Lens L1 — Religion

Scoring of A1 (religious saturation) for every non-Islamiyat book in the corpus, and A1b (doctrinal vs ethical register) for every Islamiyat / Mutalea-e-Quran volume, per the rubric.

Builds on track islamiyat, track sciences, track english, track urdu, track pakistan studies, track social studies and the evidence bank in raw-religion-social-studies. Findings there are not re-derived; everything new below was read directly from extracted-text/.

Citations are file · PAGE n (===== PAGE N ===== markers = PDF page index). Three files are native text with no markers (history-8-en, pakistan-studies-10-en, biology-9-en, plus the Sindh English 6/7/8 and Punjab/Sindh GS7 natives) and are cited by printed page, chapter or section. Verbatim quotation is from English-medium sources only. All Urdu Nastaliq OCR is theme-grade.

1. How A1 was operationalised

The rubric’s five bands were applied through four observables, in this order of weight:

Two traps, both documented in raw-religion-social-studies, are honoured throughout:

1. Communal vocabulary is not devotional vocabulary. Sindh Social Studies 8 has 167 “Muslim” and 0 “Quran”/“Prophet”/“Hazrat”; Punjab History 8 has 282 “Muslim” and 0 “Hazrat”/“Prophet”. In both, religion is a national-identity marker, not a devotional subject. Chapter fraction and register are scored on separate evidence, never from a single keyword count.
2. KP term-frequency counts are unusable. حضرت returns one hit in a 104-page KP file that demonstrably contains a chapter of Sufi hagiography. KP Urdu files are scored on structure and proper nouns only, and Quranic-verse counts for them are recorded as undetermined, never zero.

A whole-corpus religious-term density scan (26 English and Urdu markers, per 1,000 words) was run as a screen to locate material, not as a score. It is reported in §7 for transparency; it is diagnostic within a medium and worthless across media.

2. The scored table

conf = high (English text, direct evidence) · med (English OCR or structural inference) · low (Urdu OCR, or absence-based). Islamiyat rows carry A1b, not A1, and are shaded by the ‡ marker.

2.1 PCTB Punjab — SNC 2022 / NC 2023 throughout

Punjab A1 mean — English-medium non-Islamiyat books (n=18): 2.56. Including the four Urdu-medium readers (n=22): 2.68.

2.2 STBB Sindh — NC 2006 (2015–2020 editions); SS 6–7 = Sindh Curriculum 2015; Pakistan Studies = 2004

Sindh A1 mean — English-medium non-Islamiyat books (n=13): 2.31. Including the Urdu-medium readers and Urdu GS6 (n=18): 2.50.

2.3 KPTBB KP — NC 2006 (NOCs 2010–2018, reprinted 2018–2026); History 8 / Geography 8 = 2022

KP A1 mean — English-medium non-Islamiyat books (n=12): 2.92. Including the ten Urdu-medium books (n=22): 2.91.

2.4 BTBB Balochistan — NC 2022-23 pilot editions (classes 6–9) and NC 2006 (class 10)

Vintage — corrected, and it matters. This project’s catalogue originally recorded Balochistan as running the federal NCC Inclusive Scheme of Studies 2024, which is the scheme the board publishes on its own site. The books themselves say something different, and the books win. Reading the front matter of all 28 volumes shows the phrase “Inclusive Scheme” in none of them, and instead two generations running side by side:

• Classes 6–9 — National Curriculum 2022-23, printed as “Pilot Edition for a period of one academic year”. These are vintage-matched to Punjab’s SNC 2022–2023 books and to nothing else in the corpus.
• Class 10 — National Curriculum 2006 (Chemistry 10, English 10, Pakistan Studies 10). These are vintage-matched to KP’s and Sindh’s NC 2006 volumes.

So Balochistan is not one board for comparison purposes but two, and any statement that mixes its class-10 books with its class 6–9 books is confounded on vintage within the board. Where a claim below depends on that split, it is separated.

