Sunday, June 16, 2024

dayal, doula, and degel - update

 I have a long list of words to write about. Today I started looking into one of them, and then only later did it turn out that I had already written about it. This happens occasionally - with nearly 700 posts, and often several words beyond the primary word discussed in each, I suppose it’s to be expected. I’m just relieved when I discover it before I write the whole thing. 

This time, however, I found some new information, so I thought I’d write a post that updates the earlier one.

I had intended to write a post about the words dayal דַּיָּל - “steward” and “doula.” I was planning on pointing out how they share a common origin. But I had already discussed it in my post on meltzar מֶלְצַר, another word meaning “steward”:

However, as Elon Gilad writes here, Ben Yehuda did not want the word meltzar used for "waiter" in Modern Hebrew. He preferred dayal דייל (feminine dayelet דיילת). He coined dayal on the basis of the Talmudic Aramaic word dayala דיילא - "attendant", which in turn derives from the Greek word for slave or servant - doulos. Doulos is also the root of the English word doula, which literally means "female slave".

However, as happened on more than one occasion, Ben Yehuda's plans did not win out, and people continued referring to waiters as meltzarim. But his word dayal was eventually redeemed - when El Al airlines was founded in 1948, they needed a specialized word for someone attending to passengers - and so a few years later, dayal became the Hebrew word for steward. Quite the journey for these words!

But as noted, I forgot that post, and began to research. I found Klein’s entry for dayal:

NH waiter, steward (on an airplane). [Nomen opificis coined by Eliezer ben Yehudah (1858–1922) from JAram. דַּיָּלָא (= attendant, waiter), which derives from Gk. doylos (= slave), a word standing for doelos and derived from Aegean doëro (= slave).]

As well as the Online Etymology Dictionary entry for doula:

"woman trained to assist another woman during childbirth and provide support to the family after the baby is born," by 1972, a coinage in anthropology, from Modern Greek doule, from ancient Greek doule "servant-woman," fem. of doulos "slave, servant," which probably is a word of Pre-Greek origin.

That last sentence was interesting. When Etymonline says “Pre-Greek,” it sometimes refers to a Semitic etymology. But could that be the case here?

It turns out that it just may be. The Wiktionary entry for δοῦλος (doulos) has this interesting etymology:

Related to Mycenaean Greek 𐀈𐀁𐀫 (do-e-ro /⁠dohelos⁠/),[1] possibly from Canaanite *dōʾēlu “servant, attendant” (compare Late Babylonian 𒁕𒀝𒂵𒇻 (daggālu, “subject, one who waits on another, does their bidding”), Aramaic דַּיָּילָא (dayyālā), Hebrew דייל (dayyal, “flight attendant, store clerk”)).[2]

According to Parpola,[3] the word δοῦλος is related to the ethnonym Dahae (found as Δάοι, Δάαι, Δαι or Δάσαι in Greek sources) and thus related to Sanskrit दस्यु (dasyu, “bandit, brigand”) and Sanskrit दास (dāsa) which originally meant 'demon' and later also 'slave' or 'fiend'. 

The first theory is the one that interested me - since it proposes a Semitic origin. However, it seemed rather mixed up, giving the anachronistic impression that the Greek doulos derived from not only the Aramaic dayala (which we had already seen is purported to derive from the Greek, not the other way around), but also mentions the Modern Hebrew dayal, which certainly couldn’t have influenced any Ancient Greek words. 

But I thought I’d try looking around a bit more. I couldn’t find anything of note about the Canaanite *dōʾēlu, other than websites quoting or referring back to this Wiktionary page. But the Babylonian daggālu had more promise. Since Late Babylonian is another word for Akkadian, I looked in Tawil’s dictionary of Akkadian. In the “Akkadian to Hebrew Concordance,” under dagālu, he points back to his entry for דגל. In that entry he writes:

Akkadian dagālu … to look (at) …

Akkadian dagālu in the G-Stem and S-stem has a wide variety of nuances and meanings, including “to own” and “to be a subject.” With the prepositions ana, pan, and ina pan, it means “to wait for.”

This fits what I wrote in an even earlier post on the Hebrew word דגל degel. I quoted Milgrom on Bamidbar 2:2

Hebrew degel possibly originally meant a military banner. This is supported by the Akkadian dagalu, "to look", and diglu, "sight"

But while dagālu could mean “to be a subject,” is there further evidence that it’s related to dayal and doula?

