What_Do_We_Actually_Know_about_Ancient_H.pdf

What Do We Actually Know aboutAncient Hebrew

Ian Young

1. The Three Stage Theory of the History of AncientHebrew

Using the sources available to them, by the end of the nineteenthcentury scholars had been able to construct an elegant theory of thehistory of Hebrew in the ancient period. The main evidence thatscholars used was the Hebrew Bible. More specifically, for them, asis still the case with language scholars today, “the Bible” meant thetraditional Masoretic Text of the books of the Tanach, the one readilyavailable to everyone who has a Hebrew Bible. The Bible was,however, not the only evidence that scholars in the nineteenthcentury had available to them for ancient Hebrew. They were alsoaware that early rabbinic works like the Mishnah used a noticeablydifferent sort of Hebrew, which is often called “Mishnaic Hebrew.”Third, these early scholars were aware that from the late biblicalperiod on, Hebrew competed in the Jewish homeland with theAramaic language. Various forms of evidence in rabbinic sources orin other sources like the New Testament Gospels (see, for exampleJesusʼ quoted use of Aramaic in the Greek text of Mark 5:41 andelsewhere) indicated that Hebrewʼs battle against Aramaic in theJewish homeland was often a losing one.

According to this scholarly theory, there were three main periodsin the history of ancient Hebrew. First, before the exile to Babylon inthe sixth century BCE was the “Golden Age” of Hebrew literature. Earlyworks of literature such as the story of King David in the Book ofSamuel, or the early parts of the Mosaic Pentateuch were written inpure Classical Hebrew. After the Babylonian exile, however, accordingto this theory, Hebrew lost ground to Aramaic. Factors such as the risein non-native speakers of Hebrew, and the breaking of the continuityof the pre-exilic educational system meant that Hebrew declined.

Australian Journal of Jewish Studies (2013) 27: 11 – 31

Evidence for this was found in the Bible. The books of Esther, Daniel,Ezra, Nehemiah and Chronicles, by their contents, are among thebooks that must have been composed after the return from exile. Ithappens that each of these books shares a peculiar style of Hebrew,which sets them clearly apart from works of the Golden Age, such asSamuel. The main characteristic of this style is actually simply agreater openness to linguistic variety than is typical of the languageof a book like Samuel.

However, since some of the additional linguistic features whichthese books like to use are found in Aramaic or “Mishnaic Hebrew”,and since some passages in these books seem rather difficult tocomprehend (try Daniel chapter eight, for example), these books weretaken as a symptom that Hebrew in the post-exilic period was in astate of decline. Scholars therefore called this second period the“Silver Age” of Hebrew literature. The idea of decline was also usedto explain the third era of the history of ancient Hebrew. Under theinfluence of Aramaic and other factors, Hebrew continued to declinein the post-exilic period, according to this theory, until by the CE periodthe Mishnaic Hebrew of the Rabbis represents a thoroughlyunclassical form of Hebrew, perhaps an Aramaised form of Hebrew.

The three-stage linear model of the development of ancientHebrew was a reasonable and logical deduction from the evidence asit was available and understood at that time. The three-stage model,of a pre-exilic Golden Age down to about the sixth century BCE, apost-exilic Silver Age, and a post-biblical, post-classical age, wasfurthermore of great use to scholars when discussing biblical books.Since Hebrew was understood to have developed in a linear fashion,away from early, pure, classical Hebrew under Aramaic influencetowards the form of language found in the Mishnah, it was possible toargue where on that linear development works of uncertain date wereto be put. Note the following paragraph on the language of Qoheleth(or Ecclesiastes) from the authoritative Encyclopaedia Biblica,published in 1899, which is a good illustration of the results ofnineteenth century scholarship:

[T]he language...[of]...Ecclesiastes suggest[s] that it is one of thelatest books in the canon. The language has the peculiarities ofsuch late books as Chronicles-Ezra-Nehemiah, and Esther.Indeed it belongs to a much more degraded stage of Hebrew thaneither of these books1 exhibits; and in the forms of words, in thenew senses which older words are used, and in the many new

12

What Do We Actually Know about Ancient Hebrew

words employed, it has many similarities to the Targums andSyriac [i.e. Aramaic sources, IY2], [and] especially to the Mishna(circa 200 A.D.) (Cheyne 1899:1161).3

This quote illustrates how books such as Esther which are definitely“late,” that is, post-exilic, are considered to exhibit linguistic“peculiarities”; how the language of these silver-age books is“degraded” (note the pejorative term indicating degeneration from ahigher state); how books even later than these “late” books exhibit aneven more “degraded” form of Hebrew; how the late books areconsidered to be under Aramaic influence; and how the terminus of(utter?) degradation is considered to be the rabbinic Hebrew of theMishnah.

Thus, the logic of the theory meant that, even though traditionallyQoheleth/Ecclesiastes has been associated with King Solomon, andhence a very early date, even a very conservative Christian scholarlike Franz Delitzsch had to concede in 1877 that such a date wasimpossible, with the famous statement: “If the Book of Koheleth wereof old Solomonic origin, then there is no history of the Hebrewlanguage” (Delitzsch 1877:190).

2. InscriptionsFrom the latter part of the nineteenth century, archaeologicaldiscoveries meant that scholars of the Hebrew language began tohave access to an increasing amount of new evidence for ancientHebrew. On the one hand, from 1880 significant inscriptions datingfrom the pre-exilic, monarchic era were uncovered. Then, around1950, the Dead Sea Scrolls were discovered. It is very common inscholarship of all fields that, when new discoveries are made, scholarsnaturally view them in line with the prevailing scholarly theories. Itusually takes a while before new knowledge leads to radically newreconfigurations of all the evidence. This is no different in the case ofthe Hebrew language. It is still the case these days that many scholarscontinue to try to explain all the evidence in terms of the linear, three-stage history of Hebrew that we have just mentioned.

