Thread: Iconoclasm — is there an equivalence chain from Makkan idols to Hindu India?

Research question. Do these textbooks construct — explicitly, or by shared vocabulary, or by structural juxtaposition — a chain linking (a) the mushrikeen of Makkah and the destruction of the Kaaba idols; (b) Ibrahim smashing his father’s idols; (c) Muslim conquerors destroying idols and temples in Hindustan; (d) Hindu religious practice, and by extension contemporary Hindus and India?

Corpus. extracted-text/{pctb-punjab,kptbb-kp,stbb-sindh}/class-{6..10}/. All quotations below are from English-medium files and have been re-verified by exact string match against the source. Urdu-medium files are reported at structure/topic level only, at low confidence, and are never quoted.

1. Answer

No board reaches Tier 1. Nowhere in the English-medium corpus does a sentence, caption, exercise or learning outcome state that Hindu idol-worship is like Makkan idolatry, or frame a subcontinental conquest as continuing the Prophet’s or Ibrahim’s iconoclasm. The inference the question worries about is never asserted.

Sindh reaches Tier 2, fully and tightly. The Sindh board is the only one that teaches Islamiyat in English, and its English Islamiyat volumes apply one evaluative pair — monotheism against idolatry / polytheism — to Makkan paganism and to pre-Islamic Sindh inside a single volume. Islamiyat 8 says of Arabia that “the sun of monotheism began to shine on the Arabian Peninsula, and idolatry came to an end” (PAGE 13), and, sixty-two pages later, of Muhammad bin Qasim’s conquest, that “People of Sindh became aware of monotheism instead of idolatry and superstition” (PAGE 75). The same word does the same evaluative work on both sides of the chain. Class 7 Islamiyat had already generalised the mission — “to eliminate disbelief and polytheism” — beyond Arabia (PAGE 47), and class 10 Pakistan Studies applies “Polytheism” to a target in the subcontinent, coupled with Jihad (PAGE 15). Class 8 Social Studies then supplies the contemporary node: Hindu idol-worship as an imposition on Muslims in 1937 (PAGE 57), and an explicit claim that “the present behaviour of Indian Hindus” continues that of 150 years ago (PAGE 52). Every link of the a→c→d chain exists in Sindh’s own books in shared vocabulary, across classes 6 to 10.

Punjab reaches Tier 3 only. Punjab History 6 restricts the idol lexicon strictly to Makkah — “idols”, “false gods”, “infidels of Makkah” — and gives Vedic religion a different, milder register. (The Vedic passage sits in a two-column layout that the OCR interleaves with an image caption, so it is quotable only as verbatim fragments: “They started worshipping anything” / “asagod.” on the following line, and “In their religious books Vedas,” / “hundreds of gods are mentioned.” — pctbhistory-6-en · PAGE 65.) No shared word. But the two sit in consecutive chapters of one book: Chapter 3 (Aryans, PAGE 59 ff.) and Chapter 4 (Rise of Islamic Civilization, PAGE 81 ff.), with the Vedic passage at PAGE 65 and the Kaaba idols at PAGE 84 — nineteen pages apart in one school year. That is juxtaposition without lexical linkage. Punjab then actively removes the middle term: Mahmud of Ghazni’s seventeen invasions are given purely political motives and Somnath is omitted (History 7, PAGE 60), and Muhammad bin Qasim is narrated with no idol or temple vocabulary at all (History 7, PAGES 31–32).

KP: not establishable in English medium, but it is the only board holding both (b) and (c). KP teaches Islamiyat, History, Geography and Urdu only in Urdu; its English-medium books contain no element of the chain, and KP English 7 gives Mahmud purely geographically. In Urdu, however, KP is the only board where a pupil meets both the Ibrahim idol narrative — class 9 Quranic studies, mutalea-quran-9-ur, via Surah Maryam (pp.17–29, including an explicit note on Azar at p28) and Surah Al-Anbiya (pp.41–56, with a p55 exercise asking what Ibrahim explained to his people about but-parasti) — and the Somnath raid (History 6, pp.53–56). The two are three years and two subjects apart, both theme-grade, neither quotable.

