Saturday, July 16, 2011

"My Lady Rhea"

It happens every day. A celebrated scientist discovers something, makes up a daring theory and finds initial proof. Then the theory begins to spread to all corners of the world. More and more scientists support it. Younger generations may take it as an axiom. Everyone is happy; and then it is suddenly disproven. In my area of expertise: molecular biology, I have seen countless theories born, flourish and finally crumble to dust, even in our lab. And no one is immune to such fallacies. In fact, I would be overjoyed to get 1/10 of my experiments to actually see publication. Similarly, I would be very happy if I knew that just 10% of the things I stated in this blog would stand the test of time. But without preconceptions, no research can be truely successful; and they say no pain, no gain. So here am I again to share some bits of my research with you.

Eteocretan language is a difficult subject to study. Partly because there are very few inscriptions known, most of them are fragmentary, and both the vocabulary and grammar are quite exotic. The hardships of understanding even bits of it even prompted some scholars to reject any connections between Minoan (i.e. the language underlying Linear A) and Eteocretan. This was an utterly unwise idea: after all, where could Eteocretan stem from if not from the earlier, Pre-Greek inhabitants of the Aegean? And indeed, I keep stumbling upon more purely Minoan phrases in Eteocretan inscriptions, the more I look. The very provocative title I gave to this post is none other than the preliminary translation of the Eteocretan phrase TUPRMĒRIĒIA. I shall show in my current post that each and every part of the phrase Tupṛ mē-Riēya is of Minoan stock, and - what is more - clearly attested in Bronze-age Cretan scripts!

There are less than a dozen Eteocretan inscriptions, and none of them are complete. All of them come from the eastern half of Crete, from Dreros and Praisos. It is likely that Eteocretan was already a language in decline by the time these texts were composed, as many are bilingual in Greek. But the Greek halves of the Drerian inscriptions are actually a great help to understand (at least roughly) the meaning of these texts, even the Eteocretan parts. Although I am referring to these texts as being bilingual, this does not imply that the two halves are always word-by-word translations.

My attention was drawn to one of the Drerian inscriptions (#2). These texts are all written in an archaic Greek alphabet, and thus easily legible. Unfortunately, this small piece of stone is weathered and the letters are heavily damaged. Only one line of the Eteocretan text survived: this ends with the phrase TUPRMĒRIĒIA. We know that this is a complete Eteocretan word, because the inscription does contain word-dividers. Dreros #2 is actually a bilingual inscription, the lower half being composed in Doric Greek. The Greek part is also damaged, but still largely legible; according to van Effenterre, it is a religious oath. The Greek formula does not contain the name of any deity; yet that might be present in the Eteocretan text, since they are not necessarily strict translations of each other (the Eteocretan one looks much shorter). After doing some research and extensive comparisons with Minoan finds, I came to the conclusion that the word TUPRMĒRIĒIA might be this missing invocation.

Let us consider the first half of this phrase. We may separate the part Tupṛ - as we know from other examples, that Eteocretan allowed syllabic sonorants. The reason to do so is that there exists a similar word in Minoan libation formulae, traditionally transliterated as DU-PU2-RE. There are many ambiguities with its reading: the precise value of the Minoan D-series is disputed (*d? *t? *th? *dh?), and we know that RE can stand for both *-re and *-le. But we also know that PU2 certainly had a special value, as it stands for *phu or *bu in Linear B, but never for a simple *pu. It is also possible, that the sound it used to mark was partially voiced in Minoan as well, explaining its use in the word reconstructed as *duphre.

In his fundamental article about the Minoan language, the Portuguese scholar Miguel Valério made a crucial discovery about the phrase DU-PU2-RE. Several Anatolian languages used similar words: *tapar(riya)- meant 'to rule' in Luwian and taparnas (or dabarnas) was the title of Hittite kings. The latter also comes in a variant laparnas (labarnas): both the stem and its mutations are similar to the one observed in Greek λαβύρινθος (vs. Mycaenean DA-PU2-RI-TO-JO • PO-TI-NI-JA, referring to a sanctuary near Knossos). Since the Labyrinth is consistently associated with Crete, we may safely assume that this word entered the Mycaenean language as a loan-word form Minoan. Thus it is not just possible, but outright expected to see related stems in Minoan texts. Therefore Valério's explanation for the phrase DU-PU2-RE as a form closely related or identical to Anatolian *tapar- (or *dabar-) is likely correct.

