finiteness conference, Konstanz, 10-12 May 2001
Frans Plank
(Universität Konstanz)

Questions about "finiteness"

1. Can it meaningfully be considered a property? If not, what can it be considered as?

2. What is its logical structure (assuming finiteness is a property or at any rate a category)?

3. What is finiteness a property of? 4.a. Is finiteness a basic-level or a higher-level property?

4.b. If higher-level: In terms of which more basic properties is it defined?
Is it merely abbreviatory of (structured) sets of more basic properties, or does it add something that can claim (more-than-terminological) existence in its own right (and can non-circularly be referred to by rules of grammar)?

5. What purposes is finiteness used for?

6. Can finiteness be acquired, altered, and lost in change (by whatever it is a property of)?

7. Assuming finiteness is variable in time and across languages, is such variation random or subject to limitations?

8. Assuming there are such limitations, are they due to timeless laws (constraining change) or do they reflect laws of change (forcing or constraining reanalyses by successive generations of learners)?

9. Why are finiteness things the way they are, language-particularly and crosslinguistically?

Where finiteness has been considered to be implicated:

Which morphological and syntactic categories have been considered to be implicated in statements made in terms of finiteness:

finiteness as difference: a consensus view of what is at issue

independent and less or more dependent predications (sentences, clauses, verbs) are structurally similar but not identical:
(non-)finiteness is a label for the difference between them, of whatever extent and kind.

Presumably all languages grammatically distinguish less/more dependent predications from independent ones; hence finiteness, as indep/dep difference, is a universal dimension.

Where languages can conceivably differ from each other is in the extent and kind of the relevant differences:

If indep/dep difference is all there is to finiteness, then language-particular rules can hardly make reference to such a notion. It is only what indep/dep difference consists in, not finiteness in that sense itself, that can play a role in language-particular grammar.

But then, the notion of indep/dep difference can be provided with more substance:
difference consists in independent predications having something that dependent predications have to a lesser extent or lack ?and not surpluses or deficits in any arbitrary respects but in these three:

  • independent speech-event anchoring (temporal, epistemic-attitudinal, local)
    [illocutionary force, assertiveness]
  • independent informational articulation (topic-comment, > subject-predicate)
  • contextual autonomy (rather than contextual adjustments: integration,subordination in clause combining)
  • Generalizing: indep/dep difference is in the main a matter of deixis-inflectionally more or less distinctive forms used in informationally more or less highly articulated and contextually more or less unrestrained constructions.

    From the point of view of dependent predications, independence is attained by adding to them, along these three major lines (rather than in any arbitrary respect?e.g. phoneme inventories, orders of numerals and nouns, cumulative/separatist exponence, ...);
    from the point of view of independent predications, dependence means loss, again along these three lines. [cf. negation: adding neg marker/omitting assertive marking]

    But can having and lacking meaningfully be looked at as gaining and losing?
    Yes, diachronically ? in case independent predications arise from dependent ones or vice versa, by grammaticalization and analogy.

    indep/dep difference thus narrowed down, the question is how particular languages code this universal dimension. Presumably there is still much room for crosslinguistic variation; but is it random?

    Here is where consensus ends: on the evidence of some lgs claims have been made that individual choices that lgs have in coding indep/dep difference are interdependent (e.g. subject case marking, person-number agreement between subject and verb, verbal tense and mood inflection); on the evidence of other lgs, such implications have been rejected.

    Assuming that crosslinguistic variation in whatever domain is limited rather than random, such limitations can be due to timeless laws (constraining change) or they may reflect laws of change (forcing or constraining reanalyses by successive generations of learners).

    Finding contradictory evidence in different languages is at odds with timeless laws, but it need not invalidate the assumption of regular change: even if possible reanalyses are highly constrained, what results from such reanalyses may still differ widely if the forms-in-constructions undergoing reanalysis have been widely different to begin with.

    What should therefore be high on the agenda of crosslinguistic research into finiteness are diachronic scenarios of gains and losses of predicational independence, with a view to interdependencies of reanalyses along individual parameters of variation, to do with speech-event anchoring, informational articulation, and clause-combining adjustments:

    How (from which sources and by what mechanisms) do forms in consrtructions come to have the properties that they (synchronically) have?

    Direction and pathways of development in these domains:
    always, or mainly, from dependent to independent (upgrading rather than downgrading), by means of insubordination, analogical extension of inflections, other mechanisms?