Medieval Greek (also known as Middle Greek, Byzantine Greek, or Romaic; Greek: Ῥωμαϊκή) is the stage of the Greek language between the end of classical antiquity in the 5th–6th centuries and the end of the Middle Ages, conventionally dated to the Ottoman conquest of Constantinople in 1453.
From the 7th century onwards, Greek was the only language of administration and government in the Byzantine Empire. This stage of language is thus described as Byzantine Greek. The study of the Medieval Greek language and literature is a branch of Byzantine studies, the study of the history and culture of the Byzantine Empire.
The conquests of Alexander the Great, and the ensuing Hellenistic period, had caused Greek to spread throughout Anatolia and the Eastern Mediterranean. The beginning of Medieval Greek is occasionally dated back to as early as the 4th century, either to 330 AD, when the political centre of the Roman Empire was moved to Constantinople, or to 395 AD, the division of the empire. However, this approach is rather arbitrary as it is more an assumption of political, as opposed to cultural and linguistic, developments. Indeed, by this time the spoken language, particularly pronunciation, had already shifted towards modern forms.
Medieval Greek is the link between the older vernacular, known as Koine Greek, and Modern Greek. Though Byzantine Greek literature was still strongly influenced by Attic Greek, it was also influenced by vernacular Koine, which is the language of the New Testament and the liturgical language of the Greek Orthodox Church.
History and development
Constantine the Great moved the capital of the Roman Empire to Byzantium (renamed Constantinople) in 330. The city, though a major imperial residence like other cities such as Trier, Milan and Sirmium, was not officially a capital until 359. Nonetheless, the imperial court resided there and the city was the political centre of the eastern parts of the Roman Empire where Greek was the dominant language. At first, Latin remained the language of both the court and the army. It was used for official documents, but its influence waned. From the beginning of the 6th century, amendments to the law were mostly written in Greek. Furthermore, parts of the Roman Corpus Iuris Civilis were gradually translated into Greek. Under the rule of Emperor Heraclius (610–641 AD), who also assumed the Greek title Basileus (βασιλεύς, 'monarch') in 610, Greek became the official language of the Eastern Roman Empire. This was in spite of the fact that the inhabitants of the empire still considered themselves Rhomaioi ('Romans') until its end in 1453, as they saw their State as the perpetuation of Roman rule. Latin continued to be used on the coinage and in certain court ceremonies up to the 11th century.
Despite the absence of reliable demographic figures, it has been estimated that less than one third of the inhabitants of the Eastern Roman Empire, around eight million people, were native speakers of Greek. The number of those who were able to communicate in Greek may have been far higher. The native Greek speakers consisted of many of the inhabitants of the southern Balkan Peninsula, south of the Jireček Line, and all of the inhabitants of Asia Minor, where the native tongues (Phrygian, Lycian, Lydian, Carian etc.), except Armenian in the east, had become extinct and replaced by Greek by the 5th century. In any case, all cities of the Eastern Roman Empire were strongly influenced by the Greek language.
In the period between 603 and 619, the southern and eastern parts of the empire (Syria, Egypt, North Africa) were occupied by Persian Sassanids and, after being recaptured by Heraclius in the years 622 to 628, were conquered by the Arabs in the course of the Muslim conquests a few years later. The empire lost its predominantly non-Greek-speaking provinces (Syria, Egypt, North Africa) by the 7th-century Muslim conquests, and its population was overwhelmingly Greek-speaking by the 8th century.
Alexandria, a centre of Greek culture and language, fell to the Arabs in 642. During the seventh and eighth centuries, Greek was gradually replaced by Arabic as an official language in conquered territories such as Egypt, as more people learned Arabic. Thus, the use of Greek declined early on in Syria and Egypt. The invasion of the Slavs into the Balkan Peninsula reduced the area where Greek and Latin was spoken (roughly north of a line from Montenegro to Varna on the Black Sea in Bulgaria). Sicily and parts of Magna Graecia, Cyprus, Asia Minor and more generally Anatolia, parts of the Crimean Peninsula remained Greek-speaking. The southern Balkans which would henceforth be contested between Byzantium and various Slavic kingdoms or empires. The Greek language spoken by one-third of the population of Sicily at the time of the Norman conquest 1060–1090 remained vibrant for more than a century, but slowly died out (as did Arabic) to a deliberate policy of Latinisation in language and religion from the mid-1160s.