Sokoloff, in his Dictionary of Jewish Babylonian Aramaic, does make such a connection. In his entry for the Aramaic dayala דַּיָּילָא, he provides two definitions. A type of official (as in Yoma 18a) or a servant (as in Pesachim 86b). And for the etymology, he says it derives from the Akkadian dajalu - “inspector.” 

So Sokoloff says that dayala (the source of dayal) derives from dagalu. He makes no mention of the Greek doulos, but there’s nothing in what he wrote that would contradict doulos deriving either from the Akkadian or Aramaic.

Where then did Klein get his idea of a Greek origin for dayala? I assume this Ben Yehuda entry:

Ben Yehuda defines a dayal as someone who serves food (i.e., a waiter) and says it was common in Hebrew speech and also used in newspapers. In the footnote, after quoting Pesachim 86b (see above), it notes that there are those who say it is borrowed from Greek. 

I don’t know who are “those who say” but I imagine it’s possible that Akkadian scholarship at that point had not advanced to the level it later did, and so the dictionary editors weren’t aware of the possibility of an Akkadian origin. 

Sunday, June 09, 2024

intifada and pitzutz

In the past, I've talked about how I'm a fan of Mike Pesca's podcast, The Gist. Recently, Pesca had a segment, “Intifada Revolution? Or is that a linguistic delusion?” (starting at 26:30) where he railed against those claiming that because the Arabic word intifada (“insurrection, uprising”) originated in a more gentle meaning of “shaking off,” the protesters calling for an intifada today aren’t really inciting for violence. 

It’s a great segment, where Pesca skillfully explains how words change meaning, and how we need to be honest about how the words are used today. The Palestinian expressions of intifada have been very violent, with thousands killed, and it is disingenuous to claim that calls for further “global” intifada would be any less violent. 

I won’t repeat all of Pesca’s arguments here - it’s really worth listening to. But the essence of his position is against what is known as the etymologically fallacy - that a word’s meaning is determined by its etymology. On this site, I implicitly campaign against that approach constantly. By showing the development of words over time, even those with weighty religious usage, I try to show that words change, and we need to understand how they were used at the time they were said or written.

The segment on the Gist did get me thinking. Does the Arabic word intifada have any cognates in Hebrew? It took a little digging, but it certainly does. 

As noted, the word intifada did mean “shaking off.” As noted here, it comes 

from the verb intafada "to be shaken, shake oneself."

The verb intafada in turn is the reflexive form (similar to hitpael in Hebrew) of the verb nafada - “to shake, shake off.” Klein notes that nafada is cognate to the Hebrew verb נפץ - “to shatter, scatter”:

Prob. a secondary base derived from פוץ ᴵ. cp. Aram.-Syr. נְפַץ (= he shook out, emptied), Arab. nafaḍa (= he shook), Akka. napāṣu (= to shatter).

From the root נפץ, we get many words relating to shattering or exploding such as napatz נַפָּץ - “detonator” (or in modern Hebrew slang “firecracker”), mapatz מַפָּץ - “bang, explosion” (as in “the Big Bang” hamapatz hagadol הַמַּפָּץ הַגָּדוֹל), and the verb hitnapetz הִתְנַפֵּץ - “to shatter, disintegrate, crash.”

Klein noted that נפץ is probably a secondary form of the root פוץ. This root has a similar meaning: “to disperse, scatter, spread.” It’s most commonly seen as a verb in the form הפיץ - “to scatter, spread, disseminate, propagate.” As an adjective, it gives us the word nafotz נָפוֹץ - “widespread”, and as a noun tefutza תְּפוּצָה - “dispersion, diaspora.”

Another root that comes from פוץ is פצץ - “to break, to shatter.” In Biblical Hebrew it could refer to such actions as breaking rocks, like in Yirmiyahu 23:29

  וּכְפַטִּישׁ יְפֹצֵץ סָלַע - “as a hammer that shatters rock” 

In modern times, that verb was borrowed to mean “to explode, detonate,” giving such words as petzatza פְּצָצָה - “bomb” and pitzutz פִּיצוּץ - “explosion.” 