However, other scholars, myself among them, argue that theevidence that has been uncovered sheds a completely new light onthe history of Hebrew. Most importantly, as we shall discuss below, itmakes us return to the sources we have had all along, in particularthe Masoretic Text of the Bible, and realise that we have completelymisunderstood the nature of that evidence. This article is an attempt

13

Ian Young

to sketch some of the main issues in the debate about the three-stagetheory of the development of ancient Hebrew over the last decade ofscholarship.

The first Hebrew inscription of any length that was discovered wasthe Siloam Tunnel inscription, found in 1880 inside a water tunnel inJerusalem (Dobbs-Allsopp et al. 2005:499–506). It is a rather amazingAncient Near Eastern inscription since it focuses exclusively on theachievement of the stone masons in cutting the tunnel deepunderground from both ends and managing to successfully meet inthe middle with only a tiny dog-leg at the end. This was a greatengineering achievement, but what is amazing about the inscriptionis that that is all it says.

…the boring through. And this was the manner of the boring through.While [the hewers were wielding] the pick-axe each toward his fellowand while three cubits [remained yet] to be bored [through, there washeard] the voice of a man calling to his fellow – for there was a fissurein the rock on the right hand. And when the tunnel was driventhrough, the hewers struck, each man toward his fellow, pick-axeagainst pick-axe. And the water flowed from the spring toward thereservoir for twelve hundred cubits. And a hundred cubits was theheight of the rock above the head of the hewers.

The inscription mentions neither the king (often thought to be KingHezekiah, around 700 BCE), nor does it thank God. A normal ancientNear Eastern monumental inscription would start something like: “I amHezekiah, king of Judah. The LORD chose me to be king and I cutthis tunnel.”4

In regard to the language of the Siloam Tunnel inscription, it isinteresting in hindsight to look at the way scholars have, almostwithout thinking, assimilated it into the three-stage theory of thedevelopment of Hebrew. It is testimony to the way that a theory willshape the way we see the evidence. As I mentioned, the SiloamTunnel inscription is usually dated to the time of King Hezekiah,around 700 BCE, which would also be the time of prophets like Isaiahand Micah. That date puts it firmly in the first, Golden Age of ancientHebrew literature. Thus it has usually been understood as being in thesame sort of Hebrew as those biblical texts from the Golden Age.

In fact, the argument has been reversed. There was in the 1990sa vigorous discussion about whether some of the biblical compositions

14

What Do We Actually Know about Ancient Hebrew

which had previously been thought to be early, and were in theStandard, or Classical Biblical Hebrew of the Golden Era, couldactually have been composed later (see, for example, Davies 1995).No, said some prominent language scholars, that is impossible. Textswritten in Standard Biblical Hebrew could not have been written, say,after the exile, because we know that Standard Biblical Hebrew isearly, and later developed into the later Hebrew of the Silver Age. Buthow do we know that Standard Biblical Hebrew was (only) writtenearly? These scholars argued that we knew this because the languageof the pre-exilic inscriptions—and the Siloam Tunnel inscription wasone of the first mentioned—is identical to the language of the earlybooks, while post-exilic Hebrew is paralleled by the language of latesources like the Dead Sea Scrolls (e.g. Hurvitz 1997:307–311). A textlike the Siloam Tunnel inscription, it was explicitly said, was by andlarge in a form of language identical to that of the early, classicalbiblical books, and had no features characteristic of the later Hebrewof the post-exilic Silver Age.

In a couple of studies published in the last 10 years, I investigatedthe evidence of the Hebrew inscriptions to see whether they backedup the assertions scholars were making (Young 2003; Young, Rezetkoand Ehrensvärd 2008 [henceforth: LDBT], 1:143–200). I came to avery different conclusion. What I found was that it did indeed happenthat on occasion the inscriptions had linguistic forms rare in late, SilverAge Hebrew but more common in Golden Age Hebrew. In the SiloamTunnel, for example, the use of rwxb for “while, still” is not found in thetexts taken as typical of the Silver Age, that is Esther, Daniel, Ezra,Nehemiah and Chronicles. But also, surprisingly often, there wereother linguistic forms in the inscriptions which scholars had concludedon the basis of the biblical evidence were characteristic of the lateHebrew of the Silver Age. I will return to them shortly, because what Iwant to emphasise now is that by far the largest group of significantlinguistic forms in the pre-exilic Hebrew inscriptions is not those fewcharacteristic of Golden Age Hebrew, nor those few characteristic ofSilver Age Hebrew, but rather the category of forms that are eitherunattested in any biblical text, or if attested, are extremely rare.

The most significant finding of my studies of the inscriptions wasthat, although attesting the same general type of Hebrew as thebiblical texts, that is, a general literary type of Hebrew, the inscriptionsexhibit a large number of linguistic forms which were non-biblical. First,and this was something that scholars had long realised, the spellingof the inscriptions was different to any known biblical text. Thus, for

15

Ian Young

example, there is much less use of waw and yod to mark vowels inthe inscriptions (e.g. abn “prophet,” Bible aybn). Furthermore, the thirdperson suffixes are a completely different system to all biblical texts.Therefore, it has long been acknowledged that all biblical texts in ourpossession have a system of spelling that is of a later type to theinscriptions, which cover the period down to the exile, around 600BCE. All biblical texts are thus written in a form that is no earlier thanthe post-exilic period. This does not mean that no biblical text wascomposed earlier than the exile, but it seems to mean that earlycompositions have had (at least) their spelling updated.

