The death of a virtuous woman? Proverbs 31.10-31,gnomic qatal,and the role of translation in the analysis of the Hebrew verbal system

1. The אשת חיל‎ of Prov 31

In his recent commentary, Fox provides a good overview of different interpretations of the woman in Prov 31.10-31. These range from traditional Jewish ones that see the woman as a symbol of the Torah or the Shekinah, through Christian readings of the woman as the soul, the Virgin Mary, or the Church, to a widespread modern belief that she is a personification of wisdom (in addition to the ones found in earlier parts of Prov). ⁴ However, Fox’s own conclusion is that the description lacks those traits that are present in other allegories, which would ‘disturb the surface reading and reveal its inadequacy’ and thus make the reader ‘shift to a different level of interpretation’. ⁵ He states that ‘the Woman of Strength is not a figure for wisdom. While Lady Wisdom [in Prov 8 and 9] personifies wisdom, the Woman of Strength typifies it’, and further, ‘she instantiates it’. ⁶ Fox seems to reflect a widespread view when he claims that the woman is ‘not a particular woman, but an ideal’. ⁷ As Crook puts it, ‘The Woman of Worth is an ideal; there never was any such person’. ⁸ Thus, according to this view, the woman is not a symbol of some metaphysical phenomenon, but an ideal model for real, actual women. Indeed, the initial question in v. 10 points in the direction of a general application of the things said about the woman.

The widespread use of the present tense in translations seems to be the result of this view, or, conversely, maybe the view of the woman as an ideal rather than a particular woman stems from the present tense translation of the verbal forms. The point seems to be that an ideal woman, in general, does the things described in the poem and that the woman reading the text should do likewise—or, maybe more likely in the original context, the man reading the text should strive to have a wife of this kind. However, while it seems reasonable to see this as a main concern of the text, it does not automatically follow that such content must also be expressed explicitly at the surface level. Contrary to what Fox asserts, if we look at the Hebrew text with the knowledge that we have about the functioning of Hebrew verbs in other texts, it becomes clear that the general ideal of a woman finds expression through a description of one particular woman. At least in the fictional world of the text, the אשת חיל‎ is a particular woman, who at some point in the past has done or did all the magnificent things mentioned. Obviously, this does not invalidate the interpretation of the content of the poem as a general call for a certain type of behavior. It merely means that we have to infer this call from a narrative description of a particular instance of this behavior, just as general truths or behavioral appeals are presented in other types of narrative guise like fairy tales or parables. How, then, can the poem be interpreted if the widespread present tense reading of the verbs should be discarded?

Joüon already answered that question in 1922. ⁹ He argued that the woman must be dead and pointed to several verses that seem to support this. Thus, in v. 25, it is possible to interpret the words יום אחרון‎ as referring to ‘the last day’ of her life rather than ‘the following day’, as many translations would have it. Similarly, Joüon notes the words כל ימי חייה‎ in v. 12, which might also indicate that her life has ended, and the final praise offered by her sons and husband would be unexpected if the woman was still alive. ¹⁰ Thus, the poem is a sort of eulogy. His main argument for this interpretation is the use of the verbal forms: ‘En traduisant au present, on fausse la valeur de tous les wayyiqtol et de presque tous les qatal. En traduisant au passé, au contraire, il reste un très petit nombre de difficultés’. ¹¹ These difficulties are the five yiqtols occurring in the poem. Joüon does not treat the sixth one (in v. 10) since it constitutes no problem at all because it occurs in the introduction outside of the narrative indicating a general question (אשת־חיל מ ימצא‎). ¹² Similarly, the seventh one (in v. 30) is outside of the description of the woman. Joüon interprets these five yiqtols as instances of what he terms ‘yiqtol d’action’ (vv. 11 and 21) and ‘fréquentatifs d’action’ (vv. 14, 18, and 27), that is, different shades of non-anterior meaning in a past context (durativity—which must be the meaning of ‘yiqtol d’action’ in this connection—and habituality). Such usage of yiqtol is comparatively frequent and thus constitutes a ‘problem’ of far less magnitude than the interpretation of multiple qatals and wayyiqtols as general or habitual present tense. In discussing the yiqtol in v. 21 (לא־תירא‎), Joüon mentions the possibility of understanding the verb as an ordinary future tense. ¹³ Presumably, this means that the woman (though dead) need not worry for the members of her household, since she already took care of their future needs while she was alive. Whether this indicates a belief in some sort of afterlife or merely a figurative use of words is another matter. Similarly, it could be argued that the yiqtol in v. 11 (יחסר‎) is a simple future. Joüon finds this unlikely, but in fact such an interpretation seems more appropriate here than in v. 21, since the subject is the husband who is still alive and, though his wife is dead, he will not lack anything because she has made sure that the house is in proper order. As will be argued below, the Hebrew verbal system does not distinguish grammatically between the possible readings (future/general present vs habitual or durative action in the past). However, when we translate into a language that makes such a distinction, we have to choose. Either way, the main point is that it is possible to read the few yiqtols as part of a past narrative.

