The pronunciation of the old Greek cannot be dealt with block: indeed, rich person of a long past, this language was not always marked same manner. It is thus advisable to specify of which state of the Greek one speaks, while keeping in mind that this cutting is artificial: indeed, no language evolves/moves stage by stage, the continuous and progressive phonetic Modifications being (cf Phonétique history ). One stopped with the Greek of the Koinè; indeed, the medieval Greek is already too close to the modern Greek .

It will be considered that the reader can read the Greek alphabet

Characteristics

While passing from the Indo-European to the Greek, the language underwent many phonetic Modifications of which most obvious are described by the Loi of Grassmann, the Loi of Osthoff and the Loi of Rix. It is noted in addition that the Greek allows to restore in many cases the coloring of the Laryngale S IE. Lastly, it is a language '' centum ''.

The old Greek is rich in consonants Occlusive S deaf persons, low in Fricative S and in Spirante S (/w/ and /j/ are rare and are very often amuïes, just as /s/ in many positions), rich in vowels (short or long) and in Diphtongue S. One notes the presence of aspired consonants deaf, vestige of aspired sound of the Indo-European which were maintained only in Sanskrit. Together, these languages are among the others Indo-European Langues to be only had aspired, whatever they are.

Sometimes in general, the alphabet notes in a vague way the language, even if the result is much less ambiguous than in Latin. The diacritics, which currently allow a finer reading, are of later invention: a text epigraphic will generally carry from there no (at most of the separating points of words).

Step

Our knowledge of the pronunciation of the Greek is limited by two big factors: the old language is not spoken any more and it did not cease, during its history, evolving to the modern language. To restore the pronunciation of the old Greek can thus be done only starting from written sources whose chronology imports. The result will be only one restitution, limited as for the precision of the real articulations.

One will leave the sources direct and indirect allowing to restore the pronunciation of the old Greek and, from there, will try to propose of it a phonetic model (inventory of the sounds used) then phonological (i.e. a list of the relevant Phonème S).

The independent sources making it possible to restore the pronunciation are the following ones:

  • the Greek alphabet;
  • testimonys of the grammairiens;
  • loans.

The Greek alphabet

The first source is the alphabet itself. Firstly, the process of adjustment which it had to follow to pass from the Alphabet phenician, a Abjad (noting the consonants especially), support of a Semitic Langue (thus having unknown articulations of the Indo-European Langues as the emphatic consonants) with a Alphabet equipped with autonomous vowels being used for a language with the different phonological system is revealing of what the Greek phonological system was or, especially, was not. For example, the fact that one distinguished the/ĕ/from the/ē/and the/ŏ/of the/ō/are the index of the relevant character of the vocalic Quantité; the fast elimination of a letter being used for /w/, the Digamma, and of another for /k/ in front of /o/ and /u/, the Koppa is that initially of the disappearance of the /w/ phoneme in the language, then of the nonrelevant character of a /k/ deeper (the letter noting /q/ as a phenician).

Moreover, the “orthographical” uses fixed that one can meet in French for example (as Mister being read /m əsjø/) are at the beginning absent from the writing: one so to speak almost directly noted the words such as one heard them (within the limit of the letters available). They appear however and vary with the wire of time (the language evolving/moving), of the places (because of a rather large dialectal scattering) and Greek History: in -403, by decree, Athens adopts a precise alphabet, coming from Ionie, for the official texts. From there, this model will replace gradually the others in all Greece. However, once this standardization of the writing reached, many features of the pronunciation are masked to us: one does not write more as one hears but as one must write. For example, in the language of Athens was preserved longer than in the other dialects a /h/ phoneme in front of vowel at the beginning of the words. However, the selected Ionian model did not make it possible to represent it.

The writing thus announces us more or less well the evolutions and the differences in realization of the letters and the words; one can notice in this respect the epigraphy proves much more relevant than the literary texts: whereas those are generally known to us by medieval Manuscrit S which follow fixed orthographical uses and generally in shift with the real pronunciation (for example, the indication of the initial aspiration, represented by the hard Esprit, was noted there whereas it had ceased being marked in Greek), these, the texts epigraphic, are more faithful representatives by their “spontaneousness”, of the spoken language because they inevitably do not follow the written standards installation sometimes of the centuries after the drafting of the texts.

Testimonys of the grammairiens

The Greek grammairiens informed us about their own language, about dialectal differences of pronunciation located or those of grammairiens Roman on the Greek language; the Western grammatical analysis, let us not forget it, was born in Greece (that whose the vocabulary Métalinguistique attests). These testimonys many and invaluable but, generally, are written in an impressionist vocabulary and little scientist, which does not make it possible to restore with precision the articulation of the sounds.

