Alphabets, Coptic

              [Editorial note: [...] indicates use of Coptic text. Original script is available for viewing in the PDF format of this article.]

(CE:A32a-A41a)

ALPHABETS, COPTIC. Attentive study of the alphabet used in each of the various Coptic DIALECTS and subdialects obliges one to recognize that there was not a single Coptic alphabet, as is often believed, but several Coptic alphabets (or to put in a slightly different way, several varieties of the Coptic alphabet). Certainly, if one examines the Coptic texts themselves in their manifold variety, one finds that one of these alphabets, the of S etc., is employed almost everywhere (it is that of almost 92 percent of the total); it is therefore comprehensible enough that the alphabet of S should practically always be called “the Coptic alphabet,” without further explanation, while the alphabetic variety of B etc. (almost 7 percent) is only very seldom mentioned, the existence of the twelve other varieties (about 1 percent only) being entirely neglected. However, all these fourteen alphabets, major and minor, will be accorded the place which they have a right herein.

It will be convenient to recall in the first place that each type of Coptic dialect—whether a dialect in the narrow sense of the term or a PROTODIALECT, a MESODIALECT, a METADIALECT, or even a subdialect—is defined first and foremost as a phonological system, while morpho(phono)logy and morphosyntax intervene only secondarily in its definition (for want of evidence sufficiently frequently attested).

Coptic has been a dead language for several centuries, and its demise preceded the beginnings of Coptology as a modern discipline. Hence, Coptic PHONOLOGY can only be known today through the orthography of the Coptic tetxs that have survived, a very small number in comparison with the immense quantity of those that perished in the tempestuous and painful course of Coptic history; through their regular and systematic orthography only (that which we find in the texts of “good” quality), and not through the irregular and disordered graphical manifestations that may be observed in all sorts of careless and orthographically undisciplined copies.

Prudence certainly obliges one to remember that the analysis of a Coptic orthographic system is not automatically the analysis of a Coptic phonological system. One must always reckon with the possibility, however weak it may be, that the difference between the various Coptic alphabets may be not only a difference of quantity (phonemes and graphemes in more or less large numbers) but also, on some particular point, a difference of quality (a given grapheme rendering a given phoneme in one idiom, and the same grapheme rendering another phoneme in another idiom; or a given phoneme being rendered by a given grapheme in one idiom and by another grapheme in another idiom). However that may be, if one may sometimes doubt that such an original alphabetic system really attests a particular Coptic dialect, it nonetheless remains that each particular Coptic alphabetic system is a piece of evidence; and this reality, although superficial in relation to phonology, deserves to be recognized as such on its own (the alphabetic) level; this recognition in fact, is not bound to the phonological interpretation, sometimes uncertain, of the graphemes that compose the alphabets.

The uncertainty of this interpretation at any given point often derives from a question of principle and from a methodological alternative of which one must be very conscious. In fact, the investigator who strives to rediscover and analyze the phonological systems of the Coptic dialects and subdialects through their orthographical systems is soon constrained to choose between two preliminary working hypotheses: each of these presents substantial advantages, but even in the better hypothesis, they remain limited and weakened by important disadvantages.

The first hypothesis consists in postulating a priori a phonological unity of the Coptic language, a unity practically absolute. In pushing this hypothesis to its extreme consequences, one would have to admit that despite the orthographical appearances, this language is by no means divided into a plurality of dialects. The differences that orthography seems to manifest would be only superficial, or to put it simply, the various schools of scribes would make use, in certain cases, of different graphemes to express the same phonemes. One should then observe in Coptic not various dialects but various “orthographic codes” applying to a language that is “one” and not divided on the phonological level.

This hypothesis is very seductive because, over against the various earlier stages of the Egyptian language (apparently homogeneous because dialectal multiplicity does not appear in it, or practically not), it sets not a group of Coptic phonological systems but a single Coptic phonological system (or, at the very most, a group of systems that differ among themselves only very rarely and on details that are truly exceptional). It seems to be confirmed by the fact that, if one compares with one another the different orthographical forms of the same autochthonous Coptic words, it appears evident that the [...] of the majority of dialects and the [...] of [...] render [...]; the [...] of the majority and the [...] of G render / /; the [...] of the majority and the [...] of [...] likewise render [...]; the [...] of [...] and [...], the [...] of [...] and [...], and finally even the [...], of B7 and G (see below) render /x/; the [ of the majority and the [...] of [...] render /c/; and so on.

