Diglossia is the phenomenon where two varieties of a single language exist simultaneously, one for informal colloquial use and one for formal literary use. Such a phenomenon has been studied for spoken Arabic vs classical Arabic, Haitian Creole vs standard French, etc. Rendsburg, in his doctoral thesis, attempts to establish such a bifurcation for ancient Hebrew – a written dialect used for literary works and formal language and a spoken dialect used for everyday communication. Note that contrasting the two as written vs spoken dialects is slightly inaccurate since in certain settings the formal language is used for speaking while in others the colloquial language may be used for writing. Additionally, in light of sociolinguistics, even a bipartite division into formal and colloquial language may be too simple. For example, M Joos distinguishes five levels of language: intimate, casual, consultative, formal, and frozen (The Five Clocks, 1967).
Some earlier studies of ancient spoken Hebrew focused on the direct quotations in the Bible (MacDonald, 1975), however this method is flawed since in general biblical authors maintained the formal language even when reproducing speech. This is paralleled by the Quran, in which classical Arabic is used for direct speech rather than the colloquial. Instead, Rendsburg relies largely on the comparison to Arabic diglossia, as well as other living Semitic languages such as Ethiopian, modern South Arabic, and Neo-Aramaic (though these have no corresponding synchronic literary variety). Most importantly, however, is the Hebrew of the post-biblical period in which Mishnaic Hebrew (MH) represents the spoken dialect and Qumran Hebrew (QH) the literary dialect. The Bar Kokhba letters from the 2nd century have demonstrated that MH was indeed a spoken language, though Rendsburg notes the importance of studying the text of manuscripts rather than the printed editions of the Mishna (in which the language tends to be leveled to biblical Hebrew). Analysis of QH has demonstrated that it developed naturally from Late Biblical Hebrew (LBH). Further, there is no significant break in the literary tradition between the end of the later books of the Bible (c. 400 BCE), Ben Sira, Jubilees, and Daniel (c. 2nd century BCE), and then the DSS (c 150 BCE to 70 CE).
Rendsburg argues that the diglossia between MH and QH can be retrojected to biblical times. The literary dialect of post-biblical Hebrew is a natural development from classical biblical Hebrew (through LBH), and it can be theorized that MH is similarly the continuation of the colloquial dialect. While the literary dialect was probably more-or-less stable throughout the land, undoubtedly the spoken dialect varied. Most scholars recognize the differences between northern and southern Hebrew as reflected in the Samaria ostraca, for example, however for the purposes of this study Rendsburg has assumed a dichotomy only between formal and colloquial language.
For the purpose of establishing diglossia, it is obviously the informal language which must be defined. Rendsburg’s method is twofold. First, he suggests that those BH features which anticipate MH developments are possible colloquialisms which have infiltrated the classical language. As a test he compares the post-biblical literary language of Ben Sira, Daniel, and the DSS. If the features are absent or rare here, then he assumes that they are indeed spoken forms. If not, they may be features of later language. Second, those features of BH which run counter to the accepted grammar and have parallels in spoken Arabic (against classical Arabic) or the other spoken Semitic dialects may also be considered colloquialisms.
The data for the study is mainly morpho-syntactic rather than phonological or lexical, since dialectal differences are best measured by morphology. While phonological distinctions no doubt existed, there is little evidence for the actual pronunciation of spoken Hebrew in ancient times. As for lexicon, vocabulary differences are quite small between formal and colloquial dialects. An analysis of Arabic using the Swadesh list found only 24 distinctions out of 200 words, or 12%.
Rendsburg isolates twelve features of the colloquial dialect:
1. Gender Neutralization – the use of epicene forms where the standard dialect employs separate masculine and feminine forms, usually the masculine supplanting the feminine. For example, the 2mp אתם and 3mp הם for the 2fp אתן and 3fp הנה.
2. Incongruence – the lack of grammatical agreement (gender, number, or both) between the parts of a sentence (ie noun and adjective, subject and verb).
3. The merger of ל”א and ל”י verbs, such as נָשִׁיתִי for נָשָׁאתִי.
4. The elision of ה in the nifal, hifil, and nitpael infinitive construct with ל, such as לִיקָּטֵל for לְהִקָּטֵל.
5. Lack of agreement in definiteness between a noun and its adjective, either איש הישראלי “the Israelite man (Lev 24:10)” or פלשתים הערלים “The uncircumcised Philistine (Judg 14:3)”.
6. The use of the relative pronoun ש instead of אשר.
7. The use of the independent possessive pronoun של rather than the construct state or a circumlocution such as אשר ל.
8. The use of the anticipatory pronominal suffix, ותראהו את הילד “And she saw (him) the child (Ex 2:6)”.
9. The use of זֹו/זֹה as f.s. demonstrative pronoun rather than זֹאת and the use of אֵלּוּ as the c.p. form rather thֶּan אֵלֶּה.
10. The use of a second set of demonstratives – m.s. הַלָּזֶה, f.s. הַלֵּזוּ, c.p. הַלָּלוּ, etc.
11. The use of the 1 c.p. independent pronoun אנו rather than אנחנו.
12. The construction היה + participle.
In conclusion, Rendsburg analyzes the distribution of these features within the biblical corpus. First he looks at the effect of literary genre (prophetic, poetic, legal/cultic, narrative, and orational – sections of the Prophets which fall into neither the prophetic nor narrative genres). Of the 598 colloquialisms, 9.7% occur in prophetic sections, 23.7% in poetic, 6.8% in legal/cultic, 46.0% in narrative, and 13.7% in orational. Correcting for the distribution of each genre within the Bible, he calculates the following deviations (lower numbers indicating less likelihood of the existence of colloquialism): prophetic -35.3%, poetic -12.5%, legal/cultic -30.6%, narrative 16.2%, and orational 59.3%. This suggests that prophecy, poetry and legal/cultic language are the least likely to admit colloquialism (not surprising since these are all formal contexts), while prose composition such as narrative and oration admit colloquialism more freely.
Rendsburg next analyzes a sample of passages within the narrative corpus, finding that sections containing 3rd person narration occur 57.5% of the time and contain 58.6% of the colloquialisms in narrative. Similarly, sections containing direct quotation occur 42.5% of the time and contain 41.4% of the colloquialisms. This confirms that direct quotation is no more likely to contain colloquial language then the rest of the narrative.
Rendsburg also investigates whether the date of composition has an effect, finding that pre-Exilic literature shows the least influence of colloquialisms, then Exilic, and finally post-Exilic the most. This may be explained by the loss of an official literary circle with the fall of Jerusalem in 586 BCE.
Lastly, Rendsburg points to the study of Kurt Sethe who analyzed the history of Egyptian, concluding that each successive stage of the literary dialect is actually the spoken dialect of the previous period. Applying Sethe’s model to ancient Hebrew, Rendsburg suggests that the standardization of Classical Hebrew, especially from c. 1000 BCE onward in Jerusalem, initiated the distinction between the formal and informal dialect. The colloquial language continued to develop, widening the gap with the more conservative literary language. After 586 BCE, written Hebrew is less standardized and more susceptible to spoken idiom. Thus there is more development in LBH and QH than CBH. The events of 70 CE ended the literary Hebrew of BH and QH, and the spoken dialect of MH took over as the literary dialect. Presumably spoken Hebrew continued to develop until c. 200 CE.