Balochistan A1 mean, English-medium non-Islamiyat: 2.00 (n = 15). That is the lowest of the four boards, and the result is driven almost entirely by its science books, which are the most secular in the entire corpus: General Science 6, General Science 7, Chemistry 9 and Chemistry 10 return zero religious tokens across a combined 188,923 words.

The distribution matters more than the mean. Balochistan does not spread religion thinly across every subject; it concentrates it. Eleven of its fifteen English-medium non-Islamiyat books score 1 or 2, while History 8, Pakistan Studies 9 and English 10 score 4, 4 and 3. This is a different structure from Punjab’s, where religious framing recurs in low doses across many subjects. A Balochistan pupil meets almost no religion in the science classroom and a great deal of it in history, Pakistan Studies and the English reader.

Counter-evidence, given equal prominence. The English 10 sirah unit is the clearest case in the Balochistan set of religion structuring a non-religious subject, and it also carries the corpus’s familiar pre-Islamic-Arabia framing — “Women were treated like animals. Ignorance ruled everywhere and in every sphere of life” (english-10-en · PAGE 9). But the same board’s Biology 9 contains no creationist framing, and its four chemistry and general-science volumes contain no religious token at all. Any claim that Balochistan’s curriculum is uniformly religiously saturated is refuted by its own science shelf.

2.5 Class 10 completion — the matriculation year, all boards

The class-10 volumes were acquired after the first scoring passes and are added here so the study covers the full 6–10 arc for every board. Matriculation matters disproportionately: it is the terminal year, the examined year, and the last curriculum most Pakistani pupils ever meet.

The class-10 science result is the cleanest in the study. Across eleven matriculation science volumes from four boards — biology, chemistry and physics, roughly 530,000 words — religious framing is effectively absent, and five of them return zero religious tokens. Whatever these curricula do elsewhere, they do not put religion into the matriculation science classroom. This holds across three curriculum generations and all four provinces, which makes it one of the few findings in this study that no vintage control disturbs.

The divergence is in the English reader, and it is wide: Sindh 0.3, Punjab 1.1, KP 2.2 tokens per thousand words on the same subject in the same year.

2.6 Corpus expansion — the Federal board, and 27 further books

Added 23 Aug 2026. The corpus grew from 120 to 165 current books and from four boards to five. The Federal Board / National Book Foundation was added at the reader’s request: it teaches Islamabad and the federal-system schools in the provincial capitals and cantonments, and it is the baseline the provinces are nominally working from — which makes provincial divergence measurable rather than asserted.

FBISE — Federal Board / National Book Foundation

FBISE A1 mean, English-medium non-Islamiyat: 1.80 (n = 10) — lower than every provincial board except Balochistan, and driven by the same thing: five of its six science volumes return zero religious tokens, across a combined 320,000 words.

The federal pattern is concentration, not saturation. Religion sits in Pakistan Studies 9 and almost nowhere else. A federal-system pupil meets no religious framing at all in physics, chemistry or biology at either class.

New books at the provincial boards

The class-6 History result deserves separating out. Punjab’s History 6 at 13.1 tokens per thousand words is the most religiously saturated non-Islamiyat book in a corpus of 165. It is also the first history a Punjabi eleven-year-old meets. The comparison is not with Islamiyat, which is religious by definition, but with the same board’s Geography 6 at 0.1 and its General Science 7 at zero — the same board, the same pupils, the same year, three orders of magnitude apart. Punjab does not saturate its curriculum with religion; it concentrates it, very heavily, in history.

Balochistan’s history is now quotable at every class 6–8. Its History 6 and 7 arrived in genuine English medium (verified at 233 function words per 1,000 — real prose, not OCR noise), joining the class-8 volume already held. That closes a gap this project had described as structural.

