See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/290460594

Non-argument agreements: An approach to Basque allocutivity

Article · January 2015

CITATIONS READS
4 134

2 authors, including:

Aitor Lizardi Ituarte

Universidad del País Vasco / Euskal Herriko Unibertsitatea
5 PUBLICATIONS 7 CITATIONS

SEE PROFILE

All content following this page was uploaded by Aitor Lizardi Ituarte on 18 January 2018.

The user has requested enhancement of the downloaded file.

 NON-ARGUMENT AGREEMENTS:
AN APPROACH TO BASQUE ALLOCUTIVITY1
Aitor Lizardi (EHU) & Manex Munduate

0. Objectives

The aim of this article is to offer an overview of a phenomenon in Basque called

allocutivity (a special agreement morpheme which references the person hearing the
sentence, who is not an argument of the sentence) and to try to find an explanation for
this or, at least, to show the key features of this phenomenon which must be taken into
account when working on an approach to any such explanation.
To this end, in the first part we will introduce allocutivity, starting with its general
properties (1); then we will look at morphological issues (2) – both the position (2.1.) and
the form (2.2.) of the morpheme – and we will also discuss the ways in which allocutive
morphemes are introduced in inflections (2.3.), paying special attention to morpheme -i-
(2.3.1.); we will then address its syntactic properties (3) and its dialectal variations (4).

1. Basque Allocutivity: An overview

The allocutive is found in second person singular morphemes which appear in verb
inflections. Their distinctive feature is that, although they have the same form of the
dative and ergative agreements, they appear when there is no second person singular
argument.
(1) a. Itziar heldu da
arrive 3SG.AUX2
b. Itziar heldu du-n/-k
3SG.AUX-ALLO.FEM/-ALLO.MASC
c. Itziar heldu du-zu
3SG.AUX-ALLO.FORM

1 We are grateful to Beatriz Fernández for giving us the opportunity to take part in Wedisyn 2013 workshop
and to Gillen Martinez de la Hidalga and Igor Rueda Arego for their advices and corrections. All errors are
ours.

2 In this paper we will use the following abbreviations: ABS: absolutive; AFF: affective; AF: allocutive flag,
ALLO: allocutive; ASP: aspect; AUX: auxiliary verb; C: complementizer; Cen: central dialect; Cond:
conditional; DAT: dative; DF: dative flag; DP: determiner phrase; EN: eastern Navarrese dialect; EP:
epenthesis; ERG: ergative; Err: variety of Erronkari; FEM: femenine; FORM: formal; GEN: genitive; HN: high
Navarrese dialect; L: Labourdin; LN: low Navarrese dialect; MASC: masculine; MOD: modal; NALLO: non-
allocutive; NREF: non-reflexive; non-res: non-restricted allocutive; P: adposition; part-res: partially
restricted allocutive; PL: plural; PLZ: pluralizer; REF: reflexive; REL: relative; ROOT: verb root; S: Souletin
dialect; SG: singular; T: tense; TP: tense phrase; W: western dialect; Za: variety of Zaraitzu.
Non-argument agreements: an approach to Basque allocutivity

d. Itziar heldu du-xu

3SG.AUX-ALLO.FEM/-ALLO.MASC
‘Itziar has arrived.’

All the above sentences are equivalent in terms of meaning. The difference between them
is the use of different morphemes in the verb inflections, morphemes which reference the
receivers of each sentence. In (1a) the sentence is a neutral statement, while in (1b, c, d)
different morphemes have been added to the neutral inflection: (1b) uses -n to express an
informal relation with a female hearer, and -k with a male hearer; in (1c) -zu is used to ex-
press a formal relationship and in (1d) -xu is used to express an intermediate relationship
in terms of intimacy.
Before exploring the allocutive in more detail, we will examine all the second per-
son singular treatments and the agreement system in Basque (see Table 1).

Table 1. 2nd person morphemes in Basque verb agreement.

Pronoun Gender Word internal Word final
masculine -a- -k
Hi
femenine -na- -n
Xu both -xu(-)
Zu both -zu(-)
(Ber)ori both Ø (3rd person)

As can be seen in table (1), there are three types of second person singular in Basque.
First of all, hi, 'thou', the informal pronoun, which is used in colloquial contexts. Zu,
'you', is a more formal pronoun, which is used nowadays as a neutral treatment and is
replacing the pronoun hi in many varieties of the language. Some eastern dialects have
also an intermediate pronoun, xu, a palatalised version of zu, which is used in intimate
contexts, such as children's speech. Finally, in southern dialects (ber)ori can be found, a
third person pronoun with corresponding third person agreements; however, we will
leave this to one side because it does not lead to using the allocutive in any variety of
Basque.
As it is well known, Basque is an ergative-absolutive language. Furthermore,
there are multiple agreements in its verb inflections, so every argument has its equivalent
morpheme inside the inflection. So all three arguments agree in the verb inflection: first
the absolutive, next the dative, and, finally, the ergative.
(2) Ni etorri n-aiz
I.ABS come 1SG.ABS-AUX
(3) Ni-ri esan di-da-te
I-DAT tell AUX-1SG.DAT-3PL
(4) Ni-k egin du-t
I-ERG do AUX-1SG.ERG
‘You have done it.’