There is, however, much more than spelling involved in thedifference between the language of the pre-exilic inscriptions and allbiblical texts. In the Siloam Tunnel, I count 48 preserved words or“graphic units”. Of them, 12 are linguistic forms (not counting spellingvariants) that differ from what we would normally find in the Hebrewof the Bible. In other words, 25%, or a quarter of the Hebrew words inthe Siloam Tunnel inscription, are linguistic forms not paralleled in theBible. For example, over 100 times in the Bible we have theexpression wh[r la çya “a man to his neighbour.” We have the sameexpression in the Siloam Tunnel inscription, but there the suffixed formof “his neighbour” is Wh[erE but apparently w[;rE possibly even meaning“his neighbours” (plural). So, our only texts actually from the pre-exilic,monarchic era, are in a form of Classical Hebrew that is independentof any of the sorts of Hebrew found in the Bible, even if it has someimportant links with them.

The next two major inscriptions found after the Siloam Tunnelinscription fit in even less well with what had previously been known aboutHebrew. In 1908 a small limestone tablet was discovered in theexcavations of Gezer, and is called the Gezer Calendar since it talks ofthe agricultural activities of the year (Dobbs-Allsopp et al. 2005:155–165):

Months of harvest, months of sowing, months of spring pasture,month of flax pulling, month of barley harvest, month of (wheat)harvest and measuring, months of pruning, month of summer fruit.

The question arises what this tablet is for, since in traditionalagricultural societies farmers usually cannot read, and in any case,they hardly need to read a text to remind them when to plant andharvest things. If they do, they are in real trouble! One plausibleexplanation, based on other similar tablets, is that it is intended as ablessing tablet, a sort of talisman, to ensure that the agricultural

What Do We Actually Know about Ancient Hebrew

16

seasons run their course and are blessed with fertility (Wirgin 1960;see also Young 1992:367). This would explain why it was found nearthe city granary. In regard to its language, it exhibits a number ofarchaic and dialect forms unattested in the Bible, or it turns out, evenunattested in other inscriptions as well. If it is Hebrew, it is our oldestHebrew inscription with a connected sense, dating to the tenth centuryBCE, but scholars debate whether, for example, Gezer was anIsraelite site at that time (see LDBT 1:187), and so we either have aweird sort of Hebrew, or a weird sort of not-Hebrew!

The next inscriptions that were found were definitely from Israeliteterritory. Excavations at the site of ancient Samaria in 1910 discovered63 inscribed ostraca, potsherds written on in ink, and more were foundlater (Dobbs-Allsopp et al. 2005:423–497). Samaria was the capital ofthe northern kingdom, Israel, that broke away from the southernkingdom, Judah, with its capital at Jerusalem, after the time of Davidand Solomon according to the biblical story (1 Kings 12). The textsthemselves were probably about as boring as possible in their day.They seem to be brief dockets recording deliveries to the royal palacein Samaria. They would probably have had their informationtransferred to a general ledger, and then have been thrown out.Ironically, the throw-away ostraca survived, whereas the importantgeneral ledger, probably written on papyrus, perished. Now scholarspore over these temporary records trying to fill in the gaps. We couldcompare this to finding someoneʼs list thrown in the garbage and tryingto work out what it was used for. All the texts are along the lines of thisone, ostracon 6 with questions that scholars debate added below insquare brackets:

In the ninth year [of some king—but who? We think nowadays thatthe texts date to around 780 BCE], from Qoṣeh [a town in theregion of Samaria], to Gaddiyaw [sent to? Is he a tax officer? Orsomeone in the palace getting materials from his estates? Or doesit mean “for” his personal use?]. A jar of old wine.

There are only about eight different words used in the texts originallydiscovered, plus prepositions, place names, and personal names.However, several of the words are different from Biblical Hebrew, orindeed different from what we get in inscriptions from the southernkingdom. This confirms what we would already expect, that ancientHebrew had regional dialects. Thus in the northern Hebrew of theSamaria Ostraca, the word for “year” is tç, whereas in the Bible it is

Ian Young

17

hnç. We also saw the contracted divine element in the name Gaddiyaw,as opposed to what is found in southern inscriptions and the Bible,Gaddiyahu. Most scholars have thought that there are compositionsin the Bible that either came from the northern kingdom (like the bookof the northern prophet Hosea) or were based on sources, some ofwhich came from the north, such as the stories of northern judges,prophets and kings in the Books of Judges, Samuel and Kings. Thequestion, which we shall leave for the moment, is: If there are northerncompositions in the Bible, why do the specific, distinct features of ournorthern inscriptions never turn up in any biblical text?

After the weirdness of the Gezer Calendar—and the SamariaOstraca, and to be honest, the Siloam Tunnel—things settled down (abit) for Hebrew language scholars with some (relatively) substantialdiscoveries during the course of the twentieth century. The two mostsubstantial are the Lachish Ostraca, discovered in the 1930s (Dobbs-Allsopp et al. 2005:299–347), and the Arad Ostraca, found in the1960s (Dobbs-Allsopp et al. 2005:5–108). These are both relativelysubstantial collections of ostraca emanating from military bases atLachish and Arad. Mostly, the texts come from the last years of thekingdom of Judah before the Babylonian exile, around 600 BCE. Alarge proportion of the texts are letters, or the drafts of letters, whichperhaps would have been transferred onto papyrus in the final draft.Perhaps since they come from late in the history of the southernkingdom Judah, they are not as weird linguistically as texts like theSamaria Ostraca, but they do exhibit a good selection of linguisticforms unparalleled in biblical texts.

In addition, as I mentioned earlier, considering the expectationscreated by the three-stage history of Hebrew, the inscriptions, textsactually from stage one, the Golden Age, have a surprising number oflinguistic forms considered characteristic of stage two, the post-exilicSilver Age. For example, among the various biblical words for “togather,” snk is quite rare and generally its occurrences are restrictedto those five books which are considered to typify the post-exilic SilverAge of Hebrew: Esther, Daniel, Ezra, Nehemiah and Chronicles.Despite this, a recently published inscription from Jerusalem dating tothe early period uses that supposedly exclusively “late” word (Young2003:193).

