The Linguistic Periodization of the Book of Jonah: E.B. Pusey and Modern Scholarship

Abstract

This article revisits the linguistic periodization of the book of Jonah, focusing on E.B. Pusey’s 1860 commentary and its modern relevance. Pusey challenged claims that Jonah’s unusual lexicon and grammar required a post-exilic date, arguing instead for earlier, non-diachronic explanations such as dialect, foreign influence, and contextual usage. His nuanced treatment anticipated later methodological developments, especially the rule-governed approach of Avi Hurvitz, which identifies diagnostically late linguistic features through late distribution, classical opposition, and extrabiblical confirmation. Applying these criteria, the article surveys more than fifty features in Jonah deemed late by various scholars. Eleven emerge as strong indicators of lateness, while many others show partial or ambiguous significance, often explainable by genre, style-switching, or Phoenician/Aramaic influence. Taken cumulatively, the evidence suggests Jonah’s Hebrew belongs to a late stratum, most plausibly the sixth–fifth century BCE, within the Persian Period, though affinities with Rabbinic Hebrew complicate precise placement. While modern scholarship generally rejects Pusey’s pre-exilic dating, his sensitivity to methodological caution and non-diachronic variety remains instructive. Jonah thus stands as a linguistically peculiar text, chiefly Classical Biblical Hebrew, but with links to Late Biblical Hebrew, Rabbinic Hebrew, and Aramaic, that offers a valuable test case for theories of linguistic periodization.

In 1860 E.B. Pusey published what is generally considered his most consequential contribution as a professor of Hebrew, his commentary on the Minor Prophets.¹ In his introductory remarks on the book of Jonah Pusey takes issue with linguistic arguments raised by certain scholars that undermined traditional views on the book’s authorship and historicity: “As for the few words, which persons who disbelieved in miracles selected out of the book of Jonah as a plea for removing it far down beyond the period when those miracles took place, they rather indicate the contrary.”²

Whatever its literary, historical, or theological significance, Jonah represents something of a linguistic peculiarity. In its lexicon and grammar scholars have long detected indications of lateness, but the diachronic value of these features has often been problematized. Pusey is an early example of this. He treats, point by point, a series of eight linguistic elements that purportedly evidence Jonah’s late provenance (listed here alphabetically): הִתְעַשֵּׁת ‘think, purpose’ (1.6), טַעַם ‘decree’ (3.7), מַלָּח ‘sailor’ (1.5), מִנָּה ‘appoint’ (2.1; 4.6, 7, 8), סְפִינָה ‘vessel’ or ‘deck’ (1.5), רִבּוֹ ‘myriad’ (4.11), רַב חֹבֵל ‘chief sailor’ (1.6), and the relative particle שֶׁ- ‘that, which’ (1.7, 12; 4.10).³ Countering claims that, as late Aramaic penetrations into Hebrew, these betray composition in the Second Temple Period, Pusey is at pains to demonstrate that they are either authentically Hebrew or, if not, reflect a foreign context authentic to the account, thereby negating the purported linguistic evidence against a First Temple Period date of composition.

What is instructive for the modern reader contemplating Pusey’s treatment of the linguistic data over 160 years hence is the extent to which he anticipates similar considerations that have resurfaced over the years and, indeed, persist to this day. Whether centring on Jonah, specifically, or on the linguistic periodization of biblical and extrabiblical texts, more generally, much of the theoretical and methodological reasoning raised long ago by Pusey retains its relevance.

This study is divided into two parts. The first reviews Pusey’s approach to Hebrew diachrony, especially the non-diachronic factors for linguistic variety that he cites, in the light of modern theories and methods. The second part focuses on the linguistic periodization of Jonah specifically, evaluating the periodization of Pusey and others in light of the combined significance of proposed late linguistic features in the book.

Part 1: Assessing Pusey’s Arguments

Data

Like many of his contemporaries, Pusey availed himself of an impressive array of relevant data sets, searching beyond the narrow confines of Biblical Hebrew to draw useful comparisons with cognate languages: namely, Aramaic, Syriac, Phoenician, and Arabic. He can be excused for excluding other material. Akkadian was deciphered in the decade preceding the publication of his commentary. Ugaritic would not be discovered till long after Pusey’s death. The Mesha Stele, inscribed in what is perhaps ancient Hebrew’s closest known cognate, Moabite, was found in 1868, a few years after the commentary’s publication. Little comparative Hebrew material was available: the Dead Sea Scrolls, Hebrew Ben Sira, and the Iron Age II epigraphic sources were unknown (the Siloam Tunnel inscription was discovered in 1880). The written tradition of the Samaritan Pentateuch was recognised for its textual value, but the reading tradition, with its immense diachronic value, would not be studied in detail for almost a century. Pusey’s neglect of Rabbinic Hebrew, while understandable in the contemporary scholarly climate (and even within Judaism)—in which post-biblical Hebrew was widely considered unworthy of serious academic enquiry—is more problematic, especially given W. Gesenius’s earlier reference thereto,⁴ not to mention Pusey’s own recourse to Talmudic Aramaic to supplement Biblical Aramaic.

Pusey’s attention to comparative cognate and Hebrew material remains pertinent, as one still frequently encounters well-intentioned forays into ancient Hebrew periodization that neglect important extrabiblical and non-Masoretic data sets, with the result that Biblical Hebrew is dealt with as a sort of temporal isolate. In diachronic studies Biblical Hebrew in its Tiberian Masoretic manifestation must not be so treated.

