Abstract
Paper 3 — Counting
This paper establishes two empirical, observational findings. First, the counting apparatus in this genre of fatwā sources uses multiple units that cannot be converted into one another internally. Second, enumerability and the capacity for exhaustive numerical enumeration form part of the legal characterisation of a right, rather than standing outside it. These findings rest on three comprehensive surveys of the frozen release (WEC-432/1.0) of the corpus on entitlement under waqf (a legal-institutional arrangement in which property is dedicated to a designated purpose under a founder’s stipulations and managed by an administrator), comprising three hundred and ninety-six fatwā texts issued by Dār al-Iftāʾ between (1313 and 1400 AH). The surveys found that the term “mustafīd” (“beneficiary”) does not occur in this corpus, while “muntafiʿ” (“one who derives benefit”) occurs in two records without denoting a category. By contrast, its two established concepts occur as follows: “mustaḥiqq” (“entitlement-holder”) in one hundred and sixty-six records, and “mawqūf ʿalayhi” (“waqf recipient”) in one hundred and thirty.
A numerical survey conducted independently of counting vocabulary also identifies observed counts of recipients in seventeen records. Their basic unit is the “designated name”, and their governing scope is a “single waqf”. They contain no figure aggregating recipients within a statistical denominator extending beyond an individual waqf.
The paper also examines the Ḥanafī juristic question of the “threshold of enumerability”, which Egyptian muftīs applied in ways that made the specified count and the capacity for exhaustive enumeration structural conditions of a right’s validity and creation.
The argument was tested by replacing the waqf with the German purpose foundation: the waqf’s counting apparatus does not distinguish it from its comparator within the family of purpose institutions. Replacement fails, however, at the condition of enumerability and exhaustive enumeration within the structure of the right. This finding is subject to its comparative limit: the comparator from the third family has an operative criterion external to the instrument that changes with circumstances. What remains distinctive is that the criterion is numerical and that its effect reaches the validity of creation under sharia. The transition of the unit of counting to a programme category remains a proposed hypothesis whose sources have yet to be examined.
Keywords: unit of counting; threshold of enumerability; waqf entitlement; stipulated person-slot; legibility; statistical category; regular provision; Egypt.
Statement
- Type of contribution: a completed test of one bounded component of the thesis, with a concept and a hypothesis beyond it:
The completed work consists of three surveys of a corpus whose size, scope and hash are disclosed. This is the frozen release WEC-432/1.0, comprising three hundred and ninety-six texts. The first survey examines vocabulary describing the receiving side; the second examines counting and enumeration vocabulary; and the third is a numerical survey independent of that vocabulary, capturing both numerals and number words.
They establish three findings: first, the complete absence of “mustafīd” from the corpus; second, the restriction of observed counting to a single category, verification of actual presence within an individual waqf; and third, the absence of any aggregate figure extending beyond that scope. These are supplemented by an arithmetic audit of the table of al-Azhar statistics for 1875 and a source-based examination of the “threshold of enumerability” in Ḥanafī jurisprudence. The broader claim—the transition of the unit of counting to a programme category within the planning apparatus—is not tested here. It should therefore be read as a proposed hypothesis accompanied by its falsifier. It is not expressed in the language of findings, as required by the third tier of the shared inference protocol.
- Sample selection rule and direction of bias:
The primary source of evidence is the frozen release of the entitlement corpus; its inclusion rule and the stages of its selection funnel are set out in §3 and Appendix B. All three surveys cover the entire corpus, rather than a sample. There is therefore no selection on this side. The cases presented in the main text, however, were selected according to two conditions: the text must contain an explicit rather than inferred unit of counting, and its question must arise from an actual dispute. This is selection for substantiability, with a known direction of bias: it favours a picture of the waqf system as one of rigorous record-keeping. A text is produced only when a dispute reaches a body that records it. The awqāf (plural of waqf) in which record-keeping was neglected or the register was lost do not appear in the fatwā corpus at all. These cases consequently show that accurate recording was available to those who pursued litigation, rather than establishing its prevalence across awqāf as a whole. This bias favours only the waqf side of the comparison. On the statistical apparatus side, there is no equivalent corpus, only two figures reported in a secondary source. The difference presented between the two systems is therefore an upper bound, not an estimate.
- Condition for falsification and the result after testing:
Before the survey, the falsifier was formulated as follows: the entitlement corpus would contain a single figure counting recipients as an observed population. This meant a figure derived from counting persons, rather than a quota stipulated in a deed or a denominator for distribution per head. Nor would it be a count of papers and documents. The test was conducted across the entire corpus using two procedures with different starting points. It found that the falsifier was met in a restricted form. The lexical survey of counting expressions produced nothing that satisfied it. The independent numerical survey, however, identified seventeen records in which a body states how many waqf recipients actually existed at a known date. The claim must therefore be narrowed to what was counted: observed counting is present in the corpus. Its unit is the designated name, its scope an individual waqf, and its purpose to establish or deny entitlement. What was not found was a figure aggregating recipients within a denominator extending beyond an individual waqf to a residential hall, institution or sector. This finding is limited by the fact that the source comprises a single documentary genre. It attests to waqf fatwās, rather than to all genres of administrative writing.
- Strongest competing explanation:
The standing competing explanation is that the limited scope of observed counting in the corpus reflects the documentary genre rather than the structure of the system. A fatwā answers a question arising in a particular waqf; it cannot be expected to contain a population survey. This remains a viable explanation that the paper’s design does not exclude. It is more necessary to state it explicitly than merely to qualify the argument against it. The material constrains it in three respects, set out in section 12 together with a future test and the limits of what would remain viable afterwards.