However, the limits of the efficacy of this explanation are reached when one is faced with problems such as these: when [...] in F corresponds to [...] in S, the phoneme cannot be either /l/ or /r/ uniformly for [...] and [...] at the same time; and likewise, when [...] in A corresponds to [...] in S etc., the phoneme cannot be either /x/ or / / uniformly for [...] and [...] at the same time. By themselves alone, these exceptions prove that there is in Coptic a dialectal plurality.

The second hypothesis consists in postulating that in Coptic, according to the unanimous intention of the creators of its alphabet, there is for each phoneme (or each combination of phonemes, should occasion arise; cf. /th/ etc. below) a single corresponding grapheme that can never serve to express another phoneme. This “law of exclusiveness” thus does not allow of any plurality of Coptic alphabets based on a difference of phonological “quality”; this plurality can only exist as a consequence of the “quantity” of the indissoluble phoneme-grapheme unities: certain Coptic idioms would make use of the Coptic series to the full (or nearly so), while other idioms would content themselves with a very diminished series (loss of /x/ or even [...], /h/, or /c/, for example).

However, if this hypothesis gives very satisfactory results in regard to F [...] versus S etc. [...] or A [...] versus S etc. [...], for example, it seems to fail in other cases already mentioned above: in comparing the different orthographical forms of the same Coptic word, it is difficult to see what phoneme P could render if not /k/ (/q/ seems excluded for solid reasons; cf. Kasser, 1980b, pp. 244-48); in the same way, it seems unlikely that, corresponding regularly to [...] in B etc., the combination of graphemes [...] in G should render /sz/ rather than [...]; and when the [...] of G corresponds to [...] /f/ in B etc., would it be equivalent to /ph/ rather than to /f/? The rigid application of the law of exclusiveness would entail other phonological solutions that would be bizarre and difficult to accept. One must then resign oneself to some compromise between these two extreme hypotheses, a compromise to be negotiated and determined from case to case.

One particularly troublesome alphabetic fact (above all, in consequence of the second hypothesis and its law of “exclusiveness”) is the use of (apparently) the same grapheme to express two different phonemes. When this phenomenon coincides with the opposition of two Coptic idioms, as with [...] versus B [...], one may attempt to explain it by referring to the divergent principles applied by two schools of scribes belonging to two different cultural ambiences. But what is one to say of this ambiguity when it appears within one and the same Coptic orthographical system (and probably idiom)?

Thus, in P (compared to S etc. in the orthography of their common vocabulary) [...] is assuredly /k/ in the Copto-Greek words, but it is /c/ in the autochthonous Coptic words. (It is difficult to imagine that P, by some palatalization, or “damping,” applied to the Greek words, should have systematically replaced by /c/ all the /k/ in its Copto-Greek vocabulary.) One sees the same ambiguity in G, where one finds [...] both in the Copto-Greek words (for /ph/) and in the autochthonous Coptic words (for /f/, which could, however, at least locally, have become the articulation of v even in Copto-Greek). The same confusion is manifest in the Dublin L5 (cf. below), J (sporadically); and F9, where [...] appears both in the Copto-Greek words (for /kh/) and in the autochthonous Coptic vocabulary (for [...]); there is furthermore the problem of the use of [...], even for /x/ or that of the inverse use of [...] for [...] in some Copto-Greek words in L5. (The relation of [...] and [...] or /x/ is probably of another order, or at least too subtle and complex to be summarily expounded here.) Finally, a similar ambiguity appears in F7 and in some L6 texts (cf. Kasser, 1984-85), where the grapheme [...] is both the “normal” Coptic [...] /ps/ and the “normal” Coptic [...]. These are texts attested by manuscripts nearly all particularly ancient: F7 is the language of the bilingual Papyrus No. 1 of Hamburg (Coptic and Greek, end of the third century). F9 is the language of Coptic glosses in a manuscript in the Chester Beatty collection (third century; cf. Kasser, 1981a, pp. 101-102). J is the language of a Coptic schoolboy’s tablet (end [?] of third century; cf. ibid., pp. 113-15). An unpublished papyrus in Dublin (cf. Kasser, 1984), p. 274) seems to be contemporary with the preceding two or scarcely much later; it contains John 10:18-11:43 and 12:14-39 in a variety of L5 with very particular orthographicophonological characteristics (Kasser 1981b, pp. 27-29).

