47 Highpoint,
North Hill,
Highgate, 
LONDON N. 6.

17 August 1949.


Dear Mr Bennett,

	Thank you very much for your letter. Please don't worry about the 
time-lag! -Actually, it formed part of a very enjoyable batch waiting for 
me when I got back yesterday from a month's motoring & camping trip in 
Sweden.
	I expect we can leave the discussion of frequencies for the moment, as 
the general approach is the same, and the various personal refinements can't 
really be weighed until they do useful work in helping the actual decipher-
ment. But I'd very much like to go on discussing with you the question of 
vocabulary.
	To my mind there are 2 main working hypotheses for the relationship of 
the "B" language, which one might call the "P"-alternative and the "L"-alter-
native; - the first assuming a relationship through a Pelasgian/Etruscan group 
of dialects, the second through a Lycian/Lelegian/Luvian/Hittite group of 
strictly Indoeuropean languages. Other alternatives, such as relationship with 
no surviving languages, are possible too, but don't provide the material for 
any comparative tests. I've been concentrating on material for the "P"-alter-
native:
	
	I was interested that you suggested the function of ad-ac (_-_) might even 
be prepositional (I hadn't seen it on any of the Knossos material available to 
me: where does it occur?): the "P"-analogy I've had noted for the last year 
or s o is with the Etruscan preposition/postposition

		p u l	,	e p 1_	(* epul ).

Pallottino (Elementi p 63) regards it as probably a preposition of place 
("in, on, towards" perhaps "for") which governs either the nominative or 
the locative:-
		epl tularu	"within the boundary"
		rut alumna-_	"in the sacred place"

	If ad-ac _-_ actually is a preposition at pylos, I suppose its exact 
meaning depends rather on whether one considers the groups which follow and 
precede it to be place-names, proper names or divinities. And then again, one
wonders why these groups are picked out with it, when in other tablets groups 
must occur which are also in some sort of prepositional relationship to the 
transaction involved.
	If ad-ac is to be read pa-lo, then the comparison would-depend on the 
following assumptions:-
1.	The U in p u l is the Etruscan spelling of a colourless vowel 
	forming part of a sonant -l, -m, -n, -r, as in:-
	-(u)m			"and"
	mulaX / mlaX		"offering"
	muli / male / mle	"offers"
	lavtun / lavtn		"family", etc.

2.	"B" final -lo corresponds to Etruscan -l in its possible uses as
	a genitive-dative; collective plural; metronymic; ethnic etc; - the 
	final -o being explained:-
	a)	as a fuller vocalisation lost in Etruscan.
	b)	due to a velarisation of a closing L similar to that in Cretan 
		Greek.
	c)	as the normal Minoan spelling of a final "dead" vowel (final 
		consonant).

	Final -l as an apparent grammatical ending occurs in the Lemnos in-
scription (vanalasial, morinail); the word ookiasiale, if it is connected
with Phocaea, is presumably not a single ending but a double one: either
	a)	the dative of the ethnic "to the Phocaean", or
	b)	a 'redetermined' locative "at Phocaea" like E: tarXna-l-_i "at
		Tarquinii".