- Result of testing the competing explanation across the programme, and its effect on the scope of the claim:
The explanation that the programme’s series measure bias in the corpus rather than structural transformation was tested along four axes. The results appear in the Common Appendix. This paper relates to that test in three respects, detailed in section 12: the central finding survives, since “mustafīd” is entirely absent across all nine decades; a claim of increasing lexical usage cannot be sustained and is therefore not made; and the increased share of the verification-of-presence category after 1946 is reduced to an indication that is reported without favouring an explanation.
- Position of the paper within the programme’s variable matrix:
The dependent variable in this paper is the unit used to count the receiving side, in its three dimensions: the object counted, the unit of counting and the scope of aggregation. Its falsifier is the existence of a figure counting recipients as an observed population. As noted above, this was met in a restricted form. Five bodies of evidence distinguish this paper from its three companions: the three lexical and numerical surveys; the absence of “mustafīd” across all nine decades; the al-Azhar statistics for 1875 and their arithmetic audit; the Ḥanafī question of the “threshold of enumerability”, with its sources; and detailed accounts. The complete matrix appears in the Common Appendix.
Its overlap with the first paper of this series is acknowledged explicitly: the first paper counts instances and assesses the reference point of their answers, while this paper counts words and numbers within the texts. The source is the same. Agreement between the two papers should therefore not be read as independent corroboration; that requires a source external to this corpus.
Methodological Disclosure
This paper was prepared with the assistance of “al-Warrāq” in retrieving texts, conducting the surveys and constructing the tables. The author alone is responsible for the design, definition of units of analysis, assessment of evidentiary grades and writing. The paper observes six constraints established in advance. Every measure of text length is calculated from the cleaned text, rather than from the corrupted recorded field. Every figure whose unit and denominator are unclear is excluded. Information taken from a source whose original has not been examined is tagged as secondary evidence. The collection editor’s layer is separated from the question-and-answer layer. What the corpus yields is distinguished from what it does not yield. The central finding is restricted to what the completed surveys establish. Legislative detail and the channel for claims belong in “Did Charity Become a Right?” They are invoked here solely insofar as they affect the unit of counting.
1 · Introduction: Two Institutions and Four Counts with No Common Denominator
In 1875 (corresponding, according to ʿAlī Mubārak, to 1293 AH), the numbers of teaching shaykhs, students and residents in al-Azhar’s riwāqs (residential halls) were recorded. There were three hundred and sixty-one shaykhs (361) and ten thousand seven hundred and eighty resident students (10,780), distributed across fifteen quarters and thirty-eight riwāqs1.
In the same year, the number of resident students at al-Aḥmadī Mosque in Ṭanṭā reached three thousand eight hundred and seventy-four (3,874), according to a figure Ibrāhīm al-Bayyūmī Ghānim reports from a preserved statistical digest2. In a footnote to the same passage, Ghānim gives a fourth figure in these words: “At the beginning of the twentieth century, however, the number at al-Aḥmadī Mosque reached 400 students”3.
The four figures appear to belong to a single statistical category: persons studying in religious educational institutions supported by waqf. Yet direct arithmetic comparison fails at the first question. A decline from three thousand eight hundred and seventy-four (3,874) resident students to four hundred (400) students within the same institution over twenty-five years admits more than one historical reading. It may indicate a real decline in attendance. It may also reflect a change in the unit: the first figure counts “resident students”, the second “students”. In the established usage of al-Azhar and al-Aḥmadī, a resident student lived in a riwāq and was entitled to regular provision and accommodation; a student in the Public Instruction table was registered at an educational institution. Only a conversion bridge establishing each term’s definition in its source could distinguish between these readings, and none is available in the circulating version.
An internal arithmetic audit of the first table reveals a more precise limitation. The breakdown of shaykhs by the four legal schools falls thirty-six short of their total, whereas the breakdown of resident students matches their total exactly. The discrepancy is therefore confined to a single row. The calculations, the two possible explanations and the restriction on using the total appear in Appendix (D).
The text itself imposes a third constraint on its reader, and this is its most valuable feature. After presenting the figures, ʿAlī Mubārak says that those in Kitāb al-Natīja al-iḥṣāʾiyya li-l-madāris wa-l-makātib bi-l-quṭr al-Miṣrī [Book of Statistical Results for Schools and Qurʾān Schools in Egypt] are close to them: “This is approximate; there are students at al-Azhar who are not entered in its records, while its registers contain persons who do not attend lessons but instead practise trades”4. The distance between the “registered person” and the “person actually present” is thus acknowledged in the source itself in 1875, in both directions: (present but unregistered, and registered but absent). This is not a later researcher’s correction of an old table, but the table’s author qualifying and criticising his own table. He goes further by correcting a particular figure, stating that the reported count of three Ḥanbalīs does not match those actually present, since for several years al-Azhar has had only one Ḥanbalī teacher. The source thus contains a count, a critique of that count and a correction to one of its cells at the same time.
This is the starting point of the third paper’s argument: the problem was not that the beneficiary went uncounted, but that each apparatus counted the beneficiary under a different description and concept. When that description and its institutional constitution changed, so did what the institution could discern about the beneficiary and what it could do in relation to them. The individual name carrying a share in a waqf distribution is not the “pupil” entered in the Public Instruction table; a loaf distributed or a regular provision is not an abstract person; and a decision appointing a nāẓir (trustee-administrator) is not a beneficiary at all. Yet these figures are assembled into a single narrative of expanding beneficiary numbers, obscuring the distinctions that produced each figure in the first place.