As we can see, many derivatives of these related roots refer to volatile acts of explosions, detonations and shattering. I will remain consistent with my approach, and will point out that these words have also changed meanings over time (such as the coining of petzatza by Ben Yehuda). Even if the original meanings were more violent, that doesn’t mean that the original Arabic nafada had that connotation. But likewise, the meanings of those Arabic words have also changed, and so intifada cannot be divorced from its more recent associations with terrorism. 

Tuesday, June 04, 2024


The background to how I began exploring the etymology of tallit טַלִּית is complicated, but perhaps more interesting than simply the bottom line. Therefore, I’ll tell it more like a story, and hopefully it will be fascinating to you as well.

It began when I was watching a video from the wonderful Jewish history YouTube channel by Sam Aranow. This video is called The Revival of Hebrew? (1879-1908), which focuses on the contributions of Eliezer Ben-Yehuda and others in the modern revival of Hebrew.

To get a better perspective on the Hebrew language, Sam brought on Yair from the Che Languages YouTube channel. At 2:20, they mention words originating in Egyptian, Persian, Greek, and Latin. Many of those words I’ve written about here before. But one word caught my eye - tallit, which they claimed had a Greek origin.

I had never heard such a suggestion before. I was only familiar with the etymology provided by Klein:

טַלִּית f.n. Post-Biblical Hebrew 1 cloak. 2 prayer shawl. [Prob. from טִלֵּל (= he covered, roofed).]

In fact, I had mentioned that as a tangent on a post a while back:

The letters tzade and tet can switch between Hebrew and Aramaic, as can also be seen in the words tzel צל - "shade" and טלל - "to overshadow", the root of talit טלית.

But where was this Greek origin theory from?


It turns out that Yair had found the Greek origin for tallit on the Wiktionary page for טלית. The page provides two possible etymologies. One the one that I had heard, and the other claiming Greek origin:


Unclear. Suggestions include:

  • From the Aramaic root ט־ל־ל (t-l-l) (related to the Hebrew root צ־ל־ל (ts-l-l); compare צל (tsél)).

  • From Ancient Greek στολή (stolḗ, “garment”).

The only source given on that page was a 2001 post on the Avodah email discussion list by Rabbi Dr. Seth Mandel entitled Tallit/talles (a follow up from an earlier post of his and in response to a Philologos column). This is the relevant section from his post (links and italics are mine):

Furthermore, there is another, less well-known word in the Mishna which is certainly associated in meaning (remember, in the Mishna, tallet/t'lit does not mean exclusively "prayer" shawl). The word in most modern printed editions is vocalized itzt'lit: aleph, tzadi, lamed, tav, with some immot qriah thrown in as well. Look in Yoma 7:1 and Gittin 7:5. That word in the manuscripts is written in various ways: the Kaufmann ms. has estalet, with no yod at all, a segol under the aleph, then a samekh, then a tzere after the lamed (which has no dagesh). The Rambam own hand ms. of the Mishna also has the word without a yod before the tav, indicating the vowel is not a hiriq. That word, as the various spellings give away, is the Greek word stolé (also borrowed in English, by way of Latin, as meaning robe, commonly used as in mink stole). Aramaic and Leshon Hazal could not tolerate two consonants together beginning a word, and so a proclitic vowel was added to such Latin and Greek words, as also in words like itztadion (stadium) and many others. That Greek word, with the feminine Aramaic ending, was then estaleta/estalet or estalit. It seems clear that tallet was either a shortened form of this loan word, or some original Aramaic word from the root tll (which word is unattested) became influenced by the Greek loanword and its pronunciation. That would explain both tallet and t'lit: the Greek loanword had a short a vowel (commonly used as a reflex of the Greek omicron), so it either became a shva in Hebrew (and hence the Teimani form) or a pasah (which would require doubling of the lamed with a dagesh). Not only does a foreign origin explain the varying forms of tallet/t'lit

[He then goes on to explain why tallit is often pluralized as talleisim in Ashkenazic Hebrew, for more details, read the rest of his post.]

Mandel is claiming that the word tallit derives from the from the Talmudic word itztela אִצְטְלָא meaning “robe, cloak,” which in turn comes from the Greek “stole.” The English word “stole” has the same origin. It either means a long scarf or shawl, particularly used by women, or it can  refer to a liturgical vestment worn by Christians, which some say was influenced by the tallit.

