Semantic decomposition of four Quranic words

Full Text

1. Introduction

1.1. In Honour of Igor Mel’čuk

When I was invited to contribute to this special edition of the Russian Journal of Linguistics to honour Igor Mel’čuk’s 90th birthday, my affection for Igor made me very keen to share in the privilege of honouring him. Although I had been out of professional academic linguistics for more than two decades, heartfelt affection prevailed, so here is my offering.

I first had the pleasure of meeting Igor in the early 1980s, when he was visiting the Australian National University to give a workshop on Meaning–Text Theory (MTT). To this day I have a vivid memory of Igor’s impassioned appeal at the end of his presentation, calling for co-workers to join him in his linguistic mission. I was a graduate student at the time, and although I did not feel ready to sign up on the spot to devote my life to MTT, I was deeply impressed by Igor’s humanity, creativity and joy, and inspired by his love for words. He won my heart as well as my head, and a friendship formed which has endured beyond many others.

Twenty years after meeting Igor, I retired from academia to serve as an Anglican priest. If anything, our friendship grew stronger after this. He would address me as Saint Mark, and I would address him as Jesus, in honour of his father’s thwarted desire to name him Yehoshua.¹

The last time I checked, Igor Mel’čuk was an atheist. Yet when he quoted Anna Wierzbicka’s (2001: 21) semantic explication of God, he reminded us that God is good (Mel’čuk 2018: 536). Given Igor’s admiration for Wierzbicka, and his celebration of her “gift to the world” of semantic decomposition (Mel’čuk 2018: 522), he will appreciate why, although my height may be enough for most purposes, I have chosen for the best of reasons to stand on Anna Wierzbicka’s excellent shoulders to deliver this homage to ‘Jesus of Montréal’.

1.2. The Challenge of Translating Distant Texts

In my work as an Anglican priest I developed an interest in comparative theology and especially in the relationship between Islam and Christianity. I have long been fascinated by the ways in which differences in the meanings of key words frame how people of different faith traditions understand – and misunderstand – each other. Ludwig Wittgenstein (Wittgenstein 1958: §114) once wrote: “One thinks that one is tracing the outline of the thing’s nature over and over again, and one is merely tracing round the frame through which we look at it.” His point was that, far from merely describing the nature of reality, logical propositions tell us as much about the language they are couched in as about the thing they purport to describe. So, too, do the words we use: they tell us as much about our thought-culture as they do about the world they point us to.

We are imprisoned by the frames we peer through, above all by those familiar friends, the words we use. This has certainly been evident in the long history of European engagement with Islam, in which Christian scholars have viewed Islam and its scripture, the Quran, through the lens of biblical concepts. Over the centuries the biblical framing of Islam has embedded itself deeply into Western scholarship on Islam, including translation practices. This paper seeks to shine light on this framing, and challenge it by exploring the meanings of a handful of key Arabic terms using Natural Semantic Metalanguage (Goddard & Wierzbicka 2014). These terms have been chosen because they all present significant obstacles to translation and they are theologically central concepts for the belief and practice of Islam.

The challenges that these words present are not all the same. In some cases, translators have no alternative but to shatter a word into pieces, deploying seemingly disconnected terms in English to render a single Arabic word. In other cases, there is an obvious choice to translate a term, but that choice misrepresents the original meaning to a considerable extent.

In the first case, an English-speaking student of the Quran might struggle to grasp a concept because of the seeming lack of coherence of its many translational equivalents. In the second case, the same student might innocently yet falsely assume they have understood a word because one consistent translational equivalent has been deployed by the translators.

The task of translation is contradictory. One’s initial intention, elucidating the text in another language, is limited by unsatisfactory approximations which say both more and less than the original text, the translation both subtracting and adding meaning. This difficulty is amplified when dealing with more distant texts. To translate a piece of French journalism into English is to plunge into a veritable sea of exuberance and deficiency (Ortega y Gasset 1959), but this pales into insignificance compared with the difficulties that arise when the Arabic Quran is translated into languages that have been formed by a biblical tradition. Despite much talk over the past 60 years of ‘Abrahamic religions’ (Hughes 2012), the biblical and quranic traditions are more different than they might at first appear, and key concepts often do not translate readily from one faith language into another.

One of the great conceptual divides in the world today is the gulf that exists between the cultures and languages that have been shaped by the Quran, and those that have been shaped by the Bible. Viewed through the eyes of a modern Western reader, the faith of Islam is replete with foundational texts that are conceptually distant, being set far apart in time, place and conceptual worldview from Europeans, whose native languages have been shaped by a biblical tradition. For someone who comes to the study of Islam from the vantage point of a European cultural background, to understand the core concepts of Islamic texts requires patience and skilful vigilance.

1.3. Islam, Europe and Words

In 2001, when the September 11 atrocity took place, the eminent scholar of Islam Bernard Lewis was working on the page proofs of his latest book. In What Went Wrong: Western Impact and Middle Eastern Response, Lewis was tracing the crisis of the Muslim world’s engagement with the West and with modernity. This engagement had made some Muslims aware that all was not well in the House of Islam. In What Went Wrong, Lewis makes many wise observations about differences in the conceptual worlds of Islam and of Europe.

The differences Lewis points out in this handy book are embodied in the meanings of words. For example, he observed that “in Middle-Eastern usages, liberty or freedom was a legal not a political term. It means one who was not a slave, and unlike the West, Muslims did not use slavery and freedom as political metaphors” (2002: 54). In another example, Lewis commented that “secularism in the modern political meaning … is, in a profound sense, Christian” (2002: 96). In support of this statement he observed that “in the course of the centuries, Christian jurists and theologians devised or adapted pairs of terms to denote this dichotomy of jurisdiction: sacred and profane, spiritual and temporal, religious and secular, ecclesiastical and lay” (2002: 98). (One could add church and state to this list). In contrast, as Lewis reports, the Arabic of medieval Islam had no comparable terminologies: it simply lacked the lexicon to make any clear separation between the religious and the secular.

Lewis can be faulted over the way he makes his points. He says that “in Middle-Eastern usages, liberty or freedom was a legal not a political term”, but it will not do to treat freedom and liberty as meaning the same thing (Wierzbicka 1997: 25ff). It is an even worse error to describe freedom-or-liberty as a “legal term” in “Middle-Eastern usages”, for this affords a universal status to English which it does not merit. The English words freedom and liberty are not universal concepts to be realised in different languages. As Anna Wierzbicka (1997: 138ff) has shown, even the closest translational equivalents in other European languages of the English word freedom have distinct meanings, which reflect the differing cultural histories of the speakers of those languages. It would have been more precise if Lewis had said that the usual suspects for translating freedom into the major Middle Eastern languages of Islam – Arabic ḥuriyah, Persian āzādi and Turkish özgürlük – are legal and not political terms.²

Setting such niceties aside, Lewis’s point is nevertheless both insightful and sound, that the lexicalised concepts of Europe and of the Islamic world have been deeply shaped by the respective religious traditions of Christianity and Islam, and differ considerably as a result. Wierzbicka (2019: 295–299) has argued that English love, German lieben, French aimer and Russian ljubit’ have all derived their meanings from Greek agapao, as used in the New Testament, and ultimately from the Hebrew verb ‘āhēb. This concept of love, she observes, is peculiar to cultures influenced by Christianity. In contrast, in recent years I have been serving as a pastor to Iranian converts from Islam to Christianity, and have indeed found that there is no single Farsi word that can be used to translate agapao of the Bible. This means that to preach to Iranians about the biblical concept of the love of God requires careful cross-cultural semantic reflection.

Before we explore the meaning of our handful of quranic Arabic terms, we first need to lay a foundation of some fundamental Islamic theological concepts.

2. Foundations of Islamic Theology

The message of the Quran is grounded in certain beliefs about humanity, Allah, and the relationship between them. According to the Quran, Allah created the world and as creator he stands in relation to human beings as a master to slaves. This type of relationship was familiar to the community within which the Quran was first recited.

In exercising his mastery, Allah commands people to perform certain actions and to refrain from others. However, the human capacity to follow Allah’s directives is imperfect, for, as the Quran explains, “the human was created weak” (wa‑khuliqa l‑insānu ḍa‘īfan, Q4:28;³ Q30:54) and can easily “go astray” (ḍalla and aḍalla).

To address this fundamental human deficit, Allah provides means of correction. These means are referred to as hudā, a nominal from the verb hadā «guide» (root h-d-y).⁴ Hudā is normally translated in English as ‘guidance’. Because of the centrality of the concept of hudā in Islam and in the Quran, we will briefly clarify its meaning before moving on to discuss our key terms.

The English verb guide is polysemous. It has certain distinct but related meanings in reference to physical movement. One is that someone accompanies another on a journey. As the OED puts it, to guide is “to go with or before for the purpose of leading the way.”⁵ The one who does not know the way can be assisted by a guide, who accompanies and shows the way. A guide is not in a position of superiority or command over the person guided, and neither do they take responsibility for transporting the person: the person being guided moves freely of their own volition, assisted by the presence of the guide.

