Monday, June 27, 2011

'Governor' in Minoan - the origins of Greek Βασιλεύς

Hello again, dear readers! To keep the interest in this little blog, I have decided to leave the topic of toponyms for a while, and cruise into foreign waters. One thing the classification of Linear A terms was exceptionally good for - to get those words, which are (with high probability) not toponyms. But it is not chaff that remained in our hands after gleaning out the place-names. Rather, it is a handful of gems. We shall see, that some of these terms turn out to be administrative titles, that - by finding their original phonetic values - can be identified with Mycaenean and even Modern Greek words!

One term in question traditionally reads as QA-*118, and has a clearly related word in the form QI-*118. Although these terms are mentioned all across the island (Haghia Triada, Khania, Archanes, Zakros), they never occur on place-name lists. Instead, they look a lot like titles, especially QI-*118, that typically stands alongside hapax legomena (one-time terms, highly likely personal names).

But how do these terms read? To get an idea, we have to go back to one of the earlier posts, where I suggested the Lin A *118 / Lin B *83 sign to be read with a value ZI (based on the identity of place-names DA-SI-*118 / DA-*83-JA). Plugging this value into the cited titles yields readings QA-ZI and QI-ZI. This is nice, but we are not done yet: There are still a few mentions of derived cases from QA-ZI - with some strange extensions. They had to be set apart from the common *i- prefix that likely denotes a deictic or connector ('that', 'which', 'what') - frequently seen on initial words or in longer phrases.

On HT70 we can read a form QA-*118-[*]: the sign on the place of the asterisk could equally have been SA or RE (undecidable, since a breakage line runs straight through it). On the other hand, HT96 clearly gives a form QA-*118-RA-RE. Here the sudden occurrance of an -R- after QA-ZI implies that the stem word ended with a consonant (*-r or *-l), simply ommitted due to the Linear A writing conventions. It is only seen here because of the addition of the *-(a)le suffix onto the stem: that case-ending is supported by a number of other Minoan words (e.g. compare JA-MI-DA-RE [HT122, toponym on a list] with A-MI-DA-U [ZA10, the same toponym on another list]).

This is the point when one would suddenly feel enlightned: After all, the form QA-ZI-R is almost exactly the same as the Mycaenean (Linear B) QA-SI-RE-U, meaning 'village chieftain' or 'governor'. Actually, it is also the same stem what the Greek word for 'king': βασιλεύς (or modern Greek βασιλιάς) shows! The Pre-Greek origin of basileus / QA-SI-RE-U was already suggested by many linguists, yet no one was able to pin-point the origins of the stem. Now we have a plausible ancestor, for the first time!

It has to be mentioned though, that there are a lot of phonological ambiguities regarding both QA-SI-RE-U and QA-ZI. The Z-series probably expressed affricates (*ts), however, this is not a 100% proven fact. The interpretation of the Q-series is even more difficult: while they probably stand for *kw or *gw in the Mycaenean texts, no one can be sure of their Minoan values: this could theoretically be *kw, *gw, *g, *ḫw or even *ḫ. Therefore I will henceforth render these terms in my article with their traditional values (e.g. *qasileus instead of *gwasileus and *qazil instead of any other speculative value).

Going back to the Linear A tablets, the lists clearly support the important role of *qazil: for example, on HT96, more than 40 units of grain are noted (approx. 1300 litres, if measured by volume: a high quantity for a person compared to other tablets) as being donated by the *qazil. The first term in the sequence A-PA-RA-NE • QA-ZI-RA-RE specifies the circumstance or the recipient of this donation, and may possibly be connected with the theonym Apollon (who might have received gifts from the *qazil). HT131 also reports a considerable quantity (58 units) of grain as paid by the QA-ZI. On ARKH2 (see figure), a particularly concise list can be found: the first two entries probably refer to inhabitants of a place *Sidata, that returns in the second line in the form A-SI-DA-TO-I. It is speculative, but based on the form of toponyms found on jars (e.g. A-[WO]-KI-TA-A  vs. WO-KI-TA or A-TU-RI-SI-TI  vs. TU-RU-SA) it could denote an ethnic in a grammatically strange way (by the addition of an *a- prefix ('who'?) and a locative or similar suffix simulateously). It is the 3rd line where we see a combination of a personal name and the title QI-ZI.

