What the Pupil Must Write

Assessment demand as an instrument for reading ideology in Pakistani school textbooks, classes 6–10

1. Abstract

Textbook criticism has overwhelmingly analysed what school books say. This article analyses what they require pupils to produce: exercises, keyed multiple-choice items, fill-in-the-blank and true/false questions, short- and long-answer prompts, chapter activities, printed teacher notes and specimen board papers. The argument is that an examinable proposition is categorically different from a stated one. A passage offers a reading a pupil may accept, ignore or resist; an exercise compels the pupil to write the proposition in their own hand and rewards them for it. Where ideology enters assessment, it stops being available and becomes required.

The article builds a typology of assessment demands by the cognitive act each requires — recall, comprehension, justification of a supplied conclusion, unguided transfer, and evaluation — and argues that the third and fourth do the ideological work. It applies the typology to a corpus of 112 textbooks from four Pakistani provincial boards (Punjab, Sindh, Khyber Pakhtunkhwa, Balochistan), classes 6 to 10, quoting only from English-medium sources verified by exact string match. It catalogues twenty assessed items across all four boards requiring pupils to state the moral condition of pre-Islamic Arabia, against none requiring them to state anything comparable about pre-Islamic or Hindu India; it identifies the rescue-from-ignorance template being examined with Europe, not India, as its named beneficiary; and it gives full weight to Sindh’s evaluative civics assessment, the corpus’s strongest counter-case. The method observes demand, not uptake, and cannot show what any pupil believes.

2. Introduction: the understudied half of the book

The literature on Pakistani school textbooks is a literature about passages. It quotes what the books assert, counts how often they assert it, and traces the assertions across editions and provinces. That work is necessary and it has been done well. But it reads only half of each book. The other half — exercise sets, review sections, activity boxes, teacher instructions, specimen examination papers — is where the book stops describing and starts requiring.

The distinction is not merely one of genre. Consider two sentences from facing regions of one volume. The first: Before the advent of Islam, Arabian society was marked by a strong tribal structure, patriarchal norms, and social inequality. The second: What was the social condition of women, orphans, and slaves in pre-Islamic Arabia? Both concern the same content. But the first is a claim a reader encounters, and encountering a claim is compatible with a wide range of subsequent states — assent, indifference, forgetting, private disagreement. The second is an instruction. It has a correct answer, that answer is the content of the first sentence, and the pupil who supplies it is marked correct while the pupil who does not is marked wrong. Both sentences appear in Punjab, English 10, at pp. 6 and 10 — and the question appears a third time, at p. 123, inside a printed model paper for the board examination.

That assessment governs what is taught and learned — that examinations exert backwash on the classroom, and that retrieving and producing a proposition consolidates it far more durably than reading it — is well established in the study of learning and curriculum. What has not been done is the systematic application of that fact to the reading of ideology in textbooks. If assessment governs attention, the ideological content of a curriculum is not distributed evenly across its pages; it is concentrated, disproportionately and measurably, in the small fraction of pages that carry demands.

There is a further reason to look at the demand side here: the mechanism was documented in Pakistan thirty years ago, and from the examiner’s end rather than the author’s. K.K. Aziz, examining the March 1989 Punjab Textbook Board reprints, put it with a precision no amount of content analysis reaches:

“If the girl said in the examination script that the Aga Khan led the Simla Deputation, not Nawab Muhsin-ul-Mulk, the examiner would fail her, because the Punjab Textbook Board said it was Muhsin-ul-Mulk. The matter could not be debated.” (Aziz 1993, p. 189)

Aziz is describing a factual error, not an ideological claim. But the structure he identifies is exactly the one at issue: a proposition whose status as knowledge is guaranteed not by evidence but by the board, and whose non-reproduction carries a penalty. The book’s assertion and the examiner’s mark are two halves of one instrument. Content analysis sees only the first half.

Section 3 makes the methodological case and builds the typology; Section 4 describes the corpus and the evidentiary rules; Sections 5 to 7 present the findings. Section 8 gives counter-evidence at length, because the corpus contains a great deal of it and an argument of this kind is worthless if it cannot survive the material that cuts against it. Sections 9 and 10 set out implications and limits.

3. The argument: four properties, and a typology of demand

3.1 Availability, repetition, sequencing, demand

A proposition in a textbook can have any of four distinct properties, and analyses that conflate them reach unreliable conclusions.

Availability — the proposition is in the book. This is what content analysis measures and it is the weakest of the four. A sentence no exercise touches and no teacher note flags is available and almost nothing else. Punjab’s History 7, p. 32, contains that board’s only rescue-frame claim about Hindu India — lower-caste Hindus “lived a miserable and socially segregated life” and were “impressed by the equality, caste-less system and social justice of Muslims.” Nothing in the volume’s apparatus attaches to it.

Repetition — the proposition recurs, within a volume or across subjects, raising the probability of encounter and creating cross-subject reinforcement.

Sequencing — the proposition arrives at a particular age and in a particular order relative to others. Punjab’s History 8 applies “the spirit of Jihad” to the conventional inter-state war of 1965 a full year before Islamiyat 9 teaches that jihad is not a synonym for war and belongs to the state alone. The looser usage is taught first and never retracted.

Demand — the pupil is required to produce the proposition, or something that entails it, in assessable form. This is the only one of the four that changes the pupil’s relation to the proposition from receipt to production, and the only one that attaches a reward and a penalty.

The four are logically independent and here come apart in both directions. The Indian-side rescue claim is available and not demanded. The Arabian-side rescue claim is available, repeated, sequenced across two subjects in one school year, and demanded twenty times in English-medium books alone.

3.2 A typology of assessment demand

Sorting assessment items by the cognitive act each requires yields five types, of which two are of interest.

Type 1 — Recall. The pupil returns a discrete fact: a date, a name, a number. “When did World War II begin?” (Balochistan, History 8, p. 96). Recall items are the bulk of the corpus’s load and are ideologically consequential only where the fact recalled is itself contested, in which case the item settles the contest by fiat — the Aziz mechanism.

Type 2 — Comprehension. The pupil restates what the text said. “What was the condition of Arabs before spread of Islam?” (Balochistan, English 10, p. 11). These items reproduce the text’s evaluative content along with its informational content, because in these books the two are not separable. The pupil who answers correctly has written the evaluation.

