Monday, 19 April 2010

sewer and island

Proto-Indo-European *akʷ-ā- "water" became Latin aqua. Combined with ex- "out", it became *exaquāria, becoming Old French sewiere "channel to carry off overflow from a fishpond".

*akʷ-ā- became Proto-Germanic *aǥwiō (AHD) or *ahwiō- (OED) then *aujō- "thing on the water". This became Old English īeġ "island". This combined with land to form īeġland - the s was added to the word sometime after the 15th century because it was thought to be related to isle.

*akʷ-ā- is also found in the name of the glacier covering the troublesome Icelandic volcano, Eyjafjallajökull. Ey is "island", from Proto-Germanic *aujō-. Fjöll is "mountain", from PIE *peli-s- and cognate with English fell "mountain". Jökull means "glacier" and is from PIE *i̯eg- "ice", and I think is exactly cognate with Old English ġicel "ice", which became the -icle in icicle.


Glen Gordon said...

I think you mean *h₁ekʷ-, not *akʷ-, considering Hittite ekuzi 'he drinks'.

goofy said...

hm... how do you explain the initial a in Latin and Germanic?

Glen Gordon said...

I screwed up and I realize now that I've been thoroughly confused by the controversy among IEists. Schrijver reconstructs *h₁eh₁gʷʰ- and Lubotsky is unsure between *h₁h₁kʷeh₂ and *h₂ekʷeh₂. Some deny the connection with aqua and reconstruct a Narten root *h₁ēgʷʰ- or *h₁ēgʷ-, thereby explaining Hittite ekuzi, Tocharian yok and Latin ēbrius instead.

Very confusing.

Whatever the case, *akʷ- requires a laryngeal onset according to modern IE theory.

goofy said...

So you can't have a word beginning with *a? I wasn't aware of that.