3. Per-board assessment

Punjab (PCTB) — religion is placed by design, in fixed slots, and is absent everywhere else. Punjab is the only board where every book in the corpus opens with a translated bismillah page — 20 of 22 English-medium files carry “(In the Name of Allah, the Most Compassionate, the Most Merciful.)” on page 1 or 2. Beyond that formula the distribution is unusually sharp: the four English readers all open with a Seerah or Sahaba unit (Seerat-un-Nabi 6, Last Sermon 7, Tolerance of the Rasoolullah 8, The Saviour of Mankind 9) and carry a second religious unit each; History 6 quarantines an Islamiyat-grade chapter in its last 24 pages; Pakistan Studies 9 opens with the ideology chapter and proof-texts it from Quran and hadith. Against that, the three Geography books, General Science 7 and 8, Chemistry 9 and Physics 9 are as secular as anything in the corpus. The “Hazrat cliff” — 180 tokens in History 6, 6 in History 7, 0 in History 8 — shows the devotional register being switched off rather than tapering. Punjab’s distinctive contribution to A1 is providential causation in the book’s own voice: History 8 attributes the creation of Pakistan to divine favour three separate times, and Pakistan Studies 9 and 10 continue it. That is the rubric’s own level-4 indicator, and Punjab is the only board that supplies it in English at this density.

Sindh (STBB) — the most secular lower secondary and the most religious upper secondary in the corpus. Sindh English 6, 7 and 8 contain zero religious selections between them; Social Studies 6 runs a genuinely plural religions-and-festivals table and uses the ecumenical “God” rather than “Allah”; General Science 7 and Physics 9 carry no religious content at all, and Physics 9 has no bismillah. This is not an artefact of the books being thin — Sindh’s Islamiyat is the heaviest in the corpus (40 of 100 marks for Quran recitation alone) and quotably the most doctrinally assertive (the Al-Tawbah 33 dominance passage). The religious load is concentrated rather than distributed: Islamiyat carries it at 6–8, then Social Studies 8 and Pakistan Studies 9–10 carry it at the top. Two counter-currents cut against the low score and should be reported with it: Social Studies 6 · PAGE 12 is the only place in any English-medium file in the corpus where Quranic verses are quoted verbatim outside Islamiyat, and it uses them to set creation alongside evolution as coequal theories; and the Islamiyat IX–X volume is the only book that both grants a named minority-rights framework and teaches striving for the political dominance of Islam.

KP (KPTBB) — religion is distributed rather than placed, and the class-9 load is roughly double. KP has no bismillah convention that survives OCR, yet it is the board where a pupil meets religious material in the widest range of subjects: a Prophet lesson opens all four English readers; a hadith epigraph sits in General Science 7 and in Pakistan Studies 10; Muslim-scientist sections sit in General Science 6, Biology 9, Chemistry 9 and Physics 9; History 6 devotes a whole chapter to tasawwuf and closes with a Khatm-e-Nabuwwat declaration; Geography 6 opens its population chapter theologically. KP Physics 9 is the only physics book in the corpus with Quranic verses. Decisively, KP is the only board running two compulsory religious subjects at class 9 — Islamiyat IX–X (126 pp.) and Mutalea-e-Quran (245 pp.), 371 pages against Punjab’s 170 and Sindh’s 176. On instructional volume alone, before any A1 judgement, KP’s class-9 religious provision is about 2.1× either other board’s. Two offsets are real and should be reported: KP English 6 carries the corpus’s only non-Muslim child protagonist, and KP English 9 pairs the Madina Charter with the 1948 UDHR.

4. Cross-board comparison, vintage-controlled

Ranking: KP most saturated, Sindh least, on both measures. The English-medium figure is the one to quote — it rests on high/medium-confidence evidence; the all-books figure imports twelve low-confidence Urdu-reader scores and is offered only to show the ranking is not an artefact of medium.

