Coptic Language

Historical setting and sources

The encounter of Egyptian natives with the Arabs and Arabic in 641 C.E. was the starting point of cultural as well as linguistic developments resulting in the loss of the Egyptian (Coptic) language and the shift of Christian Egyptians to the use of Arabic before ca. 1300 C.E. The socio-linguistic and chronological details of this process are largely unknown and still debated (see Björnesjö 1996; Décobert 1992; Helderman 1997; MacCoull 1985, 1989; Richter 2001; Rubenson 1996). The phenomenon under discussion is based on evidence of a dead written language (on the relevant methodological issues, see Adams a.o. 2002). Language contact phenomena, however plausible in the spoken language, are scarcely visible in the written texts, if at all. There are only two kinds of linguistic interference phenomena that can be observed through the mirror of written evidence: ‘hybrid’ combinations of linguistic and graphical codes not matching each other (‘Karshuni’) and lexical code-switching (borrowing).
Parts of a large 13th-century manuscript written in the Arabic language, but with Coptic signs, have been published (Blau 1988; Burmester 1965–1966; Casanova 1901; Satzinger 1972, 1991; Sobhy 1926; Worrell 1934:134–143). Its Coptic spellings of Arabic words, although based on the phonological system of Bohairic (Lower Egyptian) Coptic, are of some value for comparison with the mostly Fayyumic (Middle Egyptian) and Sahidic (Upper Egyptian) spellings of loanwords (see section 2 below). The same phenomenon of Arabic texts spelled in Coptic letters is further attested in a bilingual curse (Crum 1902a; Blau 1988:189). The reverse phenomenon – the use of Arabic characters to transcribe a Coptic (Bohairic) text – is attested in a manuscript of hymns in honor of the holy virgin (Blau 1988; Galtier 1906; Satzinger 1972, 1991; Worrell 1934:134–143).
Coptic loanwords (and some cases of phonological and structural borrowing) in Egyptian Arabic have been dealt with several times over the last century (Praetorius 1901, 1902; Galtier 1902; Sobhy 1950; Bishai 1960, 1961, 1962, 1964; Ishaq 1975, 1991; Behnstedt 1981, 1997; Vittmann 1991; Peust 1999:321–323; Schenkel 2002), although the estimated quantity of these borrowings varies greatly. In contrast, the occurrence of Arabic loanwords in Coptic texts thus far remains almost entirely unresearched (cf. Stern 1885:117–119; von Lemm 1903:xvii.34–36; Chassinat 1921:21–47; Worrell 1934:122–133; Vycichl 1991). In particular, there exists no modern reference tool listing either Arabic loanwords in Coptic or Coptic texts containing them. Both Worrell (1934:122–133) in his phonological study of Coptic and Vycichl (1991) in his brief entry on Copto-Arabic vocabulary dealt with only two Coptic texts. At present, we know more than 90 published Coptic texts and a few unpublished ones which contain Arabic loanwords. It is a very striking fact that almost all of these texts belong to two genres:
i. scientific texts, 18 manuscripts from the 9th to 11th century, among them four large alchemistic treatises (Stern 1885 and the hitherto unpublished Bodleian manuscripts (P)a 1, 2, and 3), a manual providing arithmetical and metrological problems (Drescher 1948–1949), a page dealing with astronomy (Bouriant 1904; von Lemm 1903:35–36), and a few collections of magical (in particular Chassinat 1955) and medical (in particular Chassinat 1921) receipts. 
ii. about 80 documentary texts from the 8th to the 12th centuries, including legal documents (Richter 2001, 2003), letters, lists, and accounts, in particular a large 11th-century account book recording income and expenditure of a middle Egyptian monastery (British Library Or. 13885, unpublished).
These texts comprise altogether about 400 transcriptions of Arabic words, which will be called ‘loanwords’ here, regardless of whether they might have been well-established parts of Coptic speech or rather, as is more likely in many cases (see section 5 below), ‘one-time’ or ‘nonce’ borrowings (Field 2002:9–10; van Hout and Muysken 1994:40). The great bulk are nouns; only 19 verbs (about 5 percent) have been identified thus far, and one single function morpheme may be attested. It should be borne in mind, however, that this evidence is not certain, but depends on further philological study of Coptic texts. Linguistic conclusions are necessarily only provisional. Almost all Coptic transcriptions of Arabic words occur in Coptic texts written in a relatively informal orthography and with a relatively low degree of linguistic perscriptivism, closer to the vernacular than any literary composition. The great bulk represents the Sahidic (Upper Egyptian) dialect; a few are Fayyumic (Middle Egyptian) or Bohairic (Lower Egyptian). The linguistic decorum appropriate to the vast number of Coptic (semi-)literary texts demanded the suppression of phonetic, grammatical and, in particular, lexical innovations, so that borrowings from Arabic do not occur in them at all. There are only a few exceptions to this rule, e.g. a magical spell invoking the roh n-alla (< rūḥ allāh) ‘spirit of god’ (Beltz 1983:63), a 13th-century hagiographical text (Amélineau 1887), and a 13th-century scribal colophon (Crum 1905, no 726).