One thing remains problematic, though: In linear B, we have the terms QA-SI-RE-U (*qasileus, 'governor') and QA-SI-RE-WI-JA (*qasilewiya, 'governance'), while Linear A shows the forms QA-ZI (*qazil) and QI-ZI (*qizil). The variation of the Linear A forms looks pretty regular and they are clearly not dialectal variants (e.g. both forms are attested from Khania, compare KH10 with KH88), just like their Linear B parallels. Now, which one is which? The troblesome question of assignment is fortunately eased by the context the terms are mentioned at. In Linear B, QA-SI-RE-U would normally attract a nominative case, if mentioned together with the name of the official (which is rare), while QA-SI-RE-WI-JA typically stands alongside personal names in genitives (as the example from Knossos: SE-TO-I-JA • SU-KE-RE-O • QA-SI-RE-WI-JA [As(2)1516] shows).

In Linear A, QA-ZI mostly stands alone, without a name (on all mentions at HT). Only a single tablet from Khania [KH10] mentions a longer statement, namely: I-PA-SA-JA • QA-ZI • A-KI-PI-E-TE. While I-PA-SA-JA could be an adjectival term of PA-SE - a word common at Haghia Triada (with the *i- prefix and a *-ja suffix added, resulting in a phrase 'that(the)-[PA-SE]-ian') - it is unlikely that this would refer to a place. Yet it could easily parallel the names standing in genitive as seen in Linear B. At the same time, the hapax term A-KI-PI-E-TE could be a toponym by the virtue of its elative *-(a)te suffix. On the other hand, the phrase QI-ZI regularly (always) stands alongside personal names of various endings, as another tablet-header from Khania [KH88] illustrates: QA-NU-MA • QI-ZI. Since there is no trace of any grammatical ending on those names, it is tempting to believe that QI-ZI was the title itself (= QA-SI-RE-U), while QA-ZI refers to the office (= QA-SI-RE-WI-JA). This is also what the Greek terms would suggest; so the person who led a *qazil was simply called *qasileus.

Our identification also has far-reaching ramifications regarding the origins of the Greek suffix *-εύς. Here, the opinion of the scholars is still deeply divided: some cling to it being a genuine Greek grammatical element, while others (especially Beekes) proposed it to be a loaned structure. But we now see that - while the stem of words in *-εύς might be of foreign origin - this suffix appears to be a normal part of the Myceaenean Greek (but not of the Minoan) language.

A very interesting parallel to the word βασιλεύς could be the stem of the name Ὀδυσσεύς. Although Classic Greek may also offer (a somewhat artificial) etymology for Odysseus, it is widely believed to be a Pre-Greek loanword. The theories on its origins have not yet reached a conclusive result. As for me, I find it interesting to compare the name Ὀδυσσεύς with the prehellenic place-name *Udweza - found in Linear A as U-DE-ZA or U-DWE-ZA (there were likely multiple towns by this name across the Aegean). The *-εύς ending could have been added by the Achaian Greeks, then. Yet it is nothing but a weak parallel - because frankly, we know nothing of the true meaning of *Udweza, and thus cannot fit it with the usual 'agentive' meaning of *-εύς. I can only hope that more examples of hellenized Minoan words will be uncovered in the future, to enlighten us in these matters.


  1. The -zil in qazil- and qizil looks suspiciously like the Etruscan root zil "to oversee; supervision". This might just be a coincidence (compare *nuzil "to feast"), but the semantic similarity is certainly intriguing.
    (Thanks go to Glen Gordon for the Online Etruscan-English Dictionary)

  2. It is definitely a possibility to compare this to Etruscan phrases, but the initial syllable remains hard to explain (and its alternation is even harder to give an explanation for). As for me, I also find the possibility of an Afro-Asiatic loanword enticing. For example, there is the Old Egyptian *wsr = 'strong' (see it here). And of course, it resembles "another" Wsr, the name Osiris, mythic king of all gods and mortals alike. Just keep in mind, that there was no r/l distinction in most dialects of Egyptian.

  3. This can be a finest article for this subject matter I've ever read. I'm really quite satisfied with it. Maintain running a blog!