The Basque tense and aspect system is complicated by the co-existence of two conjugation types which allow for different oppositions. One is the conjugation of analytically (periphrastically) construed verbs, the default case, the other involves a small number of synthetically construed verbs, which can optionally be construed periphrastically as well.

Here is an example for both conjugation types. The analytical construction consists of a main verb in a non-finite form (here the verbal noun in the inessive case) and a tense-aspect auxiliary (here the present tense of the transitive auxiliary ukan):1

(1)    Egi-te-n    dut.                                 'I do it.'
       do-NOM-IN   PRS.3S<1S

Synthetically construed verbs are directly conjugated, i.e. person and tense markers are attached to the verbal root:

(2)    Ba-2  d-a-ki-t.3                               'I know it.'
       ENC-  3S-PRS-know-1S.ERG
             ABS
             PRS

The following verbs can be construed synthetically:

izan (intransitive tense-aspect auxiliary, 'to be')
ukan (transitive tense-aspect auxiliary, 'to have')
egon 'to stay, to be'
eduki 'to hold, to have'
jakin 'to know'
ekarri 'to bring'
eraman 'to take (to)'
joan 'to go (directed)'
ibili 'to go (undirected), to wander'
etorri 'to come'
etzan 'to lie'
eritzi 'to think'
irudi 'to seem' (third person only)
(jaugin 'to come', in the imperative only)
(very rare: ikusi 'to see', erabili 'to conduct, use')

In comparison with early records (16th century) the number of synthetically construed verbs has significantly decreased.4

1. Tense-aspect oppositions with analytical verbs

The analytical construction is the only productive pattern, i.e. the analytical verb represents the normal case.

1.1. Using present-tense auxiliaries
1.1.1. Present
By combining the auxiliary in the present tense with the nominalized main verb (verbal noun) in the inessive, we get the normal present tense, as can be seen in example (1). Example (3) shows the case of an intransitive verb:

(3)    Mintza-tze-n   naiz.                             'I speak.'
       speak-NOM-IN   PRS.1S

(4)    Egin     dut.                             'I have done it.'
       do:PCP   PRS.3S<1S

(5)    Mintza-tu   naiz.                          'I have spoken.'
       speak-PCP   PRS.1S

It may be labelled "near past" and is incompatible with adverbials, referring to long periods of time or remote past tense:

(6)    *Aspaldi-an    egin     dut.     'I have done it long ago.'
       long_ago-IN    do:PCP   PRS.3S<1S

(7)    *Aspaldi-an    mintza-tu   naiz.  'I have spoken long ago.'
       long_ago-IN    speak-PCP   PRS.1S

1.1.3. Resultative
Such adverbials are possible, if the participle is used like an adjective with the individualizer -a (usually called 'article'),6 which makes it accessible for number agreement with the head NP:7

(8)    Etxe-a(-k)     sal-du-a(-k)      d(it)u   aspaldi-an.      
       house-ART-P    sell-PCP-ART-P    PRS(PL)  long_ago-IN

'S/he sold the house(s) long ago.'

The resultative is used, when the speaker wants to underline that an action or situation in the past is still relevant at the time of speech.

1.1.4. Future
Future is expressed by adding the delimitative case ending -ko8 to the participle. After consonants, some dialects prefer the (possessive) genitive -en instead (I gloss them as future markers, when attached to participles):

(9)    Egin-go/-en    dut.                        'I shall do it.'
       do:PCP-FUT     PRS.3S<1S

(10)   Mintza-tu-ko   naiz.                       'I shall speak.'
       speak-PCP-FUT  PRS.1S

1.2. Using past-tense auxiliaries
In its turn the past-tense auxiliary can be combined with the same range of non-finite forms as the present tense. This results in the following tense-aspect meanings:

1.2.1. Imperfective past
When combined with the verbal noun in the inessive (the combination denoting simple present above), the past auxiliary is used to express imperfective past tense:

(11)   Egi-te-n    nuen.                'I did it / was doing it.'
       do-NOM-IN   PRT.1S>3S

(12)   Mintza-tze-n   nintzen.           'I spoke / was speaking.'
       speak-NOM-IN   PRT.1S

1.2.2. Perfective past
When combined with the participle (the combination denoting perfect above), it is used to express perfective past tense:

(13)   Egin     nuen.                                  'I did it.'
       do:PCP   PRT.1S>3S

(14)   Mintza-tu   nintzen.                             'I spoke.'
       speak-PCP   PRT.1S

Thus, the formal opposition between verbal noun and participle distinguishes imperfective and perfective aspect in the past, whereas in the present tense it is used for distinguishing simple present and perfect. This is summarized in the following table:

Table 1:
                      present auxiliary      past auxiliary

inessive verbal noun  simple present         imperfective past
                      (aspectually neutral)

participle            present perfect        perfective past

The perfective/imperfective distinction is the same as in the Romance languages.