In the same way, the single active participle in the passage (v. 27 צופיה‎) does not pose a problem, since participles occur (much like yiqtol) as expressions of simultaneity or durative action in any temporal setting. Wolters has argued that there is a special reason for the use of the lone active participle. The preservation of the final radical in this III-y verb produces a form that might be a conscious wordplay on the Greek word for ‘wisdom’, σοφία. According to Wolters, this is a ‘cleverly veiled barb in a religious polemic’ against Greek intellectualism. ¹⁴ While other scholars have questioned this view, ¹⁵ Wolters does seem to have a case when he argues that the expected form would be a qatal, like the majority of forms in the poem. ¹⁶ Thus, the choice of a participle might serve a special purpose. While questions about authors’ intentions and conscious use of language are difficult to answer definitively, the point to note is that if there is a pun, we have a reason for the use of the participle, since there would be no direct similarity to the Greek word in any other verbal form.

The acrostic principle followed in the poem is another factor to be considered in connection with the choice of verbal forms. Cholin (though using Joüon’s past tense rendering as an argument for the view that the author refers to ‘un exemple réel’) puts forward a possible objection to Joüon’s approach. ¹⁷ He states that the acrostic character of the poem might have been the reason for the use of qatal forms at least in the verses that start with verbs, and from these the use of qatal could have spread to other verbs.

Indeed, there seems to be a connection between qatals and acrostics. Giffone points out that acrostic poems tend to use qatals rather than yiqtols more frequently than non-acrostic poems. He also notes that the acrostics tend to use the qal conjugation more than nif., hif., or hof., in comparison with non-acrostics. ¹⁸ To some degree, he argues, this is a reflection of ‘the alphabetical restrictions on the first word’, though these do not account for all the extra qatals. ¹⁹ Furthermore, he notes that Ps 145 is also an acrostic but does not include any qatal forms. ²⁰ This psalm also shows that there is no necessary connection between acrostics and the placement of verbs at the beginning of verses. Wolters’ characterization of Prov 31.10-31 as ‘poetry of action’ might help explain why many of the verses begin with verbs: ‘the Song of the Valiant Woman is a portrait in verbs’. ²¹ Thus, the verb-first order in many verses serves to bring the actions of the woman into focus. As noted by Giffone, wayyiqtols and yiqtols, as well as qatals in the nif., hif., or hof., are difficult to use extensively as initial forms in an acrostic poem since they would fit with a very restricted number of verses. However, since all the initial qatals in the poem are in the qal conjugation, it is clear that the corresponding participles would also fit the acrostic principle. The fact that only one is used seems to indicate, first, that the single occurrence of an active participle might have a special purpose (Wolters) and, second, that the intention behind the use of qatals really is to express the ordinary meaning of the qatal form, that is, anteriority. If the intention was to indicate present tense, why not use participles in all these verses? In short, the acrostic restrictions do not constitute an argument that the qatals do not have their ordinary anterior meaning. Quite the opposite, the focused position of many of the verbs would seem to underline the usual semantic content of these forms. ²²

Cholin notes another point in connection with the verbal forms in the poem. He states that there is nothing wrong with rendering the passage in the present tense, since the intention is clearly generic. ²³ However, a translation that aims to represent the style of the Hebrew poem should take into account the difference between the derived meaning of the text and the expression of this meaning as a past narrative. Furthermore, a serious problem appears when scholars use such translations in their interpretations of the verbal system, as I will illustrate in more detail below.