Footnotes
-
ʿAlī Pasha Mubārak, al-Khiṭaṭ al-Tawfīqiyya al-jadīda li-Miṣr wa-l-Qāhira wa-mudunihā wa-bilādihā al-qadīma wa-l-shahīra [The New Tawfīqī Topography of Egypt, Cairo, and Their Ancient and Famous Cities and Towns], Cairo: Dār al-Kutub wa-l-Wathāʾiq al-Qawmiyya, second revised edition, 1425 AH / 2004, twenty volumes, p. 36. The text reads: “The number of teaching shaykhs, students and residents in the riwāqs of this school was recorded in the year eighteen hundred and seventy-five of the Christian era (corresponding to twelve hundred and ninety-three of the Hijra). There were three hundred and sixty-one shaykhs: one hundred and forty-seven Shāfiʿīs, ninety-nine Mālikīs, seventy-six Ḥanafīs and three Ḥanbalīs. There were ten thousand seven hundred and eighty resident students in fifteen quarters and thirty-eight riwāqs: five thousand six hundred and fifty-one Shāfiʿīs, three thousand eight hundred and twenty-six Mālikīs, one thousand two hundred and seventy-eight Ḥanafīs and twenty-five Ḥanbalīs. Towards the end of eighteen hundred and seventy-five, their number increased by approximately five hundred and sixty-four students”. The breakdown of shaykhs totals 325 against a stated total of 361, a difference of 36. The breakdown of resident students totals 10,780, exactly matching the stated total. ↩
-
Ibrāhīm al-Bayyūmī Ghānim, al-Awqāf wa-l-siyāsa fī Miṣr [Awqāf and Politics in Egypt], section “Taḥlīl bayānāt al-waqfiyyāt” [“Analysis of Waqf Deed Data”], visible pagination in the copy examined: 230. The main text reads: “The number of its resident students in 1292 AH / 1875, for example, reached 3,874”. Its footnote refers to “Book no. (370) at the office of the Ministry of Education in Cairo”, with the title given as “Fadhlaka iḥṣāʾiyya ʿumūmiyya li-l-madāris wa-l-katātīb bi-l-quṭr al-Miṣrī min sanat 1847 [General Statistical Digest of Schools and Qurʾān Schools in Egypt from 1847] (printed at Maṭbaʿat al-Madāris al-Malakiyya: 1942), p. 303”. The wording in the digital version contains an evident transcription error, recorded here rather than corrected by conjecture. The same footnote states: “At the beginning of the twentieth century, however, the number at al-Aḥmadī Mosque reached 400 students, while the number of students at al-Azhar Mosque was approximately seven thousand, later rising to 15 thousand students in 1918”. The source says “the number at al-Aḥmadī Mosque, 400 students”, without “students” after “number”; supplying that omitted noun is this paper’s interpretation, not Ghānim’s wording. Note that the main-text sentence breaks off in the copy examined at “and in that year the number of resident students at”, so al-Azhar’s figure for that year cannot be recovered from it. Note also that Ghānim pairs 1875 with 1292 AH, while ʿAlī Mubārak pairs it with 1293 AH. A contemporary estimate from outside the corpus provides a counterpart to Ghānim’s figure for al-Azhar at the beginning of the twentieth century. It is recorded here as a qualification of the reported figure, rather than evidence for a claim. The journal al-Manār published an unsigned item entitled “al-Jāmiʿ al-Azhar” [“Al-Azhar Mosque”], no. 30, 7 January 1901, pp. 743–745; the article-author field in the copy examined is blank. It summarises an article by the proprietor of the Indian newspaper Paisa Akhbār (Lahore) about his visit to al-Azhar. The counting passage reads: “I estimated those studying there at ten thousand and the teachers at a hundred or more (the correct figure is ahum one hundred), and the students range from 7 years to seventy years”. This passage imposes three constraints on using figures of this kind. First, the verb “I estimated” makes it an estimate rather than an observation, so it cannot be compared with ʿAlī Mubārak’s recorded figure for 1875. Second, the editor challenges the number within the text by saying “the correct figure is ahum one hundred”: correcting a count within the place where it is published is therefore not peculiar to ʿAlī Mubārak, but a practice within this genre of writing. Third, the unit “student” encompasses persons from seven to seventy years old, providing lexical evidence for the non-equivalence of units discussed in section 8. No calculation is based on this estimate, and it is not included in comparisons with figures from the corpus. ↩
-
See note 2. ↩
-
ʿAlī Pasha Mubārak, al-Khiṭaṭ al-Tawfīqiyya al-jadīda [The New Tawfīqī Topography], p. 36, immediately after the figures: “The figures in Kitāb al-Natīja al-iḥṣāʾiyya li-l-madāris wa-l-makātib bi-l-quṭr al-Miṣrī [Book of Statistical Results for Schools and Qurʾān Schools in Egypt] are close to these. This is approximate; there are students at al-Azhar who are not entered in its records, while its registers contain persons who do not attend lessons but instead practise trades. It also includes children of the Qurʾān schools. Its statement that there are three Ḥanbalīs does not match those actually present, for there has been only one Ḥanbalī teacher there for several years up to the present”. ↩
2 · The Number as an Institution: The Paper’s Place in the Literature
Studies in the history of statistics and the sociology of quantification treat the statistical figure as the outcome of classificatory work that determines what can be aggregated and compared. Alain Desrosières links the construction of statistical categories to both the logic of the state and technologies of knowledge, while Theodore Porter explains the appeal of numbers through their ability to present decisions as impersonal and transferable across administrative distances. Espeland & Stevens describe commensuration as the conversion of different qualities into a common metric that permits comparison while stripping things of some of their particularity. These references concern the works’ general arguments collectively, since their editions were not examined for this paper5.
