Problems in Bronze Age Aegean Studies

I have recently been working on a number of research projects, some ambitious in scope, and some extremely limited in scope, all of which are more or less centered around the Aegean in the late Bronze Age and early Iron Age. I will attempt to discuss some of the problems I have encountered while trying to work on them. These problems confound our ability to unravel history, but ultimately, I have found them rather interesting in and of themselves, and perhaps worth sharing.

For some context, in 1997 Martin Litchfield West, a classicist by trade, and a genius in my eyes, published a book, building, in a fashion fit for Babel, on earlier, rather earth-shattering discoveries of his on Hesiod, called “The East Face of Helicon” (henceforth “EFH”), wherein he detailed a colossal number of correspondences, both in language and theme, between Near-Eastern poetics, eg. Hittite, Akkadian etc., and those of the archaic and classical Greeks, eg. Homer, Hesiod, the lyric poets, and even the Tragedians. Unlike those building the tower of Babel, he was largely successful in demonstrating an irrefutable current of influence, flowing from the old literature of the Near East into the new literature of the Greeks, as if the Tigris and Euphrates themselves had turned back on their heels, flooded up and over the highlands of eastern Anatolia, and washed over the great saline plains of central Anatolia, and finally joined with the Gediz, emptying out into the Aegean with all their rich jetsam.

The question I have asked myself, and the question many others have asked themselves, ever since I acquainted myself with his work and subsequent studies by others in the same vein, is “how did this happen?"

Various theories have been proposed, of course. West himself suggests a true smorgasbord of options, though he favors a Phoenician intermediary.¹ Mary Bachvarova has, more recently, suggested Anatolian bards.² I myself have vacillated in what seems most plausible, and likewise have attempted to make my own Westian comparisons where I have seen some sort of affinity between East and West.

In the Bronze Age Aegean, however, nothing is ever simple or easy. For example, when West composed EFH, it was not yet clear that, for example, we had an apparently Greek polity in southeastern Anatolia, probably set up by some form of Greek pirates, which later came to write its inscriptions in Luwian, an Anatolian language, and Phoenician.³ This again would add another possible route of transmission, for even Herodotus tells us that even in the time of the Persian Wars, the Cilicians, ie. the people of this very region, had taken after Phoenician customs and name but were previously known as “Hypachaeans,”⁴ that is, “Sub-Achaeans”.

And interesting problems are presented here too. For this state was named “Hiyawa”, on the basis of all available evidence, by an aphaeresis of “Ahhiyawa”, the Hittite term for “Achaea.”⁵ In Herodotus, this “hypo" might simply mean that we should understand these people as having their ethnogenesis by way of the Achaeans, but it occurs to me that there could quite plausibly, though it is somewhat less likely, be formed from folk-etymological reanalysis within Greek of some later, inherited form of “Hiyawa.”

Folk etymology is the process, in this case, where an obscure or foreign word is re-analyzed in terms of more familiar terms or morphemes in use in the language in question. A decent example might be the spelling liquorice for licorice. This is actually a folk etymologizing which perhaps goes back to (Late) Latin, which borrowed Greek γλυκύρριζα (glukurrhiza, i.e. ‘sweet-root’), but re-analyzed it according to the Latin word liquor ‘fluidity’, and thus the spelling was changed. Of course, even in this case it is probably more complicated than that. The spelling in question may just as easily have been a scribal aberration, and the influence of French orthographic practices and of course the English term “liquor” have all contributed at various stages. Indeed, this helps illustrate the sort of headaches we might encounter with such problems.

This is contrast to the “Neogrammarian” theory (a theory like gravity is a theory) of “exceptionless sound change”, that is, that sounds change across time in a language consistently, across all words, all at roughly the same time, at least where words have been inherited. Here, we are dealing with borrowings — which complicates the picture somewhat, but as we will see, even in deeply ancient borrowings, consistency prevails.

Folk etymologizing is quite a common theme, particularly in onomastics, that is, the study of names. Toponyms in particular are vulnerable for all the reasons one might expect: for one reason or another, the common language of some town or places changes, and yet the old name for the town remains. Yet, at a certain point, no longer does any resident or local speak the language whence that toponym came and people, especially if they are not literate although that is hardly a requirement, start to get funny ideas about what that place’s name actually is, or perhaps what words it comes from. There are some great examples of this in American toponyms. For instance, “Key West”, originally in its Spanish colonial period, but even today in Spanish, is Cayo Hueso “Bone Cay.” It was, however, re-analyzed as “Key West” in English, because Hueso bore a resemblance to “West”, and indeed, it lies on the western extreme of the Florida Keys.⁶ A much more arbitrary example would be that the Purgatoire River of Colorado, so named by French trappers, was known by the locals in more recent times as the “Picketwire.”⁷

The reason for such emphasis on this concept of reanalysis is that the turn-of-the-millennium Aegean, and especially Anatolia, were not terribly different from colonial North America. There was an abundance of languages active in both places, some more transient than others. And yet, to build any coherent bridge between the literature of the Near East and received Greek mythology, one must connect names like Ahhiyawa and Achaea, or Ilion and Wilusa.