Most translators and commentators seem to have ignored or forgotten Joüon’s approach. Practically no one has followed Joüon’s lead, with Rogland as a positive exception. Like Joüon, he translates the passage as simple past tense while interpreting the yiqtols as ‘habitual-iterative’ or ‘modal pasts’. ²⁴ His treatment of Prov 31.10-31 is part of a general chapter on the gnomic perfect, which Rogland shows to be an unnecessary category since such usage can be understood as ‘the report of a particular, and often extraordinary, experience or observation’, or as generalizations based on the way ‘things have typically occurred in the past’. ²⁵ The latter explanation is, of course, similar to the way many earlier scholars have treated this alleged use of qatal, namely, as a type of experiential perfect, where the general present sense stems from the ordinary anterior meaning of the verbal form—something has always or often been the case, therefore this is how it is. For example, Ewald spoke of ‘allgemeine Wahrheiten, die aus der Erfahrung deutlich sind und sich schon bestimmt so bewährt haben[…] oft in Vergleichungen und Sprichwörtern’. ²⁶

In addition to the gnomic perfect, Rogland deals with two other types of alleged non-past use of qatal, the so-called prophetic and performative perfect. He argues convincingly that it is possible to interpret all such cases with the ordinary past tense meaning found in other qatal forms, if we understand this as a relative past. ²⁷ Thus, qatal does not express an absolute past tense, but rather anything past relative to the context, that is, anteriority. Other cases of supposedly non-past qatal not treated in detail by Rogland, like the so-called precative perfect, are also interpretable as anterior tense.

Most scholars would probably agree that the ordinary meaning of qatal is some kind of past tense. If the rather restricted number of cases that seem to express a meaning with more or less the opposite temporal reference can in fact be understood in a way that fits the general description of the form accounting for the vast majority of occurrences, this should be given serious consideration. Thus, instead of claiming that future tense (in the prophetic perfect), general or habitual present tense (in the gnomic perfect), and actual present (in the performative perfect) are part of the meaning of qatal as such, it should be recognized that this is not a question of the intrinsic semantics of the verbal form but rather a question of stylistics. In the case of the prophetic perfect, there is no need to assume that qatal sometimes explicitly indicates an absolute future tense. In some cases, the reason for assuming such a meaning is simply the general interpretation of the passage, which situates it in the future. Accordingly, the verbal forms are translated as future tense and consequently it must be claimed that qatals in such a passage express this sense. However, a different exegetical approach to the passage might indicate a past setting and therefore another interpretation of the general semantics of the verbal forms, for example, in Isa 5.13 (לכן גלה עמי‎). In other cases, the use of a past tense form is explainable as a report of a vision, which has already occurred, though its content might be future. ²⁸ Similarly, Rogland describes the performative perfect as a sort of stylistic convention that should not influence our analysis of the semantics of qatal. ²⁹ However, considering the available forms, the use of an anterior form to express the notion of performativity is not odd, since Hebrew has no separate form for a present perfect. Thus, qatal can denote actions that have just occurred through the speech act itself, while the yiqtol would tend to give the statement a future (or modal) connotation and the participle would mark the event as ongoing.

A most important point to note is the restricted repertoire of forms available in Hebrew. While English has several forms to indicate different types of anteriority (did, has done, had done, will have done, etc.), Hebrew only has one, namely, qatal (with its positional variant wayyiqtol). ³⁰ These forms express anything that is anterior to a contextually defined reference point, simple past, present perfect, past perfect, or future perfect. ³¹ Similarly, Hebrew only has a general category of non-anteriority (expressed by yiqtol and its positional variant weqatal) for which English has a bigger repertoire of forms, indicating present, future, future in the past, simultaneity or habituality in different time frames, and so on. ³²