Vintage control

The corpus is not one generation, and the ranking must be read against that:

The Punjab-vs-anyone comparison is confounded end to end. Punjab is the only board with a complete post-2022 series; nothing in Sindh or KP is contemporaneous with it except four books. On the one vintage-matched slice available — the 2022-era social-studies books — Punjab History 8 (4) and Geography 8 (1) average 2.0, KP History 8 (3) and Geography 8 (1) average 2.0, and Sindh Social Studies 6 (2) and 7 (2), on Sindh Curriculum 2015, average 2.0. On that slice the three boards are indistinguishable, n=2 per board. This is underpowered and is flagged as such. The honest statement is: Punjab’s headline position between KP and Sindh cannot be vintage-controlled.

The Sindh-vs-KP comparison can be vintage-controlled, and it survives. Both boards run NC 2006 books of comparable vintage across sciences, English and Islamiyat. Within that matched set KP scores 2.92 against Sindh’s 2.31 on English-medium books — a real board difference of ~0.6, not a generational one. It is concentrated in exactly the places where the two boards are not running the same book:

• Where they run the same book, they score the same. Islamiyat 6 and 7 are one syllabus in two languages (established in track islamiyat §6.1) — A1b 3, reported once. The Biology 9, Chemistry 9 and General Science 7/8 architectures are effectively identical, and their A1 scores differ by at most one point, driven by the length of the Quranic section, not its presence.
• Where they diverge, KP is higher every time. English readers: KP 3/4/4/4 against Sindh 2/2/2/3. Physics 9: KP 3 against Sindh 1. Class-9 religious-subject page count: KP 371 against Sindh 176.

One widely-repeated claim needs correcting. It is not true that Punjab’s NC-2023 books simply dropped Muslim-scientist sections. They dropped them from the science books — Biology 9, Chemistry 9, Physics 9 and General Science 6/7/8 contain not one reference to Jabir, Al-Razi, Ibn Sina, Al-Biruni or Ibn al-Haytham — and relocated the strand to History 7, where a Golden Age of Islam section (history-7-en · PAGE 35–39) profiles Ibn-al-Haytham and Al-Kindi and sets a chart activity on “the famous Muslim scientists of the Golden Age of Islam” · PAGE 39. Sindh and KP keep the strand inside the science books themselves. So the Punjab move is a relocation from science to history, not a deletion — a subject-boundary decision, not a secularising one.

5. Vertical progression, class 6 → 10

Per-class A1 means. English-medium non-Islamiyat books first (higher confidence), then all non-Islamiyat books.

Punjab: shallow U, ending higher than it starts. Class 6 is elevated by History 6’s Islamic- civilisation chapter and the Seerat-un-Nabi opener; classes 7 and 8 dip as Geography and General Science dominate the denominator; classes 9 and 10 rise as Pakistan Studies enters. Net movement 6→10 is +0.25 (EM). The English readers hold flat at 4 across all four grades — the constant.

Sindh: the steepest climb in the corpus, +2.00 from class 7 to class 10. Sindh’s classes 6–7 are the most secular books anywhere in the corpus (English readers with zero religious selections; Social Studies 6 running a plural religions table). Then Social Studies 8 opens with an argument from design, and Pakistan Studies 9–10 grounds the state in Quran and Sunnah and expounds the five pillars. A Sindh pupil’s non-Islamiyat exposure to religion roughly doubles between class 7 and class 10.

KP: highest and flattest — a plateau, not a ramp. KP never drops below 2.40 on any measure and never exceeds 3.33; the class-9 peak comes from Physics 9’s Quranic opening plus Pakistan Studies 9, and the all-books class-8 trough reflects the 2022-vintage History 8 and Geography 8 being less religious than the 2006-vintage class-6 books they sit above. This is the one place where a within- board vintage effect is visible: KP’s newer books score lower than its older ones. Net movement 6→10 is +0.50 (EM), 0.00 (all books).