Balochistan reaches Tier 4 — not established — and it is the cleanest negative of the four. Its English 10 carries the strongest Makkan iconoclasm passage in the corpus — the idols “destroyed and Muslims started offering their prayers in the ‘Harem-e-Kaaba’” (english-10-en · PAGE 10) — while Somnath returns zero occurrences across all sixteen of its English-medium books, and Mahmud Ghaznavi appears not as a temple-destroyer but as a patron of Pashto letters: “During the reign of Mahmood Ghaznavi, a man named Saifullah, formally devised Pashto alphabets” (pakistan-studies-10-en · PAGE 105). The only charged use of “temple” in the whole board runs the other way — India “announced to build Ram Temple on the site of Babri Mosque” (PAGE 18), a claim about Hindu iconoclasm against a mosque. See §3 for the full probe.

Element (b) — Ibrahim and the idols — is absent from the entire English-medium corpus, and present in exactly one book in the corpus. In Sindh’s four English Islamiyat volumes Ibrahim appears only as the Hajj exemplar and as the Prophet’s infant son: no Azar, no Nimrod, no idol-smashing. Across all eight Urdu Islamiyat volumes (KP and Punjab, classes 6–9) there is no dedicated Ibrahim lesson at all — he surfaces only inside Hajj, prophet-list and suhuf lessons. The narrative exists in one place only: KP’s class-9 Quranic studies volume. So the premise that pupils meet Ibrahim-as-iconoclast as a recurring sacred exemplar is false for every board except KP, and in KP it is a single class-9 encounter.

A structural finding that cuts hard against the chain: the Conquest of Makkah is not established in any of the eight Urdu Islamiyat volumes. The token فتح occurs eleven times across all eight files and every instance is a false positive; there is no Fath Makkah chapter in Punjab or KP Islamiyat at classes 6–9, and the ghazwat taught are Badr, Uhud, Khandaq, Khaybar, Hunayn and Tabuk. Element (a) as an act of iconoclasm is therefore taught in only one board’s religion syllabus in this corpus — Sindh’s, in English, at class 7.

2. Tiered evidence table

Tier 1 — Explicit linkage

Tier 2 — Shared vocabulary

Tier 3 — Structural juxtaposition

Tier 4 — Not established / absent

3. Per-board assessment and the class-6→10 sequence

Sindh — highest tier reached: Tier 2 (at maximum strength)

Sindh is the only board where a pupil meets every surviving element of the chain in English, in books they are examined on. The sequence:

The materials for the chain are all present. What is missing is the sentence that joins them, and — importantly — the middle term of destruction: no Sindh book describes any temple or idol in Hindustan being destroyed. bin Qasim’s conquest is narrated as replacing idolatry by preaching and example, not by smashing.

Punjab — highest tier reached: Tier 3

The within-one-book adjacency flagged as the priority test is real and tight (History 6, Chapters 3 and 4, PAGES 59–101). But it is only adjacency. Punjab’s authors reserve “idol”, “false god” and “infidel” for Makkah and give the Aryans a separate, weaker register. And Punjab then breaks the chain at the point where it would matter most: it removes Somnath, removes the idol-breaker epithet, and narrates bin Qasim with no religious-destruction vocabulary at all. Punjab’s counter-evidence is stronger than its evidence.

The one genuinely religious motive Punjab gives a conqueror is Ghori’s, not Mahmud’s: “Being a Muslim, he might like to conquer the Hindus of India and spread Islam in that country” (History 7, PAGE 61). That names Hindus as the object of conquest, but says nothing about idols.