But these are relatively easy. Not every phoneme lines up perfectly, but they’re quite explainable. There are slightly more difficult equations which are now largely accepted, such as Asia=Assuwa, which are less clear⁸ — how does one explain, if Mycenaean Greek here borrowed from Hittite, the change of u > i? One can really only guess. Since we have Mycenaean a-si-wi-ja, perhaps the vocalism changed to better accommodate the very common Greek suffix -ia (to be written i-ja in Mycenaean), but then this doesn’t quite explain why a-si- and not a-so- or similar. Perhaps it was a matter of orthography, since the Linear B writing system was limited to open syllables, sometimes where a consonant cluster existed, the “vowel” would be duplicated in the writing, so we would have /aswija/.⁹ One might think this gets us closer, but intervocalic -sw- is regularly lost in Greek, so if this were the case, we would call our largest continent “Aia” instead of “Asia”. Of course, one then thinks of the mythical island of Circe in the Odyssey, Aiaia, and her brother, king Aiētēs, and his kingdom Aia, which is interesting, though probably coincidental — but now we are far from the matter at hand. Luckily, the loss of sibilants, that is, s-sounds, almost certainly predates Mycenaean Greek. So, if the term was borrowed after this sound change, we should never have needed to bother with Aia for even a moment. However, we have, at this point, forgotten where we even began or what problems we set out to solve.

Let’s look at this from another angle. To return to reanalysis, perhaps Assuwa was reanalyzed in light of Greek asis ‘mud.’ To think of “Asia” (properly at this time, merely Western Anatolia), as the “muddy land” seems like a more reasonable misunderstanding than the “Picketwire River”. We might say “but even asis is not clearly inherited, and may have an Asiatic source and it’s not clear why we’d get a-si-wi-, but OK.” Then we go on and think a little more, and stupidly, for we realize “now a synonym for āsis is īlūs…” And now, we take note, after some research, of the vowel length there, “Now that’s very interesting, the initial i is long, that would be a regular outcome of something like wil-… like Wilusa? Could this have influenced the outcome of ‘Asia?’”

And now you’ve gone completely insane. Because there are decent cognates in other Indo-European languages both for Aiaia (not to mention that here any such connection was spurious from the outset) and īlūs. And those of īlūs contain no /w/ to speak of. “And yet”, you might even think to yourself, “these might only be chance.”

It is at precisely this point that you should be institutionalized. Furthermore, they should throw away the key, for your inversion of the principle of parsimony is now outrageous to public morality at best, and violently dangerous at worst. And you haven’t even finished explaining a basic equation, hardly controversial, though its motivations remain unclear, such as Assuwa > Asia, a safe assumption we can make on the basis of geographical, archaeological, comparative, and indeed, supplied with the right chronology of sound change and Mycenaean onomastics, linguistic evidence.¹⁰

Then there are more controversial problems. Mount Sipylus appears to be in the same area as the the Hittite Mount Zippasla (probably realized /tsipasla/). There is a resemblance here which is impossible to ignore, and yet the possible equation is controversial. One might at first think that the regular change of -sl- to -ll- intervocalically in Greek should help bring us closer, but Lejeune believes such a change occurred before Mycenaean,¹¹ so a borrowing at that the stage of Mycenaean may not have obeyed such a sound law. Yet, as he points out,¹² any such cluster of a sibilant and liquid would be impossible to express in Linear B, and the same goes for gemination, as the script, as we have already mentioned, with very limited exceptions, only had the capacity to express, again, open syllables of the type CV.

In fact, Neogrammarian exceptionless sound laws cannot get us terribly far in this matter, because, while it could be a complete coincidence that Sipylus and Zippasla happen to be in the same area but are either slightly different mountain peaks or indeed refer to the same mountain accidentally, Sipylus is not exactly clear in etymology either. To add to our headaches, we find it difficult to escape the problem because Zippasla was the seat of a Hittite province or vassal state, and Sipylus is where Tantalus is said to have ruled (more on this later).