From the late 11th century onwards, the interior of Anatolia was invaded by Seljuq Turks, who advanced westwards. With the Ottoman conquests of Constantinople in 1453, the Peloponnese in 1459 or 1460, the Empire of Trebizond in 1461, Athens in 1465, and two centuries later the Duchy of Candia in 1669, the Greek language lost its status as a national language until the emergence of modern Greece in the year 1821. Language varieties after 1453 are referred to as Modern Greek.
Diglossia
As early as in the Hellenistic period, there was a tendency towards a state of diglossia between the Attic literary language and the constantly developing vernacular Koine. By late antiquity, the gap had become impossible to ignore. In the Byzantine era, written Greek manifested itself in a whole spectrum of divergent registers, all of which were consciously archaic in comparison with the contemporary spoken vernacular, but in different degrees.
They ranged from a moderately archaic style employed for most every-day writing and based mostly on the written Koine of the Bible and early Christian literature, to a highly artificial learned style, employed by authors with higher literary ambitions and closely imitating the model of classical Attic, in continuation of the movement of Atticism in late antiquity. At the same time, the spoken vernacular language developed on the basis of earlier spoken Koine, and reached a stage that in many ways resembles present-day Modern Greek in terms of grammar and phonology by the turn of the first millennium AD. Written literature reflecting this Demotic Greek begins to appear around 1100.
Among the preserved literature in the Attic literary language, various forms of historiography take a prominent place. They comprise chronicles as well as classicist, contemporary works of historiography, theological documents, and saints' lives. Poetry can be found in the form of hymns and ecclesiastical poetry. Many of the Byzantine emperors were active writers themselves and wrote chronicles or works on the running of the Byzantine state and strategic or philological works.
Furthermore, letters, legal texts, and numerous registers and lists in Medieval Greek exist. Concessions to spoken Greek can be found, for example, in John Malalas's Chronography from the 6th century, the Chronicle of Theophanes the Confessor (9th century) and the works of Emperor Constantine VII Porphyrogenitus (mid-10th century). These are influenced by the vernacular language of their time in choice of words and idiom, but largely follow the models of written Koine in their morphology and syntax.
The spoken form of Greek was called γλῶσσα δημώδης (glōssa dēmōdēs 'vernacular language'), ἁπλοελληνική (haploellēnikē 'basic Greek'), καθωμιλημένη (kathōmilēmenē 'spoken') or Ῥωμαιϊκή (Rhōmaiïkē 'Roman language'). Before the 13th century, examples of texts written in vernacular Greek are very rare. They are restricted to isolated passages of popular acclamations, sayings, and particularly common or untranslatable formulations which occasionally made their way into Greek literature. Since the end of the 11th century, vernacular Greek poems from the literary realm of Constantinople are documented.
The Digenes Akritas, a collection of heroic sagas from the 12th century that was later collated in a verse epic, was the first literary work completely written in the vernacular. The Greek vernacular verse epic appeared in the 12th century, around the time of the French romance novel, almost as a backlash to the Attic renaissance during the dynasty of the Komnenoi in works like Psellos's Chronography (in the middle of the 11th century) or the Alexiad, the biography of Emperor Alexios I Komnenos written by his daughter Anna Komnena about a century later. In fifteen-syllable blank verse (versus politicus), the Digenes Akritas deals with both ancient and medieval heroic sagas, but also with stories of animals and plants. The Chronicle of the Morea, a verse chronicle from the 14th century, is unique. It has also been preserved in French, Italian and Aragonese versions, and covers the history of Frankish feudalism on the Peloponnese during the Latinokratia of the Principality of Achaea, a crusader state set up after the Fourth Crusade and the 13th century fall of Constantinople.