Methodology

The scholarly oeuvre of A. Hurvitz holds special importance in the diachronic study of ancient Hebrew and the linguistic periodization of biblical texts.⁵ Building on the work of earlier Hebraists,⁶ he dedicated a career of nearly fifty years to the formulation, illustration, and promotion of a disciplined procedure for the identification of chronolectal markers and the linguistic periodization of ancient Hebrew texts. His research is especially relevant to two scholarly audiences: on the one hand, those who engage in the periodization of biblical texts with no reference to language; on the other, those who pursue the linguistic dating of biblical texts, but whose approach is impressionistic, with little regard for methodological rigour.

Crucially, Hurvitz advocated an approach based on the known dates of control material—Iron Age II inscriptions and Second Temple biblical and extrabiblical sources. In a world blessed with numerous examples of each, there might be little need for methodological strictures. But because early extrabiblical evidence remains far less plentiful than later extrabiblical material, an objective methodology is essential to avoid a situation in which arguments are based primarily on intuition. Due to the temporal asymmetry of the evidence, disproportionate weight necessarily falls on the late material.

In a nutshell, the Hurvitzian procedure calls for the rule-governed assemblage of an inventory of diagnostically late linguistic features and for the broad periodization of texts consistent with the concentration of such features found therein. For inclusion in the inventory, elements must satisfy the three criteria of (a) late distribution, (b) classical opposition, and (c) extrabiblical confirmation. Subsequently, texts are linguistically periodized in accord with their (d) accumulation of late features or lack thereof.

While these criteria may be applied to features from any linguistic domain—phonology, morphology, syntax, pragmatics, sociolinguistics, lexicon, semantics, onomastics—for purposes of illustration, a toponymic example will suffice: consider the minority form of the toponym Damascus, i.e., דַּרְמֶשֶׂק darmɛśɛq.⁷

(a) In terms of biblical distribution, this form is restricted to the late biblical text of Chronicles (1 Chron. 18.5, 6; 2 Chron. 16.2; 24.23; 28.5, 23).⁸

(b) Its classical alternative is דַּמֶּשֶׂק dammɛśɛq (with no resolution of the gemination -mm- to -rm-). This form’s fulfilment of the criterion of classical opposition is clearly illustrated by passages in Samuel and Kings parallelled in Chronicles, e.g.,

‘And when the Syrians of Damascus (דַּמֶּשֶׂק dammɛśɛq) came to help Hadadezer king of Zobah…’ (2 Sam. 8.5)

‘And when the Syrians of Damascus (דַּרְמֶשֶׂק darmɛśɛq) came to help Hadadezer king of Zobah…’ (1 Chron. 18.5)

and

‘And King Asa sent them to Ben-hadad… king of Syria, who lived in Damascus (בְּדַמֶּשֶׂק dammɛśɛq)’ (1 Kgs 15.18)

‘Then Asa… sent them to Ben-hadad king of Syria, who lived in Damascus (בְּדַרְמֶ֖שֶׂק darmɛśɛq)’ (2 Chron. 16.2)

(c) The posited late character of the form דַּרְמֶשֶׂק darmɛśɛq receives corroboration thanks to its appearance in late extrabiblical Hebrew and Aramaic, including the Dead Sea Scrolls, the Targums, and rabbinic literature. Contrast the following cases from MT Isaiah and the Great Isaiah Scroll from Qumran (1QIsa^a) as well as from the Masoretic Torah and Targum Neofiti.

‘An oracle concerning Damascus (דַּמָּשֶׂק dammɛśɛq). Behold, Damascus (דַמֶּשֶׂק dammɛśɛq) will cease to be a city’ (MT Isa. 17.1)

‘An oracle concerning Damascus (דרמשק drmśq). Behold, Damascus (דרמשק drmśq) will cease to be a city’ (1QIsa^a 14.3)

and

‘And he… defeated them and pursued them to Hobah, north of Damascus (לְדַמָּשֶׂק lǝ-dammɛśɛq)’ (MT Gen. 14.13)

‘And he… defeated them and pursued them to Hobah, north of Damascus (לדרמשק l-drmśq)’ (TgN Gen 14.13)

It is unsurprising that the biblical book of Chronicles—composed partly of late renditions of material found elsewhere in the Bible and partly of late independent material—should witness the replacement of classical דַּמֶּשֶׂק with late דַּרְמֶשֶׂק. Though its Hebrew is largely classical, it is nevertheless marked by an accumulation of diagnostically late linguistic features.

By contrast, a useful example of a text of unknown chronological provenance proven late by an accumulation of late features that includes דַּרְמֶשֶׂק is the aforementioned 1QIsa^a.

(d) While MT Isaiah generally profiles as Classical Biblical Hebrew (with chs 40–66 characterized by a somewhat later, though still rather classical, concentration of features⁹), 1QIsa^a shows a contemporized edition of the book of Isaiah, the language of which regularly deviates from classical norms in favour of Second Temple alternatives, including characteristically late דרמשק for classical דמשק.¹⁰

Some notes are in order. It is true that the stipulation that the inventory of diagnostic Second Temple linguistic features be based solely on biblical and extrabiblical material unanimously and unimpeachably dated to that period risks the exclusion of pertinent characteristics limited to texts of less certain, just probable, late provenance. For example, any allegedly late linguistic items limited in the Bible to Jonah are ineligible for the inventory. This is because, though suspected by many to be a late composition, it does not definitively date itself as such—unlike the core Late Biblical Hebrew books of Esther, Daniel, Ezra, Nehemiah, and Chronicles. Only with a completely undisputed starting point can a method really claim to be rule-governed and, thus, reduce subjectivity.