Type 3 — Justification of a supplied conclusion. The pupil is given the conclusion and asked to produce the reasons for it. The grammatical signature is Justify that X, Prove that X, Conclude that X, Debate to conclude X. It borrows the surface form of argument while removing the one thing that makes an argument an argument — the possibility that it fails. The exercise cannot end in a different conclusion, because the conclusion is printed in the instruction. This is worse for ideological transmission than bare assertion. An assertion invites the reader to accept or not; a justify-that prompt recruits the pupil’s own reasoning capacity into the service of a conclusion they did not reach, and rewards the recruitment. The pupil ends holding a written set of reasons, in their own handwriting, for a proposition they were not permitted to test. What has been practised is not inquiry but advocacy on assignment.

Type 4 — Unguided transfer. The pupil applies a frame from the text to a target the item does not name: What parallels can you draw…, …in the present age. This is rarer and more permissive than Type 3, not less. Type 3 fixes both frame and target; Type 4 fixes the frame, licenses the operation, and leaves the target to the pupil, the teacher, or the classroom. It is the only form in the corpus that requires generalisation of an evaluative schema as such.

Type 5 — Evaluation. The pupil reaches a judgement the item does not supply, on evidence it does not pre-select, with more than one defensible outcome: “Critically evaluate the State of human rights in Pakistan and suggest measures to safeguard them” (Sindh, Pakistan Studies 9–10, p. 186). Type 5 is the negation of Type 3 — same verb family, opposite epistemic structure — and its presence is the clearest single indicator that a board will let a pupil arrive somewhere the book did not.

3.3 Why verb-counting fails

A tempting shortcut is to count “higher-order” verbs — analyse, evaluate, justify, assess, critically — and treat density as a proxy for inquiry. In this corpus that measure inverts. Deduplicated across all English-medium files, evaluation-family verbs occur at 9.1 per 100,000 words in Punjab and 13.1 in Balochistan, against 3.6 in Sindh — and Sindh is the only board that asks a pupil to judge a government. Punjab’s History 6 carries the corpus’s highest inquiry-verb density and its clearest received-truth instruction. The verb does different work in each case: in Sindh’s Social Studies 6 the object of assess is the sitting government’s performance; in Punjab’s History 6 the object of justify is a conclusion the same page has just hedged. The equivalent trap on the civic side is lexicon density: KP’s Pakistan Studies 9 has the corpus’s highest civic-term density and teaches the least usable civics, because that density is entirely constitution-making history ending in 1971. Density measures how often something is named; the typology measures what the pupil must do.

3.4 The demand mechanism has a documented policy ancestry

The forms catalogued here are not free-floating pedagogical accidents. Federal curriculum documents of the 1980s and 1990s stated the demand explicitly, as an objective rather than a by-product. The following are quoted at second hand, from Nayyar and Salim’s compilation, and are cited as such:

• Social Studies Curriculum, Classes VI–VIII (National Bureau of Curriculum and Textbooks, Islamabad, 1984), p. 16, lists under the concept “Jehad” the affective objective “Aspiration for Jehad.”
• Urdu Curriculum, Classes VI–VIII (NBCT, 1986), p. 13, requires that the pupil “must be aware of the blessings of Jehad, and must create yearning for Jehad in his heart,” repeated for classes IX–X in the 1988 document, p. 8.
• The Primary Curriculum of 1995, p. 41, directs that “The Ideology of Pakistan be presented as an accepted reality, and be never subjected to discussion or dispute.”

The last is the purest possible statement of assessment-as-compulsion. It does not instruct the author to make the ideology persuasive or to marshal evidence for it; it removes the proposition from the class of things that can be argued about. A proposition that may never be disputed cannot be the object of a Type 5 item; the only forms available to it are Types 1 to 4. The modern Justify that… prompt and the modern unguided-transfer question are legible in that light as the surviving instruments of an older and more explicit policy — the mechanism outliving the directive that created it.

4. Corpus and method

The corpus comprises 112 catalogued textbooks from four provincial boards — the Punjab Curriculum and Textbook Board (PCTB), Sindh Textbook Board (STBB), Khyber Pakhtunkhwa Textbook Board (KPTBB) and Balochistan Textbook Board (BTBB) — covering classes 6 to 10 across English, Urdu, Islamiyat, Mutalea-e-Quran, History, Geography, Social Studies, Pakistan Studies and the sciences. Extraction yielded 110 usable text files (65 English-medium, 45 Urdu-medium) plus two that are empty.

Four curriculum generations are in simultaneous use, and this confounds naive board comparison throughout. Punjab runs a uniform post-2022 series (SNC 2022 / NC 2023). Sindh has not adopted the SNC: its Social Studies 6–7 are on the Sindh Curriculum 2014/2015, its Social Studies 8 an early-2000s federal reprint, its Pakistan Studies a 2004 approval still in print. KP is predominantly National Curriculum 2006, approved 2010–2018 and reprinted through 2025–26, with an SNC-era exception at class 8. Balochistan’s holdings are recent pilot editions on the federal scheme. Where a comparison cannot be vintage-matched it is flagged confounded.

The English-only quotation rule. Urdu Nastaliq optical character recognition in this corpus is theme-grade: it recovers chapter structure, proper nouns, award names and Arabic-numeral dates, and does not recover wording. Ligatures break mid-token unpredictably, producing both false negatives and false positives. Accordingly no Urdu-medium wording is quoted anywhere in this article. Urdu findings are reported at structure and topic level only and marked low confidence.

The cost is unevenly distributed. Punjab and KP teach Islamiyat entirely in Urdu; KP teaches History and Geography in Urdu; Balochistan’s History 6 and Islamiyat volumes are Urdu. Sindh is the only board teaching Islamiyat in English, and is therefore the most visible board to this method — any finding that Sindh carries more of something must be read against that. The corpus’s two most structurally interesting Urdu findings — Punjab’s Urdu 8 four-page exercise block on the pre-Islamic condition of women (pp. 95–102), and Balochistan’s History 6, the only volume carrying both the Vedic-caste and pre-Islamic-Arabia clusters — are precisely the ones that cannot be quoted. The consequence for the count in §7 is that the finding is conservative: twenty assessed items is the English-medium floor, not the total.

Evidence tiers. Findings are graded on the project’s four-tier scheme: Tier 1, explicit linkage in a sentence, caption, exercise or learning outcome; Tier 2, shared evaluative vocabulary applied to two targets; Tier 3, structural juxtaposition without lexical linkage; Tier 4, not established, or counter-evidence. Absence claims are the weakest class permitted; for Urdu files a term not found is reported “not established”, never “absent”.

Verification and citation. Every English-medium quotation was located by exact substring match against its extracted source after whitespace normalisation, and its page marker recorded. Citations take the form Board, Title, class N, p. X, where X is the PDF page index of the scan, typically one to six pages ahead of the printed folio. Three files are native digital text without page markers and are cited by printed folio, marked f. OCR damage in quoted strings is preserved and flagged in place.

