Thursday, 14 February 2008

sambar, bring, metaphor

Idli sambar is a yummy Indian dish sometimes eaten for breakfast. I'll leave the recipe to the cooks and just discuss the etymology. The word for the rice cake seems to be of Dravidian origin (Tamil இட்லி iṭli), while the word for the lentil stew, sambar, is from Tamil சாம்பார் cāmpār, borrowed from Prakrit saṃbhārei "to gather", related to Sanskrit संभारयति saṃbhārayati the causative of संभरति saṃbharati "to bring together". This verb is a combination of sam "together" (from Proto-Indo-European *sem- "together, with") and bharati "to bear, support" from PIE *bʰer- "to carry; to give birth".

*bʰer- combined with *h₁neḱ- "to bring", to form *bʰrenk- (presumably after the centum-satem split? I don't know where the laryngeal went), which became Proto-Germanic *ƀrenǥan, *ƀranhta, then Old English brinȝan, brōhte, ȝebrōht, then English bring and brought.

*bʰer- became Greek φέρω pherō "to carry", which combined with μετά meta "with" (from *me- "in the middle of") to form metaphora "transference, metaphor" - that is, "carrying across".

*bʰer- gives us many other words:
fertile, prefer, differ from Latin fertilis, præfero, differo, all from fero "to carry"
bear from Old English beran "to carry"
birth from Old Norse byrð
fortune from Latin fortūna from the suffixed zero-grade form *bʰr-tu-.

In other news, I talk about the English subjunctive on Write Wow.


Glen Gordon said...

Whoa, I see an issue. You say that *bʰer- 'to bear, to carry' combined with *h₁neḱ- 'to bring' forms *bʰrenk-. That doesn't make sense for a number of reasons, so let me ask you some hard-hitting questions! :)

Where did the laryngeal go? We should expect **bʰr̥h₁énḱ- if this were true, unless perhaps this is some sort of pre-IE phonotactic reshuffling in verbal stems? Regardless, this "reshuffling" would have to be explained. Are there other examples of verbs being strung together in this way, one directly after the other, in PIE? And we can't just switch one velar stop for another (*ḱ with *k). Surely you forgot the traditional "palatal" *ḱ in what you write as *bʰrenk-.

Sorry I'm nitpicky, but I get a sick thrill out of these details.

goofy said...

Let me link to where I got this info from:
Compound root *bhrenk-, to bring (< *bher- + *enk-, to reach; see nek-2)

I left out some details, so thanks for keeping me on my toes. Since I'm not an IE scholar I'm basically trusting Watkins. If you disagree with this analysis, I'd love to know what you think!

goofy said...

I should add that I actually got the info from the actually print copy of the American Heritage Dictionary of Indo-European Roots; the online version does not include as much detail.

mahendra singh said...

If only I had some sambhar and idlis right now, were it ever so 'umble I shouldn't mind, and with that nice green chutney and I would be ever so grateful. Or a dosa would be nice.

Thanks for whetting my appetite for things that cannot be, alas.

goofy said...

OK, there seems to be a mistake on the online AHD. The entry for *bher- links to *nek-2 "to reach, attain", but that is the wrong root. It's actually *h1neḱ- "to bring", variant form *nek, and it's not in the online AHD. FWIW, both are referred to Pokorny eneḱ 316, but the printed AHD treats them as two separate roots.

Yes, I could go for some idli, sambar and coconut chutney about now.

Glen Gordon said...

The same etymology of *brenk- is mentioned in Lehmann's book as well: here. You'll notice at the top of page 80 though, an alternative view is mentioned: "More likely, PIE bher- carry + -en- with velar suffix." Now that sounds more natural for PIE than agglutinating two verbs together like this. I would think that it's derived from *bʰer- by way of a velar extension first (*bʰreḱ-), then secondarily enlarged with the well-attested nasal infix *-n-.

But whatever. All this talk makes me want to boil myself a soothing pot of chai. Mmmm.