Hurvitz’s methodology is widely accepted as the gold standard among Hebraists.¹¹ This does not mean, however, that its employment inevitably leads to definitive results. It is designed broadly to distinguish pre-exilic from post-Restoration texts. Ambiguity ensues in the case of liminal texts, i.e., those on the fuzzy edges of chronolectal periods established on the basis of more prototypical exemplars. Moreover, certain texts, by their very nature, defy straightforward linguistic periodization. This is due in part to departures from classical norms that are given to non-diachronic explanations. It has long been recognized that temporal language change is not the sole factor contributing to ancient Hebrew diversity; variation can also reflect geographic, social, and literary differences. Genre and brevity can also complicate matters. The poetry in much of the Latter Prophets, Psalms, Proverbs, Job, and Song of Songs makes them difficult to date linguistically. Though a narrative, the short book of Ruth presents a constellation of non-standard features that some, but not all, attribute to lateness. Somewhat akin to Ruth, chiefly narrative Jonah’s idiosyncratic methodological resistance to straightforward diachronic classification is especially acute and to date has not received the detailed, comprehensive, and methodologically robust treatment it deserves.¹²

Pusey on Jonah’s Language

The peculiarity of Jonah’s Hebrew has often been discussed.¹³ Its departures from classical norms have been attributed to both chronolectal and non-diachronic factors, such as foreign influence and regional dialect. And it is precisely on these matters that Pusey’s brief discussion foreshadows current arguments. As noted above, in the case of most of the eight non-standard linguistic features that Pusey treats, he explains away a purported late origin in Aramaic, opting instead for a non-diachronic explanation for Jonah’s linguistic diversity. His argumentation is sophisticated, heralding the methodological strictures eventually codified by Hurvitz as well as impressive linguistic nuance.

Consider his approach to so-called ‘Aramaisms’. It is well known that Second Temple biblical and extrabiblical Hebrew texts show a relative proliferation of Aramaic, Akkadian, and Persian loans resulting from contact with Aramaic, which enjoyed the status of an administrative language and lingua franca during that period. But some suggested Aramaisms have proved to be ‘false positives’, rare but classical Hebrew features found more commonly in Aramaic which, however, need not be attributed to language contact, late or otherwise. For Pusey a given feature’s Aramaic-like appearance is insufficient to mark it as late without first excluding other possibilities. While some elements of Pusey’s arguments have not stood the test of time,¹⁴ his approach presages the caution advocated by Hurvitz and others regarding the chronological significance of Aramaisms, the identification of which can be definitively established only through careful application of the methodology described above.¹⁵

Also exemplary is Pusey’s rejection of a methodology limited to the criterion of late distribution, especially his awareness of the importance of the principle of classical opposition. For example, he reasonably doubts the significance of the late distribution of רַב חוֹבֵל ‘chief sailor’, which includes the component רַב in the late function of an honorific, on the grounds that early biblical material furnishes no context calling for its use.¹⁶

His discussion of the relativizer שׁ- is also ahead of its time. He notes, inter alia, a biblical distribution that both includes classical texts and excludes most late texts, the probable connection with Phoenician ʾiš, as well as several considerations arguably favouring the employment of שׁ- in specific passages (see below).

Even so, his line of reasoning is not without its methodological weaknesses. He overlooks the significance of genre: Pusey seems to ignore the possibility that linguistic evolution might involve the growing prose use of a feature formerly characteristic of poetry. The parade example of this is מַלְכוּת ‘kingdom, kingship, reign’, which was clearly a perfectly legitimate, but rare and mainly poetic, lexeme in Classical Biblical Hebrew, until its post-exilic promotion to more widespread usage, presumably by writers familiar with its extremely frequent Aramaic cognate. Such complexity notwithstanding, the lexeme מַלְכוּת is widely recognized today as a useful, though by no means decisive, indicator of post-exilic composition. Due, however, to its sporadic early appearances, only as one among many pieces of linguistic evidence for lateness does it acquire diachronic significance. In Pusey’s discussion, the failure to distinguish early, specifically poetic usage from late more general employment is conspicuous in his discussion of רִבּוֹ ‘myriad’,¹⁷ where he meticulously summarizes a distribution beyond Jonah consisting of works either late, poetic, or both, but takes the apparently early poetic use as broadly representative of the classical chronolect. Similar criticism may be levelled regarding his treatment of מִנָּה ‘appoint’.¹⁸ Since outside Jonah and biblical poetry both רִבּוֹ ‘myriad’ and מִנָּה ‘appoint’ have chiefly late biblical distributions, exhibit classical alternatives, and are widely used in Second Temple extrabiblical Hebrew and Aramaic sources, they may safely be classified as diagnostically late linguistic features (see below).

