Open main menu

Gesenius' Hebrew Grammar/159. Conditional Sentences

§159. Conditional Sentences.

Cf. H. Ferguson, ‘The Use of the Tenses in Conditional Sentences in Hebrew’ (Journal of the Society of Bibl. Lit. and Exeg., Middletown, Conn., June and September, 1882).—P. Friedrich, Die hebr. Conditionalsätze, Königsberg, 1884 (Inaug.-Diss.).—Driver, Use of the Tenses, 3rd ed., p. 174 ff.

159a 1. The great variety of construction in conditional sentences is owing to the fact that it frequently depends on the subjective judgement of the speaker, whether he wishes a condition to be regarded as capable of fulfilment (absolutely, or at least possibly), thus including those already fulfilled, or as incapable of fulfilment. On this distinction depends the choice both of the conditional particle to be used (see below), and especially (as also in Greek and Latin) of the tense. The use of the latter is naturally determined according to the general principles laid down in § 106 ff.[1] In the following sketch, for the sake of clearness, conditional sentences without conditional particles will be first discussed (under b), and afterwards sentences with these particles (under l).

159b 2. The relation between condition and consequence may be expressed, as in English, by the simple juxtaposition of two clauses. At the same time, it is to be observed in general as a fundamental rule (in accordance with the original character of the two tenses), that the imperfect, with its equivalents (the jussive, cohortative, imperative, perfect consecutive, and participle), is used to express a condition and consequence which are regarded as being capable of fulfilment in present or future time, while the perfect represents a condition already fulfilled in the past, and its consequence as an accomplished fact. The other use of the perfect—to represent conditions regarded as impossible—occurs only in connexion with particles.


159c (a) Imperfect (cf. §107x) in protasis and apodosis, Jos 2218, ψ 10428 ff. יִלְקֹם֑וּן[2] תִּתֵּן לָתֶם (if) thou givest unto them, they gather, &c.; ψ 13918, Pr 1217, Jb 2024, Ec 118, Neh 18; with an interrogative imperfect in the apodosis, Ju 1312; with the jussive, Jb 1016; with the cohortative, Pr. 1:23; with the perfect, Is 2610 (yet will he not learn righteousness; the apodosis forcibly denies what the imperfect in the protasis had represented as still conceivable; cf. Ho 812); with the perfect consecutive, Gn 4725, Ex 335; with the protasis suppressed, Jb 58 (see §107x).

159d (b) Jussive in protasis (cf. §109h, i) and apodosis, ψ 10410 תָּֽשֶׁת־ח֫שֶׁךְ וִיהִי לָ֑יְלָה (if) thou makest darkness, it is night; imperfect in the apodosis, ψ 10429 b; cohortative Pr 123. Also in Ex 79 יְהִי לְתַנִּין it shall become a serpent, is the apodosis to a suppressed protasis if thou cast it down; so in 2 K 510 וְיָשֹׁב is the apodosis to a protasis if thou wash, contained in what precedes.

159e (c) Cohortative (see §108e) in the protasis; perfect in the apodosis, ψ 406; imperfect consecutive, Jb 1918 אָק֫וּמָה וַיְרַבְּרוּ־בִי (if) I arise, they speak against me; on the cohortative in the apodosis, cf. §108f.

159f (d) Imperfect consecutive in the protasis (§111x), ψ 13911 וָאֹֽמַד if I say, &c. (with a noun-clause as the apodosis); with a frequentative perfect consecutive in the apodosis, 1 S 216.

159g (e) Perfect consecutive in the protasis and apodosis (see the examples, §112kk and ll), Gn 4422 וְעָוַב אָבִיו וָמֵת and should he leave his father, his father would die; 9:15, 44:29, Ex 414, 1213, 1 S 162, 193, 2 S 1328, 1 K 830; with frequentative perfects, Ex 1621 (referring to the past, Jer 209); with imperfect in the apodosis (being separated from the Wāw by לֹא), Nu 2320, Jb 524; introduced by an infinitive absolute, 1 K 237; an interrogative clause in the apodosis, Lv 1019; a noun-clause, ψ 3710, Jb 721.

159h (f) A simple perfect (to represent actions which are to be regarded as completed) in the protasis and apodosis, Pr 1822 מָצָא אִשָּׁה מָצָא טוֹב has one found a wife, he has found a good thing; an imperfect in the apodosis, Jb 194, 2310; an imperfect consecutive, Ex 2025, Pr 112, Jb 325, 2313 b, 29:11; an interrogative clause, Nu 1214, Jb 720 if I have sinned (prop., well, now I have sinned!) what can I do unto thee? 21:31, 35:6, Am 38; a noun-clause, Jb 2719.