The Arabic underlying these Coptic transcriptions was roughly identified by Worrell (1934:123) as spoken (or heard) colloquial, rather than Classical Arabic. There are strong affinities between the phonemic correspondences prevailing in the loanwords and those attested in the homogeneous transcription system of a Copto-Arabic Karshuni text (see above). The language of the latter was studied by Satzinger (1972) and Blau (1988), who described it as follows: “Like Middle Arabic texts in general”, this one too “is characterized by freely alternating features of Classical Arabic, Neo-Arabic and pseudo-corrections” (Blau 1988:145).
Coptic spellings of Arabic phonemes and related issues
Some Arabic phonemes have close equivalents in Coptic (e.g. the sonorants l, m, n, r); hence the same graphemic correspondences always occur. More commonly, different ways to transcribe a single consonant phoneme are attested even in the same position (note, e.g., the three variants of f in ʿalaf ‘fodder’: alêf, alêb, alêou). But usually one of these varieties proves to be the most common, regular one.
i. Consonants: ‘ = mostly ø, seldom a (alaasaat < al-'asad), perhaps consonantal gemination (ammour < a'mur?), perhaps h (khithirh < kaṯīrā'?); b = p (cf. Hintze 1947b); t = d or t, seldom th (alkhiprith < al-kibrīt); ṯ = th , seldom t (almatkal < al-miṯqāl); j = Sahidic c (a palatalized velar), Bohairic č (palatal, regularly corresponding with Sahidic c); ḥ = h, seldom ø (kol < kuḥl); x = mostly kh , in Bohairic ẖ (almairêẖ < al-mirrīx), seldom š (assarnêš < az-zirnīx), perhaps h (arrôham < ar-ruxām?), perhaps ø (aulen < xawlān?), in Bohairic perhaps once k (allinek < al-līnax?); d = d or t; ḏ not attested; r = r, seldom l; z = s, seldom z (gazouan < ġazawān); s = s; š = š; ṣ = s; ḍ = t (apiat < ʾabyaḍ); ṭ = t, seldom d (hôdôt < ḥuṭuṭ); ḏ̣ = t (attaheri < aḏ̣-ḏ̣āhirī), perhaps s (naser < nāḏ̣ir?); ʿ = mostly ø, sometimes a (alaakrap < al-ʿaqrab, assalae < az-zalʿa, arrôpa < ar-rubʿ), seldom e (alceme < al-jamʿ), (e)i (assiri < az-zarʿ) or ô (arrapô < ar-rubʿ) and even k (almaksoul < al-maʿsūl), perhaps consonantic gemination (alcelle < al-jaʿāla?); ġ = g (almoulgam < al-malġam), perhaps c (alcabiri < al-ġafīr?); f = b (cf. Hintze 1947b), sometimes f, seldom ou (cf. alêf, alêb, alêou < ʿalaf), once (Bohairic) p (espêiteč < ʾisfīdāj); q = k; k = kh , sometimes k (alkous < al-kūz); l = l, seldom r; m = m; n = n, but in contact with labials, Coptic assimilation (n becomes m) occurs (assampak < azzanbaq, assoumpoule < as-sunbula); w = ou, once (Bohairic) b (iban < 'īwān); h = h (assoouhre < az-zuhara), as feminine ending ø, once h (šetineh < šādina); y = (e)i. Arabic consonantal gemination is sometimes written (almousabbi < al-muṣaffī, almoucarrap < al-mujarrab, alkhammoun < al-kammūn, alhôcce < al-ḥujja, asoukhkhar < as-sukkar, atassa < aṭ-ṭāssa, aššoukke < aš-šuqqa, ette < ʿidda), sometimes not (alcoume < al-jummāʿ, morape < murabbaʿ, oušak < wuššaq, rôman < rummān, almairêẖ < al-mirrīx), sometimes either way (alcoup(p)e < al-jubba). Gemination is never spelled in final position (alhal < al-xall, alkhas < al-xazz, alhat < al-ḥadd, almalaf < al-milaff, armôr < al-murr, arôs < ar-ruzz). In a few cases it seems to be transcribed as a vocalic ablaut, cf. aššipe < aššabb, kere < qarr, and lepe < labb; cf. the proper name Apoulase < ʾAbū Lazz. 