5. Findings, organised by demand type

5.1 Type 3 — justification of a supplied conclusion

The clearest specimen is Punjab, History 6, p. 30. The page’s narrative states an explicit historiographical uncertainty — “Decline of Indus Valley Civilization: We do not know why this civilization came to an end, but a number of possibilities have been suggested” — and lists three (climate change, seismic river diversion, invasion). Immediately below, in a box headed “Skill: Activities”, six tasks are set. The fifth is genuinely open: “Analyze reasons for the demise of the Indus Valley Civilization.” The sixth is not:

“Justify that Indus Valley people did not learn the warfare nor developed their trade, and therefore, were easily defeated by Aryans.”

Two demands sit in one box, on the page whose own narrative says the answer is unknown, and they are incompatible: one asks the pupil to weigh three hypotheses, the other requires them to argue for one and to add two causal claims — undeveloped trade, undeveloped warfare — that the narrative does not make. Thirty pages later the volume records that the Aryan arrival “was a large migration and used to be seen as an invasion” (p. 61), while the heading at p. 60 reads “Aryan Invasion of India”. The assessment demand resolves the book’s own contradiction in one direction only.

Balochistan’s History 8 runs the template as standing architecture. Each chapter opens with a two-part outcomes page — “Students Learning Outcomes” (knowledge) and “Skills” — and closes with an “Exercise”, then “Activities for Students”, then “Instructions for teacher”. The justify-that form recurs in the Skills block and is reissued as a written task in the Activities block:

The last is worth noting for its phrasing: the pupil is to share their own opinion, and the opinion is specified in advance.

Two Balochistan items show the template is applied by form, not content. At p. 32, among the World War I Skills: “Justify that the Treaty of Sevres is based on fair justice” — though the volume’s own narrative at p. 40 records that the Ottoman state “was forced to sign” that treaty under pressure and that Mustafa Kemal “refused to accept” it. At p. 82, among the World War II Skills: “Justify that the British’ decision of settling Jews in Palestine was based on fair justice” — though the narrative at p. 89 says the opposite in the book’s own voice, “The British Policy of settling Jews in Palestine was unjust for the following reasons:”, followed by four itemised reasons. (The corresponding student exercise at p. 96 is neutral.)

This is the most diagnostic pair in the corpus. Two justify-that prompts have been generated for propositions the book itself refutes. That is only possible if the prompt form is a slot into which a chapter’s topic sentence is dropped, without anyone checking whether the proposition is the book’s own position, let alone whether it is arguable. The template is prior to the content. Where the content is the national narrative it converts it into something the pupil must argue for; where it is a colonial policy the book condemns it produces an instruction nobody can carry out. That indifference is why the form should be read as an instrument rather than a series of individual editorial choices.

Punjab’s History 8 runs the same architecture on the corpus’s most consequential proposition. Its Learning Activities block for the Sir Syed chapter directs: “Generate a class discussion/debate to conclude the reason behind the Two-Nation Theory of Sir Syed Ahmed Khan” (native text, f. 27). Conclude does precise work: the class debates, the conclusion is stipulated. The chapter’s Skills outcome reads “Analyze the reasons behind the two-nation theory of Sir Syed Ahmed Khan” (f. 16) — analysis of reasons behind a theory, not of whether it holds. The same move appears a year earlier in another subject: Punjab’s English 7, p. 40, asks in its while-reading margin box how the author of the text justifies that the Hindus and the Muslims belonged to two different nations. (The box is interleaved by the OCR with body text and is not quoted as a continuous string; the reassembled fragments are “How does the … author of the … text justify that … the Hindus and … the Muslims … belonged to two … different nations?”) A Punjab pupil therefore meets the justify-a-given-conclusion move on the central national proposition at class 7 in English; again at class 8 in History, as a debate whose conclusion is named; and meets the proposition asserted without argument at class 9 in Pakistan Studies, dated to “the arrival of Muslims in the subcontinent” (p. 7). Across three passes the claim moves from argued, to stipulated, to eternal, and the apparatus thins at each step. A subtler variant sits in History 7, p. 88: “Evaluate the causes of the decline of the Muslims in the subcontinent as identified by Shah Waliullah Dehlavi” — the verb is evaluate, the object is a named authority’s list, and the pupil evaluates within a frame they may not question.

Sindh, Islamiyat 9–10, p. 34 supplies the most doctrinally loaded instance, a Type 3 item wearing Type 5 clothing. The commentary on Al-Tawbah 33 concludes: “as Muslims, we should follow the Holy Prophet … hold fast to the principles of Islam, and strive for the domination of the true religion. This is the religious duty of everyone, and bliss of the world and hereafter lies in it.” Immediately beneath, headed “ACTIVITY FOR THE STUDENTS”:

“Hold a debate on the topic ‘Suggestions for the dominance of religion in the present age’. Students will write down the points after their debate.”

A debate has two sides; this one does not. The topic is not whether religious dominance should be sought but how, and pupils must produce a written record afterwards. It is also partly Type 4: “in the present age” instructs a transfer without naming a target.

5.2 Type 4 — unguided transfer

This is the article’s central artefact. Punjab’s English 9 opens with Unit 1, “The Saviour of Mankind”, a Seerah unit framing the mission universally — “Allah raised a Rasool among themselves to lift humanity from ignorance into the light of faith” (p. 6). Its comprehension set at p. 10 contains eight questions. The fourth:

“What parallels can you draw between the state of civilization described here and the other historical periods of decline and renewal?”

The item does several things at once. It presupposes the passage describes a “state of civilization” — treating the text’s evaluative characterisation of pre-Islamic Arabia as established fact rather than the author’s view. It asserts the existence of “other historical periods of decline and renewal” without naming one. It instructs the pupil to perform the transfer operation. And it marks them for doing it. No target is named; India is not mentioned; nothing is. The pupil, or the teacher, supplies the target.

The permissiveness is the point. An item that named a target would be Type 3 and would be visible to any content analysis, because the target would appear as a string in the book. This item is invisible to string search on any term an analyst would think to search. What it demands is the schema, not the instance — and a schema, once practised, is portable to instances the book never has to print.