A second, closely related meaning of English guide is when God, providence or some other higher power or point of reference is the agent. In this case, the higher power is not said to be moving with the person, but it is as if this higher power were present with the person. Thus someone can speak about being “guided by God” or “guided by the stars”.

The Quran’s concept of hudā is different from English guidance because hudā does not imply the presence of an accompanying guide.⁶ Rather, directions are given by someone who tells another which way to go, but without accompanying them. Thus the concept of hudā is about giving directions, not accompanying guidance. In a rock inscription written in Safaitic, an Arabic dialect which predated the Quran, the word hdy appears in reference to a military commander, who is someone who gives directions to others (Al-Jallad 2015: 317).

In the worldview of the Quran, the alternative to being rightly directed is the ignorance of not knowing the way. According to the Quran, someone in this situation will “go astray”. It is bad to go astray: someone can perish in the wilderness when they wander off the track.

Consistent with the concept of the walk of faith as a journey, the Quran includes repeated references to al-ṣirāṭ al-mustaqīm ‘the straight road’. Although ṣirāt was borrowed from Roman strata ‘paved road’ via Greek and Aramaic (Jeffery 1938: 195–196), the metaphor of finding and staying on the ṣirāṭ al-mustaqīm is not about following a highway, a made road, or even a well-beaten track. The way of Islam is not so obvious that people cannot easily stray off it. Indeed, a great deal of effort is devoted in the Quran to preventing people from straying.

According to the Quran, in order to help people to avoid straying, the right way is pointed out to human beings by ʾāyāt bayyināt ‘clear signs’ (Q2:99). These are provided by Allah, but they need to be recognised for what they are and acted upon. These signs include natural features such as the sun and the rain (Q10:5, 24), as well as stories and lessons learned from people who have lived in the past, general observations about life, and even verses of the Quran itself (Q2:185; Q10:15).

3. Data and Methodology

Here we explore the meanings of four quranic Arabic words: rasūl, shirk, kāfir and ittaqā.

These terms play a central role both in the Quran and in Islamic thought, but none of them has a straightforward English translation.

For each Arabic word considered here there is an established tradition of English translation. I will call words or phrases conventionally, but not necessarily accurately, used in such translations ‘Conventional Translational Equivalents’ or CTEs. For example, the Arabic word mushrik has an actual meaning (roughly speaking) of ‘someone who wrongly attributes shared power over something to Allah⁷ and another’. ‘Polytheist’ is often used to translate mushrik, as well as ‘idolater’, ‘pagan’, ‘associator’ and ‘unbeliever’, but in reality, there is no English word that even approximates this meaning.

Here I use double angled brackets to signal a CTE; for example, mushrik «polytheist».

The Arabic words we will consider here, together with the number of occurrences in the Quran, are:

1. rasūl «messenger» (322 occurrences) and nabī «prophet» (75 occurrences)
2. shirk «polytheism» (5 occurrences) and related terms based on the root sh-r-k: ashraka «to associate» (71 occurrences) and mushrik «polytheist» (33 occurrences)
3. kāfir «disbeliever» (156 occurrences) and related terms based on the root k-f-r: kufr «disbelief» (37 occurrences) and kafara «to disbelieve» (289 occurrences)
4. ittaqā «to guard oneself» (166 occurrences)

In researching this study I have considered all instances of these forms found in the Quran.

The method of semantic analysis used here is Natural Semantic Metalanguage or NSM (Goddard & Wierzbicka 2014, Wierzbicka 2021). This method of semantic analysis uses semantic decomposition, deploying 65 semantic primes, which are postulated to exist in the lexicon of every human language, and to be sufficient for the semantic explication of all linguistic meanings. In addition, NSM postulates a universal syntax for the primes. Tables of NSM primes have been drawn up for many languages (Goddard & Wierzbicka 2014: 12).

NSM has proven extremely useful for explicating culture-specific meanings that are difficult to translate (Wierzbicka 1992, 1997, 1999, 2014, Goddard & Wierzbicka 2014, Levisen & Waters 2017, Bromhead & Ye 2020), including theological meanings (Wierzbicka 2001, 2019, 2020, Habib 2011, 2017). My purpose in presenting this research is to further test NSM by using it to explicate a key group of quranic Arabic words which are core concepts of Islam and notoriously difficult to translate.

The need to use a diversity of English words to translate certain core quranic Arabic concepts has prompted some scholars of the Quran to discern complex polysemies at every turn. As Hughes (2022: 39) puts it, “these terms can adopt a rather large range of meanings in different contexts”. This is the approach taken by Juan Cole (2020) in a recent study of quranic Arabic kafara. However, the fact that a word in one language requires a bewildering variety of contextual translations in another language can by no means be relied upon as evidence of polysemy. On the contrary, I shall argue that the difficult-to-translate quranic words considered here can be given unitary NSM explications, and the need to deploy a variety of English translations reflects the fact that parts of a complex concept, which can be easier to translate than the whole concept, come into focus in different contexts.

A key point is that it will not do to attempt to provide semantic explications of complex concepts using other equally complex concepts. Complex meanings must be decomposed using words with simpler meanings. As we shall see, the use of NSM primes allows one to cut the Gordian knot of the supposed “large range of meanings” of these theological terms, to produce compelling semantic analyses which account well for the textual data.

4. Results

4.1. rasūl

This now brings us to our first key quranic word, rasūl (pl. rusul), conventionally translated «messenger», but sometimes «apostle».

In classical Arabic, the root r-s-l is used with various derivatives related to sending a message, including the form IV verb arsala ‘send someone with a message’ and the form III verb rāsala ‘to correspond or exchange messages’. The word rsl is also attested in Sabaean (Old South Arabian) inscriptions referring to royal emissaries (Biella 1982: 490), and mursal, a nominal based on the same root, is used in the Quran with the meaning ‘ambassador’ (Q27:35), that is, someone sent by a ruler to deliver a message.

In Islamic theology, rasūl has a precise religious meaning. The Quran recites multiple stories of messengers from the past, whose biographies follow a standard pattern, summarised by David Marshall as follows:

…the messenger will typically criticize his people for not worshipping God alone, and perhaps for certain moral failings as well. However, he is rejected by most of his contemporaries, although he does have some obedient followers. The messenger also warns his people that, if they do not repent, they will suffer a great punishment from God. The story ends with a dramatic act of divine intervention: the unbelievers, as warned, are destroyed by God in a variety of ways… The completeness of the destruction of the unbelievers is often emphasized. The messenger and his followers are saved and vindicated. (Marshall 2014: viii-ix)

An example of a rasūl in this mould is Mūsā (Moses), who was sent, according to the Quran, to the Egyptians (Q7:103–105) to warn them of impeding destruction. When they refused to heed the warnings, they were destroyed and Moses was rescued.

As Marshall explains, the Quran repeatedly insists on the uniformity of the biographies of the rusul.⁸ Although there are several recurring elements in the accounts of every rasūl (for example, that they are mocked, that the people reject their message and that Allah rescues his rasūl), not all of these are defining characteristics of the office.⁹ This can be seen from the Quran’s account of the rasūl Yūnus (Jonah), who the Quran states was the only rasūl whose people heeded his warning and repented (Q10:98). This exceptional outcome does not make Yūnus any less a rasūl in the Quran’s eyes, which is evidence that the usual negative response of the people is not part of the essential meaning of rasūl.

Here is a proposed semantic explication of the meaning of rasūl:

rasūl

1. someone,¹⁰ not like many other people
2. people can say what this kind of someone is with the word rasūl
3. people can think like this about this someone:
4. “this someone is very good
5. this someone does what Allah wants
6. this someone wants other people to do what Allah wants
7. Allah says something to this someone because he wants this someone to say it to this someone’s people
8. Allah always says something like this to someone of this kind
9. Allah says to this someone:
10. “it is good if you say this something to the people”
11. this someone says this something to the people after this
12. this someone says to the people:
13. “Allah says this to you because he wants you to do something
14. it is good if you do what Allah wants
15. if you do not do what Allah wants, Allah will do something very bad to you”
16. this someone wants the people to think: “Allah says this to us” ”

Each of the elements in this explication is emphasised repeatedly in the Quran, including that a rasūl is sent to his own people, that rusul are perfect examples of faithful obedience,¹¹ and that past rusul all brought the same message. This last point is stressed, for example, in the following verse:

(1) This was our way with the messengers we sent before you: you will find no change in our ways. (Q17:77)

To call someone a rasūl is to attribute all the elements of this explication to this person. Thus if all someone did was bring messages from Allah to other individuals as personal, private revelations, this would not justify calling this messenger from Allah a rasūl in the quranic sense. To merit being called a quranic rasūl, someone must be sent by Allah to a people with a warning of future punishment and a call to repent.

The function of the elaborate religious meaning of rasūl is that the stories of past messengers are used in the Quran to establish a template which validates the mission of the quranic Messenger, named four times in the Quran as Muhammad. The semantic elements included in the meaning of rasūl function to validate the structure of Muhammad’s own mission.

