Geography, Dialectal

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

(CE:A133b-A141a)

GEOGRAPHY, DIALECTAL. A description of Egypt in terms of dialectal geography must take as its basis its physical and especially hydrogeographical characteristics. Egypt is most commonly divided geographically into two elements comparable in area, number of inhabitants, and economic and political importance, but starkly contrasted on the political and linguistic levels. On the one hand, there is the Nile Delta, the vast triangle, practically flat and often marshy, about 125 miles (200 km) on each side and bordered by the sea along its whole northern flank; its geographically open configuration favored a rather undivided (or at least not much divided) linguistic shape. On the other hand, there is the long, threadlike valley of the Nile upstream from the Delta as far as the First Cataract, a little to the south of Aswan, extending nearly 560 miles (900 km), a strip of fertile ground about 8-12 miles (12-20 km) wide in the north, but only 0.6-3 miles (1-5 km) wide in the south; it is locked between two desert plateaus of rock and sand that differ in height. The ancient oasis of the Fayyum may be considered an appendage of the valley, since an irrigation canal from the Nile could have been dug as early as the pharaonic period. Such a geographical configuration could not but favor the development of divergent dialects within the Egyptian language, whether ancient (pharaonic) or more recent (Coptic).

In traditional terminology, the Delta is generally called “Lower Egypt,” and the valley upstream from the Delta either is called as a whole “Upper Egypt” or is subdivided into “Middle Egypt” (roughly from Cairo [Heliopolis-Memphis] to al-Ashmunayn [Hermopolis] or somewhat further south) and “Upper Egypt” (comprising everything south of Middle Egypt from Dayrut or possibly Asyut [Lycopolis], or even Tima and Qaw, south of Asyut, as far as Aswan). Clearly this terminology is not without ambiguity.

It has also been suggested (Kasser, 1980a, pp. 74-76) that Egypt be divided, moving downstream, into five regions: (1) the Upper Valley, or the upper and middle parts of Upper Egypt in the strictest sense of the term (that is, from about Aswan and Philae to Tahta, north of Akhmim [Panopolis]); (2) the Middle Valley, or the lower part of Upper Egypt, in the strictest sense, and the upper part of Middle Egypt (from about al-Badari [Qaw] to north of al-Bahnasa [Oxyrhynchus]); (3) the Center (so called because its situation makes it a crossroads between the Middle Valley, the Fayyum, and the Delta), or the middle and lower parts of Middle Egypt (from south of Bani Suef, to the west of which is Heracleopolis, to north of Cairo [Heliopolis], i.e., all the territory of the Nile Valley to the east and a little to the northeast and the southeast of the Fayyum); (4) the Fayyum; and (5) the Delta.

[See PDF version of this article for table "Proposed Dialectal Geographies of Coptic Egypt."]

Since Coptic, like pharaonic Egyptian, is a dead language, it is not possible to know its dialects by direct observation of the language as it is spoken, as would be done for a living language. Only texts allow one to attain ultimately a knowledge of the dialects of a dead language. One may observe, whenever occasion affords, systematic morphosyntactical and lexical differences linked to this or that region; as regards Coptic, these differences certainly seem to exist, but for the most part they remain very modest, to the point that they do not of themselves convey the impression of true dialectal differences. But it is known that the most striking divergences between the dialects are generally of a phonological kind. To observe them in a dead language, one must admit (as the majority of linguists do) that the orthography of the language has a phonological value sufficiently precise to allow one to discover, from the various systematic graphic variants, various dialectal phonological systems. Certainly pharaonic Egyptian, in its various written forms, allows one to know the consonantal structure of the ancient Egyptian lexemes but scarcely or not at all their vocalic structure. The latter appears very clearly in Coptic, in which all the graphemes called vowels, or graphV ([...], [...], [...], [...], [...], [...], and [...]), are of Greek origin. This allows one to observe in the Coptic texts divergent orthographic systems, which have always been considered by Coptologists (with some modern exceptions) as having put into writing their phonological systems in a manner still perceptible. Without this working hypothesis, by far the most probable, it seems impossible to work out any dialectal geography of Coptic Egypt whatsoever. On this basis, various systems of dialectal geography have been elaborated; mention will be made here only of those that have been set out in extenso and chiefly the most recent among them (cf. Vergote, 1973, Vol. 1a, pp. 53-59, and the maps herein).