159i (g) A participle as casus pendens (cf. §143d, and the sections of the Grammar there cited, esp. §116w) or a complete noun-clause in the protasis; the apodosis mostly introduced by wāw apodosis, e.g. Pr 2324 Keth. יוֹלֵד חָכָם וְיִשְׂמַח בּוֹ if one begetteth a wise child, he shall have joy of him; with perfect frequentative in the apodosis, 1 S 213, &c.; but also with a simple imperfect, e.g. Ex 2112 (cf. §112n); with an interrogative imperfect, 2 K 72, 19; with an interrogative perfect, Ju 613.

159k (h) Infinitive with preposition (also as the equivalent of a conditional clause) in the protasis, and a perfect consecutive in the apodosis (cf. §112mm), e.g. 2 S 714 ff. בְּהַֽעֲוֹתוֹ וֽהֹֽכַחְתִּיו וג׳ if he commit iniquity, I will correct him; Ex 3434 f. (with imperfect, followed by perfects frequentative in the apodosis).

Rem. On the expression of condition and consequence by means of two co-ordinate imperatives, see §110f.

159l 3. Particles used to introduce conditional sentences are אִם (for which in the later and latest Books sometimes הֵן, see below, under w) and לוּ[3] (1 S 1430, Is 6319 לוּא; Ec 66, Est 74 אִלּוּ, from אִם לוּ) if, negative אִם לֹא and לוּלֵא (לוּלֵי) unless; כִּי supposing that (Lat. ut), in case that, sometimes used almost in the same sense as אִם. With regard to the difference between אִם (אִם לֹא) and לוּ (לוּלֵא), the fundamental rule is that אִם is used if the condition be regarded either as already fulfilled, or if it, together with its consequence, be thought of as possibly (or probably) occurring in the present or future. In the former case, אִם is followed by the perfect, in the latter (corresponding to the Greek ἐὰν with the present subjunctive) by the imperfect or its equivalent (frequently in the apodosis also). On the other hand, לוּ (לוּלֵא) is used when the condition is to be represented as not fulfilled in the past, or as not capable of fulfilment in the present or future, and the consequence accordingly as not having occurred or never occurring. In the former case, לוּ and לוּלֵא are necessarily followed by the perfect (mostly also in the apodosis) corresponding to the Greek εἰ with the indicative of an historic tense, and the Latin imperfect or pluperfect subjunctive. In the latter case (which is extremely rare) the perfect, or the participle, or even the imperfect, may be used.

159m Rem. Since it again frequently depends on the subjective judgement of the speaker (see under a), whether a condition is to be regarded as possible or impossible, we cannot wonder that the distinction between אִם and לוּ is not always consistently observed. Although naturally לוּ and לוּלֵא cannot take the place of אִם and אִם לֹא (on the strange use of לוּ in Gn 5015 see below), yet conversely אִם is sometimes used where לוּ would certainly be expected; cf. e.g. ψ 5012, 1375, 1398, Ho 912 (cf. verse 11). These examples, indeed (אִם with the imperfect), may without difficulty be explained from the fact that the connexion of לוּ with the imperfect was evidently avoided, because the imperfect by its nature indicates a still unfinished action, and consequently (as opposed to לוּ) a still open possibility. But אִם is also used for לוּ in connexion with the perfect, especially when an imprecation is attached by the apodosis to the condition introduced by אִם, e.g. ψ 74 ff. אִם־עָשִׂ֫יתִי זֹאת... יִֽרַדֹּף וג׳ if I have done this..., let the enemy pursue my soul, &c., cf. Jb 319 ff. The speaker assumes for a moment as possible and even actual, that which he really rejects as inconceivable, in order to invoke the most severe punishment on himself, if it should prove to be the case.

On the frequent addition of an infinitive absolute to the verb in clauses with אם see §113o above.


159n A. אִם 1. with perfect in the protasis to express conditions, &c., which have been completely fulfilled in the past or which will be completely fulfilled in the future (the perfect is here equivalent to the futurum exactum, §106o). The apodosis[4] takes—

(a) A perfect also, e.g. Pr 912 אִם־חָכַ֫מְתָּ חָכַ֫מְתָּ לָּ֑ךְ if thou art wise, thou art wise for thyself; ψ 7315 (see below on לוּ).

