Lethargic Man (anag.)

It won't come as much of a surprise to most of my readership that I am interested in philology. In particular, my engagement with German has prompted me to learn more of where German words came from and their cognates in English. Of course, the two originated as dialects of a common language, dubbed Common Germanic, spoken from about 500 BCE to the fourth century CE^*, which was ancestral to all the historical Germanic languages, from Afrikaans to Yiddish.

* Though Old Norse and the East Germanic languages (Gothic, Burgundian, Vandalian) were already splitting off by then.

Philologists talk about languages as evolving to unintelligibility over the course of a millennium, which made me wonder how different Common Germanic must have been to Old English (which is of course unintelligible to non-scholars today). So I headed off to Wikipedia to find out—and here's my findings, essentially boiling down the most interesting bits of what I read into one blog post (which I suspect will mostly be of interest just to me, but now I've got it in a form I can readily refer to in the future).

[S]omething I'd love to hear is a text read which starts in Old English, and as the reading progresses, the pronunciation, grammar and vocabulary gradually shift into Modern English, via all the intermediate stages. A quick google does not show me any such thing, though there are things like this, which have discrete texts jumping forward several hundred years between each.

I get annoyed when I hear people pronouncing קַדְּשֵׁינוּ as [kadšeinu] or קִדְשָׁנוּ as [kidšanu]. Unless you're using the Israeli pronunciation, it's [kadəšeinu] and [kidəšanu]. Now, I know most of you use a different pronunciation schema, with different rules about when a שְׁוָא is vocalised or not, but some siddurim, including the fourth edition Singer's Prayerbook, indicate this, and they bear me out.

But why is this שְׁוָא a שְׁוָא נָע ("mobile", or voiced, שְׁוָא)? My journey to understanding this starts with ewt's copy ofWeingreen's Practical Grammar for Classical Hebrew, which asserts that in Hebrew the Semitic genitive ending -i has been lost, and there is no difference between the nominative and genitive in Hebrew. Some time later, whilst davening, I noticed what seems to be a counterexample. (I'm still not sure if it actually is, because my understanding of Hebrew grammar isn't good enough.)

* Gemination is doubling of a consonant, and its length: think of the length of the N sound in 'unnamed' in English compared with in 'un-aimed'.

I still don't know why the ק is is geminated in the one example and not the other, but I now feel confident the change in the realisation of the שְׁוָא is because of this. Once I'd noticed this, I started noticing other examples, which leads me to the answer to my original question. Hebrew verbs are composed of three root letters, and in the פִּעֵל the middle letter of the root receives a דָגֵשׁ (centre dot) if it is grammatically capable of doing so (i.e. is not a guttural (א, ה, ח, ע) or ר). Nowadays this largely makes no difference to the pronunciation, but in Biblical times and later, it geminated the consonant. In the case of a few consonants (פ, כ, ב, also ת if you're Ashkenazi or Iraqi, and ג and ד if you're Yemenite—and all of these historically) it also modifies the sound of the letter, but this does not remove the gemination of the letter.

Now, in some languages you can have a geminated consonant immediately next to another consonant, but not in Hebrew. Hence when a geminated consonant takes a שְׁוָא, that שְׁוָא must be vocalised. Hence קַדְּשֵׁינוּ and קִדְשָׁנוּ, which would originally have been pronounced [qaddəšeinu] and [qiddəš̄ānu], with doubled Ds. The verb being in the פִּעֵל doubles the length of the ד, the ד having a שְׁוָא then makes that שְׁוָא voices.

Now that I understand what's going on here, I have become alert to the same phenomenon in other places, and have begun to modify my pronunciation where I can see it's wrong. The same rule applies in שַׂבְּעֵנוּ מִטּוּבֶךָ וְשַׂמְחֵנוּ בִּישׁוּעָתֶךָ: they are also verbs in פִּעֵל, so should be pronounced with [sab`əeinu] and [saməḥeinu], not [sab`einu] and [samḥeinu].

The presence of a דָגֵשׁ, forcing a שְׁוָא to be a שְׁוָא נָֹע, is not limited to this context. In the Shabbos shacharis Amida occurs the word מַתְּנַת. I'd always pronounced this [matnas]; however, there's a דָגֵשׁ in the ת as the legacy of the dropped נ (the word derives from the verb נתן, and where a letter elides out in Hebrew it normally leaves a trace in the form of a דָגֵשׁ in the following letter). Hence the word should be pronounced [matənas], and again the Singer's bears me out here.

Another example: Most people think the definite article in Hebrew is הַ־. In fact, it's actually that plus gemination of the following letter, where that letter permits it. In the Shabbos prayers, we make frequent reference to "the seventh day". "The seventh" is הַשְּׁבִיעִי; the דָגֵשׁ means it's [hašəvī`ī], not [hašvī`ī] (in Biblical Hebrew [haššəvī`ī]), and similarly וּבַשְּׁבִיעִי is [uvaš(š)əvī`ī], not [uvaš(š)vī`ī]: the דָגֵשׁ there represents the dropped ה, because בַ־ is short for בְהַ־. (Similarly מִ־ is short for מִן and the dropped נ causes gemination of the following letter.)

Now onto a different case. The Singer's Prayerbook indicates a שְׁוָא נָֹע in words like אָמְרוּ. This doesn't fit into the above categories; what's going on here? Weingreen informs me that a מֶתֶג (the short vertical line you sometimes see under a letter) occurs here; the מֶתֶג forces a pause and the end of the current syllable. (This is its use in the Bible; in siddurim it is, confusingly, often used for a different purpose, of indicating where the stress falls in a word.) This word is pointed אָ ֽמְרוּ (you may need to increase font size in your browser, using CTRL-+, to see the diacritics clearly); hence in Biblical times this would have been pronounced [ā|mərū], not [om|rū]: the מֶתֶג makes the reader pause and hence leaves the first syllable open, turning the vowel from [o] to [ā] (i.e. it does not have a קָמַץ קָטָן).