Worrell (1934) divided Egypt into six dialectal regions: (1) the Delta (at least the western Delta), the land of BOHAIRIC, or B; (2) the valley from Cairo as far as the Fayyum (to north of Heracleopolis), SAHIDIC, or S; (3) the Fayyum, FAYYUMIC, or F; (4) the valley from the Fayyhm as far as Qaw and Ichqau (south of Tima, to the south of Asyut), also S (Worrell did not yet know the existence of MESOKEMIC, or M, called by some “Oxyrhynchite”); (5) the valley from Qaw to Thebes, AKHMIMIC, or A; and (6) the valley south of Thebes, A again. As for LYCOPOLITAN, or L, Worrell placed it, rather vaguely, north of A and south of S (region of Asyut and Tima?).

The distribution proposed by Kahle (1954) is often different: (1) the Delta, land of B (properly speaking), except for Alexandria, which could possibly have been the homeland of S; (2) Worrell’s region 2 (to Heracleopolis), a variety of B particularly close to S (“semi-Bohairic”; cf. ibid., pp. 377-80; Kasser and Satzinger, 1982); (3) the Fayyum, F; (4) the valley, from Heracleopolis to the north of al-Ashmtinayn, M; (5) the valley from Ashmunayn to the north of Nag Hammadi, L; and (6) the valley from Nag Hammadi as far as the region to the south of Thebes, A (which very soon advanced toward the north, establishing itself in particular at Akhmim).

Vergote (1973, Vol. 1a) proposed a solution that on certain points may be considered a compromise between the two preceding: (1) the Delta, B; (2) the valley from Cairo to Heracleopolis, S; (3) the Fayyum, F; (4) the valley from Heracleopolis to the north of al- Ashmunayn (a little farther to the north than for Kahle), M (called O, or Oxyrhynchite, by this author); (5) the valley from al- Ashmunayn to a zone between Qaw and Akhmim, L (called A2, or Subakhmimic, by this author); and (6) the valley from Akhmim as far as the regions to the south of Thebes, A.

It can be seen that Coptology is still far from having reached any certainty concerning all points of its dialectal geography. It is therefore not unreasonable to take up again briefly the various problems of this domain. One may recall first of all (an obvious fact, the full implications of which are not, however, always drawn) that Coptic is now a dead language, so that the investigator must adapt himself to the inconvenience linked to this fact. Moreover, it has been a dead language for a very long time, unfortunately from a period largely prior to the first attempts at observation and scientific study of its philology (seventeenth century). Under these conditions, it is very often extremely difficult to localize its dialects, known almost solely from the evidence of literary manuscripts. These are liable to travel far, and since the majority have survived as the result of clandestine excavation one cannot even know exactly where they were found. (The place where they were sold is often very distant from that of their discovery, precisely to discourage investigation, whether by the police or by scholars: the “laws” of this illegal traffic require that the sources be shrouded in the most absolute secrecy, so that the stories of discovery which some inquirer thinks he has been fortunate enough to gather may well be no more than fables intended to lull his indiscreet curiosity; and if by chance one actually comes to know the place of discovery, it may well not be the place where the manuscript was copied and such an idiom was in use.)

Further, traditional data are too often vague and uncertain. Thus, the fourteenth-century grammarian Athanasius of Qus wrote of knowing the existence of three Coptic idioms: (1) “the Coptic of Misr, which is Sahidic” (Misr is Cairo, and for Arabic-speaking Egyptians the sa‘id is all Upper Egypt, in the widest sense of the term, and hence the whole Nile Valley south of the Delta as far as Aswan; the Sahidic country is thus by no means restricted to southern Upper Egypt, the region of Thebes, as numerous Coptologists have believed); (2) “the Bohairic Coptic known by the Bohaira” (this is the province occupying the greater part of the north-central western Delta); and (3) “the Bashmuric Coptic used in the country of Bashmur” (north-central eastern Delta).