There is also an account of ancient Greek authors on the temple of Zeus Labraundos in Caria, citing that λάβρυς denoted "double-axe" in the local language. I do not know if there is any connection between labrys and the modern middle-eastern (Persian, Indian) term for battle-axe: tabar, but to compare a word for 'ruler' with 'axe' could be a conflation of similarly-sounding words from the side of the Greeks. Yet this question is currently far from being solved.

Given the roots of DU-PU2-RE, Valério gave a translation 'lord', 'ruler' or 'master'. This raw translation may now be refined further. The phrase DU-PU2-RE is not an independent word: it forms a part of compound phrases like JA-DI-KI-TE-TE-DU-PU2-RE. In this word, the first half is a geographic term: Mount Dikte. But in Linear B, the term DI-KA-TA is used in a much more restricted sense: it refers to a particular sanctuary on Dikte (maybe near the cave of Psychro). So (J)A-DI-KI-TE-TE-DU-PU2-RE is more likely a religious title and not referring to a king. There is also clear evidence, that both A-DI-KI-TE-TE (e.g. there is the phrase A-DI-KI-TE-TE-?-KE-RE on PKZa11) and DU-PU2-RE (see PA-TA-DA-DU-PU2-RE on HTZb160, where PA-TA-DA is another place-name) can combine with other words. Hence there can be no doubt of their separate nature.

We also know that Minoan probably had no grammatical genders. They even used the same base logogram for men and women: this would have been impossible in Mycaenean, and indeed, Linear B has separate signs for women and men. But if there was indeed no gender distinction in Minoan, then why not read DU-PU2-RE as 'lady' or 'mistress'? And it might not be a reference to just any lady, but to a goddess. Those who were hunting for divine names in the Libation Formula can now rejoice: after all, JA-DI-KI-TE-TE-DU-PU2-RE (*i-Adiktete-Duphre) may now be read as 'that of the Lady from the Dikte'. If some titles (especially divine titles) were used in the same form for males and females alike, that could explain the confusion of Egyptian scribes, and why they referred to Minoan goddesses *Amaya and *Raziya as male gods in the Keftiu-incantations.

In correspondence with the above concept, all inscriptions that contain the term JA-DI-KI-TE-TE-DU-PU2-RE also present a chain of hapax legomena (names?) right afterwards it. This helps to explain the uncomfortable situation that the "Mistress of the Dikte" stands on a place in these formulae, which is normally occupied by toponyms (clearly donors and not recipients). But I still have to concede it to Glen Gordon, that these expressions only work if we allowed the recipient to take an *-e ending, normally expected for a donor in the original theory. Thus the expression A-PA-RA-NE • QA-ZI-RA-RE [HT96] probably has *Aplan as a recipient (but literally, it is: "of Aplan, from the chieftaindom"): only this proposition would allow to identify that name with the Greek theonym Ἀπόλλων (or Roman Apollo).