1.2.3. Pluperfect
The resultative combined with the present auxiliary turns out to function as pluperfect, when used with the past auxiliary.9 The 'article' -a agrees in number with the head noun.

(14)   Egin-a      nuen.                          'I had done it.'
       do:PCP-ART  PRT.1S>3S

(15)   Mintza-tu-a    nintzen.                     'I had spoken.'
       speak-PCP-ART  PRT.1S

Of course, the pluperfect can be interpreted as a resultative in the past, as the event in question is always seen to be relevant to another (later) event in the past.

1.2.4. Future of the past
The auxiliary can be combined with the non-finite future form (i.e. participle + delimitative or genitive) to form a secondary future tense which is used in subordinate clauses, when the main clause is in the past (consecutio temporum):10

(16)   Pentsa-tu   zuen  mintza-tu-ko   nintze-la.                
       think-PCP   PRT   speak-PCP-FUT  PRT-SR
                   3S>3S                1S

'He thought that I would speak.'

1.3. Summary
The following table summarizes the tense-aspect oppositions with analytical verbs, i.e. verbs having no synthetical forms:

Table 2:
                      present auxiliary      past auxiliary

inessive verbal noun  simple present         imperfective past
                      (aspectually neutral)

participle            present perfect        perfective past


participle + -ko(-en) future                 future of the past


participle + -a(k),   resultative            pluperfect
(+ -rik / -ta)

Note that the opposition between simple present and present perfect (a temporal opposition) is parallelled by the opposition of imperfective and perfective aspect in the past.

2. Tense-aspect oppositions with synthetical verbs

We can conjugate synthetical verbs (enumerated above) directly, by attaching person and tense markers to the verbal root. They may also be analytically construed. That means that we have an add- itional formal and functional opposition between such verbs con- strued synthetically and the same verbs construed analytically. As we shall see, some tenses and aspects cannot be expressed synthet- ically.

2.1. Synthetical construction
2.1.1. Present
The synthetical verb has a present-tense form, which (just like the present tense of analytical verbs) is neutral with respect to the aspect opposition:

(17)   Ba-nago.                                          'I stay.'
       ENC-stay:PRS.1S

The unmarked case of a synthetical verb in the present tense would be to construe it synthetically.

2.1.2. Past
Analogously, a past tense can be formed, which (contrary to the analytical form) is aspectually neutral, i.e. it can always be used, when we would otherwise have to decide between perfective and imperfective forms in the analytical construction.

(18)   Ba-nengoen.                                     'I stayed.'
       ENC-stay:PRT.1S

Especially those dialects that have been under long and heavy Romance influence (e.g. the Eastern dialects of Lower Navarra) prefer the analytical construction in the past, whereas in the present tense, synthetically construed verbs may be found to some extent (even in potential and conditional forms). The reason for this may be that in Romance the distinction of perfective and imperfective aspect is obligatory in the past tense. In general, synthetical forms tend to disappear first for the past tense and last in the imperative. Note also that the verbs still showing synthetical conjugation are movement verbs or verbs with stative semantics, i.e. verbs that do not express actions or - more generally speaking - a change of situation. They are verbs where the distinction of perfective and imperfective aspect only makes sense, if special meanings are involved (e.g. 'to know' -> 'to come to know').

2.2. Analytical construction
Formally, the analytical construction of synthetical verbs is the same as for analytical verbs. As analytical forms and synthetical forms stand in opposition to each other, their semantic value (or 'function') is different.

2.2.1. Using present-tense auxiliaries
2.2.1.1. Present progressive
If a synthetical verb would normally be used synthetically, the use of the analytical construction in the present tense (present auxiliary + verbal noun) will express a specific aspectual meaning, in most cases progressive:

(19)   Ego(i)-te-n    naiz.        (instead of: ba-nago 'I stay.')
       stay-NOM-IN    PRS.1S

'I am staying.'

