

Placed directly after the relative clause's verb. The relative clause's verb is in the indicative.

The flexible option: siim-

Declined according to case within relative clause.

Placed within the relative clause itself, i.e. somewhere before the relative clause's verb.

The relative clause's verb is in put into the infinitive.

Some example sentences to illustrate the above principles

Kuubu lo deela les cecca sen paa. The slave’s enemy is a soldier.

Kossakaidu les haiga lo mennu lon raazu. The jeweller criticises the movie’s name.

Kossakaidu les siimo mennu lon raazuce haiga le nauzu. The movie whose name the jeweller criticised is original.

Haiga lo mennu lo raazu de kossakaidu le cooqi. The jeweller who criticises the movie’s name is silent.

Cooqi de kossakaidu les haiga lon raazu. The silent jeweller critises the movie.

Kossakaidu les haiga lok cooqi den raazu. The jeweller criticises the movie silently. (Can be read as: "There is some criticising, which is (a) by the jeweller, (b) of the movie, and (c) silent.)

Lexicon

See Gaaziketti: Lexicon

Background

I invented Gaaziketti for the fun of it. I wanted it - among other things - to have a grammar which parses unambiguously, i.e. so that you always know what qualifies what in a sentence, and furthermore so that (even ignoring the written spaces between words) it is always possible to figure out where the boundaries between morphemes are (say if you are a computer). (Basically, i, a and u occur as single vowels only in the second and subsequent syllables of a morpheme, and no other vowel can occur in such a syllable.) I also wanted the grammar to be flexible and uniform, i.e. to be very general, with few or no exceptions and few "sub-rules". Also it had to be reasonably concise, but while being fairly easily pronounceable, having few consonant clusters, few vowels and diphthongs, and few rare consonants. Also, I wanted to minimise the number of pairs of words in the lexicon that differed only slightly, i.e. I built redundancy into it, as in real languages (although the way I built it in was somewhat artificial and "automatic"). Also I wanted to respect as far as possible the language "universals" that are true of most natural languages. And I wanted easily to be able to make new words by just conjoining two existing ones. On the other hand, I didn't care about resemblance between the Gaaziketti lexicon and that of natural languages.

Trying to reconcile all these goals was something of a challenge, but I think I've got something I'm reasonably happy with.

There is a bit more to it than what is above, incl. a (lazily auto-generated) provisional vocab of 1600 words. But I've yet to finalise a lot of the "small" words like personal pronouns, postpositions, conjunctions, tense-markers and so forth.

I hope to add to this page gradually as I work out more of the details.