|Genealogical classification:||a priori experimental language|
|Basic word order:||stack-based (superficially, VSO)|
The syntax of X-2 is not based on tree structures as in human natural languages (as well as in the vast majority of conlangs), but on a LIFO stack, similar to the Forth programming language and Jeffrey Henning's Fith. Its syntactic approach, however, is quite different from Fith, despite using a stack like Fith does. X-2 does not distinguish between nouns, verbs, adjectives and adverbs, but (like X-1) has only one single open word class encompassing all of these. It also features (again like X-1) a self-segregating morphology.
Credits go to Raymond A. Brown for suggesting the kind of syntax used in X-2 in a discussion in the CONLANG mailing list in February 2005.
X-2 is still very much under construction.
X-2 lacks labial consonants and rounded vowels and thus can be spoken without moving the lips. The orthography is strictly phonemic.
|Voiceless fricatives||s||c [ʃ]||x||h|
|Voiced fricatives||z||j [ʒ]|
All vowels (including the back ones) are unrounded. There are no length dístinctions and no diphthongs.
X-2 is an isolating language; there are no derivational or inflectional affixes. The morphology used is self-segregating. Each morpheme is an alternating sequence of consonants and vowels beginning and ending with a consonant. Thus, there are no vowel clusters and no consonant clusters within morphemes. On the other hand, all morpheme boundaries are marked by a consonant following a consonant. For example, the phoneme sequence galakturjeristol can only be broken up into morphemes as galak-tur-jeris-tol.
There are three classes of words in X-2. One are the lexemes, or content words. These have two or more syllables. The lexemes of X-2 correspond to nouns, verbs, adjectives and adverbs in natural languages. Lexemes are pushed onto the stack. The lexemes are the only open word class in X-2.
The second class of words are the relationals, which express relations between lexemes. A relational pops expressions from the stack, combines them into a complex expression, and pushes that expression back onto the stack. The relationals can be likened to case endings and prepositions in human languages. However, they encode semantic roles rather than notions such as 'subject' or 'object'.
The third (and smallest) word class in X-2 are the stack operators. These do such things as duplicating the top item of the stack, swapping the topmost two items, and similar things.
Morphemes of the shape CVCVC or longer are lexemes. Morphemes of the shape CVC are grammatical morphemes (relationals and stack operators). There is a single morpheme consisting of a single consonant: q, which finishes a sentence.
In the following definitions, 'E1' refers to the top expression on the stack and 'E2' to the expression immediately below it.
cil E1 is attribute of E2.
dag E1 is agent of E2.
les E1 is patient of E2.
rut E1 is location of E2.
set E1 is source of E2.
tut E1 is goal of E2.
xaz E1 is part of E2.
zil E1 is possessor of E2.
dud doubles the top element of the stack.
kik discards the top element of the stack.
lit swaps the top two elements of the stack.
q finishes the sentence (popping the top expression from the stack).
The syntax of X-2 is based on a stack, i. e. a data structure supporting only two basic operations. Imagine, for example, a stack of cards. You can either push an item onto the top of the stack, or pop the top item from the stack. The stack used in X-2 holds expressions which are either single lexemes or multiple lexemes connected by relationals (see below). An expression can be the equivalent of a word, a phrase, a clause or a whole sentence.
The stack-based syntax allows for many complicated syntactic rearrangements. Lexemes can be pushed onto the stack and left there for a long time before anything is done with them.
salat kakal cixil cil talag cil dag kahak jujur cil tohok xaz rut q
jump fox quick ATTR brown ATTR AGT dog lazy ATTR above PART LOC .
'The quick brown fox jumps over the lazy dog.'