14 Jan 2011

Back to 'back'

Concerning the etymological 'back' problem I've been having since December, I might have found a decent cure. I had elaborated before that it's long been known that there appears to be a common word for 'back' or 'hip' across ancient Greece and Turkey: Classical Greek ischíon 'hip-joint' and Hittite iskis- 'back'. This pair just can't be a coincidence and an underlying Proto-Aegean term *iskʰis(a) seemed like a plausible fit to me.

However, I kept on feeling that unlike the previous words I've suspected to be Proto-Aegean, this one comes across as a little extra odd. Firstly I can't find a way of analysing the expected Aegean root into smaller meaningful morphemes and secondly the structure of the root seems unexpected for Proto-Aegean (eg. *s in syllable codae). Yet I know that this is at least better than the horrid attempts by Indo-Europeanists to reconstruct *h₁isgʰís- based only on two items from a very restricted geographical area. Surely this can't be correct either.

It was a tough problem so I did some yoga, smoked a spliff, watched some TV and then once my mind was distanced from the problem, I experienced a profound synaptic event. I realized that my subconscious mind had been wrestling with that initial i- for some reason. I was slow to heed my inner eye telling me of a common Hittite pattern. There's a long list of Hittite words which are the products of prothetic i- breaking up original clusters of the *sC(C)- sort. For example, ispant- 'to libate' < *spend-. So why then wouldn't iskis- be approached by IEists this way too? I suspect the answer is disturbingly circular since if one is hell-bent to deny the probability of a Greek loan from Hittite and is equally determined to make this a common IE root at all costs, then one must reconstruct this silly onset, *h₁i-.

Brainstorm time! Let's start from scratch and try this again. We have a common Greek and Hittite term for 'back' or 'hip'. Let's now just assume that the Greek word is a loan from Hittite, leaving only a single term to play with. Let's also assume that the initial i- in Hittite is prothetic like these other words. This gives us a Pre-Hittite term *skis-. Let's analyse this term as a native s-stem like some other body part terms implying that it's built on a verb stem *skei-. It just so happens that there is an identical IE root *skei- 'to cut, split'. Now, if this term originally referred to the 'spine' then it indeed 'splits' the back into two halves. Thus Pre-Hittite *skei-s- > iskis- would be 'that which divides' or 'that which is divided'. I suppose then that an Aegean or Minoan intermediary is unnecessary if the loan happened towards the closing of the 2nd millennium BCE.


  1. Now, if this term originally referred to the 'spine' then it indeed 'splits' the back into two halves.

    Alternatively, it could refer to the rump, which is rather prominently divided into two cheeks.

  2. Yes, we humans are indeed split all the way down, aren't we? I don't know why but this reminds me of a quote from Chris Griffin on Family Guy:

    "Um, okay. You know Anna, when I first saw you, I thought you were the most beautiful girl in the world, and now all I wanna do is show you my inner most self, but I'm afraid you'll reject me because you won't like what you see. Or that you'll see my scrotum, and see that it has a seam on it, and then you'll think I'm made up of two different guys that were sewn together, cause that's what I think happened."

  3. This reminded me of the Spanish and Occitan esquina, which moved me to look up some etymologies that fall right in line with your hypothesis:


    It seems the Spanish and Occitan word (there is also the Italian schiena) derives from a Germanic source related to 'shin' and, further back to other words emerging from the concept of 'split'.

    Italian etymological dictionary entry:


  4. Nifty! It's also interesting that it doubly looks like a contamination of Latin spīna 'thorn; spine'.

  5. That seems a very plausible etymology. This means that there is no genetic link between ἰσχίον and ὀσφύς. Do you have any ideas on the etymology of ὀσφύς and ὄσχη as well?

  6. A rather odd word to to loan, but other than that, definitely nicer than a PIE reconstruction.

    Glad to see my suggestion forced you to go into deep meditation ;-)

  7. Hans: "Do you have any ideas on the etymology of ὀσφύς and ὄσχη as well?"

    It seems that ὀσφύς 'hips' has already been attributed to PIE *h₂ost-bʰu- which I assume would literally mean 'bone (out)growth'.

    I still must track down the history of ὄσχη 'scrotum'.

  8. Phoenix: "Glad to see my suggestion forced you to go into deep meditation ;-)"

    Nay, young grasshopper. Words cannot force deep meditation. One need only untie one's mind from the shore and let it drift down the river of thought like a loose twig carried along by the current. So says the lord Buddha. ;o)

  9. @ Phoenix: "A rather odd word to to loan,"
    If we assume that the original meaning was "backside", this could be either an expressive or euphemistic loan (depending on the status of the source language in relation to Greek). Or it could have been a medical term. For euphemistic or medical use, that would be comparable to the use of Latin names of body parts in English (e.g. anus, for a word in the same region).