Search

Abstract The process by which a monophthong like i or u becomes a diphthong like ai or au (sometimes known as vowel breaking). Diphthongization is the process by which a monophthong becomes a diphthong. There are two processes of diphthongization in Ancient Greek, both of which are diachronic (for a general discussion of the phenomenon, see Andersen 1972).  The first results from the intervocalic loss of w, y, or s, which results in hiatus, i.e., two adjacent vowels in distinct syllables. They then fuse together to form one syllable, as illustrated by the word for ‘child,’ páis > paîs (see f…

Abstract A split arises when allophones of a phoneme cease to be in complementary distribution and thereby become distinct phonemes. A phonemic split occurs when two allophones of a phoneme cease to be in complementary distribution: each takes on a life of its own and the original phoneme ‘splits’ into two over time. The process is also called ‘phonologization’, since an allophone becomes its own phoneme over time. Split sounds themselves do not change. Rather it is the merger of other sounds in their environment that causes the phonemic status of the sounds inv…

Abstract Consonants are segments whose articulation involves total or partial obstruction of the airstream at some point within the vocal tract. Greek had the consonants / p t k b d g p ʰ t ʰ k ʰ m n s l r h/. 1. Introduction The Classical Attic consonant inventory is comprised of the following 15 phonemes (Lejeune 1955 and Brixhe 1996 are more detailed studies; Lejeune 1972, Meier-Brügger 1992, Rix 1992, and Sihler 1995 present the diachronic developments; Sturtevant 1940 and Allen 1973 present a wealth of evidence for pronunciation): 9 stops / p t k b d g pʰ tʰ kʰ/; 2 nasals / m n/; 2 liquids / l r/; a f…

Abstract Wackernagel’s Law is a generalization referring to the tendency of certain enclitics and postpositives to occur second within their clause or sentence. 1. Introduction Wackernagel (1892) observes that, across archaic Indo-European languages (Greek, Latin, Sanskrit, Gothic, etc.), enclitic and postpositive items tend to occur second in their clause or sentence (depending on the clitic), as in the following example from Herodotus ( = marks the host-clitic relationship; the relevant clitic is in bold): 1. eíretó=min ho Astuágēs asked-him the Astyages ‘Astyages asked hi…

Abstract The syntax-phonology interface refers to the correspondence between the syntactic structure of an utterance and its phonological encoding. Most linguistic theories divide language into computational systems, such as the phonological, morphological, and syntactic (a practice that goes back at least to Morris 1938 within the linguistics literature). The syntactic component is responsible for the construction of sentences, the phonological component for how those sentences are pronounced. The syntax-phonology interface refers to t…

Abstract The utterance is the largest recognized prosodic unit, roughly equivalent to a sentence in syntax. The utterance (υ) is the largest recognized unit of prosody (Nespor and Vogel 1986:221-248) and is usually bordered on either side by silence. Part of the syntax-phonology interface, it tends to correspond with the syntactic notion sentence (simple, complex, or compound), but often consists of less (Selkirk 1980; Nespor and Vogel 1986:221-223), as the beginning of Plato’s Crito (43a) illustrates. The dialogue begins with two utterances (marked off in square bra…

Abstract Syllabic consonants are segments articulated as consonants that nonetheless occupy the nucleus of a syllable, as for instance in the final segment of English even [ˈiːvn̩]. Although syllabic consonants are not unambiguously present at any stage of attested Greek (but see below for the proposal of Heubeck 1972), their development from Proto-Indo-European has received an enormous amount of attention. A set of four syllabic consonants (two nasals, [ṃ] and [ṇ], one lateral [ḷ], and one rhotic [ṛ]) are reconstructed for Proto-Indo-European as…

Abstract The phonological phrase is a constituent in the prosodic hierarchy that, prototypically, corresponds to sub-clausal noun-, verb-, and adjective-phrases in syntax (XPs). The phonological phrase (φ), also known as the minor phrase or minor phonological phrase, is the layer of prosody immediately beneath the intonational phrase that combines prosodic words and clitic groups into a single prosodic unit (Nespor & Vogel 1986:165-186; Selkirk 1980, 1986; Horne 1986; Devine & Stephens 1990; Truckenbrodt 1995, 1999; Devine & Stephens 1994:377).  Metrically, the phon…

Abstract A dipthong is a pair of vowels that occupy the same syllable. Ancient Greek had a number of diphthongs, including ai and oi. A dipthong is a pair of vowels that occupy the same syllable. Thus two-syllable diá ‘through’ does not have a diphthong but one-syllable paîs ‘child (nom.)’ does. Classical Attic has an inventory of eleven diphthongs (see generally Allen 1987:79-88; for a diachronic overview see Rix 1992:46-49, 51-52):   “ Short” Diphthongs   “ Long” Diphthongs     /yi/ <υι>   /εːi/ <ηι> /εːu/ <ηυ>   /oi/ <οι> /εu/ <ευ> /ɔːi/  <ωι>   /ɔːu/ <ωυ>   /ai/ <αι>  /au/ <αυ> /aːi/  <αι> /aː…

