PronunciationHow close is the sound of Modern Greek to that of Classic Greek? Phonetically, Classic Greek would sound rather alien to contemporary Greeks, but don’t ever say this to them! It is an issue that most Greeks, even educated ones, ignore.(1) I suspect it is because the alphabet has stayed unchanged, so Greeks can read classic texts with no trouble at all (pronouncing in Modern Greek). After all, it all looks Greek to them! If any (non-Greek) scholar attempts to pronounce classic texts in the reconstructed(2) pronunciation, that, to Greeks is tantamount to sacrilege. As a contemporary Greek myself, I can give you my personal feeling for how the reconstructed pronunciation sounds: it is as if a barbarian is trying to speak Greek.(3) For example, take the word “barbarian” itself (which is of Greek origin): in Classic Greek it would be pronounced [bár·ba·ros]. In Modern Greek, it is [vár·va·ros]. In general, the second letter of the alphabet, beta, was pronounced as [b] in Plato’s time, but was changed to [v] by the time the Gospels were written. Now, to the modern Greek ear, [v] is a soft sound (a “fricative” in linguistics), sort of smooth and gentle, while [b] is a hard one (a “plosive”), kind of rough and crass. The same can be said about the letter delta, which was pronounced as [d] by Plato, and as [ð] (as in this) since around Christ’s time; and the letter gamma ([g] in Classic Greek, [γ] later the latter sound is a “voiced velar fricative”; click here to see the full repertoire of Modern Greek sounds). Greek readers of this text who do not believe that Plato, Socrates, etc., were sounding so barbaric, may take a clue from this very word: “barbaros” was coined after somebody who, as a non-native speaker of Greek would produce incomprehensible speech, which sounded like... well, what? Could it be “var-var-var”? Wouldn’t it sound more barbaric if it were like “bar-bar-bar”? Besides this word, direct evidence for beta comes from a fragment of Attic comedy where it is said that the voice of the sheep is BH-BH.(4) In Modern Greek this would read as “vi-vi”, rather un-sheepish-like; while in the reconstructed way it would be “beeh-beeh”, exactly the sound that we, contemporary Greeks, attribute to the animal. (If the reader would like to make a comment on the above issues, email to me, and let me know what you think; but please make sure to have first read the links that say “Evidence” on the rightmost column of the table, below.) However, the truth is when non-Greek scholars attempt to pronounce Classic Greek in the reconstructed way, they think they pronounce accurately. To me, American scholars sound distinctly American (like Platos with spurs and cowboy hats), Germans sound German, etc. Probably nobody can reproduce exactly the Classic Greek pronunciation: we might know the rules of the reconstructed system, but when it comes to moving our jaws, tongue, and lips, something different comes out of our mouths. As native speakers of this or that language we necessarily carry over our native phonology. Finally, let it be noted that Classic Greek used pitch to differentiate vowels in words, while nearly all modern European languages (including Modern Greek) use stress instead.(5) The Alphabet(Click on the speaker icon, next to the letter name, to hear the pronunciation in Modern Greek)
|
Letter | Name & Sound | Modern Greek pronunciation | Classic Greek Pronunciation (Attic) | |||
1 |
|
[a], as in “father”. Same as [a] in Spanish and Italian. Phonetically, this sound is: open, central, and unrounded. | As in Modern Greek | |||
2 |
|
[v], as in “vet”; a voiced labiodental fricative. | [b], as in “bet”; a voiced bilabial plosive. Evidence | |||
3 |
|
[γ], a sound that does not exist in English. If followed by the sound [u] then it sounds almost like the initial sound in “woman”, but with the back of the tongue touching more to the back (soft) palate. To pronounce [γa], try to isolate “w” from “what” without rounding your lips, and then say [a]. In Castilian Spanish this sound exists in “amiga”. Same is true for [γo]: try eliminating the [u] sound from “water”. (C. Spanish: “amigo”.) On the other hand, due to a phonetic phenomenon called palatalization, [γe] sounds a bit like “ye” in “yes”, and [γi] sounds a bit like “yi” in “yield”. Phonetically, gamma is a voiced velar fricative. (Its palatalized version is a voiced palatal fricative.) |
[g], as in “got”; a voiced velar plosive. Evidence | |||
4 |
|
[th], as in “this”; a voiced dental fricative. | [d], as in “do”; a voiced alveolar plosive. Evidence | |||