Loans

The loans made by the other languages, finally, are significant that these loans is words or only letters (for the alphabets drawn from the Greek).

The Latin , in this respect, is very useful by its constant bonds with the Greek world. For example, the fact that the Romans needed to supplement their alphabet by means of two letters taken to the Greek directly, i.e. Y and Z , is important. Indeed, it announces that the Romans did not have any letter to represent these foreign sounds; in this case, one can exclude the Latin phonemes to restore those of the Greek.

Another convincing example: the fact that Greek words containing Φ borrowed in Latin were initially noted with a P , pH in the erudite terms or proper names (what once more there implies that the Romans tried to note a its unknown of their language), then with F as from the Christian era announces the probable period during which the letter Φ passed from a pronunciation Occlusive to Fricative.

For the more recent states of the Greek, the bonds between the East and Occident having been relatively crossed during the Middle Ages, testimonys are thus limited to those of the Greek himself and to those of the languages strongly influenced by him like writing and liturgical Langue of the orthodoxe Église (Slavic Langues, Copte, Gotique, etc). For example, the Cyrillic alphabet, in which the letter В ve (resulting from the B) is worth /v/ and not /b/ makes it possible to confirm that in IXe century the Greek beta had come well from there to decide thus, from where need for the Slavic ones for creating a new letter for /b/: Б (conversely, in the gotic Alphabet the letter 𐌱 resulting from beta notes /b/ well: one can deduce from it that in IVe century the value of the letter was still intact). In the same way, the old letter Ѳ , fita , deciding /f/, watch that Θ Greek theta was well spirantized. The Slavic languages however did not go until importing its foreigner/θ/whom note the Greek letter of the time and were satisfied with /f/, labiodental Fricative and not interdentale (it is noted that the Romans, of the centuries earlier had not had these scruples into important the /y/ absent from their language but which they them also sometimes noted by F Greek theta).

States of the language

Three representative stages were chosen:
  • the “antiquated language” (VIIIe S. → IVe century before our era). The alphabetical antiquated Greek language is thus indicated, whose Homère is one of oldest testimonys (although the text was fixed in writing only at Ve century before our era);
  • the traditional language of Athens (Ionian-attic; IVe century before our era);
  • the Koinè (IVe S. before → IVe S. after our era). It is about the Greek language hellenistic then Roman;
  • the medieval Greek (394 - 1453).
The Mycénien and the modern Greek are treated separately.

One will not lose sight of the fact only the data which follow are artificial and synthetic: indeed, it would not have to be believed that a θ, for example, became/θ/exactly at the beginning of the medieval period, the passage being made gradually and first modification indices phonetic (a Spirantisation) being able to be former, which is the case here, since the first notations of θ by a F Latin (which inform about the fricative character but not about the Joint of the consonant) begin at the beginning of the Christian era. Conversely, the Simplification of the geminated or the Spirantisation of δ and γ was done during the passage of the medieval Greek to the modern Greek, which the table does not indicate. One, moreover, generally represented the pronunciation of Athens to the detriment of the others Greek Dialectes.

Is finally indicated as an indication the pronunciation known as érasmienne because developed by the humanistic Érasme for teaching reasons. It is here about the pronunciation to the Frenchwoman, who differs somewhat from that recommended by the author, who tried to approach an ancient pronunciation, deliberately distant from that one practiced at the time, copied on the model of the Greek of then (almost of the modern Greek ). The result is however not very conclusive: one mingles with the ancient and modern pronunciations with others which never had course. It is however this pronunciation which one uses in France since the Rebirth in the teaching of the old Greek. One will follow in other countries of other uses, more the current being that of a reading close to the modern pronunciation (thus one currently proceeds in Greece).

Inventory of the phons and phonemes

Consonants

To note conventions of phonetic Transcription:
  • one followed overall the International Phonetic Alphabet;
  • one mainly followed interpretations which Michel Lejeune gives, in its Phonétique history of the mycénien and the old Greek ;
  • the Rond subscribes diacritic indicates a soft realization of a Occlusive: thus, includes itself/understands like a soft aspired with less a tension than it. With the other symbols, it announces a phoneme devoized;
  • Ø (“empty set”) indicates a Amuïssement;
  • the Croisillon ( # ) announces, in front of a character, that the letter represented is at the beginning of word;
  • the consonants Géminé be (whose Simplification took seat at various times, generally between the medieval Greek and the modern Greek) can be returned by a redoubled symbol, which is the case for.

Phonetic inventory

Are taken into account all the Allophone S of the phonemes.