KP — highest tier reached: Tier 3, at low-to-moderate confidence; Tier 2 not establishable

KP’s English-medium books contain none of the chain, so nothing here is quotable and no Tier 2 finding can be made to the standard applied to Punjab and Sindh. What the Urdu structural survey does establish is that KP is the only board in the corpus that holds both element (b) and element (c):

Two qualifications keep this at Tier 3. First, the two nodes are three years, two subjects and two book-series apart, and nothing in either text refers to the other. Second, KP’s own Islamiyat volumes supply neither the Conquest of Makkah nor a dedicated Ibrahim lesson, and its History 7 and 8 do not cover the conquest period at all — History 7 opens with Babur, History 8 with Plassey. So the KP pupil’s encounter with iconoclasm is not a sequence at all; it is two isolated encounters in unrelated books. The idol-smashing pericope itself (Al-Anbiya 21:51–75) falls in an OCR-blank block, pp.46–53, and is inferred from the p55 exercise rather than read.

KP’s position therefore cannot be settled at this analysis’s evidentiary standard. That is a limitation of the corpus, not a finding about KP.

Balochistan — highest tier reached: Tier 4, not established

Probed on the same terms across all sixteen of its English-medium volumes.

The Makkan half of the chain is present, and it is the strongest single statement of it in the corpus. Balochistan’s English 10 — a language reader, not a religious-studies book — narrates the destruction of the idols and the cleansing of the Kaaba directly: the Prophet “asked the people to worship one and the only Allah and to give up worshipping idols” (english-10-en · PAGE 9); the Makkans who resisted are “the kafirs of Makkah” (PAGE 9); and the idols are “destroyed and Muslims started offering their prayers in the ‘Harem-e-Kaaba’” (PAGE 10).

The Indian half is absent, and absent in an unusually clean way.

Two of those rows carry the finding.

Mahmud Ghaznavi appears in Balochistan’s corpus as a patron of literature, not as an iconoclast. His single substantive appearance is in a Pashto literary-history passage: “During the reign of Mahmood Ghaznavi, a man named Saifullah, formally devised Pashto alphabets which are in use till today” (pakistan-studies-10-en · PAGE 105). The figure who carries the temple-destruction narrative in most South Asian textbook traditions is present in this corpus and is doing something else entirely.

The word “temple” reaches Balochistan’s pupils through contemporary politics, not medieval history. Its only charged occurrence is the Babri Mosque dispute: India “announced to build Ram Temple on the site of Babri Mosque” (pakistan-studies-10-en · PAGE 18). That is a claim about Hindu iconoclasm directed at a mosque in the 1990s — the inverse of the chain this thread was testing for, and it is worth recording that the reversal appears where the original does not.

Verdict: Tier 4. The Makkan iconoclasm narrative is taught, the Indian-side counterpart is not constructed, and no vocabulary bridges them. Balochistan does not weaken the corpus-wide negative; it strengthens it, and it removes the objection that the finding rested on an incomplete corpus.

One caveat, stated at full strength. Twelve of Balochistan’s twenty-eight books are Urdu-medium and were read at theme grade only. تاریخ and ہندو clusters were checked structurally in History 6, History 7 and History 8 (Urdu), and no Somnath or Ghaznavid conquest chapter is legible in their structure — but this is a “not established”, not an “absent”, and it covers precisely the volumes where such material would most plausibly sit.

4. Counter-evidence

Given equal weight. Several of these are stronger than anything on the positive side.

1. Punjab omits Somnath and the idol-breaker epithet entirely. Mahmud’s seventeen invasions are motivated by Central Asian power politics and dynastic expansion (History 7, PAGE 60). The single most iconoclasm-loaded episode in subcontinental Muslim history is simply not taught in Punjab. Confirmed in [track social studies](/studies/track-social-studies/); re-verified here.

2. Punjab History 6 attributes the judgment rather than asserting it. The wording is “pilgrims pouring in from all corners of the Arabian Peninsula to worship these idols, whom Hazrat Muhammad … considered false gods” (PAGE 84). “False gods” is placed inside the Prophet’s view, in reported speech, not in the textbook’s own voice. Compare Sindh English 9, which asserts it directly (“asked the Quraish not to worship their false gods”). A small authorial choice, but it is the difference between narrating a belief and endorsing a verdict, and it runs against the chain.