The following hypothetical explanation could however, to some extent resolve these diverse enigmas, except for the ambiguity of [...] in G. One should have in each instance, for two phonemes undoubtedly different, not a single grapheme considered (except for [...]) as of Greek origin but two graphemes to be distinguished from one another, the one of Greek origin, the other of demotic—two graphemes distinct in their origin but which the autochthonous Egyptian sign has gradually been so strongly influenced in its form by the Greek grapheme that it has become practically identical to the latter (whence the confusions that ensue). In what follows, except for special mention, references are made to du Bourguet (1976, p.75), where the demotic signs are presented in their “usual” form on the left and then in certain of their “variants on the right.

In P, the autochthonous [...] could be derived from a sign for g, the first of the variants, resembling a very “flattened” [...]; this ambiguous usage, alongside [...] /k/ in the Copto-Greek words, will have led to the usage in S of [...] (no longer [...]) for /k/ throughout, and [...] (no longer [...]) for /c/.

In the Dublin L5, with J and F9, the autochthonous [...] could be descended from a sign for d, the first of the variants (it has the look of a slightly upturned , of which the first stroke, which in Greek goes from top left to bottom right, is near the vertical and the second stroke is consequently near the horizontal); confusions between the [...] /kh/ and this [...] will have led to the grapheme [...] being soon preferred to it; this is also descended from a sign for d, either the first usual sign (vaguely resembling an a the loop of which has been completely flattened) or the last of the variants (resembling a bulging [...] with the rounded part at the bottom and the two horns at the top).

In F7 and some L6 texts, the [...] /ps/ is exactly identical with this grapheme as one sees it in the contemporary Greek manuscripts, which also makes it unfortunately almost identical with the autochthonous [...] /ti/, habitually considered as derived from a demotic sign (cf., e.g., Steindorff, 1951, p.12; but see also Kasser, 1984-1985); this ambiguity will have incited the Coptic scribes to modify into [...] the grapheme for /ps/ borrowed from the Greek alphabet.

With regard to the two [...] (the S one for /c/ and the B etc. one for [...]), one remains within the autochthonous Egyptian zone, without interference from the Greek alphabet . It is admitted that the [...] /c/ derives from a demotic sign for [...], the first usual on (which has the appearance of a [...] the circle of which is very small, the stroke that escapes from it leaving at the summit and stretching horizontally at length to the right). The [...] could be descended from the demotic sign for d, the last of the variants (the one that resembles a bulging [...] of which the rounded part would be at the bottom and the two horns at the top; cf. above with regard to [...] and [...] for [...]; the suppression of the left horn could well yield a kind of[...].

As can be seen, this compromise obliges one to renounce the thesis of the “absolute phonological unity” of the Coptic language; there are then several Coptic dialects, a fact that is incontestable. But this compromise probably obliges one to consent to some detractions from the law of exclusiveness, which flows from the second hypothesis: if (admitting the duality of [...], above) one is to see each grapheme always rendering the same phoneme, it may come about that a phoneme is rendered, according to the idiom, by two or even three different graphemes, as with [...] and [...] for [...]; [...] and [...] for [...] ; [...] and [...] for /f/; [...] and [...] for [...]; [...] and [...] for /c/; [...] and [...] for [...]; [...], [...], and , [...] for /x/. In this last case, one might think of the influence of a local Greek articulation [...] /x/ rather than [...] /kh/ and extending to the pronunciation of the Copto-Greek words in B7 and F9 (cf. also the eventuality of [...] /f/ in G above. There are then in Coptic not only several Coptic dialects but also several Coptic alphabets employed to render these various dialects, the limits of the field of application of these alphabets not always coinciding with the phonological interdialectal limits.

