The Psalms Scroll, one of the Dead Sea Scrolls, written in ancient Hebrew

The Psalms Scroll, one of the Dead Sea Scrolls. Israel Antiquities Authority, 1993 / Wikimedia Commons (public domain).

There are four core tenses for the Hebrew verb, that is two plus two others derived from them by adding the prefixed particle wǝ, most commonly translated: and. First we have the imperfect future tense as exemplified by : yi-qtol. This has a range of meanings from ‘he is killing’ to ‘he will kill, he may kill, let him kill’, in fact any translation in which the action is not completed, including not started and so not completed. By contrast, when wǝ is added, giving wayyiqtol, the meaning switches to perfective or completed action: and he killed. Then we have the perfect tense represented by qatal = he killed (completed action), but when preceded by wǝ as wǝqatal, it takes on the meaning of the simple imperfect/future tense: and he will kill etc. Wǝqatal is also known as the consecutive perfect and wayyiqtol as the consecutive preterite. How did such a strange situation come about?

How the Imperfect (Yiqtol) Developed

Let us start with the imperfect yiqtol. It started as a preterite tense *ya-qtul = he killed, cf Akkadian i-prus = he separated, with the same root vowel. In Central Semitic, which includes Arabic, a -u got added, so *ya-qtul-u. This was probably a nominative case marker used when the verb was in a relative clause, linking it to the head noun eg *in(ā)sum dʒū yaqtulu = a man that (dʒū) killed. When the relative pronoun dʒū dropped out, you were left with the sense : a man of the killing sort, where *yaqtulu was similar to a qualifying adjective. This adjective could then be extrapolated as a verb form: is a man of the killing sort, is a killer. (Remember there is no verb ‘to be’ used as a copula in Semitic). This could then generalize to an imperfective: is killing, will kill, may kill etc.

As we see in the first chart below, the next stage was the loss of final -u, making it largely identical with the original preterite / perfective, including in the plural, where the distinction 3 pl -una (impf) and -u (pret) broke down, and the form in -u predominated in both cases. 3 f pl *ya-qtul-nā was replaced by the 2 f pl *ta-qtul-nā by analogy with the singular, where the 2 f sg ta- derives from the Proto-Afroasiatic 2nd sg marker ‘t’, while the identical 3 f sg ta- derives from the feminine marker ‘t’. In Arabic the equivalent of *yaqtulnā is retained in the 3 f pl. Meanwhile the wayyiqtol sequential past was formed from the original preterite, while the simple preterite faded from use as its forms merged with the new imperfective. However imperfective / future yiqtol could also still occasionally be combined with wǝ as wǝyiqtol, so Aaron Hornkohl and others argue that much later in the 1st millenium AD the Masoretes restored the most ancient pronunciation of wǝ, namely wa-, and doubled the following consonant to preserve the a vowel and keep the two perfective and imperfective consecutive tenses distinct. The conjugation tables displayed by David Steinberg and published by Adath Shalom show the developments:

Proto-Hebrew and the Rise of the Sequential Tenses

Proto-Hebrew (1200 BC) (850 – 550 BC)
Indicative imperfect Imperfective / future
1 cs aqtulu iqtul
2 ms taqtulu tiqtul
3 fs taqtulin(a) tiqtuli(n)
3 ms yaqtulu yiqtul
3 fs taqtulu > tiqtul
1 cp naqtulu niqtul
2 mp taqtulūn(a) tiqtulū(n)
2 fp taqtulnā tiqtulnā
3 mp yaqtulūn(a) yiqtulū(n)
3 fp (yaqtulnā) (yiqtulnā)
Preterite (1200 BC) Sequential Past (850 – 550 BC)
1 cs aqtul wayyiqtul
2 ms taqtul wattiqtul
3 fs taqtulī wattiqtuli
3 ms yaqtul wayyiqtul
3 fs taqtul wattiqtul
1 cp naqtul wanniqtul
2 mp taqtulū wattiqtulū
2 fp taqtulnā wattiqtulnā
3 mfp yaqtulū wayyiqtulū