3. Punjab’s Muhammad bin Qasim carries no idol or temple vocabulary at all, and the text volunteers that “It is noted that Arabs did not force the locals to embrace Islam” (History 7, PAGE 32). Conversion is attributed to “the equality, caste-less system and social justice of Muslims”, and the converts are named as “Many Buddhists and lower-caste Hindus” — a caste explanation, not a theological one.

4. Sindh Islamiyat 8 grants Hindus temple-building rights in the same paragraph-block as the “idolatry” line: “He gave religious freedom to the Hindus and gave them lands to build temples there” (PAGE 75). The book that supplies the corpus’s tightest lexical link also supplies an explicit endorsement of Hindu worship rights under Muslim rule.

5. Sindh Islamiyat 6 drills, as an examinable item repeated four times, that polytheists must be treated well. “Even if the parents are polytheists, they are commanded to be treated well” (PAGE 68), then again as a multiple-choice question, a fill-in-the-blank, a true/false item and a matching item (PAGE 70). The same word that could carry the chain is used, in the same volume, to teach an obligation of kindness toward those it names.

6. Sindh Social Studies 7 treats Hinduism neutrally and positively: “Under the rule of Gupta kings, the people returned to their Hindu roots and Hinduism became the religion of the state” — followed by the golden-age framing of arts, science and lowered taxes (stbbsocial-studies-7-en · PAGE 12). Elsewhere the same book lists “the religions like Hinduism, Islam and Christianity” flatly, without evaluation (PAGE 79).

7. Sindh Social Studies 6 describes a Hindu deity descriptively, without evaluation: of an Indus Valley seal, “This god is a bit like the Hindu god Shiva (who also has three faces)” (PAGE 35).

8. Punjab History 6 calls the Gupta period “the Golden Age of Hindu rulers” — and sets it as a learning outcome to examine why (PAGE 77). Same book, four pages before the Islamic Civilization chapter opens.

9. Jinnah’s 11 August speech appears in all three boards and protects temples by name. “You are free to go to your temples” or near-identical wording appears in kptbbenglish-8-en · PAGE 65, kptbbpakistan-studies-9-en · PAGE 11, kptbbpakistan-studies-10-en · PAGE 125, stbbenglish-9-en · PAGES 87 and 93, and pakistan-studies-10-en (this file is a native text extraction and carries no page markers; the passage is at line 7369). A pupil in every board meets a foundational national text that names temples as protected.

10. Punjab Pakistan Studies 10 lists Katas Raj Temple, Tilla Jogian, Nankana Sahib and Kartarpur Sahib as national religious tourism heritage, unremarked, in a list that also contains Badshahi Masjid (pakistan-studies-10-en, line 6604; no page markers in this file).

11. No English-medium textbook in the corpus narrates the destruction of a temple in Hindustan. Not one. The (c) node of the chain — iconoclasm as heroic history — is, in English-medium schooling, essentially missing.

12. Element (b) is absent. Ibrahim is taught as the Hajj exemplar, never as an iconoclast. The question’s assumption that a pupil meets idol-smashing twice — once as prophetic exemplar in Ibrahim, once in Muhammad — does not hold.

13. KP English 9 names idol-worshippers as a protected community, not an enemy. In its Charter of Medina lesson: “The other communities living in Medina were the idol worship…” — the surrounding sentence, though badly OCR-corrupted, turns on “harmony” and the Prophet’s charter (kptbbenglish-9-en · PAGE 50; confidence medium, OCR degraded, only the verified fragment quoted). This is the corpus’s one English-medium instance of idol-worshippers appearing as neighbours under a compact rather than as adversaries.

14. Sindh Islamiyat 8’s own reform narrative points inward, not at Hindus. Shah Waliullah’s mission is framed against Muslims’ own decline: “Bad customs and foul beliefs were becoming part of Islam” (stbbislamiyat-8-en · PAGE 83). Purification is directed at the Muslim community, not at neighbours.

