Summary so far: The date for the events of the Cad Goddeu ('The Battle of the Trees') suggested by Graves seems too early for any supposed event involving replacement of an older alphabet with a newer alphabet in Britain to have taken place. However, the events may have happened earlier.

KEY QUESTION FOUR:
Was there an alphabet in existence in 1700 BCE that had 13 consonants?

A REVIEW OF ALPHABETS IN EXISTENCE IN 1700 BCE:

INDO-ARYAN LANGUAGE? (NO ALPHABET IN 1700 BCE):
"Indo-Aryan languages, including Sanskrit, attested (via oral tradititon) from about the mid 2nd millennium BC (Rigveda)." No evidence of an alphabet that early, but, just for the sake of the argument, it (at least later) had:
"36 phonemes . . . 48 sounds. http://en.wikipedia.org/wiki/Sanskrit
(So it's like Chinese, depending on tone for meaning?)
14 vowels? so 22 consonants? Too many.



THE EGYPTIAN ALPHABET? (24 CONSONANTS)
http://en.wikipedia.org/wiki/Middle_Bronze_Age_alphabets#Origin_of_the_alphabet
shows an Egyptian alphabet with 22 to 26 letters. Only 3 of them vowels, so 19 to 23 consonants. And points out that there we have no knowledge that they were grouped together as an "alphabet" at that early a date.

> how many consonants there were in ancient Egyptian?
I've now read up about "triliterals", so I now realize that is an over-simplified question. (Triliterals are groups of 3 consonants to which different combinations vowels were added for different meanings, or sometimes other types of signs, to help clarify, leading to a rather confusing mess. We still have the echoes of that today: sing, sang, sung.)

http://www.friesian.com/egypt.htm -
The Pronunciation of Ancient Egyptian (an awesome page) explains a bit: "Besides phonograms that stand for two or three consonants, there are also 24 (or 25) signs that represented single ("uniliteral") sounds, the Egyptian "alphabet." These were originally ideograms also, and some continued to stand for common words. For instance, is the picture of a mouth, is used to mean "mouth," "language," etc., and is a uniliteral sign. These alphabetic signs were frequently written with ideograms or pictograms as "phonetic complements," both to provide reminders about pronunciation and to distinguish meanings, as when grammatical endings differentiate between nouns and verbs, or between singular and plural."

If you google "Egypt hieroglyph 24 letter", you get a lot of agreement on that, but no clear opinions on how many letters would be "consonants."
Even http://en.wikipedia.org/wiki/Egyptian_language
http://en.wikipedia.org/wiki/Egyptian_hieroglyphs

Here's a definite opinion:
"The Egyptian language had 24 consonantal sounds, and there was a hieroglyphic symbol for each of these. Many had two symbols, either of which could be used, making about 30 symbols representing single consonants."
http://www.bbc.co.uk/dna/h2g2/A20725742
y, y, ayn, w, b, p, f, m, n, r, h, hh, kh, ch, s, z, sh, q, k, g, t, tsh, d, j
("The 'Ayn sound exists in modern Arabic but not in European languages. It is the initial sound of the Arabic name 'Ali' and sounds like a growled 'ah'. The apostrophe at the beginning of the word 'ayn is part of the word and represents a glottal stop.")



THE HEBREW ALPHABET? (17-20 CONSONANTS):
Hebrew has 3-4 vowels and 17-20 consonants, depending on how one counts (also, two consonants are used for a second sound by addition of a dot, S and Sh (Seen/Sheen) and Taf XXX). Sticking with ancient Hebrew, with vowels in capitals:

ALEPH bet gimmel dalet heh = 4 consonants + 1 vowel
VAV zion chet tet YUD = 3-4 c + 1-2 v (yud could be both)
kaf lamed mem nun = 4 c
samech AYIN pey/fey tzadi = 3-4 c + 1 v
kuf resh seen/sheen taf = 3-4 c

http://www.ancient-hebrew.org/28_chart.html
http://en.wikipedia.org/wiki/Hebrew_alphabet

THE UGARITIC ALPHABET? (31 LETTERS AND NOT IN EXISTENCE YET)
The Ugaritic alphabet, a later Semitic alphabet, has 31 letters. http://en.wikipedia.org/wiki/Ugaritic_script

http://en.wikipedia.org/wiki/Genealogy_of_scripts_derived_from_Proto-Sinaitic
has a tree with Wadi el-Hol and Proto-Sinaitic at the root, and three first-level branches: Ugaritic, Proto-Canaanite, and South Arabic. Ugaritic is in the north (Syria) but at least in some cases its alphabetic order is attested as the South Arabic order, which looks nothing like the Canaanite-descended order we are used to. The only descendant of the now-extinct South Arabic scripts is Ethiopic script.
The tree contains abjads (consonantal alphabets) as well as alphabets in the narrow sense, syllabaries or semi-syllabaries, and abugidas (Indian and Ethiopian scripts where vowels are written as diacritic marks around the consonants, instead of inline).