In an appropriate context it may also denote habituality.11

2.2.1.2. Present perfect
The present auxiliary combined with the participle is again used as a perfect:

(20)   Egon        naiz.                            'I have come.'
       stay:PCP    PRS.1S

The perfect can only be formed analytically.

2.2.1.3. Resultative
The same holds for the resultative:

(21)   Joan-a      da.                             'S/he is gone.'
       go:PCP-ART  PRS.3S

Compare this sentence with the following in the present perfect:

(22)   Joan     da.                          'S/he has (is) gone.'
       go:PCP   PRS.3S

This sentence can have a resultative meaning, but normally refers to an event in the near past (meaning e.g. 'S/he has gone (there) this morning.'), whereas joana da means something like: 'S/he is in the state of being gone.'

2.2.1.4. Future
In the modern language, the future can only be formed periphrastically by attaching the delimitative (or genitive) to the participle, as has been illustrated above:

(23)   Joan-go  /  (joan-en)      naiz.              'I shall go.'
       go:PCP-FUT  go:PCP-FUT     PRS.1S

2.2.2. Using past-tense auxiliaries
When synthetically used, the synthetical verb is aspectually neutral in the past tense, as it is in the present. It does not interfere with the aspectual opposition between perfective and imperfective past, expressed by the inessive verbal noun and the participle respectively, both combined with a past tense auxiliary.

2.2.2.1. Imperfective past
Instead of the aspectually neutral form (ba-)nindoan ('I went.'), the following form is usually preferred, when the event is viewed as on-going at some point of reference or as iterative, i.e. imperfective:

(24)   Joa(i)-te-n    nintzen                'I went (was going).'
       go-NOM-IN      PRT.1S

2.2.2.2. Perfective past
If the event is viewed as completed and punctual, i.e. perfective, the following analytical construction (the same as with analytical verbs) is preferred over the simple past:

(25)   Joan     nintzen.                        'I went / I left.'
       go:PCP   PRT.1S

2.2.2.3. Pluperfect
The pluperfect is always analytical:

(26)   Joan-a      nintzen.                     'I went / I left.'
       go:PCP-ART  PRT.1S

2.2.2.4. Future of the past
So is the future of the past:

(27)   Pentsa-tu   zuen  joan-go     nintze-la.                   
       think-PCP   PRT   go:PCP-FUT  PRT-SR
                   3S>3S             1S

'He thought that I would come.'

2.3. Summary
The following table summarizes the tense-aspect oppositions of synthetical verbs. The differences with analytical verbs are highlighted:

Table 3 (for synthetical verbs only):

                      present                past

synthetical constr.   simple present         simple past
                      (aspectually neutral)
analytical constr.:

inessive verbal noun  present progressive    imperfective past
                      (habitual)

participle            present perfect        perfective past


participle + -ko(-en) future                 future of the past


participle + -a(k),   resultative            pluperfect
(+ -rik / -ta)

3. A word on mood

Although this article is primarily about tense and aspect in Basque, I shall mention mood as well, without going into detail, just to give a view of the whole.

3.1. Subjunctive
Except for the third person, the subjunctive is formally identical with the imperative. Basically, it is used in subordinate clauses that depend on a main clause with a verb of wishing or order. The event for which it is used is presented as non-real, but desired. Only the tense-aspect auxiliaries (izan and ukan) have special subjunctive forms. They are suppletive: for izan, the root -di- is used (grammarians speak of a subjunctive verb *edin, which, of course, has no participle), for ukan the root -za- (grammarians speak of a subjunctive verb *ezan). Here is an example:

(28)   Nahi  dut   (bera/hura)    joan  dadi-n.                   
       want  PRS   s/he/that one  go    PRS.SBJ-SR
             3S<1S                      3S

'I want him/her to go.' or lit.: 'that s/he goes.'

In this example dadi-n (subjunctive + subordinator) is used instead of de-n (da + -n, indicative + subordinator).

As for transitive verbs, here is an example for the use of deza-n (subjunctive + subordinator) instead of du-en (indicative + subordinator):

(29)   Nahi  dut   (bera/hura)    egin  deza-n.                   
       want  PRS   s/he/that one  do    PRS.SBJ-SR
             3S<1S                      3S<3S

'I want him/her to do it.' or lit.: 'that s/he does it.'