Foreign Influence

Returning to the title רַב חֹבֵל, the honorific use of רַב has long been seen as a post-classical feature.¹⁹ As a part of a compound title, the word is used primarily of foreign officials in the later manifestations of Classical Biblical Hebrew and in Late Biblical Hebrew. Its most conspicuous late usage is in reference to Israelites, a trend culminating in Rabbinic Hebrew. This expanded use in Hebrew is said, quite plausibly, to have arisen as a result of foreign influence, from or via Aramaic. Its classical synonyms include such forms as שַׂר and גָּדוֹל. Jonah’s רַב חֹבֵל is often listed as such a late usage.

For his part, Pusey (1860, 250) objects, and this on several grounds. As already noted, he compellingly accounts for the collocation’s absence from earlier Hebrew by noting that no other context calls for its use. He also questions the supposed semantic and functional equivalence of רַב and its proposed classical synonym שַׂר. More relevant to the issue of ‘the foreign factor’ are two further considerations raised by Pusey: (1) the honorific usage of רב exists in Phoenician and (2) the ship in question was Phoenician. Pusey does not argue here for style-switching, i.e., the conscious adoption of foreign (but intelligible) language in support of the literary depiction of foreigners and/or a foreign context.²⁰ He rather sees use of the title רַב חֹבֵל as an authentic reference to the Phoenician that Jonah himself heard in Joppa.

It is worth highlighting the compelling force of Pusey’s argumentation here. Whatever the overall diachronic profile of the Hebrew of the book of Jonah, the honorific use of רַב in the title רַב חֹבֵל per se seems a linguistic feature of rather dubious chronological weight, since even in the Bible use of רַב in foreign honorific titles seems to predate the Persian Period. All other things being equal, רַב חֹבֵל is not obviously later than רַב־שָׁקֵה ‘Rabshakeh’, who appears in the Classical Biblical Hebrew of 2 Kings 18–19 and Isaiah 36–37.²¹

Pusey makes a similar argument for the only form he does deem authentically Aramaic in Jonah, טַעַם in the specific meaning ‘decree’ (rather than standard ‘sense, taste’), the issuing of which, Pusey opines, the prophet witnessed first-hand. Pusey’s argument for the possibility of a foreignism in the case of טַעַם in the sense ‘decree’ is only partially valid. On the one hand, since Akkadian ṭēmu is likely the origin of both Aramaic טְעֵם and Hebrew טַעַם in the meaning ‘decree’, the etymon in the relevant sense was in theory available in the eighth century bce. On the other hand, when it comes to foreign borrowings into Hebrew, the most important consideration is not their abstract availability, but their concrete currency among the readership. Since טעם ‘decree’ is late in Aramaic itself, being attested no earlier than Imperial Aramaic,²² it is not clear how Hebrew users could have been expected to be acquainted with the usage prior to the fifth century bce.

Broadening the focus, it is worth noting that consideration of ‘foreignness’ has long played a role in arguments concerning linguistic periodization. Aramaic, Akkadian, and Persian loanwords are associated with Late Biblical Hebrew, whereas Egyptian and Phoenician have sometimes been invoked to explain ostensibly early non-standard Hebrew. Consider Ohad Cohen’s recent call for scholars to acknowledge the persistent role of Phoenician in the linguistic milieu of the Persian Period.²³ On the one hand, Cohen’s research successfully substantiates the presence of Phoenician (or, more generally, Canaanite) culture and language into the Persian Period from Tyre and Sidon, down the coast, and all the way to Idumaea. If Cohen is correct, namely, that Phoenician continued to penetrate Hebrew well into the Persian Period, then in the present connection, Phoenician influence cannot be considered a sure sign of Iron Age periodization. On the other hand, Cohen’s contention that contact with Phoenician better accounts for certain Second Temple Hebrew developments than does contact with Aramaic must be assessed on a case-by-case basis. In some instances, a Phoenician explanation is either no more compelling, or seems considerably less compelling, than an Aramaic explanation.

For example, though Cohen sees Hebrew שׁ- as a marker of Late Biblical Hebrew, he largely rejects Aramaic influence as a factor in its late proliferation, preferring to attribute it to late contact with Phoenician—this despite his recognition of the functional overlap of Aramaic ד(י)-.²⁴ The chief reason for Cohen’s attribution to Phoenician, rather than Aramaic, contact is apparently etymological. Whereas the archaic Phoenician of Byblos represented proto-Semitic ḏ- with ז, by the early ninth century bce, all the Phoenician dialects had replaced that latter with ʾiš.²⁵ Since Phoenician ʾiš is etymologically closer than Aramaic ד(י)- to Hebrew שׁ-, Cohen reasons that Phoenician contact is responsible for LBH שׁ-.

Beyond the important objections to a superficial association between Late Biblical Hebrew and the particle שׁ- raised by Pusey—especially, a biblical distribution that includes classical texts and excludes most late material—it is necessary to call into question a few of the planks in Cohen’s argument.

First, whatever the etymological link between the Phoenician and Hebrew particles, as Cohen himself observes, there exists a high degree of functional equivalence between the particles in all three languages. This arguably obviates the need for shared etymology. On the assumption of Aramaic influence on Late Biblical Hebrew Cohen seems to opine that we should expect Hebrew ז- or ד-, rather than שׁ-. But a strict lexical borrowing is not the only possibility. The expanded use of a native feature formerly characterized by more restricted use due to its formal and functional similarity to an etymologically unrelated foreign parallel is also possible. Here Pusey’s contention, shared by Cohen,²⁶ that שׁ- was a vernacular feature subsequently promoted to literary use in works characterized by non-prototypical biblical style, tallies with the notion that multilingual users may have calqued a functional, rather than etymological equivalent.