There is, however, the fact that the second component in the Greek looks like a Greek placename, Pylos. Pylos is derived from pulē ‘gate’ (though even this word has an unclear etymology), and we know it was quite the place-to-be in the Mycenaean period. Whether from Pylos or pulē, one explanation is that Greek speakers, once they had arrived in the area, at some point came to call the mountain Sipylus. Perhaps the -sl- cluster evaporated first, as it would have been an unfamiliar cluster to these Greeks, and perhaps the vocalism changed by folk etymology as time went on. It’s hard to say, but it leaves us rather more satisfied than considering the matter coincidental.

What of Tantalus then? Well, West suggests, tentatively though tantalizingly, that we might derive his name from the name of various Hittite kings named “Tuthaliyas”.¹³ Plato in the Cratylus suggests that Tantalus’ name is to be derived from the adjective talās (superlative: talāntatos) “wretched”, but of course one can be quite sure that an unexplained metathesis whereby talāntatos > Tantalos (n.b.: there is no long a in the latter term) is quite unlikely. Yet, on the other hand, we are presented with very many problems with the derivation from Tuthaliyas (perhaps pronounced something like /Tɔtɣalijas/). On the whole however, it is not even clear that the tool of re-analysis can help us much here, as, as far as I am aware, there is no Greek lexeme besides perhaps the one Plato adduced (which only clumsily resembles Tantalos), which could explain fully the transformation. However we can be quite sure that a cluster like /tɣ/ would be fairly unfamiliar to the Greek-speakers of this time. Furthermore, [ɣ] is the sort of phoneme which could easily be left unnoticed by a non-native speaker in a cluster such as this. It is quite subtle and historically in other languages frequently is weakened to a glide of some sort (compare English yield with German geld or yellow with gelb.)

To return to re-analyzed rivers, we can confidently adduce the modern Sakarya river as a comparandum, which in Greek was apparently Σαγγάριος (Sangarios), and in Latin, curiously, Sagaris. The Latin is curious because it is much closer to the Hittite Saḫiriya (pronounced /saɣirija/), without the intrusive nasal, and one does wonder if this preserves an older form. However, given that this spelling with the nasal is already there in Homer, unless we are to suppose that this nasal is somehow an interpolation, it would seem extremely unlikely that the Latin is preserving anything here. Greek orthography stipulates that a doubled gamma is to be read as a voiced velar nasal, but one wonders if there is some funny business here. In any case, if West’s etymology is correct, we of course do see an intrusive nasal in Tantalus, much like Sangarios, which is interesting. At the same time, the very few unambiguously borrowed toponyms with the -h- have a gamma replacing it.¹⁴ For the Turkish, I am no expert but sakar in Turkish is an adjective meaning unlucky or clumsy, which may have been applied to the Sakarya by reanalysis, especially since it is a river which makes a huge pair of >180 degree turns in its course from the Bayat Plateau to the Black Sea. At the same time, sakar has a dialectical meaning of ‘cliff’ which is perhaps even better fitting, so at least in the Turkish case, we have some idea of what was going on.¹⁵

One wonders whether much of a science can be made out of this whole business at all. The ghost of reanalysis reduces our historical linguistics at times here to a folk etymology of its own sort. “Tuthaliya vaguely looks like Tantalus and sort of vaguely semantically works so perhaps they’re related” — is this really any better than Plato’s etymology? Perhaps, though only because we know that this is how other names of this sort are transmitted, and there’s no real precedent for the transformation which Plato would ask us to believe. Yet, in Plato’s own time, there may well have seemed precedent, since, as we’ve just alluded to, the process of finding etymologies was hardly distinguishable from our own etymological practice of assuming folk etymology. That is, folk etymology as an explanation for how folk etymologizing can change (qualified in our time by borrowed) words can become in and of itself a sort of folk etymology.