The earliest evidence of prose vernacular Greek exists in some documents from southern Italy written in the tenth century. Later prose literature consists of statute books, chronicles and fragments of religious, historical and medical works. The dualism of literary language and vernacular was to persist until well into the 20th century, when the Greek language question was decided in favour of the vernacular in 1976.
Dialects
The persistence until the Middle Ages of a single Greek speaking state, the Byzantine Empire, meant that, unlike Vulgar Latin, Greek did not split into separate languages. However, with the fracturing of the Byzantine state after the turn of the first millennium, newly isolated dialects such as Mariupol Greek, spoken in Crimea, Pontic Greek, spoken along the Black Sea coast of Asia Minor, and Cappadocian, spoken in central Asia Minor, began to diverge. In Griko, a language spoken in the southern Italian exclaves, and in Tsakonian, which is spoken on the Peloponnese, dialects of older origin continue to be used today. Cypriot Greek was already in a literary form in the late Middle Ages, being used in the Assizes of Cyprus and the chronicles of Leontios Makhairas and Georgios Boustronios.
Phonetics and phonology
It is assumed that most of the developments leading to the phonology of Modern Greek had either already taken place in Medieval Greek and its Hellenistic period predecessor Koine Greek, or were continuing to develop during this period. Above all, these developments included the establishment of dynamic stress, which had already replaced the tonal system of Ancient Greek during the Hellenistic period. In addition, the vowel system was gradually reduced to five phonemes without any differentiation in vowel length, a process also well begun during the Hellenistic period. Furthermore, Ancient Greek diphthongs became monophthongs.
Vowels
The Suda, an encyclopedia from the late 10th century, gives some indication of the vowel inventory. Following the antistoichic system, it lists terms alphabetically but arranges similarly pronounced letters side by side. In this way, for indicating homophony, αι is grouped together with ε /e̞/; ει and η together with ι /i/; ο with ω /o̞/, and οι with υ /y/. At least in educated speech, the vowel /y/, which had also merged with υι, likely did not lose lip-rounding and become /i/ until the 10th/11th centuries. Up to this point, transliterations into Georgian continue using a different letter for υ/οι than for ι/ει/η, and in the year 1030, Michael the Grammarian could still make fun of the bishop of Philomelion for confusing ι for υ. In the 10th century, Georgian transliterations begin using the letter representing /u/ (უ) for υ/οι, in line with the alternative development in certain dialects like Tsakonian, Megaran and South Italian Greek where /y/ reverted to /u/. This phenomenon perhaps indirectly indicates that the same original phoneme had merged with /i/ in mainstream varieties at roughly the same time (the same documents also transcribe υ/οι with ი /i/ very sporadically).
In the original closing diphthongs αυ, ευ and ηυ, the offglide [u] had developed into a consonantal [v] or [f] early on (possibly through an intermediate stage of [β] and [ɸ]). Before [n], υ turned to [m] (εὔνοστος ['evnostos] → ἔμνοστος ['emnostos], χαύνος ['xavnos] → χάμνος ['xamnos], ἐλαύνω [e'lavno] → λάμνω ['lamno]), and before [m] it was dropped (θαῦμα ['θavma] → θάμα ['θama]). Before [s], it occasionally turned to [p] (ἀνάπαυση [a'napafsi] → ἀνάπαψη [a'napapsi]).
Words with initial vowels were often affected by apheresis: ἡ ἡμέρα [i i'mera] → ἡ μέρα [i 'mera] ('the day'), ἐρωτῶ [ero'to] → ρωτῶ [ro'to] ('(I) ask').