The paper draws on James C. Scott’s concept of legibility, with two explicit modifications. First, centralisation does not begin from an illegible local vacuum: waqf material reveals precise systems of registration, distribution and waiting. Second, legibility is distinguished from equivalence of units: two systems may both be legible while their units remain inconvertible6. What is new is therefore not legibility itself, but the change in its scope and unit of calculation: (from a waqf or riwāq to an institution, province and sector).
For Egypt, recent studies have reconstructed the history of census-taking and planning as a history of struggles over the definition of resources, labour and population. Malak Labib’s work situates the politics of quantification in Egypt in relation to public debt and the construction of knowledge usable by government, and then in relation to counting workers and formulating concepts of underemployment7. This reference must be read within its limits: neither edition was examined, so this paper draws no dating to a particular decade from them. These studies, however, concentrate on population, labour and resources. This paper turns to the receiving side of charity, asking how those receiving revenue, regular provision or education were counted. It asks what distinguishes counting someone as an entitlement-holder from counting them as an affiliate or a member of a category.
The closest precedent within this research programme that threatens the claim of novelty here is “When the Waqf Became Legible”. The distinction is precise: that paper traced the shift in the unit of accounting from the individual waqf to the aggregate domain of assets and revenues, taking the structure of production as its subject. This paper examines the legibility of the receiving side, taking persons, shares and services as its subject. It follows that more figures do not necessarily mean more knowledge of the beneficiary: the number of tables may increase while the information they contain about the beneficiary’s identity and particular legal position declines.
An explicit qualification of the claimed gap is necessary here. This paper has not undertaken a systematic survey of studies of charitable quantification in the Arab world. The claim of a gap is limited to the scholarship surveyed and is not elevated into a claim that the field as a whole is empty.
Footnotes
-
Alain Desrosières, The Politics of Large Numbers (1998); Theodore M. Porter, Trust in Numbers (1995); Wendy Nelson Espeland and Mitchell L. Stevens, “Commensuration as a Social Process,” Annual Review of Sociology 24 (1998): 313–343; idem, “A Sociology of Quantification,” European Journal of Sociology 49, no. 3 (2008): 401–436. The reference is to the works as a whole; their editions were not examined here. ↩
-
James C. Scott, Seeing Like a State (New Haven: Yale University Press, 1998). The reference is to the work as a whole. ↩
-
Malak Labib, Recenser l'Égypte (Cairo: IFAO, RAPH 47, 2024); idem, “An Unlimited Supply of Labor?” International Labor and Working-Class History 108 (2025): 1–17. The reference concerns the general arguments of the two works as a whole; neither edition was examined. This paper therefore attributes no dating to a particular decade to them and mentions only their general subject. ↩
3 · Concept and Method: An Instance of Counting the Receiving Side
3.1. Defining the Unit of Analysis and Disaggregating Receipt and Production
The smallest unit of analysis in this paper is an “instance of counting the receiving side”, defined methodologically as: “Any instance in which a body produces a count of persons or of benefits directed towards persons, according to a known inclusion rule, time and scope, for the purpose of distribution, accounting, regulation, reporting or planning”.
This unit is distinguished from an “instance of counting the structure of production”, whose object is the waqf itself, an institution, a document or an employee. Although figures concerning the structure of production may help explain the environment that generated beneficiary counts, they do not enter the denominator for beneficiaries themselves.
Each instance is coded across twelve fields, detailed with their coding questions in Table (M1) of Appendix (A). The mere existence of a figure does not make comparison possible: its denominator, inclusion rule and date must also be identified.
3.2. Sources of Evidence and the Limits of Archival Coverage
The primary source of evidence is the frozen release (WEC-432/1.0) of the Waqf Entitlement Corpus, a complete automated enumeration of the published collection of Dār al-Iftāʾ fatwās. Its inclusion rule is explicit and operational: the mention of “waqf”, “founder” or “dedicated as waqf” at the beginning of a text together with an entitlement term from a closed lexicon, or a title concerning waqf. This is followed by two stages of manual review. The frozen release consequently comprises three hundred and ninety-six texts with stable identifiers in the form F432-NNNNN. The surveys published here were conducted on that release, and all their figures refer to it unless another layer is explicitly identified. The stages of the selection funnel, the error rates in both directions and coding accuracy on a sample of sixty records are all given in Appendix B. The complete record, with its hashes, is in the unified corpus selection funnel file8. The corpus extends from 1313 AH to 1400 AH; five texts have no date in their headings. Removing those five leaves three hundred and ninety-one texts included in the temporal count using the corrected dating key described below. The two keys must be distinguished: under the heading-based key, five records lack a year and the denominator is three hundred and ninety-one; under the year extracted from the body of the text, six lack a year and the denominator is three hundred and ninety. These are the denominators of two different keys, not two conflicting figures.
The second source of evidence consists of descriptive and statistical texts examined directly. These include al-Khiṭaṭ al-Tawfīqiyya on al-Azhar, its riwāqs and its statistics for 1875; two sections of Ghānim’s book, one analysing waqf deed data and the other describing registers of nāẓirs’ reports; and, third, works of Ḥanafī jurisprudence on the threshold of enumerability.
Each item of evidence is tagged on a four-grade scale, set out in Appendix D. Two constraints govern the tagging. First, a source’s silence is recorded as “unknown” and is not treated as evidence supporting the thesis. Second, the original statistical digest has not been examined. Every figure taken from it therefore remains at the grade of reported evidence and cannot be promoted to directly examined evidence.
3.3. Technical Controls: Separating Layers and Correcting Keys
Separating the editorial layer is a condition of every count published here. A record in this collection has three layers, not one: first, a title supplied by the collection’s editor; second, a “Principles” block summarising the fatwā’s implications; and third, the question and answer, which contain the questioner’s and muftī’s language. Conflating these layers would attribute a later editor’s wording to the muftī. The title-and-principles block is therefore separated from the question-and-answer block, and every count is published in two columns. This separation has a substantial effect: “mustaḥiqq” falls from one hundred and seventy-four records to one hundred and sixty-six; “mawqūf ʿalayhi” falls from one hundred and forty-four to one hundred and thirty; and occurrences of “iḥṣāʾ” (“enumeration”) fall from two to one.