In (2, 3, 4) the verbs agree with the arguments, a rule which all verbs have to comply
with. The previous examples had no second person argument: however, such sentences

2
 Lizardi & Munduate 2016

can contain second person morphemes in the inflection and those morphemes are the
allocutive. In (2') we take example (2) and add the allocutive morpheme.
(2') a. Ni etorri n-au-n/k
I.ABS come I.ABS-AUX-ALLO.FEM/MASC

b. Ni etorri n-au-zu
I.ABS come I.ABS-AUX-ALLO.FORM
c. Ni etorri n-au-xu
I.ABS come I.ABS-AUX-ALLO.AFF
‘I have come.’

The same sentence can be built in three different ways. The use of each allocutive
depends on the way in which the sender of the sentence addresses the hearer during the
conversation.
However, different ways of addressing hearers do not affect the language in the
same way. Firstly, the use of hi during a conversation means that the allocutive has to be
used in every case possible, including sentences in which there are not second person
arguments. This feature is common to all Basque dialects and varieties. On the contrary,
zu and xu do not have to follow this rule in most of the language's varieties, although in
some of them the use of the allocutive is mandatory with these pronouns too.

Table 2. Allocutive forms of addressing hearers in different dialects.

Dialects HI ZU XU
Allo Nallo Allo Nallo Allo Nallo
W-C-HN-L ✓ X X ✓ X X
LN-Za ✓ X ✓ ✓ ✓ ✓
S-Err ✓ X ✓ ✓ ✓ ✓

As table (2) shows, hi is the only pronoun which requires allocutive in all dialects.
Moreover, hi cannot be used in any variety without allocutive agreement. Therefore,
hereafter we will use hi and its corresponding female morpheme in every example as all
allocutive ways of addressing hearers follow the same patterns.
It should be pointed out that there are no forms of address in any single variety in
which both allocutive and non allocutive uses are possible. A sender who has decided to
use an allocutive form of address must be coherent and if he/she does not use the
allocutive later, in a single sentence or a subsequent one, the sentences in question
become ungrammatical. (3) is an example of this.
(3) *Etxera noa-n ./eta ondoren, lan egingo dut.
Home.to go-ALLO ./and then work do AUX.NALLO
‘I have gone home ./and then, I worked.’

After this general view of the phenomenon, we will now examine it in greater depth. First
of all, we will look at its morphology.

3
Non-argument agreements: an approach to Basque allocutivity

2. A Morphological Approach

Our morphological description will focus on the allocutive morphemes. First we will
analyse morpheme forms and, then, their position inside verb inflections.

2.1. Form

One of the most significant facts about allocutive morphemes is that they are identical in
form to second person agreement marks. Here is an example of this.
(4) a. Eskatu dio-na-n faborea handiegia dun
ask AUX-2SG.ERG-REL favour big.too AUX.ALLO
‘The favour that you asked him/her for is too big.’
b. Eskatu di-na-n faborea handiegia dun
ask AUX.-2SG.DAT-REL favour big.too AUX.ALLO
‘The favour that he/she asked you is too big.’
c. Faborea eskatu ø-zi-o-na-t
Favour ask (3SG.ABS)-AUX-3SG.DAT-ALLO-1SG.ERG
‘I have asked him/her a favour.’

Here we are going to focus on the morpheme -na-. As the notes show, the same -na- form
has different values in each sentence. In all three sentences it is a second person feminine
mark, but in (4a) it is an ergative agreement, in (4b) a dative agreement and, finally, in
(4c) it is an allocutive mark.
This syncretism can also be seen in word-final positions.
(5) a. Faborea eskatu dio-n
favour ask AUX-2SG.ERG
‘You have asked her/him a favour.’
b. Faborea eskatu di-n
favour ask AUX-2SG.DAT
‘She/He has asked you a favour.’
c. Faborea eskatu zio-n
favour ask AUX-ALLO
‘She/He has asked her/him a favour.’

Again, the ergative (5a), dative (5b) and allocutive (5c) morphemes take the same form.
Only the word-initial position breaks this law.
(6) a. Etxera joan h-aiz
home.to go 2SG.ABS-AUX
‘You have gone home.’
b. Lan egin h-uen
work do 2SG.ERG-AUX
‘You worked.’
c. Etxera etorri zu-na-n
home.to come AUX-ALLO-T
‘He/She came home.’

4
 Lizardi & Munduate 2016

In fact, only the absolutive agreement and, under specific conditions 3, the ergative
agreement are in word-initial positions. As the allocutive cannot appear in word-initial
positions, the forms are not the same in these cases.
In summary, the following scheme:

Table 3. Second person inflection morphemes.

Person Word-initial Word-internal Word-final
Absolutive h-, z- - -
Ergative h-, z- -a-, -na-, -zu-, -xu- -k, -n, -zu, -xu
Dative - -a-, -na-, -zu-, -xu- -k, -n, -zu, -xu
Allocutive - -a-, -na-, -zu-, -xu- -k, -n, -zu, -xu

There is considerable evidence that all word-internal and word-final morphemes have the
same origins (Gómez & Sainz 1995: 249-260). Looking at Table (3), it seems that all the
word-internal and word-final morphemes are the same and their function (ergative, dative
or allocutive) is determined by their position inside the inflection.

2.2. Position

We have already seen that the function of the morpheme is linked to its place in the
inflection more than to the morpheme itself. So in this section we will try to examine in
detail what that place is, taking into account that the allocutive morpheme always takes
the same position inside the verb inflection.
When discussing person agreements, some other linguists have already drawn this
conclusion. Let us explain it step by step with an example.
(7) Nik amari dirua osten zionat
I.ERG mother.DAT money.(ABS) steal.ASP AUX
‘I steal money from my mother.’