Staying by divine names, this is where the Phaistos Disc comes into sight. We cannot make out much yet of its very unique-styled Minoan inscription. But one thing seems probable: certain words that are marked with an additional wedge, seem to be names. One of such terms reads as RA2-*07 (the sign Pha *07 cannot be read with any certainty). It also returns in the form MA-RA2-*07, where *MA- seems to be a prefix element (it is seen on other words on the Disc and - albeit very rarely - in Hieroglyphics). This prefixed form is actually very similar to the second half of TUPRMĒRIĒIA: which is probably *mē-Riēya. The correspondence of Eteocretan to Minoan could be pretty regular, if its vowels developed similarly to early Greek ones. The core stem (Rya-*07 on the Disc and Rieya here) also displays a high similarity with the classic Greek theonym Rhea. Two things are worth noting: first of all, the Cretan forms have a consistent *ry- cluster in their stem, what the Greek version lacks. Second, we cannot make out the value of that missing Pha *07 sign that easily, as the contemporary Egyptian rendering of this name: R-ṯ3-jj (perhaps *Ratsiya) warns us of a potential stop consonant in the original stem. This was probably lenitioned out and lost in later stages of the Minoan language, but might still be seen on Middle Minoan relics. I am not going into irresponsible guessings here, but the value of Pha *07 could be either 'TA', 'TI' or even 'SI'.

Prefixes are frequently seen in Linear A and seem to be an intergral part of the Minoan language. These might also be clitics or simply, irregularly-written short particles. Nevertheless, they are often difficult to interpret. Because prefixes combine freely with suffixes (e.g. locative, elative or ablative cases), any theory that seeks to explain these as a marker of just another regular case (say, dative) runs into a serious trouble. Based on extant languages that use both prefixes and suffixes at the same time on nouns (e.g. the Mesoamerican Nahuatl language, or certain Caucasian languages), a separation of roles is expected: if suffixes express directionalty and location, it is probable that prefixes would instead be pronominal in nature.

This is also what the study of potentially related languages hints at. While Minoan Linear A shows prefixes *i- (that looks like a generic deictic / connector) and *a- (that seems to refer to persons only), Etruscan has third person pronouns in the forms in (inanimate version, 'it') and an (animate version,'he'/'she'). Even so, the reading of rare prefixes, like *ma- remains uncertain. But if all prefix-like elements are indeed pronouns (which is a big assumption), then *ma- could plausibly be a first person possessive pronoun ('my') in an enclitic form. Note that verbs take personal markings as suffixes (e.g. KA-NI-JA-MI [CRZf1] or KA-NI-JA-SI [PKZa12] - both verbal forms of KA-NA [HT23], 'gift'), not as prefixes. As with all novel decipherment attempts, the reading of Tupṛ mē-Riēya remains unconfirmed: we still need more insight, especially into the nature of prefix elements before we can either confirm or reject this explanation.

I realized that I was not paying enough attention to a particular detail. Even if we posit a form *dabrwintha as ancestral to λαβύρινθος, we still have to count with two different, but related Cretan stems: *daphr- and *duphr-. While this was not impossible in Minoan (e.g. *qazil vs. *qizil), it adds an unnecessary level of complexity to the problem. By the same virtue, we may also suppose that these two stems were unrelated to each other. Note that other explanations also exist for *duphr-: for example, Glen Gordon has equated the Eteocretan tupr with the Etruscan word θuφ - which he proposed to read as "oath". Although I cannot say that the meaning of θuφ is certain to any extent, it is still insteresting to observe that θuφ also - very characteristically - stands as an epithet to the sky-god Tinia in one of his many roles ("Tin θuf").

Tuesday, July 5, 2011

Those "bloody" Minoans...

For my next post, we are going to discuss yet another popular topic: the forms and possible meaning of word A-SA-SA-RA-ME. It is also time for me to get my hand dirty, since I have never seen so many ill-fitting translations of Minoan phrases that was given for this poor little word. Now is the time to try and find better explanation(s).

Up to date most (all?) theories aspire to read (J)A-SA-SA-RA-ME as some sort of divine name. Apart from some (phonologically and semantically) really implausible explanations, there are two main problems with this theory. First of all, the term (J)A-SA-SA-RA-ME is very common: found on almost all objects of ritual context: on libation tables, on sacrificial vessels, on a statue and even on a silver pin, a total of at least 16 times. A little bit too common, if you ask me. Given the highly polytheistic nature of almost all bronze-age religions (as also evidenced by both Linear A and B tablets), this is not what we would expect if it were a theonym. The only religious term in Linear B that could match this high frequency would be PO-TI-NI-JA. This is, however, not a proper name, but a title ('lady', 'mistress') appended to the names of most if not all the goddesses invoked.