Cf the Akkadian preterite:

1 cs a-prus 1cp ni-prus
2ms ta-prus 2cp ta-prus-ā
2fs ta-prus-ī 3mp i-prus-ū
3cs i-prus 3fp i-prus-ā

(expected 3fs i-prus-ī and 2ms ta-prus-ū not used)

The Hebrew Perfect (Qatal) and Its Semitic Relatives

We now turn to the Hebrew past tense (qatal):

1cs qatal-tti 1cp qatal-nu
2ms qatal-tta 2mp qatal-ttem
2fs qatal-tt 2fp qatal-tten
3ms qatal 3cp qatl-u
3fs qatl-ah

Cf Classical Arabic (verb ktb = write):

1cs katab-tu 1cp katab-nā
2ms katab-ta 2mp katab-tum
2fs katab-ti 2fp katab-tunna
3ms katab-a 3mp katab-ū
3 fs katab-at 3fp katab-na

Cf the Akkadian stative:

1cs parsā-ku 1cp parsā-nu
2ms parsā-ta 2mp parsā-tunnu
2fs parsā-ti 2fp parsā-tina
3ms paris 3mp pars-ū
3fs pars-at 3fp pars-ā

The endings of the Ancient Egyptian stative conjugation (the stative becomes a resultative in Egyptian, where the suffixed conjugation becomes a perfective in Western Semitic):

1cs .kw 1cp .wyn
2ms .tj 2cp .twny
2fs .ty
3ms .w 3cp .w
3fs .ty

From Stative to Perfective: The Qatal Tense

At the same time as nominal *yaqtulu was evolving into a verb form, if not earlier, the stative form *qatal-ā = is a killer, evolved into the perfective qatal = has killed, killed, had killed, thus moving in the opposite direction. The two forms moved in tandem: as *qatalā became more perfective, *yaqtulu became more imperfective, according to Ola Wikander (who produced some of the earlier material I mentioned as well). The tables above show the evolution of the endings (Arabic is more conservative than Hebrew and Akkadian even more so) and their link with the Ancient Egyptian stative. The -ā in the 3 m sg in Arabic and in the 1st and 2nd person in the Akkadian stative might go back to an absolutive case marker. Semitic or its predecessor Proto-Afroasiatic may have had ergative alignment before it changed to nominative / accusative alignment. The ergative case marked the agent or doer of the action on a direct object, such as ‘the boy’ in the sentence ‘the boy hit the ball.’ The absolutive marked the object: ‘the ball’ in the example, and from such a use would evolve into the accusative marker. But it was also used to mark the subject where an intransitive verb was used and there was no direct object. This is typically the case where a stative tense is used, eg ‘he is lying down / is a killer.’ Akkadian pars- or its predecessor would have been treated as a nominal form or participle and marked with the case ending -ā and used with a subject pronoun or noun: it (is) separate.

Understanding Wǝqatal

Turning to the wǝqatal, Bo Isaksson argues that it should not be seen as a separate conjugation with future meaning. It always follows another clause and may be related to it as a consequence (and so…, thereby…) or simply relate an accompanying action. Its tense is determined by that of the preceding clause, whether past (did), durative (is/was doing), past anterior (had done), modal (may/might do), future (will do) etc. Its original past / perfective function was expanded to include other temporal functions as it took over as it took over from imperfective / future *yaqtulu (later yiqtol) in initial position. It is therefore incorrect to see wǝqatal as a sequential future tense complementing sequential past wayyiqtol, which, deriving from a preterite yaqtul, has retained strict past reference.

It takes a while to understand what is going on with these tenses, so don’t worry if you don’t get it all at once. If you would like to learn some Hebrew or another Near Eastern language such as Aramaic, Ancient Egyptian, Sumerian or Akkadian, please contact me on this site.

ND

Nigel Dickey

MA, University of Auckland. Nigel spent seven years attending degree-level classes across the Indo-European language family at Oxford University, and now teaches ancient languages to students worldwide from Auckland, New Zealand.

Read Nigel's full story →