(7) is a ditransitive sentence and, as a consequence, the inflection has all three argument
agreements as well as the allocutive. We will examine this by dividing the inflection into
morphemes:
(8) z-: third person absolutive.
-i-: dative flag, DF, a morpheme preceding a dative agreement.
-o-: third person dative.
-na-: feminine informal allocutive.
-t: first person ergative.

As (8) shows, the allocutive's position between the argument agreements is after the
dative (-o-) and before the ergative (-t). Some linguists, such as Eguren (2000) and Albizu
(2002) have concluded this and have defined the Basque agreement system as
(9) Absolutive-Root-Dative-Allocutive-Ergative

3 Ergative displacement. In non-present tense transitive and ditransitive inflections, when the absolutive is
third person and the ergative is first or a second person, ergative agreement happens in the first position,
using word-initial agreement marks. For more on this phenomenon see Hualde & Ortiz de Urbina (2003:
222-223); Laka (1993a)

5
Non-argument agreements: an approach to Basque allocutivity

But there are more morphemes than argument agreements in Basque verb inflections and
the description in (9) is insufficient for our analysis. So we will try to define the position
of the allocutive taking into account all Basque verb inflection morphemes.
One of the morphemes that the allocutive is used with is the modal -(te)ke-. It
takes the position between the dative (or the root, in its absence) and the allocutive. In
(10) the modal -(te)ke- is shown italics and the allocutive is underlined.
(10) a. nai-teke-n
b. zeza-ke-na-t
c. litzaio-ke-n
d. zakio-ke-na-n
e. niezaio-ke-na-n

Nevertheless, there is also another morpheme which precedes the allocutive. That is the
-te pluralizer. However, three allomorphs of -te pluralizers can be observed in the
inflection.
1. An absolutive pluralizer, whose form is -de, which appears just in a few cases.
(11) a. zau-de-n
b. geun-de-na-n

2. A dative pluralizer, whose form is a -e, -te in some central varieties.

(12) a. zai-ø-e-n
b. zai-o-te-n (some central varieties)

3. An ergative pluralizer, whose form is a -te, -(d)e in non-central dialects.

(13) a. di-te-n
b. zezake-te-na-n

The difference in all of them is that the modal -(te)ke- goes after the absolutive and dative
pluralizers but before the ergative pluralizer.
(14) a. leu-de-ke
ROOT-ABS.PLZ-MOD
b. litzai-e-ke
ROOT-DAT.PLZ-MOD
c. lu-ke-te
ROOT-MOD-ERG.PLZ

And the allocutive goes after the ergative pluralizer.

(15) li-ke-te-n
AUX-MOD-ERG.PLZ-ALLO

In summary, the order of the elements in Basque verbal inflection is

(16) ABS/ERG4-ROOT-DF-DAT-e-(te)ke-te-ALLO-ERG-T

4 Ergative mark just in case of ergative displacement. See note 2.

6
 Lizardi & Munduate 2016

2.3. Changes in Inflection

The insertion of the allocutive usually leads to some type of change in the verb inflection.
First of all, in order to get a better understanding of this, the Basque auxiliary verb
system must be explained in detail (see Table 4).

Table 4. Auxiliary system in Basque.5

Intrasitive Transitive
Indicative izan ('be') *edun ('have')
Non-indicative *edin ('come') *ezan

Taking Table (4) into account, there can be three types of changes in the inflection.

1. Intransitive → transitive → ditransitive

Indicative auxiliaries, if they do not have a dative agreement, change in the following
ways:
a. The intransitive auxiliary izan becomes the transitive *edun. In this case, the
allocutive morpheme occupies the same place of the ergative, compared with real
transitive verbs, most of the time.
(17) a. na-iz → na-u-n
I.ABS-izan I.ABS-edun-ALLO
b. na-iz → na-u-n
I.ABS-izan I.ABS-edun-2SG.ERG

b. The transitive auxiliary *edun takes the form of a ditransitive auxiliary: the root
-u- disappears and an -i- appears instead. This change happens both when the dative is
inserted and also when the allocutive is inserted. So the allocutive morpheme seems, at
first sight, to take the place of the dative.
(18) a. d-u-t → d-i-na-t
(3SG.ABS)-edun-1SG.ERG (3SG.ABS)-i-ALLO-1SG.ERG
b. d-u-t → d-i-na-t
(3SG.ABS)-edun-1SG.ERG (3SG.ABS)-i-2SG.DAT-1SG.ERG

However, this change only occurs in central dialects.

2. No change

There is no change with the auxiliary *edin (19) or with izan when there is dative
agreement (20)

(19) a. na-i-teke → naiteke-n

1SG.ABS-edin-MOD -ALLO

5 Here the asterisk ‘*’ does not make reference to any ungrammaticality, but to the fact that these verb
forms have been reconstructed; that is, they have not been attested in the before mentioned forms.