Phonological inventory

Only the Phonème S. are listed.

Vowels

Phonetic inventory

Phonological inventory

The phonemes are noted per pairs: initially the nonround version then the round version . One gives the system of the Ionian-attic here, one of richest.

Simple vowels
The vowels can be short or long, without all the combinations being possible. See low for other details.

Traditionally, one uses an alternative of the Transcription of the novelists to describe the vowels: thus, is rendé by ẹ̄ , ǭ etc

Diphthongs

Sandhi

Feature suprasegmental

The suprasegmental milked which knows the old Greek are:
  • the vocalic quantities (opposition between the short and long vowels) and consonant, or, at least, the Gémination (the consonants are differentiated according to whether they simple or are doubled);
  • the Tonème (stressing height);
  • to a lesser extent, the consonant Tension (certain consonants can be marked with more or less of muscular tension).

Vocalic quantity

Simple vowels
The vowels exist in two versions: one short, the other longer. The possible combinations differ according to the dialects. One finds in Ionian-attic the following cases of figure:
  • /a/ ~ /a ː/= α;
  • /e/ = ε ~ /e ː/= ει;
  • /ɛː/= η;
  • /i/ ~ /i ː/= ι;
  • /o/ = ο ~ /o ː/= ου;
  • /ɔː/= ω;
  • /y/ ~ /y ː/= υ.

To note the use of the false diphthongs to note /e ː/and /o ː/, which is explained by the appearance in Ionian-attic of these two phonemes absent from the Greek commun run in a writing which does not have signs to note them: before the adoption of the Ionian alphabet in -403, Ε and Ο were used for /e/ and /o/ as well as with /e ː/and /o ː/. It after -403 and is once carried out confusion between old diphthongs /ei ̯/and /ou ̯/(monophtonguées in /e ː/and /o ː/) with old /e ː/and /o ː/that one could make use of combinations ΕΙ and ΟΥ to note all the cases of figures, whether they result old or from /ei ̯/and /ou ̯/.

The writing is ambiguous for α, ι and υ, which has only one sign to represent the two alternatives. In the philological editions and the didactic works, one uses the Brève and the Macron (or the macron only): /a/ = ᾰ ~ /a ː/= ᾱ or /a/ = α ~ /a ː/= ᾱ, etc

Diphthongs
The diphthongs are invariably considered long, except οι and αι when they are Désinence S (counted however like long for the stressing with the optative and in the adverbs in οι by remembering time when these diphthongs resulting from a Coalescence were worth two syllables). Thus, in the forms λῦσαι and λύσαι, the first - αι final is short (infinitive aorist of the verb λύω), the second length (form of optative, 3rd p. of the sg.).

Consonant quantity/gemination

Pitch

The accent of the Greek, noted sporadically since the hellenistic time then more regularly starting from the medieval Greek (Diacritic cf of the Greek alphabet ), is purely height and not of intensity: indeed, the accent so to speak did not have any modifying action on the vowels (Apophonie, for example) or the words (Métaplasme S) as long as it remained of this nature. It does not have either any influence on the Scansion. On the other hand, becoming a tonic stress in modern Greek, it modified pronfondément the aspect of certain terms.

Acute accent and circumflex accent
Only one vowel of a word (except the cases of Enclise) can receive the pitch, which appears like a rise of a fifth according to Denys d' Halicarnasse. According to the quantity of the vowel and the place of the accent compared to the more S of a long vowel, one obtains several intonations:
  • δό=: intonation (or accent) acute → the accent falls on short;
  • δού=: acute intonation → on the second more long;
  • δοῦ=: intonation (or accent) circumflexe → on the first more long.
Thus, only long can carry a circumflexe intonation.

Barytonèse of the oxytones
When the use of diacritic to note the accents was systematized, an accent, the grave accent, was used to indicate the dull vowels. Quickly, only the intonée vowel was marked; the grave accent should thus have disappeared. However, it was preserved to replace the accent of the vowels accentuated of acute on the final (of the words oxytones ) when the words are followed of another tonic word (or, formulated differently, any final acute accent becomes serious except in paused . Thus, one must write “Πέφευγε λαθ ώ ν. ” and “Λαθ ὼ ν πέφευγε. ” (“he fled in hiding-place”). It is said that the oxytone λαθών becomes the baritone λαθὼν.

The significance of this convention is not clear. Four interpretations were proposed:

  1. it is only one orthographical convention without phonetic reality, i.e the grave accent decides like the acute one;
  2. the grave accent indicates an intermediate tone between the acute accent and the absence of accent;
  3. the grave accent indicates an absence of tone;
  4. the grave accent indicates one 3rd tone, low.

