Sum: Basque modal particle

[Larry Trask]

----------------------------Original message----------------------------
The other day I posted a question asking for possible parallels for the
apparent development of the Basque noun <bide> 'road, way' into a modal
particle meaning 'apparently, probably, undoubtedly'.  I received a number
of interesting replies.  So far, no one has reported a strong parallel for
such a development, but several people have drawn attention to other
developments which are not dissimilar.

To start with, one respondent asked me to clarify the behavior of modal
<bide>.  OK.  This is a particle, with three noteworthy properties.

(1) It is invariant in form, and it can take no affixes.  This is not an
especially common property in morphologically rich Basque.

(2) Its presence has no morphological or ordering consequences for other
elements in the sentence.  It's just "plonked in" directly before the
finite verb or auxiliary.

(3) When the verb-form is periphrastic, <bide> stands between the
non-finite and finite parts of the verb.  This is a position which can be
occupied only by a handful of particles.  No other elements, including
adverbials of any kind, can stand in this position.

So, to recap a bit.  My base sentence was <Zoratu da> 'He has gone crazy',
where <zoratu> is a participle of the verb meaning 'go crazy' and <da> is a
finite auxiliary, third singular.

With <bide>: <Zoratu bide da> 'He has apparently/probably/undoubtedly gone
crazy.'

With the reportative particle <omen> 'they say', 'I hear': <Zoratu omen da>
'They say he has gone crazy', 'I hear he has gone crazy'.

With the yes-no question particle <al>: <Zoratu al da?> 'Has he gone crazy?'

With the "doubtful" or "surprised" question particle <ote>: <Zoratu ote
da?> 'Do you suppose he's gone crazy?', 'Has he really gone crazy?'

Very few other items can behave like this.  The negative particle <ez>
'not' almost does, but it induces a change in word order: <Ez da zoratu>
'He hasn't gone crazy'.

Now, several respondents pointed to the use of what is etymologically a
noun meaning 'way' in various languages -- English, German, French,
Friulian -- to express things we might regard as modalities.  All the
examples are interesting, but all involve the presence of additional
morphology or syntax, rather than the use of the bare noun.  Two
respondents observed that the passive of Latin <videre> 'see' has developed
modal or hedge-like functions in Occitan and Catalan (at least), and they
wondered whether the Basque particle might have been borrowed from one of
these Romance items and then folk-etymologized to the phonologically
similar <bide> (Latin /w/ and Romance /v/ are always borrowed into Basque
as /b/; Basque has neither /w/ nor /v/).

All this is fascinating, but I was really hoping to find something that
fitted well with what little historical evidence we have.  Recall that
eastern Basque has a compound verb <bide izan> 'be proper, be licit', and
that modal <bide> is very sparsely attested in early texts with the
functions of deontic 'must' and epistemic 'must'.

My favorite of the proposals so far is therefore this one.  In English,
'way' can have the sense of 'proper or conventional way of doing things',
as in that fine phrase 'the American way'.  If we assume that the same
semantic development might have occurred in Basque, then we have a sequence
of semantic developments which is appealing and which moreover accords with
the historical data.  It goes like this:

  <bide> 'road, way' (the source word) >

  'proper way of doing things' (well attested in the compound verb <bide
     izan> 'be proper') >

  deontic 'must' (sparsely recorded) >

  epistemic 'must' (sparsely recorded) >

  'undoubtedly, probably' (the target sense)

This looks good to me.  Strictly speaking, 'it's the way' is <bidea da> in
Basque, with the article attached to the noun, but the absence of the
article inside a compound verb is perfectly normal.

So, the syntactic development required to make all of this work is as
follows.  The invariant form <bide da> 'it's proper', with a third-singular
auxiliary <da>, must have occurred often enough with a third-singular
subject (sentence-initial in verb-final Basque) that this subject NP came
to be interpreted as the subject of <da>, rather than of <bide da>, with
<bide> then taken as a modal particle.  This development then allowed a
non-third-singular NP to occur in the same construction, with the finite
copula now required to agree with it normally -- the position in the modern
language.

My thanks to Alexis Manaster Ramer, Ton van der Wouden, Miguel Carrasquer
Vidal, John Hewson, Elizabeth Traugott, Paolo Ramat, Benji Wald, and Mark
Jones.


Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk

Tel: (01273)-678693 (from UK); +44-1273-678693 (from abroad)
Fax: (01273)-671320 (from UK); +44-1273-671320 (from abroad)