Infidel or Paganus? The Polysemy of kafara in the Quran

This article explores the meaning of the root k-f-r in the Quran, questioning the practice of translating the noun kāfir as “infidel.” It argues for a distinction between the idiomatic phrasal verb kafara bi-, which does mean to reject or disbelieve, and the simple intransitive verb kafara and its deverbal nouns, which are used in the Quran in a large number of different ways. This polysemy is explored through contextual readings of Quran passages. It is argued that the noun kāfir, unlike the verb kafara, is used only with regard to adherents of traditional polytheism and is not deployed in an unmodified way with regard to Jews and Christians. The possible influence on the Arabic kafara of Greek and Latin conceptions is also broached.

developed and widespread tradition in places like Najran, Petra, and Nessana. 9 The educated used Greek and Syriac for theological purposes, but the large communities of Arabophone Christians in Transjordan and Syria over centuries would have developed their own neologisms and culture. 10 As they adopted Christianity, Arab preachers would have needed Arabic technical terms for theology and homilies. 11 Texts preserving this sophisticated Christian Arabic of the Levant have not survived from the fifth and sixth centuries, but it is likely that its vocabulary is visible for the first time in the Quran, given the trade and cultural ties that bound the Hijaz and the Eastern Roman empire. Arabic theological vocabulary, then, developed in part because it was influenced by Greek, Aramaic, and Middle Persian.
the Cover-up Although many contemporary translators and commentators have lost sight of the polysemy of k-f-r and its derivatives, it was recognized by medieval Muslim thinkers concerned with word meaning. One of the first Arabic dictionaries, Kitāb al-ʿAyn of al-Khalīl b. Aḥmad, begins by defining k-f-r as the opposite of faith (īmān) and as the opposite of gratitude (shukr). 12 He goes on, however, to give many other meanings, though in a piecemeal fashion-the root can have to do with, for instance, hypocrisy or the coronation of a king. It also refers to villages (sing. kafr) in thinly populated terrain. Hundreds of years later, the North African lexicographer and court judge Muḥammad Ibn Manẓūr (d. ca. 1312), who settled in Mamluk Cairo, made a sophisticated linguistic argument for k-f-r and its derivatives as polysemous terms and put forward the principle by which its various forms are semantically related. 13 He saw the root's different senses as issuing from the notion of "covering up" (taghṭiya, satara). In a concrete sense, the progressive particle kāfir, he says, can refer to a peasant farmer, who covers seeds with earth after planting them. A kāfir can also be a carrier of a concealed weapon, hidden beneath his robes. The verb kafara can mean to reject, and he argues that this sense derives from the action of covering up the truth of an assertion. It can mean to be ungrateful for a gift, inasmuch as the ingrate covers up the obligations of the Identity," in From al-Andalus to Khurasan: Documents from the Medieval Muslim World, ed. P. Sijpesteijn et al. (Leiden: Brill, 2007), 219-42, esp. 231-32. 9. R. Hoyland, "The Language of the Qur'an and a Near Eastern Rip van Winkle," in A Life with the Prophet? Examining Hadith, Sira and Qurʼan, ed. A. Fuess and S. Weninger (Berlin: EBVerlag, 2017), 17-43, at 39;idem, "Mount Nebo, Jabal Ramm, and  heart. With a preposition, the intensive form kaffara ʿan can mean "to absolve," i.e., when God covers up the past sins of the penitent. The comparative religionist Toshihiko Izutsu accepted the underlying senses of "covering up" and "ingratitude" for these words. 14 The association of words deriving from this root with covering up obtains in other Semitic languages as well, suggesting that Ibn Manẓūr was on to something. The noun kafr occurs in Dadanitic inscriptions from around Ula in northern Arabia (now Saudi Arabia), likely dating to the last centuries of the first millennium bCe. The OCIANA database contains several instances of this word, meaning in these inscriptions "tomb," so that it is cognate to the Arabic qabr. Since a tomb is a means of covering up a corpse, this sense of the term is understandable. One Dadanitic inscription found in the early twentieth century on a stone in the vicinity of Ula records that one ʿAbd Kharag "built this tomb for him and for his descendants, the whole of this tomb." In both occurrences, "this tomb" is inscribed h-kafr. 15 The root appears in other Semitic languages. The Akkadian kapāru can mean "to efface" or "to cleanse." In Hebrew the root can mean to cover over, but also to propitiate (hence Yom Kippur or the day of atonement). It exists in Aramaic and Syriac in the sense of effacing or wiping clean. 16 In Christian Syriac works it can mean "to reject," "to be ungrateful," and "to blaspheme," covering some of the same terrain as the Arabic words from this root, though the Arabic family of senses does not match that of the Syriac exactly, contrary to what Édouard-Marie Gallez has alleged. 17 I concur with Walid Saleh that knowing the etymology of a word, and awareness of its cognates, does not provide us with its exact meaning at any particular time and place, and for this reason the below examines Quran passages contextually. On the other hand, as Saleh admits, historical linguistics can often offer useful insights. 18 denial Let us begin by considering how and where the root gives the sense of "to reject, deny" in the Quran. The early sura al-Balad (90) condemns the Meccan elite's arrogance, pride in wealth, and disregard for the needy. It contrasts these heartless persons with moral exemplars who have believed and know the way through the difficult pass of high ethics, who free slaves, feed orphans, and provide nourishment to the poor. In contrast, "Those who have denied (kafarū bi-) our verses are companions of the left hand" (90:19). Note the preposition bi-. We have here a two-part or "phrasal" verb, which is idiomatic in that it requires for its meaning a preposition that takes an oblique object. This phrasal verb is not polysemous in the Quran for it always means to deny or reject.
The simple verb kafara has a much wider range of signification, but it does occasionally overlap in meaning with the phrasal verb, despite being intransitive, inasmuch as it is contrasted with belief. For instance, al-Kahf 18:29: "And say, the truth is from your lord. So, let the one who wishes to, believe, and let the one who wishes to, decline (yakfur)"; al-Ghāfir 40:10, addressing the pagans: "When you are called to faith, you decline (takfurūna)"; and al-Ḥajj 22:57: "Those who kafarū and impugned the veracity of our signs, a humiliating torment awaits them." 19 Even where words from k-f-r are contrasted with the verb "to believe" and nouns from that root, however, it cannot be assumed that they are always used as an exact antonym. After all, the impious and those who blaspheme are also the opposite of those who believe, but an impious person or blasphemer might not so much deny the existence of the sacred as belittle it or rebel against it. An example is Muḥammad 47:3, where it is written: "Those who kafarū followed falsehood (bāṭil) and those who believed (āmanū) followed the truth from their lord." Kafara here has to do not with denying something but in positively upholding something that is wrong. This verse might be compared to Eph 4:27: "So then, putting away falsehood (pseudos), let all of us speak the truth to our neighbors, for we are members of one another."