After installing the akkadian.otf font from: http://users.teilar.gr/~g1951d/download.html
one can see all the syllabic characters in: http://en.wikipedia.org/wiki/Cuneiform#Syllabary (At least in Firefox.)
There are 4 x 14 = 56 syllables each for CV and VC. However most syllables have more than one character that can be used to represent them, the highest being GU with 7; but some characters are reused to represent multiple syllables. Perhaps 200-300 characters total in the table.

Ugaritic also differs from the other Semitic alphabets in having three letters usually transcribed as 'a, 'i, 'u instead of the single Semitic aleph for '; they were not used for post-consonantal vowels in Ugaritic, but the Greeks and other westerners could easily have gotten the idea of vowel letters here.

Another thing that changed around then, i.e. with Ugaritic, is the beginning of writing left to right horizontal. http://www.omniglot.com/writing/direction.htm
(The Indus Valley script was usually written from right to left - http://ourworld.compuserve.com/homepages/dp5/easter3.htm )

THE HURRIAN (CUNEIFORM) ALPHABET:
Hurrian didn't have many consonants (12 shown): http://en.wikipedia.org/wiki/Hurrian_language

Urdu/Hindi By Abdul Jamil Khan:
"The Hurrian language appears to be related to Uratrian, an ancient language of the Caucasus region, and it might have interacted with Elamite, which had extended into the Turkmenistan area. . . .

http://www.premiumwanadoo.com/cuneiform.languages/en_phonetique.htm
"The Hittites apparently borrowed the cuneiform writing from the Hurrians who themselves had borrowed it from the Old-Akkadians."



THE HITTITE (CUNEIFORM) ALPHABET:
Urdu/Hindi By Abdul Jamil Khan - Page 73 says the Hittite alphabet had 13 consonants.

"Related more to Urdu in this theater is the Hittite language, with a resemblance to Persian and Sanskrit. Hittite, covering a vast territory, had intermingled with Sumerian, Akkadian, Assyrian, Aramaic, Hattian, Hurrain, Lycian, Lydian, and others, and must have exchanged words and idioms. The Hittite king, Suppliliuma, boasted of knowing twelve languages. Hittite vocabulary is thus quite mixed, but its verb flexion bears a resemblance to some Indo-European languages. It reveals the guttural, or laryngeal, sounds as well as the aspirant h found in other Indo-European languages."

The Hittite cuneiform alphabet, from http://www.ancientscripts.com/hittite.html

(arranged to match the BLNF alphabet order of Graves' 13-consonant alphabet, described above):
b l n [no F] š ḫ d t k [instead of C?] m g [no P] r + z

(arranged to match the BLFS alphabet order of Graves' 15-consonant alphabet, described above):
b l [no F] š n ḫ d t k [instead of C?] q m g [no Ng] [no FF] z r

+ p w y [which (explained above) went into the Semitic alphabet].

http://en.wikipedia.org/wiki/Hittite_language
Hittite was spoken from 1800. Their alphabet was cuneiform, a syllabary with 375 signs.
http://en.wikipedia.org/wiki/Hittite_cuneiform & there is agreement at: http://www.ancientscripts.com/hittite.html

[Also note: http://en.wikipedia.org/wiki/Anatolian_hieroglyphs (re: Luwian)]

But the Hittite signs were made of combinations of a consonant plus a vowel.
However, the limited number of consonant-sounds matches Graves' 13-consonant alphabet.

The large number of (cuneiform) signs in the Hittite cuneiform alphabet may not conflict with Graves' postulate: Graves never says there was an actual set of visual signs assigned letter values, only a limited number of consonants. The poems could have been memory devices for remembering some other ordering system. Actually, the letters could have been cuneiform, or ogham, or whatever, as long as they had those 13- and 15-consonant values assigned to them.



THE SUMERIAN (CUNEIFORM) ALPHABET:
Sumerian, the original cuneiform, had syllabic signs that were used heavily.
http://en.wikipedia.org/wiki/Cuneiform#Syllabary

In fact if you take the DeFrancis viewpoint, invention of syllabic signs was what created full writing as opposed to proto-writing, around 3200 BC.

b, d, g, ḫ, k, l, m, n, p, r, s, š, t, z = 14 consonants.

What was missing: B L N F S H D T C M G P R


B L F S N H D T C -Q- M G -Ng- -(FF or Z)- R
[added letters are shown like -this- ; and P is missing]

Compared to these, Sumerian is missing: C, Q, Ng, FF

Sumerian is an even closer match to Graves' alphabets than Hittite.



In adddition, the sounds (C, Ng, F) are missing from both Hittite and Sumerian,
and these sounds are similar to what was introduced with the BLFS (Q, Ng, FF or Z, and removal of P).

Conclusion: Sumerian, Hittite, or another Semitic alphabet such as Hebrew, or even Hurrian, could have been the 13-consonant "Beth-Luis-Nion" language that was being referred to in Amergin's poems if the events described by them took place in 1700 BCE.


CONTINUE ON TO MORE DISCUSSION OF ROBERT GRAVES' THE WHITE GODDESS