Re: Did ancient Greeks pronounce medial 'h'?

Carl W. Conrad (
Sun, 24 Nov 1996 06:33:31 -0600

At 12:49 AM -0600 11/24/96, Peter Wise wrote:
>Does anyone know if the speakers of ancient Greek pronounced the 'h' sound
>in the middle of words-even if they did not indicate it orthographically?
>(cf. BHQLEEM and GEENNA, etc.) It seems enigmatic to me that rough
>breathings occur only on initial letters. Did Greek have a 'h' sound only
>at the beginning of letters? Or are only the initial ones orthographically
>Does anyone have any light to shed on this?

This is a good question,and one that brings to light how much we've
absorbed from what we were taught without questioning it; ultimately you'll
need to consult the chief authority on the subject (Sullivan, _The
Pronunciation of Ancient Greek_ --or is it Sandford? My memory for names,
even those I know, gets fuzzier!). Although we've had many discussions of
Greek pronunciation on the list, I don't think this one has ever come up.

(1) My understanding of the situation is that the aspiration was definitely
HEARD in Chi, Phi, and Theta; the best evidence for this is that the Romans
added the H when transliterating Greek words containing them,e.g. QEATRON
--> THEATRUM, KAQEDRA --> CATHEDRA, etc., etc. Hence beginning Greek texts
that teach what is thought to be an accurate ancient pronunciation say one
should pronunciation Q like the TH in "hothouse," F like PH in "cuphook," X
like KH in "bakehouse." Another pretty clear proof that this is so is that
the spelling of compounds and contracted syllables is affected by the
juxtaposition of a stop with an initially aspirated syllable, e.g. KAQEILON

(2) I believe that the standard phonological understanding of the fate of
original initial and medial S is that they became H, that the initial H was
retained (e.g. SALS --> hALS, "salt", SEMS --> hEMS --> hEIS, "one"), but
that medial S passed through a stage of pronunciation of H which eventually
evanesced, leaving adjacent vowels in hiatus, which eventually were
contracted: ESEXON (impf. 1sg. of EXW) --> EhEXON --> E-EXON --> EIXON.

>While I am at it, does anyone have any insight on the rationale behind the
>development of smooth breathing marks?

The account of this that I have learned, whether or not it's valid, is that
the character "H" which retained its aspirate value in Western Greek
alphabets and entered thence into the Roman alphabet, was used in the Ionic
alphabet for Eta, but was also split into halves ( |-, -|), the first of
these being used to indicate aspiration ( |-ALS for hALS ) the other to
indicate NON-aspiration ( -|ALLOS for ALLOS ); then the two marks |-
and -| lost the upper half of the vertical: i- and -i (can I get away with
this ASCII representation?) and finally, the sharp perpendicular angles
became rounded into the cursive rough and smooth breathing marks that we
still use for writing ancient Greek. Of course it is true that the smooth
breathing never made any sense--why designate the "unmarked for aspiration"
at all? For this reason, I believe, the smooth breathing went out of use in
modern Greek spelling, and, since it is not pronounced in modern Greek, the
rough breathing was also dispensed with.

Bear in mind when reading this that it is a second-hand account of these
phenomena. I am reasonably confident about the pronunciation of initial and
medial sigma; but the account of the breathing marks is one I've accepted
without ever seeing the source from which it was supposedly derived.

Carl W. Conrad
Department of Classics, Washington University
One Brookings Drive, St. Louis, MO, USA 63130
(314) 935-4018 OR