Based on such a general view of the verbal system, we might argue that the verbal forms in Prov 31 are to be understood as present perfects referring to a person still living. However, the arguments brought forward by Joüon seem to favor the interpretation of the woman as deceased. ³³

Although Fox does not even suggest the possibility of translating the poem using past tense forms, he does note that the Septuagint renders most of the text in this way, giving ‘the poem the sound of a eulogy, though this is not carried through consistently’. ³⁴ However, as described above, such a reading of the Hebrew text is actually quite straightforward, since the yiqtols and the participle are unproblematic in a past context, while the Greek text, which uses future and present tense in several verses, is more problematic. Thus, the use of present tense in v. 21 with the husband as subject and the addition of the words ‘those who are with her’ make it an unlikely eulogy (οὐ φροντίζει τῶν ἐν οἴκῳ ὁ ἀνὴρ αὐτῆς ὅταν που χρονίζῃ πάντες γὰρ οἱ παρ᾽ αὐτῆς ἐνδιδύσκονται).

Note how Fox treats the aorist in v. 25 (Greek 26): ‘“(b) and she rejoiced [aor] in the last days.” The last days are the end of her life. M’s “she laughs at the latter day” must have seemed overconfident’. ³⁵ However, the Hebrew text does not say ‘laughs’. The form is a wayyiqtol, ‘and she laughed’. Oddly, in his commentary, Fox actually expresses doubt about the existence of gnomic wayyiqtols (without mentioning qatals) and claims that he has ‘tried to preserve something of its narrative and past-tense implications, so as to allow recognition of little anecdotes’. ³⁶ Apparently, he did not follow this approach in chapter 31.

2. Recent discussion of the gnomic perfect

In his recent treatment of the use of verbal forms in Proverbs, Cook praises Rogland’s approach to the question and argues that the widespread use of present tense in rendering practically every verb form in proverbial texts ‘begs the question why different verb forms have been employed’. ³⁷ He does not treat Prov 31.10-31, but bases his investigation on 10.1-22.16 and 25.1-29.27, collections of proverbs that, according to Cook, consist mostly of generic sentences. ³⁸

While Cook’s approach to the use of qatal is clearly on the right track, it should be noted (in anticipation of the general discussion of translation-based interference) that his readings of verses with yiqtol forms introduce distinctions into the Hebrew system that are based solely on interpretative translations. Thus, he distinguishes between three categories of yiqtol in the passages under consideration and translates them with added expressions such as ‘typically’, ‘inevitably’, with ‘should’, or as what he terms ‘contingent modality’. ³⁹ He admits that the distinction between the first two categories is ‘an issue of interpretation’, ⁴⁰ but this is the case with all of them. Hebrew uses the same verbal form, which simply expresses non-anteriority. The rest he derives from his interpretation of the context. However, concerning the main question, the use of qatal and wayyiqtol, he correctly states that most often ‘the usual translation’ (the way we would usually translate such forms in other texts) is acceptable. ⁴¹ Often the passages in question are ‘past tense anecdotes whose moral message is left for the reader to discern’. ⁴² This description is very fitting. The only objection is Cook’s continued use of the terms ‘genericity’ and ‘generic statements’ for such passages, despite his conclusion that the genericity is not a property of the verbal forms themselves. Such terms might obscure the need to abandon the notion of the gnomic qatal. As Cook notes, the past tense, anecdotal explanation does not fit all cases. Some qatals are present statives or present perfects. ⁴³ However, in the analysis presented here, these different meanings are merely interpretations that we need to make when we translate into a language with more categories than Hebrew. In Hebrew itself, qatal (and wayyiqtol) simply expresses anteriority.