5 |
|
[e] as in “pet”, except that the [e] in “pet” (and other similar English words) is lax, whereas in Greek it is tense. To pronounce a tense [e] pull the edges of your lips to the sides a bit more than when you say “pet”. (We pull the edges of our lips to the sides when we smile; but I don’t mean you need to smile every time you pronounce the Greek epsilon, OK? ) |
As in Modern Greek | |||
6 |
|
[z], as in “zone”, a voiced alveolar fricative. Actually, the remark for sigma (see below) applies to zeta as well (it is shifted a bit toward [ʒ], as in “pleasure”). Read the remark for sigma to understand why, and how to pronounce it. |
[zd], as in “Mazda”. Also: [z], and even: [dz]. Evidence | |||
7 |
|
[i], as in “meet”, but shorter, not so long. This is one of the three [i]’s in the Greek alphabet; they all have identical pronunciation. The reason for this redundancy has to do with Classic Greek, where they were not redundant. |
long open mid-[e], as in “thread” (but long). Evidence | |||
8 |
|
[θ], as in “think”; a voiceless dental fricative. In Castilian Spanish: “zorro”. | [th], as in “top”, but more aspirated. Evidence | |||
9 |
|
[i], exactly like eta (see above). The name of the letter is pronounced “yota” in Modern Greek. (The reason for the y-sound in front of the letter’s name is due to the phonetic transformation of [io] into [yo]). |
As in Modern Greek | |||
10 |
|
[k], as in “skip”. Notice that in English [k] is aspirated if it appears word-initially; Greek makes no such distinction. When followed by the vowels [e] or [i] it becomes palatalized — for the exact pronunciation please check the page on palatalization. Phonetically, it is a voiceless velar plosive. (Its palatalized version is a voiceless palatal plosive.) |
As in Modern Greek | |||
11 |
|
[l] as in “lap”. When followed by the vowel [i] it becomes palatalized, turning to a sound that does not exist in English (check the page on palatalization). The name of the letter is pronounced [lamða] ([b] is eliminated because it is difficult to pronounce it between [m] and [ð]). A voiced alveolar lateral approximant. |
As in Modern Greek | |||
12 |
|
[m], as in “map”; a voiced bilabial nasal. Notice that the name of the letter is pronounced [mi] (mee), not “mew” as in American English. |
As in Modern Greek | |||
13 |
|
[n], as in “noble”; a voiced alveolar nasal. When followed by the vowed [i] it becomes palatalized, turning to a sound that does not exist in English (but exists in Spanish, written as ñ; see the page on palatalization). Notice that the name of the letter is pronounced [ni] (ñee), not “new” as in American English. |
As in Modern Greek | |||
14 |
|
[ks] as in “fox”. Contrary to the English “x”, the letter ksi does not change pronunciation at the beginning of a word (it does not become a [z]; Greeks have no trouble starting a word with [k]+[s]). For example, in the word ksenofovia (ξενοφοβία = xenophobia) the initial sound [k] is not omitted. Don’t put any aspiration between [k] and [s] when pronouncing this letter. The remark for sigma applies to the [s]-sound of ksi, too. |
As in Modern Greek | |||
15 |
|
Same like [o] in “got” the way it is pronounced in British English. Notice how the vowel in British “got” is tense, which means that you should really round your lips when you pronounce the Greek [o]. A mid-close back rounded vowel. |
As in Modern Greek | |||
16 |
|
[p], as in “spot”; a voiceless bilabial plosive. Notice that in English [p] is aspirated if it appears word-initially; Greek makes no such distinction. |
As in Modern Greek | |||
17 |
|
[r]: between vowels it is a sound that exists in American English in the pronunciation of “tt” in “butter” (but not in Brittish English). Sounds like the Spanish [r] in “pero”. (Spanish speakers: in Greek there is no difference in whether you trill your rho as in “perro” or not; but normally Greeks pronounce it more as in “pero” than as in “perro”.) Otherwise it’s a trill, like the Italian [r]. Almost every Greek can pronounce rho as a long trill if they wish (like the Russian [r]), and you will hear it pronounced like that in some Greek songs. Phonetically, it is a voiced alveolar tap (and occasionally a trill). |
Probably as in Modern Greek when single, and as a trill when double. Word-initially: aspirated: [hr] | |||
18 |
|