The lexical matching protocol was declared before reading, and its specification appears in Appendix (B). Its word boundary is not a procedural detail: matching “nafar” (“person”) without a word boundary raises the count from thirteen occurrences to more than a hundred. This happened in an initial trial and was corrected before the results were used. The difference between the two counts is not a difference in precision, but the difference between measuring a phenomenon and measuring an illusion. The procedure and the incident are detailed in Appendix (D).
The dating key has been corrected in this version. The automated Gregorian-year field is corrupted in one hundred and seventeen records, where the year is calculated by a fixed offset rather than extracted; in fifteen records, it overlooks an explicit Gregorian date in the heading. The key adopted here is constructed from the heading itself: the Gregorian year stated there, where present; otherwise, a year converted from the Hijri month and year using the tabular calendar. All the figures are given in Appendix (B). Hijri dates are retained in the main text because they are expressly stated in the sources; a Gregorian date is added only where the heading supplies one.
Three tests were completed. Anything not carried out is not a condition for falsification in this paper: an announced but untested condition is a promise of research, not a falsification test. This account is therefore restricted to what was actually done, with the remaining items transferred to the conditions for upgrading claims at the end of the paper.
The first test surveys vocabulary describing the receiving side across the entire corpus and is presented in section 4. The second surveys counting and enumeration vocabulary throughout it. The third is a numerical survey independent of counting vocabulary, serving as the alternative test of the stated falsifier. The latter two appear in section 5.
Footnotes
-
The frozen release WEC-432/1.0 contains 396 texts. Its manifest includes the hash of the release file, the combined hash of the texts and the hash of the governing source on which the identifiers are fixed. The five records that the automated key placed in 1997, all by ʿAṭiyya Ṣaqr on non-waqf subjects, were excluded from the release through manual coding. They are therefore no longer in the file and require no further subtraction. Details of the count, inclusion rule and stages of the selection funnel appear in Appendix B; the complete record is in the unified corpus selection funnel file. ↩
4 · First Test: The Lexicon of the Receiving Side in the Corpus on Entitlement
4.1. The Explicit Statistical Finding and Lexical Refutation
“Mustafīd” does not occur even once in the entitlement corpus, either in the fatwā texts or in the editorial layer, across (396) texts spanning (87) Hijri years. “Muntafiʿ” occurs in only two records, in neither of which does it name a category of recipients. By contrast, “mustaḥiqq” and its derived forms occur in one hundred and sixty-six records (166) in the fatwā texts, with five hundred and sixty-three occurrences. “Mawqūf ʿalayhi” occurs in one hundred and thirty records (130), with three hundred and sixty-seven occurrences9.
The totals for these terms in both layers, together with each term’s location in the corpus, appear in Table (D1) of Appendix (D). The figures count records and occurrences, not persons.
4.2. Close Analysis of the Two Occurrences of “Muntafiʿ”
The two records containing “muntafiʿ” merit examination to establish their contexts:
-
In the first (Ṣafar 1387 AH / 11 May 1967), the founder requires “the descendants residing in the aforementioned house and benefiting from it” to restore and repair it, each according to their ability. The word here applies to the user of an asset, and its context is an obligation rather than an entitlement.
-
In the second (Ṣafar 1322 AH / May 1904), a partner in a company is described as “in possession of and benefiting from his share in the steam engine”. The word applies to the holder of a share in a company. “Muntafiʿ” in this corpus is therefore not the name of a recipient category, but a description of someone possessing an asset or a share. The corpus thus contains no single term encompassing recipients as a category targeted by a programme.
4.3. Distribution of Terms over the Decades and Its Limits
The distribution of terms across periods reveals no lexical transition within the corpus. The figures appear in Table (D2) of Appendix (D). The column total differs from the overall total for a term because some records fall outside the dating key. The upper bound compatible with zero occurrences varies with the period’s denominator: three per cent in the first period (3%), two and a half in the second (2.5%), six in the third (6%), seven in the fourth (7%), six in the fifth (6%), and thirteen and a half in the sixth (13.5%). Across the corpus as a whole, it is approximately three-quarters of one per cent (0.75%), certainly below one per cent (1%).
“Mustaḥiqq” and “mawqūf ʿalayhi” remain dominant in all six periods, while “mustafīd” remains at zero throughout. The corpus therefore does not record the lexical transition assumed in contemporary discourse. Yet this finding cannot be applied to the final period alone: that inference was attempted and failed. The denominator for that period is twenty-two records. Zero occurrences are compatible with a true frequency of approximately thirteen and a half per cent (13.5%). A term occurring in one record out of every seven could plausibly have been absent from this set. The finding is therefore stated for the corpus as a whole: zero out of three hundred and ninety-six. The inference specific to the period following the establishment of the Waqf Affairs Committee is recorded as inconclusive and left for a larger corpus.
This distribution was reconstructed using a second, exclusively Hijri key as a check. That exercise was conducted on the third layer before freezing and is attributed to that layer; its figures appear in Appendix (B). The zero count for “mustafīd” holds across all six periods under both keys, as does the dominance of “mustaḥiqq” and “mawqūf ʿalayhi”. The difference between them is confined to the size of the final period. The weakness of that period therefore reflects the material itself, rather than the choice of key. Moreover, the zero count for “mustafīd” does not depend on any key, since it is zero across the corpus as a whole.