The word snk is in fact an unusual example of a “late,” Silver Agelinguistic form, since it is never found even once in one of the maintexts that are considered typical of the early, Golden Age Hebrewliterature, such as the Pentateuch or Former Prophets. One amazing

What Do We Actually Know about Ancient Hebrew

18

fact that my co-researchers and I discovered was that there are almostno well-attested features supposedly characteristic of late, Silver AgeHebrew which are not also found in texts considered to representearly, Golden Age Hebrew (LDBT 1:83–87, 111–119). This isconnected to another remarkable fact, namely that no biblical text ofany type or supposed era, is without supposedly “late” Hebrewfeatures (LDBT 1:132–136). So, we have the situation that in fact,contrary to what might have been expected from the three-stagemodel of Hebrew, all biblical texts, whether considered “early” or “late,”use the same linguistic forms. This phenomenon raises someinteresting questions, such as: If the supposedly “late” linguisticfeatures are (almost) all found in supposedly “early” texts, in whatsense are they late? Another amazing discovery is that the corpus ofpre-exilic inscriptions from Arad, when I investigated them, had ahigher proportion of supposedly “late,” post-exilic, Silver Age linguisticforms than most biblical books whether usually dated to the pre-exilicor post-exilic period (LDBT 1:129–139; 163–168). This includes textsof uncertain date which language scholars had considered must bepost-exilic because they had, in their opinion, such a significantproportion of “late” linguistic forms in them.

To summarise where we are so far in our discussion: The three-stage model of ancient Hebrew has difficulty coping with the evidenceof the Hebrew inscriptions, texts contemporary with the period whenbiblical compositions of the Golden Age are generally considered tohave been written. The inscriptions evidence rather a high number oflinguistic features supposedly characteristic of a later period. They aredistinguished from all biblical texts in quite a significant number oflinguistic features. Even though scholars have suggested that thereare northern texts among the biblical texts of the Golden Era, there isno trace of any of the distinctive features of the northern inscriptionsin any biblical text.

However, there is still hope for the three-stage theory whereHebrew degenerates from the Golden, pre-exilic Age to the Silver,post-exilic Age, to the post-biblical Mishnaic Age. This is because it isstill the case that our five core “late” texts, Esther, Daniel, Ezra,Nehemiah, and Chronicles, do definitely have much higherconcentrations of these late features than any Hebrew inscriptions orother biblical texts. If they diffuse at all, it is normal for new linguisticforms to steadily increase in prominence in a language. Even if theinscriptions demonstrate that the late linguistic forms were a lot earlier,and a lot more frequent earlier than was thought, we could still trace

Ian Young

19

a growth in their use from the pre-exilic period to those five books ofthe post-exilic era, and on to the ultimate degeneration in MishnaicHebrew. The fact that we cannot accept this solution is due to the othernew source of linguistic evidence, from the end of the biblical period,the Dead Sea Scrolls.

3. The Dead Sea ScrollsThe Dead Sea Scrolls were found in caves at Qumran in the Judeanwilderness in the late 1940s and in the 1950s. It took a long time, butsince the 1990s, all the scrolls have been available to everyone. Withthe scrolls, as is typical of new discoveries, scholars initially attemptedto place what was found within the existing models in scholarship.Now, 60 years later, it can be definitely stated that the Scrolls haveutterly revolutionised our understanding of the Hebrew Bible, ancientJudaism, and the Hebrew language.5

The Qumran scrolls are fragments—usually very, veryfragmentary—of over 900 scrolls, all or almost all of them literary texts.This means that Qumran is a very unusual find, both for the numberof manuscripts discovered, and for the predominance of literary—meaning religious—texts in it. In regard to date, the earliestmanuscripts are considered to date to the third century BCE, althoughof course the compositions represented in the manuscripts could havebeen written much earlier than that. Our earliest physical attestationof biblical texts—regardless of when they were written—is thesemanuscripts from the third century BCE and following centuries.

According to most scholars, the scrolls were put in the cavesduring the revolt against Rome around 68 CE. A small number ofscholars, myself included, think this is wrong, and that the scrollswere deposited in the first century BCE not CE (Young 2002; 2005;2013, building on Doudna 2001; 2004; 2006; Hutchesson 1999). Onestrong indication for this dating is the fact that while there is aclustering of historical references and allusions to people and eventsfrom the first half of the first century BCE, there are no references toanything or anyone after a certain point, around 40 BCE. The mostobvious explanation of that striking fact is that the scrolls were putin the caves somewhere around 40 BCE. Another strong piece ofevidence for the early dating is the utter contrast between the sort ofbiblical texts at Qumran (which I will discuss shortly) and all otherHebrew biblical texts deposited in the first or early second centuriesCE, the supposed date of the Qumran scrolls, such as the scrollsfound at Masada.

What Do We Actually Know about Ancient Hebrew

20

Even more disputed is the question of who owned these scrolls. Itis still common these days to find scholars who agree with the earlierdominant theory that the scrolls belonged to a Jewish group called theEssenes, and that the Essenes had an important base at Qumran.However, even those committed to this theory have abandoned anumber of ideas that were fashionable among older scholars. Olderscholars sometimes gave the impression that the scrolls were alldocuments produced by a sectarian community of Essene monks atQumran. Nowadays it is accepted that the majority (at least) of thescrolls were not actually produced at Qumran, but were brought therefrom outside. Also, it is widely believed these days that even if Qumranwas an Essene settlement, it was just one of a number of suchsettlements. Other scholars, of course, such as the late Alan Crown,deny the whole theory that the scrolls were connected with theEssenes, or produced in relation to the Qumran settlement at all(Crown 2005). In the face of this scholarly turmoil, I was led to remark:“The only really certain thing about the Qumran texts is that somepeople put some scrolls into some caves” (LDBT 1:250).