[s], as is “soap”; a voiceless alveolar fricative. Actually, if you listen carefully to native Greek speakers, it sounds a bit like between [s] and [sh] (probably because there is no [sh] in Greek, so the sound is somewhat shifted in the phonological space). However, to the native English ear it sounds much closer to [sh] than to [s], whereas every native Greek speaker would swear they pronounce it exactly like the English [s], unless forced to admit the difference by looking at spectrograms. In reality, you can produce it like this: feel where your tongue is when you say [s]: very close to the front teeth, right? Now feel where it is when you say [sh] (far back). Place it somewhere midway, and you will produce the Greek [s]. You’ll find that you’ll need to make a similar adjustment to the shape of your lips, midway through rounded for [sh] and tense for [s]; in the Greek sigma the lips are relaxed. This is the way “s” is pronounced in Castilian Spanish (as opposed to Latin American Spanish). Notice that the second way of writing the lower case sigma is used exclusively when the letter appears at the end of a word (there is only one capital form). |
Probably as in Modern Greek | |||
19 |
|
[t], as in “stop”; a voiceless alveolar plosive. Notice that in English [t] is aspirated if it appears word-initially; Greek makes no such distinction. |
As in Modern Greek | |||
20 |
|
[i], exactly like eta and iota (see above). The name of the letter is pronounced [ipsilon] (ee-psee-lon), not “yupsilon” as it is called in American English. |
Rounded [i], as in French “une”. Evidence | |||
21 |
|
[f] as in “fat”; a voiceless labiodental fricative. | [ph], as in “pit”, but more aspirated. Evidence | |||
22 |
|
[x], a sound that does not exist in English (but exists in Scottish, as in “loch”; German: “Bach”; Spanish: “Jorge”). When followed by vowels [e] or [i] it is pronounced as in German “ich”. For the exact pronunciation in this case, please check the page on palatalization. Phonetically, it is a voiceless velar fricative. (Its palatalized version is a voiceless palatal fricative.) |
[kh], as in “cut”, but more aspirated. Evidence | |||
23 |
|
[ps] as in “lopsided”. Contrary to English, the sound of the letter does not change at the beginning of a word (it does not become a [s]; Greeks have no trouble starting a word with [p]+[s]). For example, in the word psychologia (ψυχολογία = psychology) the initial sound [p] is not omitted. Don’t put any aspiration between [p] and [s] when pronouncing this letter. The remark for sigma applies to the [s]-sound of psi, too. | As in Modern Greek | |||
24 |
|
[o], exactly like omicron. (Once again, the reason for the redundancy is to be found in Classic Greek.) | Long open mid-back [o], as in “law”. Evidence |
Phonology and OrthographyOops! Twenty-four letters only? Surely some sounds must be missing? That’s correct. There are sounds common in other languages that do not exist in Greek. Such sounds are all the postalveolar fricatives and postalveolar affricates ([ʃ] as in “shop”, [ʒ] as in “pleasure”, [tʃ] as in “church”, and [dʒ] as in “job”). So what do Greeks do when they want to pronounce foreign words with these sounds? If they are not trained to pronounce correctly, they simply transform these postalveolar sounds to their corresponding alveolar ones: [ʃ] → [s], [ʒ] → [z], [tʃ] → [ts], [dʒ] → [dz]. Ask a Greek to pronounce “fish ’n chips” next time you want to have some linguistic fun. And what about other very common sounds, like [b], [d], [g], etc.? These seem to be missing from the alphabet, too! Are they also missing from the repertoire of the sounds of the language? No! These are existent as sounds in the language. It is just that there are no single letters to denote them. When Greeks want to write those sounds they write them as two-letter combinations: [b] is written as μπ (mu + pi), [d] as ντ (nu + tau), and [g] as γκ (gamma + kappa), or as γγ (double gamma). Why all this trouble? Remember, as explained in the introductory paragraph on this page, the sounds [b], [d], and [g] used to exist in Classic Greek. Later, probably some time after the New Testament was written in the so-called Koine (common) Greek, these three sounds had shifted in pronunciation to the corresponding “soft” ones ([v], [ð], and [γ]). This left a void in the phonological space. Words that contained combinations like “mp” and “nt” started being pronounced as [mb] and [nd], respectively. So the “plosive” sounds were re-introduced, but pairs of letters were used now to denote them. There is one more sound in the language which is absent from the alphabet: it is [ŋ], the “ingma”, the last consonant in “king”. This sound is very