15. The jahiliyya frame is never transferred to India. “The age of ignorance” / “days of ignorance” is the classic vehicle for exactly the equivalence this thread tests — it would let a text label pre-Islamic Hindustan with the term reserved for pre-Islamic Arabia. In every board it is applied only to Arabia: stbbislamiyat-9-10-en · PAGE 146; stbbislamiyat-7-en · PAGE 45; pakistan-studies-9-en (“In Arab society, in the age of ignorance before the advent of Islam…”); kptbbenglish-7-en · PAGE 10. There is not one instance of it being applied to the subcontinent. The one near-miss is Sindh Islamiyat 8’s phrase “the early Brahmin era” (PAGE 76) — a period label, not the theological one.

16. Neither Punjab nor KP teaches the Conquest of Makkah in Islamiyat at all, classes 6–9. On the Urdu structural survey, no Fath Makkah chapter is established in any of the eight Urdu Islamiyat volumes; Punjab class 6’s own contents page names a chapter of “four ghazwat” that excludes it. The one node of the chain that is unambiguously scriptural and unambiguously iconoclastic — the clearing of the Kaaba — is taught as such in exactly one board’s religion syllabus in this corpus, Sindh’s, in English, at class 7. Confidence on the absence is low (OCR), but the negative evidence is unusually clean: eleven occurrences of فتح across eight files, all false positives.

17. No مشرک / کافر ↔ ہندو / ہندوستان co-occurrence survives inspection. Four same-page candidates across eleven Urdu files; all four resolve to OCR artefacts (مشترک read as مشرک, کے افراد read as کافراد) or to thematically unrelated content. The one genuine co-occurrence is the Chishti-order lesson in Punjab Islamiyat 8 (PAGE 82), where kafir appears inside a hadith forbidding Muslims to call one another kafir. This is the direct Tier-2 test on the Urdu side, and it comes back empty.

18. KP’s History 7 and 8 do not cover the conquest period at all. History 7 opens with Babur and the 1526 Panipat campaign; History 8 begins at Plassey. Somnath, Mahmud and Muhammad bin Qasim are not established in either. Whatever iconoclasm content KP carries is confined to the single class-6 volume — it is not reinforced anywhere later in the series.

19. No comparative construction exists anywhere in the English-medium corpus. A targeted search for “compare / contrast / just as / similar to / in the same way” within 120 characters of “idol”, “polytheist”, “infidel”, “Makkah”, “Kaaba” or “Hindu” returns exactly one hit across all three boards, and it is the neutral Indus Valley seal description in Sindh Social Studies 6 (PAGE 35). This is the direct negative test for Tier 1, and it comes back empty.

5. Limits

What this method can show. It can show, objectively and reproducibly, which evaluative words appear in which books at which grades, applied to which referents; whether the same word does the same work on both sides of a proposed equivalence; and how the elements are sequenced within and across volumes. Tier 2 is checkable by string match. Tier 3 is checkable by page range. Both are verifiable by anyone with the same files.

What it cannot show. It cannot show that any child draws the inference. Reading is not determined by text. A pupil may never connect a class-7 Islamiyat chapter to a class-10 Pakistan Studies paragraph; a teacher may draw the connection explicitly even where the book does not; a teacher may equally foreclose it. Nothing here is evidence about classroom practice, examination papers, teacher guides, or what pupils actually believe. The finding is about the availability of materials for an inference, not about the inference.

Evidentiary asymmetry by medium. Quotation was restricted to English-medium files. This systematically under-samples KP, which teaches Islamiyat, History, Geography and Urdu only in Urdu, and under-samples Punjab’s Islamiyat, which is Urdu-medium at every grade 6–9. Sindh is the only board whose Islamiyat is in English. Sindh therefore appears strongest partly because Sindh is the most visible to this method. A finding of “Tier 2 in Sindh, Tier 3 in Punjab” must be read with that bias stated: it is possible that Punjab’s Urdu Islamiyat carries the same vocabulary as Sindh’s English Islamiyat, and this analysis would not see it. What can be said without that caveat is narrower and still substantive: Punjab’s English-medium History deliberately removes the iconoclasm content that KP’s Urdu History retains. That is an authorial choice visible in both directions.

OCR. Nastaliq OCR in this corpus is theme-grade. Urdu findings are marked low confidence throughout and are never quoted. Where a term could not be found in an Urdu file, this report says “not established”, not “absent”.

