Category: Ancient Learning > Writing Systems
1 Introduction
Many people who know only English or other languages written in the Roman alphabet believe that learning a new script is one of the most difficult aspects of learning an ancient language. This is true for some scripts and for some people, but it is hardly the case for most when it comes to the Greek alphabet, since it is very closely related to the Roman letters. In any case, it is useful when reading about ancient Greek matters to know at least a little bit about the alphabet, even if one does not have the leisure to learn reading it fluidly.
It may be less obvious why people who can already read this would benefit from reading something about the Roman alphabet here, but there are some important differences between the ancient Roman script and its modern incarnations, and it pays to be aware of these.
[I hope to eventually rearrange the information here in a more easily usable format, but do not intend to change the content, except to add links to texts from ancient grammarians discussing the alphabets.]
2 The Greek Alphabet
(Also see On Romanizing Ancient Greek.)
The Greek alphabet was adapted from the Phoenician script around 800 BCE, even if it possesses several novel features. Its name openly shows that dependence, since the names of the letters Alpha and Beta are not Greek in origin, but indeclinable Phoenician loanwords. They are therefore almost identical to the names of the letters of the Hebrew alphabet, since Hebrew and Phoenician are dialects of the same language (conventionally called Canaanite).
The most important change relative to Phoenician is the repurposing of a number of letters as vowels; Phoenician only wrote the consonant sounds, omitting vowels. But this innovation is less radical than often claimed, as earlier syllabary scripts already included vowels, and Phoenician writing is much less dependent on vowel sounds to be comprehensible. It is really only a sign of accomodation to a new language. In other words, the script suits the language it was designed for very well – unlike the Linear B script which had been used for the Greek language in a much earlier period (15th–13th cents. BCE), and which was not able to record Greek words with the same precision – but it did not represent a great leap forward in the intellectual history of humanity, as European chauvinists hold forth.
The standard Greek alphabet, which originated in Ionia (Asia Minor), was officially adopted in Athens in 403 BCE (replacing a slightly different Attic variant of the script), and then slowly replaced local variants across the Greek-speaking world, runs as follows (for help with pronunciation, see the Wikipedia pages on the respective International Phonetical Alphabet = IPA symbol):
- Α α, called Alpha (ἄλφα), is pronounced like continental A (IPA: /a/).
It could be either short or long. - Β β, called Bēta (βῆτα), was classically pronounced like B (IPA: /b/).
Its modern Greek pronunciation is like V (IPA: /v/). - Γ γ, called Gamma (γάμμα), was classically pronounced like a hard G (IPA: /g/).
Its modern Greek pronunciation is like Dutch G or Arabic GH (IPA: /ɣ/) - Δ δ, called Delta (δέλτα), was classically pronounced like a D (IPA: /d/).
Its modern Greek pronunciation is like the soft TH sound in ‘the’ (IPA: /ð/). - Ε ε, called Epsīlon (ἒ ψῑλόν), is pronounced like continental E (IPA: /e/).
In the classical alphabet, it was always short. - Ζ ζ, called Zēta (ζῆτα), was pronounced in different ways:
The classical Attic pronunciation was ZD (IPA: /zd/).
But in some places and times, a fricative pronunciation such as DZ (IPA: /dz/) is attested.
The Koine and modern Greek pronunciation is like Z (IPA: /z/). - Η η, called Ēta (ἦτα), was pronounced as a long vowel:
The classical Attic pronunciation was /ɛː/.
The Koine pronunciation was first /eː/, then /iː/, finally merging with Iota and Ypsilon. - Θ θ, called Thēta (θῆτα), was pronounced in different ways:
The classical Attic pronunciation was like the aspirated T in ‘tear’ (IPA: /tʰ/).
The Koine and modern Greek pronunciation is like the hard TH sound of ‘thorn’ (IPA: //). - Ι ι, called Iōta (ἰῶτα), is pronounced like continental I (IPA: /i/).
It is not consonantal in ancient Greek. - Κ κ, called Kappa (κάππα), is pronounced like K or a hard C, unaspirated as in ‘scar’ (IPA: /k/).
- Λ λ, called La(m)bda (λά[μ]βδα), is pronounced like L (IPA: /l/).
- Μ μ, called My/Mu (μῦ), is pronounced like M (IPA: /m/).
- Ν ν, called Ny/Nu (νῦ), is pronounced like N (IPA: /n/).
- Ξ ξ, called Xī (ξεῖ), is pronounced like X (IPA: /ks/).
- Ο ο, called Omicron (ὂ μικρόν), is pronounced like continental O (IPA: /o/).
It could be either short or long. - Π π, called Pī (πεῖ), is pronounced like the unaspirated P in ‘spear’ (IPA: /p/).
- Ρ ρ, called Rhō (ῥῶ), is pronounced like the R in Scottish English or Italian (IPA: /r/).