peasants and polytheists
The active participle kāfir most often functions as a noun rather than as a verbal noun in that it does not take an object, though the phrasal verbal noun kāfir bi-is attested (al-Anʿām 6:89). Where it lacks a preposition, kāfir should be seen as deriving from the polysemous simple intransitive verb kafara and therefore cannot be assumed necessarily to mean "rejecter of something" or "infidel." Rather, it has a wide range of meanings that can be discerned contextually. In al-Ḥadīd 57:20 the broken plural refers to rustic farmers: "Know that the life of this nether world is a game, a sport, a trinket, a mutual boast among yourselves and a multiplication of your wealth and children. It resembles rain whose resultant vegetation pleases the peasants (kuffār), but then it withers and you see it yellowing into chaff." As al-Khalīl mentioned, kafr means village, reinforcing the rural connotation of the root. It may be that a secondary meaning of polytheist or adherent of traditional religion emerged because the population in the countryside was more likely than its urban counterpart to have clung to the old gods and resisted accepting monotheism.
The root is also clearly associated in the Quran with polytheism. To you your religion and to me my religion." There is an admission that the pagans have a religion, but it is simply castigated as a false one, which makes translating kāfir as "infidel" seem odd. That the dispute was over Muḥammad's monotheism versus Arabian polytheism is demonstrated by Ṣād 38:4-5, which says of the pagans, "They marvel that a warner came to them from among them, and the kāfirūna said, 'This is a lying sorcerer. Has he made the gods into only one God? That is an astonishing thing'." This and many other verses demonstrate that the Quran came out at least in part of a milieu where there were adherents of traditional religion. 20 19. See also al-Baqara 2:253 and al-ʿImrān 3:106, among others. Because it would prejudice my search for the various meanings of k-f-r, I initially will not be translating words from that root.
20. This point is acknowledged by Hawting, Idea of Idolatry, but then downplayed in favor of seeing the kāfirūn and mushrikūn as in part Christians and Jews and suggesting at least obliquely that these verses were produced in a venue other than the early seventh-century Hijaz. I will argue against these theses. There is increasingly strong reason to see the Quran as early seventh century; see The identity of these pagans has been argued in the literature in recent decades, but it is increasingly clear that they are simply a provincial survival of Greco-Nabataean religion. In the Transjordan and northern Hijaz during centuries of Roman rule after 106 Ce, locals had often assimilated the North Arabian gods worshipped by Nabataeans and by Arabic speakers throughout the Near East to the Greek Olympians. The great goddess Allāt was generally identified with Athena, al-ʿUzza with Aphrodite, and Manat with Tykhe, though such identifications varied and were not uniform. Allāt-Athena or Athena Kyria is, however, widely attested in inscriptions throughout the Roman Near East and in the Decapolis. 21 An inscription at Petra is dedicated to Zeus Hypsistos, the All-High God. 22 Even as Christianity became hegemonic in the centuries after Constantine's 312 conversion, Gideon Avni has underlined that the archaeological record and literary sources make clear "the continuity of the pagan population" and the continued worship of betyls or standing stones outside the cities in the Near Eastern eparchies of the Eastern Roman empire of the sixth and seventh centuries. 23 Vocabulary in the Quran reflects this pagan remnant, which in the Roman empire was called gentiles in Latin and in the New Testament ἔθνη or ἐθνικοί (Matt 6:7, 6:32), for which I believe the Quran uses the loanshift ummī (cf. al-Jumʿa 62:2: "He it is who sent to the gentiles (ummiyyīna) a messenger from among them to recite to them his verses and purify them and teach them the book and wisdom, even though they were aforetime in manifest error"). 24 Like Paul of Tarsus, Muḥammad had a mission to the gentiles, with the difference that Muḥammad himself sprang from a pagan population.
In the late Meccan sura al-Aʿrāf, 7:65-66, God sent the messenger Hūd to the people of the ʿĀd tribe to say to them, "People, worship God; you have no god but him. Will you not be godfearing? The assemblage of those who kafarū replied" that they viewed the ancient prophet as a fool and a liar (for denying the pantheon of gods). In al-Anʿām (6) from around the same period, Muḥammad's opponents are castigated: those who kafarū dismissed the Prophet's narratives about the monotheistic prophets as "fables of the ancients" (6:25). Also castigated in the Quran is the association (shirk) of Allah with other divinities in a pantheon as a form of impiety (kufr) (Āl ʿImrān 3:151, in full: "We will cast fear into the hearts of those who kafarū inasmuch as they have associated with God (ashrakū) that for which he revealed no authority. Their refuge is the fire, and how miserable is the abode of wrongdoers"), where kafara is equated to ashraka or the attribution of divine family members or associates to God, and al-Māʾida 5:90 forbids reverencing the standing stones (nuṣub). In contrast, God in the Quran is unique and unlike anything else (al-Ikhlāṣ 112). The concept of shirk may come from Sabaic in Yemen and be based on an analogy from sharecropping (the Ḥimyarite elite had its own monotheist revolution from 380 Ce and so developed a polemic against polytheism). 25 As the Quran depicts the situation in Mecca, the pagans viewed the Kaʿba as a Pantheon, against insistence in the Quran that the temple should be dedicated solely to the creator-God, Allah, as a site of monolatry : "The temple belongs to God; do not call upon anyone else there alongside him. And when the Servant of God stood up there supplicating him, they virtually swarmed around him." This resembles the situation in Rawwāfa in the second century, where a Thamūd brigade serving as limitanei or border guards for the Roman governor at Bostra left behind bilingual Aramaic and Greek inscriptions at a temple dedicated to Alāhā/Theos. One inscription reads, "The temple which Shiddat, the priest of God [Alāhā], son of Megido, who is from Rabato, made for God . . . with the encouragement of our lord the governor;" another, "For the well-being of [Marcus] Aurelius Antoninus and Lucius Aurelius [Verus] . . . This is the temple which the brigade of Thamūd made." 26 This site is, however, the only archeological attestation of a singular temple to Alāhā (viz., Allah in the Arabic of the Hijaz) in pre-Islamic times, though hundreds of Nabataean and Safaitic inscriptions contain theophoric names that refer to him.