rare in Greek, and when it appears (as in “άγχος”: anxiety; “έλεγχος”: checking) it is denoted by the combination gamma + chi, with the gamma pronounced as [ŋ]. All of the above plus much more, including the pervasive phenomenon of palatalization, can be found in this page on the details of Modern Greek pronunciation, which includes sound samples with the author’s voice for all of the presented examples. You may also find useful this page, showing the sounds of Modern Greek against all possible sounds of any language in the world. The tables for consonants and vowels in that page are very familiar to linguists, but you don’t need to be a linguist to understand it. For your convenience, here is a table to use as quick reference, listing the two-letter clusters that result in new sounds, not included in the Greek alphabet:
Arguably, there are also the following pairs, which do not result in unique sounds but are perceived as “one thing” by native speakers of Greek:
What about vowels? Is there any similarity with the English vowels, or with those of any other language? Vowels in Greek are easy. That is, if you are not a native speaker of English! That’s because although English is very rich in vowel sounds, still, it lacks almost completely the Greek vowels. The latter are more like the vowels of Italian, Spanish, or Japanese: they are the five sounds [a], [e], [i], [o], and [u](6). Now, there are three letters for [i] in the alphabet (eta, iota, and upsilon), pronounced identically, and two letters for [o] (omicron and omega), also pronounced identically. For the sound [u] (as in “loot”) the combination ου (omicron + upsilon) is used. Here are three good rules of thumb for native English speakers:
So, sounds simple. Is there anything else about vowels? Not in pronunciation. In writing, however, there is. There are
three so-called “diphthongs”, which are not diphthongs
anymore, but digraphs. (A diphthong is a long vowel with more
than one part, each of which has a different quality, such as
the ou in “loud”,
or the oy in “boy”;
a digraph is two letters which, when put together, are read as
one unit, such as the English th in “think”,
or the ph in “graph”.)
Here are the Greek digraphs of vowels: |
Digraph | Modern Greek pronunciation | Classic Greek Pronunciation (Attic) |
Exactly like ε (epsilon, 5th letter, see above) | [ai], as in “buy”. Evidence | |
Exactly like ι (iota, 9th letter, see above) | [ei], as in “bay”. Evidence | |
Exactly like ι (iota, 9th letter, see above) | [oi], as in “boy”. Evidence | |
Exactly like ι (iota, 9th letter, see above; also read comment below) | ~[yui]. Evidence | |
[u], already explained in a previous paragraph | As in Modern Greek | |
[av] if the following sound is voiced, and [af] if the following sound is unvoiced | [au], as in “loud”. Evidence | |
[ev] if the following sound is voiced, and [ef] if the following sound is unvoiced | [eu]. Evidence | |
[iv] if the following sound is voiced, and [if] if the following sound is unvoiced | ~[e:u]. Evidence |
Notes:
If you want to know the reason why these weird-looking combinations of letters exist, once again, blame ancient Greek, in which those were true diphthongs. When later the vowel space was flattened to its present five members, and no long vowels existed anymore, the diphthongs were transformed as the table above shows. Does that mean that the sound [ai] (as in “buy”), for example, can never occur in Modern Greek, because if written as αι it would be pronounced [e]? It is possible to have the letters alpha and iota next to each other and producing the sound [ai], but then we need to show this in writing. We do this by putting a pair of dots, the diaeresis, over the iota, like this: αϊ, as in the word παϊδάκι [paithaki] (=“rib steak”). It is even possible to have the stress together with the diaeresis over the iota: παΐδι [paithi] (=“rib”). The diaeresis can “dismiss” any of the vowel-digraphs: αϊ, εϊ, οϊ, υϊ, οϋ, αϋ, and εϋ (whereas ηϋ does not occur in Modern Greek), in which case the two constituent vowels are pronounced separately. More about the diaeresis in my page on accent marks. What are those short straight lines placed over some vowels in Greek texts? That’s the stress. It shows which syllable should be pronounced slightly higher in pitch and volume than the rest. I suggest that you click here to learn all the details about how to place accent marks to show the stress in Greek. But if you want only a brief description, perhaps the following two paragraphs would suffice.