The other problem is the baffling word-formation observed on the stone vessel KNZa10, where a derived case: JA-SA-SA-RA-MA-NA can be read. The *-na ending seen here is suspiciously similar to the one seen on words expressing ethnic origin (Phraisona = 'from Praisos' in Eteocretan or PA3-NI-NA = 'from PA3-NI' in Linear A) and on innumerable Pre-Greek place-names (mostly hellenized to *-nos). This could be explained as a 'pertinentive adjective', also found (in the same form) in other Aegean languages, such as Etruscan. Such a derived form is not what we would expect if (J)A-SA-SA-RA-ME were a proper name. After all, the Greek priests would have given offerings to Zeus (or perhaps even of Zeus), but probably not Zeusian offerings. But it does not exclude a reading as a more generic term (even as an epithet). I am yet to see examples of a language that prefers to cite the very names of divinites in an adjectival form when speaking about sacrifices. If you have any good examples, I would appreciate if you could share them with me.

To this end, I looked up the word isḫassaras in the Etymological Dictionary of the Hittite Inherited Lexicon (here is a link for the online pdf version). This is the so-far best match in any other language, based on its form and meaning - if we stick with the original 'theonymy' theory. Isḫassara- is a compound stem, made up from isḫa- = 'lord' and the feminizing suffix -sara-, thus meaning 'lady'. None of its parts have a particularly good Indo-European etymology. But to derive A-SA-SA-RA-ME from this word, we have to conjecture a formative (*-ma). Does isḫassara- admit further derivations in Hittite? Much to my surprise, it does: we even have an adjective isḫassarwant- = 'lordly'. This could also potentially make our first theory work: after all, the sacrifices offered on the stone altars could have been 'lordly', 'noble' or 'divine' in a sense of either the donor (official), the circumstance (feast) or the recipient (divinity). To get this, we have to conjecture a chain of derivative suffixes on the stem of JA-SA-SA-RA-MA-NA (*-m(a)-na), not mentioning the pronominal prefix element *i- my readers are likely already familiar with.

The fact that a Minoan word can admit multiple formatives in a long chain is not an isolated phenomenon. Let us consider the word-formation in I-PI-NA-MI-NA-TE, a word seen on the sacrificial stone-vessel APZa2. This single word is enough to illustrate
the polysynthetic tendency of the Minoan language. The base stem appears to be *ip(i). Let's say (as a simple assumption - based on earlier considerations) that *ip meant 'blood'. Then *ipna would mean 'bloody' (adjective in *-na), *ipnama would be 'bloodletting' (*-ma ending: expressing action?), thus *ipnaməna 'bloodletting cup' (the same *-na formative again). and finally *ipnamənate(n) = 'from (this) bloodletting cup': a regular elative (suffix *-(a)te(n)). I wrote *-te(n) instead of just *-te, to connect this ending with the Classic Greek elative case in -θεν and the Hurrian ablative ending *-tan, as they could be related forms (due to a linguistic areal effect). Also, to explain the *-i- (*-ə-?) vowel, we have to keep in mind that many of the Minoan suffixes appear to intrude into the stem they fit on, deleting or re-colouring (i.e. *a→*i) any stem-vowels they collide with.

It is also possible that the -i- (*-ə-?) vowel was only inserted into the word I-PI-NA-MI-NA to make the *-mn- cluster easily pronounceable. In a number of cases, a (helper?) -i- vowel is seen, that has been deleted in others: the most famous example could be KU-NI-SU (= Knossos?): here, the methathesis is already seen in the Linear B version: KO-NO-SO (due to the spelling rules, this cannot be any other form, just Knossos). That is only possible if KU-NI-SU was also pronounced as *Kunəsu and even as *Kunsu in real life.

