This study examines how canonical qirāʾāt plurality is organized in five Qur’anic manuscripts preserved at the National Library of Malaysia (PNM): MSS 4333, MSS 3590, MSS 3599, MSS 5050, and MSS 3592. The manuscripts form a purposive analytical sample in which Sūrat al-Kahf (Q18) is legible at 16 loci of comparison. Main-text forms, corresponding marginal readings, and relevant visual markings were manually collated and mapped primarily against Abū ʿAmr al-Dānī’s al-Taysīr, with al-Nashr by Ibn al-Jazarī used to check riwāyah-level and ambiguous cases. Of the 80 main-text observations, 76 (95%) are compatible with the reading of ʿĀṣim, usually with a clear inclination toward Ḥafṣ. This stability does not remove other canonical readings. MSS 3590 preserves marginal alternatives at twelve of the sixteen loci, most often readings associated with Nāfiʿ and Abū ʿAmr, while retaining an ʿĀṣim-oriented main text. The sample displays two marked configurations—marginal annotation and in-line visual layering—and a residual category of unmarked main-text variation. This study argues that these pages establish a hierarchy of visibility: the main text privileges one reading, whereas selected alternatives remain available for comparison. The evidence supports comparative consultation but does not prove classroom use, local recitational prevalence, or the chronology of Ḥafṣ consolidation.
Copyrights © 2026