According to the Quran, the Meccan pagans acknowledged Allah as the creator-God ("If you asked them who created the heavens and the earth, they will say 'Allah'," Luqmān 31:25), but made him part of a pantheon with divine associates and relatives. The argument of G. R. Hawting and Patricia Crone that it is wrong to think of the Allah of the pagans as a deus otiosus or "High God" in the terms of Victorian and early twentieth-century notions about the evolution of religion may be correct. 27 He is, however, in some way extraordinary, lacking many temples but abundantly present in theophoric names. Given the long and frequent identification of Zeus with North Arabian deities such as Dushara in Transjordan, Syria, and elsewhere, it is likely that for the remaining pagans in the sixth and seventh centuries, Allah was simply the chief of the gods, by analogy to Zeus. North Arabian polytheism is denounced in al-Najm 53:19-23: "Then, have you seen Allāt and al-ʿUzza and Manat, the third, the other one? Do you have, then, males and God only females? That would be an entirely unfair division. They are only names you have given them, you and your ancestors," adding in verses 26-27: "However many angels subsist in the heavens, their intercession means nothing, except if God permits it, to whomever he wills and pleases. Those who do not believe in the afterlife give female names to the angels." As I have argued elsewhere, this is the same sort of late antique argument that Augustine of Hippo made in his De civitate Dei contra paganos: that pagans such as Neoplatonists who recognized a first principle and who referred to the gods as angels could and should become monotheists by adopting a Christian angelology that reduced them to beings lacking autonomy: "If the Platonists prefer to call the angels gods rather than demons and to number them among those whom their founders and master Plato asserts were created by the supreme God, let them say this if they wish; for we must not labour over a merely verbal controversy." 28 Crone suggests that al-Najm 53:19-27 depicts its opponents as "pagan monotheists" because it says they configured the goddesses as angels. 29 If she is correct, this is another reason not to translate kāfir as "infidel" or "unbeliever." We may have to nuance her argument, however. For gods to have angels in a polytheistic environment is not necessarily a step toward monotheism. Adherents of Near Eastern religions in the Hellenistic and late antique periods routinely asserted that the gods had divine envoys or angels, who were themselves objects of devotion. In pre-Christian Palmyra, the god Bel had a Malakbel or angel, whom locals made an object of worship. Likewise, the largely Syrian god Baalshemim or Lord of the Heavens had an angel. In the Syrian town of Maloula, an Aramaic inscription transliterated into Greek script and dated 107 Ce speaks, Teixidor says, of "mal'ak 'el-'aliyan, i.e., the 'Angel of god the Most High'." 30 It is implied in the Quran that the pagans made Allah the father of the three goddesses. Such family relationships are typical of pantheons, as with the Dodekatheon at the Parthenon in Athens, where Athena herself was held to be the daughter of Zeus and Poseidon to be his brother. Arabic speakers of the Transjordan, which was culturally linked to the Hijaz, might still have known a remnant of Grecophone worshipers of Zeus and the other Olympians in the sixth century, as an overlay on local figures. As Ahmad Al-Jallad has shown, the trope of the daughter of God existed in ancient Arabia, where an inscription calls Allāt the daughter of the martial star-god, Roḍaw. Given that in Roman times in Transjordan and Syria Allāt was identified with Athena, the daughter of Zeus, and Zeus, in turn, would have been identified with Allah, Hellenic influence may have been in part responsible for the belief, implied in the Quran, that Allāt was the daughter of Allah. The sense of "to worship the gods" for k-f-r is underlined in al-Baqara 2:257: "God is the patron of those who believe, bringing them out of darkness into the light. And those who kafarū, their patrons are Ṭāghūt, who bring them out of the light into darkness." Ṭāghūt is a loan from Geʿez that means "new or alien god" or "idol," and, interestingly, is treated as a plural in this quranic verse, corresponding to numerous patrons. 32 Belief in polytheistic religion is not, properly speaking, disbelief but the wrong sort of belief, from the point of view of the Quran. It is not a charge of atheism. Not only are such believers committed polytheists but they are also militant: "Those who believed fight in the path of God, and the pagans (al-ladhīna kafarū) fight in the path of Ṭāghūt, so fight the associates of Satan, for the guile of Satan is feeble" (al-Nisāʾ 4:76).

rebellion and libertinisM
The pagans with whom Muḥammad debated are depicted as denying the bodily resurrection and the Eschaton: "They swear the most strenuous of their oaths by God that he will not resurrect the dead. Rather, it is a true promise that he has made, but most of the people do not know. He will show them that about which they differed and will teach those who kafarū that they are liars" (al-Naḥl 16:38-39). With regard to the Quran's prediction of the end of days, it is reported (al-Anbiyāʾ 21:97): "And the true promise approached, and behold, the eyes of the pagans (al-ladhīna kafarū) stared fixedly: 'Woe to us, we were heedless of this, rather, we were wrongdoers'." Elsewhere, it is admitted that they are believers in their own tradition; when they question the eschatological opening or grand success, the verse reads: "Say: On the Day of the Opening, the faith (īmānuhum) of those who kafarū will not benefit them, nor will they be granted a respite" (al-Sajda 32:29). Since it is allowed that they have faith, they are not unbelievers strictly speaking and translating this phrase as "the faith of the infidels will not benefit them" would be self-contradictory. While they are not accused of disbelieving, they are, however, liars and wrongdoers, dishonest and workers of evil (cf. al-Nisāʾ 4:167-68). As well as labeling them "wrongdoers" (sing. ẓālim), they are "morally dissolute" (fāsiqūna) for responding incorrectly to God's proverbs (al-Baqara 2:26). Along the same lines, it is said of Muḥammad's monotheistic followers: "God has caused you to love faith, rendering it beautiful in your hearts, and he has caused you to abhor impiety (kufr) and ungodly behavior (fusūq) and rebellion" (al-Ḥujurāt 49:7).