The 900 manuscripts found at Qumran can be conveniently sortedinto three groups. First, there are copies of books found in the HebrewBible. These account for over 200 of the manuscripts. Second, thereare copies of books we knew previously, but which are not in theHebrew Bible, about 100 of them. Thus, for example, there are severalcopies of the Book of Tobit, found in the larger traditional Christian OldTestament. Third, there are copies of compositions that we did notknow before they were found at Qumran. These include rules ofcommunities, biblically-inspired rewrites, calendrical texts, poetic texts,wisdom literature, commentaries on prophetic books, and so on.

In regard to language, Qumran was a complete surprise for thethree-stage theory of the development of Hebrew. The overwhelmingnumber of the 900 manuscripts from Qumran are in Hebrew. Thereare only about 120 manuscripts in Aramaic and about 20 in Greek; therest are in Hebrew. That already is an interesting observation in lightof the three-stage theory of ancient Hebrew. We might have expectedAramaic to be in a position of dominance.

The next question is: What sort of Hebrew are the Dead Sea Scrollsin? Let us remind ourselves what the three-stage theory would predictas the answer to that question. We think the large majority of previouslyunknown compositions from Qumran were actually composed in thelast couple of centuries BCE. Thus, for example, we are sure that thePesher commentary on Habakkuk is from that period, since it refers to

Ian Young

21

the Roman intervention in Jewish affairs in the first century BCE. Thethree-stage theory of Hebrew, we recall, saw Hebrew degenerating froma Golden Age before the Babylonian exile in the sixth century BCE, toa Silver Age in the post-exilic period, exemplified by the Hebrew ofEsther, Daniel, Ezra, Nehemiah and Chronicles, getting progressivelymore degenerate until you get monstrosities like the books of Qohelethand Song of Songs, which are just a step away from the utterdegeneration that is Mishnaic Hebrew. So if a text like the Pesher onHabakkuk dates to the first century BCE, it would evidently be expectedto be the linguistic equivalent of a two-headed monster. It should beeven more full of late Hebrew forms than Esther, Daniel, Ezra,Nehemiah, Chronicles, which are all earlier in date; should be more“degenerated” than Qoheleth or Song of Songs, since it is later thanthem, and hence closer to “degenerating” into Mishnaic Hebrew; andshould be in a non-classical form of Hebrew, full of Aramaic, MishnaicHebrew and other late features.

The first generations of scholars did, in fact, try to fit the evidenceof Qumran into the old three-stage model. They were of courseconstrained in that it is blindingly obvious that regular Qumran Hebrewis nothing like the language of books like Qoheleth and Song of Songs:the Qumran texts are in a very classical form of Hebrew.6

Nevertheless, the most common way that Qumran Hebrew isdescribed was as a development of Late Biblical Hebrew, meaning adevelopment of the language of those books considered typical ofSilver Age Hebrew (Esther, Daniel, Ezra, Nehemiah, Chronicles). Thiscontinuity was demonstrated by citing odd forms here and there whichwere typical of those five books. However, we recall that we found outearlier that all biblical texts, even those representing the Golden Age,have a few of these supposedly late forms. All biblical texts, whetherclassified as early or late, use the same linguistic forms. What setsthe core “late” books of Esther, Daniel, Ezra, Nehemiah andChronicles apart is their greatly increased proportion and preferencefor these forms. It is not enough, therefore, to point out that Qumrantexts have occasional supposedly “late” linguistic forms. If there reallyis a development, if Qumran Hebrew really is a later development ofthe five late booksʼ language, it should show the same or more likelya larger proportion of the late forms than the five late books.

I have investigated a number of Qumran texts, including a detailedstudy of the Pesher Commentary on Habakkuk (Young 2008; LDBT1:250–279). What I found was that, for example, the Peshercommentary on Habakkuk had no greater proportion of “late” linguistic

What Do We Actually Know about Ancient Hebrew

22

forms than a typical composition from the early, pre-exilic Golden Ageof biblical literature. The author(s) of Pesher Habakkuk clearly had amastery of Classical Hebrew style that was supposed to have died out600 years before his time. It is clear that the author was not struggling,and for example, relying on the biblical text of Habakkuk for hisclassical linguistic forms. I found places where in his commentary theauthor of Pesher Habakkuk decided not to use the perfectly goodClassical Hebrew form in the Habakkuk passage he was commentingon, but instead chose to use a sometimes even more typically GoldenAge synonym to express what he wanted to say (Young 2008:34–35).7

We recall that the corpus of pre-exilic inscriptions from Arad that Iinvestigated caused problems for the theory of Hebrewʼs degenerationby having too many supposedly late forms too early. A compositionlike Pesher Habakkuk causes problems by having far too few “late”forms when it should be full of them and its Hebrew should be in anadvanced stage of degeneration. Yet another amazing observation isthat the corpus of pre-exilic inscriptions from Arad (dated around 600BCE) has more supposedly post-exilic late Hebrew linguistic featuresthan Pesher Habakkuk, a work from the first century BCE. So muchfor a progressive degeneration of Hebrew! No Qumran work that Ihave looked at has the concentration of “late” linguistic features foundin the supposedly typical late books of Esther, Daniel, Ezra, Nehemiahand Chronicles.

These observations make us come back and ask: What made usthink that these five books were typical, normal, or indeed representedthe only sort of Hebrew in the post-exilic period? These books share adistinctive style in that they all have a stylistic preference for certainlinguistic forms which are generally, but not totally, avoided by otherHebrew texts. In other words they are characterised stylistically by agreater openness to linguistic variety than other Hebrew literary texts.The fact is that no other Hebrew literary text shares the style of thesefive books. There are in fact a number of other biblical compositionswhich scholars date to the post-exilic period, for example prophets likeHaggai, Zechariah or Joel. None of them are in the same style as thefive books, not even close, in fact. In the second century BCE a Wisdomwriter in Jerusalem whom we refer to as Ben Sira wrote a massive workof Wisdom, preserved in Greek in the traditional Christian Old Testament(where it is often referred to as Ecclesiasticus or Sirach). Fragmentswere found at Qumran, but a much more substantial section of theHebrew text was found at Masada. Ben Sira is also very short on thesesupposedly characteristic “late” Hebrew elements.