It’s certainly an interesting theory, and Mandel’s explanation is certainly detailed and well thought-out. The words itztala and tallit could be connected linguistically, and certainly have a similar meaning. And the tallit was, as Mandel notes, a garment for general use - not only for prayer as it is today. But is the etymology his original idea, or did he base it on previous scholarship?

I intended to ask Rabbi Mandel directly, but sadly he passed away while I was researching the issue. But despite that loss, and hopefully to perpetuate his memory (and our shared love of Hebrew language history), I felt it was important to continue the search.

I first noticed that Rashi (Sanhedrin 44a, ד”ה איצטלא דמילתא) explains that itztela means tallit. That helps in identifying that the words have a shared meaning, but doesn’t necessarily speak to the etymology (and Rashi certainly does not make such a claim.)

I looked at more recent scholars for clues. Jastrow supports the טלל origin (not surprisingly, since he generally leans in the direction of Semitic roots for Hebrew and Aramaic words, even when more recent scholarship doesn’t justify it.) Kohut prefers a Persian origin (again, that seems to be his default preference.)

Ben Yehuda’s entry for טלית is interesting. The footnote (likely edited by Tur-Sinai) says that the origin of the word is unclear. It brings the טלל theory, but rejects it. (This makes Klein’s adoption of the theory surprising, since in general he follows the Ben Yehuda dictionary.) 

In support of this rejection, he quotes an 1890 article by the scholar David Günzburg in the journal Revue des Études Juives. In the article, “Origine du mot Talit”, Gunzburg explores a number of possibilities as to the etymology of tallit

Full disclosure, I wasn’t able to get a fully readable English version from online translation tools. If I’m not mistaken, he suggests a possible connection to the Latin trilix, meaning a three-threaded garment. That seems far-fetched, and I can see why Ben-Yehuda didn’t quote it.

But one claim of his did draw my attention. Just as Mandel had argued, Gunzburg also writes (page 18) that tallit is a masculine noun (not feminine as we use it today), which proves that it is a non-Semitic root. (In other words, the ת at the end is not a suffix, but part of the word.) But neither Gunzburg, nor Ben-Yehuda, offer a conjecture as to what that non-Semitic word might be.

Gunzberg didn’t, but Ginzberg did. I’m referring to the scholar Louis Ginzberg. I found a 1916 Festschrift for Adolf Schwarz, edited by the linguist Samuel Krauss. In the book, Ginzberg has a long essay (329-361) where he discusses the etymology of various Hebrew and Aramaic words. On page 359, he has a paragraph about the origin of tallit. He quotes Gunzburg, and accepts the non-Semitic origin of the word. He then goes on to propose that tallit derives from itztala

Here’s where it gets a little confusing. Ginzberg cites Krauss (the editor of the Festschrift). Once again, full disclosure - I’m relying on online translation for Ginzberg’s German. Here’s the original:

The translation seems to indicate that Krauss a) acknowledged the derivation from itztala but rejected it, and also b) accepted that explanation in another source. 

However, in Tosefet HeArukh (of which Krauss was one of the editors) the entry for tallit quotes Ginzburg in the Festschrift:

This entry also quotes Krauss (so maybe he didn’t write it? or it was written by committee?), and it says Krauss gives it an entirely different origin (the root טלא - “to patch”.) It also goes on to reject Kohut’s Persian origin (significant because one of the other editors was George Alexander Kohut, the son of the author of the Arukh HaShalem).

So it seems that Krauss was aware of the idea before Ginzburg. Did he come up with it only to reject it (a hava amina of sorts?), only to have Ginzburg remain convinced? Or was he quoting someone else?

I suppose more research needs to be done. And of course, the question remains: did Rabbi Mandel know about the theories of Krauss and Ginzberg? Either way, his scholarship is impressive. If he found their research, that must have taken a good deal of effort (certainly prior to the ease of internet searches which I benefited from). If he came up with it on his own, then he displayed creativity combined with serious intellect. 

And now I’m at the end of my journey. I started off being surprised by a claim in a video from 2023, and ended up finding debates from the turn of the 20th century. I hope you enjoyed the ride!