L2/05-241

Subject: "Old Anatolian scripts"
Source: Michael Everson
Date: 2005-08-31 17:17:56 -0700

As part of the action Debbie Anderson and I have on Lycian, Lydian, 
and Carian, I've been trying to put together a chart that "unifies" 
the three.

It just doesn't work.

Lycian has a triangle like DELTA which is used for /d/. This is /l/ 
in Carian. The Lydian /d/ looks like Runic CEN.

The letter answering to DIGAMMA is /r/ in Carian, /v/ in Lydian, /w/ in Lycian.

The letter answering to LAMBDA is /b/ in Carian, /l/ in Lydian, /l/ 
in Lycian, though their shapes are different.

The letter answering to QOPPA is /t/ in Carian and absent in the other two.

Each of the scripts has a whole series of additions which don't match 
each other.

Lydian is strong RTL as can be seen in Figure 2 of L2/05-102. Having 
said that its first 19 characters are fairly equivalent to the Greek 
prototype and the last seven are unique additions. 26 characters in 
all.

Lycian is strong LTR. It more or less follows the Greek model, but 
inserts new characters into the sequence after the characters 
answering to ALPHA, BETA, KAPPA, NU, and RHO and then has 5 additions 
at the end. 29 characters in all.

Carian has 45 identified characters, 17 (or 19?) of which are not 
deciphered. The values of all of them are completely unrelated to the 
Greek prototype. While ALPHA is /a/. GAMMA is /d/, DELTA is /l/, 
EPSILON is /u/, DIGAMMA is /r/, ZETA is [ly], THETA is /q/, LAMDA is 
/b/, NU is /m/, OMICRON is /o/ (amazingly), QOPPA is /t/, RHO is 
/sh/, SAN is /s/ (amazingly), and the rest are additions, apart from 
PHI which is /ny/ and PSI which is /n/. Carian in Egypt follows the 
RTL directionality of Egyptian (but we will encode Egyptian as LTR) 
and in Caria it's LTR anyway.

Moreover, even the canonical shapes really aren't shared for an awful 
lot of the "matches" anyway.

The glyphs in my three documents were just rectified Everson Mono 
forms based on code charts, which may account for more superficial 
"equivalence" than actually occurs.

Trying to draw up unification tables has driven me to distraction. 
The puzzle pieces simply don't fit. This is completely different from 
the Old Italic unification, which was trivial. All the pieces fit.

Lydian and Lycian are similar, but they have different 
directionality. Carian is not in any way unifiable with these by any 
criterion that I recognize as sensible.

All of the existing documentation I have seen treats the three 
scripts as unique (something which was not true of Old Italic, where 
comparison charts were common). I have yet to see, in the literature, 
a comparison chart for Lydian, Lycian, and Carian. As I have just 
tried -- and failed -- to make one that was meaningful, I can 
understand why.

I will proceed now on the assumption that the Action is completed, 
whether or not we have done the exercise of asking the experts if 
they think an "Old Anatolian script". The answer to that is a 
foregone conclusion, and not a good use of our time.

I'm glad I tried the unification. It was useful to see it fail.

-- 

Michael Everson * http://www.evertype.com