From this perspective, a search through the most diverse Coptic texts finally ends in the identification of at least fourteen different Coptic alphabets. As was underlined at the beginning of this article, they vary considerably in their relative importance if one takes account of the number and the extent of the texts that employ each of them. One of these alphabets is supported by S, the vehicular language of the whole of the valley of the Egyptian Nile (the Delta excepted); it was also utilized by a large number of dialects and subdialects in the valley and even in the Fayyum. Another alphabet is supported by B, the vehicular language of the Egyptian Delta. These are, one might say, the “classic alphabets” of the Coptic language. The other alphabets are supported only by a small number (on occasion even by a trifling number) of texts slight in extent and variety or even by a single small text, the sole representative of an idiom whose original character, on the level of the dialect and not simply on that of the IDIOLECT, remains open to discussion: hence, one may call them, respectively, “marginal” and “very marginal” alphabets.

An alphabet would be marginal for various reasons. It could be situated in the “preliminary historical margin” of Coptic literary life; this would be one of the various alphabets created by way of essays at the time when the pioneers of the literary Coptic were attempting as individuals or in small isolated groups to forge the instruments indispensable for the realization of their enterprise; some of these alphabets would not have obtained the adhesion of a social or cultural group, so that they would very quickly have been abandoned, even by their few partisans. Or again, an alphabet could be marginal because it was descended from the inventive spirit of an individual or a small group living somewhat on the margin of the Society that was contemporary with them, and this at least on the cultural or eventually the religious level; this alphabet would have known only an extremely restricted diffusion and an existence probably all too short. But whether “classic” or more or less “marginal,” all the Coptic alphabets thus attested by the texts of this language will be of interest for the researcher.

A synoptic view of the fourteen Coptic alphabets mentioned above will be found in Table 1. Each of the alphabets is indicated either by the unique dialect or subdialect that attests it or by the principal idiom (language or dialect) that attests it. These are as follows:

(I) P (= a PROTODIALECT remarkably similar to *pS, the latter being the tentatively reconstructed proto-Sahidic idiom; cf. DIALECT P), 0.1 percent of the whole Coptic textual surface.

(II) i (= pL, proto-Lycopolitan dialect; cf. PROTODIALECT and DIALECT i), 0.01 percent.

(III) A etc. (= A, Akhmimic dialect, with i7 = p’L, an evolved proto-Lycopolitan dialect; cf. AKMIMIC and PROTODIALECT with DIALECT i), 0.6 percent.

(IV) B etc. (=B5, the “classic” and relatively late BOHAIRIC language, commonly designated by B, with B4, the Bohairic dialect of “ancient” attestation, cf. Kasser, 1981a, pp. 92-93; B4 texts published in Husselman, 1947; Quecke, 1974; Kasser, 1958, only p. 53 of Papyrus Bodmer III), in all, more than 6.5 percent (cf. LANGUAGE(S), COPTIC).

(V) S etc. (= S properly so called, the SAHIDIC language, with [a] the variety of the Fayyumic dialect showing lambdacism most recently attested, F5 [cf. FAYYUMIC]; [b] V5, the least widely attested of the two mesodialectal varieties of the Fayyumic dialect without lambdacism (cf. Fayyumic); and [c] the three subdialectal varieties [or even dialects entirely apart, according to Funk, 1985] of L, the LYCOPOLITAN or LYCO-DIOSPOLITAN dialect), as a whole nearly 92 percent.

(VI) M etc. (= M properly so called, the MESOKEMIC dialect, with [a] the crypto-Mesodialect called dialect W [published in Husselman, 1962]; [b] V4, the most widely attested of the two mesodialectal varieties of the Fayyumic dialect without lambdacism; [c] F4, one of the two varieties of the Fayyumic dialect with lambdacism and of ancient attestation [F4 and F7]; and [d] the two very small mesodialects K and K7(1) [cf. Kasser and Satzinger, 1982]), in all 0.7 percent.

(VII) B7 (cf. Kasser, 1981a, p.93; subdialect), a little less than 0.001 percent.

(VIII) L5 (Dub.) (= the particlar subdialect of the Johannine fragment L5 in Dublin), 0.015 percent (cf. LYCOPOLITAN or LYCO-DIOSPOLITAN).

(IX) F7 (the FAYYUMIC subdialect F7, of ancient attestation; cf. Kasser, 1981a, pp. 97-100), 0.05 percent.

