Bunis on the survival of Hebrew

Hebrew is somewhat famous as the most successful attempt to bring a formerly dead language back to life. But when did Hebrew die out in the first place? Up to the early 20th century, scholars thought that basically all speakers switched to Aramaic during the Babylonian Exile (sixth century BCE), killing off Hebrew in the process. The Hebrew literature written after that point would then have been composed in a language that was no longer in spoken use. Since then, however, a better appreciation of the vernacular status of Rabbinic Hebrew as well as epigraphic finds have shifted the consensus. Today, most scholars believe Hebrew died out as a spoken language after another dramatic event: the failed Bar Kokhba Revolt (second sentury CE) and subsequent deportation of most Jews from Jerusalem and the Roman province of Judaea (renamed Aelia Capitolina and Syria Palaestina for good measure).

Coin from the Bar Kokhba revolt (specifically 134/5 CE) depicting the Temple with the name 𐤔𐤌𐤏𐤅𐤍 ‘Shim’on’ (left) and a lulav and etrog and the inscription 𐤋𐤇𐤓𐤅𐤕 𐤉𐤓𐤅𐤔𐤋𐤌 ‘for/of the freedom of Jerusalem’ (right).

In two recent papers (and more to come?), Ivri J. Bunis argues that Hebrew persevered even longer, surviving in Palestine into the late Roman and (say it with me) Eastern Roman period. Both papers focus on Palestinian Amoraic (not Aramaic!) Hebrew, that is: Hebrew from Palestine (mostly Galilee) from ca. 200–500 CE. A summary of his arguments:

In his 2020 paper, Bunis shows that sequences of imperatives, especially with verbs of movement, are mostly expressed without ‘and’ (e.g. “go see”) in both Biblical Hebrew and Jewish Palestinian Aramaic, while Palestinian Amoraic Hebrew mostly does use ‘and’ (“go and see”), like the earlier phase of Rabbinic Hebrew, Tannaitic Hebrew, which is now generally considered to reflect a living language. This shows that the people writing Palestinian Amoraic Hebrew weren’t just writing Aramaic with a thin Hebrew veneer or basing themselves on Biblical Hebrew. In my opinion, this mainly shows that Palestinian Amoraic Hebrew is indeed a form of Rabbinic Hebrew, continuing the earlier, Tannaitic stage of the language. But especially with subtle, syntactic things like this, it would be very easy for a scribe who was dominant in Aramaic to slip up and get it wrong. This paper also has a detailed review of the scholarship on the survival of Hebrew in previous literature.

A stronger case is made in Bunis’s 2022 paper on Rabbinic Hebrew hallā– pronouns. (Conflict of interest: I hope to present something on these pronouns together with Ivri at a conference this summer.) Bunis does a really quite excellent job of showing that the use of these demonstratives evolves throughout the attested stages of Rabbinic Hebrew. Hopefully without oversimplifying or getting too much wrong, the picture looks something like this :

Bar Kokhba Letters (some of these forms aren’t attested as such and some other variants occur)

	‘this/these …’	‘this/these one(s)’	‘that/those …’	‘that/those one(s)’
m.sg	hzh	zh	hlz	hlz
f.sg	hzw, hzʔt	zw, zʔt	hlzw	hlzw
pl.	hʔlh, hʔlw	ʔlh, ʔlw	hlww	[h]llw

Eearlier Tannaitic Hebrew

	‘this/these …’	‘this/these one(s)’	‘that/those …’	‘that/those one(s)’
m.sg	(haz)ze	ze	ʔōṯō, hallāz	ʔōṯō, hallāz
f.sg	(haz)zō	zō	ʔōṯāh, hallāz	ʔōṯāh, hallāz
pl.	(hā)ʔḗllū	ʔḗllū	ʔōṯān, hallā́lū	ʔōṯān, hallā́lū

Later Tannaitic Hebrew

	‘this/these …’	‘this/these one(s)’	‘that/those …’	‘that/those one(s)’
m.sg	(haz)ze	ze	ʔōṯō	ʔōṯō, hallā
f.sg	(haz)zō	zō	ʔōṯāh	ʔōṯāh, hallāz
pl.	(hā)ʔḗllū	ʔḗllū	ʔōṯān	ʔōṯān, hallā́lū

Palestinian Amoraic Hebrew

	‘this/these …’	‘this/these one(s)’	‘that/those …’	‘that/those one(s)’
m.sg	hazze	ze	ʔōṯō	ʔōṯō
f.sg	hazzō	zō	ʔōṯāh	ʔōṯāh
pl.	hallā́lū	ʔḗllū	ʔōṯān	ʔōṯān

So: the hallā– pronouns are quite common in the 2nd-century Bar Kokhba Letters; Tannaitic Hebrew innovates a much more productive set of distal demonstratives based on ʔōṯ- and later restricts the hallā– pronouns to substantive use; and Palestinian Amoraic Hebrew loses them altogether, except for plural hallā́lū, which gets repurposed as the adnominal counterpart of proximal plural ʔḗllū for morphological reasons. All of those developments, including the ones in Palestinian Amoraic Hebrew, look perfectly natural, like the kind of thing speakers would do in a living language. And while Bunis doesn’t spell this out as much as in the 2020 paper, there’s nothing at all in Jewish Palestinian Aramaic or Biblical Hebrew that would motivate this.

Based on the second paper, especially, I think I’m convinced that Palestinian Amoraic Hebrew had some kind of living status. I’m not sure Bunis has shown that it was a living language in the strict sense, with children acquiring it as their first language; it could still be that it was only used in an academic setting, so to speak. But either way, it clearly had some kind of life of its own: speakers were apparently still learning to produce Hebrew very fluently, whether as a first or second language.

So, should we revise the date of the death of Hebrew once again? Both papers can be read for free on Bunis’s Academia page, and while they’re fairly long, I think they’re well written and easy to follow. So if you want to form an opinion on what might be a big development in the historiography of Hebrew, do head over there and give hammaʔămārīm hallā́lū a read.