THEORY: Diachronic Origin of Verbals and Dependent Clauses

(BTW Is there a connection between the dative use of "to" and the
infinitive use of "to"?  Lehmann seems to suggest that in some Indo-
European languages there is a connection between datives and infinitives.)

I have been reading two books edited by Charles N. Li; one is "Subject and
Topic" and the other is "Word Order and Word Order Change".

The "Word Order and Word Order Change" book contains an article by Winfred
P. Lehmann, "A Discussion of Compounds and Word Order".  It suggests some
ideas that might be useful to conlangers when inventing verbals such as
infinitives and gerunds and participles.  It also would be useful in
inventing the diachronic background for a change from stative verbs to
adjectives, for example, and for the rise of complex sentences with
subordinate and embedded clauses.

Lehmann suggests that a language could begin with consistent VO or
consistent OV order.
A "conistent" VO language is much like Tesniere's (sp?) "head-first"
or "centrifugal" language, save that adverbs and other modifiers of verbs
precede the verb. (Adjectives and modifers of nouns follow the nouns, and
adpositions are prepositions.)
A "conistent" OV language is much like Tesniere's (sp?) "head-last"
or "centripetal" language, save that adverbs and other modifiers of verbs
follow the verb. (Adjectives and modifers of nouns precede the nouns, and
adpositions are postpositions.)

In this language, "dependent" clauses are always the first or last element
of the "main clause".  The question whether or not the "dependent" clause
is actually a constituent of the main clause, or is merely chained to it,
is fuzzy, and is more one of semantics and/or pragmatics than one of
morphosyntax.  Furthermore, the distinction between co-ordination of
clauses and sub-ordination of clauses, is fuzzy, and is also more semantico-
pragmatic than morphosyntactic.

Now imagine that the word-order changes; whether from VO to OV or from OV
to VO.

Among the things that may happen;

o It becomes necessary to position some relative clauses within the main
clause, so that they will be in the proper position relative to the things
they modify.  In other words, "centrally embedded" subordinate
clauses "spring into existence", or at least cease to be rarities.

o It starts to be quite important to decide
- whether a clause is dependent or not on another clause;
- whether two clauses are coordinately conjoined or one is subordinated to
the other;
- whether or not one clause is embedded in another.

o Some stative verbs may become adjectives instead.  The path of change
will take them through relative clause, then participle (stative
participle? rather than active or passive participle?), then plain old
nomina adiectiva.

o A purpose clause, nominalized and inflected in the dative, becomes an
infinitive.

o Various kinds of subordinate clauses evolve into various kinds of verbals.

-----

What do the rest of the list think about these things in relation to;

o How true it all is?

o How useful it might be in conlanging, even if it isn't "true" in real
life?

o How well it applies to natlangs?

o Which natlangs it applies well to?

o How well, in particular, it applies to the history of the Indo-European
family and to P.I.E.?

----

Some conlangers on list don't try to place their conlangs diachronically --
neither w.r.t. other conlangs (for example, "sketchlangs" invented purely
to give the conlang diachronic depth) nor w.r.t. natlangs.

But some do.  Those of you on-list who read this -- if you have given (some
of) your conlang(s) diachronic history, have you used ideas like the
above?  Have you used word-order change?  If you haven't, do you think
you'd like to in future?

-----
eldin