"Rebel" is one meaning of the root k-f-r. In the story of how Lucifer fell (al-Baqara 2:34) it is reported: "And when we said to the angels, 'Bow down to Adam', they prostrated themselves, save the Devil; he refused, and grew haughty, and so he became one of the rebellious (kāfirīna)." The active participle here does not involve disbelief but disobedience. The Devil (Iblīs, < Gk διάβολος) is not accused of rejecting the existence or oneness of God but of refusing the divine order to bow down to the first human being. Indeed, in 2:30 the angels are depicted as arguing with God that creating Adam would lead to turmoil, and the implication is that Satan parted ways with God not because he disbelieved but because he had a positive if misguided motive-he differed with him on the wisdom of opening Pandora's box. 33 The association of the term with impiety or libertinism is clear in sura Nūḥ, where 71:27 says of opposers of the patriarch, "Truly if you leave them alone, they will misguide your servants, and will give birth only to a dissolute libertine (fājiran kaffāran)." Here kufr is equated with impiety, which Grecophone Christians in their polemics against the pagans called ἀσέβεια. 34 Likewise, in al-Taḥrīm 66:10 God had made the wives of Noah and Lot an object lesson for those who kafarū because of these women's preference for pagan society over their husbands. The reason given in 2 Pet 2:6 for the calamity that befell the people of Sodom and Gomorrah is that they lived impious lives (ἀσέβεσιν), which seems roughly the meaning of kufr in Q 66:10.
As al-Khalīl and Ibn Manẓūr noted, one meaning of the simple verb kafara is "to be ungrateful," and in this it resembles its Syriac cognate. In al-Shuʿarāʾ 26:18 Pharaoh upbraids Moses for his opposition, saying "Did we not bring you up in our home as a child and did you not live with us for many years?" He adds, "Then you carried out the deed that you committed, and you are among the ungrateful (kāfirīna). 35 God addresses humankind: "Remember me and I will remember you, and be thankful to me, and do not be ungrateful (takfurūni)" (al-Baqara 2:152). Here the connotations of the word are "heedless" and "unappreciative," the antonyms of the virtues praised. This sense of the term might be compared to the New Testament verse Luke 6:35: "But love your enemies, do good, and lend, expecting nothing in return. Your reward will be great, and you will be children of the Most High; for he is kind to the ungrateful (ἀχαρίστους)." In the Peshiṭṭa, "ungrateful" is translated with a cognate of the Arabic k-f-r. blaspheMy A controversial passage in al-Baqara 2:102 provides a further sense of the verb. The Quran condemns those in the era of Solomon who followed demons (shayāṭīn) that taught magic. It goes out of its way to underline that Solomon himself did not commit kufr, even though in late antique folk tradition he was held to be able to control sprites and demons. The demons were guilty of putting otherwise inoffensive teachings to evil purposes, turning them into black magic, so that they kafarū (A. J. Arberry translates this as "disbelieved"). 36 Of what, however, did this act consist? It does not appear to have been a denial of anything, but rather was a blasphemous activity. The humans were eager to have the teaching of the two angels of Babylon, Hārūt and Mārūt, which they then desecrated by turning it into dark arts so as to separate spouses from one another. The demons' instruction harmed people rather than benefited them, and turning to the occult deprived these individuals of any portion of heaven.
Hārūt and Mārūt are two of the Zoroastrian celestial spirits, Haurvatāt and Ameretāt. These emanations of the supreme deity, Ahura Mazda, symbolize wholeness and immortality. 37 For instance, in the Younger Avesta, Yasht 19.95-96, the last days during which the world will be renovated are described thus: "Evil thought will be overcome, good thought will overcome it . . . The celestial spirits Integrity (Haurvatāt) and Immortality (Ameretāt)  Iranica, 2012Iranica, (2003, http:// www.iranicaonline.org/articles/harut-and-marut. Both of these articles stress the "fallen angel" interpretation, which I do not believe the actual text of the Quran upholds. will defeat the demons of Hunger (Shud) and Thirst (Tarshna)." 38 The two celestial spirits associated with nemeses among the demons symbolizing bodily human cravings like hunger and thirst may have inspired the Quran's motif that devils misused their teachings to satisfy lust. Moreover, Ameretāt is associated with plants, fertility, and the tree of life. The Quran could be projecting into the time of Solomon a contemporary set of Zoroastrian ideas. The retrofitting of this motif to the time of the Hebrew monarch may in turn have come about because of the association in late antiquity of Solomon with mastery of the sprites or demons, which is reflected in quranic passages. 39 In late antique Greek Christian authors, black magic was associated with blasphemy (which originally meant slandering [God]). In his "Homily 10 on 2 Timothy," John Chrysostom (ca. 349-407 Ce) wrote, "Let us then so live that the name of God be not blasphemed (βλασφημεῖσθαι)." Among the many examples he gave of Christians blaspheming in failing to live up to their ideals were "your auguries, your omens, your superstitious observances . . . your incantations, your magic (μαγείας) arts." 40 What if we translated al-Baqara 2:101 this way?
They followed what the demons recited over the realm of Solomon. Solomon himself was not a blasphemer, but the demons were blasphemers, teaching the people magic and what was revealed to the two archangels of Babylon, Haurvatāt and Ameretāt. But these two had been careful not to teach anyone without warning them, 'We are a potential disturbance of faith (fitna), so do not fall into blasphemy.' From them they learned how they might divide a man and his wife [. . .].
Here is a condemnation of warlocks and witches who engage in what is seen as necromancy, which apparently enables those who covet married persons to cast spells to separate them from their spouses. They are instructed by demons who pervert and misuse the teachings of divinely inspired Zoroastrian angels.
Later Muslim commentators on this text are divided over its meaning. Some saw the anecdote as concerning fallen angels. 41 Others defended the angels as having been sinless, and held that while they performed licit miracles, the demons turned their teachings to the purposes of thaumaturgy. 42 As I read the text, the teaching of the angels itself is not being condemned here. Solomon, the verse says, bore no blame for his mastery of the spirits. The Zoroastrian celestial spirits are spoken of with reverence, called angels rather than demons, and are depicted as having been given inspiration (unzila) by God. The angels act responsibly inasmuch as they give disciples an explicit warning that learning their esoteric teachings Instead the problem lies with the demons, who announce no such alert about how this arcane teaching could result in a departure into impiety or blasphemy, and who appear to encourage people to misuse the spells. The passage does not refer to the rejection of God or of monotheism, but to a set of beliefs and practices that have their origin in angelic inspiration but were then perverted for satanic purposes. To the already-mentioned synonyms of "to be morally dissolute" and "to disobey," we may thus add "to blaspheme" as a connotation of kafara. This conclusion is bolstered by al-Māʾida 5:103, "God has not prescribed Baḥīra, Sāʾiba, Waṣīla, or Ḥām; but those who kafarū slander (yaftarūna ʿalā) God, and most of them have no understanding." The verse denies that God had ordered camels to be used in pagan sacrifice, and accuses these pagans of libeling the supreme deity. To defame God is the original meaning of blasphemy in Greek (see Matt 12:31), and it is possible that the Arabic phrasal verb iftarā ʿalā is a calque on the Greek βλασφημία here; it is being equated with kafara.