Second, it is worth pointing out that not all Biblical Hebrew uses of שׁ- line up with its Phoenician employment. Consider such complement or content clauses as יָדַעְתִּי שֶׁגַּם־זֶה הוּא רַעְיוֹן רוּחַ ‘I knew that this too is chasing the wind’ (Qoh. 1.17). This is not a function of Phoenician ʾiš, but is characteristic of Aramaic ד(י)-.²⁷ Moreover, several of the compound expressions with שׁ-, including those in Jonah, have obvious parallels in Aramaic, but not in Phoenician.²⁸ It is not impossible that Phoenician influence is behind some Late Biblical Hebrew usages and Aramaic influence behind others, but beyond the matter of methodological parsimony, the usage of the שֶׁ- compounds in Jonah is consistent with Aramaic, not Phoenician influence.

Additional Non-diachronic Factors Contributing to Biblical Hebrew Diversity

Similar to arguments for the foreign rather than late character of non-standard linguistic features are alternative or complementary hypotheses according to which they represent regional diversity, vernacular departures from an otherwise literary register, or deliberate literary strategies, such as style-switching or addressee-switching. Before proceeding, it is worth highlighting that sensitivity to non-diachronic linguistic variety does not necessarily entail wholesale rejection of linguistic periodization. A useful case in point is the scholarly output of Gary Rendsburg, who has famously written on all the aforementioned non-diachronic factors contributing to Biblical Hebrew diversity, and yet remains convinced of the possibility of detecting chronolects in Biblical Hebrew.²⁹

Since Pusey considers Jonah more of a report than a piece of literature, he does not cite style-switching as an explanation for non-standard language; rather, he identifies genuine foreign forms, perhaps adjusted for Hebrew pronunciation. He does occasionally, however, note the possibility of regional and register variation. Thus, when it comes to רִבּוֹ ‘myriad’, appearances in Hosea and Jonah lead Pusey to entertain the possibility that the form might be characteristic of northern speech—though, in his mind, this fails as a comprehensive explanation, since the word also appears in a psalm of the southerner David. As already noted, one might have wished in this connection for greater sensitivity to genre distinctions.

Pusey’s aforementioned inattention to genre distinctions in the case of רִבּוֹ ‘myriad’ and מִנָּה ‘appoint’ is all the more glaring when compared to the sophistication of his discussion of שׁ-, where he considers, in addition to foreignness, a multiplicity of factors, including regional variation, register, and even rhythm. While Pusey is too forceful in the complete dismissal of attaching any diachronic weight to שׁ-, his arguments are creditably nuanced, more so than many modern approaches, which rather uncritically take שׁ- as evidence of late slips into Rabbinic-style Hebrew. Methodologically, שׁ- should be assessed as diachronically meaningful only in accord with the general linguistic profile of the work in which it is found, which must be assessed based on an accumulation of diagnostically late features or the lack thereof. In other words, notwithstanding its association with Rabbinic Hebrew in the minds of scholars, שׁ- is not itself an innovation of post-exilic Hebrew. Even so, contrary to extreme alternative positions, there can be little doubt that its use in some late material, couched in Hebrew replete with other late features, is due to its expanded late usage, in which case it, too, joins them as an indicator of lateness. The question remains: Is Jonah a work characterized by such an accumulation?

Part 2: Reassessing Jonah’s Hebrew

To this point, the discussion has exploited Pusey’s treatment of Jonah’s language to explore broader issues of theory and method. Various strengths and weaknesses in Pusey’s approach have been identified, yet no verdict on Pusey’s general conclusion—that Jonah is pre-exilic despite its apparent deviations from Classical Biblical Hebrew—has been offered. This is because, while Pusey’s treatment is useful as far as it goes, his coverage of a mere eight features omits many potentially significant elements. Indeed, according to the representative survey of treatments of Jonah’s language summarized, there are some 55 features deemed late by one scholar or another.

Potentially Late Linguistic Features in Jonah

Applying Hurvitz’s procedure to the late features in Jonah variously proposed by scholars, one arrives at mixed results. The features broadly divide into three tiers.³⁰

Tier 1

Tier 1 includes eleven features whose late character should be judged between highly probable and nearly certain, in that they satisfy all three of Hurvitz’s criteria of late distribution, classical opposition, and extrabiblical confirmation.

• גּוֹרָלוֹת ‘lots’ (1.7, 7); cf. גּוֹרָל

• הִפִּיל גּוֹרָלוֹת ‘cast lots’ (1.7, 7); cf. הִשְׁלִיךְ/יָדַד/יָרָה גּוֹרָל

• הַרְבֵּה מִן ‘more than’ (4.11); cf. רַבִּים/רַבּוּ מִן

• (יָדַעְתִּי כִּי אַתָּה אֵֽל־)חַנּוּן וְרָחוּם ‘(I knew that you are a) gracious and merciful (god)’ (4.2); cf. רָחוּם וְחַנּוּן ‘merciful and gracious’

• מִגְּדוֹלָם וְעַד־קְטַנָּם ‘from their greatest to their least’ (3.5); cf. (לְ)מִקָּטֹן וְעַד גָּדוֹל ‘from least to greatest’