7
Non-argument agreements: an approach to Basque allocutivity

b. da-ki-o-ke → dakioke-n
(3SG.ABS)-DF-3SG.DAT-MOD -ALLO

(20) za-i-o → zaio-n

(3SG.ABS)-izan-DF-3SG.DAT -ALLO

3. The insertion of the morpheme -i-

In the remaining cases, an -i- is inserted before the root, giving the verb an allocutive
morpheme. We will call this the allocutive flag, AF. This happens with the auxiliary
*ezan (21) and with synthetic verbs (22).
(21) d-eza-ke-t → d- -i- -ezake- -na- -t
(3SG.ABS)-ezan-MOD-1SG.ERG -AF- -ALLO6-

(22) da-tor → d-i-a-torr-e-n

(3SG.ABS)-come (3SG.ABS)-AF-come-EP-ALLO

Therefore, the -i- of the pseudo-ditransitive verbs with the *edun root can also be
interpreted as an allocutive mark.
(18a') d-u-t → d-i-na-t
(3SG.ABS)-AF-ALLO-1SG.ERG

The use of the AF is different in each dialect, while there are some verbs whose AF is the
same in all dialects.
(23) n-i-ai-teke-k (Oyharçabal, 1993: 17)
1SG.ABS-AF-edin-MOD-ALLO

(24) Ene buruya c-i-ay-da-çu harendaco abastu

My head (3SG.ABS)-AF-izan-1SG.DAT-ALLO that.to sufficient
‘My person will suffice me for that.’ (Etxepare 1995: Amoros Jelosia)

We have already seen ways of inserting the allocutive in the verb inflection. The two first
ways do not need any further explanation but the last one, the insertion of another
morpheme, requires an explication with regards to the nature of that morpheme. We will
address this in the following section.

2.3.1. -i-: dative flag (DF) or allocutive flag (AF)?

First of all, before a more in-depth analysis of this question, two concepts must be
understood clearly: the DF and what we have called the AF.

1. The Dative Flag (DF)

We will examine this by looking at example (11) once again.

6 As Basque's verbal morphology is rich, we did non see necessary to explain every morpheme in the
glosses. In our opinion, that would be unnecessary and too complicated and therefore we prefer to focus on
the meaningful parts.

8
 Lizardi & Munduate 2016

(8) z-: third person absolutive.

-i-:DF, a morpheme preceding a dative agreement.
-o-: third person dative.
-na-: feminine informal allocutive.
-t: first person ergative.

In this verb there is an -i-, a morpheme which does not mark any agreement or other
functional head. It is the DF, that is, a morpheme that marks the presence of a dative in the
verb inflection.
The dative flag has two allomorphs: -i- and -ki-. It is always positioned before the
dative agreement mark.
(25) a. za-i-o
AUX-DF-DAT
b. dator-ki-t
come-DF-DAT

There are two exceptions regarding its position: in ditransitive verbs with the auxiliary
*ezan (26a) and with the -zki- pluralizer (26b). The latter case can be explained as
follows: -z- (pluralizer) and -ki- DF, although there is also another DF before it.
(26) a. d-i-eza-zu-ke-t
(ABS)-DF-AUX-DAT-MOD-ERG
b. d-i-z-ki-o-t
ABS-DF-PLZ-DF-DAT-ERG

Some linguists consider that the DF is actually an applicative (Elordieta 2001: 61-64;
Fernández 2012, 2014) or a preposition (Fernández 2013).

2. The Allocutive Flag (AF)

We call the morpheme used for the same function as the allocutive the AF. It takes the
position before the root and its only allomorph is -i-.
(27) a. d-i-auka-na-t
(ABS)-AF-ROOT-ALLO-ERG
b. d-i-ago-n
(ABS)-AF-ROOT-ALLO

This is the first time that the term allocutive flag has been used. The fact that the DF and
the AF may be considered to be a single morpheme or not has been discussed at length. In
this article our position is that, in fact, we should talk about two morphemes.
There is more than one reason to consider this analysis. First of all, the position:
(28) a. Dator-ki-t a'. d-i-atorre-n
b. D-i-eza-i-o-kegu b'. z7-i-eza-i-o-ke-na-gu

7 As we will point out in section 4, in both central dialects and in standard Basque the union of the initial
d- and the posterior -i-, the AF, leads to z-: di- > z-. In the following examples a z- indicates the union of
both morphemes and, so, an AF.

9
Non-argument agreements: an approach to Basque allocutivity

As we have said before, the DF appears before the dative (28a) while the AF's position
precedes the root (28a'). We said that there is one case in which the DF precedes the root
(28b); but there, too, the allocutive can appear and the AF's position is before the DF. The
syncretism between the indicative ditransitive and the transitive with the allocutive could
be solved explained in this way.
(29) a. d-ø-i-na-t (with dative)
(ABS)-edun-DF-DAT-ERG
b. d-i-ø-na-t (with allocutive)
(ABS)-AF-edun-ALLO-ERG

In fact, in most dialects there is no such match even between those forms because they
use other mechanisms to mark the insertion of the dative.
(30) a. Deu-na-t (+dat) a'. J-o-na-t (+allo)
(Badihardugu elkartea, 2005b: 4-5) (W)
b. Da-na-t/d-i-na-t (+dat) b'. D-i-ne-t (+allo)
(Salaburu, 2005: 117, 121) (Central-eastern varieties)

Secondly, the form of both morphemes is also different. While the DF has almost three
allomorphs, the AF remains the same in all contexts.
(31) a. d-i-o-t a'. z-i-o-na-t
b. dator-ki-t b'. d-i-atorre-n
c. dio-ts-o c'. z-io-na-t

Thus, if both were a single morpheme the verbal forms which differ only in the
allocutive and the 2nd person dative would be syncretic. However, this does not happen in
most cases.
(32) neutral +dat +allo
a. dator a'. dator-ki-n a''. d-i-atorre-n
b. dauka b'. deko-s-ta-zu b''. d-i-auka-n

Finally, the assumption that both morphemes are one means that actually the allocutive is
a kind of dative. We do not see this as a possibility. With regards to their form, it is true
the morphemes are identical, but it should be born in mind that all word-internal and
word-final morphemes take the same form, as we explained in Table (3). With regards to
position, they do not take the same place. The -te pluralizer divides both marks. In (33),
take into account the position of -n(a)-.
(33) a. di-na-te (+dat)
b. d-i-te-n (+allo)

So all those arguments lead us to think that the allocutive and the dative – and, as a
consequence, the DF and the AF– are two different morphemes that have their own
functions inside verb morphology. We understand that both morphemes may have the
same origin, but believe that they have had different grammaticalizations and that this has
led to two different morphemes, with different specializations as a result. Rebuschi
(1984), for example, reached a similar conclusion and Lakarra (2008) also points out in
the same direction.