As is well known, the qatal of so-called stative verbs often has present reference in Hebrew (and other West Semitic languages). West Semitic *qatala (whence Hebrew qatal) derives from an originally atemporal, nominal construction like the Akkadian stative, which, when formed from a verbal root with resultative meaning, developed into a proper tense form with anterior meaning, gradually replacing the older anterior form *yáqtul (Akkadian iprus). Seen in this light, stative qatals are relics of the older stage where the form merely signified that the subject is (or was or will be, etc., according to context) the notion expressed by the root (e.g. ‘I am old’). However, these verbs were probably reanalyzed as part of the new system when the form became an anterior tense form. Thus, ‘I am old’ could be reinterpreted as derived from ‘I have become old’, though this is not the origin of the form, and, similarly, ‘I know’ = ‘I have come to know’, and so on. Note that stative verbs in qatal are also interpretable as past, that is, the same form also means ‘I was old’ or ‘I became old’ or ‘I knew’. In this way, the ambiguity is the same as the one in non-stative verbs, which are always interpretable as either present perfect (with present reference) and simple past (with past reference). The choice between the two (or between other types of anteriority) is only relevant when we translate into a language that makes more distinctions than Hebrew in this area. In Hebrew itself, all these forms belong to the general category of anteriority.

The exact interpretation depends on the context—and on the repertoire and usage of verb forms in the language into which we are translating. For example, Cook’s example of a present perfect (Prov 20.12, ‘Yhwh has made both of them’) might also be rendered as a simple past in English, and many languages, like Hebrew, do not possess a separate present perfect form. This underlines the fact that the present perfect versus simple past distinction is not inherent in the Hebrew system itself. ⁴⁴

Another recent treatment of the gnomic qatal is by Andrason, who discusses the phenomenon in one of his many articles based on a ‘dynamic’ approach to the Hebrew verbal system. ⁴⁵ The basis for his approach is grammaticalization and the so-called path theory that claims that linguistic forms follow universal paths of development. A ‘gram’, a grammatical morpheme, starts out as a lexical word used in a periphrastic construction and gradually loses its independence and becomes a part of other words as a grammatical affix. The meaning of a gram evolves as it develops new senses in new contexts and these senses can be mapped onto paths, which are considered universal and irreversible. For example, an original resultative construction is presumed to develop along the so-called anterior path toward becoming a simple past tense, acquiring meanings from the categories of taxis, aspect, and tense (in that order) along the way. ⁴⁶

In his discussion of gnomic qatal, Andrason mentions Rogland and Cook, ⁴⁷ but he does not follow their conclusions. Although he correctly states that there is an important difference between the general meaning of a proverb and the semantics of the verbal forms used, and though he derives the supposedly gnomic usage from different perfect values of the form, he still maintains that there is such a thing as a gnomic qatal. ⁴⁸ He claims that his ‘data demonstrate that the gnomic sense is commonly conveyed by formations located on the anterior path’. ⁴⁹ However, none of his examples from other languages conclusively demonstrate the so-called ‘gnomic branch’ on the anterior path of development. Some of the cases cited include a present tense form of a copula verb plus an adjectival element, in which case the gnomic interpretation is completely straightforward (since the form is actually in the present tense), or, in the other types, they are interpretable as present perfect or past tense. The gnomic interpretation is only Andrason’s interpretation as indicated in his translations. Similarly, when it comes to the Hebrew cases, the gnomic interpretation is clearly just an interpretation and not part of the semantics of the verbal form. He cites examples from Proverbs such as ‘whoever finds me finds [<has found] life’ (Prov 8.35). ⁵⁰ Nothing compels us to choose the present tense translation instead of the one in the brackets and none of the cases should be thought of as inherently gnomic. He uses Prov 31.11-29 as ‘an exemplary fragment’ showing the supposedly gnomic usage, since, according to Andrason, the qatals in this passage describe ‘constant, generic, habitual, typical properties of a good wife’. ⁵¹ However, as argued above, while this might be the derived meaning of the entire passage, there is nothing to indicate that the verbal forms as such carry this meaning.

At the end of the article, Andrason asks the question, ‘how should the gnomic qatal be translated?’, arguing that if the language into which it is translated allows this, an anterior-type form could be used (like the English present perfect), but most often in a language like English, the more usual gnomic forms (simple present or resultative proper) should be used. ⁵² However, it is clear that the present tense translation of supposedly gnomic qatal is actually the starting point of his entire investigation (and not a question to be answered after the investigation), since the idea that there is such a thing as a gnomic qatal is derived precisely from such renderings of specific passages. By simply presupposing that these translations are a correct expression of the sense of the verbal form used, he has not come any closer to a demonstration of the existence of a gnomic qatal in Hebrew.