• מַהֲלָךְ ‘distance’ (3.3, 4); cf. דֶּרֶךְ

• מִנָּה ‘appoint’ (2.1; 4.6, 7, 8); cf. הִפְקִיד, פָּקַד, שָׂם, שָׁת, נָתַן

• נָפַל הַגּוֹרָל ‘the lot fell’; cf. הָיָה/יָצָא/עָלָה הַגּוֹרָל

• עִמָּהֶם ‘with them’ (1.3); cf. עִמָּם

• רִבּוֹ ‘myriad’ (4.11); cf. רְבָבָה

• verbs taking object suffixes (1.12, 12, 15; 2.3, 4, 4, 6, 6; 4.10); cf. verbs with declined forms of אֵת

In the case of a few of the above, fulfilment of the criterion of late biblical distribution might be questioned based on their sporadic appearance in Wisdom literature and/or biblical poetry: הִפִּיל גּוֹרָלוֹת ‘cast lots’, מִנָּה ‘appoint’, רִבּוֹ ‘myriad’. However, as argued above in the case of מַלְכוּת ‘kingdom, reign’, the late penetration into prose of features formerly restricted to non-prose genres, often explicable as the expanded use of an existing Aramaic-like feature due to late Aramaic contact, is a recognized developmental path for diagnostically late linguistic features. Thus, a feature’s potentially early biblical attestation in poetic or Wisdom literature does not necessarily disprove a generally late biblical distribution.

Tier 2

Tier 2 includes 28 features consistent with a late date of composition, but which, for various reasons, must be deemed somewhat less certain in terms of their late diagnostic value.

• אֱלֹהֵי הַשָּׁמַיִם ‘God of Heaven’ (1.9); cf. יְהוָה, אֱלֹהִים, יְהוָה אֱלֹהִים

• אָמַר ‘command’ (2.11); cf. צִוָּה

• אֲשֶׁר יֵשׁ בָּהּ ‘in which there are’ (4.11); cf. אֲשֶׁר בָּהּ

• בִּן לַיְלָה ‘overnight’ (4.10); cf. בַּלַּיְלָה הַהוּא, *בְּעֶצֶם הַלַּיְלָה

• בָּרַח מִלִּפְנֵי ‘flee from, before’ (1.3, 10); cf. בָּרַח/נָס מִפְּנֵי.

• בְּשֶׁל, בַּאֲשֶׁר לְ- ‘because of’ (1.7, 8, 12); cf. בִּגְלַל-, מִפְּנֵי, עַל אֹדוֹת

• הִצִּיל לְ- ‘save, rescue somebody’ (4.6); cf. הִצִּיל אֵת, הִצִּילוֹ

• הִתְעַשֵּׁת ‘think, purpose’ (1.6); cf. חָשַׁב

• זָעַף ‘rage, storm’ (1.15); cf. סָעַר

• חָס ‘pity’ (4.10, 11); cf. חָסָה עַיִן, חָמַל

• טַעַם ‘decree’ (3.7); cf. דְּבַר (הַמֶּלֶךְ),צָו

• כַּאֲשֶׁר חָפַצְתָּ עָשִׂיתָ ‘as you have desired, you have done’ (1.14); cf. עָשָׂה הַטּוֹב בְּעֵינֵי- ‘do what is good in one’s eyes’

• לַמָּחֳרָת ‘on the following day’ (4.7); cf. מִמָּחֳרָת

• מַה לְּךָ נִרְדָּם ‘why are you sleeping?’ (1.6); cf. *מַה לְּךָ כִּי תֵרָדֵם/נִרְדַּמְתָּ

• נָגַע אֶל ‘arrive to’ (3.6); cf. הִגִּיעַ אֶל

• נָקִיא ‘innocent’ (1.14); cf. נָקִי

• סְפִינָה ‘vessel’ (1.5); cf. אֳנִיָּה, אֳנִי

• עָמַל ‘toil’ (4.10); cf. יָגַע

• קִדַּמְתִּי (לִבְרֹחַ) ‘arise early (to flee)’ (4.2); cf. *הִשְׁכַּמְתִּי (וָאֶבְרַח)

• קִדַּמְתִּי לִבְרֹחַ ‘arise early to flee’ (4.2); cf. *הִשְׁכַּמְתִּי וָאֶבְרַח

• קָרָא עַל ‘call to’ (1.2); cf. קָרָא אֶל

• קָרַב ‘approach’ (1.6); cf. נִגַּשׁ

• קְרִיאָה ‘call’ (3.2); cf. דָּבָר (?), קָרוֹא (?)

• רַב (חֹבֵל) ‘chief sailor’ (1.6); cf. שַׂר, גָּדוֹל

• רַע אֶל ‘displease’ (4.1); cf. רַע בְּעֵינֵי- ‘evil in the eyes of’

• שֶׁ- ‘which, that’ (1.7, 12; 4.10); cf. אֲשֶׁר

• שָׁתַק ‘become quiet, still’ (1.11, 12); cf. חר"ש, דמ"ם/דו"ם, חש"י, שק"ט

• verbal emphasis by means of cognate/internal object (1.10, 16; 3.2; 4.1, 6); cf. paronomasia with a cognate infinitive absolute

Some features fulfil all three procedural requirements, but must be considered questionable in terms of diagnostic significance on other grounds. In some cases the ‘foreign factor’ may have conditioned non-standard usages. The title רַב (חֹבֵל) ‘chief sailor’ likely refers to a non-Israelite, whereas both אֱלֹהֵי הַשָּׁמַיִם ‘God of Heaven’ and שָׁתַק ‘become quiet, still’ come in conversation with foreigners. Doubt also attaches to the significance of Jonah’s consistent use of cognate/internal objects for verbal emphasis, as this involves the eschewal of a classical feature in favour of an alternative which is itself classical.