10
 Lizardi & Munduate 2016

3. Syntactic behavior

From a syntactic point of view, the allocutive shows some distinctive features which
differentiate it from argumental agreements. Its non-argumental nature is summarized by
Oyharçabal (1993) and we will use that article to mark its most significant features.
Firstly, the allocutive can be found in ditransitive sentences. This shows that the
allocutive is independent from the arguments of the sentence.
(34) Nik Joni sagarra erosi z-i-o-na-t
I.ERG Jon.DAT apple.ABS buy (ABS)-DF-DAT-ALLO-ERG
‘I have bought an apple from Jon.’

Secondly, in contrast to argumental agreements, there cannot be any DP sentences with

allocutive. In (35), the ungrammaticality of the sentence is caused by the apparition some
DPs which refer to the allocutive.
(35) *pro Hi/hik /hiri mintza niaiteke-k (Oyharçabal 1993: 17)
proI you.ABS/ERG/DAT speak AUX-ALLO
‘I can speak.’

Finally, some genitive pronouns give rise to the same point of view. Until almost the 18 th
century, there were two separate types of personal genitive pronouns in Basque: reflexive
and non-reflexive pronouns. If a sentence had a genitive with reference to an argument, it
had to take the reflexive form. This phenomenon is called the Aresti-Linschmann law 8.
Nevertheless, the allocutive did not activate the reflexive genitives and the genitives
which made reference to the allocutive had to have the non-reflexive form.
(36) Hire / *heure ama ikusi di-a-t
NREF.GEN REF.GEN mom see AUX-ALLO-ERG
‘I have seen your mother.’

There is another important syntactic feature to the allocutive: its syntactic restrictions. As
we examine in depth in section 4, nowadays there is considerable variation even between
local varieties, but traditionally in Basque allocutive has been restricted to main clause
statements. So its use in embedded (37a), interrogative (37b), imperative (37c) and
exclamative (37d) clauses was ungrammatical and it still is nowadays in some varieties.
(37) a. * Ez dakinat zer gertatu dunan (Oyharçabal, 1993:21)
b. * Lan egin dina hire lagunak? (Oyharçabal, 1993:21)
c. * Betorren
d. * Zer polita dunan!

This point leads us to think that this restriction is influenced by complementizers. This
point is essential in order to approach a syntactic explanation of the phenomenon. In this
article we will restrict ourselves to explaining two theories about it: Oyharçabal (1993:
24) and Miyagawa (2011: 14).

8 For more information about the law and its application, we recommend Sarasola (1979), specially the
third section.

11
Non-argument agreements: an approach to Basque allocutivity

The difference between the two proposals is the position of the allocutive in the
inflection. Oyharçabal (1993) believes that the allocutive is an operator adjuncted to the
TP. On the other hand, Miyagawa (2011) prefers to think of it as a complementizer, trying
to explain its syntactic constraints.
Having seen both proposals, we think that Oyharçabal, despite not being able to
give a detailed explanation of the syntactic constraints, is the one which fits best with our
hypothesis. The allocutive does not seem to be a funtional head, as its non-argumental
nature shows. Furthermore, Miyagawa's proposal does not take into account the order of
the elements in the verb inflection. However, Oyharçabal does respect the order of
elements and this matches perfectly with Laka's (1993) description of the building of
Basque verb inflections.

Therefore we think that Oyharçabal's (1993) proposal is better, although we know that it
does not explain all the syntactic constrains. However, we believe that analysis can be
developed to achieve a more complete syntactic explanation of the phenomenon.

12
 Lizardi & Munduate 2016

4. Dialectal Variation

The allocutive is highly varied in different Basque dialects. This variation will be studied
from two points of view. On the one hand, we will describe morphological variations; on
the other, we will examine syntactic variations.

4.1. Morphological Variation

Within morphological variation, we will specify morphological processes produced by

the AF in the allocutive verbs. This phonological evolution of the AF with the preceding
morpheme d- has developed in different ways in each dialect. As a consequence, a rich
morphological variation is to be found. There are two main features which must be
underlined.
Firstly, there is a difference between western Basque and the other varieties in the
morphological visibility of the root of edun 'have'. In most dialects, the presence of the AF
masks the auxiliary root *edun ‘have’:
(40) d-u-Ø → d-i-Ø-n
(3SG.ABS)-ROOT-3SG.ERG (3SG.ABS)-AF-ROOT-ALLO

In the western dialect, however, the AF has been located just after the initial d-, causing a
morphonological process which leads to j- ( di->j-). So the root of the auxiliary *edun
‘have’ can be morphologically identified by means of another allomorph (41b)
(41) a. d-au-Ø → *d-i-au-n-Ø9
(3SG.ABS)-ROOT-(3SG.ERG) (3SG.ABS)-AF-ROOT-ALLO-(3SG.ERG)
b. *d-i-au-n> *j-au-n > j-o-n

(40) is to be found in non-western dialects. The insertion of the AF (-i-) in the verb makes
the root -u- of the auxiliary *edun ‘have’ invisible: it is not produced phonetically. In the
second one (41b), if the initial j- is accepted as the result of the process which happens
with di- (palatalization: [d+i]> [ji]) and the final -k/n morpheme belonging to the
allocutive, the intermediate vowel -o-could be identified as the root of the auxiliary *edun
‘have’. So, in this case, it remains visible, although under another form.
Another development has produced a further rich dialectal variation. The union of
d- and the AF has resulted in different forms, which varies in each dialect and subdialect
(see Figure 1).