Athanasius located B with relative precision. (Bohairic is a well-known idiom, whose localization is now confirmed by hundreds of parietal inscriptions found in the monastic site of Kellia, some of which are also in Greek; no one attests S or any dialect of the Nile Valley above the Delta.) He also located, to a certain degree, BASHMURIC (of which unfortunately practically nothing remains: two perhaps doubtful words in all). As for S, he said only that its territory is somewhere to the south of the Delta.

One must therefore have recourse to other means of locating the majority of the Coptic dialects, but unfortunately such indicators are often lacking. Certainly, the large quantity of nonliterary F texts found in this region allow one to believe that the cradle of his dialect is the Fayyum. Various phonemic and graphic resemblances between an Old Coptic manuscript (Osing, 1976) and the M texts have led some to locate the land of M in the region of Oxyrhynchus and thus call this dialect “Oxyrhynchite,” but this too-precise localization has been contested with serious arguments (Osing, 1978), which leads one to think that M should perhaps be located not exactly at Oxyrhynchus but a little farther north (or south?). Finally, it does indeed seem from graffiti found in situ that A was spoken very early, and probably from the beginning, at Akhmim (whose name in its present Arabic form with /x/ after the initial /a/ seems to reflect an ancient dialectal orthography [...], differing from S [...], but also written [...] or [...] cf. Westendorf, 1977, p. 481; B [...] is certainly a simple orthographical revival of S [...]). But the arguments (e.g., Kahle, 1954, pp. 198-99) according to which A was at first the idiom of Thebes, before being driven out, especially by S, are not entirely convincing. (Crum and Kahle knew neither DIALECT i as a dialect, proto-Lycopolitan, nor DIALECT P, also a typical PROTODIALECT, which often looks like what can be known about the logical predecessor of S, a tentatively reconstructed proto-Sahidic; a protodialect that, Nagel, 1965, has shown, had some affinities with the language of Thebes and which could, as much as A or L, or at least along with them, have influenced the orthography of the local nonliterary texts.)

There remain L and especially S, the most neutral classical Coptic idioms, the localization of each being particularly difficult to determine. The area within which L manuscripts have been found extends apparently from the Fayyum to the region near Thebes (perhaps even farther south, as far as Aswan; Worrell, 1934, p. 74). The area of the attested existence of S is even larger, since it is certain that it covers all the Egyptian Nile Valley above the Delta. One must therefore have recourse to other methods, especially the comparative analysis of isophones (phonemic isoglosses), in the attempt to locate L and S in relation to the dialects already more or less exactly localized: B in the Delta (probably at first the western Delta); F in the Fayyum (and V, or “Fayyumic without lambdacism,” a semineutralized variety of F or a MESODIALECT between dominant F, and W and M, probably in the east or southeast [?] of the Fayyum and somewhere in the Nile Valley immediately to the east or southeast [?] of the Fayyum); M in the neighborhood of Oxyrhynchus in Middle Egypt; A in the center (and south) of Upper Egypt.

To turn to the phonemic isoglosses (or isophones) is to admit as a general principle (Vergote, 1973, Vol. 1a, p. 56) that “the numbers of isophones are proportional to the distances between the dialectal areas” (the greater the distances between two dialects, the smaller the number of isophones shared by them). One must note the use of the same principles and similar methods in Hintze (1984) and Kasser (1981, pp. 124-31), and a more developed process (with a copious set of phonemic isoglosses completed with various morphophonological and morphosyntactic isoglosses) in Funk (1988).

