In a collection of essays edited by Young (2003a), the chronological model was challenged by the claim that all biblical literature has its origin in the Persian era or later. Davies (2003) argues that Persian-period scribes wrote several varieties of Hebrew, and therefore it is conceivable that classical Hebrew was one of these.

Young (2003b:314–317) provides the following outline based on the work of Talshir (2003:251–275): a) Early Biblical Hebrew continued to be the language of Yehud until the Persian period, especially in those sources without an eastern bias, such as Haggai, Zechariah and Malachi; b) Late Biblical Hebrew is connected with the eastern diaspora; proto-Late Biblical Hebrew features first began to make their presence felt strongly in literary Hebrew associated with the exiles in the eastern diaspora (Ezekiel being the first example); c) in the days of the Second Temple period, political separation saw the development of a separate dialect, Tannaitic (Mishnaic) Hebrew in the lowlands, whilst in Yehud proper, Hebrew remained more conservative. Although neither Qumran Hebrew nor Mishnaic Hebrew is identical to Late Biblical Hebrew, there are important isoglosses which they share with Late Biblical Hebrew in opposition to Early Biblical Hebrew (see also Rezetko 2003).

Young et al. (2008; see Naudé 2010) argue that Early Biblical Hebrew and Late Biblical Hebrew do not represent different chronological periods in the history of Biblical Hebrew, but instead represent coexisting styles of literary Hebrew throughout the biblical period and are best taken as representing two tendencies among scribes of the biblical period: conservative and non-conservative. The authors and scribes who composed and transmitted works in Early Biblical Hebrew exhibit a tendency to conservatism in their linguistic choices, in the sense that they only rarely use forms outside a narrow core of what they considered literary forms. At the other extreme, the Late Biblical Hebrew authors and scribes exhibited a much less conservative attitude. Between extreme conservatism (e.g. Zechariah 1–8) and extreme openness to variety (e.g. Ezra), there was a continuum into which other writings may be placed (e.g. Ezekiel).

At Qumran too, the Damascus Document and the Temple Scroll fall somewhere in the middle between Early Biblical Hebrew and Late Biblical Hebrew. The other samples of Qumran (especially Pesher Habakkuk) and Ben Sira studied by Young et al. (2008:250–279) fall in the mid to high end of the Early Biblical Hebrew scale, even further from the core Late Biblical Hebrew books.

This is a reaction to viewpoints in Miller-Naudé and Zevit (2012) Diachrony in Biblical Hebrew. It is clear that many of the suggestions and ideas of the authors find a place in the volume of Rezetko and Young (2014) (see Rezetko & Young 2014:593–599), for example, their integrated approach versus the idea of complexity theory (Naudé 2012) as well as the ideas concerning the terms change, diffusion, variation, idiolects, etc. as explicated in Chapter 1.

Rezetko and Young (2014:10–11) now use the phrase ‘Classical Hebrew’ for the four premishnaic corpora: the Hebrew inscriptions, Ben Sira, the Dead Sea Scrolls, and the Hebrew Bible. Instead of the two main types or periods of Biblical Hebrew (Golden Age, Early, Classical, or Standard Biblical Hebrew on the one hand, and Silver Age or Late Biblical Hebrew on the other hand) they use the term ‘Standard Classical Hebrew’ and ‘Peripheral Classical Hebrew’ respectively for linguistic forms/uses that are ‘standard’ and that are ‘peripheral’ or ‘non-standard’ in the same corpus. They use these terms in ways that differ from their usage in dialect geography or historical dialectology, using them as descriptive labels for linguistic items which occur more or less frequently in the surviving written specimens of ancient Hebrew.

Rezetko and Young (2014:56–57) argue that cross-textual variable analysis and variationist analysis can help scholars of Biblical Hebrew to get a much better grasp of the linguistic facts of Biblical Hebrew and therefore to formulate eventually a better history of ancient Hebrew. Their main contention is that historical linguistic study of Biblical Hebrew should aim to target, record, organise, and evaluate individual linguistic items, their processes of variation and change in specific compositions and manuscripts, not only or mainly in the Masoretic Text or in assemblages of biblical books or from the perspective of the conventional (or any other) periodisation of Biblical Hebrew.

Rezetko and Young (2014:115–116, 210) further claim that it should not be postulated that the Masoretic text reflects the original text of the biblical books better or more frequently than any other text. As a result, it also should not be postulated that the language of the Masoretic text reflects the original language of the biblical authors better or more frequently than any other text. For them, neither the Masoretic text, nor any other biblical text is likely to preserve the authentic details of the language of any biblical author (Rezetko & Young 2014:406). They claim that there are many late adjustments and additions to biblical writings. For example, by taking into account Qumran Samuel, the book of Samuel must have a complex history of production that lasted from early in the First Temple period until late in the Second Temple period, with the implication that the language of the book is a witness to written Hebrew throughout this entire extended period of time (Rezetko & Young 2014:210). Their interpretation is that the largest proportion of linguistic variations between the Masoretic text and Qumran Samuel are individual variants, as opposed to large-scale systematic variations which do not support the conventional historical linguistic perspective of viewing each corpus as a coherent whole in its own right with an unambiguously distinctive linguistic profile (Rezetko & Young 2014:210, 328). However, Rezetko and Young (2014:403) acknowledge that it is difficult or impossible to know the precise reasons behind the linguistic variations in the Hebrew Bible. They do not attribute all variation to style, but they keep the explanation on the table (Rezetko & Young 2014:408).