Add to this the Islamiyat instructional load, which the A1 table deliberately excludes: at class 9 KP requires 371 pages of compulsory religious text across two subjects, Punjab 170 across one, Sindh 176 across one. On that measure the class-6→10 trajectory is steeply upward for KP and flat for the other two.

6. Cross-subject theme traces

6.1 The Farewell Sermon (Khutbat-ul-Wida) — the corpus’s most-reinforced text

The equality clause — “An Arab has no superiority over a non-Arab… nor a white over a black” — is the single most-repeated sentence in the corpus, appearing in seven books across three subjects. The boards reinforce it in different registers. Sindh and KP teach it twice, both times as ethics: in Islamiyat as a seerah episode and in the language reader as a comprehension text. Punjab teaches it three times and moves it into politics — the same hadith that anchors the English reader’s Unit 1 at class 7 reappears at class 9 as the scriptural proof-text inside the Ideological Basis of Pakistan chapter, immediately before “Almighty Allah is the source of law. The basis of law is the Holy Qur’an and Sunnah” (pakistan-studies-9-en · PAGE 10, 12). That is the clearest instance in the corpus of a religious text being re-purposed as constitutional argument across a subject boundary.

6.2 Scripture inside the science classroom

All three Biology 9 books carry a Chapter-1 Quranic section quoting the same core verses and all three examine it — this is the most aligned feature in the entire corpus and it survived Punjab’s complete NC-2023 rewrite. It is not a board difference. The divergence is at the edges: Punjab uses the verses to support a common-descent reading (“Quran also describes the common origin and modification of animals”, · PAGE 13) where Sindh reads Al-Anbiya 30 materially (“water is symbolized with the protoplasm”); and only KP extends scripture beyond biology. Chemistry is the one subject where no board puts scripture in the book. A KP pupil meets Quranic text in two science subjects plus a hadith in a third; a Sindh pupil meets it in exactly one; a Punjab pupil in exactly one.

6.3 The Muslim scientist

This is the sharpest divergence on L1. In Sindh and KP the Muslim scientist is a cross-curricular motif that a pupil meets in science, in the language readers and in Islamiyat, reinforcing at every turn. In Punjab it is a single history lesson and nothing else. Note what this does to the religion score: it lowers KP’s and Sindh’s science books only slightly (these are A1 2 features, history-of-science rather than doctrine) but it is the mechanism by which religion stays continuously present in the KP curriculum rather than being confined to designated slots.

6.4 Providential and ideological causation

Two distinct mechanisms, and the boards use different ones.

Providential — history attributed to divine agency in the textbook’s own voice — is a Punjab signature and is essentially Punjab-only in English: History 8, three times (“with the grace of Allah Almighty, made the impossible possible”, p.67; “with the blessings of Allah Almighty”, p.74; “By the grace of Allah Almighty… Pakistan had been created”, p.81); Pakistan Studies 9, “Pakistan is a fortunate country, in the sense, that Almighty Allah has blessed it” · PAGE 61; Pakistan Studies 10, “with the grace of Allah Almighty and the determination of the Pakistani nation and armed forces, this country has always been … protected”. No Sindh or KP English-medium file uses this construction.

Ideological — the state’s existence explained by Islam being a complete code of life — is universal. All three boards open Pakistan Studies with an Ideological Basis / Ideology of Pakistan chapter: KP · PAGE 6–7 (“Pakistan is called an ideological state… the ideology of Pakistan was simultaneously a religious, political and economic ideology”); Sindh · PAGE 5 (“Islam is a universal Deen i.e., it is a complete code of life”); Punjab · PAGE 9–12. Sindh additionally applies the argument-from-design move at the head of Social Studies 8 · PAGE 4, and KP at the head of Geography 6’s population chapter · PAGE 101. Punjab’s geography books make no such move anywhere — the clean control that shows all three boards can and do write secular subject matter when they choose to.

7. Density screen (diagnostic only, not a score)

Religious-marker hits per 1,000 words, English markers for English files and Urdu markers for Urdu files. Not comparable across media — Nastaliq OCR suppresses Urdu counts by roughly an order of magnitude. Reported because it is what located most of the material above.