(b) Imperfect, e.g. Dt 3241 אִם־שַׁנּוֹתִי if I whet my glittering sword... אָשִׁיב I will render vengeance, &c.; Jb 915 f.30 (in both cases we should expect לוּ rather than אִם־; so also in ψ 4421 f., with an interrogative imperfect in the apodosis); Jb 1113 (the apodosis is in verse 15).

(c) Jussive (or optative), e.g. Jb 319 ff. (see m above); Gn 183. 159o (d) Perfect consecutive (see the examples in §112gg), e.g. Gn 439 אִם־לֹא הֲבִֽיאֹתִיו וג׳ if I bring him not... then I shall have sinned, &c.; Ju 1617, 2 S 1533, 2 K 74. On the other hand, e.g. Gn 476, Mi 57, Jb 74 refer to actions already completed; in Gn 389 and Nu 219 the perfect with וְ is a perfect frequentative and refers to past time.

(e) Imperfect consecutive (see §111q), e.g. Jb 84 if thy children have sinned (חָֽטְאוּ)..., וַיְשַׁלְּחֵם he has delivered them, &c.

(f) Imperative, e.g. Gn 504 אִם־נָא מָצָ֫אתִי חֵן בְּעֵֽינֵיכֶם דַּבְּרוּ־נָא לג׳ if now I have found grace in your eyes, speak, I pray you, &c.; the imperative precedes in Gn 4716 and Jb 384, 18.

159p (g) A (complete or incomplete) noun-clause, e.g. Jer 1418 (a vivid realization of the future) if I have gone forth into the field (= if I go, &c.), then, behold, the slain with the sword! &c.; Pr 2414 (apodosis with wāw apodosis).

159q 2. אִם with imperfect in the protasis, to express what is possible in the present or future, as well as (according to §107b) what has continued or been repeated in the past. The apodosis takes—

(a) The perfect, e.g. Nu 3223 וְאִם־לֹא תַֽעֲשׂוּן כֵּן הִנֵּה חֲטָאתֶם but if ye will not do so, behold, ye have sinned; here the apodosis represents the time when the consequence has already taken place; so also Jb 2012–14. On the other hand, Nu 1629 (as also 1 S 69 and 1 K 2228) is a case of a pregnant construction, if these men die as all men die, then (it will follow from this) the Lord hath not sent me.

159r (b) The imperfect, e.g. 2 K 74 אִם־יֶחַיֻּ֫נוּ הִֽחְיֶה if they save us alive, we shall live, &c.; Gn 1316, 1828, 30, 28:20 ff., Ex 2025 (the second imperfect is equivalent to a jussive); Is 118, 1022, Am 92–4, ψ 5012 (where אִם ironically represents an impossibility as possible); Jb 85 f. (with the insertion of a second condition in the form of a noun-clause); 9:3, 20, 14:7; a frequentative imperfect referring to the past, Gn 318 אִם־כֹּה יֹאמַר if (ever) he said thus..., וְיָֽלְדוּ then they bare...; Ex 4037. In Gn 4237 the consequence (on תָּמִית cf. §107s) precedes the condition.

(c) The jussive (or optative), e.g. ψ 1375; cf. §109h.

(d) The cohortative, e.g. Gn 139, Jb 317; cf. §108f.

159s (e) The perfect consecutive (see the examples in §112ff and gg), e.g. 1 S 206 אִם־פָּקֹד יִפְקְרֵ֫נִי אָבִ֫יךָ וְאָֽמַרְתָּ if thy father miss me at all, then shalt thou say, &c.; Gn 2441, Ju 420; with a frequentative perfect consecutive, Gn 318 if he said (as often happened)..., then, &c.

(f) The imperfect consecutive; so perhaps ψ 5916, if וַיָּלִ֫ינוּ is to be explained according to §111t.

(g) The imperative, e.g. Gn 3150, 1 S 2021 (with wāw apodosis, but in verse 22 simply לֵךְ), 21:10, Jb 335.

159t (h) A noun-clause, e.g. Gn 47, ψ 1398, Jb 86, 3126 f.

3. אִם with cohortative, e.g. Gn 3031; cf. the passages in §108e.

159u 4. אִם with infinitive, Jb 927 אִם־אָמְרִי prop. if my saying is (but probably we should read אָמַ֫רְתִּי).