The Quran also uses a related term, nabī «prophet» (root n-b-y). Scholars have pondered how and whether nabī is different from rasūl. Jeffery (1950: 115) concluded that Muhammad “made no special distinction between the two names rasūl and nabī ”. Wansbrough (1977: 54) came to the same view: “rigorous and consistent distinction between the designations nabī and rasūl is not justified by quranic usage”.

In essence, a quranic nabī is a particular flavour of rasūl. In the Quran, the two characteristic features of a nabī, in contrast to the rasūl, are first, that the nabī can be a hereditary office – “some of them are descendants of others” (Q3:34) – and second, that this hereditary office was granted to the “sons of Israel” (Q5:20; Q29:27). However, the Quran reports that this office has been taken away from the Jews, who “disbelieve in it”, and given to the Arabs, because they are “a people who do not disbelieve in it” (Q6:89).

It is the orthodox belief of most Muslims that Muhammad was not only the rasūl of the Quran; he was also the last nabī, the final holder of the office of nabī. The Quran discounts any further continuation of a hereditary prophetic office when it declares that “Muhammad is not the father of any of your men but he is Messenger of Allah and the seal of the prophets” (Q33:40).

Let us now contrast the explication of the quranic rasūl with that of the biblical Hebrew nābī’, bearing in mind that the quranic nabī differs only slightly in meaning from rasūl.

Wierzbicka has given a description of what a biblical prophet is in What Christians Believe. For our purposes, the key component in Wierzbicka’s explication is the following:

When God wanted to say something to the people of Israel,

God said it to someone not like many other people,

after this, this someone said it to the people of Israel.

Someone like this was called a prophet.

Many prophets spoke to the people of Israel¹² at many times (some were women).

They spoke not like other people. When they said something, they wanted people to think: “God says this to us.”¹³ (Wierzbicka 2019: 80)

This semantic explication needs further refinement.

In the Hebrew Bible, the verb nābā’ ‘prophesy’ (root n-b-’) is the basic concept upon which nābī’ ‘prophet’ is built. This can be seen, for example, in the fact that people can prophesy without being prophets. King Saul, who was not a prophet, prophesied on two occasions (1 Samuel 10:11, 19:23–24); hence the proverb, “Is Saul also among the prophets?”, which is given as commentary on both of these incidents. The intended answer is “No, Saul is not a prophet.” On the second occasion, three groups of messengers, who had been sent by Saul to take David, also prophesied, likewise without being considered prophets (1 Samuel 19:20–21). In an earlier incident, seventy elders of Israel prophesied once only at their commissioning under Moses, inspired by the Holy Spirit (Numbers 11:25–29), yet this did not make them prophets either, for “they did so no more” (Numbers 11:25). Another example is when David, who is not referred to in the Bible as a prophet, sang a prophetic song (2 Samuel 23:2–7).

Here is a proposed NSM explication of Hebrew nābā’ ‘prophesy’, which develops Wierzbicka’s Minimal English account:

X nābā’ ‘X prophesied’

1. Someone (X) said something to someone else at that time
2. this someone (X) said it like someone can say something like this to someone else when someone thinks like this:
3. “God said something to me
4. God wants me to say it to someone else after this
5. God wants that someone else to think like this:
6. “God is saying this to me” ”

Components (b–f) of this explication capture the idea that this speech act has a conventional character in which God speaks to someone wanting them to pass it on to another person so that the other person can know God is saying this to them.

In Exodus 7:1, Aaron is described as Moses’ “prophet” to Pharoah: “See, I have made you like God to Pharoah, and your brother Aaron will be your prophet.” Earlier, in Exodus 4:15–16, it is said that Aaron was to function as Moses’ “mouth”, and Moses was to “serve as God” to Aaron. This means that when Aaron was acting as Moses’ mouthpiece to Pharoah, people should think this:

Moses said something to Aaron

Moses wanted Aaron to say it to Pharoah after this

Moses wanted Pharoah to think like this:

“Moses is saying this to me”

Many biblical passages depict prophesying as a two-stage process, as described in the semantic explication. The first stage, when God speaks to the one prophesying, is often described as “the word of the LORD came to X”, where X is the prophet’s name (e.g. 1 Samuel 15:10). In Ezekiel 3:1–4, the reception of the message is enacted graphically through a vision in which the prophet eats a divine scroll, after which he is instructed to take these consumed words to heart and then go and repeat them to Israel: “Mortal, all my words that I shall speak to you receive in your heart and hear with your ears; then go to the exiles, to your people, and speak to them.” (Ezekiel 3:10–11).

The semantic explication proposed here of nābā’ implies that a false prophecy can still be called an act of prophesying, as indeed happens in Jeremiah 23:21: “I did not send the prophets, yet they ran; I did not speak to them, yet they prophesied.” It is consistent with the explication of nābā’ that someone could prophesy lies, without actually thinking that “God said something to me”. This appears to be the case in Zechariah 13:3: “If anyone continues to prophesy, his own father and mother will tell him, ‘You must die, for you have prophesied lies in the name of the Lord.’” The way the explication is worded, it suffices that the one prophesying is doing this in the manner of someone who thinks God has said something to them.

Note also the comparative simplicity of the meaning of biblical Hebrew nābā’, which does not require that the prophesying be directed to the nation of Israel: it could be directed to individuals, as often happens in the Bible (e.g. the prophecy of the old prophet of Bethel in 1 Kings 13:20–22). Moreover, the prophetic message need not include a warning: there is no restriction on its contents.¹⁴

Here is a proposed explication of the biblical Hebrew nābī’ ‘prophet’:

nābī’

1. someone, not like many other people
2. people can say what this kind of someone is with the word nābī’
3. people can think like this about this someone:
4. “God can say something to this someone
5. God does it because he wants this someone to say it to someone else after this
6. when this someone says it to that someone else God wants that someone to think like this:
7. “God is saying this to me” ”

As we have seen, the meaning of prophesy allows for the possibility that someone could prophesy without being considered a prophet, as well as for an act of prophesying to be insincere or false. What qualifies someone as a prophet is whether people think that the person can prophesy: this is a socially recognised role. If someone like Saul, who was not thought of in that way, did prophesy on occasion, this was insufficient in itself to justify speaking of him as someone who could prophesy: an isolated act of prophesying does not make someone a prophet.

The explications I have proposed here differ from Wierzbicka’s characterisation in that she has the prophets addressing the people of Israel, and sets this in the context of Israel’s overall salvation history. While this is a valid attribute of the New Testament concept of hoi prophetai ‘the prophets’ (e.g. Matthew 2:23; Acts 3:25; 1 Peter 1:10), it is not a necessary part of the meaning of the Hebrew nābī’ as it is used in much of the Hebrew Bible.

It should be apparent that there are major differences between the biblical nābī’ and the quranic rasūl / nabī. Whereas in biblical Hebrew there is a distinct speech act of prophecy, there is no such speech act for the quranic rasūl: there is no Arabic verb meaning ‘to prophesy’. The phrase al-balāgh ‘the reaching, the attaining’ can refer to the delivery of a message from Allah by a rasūl; the verb talā ‘recite’ can refer to delivery of verses from the Quran; the verb arsala ‘send’ can refer to Allah’s act of sending a messenger or signs to a people; and the verb awḥa ‘suggest, inspire’ can refer to the process in which verses are send down to the rasūl. However, no Arabic verb is attested in the Quran that describes the whole prophetic process of a rasūl receiving and then delivering a message from Allah to others.

There are other differences. Biblical prophecy knows no limitations on the kind of message brought, but the quranic prophet brings an unchanging message that warns of imminent punishment. The biblical prophet can bring a message for an individual, group or nation, but the quranic prophet is always sent to a community, such as a tribe or a town. Biblical prophecy takes place in the context of a long history of God’s communications with the people of Israel, in which each individual prophet contributes to a conversation spanning centuries – this is reflected in Wierzbicka’s explication of the role of the prophet – but the quranic rasūl is sent to their own people to give them a unique, one-time opportunity to repent before they are destroyed. In the Bible, there are no examples of a hereditary biblical prophetic office, unlike the other two anointed offices of priest and king which are hereditary, but the quranic nabī can be a hereditary office. The quranic rasūl is righteous by definition, but in the Hebrew Bible a nābī’ is not necessarily attributed with righteousness.

These many differences point to difficulties for translation. The question arises: Is it even appropriate to use a biblical term, such as prophet or apostle, for the quite distinct quranic office of rasūl (or its variant, the nabī)? In the case of rasūl, the majority solution, adopted both by scholars and by English-speaking Muslims, is to use ‘messenger’ as a calque for rasūl and ‘prophet’ as a calque for nabī. Thus Muhammad’s title, al-rasūl Allah, is customarily rendered in English as “the Messenger of God”. A less frequent translation is “the Apostle of God”, no doubt chosen because the Greek apostolos is derived from apostello ‘send’, which is similar to the core meaning of the root r-s-l. However, the New Testament Greek apostolos differs even more in meaning from rasūl than does Greek prophetes.¹⁵

In the discussion of rasūl, we have considered a concept that has some degree of similarity to a biblical concept, and a stable CTE of «messenger», but this translation is inaccurate, because a quranic rasūl is much more than just a messenger. At the same time, the quranic nabī, although translated as «prophet», is very different from the biblical prophet, having a meaning similar to and based on that of rasūl.