The subjunctive is not used with the participle, but with the non- finite verbal stem, which is homophonous with the participle, if it ends on -n. If the verb has an -i- or -tu-participle (only the latter being productive), the difference becomes evident:

(30)   (i)   Ikus-te-n   du.                     'S/he sees (it).'
             see-NOM-IN  PRS.3S<3S

       (ii)  Ikus-i   du.                    'S/he has seen (it).'
             see-PCP  PRS.3S<3S

       (iii) Ikus  deza-n.       '(I wish) that s/he sees (it).'12
             see   PRS.SBJ-SR
                   3S<3S

This means that there is no aspectual opposition possible with the subjunctive.

If the main verb is in the past tense, the subjunctive in the subordinate clause is used in the past as well:

(31)   Nahi  nuen     ikus  zeza-n.                               
       want  PRT      see   PRT.SBJ-SR
             1S>3S          3S>3S

'I wished s/he would see / saw it.',

'I wanted him/her to see (it).'or 'I wanted that s/he saw (it).'

3.2. Potential and conditional
If an event is viewed as only possibly occurring, the potential is used, i.e. the potential marker -ke- is inserted into the verb (synthetical or analytical):

(32)   (i)   Ba-duzu.                               'You have it.'
             ENC-PRS.3S<2S

       (ii)  Ba-dukezu.                'You may possibly have it.'
             ENC-POT.3S<2S

Inserted into past tense forms, -ke- marks the conditional, which is mainly used in the apodosis of unreal conditional clauses and thus corresponds to the conditional of Romance languages, or of English; note that additionally in modern Basque, the non-finite part of the analytical construction is marked for future:13

(33)   Zu-re plaza-n  (baldin) ba-nintz,14   (if-clause, protasis)
       2-GEN place-IN (if)     ENC-SUP.1S

       egin-go     nuke.                   (then-clause, apodosis)
       do:PCP-FUT  COND.1S>3S

'If I were in your place, I should do it.'

It is commonly thought (and philologically substantiable) that -ke- was originally a future marker. Due to the modal component of the future, it could develop into a mood marker, when a new analytical future came up.

The semantics of the delimitative makes us think that the analytical future was originally restricted to near future (event 'on the edge of' (delimitative!) occurrence).

4. Conclusion

We have seen that synthetical verbs show a set of tense-aspect oppositions that differ from the tense-aspect oppostions of analytical verbs. The former oppose an aspectually neutral, simple present to a progressive (habitual) one and have an aspectually neutral past tense that does not exist for analytical verbs.

All synthetical verbs are old ones. The number of such verbs has been decreasing over the last centuries up to the point that in some dialects they have almost disappeared altogether. It is plausible to assume that before the first written sources, Basque had had more synthetical verbs. It is even probable that the analytical construction is an innovation that came about in a (pre-) romance language contact situation, most probably originating from spoken Latin. The reason for such an assumption is the structural similarity with the habeo-factum perfect which is typical of vulgar (or spoken) Latin. The Latin construction, which is used instead of the simple past for resultative events, consists of the verb 'to have' (habeo 'I have', at that time pronounced something like /ajo/) and the participle (e.g. factum 'done', habeo factum 'I have done'). Due to language contact this construction has spread into adjacent Germanic and Slavic languages, being otherwise uncommon among the languages of the world which, so far, have not inherited it from European languages. The productive participle ending -tu corroborates the assumption that the Basque perfect construction comes from Latin.

The evolution of the analytical present is comparable to what happens in quite a number of languages:15 In order to denote progressive aspect, the verbal noun is used in a locative case (here the inessive) together with a verb meaning 'to be', i.e. 'to be in/at an action'. The specialty of Basque is that both 'to be' and 'to have' are used with the inessive verbal noun, according to the actant structure of the verb.16 This 'anomaly' with respect to other languages with a structurally similar progressive can be explained by the fact that the distinction of the two auxiliaries is necessary to clearly mark the actant structure.

The progressive meaning gets lost with the decline of a regular opposition between analytical and synthetical forms.