The remaining Tier 2 features satisfy only two of Hurvitz’s criteria. In most of these cases, the requirement unmet is that of late biblical distribution, the feature either unique to Jonah in the Bible or occurring elsewhere in biblical literature, but not in Late Biblical Hebrew proper.³¹ Some of these, however, are arguably rather strong indicators of lateness, their exclusion from Tier 1 due to what amounts to a technicality. For example, as a hapax legomenon, טַעַם in the meaning ‘decree’ is by definition disqualified. But as an indirect calque from Akkadian mediated by Aramaic, in which the specific meaning is also late, its diagnostic value is arguably strong.³² Furthermore, the absence of such features as כַּאֲשֶׁר חָפַצְתָּ עָשִׂיתָ ‘as you have desired, you have done’ and עָמַל ‘toil’ from the core Late Biblical Hebrew books seems accidental, as they appear in other texts with widely acknowledged late linguistic profiles, e.g., Isaiah 40–66, late psalms, and Qohelet.

Tier 3

Tier 3 comprises sixteen features identified by one scholar or another as characteristically late, the methodological significance of which is highly dubious.

• אָנָּה ‘please’ (1.14; 4.2); cf. אָנָּא

• אֲנִי ‘I’ (1.9, 12; 2.5, 10; 4.11); cf. אָנֹכִי

• דָּגָה ‘(individual) fish’ (2.2); cf. דָּג

• הֵטִיל ‘cast, throw’ (1.4, 5, 12, 15); cf. הִשְׁלִיךְ

• הֶעֱבִיר מֵעַל ‘remove (a garment) from upon’ (3.6); cf. הֵסִיר מֵעַל

• חֹבֵל ‘sailor’ (1.6); cf. *אִישׁ אֳנִיָּה, *יֹדֵעַ יָם, *יֹרֵד יָם

• הֶחָמָס אֲשֶׁר בְּכַפֵּיהֶם ‘the violence that is in their hands’ (3.8); cf. ?

• חֲרִישִׁית ‘silent’ or ‘scorching’ (4.8); cf. ?

• (וְהָ֣אֳנִיָּ֔ה) חִשְּׁבָה לְהִשָּׁבֵר ‘(and the vessel) threatened/was about to break apart’ (1.4); cf. ?

• יָרֵא ‘fear (with religious reverence)’ (1.9, 10, 16); cf. ?

• (יָדַעְתִּי) כִּי (אַתָּה אֵֽל־חַנּוּן וְרָחוּם) ‘(I knew) that (you are a gracious and merciful)’; cf. ?

• (יָדַעְתִּי) כִּי אַתָּה אֵֽל־חַנּוּן וְרָחוּם ‘(I knew) that you are a gracious and merciful god’ (4.2); cf. *כִּי אֵל רָחוּם וְחַנּוּן אַתָּה

• הַמֶּלֶךְ וּגְדוֹלָיו ‘the king and his nobles’ (3.7); cf. ?

• מַלָּחִים ‘sailors’ (1.5); cf. אַנְשֵׁי אֳנִיּוֹת ‘boatmen’, יֹדְעֵי יָם lit. ‘(those) familiar with the sea’, יוֹרְדֵי הַיָּם lit. ‘(those who) descend to the sea’

• מִן ‘from’ (3.8; 4.5); cf. -מֵּ

• קִיקָיוֹן ‘castor-oil plant (?)’ (4.6, 6, 7, 9, 10); cf. ?

In several instances, e.g., דָּגָה ‘(individual) fish’, חֲרִישִׁית ‘silent’ or ‘scorching’, and קִיקָיוֹן ‘castor-oil plant (?)’, the feature may indeed be late, but rarity in both biblical and extrabiblical sources leaves it with no clear late distribution. Such infrequency also often results in semantic ambiguity, on account of which it is difficult to identify a classical alternative.

In other cases, the available information militates against classification as a late feature. It is true that Jonah’s use of אֲנִי ‘I’ is in line with late tendencies, but more characteristically classical אָנֹכִי appears in the same verse. Mixed usage of the two first person subject pronouns is atypical of Late Biblical Hebrew works. Despite some numerical fluctuation, the verb הֵטִיל ‘cast, throw’ is not obviously more characteristically late than הִשְׁלִיךְ in either biblical or extrabiblical Hebrew. The verb יָרֵא ‘fear (with religious reverence)’ is abundantly documented in classical texts and its sense in Jonah is not obviously more religious than in other cases. Bendavid holds that כִּי ‘that’ in Jon. 4.2 is a pseudo-classical reflex of lateשׁ-,³³ but the usage is utterly classical. Jonah’s retention of the nun in מִן ‘from’ is in line with classical patterns.