However, one must also “take account of the relative importance of the phenomena” (Vergote, 1973, Vol. 1a, p. 56), particularly the isophone (which is the most convenient and generally used criterion; see DIALECT, IMMIGRANT); there (Kasser, 1987) some priority might be reasonably granted to the consonantal (and among the consonantal to the general) variables, excluding the cases in which the oppositions are not synchronic but diachronic (e.g. [...] /ç/ P versus [...] /[...]/ S, L, M etc., according to the consonantal late Egyptian evolution [...] > /ç/ > [...]/; [...] /x/ P versus [...] /h/ S, L, M etc., according to h > or = /x/ > /h/; see Vergote, 1945, pp. 122-23).

An additional restriction may be added here: the preceding rules only have their full value if one compares idioms that are really all of the same nature—that is, local dialects (and not, like S and B, “vehicular,” supraregional common languages; see DIALECTS, GROUPING AND MAJOR GROUPS OF), these local dialects being generally not neutralized (or only slightly neutralized)—for it is evident that an idiom whose expansion always remained strictly limited to its local area will normally have undergone only the influence of similar and neighboring local dialects, those who speak it belonging chiefly to a social level where professional occupation (agriculture, minor trades, etc.) and often modest way of life do little to encourage travel. This is a social aspect of the Coptic languages and dialects, which no doubt existed, frequently and secondarily, alongside their geographical aspect. An idiom of strong expansion, a vehicular language, will be much more neutralized by its repeated contacts, accompanied by reciprocal influences, not only with neighboring local dialects but also with more distant regional dialects and probably with one or another common language from even farther afield (cf. Chaîne, 1934, pp. 17-18), because those who speak a vehicular language, normally rather neutralized in its zone of expansion (hence outside its region of origin), belong chiefly to a social level where professional occupation (major trades, industry, commerce, higher administration, etc.) and a relatively comfortable way of life encouraged travel.

Consider now a semitheoretical example. Suppose the geographical chain of idioms 1, 2, 3, 4, 5, and 6, moving from south to north in this order the length of the Nile Valley. Of these, 2 and especially 1, 4, and 5 are typically regional dialects, not neutralized (or only slightly neutralized); and 3 and 6 are clearly supraregional idioms and are neutral (or at least more neutralized than the local dialects with which they are in touch, as immigrant dialects or vehicular languages). Of course, 3 will have isophones, among other things, in common not only with its neighbors 2 and 4, but also with the farther idiom 6, because, in spite of the remoteness of their geographical origin, both “common languages” are in touch on a higher (supralocal, social, etc.) level. Thus, it would be incautious to draw geographical conclusions too mechanically, by only counting the various isophones, many of which seem to locate 3 near 6. Both vehicular languages, as a result of their wide expansion, have been brought into contact, and this contact has made them influence one another, even if they may have been in their origin very far apart; in fact, their isophones (at least) bring them notably closer. One might easily take for geographical proximity what is probably no more than a sign of their similar nature as common languages and as neutral (or semineutral) idioms.

Consequently, it will be prudent to submit to critical reexamination the conclusion (Worrell, 1934; Kahle, 1954; Vergote, 1973, Vol. 1a, p. 59) that sought to locate the dialects of the Nile Valley and the Delta by their isophones in the following order, moving downstream and leaving F aside in its corner (and remembering that Worrell did not know M): A, L, (M), S, B. This order has today become almost conventional, but one may prefer an order more in conformity with the theoretical schema set out above, placing A in 1, L in 2, but S (not M) in 3, M (not V) in 4, V (not S) in 5, and finally (as in every scheme of dialectal geography) B in 6.

But perhaps such a division of Coptic Egypt is still too detailed and too precise according to the present state of knowledge? One way of doing justice to the reservations that this skepticism implies would be, for example, to classify the dialects not in groups (six in number) but in “major groups” (the number limited to three; see DIALECTS, GROUPING AND MAJOR GROUPS OF) and, in dialectal geography, to divide Egypt into three main regions only (cf. Kasser, 1982); this would be a way of returning, by and large, to the tripartite division most commonly accepted as regards Egypt in general.