This is a general weakness in Andrason’s approach. He presupposes that the verbal forms that he treats have a very large polysemy, ⁵³ that is, he assumes that the translations and traditional readings of various passages are correct. Concerning the so-called precative perfect, he simply takes for granted that such a usage exists even though all the alleged cases are interpretable as ordinary anteriors with either past or present perfect sense. ⁵⁴ In general, when he speaks about the ‘individual senses (shades of meaning) empirically “recorded” in specific realistic cases, i.e. in concrete contexts’, ⁵⁵ he fails to see that there is nothing empirical about this, since we first need to interpret the texts. The point is to find the senses expressed by the verbal form, not just to assume that the translations reflect a correct analysis of the Hebrew system.

3. Hebrew verbs and translation-based interference

Thus, it seems that the present tense used in most translations of Prov 31.10-31 really is the result of the idea that a passage like this must be generic/habitual, and therefore it must be put into the present tense, no matter what the Hebrew forms are. When scholars use such translations as the basis for analysis of the Hebrew verbal system (and in general of the Hebrew language), problems arise. In fact, sometimes scholars seem to be analyzing their English translations (or whatever other language they are using) instead of the Hebrew text itself. This is a widespread problem, not just in connection with supposedly gnomic qatal. ⁵⁶ Here I will review a few examples of this kind of interference, but many more could be adduced.

The first case is the work of Ken Penner on the verbal system in Qumran Hebrew. ⁵⁷ He uses a so-called ‘empirical method of statistical correlation between form and meaning’, and his aim is to decide which of the categories tense, aspect, or mood is most prominent in Qumran Hebrew. ⁵⁸ To this end, he presents a database (available as an online supplement to the book), with grammatical tagging of all the finite verbs from all the least fragmentary scrolls, including information about temporal, aspectual, and modal reference. Based on such statistics, Penner concludes that the parameter most correlated to verb form is not aspect but absolute tense or modality. ⁵⁹ In the following discussion, he argues that the category of absolute tense is in fact the most prominent one. ⁶⁰ In the present context, the central problem is that this method is far from ‘empirical’ since the statistics on which Penner builds his conclusions are derived solely from his own interpretation of the texts. ⁶¹ Often, multiple interpretations are possible, and therefore the value of statistical studies is very dubious. This kind of statistics is not objective at all but simply reflects the interpretation and translation of the researcher. In fact, absolute tense as a separate category opposed to relative tense is a distinction that is simply not present in the Hebrew verbal system, but imported from the English translational equivalents, contrary to Penner’s stated aim of avoiding the imposition of foreign categories. ⁶² Hebrew uses the same forms (qatal and wayyiqtol) to express all kinds of anteriority, which in English (or similar languages) can or must be expressed using specific forms such as simple past, present perfect, past perfect, and so on. Statistics might show that, in the chosen corpus, a qatal or wayyiqtol is most often translated into English as a simple past expressing absolute tense, but that does not mean that we can make a majority ruling and conclude that the basic semantic category of the forms is absolute tense. The exact numbers (though interesting for other reasons) are not decisive in this connection. The only thing that really matters is the fact that we regularly find the forms in question in sentences where we need to translate them as past perfect, present perfect, or future perfect, that is, the same Hebrew verbal forms can indicate both absolute and relative tense. The occurrence of qatal forms in clearly future contexts should immediately alert us that there must be something wrong with a method that leads to the conclusion that absolute tense is the main parameter in the system. On the contrary, if we find relative as well as absolute uses of the same verbal form, and the system does not possess separate specifically relative tense forms, the system cannot be termed an absolute temporal system. If it is temporal, it must be relative. ⁶³