ii. Vowels: ā regularly occurs as a or e, sometimes as ê, seldom as ee (alpeep < al-bāb), i (alkili < al-qily), or ôe (almôes < al-mās). There is thus strong evidence of ʾimāla (as in Copto-Arabic Karshuni, cf. Blau 1988:152). ū occurs as o, oo, ou, ô, ôô; ī occurs as i or ê. The feminine ending (tā' marbūṭa) is almost always spelled e (again clear evidence of ʾimāla), sometimes a (a and e also in Copto-Arabic Karshuni, see Blau 1988:176), seldom ai (almešmelai < al-mišmala), i (almanari < al-manāra), ø (alpourat < al-burāda), once eh (šetineh < šādina). Other short vowels must be left out of consideration, since both their quality in colloquial Arabic and their Coptic transcriptions show a great deal of variation, so that correspondences remain unclear.

Morphology of Arabic words in Coptic

3.1 Nouns
As a rule, Arabic nouns taken into Coptic are borrowed in a form beginning with al-, less often spelled ar-, el-, or er-. Before the ḥurūf šamsiyya, assimilation usually occurs: an-n…, ar-r…, as-s…, aš-š…, at-t…, although often spelled haplographically with no gemination: an…, ar…, as…, aš…, at…. As in Spanish, this Arabic article does not function as a determiner. Every borrowed Arabic noun, whether prefixed with al- or not, was subject to the elaborate Coptic determination system (cf. Layton 2000:35–53), distinguishing, e.g.,
i. definite articles: masc. sg. p-, e.g. taau ehoun e-p-alhal ‘add them to the (p-) vinegar (al-xall)’; fem. sg. t-, e.g. ci…n-t-alpourate ‘take the (t-) filings (al-burāda)’; and pl. n-, e.g. etbe n-alhecos euhiptôou ‘because of the (n-) barriers (al-ḥajz) which are on the mountains’ (Crum 1902b:no. 290); 
ii. indefinite articles: sg. ou-, e.g. eišouei mmof hn-ou-alkous ‘heat it up in a (ou-) jug (al-kūz)’ (Bodl. ms (P)a2, 26) and pl. hen-, not attested; 
iii. zero-article: ø – as in ou-alkapele n-at-ø-almisahe ‘a tenancy (al-qabāla) without survey (al-misāḥa)’ (Richter 2003).
These determiners were applied to Arabic nouns according to both the semantic and the syntactic demands of the Coptic language. The use of Arabic nouns without the article al- is far less frequently attested. However, this is the standard in a large medical manuscript (Chassinat 1921), and it often occurs in rather early cases of borrowing, as can be shown by the word (al)para (< barāʾa) ‘receipt’ which is spelled without al- in a number of 8th-century documents (e.g. Kahle 1954:no. 291,5.29, t- or p-para), while in later (9th- and 10th-century) documents it is always written t-alpara (e.g. Crum 1902b:no. 377,9; Crum 1939:no. 49,11.13).
Unlike Arabic nouns, which are subdivided into unmarked masculine vs. marked feminine forms, Coptic nouns have an associated (inherent) gender, which is expressed not by special forms but by masculine vs. feminine determination morphemes (Layton 2000:85–86). In most instances, the choice of a Coptic article matches the respective grammatical gender of the Arabic noun. There are only a few cases of discrepancy between the gender of the noun in Arabic and of the article used in Coptic. The gender of borrowed nouns is often influenced by target language nouns of similar meaning. This may be the case with p-athaskieie (< at-tazqiya) ‘purification’ (Bodl. ms (P)a1, g11; masculine Coptic equivalent tbbo), p-para (< barāʾa) ‘receipt’ (Kahle 1954:no. 291,29; masculine Graeco-Coptic equivalent entagion), t-almiret (< al-mīrāṯ) ‘heritage’ (Richter 2001:80; feminine Graeco-Coptic equivalent klêronomia), or t-almisan (< al-mīzān) ‘scales’ (Bouriant 1904; feminine Coptic equivalent maše). Some words are treated as either masculine or feminine, e.g. p- or t- (al)para (< barāʾa) ‘receipt’, p- or t- alpourate (< al-burāda) ‘filings’.
