number marking in Basque-2

Alan R. King mccay at REDESTB.ES
Sun Nov 8 16:11:52 UTC 1998

In the second instalment of his observations on number marking in Basque,
Gontzal Aldai wrote: 

>> (a) Is the plural MARKED vis-à-vis the (relatively UNmarked) singular?
>> I think so; Lourdes' remarks may suggest otherwise.
>(Lourdes was only talking about the genitive case, if I'm right. So she
>never claimed that the plural is in general unmarked).

Precisely.  However, I thought that Lourdes' remark on the genitive might
potentially have led the non-Bascologist readers of the list to think
otherwise, and in the interest of Truth I felt obliged to clarify the point.

>> The above examples are all in the absolutive (unmarked) case, and when case
>> morphemes are added (suffixed to the above), the surface forms differ.
>gizon-AREN 'of the man'
>gizon-EN 'of the men'
>"Morphemes are added"? Added to what? Suffixed to the absolutive case?
>Does this mean we are to consider an underlying form and a derivation? Or
>is it that we are to consider Basque as strictly agglutinating?

I thought it was clear what I meant.  The absolutive case in Basque has no
morphological case marker, it's the "zero case".  Other cases are marked by
morphemes added to the NP (which, in the absence of said morpheme, would
have been in the zero/absolutive form).

However, the tricky bit lies in the morphology of the suffixes.  In the
statistically most usual case, a Basque NP contains a determiner, and the
most usual determiner is the article (the default determiner), and the
article takes the form of a suffixed morpheme attached at the end of the NP
(i.e. affixed to whichever word, other than the determiner, comes last in
the NP).  So more often than not we encounter the sequence of
affix-morphemes: + ARTICLE + CASE suffixed to a word at the end of the NP
(except in the absolutive case, where we just get + ARTICLE).  These
ARTICLE + CASE sequences constitute morphological paradigms (whose
dimensions include number and, of course, case - number in the article,
that is), and the paradigms exhibit a degree of fusion between article (+
number) and case morphemes, and a limited degree of (systematically
distributed) allomorphy and homonymy besides.  I do say *limited*
allomorphy to make it clear we're not looking at anything as chaotic as,
e.g., Latin declension paradigms.  (On the other hand, there can be a fair
amount of *phonological* reprocessing in some dialects; see for an example
Karmele Rotaetxe's dissertation on Ondarroa Basque.)  An example of these
morphological complications is our now familiar one, which I repeat for the
sake of the non-Basco's here (whose attention I'm trying valiantly not to

gizonak  '(the) men'

gizonen  'of (the) men'

which contrasts with the morphological transparence of the singular forms:

gizona  '(the) man'

gizonaren  'of (the) man'

where -r- is merely an epenthetic consonant separating two vowels, not a

>> Personally I don't have much of a taste for the generalized positing of
>> abstract underlying forms such as *-ag unless highly motivated in
>> particular instances, and I don't believe this Basque underlying form
>> specifically is all THAT strongly motivated synchronically, but it gets
>> repeated in the literature, so there we are.
>Well. It seems it's time to get rid of this kind of reasoning. If we don't
>believe in abstract underlying forms, let's forget about them altogether.

Actually, *this* is the kind of reasoning *I* would like to get rid of: the
idea that a concept such as abstract underlying forms must EITHER be
applied across the board in grammar OR must be applied nowhere.  I suspect
that the human brain is versatile enough (unlike, perhaps, the current
implicit computer-like processing models we tend to gravitate towards in
our times) to be quite capable of using a given processing strategy when
it's most useful and not bothering when it's not needed.  And I feel that
in our metalinguistic role as describers of languages, we should emulate
that capacity.  But I REALLY didn't want to get into those deep waters (I'm
already repentant, though not enough to delete the previous sentence, I

>Specially the one Alan presents doesn't have any motivation at all. The
>motivation, of course, would have been diachronic. But in this case even
>the diachronic motivation is absent. (I cannot explain here the details).

