Cuneiform | What Was the First Alphabet? | The Development of the Greek Alphabet: The letters, their assignment to Greek sounds, and the style of writing
Like so much of ancient history, we only know so much. Beyond that, scholars specializing in related areas make educated guesses. Discoveries, usually from archaeology, but more recently from x-ray type technology provide us with new information that may or may not substantiate previous theories. As in most disciplines, there is rarely consensus, but there are conventional approaches and widely held theories, as well as intriguing, but hard to verify outliers. The following information on the development of the Greek alphabet should be taken as general background. I've listed some books and other resources for you to follow if, like me, you find the history of the alphabet especially fascinating.
It is currently believed that the Greeks adopted a West Semitic (from an area where Phoenician and Hebrew groups lived) version of the alphabet, perhaps between 1100 and 800 B.C., but there are other points of view [see: Ancient scripts and phonological knowledge, by D. Gary Miller (1994)]. The borrowed alphabet had 22 consonantal letters. The Semitic alphabet wasn't quite adequate, though.
VowelsThe Greeks also needed vowels, which their borrowed alphabet didn't have. In English, among other languages, people can read what we write reasonably well even without the vowels. There are surprising theories about why the Greek language needed written vowels. One theory, based on events contemporary with possible dates for the adoption of the Semitic alphabet, is that the Greeks needed vowels in order to transcribe hexametric poetry, the type of poetry in the Homeric epics: The Iliad and The Odyssey. While the Greeks may have been able to find some use for around 22 consonants, vowels were essential, so, ever resourceful, they reassigned the letters. The number of consonants in the borrowed alphabet was roughly adequate to the Greeks' need for distinguishable consonantal sounds, but the Semitic set of letters included representations for sounds the Greeks didn't have. They turned four Semitic consonants, Aleph, He, Yod, and Ayin, into symbols for the sounds of the Greek vowels a, e, i, and o. The Semitic Waw became the Greek Digamma (voiced labial-velar approximant), which Greek eventually lost, but Latin retained as the letter F.
When the Greeks later added letters to the alphabet, they generally put them at the end of the alphabet, maintaining the spirit of the Semitic order. Having a fixed order made it easier to memorize a string of letters. So, when they added a u vowel, Upsilon, they placed it at the end. Long vowels were later added (like the long-o or Omega at the very end of what is now the alpha-omega alphabet) or made long vowels out of existing letters. Other Greeks added letters to what was, at the time and before the omega's introduction, the end of the alphabet, to represent the (aspirated labial and velar stops) Phi [now: Φ] and Chi [now: Χ], and (stop sibilant clusters) Psi [now: Ψ] and Xi/Ksi [now: Ξ].
Variation Among the Greeks
Eastern Ionic Greeks used the Χ (Chi) for the ch sound (aspirated K, a velar stop) and the Ψ (Psi) for the ps cluster, but Western and mainland Greeks used Χ (Chi) for k+s and Ψ (Psi) for k+h (aspirated velar stop), according to Woodhead. (The Χ for Chi and Ψ for Psi is the version we learn when we study ancient Greek today.)
See Latin Changes to the Alphabet to find out why we have the redundant letters c and k.
Since the language spoken in different areas of Greece varied, the alphabet did so, as well. After Athens lost the Peloponnesian War, and then overthrew the rule of the thirty tyrants, it made a decision to standardize all official documents by mandating the 24-character Ionic alphabet. This happened in 403/402 B.C. in the archonship of Euclides, based on a decree proposed by Archinus*. This became the dominant Greek form.
Direction of the Writing
The writing system adopted from the Phoenicians was written and read from right to left. You may see this direction of writing called "retrograde." It was how the Greeks first wrote their alphabet, as well. In time they developed a system of circling the writing around and back on itself, like the course of a pair of oxen yoked to a plough. This was called boustrephedon or boustrophedon from the word for βούς bous 'oxen' + στρέφειν strephein 'to turn'. In alternate lines, the non-symmetric letters usually faced the opposite way. Sometimes the letters were upside down and boustrephedon could be wrtten from up/down as well as from left/right. Letters that would appear different are Alpha, Beta Β, Gamma Γ, Epsilon Ε, Digamma Ϝ, Iota Ι, Kappa Κ, Lambda Λ, Mu Μ, Nu Ν, Pi π, Rho Ρ, and Sigma Σ. Note that the modern Alpha is symmetrical, but it wasn't always. (Remember the p-sound in Greek is represented by a Pi, while the r-sound is represented by the Rho, which is written like a P.) The letters that the Greeks added to the end of the alphabet were symmetrical, as were some of the others.
There was no punctuation in early inscriptions and one word ran into the next. It is thought that boustrophedon preceded the left to right form of writing, a type that we find and call normal. Florian Coulmas asserts that the normal direction had become established by the fifth century B.C. E.S. Roberts says that before 625 B.C. the writing was retrograde or boustrephedon and that normal facing writing came in between 635 and 575. This was also the time the iota was straightened to something we recognize as an i vowel, the Eta lost its top and bottom rung turning into what we think looks like the letter H, and the Mu, which had been a series of 5 equal lines at the same angle top and bottom -- something like:
\/\/\ and thought to resemble water -- became symmetrical, although at least once on its side like a backwards sigma. Between 635 and 575, retrograde and boustrephedon ceased. By the middle of the fifth century, the Greek letters we know were pretty much in place. In the later part of the fifth century, rough breathing marks appeared.
* According to Patrick T. Rourke, "The evidence for Archinus' decree is derived from the fourth-century historian Theopompus (F. Jacoby, *Fragmente der griechischen Historiker* n. 115 frag. 155)."