These translation difficulties are compounded by the fact that the Quran considers the faith it preaches to be the true Judaism and the true Christianity, stating that Abraham was neither a Christian nor a Jew but a Muslim (Q3:67), so its concept of a rasūl is believed to be original and authentic to both Christianity and Judaism.

Now we shall consider two terms that present a different kind of challenge for translation.

4.2. shirk

The Quran uses two primary lexical concepts to refer to someone who is not a Muslim. For each concept there is an abstract noun, a verb, and an agent nominal:

We will be proposing NSM analyses of the abstract noun shirk and the plural of kāfir.

The concepts of shirk and kufr are distinct but complementary, referring to different dimensions of disbelief. In a nutshell, a person who is a kāfir rejects Allah, his signs and his messengers while concealing or denying the truth, while a person who is a mushrik transgressively claims that another being shares in Allah’s unique powers and prerogatives.

Derivatives of the Arabic root sh-r-k are grounded in a core meaning of shared ownership. A co-owner of an animal or a slave is a sharīk. The form III verb shārika, which does not occur in the Quran, means ‘he is a co-owner of something with someone’ and the causative form IV verb ashraka means ‘he makes or treats someone as a co-owner’. As is the case for some other form IV verbs, for ashraka the causative function includes declaring someone a co-owner.¹⁶

In the Quran, the noun shirk refers either to a share or part ownership of something (e.g. Q35:40, “Do they have a share (shirk) in the heavens?”) or to the sin of attributing “partners” to Allah (e.g. Q31:13, “O my son, do not attribute partners (ashraka) to Allah, for shirk is the worst of wickedness”).

Shirk is the “foremost religious crime in Islam” (Böwering 2002: 329), and is considered a gross, uniquely unforgivable sin. In the understanding of the Quran, it is a terrible transgression to attribute “partners” to Allah. Q4:48 (repeated word-for-word in Q4:116) states that Allah can forgive any sin except attributing partners [ashraka] to him.

The term mushrik is used frequently in the Quran for someone who calls someone an associate or ‘co-owner’ with Allah. Although the root meaning of sh-r-k is co-ownership, this is broadened to include the attribution of the powers and prerogatives of Allah to other beings, in particular as pertains to Allah’s benevolence. Thus a mushrik is someone who “in his behavior and attitudes … proceeds as if other beings, supernatural or perhaps sometimes human, have powers which a true monotheist would recognize as belonging to God alone” (Hawting 2002: 477). An example is someone who calls upon someone other than Allah for help of a kind that only Allah can provide.

Underlying the concept of shirk is the conviction that, as the sole creator, Allah is the only ‘owner’ and disposer of human beings. That it is a logical impossibility for Allah to enter into any kind of partnership over his creation is explained by means of an analogy with the plight of a slave owned by two masters, which compares unfavourably to the state of a slave owned by just one master:

(2) Allah presents a parable: a man was owned by several quarrelling partners (shurakā’u) and a man belongs exclusively to one man. Are the two equal? (Q39:29)

This analogy takes it to be obvious that the condition of a co-owned slave is greatly to be pitied. Such a slave will be pulled between two masters quarrelling over the slave. The slave, unable to meet the conflicting commands of the two masters, can please neither of them, for “he would be confused as to whom of them he should serve” (al-Maḥallī & al-Suyūṭī 2007, commentary on Q39:29). For humans to imagine that they serve several gods when there is in fact only one God, Allah, would be calamitous indeed for them. For Allah’s part, the Quran goes on to declare that those who make such a claim are “telling a lie about Allah” and for these a special place in hell has been preserved (Q39:32). In another parable, the Quran states that for Allah to have a sharīk would be like a slave owner co-owning wealth with his slave (Q30:28). Just as no slave owner would tolerate this, neither will Allah.

Aversion to co-ownership of creation is spelled out in another passage, where it is claimed that if there were more than one creator god, each would have tried to dominate the other, fighting the other over creation (Q23:91). For this reason, Q21:22 concludes that the creation would have been wrecked if there had been more than one creator god (Mir 2004: 161).

This argument against polytheism appears to be original to Islam: such arguments are not found in the Bible. When Jesus tells a parable of a slave with two masters (Luke 16:13, Matthew 6:24), he uses the analogy to warn that a slave with two masters will have divided loyalties, and a person must choose whom they will serve. This is an argument for the exercise of freedom of choice by human beings: as Joshua said to Israel, “choose for yourselves this day whom you will serve” (Joshua 24:15). In contrast, the Quran uses the analogy of a slave with two masters to make a strikingly different point, namely that, just as it is a terrible thing to have two masters because they would compete with each other and make life a torment for the slave, so it would be a disaster for humanity if Allah had co-owners of creation.

The Quran’s polemic against shirk focuses on the impossibility that another being could exert the power of Allah. It asserts that no-one but Allah could raise the dead (Q21:21); no-one has the right to question Allah about anything he does (Q21:23); those whom some call ‘sons’ of Allah are but Allah’s slaves, who can only speak or act by his command (Q21:27); and no other ‘gods’ can defend their servants from Allah, who can do what he wants with such ‘gods’ (Q21:43), including sending them to hell (Q21:29). The command to serve or worship Allah alone (Q21:25) derives logically from these considerations, since someone should only serve another if they have power to do with you what they want.

It seems also that the concept of shirk is invoked in the Quran whenever a beneficial power is attributed to another, of a kind which only Allah can exert. To seek help from Allah alone protects against committing shirk:

(3) Say: “I call only upon my Lord, and I do not attribute partners (ushriku) to him.” (Q72:20)

Shirk is thus an error of attribution, saying that another has a beneficent power to command creation which in reality only Allah has (Q18:26; Q30:40). Various verses describe shirk in terms of saying falsehoods about Allah:

(4) Say: “My Lord has only forbidden … that you attributed partners (tushrikū) to Allah without his authority, and you said things about Allah of which you had no knowledge.” (Q7:33)

Shirk can manifest in a variety of different ways, such as misdirected prayer or worship, verbal confession, or reliance on some other ‘god’ for aid. The Quran calls it shirk when someone looks to anyone, whether human or a spirit, to do something for them which only Allah can do. An example is taking someone other than Allah as a patron or protector:

(5) Shall I take for my patron (walī) any other than Allah, the maker of the heavens and the earth? … Say, “No! I am commanded to be the first to bow to Allah. Do not be one of the associators (mushrikīna).” (Q6:14)

Another example of shirk is a confession of faith that conflicts with Allah’s prerogatives; for example, saying that Jesus Christ is divine (Q5:72) or talking about Allah as having sons or daughters. On judgement day, Allah will say to all who do this, “Where are the partners you talked about?” (Q6:22) The point of this statement is that these alleged “partners” will be powerless to aid their devotees at the Last Day, and thus prove not to be partners of Allah in his rule. Another example of shirk is giving thanks to someone else besides Allah for the birth of a child (Q7:190), which is wrong because only Allah can give life.

Although some translators render shirk as «idolatry», the word ṭāghūt ‘idol’ is never mentioned in the Quran in the same verse as an instance of the root sh-r-k. Where the concept of sh-r-k is invoked is in contexts when the Quran emphasises that there is only one creator (Q35:40) ruling over the creation, and particularly when the focus is on Allah’s beneficial mastery over everything. In such contexts, the Quran repeatedly reminds us that those who look to others besides Allah for aid will find their alleged protectors to be powerless before Allah (Q10:28; Q28:64), for “he has no partners in his rule” (Q18:26). Thus the focus in contexts where shirk is invoked is not on idolatry per se, but on the exclusivity of Allah’s rule and the imperative for human beings to rely only on the benevolence of Allah.

The being to whom Allah’s attributes are falsely attributed in an act of shirk can be an angel or an (imagined) god, but it can also be human, as reflected in the following verse naming priests, monks and Jesus (understood by the Quran to be no more than a human being) as objects of ‘association’:

(6) They take their priests and their monks as their lords to the exclusion of Allah, and Christ the son of Mary; yet they were commanded to worship but one god: there is no god but he. Praise and glory to him. He is glorified above what they associate (yushrikūna). (Q9:31)

Finally, we note that in the Quran, shirk is often said to lead to divine punishment:

(7) We will cast terror into the hearts of those who disbelieved (al-ladhīna kafarū), because they attribute partners (ashrakū) to Allah, for which he has sent down no authority. Their refuge is the Fire: evil is the lodging of evildoers. (Q3:151)

In the light of all these considerations, it seems unnecessary to define shirk in terms of ownership: it is enough to focus on the unique power of Allah as benevolent master over his creation. The offense of attributing ‘associates’ to Allah consists in looking to other beings for help that only Allah can provide. This is bad for the person, because such help is illusory, and it denies Allah his due.