3.2 Verbs
Almost all verbs borrowed from Arabic come from alchemistic treatises, where not only concrete objects like ingredients, utensils, etc., but also certain procedures are designated by technical terms. However, unlike nouns with their common al-‘prefix’ clearly pointing to Arabic etymology, it is not always so easy to make a decision on whether a Coptic-written verbal lexeme comes from Arabic or not, the more so as the morphological richness of the Arabic verb, with its breakdown into stems, conjugations, and verbal nouns, can complicate the identification. Coptic verbal syntax requires only two verbal forms, both operating without inflexion. Verbs borrowed from Greek into Coptic are even restricted to a single basic form: they occur in a non-Classical (Greek) infinitive form and operate within Coptic syntax as (Coptic) infinitives. Similarly, Arabic verbal forms seem to be used in Coptic sometimes in their infinitive forms (see examples [1]–[5] below), although the difficulty of determining vowel qualities (see section 2 above) leaves some uncertainty:
Strong verb, Form IV
(1) akêt (< ʿaqada IV ‘to boil down, to thicken’) – infinitive: ʾiʿqād (cf. imperative ʾaʿqid)
(2) elhêf (laḥafa IV ‘to cover’) – infinitive: ʾilḥāf (cf. imperative ʾalḥif)
Geminated verbs
(3) kera, kere (< qarra ‘to be cold, to be cool’) – infinitive: qarr 
(4) lepe (< labba ‘to stay’) – infinitive: labb
(5): gazouan (< ġazā ‘to conquer, to capture’) – infinitive ġazawān
In other cases, however (see [6]–[12]), forms similar to the imperative, or even the apocopate imperfect (but without subject prefixes), seem to underlie Coptic transcriptions:
Strong verb, Form II
(6) saeid (< ṣaʿida II ‘to sublimate’) – imperative: ṣaʿʿid (imperfect yuṣaʿʿid(u), but infinitive taṣʿīd)
(7) taperi (< dabara II ‘to prepare’) – imperative dabbir (imperfect yudabbir(u), but infinitive tadbīr)
Geminated verbs, Form VII
(8) nhal (< ḥalla VII ‘to dissolve’) – imperfect: yanḥall(u) (imperative inḥalil, colloquial also inḥall?, but infinitive: inḥilāl)
IIIw, Form II
(9) safbi, sabbi (< ṣafā II ‘to clean’) – imperative: ṣaffi (imperfect yuṣaffi, but infinitive taṣfiya)
IIIy, Form IV
(10) eišouei (šawā IV ‘to roast, to fry’) – imperative: ʾašwi (or infinitive ʾišwāʾ?)
(11) eicri (< jarā IV ‘to carry out’) – imperative: ʾajri (or infinitive ʾijrāʾ?)
(12) am(m)our (<ʾamara ‘to command’) – imperfect: yaʾmur(u) (but infinitive ʾamr)
3.3 Function morphemes
Only one function morpheme probably borrowed from Arabic has been identified thus far. In two Coptic alchemistic treatises, a morpheme ô- linking entity terms to each other occurs (Stern 1885 passim; Bodl. ms (P)a2, 70), which is considered to be identical to the Arabic conjunction wa-, e.g. cop p-ašêlas ô p-almêstikhe ô p-assampak ‘take the whey (aš-šīrāz) and (wa-?) the mastic gum (al-maṣṭakā) and (wa-?) the lily (az-zanbaq)’ (Stern 1885:VII, 18–19). In an amazing example of written code-switching, the same conjunction, although now written in Arabic script, is used elsewhere to link Coptic-written Arabic nouns: sincipil wa-houlincan wa-kalanfour wa-soumpoul ‘ginger (zanjabīl) and alpinia officinarum (xūlanjān) and (wa-) clove (qaranful) and (wa-) nard (sunbul)’ (Chassinat 1921:155).
Insertion of Arabic words into Coptic syntactic structures
4.1 Nouns
Due to the strong analytic type of Coptic syntax (cf. Hintze 1947a; Loprieno 1995:7), the embedding of Arabic words into Coptic syntactic structures works rather easily. All grammatical categories having to do with entity terms, like gender, number, and determination, are marked exclusively by morphemes belonging to determiner paradigms (see above), while the grammatical function of nouns is indicated by distinctive sentence patterns and function morphemes.
(13) P.Lond.Copt. I 487 (Richter 2003): ai-ti nak ou-alkapele n-at-almisahe 
a=i- [perfective conjugation base + 1 sg. pronoun] ti [predicative infinitive] na=k [dative preposition + 2 sg. pronoun] ø- [object position] ou-alkapele [indefinite sg. article + noun] n- [attributive modifier] at- [privative nominal base] ø -almisahe [zero-article + noun] ‘I gave you a tenancy (al-qabāla) without survey (al-misāḥa)’.
