Word Order
I. Basque has a flexible constituent word order, i.e. the arguments can occur in any order. The only restriction on the clause units is that no constituent can intervene between the verb and the auxiliary in non-negative sentences.
Here is an example of the sentence in Basque with all the possible alternations in the word order (J.Ortiz de Urbina 1989:85):
Jon-ek askutitz-a bidali z-u-en
John-Erg letter-def.sg.Abs send 3sg.Erg-(3sg.Abs)-aux2-pst
‘John sent the letter.’
The sentence above can appear with any of the listed below alternations in the word order:
Askutitza Jonek bidali zuen.
Askutitza bidali zuen Jonek.
Jonek bidali zuen askutitza.
Bidali zuen askutitza Jonek.
Bidali zuen Jonek askutitza.
It is not strange that Basque has flexible word order. Due to its morphologically rich case marking system it doesn’t have to identify grammatical functions by positional relations among verb and arguments.
At the same time the precise analysis of Basque grammar and discourse functions reveals that different word orders convey different informational import and, especially, that particular semantic differences can be associated in a principled way with particular changes from what speakers seem to consider ‘neutral’ order. All this clearly indicates that permutations in the word order in Basque are not equal. Thus, let’s consider the two sentences from the example above, which differ only in the relative order of subject and object:
a. Jon-ek askutitz-a bidali z-u-en
John-Erg letter-def.sg.Abs send 3sg.Erg-(3sg.Abs)-aux2-pst
‘John sent the letter.’
b. Askutitza Jonek bidali zuen.
If Basque word order were actually ‘free’, we would expect these sentences to be equivalent, while in fact they differ from each other in some important ways (J.Ortiz de Urbina 1989: 85). While (a) can be given a neutral intonation similar in the relevant aspects to the one in its English gloss, (b) with a word order differing from neutral SOV, must be assigned a marked intonation. In (b) either eskutitza ‘letter’ and Jonek ‘John’ are given a listing intonation and the verb is assigned primary stress, bearing the focus of information; or, more likely, eskutitza ‘letter’ is given neutral stress, while a pause separates both constituents. In the latter case we are more likely to be dealing with topicalization of the first constituent.
II. Thus the basic constituent word order is considered as being Subject – Object – Verb.
In order to support the SOV basic order we can apply various criteria:
According to de Rijk (1969:321-325) for a sentence with transitive verb, the SOV order is statistically as frequent as the other five taken together. Although based on observation of Rotaetxe (1978) this might not be the case for all dialects where in some varieties of Basque SOV and SVO have the same relative frequency.
The basic constituent order of a language typically occurs in utterances that have the least formal marking.
As have shown in the two sentences above (repeated here fore convenience):
a. Jon-ek askutitz-a bidali z-u-en
John-Erg letter-def.sg.Abs send 3sg.Erg-(3sg.Abs)-aux2-pst
‘John sent the letter.’
b. Askutitza Jonek bidali zuen.
(a) can be given a neutral intonation, while (b) with a word order differing from neutral SOV, must be assigned a marked intonation adding some extra phonological material to the neutral intonation.
It is also argued that if native speakers were asked to translate without context the English sentences ‘I can give the book to John’, or ‘Michael made the bomb’ they would propose translations with SOV constituent order (De Rijk 1968). This fact shows that SOV constituent order could be considered as prototypical.
Pragmatically neutral contexts shape into SOV pattern
The analysis of discourse functions in Basque reveals that the clearest distinction between possible combinations of word order derives from certain pragmatic factors. One process that influences the word order is the focus position. Thus we find that whenever some element is emphasized as the focus or new information element, it will occur in the position immediately preceding the verb (J.Ortiz de Urbina 1989: 205-447). Another process that affects the word order in Basque is topicalization that can take any element and place it at the beginning of the clause, separated from the rest of the clause by a pause (de Rijk 1978).
It is important to notice that not all sentences have focused or topicalized constituents and that it is perhaps in these more ‘neutral’ sentences where order constraints may reflect ‘grammatical’ as opposed to ‘pragmatic’ restrictions on word order.