Here is a proposed semantic explication of shirk, which gathers these insights together. In this explication the container for the explication (lines a–d) follows the semantic analysis of abstract nouns in Goddard and Wierzbicka (2014: 205–237).

shirk

1. something
2. people can say what this something is with the word shirk
3. someone can say something about something with this word when
          someone thinks like this:
4. “it can be like this:
5. someone says about someone else:
6.                 “this someone can do something very good for me”
7. at the same time, people can know that it is like this:
8. Allah can do something like this for people, no-one else can
9. this someone is not Allah
10. it is very bad if it is like this
11. people can know that if someone says something like this, after some time Allah will do something very, very bad to them because of it”

4.3. kuffār / kāfirūn

We will now consider the concept of kufr, proposing an explication of the plural of kāfir «disbeliever». This noun has two main plural forms, kuffār and kāfirūn, which are identical in meaning.¹⁷

The concept of kufr is even more frequently invoked in the Quran than shirk. As Toshihiko Izutsu has explained in Ethico-Religious Concepts in the Qurʾān:

“Even a cursory reading of the Scripture [i.e. the Quran] will convince one that the role played by the concept of kufr is so peculiarly influential that it makes its presence felt well-nigh everywhere in sentences about human conduct or character.” (Izutsu 1966: 119)

In its non-religious sense, the verb kafara means ‘cover, conceal’ (Adang 2001: 220); for example, clouds covering the sky or the earth covering seed.¹⁸ While we are interested here in derivatives with k-f-r that refer to disbelief and rejection of faith, there are k-f-r forms in the Quran that do not have this meaning. The form II verb kaffara means ‘to grant a pardon’ (i.e. to cover over someone’s transgression); in Q57:20, kuffār is used with the sense ‘sower’ (someone who covers over seed with earth); the noun kaffārat is used with the meaning ‘atonement’ in the fifth Sura (Q5:45, 89, 95); and in the following verse the noun kufrān,¹⁹ a hapax legomenon in the Quran, is used in the context of affirming that a person’s good deeds will not be omitted from their record:

(8) Whoever does righteous deeds, as a believer, (there will be) no denying (kufrān) his striving (by Allah). We shall record it to his account. (Q21:94)

Kāfir as a non-religious term can also mean ‘ingrate’. In the following example, Pharoah rebukes Moses, presumably for killing an Egyptian (Exodus 2:11–12). Moses’ crime is not specified in the Quran. The point of the rebuke is that Moses should have been grateful to the Egyptians for raising him:

(9) He (Pharoah) said (to Moses): “Didn’t we bring you up among us, and weren’t you among us for some years of your life? Yet you did what you did, and are one of the ungrateful (al-kāfirīna²⁰).” (Q26:18–19)

Setting aside these non-religious meanings, we will focus on kafara «disbelieve», kāfir «disbeliever» (pl. kuffār / kāfirūn), and kufr «disbelief», as used to characterise disbelief in and rejection of the message of Islam. For the purposes of discussion here, and in the absence of any convincing evidence to the contrary, I will treat these three forms as semantically equivalent. Thus a kāfir is someone who does kafara, and kufr refers to the act of committing kafara or being a kāfir.

In the Quran, the verb kafara is often used intransitively, but it can also take a second argument introduced by the preposition bi- (e.g. ‘they rejected (kafarū bi-) Allah and the Messenger’ (Q9:54)). As a religious term used for rejection of the Quran’s message, kafara receives a diversity of renderings in the English translations, including ‘disbelieve’, ‘be an unbeliever’, ‘misbelieve’, ‘be an infidel’, ‘be ungrateful’, ‘deny’, ‘knowingly reject the truth’, and ‘be without faith’.

One of the most salient aspects of the religious meaning of k-f-r is ingratitude: indeed, it was this meaning that the classical Muslim lexicographers considered to be kufr’s fundamental meaning (Adang 2001: 221). For example, in this next verse the people of the town are criticised for being ungrateful for Allah’s good deeds which have benefited them:²¹

(10) Allah tells a parable: a town was secure and at rest, with abundant provision coming to it from every side, but it was ungrateful (kafarat bi-) for Allah’s favours, so Allah clothed it with hunger and fear for what they had been doing. (Q16:112)

The Quran speaks of kufr as the opposite of thankfulness:

(11) If you remember me, I will remember you. Be thankful to me (to Allah), and do not be ungrateful to me (takfurūnī). (Q2:152)

The Quran also often refers to a human tendency to ingratitude, and contrasts this with Allah’s many acts of benevolence to humankind, which ought to evoke gratitude:

(12) It is Allah who created the heavens and the earth, and sends down water from the heavens, and brings forth fruits to provide for you. And he has subjected ships to you, to pass through the seas by His command, and he has subjected the rivers to you. And he has subjected the sun and moon to you, both constant (in their courses), and he has subjected night and day to you. He has given you some of all that you have asked for. If you (attempt to) count Allah’s favours (to you), you will not be able to. Surely humans are unjust and ungrateful (kaffārun – emphatic form of kāfir). (Q14:32–34)

Such instances of Allah’s goodness are included among the “signs” of Allah, which those who commit kafara call a lie:

(13) But those who are ungrateful (kafarū) and call our signs a lie shall be companions of the Fire. They will stay there in it. (Q2:39)

The concept of kufr is not only about ungrateful denial: it also conveys the idea of disbelief.

In several passages, forms with k-f-r and ’-m-n ‘believe’ are contrasted. For example, in the following verse the kāfirūna are contrasted with the mu’minūna ‘believers’:

(14) Oh you who believe! If any of you turn back from his faith, Allah will raise up a people whom he will love as they love him: (acting) humbly towards believers (mu’minīna), and powerful against disbelievers (kāfirīna), fighting in the way of Allah … (Q5:54)

The following passage also contrasts kufr and belief, and associates kufr with rejecting Allah’s signs and preferring a different way:

(15) Do those who disbelieve (kafarū) not see that the heavens and the earth were joined together, and we separated them, and made all living things from water? Will they not believe (yu’minūna)? And we set firm mountains on the earth, so that it does not shake, and we placed passes in it (between the mountains) as pathways, so that they might be guided. And we established the sky as a guarded roof. Yet they turn away from its signs. (Q21:30–32)

The idea of rejection – that the person who practices kufr knows full well the good things Allah has done, and not only dislikes this, but rejects this knowledge and what it implies – is apparent in the following verse in which the ‘People of the Book’ (Jews and Christians) are said to deliberately conceal the truth:

(16) You People of the Book, why reject (takfurūna) the signs of Allah, of which you are witnesses? You People of the Book, why do you mix truth with falsehood, and hide the truth, while you know (what the truth is)? (Q3:70–71)

The rejection can be of a teaching, such as that of a future resurrection, in which case kafara could be translated as ‘disbelieve’. Note also in the following verse the threat of punishment for those who practise kufr:

(17) If you are astonished (at their lack of faith), it is astonishing that they say, “When we have turned to dust, shall we really be in a new creation?” Those are the ones who have disbelieved (kafarū) in their Lord, and those – iron chains will be on their necks – those are the companions of the Fire, where they will remain. (Q13:5)

Another aspect of kufr is rejection of Allah’s messengers and their unvarying message of the unity of Allah:

(18) They are amazed that a warner has come to them from among themselves. The unbelievers (kāfirūna) say “This is a sorcerer, a liar. What! Has he made all the gods into one Allah? That is amazing!” (Q38:4–5)

(19) Praise be to Allah, who created the heavens and the earth, and made darkness and light. Yet those who disbelieve (kafarū) treat others as equal to their Lord. (Q6:1)

In the following verse this rejection of the Messenger is coupled with the accusation of concealing what they had known to be true:

(20) How will Allah guide a people who have disbelieved (kafarū) after they believed and have borne testimony that the Messenger is true and that clear signs have come to them. … Surely those who disbelieve (kafarū) after they believed and then increase in disbelief (kufr) – their repentance will never be accepted. They are ones who have gone astray. (Q3:86, 90)

Allah does not love someone who chooses the path of kufr, rejecting Allah and the Messenger:

(21) Say, “Obey Allah and his Messenger.” But if they turn back, Allah does not love disbelievers (kāfirīna). (Q3:32)

The commission of kufr is associated with future punishment:

(22) The penalty of those who deny (kafarū) their Lord is hell, an evil destination. (Q67:6)

The idea that the k-f-r forms express a range of distinct meanings in reference to rejectors of Islam is commonplace in Islamic studies scholarship. For example, Camilla Adang (2001: 220) distinguishes four distinct meanings for kafara: ‘to ignore or fail to acknowledge’, ‘to reject’ (or ‘spurn’), ‘to be ungrateful’ (or ‘thankless’) and ‘to disbelieve’.

