Synthesis — scored social-angle comparison of five Pakistani curricula

Aggregates the six lens reports scored against the rubric. Every figure here is a mean of book-level scores; the underlying per-book tables with page-cited evidence live in the lens files. Read this document with §4 (vintage control) in hand — several raw rankings reverse under it.

1. Board means by axis

All axes run 1–5. 1 = secular / pluralist / open / liberal pole. 5 = religious / exclusivist / closed / conservative pole. A2 is reverse-scored by construction (1 = thickest civics).

A6 target-by-target rows (Hindus, Indian state, Pakistani minorities, the West, internal others) were scored before the corpus expansion and cover the original three boards only. They are retained in §4 rather than here, because they cannot be compared with the five-board combined row above.

The Federal Board (FBISE / National Book Foundation) serves Islamabad and the federal-system schools in the provincial capitals and cantonments. It is the baseline the provinces are nominally working from, which makes provincial divergence measurable rather than asserted.

‡‡ A4 is the weakest row in this table and must not be quoted as a board figure. It rests on one Sindh book and two KP books. National-narrative closure can only be scored where a book carries the national narrative, which is a handful of volumes per board; the axis is real but its denominators are too small to rank provinces on. Every other row in this table rests on 6 or more books per board.

Class-10 science pulls every board’s A1 down, because matriculation science is close to secular in all five. Read the A1 column with that in mind: it is an average over a corpus in which the science books are numerous and almost uniformly secular.

† The A3 figure for Balochistan is the one number in this table that must not be read at face value. Balochistan carries more inquiry vocabulary than any other board — History 8 alone sets 32 evaluation prompts — and almost no open-opinion demand. Its history and Pakistan Studies books run a justify-that template on contested propositions (“Justify that the Treaty of Sevres is based on fair justice”), while its geography and science books set genuinely open tasks. The 2.92 mean averages those two opposite practices and describes neither. See lens history §2.4.

‡ Balochistan’s A6 was scored as a single axis, not disaggregated by target, because its out-group material is concentrated in three books rather than spread across the corpus. The Hindus-row figure is the mean of its two Pakistan Studies volumes and History 8; the combined-target row is the comparable cross-board figure.

Balochistan is scored on a smaller and unevenly distributed base than the other three boards — 16 English-medium books, and only 3–4 of them carry A4. Its 13 Urdu-medium volumes are theme-grade and are excluded from every mean above, as they are for the other boards.

The two composite indices — and why they disagree

Secularity index (= A1, lower is more secular) and conservatism index (= mean of A1, A2, A4, A5, A6 Hindus, A7) do not produce the same ordering. That divergence is the single most useful result in this document.

The result the fifth board produced

The federal board is the most secular and the least conservative of the five, on both indices at once. It is also the board the provinces are nominally working from. Every provincial version of the same curriculum is more religiously saturated, thinner on civics and more gender-conservative than the federal text it derives from.

That is a different claim from anything this study could make with four provincial boards, and it reframes the central question. The variation this project has spent its length measuring is not provinces diverging from each other around a neutral centre. It is provinces diverging from a common centre in a consistent direction — and Punjab, which prints the newest books in the country, is furthest from it on the conservatism index.

Two cautions, both material:

• Federal is scored on the smallest base. Ten English-medium books against Punjab’s twenty-two, and only classes 9 and 10 — its 6–8 volumes were not available. Its A2 rests on four books and its A4 on two. The direction is consistent across every axis, which is what makes it worth stating, but the margins are not.
• The two indices now agree, which they did not when the study had three boards. Federal and Sindh occupy the top two places on both. The divergence that made the earlier finding interesting — Punjab more secular than KP and more conservative than KP — survives between those two boards and no longer holds across the table.

Balochistan enters as the most secular board on A1 and the second-least conservative overall. Before that is quoted, one correction and one split have to be carried with it.

The correction. This project originally recorded Balochistan as running the federal NCC Inclusive Scheme of Studies 2024, which is the scheme the board hosts on its own website. Reading the front matter of all 28 volumes shows that phrase in none of them. Balochistan is running two generations simultaneously: classes 6–9 on National Curriculum 2022-23, printed as pilot editions for a single academic year, and class 10 on National Curriculum 2006 (Chemistry 10, English 10, Pakistan Studies 10). The catalogue has been corrected and the board-level statement “Balochistan’s books are the newest in the corpus” is withdrawn — a third of its scored English-medium books are on the corpus’s oldest curriculum.