The same page reinforces the evaluative half, in Question 8: “How does the author view the pre-Islamic Arabs? Write down the examples and the evidences supporting your inferences.” This is a well-formed comprehension item — it correctly locates the characterisation in the author rather than in the world, and asks for textual evidence — but it requires the pupil to write out the characterisation and marshal support for it, and the book repeats it near-verbatim as the first question of Review 1 (p. 87). The unit’s pre-reading prompts (p. 6) have already primed both halves: “Mention the condition of Arabia before Islam.” / “How did Islam change the lives of the people?”

The second transfer specimen is a negative one. KP’s English 6, p. 9, sets as pre-reading: “Can you tell about the condition of Arabia before Islam?” The reading passage that follows is a biographical sketch of the Prophet containing no jahiliyya material whatsoever; the question is unanswerable from the book. It requires the pupil to import the condition-of-Arabia narrative from outside the volume — in KP, from Urdu-medium Islamiyat, which this method cannot read. An item demanding content the book does not supply is a demand on the pupil’s wider curricular formation.

5.3 Types 1 and 2 — keyed items, and what a distractor set does

Recall and comprehension items are the corpus’s bulk. Their ideological significance lies in a property peculiar to closed-form assessment: a multiple-choice item does not merely test a proposition, it fixes its truth-value and prints its alternatives.

The sharpest instance is Punjab, Pakistan Studies 9, p. 18, where the stem is not neutral:

“In 1867, while exposing the enmity of Hindus against Muslims, Sir Syed Ahmad Khan had clearly declared: (a) Muslims and Hindus are separate nations. (b) Muslims should remain separate from politics. (c) Hindus are not our friends. (d) Muslims should learn English.”

The stem asserts “the enmity of Hindus against Muslims” as background fact before the pupil reaches the options; the keyed answer is (c). A pupil answering correctly writes the sentence “Hindus are not our friends” in an examination script, having first been told the enmity was exposed rather than alleged.

The same volume keys the Arabian moral frame nine pages after the narrative that supplies it:

“In Arab society, before the advent of Islam, in the age of ignorance, girl was: (a) burnt (b) vani (given in marriage) (c) buried alive (d) respected” (Punjab, Pakistan Studies 9, p. 107)

Distractor (a), burnt, is the only appearance of anything resembling widow-burning anywhere in the corpus, and it appears as a wrong answer about Arabia.

Distractor sets carry content of their own. Punjab’s Geography 6, pp. 39–40, describes equatorial and Amazonian peoples in colonial racial-ethnographic register (“Natives are black people… People are uncivilized”; “These people live in tribes and are known as Red Indians”). At p. 47 the material becomes examinable: “People living in Northern Canada are known as: a. Pathan b. Red Indians c. Eskimos d. African.” The distractor set places an ethnic group of Pakistan in a taxonomy of exotic peoples and requires the pupil to sort them correctly.

Sindh’s Islamiyat 6 converts a collective allegation into a true/false item. The narrative at p. 42, under the heading “Hostility of Jews and Banu Ghatfaan tribe to Islam”, reads: “The Jews inhabited in this town. They were always busy in anti-Islamic activities and conspiracies. They did not miss any opportunity to weaken and harm Islam.” At p. 46 the pupil must mark true or false: “The Jews of Khyber were always busy in anti-Islamic activities and conspiracies.” The keyed answer is true. This is a class-6 book, English-medium, and among the corpus’s most overlooked adversarial texts precisely because the allegation lies in an exercise rather than a passage.

Balochistan’s English 10 runs the Arabian frame through five assessment forms in one volume — material acquired after the project’s earlier thread analysis, which reported Balochistan as not examining the Arabian side at all. The unit narrative at p. 8 reads: “Before him, Arabs had indulged in all sorts of evils. They did not worship Almighty Allah but gods and goddesses of various kinds. They gambled and drank wine.” The apparatus then requires the proposition: as a Type 3 pre-reading prompt (p. 8, “Hazrat Muhammad-ur-Rasool Ullah Khatam-un-Nabiyeen’s … life is role model for all of us. Discuss it.”); as comprehension (p. 11, “What was the condition of Arabs before spread of Islam?”); as a keyed MCQ, twice (pp. 11 and 47, “Before Islam, Arabs had a. indulged in all sorts of evils. b. helped one another. c. worshipped Allah. d. were very civilized.”); as short answer (p. 13); and as a revision true/false item (p. 46): “Before Islam, the Russians had indulged in all sorts of evils.” The last deserves a moment. Its pedagogical function is to test attention to a single substituted word. Its incidental effect is to hold the predicate constant across two subject terms and require the pupil to affirm that it attaches to one of them.

5.4 The board-paper level, and production tasks

Two forms sit above the chapter exercise in authority.

The specimen board paper. Punjab’s English 10 prints both an examination blueprint and a full model paper. The blueprint (pp. 120–121), headed “ESSENTIAL INSTRUCTIONS FOR PAPER SETTERS”, specifies a 75-mark paper, its objective/subjective split, its timing and the bifurcation of question choice across units. The model paper follows; its short-answer section offers eight questions of which five are answered, and the first is “What was the social condition of women, orphans, and slaves in pre-Islamic Arabia?” (p. 123). The paper’s largest single mark allocation, fifteen marks, is: “Write an essay of 150-200 words on any ONE of the following topics: (15) i. A True Muslim ii. Health iii. Tolerance”. An English-language board examination whose highest-weighted item offers, as one of three topics, the composition of a normative account of what a true Muslim is, is an unusually clean instance of the argument. Nothing is asserted; a form of writing is required and remunerated.

The competitive production task. Punjab’s Pakistan Studies 9 closes its Ideological Basis chapter (p. 19) with: “Activity — Conduct a speech competition regarding the ideology of Pakistan. Instructions for Teachers — Explain to the students the concept of two nations in the subcontinent.” The pupil speaks the ideology aloud, in competition, and is ranked for it; the teacher is instructed to supply the content beforehand. Punjab’s Pakistan Studies 10 sets “Organize an essay competition with reference to Yaum-e-Takbir”, the anniversary of the 1998 nuclear tests. Sindh’s Islamiyat 8, p. 76, closing the lesson that canonises Muhammad bin Qasim inside a chapter of “Famous Personalities of Islam”, directs: “Write an essay on the personality of Muhammad bin Qasim and submit it to the teacher.” Punjab’s English readers use the device on the martial frame: “Write about the heroic deeds of an awardee of Nishan-e-Haider who sacrificed his life for the nation” (English 7, p. 36) and, in a Review section, “Why should we not cry over the death of a soldier?” (English 6, p. 128) — a question whose form forecloses the alternative.

