What is al-raqim (Qur'an 18:9) — "the inscription" or the place-name ar-Raqim?

Type: supporting argument (off-timeline) · Assessment: contested — a genuine lexical crux over a word that occurs only here in the Qur’an and was opaque to the earliest exegesis. The reading “the inscription/tablet” has the stronger morphology (the root means “to write”), was al-Ṭabarī’s own choice, and matches a real detail of the Christian legend; but a place-name reading — now seriously revived as Petra — is attested from the first generations and cannot be disposed of · debated by the classical commentators and by modern scholars.

The claim

Sūrat al-Kahf opens the story by pairing two things: “the companions of the cave (al-kahf) and al-raqīm” (Q 18:9). Al-raqīm (الرقيم) is a hapax legomenon — it appears nowhere else in the Qur’an — and it was obscure to the tradition from the start: Ibn ʿAbbās is famously said to have listed it among the handful of Qur’anic words he could not explain, and the commentators duly produced a spread of competing guesses. Two families of reading are respectable. On one, it is a common noun — “the inscription” or “the tablet,” a written record of the sleepers, so that they are literally “the People of the Inscription.” On the other, it is a proper name, “ar-Raqīm” — a valley, a mountain, a town, or (on the strongest modern form) the Nabataean city of Petra. The stake is not small: a place-name reading is what later underwrites the identification of a specific cave — at ar-Raqīm/al-Rajīb near Amman — as “the” cave.¹

Who advances it

For “the inscription/tablet.” The commentators gathered by al-Ṭabarī (d. 923) record this from Saʿīd b. Jubayr (“a tablet of stone on which they wrote the story of the People of the Cave, then set it at the entrance”) and from strands attached to Ibn ʿAbbās through al-Ḍaḥḥāk and al-Suddī; Kaʿb al-Aḥbār is cited for a lead tablet inscribed with the sleepers’ names, genealogies, and religion. Al-Ṭabarī, having stacked the views, chose this one, glossing al-raqīm as “a tablet, a stone, or something with writing inscribed on it”; al-Jalālayn and Maududi follow suit, and it is the default among modern Western readers of the sura because it fits the Syriac source-legend (below). For a place-name. An equally old strand — Qatāda, Mujāhid, al-Ḥasan al-Baṣrī, and further Ibn ʿAbbās transmitters — reads it as a valley, a mountain, or a town (one report: “a town in Byzantine territory”). In the modern period this pole was revived decisively by Mehdy Shaddel (2017), who identifies al-raqīm with Petra — the Nabataean capital whose indigenous Aramaic name was Raqmū/Reqem. Two strands sit at the weak edges: that al-raqīm is the dog’s name (a curiosity — the dog’s later name Qiṭmīr is itself a product of tafsīr), ʿIkrima’s oddity that it means “inkwell” in the Byzantine tongue, and the modern conjecture, below, that the word is a scribal error.²

The evidence for

For “the inscription.” The morphology points this way. The root r-q-m (ر-ق-م) means to write, mark, or inscribe; raqīm is a faʿīl form with passive sense — “that which is written” — on the pattern of marqūm, and the Qur’an supplies its own gloss: the heavenly ledger of Q 83:9, 20 is a kitāb marqūm, “an inscribed book,” from the very same root. And the sense answers to a concrete detail of the Christian Sleepers legend: in the Syriac and Latin recensions two Christians enclose a sealed leaden tablet inscribed with the sleepers’ names and story in the wall that seals the cave, so that they can be identified when found. Jacob of Serugh’s Syriac homily says it plainly — “they made tablets of lead and placed them beside them; on them they wrote down the names of the children of light” — and the same sealed tablet passes into Gregory of Tours’ Latin. On this reading al-raqīm is not an obscure toponym but exactly the tablet the source-tale carried, which is why Sidney Griffith and others, reading the sura against the living Syriac tradition, take the word within its plain lexical range and need no emendation.³

For a place-name. The geographic reading is not a late invention: it is carried by some of the oldest transmitters and locates the cave in a real corridor of the Roman Near East. Its strongest form is Shaddel’s: the Nabataean name of Petra is attested as Raqmū in Nabataean Aramaic, and both Josephus (Petra “called by the whole Arabian nation Rekeme”) and Eusebius’ Onomasticon (Petra “is also called Rekem”) preserve the equation — so a proper name al-raqīm would name a real, monument-filled city fit to stand beside the cave as one of God’s “signs.” A hapax can perfectly well be a proper name the first audience knew and later readers did not, and nothing in the grammar forbids reading “the companions of the cave and [of] ar-Raqīm” as two coordinated names.⁴

Adversarial assessment

What holds on each side. The “inscription” reading has the better of the morphology and a striking external corroboration: the root means “write,” the passive faʿīl sense is regular and fixed by kitāb marqūm, al-Ṭabarī judged it the most fitting view, and the source-legend really did carry an inscribed tablet of the names. That is a strong, convergent case. But it does not refute the place-name reading, which is anchored in the earliest tafsīr and, in Shaddel’s Petra form, rests on a genuine epigraphic-and-historical chain rather than mere folk etymology. The two have coexisted since the first generations precisely because the bare word does not decide between them.

What is overstated or imprecise. Three claims should not be leaned on. First, the “inscription as certainly the original sense” overreaches: the Qur’an never returns to the tablet, so its link to the Syriac motif, though attractive, is inference rather than statement. Second, the Petra identification, strong as it is, depends on the Rekem = Petra equation, which may itself rest on an ancient conflation (in the Aramaic Targums “Rekem” glosses Kadesh, not Petra) — so it is the leading modern reading but not proven. Third, and most clearly, James A. Bellamy (1991) proposed that al-raqīm is a copyist’s error (taṣḥīf) for al-ruqūd (“those asleep”) — a word that does occur later in the sura (18:18, wa-hum ruqūd) — so that the phrase originally read “the companions of the cave and of sleep.” It is ingenious, but a conjectural emendation with no manuscript support, widely judged unnecessary and now superseded by the toponym solution; the received text stands, and the dispute is over its sense, not its letters. The dog’s-name and “inkwell” glosses are curiosities.⁵

Assessment

Contested. A real, unresolved crux of a single opaque word. The reading “the inscription/tablet” carries a strong case — the root means “to write,” the form is regular and fixed by kitāb marqūm, al-Ṭabarī chose it, and it matches the inscribed name-tablet of the Christian source-legend — and is a fair modern default; but the place-name reading is attested from the earliest commentary and, in Shaddel’s identification with Petra, now has a serious epigraphic case that no one has closed off. Bellamy’s emendation to al-ruqūd and the dog’s-name gloss are the weak ends and change nothing. Which way one reads the word matters downstream: a place-name sense is the only textual peg for the further claim that a particular cave — ar-Raqīm/al-Rajīb near Amman, not the Petra of Shaddel’s reading — is the sleepers’ own, a claim weighed on its own page.

Sources