SOV has rather unanimously been recognized as the pragmatically ‘neutral’ word order in Basque and so it is proclaimed in the official grammar issued by the Academy of the Basque Language Euskaltzaindia (Euskaltzaindia 1985: 43).
III. Typologically, the language fits well into the SOV pattern. Let’s consider the noun with all its possible modifiers and relative position of main verb and auxiliary that according to Greenberg’s universals classify Basque along with SOV language.
The analysis of affixation in Basque reveals an interesting property grouping Basque with exclusively suffixing languages. In addition, all Ps appear to the right of the head, identifying themselves as postpositions. This is shown in the following examples (from J.O. de Urbina 1989: 74-75):
zu-re aurka
you-gen against
‘against you’
etxe-tik at
house-abl. outside
‘outside the house’
bendurr-ik gabe
fear-part without
‘without fear’
mendi-an zehar
mountain-loc. across
‘across the mountain’
Therefore, Basque falls into the pattern described by Greenberg’s Universal 27 "if a language is exclusively suffixing, it is postpositional". The postpositional character of Basque is supported by Greenberg’s Universal 4 that shows the following typological tendency: "with overwhelming greater chance frequency, languages with normal SOV order are postpositional.
In Basque genitives occur to the left of the head, marked by the genitive case – (ar)en, as shown in the following examples (J.O.de Urbina 1989: 68):
ama-ren diru-a
mother-sg.gen money-sg.Abs
‘mother’s money’
zakurr-en buztan-ak
dog-pl.gen tail-pl.Abs
‘the dogs’ tails’
problem-en azalpen-a
problem-gen explanation-sg.Abs
‘the explanation of the problem’
Chomsky-ren azalpen-a
Chomsky-gen explanation-sg.Abs
‘Chomsky’s explanation’
The described head final position of genitive constructions in Basque is supported by a typological generalization demonstrated by Greenberg’s Universal 2 where in languages with postpositions the genitive almost always precedes the governing noun.
Adjectives appear immediately to the right of the head noun and bear a determiner or case marker since only noun phrases can be inflected in Basque. The fact that all inflectional categories are marked by clitics attached to the whole noun phrase, and in particular to the adjective following the noun is described in Greenberg’s Universal 40 stating that "when the adjective follows the noun, the adjective expresses all the inflectional categories of the noun. In such case the noun may lack overt expression of one or all of these categories." The adjective position and NP marking is shown in the following examples (from Saltarelli, M 1988: 81-82):
etxe zuri-a
house white-def.sg.Abs
‘the white house’
exte zahar haundi-etan
house old big-def.pl.loc
‘in the large old houses’
gona gorri estu luze-ak
skirt red tight long-def.pl.Abs
‘the long tight red skirts’
Examples (from Saltarelli, M 1988: 75)
ume txiki-ek lan haundi-a ema-ten d-u-te
child small-pl.Erg work big-sg.Abs give-hab 3Abs-(prs)-aux2-3pl.Erg
‘Little children make for a lot of work’.
There is no difference between the form of a given adjective used predicatively as opposed to attributively (from Saltarelli, M 1988: 247):
mutiko-a bihurri-a d-a
child-sg.Abs naughty-sg.Abs 3Abs-prs-(be)
‘The child is naughty’.
mutiko bihurri hori ezin d-a geldi-rik ego-n
child naughty that unable 3Abs-prs-(aux1) quiet-adv be-prf
‘That naughty child cannot keep sleep’.
As we see from the examples the preceding head position of adjective in the adjective phrase do not support the head final pattern of the language but at the same time doesn’t disapprove it since based on the discussion by Greenberg in SOV languages adjective may either follow or precede the head noun with more or less equal distribution.
We also find in the data that a few adjectives may appear to the left of the head noun (from Saltarelli, M 1988: 75):
amerikar hiri-a / hiri amarikar-ra
American city-def.sg.Abs city American-def.sg.Abs
Hence, Basque contributes evidence to the first part of Greenberg’s observation formalized in his universal 19: "when the general rule is that the descriptive adjective follows, there may be a minority of adjectives which usually precede, but when the general rule is that descriptive adjectives precede, there are no exceptions".