Inasmuch as the Quran does condemn the kāfirūn on doctrinal and moral grounds and directs them to abandon their pagan beliefs and practices, there is, of course, a sense in which it views them as outside of and antagonistic to the true faith, part of what translators who used the term "unbeliever" wished to convey. I would argue, however, that there is a key lexical difference between a denier of God and an affirmer of God who gets God wrong. apostasy Another connotation of kafara is "to apostatize." Al-Tawba 9:74 remonstrates with those who had covertly rejoined the pagans in Mecca: "They swear by God that they did not say it, but they indeed uttered the word of kufr, and kafarū after their acceptance of the monotheistic tradition (islāmihim)." In the Quran, islām is not used to refer to the religion of the Prophet in particular, but rather to the prophetic monotheistic tradition that it holds runs through all the valid religions. 43 In late antiquity, with the vast influence of Greek, a "word" or λόγος implied a system of religious belief. Thus, the pagan Celsus had entitled his defense of Hellenic religion against Christianity Λόγος ἀληθής (The true word). The Aramaic for λόγος is meltā, taken into quranic Arabic as milla. In Ṣād 38:7, the pagans are represented as rejecting Muḥammad's message and saying, "We have not heard of this in the ultimate Word (al-milla al-ākhira); this is surely an invention" (presumably the odd Arabic phrase is a calque on a Greek usage such as τέλειος λόγος).
Likewise, Christians who responded to pagan critique accused the latter of adhering to a false logos. 44 After his conversion to Christianity, the emperor Constantine urged his subjects to turn to the worship of the one God and obedience to his laws, but allowed them, if they so desired, to retain their "sanctuaries of the false Word (pseudologias témeni)." 45 In al-Tawba 9:74, adhering to "the word of kufr" is made a synonym with "to practice polytheism," and in context "to revert to paganism." It is instructive to compare it to 1 Tim 4:1-2: "Now the Spirit expressly says that in later times some will renounce the faith (ἀποστήσονταί) by paying attention to deceitful spirits and teachings of demons, through the hypocrisy of liars (ψευδολόγων) whose consciences are seared with a hot iron." The Greek ψευδολόγος, richer than the English "liar," involves following or speaking a false logos, a way of life or structure of thought that is distorted by its untruth. It is worthwhile comparing ψευδολόγος to the quranic phrase "word of kufr" (kalimat al-kufr), which brings out the implication of the Arabic word of falsehood. Like the Hijazis castigated in the Quran, the early Christians mentioned in 1 Timothy affirmed this false Word after having earlier been believers but then apostatizing. Likewise, the Middle Persian agdēn, "without religion" or "infidel" was not simply a term denigrating non-Zoroastrians but tended to be deployed especially with regard to apostates from the faith of Zarathustra. 46 It is not only Muḥammad's believers who can apostatize. The Quran complains of Jewishpagan syncretism in Medina and of the wish of that faction of Jews that his community should revert to the worship of the old gods. Al-Nisāʾ 4:51 asks: "Have you not seen those who were given a portion of the Book believing in Jibt and Ṭāghūt and saying to those who kafarū that they are better guided to the path than those who have believed?" I conclude that those given a "portion" of the Bible are the Jews, as opposed to Christians who had both the Old and New Testaments. Such hybrid Jewish-pagan practices have been documented in earlier centuries in the Roman-ruled Levant, and come as no surprise in Medina, on the stillpagan fringes of the eastern Roman empire. 47

doCtrinal sins of the Monotheists
Monotheists are for the most part contrasted in the Quran with kāfirūn as a noun, as two different sociological and theological communities. The verb kafara, however, is more fluid and is sometime applied to monotheists. Āl ʿImrān 3:167 complains about those of Muḥammad's believers who declined to go out to defend the city (later commentators say the verse concerned the battle of Uḥud in 625): "They were told, 'Come, fight in the path of God, or at least take a defensive position'. They replied, 'If we knew how to fight, we would have followed you'. That day, they were closer to kufr than to faith, inasmuch as they said with their lips what was not in their hearts. God knows best what they are concealing." The deverbal noun kufr here clearly means hypocrisy or dishonesty rather than disbelief. It is not, as with the distinction in the active participle between kāfir and believer (muʾmin), fixed or black-and-white or serving to demarcate reified social groups (compare al-Ghāfir 40:14, "So call upon God in sincere service to him, even if the kāfirūn hate it"). The verb kafara and the abstract noun kufr rather exist on a spectrum and can characterize Muḥammad's believers, at least briefly, when they make the wrong moral decision. Kufr in 3:167 means "inauthenticity" or "bad faith." In contrast, the believers cannot be kāfirūn, the simple noun. This distinction between the hard, inflexible noun kāfir and the fluid verb kafara is an example of idiosyncratic polysemy, where the verb and its deverbal nouns have evolved in different directions. 48 A key attribute of the kāfir, as we have seen, is that such a person is damned to hell. Al-Mulk 67:6 reads: "And for those who denied (kafarū bi-) their Lord, there awaits the torment of hell, and a wretched destination!" In contrast, in speaking of Jews and Christians we find in al-ʿAnkabūt 29: "Debate the scriptural communities only in the best of ways, except for those who do wrong. Say 'We believe in the revelation sent down to us, and the revelation sent down to you; our God and your God is one, and to him we have submitted'." According to a late seventh-century Christian author, Jacob of Edessa, "Muḥammad went down for trade to the lands of Palestine, Arabia and Syrian Phoenicia." 49 Muḥammad thus lived and worked in Christian societies and would have had a fair experience of contemporary Christian practice and belief. The Quran shows positive attitudes throughout to Christians and al-Baqara 2:62 admits Christians to heaven ("Those who believed, and the Jews, and the Christians, and the Sabians, and whoever has believed in God and the Last Day and performed good works, they shall have their reward with their Lord"). To underline the difference, the Quran shows God pledging to Jesus regarding future Christians in Āl ʿImrān 3:55: "God said, 'Jesus, I will take you to me and will raise you to me and I will purify you of those who kafarū and will render those who follow you superior to those who kafarū until the judgment day'." Likely it is distinguishing between the old pagan Romans, who had persecuted Jesus and his faithful, and the Christians themselves. There will always be, the Quran vows, a difference between followers of Jesus and the kāfirūn. This and other passages suggest to me that the deverbal noun kāfir is never used tout court for Jews and Christians.
