[Editorial note: [...] indicates use of Coptic text. Original script is available for viewing in the PDF format of this article.]
(CE:A131a-A133b)
GEMINATION, VOCALIC. Fairly frequently, Coptic manuscripts present examples of graphic vocalic gemination (duplication of various graphV= graphemes called “vowels”: [...], [...], [...], [...], [...], [...], and [...]; to be distinguished from V = vocalic phonemes, which, in addition to /a/, /e/, /[...]/, /[...]/, /i/, /u/, and /[...]/, include the sonants /[...]/, /[...]/, /[...]/, /[...]/, and /[...]/). This article will ignore the nonsystematic cases, arising from causes that produced the IDIOLECT of some scribe insufficiently trained and incapable of adhering unreservedly to one well defined orthographic system, to one dialect (cf. Worrell, 1934, pp. 110-11; Kasser, 1980, pp. 78-82). The discussion here is restricted to the series of cases that have a systematic cause.
Two types of graphic vocalic gemination are found (graphV + same graphV), each belonging to certain particular dialect(s) or subdialect(s), of which it is one of the fundamental characteristics.
The first of these types of gemination is exceptional. It may be observed in the examples of the most archaic orthography of the text of P. Bodmer VI, the only existing witness of DIALECT P, which is either possibly a kind of proto-Sahidic according to Vergote (1973b), or (looking indeed often very like a kind of reconstructed proto-Sahidic) is rather another variety of Coptic PROTODIALECT, perhaps some sort of proto-Theban (Nagel, 1965; Kasser, 1982a), if not kind of proto-Sahidic immigrant in the Theban Region and strongly influenced by the local (nonliterary) Theban idiom (see Kasser, 1985, and DIALECT, IMMIGRANT). Whatever it may be, that type of gemination only appears sporadically in these examples, being always under strong competition from examples of a more evolved orthography (see below).
There, the tonic vowel of the lexeme is systematically duplicated. (These forms remain very much in the minority, about 2 percent of the whole, to which may be added 3 percent of forms in which an unusual gemination of the consonant immediately preceding or following the tonic vowel or the first consonant of the syllable containing the tonic vowel seems to have been produced through negligence instead of the vocalic gemination mentioned above.) Contrary to what happens in regard to the second type of vocalic gemination (see below), the first is not motivated by any etymological factor; the formula for this first type is thus graphV + same graphV = tonic V. Here are some characteristic examples: P [...], existence; [...] (sic), loss; [...]. [...] (sic), houses; [...]. [...] (sic), fathers, parents; [...], dew; [...], pretext; [...], to smooth; [...] (sic) or [...], head; [...], wages; [...] (sic), voice; [...] (sic), road, way; [...] (sic), to produce; [...] (sic), to empty; and even the Copto-Greek [...] (= [...]), breath.
Readers will have noted the supralinear stroke that quite often joins the top of these graphV; appearing perhaps at a late stage to distinguish (in P) these geminations from those of the second type (see below), which are phonologically very different, this stroke could well have here, as it usually does elsewhere in Coptic, a syllabic significance, indicating that this gemination of graphV does not express a “broken vowel,” a “hiatus,” or any analogous phenomenon, and this no more in bradysyllabication (corresponding to slow speech) than in tachysyllabication (corresponding to natural, rapid speech; see SYLLABICATION); it expresses a single V ( vocalic phoneme) simply tonic (= stressed).
The second, much more common type of vocalic gemination appears systematically (or nearly so) in the least archaic forms of the dialect P, as well as the dialect i (= pL, protodialect of L), and especially the idioms A, L4, L5, L6, V5, F5, and S (but not in M, W, V4, F4, B and its subdialects, and G). The purely vocalic aspect of this gemination is deceptive, since its formula is graphV + same graphV = tonic vowel + consonant (the demonstration will be found under ALEPH). Those who have sought to analyze this gemination phonologically have in fact very soon realized that its appearance in Coptic most often coincides with the disappearance of an older Egyptian radical consonant. (Some lexemes not affected by this disappearance later took on the vocalic gemination by simple analogy with lexemes that were superficially similar.) However, the conclusions of these investigators have not, from the outset, been unanimous (Kasser, 1982c).
Stern (1880, p. 54): “We understand it [the duplication] as a breaking of the vowel, and compare the stem affected with the Semitic roots expanded by [...] (mediae quiescentis).” But if one notes that for this author the Coptic “breaking of the vowel” seems indeed to be a kind of diphthongizing; that for him (pp. 34-35) the diphthong is the (syllabic) combination of a vowel with (after it) a semivowel (= glide); that he states (pp. 29-30) that ancient Egyptian had three semivowels (j = Coptic /j/ = [...], w = Coptic /w/ = [...], and finally ‘), which often became the second element of a vocalic gemination in written Coptic; and that, broadly speaking, one may thus say that all these semivowels appear as graphV in Coptic, then one may suppose that Stern tended, if not always, at least frequently, to identify the second element of the vocalic gemination in Coptic phonologically as an /‘/. Lacau (1910, pp. 77-78), while analyzing the phenomenon with much finesse and perspicacity, nevertheless seems to have admitted tacitly that the duplication of the vowel caused by the dropping of the consonant i, ‘, r, or t is equally a vocalic duplication on the level of superficial phonology, the second V of this gemination replacing those consonants which have effectively disappeared and for which nothing has been substituted, not even some /‘/ derived from them (which seems acceptable in bradysyllabication, but debatable in tachysyllabication).