Why not, then, also connect a name like Tuthaliya with others. For instance, Teucer (Teukros), the legendary king of Dardania in the Troad, might we not make a similar argument for him, perhaps even more cleanly? We might speculate that in some intermediary, the l/r distinction was weak, stress dragged the latter vowels into the abyss of syncope, and here the semantics are quite strong. It is from the union of Teucer’s daughter Batea and Dardanus that the entire Trojan (and thus, perhaps, Anatolian, as far as the Greeks of this time knew) race sprung. His name has been associated with the Anatolian storm god Tarhunt via some language (my money would be on Lycian trqqas), but still we’d need some sort of metathesis. Or we might even be bold enough to bring Deucalion, the survivor of “The Deluge” in Greek mythology, into the mix. His name’s phonology is nigh a perfect match for Tūthaliyas, especially if we handwave a bit on the voicing since it’s quite unclear what exactly was happening there in Hittite, anyway. All we need is the quite justifiable assimilation of *-tg- > -kk- or some such, and it all becomes quite manageable. The only problem there is the semantics. Was Tuthaliya’s name linked with Deucalion because this was the oldest name people could remember? Was the flood story transmitted from Hittite and thus carried Hittite names? The latter seems more likely, and I have some additional notes I’ll share in the future on this.

Nonetheless, it is clear that there are plenty of onomastic equations which are now more or less a matter of consensus, even where phonology is not immensely favorable. For such examples, we might consider Apasa=Ephesus and Miletus=Millawanda,¹⁶ while Lazpa = Lesbos is not even necessarily straightforward, and yet is quite accepted indeed.¹⁷ Then there are the examples where a sound knowledge of the corpora and phonological development yields quite perfect correspondences. Asia=Assuwa is one we’ve already discussed, but there is also the satisfying Mopsus=Muksu- (via Myc. Greek /⁠mokʷsos⁠/) and Pegasos=pihassi-.¹⁸ I could extend this list for some time. Yet, across all of these, the similarities far outweigh the differences, and we seem to be able to observe some sort of a/e alternation for some, for example. It’s not even particularly clear what role folk etymology or reanalysis actually plays in the most secure of our examples. One desires to make a science out of the whole thing, and accordingly, one must either reject or bring overwhelming evidence to the table when it comes to more exotic formulations. Yet, we occasionally arrive at a name like Sangarios=Saḫiriya which entirely defies the usual explanations. Zippasla is another, where I am quite convinced it can only refer to Sipylus. This is all quite frustrating, and yet, of great importance to our understanding of history.

Which makes it all the more frustrating indeed.

---

Bibliography

AhT = Beckman, Gary M., Trevor R. Bryce, and Eric H. Cline. The Ahhiyawa Texts. Writings from the Ancient World 28. Atlanta: Society of Biblical Literature, 2011.

AI = Collins, Billie Jean, Mary R. Bachvarova, and Ian C. Rutherford (eds.). Anatolian Interfaces: Hittites, Greeks and Their Neighbours. Oxford: Oxbow Books, 2008.

EFH = West, M. L. The East Face of Helicon: West Asiatic Elements in Greek Poetry and Myth. Oxford: Clarendon Press, 1997.

PH = Lejeune, Michel. Phonétique historique du mycénien et du grec ancien. Paris: Klincksieck, 1972.

Bachvarova, Mary R. From Hittite to Homer: The Anatolian Background of Ancient Greek Epic. Cambridge: Cambridge University Press, 2016.

Cline, Eric H. “Aššuwa and the Achaeans: The ‘Mycenaean’ Sword at Hattušas and Its Possible Implications.” Annual of the British School at Athens 91 (1996): 137–151.

Hawkins, J. David. “Tarkasnawa King of Mira: ‘Tarkondemos,’ Boğazköy Sealings, and Karabel.” Anatolian Studies 48 (1998): 1–31.

Mason, Hugh J. “Hittite Lesbos?” In AI, 57–62.

Nikoloudis, Stavroula. “Multiculturalism in the Mycenaean World.” In AI, 45–56.

Oettinger, Norbert. “The Seer Mopsos (Muksas) as a Historical Figure.” In AI, 63–66.

Yakubovich, Ilya. Sociolinguistics of the Luvian Language. PhD diss., University of Chicago, 2008.

1. EFH 629.

2. Bachvarova, 2016

3. AhT 265ff.

4. Herodotus, 7.91

5. Yakubovich, 2008, 191f.

6. https://fcit.usf.edu/Florida/docs/k/keys03.htm

7. https://edits.nationalmap.gov/apps/gaz-domestic/public/gaz-record/201784

8. Cline, 1996

9. In the course of writing this, I found that this indeed is what the evidence points to cf. Nikoloudis 2008, 48.

10. ibid., of course!

11. PH §112

12. ibid.

13. EFH 472-473

14. Sakarya is one. Another secure example would be Parha > Perge.

15. Nişanyan, Sevan. "Sakar." Nişanyan Sözlük. https://www.nisanyansozluk.com/kelime/sakar

16. Hawkins 1998, 1-2.

17. Mason 2008, 57–62

18. Oettinger 2008, 63–66.