(X) F8 (a very small subdialect; cf. Kasser, 1981a, p. 101), a little less than 0.001 percent.

(XI) J (very small subdialect; cf. Kasser, 1981a, pp. 113-15), 0.001 percent.

(XII) G (=DIALECT G or mesodialect [?]; cf. Crum, 1939, Kasser, 1975, and 1981a, pp. 102-103), 0.005 percent.

(XIII) F9 (very small subdialect; cf. Kasser, 1981a, pp. 101-102), a little more than 0.0005 percent.

(XIV) H (=DIALECT H or metadialect; cf. Kasser, 1966; 1975-1976; 1981a, pp. 104-112), 0.03 percent.

In the synoptic table (Table 1), everything has been grouped around S etc., the attestation of which, in relation to the other Coptic idioms, is very amply preponderant (92 percent). This is why (always with the exception of [...]; see below) the order of the phonemes (or combinations of phonemes, should occasion arise), along with the alphabetic order corresponding to them, is first of all the one habitually found in the Coptic grammars and lexicons (or dictionary elements) limited to Sahidic, S: (1) First come letters of Greek origin. (2) Next come letters of demotic origin (to the extent that they are in use in S; with regard to the debated origin of [...], see below; for details on the origin of these non-Greek graphemes, see ALPHABETS, OLD COPTIC). As regards the alphabetical order of the Coptic letters of demotic origin, it should be mentioned here than in some ancient documents showing that part of the Coptic alphabet, [...] is placed before [...] (e.g., Hall, 1905, pp. 35-36; Krall, 1988, pp. 129-30; question raised in di Bitonto Kasser, 1988). (3) Last come various “supplementary” phonemes (or combinations of phonemes), almost all descended from demotic (or their precise origins, see ALPHABETS, OLD COPTIC): deriving from pre-Coptic Egyptian, ALEPH = /’/ is a CRYPTOPHONEME in S, as elsewhere in Coptic (except in P), and is therefore not rendered by any grapheme exclusively its own; also of autochthonous origin and pre Coptic, [...], /x/, and eventually also [...] were abandoned in S etc., but have been preserved in other dialects or protodialects; [...] is specific to the Bohairic domain (B etc. and probably also B7); /v/ is the characteristic of a tendency that results in various manifestations of Coptic metadialects. In regard to the particularly varied graphemes that correspond to these supplementary phonemes, several will be noted in P that can be described as Old Coptic (so [...], in addition to the [...] and [...] , already mentioned above, which respectively in P alone replace the [...] and [...] of Greek origin); others ([...] and [...], or [...] and [...] with their diacritical signs, which, respectively, mark them off from, and oppose them to, [...] and [...], which themselves belong in group 2) are simply of demotic origin; others finally ([...] and [...]) are, or seem to be, Greek graphemes, but here play an unaccustomed role.

As regards the correspondence between the phonemes and he graphemes of the various Coptic alphabets, it will be noted that in this area the situation in Coptic is very similar to that in Greek. Thus, broadly speaking, for each Coptic alphabet there is a corresponding phonemic series made up of either (most frequently) isolated phonemes (e.g. /a/, [...], /g/, etc., the affricate [...] = [...] also being considered as a “single” phoneme) or of combinations of phonemes (/th/, /ks/, /ph/, /kh/, /ps/, /ti/).

Taking into account the total phonemic series resulting from the addition of all the particular phonemic series, each of which corresponds to one of the fourteen individual Coptic alphabets, one may establish, from case to case, the existence of one or another of the five following possibilities:

(1) In a given Coptic alphabet, a given phoneme (or combination of phonemes) is rendered by a grapheme that, according to a system of correspondence usual to this alphabet as well as to (almost) the whole of the Coptic dialects and subdialects, is peculiar to it and serves for its exclusive use: thus, for example, in S etc., [...] for /a/, [...] for /b/ etc., [...] for /th/ etc., [...]for /k/ etc. (it will be noted in this regard that, according to rules which cannot be set out here [cf. Kasser, 1983], /i/ may be rendered either by the grapheme [...] or by the combination [...], although the same [...] or [...] may equally render /j/; and if /u/ is almost always rendered by the combination [...], nevertheless /w/, normally rendered by the same combination, often sees its initial o disappear in orthography after [...], [...], or [...], which conveys the illusion of a correspondence [...] = /w/). This type of peculiar and exclusive phoneme-grapheme relationship is in the Coptic alphabets the most normal and widely favored possibility; every grapheme that attaches to it—and likewise the eventual combination [...] or [...]—is then noted, just as it is in the synoptic table herein.