At the beginning of a word, and when geminated, it was aspirated (IPA: /ʰr/). - Σ σ, called Sīgma (σῖγμα), is pronounced like S (IPA: /s/).
- Τ τ, called Tau (ταῦ), is pronounced like the unaspirated T in ‘steer’ (IPA: /t/).
- Υ υ, called Ȳpsīlon (ὖ ψιλόν), was pronounced as a short or long vowel:
The archaic pronunciation was /u/.
The classical Attic pronunciation was /y/.
The modern Greek pronunciation is /i/ (i.e., the same as Eta and Iota). - Φ φ, called Phī (φεῖ), was pronounced in different ways:
The classical Attic pronunciation was like the aspirated P in ‘peer’ (IPA: /pʰ/).
The Koine and modern Greek pronunciation is like F (IPA: /f/). - Χ χ, called Chī (χεῖ), was pronounced in different ways:
The classical Attic pronunciation was like K or a hard C, aspirated as in ‘car’ (IPA: /kʰ/).
The Koine and modern Greek pronunciation is like Arabic KH / Scottish or German CH (IPA: /x/). - Ψ ψ, called Psī (ψεῖ), is pronounced as PS (IPA: /ps/).
- Ω ω, called Ōmega (ὦ μέγα), was pronounced as a long vowel:
The classical Attic pronunciation was /ɔː/.
The Koine pronunciation was /oː/, and in modern Greek has merged with Omicron.
(Here, archaic means pre-classical Attic or non-Attic dialects, and Koine means the Greek spoken across the Mediterranean in the Hellenistic and Roman imperial periods, which then blends into modern Greek.)
In antiquity, only the capital letter forms were used (though not always in the shape used today), and a few of the vowels had different names:
- Ε ε was called Ī or Ei (Εἶ), as the spelling ει in classical Attic represented /eː/.
- Ο ο was called Ū or Ou (Οὖ), as the spelling ου in classical Attic represented /oː/.
- Υ υ was called Hȳ (Ὗ), since words beginning with Ypsilon are (classically) all preceded by h.
- Ω ω was simply called Ō (Ὦ).
In addition to the 24 letters, there is a sound /ŋ/ that occurs before the consonants Γ γ, Κ κ and Χ χ, and is spelled as Γ γ (hence sometimes misleadingly transcribed as gg, gk, gkh), and an /h/ sound indicated only by a difference of accent at the beginning of a word (ἀ = a vs. ἁ = ha), and not indicated at all within a word (e.g. Εὐήμερος = Euhḗmeros, not *Euḗmeros). In antiquity, these and other accents were usually left unwritten.
[I will at some point address the accents elsewhere, gods willing.]
Furthermore, there are the following diphthongs:
- αι is /ai/.
In modern Greek /e/. - Αυ αυ was /au/.
In modern Greek /av/ or /af/. - Ει ει was /eː/ in classical Attic (and in some dialects /ei/),
/iː/ in Koine.
In modern Greek, /i/. - Ευ ευ was /eu/.
In modern Greek /ev/ or /ef/. - Οι οι was /oi/ in classical Attic,
/yː/ in (late) Koine.
In modern Greek, /i/. - Ου ου was /oː/ in classical Attic (and in some dialects /ou/),
/uː/ in Koine.
In modern Greek, /u/. - Ηυ ηυ was /ɛːu/ in classical Attic,
/eːu/ in Koine, then /iːu/.
In modern Greek /iv/ or /if/. - Υι υι was /ui/ in archaic Greek,
/yi/ in classical Attic.
In modern Greek /i/. - Αι/ᾼ ᾳ was /aːi/ in classical Attic,
/aː/ in Koine.
In modern Greek /a/. - Ηι/ῌ ῃ was /ɛːi/ in classical Attic,
/ɛː/ or /eː/ in Koine, then /iː/.
In modern Greek /i/. - Ωι/ῼ ῳ was /ɔːi/ in classical Attic,
/ɔː/ or /oː/ in Koine.
In modern Greek /o/.
In the last three cases, the originally ‘adscript’ Iota (e.g., ΑΙ αι) was reduced to a ‘subscript’ Iota (ᾼ ᾳ) when the diphthongs were simplified to long vowels in pronunciation.
3 Relationship of the Greek to the Roman Alphabet
The Roman alphabet derives from the Greek, but (A) indirectly and (B) from an earlier form than the one given above. In a sense, therefore, they are sibling writing systems more than model and derivative.
This is irrelevant as far as the identical or near identical letters go; I mean:
- Α α = A a
- Β β = B b
- Ε ε = E e
- Ζ ζ = Z z
- Ι ι = I i
- Κ κ = K k
- Μ μ = M m
- Ν ν = N n
- Ο ο = O o
- Τ τ = T t
- Υ υ = Y y
In these cases, only the lower-case letters, which developed later, are different in appearance.