Note what this screen does not show. KP English 9 (6.22) and KP Pakistan Studies 9 (15.11) sit high because their religious content is in English; KP History 6 registers 1.38 on the Urdu scale despite containing a whole chapter of Sufi hagiography and a Khatm-e-Nabuwwat declaration page. The screen found material; it did not rank books, and no score above was set from it.

Naming register differs by board and is a usable A1b/honorific signal: Punjab uses Rasoolullah / Hazrat (English 8: 30 Rasool, 34 Hazrat); Sindh uses the Holy Prophet / Holy Last Prophet (Islamiyat IX–X: 289) and never Rasoolullah; KP mixes Hazrat, Prophet Muhammad and Holy Prophet.

8. Caveats

1. The Punjab comparison cannot be vintage-controlled. Punjab’s series is uniformly SNC 2022 / NC 2023; almost nothing in Sindh or KP is contemporaneous. Punjab’s mid-table position between KP and Sindh is therefore confounded and is reported as such. Only the Sindh–KP gap is vintage-controlled and survives.
2. Twelve of the 62 non-Islamiyat books are Urdu Nastaliq OCR and are scored theme-grade, on chapter headings and proper nouns alone. Every such score is marked low confidence. Board means are therefore given twice — English-medium only, and all books — and the ranking is stated on the English-medium figure.
3. KP term-frequency counts are unusable and no KP claim above rests on one. حضرت returns one hit in a 104-page file containing a chapter of Sufi hagiography; ligature breakage produces both false negatives and false positives. Quranic-verse counts for KP Urdu files are undetermined, never zero. Absence of a bismillah in KP files cannot be confirmed — title pages are scanned artwork.
4. general-science-6-ur could not be theme-graded at all and its A1 is recorded as not established, with a provisional 1 used only in the all-books mean. Its inclusion moves the Sindh all-books mean by less than 0.1.
5. Punjab History 8’s native extraction strips 100% of Arabic script, so all ﷺ honorifics and the probable title-page bismillah are invisible. Its zero honorific count is an artefact. Its English providential framing is the reliable signal, and that is what the score rests on.
6. english-9-en (native) is a partial extraction beginning at Unit 2 — it omits Unit 1, The Saviour of Mankind, which is why its religious-marker density (0.47) is an eighth of the complete OCR file’s (3.07). The OCR file was used for scoring. Anyone re-running the density screen must exclude the native duplicate, and likewise the Sindh English 6/7 and Sindh General Science 7 duplicate pairs.
7. Two KP class-9 files post-date track islamiyat and correct it. That report states the KP Islamiyat 9 scan is incomplete (61 pp.) and that the KP Mutalea-e-Quran title is “not in this corpus”. Both are now present: islamiyat-9-10-ur (126 pp., complete, with the index showing a Hijrat-o-Jihad lesson — closing the “jihad framing unverifiable” gap at the level of presence) and mutalea-quran-9-ur (245 pp.). track islamiyat §7 should be updated. This materially changes KP’s class-9 religious load and is the single most consequential new fact in this lens.
8. The A1 bands are ordinal, and the means are means of ordinals. A board mean of 2.92 against 2.31 should be read as “KP’s typical book sits between bands 2 and 3, KP’s noticeably closer to 3” — not as a 26% difference in anything. Differences below ~0.3 between boards are not interpreted.
9. Absence is the weakest claim throughout. “Zero religious selections” is asserted only for the three Sindh English natives, where the text is exact. Everywhere else, and universally for Urdu files, absence is reported as not established.
10. Page markers are PDF indices, not printed folios, and typically run 2–5 pages ahead of the book’s own numbering. The banked social-studies profile cites Punjab History 8 and Punjab Pakistan Studies 10 by printed page because those files are native text; the two conventions are not interchangeable.