One of the more methodologically interesting features in Tier 3 is מַלָּחִים ‘sailors’. The lexeme’s only biblical appearances outside Jonah are in exilic/early post-exilic Ezekiel. Persisting in Qumran Hebrew, in Rabbinic Hebrew it is largely replaced by סַפָּן. Despite the rarity of nautical terms in biblical literature, it seems—contra Pusey—to be paralleled by classical alternatives. As such, it is sometimes regarded as a late linguistic feature.³⁴ Even so, its absence from Classical Biblical Hebrew may be accidental, since—in agreement with Pusey (see above)—it exists in Iron Age Phoenician³⁵ and is regarded by Mankowski as a non-trans-Aramaic Akkadian loan.³⁶ It is also not impossible that its employment in Jonah is due to literary artistry, i.e., style-switching to depict Phoenician sailors.

Summary of Findings

Coming to the matter of Jonah’s overall linguistic periodization, the eleven features included in Tier 1, representing 25 tokens, constitute a meaningful accumulation relative to book length (48 verses: forty prose, eight poetry). Though the features in Tier 2 do not reach the same level of probability, they are plausibly late. Thus, a rather impressive array of probative data combines with a preponderance of circumstantial evidence to point to a probable late date for Jonah.

But how late? Is Jonah reflective of the early or late Persian Period (550–330 bce) or of Hellenistic (post-330 bce) language use? Should it be considered on a par with Late Biblical Hebrew proper or something of a transitional stage linking Classical and Late Biblical Hebrew? While the evidentiary grounds for Jonah’s broadly late linguistic periodization are relatively firm, finer chronological conclusions are more difficult to prove. The book’s language remains exceptional within Biblical Hebrew. Viewed from a narrow biblical perspective, the lack of greater affinity with Late Biblical Hebrew proper seems weighty. Adopting a broader perspective, however, Jonah’s numerous affinities with Rabbinic Hebrew and late Aramaic are hard to ignore, even if the majority are not shared with the core Late Biblical Hebrew books. Indeed, a precise diachronic assessment of the book’s language ultimately hinges on the significance of its frequent Tier 2 features. They are either diachronically meaningful isoglosses with Rabbinic Hebrew and late Aramaic or represent a casual concentration of such features. Some of these features may be early and only accidentally omitted from Classical Biblical Hebrew, but in the context of an accumulation of Tier 1 features, it beggars belief to view such frequent affinity as random and devoid of temporal significance.

Along with several scholars, Pusey among them, the approach adopted here considers non-diachronic explanations for Jonah’s departures from classical norms.³⁷ According to the foregoing argumentation, accounts attributing non-standard usages to genre or literary strategy are deemed more persuasive than vague allusions to northern dialectal variation. It should be noted that even the most sceptical of Jonah’s late linguistic periodization among the modern scholars surveyed, i.e., Landes, finds sufficient evidence to date Jonah somewhere in the sixth century bce, in near agreement with Brenner and Golan Ben-Uri, who accept the late diagnostic significance of a greater number of features and more precisely date the book to the second half of the sixth century/beginning of the Persian Period.³⁸ Rofé argues for a mid-fifth century date on linguistic and other grounds,³⁹ which is basically in line with Driver’s view.⁴⁰ Bendavid likens Jonah’s Hebrew to that of Esther, in no small part because each achieves linguistic variety via what he considers the artistic mixing of early and late features, situating both in the Second Temple Period.⁴¹ Almbladh can be no more precise than ‘post-exilic’,⁴² while Dan cautiously avers that the linguistic and non-linguistic evidence tips the scales in favour of a late date.⁴³

It is difficult to be more precise than sixth–fifth century bce. Arguable affirmative evidence in favour of situating the book earlier in the Persian Period includes the absence of Persian words;⁴⁴ the mixed usage of אֲנִי and אָנֹכִי ‘I’ (with no apparent conditioned use of the latter);⁴⁵ and the use of נָא.⁴⁶ A later placement would seem commensurate with numerous features shared with Rabbinic Hebrew. Again, Jonah’s unique linguistic combination of classical lexemes and grammar, unambiguously late features, and Mishnaisms and Aramaisms not found in Late Biblical Hebrew proper render it methodologically challenging in terms of linguistic periodization.

The results insofar as they pertain to E.B. Pusey are mixed. On the one hand, those modern scholars who have dealt with Jonah’s language in detail generally reject the pre-exilic date that Pusey championed. On the other hand, the suggested late significance of features is often still problematized on the very non-diachronic grounds that Pusey raised, e.g., dialect, foreign context, or otherwise disqualify alleged diagnostic features on similar methodological grounds. Further, several of them distinguish Jonah’s approximation of classical style from Late Biblical Hebrew proper, partially vindicating Pusey take on the book’s language.

As noted at the outset, Pusey could not exploit significant data sets available to modern scholars and treated just eight of the 55 features listed above. Even so, it is by no means certain that, afforded access to more extensive Hebrew and comparative material, a scholar of Pusey’s traditionalist bent would have concluded that Jonah was anything other than an eighth-century bce composition. Diachronic linguistics is sometimes weaponized by those who seem to have predetermined the date of chronologically disputed material. In such cases, it is easy to cherry pick diachronically meaningful data to confirm a presumed date of composition. This is a misuse of the proposed methodology. While it is wholly legitimate and desirable to combine linguistic and non-linguistic approaches to literary periodization, the linguistic data should, in the first instance, be assessed independently of other considerations. Ideally, maximal recourse to data, sound procedures, and judicious and nuanced interpretation feed into a highly explanatory framework characterized by the weighing of diachronic and non-diachronic explanations that are complementary rather than mutually exclusive.