Two specific hazards, both of which produced false results during this analysis before being caught, and which anyone repeating this work must control for:

• False negatives from clean-string search. ابراہیم never appears as a clean string in any Urdu file in the corpus. The OCR renders it as ابرامیم / ابراییم / ابراڈیم / ابرہیم / ابرھیم and worse. A naive grep returns zero everywhere — which is exactly what my first pass returned, and it was wrong. The Ibrahim material in mutalea-quran-9-ur was found only by space-normalised fuzzy matching. Any Urdu absence claim in this report rests on fuzzy variants, not on exact tokens.
• False positives from short tokens. بت (idol) matches inside بتایا and مصیبتوں; مشرک matches مشترک (“joint”); کافر matches کے افراد; مندر matches مندرجہ ذیل (“the following”); آزر matches ہزار and بازار; نمرود matches نماز and نمبر. Raw counts on these tokens are meaningless and every candidate hit in this report was inspected in context.

Three English-medium quotations carry OCR damage and are marked in place: Sindh Social Studies 8 PAGES 52 and 57, and KP English 9 PAGE 50. Two Punjab History 6 passages (PAGE 65) sit in a two-column layout that the OCR interleaves with an image caption; they are given as verbatim fragments rather than reconstructed sentences. Every other quotation in this report was re-verified by exact string match after whitespace normalisation — 47 strings checked, 47 confirmed.

Known gap, worth closing. The largest single hole in the evidence is mutalea-quran-9-ur pages 46–53, an OCR-blank block that contains Al-Anbiya 21:51–75 — the idol-smashing pericope itself. Its presence is currently inferred from a PAGE 55 exercise question rather than read. A higher-quality re-OCR of those eight pages would settle whether KP’s class-9 Quranic studies volume narrates the smashing directly and with what framing, which is the one question this analysis leaves genuinely open. Two smaller gaps: the surah sequence in that volume is unresolved at pp.67–83 and pp.113–131, and no KP class-6 History volume in Urdu was surveyed for the Ghaznavid material beyond the Somnath passage itself.

Single-edition snapshot. All findings are against the editions catalogued in the edition audit. Punjab’s History 6 is an Experimental Edition based on Single National Curriculum 2022, and History 7 an Experimental Edition dated January 2023; neither may reflect what is currently in classrooms. The Punjab omissions reported in §4 are therefore statements about the SNC-era Punjab books specifically, and should not be generalised backwards to earlier Punjab editions.

Follow-up attempted: re-OCR of the blank Quranic block (negative result)

§5 flagged mutalea-quran-9-ur pp. 46–53 as an OCR-blank block believed to contain Al-Anbiya 21:51–75 — the idol-smashing pericope itself — with its presence inferred from a p.55 exercise rather than read. That re-OCR was attempted and did not succeed.

Method. The eight pages were re-rendered from the source PDF at 450 dpi greyscale (up from the 300 dpi used corpus-wide) and re-run through tesseract with the Urdu model at --psm 6 (uniform block), which suits dense scriptural typesetting better than the default segmentation.

Result. The pass recovered far more characters — 31,056 across the eight pages, against effectively none before, with individual pages yielding 5,000–9,800. But the recovered text is not readable. Searches returned zero hits for ابراہیم and its OCR variants, زero for آزر (Azar), zero for اصنام, zero for توڑ (to break), and zero for شرک. The seven apparent hits for بت resolve on inspection to character noise inside malformed strings, not the word for idol — the same class of false positive documented in §2.

What this means. The block remains not established. Higher resolution increased the OCR’s output volume without increasing its fidelity, which is consistent with the corpus-wide finding that Nastaliq OCR fails on this material for reasons of script and typesetting rather than image quality. The inference from the p.55 exercise is unchanged and remains an inference.

What would actually resolve it: a human Urdu reader, or a Nastaliq-specific OCR engine, applied to those eight pages. Rendering the pages as images and reading them directly — the method used elsewhere in this project for load-bearing Urdu claims — would settle it in minutes and is the recommended next step.