I’ve been reading Anne Carson’s Eros the Bittersweet: An Essay (Princeton, 1986) the past few days as something to reflect upon in the evenings when I’m tired of doing linguistics, and also with the secondary ambition of finding ways to articulate why it is that I love Greek melic poetry and make it relevant to others by finding other peoples’ opinions on it that I can relate without having to come up with original thoughts on it myself. (I am, after all, by no means a literary or cultural theorist.) One of the ideas she elaborates in this essay that I like very much looks at ἔρως as a problem of demarcating and defining boundaries, edges, and the tension at these borders between self and the other that would only be annihilated by their very transgression. Her section ‘Finding the Edge’ sketches this idea out in a beautiful way:
Eros is an issue of boundaries. He exists because certain boundaries do. In the interval between reach and grasp, between glance and counterglance, between ‘I love you’ and ‘I love you too,’ the absent presence of desire comes alive. But the the boundaries of time and glance and I love you are only aftershocks of the main, inevitable boundary that creates Eros: the boundary of flesh and self between you and me. And it is only, suddenly, at the moment when I would dissolve that boundary, I realize I never can.
Infants begin to see by noticing the edges of things. How do they know an edge is an edge? By passionately wanting it to be. The experience of eros as lack alerts a person to the boundaries of himself, of other people, of things in general. It is the edge separating my tongue from the taste for which it longs that teaches me what an edge is. Like Sappho’s adjective glukupikron [γλυκύπικρον], the moment of desire is one that defies proper edge, being a compound of opposites forced together at pressure. Pleasure and pain at once register upon the lover, inasmuch the desirability of the love object derives, in part, from its lack. To whom is it lacking? To the lover. If we follow the trajectory of eros we consistently find it tracing out the same route: it moves out from the lover toward the beloved, then ricochets back to the lover himself and the hole in him, unnoticed before. Who is the real subject of most love poems? Not the beloved. It is that hole.
When I desire you a part of me is gone: my want of you partakes of me. So reasons the lover at the edge of eros. The presence of want awakens in him nostalgia for wholeness. His thoughts turn toward questions of personal identity: he must recover and reincorporate what is gone if he is to be a complete person. The locus classicus for this view of desire is the speech of Aristophanes in Plato’s Symposium. Here Aristophanes accounts for the nature of human eros by means of a fantastic anthropology (189d-93d). Human beings were originally round organisms, each composed of two people joined together as one perfect sphere. These rolled about everywhere and were exceedingly happy. But the spherical creatures grew overambitious, thinking to roll right up to Olympus, so Zeus chopped each of them in two. As a result everyone must now go through life in search of the one and only other person who can round him out again. “Sliced in two like a flatfish,” says Aristophanes “each of us is perpetually hunting for the matching half of himself” (191d).
Most people find something disturbingly lucid and true in Aristophanes’ image of lovers as people cut in half. All desire is for a part of oneself gone missing, or so it feels to the person in love. Aristophanes’ myth justifies that feeling, in typical Greek fashion, by blaming the whole situation on Zeus. But Aristophanes is a comic poet. We might look, for a more serious exegesis, to more serious lovers. A feature of their reasoning will at once strike us. It is outrageous. (pp.30-31)
Surely anyone who has been in love cannot deny the truth in these observations. Be that as it may, I would however like to turn to a more technical, perhaps dispassionate, linguistic angle that may even lend some support to this analysis of ἔρως as a problem of boundaries.
Some years ago, in an influential study published in Harvard Studies in Classical Philology (Vol. 98, 1998 pp.31-61; article on JSTOR), Michael Weiss attempted to etymologize the three principal words used in Ancient Greek for ‘desire’: πόθος, ἔρως, and ἵμερος. Weiss takes Pausanias’s description of three sculptures in the temple of Aphrodite at Megara depicting the three, although at the time of Pausanias’s writing in the second century AD, these three concepts seem to have become synonymous in the Greek language, as he remarks (1.43.6) Ἔρως καὶ Ἵμερος καὶ Πόθος, εἰ δὴ διάφορά ἐστι κατὰ ταὐτὸ τοῖς ὀνόμασι καὶ τὰ ἔργα σφίσι. “[Scopas made] Eros, Himeros, and Pothos, if indeed there is a difference according to the function of each by name and in their deeds.” Although it would be a bit much to summarize the etymologies reached for all three, Weiss’s proposal for the etymology of ἔρως is ingenious. Ἔρως traditionally has not had an etymology listed in any of the classic Greek etymological dictionaries, such as that of Hjalmar Frisk (1960-1972) ‘ohne Etymologie’ or Pierre Chantraine (1968-1980) ‘inconnue’. Regrettably Weiss’s proposal is also overlooked in the newest (2010) Etymological Dictionary of Greek by Robert Beekes (s.v. ἔραμαι) as well. The morphological facts of the complex of vocabulary relating to ἔρως however strongly suggest (Weiss goes on so far to say ‘guarantee’) an Indo-European origin for them. There is an athematic medio-passive verb ἔραμαι, o-stem noun ἔρος, -ου, t-stem noun ἔρως, ἔρωτος, various adjectives ἐραννός, ἐρατός, ἐρατεινός, all ‘lovely’, ἐρόεις, -εσσα, -εν ‘having ἔρος, charming’, and the name of the muse Ἐρατώ. Circumambulating all the technical argumentation within the article, the facts conclude that ἔραμαι must be from a hypothetical Indo-European root *h₁erh₂- (pp.36-40).
The semantics are less obvious, as it is not always clear where verbs for ‘love’ originate, although through a comparison with the development of two other verbs for ‘love’ in other Indo-European languages, Lat. deligo and Skt. bhajate, Weiss makes the case for a possible semantic development from the idea of ‘divide for oneself, take one’s share’ to ‘enjoy’ (p.40ff.). Using these semantics as a potential parallel, Weiss then goes to try to find comparative evidence for the hypothetical root *h₁erh₂- with a meaning ‘divide’ (p.42ff.). There is, in short, seemingly sufficient external evidence to support a reconstruction of a root *h₁erh₂- in Indo-European with that meaning via Hittite arḫāš ‘border’ (Cuneiform Luwian irḫa-, Hieroglyphic Luwian irha-), Latin ōra ‘border, brim, edge, margin’, Old Irish or ‘border’ all as nominal reflexes, and Lithuanian ìrti as a primary verb. Admittedly, I have greatly simplified his argumentation in summary and is only one small part of a much longer article, but between the semantic investigation and the formal comparative reconstruction, Weiss gives, I think, a persuasive argument for the etymology of the complex of vocabulary surrounding ἔραμαι / ἔρως etc. as that which ‘divides’, and so one desires that from which one is divided. This etymology, reached from as scientific grounds as one can get in linguistic analysis, also fits in nicely with Anne Carson’s views on Ἔρως as a problem of ‘boundaries’ as discussed in the quotation I’ve reproduced from her essay above.
Postscript: I do encourage people to read both Carson’s essay and Weiss’s article. They are illuminating. I first read “Erotica” as an undergraduate (date of retrieval from JSTOR on my .pdf of it states Tue Dec 18 2007), and it quite understandably confused me quite a bit back then as it is a very technical piece of Indo-Europeanist literature. But as I’ve come back to revisit it over the years, I’ve come to understand and appreciate its genius ever the more, not just out of understanding the technical details, but also personal experience with the semantic argument that, perhaps, seemed somewhat further fetched when I was younger, seems less and less so as time goes on.