What this buys. Balochistan is the only board that supplies vintage-matched partners for both of the other generations, which makes it the most useful board in the study for separating era from province:

The second row is the more valuable of the two, because it is a like-for-like comparison on the same curriculum document. On identical curriculum instructions, two boards produced materially different books. That is the strongest single piece of evidence in this study that provincial choice — not the federal curriculum alone — determines what a pupil reads. It cuts directly against the finding elsewhere in this document that vintage dominates province; both results are real, and they bound each other.

The divergence between the two indices survives the completed corpus: KP is the least secular board and the second-least conservative one, separated from Balochistan by 0.05. Punjab is third on secularity and clearly last on conservatism. Neither ranking should be quoted alone, and the gap between the two Punjab positions is the reason why.

Punjab is more secular than KP and more conservative than KP. The two properties come apart because they are carried by different axes: KP’s higher religious saturation (A1 2.62) sits alongside the corpus’s most open national narrative (A4 3.00, though on only two books), the lowest othering score of any board (A6 1.71) and low militarism (A5 1.43). Punjab’s lower religious saturation (A1 2.32) sits alongside the highest narrative closure (A4 4.50, n4), the highest militarism (A5 2.46) and the highest othering of the four boards (A6 2.46).

The practical implication: “how religious is this curriculum” and “how conservative is this curriculum” are separate questions about these books, and a single ranking that conflates them will mislead. Anyone using one number should say which.

Do not stop reading here. This raw ordering is partly an artefact of publication era, and §4 shows which parts survive controlling for it and which invert.

2. What separates the boards — the honest summary

Sindh leads on five of the seven liberal-pole measures, and its lead is real on civics: it is the only board that teaches citizenship as a practice rather than an institutional description — the franchise, the voting age, how to register, protest and boycott as legitimate acts, civil society existing “to limit and control the power of the state,” and the corpus’s only statement that “political control of the armed forces is with the civilian institutions.”

Punjab scores highest on closure, militarism and hostility to Hindus and India — and does so in its newest books, which is the finding that resists every attempt to explain it away as legacy material.

KP is the most internally consistent and the most open national narrative (A4 3.00, on two books): it is the only board that names its own military rulers’ overreach, prints the 1971 POW figure, treats Kargil as a loss of face, and distinguishes historical Hindus from Pakistani Hindu citizens inside a single book. It is simultaneously the worst board on gender.

No board is uniformly anything. Punjab is the most hostile board on Hindus and among the most inclusive on Pakistani religious minorities — its Pakistan Studies 9 (A6 5) and 10 (A6 1) are consecutive volumes of one course. The pluralism is genuine and compartmentalised into the citizenship slot; it never touches the historical narrative.

3. Progression, class 6 → 10

The vertical picture contradicts the intuition that these curricula radicalise with age.

The single most useful progression result: the class-9 gender collapse is not a pedagogical decision about adolescence. It is what happens when nobody revised the class-9 book. Where a board did revise it, the curve inverts.

4. Vintage control — where the raw ranking survives and where it breaks

The corpus runs four curriculum generations simultaneously. Uncontrolled board comparison is not meaningful, and the controls cut differently for each lens.

Survives — these are genuinely provincial

• Othering, at matched vintage. Punjab PS 9 (2023) vs KP PS 9 (2024–25): same subject, same grade, same curriculum scheme, descended from the same master text. They differ by a full band on Hindus (5 vs 4) and India (4 vs 3), and invert on minorities (2 vs 1). Punjab uses “enemy” 13 times; KP once, of the Pakistan Army. Editorial, not generational.

• Narrative closure — the finding survives three clean vintage-matched pairs. Punjab History 8 (SNC 2022) vs KP History 8 (2022–23): A4 5 vs 3. Punjab PS 9 (2023) vs KP PS 9 (AY 2024–25): 5 vs 3. Old-generation Sindh vs KP Pakistan Studies: 3 vs 3, no gap. The boards’ older books converge on A4 ≈ 3, so it is Punjab’s recent revision that diverges upward. Punjab runs the newest books in the corpus and produces the most closed narrative in it — the opposite of what a “legacy material” explanation predicts.