4.2 Verbs
Native Coptic verbal lexemes can be realized in two forms, the infinitive (including the status absolutus and two distinct forms indicating close connection with a nominal or a pronominal direct object) and the stative (Layton 2000:124–157). However, Coptic verbs borrowed from Greek are restricted to the basic form, the infinitive status absolutus, a form which can function as a verbal predicate of any conjugation pattern, as a verbal noun, and as an imperative, depending on the grammatical context. In the few cases of verbal lexemes borrowed from Arabic, the same technique occurs, as is demonstrated by examples 14–16, each showing an Arabic verbal lexeme used as a (Coptic) infinitive in two functions, the imperative and the verbal predicate of a conjugation pattern:
(14)akêt (ʿaqada IV) ‘to boil down, to thicken’
akêt mmo=ou hičô-ou-kôht e=f-kere… e=k-šan-bol=f ebol n-3 n-sop k-akêt nmo=f ša=f-rôše ‘boil [imperative] them down on a fire which is cold (qarra)… if you dissolve it 3 times (and) you boil [conjunctive conjugation] it down, it will be enough’ (Bodl. ms (P) a3, 28–30) 
(15)eišouei (< šawā IV) ‘to roast, to fry’
eišouei mmo=ou tso=ou kata 3 n-hoou šante=k-eišouei mmo=ou ‘fry [imperative] them (and) water them during 3 days, until you have fried [limitative conjugation] them’ (Bodl. ms (P)a 1, f 12) 
(16)saeid (ṣaʿida II) ‘to evaporate, to sublimate’
saeid mmo=ou ‘evaporate [imperative] them’ (Bodl. ms (P) a1, a11)
nta=f-saeid n-p-assipak n-7 n-sop ‘(I saw the master), who evaporated [relative converter + perfect conjugation] the quicksilver (az-zībaq) 7 times’ (Bodl. ms (P) a1, g1)
Semantic issues
In those genres of Coptic texts providing Arabic words at all, a great many of the borrowed terms are in some way technical. In Coptic scientific treatises, we encounter names of planets (e.g. as-soouhre < az-zuhara ‘Venus’), constellations (e.g. assarataan < as-saraṭān ‘Cancer’), plants (e.g. alkha-bôôr < al-kāfūr ‘camphor’), spices (e.g. alboulboul < al-fulful ‘pepper’), minerals (e.g. assipak < az-zībaq ‘quicksilver’), chemicals (e.g. alkhiprit < al-kibrīt ‘sulphur’), diseases (e.g. annikrês < an-niqris ‘gout’), and mathematical terms (e.g. alkhousôr < al-kusūr ‘fraction’). Although it is difficult to estimate their linguistic significance, there is good reason to doubt the conclusion drawn by Vycichl (1991:215): “The spoken language was full of Arabic words, as one can see from a medical papyrus or a treatise on alchemy”, for these terms are not part of the vernacular vocabulary. Rather, they belong to specialized taxonomic vocabularies, which in general are subject to special rules of borrowability.
In documentary texts, we meet titles (e.g. amira < ʾamīr ‘commander’), weights (e.g. almatkal < al-miṯqāl ‘weight of one dinar’), measures (e.g. arrôpa < ar-rubʿ ‘quarter’), coins (e.g. derham < dirham ‘dirham’), book-keeping terms (e.g. nabaka < nafaqa ‘expenses’), and legal words (e.g. dyn < dayn ‘debt of money’). Further, there are designations for diverse objects, especially vessels (e.g. alkaroore < al-qārūra ‘flask’), clothes (e.g. almicar < al-miʿjar ‘cap’), and textiles (e.g. alkhas < al-xazz ‘silk fabric’), probably referring to specific qualities of the respective categories flask, cap, etc. in a genus-pro-specie way. At any rate, Arabic nouns tend to be used in a specialized, narrower sense when taken into Coptic, e.g. alkapele (< al-qabāla) ‘obligation, contract, etc.’ in the meaning ‘tenancy’, or alhat (< al-ḥadd) ‘border’ in the sense of ‘bordering estate’, in keeping with the technical use of these words in corresponding Arabic texts of the same genres.
Although the total amount of Arabic loanwords in Coptic is rather low, there is a conspicuous accumulation of Arabic words in two semantic fields: sciences and economy. The first might point towards a high esteem for Arab natural sciences, established in educated circles of Egyptian Christian society, as an eastern counterpart to the well-known reception of Arabic sciences in medieval Spain (cf. Gallego 2003; Burnett 1997). The latter may indicate widespread commercial transactions between Arabic and Coptic speakers. But this sociolinguistic conclusion remains to be proven by broader evidence.
Tonio Sebastian Richter (University of Leipzig)
  • Recommend Us