Hawting's proof text against this proposition, al-Anʿām 6:89, whereby he attempts to show that Jews and Christians can be kāfirūn, is not persuasive. In this verse, God has bestowed scripture, wisdom, and prophecy on a people, but if they prove ungrateful for them (fa-in yakfur bihā), God will delegate these gifts to another people who are not ungrateful for them (bihā kāfirīna). 50 But kāfirīn bihā, taking an oblique object, is a phrasal verbal noun. Using a verbal noun for ingrates among the people of the Book is no different from using the verb itself, which the Quran sometimes does. This passage is hypothetical and hyperbolic, warning the biblical communities that they are not indispensable, but not categorizing them as kāfirūn, which as a simple noun refers to the damned pagans and not to saved Jews or Christians who worship the same God as Muḥammad and his believers. Fred Donner has argued persuasively, in my view, that Muḥammad's movement was ecumenical, including nonconvert Jews and Christians, and that righteous Jews and Christians clearly were not categorized as kāfirūn, and I see an overlap between the political alliances in Medina and this soteriological pluralism. 51 Scholars have attempted to explain why the Quran for the most part demonstrates a positive attitude toward some Christians (naṣārā) but uses the verb kafara to denounce others. François de Blois suggests that the Quran's naṣārā were not Christians in general but rather were a sect of Jewish-Christians of a sort of which the Unitarian Muḥammad might approve. 52 Iranian Zoroastrians, however, used a similar word, equating to Nazarene, and made a distinction between local Christians (nāsrāye) of the Church of the East and western Catholics (Christians). 53 The problem is that Christians in the Quran, unlike the Zoroastrian usage, are never distinguished by terminology and there is no reason to think that a Jewish-Christian sect existed in the early seventh century. 54 If Muḥammad was as peripatetic as both seventh-century Christian and later Muslim sources allege, it is impossible that he had a narrow, provincial view of Christianity and was only familiar with some Judaizing Hijazi version. Further, the Quran (al-Rūm 30:1-6) evinces hope that the Roman emperor Herakleios would defeat his Iranian foes and identifies that prophesied victory as the triumph of God himself, which makes no sense if Constantinople was seen as a center of infernal infidels. 55 Rather, Muḥammad appears to have been satisfied that conventional Chalcedonians and Miaphysites were still monotheists, even if they had departed somewhat from the Abrahamian λόγος.
It should be remembered that when Roman Christians distinguished themselves from pagans, they underlined God's unicity, and inscribed εἷς θεός (one God) on doorways and lintels. Harold Remus notes that Ambrose contrasted the Christian deity, "Almighty, One, only and True (omnipotenti Deo . . . unusquisque Deum verum . . . solus verus; Ep 17.1)" to the images (simulacrum) of the pagans. 56 Zachariah Scholasticus told the story of a punitive expedition ordered by patriarch Peter III of Alexandria (482-489) ninety years before Muḥammad's birth, to the town of Menouthis to the northeast of the metropolis, which was still a center of Isis worship. Some Tabennesiote monks from Canopus came along. At one furtive temple, worshippers of the old Egyptian gods had attempted to hide the entrance and conceal their idols and sacrifices within. They were nevertheless unmasked, and one of the monks entered to find the idols and a blood-stained altar. As Trombley notes, Zachariah wrote that the monk exclaimed, "'One God!' (had ʿAlāhā), as he wished to say by that that it was necessary to extirpate the error of polytheism." 57 This anecdote could just as well be a later Muslim one, with the same diction.
In the Medinan period, the Quran uses the verb kafara when it begins speaking of an antagonistic group from among the other monotheists: "Neither those who kafarū from among the people of the Book, nor the polytheists (mushrikūna) themselves, desire that good from your lord descend upon you" (al-Baqara 2:105). Some groups from among the biblical communities had allied politically with the militant pagans. A hypernym-for instance, "tree"-is lexically superordinate to hyponyms, another set of nouns or phrases under its rubric (e.g., "juniper" and "acacia"). Here the phrase "people of the Book" functions as a phrasal hypernym, which is lexically superordinate to the hyponym "Those who kafarū from among the people of the Book." 58 Logically speaking, the need to identify this subset of believers in the Bible as those who kafarū proves that kāfir does not ordinarily refer to Jews and Christians. That is, if all Jews and Christians were always kāfirūn, it would be redundant to identify this group "from among the people of the Book" as "those who kafarū." Moreover, if all Jews and Christians were always kāfirūn, it would make nonsense of God's pledge to Jesus (Āl ʿImrām 3:55) that he "will render those who follow you superior to those who kafarū until the judgment day." Christians are not kāfirūn under ordinary circumstances, just as they are not doomed to hell under ordinary circumstances. Still, just as they can commit mortal sins and so depart from righteousness into perdition, so they can throw in with bellicose polytheists against Muḥammad and his cause, and likewise join the damned.
Al-Bayyina 98:6 warns: "Those who paganized (kafarū) from among the people of the Book and those who make God part of a pantheon shall be in the fire of Gehenna, dwelling therein forever; those are the worst of creatures." Although it has been common for the later Muslim commentary tradition to identify this group as Jews, there is no a priori reason for this interpretation, and it is just as possible that they were Christians or pagan monotheists. In translating kafara in this phrase as "to paganize," I am suggesting that in context it is speaking of a biblical community that allied politically with Muḥammad's pagan enemies, and I suspect that the Quran views it as treason rather than as heresy.