A regular phenomenon in most dialects is synizesis ("merging" of vowels). In many words with the combinations [ˈea], [ˈeo], [ˈia] and [ˈio], the stress shifted to the second vowel, and the first became a glide [j]. Thus: Ῥωμαῖος [ro'meos] → Ῥωμιός [ro'mɲos] ('Roman'), ἐννέα [e'nea] → ἐννιά [e'ɲa] ('nine'), ποῖος ['pios] → ποιός ['pços] ('which'), τα παιδία [ta pe'ðia] → τα παιδιά [ta pe'ðʝa] ('the children'). This accentual shift is already reflected in the metre of the 6th century hymns of Romanos the Melodist. In many cases, the vowel o disappeared in the endings -ιον [-ion] and -ιος [-ios] (σακκίον [sa'cion] → σακκίν [sa'cin], χαρτίον [xar'tion] → χαρτίν [xar'tin], κύριος ['cyrios] → κύρις ['cyris]). This phenomenon is attested to have begun earlier, in the Hellenistic Koine Greek papyri.
Consonants
The shift in the consonant system from voiced plosives /b/ (β), /d/ (δ), /ɡ/ (γ) and aspirated voiceless plosives /pʰ/ (φ), /tʰ/ (θ), /kʰ/ (χ) to corresponding fricatives (/v, ð, ɣ/ and /f, θ, x/, respectively) was already completed during Late Antiquity. However, the original voiced plosives remained as such after nasal consonants, with [mb] (μβ), [nd] (νδ), [ŋɡ] (γγ). The velar sounds /k, x, ɣ, ŋk, ŋɡ/ (κ, χ, γ, γκ, γγ) were realised as palatal allophones ([c, ç, ʝ, ɲc, ɲɟ]) before front vowels. The fricative /h/, which had been present in Classical Greek, had been lost early on, although it continued to be reflected in spelling through the rough breathing, a diacritic mark added to vowels.
Changes in the phonological system mainly affect consonant clusters that show sandhi processes. In clusters of two different plosives or two different fricatives, there is a tendency for dissimilation such that the first consonant becomes a fricative and/or the second becomes a plosive ultimately favouring a fricative-plosive cluster. But if the first consonant was a fricative and the second consonant was /s/, the first consonant instead became a plosive, favouring a plosive-/s/ cluster. Medieval Greek also had cluster voicing harmony favouring the voice of the final plosive or fricative; when the resulting clusters became voiceless, the aforementioned sandhi would further apply. This process of assimilation and sandhi was highly regular and predictable, forming a rule of Medieval Greek phonotactics that would persist into Early Modern Greek. When dialects started deleting unstressed /i/ and /u/ between two consonants (such as when Myzithras became Mystras), new clusters were formed and similarly assimilated by sandhi; on the other hand it is arguable that the dissimilation of voiceless obstruents occurred before the loss of close vowels, as the clusters resulting from this development do not necessarily undergo the change to [fricative + stop], e.g. κ(ου)τί as [kti] not [xti].
The resulting clusters were:
For plosives:
[kp] → [xp]
[kt] → [xt] (νύκτα ['nykta] → νύχτα ['nixta])
[pt] → [ft] (ἑπτά [e'pta] → ἑφτά [e'fta])
For fricatives where the second was not /s/:
[sθ] → [st] (Μυζ(η)θράς [myz(i)'θras] → Μυστράς [mi'stras])
[sf] → [sp] (only occurred in Pontic Greek)
[sx] → [sk] (σχολείο [sxo'lio] → σκολειό [sko'ʎo])
[fθ] → [ft] (φθόνος ['fθonos] → φτόνος ['ftonos])
[fx] → [fk]
[xθ] → [xt] (χθές ['xθes] → χτές ['xtes])
For fricatives where the second was /s/:
[fs] → [ps] (ἔπαυσα ['epafsa] → ἔπαψα ['epapsa])
The disappearance of /n/ in word-final position, which had begun sporadically in Late Antiquity, became more widespread, excluding certain dialects such as South Italian and Cypriot. The nasals /m/ and /n/ also disappeared before voiceless fricatives, for example νύμφη ['nyɱfi] → νύφη ['nifi], ἄνθος ['an̪θos] → ἄθος ['aθos].