But a few Greek letters never made it into the Roman script (Θ θ, Ξ ξ, Φ φ, Ψ ψ, Ω ω); and two that are shared have different values, because the Roman alphabet followed a different Greek model from the one that became standardized:
- Η η ~ H h
- Χ χ ~ X x
Finally, some letters diverge in form while retaining (more or less) the same sound value because they developed in different directions from a common ancestor:
- Γ γ and C c both descend from a character similar to <. In Greek, the angle was shifted, in Latin, the letter was rounded.
- Δ δ and D d both descend from a letter shaped somewhat like ▷.
- Λ λ and L l both descend from a letter similar to ᛚ, which could stand at various angles.
- Π π and P p both descend from something between a Π with the right ‘stem’ only reaching down to the middle and a P whose ‘bowl’ is not closed (somewhat like a rounded ᛚ).
- Σ σ and S s both descend from something like ᛊ.
Only in one case could it be said that the Greek looks (more or less) like the more ancient form behind a Roman letter:
- A letter much like Ρ ρ was changed to R r by adding a ‘leg’ to differentiate it from Roman P p (~ Greek Π π).
Other Roman letters (G g, F f, J j, Q q, U u, V v, W w) do not have an exact equivalent in the standard Greek alphabet; their origins will be explained in the next section.
4 The Roman Alphabet
Note that the Roman vowels were all pronounced in the continental manner, i.e., as in Italian, not as in English, and can be either short or long (if long, written with a macron, e.g. ā).
- A a was called ā.
- B b was called bē.
- C c was called cē.
It was pronounced as a hard C /k/. - D d was called dē
- E e was called ē.
- F f was called ef.
- G g was called gē.
It was pronounced as a hard G /g/. - H h was called hā.
It was often omitted, but if written, always pronounced /h/. - I i was called ī.
It can function as a consonant /j/ or a vowel /i/. - K k was called kā.
Only used before A, at the beginning of a word, and only in certain words. - L l was called el.
- M m was called em.
- N n was called en.
- O o was called ō.
- P p was called pē.
Unaspirated except when written PH (which can however also stand for /f/, depending on the time). - Q q was called qū (i.e., cū).
It was always combined with V and sometimes pronounced as K (e.g., quom = /kum/), sometimes as something like modern QU (e.g., qui = /kʷiː/). - R r was called er.
It was prononuced like the R in Scottish English or Italian. - S s was called es.
- T t was called tē.
Unaspirated except when written TH (which can however also stand for /θ/, depending on the time). - V v was called ū.
It can function as a consonant /w/ or a vowel /u/ (in the latter case, it is usually transcribed u today). - X x was called ix.
It stands for /ks/. - Y y was called hȳ or ȳ Graecum (i.e, ī Graecum).
It stands for Υ υ = /y/ in Greek loanwords, and sometimes /y/ in other contexts (but was also often pronounced /i/) - Z z was called zēta.
It stands for Ζ ζ = /z/ in Greek loanwords, and can also represent /z/, /dz/ or /ts/ in other contexts.
Only the capital letters were used in antiquity.
In addition to the letters, one should be aware of the diphthongs: ae (~ ‘aye’) and oe (~ ‘oy’), au (~ ‘ow’) and eu (like ‘eh’ running into ‘oo’), ei (‘ey’) and ui (‘uy’). The combination ng was pronounced like English ng plus a hard G, i.e. as in ‘anger’, not as in ‘hang’. There are other niceties, but they are too marginal to need addressing here.
Of the 23 letters, all but G, Y and Z were inherited; G was created as a modification of C, while Y and Z were borrowed from the classical Greek alphabet:
- C, K and Q were originally all used for both /k/ or /g/, but used in different contexts: K before A, Q before V, C elsewhere. As there was no separate letter for the voiced sound /g/, a new letter G was developed from C by the freedman Spurius Carvillus Ruga in the 3rd century BCE. The new letter was slotted in after F and before H, replacing inherited Z, which had fallen out of use.
- V in its pronunciation as a vowel /u/ was originally equivalent to Greek Υ, which had also been pronounced /u/; but as its pronunciation shifted to /y/, the Greek letter Y was borrowed anew and appended at the end of the alphabet, to be used in loanwords.
- Similarly, Z was re-borrowed (after previously having been removed from the alphabet, as noted above) to represent Greek Ζ in loanwords. In antiquity, both Y and Z were considered Greek letters used in Latin out of necessity, not Latin letters properly speaking.
Other letters subsequently developed from I and V long after the end of antiquity:
- J was originally a medieval graphical variant of I, but came to be used for consonantal uses of the letter. In English, it now represents the affricate /dʒ/, while in German (for instance) it has the same value as ancient consonantal I, namely /j/.
- U likewise was originally a graphical variant of V, but came to be used specifically for the vowel sound.
- W is in origin two V’s. Its pronunciation in English, /w/ rather than /v/, is equivalent to the ancient pronunciation of consonantal V.
SARTRIX 