Al-Ḥashr 59:2 appears to refer to a conflict with Muḥammad's believers provoked by a monotheistic village throwing in with pagan Mecca. It addresses the Medinans and says of God, "He it was who expelled from their homes the scriptural community that paganized, at the first gathering." The following verse, 59:3, however, seems to see this military defeat of monotheistic traitors as having been salutary for them, inasmuch as their exile spared them from being in the crossfire between Medina and Mecca, and it appears to imply that they had been removed from the temptation of paganizing and so reverted to being among the saved: "If God had not prescribed for them exile, he would have tortured them in this world and they [would have] undergone the torments of fire in the next." As with the verb kafara in general, to "paganize" is an action rather than an essence, and those who commit the act can be redeemed. In this it differs from the noun kāfir, which refers to a person who must change (i.e., convert) in order to escape damnation.
Kafara is used of Jews and Christians, as it is of Muḥammad's believers, to describe wayward actions other than political treason, as well. Al-Baqara 2:253 asserts, "Those Messengers-some we have preferred above others; among them are some to whom God spoke, and some he raised up through levels. And we gave Jesus the son of Mary clear signs, and supported him with the Holy Spirit. And had God willed, those who came after him would not have fought against one another after clear signs had come to them; but they differed with one another, and some of them believed, and some kafarū; and had God willed they would not have fought against one another; but God does whatever he desires." Here violence 58. "Hyponymy and Hyperonymy" (M. L. Murphy), in Encyclopedia of Language and Linguistics, ed. K. Brown, 2nd ed., 2006, https://www.sciencedirect.com/referencework/9780080448541/encyclopedia-oflanguage-and-linguistics among Christian sects in the sixth and seventh centuries is denounced. 59 The Muslim scripture is unlikely to have seen either Miaphysites or Chalcedonians as doctrinally superior and so is probably not distinguishing between them by using the terms "believed" and kafarū. It may be using the latter term to describe Christian sects that departed too far into a form of paganism to be acceptable to the Unitarian Prophet, such as the Gnosticism that denied that Jesus and Mary ate food or forms of Tritheism. It may also be that during Sasanian rule in the Levant during the first third of the seventh century, some pastoralist Arab nominal Christians reverted to paganism. The imperial poet Georgios of Pisidia wrote that in 622 "a battalion of long-haired Saracens"-presumably pro-Sasanian Arab foederati or allied cavalry-came up from Syria to attack the army of Emperor Herakleios. 60 They were either traditionalists who worshiped North Arabian deities such as Allāt, or Miaphysite Christians who held that Constantinople had veered into heresy and who therefore allied with the Sasanians. Users of Syriac among the latter referred to their doctrinal enemies with words from the root k-p-r. 61 While Christians as such are never called kāfirūn in the Quran, then, they are capable of engaging in acts for which it uses the verb kafara and which are on a sliding scale ( just as with Muḥammad's cowardly believers who avoided Uḥud). Many of these acts appear to be venial sins. The processual character of the verbal form of the root is appropriate to sinful actions taken by a saved community in a way that the fixed noun is not. a "Collyridian" heresy?
The Quran deploys the verb kafara, as well, in a manner similar to the use of αἵρεσῐς (heresy) among Christian writers (e.g., Gal 5:20, 1 Cor 11:19). Late passages of the Muslim scripture denounce a group that it appears to view as a syncretic Christian-pagan sect. Al-Māʾida 5:116 reads, "And God said to Jesus the son of Mary, 'Did you say to the people, "Take me and my mother as two gods other than God?"' He replied, 'Praise be to God, it is not for me to say what I have no right to say. Had I said it, you would have known. You know what is in my soul, but I do not know what is in yours. In truth, you know things unseen'." The Quran does not refer to these Mariolaters as Christians, but is apparently describing a faction of Christianity that held that Mary and Jesus were gods, possibly having assimilated them to the Arab goddess Allāt and to one of the male North Arabian gods, respectively.
This group resembles the sect that Epiphanius in his Πανάριον (Lat. title, Adversus haereses) calls Arab Collyridians, who allegedly mixed Christian motifs with Nabataean religion in the fourth century. 62 Some have challenged this interpretation, whether because Epiphanius appears to have had an active imagination or because they wished to emphasize the Quran's polemical techniques. 63 Recent work on the apocrypha of the Dormition of the Virgin Mary, written in the sixth and seventh centuries but incorporating much earlier material, has, however, lent credence to some of Epiphanius's assertions. 64 If, moreover, the Quran is speaking of a pagan-Christian sect in sura al-Māʾida (5) rather than of mainstream Christianity, that would help resolve the tension in the text between its pluralistic soteriology for righteous Christians (al-Baqara 2:62) and its use in al-Māʾida 5:73 of k-f-r to describe theologically extreme followers of Jesus. While the sect the Quran denounces is not identical to that described by Epiphanius, the two do have some similarities, and his epithet of "Collyridians" is as good a name for them as any other.
The Quran's denunciation of this religious group parallels that of Epiphanius himself, who wrote: "Yes, of course Mary's body was holy, but she was not God. Yes, the Virgin was indeed a virgin and honored as such, but she was not given to us to worship . . . For the age-old error of forgetting the living God and worshiping his creatures will not get the better of me. They served and worshiped the creature more than the creator . . . If it is not his will that angels be worshiped, how much more the woman born of Ann . . . She was surely not born other than normally. . . the Word . . . had assumed flesh from a holy virgin. But certainly not from a virgin who is worshiped, or to make her God, or to have us make offerings in her name, or again, to make women priestesses after so many generations." 65 The Melkite (Chalcedonian) Patriarch of Alexandria, Eutychius (d. 940), also took up this theme, alleging that at the Council of Nicea Marianites were condemned for holding that aside from the supreme God there were two other gods, Jesus and Mary. 66 Muḥammad appears to have encountered such a sect only at the end of his life, since such beliefs are not mentioned until the very late sura al-Māʾida.