A new set of voiced plosives [(m)b], [(n)d] and [(ŋ)ɡ] developed through voicing of voiceless plosives after nasals. There is some dispute as to when exactly this development took place but apparently it began during the Byzantine period. The graphemes μπ, ντ and γκ for /b/, /d/ and /ɡ/ can already be found in transcriptions from neighbouring languages in Byzantine sources, like in ντερβίσης [der'visis], from Turkish: derviş ('dervish'). On the other hand, some scholars contend that post-nasal voicing of voiceless plosives began already in the Koine, as interchanges with β, δ, and γ in this position are found in the papyri. The prenasalised voiced spirants μβ, νδ and γγ were still plosives by this time, causing a merger between μβ/μπ, νδ/ντ and γγ/γκ, which would remain except within educated varieties, where spelling pronunciations did make for segments such as [ɱv, n̪ð, ŋɣ]
Grammar
Many decisive changes between Ancient and Modern Greek were completed by c. 1100 AD. There is a striking reduction of inflectional categories inherited from Indo-European, especially in the verbal system, and a complementary tendency of developing new analytical formations and periphrastic constructions.
In morphology, the inflectional paradigms of declension, conjugation and comparison were regularised through analogy. Thus, in nouns, the Ancient Greek third declension, which showed an unequal number of syllables in the different cases, was adjusted to the regular first and second declension by forming a new nominative form out of the oblique case forms: Ancient Greek ὁ πατήρ [ho patɛ́ːr] → Modern Greek ὁ πατέρας [o pa'teras], in analogy to the accusative form τὸν πατέρα [tom ba'tera]. Feminine nouns ending in -ις [-is] and -ας [-as] formed the nominative according to the accusative -ιδα [-iða] -αδα [-aða], as in ἐλπίς [elpís] → ἐλπίδα [elˈpiða] ('hope'), πατρίς [patrís] → πατρίδα [paˈtriða] ('homeland'), and in Ἑλλάς [hellás] → Ἑλλάδα [eˈlaða] ('Greece'). Only a few nouns remained unaffected by this simplification, such as τὸ φῶς [to fos] (both nominative and accusative), τοῦ φωτός [tu fo'tos] (genitive).
The Ancient Greek formation of the comparative of adjectives ending in -ων, -ιον, [-oːn, -ion] which was partly irregular, was gradually replaced by the formation using the more regular suffix -τερος, -τέρα (-τερη), -τερο(ν), [-teros, -tera (-teri), -tero(n)]: µείζων [méːzdoːn] → µειζότερος [mi'zoteros] ('the bigger').
The enclitic genitive forms of the first and second person personal pronoun, as well as the genitive forms of the third person demonstrative pronoun, developed into unstressed enclitic possessive pronouns that were attached to nouns: µου [mu], σου [su], του [tu], της [tis], µας [mas], σας [sas], των [ton].
Irregularities in verb inflection were also reduced through analogy. Thus, the contracted verbs ending in -άω [-aoː], -έω [-eoː] etc., which earlier showed a complex set of vowel alternations, readopted the endings of the regular forms: ἀγαπᾷ [aɡapâːi] → ἀγαπάει [aɣaˈpai] ('he loves'). The use of the past tense prefix, known as augment, was gradually limited to regular forms in which the augment was required to carry word stress. Reduplication in the verb stem, which was a feature of the old perfect forms, was gradually abandoned and only retained in antiquated forms. The small ancient Greek class of irregular verbs in -μι [-mi] disappeared in favour of regular forms ending in -ω [-oː]; χώννυμι [kʰóːnnymi] → χώνω ['xono] ('push'). The auxiliary εἰμί [eːmí] ('be'), originally part of the same class, adopted a new set of endings modelled on the passive of regular verbs, as in the following examples:
In most cases, the numerous stem variants that appeared in the Ancient Greek system of aspect inflection were reduced to only two basic stem forms, sometimes only one. Thus, in Ancient Greek the stem of the verb λαμβάνειν [lambáneːn] ('to take') appears in the variants λαμβ- [lamb-], λαβ- [lab-], ληψ- [lɛːps-], ληφ- [lɛːpʰ-] and λημ- [lɛːm-]. In Medieval Greek, it is reduced to the forms λαμβ- [lamb-] (imperfective or present system) and λαβ- [lav-] (perfective or aorist system).