As it happens, I tend to agree with you, Gontzal, as I think was made
sufficiently clear in my original exposition: I did not "present" the
abstract analysis (of underlying or reconstructed *"gizon-ag-en" --> or >
surface or modern "gizon-en") as my favourite theory, but as a hypothesis
widely taught and referred to (by others) in the discipline, so that it
deserves at least a passing mention on this forum, I should think, even if
you and I choose to reject it.  (It would be interesting to know if we both
reject it for similar reasons; but I also agree with you that that should
not be discussed here and now.)

>In any case, why do we need to posit one only "origin/underlying form" for
>all plural cases? Are we forced to do so?

No!  Sorry to be so boring, but I once again agree with you.

>[By the way, what is the dialectal variant that shows -AN as the genitive
>plural marker?]

This dialectal form (like several others) doesn't seem to be a favourite in
the linguistic literature, for some reason.  Perhaps because it is not one
of those dialectal features that ever found its way into one of the
standard "literary dialects" that were the vehicle of most written Basque
(and thereby became familiar to most readers) before the advent of modern
unified literary Basque.  (Even when discussing dialectal features, the
philological tradition that STILL predominates in native Basque language
study tends to emphasise the things which finds its way into writing over
those which do not.)

To answer your question: In some western varieties the genitive plural
takes the form -AN.  In some (or all?) of those varieties, the genitive
singular is also -AN.  So what we sometimes find is singular/plural
syncretism in the genitive case (overlooking possible contrasts of
accentuation in certain varieties, that is).  It should be noted that some
of the varieties in question are among those varieties with the
phonological process I call vowel harmony (I understand that our
phonologist Lourdes Oñederra is not in favour of this application of the
term, though), where A becomes E if the preceding syllable contains a high
vowel; therefore one must be prepared to find, in such an environment, the
(surface) form -EN rather than -AN, (e.g. gizonan but lagunen) *in both
singular and plural*.  I'll look up some references for you if you're

Before leaving this point (which is not as immaterial to the original
thread of discussion as might first seem, since we're looking here at an
instance of (limited) neutralization of number marking, though
morphological rather than syntactic neutralization; furthermore I happen to
harbour the suspicion that the existence of these forms might have
interesting implications for diachronic morphological reconstruction - but
that most certainly IS another story!), notice, Gontzal, that the existence
here of -A- as both singular and plural article allomorphs is not an
isolated case; compare, in western dialects once more, and in some cases
outside of widely *written* dialect forms but very much alive in the spoken
language, these other instances of singular-plural syncretism in the form
of -A-:

ergative -A-K (cf. standard sing. -ak, plur. -ek)
dative -A-I (cf. standard sing. -ari, plur. -ei)
comitative -A-KIN (cf. standard sing. -arekin, plur. -ekin)

(By the way, I use "western" in a Basque dialectological context to mean
"Bizkaian yes, and Gipuzkoan maybe", i.e. all the above phenomena exist
west of the Bidasoa.)

I do hope this discussion is not getting too far out of hand for LINGTYP,
and that if it is, someone will put a stop to it.


Sources mentioned in this and my previous post in response to Gontzal Aldai:

Azkue, R.M. (1905).  Diccionario vasco-español-francés.  Reprinted: Bilbao:
Gran Enciclopedia Vasca, 1969.

Euskaltzaindia/Basque Language Academy (1991).  Euskal gramatika: lehen
urratsak [EGLU], vol. 1.  Bilbo: Euskaltzaindia.

Rotaetxe, Karmele (1978).  Estudio estructural del euskara de Ondárroa.
Durango: Leopoldo Zugaza.

Villasante, Luis (1972).  La declinación del vasco literario común.
Aranzazu: Editorial Franciscana.

The view of the Basque noun phrase that I express here is also reflected,
to the extent that it is relevant, in my didactic grammars in English:

Alan R. King, The Basque Language: A Practical Introduction, University of
Nevada Press, Reno, 1994.

Alan R. King & Begotxu Olaizola Elordi, Colloquial Basque, Routledge,
London, 1996.

Alan R. King, Ph.D.
alanking at
- - - - - - - - - - - - - - - - - - - - - - - -  
SNAIL: Orkolaga plaza 3 1A, 20800 Zarautz, Basque Country, Spain.
PHONE: +34-943-134125   /   FAX: +34-943-130396
Alternative email addresses:   mccay at, a at,
70244.1674 at
Internet:   <>

More information about the Lingtyp mailing list