Al-Māʾida 5:75 also seems to critique a similar heresy: "Christ, son of Mary, is only a messenger, and messengers before him passed away. His mother was upright. They both used to eat food. Behold how we explain the signs to them, then behold how deluded they are." This verse rebuts a theology that Jesus and Mary were immaterial and therefore did not need bodily nourishment and that Jesus was immortal and incapable of actually dying. Likely addressing these "Collyridians," al-Māʾida 5:73 asserts: "They have committed heresy (kafara), who said that God is the third of three." The idiosyncratic Trinity here, where God comes last after two other figures, is likely the one where Mary and Jesus are both gods. If this sect was a Nabataean matriarchal tradition that identified Mary with the goddess Allāt, it may well have exalted Mary above the other two. In any case, the verse cannot be aimed at Chalcedonian or Miaphysite Christians of the Roman empire, and the Quran is certainly denouncing a group that the mainstream Christians of that era would also have viewed as heretical, whether Collyridians or Tritheists or some other. 67 the sixth-century Code of Justinian were still issued in Latin as well as Greek in the sixth century. 73

ConClusion
I have argued that kāfir in the Quran for the most part does not mean "unbeliever" or "infidel." In most of our examples, a lack of belief is not at stake. Rather, kāfir is a polysemous term that has a wide range of meanings, including "peasant," "pagan," "libertine," "rebel," and "blasphemer." These are discernible if we look at the parallelisms, synonyms, and antonyms with which quranic verses surround this noun. I understand the impulse of translators to use "unbeliever" for kāfir, and, of course, the term sometimes does mean just that. Moreover, the condemnations of pagan belief and practice, while often made with other terms, could be seen to imply unbelief at some meta level. I argue, however, that limiting the meaning of the root so severely causes us to miss a rich set of other connotations that give us a rounder idea of the Quran's intent.
The widespread assumption that k-f-r primarily means "to reject, deny, disbelieve" derives from a failure to distinguish between two distinct forms of the verb, the idiomatic phrasal verb kafara bi-and its verbal noun kāfir bi-and the simple intransitive verb kafara and the noun (which does not take an object) kāfir. Kafara bi-and its verbal noun do indeed have to do with denying, rejecting, and disbelieving. The simple intransitive verb kafara and its deverbal nouns, however, exhibit an extensive polysemy with "to reject" being only one of the meanings. This finding is of great importance to commentators on and translators of the Quran.
I have suggested that the bilingual lives of many Arabic speakers in and on the fringes of the Roman empire over hundreds of years (Arabic-Aramaic and Arabic-Greek) contributed to this polysemy, through the phenomenon of the loanshift. The Latin paganus, which came to have the connotation both of "rustic" and "polytheist" in the fifth and sixth centuries, may well lie behind al-Ḥadīd 57:20, which refers to kuffār as peasants happy to see rain and greenery. At the same time, the quranic term is clearly also used to refer to polytheists. Ṣād 38:5 reports of the kāfirūn that they rejected the notion that the many gods could merge into only one, while al-Baqara 2:257 says that those who kafarū had taken the deity or idol Ṭāghūt for their patron instead of God. Āl ʿImrām 3:151 menaces these pagans with hellfire for having made God part of a pantheon (ashrakū). While it is not impossible that Arabic independently invented a connection between farmers and polytheists, Occam's razor would suggest that we instead posit that Arabic was influenced by late antique Roman Christian usage, which was embedded in imperial laws applying to Arabophone citizens of the empire. In any case, far from being deniers or nihilists, the pagans are admitted to believe in their own religion (dīn) and to have faith (īmānuhum) in it. It is simply a false religion. Kafara thus has a positive valence that "to disbelieve" does not capture, even if the latter is not ultimately an incorrect characterization of the quranic view of the pagans.
Of course, the term most often has negative connotations. It is equated in several verses to moral turpitude. Where the root k-f-r has to do with impiety or impious, immoral actions, it may be a loanshift for the Greek ἀσέβεια or impiety, a term often applied to pagans by Christian polemicists in late antiquity. I argue that in the story of the demons' perversion of the divine teachings of Haurvatāt and Ameretāt, the root k-f-r means "to blaspheme." This sense of "to blaspheme" or "to rebel" is also apparent in al-Baqara 2:34, which tells the 73. The Codex of Justinian, esp. book one. story of the Devil's refusal to obey God's order that he bow to Adam. In neither instance are the miscreants guilty of disbelieving anything. In al-Shuʿarāʾ 26:18, this term is put into the mouth of Pharaoh when he is clearly scolding Moses for his ingratitude. In some instances, those characterized by k-f-r are accused of ingratitude to God. This connotation has parallels with the Syriac use of the root, though the polysemy of the Arabic terms derived from this root is so great as to argue against seeing it solely through the lens of Syriac Christian texts.
I have argued a further distinction, one between the verb kafara (which can be used with regard to any group of human beings, including monotheistic communities) and the plural deverbal noun kāfirūn. This noun refers to a reified social group, that is, the pagans. Thus, in al-Kāfirūn 109:6, Muḥammad addresses the kāfirūn, saying that he does not worship what they worship and that they have their religion and he has his. In contrast, al-ʿAnkabūt 29:46 instructs the believers in Muḥammad's mission to address the people of the Book and say, "We believe in the revelation sent down to us, and the revelation sent down to you; our God and your God is one, and to him we have submitted." Whereas the noun has a sociological function in demarcating a group of people with a polytheistic faith, the verb is more analogous to "to sin," in involving actions that even monotheists can commit. Idiosyncratic polysemy of this sort is produced by the divergence of the verb from its deverbal noun. Thus, when some of Muḥammad's believers declined to come out of Medina and defend the city at Uḥud, they had committed kufr (Āl ʿImrān 3:167) or shown bad faith on that occasion, but it would be nonsensical to call them infidels. Likewise, in the same sura God promises Jesus (3:55) that his followers, the Christians, will always be superior to the pagans right until the judgment day.
There are three main circumstances in which the verb kafara characterizes monotheists in the Quran. The first is where a subset of them ally politically with the Meccan pagans and become military foes of Muḥammad's believers, becoming "those who paganized from among the people of the book." The second is where monotheists have committed a serious offense, such as disobedience, blasphemy, corruption, impiety, or another moral lapse. The third is where they apostatized, reverting to paganism, or where they adopted beliefs so incompatible with monotheism as to have departed into a paganizing heresy, as with the "Collyridian" Christians who made Jesus and Mary gods and appear to have exalted them above God the father.
I conclude that the noun kāfir is in general best translated as "pagan" rather than "infidel," in acknowledgment that it is used in the Quran in the main to refer to provincial polytheists. My finding of substantial polysemy in the root k-f-r, however, also suggests that in many instances the best translation will derive from context, and that derivatives should not be rendered uniformly. Given the new understanding of the linguistic situation in Transjordan afforded us by the Petra papyri and rock inscriptions, it is now important to look for Greek, as well as Aramaic, loanshifts when considering the meaning of Arabic theological terms, and I have suggested that some of the polysemy we see in k-f-r may derive from the other languages late antique Arabic speakers cultivated.