Two Urdu-medium production blocks are recorded at structure level only. Punjab’s Urdu 8, pp. 95–102, closes a lesson on the status of women with four consecutive exercise pages: short and long answers, two concerning the pre-Islamic condition of women and of daughters (p. 98); an MCQ and fill-in-the-blanks whose blanks are those propositions (p. 99); a vocabulary list including the term for live burial (pp. 100–101); an essay task (p. 102). And Punjab’s Islamiyat 9, p. 111 — verified by direct image reading of the source PDF rather than by OCR — carries, in one exercise set on one printed page, multiple-choice items testing the Paigham-e-Pakistan counter-extremism doctrine and the distinction between jihad and fasad fil-ard, a short-answer item on Paigham-e-Pakistan, and an activity directing pupils to write into their diaries the names and deeds of the martyrs awarded the Nishan-e-Haider. The restraint doctrine and the military-decoration memorial are not merely in the same chapter; they are in the same exercise set on the same page.

6. Per-board assessment profiles

The table reports English-medium markers only, deduplicated, per 100,000 words. These are ordinal screening figures, not measurements.

Read alongside §3.3: the evaluation-verb column and the open-opinion column point in opposite directions, and it is the second that tracks the presence of Type 5 items.

Punjab (PCTB) — demand distributed across five subjects, at the highest authority level. Punjab’s regime is the corpus’s most extensive and most vertically integrated. It is the only board printing an examination blueprint and a specimen board paper, and the only board running the same ideological content through assessment in two media at once: at class 6 the Nishan-e-Haider roll is assessed in the English reader and the Urdu reader in the same year. Its distinctive form is the justify-a-given-conclusion prompt inside a “Skill: Activities” box; its distinctive placement is the English reader, which delivers the national narrative a year ahead of the History syllabus and does the justify-that work there. At class 9 the demands converge: in one year the pupil must draw parallels between pre-Islamic Arabia and other periods of decline and renewal, characterise the pre-Islamic Arabs with textual evidence twice, key “Hindus are not our friends” and “buried alive”, explain women’s rights in Islam from Qur’an and Sunnah, compete in a speech contest on the ideology of Pakistan, and — in Urdu-medium Islamiyat — answer items on the state’s counter-extremism doctrine while recording martyrs’ names in a diary. Punjab’s high evaluation-verb density is almost entirely Type 3; its one genuine Type 5 item about a Pakistani government, across twelve volumes, compares two terms of the same civilian prime minister (Pakistan Studies 10, f. 25).

Sindh (STBB) — the heaviest apparatus, and the only evaluative civics. Sindh’s architecture is by some distance the most elaborate: teacher-instruction blocks at 42.0 per 100,000 words against Punjab’s 6.5 and KP’s 0.1, and fill-in-the-blank items at roughly three times any other board. Its Islamiyat volumes attach a standing “Instructions for the Teachers” box to nearly every lesson, and the modal instruction is arrange a debate — on the rights of children (Islamiyat 6, p. 73), the requisites of prophethood (Islamiyat 7, p. 15), zakat (p. 27), the Prophet’s moral character (Islamiyat 8, p. 11), lawful earning (p. 46), respect for law (p. 53), the rights of orphans, widows, the disabled and travellers (p. 65). Almost all are debates on propositions with one admissible side. Against that, Sindh’s 2014/2015-curriculum Social Studies volumes carry the corpus’s only sustained Type 5 regime (§8), and its English 9 teaches two-sided debating as an explicit protocol. Sindh is also the only board whose doctrinal register peaks inside an assessment item rather than a passage.

Khyber Pakhtunkhwa (KPTBB) — the thinnest apparatus, and the widest gap between narration and demand. KP’s English-medium books carry effectively no printed teacher layer and the corpus’s lowest density of open-opinion prompts. Its assessment is overwhelmingly Types 1 and 2: the Pakistan Studies 9 chapter-end apparatus (pp. 96–98) is twenty recall MCQs on dates and offices, ten short answers, six long answers, and two activities. Yet KP’s Pakistan Studies volumes contain the corpus’s frankest political history — Zia’s total ban on expression and political activity, Ayub’s manipulation of the 1965 election, 1999 named as “the overthrow of the elected civilian government”, the 1971 military action having made “people of East Pakistan … enemies of the Army”. None of it is assessed as judgement: across both volumes, evaluate, assess, in your opinion and do you think return zero. KP tells the pupil, frankly, and asks the pupil nothing. Two exceptions sit in the English reader rather than the social-studies spine: the class-6 Democracy comprehension set — “What is the advantage of democratic form of government?”, “What is the disadvantage of monarchy?” (p. 54) — and “Compare Madina Charter with the UN Human Rights Charter” (English 9, p. 55). The largest caveat on KP is medium: Islamiyat, Mutalea-e-Quran, History, Geography and Urdu are taught only in Urdu, so most of its apparatus is invisible here.

Balochistan (BTBB). Its books are recent pilot editions printed for single academic years and explicitly soliciting revision feedback. Within those bounds: its History 8 is English-medium and quotable and runs the justify-that template as standing architecture, including the two prompts asking pupils to justify propositions the book refutes; its evaluation-verb density is the corpus’s highest and its open-opinion density the lowest, the sharpest instance of the divergence in §3.3; and its English 10 examines the Arabian moral frame five times in one volume. This last corrects the project’s earlier finding that Balochistan examines the Indian-side material and not the Arabian-side, which rested on the class-6 Urdu History volume alone before the class-10 English reader entered the corpus. Balochistan also supplies genuine counter-material: its English 7, p. 112, teaches the fact/opinion distinction with a worked local example, and its Geography 8 teacher blocks direct open debates, including “organize a debate among the students on the reasons why Pakistan’s exports are less than other South Asian countries” (p. 59).

7. The asymmetry, and what it means

Across the whole English-medium corpus, twenty distinct assessed items require a pupil to state something about the moral or social condition of pre-Islamic Arabia, distributed across all four boards:

One is a printed board model paper; two are questions repeated in review sections; three are keyed multiple-choice or true/false items. To these must be added, at structure level only, Punjab’s four-page Urdu exercise block at class 8.

Zero assessed items anywhere in the corpus require a pupil to state that pre-Islamic or Hindu India was barbarous, that Indians killed girls, that widows were burned, or that Muslim rule ended any named Indian social evil. The single Indian-side assessed item found, in Balochistan’s Urdu-medium History 6 (p. 53), asks for the structure of the caste system and the origin of the Aryans, descriptively, and draws no comparison to Arabia.