In the following sections, we examine one aspect of Hebrew syntax within Qumran Hebrew, namely the syntax of the negation of the participle in order to discover to what extent Qumran Hebrew is the same or different from Biblical Hebrew (see Muraoka 2000 for a similar approach) and whether linguistic variation should be attributed solely to style.

The most common construction in Biblical Hebrew is also attested at Qumran, namely, ʾên followed by a pronominal suffix as the subject of the participle, as illustrated in (3):

(3) 4Q396 f1–2i:1

The scope of negation extends to the entire sentence.

The usual word order in this construction is ʾên with a pronominal suffix (the subject) and the participle, followed by possible objects and adjuncts (e.g. prepositional phrases). In two variants to this word order, a non-verbal constituent occurs in a position before the negative particle. In the first construction, the constituent is moved to the front of the sentence, but remains within the sentence. This construction is referred to as the topicalisation of the constituent (Holmstedt 2009, 2014). Topicalisation is typically used either to highlight the informational topic of the sentence, which orients a reader or hearer to the theme from the context, or to the focus of the sentence, which instructs a reader or hearer to contrast information with other alternatives (Holmstedt 2009). In the second construction, the constituent occurs outside of the boundary of the sentence and a resumptive pronoun within the sentence refers to it. This construction is referred to as left dislocation (see Holmstedt 2014; Naudé 1990).

Both topicalisation and left dislocation occur with this construction at Qumran. In (4) the object constituent is topicalised before the negative marker:

(4) 4Q394 f8iv:1 (= 4QMMT)

The object of the sentence (אשׁם) occurs before the negative particle and participle; there is no resumption of the topicalised object in the sentence (contra Geiger 2012:306).

There are two very interesting diachronic facts concerning this construction. Firstly, in a number of cases involving parallel texts in Biblical Hebrew, the earlier text uses lōʾ followed by a yiqtol form, whereas the later text uses ʾên with a pronominal suffix followed by the participle (see Geiger 2012:303). Compare the examples in (5a) and (5b):

(5a) 1 Kings 22:8

And the king of Israel answered Jehoshaphat, ‘There is one more man through whom we can inquire of the LORD; but I hate him, because he never prophesies anything good for me, but only misfortune – Micaiah son of Imlah.’ But King Jehoshaphat said, ‘Let not the king say so!’

(5b) 2 Chronicles 18:7

And the king of Israel answered Jehoshaphat, ‘There is one more man through whom we can inquire of the LORD; but I hate him, because he never prophesies anything good for me but always misfortune. He is Micaiah son of Imlah’. Jehoshaphat replied, ‘Let not the king say so!’

Secondly, this syntactic construction continues into Mishnaic Hebrew, where the ordinary pattern is ʾên followed by an independent or enclitic pronoun and the participle (Geiger 2012:305). What is different about Mishnaic Hebrew is the fact that the pronoun may be an independent subject pronoun. This structure is not attested in Biblical Hebrew,¹¹ but it does occur at Qumran:

(6) 4Q372 f1:17–18

In (7), the same construction occurs, but now the subject pronoun is left dislocated and resumed in the main sentence with a second subject pronoun; this is not attested in Biblical Hebrew at all:

(7) 11Q 19 35:6 (= 11QT)

We can therefore see development in Qumran Hebrew from Biblical Hebrew, as a new construction is introduced, namely, ʾên with an independent subject pronoun, which becomes very common in Mishnaic Hebrew (Segal 1927/1970:162).

In the second construction ʾên is followed by an explicit noun phrase subject and then the participle, as in (8):

(8) 4Q277 f1ii:11

As in Biblical Hebrew, this construction involves sentential negation and the basic word order is ʾên, subject, participle.

In the first two constructions with ʾên, there is an overt subject, either a pronoun as in the first construction, or an explicit noun phrase as in the second construction. Syntactically, participles functioning as verbs need to have a subject, since, unlike finite verbal forms, the participle does not indicate a subject in its morphology.

In the third construction ʾên negates a participle with no overt subject, as in (9):

(9) 11Q19 LIX:8 (= 11QT)

The participle functions syntactically as the predicate of the clause in the sense that it governs an adjunct, the prepositional phrase מפני רעתמה (‘because of their wickedness’). Semantically, the subject of the predicative participle must be interpreted as ‘no one’. Syntactically, we understand the negative marker ʾên to be in construct with the zero subject of the participle (i.e. a zero noun phrase). Evidence for this analysis comes from examples in which there is a constituent which modifies the zero noun phrase, as in (10):

(10) 4Q405 f23i:10

In this example, the prepositional phrase במה (‘among them’) is dependent upon the zero noun phrase.