In his most recent article, Andrason uses a statistical approach similar to the one used by Penner. ⁶⁴ The aim is to compare the senses most typically expressed by qatal and wayyiqtol and, like Penner, he seems to think that counting the verbal forms in their different uses is an objective and empirical way of doing this. However, the input for the large number of figures and tables in the article is his interpretation/translation of Hebrew sentences, and, in some cases, his actual interaction with the Hebrew text is quite odd. ⁶⁵ A rather absurd example of translation-based interference is his treatment of Gen 4.8, which is supposed to show a modal cohortative use of wayyiqtol. ⁶⁶ The verb in question is ויהי‎, which quite clearly is a regular anterior tense introducing an infinitival construction (‘and it happened when they were in the field’). However, probably based on emendations in some modern translations and commentaries, Andrason translates this as ‘Let us go out to the field’, and therefore concludes that wayyiqtol has cohortative as one of its senses! In a less grotesque way, many of the other textual examples of allegedly different senses expressed by the Hebrew forms are also highly dependent on the chosen translations. A case in point is Gen 32.6, where Andrason claims that the wayyiqtol (again ויהי‎) has a present sense (‘I have oxen’). However, this verb clearly does not mean ‘have’, which shows that this is a problem involving quite different ways of expressing things in Hebrew and English. A literal translation would be ‘And oxen have become for me’, which means ‘I have acquired oxen’. In other words, there is no need to see a specific present tense meaning here. Andrason distinguishes between several senses in the wayyiqtol form, to which he assigns very precise percentages, 89.7% perfective past, 4% various types of perfect, 6% durative past, 0.15% present, 0.05% performative, and 0.05% cohortative (the case from Gen 4.8 just cited). ⁶⁷ As indicated, such distinctions are primarily the result of the fact that the English renderings of the sentences employ different verbal forms. Thus, he chooses to translate the wayyiqtol in Gen 3.17 (ותאכל‎) as a present perfect, but other translators would use a simple past tense here (e.g. NIV). In Hebrew, of course, the form is the same.

Another type of translation-based interference in the analysis of the Hebrew verbs—and a very widespread one—is the idea that the verbal forms have special meanings in poetry as opposed to prose, or rather, that any verbal form can mean anything in poetry. ⁶⁸ However, unless it can be argued that a present or future tense interpretation of a qatal or a simple past tense interpretation of a yiqtol is the only possible one, we should not let our analysis of the semantic potential of the Hebrew verbal forms be influenced by the fact that such renderings might sound more idiomatic and stylistically well polished in a translation. The alleged equivalence of all the verbal forms in poetry seems to be the result of influence from translations into languages that do not function in the same way as Hebrew and do not share the same stylistic conventions. If the purpose is to describe the meaning of the Hebrew verbs themselves rather than the best way to translate them into what is considered good poetic style in another language, the starting point should be the one that Alviero Niccacci has spelled out (after abandoning his earlier position). According to this view, the verbal forms in poetry function in the same way as in prose and we ought to proceed from this basic position and try to read the forms as we would in prose, more specifically in direct speech. ⁶⁹ Joosten, like Niccacci, assumes that ‘[b]iblical poets did not select their verbal forms at random’, and ‘verbal meanings in poetry are basically the same as in prose’. ⁷⁰ He correctly points out that ‘linguistic differences between poetry and prose belong for the most part to the domain of style, not grammar—parole, not langue’. ⁷¹ Similarly, Rogland talks about ‘poetics, not grammar’. ⁷²

Thus, when we observe a shift, for example, from qatal to yiqtol, we should try to make sense of this as a real shift of temporal reference, but we should do this with the awareness that the Hebrew tenses are relative, not absolute. If the verbal forms were markers of absolute tense, there would be many instances where the widespread type of parallelism with one of each form would seem odd (hence the desire to translate everything as present tense). However, in a system of relative tense, such shifts are readily understandable. First, yiqtol as a non-anterior tense can occur in a past setting as an expression of iterativity, habituality, future in the past, as well as an ordinary present or future. In the Hebrew system, these meanings are the same thing, belonging to one category. Second, qatal and wayyiqtol are not just simple past tense but often occur when the temporal reference is the ‘now’ of the poet (cf. Niccacci’s view that poetry is a type of direct speech). They are translatable into English as present perfects. This means that we do not get the unwanted clash between absolute past and future in the same sentence. Thus, to cite an example from Proverbs: ‘They have eaten the bread of wickedness, and they drink the wine of violence’ (Prov 4.17), or, similarly, Ps 2.2 (‘they take their stand’, ‘they have taken counsel together’).