In a recent article, Juan Cole (2020) has argued that kāfir should not be translated as ‘infidel’ or ‘unbeliever’. Instead, he proposes multiple polysemies for k-f-r forms. When kafara is used intransitively, Cole proposes that it has a polysemous range of meanings which he described as ‘fluid’ (2020: 627). The distinct meanings he distinguishes for kafara include ‘be ungrateful’, ‘reject, deny, disbelieve’ (these three glosses are intended to point out a single meaning), ‘worship the gods’, ‘rebel’, ‘be impious’, ‘be morally dissolute, be a libertine’, ‘disobey’, ‘blaspheme’, ‘become apostate’, ‘paganise’, and ‘commit a heresy’. Cole also proposes polysemous meanings for kāfir, which include ‘peasant’, ‘pagan’, ‘rebel’, ‘blasphemer’, and ‘libertine’, while kufr can, he suggests, in addition to ‘disbelief’, also mean ‘inauthenticity’ and ‘bad faith’.

One can agree wholeheartedly with Cole’s conclusion that ‘infidel’ is an inadequate all-purpose translation for kāfir, but without being persuaded by his reasoning.

Cole’s main argument against translating kāfir as ‘infidel’ or ‘unbeliever’ is that those who are called kuffār / kāfirūn in the Quran did believe in a god or gods, so it is wrong to call them unbelievers. For example, he writes, “… the pagans have a religion, but it is simply castigated as a false one, which makes translating kāfir as ‘infidel’ seem odd” (Cole 2020: 619). However, this is just what the English words unbeliever and infidel mean: they do not mean ‘someone who does not believe in a god’, but are insider terms for those who do not believe as the insiders believe. To illustrate, one of the meanings the OED gives for infidel is: “From a Christian point of view: An adherent of a religion opposed to Christianity; esp. a Muslim, a Saracen (the earliest sense in English); also (more rarely), applied to a Jew, or a pagan.”²² Cole seems to be objecting to what these English words actually mean, and wants to use them in accordance with what he believes they ought to mean.

A more serious difficulty in Cole’s approach to lexical semantics is his method of semantic analysis. He uses English as his semantic metalanguage, but offers neither semantic decomposition nor stable definitions of individual meanings. Whenever the context suggests a different English translational equivalent for kafara, Cole takes this as evidence that yet another polysemous meaning has been unearthed. For example, in Q2:102, where devils are said to commit kafarū by diverting people into the practice of magic, Cole rejects Arberry’s translation, ‘disbelieve’, and proposes ‘blaspheme’ instead:

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. (Cole 2020: 624).

Here Cole offers no evidence that kafarū, in context, means ‘blaspheme’, other than his judgement that what the demons did was sacrilegious.

In another example, Cole suggests that kafara in Q9:74 means ‘commit apostasy’. This verse speaks of people who had formerly accepted Islam but later disbelieved (kafarū). One might just as well conclude that kafarū here means ‘make a bad decision’, ‘act unwisely’, ‘be double-minded’, or ‘be losers’.

The other problem with Cole’s semantic approach is the lack of stable, testable definitions. This is apparent in Cole’s discussion of the phrasal expression kafara bi-, which he claims has a single, stable meaning: “this phrasal verb is not polysemous in the Quran for it always means to deny or reject” (Cole 2020: 618). Later, Cole adds ‘disbelieve’, glossing kafara bi- as ‘deny, reject, disbelieve’ (2020: 634).²³ This scattergun approach to semantic analysis veils the meaning of kafara behind a cluster of English words, deny, reject and believe, each of which means something different.

I agree with Cole’s conclusion that kāfir does not mean ‘infidel’ and kafara does not mean ‘disbelieve’. However, my argument is neither that to attribute insider meanings to infidel, unbeliever or disbelieve is illegitimate, nor that kafara, when used in reference to rejectors of quranic faith, is multiply polysemous. (Certainly the non-religious uses of k-f-r forms include polysemies.) Instead, my argument is that the family of k-f-r forms has a complex semantic core which needs to be precisely defined. This precision cannot be achieved by simply listing English translational equivalents.

Contrary to Cole, I propose that the Quran treats the semantic components of the religious uses of k-f-r forms as bound together into one concept, not many. It is not that there are different kinds of kufr, or different kinds of people known
as kāfir – kāfir₁, kāfir₂, kāfir₃, etc. – but that the single state of kufr has different aspects to it, and the one kind of people known as the kuffār / kāfirūn display a range of characteristics, which include not only ignorance, rejection of truth, ingratitude and disbelief, but also wilful disobedience and incurring the wrath of Allah.

In formulating a semantic explication for kuffār / kāfirūn, I have treated it as an insider term. Kuffār / kāfirūn and the other k-f-r terms divide human beings into insiders, who are grateful believers, and outsiders, who are ungrateful disbelievers. This ‘insider’ meaning can be defined in NSM using a ‘people of two kinds’ explication. An explication is provided for the plural form, because the large majority of instances are plural. Here, then, is our proposed semantic explication for kuffār / kāfirūn:

kuffār / kāfirūn

1. people of one kind
2. they are not as Allah wants people to be
3. it is like this:
4. there are people of two kinds
5. people of one kind are as Allah wants people to be
6. people of this kind think like this:
7. “I know that it is like this:
8. Allah does many very good things for people
9. Allah does not have to²⁴ do these things
10. because of this, I want to think very good things about Allah
11. at the same time, people of this kind think like this:
12. “I know that it is like this:
13. Allah says many things to people, these things are true
14. Allah sometimes says these things to some people not like other people
15. afterwards these people say these things to everyone else
16. because of this, people can know what Allah says
17. I want to know what Allah says
18. I want to do everything as Allah wants”
19. people of the other kind are not like this
20. they are not as Allah wants people to be, they are like this:
21. they don’t want to think: “Allah does good things for people”
22. they don’t want to think something good about Allah because of this
23. they don’t want to know what Allah says
24. they don’t want to do everything as Allah wants
25. they want to do other things, not as Allah wants people to do
26. when Allah says things to these people, they say

a'.           “these things are not true”
b'. it is very, very bad if people are like this
c'.  people can know that if people are not as Allah wants them to be, after some time Allah will do something very, very bad to them because of it

In this explication, kuffār / kāfirūn encompasses ingratitude, wilful rejection of Allah’s messages and his messengers, rebellion against the truth, choosing to act in ways that are contrary to Allah’s commands, and rejecting or ignoring Allah’s signs and the threat of punishment. The comprehensiveness of this explication means that a disbeliever in Islam is, as a matter of course, considered to be guilty of a wholesale rejection of Allah’s commands, of his messengers, and of the truth. This is indeed what the concept of kuffār / kāfirūn conveys.

Equivalent semantic components will need to be included in semantic explications of the verb kafara and the noun kufr.

A question that arises is whether some of these additional elements added to the meaning of kafara could be incidental inferences that might arise in particular contexts, but are not intrinsic to the semantic structure of kuffār / kāfirūn. If someone is said to do kafara or be guilty of kufr, would all the elements of this explication be attributed to them? I believe they would. Not only have the exemplar verses cited here been carefully chosen to stand for many other similar verses – they are by no means cherry-picked – but to say someone is a kāfir is indeed to accuse the person of rejecting Allah’s messengers by calling them liars, of denying the truth of Allah, of disobedience, of coming under the wrath of Allah, and of ingratitude. In the divided worldview of the Quran, anyone who refuses to embrace Islam and rejects the Messenger and his message is guilty of all this. The verses of the Quran reinforce these judgements repeatedly.

The combination of qualities in this semantic explanation reflects the way the Quran construes disbelief as a full rejection of the quranic Messenger and his message, in its entirety, by people who should know better. By attributing ingratitude and denial to anyone who does not accept the Quran’s message, all disbelievers in Islam are stigmatised.

This does not mean, however, that all elements of this semantic explication are in focus in every instance in which kāfir, kafara or kufr are used. Consider, for example, the uses of kafara in Q2:102, which Cole took to have the sense of ‘blaspheme’:

(23) …and they follow what the devils recited for Solomon’s kingdom. Solomon did not disbelieve (kafara), but the devils disbelieved (kafarū). They taught humans sorcery, and what was sent down to Babylon’s two angels, Hārūt and Mārūt. Whenever they taught anyone they said, “We are but a test; do not disbelieve (takfur).” From those two they (the humans) learned how to divide a man from his spouse. Yet they did not harm anyone in this way, except by Allah’s permission. And they learned what harmed them, and what did not profit them. And they knew full well that whoever buys it [magic] will have no share in the world to come. Evil is the price for which they sold themselves. If only they had known! (Q2:102)

What aspects of kufr are in play here? In Q2:102, I propose that the reason the devils of Solomon’s time are said to kafarū is that they were enticing people to use magic to fulfil their desires instead of looking to what Allah says, doing what Allah wants, and depending upon Allah’s benevolent provision. The human magic practitioners were not thinking, “Allah does many very good things for people” (lines h, u in the explication), “I want to know what Allah says” (lines q, w), or “I want to do everything as Allah wants” (lines r, x). Instead, “they want to do other things, not as Allah wants people to do” (line y). In all this they were acting as kuffār / kāfirūn, as defined in the semantic explication.