(2) In a given Coptic alphabet, a given phoneme is rendered by a given grapheme that, according to a system of correspondence usual to this alphabet, although not to the other Coptic dialects and subdialects is peculiar to it and serves for its exclusive use: thus, for example, [...] for /c/ in P, although everywhere else [...] = /k/; [...] for /[...]h/ in B, although everywhere also [...] = /c/ (but see earlier discussion of law of exclusiveness). This possibility is not the most normal, but it remains very widely favored; this grapheme is then also noted, just as it is in the synoptic table.

(3) In a given alphabet, a given combination of phonemes, instead of being rendered by the grapheme usually peculiar to it, is rendered by each of the graphemes that habitually render each of the components of this combination: thus [...], and not [...], for /ti/  (one could likewise imagine [...], and not [...], for /th/, etc.). This possibility, scarcely less widely favored than the preceding ones, is marked by the symbol = in the synoptic table, the complementary explanations being found in the commentary following the table.

(4) In a given alphabet, a frankly more problematic case, the proper and exclusive grapheme for a given phoneme is missing; the phoneme, however, is reputed to be present despite this and is then said to be a CRYPTOPHONEME. This is, by definition, rendered by a grapheme or combination of graphemes each of which is normally appropriated to the proper and exclusive use of another phoneme: for example, in S o = /o/ in normal usage, and yet one may note the second element of oo (not o[...]) = /o’/ in graphic vocalic gemination (cf. ALEPH); or again in G, c = /s/ and [...] = /z/ in normal usage, and yet [...] (not [...]) = /[...]/. This possibility is indicated by the symbol =/= in the corresponding box in the synoptic table, the complementary explanations being found in the commentary following the table.

(5) Finally, in a given alphabet the absence of the usual grapheme (cf. point 1) signifies the absence of the phoneme concerned; this possibility, the only one that is really and fully negative, is marked by an empty shaded cell in the synoptic table.

In this table, any grapheme between parentheses is of considerably reduced usage (because it corresponds to a phoneme that is itself also of greatly reduced usage) in all of the dialect, subdialect, or group of idioms concerned (thus, e.g., the sonants /[...]/, /[...]/, /[...]/, /[...]/ in B etc.). This obviously does not apply to [...] in [...] or to [...] in [...], which signify respectively, the simultaneous existence of spellings in [...] and in [...], in [...] and in [...]. Any grapheme between square brackets has had to be restored, taking account of the probabilities (the textual base being too narrow, there has been no occasion for this grapheme to appear). In line 31, “gem.” signifies graphic vocalic gemination (see GEMINATION VOCALIC), a way of rendering /’/ in writing as a cryptophoneme and not a phoneme in the ordinary sense.

For convenience, S has been assigned the function of a norm or standard; in relation to it the following phonemic and alphabetic differences will be noted (transformations and simplifications):

Lines 2a and 32: J, G, F9, H [...] /[...]/ > /v/ (cf. l. 26): B7 also has [...] /v/, but could well have preserved [...] /[...]/ simultaneously.

Line 3: H [...] /g/ > k /k/.

Line 4: H [...] /d/ > [...] /t/.

Line 5a: H [...] /e/ > [...] / /[...]/.

Line 6: H [...] /z/ > c /s/.

Line 10: P [...] /k/ > S etc. (and all the other Coptic dialects etc.) [...] /k/ (cf. l. 29).

Line 15: H o /o/ > [...] /[...]/.

Line 20a: H [...] /y/ > [...] /[...]/.

Line 21: See line 26.

Line 22: See lines 28 and 35.

Line 25: F9 [...] /[...]/ > c /s/.

Line 26: J, F9, H [...] /f/ > [...] /v/ (cf. l. 2a); G [...] probably /f/ rather than /ph/ (cf. l. 21).