Ps 18.5 (אפפוני חבלי־מות ונחלי בליעל יבעתוני‎) shows how different interpretations of the two categories of anteriority and non-anteriority are possible when translating into a language with more categories. The qatal can be rendered as simple past and the yiqtol as a contemporaneous/durative event in the past (‘entangled me’, ‘were overwhelming me’). Alternatively, if the temporal frame is present tense, the anterior form is rendered as present perfect and the non-anterior as future or present (‘have entangled me’, ‘overwhelm me’). The point is that the Hebrew system does not distinguish grammatically between these two alternatives.

There are many other cases of translation-based interference beyond those concerning the finite verbal forms. A case worth mentioning is the alleged use of construct infinitives with ל־‎ ‘as finite verbs’, which some scholars find in presumably late biblical Hebrew and in Qumran Hebrew. However, the fact that idiomatic English would often prefer finite verbs does not free us from the task of examining how such infinitives actually function in the Hebrew sentences. ⁷³ Similarly, the fact that the frequent ויהי‎ plus temporal phrase in biblical narrative is often rendered in English narrative style without ‘and it happened’ does not in itself prove that this construction is some kind of discourse particle rather than a verb with the same basic functions as other wayyiqtols (expressing anteriority plus the semantics inherent in the prefixed ‘and’). ⁷⁴

In a similar vein, note Muraoka’s criticism of the alleged final use of volitive verbal forms introduced by we-. Contrary to the widespread view that this type of construction is a sort of subordinate clause, Muraoka argues that it is better to recognize that in Hebrew the forms are simply coordinated volitive forms:

the syntagm in question does not have a function of formally indicating a purpose. […] The fact that in some cases the second verb can be more elegantly translated as indicating a purpose of the first is essentially a question of pragmatics and translation techniques, and not of descriptive grammar and syntax. ⁷⁵

This is a very precise description of many similar cases.

4. Conclusion

As demonstrated by the examples cited here (and it is indeed possible to cite many more), translation-based interference in the analysis of the Hebrew language, not least of its verbal system, is a widespread problem. When translating, it is often necessary to render passages in ways that are far removed from the actual Hebrew usage. This in itself is not a problem, of course, if the purpose is to produce a readable translation that conveys the general meaning of the original. However, when trying to interpret the semantic content of Hebrew verbal forms or to reconstruct the history of the verbal system, it is necessary to be constantly aware of the tendency to import categories from other languages into the analysis through the use of translations.

In this contribution, I have attempted to show that even a poem like Prov 31.10-31 (though generally rendered in the present tense and often seen as a typical gnomic passage) makes perfect sense when the verbal forms are interpreted in the same way as in other texts, that is, qatal and wayyiqtol as anterior tense and yiqtol (and weqatal) as non-anterior. The gnomic qatal only exists when we choose to translate the verbs in a certain way.

On one hand, the examples of translation-based interference have indicated that this phenomenon is quite widespread. On the other hand, it has also been argued that in all these cases, it is in fact possible to analyze the system without introducing foreign categories if we try to differentiate between translational style and the meanings carried by the Hebrew grammatical forms. In this connection, we should emphasize the role played by exegetical interpretation in any analysis of the Hebrew verbs. Before we draw too firm conclusions about the semantic potential of verbal forms and claim the existence of special usages, it is necessary to examine whether a different exegesis might enable us to understand the sentences in question in a way that would lead to a simpler overall interpretation of the verbal system.

The first task is to recognize the rather restricted repertoire of semantically distinct forms available for the expression of temporal and/or aspectual notions in the Hebrew system as compared to English or similar languages. Specifically concerning the question of the basic character of the Hebrew verbal system, we should start by asking the following questions: Which forms are available in the verbal system? Which category is the most apt designation to cover the different uses of each form? As has been argued in this contribution, relative tense, as understood in the sense outlined above, seems to be the best option for describing the semantic content of the basic opposition between the (non-volitive) finite verbal forms, namely, a simple opposition between anteriority (qatal and wayyiqtol) and non-anteriority (yiqtol and weqatal).