Before discarding this (highly hypothetic) derivation, one should also look at some Hittite grammar-books for parallels: the Anatolian stem esḫar- (= 'blood') is derivatized in a comparable way : although the formatives themselves are different, the result is fairly similar. Thus isḫarnumae- actually means 'to make bloody'. In Luwian, the related word: asḫarnummai- may translate similarly: 'be covered in blood'.

And thus we have arrived to the point to discuss a second theory about the meaning of A-SA-SA-RA-ME. It will be more in-line with the contemporary Minoan customs, but probably less pleasing to a faint-hearted reader. This possible explanation would be to compare A-SA-SA-RA-ME with the hieroglyphic Luwian word asḫarmis (plural: a-sa-ḫa+ra-mi-sa) = 'offering', 'sacrifice' (or similar). Hittitologists tend to connect this word with Luwian asḫar- = 'blood', thus *asḫar-m-is- originally meaning 'bloody sacrifice'. Whatever its orginal etymology was, it was used in a bit more generic sense in the Karkamiš inscriptions, since at least one of its mentions (see the figure) also involves sacrifice of bread, not just animals. Given the number of phrases in Minoan with possible Anatolian cognates, we should not be surprised to see yet another one added to the list. While the gemination of SA syllables is definitely problematic in Minoan (we must assume a development *-asḫa-*-asaḫa-*-asasa- upon borrowing - as Minoan might not have had the consonant ) and its stem-ending is different, a generic meaning 'sacrifice' would fit exceptionally well with *A-SA-SA-RA-M-. Should this identification be true, A-SA-SA-RA-ME could mean 'of sacrifice' and conversely A-SA-SA-RA-MA-NA 'sacrificial'. This could easily explain the universal use of these terms in religious contexts.

The fact that Minoans practiced animal sacrifices regularly, is well-evidenced by archeological finds in and around many sanctuaries. The public altar found in the courtyard of Gournia could have served a smilar purpose as Ian Swindale has suggested (and it might be true to the site of Mallia as well). The spectacular Haghia Triada sacrophagus also depicts such a scene on one of its sides (see figure): Here, the priestess - dressed in a ceremonial robe and a crown with feathers - collects the blood of the sacrificed cow into a conical vessel, quite similar to the inscribed stone cup APZa2. The blood is presumed as having been poured onto the altar-stone by the same three figures, as shown on the opposite side of the sacrophagus. The slaughter of animals was just a small part of religious feasts. The meat was likely roasted or cooked and was offered as a communal meal for all participants (gathered outside the temple - one thing the large squares in the Minoan city-centres were exceptionally good for). Ceremonies of this sort were commonplace in Classic Greece, where they lasted for multiple days, and encompassed processions, sacred chants and drama sessions (in theatre), sports competitions, etc. These festivities also appear to be very similar in core to the (pagan) Old English Blót.

There is also a "dark side" of Minoan customs we should not ommit the mention of. The deep discordance of Middle Minoan arcaeological finds (when animal sacrifice was common), and the Linear B archives of Knossos, that clearly avoids any mentions of bloody sacrifices, cries out for an answer. Because regular animal sacrifice was mentioned at Pylos, this must have been a specifically Cretan trend. One cannot dismiss the deduction that the avoidance of bloody rites was a kind of "rebound", in response to the morally repugnant practice of human sacrifice on Crete, rarely, but definitely seen in both Middle and Late Minoan archaelogical contexts. As in Pylos, this kind of action was likely a "last resort": only executed in times of great calamities. If there is any historical basis of Greek myths, in particular the legend of the Minotaur, we may assume that it were the early Greeks who put an end to the Minoans' unsavoury willingness to sacrifice young boys and girls to their gods if their outlooks on war were grim. But as it was just a far-flung extension of the annual, usual religious ceremonies involving animal slaughter and feasting, the Greek rulers might have opted to suppress these customs altogether. Though they probably did not die out, as the sacrificial scenes on the Haghia Triada sacrophagus suggest - this marvellous piece of Cretan art was clearly made under the late Mycaenean era (ca. 1370-1320 BC).