9 Note that here again the asterisk does not make reference to ungrammaticality but to reconstructed forms.

13
Non-argument agreements: an approach to Basque allocutivity

(42) di->
a. W: j-: Diaukan > jaukan
b. Cen / HN / L: z-. Diaukan > zaukan
c. Aezkoa / Oñati: x-. Diaukan > xaukan
d. S / LN / Baztan / EN / Mutriku / Legazpi / Ergoiena: di-. Diaukak.

As example (42) specifies, in the western dialect the cluster di- results in j- (diakiat >
jakiat). In the central dialect, high Navarrese and Labourd z- (diakiat > zakiat). In Aezkoa
and Oñati x- (diakiat > xakiat). In Souletin, low Navarrese, eastern Navarrese, Baztan,
Mutriku, Legazpi and Ergoiena di- does not change (diakiat > diakiat).
The morpheme order can also change. The AF -i-, in some verbs, can appear
within the root, which triggers some phonological changes (Rebuschi 1984).
(43) n-i-ago-n → n-ag-i-o-n>n-aj-i-o-n > n-atx-i-o-n
1SG.ABS-AF-root-ALLO 1SG.ABS-root-AF-root-ALLO
‘I am.’ (some central varieties)

Usually the AF appears before the verb root (-ago, ‘be’ in this case). However, in some
circumstances it can be moved into the root (-ag-i-o-), which leads to some
morphonological processes such as palatalization in the central dialect: -agi- > -aji- >
-atxi- (Rebuschi 1984)

14
 Lizardi & Munduate 2016

4.2. Syntactic Variation

When explaining the syntactic features of the allocutive it has been mentioned that
traditionally the allocutive has been used only in main declarative sentences. However,
nowadays these syntactic restrictions are not so rigid.
In the 20th century, the central-western dialects underwent modifications
compared with classical Basque and contemporary north-eastern dialects.
Rebuschi (1984) and Oyharçabal (1993) stated that traditionally the allocutive
could only be used in main declarative sentences. Therefore, it was not to be found in
embedded clauses, questions, exclamations and imperatives. This is what we find in
central-eastern dialects. See examples (44).
(44) a.*Hi, Miren etorri dun?
you Miren come ROOT.ALLO
‘You, has Miren come?’
b. *Gaur bazatorren, esaidan
today CCOND.come.ALLO say.1SG.DAT.2SG.ERG
‘If (s)he comes today, tell me.’
c. *Etorri dunalako egin dinat
come ROOT.ALLO.C do ROOT.ALLO
‘I have done it because (s)he has come.’

On the contrary, in central-western dialects, there is some variation with regards to this.
As Alberdi (1994: 516-517) points out, the speakers of these dialects use allocutive forms
in questions and in conditional clauses as well. However, the allocutive does not appear
with the same frequency in all these cases. The allocutive can be used more easily in
interrogative clauses (44a) than in conditional clauses (44b). Likewise, it appears more
frequently in conditional clauses and in embedded sentences (44c).
The investigation into address forms in Basque carried out by Alberdi (1994)
confirms these conclusions. Alberdi (1994: 516-517) claims that in central-western
dialects the use of the allocutive in questions and conditional sentences is completely
normal. Furthermore, in speakers younger than 45 years old the use of the allocutive
extends to embedded sentences. Nevertheless, this has not spread to all dialects. The
speaker’s age seems to be an important social factor for understanding the evolution of
this change, but not the only factor. This development is the result of several factors over
time, as summarised in Table (5).

Table 5. Respecting traditional constraints in different dialects.

North-east C-W C-W
Syntactic context (restricted, (partially restricted, (non-restricted,
traditional) 20th century) nowadays)
Main declarative ✓ ✓ ✓
Interrogatives X ✓ ✓
Present
X ✓ ✓
conditional
Subordinates X X ✓