Unlike in the previous example, I do not understand why this מֶתֶג is there. So how do we know that it represents a genuine indication of how the word was pronounced in ancient times, particularly given that the נִיקוּּד (system of diacritics indicating vowels, including the מֶתֶג) is less than fifteen hundred years old (though is said to encapsulate older oral tradition)?

The answer, as pointed out to me by boroparkpyro, lies with transcriptions of names into other languages. The name אָ ֽסְנַת, which is pronounced by most people nowadays "Osnat", has a מֶתֶג under the א, and, indeed, the ancient transliterations of this name, Ασεννεθ and Asenath, bear this out. (You can see the same in other similar Biblical names: (A/O)holiav, (A/O)holibamah, perhaps also Basemath.)

So, in summary, I recommend davening with a siddur that indicates where a שְׁוָא is שְׁוָא נָע, and paying close attention to where the way you pronounce something doesn't agree with what it says in the siddur. It's probable that God isn't so petty as to get narked when you pronounce something wrong, but God probably also appreciates you putting in the extra effort to ensure you do pronounce things right.

I got an unusual email today:

I stumbled across one of your blog posts while looking for how to write a birthday in Cuneiform, and was wondering if you have gotten any further along in your studies?

The reason why I ask is because I am typesetting a book for an author who would like to have his birthday (January 31, 1968) written in Cuneiform on the "About the Author" page.

I thought: either this will be a five minute google, or it will be completely beyond me. What I found, to my surprise, is that it was somewhere in between the two, so I regarded it as a challenge. Starting from an overview of the Babylonian calendar on Wikipedia, I found a calendar converter online, but it only supoprts dates up to 76 CE. So, to work out the birthday, we need to move the date back into that range. Now, the Babylonian calendar runs in a nineteen year period; this is because the lunar and solar calendars come together every nineteen years. So, if we move multiples of nineteen years into the past, we'll arrive at the same Babylonian day.

That's the theory.

In practice—warning: calendar geekery follows—the nineteen year cycle in the lunar calendar consists of 235 months of 29.53059 days, totalling 6939.68865 days; but nineteen solar years, according to the Gregorian year of 365.2425 days, comes to 6939.6075 days, i.e. the lunar year is 0.08115 days (1 hour, 56 minutes and 51.36 seconds) longer than the solar year. So, winding the birthday back by one hundred nineteen-year periods, we end up on the 31 January 68 CE Gregorian, which is 2 February 68 in the Julian calendar; but by that stage the lunar calendar is 154.185 days out of sync with the solar calendar—almost half a year!

So the question is: do we adjust the date by that period, or assume the Babylonians would have corrected their calendar to keep it synced with the solar year (like Western culture did with the switch to the Gregorian calendar)—or like the Jews, who also use a lunisolar calendar, and picked up the Babylonian month names during the Babylonian Exile. I'm going to assume the latter: after all, they did put intercalary months in to result in that nineteen year cycle, unlike the Muslims, whose calendar gradually gets earlier and earlier compared to the Gregorian calendar (which is why the Islamic calendar now says it's 1432 years after the Hejira, despite that happening 1389 common-sense years ago!).

So, if we enter 2 February 68 into the above web form, this tells us that the date then was 9 Shevat Šabaṭu of the year 378 of the Seleucid Era according to the Babylonian reckoning (and 379 according to the Macedonian reckoning), or year 314 of the Arsacid Era. Seleucus was one of the generals of Alexander the Great, who established an empire of his own in Babylonia after the death of Alexander; Arashk (or in Greek, Arsaces) was the founder of the (later) Parthian Empire. I have no idea which is more relevant, but let's go with the Seleucid reckoning, on the grounds that it's referred to more on the Net of a Thousand Lies. In reality, if cuneiform had remained in use, they would almost certainly have switched to the Islamic calendar, but by the time Islam reached Mesopotamia, cuneiform had fallen out of use in favour of the much easier to write Aramaic.

So, now how do we represent the date 9 Šabaṭu 2178? A prolonged google (via an interesting digression to the tablet showing the calculation of √2) seems to show that dates were expressed in the form:

where MU, KAM, ITI and UD are sign names referring to, I think, what they represented in Sumerian; in Babylonian MU is (I think) šattu, ITI (w)arḫu (related to Hebrew יֶרַח), UD ūmum (יוֹם). KAM I couldn't easily find a Babylonian equivalent of, but it doesn't matter, as we're just using the cuneiform sign anyway. The formula above is from a late-period text, so that suits my purposes.

Therefore what I need to write is:

(where 36,18 is 2178 expressed in sexagesimal, the base 60 system the Babylonians used), which appears to be (mouseover for more information):

Note: I'm not sure about the first sign of Seleucus's name; that appears before his name in various places, represented as a superscript m. I have no idea what it means. [ETA: Found it out: m, f and d represent signs indicating a man, a woman and a god(dess).]

So there you are, how to represent 31 January 1968 in Babylonian cuneiform. Do be warned, though, that if you use this, sooner or later you will get a knowledgeable reader coming up and pointing out that I've made horrendous mistakes, and mixed together cuneiform signs from a thousand years apart, or used the wrong sound value, or anything even worse.

But for your average reader, I could have written "I like linguistics geekery", and 99.9% of them would never know the difference. :o)