Ian Young

23

It turns out, therefore, that Esther, Daniel, Ezra, Nehemiah andChronicles, far from representing the key examples of post-exilicHebrew, are in fact isolated in their linguistic style. They were notwritten in the normal style of literary Hebrew in the post-exilic period,which remained throughout a form of Classical Hebrew much closerto the Hebrew of the Golden Age than them. Far from representingthe inevitable decline and degeneration of Hebrew in the late period,the authors of these books must have chosen to write in that distinctivestyle.8 Work is only now just beginning to figure out the reasons forthat stylistic choice, such as geographical or ideological distance fromthe other works (for some preliminary thoughts see Young 2009). TheDead Sea Scrolls tell us, therefore, what we could already haveinferred from the biblical texts themselves, that there was nodegeneration in literary Hebrew in the post-exilic period, but that formsof Hebrew with a recognisable similarity to Hebrew of the Golden Agecontinued until the end of the Second Temple period.4. Mishnaic Hebrew We now turn to the final stage in the three-stage theory of Hebrew: IsMishnaic Hebrew a degeneration of Biblical Hebrew? Scholars havein fact moved well away from that position. The current consensus isthat the roots of Mishnaic Hebrew are in a colloquial dialect from thebiblical period. Indeed, it is the consensus that Mishnaic Hebrew isnot descended from Biblical Hebrew at all, but from a related dialect(or dialects) whose roots can be traced back to the pre-exilic period(on Mishnaic Hebrew see LDBT 1:223–249). Thus Mishnaic Hebrewis not a degeneration of Biblical Hebrew, it is simply a different dialectthat co-existed with it throughout the biblical period. It did not achieveliterary status until the rabbinic period, but it did have a sporadicimpact on earlier literary texts. What this means is that just because atext has sporadic Mishnaic Hebrew features, or even those texts thatreflect a Mishnaic-like dialect, such as Qoheleth or Song of Songs,are not thereby placed in a late period. Mishnaic Hebrew or itsancestor was always around, and authors could make a stylisticchoice in any period to use Mishnaic Hebrew forms in literary Hebrew.5. Linguistic FluidityThe three-stage model of ancient Hebrewʼs degeneration is at oddswith the extra-biblical evidence for ancient Hebrew. More than this,however, the Dead Sea Scrolls have provided the key for a totalrethink of the nature of the Hebrew of the Bible. It seems blindinglyobvious that if a biblical book was written early, it is an example of

What Do We Actually Know about Ancient Hebrew

24

early language. If a book was written late, it reflects the language ofthe late period in which it was written. If the prophet Micah wrotearound 700 BCE, then the language of the book of Micah is theHebrew of 700 BCE. These blindingly obvious statements arepresupposed by almost all scholars of the Hebrew language. Moreparticularly the Masoretic Text (MT) of the Hebrew Bible is taken asreflecting in detail the language that left the pens of the original authorsof the biblical texts. But these blindingly obvious presuppositions ofthe language scholars are diametrically opposed to the currentconsensus of most mainstream Bible scholars as to the compositionhistory of the Hebrew Bible.

Scholars note that when we place the Qumran biblical scrolls, theOld Greek translations, and the Samaritan Pentateuch alongside theMT, we are faced with a rather startling variety of biblical texts. Despitethe very fragmentary nature of our evidence for the text of the HebrewBible, we nevertheless have in our possession radically different textsof most books of the Hebrew Bible. A classic example is the shorterand longer editions of the book of Jeremiah (Tov 2012:286–294). Theshorter version, attested in the Greek Septuagint and the Qumranscrolls is a sixth (17%) shorter than the longer (MT) edition, which isalso attested among the Qumran scrolls. This means that over 3,500words of the MT Jeremiah are not represented in the shorter version.Differences involve the presence or absence of some whole sections,but most commonly there is just simply more material in the parallelsections of the MT. In addition, common material appears in differentplaces in the book, most notably the oracles against the foreignnations which are at the end (Chapters 46-51) of the MT, but are inthe middle of the other version (Chapter 25).

Language scholars assume that the MT provides detailedevidence of the linguistic forms used by the original authors of biblicalcompositions. In his authoritative handbook on the textual criticism ofthe Hebrew Bible, Emanuel Tov says, in contrast:

[T]he textual evidence does not point to a single “original” text, buta series of subsequent authoritative texts produced by the sameor different authors ... the original texts(s) remain(s) an evasiveentity that cannot be reconstructed ... Some biblical books, suchas Jeremiah, reached a final state more than once ... the originaltext is far removed and can never be reconstructed ... the JudeanDesert scrolls [our earliest biblical manuscripts] reflect a relativelylate stage of the textual development (Tov 2012:167–169).

25

Ian Young

If biblical books had been composed like modern books, at one time,and thereafter remained basically the same, it is obvious that we mightexpect to detect a chronology in the way language is used in variousbooks. But in light of the evidence that texts were written and rewrittenover centuries, the idea that there is a “date” when a biblical book waswritten is anachronistic. So, too, since every biblical text containswithin it a chronology of earlier and later composition, the idea thatbiblical books or chunks of books represent the language of oneparticular time and place appears to be extremely unlikely. We in factfind that language is one of the most peripheral and hence mostchangeable aspects of the biblical texts.9

Numerous examples of systematic linguistic differences can befound in the Dead Sea Scrolls biblical texts, or in the SamaritanPentateuch (LDBT 1:341–360; documented in detail in Rezetko andYoung, forthcoming). As just one example, the preposition ˜mi “from”normally stands separate before the definite article, hence:–hæ ˜m i.However, the MT book of Samuel stands out by around 50% of thetime preferring the very rare assimilated form – hæme for this case studysee also Young, forthcoming). Is that a peculiarity of the original authorof Samuel or just a scribe in the tradition of the MT? That we do notknow, and the changeable nature of linguistic evidence in textualtransmission, is underlined by the fact that the Qumran Samuelmanuscript 4QSama always has the standard form – h ˜m, even incases where the MT parallel has the unusual – hm.