• Militarism rank order Punjab > KP > Sindh survives; the size of Punjab’s lead does not.

• Civics. Sindh’s lead comes from its older 2014/15-curriculum books, so it is confounded in Sindh’s favour, not against it — the finding is stronger than it looks, not weaker.

• Religious saturation, Sindh vs KP. Both run NC-2006 books, yet KP scores ~0.6 higher. Where they run the same book (Islamiyat 6–7, the Biology/Chemistry/General Science architectures) the scores are identical and counted once; the entire gap sits where they diverge — the English readers (KP 3/4/4/4 vs Sindh 2/2/2/3) and Physics 9 (3 vs 1).

• KP’s religious instructional load at class 9 is roughly twice either other board’s. KP requires two compulsory religious subjects — Islamiyat IX–X (126 pp.) and Mutalea-e-Quran (245 pp.) — a total of 371 pages against Punjab’s 170 and Sindh’s 176. Neither other board has any counterpart to the second subject.

Breaks — these were era, not province

• “Sindh is the harshest board.” This rests on Sindh’s 1→5 jump between classes 7 and 8, which is purely a 2015 book sitting beside an early-2000s federal reprint. Vintage-matched at 2004/2006 and at 2023–25, the ordering reverses.
• Punjab’s gender lead. PS 9’s 142 mentions of “women” against KP’s 1 is a vintage gap. Vintage-matched Pakistan Studies scores 4 for both Sindh (2004) and KP (2006).
• Punjab’s civics void is not caused by the SNC split. Punjab’s 2022 books (4.33) score identically to KP’s own 2022–23 volumes (4.33), and KP never split History from Geography.

Cannot be controlled — flagged confounded

The English readers (three genuinely independent anthologies on three different curriculum bases) and the gender reader-ratio comparison. Punjab’s religious-saturation score against either other board is also confounded — Punjab is uniformly SNC-2022/NC-2023 while almost nothing in Sindh or KP is contemporaneous, and the only vintage-matched slice (2022-era social studies) puts all three boards at 2.0 with n=2 each, which is underpowered. The Sindh–KP religious gap, by contrast, is controlled and does survive.

5. Cross-subject delivery — how a theme travels

• Punjab reinforces martial themes across four subjects at A5 ≥ 4 (English, Urdu, History, Pakistan Studies) against KP’s one (Urdu) and Sindh’s none. Punjab is also redundant across media at class 6: the same Nishan-e-Haider roll appears in both the English and the Urdu reader.
• The sharpest cross-subject contradiction is intra-board and doctrinal. Punjab’s Islamiyat 9 teaches a restrained, state-monopoly, anti-terrorism jihad framed by Paigham-e-Pakistan — and places a Nishan-e-Haider memorialisation exercise on the same printed page (verified against the source PDF, p.111). Meanwhile its History 8 has the 1965 army “filled with the spirit of Jihad” a full year before the doctrinal restraint arrives.
• Sindh delivers every civic theme in two or more subjects (some in three, including Islamiyat). KP delivers democracy and the UDHR across two, with the load carried by the English reader, not social studies. Punjab delivers no civic theme substantively twice.
• Sufism splits the boards three ways on identical saints: KP teaches it as religion (a whole History 6 chapter), Sindh as heritage tourism, Punjab barely at all.
• The Farewell Sermon travels furthest — seven books across three subjects. Sindh and KP teach it twice each, as ethics. Punjab teaches it three times and uniquely re-purposes it as a constitutional proof-text in Pakistan Studies 9. Same passage, different work.
• A theme can be relocated rather than removed. Punjab’s six science books and eight readers contain no Muslim-scientist material at all — but the strand survives in History 7’s Golden Age of Islam section. It moved subject, not curriculum. (This corrects the earlier reading that Punjab had dropped it outright.)
• A theme can outrun its own syllabus. Punjab’s English reader delivers the Two-Nation Theory, the 1940 Lahore address and the Fourteen Points at class 7 — a year before the History syllabus gets there — asking pupils “How does the author of the text justify that the Hindus and the Muslims belonged to two different nations?” Counting readers, the Pakistan Movement gets three passes in both Punjab and KP, and across those passes the central claim degrades epistemically: from argued, to stipulated (“debate to conclude”, History 8), to eternal (dated to 712 AD, PS 9).
• Punjab contradicts itself on that same claim one year apart: History 8 attributes the Two-Nation Theory to Sir Syed; PS 9 dates it to “the arrival of Muslims”. Neither is flagged as contested.