One of the numerous forms that disappeared was the dative. It was replaced in the 10th century by the genitive and the prepositional construction of εἰς [is] ('in, to') + accusative. In addition, nearly all the participles and the imperative forms of the 3rd person were lost. The subjunctive was replaced by the construction of subordinate clauses with the conjunctions ὅτι [ˈoti] ('that') and ἵνα [ˈina] ('so that'). ἵνα first became ἱνά [iˈna] and was later shortened to να [na]. By the end of the Byzantine era, the construction θέλω να [ˈθelo na] ('I want that...') + subordinate clause developed into θενά [θeˈna]. Eventually, θενά became the Modern Greek future particle θα Medieval Greek: [θa], which replaced the old future forms. Ancient formations like the genitive absolute, the accusative and infinitive and nearly all common participle constructions were gradually substituted by the constructions of subordinate clauses and the newly emerged gerund.
The most noticeable grammatical change in comparison to ancient Greek is the almost complete loss of the infinitive, which has been replaced by subordinate clauses with the particle να. Possibly transmitted through Greek, this phenomenon can also be found in the adjacent languages and dialects of the Balkans. Bulgarian and Romanian, for example, are in many respects typologically similar to medieval and present day Greek, although genealogically they are not closely related.
Besides the particles να and θενά, the negation particle δέν [ðen] ('not') was derived from Ancient Greek: oὐδέν [uːdén] ('nothing').
Vocabulary, script, influence on other languages
Intralinguistic innovations
Lexicographic changes in Medieval Greek influenced by Christianity can be found for instance in words like ἄγγελος [ˈaɲɟelos] ('messenger') → heavenly messenger → angel) or ἀγάπη [aˈɣapi] 'love' → 'altruistic love', which is strictly differentiated from ἔρως [ˈeros], ('physical love'). In everyday usage, some old Greek stems were replaced, for example, the expression for "wine" where the word κρασίον [kraˈsion] ('mixture') replaced the old Greek οἶνος [oînos]. The word ὄψον [ˈopson] (meaning 'something you eat with bread') combined with the suffix -αριον [-arion], which was borrowed from the Latin -arium, became 'fish' (ὀψάριον [oˈpsarion]), which after apheresis, synizesis and the loss of final ν [n] became the new Greek ψάρι [ˈpsari] and eliminated the Old Greek ἰχθύς [ikʰtʰýs], which became an acrostic for Jesus Christ and a symbol for Christianity.
Loanwords from other languages
Especially at the beginning of the Byzantine Empire, Medieval Greek borrowed numerous words from Latin, among them mainly titles and other terms of the imperial court's life like Αὔγουστος [ˈavɣustos] ('Augustus'), πρίγκιψ [ˈpriɲɟips] (Latin: princeps, 'Prince'), μάγιστρος [ˈmaʝistros] (Latin: magister, 'Master'), κοιαίστωρ [cyˈestor] (Latin: quaestor, 'Quaestor'), ὀφφικιάλος [ofiˈcalos] (Latin: officialis, 'official'). In addition, Latin words from everyday life entered the Greek language, for example ὁσπίτιον [oˈspition] (Latin: hospitium, 'hostel', therefore "house", σπίτι [ˈspiti] in Modern Greek), σέλλα [ˈsela] ('saddle'), ταβέρνα [taˈverna] ('tavern'), κανδήλιον [kanˈdilion] (Latin: candela, 'candle'), φούρνος [ˈfurnos] (Latin: furnus, 'oven') and φλάσκα [ˈflaska] (Latin: flasco, 'wine bottle').