According to this schema, major region I would be the land of the “major (dialectal) group I” and would correspond to Upper Egypt (including Asyut and maybe upper Middle Egypt). It would probably take in several local dialects, little neutralized, of which only A is known today, used in any case in the region of Akhmim and probably in other areas further south (e.g., the region of Thebes). In addition, major region I would include L (= L4 + L5 + L6) as a fairly neutralized cluster of (sub-)dialects, used regionally at least in the area of Asyut (= Lycopolis) and widespread as a second vehicular language by the side of S, over the same area, of course, and further south over the various areas of A, and possibly, though more discreetly and weakly, even further north (in parts of upper and a part of middle Middle Egypt?), temporarily and everywhere in rivalry with S. Gradually confronted by its most serious rival, S, invading from the north (and possibly also from Thebes, where S could have infiltrated very early, by way of the river), L finally perished, a little before A, both being stifled by S.

Major region II would be the land of the “major (dialectal) group II” and would correspond, if not to upper Middle Egypt, at least to middle and lower Middle Egypt and the Fayyum. This region would include several local dialects, little neutralized, of which the only ones known are F (located in the Fayyum and relatively little neutralized) and M (to be placed in the neighborhood of Oxyrhynchus or perhaps a little further north), in some respects a little better neutralized than F. Hemmed in between major regions I and III, and perhaps culturally less active, major region II would find itself invaded very early and traversed throughout, above all from the south, by the vehicular language S (and perhaps partially by L, the most neutralized dialect of the neighboring major region), which would leave but little chance of development for its own most, if little, neutralized Fayyumic subdialect V, which would possibly have tried to gain acceptance as the vehicular language over the greater part of this territory. Finally, V would have perished, with W and M and probably before F (which was better fitted to resist in its remote corner of the Fayyum), all stifled by S. Major region III would correspond to the Delta (or Lower Egypt). This region would probably include several local dialects, neutralized to different degrees, in which B (as a supralocal and supraregional vehicular language rather than a regional dialect; see LANGUAGE(S), COPTIC) is sufficiently known (from the first, well established in the western Delta, and then gradually penetrating throughout the Delta). Major region III would be essentially that of “major group III,” connecting dialects (those of the Delta) and, above all, two large vehicular languages, of which one (B) is that of the Delta also, and the other one (S) is used only outside of Lower Egypt, being superposed on all the regional dialects and local subdialects of the Egyptian Nile Valley above the Delta (i.e., chiefly A, L, M, W, V) in the whole of major regions I and II.

As is shown by the majority of the typical (nonvocalic!) S phonological features and by the most numerous morphosyntactical variables (see Funk, 1988), it is at least most likely that S derives from some local dialect of upper Middle Egypt (between a kind of pre-L and a kind of pre-M) in pre-Coptic times. This pre-S, whose tonic vowels were generally like those of pre-L and pre-M, became, probably very early on, the southern koine of Egypt, that of the whole Egyptian Nile Valley between the Delta and Aswan. As a vehicular language, it came in contact (near Memphis) with the second vehicular language, B, the northern koine—hence, a strong vocalic (stressed vowels) similarity between S and B (probably the influence of some pre-B on some pre-S about five centuries B.C.; cf. Chaine, 1934, pp. 13-18, and, more clearly, Satzinger, 1985). One might also suppose that S penetrated very soon, by way of the river traffic, to Thebes, where it would have established a center of expansion more and more active into the very heart of major region I (developing at the same time a variety of proto-S moving into the Theban region and bearing the phonological marks of this implantation, according to a former hypothesis; cf. Nagel, 1965, and Kasser, 1982).

[See PDF version of this article for table.]

The sequel seems better known and may be deduced from what is recorded of S in the classical Coptic period: pre-S, if not S itself, would have ended by occuping the whole Nile Valley (but not the Delta) to the detriment of its local dialects, invading in particular major region I from the north (and possibly from Thebes; cf. above), eliminating L and then A, and finally reducing the last pockets of resistance in major region II by the elimination of V. W, and M.

RODOLPHE KASSER