Rather than dwell at further length on the massive amount ofevidence for linguistic fluidity in non-MT manuscripts, I would like tomove on and suggest that we already, in the MT itself, had all theevidence we needed on the question of how fluidly the linguistic formswere copied. In the MT we have several parallel passages, in otherwords cases where we consider we have the same composition in twodifferent places. It is interesting to look at how “less common” linguisticforms, the sort of ones that are considered significant evidence oflinguistic peculiarities of the “authors” of biblical texts, are treated inthese parallel texts. What I mean by “less common” linguistic formsare those forms that are synonymous with another form, and whichare the less commonly attested one of the synonyms in the MT Bibleas a whole. These forms are not necessarily rare, and may be attestedfrequently in the MT Bible. One example is the use of ta plus suffix tomark the object of a verb (e.g. wta wl[yw “and they brought him up”: 2Kings 25:6 // Jeremiah 52:9), which is nevertheless less common thanattaching the object suffix directly to the verb (whl[yw “and they brought

26

What Do We Actually Know about Ancient Hebrew

him up”: Jeremiah 39:5). Yet even taking such frequent forms intoaccount, the results are striking (for the full study see Rezetko andYoung, forthcoming).

One of the texts I have investigated is the poem in 2 Samuel 22and its parallel in Psalm 18. What I found is that the poetic section ofthese parallel texts contains 30 of these less common linguistic forms,but that only two (or about 7%) are found in both texts. In other words,about fourteen out of every fifteen of the less common linguisticforms—the sort of forms that are taken as decisive evidence of thelinguistic peculiarities of the original author—have either been addedor removed during the scribal transmission of the composition in itstwo versions. A similar picture, if not quite so dramatic, emerges fromthe other case studies I have done. For example, there is 2 Kings 24–25 where it parallels Jeremiah 52. It is commonly thought that thepassage was taken from Kings at a fairly late stage to form a sort ofhistorical appendix to Jeremiah, so they are very closely related texts.Here my study identified 37 less common linguistic forms, and onlyten are shared between both texts, that is, about 27%. In other words,even in these closely related texts, nearly three out of every four ofthe less common linguistic forms have been added or removed inscribal transmission. When we introduce a further parallel to part ofthe passage, in Jeremiah 39, this proportion of shared forms drops toabout 20%, or just one in every five. The more texts we have, the lessof these linguistic forms they share in common, so that if we had morethan the very meagre textual evidence for the Hebrew Bible that wehave now, it is quite likely that the proportion of shared less commonlinguistic forms could approach zero, as it does with 2 Samuel 22 andPsalm 18. The MT itself, therefore, provides all the evidence we need(never mind the non-MT manuscripts) that the language of the biblicaltexts was transmitted very fluidly.

Every book of the Hebrew Bible, in whatever manuscript we haveit, is therefore a linguistically composite text reflecting language fromdifferent layers of composition, redaction and transmission. No detailor collection of linguistic details in our biblical manuscripts is likelyto represent the language of the “original author” or earliest stage ofcomposition, except in very large-scale and exceptionalcircumstances, such as the overall peculiarity of Qohelethʼslanguage, or perhaps the stylistic openness to variety of the five so-called “late” books. So we are left with a lot of linguistic evidence inthe Hebrew Bible, but we are often unable to work out what to dowith it. For example, we will not be able to unravel any linguistic

27

Ian Young

chronology out of the Hebrew Bible (i.e. what is early and late) untilwe have sufficient dated extra-biblical sources to independentlyestablish such a chronology. The inscriptions I discussed areimportant and provide some datable and localised samples oflanguage. However, all told, the corpus of inscriptions is not quite2% of the size of the Hebrew Bible (Clines 2011:9–10). So it turnsout that the answer to the question of what we actually know aboutancient Hebrew is: Quite a lot, but perhaps a lot less for certain thanwe previously thought.

AcknowledgementsI would like to thank the organizing committee of the AAJS 2013conference, in particular Michael Abrahams-Sprod, Anna Hueneckeand Suzanne Rutland for inviting to me to deliver the keynote addresson which this article is based.

Endnotes1. The author reflects a common scholarly idea of the time that

Chronicles, Ezra and Nehemiah comprised one composition.2. IY – The author of this article, Ian Young.3. For a classic work applying the three-stage theory of the

development of ancient Hebrew to the books of the Bible seeDriver 1913.

4. Perhaps the stone cutters banked on the fact that since theinscription was inside the tunnel, the king wasnʼt going to riskgetting the royal feet wet to look at it.

5. There are a number of reliable scholarly introductions to theScrolls, such as VanderKam 2010. For an excellent recent surveyof scholarship see Tzoref 2013.

6. Two compositions, MMT and the Copper Scroll are exceptions(see LDBT, 1:237–41).

7. I am not claiming that the language of a text like PesherHabakkuk is identical in all respects with that of earlier biblicalbooks. I made this clear in my article (Young 2008:38). The mainpoint being made was the failure of these texts to conform to theexpectations of what normal “late” Hebrew should look likeaccording to the predictions of the three-stage model of ancientHebrew.

8. Another possibility to be considered, however, is that theirlinguistic features are solely due to the scribal histories of thoseworks (see section 5).

28

What Do We Actually Know about Ancient Hebrew

9. This would seem to be the most likely explanation for why thereare no traces of the northern linguistic forms found in theinscriptions in the current (late) manuscripts of the biblical books,see section 2, above.