The asymmetry is near-total, and where broken it is broken in the direction that weakens any equivalence hypothesis. That deserves emphasis, because the obvious reading of a heavily assessed jahiliyya narrative in a Pakistani curriculum is that it exists to be transferred to the subcontinental past. The corpus does not support that reading: the relevant Indian-side terms are simply absent. Sati does not occur genuinely anywhere in the corpus in either medium; no English-medium book attributes female infanticide or the killing of girls to India, Hindus, Aryans or the pre-Islamic subcontinent.

Where the rescue template does get a named beneficiary and does get examined, the beneficiary is Europe. Sindh’s Islamiyat 6, p. 96, states that “before the arrival of Muslims in Europe, the people of Europe were in darkness and ignorance”, and the following page keys it: “Before the arrival of Muslims, the people of Europe were: (A) highly educated (B) in ignorance. (C) civilized (D) under developing” (p. 97). This is the template’s one examined application to a named region in the whole English-medium corpus, and India is not that region.

Three readings of the asymmetry are available and the evidence does not fully separate them. First, that the demand does only what it appears to: it teaches the standard Islamic historical narrative of the jahiliyya, which requires the pre-Islamic Arabian condition to be characterised negatively, and examines it as it would any other unit. Second, that the schema is being installed rather than the instance — that what twenty assessed items across five years and four boards practise is not a fact about Arabia but a form of historical explanation in which a named population is morally deficient until rescued; on that reading Punjab’s Q4 is not an anomaly but the schema’s explicit statement. Third, that the asymmetry is an artefact of visibility, since twelve of the corpus’s fifteen Islamiyat volumes are Urdu-medium.

The honest position is that the second and third are both live and the first is insufficient alone, because it does not explain Q4. What can be said without qualification is narrower and still substantive: the Arabian moral frame is required of pupils, repeatedly, in two languages, in four subjects, in four provinces, up to and including board-examination level; and the corresponding Indian-side proposition is not required of anybody.

8. Counter-evidence

An argument about assessment demand is worth making only if it survives the assessment that runs the other way. There is a great deal of it, and one board’s is strong enough to require argument rather than acknowledgement.

8.1 Sindh’s evaluative civics — the corpus’s strongest counter-case

Sindh’s Social Studies 6 and 7, written to the Sindh Curriculum of 2014 and 2015, do not merely contain the occasional open question; they run a complete assessment architecture built on it. Each chapter closes with standing headed blocks — “Answer the Following Questions”, “Conduct an Inquiry” / “Use Your Inquiry/Investigation Skills”, “Cooperate With Others”, “Be Creative”, “Become Technologically Smart” — and the inquiry and cooperation blocks routinely require the pupil to gather evidence from outside the book and reach a conclusion the book does not supply. The clearest instance is a demand to rate the sitting government:

“Work together in groups of four, use a scale to assess how well the current government of Pakistan is fulfilling its purpose. Suggest three things they could do to better fulfill the purposes of government.” (Sindh, Social Studies 6, p. 64)

Nothing else in the corpus resembles this. It is a Type 5 item with a live political object, set to eleven-year-olds, requiring a scaled judgement and a remedial proposal. It is reinforced rather than isolated:

“Make an inquiry to find out who is the MPA and MNA from the area in which you live. Find out which political party they belong to. Find out what is the position of their political party on education.” (Social Studies 6, p. 85)

“Make an inquiry to find out the problems women face in casting their votes and contesting the elections. Suggest five things that can be done to deal with these problems.” (Social Studies 6, p. 81)

A third item sets pupils to “Find out three Local level issues which have not yet been resolved by the Local Government” (p. 78). At class 7 the regime moves from machinery to instruments and makes the audit of rights an examinable outcome: “Choose a right or freedom and assess the extent to which it is met in Pakistan” and “Explain how rights and freedom given in the constitution are ensured and denied through citing everyday examples” (Social Studies 7, p. 46, chapter SLOs); “Survey your neighborhood and find out how many children in your area are enjoying right to free education?” (p. 51). At matriculation, in an otherwise obsolete 2004 volume, the appended rights chapters carry: “Critically evaluate the State of human rights in Pakistan and suggest measures to safeguard them” (Pakistan Studies 9–10, p. 186). Note what this does that a Type 3 item cannot: it names the object of criticism as the pupil’s own state, does not supply the verdict, and requires a remedial proposal. It is the exact structural inverse of “Justify that…”.

8.2 Sindh’s assessment of historical method

More consequential still, Sindh’s Social Studies 6 assesses historiography itself — and assesses precisely the proposition that a single official version is a problem. The book states criteria in its own voice at p. 9: history “must be: As accurate as possible. Based on evidence and logical thought. Not have falsehoods or political ideas.” It then sets, at p. 11, a group exercise in which pupils record everything they can recall of their first day in class 6, compare accounts in groups, and answer: “Does each account add different details to the story? Do some of them contradict one another? How do the accounts differ? How would you decide what to include and what to exclude?” The attached teacher’s note is the single most striking item in the corpus:

“Teacher’s Note: Help students to understand that history is always incomplete. Histories always involve interpretation and that the inclusion of diverse perspectives, adds to their completeness by including details that may otherwise have been overlooked. Ask them, what if the version told by students of Class X, was declared to be the only ‘true’ and officially accepted story?”

A provincial board directs the teacher to ask eleven-year-olds to imagine the consequences of an officially declared single true version of events. Set against the 1995 curriculum directive that the Ideology of Pakistan “be never subjected to discussion or dispute”, the two documents are answering each other.

The same book closes with an artefact-interpretation exercise in which pupils must adjudicate competing readings of an unknown object and are asked “Do you agree with any of these arguments? What do you think the cord was used for?” (p. 231); its class-7 companion runs a bagged-artefact archaeology exercise whose stated lesson is evidential underdetermination (pp. 179–180).

8.3 Two-sided argument taught as a protocol

Sindh’s English 9 teaches debating as a skill with an explicitly symmetrical procedure. Pupils are divided into pairs within groups; each group receives a topic; “One sub-group will prepare points in favour of the topic and the other against it”; the groups then swap: “Now, group B will debate in favour of the topic and group A against it” (pp. 193–194). The suggested topics are deliberately low-stakes — wedding guest lists, automatic promotion. The pedagogical content is the form: that a proposition has two arguable sides and the pupil must be able to occupy either.

This matters more than its modest content suggests. Every “arrange a debate on…” in Sindh’s Islamiyat volumes and every “generate a class debate to conclude…” in Punjab’s and Balochistan’s History volumes borrows the word debate while removing exactly the property Sindh’s English 9 teaches. The corpus contains both the genuine form and its hollowed-out homonym, which makes the distinction a finding rather than a category imposed by the analyst.