Last interpretation seems excluded. Indeed:

  1. the vowels carrying the baritone are not sung low than the others, if one believes the ancient partitions of them;
  2. the words baritones are never treated like the proclitiques ones.

The first interpretation appears doubtful: indeed, the maintenance of this notation of the grave accent by the ancient grammairiens cannot be free (considering their preoccupation with a systematization, going until the invention of a soft Esprit doubly useless).

The second and the third interpretations seem of this fact most plausible. The ancient testimony of the grammairiens is not without ambiguity, but the references to a tone engraves and with a weakening of the acute accent in grave accent seem compatible with these two interpretations. In support of the second interpretation, one can quote the fact that the syllables with the grave accent are always accentuated in modern Greek, which can suggest that the old speakers in one way or another felt and marked the accent on these vowels. In support of the third interpretation, the fact that the grave accent was initially often used to note the dull syllables. At all events, the words baritones not being compared to dull words, the lifelessness and the barytonèse are not identical.

It is however not possible to know true nature of the difference between a vowel carrying the baritone and a vowel carrying the oxytone. At most can one affirm that the baritone is only one modification secondary of the oxytone. It is not thus advisable to speak about one 3rd intonation.

Laws of limitation
In addition, there exist “laws of limitation”, which determine the possible places for such or such type of intonation, according to the quantity of the last vowel. Indeed:
  • if the last vowel is short, the acute accent cannot go up (on the basis of the end of the word) beyond the 3rd syllable and the circumflex accent beyond the 2nd syllable;
  • if the last vowel is long, the acute accent is limited to the two last syllables, the circumflex accent with the last.

Here a table of the possible placements of the two intonations. The symbols ansi will be read: O indicates a more, # an unspecified syllable (with long or short vowel). For example, “#.oó.oo” is read: “accent on the 2nd more of the penult syllable, the final being long and the vowel of the first undifferentiated syllable”.

Accentuated long penult
One marked between brackets, in the preceding table, the combination #.oó.o, i.e. when the accent finally falls on the 2nd more from a penultimate long vowel followed by a short vowel. Indeed, a law specific to certain dialects, whose attic, makes that such an accent goes up on the 1st more long syllable and gives a circumflexe intonation.

Thus, of the words accentuated like παίδες, “children”, or γυναίκες, “women”, are not possible in Ionian-attic bus they become systematically παῖδες or γυναῖκες. They exist however in the forms παίδες and γυναίκες, in dorien.

This law makes it possible moreover to know the quantity of a final: indeed, the form λύσαι (with ῡ) is allowed only if the final - αι is long; whereas it is normally counted like short, its long quantity states here that one deals with optative in short verb λύω (of which infinitive aorist is well λῦσαι, with - αι final).

Contractions
The Contraction is a consistent Métaplasme in the fusion of two vowels in Hiatus in one alone vowel whose stamp and quantity depend on the vowels of origin. In old Greek, the contractions are very frequent, the language tolerating rather badly the hiatus and often solving it by this process (with the elision and, more rarely, the opposite elision). When, during a contraction, one of the vowels carried the accent, the following rules are observed:
  • the vowel resulting from the contraction long and will consist of two mores, that the two vowels of origin were short (a more) or long (two mores);
  • the first vowel of origin will constitute, for the stressing of the resulting vowel, the first more;
  • the second vowel will constitute, in the same context, the second more.

Thus, any combination in which the first vowel is accentuated the acute one results in circumflexe, any combination where it is the second which carries the accent in an acute accent. For example, the contraction of φιλέω, “I like”, results in φιλῶ, that of φιλεόμενος, “magnet”, in φιλούμενος.

Mobility of the accent
As the quantity of the final vowel of a word varies during the Flection, the accent is thus mobile.

Accent first and variation
One names “accent first” that which carry to the Nominatif a Nom, a Déterminant, a Adjectif or a Pronom. There does not exist universal rule to determine the place of the accent first. At most it is possible to observe some particular rules. For example, in the proper names, the neutrals and the adjectives with the Superlative , the accent tends to go up highest possible in the mot.