159v 5. אִם with a noun-clause, e.g. Dt 522 (in the apodosis a perfect with wāw apodosis), Gn 2746, Ju 915 (imperative in the apodosis); 11:9 (imperfect in the apodosis); 2 S 128 (cohortative in the apodosis); Ho 1212; especially if the subject of the conditional clause be a personal pronoun. In an affirmative sentence this pronoun is often joined to יֵשׁ, in a negative sentence to אֵין (cf. on both, §100o), while the predicate (cf. §116q) is represented by a participle, usually expressing the future, e.g. Ju 636 f. אִם־יֶשְׁךָ מוֹשִׁיעַ if thou will save, &c.; Gn 2449 אִם־יֶשְׁבֶם עֹשִׂים if ye will deal, &c.; 1 S 2323. In Gn 2442 f. the condition is expressed in a more humble form by the addition of נָא. With אֵין Gn 435 וְאִם־אֵֽינְךָ מְשַׁלֵּחַ but if thou wilt not send, &c.; 20:7 (with imperative in the apodosis); Ex 817, 92 f., 1 S 1911 (all with a participle also in the apodosis). But יֵשׁ and אַ֫יִן may also be used after אִם without a suffix; thus יֵשׁ Gn 238, 1 S 208, 2 K 915, &c., אִם־אַ֫יִן (if it be not the case) Gn 301, Ex 3232, Ju 915, 2 K 210; cf. also אִם־כֵּן if it be so, Gn 2522.

159w B. הֵן if, generally supposed to be originally identical with הֵן behold![5] Probably, however, הֵן if, is a pure Aramaism, and since the Aramaic word never has the meaning behold, it is at least improbable that it had originally any connexion with הֵן or הִנֵּה. Cf. Ex 822, Lv 2520, Is 5415, Jer 31, Hag 212, 2 Ch 713, and frequently in Job, as 911, 12, 12:14, 15, 19:7, 23:8, 40:23, always with wāw apodosis following, except in 13:15, where consequently the meaning see is no doubt preferable.

159x C. לוּ if, לוּלֵא (לוּלֵי) if not.

1. With perfect in the protasis and apodosis (cf. §106p), e.g. Ju 819; אִלּוּ is used in the same sense as לוּ in Est 74, cf. Ec 66 (with a question in the apodosis).—With the perfect in protasis and apodosis after לוּלֵא Gn 3142, 4310, Ju 1418, 1 S 2534, 2 S 227, Is 19. On the other hand, in Dt 3229 לוּ with a perfect is followed by an imperfect in the apodosis, if they were wise, they would understand this; in Mi 211 by a perfect consecutive.

159y 2. With imperfect after לוּלֵא Dt 3227, אָגוּר probably as the modus rei repetitae, were it not that I ever and again feared, &c.; so also the imperfect after לוּ with the apodosis suppressed, Gn 5015 supposing that Joseph should hate us; since, according to the context, the danger was real, the use of לוּ here is strange; conversely in other cases, e.g. ψ 7315, Jb 915 f.30, לוּ would be more natural than אִם.

159z 3. A noun-clause occurs after לוּ 2 S 1812, 2 K 314, ψ 8114, all with imperfect in the apodosis; Jb 164 לוּ יֵשׁ, with cohortative in the apodosis.

D. כִּי supposing that, if:—

159aa 1. כִּי with perfect in the protasis, e.g. Nu 520 וְאַתְּ כִּי שָׂטִית but thou, if thou hast gone astray, &c.; with a frequentative perfect consecutive in the apodosis, Jb 713 f.; with an imperfect consecutive, Jb 2229.

159bb 2. כִּי with imperfect in the protasis, e.g. ψ 234 גַּם כִּֽי־אֵלֵךְ yea, though I walk (have to walk)..., I will fear no (לֹֽא־אִירָא) evil; 37:24; Ex 212 כִּֽי־תִקְנֶה עֶ֫בֶד עִבְרִי וג׳ if thou buy an Hebrew servant, six years shall he serve (but in verses 3–5 a series of definite conditions with definite consequences is introduced by אִם; so also the כִּי in verse 7 is followed in verses 8–11 by the special cases with אִם; cf. also verse 17 ff.); cf. Gn 424, 2441, Jb 385; with a perfect consecutive in the apodosis, Gn 3218 f., Ex 1816; with a noun-clause, Is 115.

3. כִּי with a noun-clause (and imperfect in the apodosis), 2 S 198.