The limit of this finding is explicitly acknowledged: the source genre is a fatwā on a waqf case, whose language concerns stipulations, legal positions and shares. The term’s absence from it does not establish its absence from the press, association regulations or social affairs reports. Only two things are established. First, the genre governing waqf entitlement did not use the term up to 1400 AH / 1979. Second, a claim of lexical transition cannot be built on this corpus; it must be tested in a different administrative corpus that has not yet been examined.
Footnotes
-
A completed survey of the entire frozen release WEC-432/1.0 using the protocol in Appendix B, after separating the editorial layer. The two records containing “muntafiʿ” are F432-05481 (Ṣafar 1387 AH / 11 May 1967, Aḥmad Huraydī, “Intihāʾ al-waqf qānūnan wa-qismatuhu” [“Statutory Termination and Division of a Waqf”]), which reads: “The descendants residing in the aforementioned house and benefiting from it must restore and repair it when necessary, each according to their ability and capacity”; and F432-06089 (Ṣafar 1322 AH, Muḥammad ʿAbduh), which reads: “Throughout that period, each of the partners was in possession of and benefiting from his share in the steam engine”. The second record’s heading supplies no Gregorian date. ↩
5 · Second and Third Tests: Does the Corpus Contain Observed Counts of Recipients?
5.1 · Survey of Counting Vocabulary
This statistical survey covered the entire Waqf Entitlement Corpus using three matching patterns: (“ʿadad” (“number”) and its derived forms; “nafar” and “anfār” (its plural); and “iḥṣāʾ” and its derived forms). Every occurrence was then examined in context.
- Results for the “ʿadad” pattern: forty-six occurrences in the fatwā texts across twenty-seven records, and three occurrences in the editorial layer. They divide into five uses, none of which constitutes observed counting:
The first counts papers and documents, taking the document rather than the person as its object: documents submitted to the muftī are “23 in number”, and papers sent are “4 in number”. The second supplies a denominator for distribution per head. The third gives a quota stipulated in the deed. The fourth expresses capacity derived from funds. The fifth is a hypothetical example used in reasoning within an answer. None of the five constitutes observed counting of recipients. The head is an arithmetic denominator in a distribution, rather than a unit for observing a population; a quota is fixed in the deed before its places are occupied; the number of patients in the waqf of al-Minshāwī Hospital, “as many as the aforementioned funds can accommodate”, is a calculation from the budget rather than an observation of those who fell ill; and the figure in the hypothetical example serves an explanatory purpose. The evidence for all five uses, with its texts, appears in Appendix D, and the identifiers are given in the footnote10.
- Results for the “nafar” pattern: thirteen occurrences across seven records, all in the fatwā texts and all expressing a stipulated quota. They include ten reciters who have memorised the Book of God, in a waqf constituted by a deed from al-Bāb al-ʿĀlī Court in 1217 AH; eleven Sufis whose payment is divided into twelve shares; and eighteen bearers of the Qurʾān from the Jabartī community residing in Medina. ⚠️ This last case is not Egyptian and should be read as evidence of the stipulated person-slot formula, rather than of conditions in Egypt. All seven appear with their texts and identifiers in the footnote11, and their enumeration is given in Appendix (D).
The nafar in this corpus is therefore a unit of design rather than a unit of observation. Its number is fixed in the founder’s deed before anyone occupies the place, and its function is to determine the denominator for distribution and the size of the share. This is clear in the case of the eleven Sufis, where the payment is divided into twelve shares for eleven persons: the figure serves explicitly as an arithmetic denominator. It is clearer still in the case of the eighteen bearers of the Qurʾān. Their number is not disputed, since it is fixed in the deed; the dispute concerns the definition of “bearers of the Qurʾān”: does it include those who read from a written copy, or only those who have memorised it? The questioner describes the replacement mechanism as a practice followed “age after age”, whereby “when one of them dies, another from the aforementioned community is appointed”. What is contested in this system is therefore the inclusion rule, rather than the result of counting.
- Results for the “iḥṣāʾ” pattern, which changed after the layers were separated: one occurrence in the language of fatwā and one in the editor’s language. An occurrence previously treated as evidence of enumerating persons reads: “The reciters of the Qurʾān reading circle become individually designated through designation owing to the possibility of their enumeration”. It appears in the “Principles” block of Bakrī al-Ṣadafī’s fatwā of Jumādā al-Thāniya 1331 AH. It is therefore the collection editor’s formulation, rather than the muftī’s answer. The answer uses different wording: “As for the reciters of the aforementioned Qurʾān reading circle, they become individually designated through designation because they can be enumerated”12. The second occurrence is in the text of Shaykh ʿAbd al-Majīd Salīm’s fatwā of Dhū al-Qaʿda 1358 AH / 3 January 1940. It concerns the enumeration of waqf receipts and expenditure in an account book, rather than the enumeration of persons13. Separating the layers yields a more precise finding than previously supposed: not a single muftī’s answer anywhere in the corpus uses “iḥṣāʾ” in relation to persons. The sole source of that association is a later editor’s wording; the muftī uses the verb “yuḥṣawn” (“can be enumerated”), rather than the abstract noun “iḥṣāʾ”.
5.2 · The Independent Numerical Survey and the Result of the Falsifier
A lexical survey examines counting vocabulary rather than numbers. This limitation must be addressed, rather than merely acknowledged: an explicit statement such as “there are eighty men in the riwāq” falls entirely outside the survey because it contains none of the words “ʿadad”, “nafar” or “iḥṣāʾ”.
