Re: S7 grammar in a nutshell (long)

That looks very interesting. I am looking at the information more closely, but as
it is it seems to be very cool. Especially thinking about how this type of
language could be applied. Anyway, congratulations on an exellent conlang!

Henrik Theiling <theiling@...> wrote:
Hi!

I'd like to present a brief overview of my current project S7. (The
language does not yet have a good name.) Please give me feedback
about what you think of this.


S7 Features
-----------

Design goals:
- polysynthetic

- rich derivational expressiveness

- optional categories instead of mandatory ones, thus
underspecification by default and precision on demand.
This is to achieve feasibility for both poetry and law.

- minimal grammar-only affix inventory, maximal degree of
lexicon ~ grammar coincedence (e.g. tenses are also
free-standing words as are aspects etc.)

Achievements:
- Only valence and degree infixes are grammar-only and thus
not found in the lexicon but in the grammar description.
(Degree will not be discussed in this short overview.)

- This seems to become the first language I create that I like
the sound of. I am happy! :-)


S7 morphosyntax in a nutshell
-----------------------------

Roots:
- Roots have the following structure:

{C1 -- C2 --}

C1: classifier consonant
C2: second consonant ('true root consonant')
--: gaps for vowels
{}: word boundaries

(But S7 is not a Semitic conlang...)

Complex roots with additional syllables exist, but will be
neglected for simplicity here.

The first consonant C1 is a *classifier*. It is only used on the
head (the first component) of a composite word. The classifier
very vaguely give a hint about the meaning.

Roots have vague meanings. They are the basis for the main
lexicon entries: stems.

Stems:
- Inserting a vowel after the C2 consonant of a root gives you
a stem:

{C1 -- C2 V}

This is a lexicon entry. Stems are semantically related to the
base root; the vowel makes a more precise meaning.

There is only one class of words in S7 -- all stems behave alike.

Derivation by Composition:
- Multiple stems can be composed. The class consonant is lost
for the second component:

Composition of {C1a -- C2a Va} + {C1b -- C2b Vb}
yields {C1a -- C2a Va C2b Vb}

Composition is head-first. E.g. 'snowman' ~
{ GAP 'person' 'snow'}

During composition, the meaning of the second part is often
generalised. Therefore, many composites are also lexicon entries.
However, new composite words can be generated freely by this
composition mechanism.

The part of a stem without the classifier consonant will be called
'clitic': it is an item that is derived from a stem, but cannot be
used independently.

Modification by Aspect/Tense/Inclination/Degree/...:
- A composite can be modified with infixed aspect, tense
clitics. The classifier consonant of the modified composite
is kept in front. The stems modifiers for aspect/tense/inclination
and whatelse are derived from are normal stems in the lexicon and
can be used freestandingly, too.

Given a composite: {C1a -- RESTa}
and a modifier stem: {C2a -- C2b Vb}
composition of them: {C1a -- RESTa} + {C2a -- C2b Vb}

yields: {C1a -- C2b Vb RESTa}

E.g.: 'the ex-snowman' ~
{ GAP 'was' 'person' 'snow'}

Phrases:
- To make a word complete, the GAP has to be filled. There
are two possibilities:

a) You want a predicate: add a valence vowel
b) You don't: add a case vowel

In case of a), the added valence vowel defines the argument
structure of the predicate. S7 marks semantical *and* syntactical
valence here (due to this, a simple vowel is often not enough,
but we'll neglect this for now).

E.g. you want to add a patient (e.g. 'Joe') to the above
composite. For the predicate, you'd get:

{GAPe GAP <'person'> <'pat' valence> 'was' 'person' 'snow'}

There are two new gaps now at the beginning of the word.
The second one is the same as before: for another valence or
case vowel. Only you cannot add two valence vowel, so this
*must* be a case vowel now.

!! Every words needs to carry a case vowel in S7,
!! even predicates.

The first gap shown as GAPe is for an evidential or mood infix.

!! Every predicate carries an evidential/mood affix.

Evidence/Mood:
To the above phrase, we'll add an evidential for perception.

We get:

{
 GAP <'person'> <'pat'> 'was' 'person' 'snow'}

This is no sentence: case and the patient are both still missing
for a complete sentence:

Core Case:

If a phrase is non-modifying, it is assigned the
*predicative case*. There is only one core case, since
the agent/patient function is marked by the valence
already.

:: Agentive/patientive case is marked on the head in the
:: valence infix.

Predicative case (PRD) marks main clause predicates and
non-modifying dependents. The above phrase as main clause:

{

<'person'> 'was' 'person' 'snow'}
~
'(to perceive that) X is an ex-snowman/was a snowman'

The patient 'Joe' is still missing. We must make a stem from a
name. S7 does it like this:

{ GAP "Joe" 'name'}

("Joe" will probably be /tsu/ in S7)

John will be the patient of a clause, so it does not modify
anything. This it will carry predicative case, too. We get:

{<'unique'>
"Joe" 'name'}

Now we can compose a complete sentence:

{

<'person'> <'pat'> 'was' 'person' 'snow'}
{<'unique'>
"Joe" 'name'}.

= '(I see/It can be seen:) Joe is an ex-snowman./was a snowman'.

Incorporation:

A different valence infix can be selected to mark that the patient
is incorporated into the predicate. Let's call this valence 'pat*'.
Then we get:

{

<'person'> <'pat*'> 'was' 'person' 'snow' "Joe" 'name'}.

This is 'Joe was a snowman.' in one word.

Only words in predicative case can be incorporated. Together with
their case infix they will drop their classifier.

Other cases:

Clauses may be modified by other clauses. For this, the clause
is assigned a different case than predicative. The modifier
precedes the modified.

Consider the sentence: 'I think, therefore I am'.

You'd put the phrase 'I think' in *causative case* (CAU) modifying
'I am' to express 'because'.

For this sentence, you'd use agent valence for 'I think' and
patient valence for 'I am'. Further, 'I think' would use
experiential evidence and 'I am' conclusive evidence. Together:

{ <'agt'> to-'think' 'I'}
{
<'pat'> 'to-be' 'I'}.

Those are the basics of S7.


What do you think? I'd very much appreciate questions and comments.
(Also about anadewisms. :-))

**Henrik

http://www.theiling.de/projects/s7/

Aaron Morse
www.diarenye.net
Man sad si i roch ah i rochon?
Do you Yahoo!?
Yahoo! Finance Tax Center - File online. File on time.