8.4 Critical apparatus applied to national material

Punjab’s English 9 supplies the most interesting internal counter-case, because it applies critical apparatus to its own patriotic content. The Patriotism unit delivers the full eleven-name Nishan-e-Haider roll (p. 23), then at p. 26 sets: “Read the following statements and identify whether each is a fact or an opinion. 1. Patriotism is a commendable quality. a. fact b. opinion” — and, immediately after, “identify whether each is a generalized statement or an evidence-based information. 1. Many soldiers showed great bravery during the wars of 1965 and 1971.” A pupil who answers the first item correctly classifies the proposition “patriotism is a commendable quality” as an opinion. Whatever the unit’s affective intent, the item requires the pupil to place the unit’s own governing claim outside the class of facts. Balochistan’s English 7, p. 112, does the same work on local material.

8.5 Assessment aimed at prejudice itself

Sindh’s Social Studies 6 media unit teaches stereotyping as a named fallacy and assesses it. It defines gender stereotypes (“Men are the ‘bread winners’… Women are ‘housewives’…”), cultural stereotypes (“Girls are not good at sports”), and — uniquely in this corpus — intra-Pakistani ethnic stereotypes, anonymised: “‘A’ ethnic group is more intelligent, hard worker and loyal to country and religious than the group ‘B’.” The demands follow: “List the gender stereotypes promoted by the media in Pakistan. Give examples from television dramas and advertisements” (p. 101); “Define the term ‘stereotyping’ and its forms in Sindh. Give three example of how women are stereotyped in Sindh. What is propaganda? How can it serve to help or harm a cause?” (p. 103). Nothing comparable exists in Punjab, KP or Balochistan at any grade.

8.6 The weight to give all this

Three qualifications keep the counter-evidence in proportion. Sindh’s evaluative regime is confined to two volumes at classes 6 and 7 and to an appended block in a 2004 matriculation text; the pupil taught at class 6 how to rate a government meets, at class 8, a reprinted federal volume with no civics at all; and Sindh’s own Islamiyat volumes run the corpus’s most doctrinally assertive assessment item in the same years.

One point survives all three: the counter-evidence cannot be dismissed as an artefact of recency or reform. Sindh’s evaluative material is the older material. Its Social Studies volumes are a curriculum generation behind Punjab’s, and the newest social-studies books in the corpus have the thinnest evaluative assessment. Whatever produced the difference was not the passage of time — and it demonstrates that the alternative regime is not hypothetical. A Pakistani provincial board wrote it, printed it, and is still issuing it.

9. Implications

What is cultivated by requiring production rather than offering a reading. The distinction between a passage and an exercise is a distinction in what the pupil practises. A passage practises reading. A Type 3 item practises the construction of supporting reasons for a conclusion supplied by authority; a Type 4 item, the extension of an evaluative schema to unnamed targets; a Type 5 item, judgement under uncertainty. These are different capacities, and there is no reason to expect that practising the second and third produces the fourth. That is the substance of the claim, and it is a claim about capacity, not belief: a pupil who has spent five years writing reasons for conclusions they were given has practised a specific operation many hundreds of times, and a pupil never once asked to reach a conclusion the book does not hold has not practised the corresponding one at all.

What makes this more than an observation about rote learning is the selectivity. The corpus is not uniformly closed. Punjab’s History 6 teaches primary and secondary sources as a formal lesson at p. 5 — and withdraws the source apparatus entirely for the one chapter, on the rise of Islamic civilisation, where it might be applied. Punjab’s English 9 teaches fact-versus-opinion and applies it to patriotism but not to the ideology. Balochistan’s History 8 asks pupils to “Draw a conclusion of who was the actual winner of World War-1” — genuinely open — on the same page as “Justify that the Treaty of Sevres is based on fair justice.” The tools of inquiry are present in these books and are switched off over designated content. That selectivity is the finding, and it is not explicable as pedagogical backwardness, because the same authors demonstrably possess the alternative.

The compounding effect of a frame met in five subjects in the same years. Repetition across subjects is qualitatively different from repetition within one. A claim repeated inside a History book is a claim History makes. The same claim met in the English reader, the Urdu reader, Islamiyat, Social Studies and Pakistan Studies is not the position of a subject; it is the shape of the world as the school describes it, and there is no subject from which to view it. Punjab at class 9 is the clearest case: in a single year the pupil is assessed on the Arabian moral frame in English, on the Two-Nation Theory and the “enmity of Hindus” in Pakistan Studies, on the state’s counter-extremism doctrine and martyr memorialisation in Islamiyat, and on the APS Peshawar attack in Urdu. None is in dialogue with the others. The corpus contains no cross-reference anywhere in which one subject notes that another treats the same material differently — including in the several cases where two volumes of the same board flatly contradict each other.

Inquiry vocabulary collapses at matriculation, exactly when pupils can handle it. The corpus’s most robust vertical finding is that the apparatus of inquiry does not grow with the pupil; it disappears. Scored against a common rubric, inquiry-versus-received-truth means run about 3.0 across classes 6–8 and about 4.5 across classes 9–10, in every board and at every vintage. Inquiry-verb density collapses correspondingly: Punjab’s History 6 runs at 10.1 occurrences per 10,000 words, Punjab’s Pakistan Studies 10 at 0.49, KP’s Pakistan Studies 9 at 0.76, Sindh’s Pakistan Studies 9–10 at 1.1. Punjab’s Pakistan Studies 9 is a 2023 book with the same near-zero density as KP’s book on a 2006 curriculum: two boards that disagree about almost everything else agree exactly here.

This inverts what a developmental account would predict. The apparatus is richest at eleven and twelve, when pupils are least equipped to use it, and is withdrawn at fourteen and fifteen, when they are most equipped — and it is withdrawn precisely as the syllabus arrives at the content most in need of it: the ideology, the Two-Nation Theory, 1971, the wars. The plain structural explanation — that classes 9 and 10 are board-examination years and board examinations reward determinate answers — does not make the effect smaller. It makes it a property of the assessment system rather than of the books, which is the more consequential place for it to live.

The intergenerational argument, stated carefully. An assessment regime is a selection mechanism: it ranks pupils, and the ranking determines who proceeds. A regime composed predominantly of Types 1 to 3 ranks pupils by fidelity in reproducing supplied conclusions and fluency in constructing reasons for them. It does not rank them by capacity to reach conclusions the syllabus does not hold, because it never elicits that capacity and therefore cannot observe it. Whatever distribution of that capacity exists in a cohort, the examination is blind to it — neither rewarding nor penalising it, which over a career of examinations amounts to selecting against it relative to the traits the system does reward. The pupils so ranked become, in some proportion, the next generation’s teachers, examiners, curriculum officers and textbook authors. Aziz’s examiner in 1989 was himself the product of an earlier iteration; on Aziz’s account the girl could not debate the point not because of the examiner’s private conviction but because the board had settled it. The 1984, 1986 and 1995 curriculum directives were written by people who had themselves been schooled and examined.