Steindorff (1930, pp. 34-35) presented a distinctly different position: “In Sahidic, in those syllables which have been opened through the suppression of a following consonant ..., the short medial vowel is frequently doubled: ... [...] eere “daughter” for [...], [...], [...] ... This process is called a compensating duplication; it is a substitute for the lengthening of a short vowel which appears in an open syllable.” Later Steindorff (1951, pp. 34-35) adopted a less clear position, apparently seeking to harmonize his earlier explanation, in modified form, with other explanations that come into play; thus, he subsequently distinguished the cases of compensating duplication from those of “vowel assimilation,” on the one hand, and those of “breaking the vowel,” on the other (see above).
Kuentz (1934) examined these various possibilities and finally proposed the idea of a compensating (vocalic) lengthening (Ersatzdehnung, not Ersatzverdoppelung); in reply to the objection that in Coptic [...] and [...] are the long forms of [...] and [...], and hence that instead of, for example, [...], pasture, feed, one ought to find [...], Kuentz supposed that “at some undetermined period the old opposition of quality became an opposition of timbre; no doubt [...] and [...], representing old long vowels, became closed vowels, while o and e were open vowels. Thenceforth the graphic duplication of the various vowels examined is naturally interpreted as a notation for long vowels, whether open or closed, at the period when this system of writing was put into use.”
Till (1929) was the first to express clearly the idea that the second element (graphV) of the graphic vocalic gemination examined here must represent a consonantal phoneme, without, however, venturing to say which. Later Till (1955, p. 46) became more precise (though still ambiguous): “‘Aleph and ‘Ajin are still present in Coptic, although no special letters for them exist. Both may have been expressed alike (probably ’), although ‘ in some circumstances has a different effect on neighbouring vowels from [...].” And Till (1961, p. 10) wrote that “the vowel written double is to be understood as a simple vowel + Aleph or Ajin.” This author (perhaps under the influence of Vergote, 1945, pp. 89-91) thus very clearly comes close to the solution most generally admitted today, according to which it is always ALEPH that the second element in the vocalic gemination renders (Till seems to have seen there sometimes /’/, sometimes /‘/, but then it is a /‘/ practically pronounced /‘/); however, the ambiguity of his position suddenly appears again in a different fashion in his suggestion (Till, 1955, p. 46) that this /‘/ “was evidently no longer felt to be a consonant” and in his transcription of [...] by [...] (p. 46) but of [...] by [...] (p. 259).
Edgerton (1957, pp. 136-37) adopted a position resolutely opposed to that of Till, refusing to admit the phonological survival of /’/ or /‘/ in Coptic, from the time when they were not represented by any grapheme of their own (the problem of the CRYPTOPHONEME): “It seems simplest to explain the non-existence of signs for ’aleph and ‘ajin in Coptic writing by assuming the non-existence of these phonemes in Coptic speech.”
Finally, Vergote (1945, pp. 87-96; 1973, Vol. 1a, pp. 12-15, and Vol. 1b, pp. 31-37) clearly demonstrated that aleph is the consonantal phoneme best suited for replacing ancient [...] , ’, r, and t (and even j or w), which have disappeared. His opinion was entirely shared by Kasser (1982c), who, however, thought that the graphic aspect (graphV + same graphV) of this gemination (the orthography corresponding largely to br/syl. = bradysyllabication, an artificially slow articulation, in which this gemination is effectively vocalic even in phonation [V tonic + same V atonic]) is to be distinguished radically from its phonetic and phonological expression in normal articulation (in t/syl. = tachysyllabication, where this gemination renders a tonic vowel followed by /’/): thus, [...], to put it, [...]. [...] (monosyllabic, cf. Vergote, 1973, Vol. 1a, p. 45), but [...]. [...] (disyllabic); or [...], ear, [...] (disyllabic) but [...] (trisyllabic). The transition from t/syl. to br./syl. would entail a kind of “echo effect” resulting in this vocalic gemination which appears in Coptic orthography: thus, for example, ‘/mt/[...]e/’> */m[...]a[...]e/> ‘/m[...]a[...]e/’ (an idea the first expression of which could already be found in Vergote, 1945, p. 91, and which, taken up a little differently, is developed in Kasser, 1981, pp. 7-9; 1982b, p. 29, n. 23; 1982c, pp. 33-34).
RODOLPHE KASSER