The analytical construction revolutionized the tense-aspect system as a whole. New tense-aspect oppositions evolved (near past vs. remote past, perfective vs. imperfective etc.); a new analytical future was introduced (perhaps a near future in the beginning), which came to stand in opposition to a potential future developing into a merely modal category (potential, conditional). Moreover, the auxiliary suppletion (izan vs. *edin (intransitive, 'to be') and ukan vs. *ezan (transitive, 'to have'), perhaps different verbs once) allowed for a regular distinction of indicative and subjunctive mood.

The following table gives a synoptic view of the tense-aspect system and its evolution. The small (and probably still decreasing) number of synthetical verbs are treated as relicts from Classical Basque and therefore neglected under "Modern Basque":17

Table 4: (S: synthetical construction, A: analytical construction)

Stage 0               Stage 1                Stage 2

*Proto-Basque         Classical Basque       Modern Basque
                                             (neglecting S-verbs)

present (S)           present (S)            present
                      progressive (A)        
                      perfect (A)            perfect
                      resultative (A)        (resultative)

past (S)              past (S)               
                      past perfective (A)    past perfective
                      past imperfective (A)  past imperfective
                      pluperfect (A)         pluperfect

future (S,-ke-)       potential fut. (S,A)   (potential)
                      near future (A)        future
                      
fut. of past (S,-ke-) conditional (S,A)      conditional
                      fut. of past (A)       future of past
                      
                      subjunctive (A)        subjunctive
                      past subjunct. (A)     past subjunctive
                      subjunct.-potent. (A)  
                      subjunct.-condit. (A)  

The present situation of Basque is somewhat in between stage 1 and 2, as it still shows some synthetically construed verbs, incidently less in the past tense. As already said, the reason for this may be the obligatory past perfective/imperfective distinction in the Romance contact languages. On the whole, the stage 2 system comes very near to the Romance system.

We can conclude that the change of the Basque tense-aspect system is contact-induced. The most significant step that made all the other steps follow, was the introduction of the analytical construction, probably on the model of spoken Latin.

If we compare this evolution with what had happened in other spheres of Basque grammar, we find that the tense-aspect system is particularly innovative, Basque morphosyntax being otherwise very conservative.18 A reason for this may be that tense-aspect oppositions have very concrete semantic meaning. This 'semantic load' makes bilingual speakers feel the necessity to transpose structure from one language into the other, all the more as they found 'meaningful' material (e.g. a verb 'to have') which could be translated without much difficulty and thus be used as a basis of structural transpositions ('calques').

Abbreviations

ABS    absolutive
ART    article
COND   conditional
ENC    enunciative
ERG    ergative
FUT    future
GEN    genitive
IN     inessive
NOM    nominalizer
P      plural
PCP    participle
POT    potential
PRS    present
PRT    preterite
S      singular
SBJ    subjunctive
SR     subordinator
SUP    suppositive

Bibliography

BREU, Walter (1988): 'Resultativit„t, Perfekt und die Gliederung der Aspektdimension' in: RAECKE, Jochen (ed.): Slavistische Linguistik 1987. Referate des XIII. Konstanzer Slavistischen Arbeitstreffens, Tbingen 22.-25.9.1987 (= Slavistische Beitr„ge 230) - Munich: Sagner: 42-74

BYBEE, Joan L. / DAHL, ™sten (1989): 'The creation of tense and aspect systems in the languages of the world', Studies in Language 13,1: 51-103

HAASE, Martin (1991a): 'Resultatives in Basque' in: Arbeiten des K”lner Universalienprojekts (AKUP) 83 - K”ln: Institut fr Sprachwissenschaft; to appear again in: Funci¢n, Guadalajara (Mexico)

id. (1991b): 'Tempus und Aspekt im Sprachkontakt (Baskisch- Gaskognisch-Franz”sisch' in: FELDBUSCH, Elisabeth / POGARELL, Reiner / WEIá, Cornelia (eds.): Neue Fragen der Linguistik. Akten des 25. Linguistischen Kolloquiums, Paderborn 1990. Band 1: Bestand und Entwicklung. - Tbingen: Niemeyer: 465-474

id. (1992): Sprachkontakt und Sprachwandel im Baskenland: Die Einflsse des Gaskognischen und Franz”sischen auf das Baskische [Ph.D. thesis] - Hamburg: Buske