What may be claimed: that a regime of this composition does not select for evaluative capacity; that the personnel staffing the next cycle are drawn from those it does select; and that this is a mechanism by which an assessment culture can reproduce itself without anyone deciding that it should. What may not be claimed, and is not claimed here: that any individual teacher, examiner or author holds any particular view; that pupils schooled this way are incapable of evaluation; that the regime determines outcomes rather than shaping incentives; or that a single mechanism explains an educational culture in which examination boards, teacher training, class sizes, language policy and political economy all operate simultaneously. The corpus is evidence about textbooks; it is suggestive about a reproduction mechanism and is not proof of one. The mechanism is also not sealed: Sindh printed a teacher’s note asking pupils to imagine an officially declared single true version and reject it, and that note is in a book currently in classrooms.

What follows for anyone revising these books. Two consequences are direct and cheap. First, the demand side is where a small edit has a large effect: changing “Justify that Indus Valley people did not learn the warfare nor developed their trade, and therefore, were easily defeated by Aryans” to “Assess the evidence for each of the three explanations given above” converts a Type 3 item into a Type 5 item on the same content, in one line, with the surrounding narrative already supporting it. Balochistan’s History 8 demonstrates that the template is applied mechanically; a template can be changed mechanically. Second, any audit that reads only passages will under-report. The three most consequential ideological items located here — the transfer question at Punjab English 9 p. 10, the keyed distractor at Punjab Pakistan Studies 9 p. 18, and the true/false item at Sindh Islamiyat 6 p. 46 — are all in exercises, and two are invisible to the keyword searches content analysis performs.

10. Limits

This method cannot show what any pupil believes. Nothing here is evidence about the mental state of any child. Reading is not determined by text, and neither is answering. A pupil may write the keyed answer without holding it, may hold it without having been taught it, or may have been taught its opposite by a teacher. The claim throughout is about what the assessment regime requires, not about what it achieves.

It observes demand, not uptake. Whether a given exercise is set, marked, or reached at all in a school year is unobserved. These books routinely carry more exercise material than any class could complete.

Classroom practice, actual examination papers and teacher behaviour are entirely unobserved. The one specimen board paper in the corpus is a specimen printed by the board in its own textbook, not a paper that was sat. No real script has been consulted, no classroom visited, no teacher interviewed. A teacher may foreclose a transfer the book licenses, or draw one it does not. Aziz’s 1989 examiner is reported at second hand from a published critique and is offered as evidence about a documented practice at a documented moment, not about current marking.

Urdu-medium under-observation biases the count downward. Forty-five readable files are Urdu-medium and none is quoted; two of four boards teach Islamiyat only in Urdu and two teach History only in Urdu. The twenty items in §7 are a floor. Conversely the absence claims are correspondingly weaker in Urdu than in English and are reported as “not established” rather than “absent” for Urdu files throughout.

Absence claims are bounded by extraction quality. For OCR’d scans a term not found may be a term not recognised. Nastaliq extraction breaks ligatures unpredictably, so a zero token count is not evidence of absence even for high-frequency proper nouns.

Vintage confounds board comparison. Four curriculum generations are in simultaneous use and Punjab cannot be vintage-matched against Sindh or KP in most subjects because the matching books do not exist. The honest form of most board statements is not “Punjab is more X than Sindh” but “Punjab’s current books are more X than the books Sindh is currently printing”.

Balochistan’s profile is provisional, for the reasons in §6: partial acquisition, two zero-byte extractions, at least one volume processed with the wrong language model, and pilot-edition status.

The typology is the analyst’s, not the books’. No textbook here labels its items by cognitive demand. Assigning a given item to Type 3 rather than Type 2, or Type 4 rather than Type 5, is a judgement. The judgements most exposed to challenge are those where an item’s verb and its epistemic structure disagree — and those are, by construction, the ones the argument turns on. They are quoted in full so a reader may re-sort them.

11. References

Primary corpus. 112 catalogued textbooks of the Punjab Curriculum and Textbook Board, Sindh Textbook Board, Khyber Pakhtunkhwa Textbook Board and Balochistan Textbook Board, classes 6–10, held as PDF scans with extracted text at extracted-text/{pctb-punjab,stbb-sindh,kptbb-kp,btbb-balochistan}/class-{6..10}/ and catalogued in analysis/books.json. All quotations are from English-medium files and were re-verified by exact substring match. Page references are PDF page indices except where marked f. for printed folio.

Project analytical documents relied on for scoring, vintage control and prior verification, and not re-derived here: the rubric; the synthesis; the edition audit; [thread moral rescue](/studies/thread-moral-rescue/); [thread iconoclasm](/studies/thread-iconoclasm/); [lens religion](/studies/lens-religion/), lens civics, lens history, lens patriotism, lens othering, lens gender.

Aziz, K.K. The Murder of History: A Critique of History Textbooks Used in Pakistan. Lahore: Vanguard, 1993. ISBN 969-402-126-X. The passage quoted in §2 is at p. 189; Aziz’s examination is of the March 1989 Punjab Textbook Board reprints. Quoted here directly from Aziz.

Nayyar, A.H., and Ahmed Salim (eds). The Subtle Subversion: The State of Curricula and Textbooks in Pakistan. Islamabad: Sustainable Development Policy Institute, 2003. Full text: https://eacpe.org/content/uploads/2014/05/The-Subtle-Subversion.pdf

The following federal curriculum documents are quoted in §3.4 at second hand, from Nayyar and Salim (2003), and have not been consulted in the original by the present author: Social Studies Curriculum, Classes VI–VIII (National Bureau of Curriculum and Textbooks, Islamabad, 1984), p. 16; Urdu Curriculum, Classes VI–VIII (NBCT, 1986), p. 13, and Urdu Curriculum, Classes IX–X (1988), p. 8; Primary Curriculum (1995), p. 41.

Concepts referred to descriptively and without attribution, because the intention is to invoke a well-established idea rather than a particular formulation of it: examination backwash on teaching and learning; the consolidating effect of retrieval and production relative to re-reading; curriculum–assessment alignment; and the hidden curriculum. No specific work is cited for these because none was consulted in preparing this article.