15
Non-argument agreements: an approach to Basque allocutivity

Table (5) shows from left to right changes to the syntactic restrictions to the allocutive.
Each column shows what has changed from the former one on its left, indicating which
restrictions have been kept and which ones have changed in each dialect.
The situation shown in the left column is the original one: as all the traditional
syntactic restrictions are in it, it is called the restricted pattern. The allocutive can be used
only in main declarative sentences, not in questions, conditionals or embedded sentences.
Although traditionally this was so in all dialects, restrictions have become less rigid
during the 20th century. Nowadays they are observed only in the north-eastern dialects.
The second column belongs to the partially restricted pattern of the central-western
dialects where the restrictions are not respected completely. In this case the use of the
allocutive has spread from main declarative sentences to interrogative and conditional
sentences, leaving some of the restrictions unused. As Alberdi explains, this happens
mainly with speakers below 45 years old. Lastly, in the third column none of the
restrictions have been maintained, resulting in the non-restricted pattern. Here the last
restriction which appears in the partially restricted pattern, the non-use of the allocutive
in embedded sentences, has been removed. So it has been seen that the allocutive has
moved from respecting all the syntactic restrictions to being used in all contexts
regardless of them. The latter case has been seen mainly amongst the youngest speakers,
as in (44c).
There is no doubt that further and deeper investigation is needed to reach
consistent conclusions. One of the possibilities is to study literary production as this
evolution can be seen there too. Until the 20 th century, canonical restrictions applied. In
the first half of the century, texts still had traditional restrictions and others followed with
the central-western restricted pattern; but there are also texts with no restrictions at all.
Nowadays, the central-western restricted way is used on television and in some literary
production, but in informal context, the use of the Basque allocutives in all contexts is
usual, above all amongst young speakers:
(45) a. Neuk bañekik!
I.ERG C.know.ALLO
‘If I knew!.’ (Bustintza 1986, Abarrak; W, part-res)
b. Valentinek ez al zian horri buruzko zerbait bidali?
Valentin.ERG no C AUX.ALLO that about.GEN something send
‘Didn't Valentin send something about that?’
(Atxaga 1988: 222, Obabakoak; Cen, part-res)

In these two sentences the partially restricted pattern is used. In the (45a) the allocutive
appears in an exclamation, and in (45b) in an interrogation.
Nowadays, as in the central-western region the most commonly-used pattern is
the partially restricted, the restricted pattern is used in more formal contexts like
television or literature. But in informal contexts, the allocutive is not restricted in any
way, and this is mainly heard in young people's speech:
(46) a. Hire maitasun habiatxoa duk hau?
you.GEN love nest ROOT.ALLO this
‘Is this your love nest?’ (Cano 2011: 429, Twist; part-res)

16
 Lizardi & Munduate 2016

b. Zer ote zakik zertan sartuta dabilen Koldo?

what know.ALLO what.in entered be.C Koldo
‘What will he/she know about what Koldo is involved in?’
(Goenkale, episode 817; partially restricted)
c. Etorri dekela esan zidak.
come ROOT.ALLO.C say ROOT.ALLO
‘(S)he has told me that (s)he has come.’
(Informal context; C, non-res)

The first two examples show that even in formal contexts (TV and literature) the
partially restricted pattern can appear, using the allocutive in interrogative sentences. The
third one belongs to an informal context where in the non-restricted pattern the allocutive
is used in an embedded sentence.
Leaving aside central-western dialects, what are the reasons behind the syntactic
restrictions in northern-eastern dialects? In order to answer this question, let us focus on
(47).
(47) Close interrogative
a. Lan egiten duia hire lagunak? (-allo)
work do root.C your friend.ERG
‘Does your friend work?’
b. *Lan egiten dina hire lagunak? (*+allo)
work do root.ALLO.C your friend.ERG
‘Does your friend work?’ (Oyharçabal 1993: 21)

But:
(48) Reforced affirmative
a. Banenki banikek
C.know.NALLO AFF.have.ALLO
‘If I knew, I would have. ’
b. *Bahenki, (*banenkike) banekikek
C.2SG.ERG.know AFF.know.NALLO AFF.know.ALLO
‘If you knew, I would knew. ’ (Oyharçabal 1993: 12)

The aim of these examples (47) and (48) is to study the role of suffixes and prefixes in
the allocutive. Is the use of the allocutive connected with the fact that the verb has a
prefix or a suffix?
The example (47b) is not grammatical because of the presence of the allocutive
together with the -a interrogative complementizer. In (48), the verb banenki does not
allow the allocutive. In this case the ba- prefix is a conditional prefix. So this means that
the position of the affix does not influence the presence of the allocutive because both
(47b) and (48b) are ungrammatical despite having a suffix and a prefix respectively.
So can we conclude that any affix prevents to use of the allocutive? That is to say,
do affixes in verbs inhibit the apparition of the allocutive and are they the main cause of
the syntactic restrictions?
The sentence (48a) Banenki banikek is grammatical, but it would not be so if the
two verbs were in the allocutive form *Banenkik banenkikek. This is why not all the

17
Non-argument agreements: an approach to Basque allocutivity

affixes work the same way, and their features must be studied. In this case, what is the
difference between the prefix ba- on these two verbs?
The prefix in banenki is a complementizer that introduces an embedded
conditional sentence. In banikek it reinforces an affirmation – something similar to I did
know – so it is not a complementizer. This makes it clear that the conditional ba- prefix
which is a complementizer blocks the apparition of the allocutive (*banenkik), but the
affirmative one (banikek) does not. In this case, as in the first verb the speaker is not an
argument, it is not obligatory to insert the allocutive and the sentence banenki banuke is
grammatical too. In the second one [bahenki, banekikek], however, bahenki has the
speaker as argument (if you knew), so that for the sentence to be grammatical the next
verb – banenkike – has to be allocutive (banenkikek). It should also be pointed about that
in this case the ba- prefix does not prevent the allocutive as it is affirmative and not a
complementizer.
As a result, it is the complementizer which leads to embedded sentences being
subject to syntactic restrictions. That is to say, it is the complementizer, overt or not,
which blocks the allocutive and not the presence of any affix (prefix or suffix). In order to
explain the evolution of these restrictions, it would be interesting to study the influence of
geographical and sociological factors on them.