A second survey was therefore conducted independently of counting vocabulary. Its procedure was declared before reading, and its stages and their counts are detailed in Appendix (B). Every numeral and explicit number word in the fatwā texts is captured; processing is then restricted to those occurring near a term from a closed list of names for the receiving side. Dates, sums of money, areas, shares, and article and law numbers are excluded according to a rule established in advance, and everything remaining is read in context. The layer used for its calculations is disclosed: its intermediate stages were calculated on the third layer, comprising four hundred and five instances, and were not rerun on the frozen release. They are attributed to their own layer and are not adjusted by estimation. The survey’s final result does not change between the two layers: all seventeen records are included in the frozen release under their identifiers, and none of the nine excluded records contains an occurrence of the sixth category.
The contextual reading found that the five categories of usage established in advance do not exhaust the material. A sixth, previously unrecognised category is required.
Sixth Category: Counting to Verify Actual Presence
This category comprises a passage in which the questioner, muftī or an administrative body states how many waqf recipients or entitlement-holders actually exist at a known date. This may take the form of an explicit figure or an exhaustive list of names. Its lexical indicators are declared: “there are no … except”, “none remain … other than”, “is confined … to”, “the only”, and “currently present”. Seventeen records in the corpus fall within this category; the supporting evidence is given in the footnote14.
Its three clearest examples are:
The first example: (Rajab 1344 AH / 30 January 1926). The question arrived in a letter from the Office of the Shaykh of al-Azhar. A sum from the waqf of Zaynab bint Saʿīd was held in al-Azhar’s treasury, representing two years’ entitlement for students from the island of Djerba. “When payment of this sum to its entitlement-holders was about to begin, it emerged that there were no students from the island of Djerba in the Maghribi riwāq … However, there is an unaffiliated student residing near Aḥmad ibn Ṭūlūn Mosque, as reported by the respected shaykh of the Maghribi riwāq”. The muftī concludes that it should be paid to that student “provided that he satisfies the aforementioned founder’s stipulations and that there is no student other than him”15. This is an instance of observation with all its components: (a counting authority, the shaykh of the riwāq; a defined spatial scope; and a known date). Its numerical result is zero within the riwāq and one outside it, and its administrative route runs upwards from the shaykh to the administrative council and then to the muftī.
The second example: (Dhū al-Qaʿda 1315 AH / April 1898). It concerns a house dedicated as waqf for the residence of freed persons. The text reads: “Since none of those freed persons now remain except three married women residing in their husbands’ houses, and none of the freed persons is in the house, is it therefore possible … to obtain rental income from it and add this to the waqf’s revenue?”16. The count is three, accompanied by an observation of residence. Its purpose is a decision on exploiting the asset, rather than distribution per head.
The third example: (Shaʿbān 1334 AH / 8 June 1916). It contains an exhaustive list of names tied to a moment identified as “now”. Only one named freedwoman remains; the muftī then lists the names of the existing children of freed persons entitled in the second tier17. Its text appears in Appendix (D).
The falsifier was therefore met in a restricted form, and the claim must be narrowed to what was counted. It is untenable to say that the corpus contains no observed counting of recipients: it contains such counting in seventeen places. Three descriptions of this counting are warranted. First, its unit is the designated name rather than the anonymous head, preserving identity fully: in most passages, those counted are named individually. Second, its scope is the individual waqf: not a single figure was found aggregating recipients within a denominator extending beyond a waqf to a riwāq, institution, province or sector. Third, its purpose is to establish or deny entitlement, rather than to describe a population or measure coverage.
The distribution of this category across periods provides an unexpected indication.
| Period | Instances | Instances verifying actual presence | Percentage |
|---|---|---|---|
| 1895–1913 | 101 | 2 | 2% |
| 1914–1933 | 124 | 5 | 4% |
| 1934–1945 | 52 | 0 | 0% |
| 1946–1953 | 42 | 4 | 10% |
| 1954–1959 | 50 | 6 | 12% |
| 1960–1980 | 22 | 0 | 0% |
Table 3: Instances of counting to verify actual presence by period, relative to each period’s denominator, in the frozen release. The zero counts in the third and sixth periods are compatible with upper bounds of six per cent (6%) and thirteen and a half per cent (13.5%), respectively. They therefore cannot be read as evidence of absence. The zero count in the sixth period should not be interpreted at all, as established in section 12. 🔴 The distribution across the final two rows had remained unresolved; rerunning the key on the frozen release resolved it, and the result is recorded here. The last of the seventeen records is F432-05835. Its Hijri field reads 1387 and its Gregorian field 1959, an eight-year discrepancy. The printed heading gives both dates in these words: “Shawwāl 1387 Hijri - 25 April 1959 CE”. Neither converts into the other. The inconsistency is in the printed source, rather than in extraction.
Two pieces of evidence from the corpus itself resolve the matter. First, Shawwāl 1378 AH corresponds precisely to April 1959; the dates align if “1378” is read in place of “1387”. Second, the muftī in the case is Ḥasan Maʾmūn, who has forty-three records in the frozen release. Forty-two fall between 1375 and 1380 AH; 1387 occurs only in this record. It is therefore likely that “1387” is a transcription error for “1378”, and that the record belongs in the fifth period, as entered in the table.
⚠️ The limit of this judgement is explicit: the paper edition has not been examined, and the two pieces of evidence are internal indications rather than inspection of an original. The zero count in the sixth period retains its status: it cannot be read as evidence of absence, as established in section 12.
The table indicates that this category’s share of instances rises after 1946: seven instances out of two hundred and seventy-seven in the preceding three periods, compared with ten out of one hundred and fourteen in the following three. This is a described association, and the paper favours no explanation for it. Two explanations remain, between which the corpus cannot distinguish internally. First, the abolition of family waqf under Law 180 of 1952 converted entitlement into ownership to be apportioned at a single date, requiring the authorities to enumerate those alive on that day. Second, the composition of cases reaching Dār al-Iftāʾ changed. This explanation is measured rather than assumed, as discussed in section 12. The small denominators impose a further limitation. The discriminating test is specified: compare this category’s share in the fatwā corpus with its share in the minutes of family waqf liquidation committees between 1952 and 1957.