It is of this accent first that one will determine the place which the accent with the other forms of the Paradigme during the inflection occupies when the basic principles apply:

  • law of the accentuated long penult;
  • laws of limitation.
The accent first guard its place and its quality during the inflection except when that would contradict the two preceding laws. For example, are the names ἄνθρωπος, “human being”, and δῶρον, “gift”, cities here with personal with their accent first: it is, in both cases, as far in the word as the laws of limitation allow it (on the antepenultimate vowel for ἄνθρωπος and the first more of the penult vowel for δῶρον). With the Genitive, the final vowel becomes long, - ου /o ː/: the accent must thus “go down again”: ἀνθρώπου and δώρου. One determines the accent to obtain by seeking the allowed position which is the least far away from the accent first. In ἀνθρώπου, the accent must go down from two mores ( ó.oo.oo → o.oó.oo ) because *ἀνθρῶπου (descent of only one more: ó.oo.oo → o.óo.oo ) would not have been possible (the properispomenes require a short final). In δώρου, the accent needs to go down only from a more. Conversely, the vocative of πολίτης (with a ῑ), “citizen”, is πολῖται because the final is short and the law of the accentuated long penult applies. The accent goes up of a more.

In addition, an accent first oxytone (acute finally) becomes perispomene (circumflexe finally) with the oblique cases (genitive and dative) when the final is long: ὁδός, “road” made, with the singular oblique cases, ὁδοῦ and ὁδῷ, and in the plural ὁδῶν and ὁδοῖς. Moreover, the topics in Occlusive S of the 3rd variation know a notable accented alternation when they are monosyllabic: the accent passes to the finale on the oblique endings and becomes circumflexe when the vowel is long. Thus, the genitive and the dative of φῶς, “light”, and φλέψ, “vein”, are φωτός, φλεϐός, φωτί and φλεϐί (singular), φωτῶν, φλεϐῶν, φωσί and φλεψί (plural).

Lastly, there exist notable irregularities for certain terms in which the accent moves in a more or less unexpected way. For example, the name πατήρ, “father”, make go up the accent with the vocative, πάτερ.

Proparoxytons irregular
In some nominal inflections, a Métathèse of quantity exchanged the quantity of the final vowel with that which precedes (according to the model ηο → εω). This process having taken seat after the application of the laws of limitation, there exist irregular maintenances of a proparoxyton whereas the final became long.

The phenomenon intervenes mainly in the topics in - ι and - υ, of the type πόλις, “city” and πῆκυς, “axe”, whose notable forms are the following ones:

  • Genitive S Singular S πόλεως and πήκεως;
  • plural genitives πόλεων and πήκεων.

Accent first and conjugation
In the conjugation, the accent conforms to a general operation different from that, very free, nominal inflection. Indeed, it is known as to go up highest possible, i.e. as far as possible of the end of the word, in the respect of the laws of limitation. Thus, the form λυομεθᾰ cannot be accentuated differently than λυόμεθα, the accent going up highest possible, i.e. here on the antepenultimate of a word to initial short.

The fact is that certain Paradigme S does not conform to this general rule. For example, the infinitive of the aorists sets of themes is perispomene. Thus, the verb λείπω, “to leave”, has an infinitive present λείπειν, whose stressing is regular, but an infinitive aorist set of themes λιπεῖν.

Consonant tension

To give an account of the evolution of certain Greek phonemes, it is necessary to postulate a consonant difference of Tension between consonants which, without that, would be identical from the point of view of the discrete units ( Joint, Mode of articulation, etc). Normally, a deaf consonant is naturally emitted with more tension or of articulatory force that sound. Indeed, a deaf person not having the vibrations glottales to accompany it, it can be less audible than sound. At ends of balance, one thus tends to pronounce sound in a way less energetic than a deaf person, and conversely (to also consult Voisement for other details). It is said that the sound ones are soft and strong deaf persons.

In old Greek, however, there exist also soft deaf persons: they are the aspired . Indeed, the emission of the breath glottal makes them more audible than a simple deaf person. The fact that aspired were soft notes various manners:

  • whereas the deaf persons were particularly well maintained during the phonetic history of the Greek, the sound ones and aspired were gradually spirantized: it is because of their soft character. These marked phonemes with less clearness were more easily modified;
  • the C-Ws communication φθ and χθ (which replacement obligatorily, except in some antiquated inscriptions, πθ and κθ) cannot be read and, sequences of sounds not very probable and made impossible by the Loi of Grassmann. One from of deduced that the use of φ and χ in front of another aspired makes it possible to not indicate the aspiration but the soft character of the consonant. Thus, φθ and χθ are interpreted and not and;
  • conversely, geminated, which is inevitably strong, is noted, for aspired, by C-Ws communication τθ, πφ and κχ, with a strong deaf consonant in the digraph.

These details arose however with phonetics: phonologiquement, there do not exist relevant oppositions between strong and soft. Each version is only a Allophone other, in complementary distribution. Thus, there does not exist Minimal pair enters and.

Random links:Cairo Montenotte | District of Montauban | Psychological operations | Tortoise of fiction | Isle lady