In this third construction with ʾên, the scope of negation is only the noun phrase subject and not the entire predication. In other words, the kind of negation is constituent negation rather than sentential negation. This construction is also found in Biblical Hebrew, as illustrated in (11):

(11) Psalm 105:37

He led them out with silver and gold and no one among his tribes was stumbling.

Again, the negative ʾên is followed by a prepositional phrase which modifies the zero subject noun phrase.

It is possible for a constituent other than the participle to be topicalised so that it appears before the negative particle:

(12) 1Q33 XIV:11 (= 1QM)

In this example, the prepositional phrases (לכול גבוריהם and לקליהם) are topicalised.

So far we have seen examples that are syntactically identical to those in Biblical Hebrew. In one example in Qumran Hebrew, this construction with ʾên followed by a zero subject and a participle is attested as the embedded object of a clause:

(13) 4Q381 f45:1

In this example, the clause ואין מבין (‘and no one understands’) is the object of the verb אשׁכיל (‘I shall teach’).

One of the most important uses of lōʾ is its use to negate a single constituent in a clause; this use is never exhibited by ʾên. Examples of lōʾ negating the predicates of a nominal clause occur twice in the following sentence:

(14) 4Q186 f2i:3–4

Note that the negative particle does not occur initially in the clause, but rather it immediately precedes the two predicates. The negative has scope only over the predicate constituents.

At Qumran the use of the negative lōʾ to negate a sentence with a participial predicate is increasing, especially when the participle is passive rather than active, as in (15):

(15) 4Q365a f5i:4

Note that the negative occurs in the initial position in the sentence; the scope of negation is thus the entire sentence. It is also important to note that, in contrast to sentences negated with ʾên, the word order is verb-subject rather than subject-verb.

An important way in which the Hebrew of Qumran has developed from Biblical Hebrew involves the absence of an explicit subject in participial sentences negated with lōʾ. This occurs in particular with passive participles, as illustrated in (16):

(16) CD V:3–4

In the history of the development of Hebrew, this is significant because pro-drop (the absence of an overt subject with a finite verb) is an important syntactic feature. However, it is not present in Biblical Hebrew on participles, since participles do not index subjects in their morphology, apart from gender and number. At Qumran, the process is not complete and often occurs in contexts where the subject can be inferred from the preceding context, as in (17):

(17) 11Q14 f1ii:11–12

‘and no one will miscarry (feminine) in your land and (she) will not be sick,¹³ blight and mildew will not be seen in its (lit. her) harvests’

In the first sentence, the negative ʾên is used to negate the zero subject noun phrase before a feminine participle. In the second sentence, the negative lōʾ is used before the participle, but no subject is expressed. However, from the context, the subject must be the same as that of the preceding sentence, namely a female animal. The import of the two sentences is: ‘No female animal will miscarry in your land, nor even be sick’.

A similar example occurs in (18), where a coordinate sentence is actually modifying a preceding noun phrase:

(18) CD IX:10–12

The clause with lōʾ could have been expressed as a relative clause modifying the preceding noun phrase, but rather than being subordinate, it is paratactic. This example illustrates that, whereas the participle in Qumran Hebrew is beginning to exhibit features that we would expect of a finite verb, it is doing so within highly constrained syntactic contexts and specifically in contexts of syntactic embedding or subordination.

The development of Qumran Hebrew from Biblical Hebrew in this regard can be seen as well in the text of Isaiah 44:12.

(19a) Isaiah 44:12 (MT)

‘(If) he drinks no water, he would grow faint’.

In the MT version, the verb שׁתָָה is a perfect form used in a modal sense. In the Qumran version of Isaiah from Cave 1, the participial form שותה is used. Again, it seems as if the participle at Qumran can be used with pro-drop; in other words, it is beginning to have the syntactic characteristics of a finite verb.

In the same way that lōʾ can negate a constituent within the clause; it can also negate a constituent within a prepositional phrase, as in (20):

(20) 1QH^a XVI:10–11¹⁴

In the two examples in this passage, a prepositional phrase takes as its object a passive participle negated with lōʾ. The following passage is similar, again with a passive participle:

(21) 1QH^a XVI:36

The passive participle is feminine and the implicit subject must be the noun ‘tongue’ from the previous sentence.

In (22), the passive participle is negated within a prepositional phrase, but there is an unmarked headless relative clause between the prepositional and the negated clause.

(22) 4Q418 f69 ii:5

We must understand that the prepositional phrase means ‘to the one who has not come into being’ even though the head of the relative clause is not overt and the relative marker is also not overt.