REFERENCESCheyne, T.K. 1899. “Ecclesiastes,” in T.K. Cheyne and J. Sutherland

Black (eds.), Encyclopaedia Biblica. London: Watts & Co.: col.1155–1164.

Clines, David J.A. 2011. The Dictionary of Classical Hebrew, VIII Sin–Taw. Sheffield: Sheffield Phoenix Press.

Crown, Alan David. 2005. “An Alternative View of Qumran,” in MosheBar-Asher and Moshe Florentin (eds.), Samaritan, Hebrewand Aramaic Studies: Presented to Professor Abraham Tal.Jerusalem: Bialik:1*–24*.

Davies, Philip R. 1995. In Search of “Ancient Israel”. Journal for theStudy of the Old Testament Supplement, 148. Sheffield:Sheffield Academic. 2nd edn.

Delitzsch, F. 1877. Commentary on the Song of Songs andEcclesiastes. Trans. M. G. Easton. Edinburgh: T.&T. Clark.

Driver, S.R. 1913. An Introduction to the Literature of the OldTestament. Edinburgh: T.&T. Clark. 9th edn.

Dobbs-Allsopp, F.W., J.J.M. Roberts, C.L. Seow, R.E. Whitaker. 2005.Hebrew Inscriptions: Texts from the Biblical Period of theMonarchy with Concordance. New Haven: Yale University Press.

Doudna, Gregory L. 2006. “The Legacy of an Error in ArchaeologicalInterpretation: The Dating of the Qumran Cave ScrollDeposits,” in Katharina Galor, Jean-Baptiste Humbert andJürgen Zangenberg (eds.), Qumran The Site of the Dead SeaScrolls: Archaeological Interpretations and Debates -Proceedings of a conference held at Brown University,November 17–19, 2002. Leiden: Brill:147–157.

2004. “Redating the Deposits at Qumran: the Legacy of an Error inArchaeological Interpretation,” [http://www.bibleinterp.com/articles/Doudna_Scroll_Deposits_1.htm].

2001. 4Q Pesher Nahum: A Critical Edition. Journal for the Study ofthe Pseudepigrapha Supplement, 35. Sheffield: SheffieldAcademic Press.

Hurvitz, Avi. 1997. “The Historical Quest for ʻAncient Israelʼ and theLinguistic Evidence of the Hebrew Bible: Some MethodologicalObservations,” Vetus Testamentum, 47:301–315.

29

Ian Young

Hutchesson, Ian. 1999. “63 BCE: A Revised Dating for theDepositation of the Dead Sea Scrolls.” The QumranChronicle, 8:3:177–194.

LDBT see Young, Rezetko and Ehrensvärd 2008.Rezetko, Robert, and Ian Young. Forthcoming. Historical Linguistics

and Biblical Hebrew: Steps Toward an Integrated Approach.Society of Biblical Literature Ancient Near EasternMonographs. Atlanta: Society of Biblical Literature.

Tzoref, Shani. 2013. “Qumran Communities—Past and Present,” inShani Tzoref and Ian Young (eds.), Keter Shem Tov:Collected Essays on the Dead Sea Scrolls in Memory of AlanCrown. Piscataway: Gorgias Press:17–55.

VanderKam, James C. 2010. The Dead Sea Scrolls Today. GrandRapids: Eerdmans. 2nd edn.

Wirgin, Wolf. 1960. “The Calendar Tablet from Gezer.” Eretz Israel6:9*–12*.

Young, Ian. forthcoming. “Patterns Of Linguistic Forms In The MasoreticText: The Preposition ˜m ʻFromʼ,” in James K. Aitken, ChristlMaier and Jeremy Clines (eds.), Interested Readers (ClinesFestschrift). Atlanta: Society of Biblical Literature:385–400.

2013. “The Contrast Between The Qumran And Masada BiblicalScrolls In The Light Of New Data. A Note In Light Of The AlanCrown Festschrift,” in Shani Tzoref and Ian Young (eds.),Keter Shem Tov: Collected Essays on the Dead Sea Scrollsin Memory of Alan Crown. Piscataway: Gorgias Press:113–119.

2009. “What is ʻLate Biblical Hebrewʼ?” in Ehud Ben Zvi, DianaEdelman, and Frank Polak (eds.), A Palimpsest: Rhetoric,Ideology, Stylistics, and Language Relating to Persian Israel.Perspectives on Hebrew Scriptures and its Contexts,  5.Piscataway: Gorgias:265–280.

2008. “Late Biblical Hebrew and the Qumran Pesher Habakkuk.”Journal of Hebrew Scriptures 8, Article 25[http://www.arts.ualberta.ca/JHS/abstracts-articles.html#A102]:1–38.

2005. “The Biblical Scrolls from Qumran and the Masoretic Text: AStatistical Approach,” in Marianne Dacy, Jennifer Dowling andSuzanne Faigan (eds.), Feasts and Fasts. A Festschrift inHonour of Alan David Crown. Mandelbaum Studies inJudaica, 11. Sydney: Mandelbaum Publishing, University ofSydney: 81–139.

30

What Do We Actually Know about Ancient Hebrew

30

2003. “Late Biblical Hebrew and Hebrew Inscriptions,” in Ian Young(ed.), Biblical Hebrew: Chronology and Typology. Journal forthe Study of the Old Testament Supplement, 369. London:T.&T. Clark:276–311.

2002. “The Stabilization of the Biblical Text in the Light of Qumranand Masada: A Challenge for Conventional QumranChronology?” Dead Sea Discoveries, 9:364–390.

1992. “The Style of the Gezer Calendar and Some ʻArchaic BiblicalHebrewʼ Passages.” Vetus Testamentum, 42:362–375.

Young, Ian, Robert Rezetko and Martin Ehrensvärd. 2008. LinguisticDating of Biblical Texts. 2 vols. Bible World. London: Equinox.

31

Ian Young

31