6. What all four boards share

These are the corpus-wide constants, and they matter more than the rankings.

1. Women are almost never agents of knowledge — but never was wrong, and is corrected here. There are four mentions, in three books, from two of the five boards: Marie Curie in KP’s Physics 10 (p. 244, with her dates) and Punjab’s Physics 10 (p. 174); Madam Curie in KP’s Chemistry 9 (p. 40); and Lise Meitner in KP’s Physics 10, credited with the theoretical explanation of nuclear fission and given a portrait at Figure 18.12 (pp. 252–253). Sindh and Balochistan name none in any science volume, and no board names one below class 9. The surviving claim is that women as agents of knowledge are vanishingly rare, wholly absent below class 9, and wholly absent from two of the four boards — not that they are absent. Rosalind Franklin still has zero occurrences corpus-wide while Watson and Crick are credited for DNA in three books. No woman is profiled as a scholar in any Islamiyat volume, and no woman is credited as author of any Urdu selection in twelve readers. A counter-intuitive corollary: the only three women credited as authors anywhere — Carney, Rossetti, Dickinson — sit in KP’s readers, the board that scores worst on every other gender measure. Its 2006 classic-anthology list admits women by accident; theme-commissioned SNC books admit them by quota, and only as subjects.
2. The “Hindu teachers” explanation for 1971 survives in all three of the boards that carry it, across 2004, 2023 and 2024–25 editions — the corpus’s most durable othering element and the strongest single proof of a shared federal ancestor.
3. Inquiry at matriculation is allocated, not absent — corrected. This list previously said inquiry collapses at matriculation everywhere. Completing the class-10 tier refutes that as stated: Punjab’s English 10 sets 22 open-opinion prompts, the most of any book in the corpus, and Sindh’s sets 5. Every one is on a personal, scientific or literary topic. The class-10 books carrying the national narrative set none — KP’s Pakistan Studies 10 contains zero evaluation verbs and zero open prompts in the whole volume. Pupils are taught to hold an opinion, on subjects where no institutional interest is at stake.
4. Partition violence is absent from every board, as is any economic critique of colonialism.
5. Pakistani religious minorities are the one out-group all three boards treat in the pluralist half (1.33–1.83) — the clearest shared positive norm in the corpus.
6. The West is not an out-group anywhere (2.00–2.50); Britain is a historical adversary only.

7. Confidence and limits

• 24 of 74 books are Urdu-medium and scored theme-grade from structure and proper nouns only. Nothing in Urdu is quoted anywhere in this analysis. KP carries the heaviest Urdu load, so KP’s scores are the least secure of the three boards — its A3/A4 means rest on 2 high-confidence and 3 low-confidence books.
• Absence-based scores are the weakest claims made. 31 books score A7 = 5 and all 31 are absence-driven; where the file is Urdu, absence may be OCR failure rather than editorial choice. Those are recorded as “not established”, not “absent”.
• Two known scoring traps, both honoured: communal vocabulary is not devotional vocabulary (Sindh SS8 has 167 “Muslim” and zero “Quran”), and nation-as-”her” personification inflates female token counts in Pakistan Studies — 24 of 43 computed ratios were flagged uninterpretable for this and related reasons.
• The largest single evidentiary hole is KP History 8 (Urdu), which carries KP’s entire colonial and Pakistan-Movement narrative at a register that cannot be resolved from the current scan. A Nastaliq-tuned OCR pass on that one file would move more of this analysis than any other acquisition.
• Quotation discipline: every verbatim quotation in the lens reports was re-grepped against source before inclusion; the patriotism lens’s two most load-bearing Urdu-derived claims were additionally verified by rendering the source pages as images and reading them directly.