Demonstratives usually appear as the final element of the NP to the right of the head following not only noun but descriptive adjectives as well. This is shown in the following example (from Saltarelli, M 1988: 77):
liburu berri hori-ek oso interesgarri-ak d-i-ra
book new that-pl.Abs very interesting-pl.Abs 3Abs-prs-be
‘Those new books are very interesting’.
Thus we can notice that demonstratives should be grouped together with adjectives in their position towards the head where both of them correspond to Greenberg’s Universal 20 stating that when any or all of the items such as descriptive adjectives, numerals, and demonstratives follow the noun, they are always found in that order. And though we will show later that numerals should be excluded from the list of items (they usually precede the head noun) still the pattern described in the universal works perfectly well.
Quantifiers may be divided into two groups on the basis of the position, which they assume with respect to the head of the NP.
bi etxe - 2 houses
hiru etxe - 3 houses
lau etxe - 4 houses
Example: zenbait etxe ‘a number of/ some houses’
Example: etxe bat ‘one house/ a (certain) house’ (depending on the stress)
Example: etxe asko ‘lots of houses’
etxe gutxi ‘few houses’
As we see the quantifiers split between two positions proving the typological tendency of numerals to precede the head noun even when the descriptive adjective follow the noun (Greenberg 1978).
Relative clauses usually appear to the immediate left of the head noun.
Examples (from J.O. de Urbina 1989: 71):
Bizi naiz etxe-a XVIII. mende-ko-a d-a
live aux1-comp house-Abs century-of-Abs 3Abs-prs-(be)
‘The house where I live dates from the XVIIIth century’.
Numerals and quantifiers which appear to the left of the head noun may intervene between the relative and the head, where no other element may appear.
Examples (from Saltarelli, M 1988; 76):
irakurr-i d-it-u-da-n zenbait berri-k kezka-tu n-a-u-te
read-prf 3abs-(prs)-Abs.pl-aux2-1sg.Erg-comp several new-Erg worry-prf 1sg.Abs-prs-aux2-3pl.Erg
‘Certain news which I have read worried me’.
The position of relative clauses supports the head final, postpositional character of Basque and is demonstrated in Greenberg’s Universal 24: "If the relative expression precedes the noun either as the only construction or as an alternative construction, either the language is postpositional or the adjective precedes the noun or both".
Thus summing up the discussion about nouns and their modifiers we can identify the basic structure of a simple NP as the following:
genitive + (relative) + (enumerator) + N + (adjective) + determiner
where determiners include demonstratives, quantifiers and an article.
7. Auxiliary verbs.
As we have shown in the first part the position of the auxiliary is strictly restricted to the position immediately following the main verb. This is shown in the following example (J.O. de Urbina 1989: 85):
Jon-ek askutitz-a bidali z-u-en
John-Erg letter-def.sg.Abs send 3sg.Erg-(3sg.Abs)-aux2-pst
‘John sent the letter.’
*Jonek askutitza zuen bidali.
The described restriction in the position of the auxiliary verb proves the typological pattern revealed in Greenberg’s Universal 16 stating that "... in languages with dominant order SOV, an inflected auxiliary always follows the main verb".
IV Typological non-markedness of Basque.
We have shown that genitives, relative clauses and enumerators, as well as suffixation, postpositions and auxiliary verbs proved the head-final character of the Basque language and consequently its unmarked nature. We also find that adjectives and determiners though they occur in the following head position still fall into the category of typologically usual pattern for SOV languages. Therefore, we may conclude that Basque is unmarked with regard to nouns, their modifiers and relative position of the main verb and auxiliary.
Abbreviations
Verb morphology
1, 2, 3 – person marker
sg. - singular number
pl. - plural number
prs - present tense
pst - past tense
prf - perfective
hab - habitual
fut - future
aux1 - intransitive auxiliary
aux2 - transitive auxiliary
mod. - modal auxiliary
comp - complementizer
Nominal morphology
Erg - ergative
Abs - absolutive
Dat - dative
gen - genetive
loc - locative
abl - ablative
all - allative
part - partative
mod - modal
rel - relative
Other
emp - emphatic marker
det - determiner
prx - proximal
quant - quantifier
inform - informal
form - formal
References
Back to languages table of contents
Back to topics
Revised April 12, 1999