HOURCADE, Andre (1986): Grammaire bearnaise. - o.O. (Pau?): Los Caminaires

ITURRIOZ LEZA, Jose-Luis (1982): 'Apprehension im Baskischen' in: SEILER, Hansjakob / STACHOWIAK, Franz Josef (eds.): Apprehension. Das sprachliche Erfassen von Gegenst„nden. Teil II: Die Techniken und ihr Zusammenhang in Einzelsprachen (= Language Universal Series 1/II) - Tbingen: Narr: 1-43

id. (1985): 'La funcion de -a y de -ta a la luz de la dimension de individuacion', Euskera 30: 175-213

LAFITTE, Pierre (1979): Grammaire Basque (Navarro-Labourdin Litteraire). 3rd edition - Donostia (San Sebasti n): Elkar

LAFON, Rene (1943): Le systŠme du verbe basque au XVIe siŠcle. 2 vols., reprinted in one volume 1980 - Donostia (San Sebasti n) / Baiona (Bayonne): Elkar

OYHAR€ABAL, Bernard (1987): Etude descriptive de constructions complexes en basque: Propositions relatives, temporelles, conditionnelles et concessives. (ThŠse de doctorat d'Etat Šs Lettres et Sciences Humaines) - Paris: Centre National de la Recherche Scientifique

Footnotes

1. ukan, the participle form of the verb (usually the citation form), is used in the North Eastern dialects only. Basque grammarians therefore reconstructed a more regular participle form *edun, which is used in some grammars to refer to that auxiliary.
2. The enunciative ba- is obligatory, if a conjugated verb stands at the beginning of the sentence. It fills the empty focus position which comes canonically before the conjugated verb.
3. d-u-t can be analyzed in the same way, -u- being the transitive root. Especially in the auxiliaries, we have to deal with a great deal of fusion, all the more as the forms become more complex. I shall therefore refrain from a detailed interlinear analysis. The actant structure will be visualized as in (1), where the ergative is indicated by the open side of the pointed bracket (<).
4. LAFON (1943) lists about 60 in texts of the 16th and 17th century.
5. i.e. what may be called a "recent perfect" (cf. BREU 1988).
6. That the 'article' is actually an individualizer is shown by ITURRIOZ (1982, 1985); unfortunately, Basque grammarians and school teachers keep on speaking of "mugagabea" ('indetermined') and "mugatua" ('determined'), in spite of better knowledge.
7. There are concurrent ways of formation: In North-Eastern dialects, the partitive -rik is attach to the participle, in the South-Western dialects a special formative -(e)ta is added, which may be etymologically related to eta 'and'. HAASE (1991a) studies the resultative in greater detail. HAASE (1991b) gives a short account of the development in contact with Gascon and French, where the resultative came to be a passive.
8. Some grammarians (cf. e.g. LAFITTE 1979) call the delimitative "locative genitive".
9. It may be reinforced by adding the participle izan(a) (or ukan(a) in transitive clauses, where this form exists). This seems to be a 'calque' from the neighboring Romance languages. The participles izan and ukan may also be added to the simple past tense, in order to signalize non-resultative anterior.
10. Sometimes normal future may be found in sentences like the following, if the event in the subordinate clause is still going to happen.
11. There are less grammaticized ways of expressing progressive and habitual aspectuality, using the operator verbs ari ('to be occupied with') and ohi ('to be used to').
12. The main clause may be left out elliptically.
13. In the dialects of the North-East and in older texts, the subjunctive is combined with -ke-, which is used with the verbal stem, if the construction is analytical. This is not the place to treat conditional clauses at length; for that matter, see OYHAR€ABAL (1987).
14. The suppositive (SUP) is a variant of the past without the ending -(e)n, used in if-clauses.
15. cf. BYBEE/DAHL (1989)
16. 'to be' with intransitive verbs, i.e. verbs requiring an absolutive only, or absolutive and dative; 'to have' with transitive verbs, i.e. verbs requiring the ergative as well
17. They are indeed more abundant in the written standard than in most spoken varieties.
18. HAASE (1992) gives an account of contact-induced change throughout the grammar of Basque, and proposes explanatory concepts, such as 'semantic load' and 'translatability'.

This page is part of Buber's Basque Page and is maintained by Blas Uberuaga. Please report any problems or suggestions to Blas. Eskerrik asko!