Abstract The intonational phrase is the prosodic constituent immediately below the utterance and immediately above the phonological phrase; it corresponds prototypically to the clause in syntax. The prosodic constituent immediately below the utterance and immediately above the phonological phrase is the intonational phrase, which has been the subject of extensive research cross-linguistically. As a result, the unit is known by a range of different terms in the literature, including intermediate phrase (Pierrehumbert 1980), tone unit (Halliday 1967), and intonation unit (Cha…

Abstract Pitch is the auditory sensation of frequency. It is used in Greek to mark accent. When we hear a sound with regular vibration (such as that of a vowel), the pitch sensation that we perceive corresponds closely to the frequency of vibration of the vocal cords: the higher the frequency of the vibration of the vocal cords, the higher the pitch we perceive; the lower the frequency, the lower the pitch (see Gussenhoven 2004:1-11, Devine and Stephens 1994:157-194). The number of cycles per second (or any other measure of time) is known as the fundamental frequency (or FØ “f-zero”) of a…

Abstract Voicing refers to an assimilative process by which a voiceless sound becomes voiced; the term can also be used in reference to the feature [voice], which involves vibration of the vocal cords. Voicing is a process whereby a consonant changes from voiceless to voiced (the reverse is called devoicing). In Ancient Greek the change is always due to assimilation with an adjacent sound, either a vowel (a, i, u), sonorant (m, l, w), or voiced obstruent (b, d, g, v, z). Greek stop clusters always agree in voicing (Probert 2010:94), so voicing assimilation can be observed whenever t…

Abstract Vowel length is the perceived duration of a vowel sound (for consonant length, see Geminates). Vowel length is phonemic in Ancient Greek, as minimal pairs like the following show: - tṓ ‘the two’ (nom.-acc. dual) : tó ‘the’ (nom.-acc. sing.) - hós ‘who’ (masc. sg.) : hṓs ‘like, as’ Since at least Hermann (1816) vowel length has been measured in Greek in terms of moras: a short vowel has a single mora, while a long vowel or diphthong has two (Allen 1987:100; 1973:161-63). Accentuation is conditioned by the moraic structure of a word and thus by length.   The actual duration of Greek v…

Abstract Dissimilation is a phonological process in which one segment becomes less similar to another segment with respect to a given feature.  Dissimilation is a phonological process in which one segment becomes less similar to another segment with respect to a given feature. Regular synchronic alternations involving dissimilation are much rarer than those involving assimilation (for a general overview of dissimilation, see Aldrete and Frisch 2007, and Bye forthcoming; Suzuki 1998 is a typological study of the phenomenon). Diachronically, dissimilation tends to …

Abstract Phonotactics define the permissible combinations of sounds within a word. Phonotactic constraints define permissible sequences of consonants and vowels (for a detailed description of licit consonantal combinations, see Lupaş 1972:133-41 and Steriade 1982:293-4; for vowels, see Lupaş 1972:141-51). We illustrate this here with consonant phonotactics in Ancient Greek. Within a word, Attic Greek permits consonants to cluster with considerable freedom (for possible segmental sequences between words see Sandhi). In this article we will loo…

Abstract Truncation is the shortening of a word or stem by removing part of it and leaving the rest intact, e.g., Pete < Peter or Pat < Patricia. Truncation is the shortening of a word or stem by removing part of it and leaving the rest intact. Simple cases from English include (personal names) Michael > Mike [maik], Christopher > Chris, Peter > Pete, Ronald > Ron, etc. There is some variation in the usage of the term; Joseph (2001:352), for instance, uses it to describe the reduction of phrases like ‘Watergate affair’ to simply ‘Watergate’. See Zwicky and Pullum (1987), Spencer (1991) for gene…

Abstract Vowels are segments articulated with an open vocal tract. Greek had twelve vowels: /i iː y yː e eː ɛː o oː ɔː a aː/. The vowel system of late fifth-century Classical Attic is well known for its asymmetry, as it has almost twice as many front vowels as back. Considerable debate surrounds this topic, and one is hard pressed to find any two (synchronic or diachronic) accounts of the Attic vowel system that agree in all their details. See Brixhe (1996); Threatte (1980) for the inscriptional evidence; Lejeune (1972), Meier-Brügger (1992), Rix (1992), and Sihler (1995) for diachronic ov…

Abstract A movable consonant is a segment occurring at the end of a word that alternates with zero under certain conditions. The term “movable consonant” refers to a set of lexically-specified consonants ( n, s, k) that alternate with zero under certain conditions at the edge of a word. In the linguistics literature, they are more often known as “latent segments” (e.g. Hansson 2005); Devine and Stephens (1994:252) also use the term “antihiatic consonant”. Of the three movable consonants, nu, which was termed nu ephelkustikón (‘attracted, suffixed’) by ancient grammarians, has b…