159cc 1. In 2 K 513 the particle אבי (Masora אָבִי, probably in the sense of my father) appears exceptionally for לוּ; its meaning here is unquestionable, but its origin is obscure. Cf. the exhaustive discussion of Delitzsch and Wetzstein on Jb 3436, where this אבי appears to be used as a desiderative particle.—Sometimes when one case has been already discussed, another of the same character is added by means of אוֹ or, e.g. Ex 2136 אוֹ נוֹדַע וג׳ or (another possible case) it is known that, &c., i.e. but if it be known, &c., LXX ἐὰν δέ, Vulg. sin autem; cf. Lv 423, 28, 5:1, 25:49, 2 S 1813; with a following imperfect, Ez 1417f.—On the hypothetical use of אֲשֶׁר (which is interchangeable with כִּי in other senses also) Lv 422 (in verses 3 and 27 אִם), Dt 1127 (verse 28 אִם), Jos 421, see the Lexicon.

159dd 2. The conditional sentence is frequently found in an abridged form, where the suppressed clauses can be easily supplied from the context; cf. Gn 139, 2449, 1 S 216 וְאִם־לֹא and if not, i.e. and if thou wilt not give it to me, then I take it (perfect according to §106n) by force; cf. 1 S 69. The use of וְיֵשׁ alone in Ju 613 is peculiar, as also וָיֵשׁ in 2 K 1015 (where read with the LXX וַיֹּא֫מֶר יֵהוּא וָיֵשׁ) in the sense of if it be so.—In 2 S 1326, 2 K 517 וָלֹא alone appears to be used in the sense of if really ... not, in each case with a following jussive equivalent to may there at least, &c. (cf. §143d); but perhaps with Matthes, ZAW. 1903, p. 122 ff., following Kuipers, we should read וְלוּ would that!—In 1 S 1313, Jb 313 the condition must be supplied from the preceding clause to complete the sentence introduced by כִּי עַתָּה, in Jb 3128 by כִּי, in 2 K 1319 by אָז.—The apodosis also appears sometimes in an abridged form (e.g. Gn 424, Is 432) or is entirely suppressed, e.g. Gn 3027, 3817, 5015 (see y above), Ex 3232, ψ 2713, Jb 385, where properly הַגֵּד must be supplied with כִּי תֵדָ֑ע as in verses 4 and 18; cf. §167a.—In ψ 84, instead of the apodosis I exclaim which we should expect, the exclamation itself follows.

159ee 3. The absolute certainty with which a result is to be expected is frequently emphasized by the insertion of כִּי Is 79; כִּי אָזָ 2 S 227, 197, Jb 1115; or כִּי עַתָּה now verily, Nu 2229, 1 S 1430 after לוּ, Gn 3142, 4310 after לוּלֵי, Jb 86 after אִם. On this corroborative כִּי cf. such passages as Gn 1820, &c., and §148d. On כִּי אִם after an oath cf. 163 d.

159ff 4. Sometimes the force of a hypothetical particle extends beyond the apodosis to a second conditional clause, as in the case of אִם Pr 912, Jb 1015, 166, 2223, and כִּי Is 432.

159gg 5. In Ex 3320 a negative statement takes the place of a condition with a negative consequence, for a man doth not see me and live, instead of for if a man sees me, he does not live; cf. the similar passages, Dt 221, 4 thou shalt not see ... and hide thyself, instead of if thou seest ... thou shalt not hide thyself.

  1. It may, moreover, happen that a different idea is introduced in the apodosis, from that with which the protasis started—a source of many further variations.
  2. On the termination -וּן cf. §47m. In verse 28 b also יִשְׂבְּעוּן is probably to be explained from its immediately preceding the greater pause. These terminations in verses 28–30 and ψ 13918 can scarcely have any connexion with the conditional sentence, although it is strange that -וּן in Nu 3223 appears after אִם־לֹא in the protasis. In Nu 1629, 3220 -וּן as before א (as in Jb 3110 in the apodosis) is to be explained from the dislike of hiatus.
  3. On לוּ cf. Kohler in Geiger’s Zeitschr. für Wiss. und Leben, vi (1868), p. 21 ff.
  4. We are not here concerned with the fact that the logical apodosis (the consequence of the condition) is sometimes mentioned before the condition; as in Gn 1828, 30, Ju 1110, ψ 636 f., 137:6, and according to Dillmann Is 44.
  5. There could be no doubt of their identity if וְהִנֵּה in 1 S 97, 2 S 1811, simply meant if. We must, however, keep to the meaning but behold.