It is also relevant to consider the immediately preceding and following verses in Q2:100–105. The preceding verses make clear that Q2:102 is about ‘People of the Book’ (in this case, Jews) who had discarded their covenant and rejected what their messenger brought, “throwing away the book of Allah” (lines z–a'). Furthermore, Q2:103 speaks of these past People of the Book abandoning faith and not refraining from evil (lines w–y); Q2:104 speaks of “people of Faith” in the present moment speaking duplicitously and disrespectfully, rejecting the quranic Messenger (lines z–a'); and Q2:105 states that such people without faith “do not want anything good to come down to you from your Lord” (line u). This surrounding context makes clear that these people were rejecting what a messenger had brought to them from Allah, which is a key element of the semantic explication of kuffār / kāfirūn.

These points all align with the semantic decomposition of kuffār / kāfirūn offered above. The humans’ resort to magic functions as but one element in an extended discussion of kufr which runs through the whole passage of Q2:101–105.

We conclude that the use of kafara in Q2:102, in its context, is consistent with key semantic elements in the proposed explication of kuffār / kāfirūn. The sustained focus throughout this passage is on the kufr of these Jews in Solomon’s time, so there is no need to posit a separate polysemous meaning ‘blaspheme’ to account for the instance in question.

Note, however, that in Q2:102 not all the semantic elements of kuffār / kāfirūn are in focus. Ingratitude (lines h–j, v) is not in focus in the surrounding passage. On the other hand, rejection of Allah’s truth and rejection of his messengers are in focus in the surrounding verses, but not in Q2:102.

Our explication of kuffār / kāfirūn is not unnecessarily overloaded. There are some aspects of kufr which have not been included in the explication because they are entailed by the meaning as defined. For example, arrogance is repeatedly attributed to the kuffār / kāfirūn, as expressed by a variety of different words such as astakbara ‘be arrogant’ and ‘alī ‘high, superior’ (Izutsu 1966: 142–152). However, in the Quran arrogance is implied by the meaning of kafara: human beings who take it upon themselves to deny the truth of what Allah has spoken are by definition arrogant, since they rate their own judgement above their all-powerful and all-knowing creator’s. Since arrogance is implied by the semantic explication of kafara, it does not need to be written into it.²⁵

Kufr is a complex and unique concept, tailor-made to fit the theologically divided worldview of the Quran. It is therefore not surprising that in the religious practice of Muslims around the world, the word kāfir is normally not translated but is borrowed directly from Arabic. The term is universally considered to be derogatory.

In English translations of the Quran, it seems impossible to avoid using a variety of terms for forms derived from the root k-f-r, depending upon the context. For any one instance, a translator must choose whether to emphasise ingratitude, rejection, denial or disbelief. The cost of having to make this choice in translations of the Quran is that the unity of the concept of kufr is concealed from the reader. This unity can only be retrieved through working with the Arabic text.

4.4. ittaqā

Our final lexical study is of the verb ittaqā (the root is w-q-y). This is one of the most difficult quranic words to translate into English. The derivative abstract noun is taqwā, which Izutsu, who produced two monographs on the semantics of quranic concepts (Izutsu 1964, 1966), translated as ‘fear of God’, a choice which is no doubt influenced by the biblical ‘fear of the Lord’:

… the idea of taqwā ‘fear of God’ … is indeed the central element of the Islamic conception of religion in general. (Izutsu 1966: 120)

A form of the verb ittaqā appears, for example, in the following verse, where it is translated as ‘protect yourselves’:

(24) “Oh People! Serve your Lord who created you, and those that were before you, so that you may protect yourselves (tattaqūna).” (Q2:21)

The agent noun formed from the root w-q-y is muttaqīn, occurring only in the plural.

The difficulty ittaqā presents for translators may be gauged by the diversity of English renditions of the final word in this single verse, which include ‘learn righteousness’ (Ali 1946), ‘attain piety’ (Munshey 2016), ‘guard against (evil)’ (Shakir 1985), ‘ward off (evil)’ (Pickthall 1976), ‘fear (Allah)’ (Palmer 1880, Sale 1734, Rodwell 1876), ‘be godfearing’ (Arberry 1998), ‘be saved’ (Khalifa 1981), ‘become Al-Muttaqoon (the pious)’ (al-Hilālī & Khān 1998) and ‘be Godwary’ (Reynolds 2018).

Most translators use a variety of English terms to translate this one Arabic word. Other translations of the same word in English Quran translations include ‘the righteous’, ‘act rightly’, ‘do right’, ‘reverence (Allah)’, ‘practise self-restraint’, ‘be careful of’, ‘be wary of’, ‘beware’, ‘do right’, and ‘keep your duty to’.

The root w-q-y occurs frequently in the Quran, mostly as the form VIII verb ittaqā, which can be used intransitively or transitively. The most basic verb derived from this root (form I) is waqā, which ditransitively means ‘protect someone against something’ and intransitively means ‘be wary, cautious’. Occurring ten times more frequently than form I waqā, the form VIII ittaqā is in the top 2% of most frequent lexemes in the Quran. Verbs of form VIII are typically reflexive or reciprocal in meaning, but they can also signify that the subject is doing an action for their own benefit (Wright 1896–1898: I:42). This suggests that ittaqā could be glossed as ‘he protected himself (from)’.

Throughout the Quran it is Allah who is almost invariably used as the object of the verb ittaqā, the one against whom people guard or protect themselves. Although some translators render the common phrase ittaqū allāha as ‘you (pl.) fear Allah’, the contexts where ittaqā is used are mostly not about fear or other negative emotions, but about being rightly directed or guided. Consider, for example, the following verses:

(25) Allah would never lead a people astray after guiding them, until he makes clear to them what they should be guarding themselves against (yattaqūna). Surely Allah knows everything. (Q9:115)

(26) Eat of what you have taken as booty, as is lawful and good, and guard yourself against (ittaqū) Allah; surely Allah is all-forgiving, all-compassionate. (Q8:69)

The focus in these verses is on following the guidance of Allah. In Q8:69, ittaqā describes the lawful eating of food taken as booty as an act of “guarding yourself against Allah”: this is guarding yourself because to do lawful acts keeps one on the right sight of Allah.

It is not that the Quran lacks vocabulary for fearing others or Allah. The two verbs khāfa (root kh-w-f) and khashiya (root kh-sh-y) can be translated as ‘fear’. In the Quran, khāfa (root kh-w-f) is a negative, undesirable emotion, which describes human feelings towards Allah only a few times (e.g. Q5:28, 94). Although occasionally it is said that believers fear (khāfa) the punishment of Allah (e.g. Q13:21), most of the instances of khāfa are affirmations that the rightly guided will be free from fear (e.g. Q3:170: “on them will be no fear”). Khāfa is thus for the most part an undesirable emotion of disbelievers. In contrast, it is stated that “the allies of Allah (there is) no fear (khawfun) on them” (Q10:62).

In contrast to khāfa, the verb khashiya (root kh-sh-y) can be used to refer to a godly fear of Allah, often in contrast to fearing people (e.g. fa-lā takhshawu al‑nāsa wa-akhshawnī, “do not fear people but fear me” (Q5:44)). A famous verse in Q33 reveals that Muhammad was right to have married Zainab, the former wife of his foster son Zaid. Ittaqā is used initially as an instruction from Muhammad to his foster son Zaid to keep his wife, which was at a time when Muhammad thought this was the right thing to do. However, this was subsequently corrected by a revelation that it was permissible for a man to marry the divorced wife of his foster son. Indeed, it was said that Muhammad himself ought to do so in order to demonstrate that this was permitted for believers in general. In this passage, Allah says that Muhammad had been holding back from marrying Zainab because he feared (khashiya) people instead of fearing Allah:

(27) (Remember) when you said to the one Allah had favoured and you had favoured (to Zaid), “Keep your wife and guard against Allah (wa-attaqi allāha).” But you hid in your heart what Allah was about to reveal, and you feared (takhshā) the people, when Allah had more right for you to fear him (takhshāhu). (Q33:37)

A fundamental problem with translations that use fear is that ittaqā is not actually an emotion at all. It is something one does, not something one feels. This contrasts with experiencing the (biblical) “fear of the Lord”, which CS Lewis has described as a feeling of “numinous awe” (Lewis 1940: 5).

If ‘fear’ is problematic in translating ittaqā, the alternative of ‘guard oneself’ or ‘protect oneself’ is equally problematic. One issue is a potential conflict with the theological principle that nothing can thwart Allah’s will. The English words guard and protect imply an effective exercise of power to thwart or defend against an attempted act of hostility: someone wants to do something bad to someone else, which is thwarted. That guard and protect combine with the preposition against is significant in this context. Thus the OED defines protect as “to support or assist against hostile or inimical action”.²⁶ However, the Arabic ittaqā is about being careful to avoid transgression of Allah’s laws. It is not about defending against and thwarting Allah’s hostility, which from a quranic perspective is a theological impossibility, due to Allah’s omnipotence.