Line 28: With regard to [...] (L5 Dub. And J, F9) for [...], see discussion above.

Line 29: P [...] /c/ > S etc. (and all the other dialects, etc., which have this phoneme) [...] /c/ (cf. l. 10); note, however, that in J and F9 [...] /[...]/ > k /k/; F7 [ /c/ > [...] /[...]/.

Line 32: See line 2a.

Line 34: P[...] /ç/ and also i[...]/ç/ > S etc. (and L and all the other Coptic dialects, etc.) [...] /[...]/.

Line 35: P (cf. B) [...] /x/ and also i (cf. A) with J[...] /x/ > S etc. (and L and all the other Coptic dialects, etc., except A etc. and B etc. with B7 and G) [...] /h/; note B7, G [...] probably /x/ rather /kh/.

TABLE 1: Synoptic Table of Coptic Alphabets, with Commentary [See PDF version of this article]

It is difficult to know with any precision the names of the letters of the various Coptic alphabets. Those proposed by modern or semimodern grammarians all rest upon relatively late tradition and represent not the primitive forms but forms already somewhat modified (Stern, 1880, p.7; Mallon, 1907, p.7; Steindorff, 1930, pp. 6-7, and 1951, p.11; Plumley, 1948, p.1; Worrell, 1942, pp. 314-27, taken up in Till, 1955, p.40). To provide a useful, rather standardized example, even if its value may be open to debate, here are presented first the names of the Coptic letters in Plumley (1948), with some adaptation; while Plumley writes the names in Coptic letters, they are here transliterated, generally in accordance with Plumley’s system of phonetic equivalents (e = short e; [...] = long e; o = short o; [...] = long o; ch = guttural ch as in German Wehrmacht, army; g’ = hard g): [...] = alpha, [...] = b[...]ta, [...] = gamma, [...] = dalda, [...] = ey (or ei), [...]z = z[...]ta, [...] = h[...]ta, [...] = th[...]ta, [...] =y[...]ta, [...] = kappa, [...] = lauda (or lawda), [...] = m[...], [...] = ne, [...] = ksi, o = ow, [...] = pi, [...] = r[...],  [...] = s[...]mma, [...] = tau (or taw), [...] = he, [...] = phi, [...] = khi, [...] = psi, w = [...], [...] = shai, [...] = fai, [...] = chai, [...] = hori, [...] = janjia (or janjya), [...] = g(y)ima, [...] = [...].

Coming closer to the testimony of the texts in their (quite confusing) manifold witness, see also, from various periods and in Coptic, the names of graphemes noted by Crum (1939), a list completed here by that of the names of autochthonous Coptic letters that appears at the end of the (unpublished) Bodmer papyrus of the Acta Pauli (fourth-fifth centuries, L5, siglum hereafter Bod.): [...]

(1) So far as the Coptic graphemes of Greek origin are concerned, one will probably be closer to their primitive names if one names the in the Greek fashion: [...] = alpha, [...] = beta, [...] = gamma, [...] = delta, [...] = epsilon, [...] = zeta, [...] = eta, [...] = theta, [...] = iota, [...] = kappa, [...] = lambda, [...] = mu, [...] = nu, [...] = xi, o = omicron, [...] = pi, [...] = rho, [...] = sigma, [...] = tau, [...] = upsilon, [...] = phi, [...] = chi,[...]' = psi, [...] = omega.

(2) For the Coptic graphemes of demotic origin in S, the preference here is for the following forms (in the S vocalism): [...] = shai, [...] = fai, [...] = hori, [...] = janja, [...] = gima (easier to pronounce than the more exact k(y)ima), [...] = ti.

(3) For the supplementary graphemes, apart from [...], their names are unknown, so that it was necessary to create them (if possible in relation with their written form, which is certain because it can be observed, rather than with their phonological value, which is sometimes less certain): [...] /k/ = zetoid kappa; - /[...]/ = hyphen-shaped nu; [...] /[...]/ (for [...]) = Chioid janja; [...] /’/ = reverse-tau-shaped aleph; [...] = kai compendium; [...] = 9-spiraled grapheme; = crossed shai; [...] = khai or [...]ai; [...] = barred-hori; [...] /[...]h/ in B etc. = aspirated janja.

RODOLPHE KASSER