5. Conclusions

These are the main conclusions to this article:

1. In verb inflections in Basque there can be allocutive morphemes which refer to

the speaker. In spite of the fact that these morphemes are structurally equal to the ergative
and dative cases, they are located outside the argumental structure of the verb. If in
speech the 2nd pronoun from the informal register is used, the allocutive must be used
thereafter.

2. This morpheme causes some changes in the verbs which can be classified in
three groups. In intransitive verbs with dative agreement there are no changes.
Nevertheless, in intransitive verbs without dative agreement the auxiliary verb changes
from izan (‘be’) to *edun (‘have’). In all the other cases, that is to say, with synthetic
verbs, and in sentences where the auxiliary verb is *edun (‘have’) or *ezan, the
morpheme -i- is inserted before the verb root.

3. In some of these cases the verb inflection contains the morpheme -i-. Although
this morpheme has been traditionally linked to the DF we believe that there are enough
reasons to treat them as two different morphemes. For one thing, they differ in position:
while the DF can appear both before the root and before the dative agreement, the AF can
only and without any exception be placed before the verb root. For another thing, there
are differences in form: the DF may change its form according to its nature or the dialects
(-i-, -ki-, -ts-, -ra-) and the AF is always a single -i- morpheme.

4. The allocutive in Basque follows some syntactic restrictions which vary from
dialect to dialect. Two main tendencies can be observed: in north-eastern dialects
traditional constraints have remained, while in central-western dialects their influence has

18
 Lizardi & Munduate 2016

been weakened. So in northern-eastern dialects the allocutive is used only in main

declarative sentences. But in central-western dialects it is quite usual to find the
allocutive in interrogative and conditional sentences, even in subordinate completive
sentences amongst the youngest speakers.

7. References

Alberdi, Xabier. 1994. Euskararen tratamenduak: erabilera. Iker 9. Bilbo:

Euskaltzaindia.
Alberdi, Xabier. 1995. “The development of the Basque System of terms of address and
the allocutive conjugation”. In Jose Ignacio Hualde, Joseba Andoni Lakarra &
Robert Larry Trask (eds.) Towards a history of the Basque language.
Amsterdam/Philadelphia: John Benjamins. 275-295.
Albizu, Pablo. 2002. “Basque verbal morphology: Redefining cases”. In Xabier
Artiagoitia, Patxi Goenaga & Joseba Andoni Lakarra (eds.) Erramu Boneta:
Festchrift for Rudolf P. G. de Rijk. Bilbo: UPV/EHU. 1-19.
Atxaga, Bernardo. 1988. Obabakoak. Donostia: Erein.
Bustintza, Evaristo. 1986. Abarrak. Itziar Laka (ed). Donostia: Euskal Editoreen Elkartea.
Digital edition: http://goo.gl/N9sDNW. [Last view: 26/08/2015].
Cano, Harkaitz. 2011. Twist. Donostia: Susa.
Eguren, Luís. 2000. “El morfema de alocutivo del euskera y el modelo de gramática”.
Hermeneus 2. 95-118.
Elordieta, Arantzazu. 2001. Verb Movement and Constituent Permutation in Basque.
Utrecht: LOT.
Etxepare, Bernart. 1995. Linguae Vasconum Primitiae. Bilbo: Euskaltzaindia.
Etxepare, Ricardo. 2003. “Valency and argument structure in the Baque Verb.” In José
Ignacio Hualde & Jon Ortiz de Urbina (eds.), A grammar of Basque. Berlin / New
York: Mouton de Gruyter. 363-425.
Fernández, Beatriz. 2012. “-(k)i eta -ra”. ASJU XLVI-2. Bilbo: UPV/EHU.
Fernández, Beatriz. 2013. “-(k)i hiztegian, eta hiztegiaz harantzago”. ASJU XLVII-2.
Bilbo: UPV/EHU.
Fernández, Beatriz. 2014. “-(k)i”. In Beatriz Fernández eta Pello Salaburu (eds.), Ibon
Sarasola, Gorazarre. Homenatge, Homenaje. Bilbo: UPV/EHU.
Gómez, Ricardo & Sainz, Koldo. 1995. “On the origin of the finite forms of the Basque
verb”, in Hualde, J.I., Lakarra, J.A., Trask, L., Towards a history of Basque
language. Amsterdam: John Benjamins Publishing Company.
Hualde, José Ignacio & Ortiz de Urbina, A grammar of Basque. Berlin/New York:
Mouton de Gruyter.
Lakarra, Joseba A. 2008. “Aitzineuskararen gramatikarantz (malkar eta osinetan zehar)”.
In Xabier Artiagoitia & Joseba Lakarra (eds.), Gramatika jaietan: Patxi
Goenagaren omenez. Bilbo. UPV/EHU. 451-490.
Oyharçabal, Bernard. 1993. “Verb Agreement with Non Arguments: On Allocutive
Agreement”. In Jose Ignacio Hualde & Jon Ortiz de Urbina (eds.) Generative
Studies in Basque Syntax, Amsterdam, John Benjamins. 89-114.
Rebuschi, Georgés. 1984. Structure de l’énoncé en basque. Paris: SELAF.

19
 Non-argument agreements: an approach to Basque allocutivity

Trask, Robert L. 1995. “On the history of the non-finite forms in Basque.” In Jose
Ignacio Hualde, Joseba Andoni Lakarra & Robert L. Trask (eds.) Towards a history
of the Basque language. Amsterdam / Philadelphia: John Benjamins. 207-234.
Zuazo, Koldo. 2008. Euskalkiak euskararen dialektoak. Donostia: Elkar.

20

View publication stats