Another difficulty with guard and protect as translations of ittaqā is that they do not capture the sense of righteousness and goodness that ittaqā projects. The positive good ittaqā conveys is reflected in translations that use the words pious and careful.

In the light of these observations, here is a proposed explication of ittaqā, third person masculine singular, perfect, used intransitively:

 X ittaqā (intr.)    

1. someone (X) thought like this:
2. “something very bad can happen to me
3. I don’t want this
4. it will not happen if I do something good
5. I want to do it”
6. because this someone thought like this, this someone did this good thing
7. people can think about it like this:
8. “because this someone did this good thing,
   this very bad thing will not happen to him”
9. at the same time, they can think like this:
10. “this someone is someone good”

Here is an explication of a transitive use of the same verb with Allah as the object:

X ittaqā Allah (tr.)          

1. someone (X) thought like this:
2. “Allah can do something very bad to me”
3. I don’t want this
4. it will not happen if I do something good
5. I want to do it”
6. because this someone thought like this, this someone did this good thing
7. people can think about it like this:
8. “because this someone did this good thing,               
   this very bad thing will not happen to him”
9. at the same time, they can think like this:
10. “this someone is someone good”

The semantic explication of ittaqā integrates a number of key components. One is the component of ‘fear’ (lines b–c): the person thinks something bad can happen to them if they do not do something. Another is the ‘piety’ element: the person wants to do something good and believes it is good to do this, and other people think the person is good for doing this good thing (lines d–e, h–j). Another is the ‘guarding’ aspect: because of what the person does, something bad will not happen to him. Note that the explication does not include the semantic prime feel: ittaqā involves doing and thinking, not feeling.

Note too that this explication does not suggest any hostility on Allah’s part, which could be implied by a translation with English protect and guard. It is not said or implied that Allah has an intent to do something bad to the person, nor that the will of Allah could be opposed or resisted in some way. Nor is it implied that the person is only acting under compulsion: rather, the person wants to do what is good. They are not just acting prudently, but acting piously.

How, then, should one render ittaqā in an English translation of the Quran? The heart of the meaning is about doing something good in order to prevent harm to oneself, in accordance with Allah’s direction. The problem with translations like ‘be pious’ or ‘learn righteousness’ is that they completely miss out the element of danger (line b). They are also too passive, downplaying the idea that the person is doing something. A possible translation is ‘be cautious (of)’, but no one translation can adequately capture all the facets of the meaning of ittaqā.

5. Conclusion

Words matter. To understand a culture one needs to rightly discern the meanings of its key words. While the practice of translation can offer a window into the world of another culture, it inevitably also frames and thus distorts the source text by means of the words of the target language. The instruments deployed to liberate meanings from the bondage of one language and make them known in another language are the very tools of their distortion and veiling.

Here we have considered certain key concepts of the Arabic Quran, a text which has exerted a profound and enduring influence to shape the languages and cultures of the two billion people in the world today who follow the religion of Islam.

The scholarly discipline of Islamic studies, as practised in the Western tradition, has relied on a lexicon infused with biblical concepts for understanding, interpreting and translating Islamic texts. To set aside the biblical frame, we have explicated a handful of Islamic terms using the tool of Natural Semantic Metalanguage, applying the technique of what Goddard and Wierzbicka have termed ‘experimental semantic analysis’ (Goddard & Wierzbicka 2014: 11). We have argued that terms that may be very difficult to translate into English can be precisely defined using NSM semantic decompositions. The results achieved are provisional and can no doubt be improved upon. Nevertheless, I hope they have demonstrated the power of NSM semantic analysis and its value for facilitating cross-cultural awareness and understanding in the important domain of religious belief.

With that conviction, and in honour of a dear friend Igor (‘Yehoshua’) Mel’čuk, I submit these findings, in the hope that others will improve what has been here begun.

^{1 English Jesus is from Latin Iesus, from Greek Yesous, from Aramaic Yeshua, from Hebrew Yehoshua.}

^{2. Today the meanings of these terms are expanding under the influence of global culture, so I am not sure that Lewis’s observation still holds true of these terms as they are currently used.}

^{3. Q4:28 refers to Sura 4, verse 28 of the Quran.}

^{4. Arabic verbs are cited in the third person singular perfect, which is the standard citation form in Arabic linguistics. Most Arabic words are formed around roots of three consonants.}

^{5. “guide, v.” OED Online. Oxford University Press, December 2021 (accessed 15 February 2022).}

^{6. From a different Arabic root, r-sh-d, the word murshid is formed, which can be an accompanying guide.}

^{7. I will use Allah to refer to God in the Quran and God for biblical references.}

^{8. I have called this the doctrine of Messenger Uniformitarianism (Durie 2018).}

^{9. The Quran attributes this uniformity to the sunna ‘way’ of Allah, which it says never changes (Q33:62).}

^{10. Although all the quranic messengers are male, it is far from clear that maleness should be part of the lexical explication: messengers, the Quran states, are chosen from among humans (al-nāsi), not from among men (Q22:75). The Quran also refers to angelic rusul: for example, Q35:1 calls angels “messengers with wings”. However, it seems that angels are considered rusul in the more general sense of ‘someone who delivers a message’, for elements of the typical rasūl biography are never attributed to angels.}

^{11. The doctrine of the moral perfection of messengers is known in Islam as ʿismah.}

^{12. The mission to Israel is characteristic of many prophets in the Hebrew scriptures, but not the post-Pentecost prophets of the early church.}

^{13. This explication is in Minimal English, which is a “highly reduced version of English” (Wierzbicka 2019: 45) built upon a core of the 65 NSM primes.}

^{14. In the Bible, prophesying is not even limited to the messages from the God of Israel (see e.g. 1 Kings 18:19; Jeremiah 23:13): people can also be said to prophesy in the name of other gods. However, for the sake of comparison with the quranic rasūl I will here limit my focus to prophesying in the name of God.}

^{15. The complexity of inter-religious cross-cultural communication is exacerbated because Arabic-speaking Christians use the word rasūl to translate apostolos.}

^{16. For example, compare the form I verb nakira ‘not know’ with form IV ankara ‘disavow, declare someone to be unknown’.}

^{17. Ambros & Procházka (2004: 239) report two other forms, kafara(t) and kāfira(t), as well as the feminine plural kawāfir, but these are all rare.}

^{18. This root is cognate with the Hebrew kāpar ‘to cover, atone’.}

^{19 Some translators render kufrān here as ‘ingratitude’, but it seems inconsistent with the Quran’s theology to attribute gratitude or ingratitude to Allah.}

^{20. This is a participial form of kafara.}

^{21 It is noteworthy that the expected gratitude is associated with awareness of Allah’s benevolent acts, not with Allah’s inherent goodness. Allah is never called ‘good’ in the Quran and ‘the good’ is not one of his ninety-nine “beautiful names” (Böwering 2002: 317). Moreover, there are references in the Quran to Allah as the source of bad as well as good. For example, Q91:8 speaks of Allah inspiring debauchery in humans and Satan also repeatedly declares that, since Allah has led him astray, he will in his turn devote himself to leading people astray (Q7:16; Q15:39). In Q38:82, Satan explains that he will do this by the power of Allah. Furthermore, there are many quranic references to Allah leading humans astray (cf. Q2:26; Q6:125).}

^{22 “infidel” OED Online. Oxford University Press, December 2021 (accessed 26 March 2022). Cole is not unaware that kāfir could be considered an insider term used for outsiders, for he writes: “there is, of course, a sense in which it [the Quran] views them [the kuffār / kāfirūn] as outside of and antagonistic to the true faith, part of what translators who used the term ‘unbeliever’ wished to convey.” His counter to this is: “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” (Cole 2020: 626). This misses the point that, in their actual lexical meaning, the English words infidel and unbeliever are insider terms used for outsiders.}

^{23 One would need to add ‘be ungrateful’ to this list because of examples like the following:}

^{Allah presents a parable: a town was at peace and content, abundant provision coming to it from every side. Yet it was ungrateful (kafarū bi-) for the favours of Allah. So Allah let it taste hunger and fear … (Q16:112; cf. also Q16:72).}

^{24. The predicate have to in X has to do Y is shorthand for X can’t not do Y. See Goddard (2014) for a broader discussion of modal verbs of necessity.}

^{25. The same can be said of the whole range of attributes of kufr which Izutsu explores in chapters 7 and 8 of Ethico-Religious Concepts in the Qurʾān.}

^{26. “protect, v.” OED Online. Oxford University Press, December 2021 (accessed 15 February 2022).}

About the authors

Mark Durie

Author for correspondence.
Email: mdurie@jefferycentre.mst.edu.au
ORCID iD: 0000-0003-2990-0937

Australian linguist, theologian and pastor. He is a Senior Research Fellow at the Arthur Jeffery Centre for the Study of Islam at Melbourne School of Theology, Founding Director of the Institute for Spiritual Awareness, and an Honorary Associate Professor at the Australian National University

References