Do Greek letters have some inherent meaning?(7) What are the dictionary definitions of words like “alpha”, “beta”, etc.? No, there is no meaning in Greek letters. You are probably thinking of Chinese ideograms, or ancient Egyptian hieroglyphs, which are symbols with some associated meaning. Alpha, beta, gamma, delta, etc., bear no more meaning in Greek than a, bee, cee, dee, etc., bear in English. Their names are just a bit longer, that’s all, that’s why they might look like meaningful words. Now, there is an English letter the name of which can be said to have some meaning, however trivial: it’s w, which we pronounce “double-u”. The name of this letter is derived from ancient forms of it, when it was written as two U’s, joined like this: UU. Similarly, there are a handful of Greek letters that can be said to have such trivial meanings, associated always with their pronunciation: epsilon (εψιλον) is really e-psilon, meaning “light e”, “bare e”, or “mere e”, a name introduced in Byzantine times, to distinguish it from the other [e], the digraph alpha-iota (see above). In ancient (e.g., classical) times this distinction was unnecessary, because ε and αι had completely different sounds, so the name of this letter was simply ε. Similarly, upsilon (υψιλον) is really u-psilon, meaning “mere u”, distinguishing it from the other two [i]’s: ι and η; in ancient Greek its name was υ. Finally, omicron (ομικρον) is o-micron, or “little o”, to distinguish it from o-mega (ωμεγα), or “great o”. Again, these names were introduced in later times, when the pronunciations of the two letters had become identical; in ancient times their names were simply ο and ω. That said, it should be mentioned that the origin of
the Greek letters, which is the ancient Phoenician alphabet, did
assign meaning to each letter. For example, the first letter of
the Phoenician alphabet (a close cousin of the Hebrew alphabet,
the first three letters of which are aleph, beth, and gimel)
was written as an inverted A, a
stylized depiction of the triangular head of an ox with its
horns. In Phoenician, the name of the letter was the
word for “ox”. Similarly, the letter gamma (Hebrew gimel)
is derived from the Phoenician word for “camel”; and so on.
But this is not specific to Greek, it is present in Semitic languages. In Greek, the
letters are just symbols, devoid of meaning.
|
Footnotes (clicking on the footnote-number, on the left, brings back to the text) (1). As I was informed recently by a Greek reader of this page, now they do learn at school that the ancient pronunciation was different. Good, that’s progress; in my high school years (late 70’s) we were left in the dark. However, I tested this reader’s information by asking a couple of young Greek students (friends’ children), and found out that just about the only thing they learn is that some vowels, such as the omega and eta, were pronounced differently by the ancients; at least that’s what was registered in those young students’ memory. (2). An earlier version of this page referred to the “Erasmian” pronunciation of Greek, after the Dutch scholar Desiderius Erasmus (1466?-1536), who established a pronunciation system for ancient Greek that was geared towards the phonological abilities of speakers of Dutch. Later, his system was adopted by speakers of various other languages, who adjusted it slightly to their own needs. Since today it is not known what exactly one refers to when one mentions “the Erasmian system”, this page refers to the “reconstructed” system of pronunciation, as shown on the rightmost column, “Classic Greek Pronunciation (Attic)”, and as explained if you follow the links that say “Evidence”. (3). Some readers (esp. fellow Greeks) took this comment to mean that I do not believe the reconstructed pronunciation is right. Wrong! I do believe it is right. However, what I believe or not is totally inconsequential. One should believe their eyes (looking at data, that is) and that is why I have collected the data (which is known to me) under those links that say “Evidence” (last column of the table; you know, those blue underlined things? Click on them!). I made this comment to explain to non-Greeks what it feels to hear the reconstructed pronunciation if you are a Greek, and if you have been educated thinking that Pericles could have said to his wife, “Ασπασία, τι γκαντεμιά! Εγώ τη ψυλλιάζομαι τη δουλειά, φιρί-φιρί το πάνε να μας την πέσουν οι Λακεδαιμόνιοι. Πρέπει να τους σπάσουμε τον τσαμπουκά!” (4). Cratinus, in Dionysalexandros: (5). Not really all modern Indo-European languages, though; Serbo-Croatian, for example, uses pitch. There are probably other I-E languages using pitch as well. (6). But, English speakers beware: the symbol [e] does not stand for the sound of “ee” as in “meet”, but for the sound of “e” as in “get”. Likewise, [i] stands for a shorter version of “ee” as in “meet”. Also note, since the Japanese language was mentioned: actually the Japanese [u] is unrounded, whereas the Greek [u] is rounded. (7). Occasionally young people from the U.S.A. who want to establish a new fraternity or sorority ask me this question, because they want to choose “meaningful” Greek letters for their organization. If you are one of those people and are disappointed by the answer of this question, there is still hope. Select two or three English words that are meaningful to you, ask me what the corresponding Greek words are, and choose the initial Greek letters for your fraternity or sorority.
|
Where to go from here: want to learn more about Greek? Back to the “main” Greek language page