The numerical survey has three limitations. First, capture was restricted to numbers near terms from a closed list naming the receiving side; a figure counting recipients under a description outside that list would escape the survey. Second, as stated in the declaration, the competing explanation based on documentary genre remains viable. Third, the survey’s intermediate stages were calculated on the third layer rather than the frozen release, as disclosed above.
Footnotes
-
Identifiers of the records cited in this subsection, in order of appearance: counting documents in F432-05309 (Muḥammad Bakhīt, Ramaḍān 1338 AH, location 2656, the case of the drinking trough allocated from public funds presented in “Did Charity Become a Right?”), F432-05483 (Muḥammad ʿAbduh, Rabīʿ al-Awwal 1314 AH, location 2743), and F432-05549 (Muḥammad ʿAbduh, Dhū al-Ḥijja 1317 AH, location 2776); the denominator for distribution in F432-05493 (Ḥasanayn Muḥammad Makhlūf, location 2748, whose heading reads “Jumādā al-Thāniya 1272 AH - 8 March 1953”, with the Hijri year corrupted from 1372 AH, as indicated by the Gregorian date), F432-05693 (Ḥasanayn Muḥammad Makhlūf, Shaʿbān 1372 AH / 27 April 1953, location 2848), and F432-06111 (Muḥammad Khāṭir, Rabīʿ al-Awwal 1391 AH / 24 May 1971, location 3057); the stipulated quota in F432-06031 (ʿAbd al-Raḥmān Qarāʿa, Ṣafar 1340 AH / 25 October 1921, location 3017) and F432-05621 (ʿAbd al-Majīd Salīm, Ṣafar 1347 AH / 31 July 1927, location 2812); capacity derived from funds in F432-05891 (Bakrī al-Ṣadafī, Rabīʿ al-Thānī 1327 AH, location 2947), the al-Minshāwī Hospital fatwā; and the hypothetical example used in reasoning in F432-05765 (Muḥammad Bakhīt, Shaʿbān 1333 AH, location 2884). Hijri dates are retained here because they are stated in the headings. A Gregorian date is added only where the heading supplies one, as established in section 3. ↩
-
The seven instances with their identifiers: F432-05471 (ʿAbd al-Majīd Salīm, Rabīʿ al-Thānī 1361 AH / 27 April 1942): “Its revenue shall be paid to ten reciters who have memorised the manifest Book of God”; F432-05965 (Ḥassūna al-Nawāwī, Rabīʿ al-Awwal 1316 AH): “And to five reciters, 300 equally”; F432-05993 (Bakrī al-Ṣadafī, Rabīʿ al-Awwal 1328 AH): “Payment shall be made to eleven Sufis, the eleventh being their shaykh. What is paid to them shall be divided into twelve shares: two for the shaykh and ten for the ten Sufis, equally”; F432-05995 (Bakrī al-Ṣadafī, Rabīʿ al-Awwal 1322 AH): “And for the remuneration of ten reciters who have memorised the Book of God and complete a recitation of the Qurʾān every Thursday eve at the al-Aḥmadī Shrine”; F432-05997 (Bakrī al-Ṣadafī, Dhū al-Ḥijja 1332 AH, location 3000): “He stipulated in his aforementioned waqf eighteen bearers of the Qurʾān from the aforementioned community, each to read a part of the glorious Qurʾān every day”. The community consists of the Jabartīs residing in Medina, so the instance is read as evidence of the stipulated-quota formula, rather than as an Egyptian case; F432-06021 (Muḥammad Bakhīt, Rajab 1336 AH / 19 April 1918): “The construction of an elementary school above the public drinking fountain … to teach twenty Muslim children the reading of the glorious Qurʾān, writing and the oneness of God”; F432-05783 (ʿAbd al-Majīd Salīm, Rajab 1351 AH / 31 October 1932): “A committee composed of five persons, two of them senior entitlement-holders of this waqf, elected by all the entitlement-holders”. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “al-Waqf ʿalā al-ʿulamāʾ wa-ṭalabat al-ʿilm” [“Waqf for Scholars and Students”], fatwā of Bakrī al-Ṣadafī, Jumādā al-Thāniya 1331 AH, location 2883 (F432-05763). The question is from Muḥammad Ḥasanayn, shaykh of al-Aḥmadī Mosque, concerning the waqf of al-Sayyid Aḥmad al-Ḥalabī, approximately 52 feddans in the al-Sanṭa area of Gharbiyya. An attribution clarification is necessary: the passages quoted in the main text come from the answer block, rather than the “Principles” block. “The reciters of the Qurʾān reading circle become individually designated through designation owing to the possibility of their enumeration” comes from the “Principles”, and is therefore the collection editor’s formulation. The answer reads: “As for the reciters of the aforementioned Qurʾān reading circle, they become individually designated through designation because they can be enumerated”. The answer also states: “There are three opinions concerning those who can be enumerated. Abū Yūsuf’s position is that those who can be enumerated are those whose exhaustive enumeration requires neither a written record nor calculation, unlike those who cannot be enumerated. Muḥammad’s position is that if they number more than a hundred, they cannot be enumerated. Others say that the matter is left to the judge’s judgement, and this is the position adopted for fatwā. Muḥammad’s position is the easier. As for the position adopted for fatwā, al-Ikhtiyār [The Choice] states that it is the preferred and more cautious position. Thus it appears in al-Durr [The Choice Pearl] and its glosses”. Earlier in the answer, there is a verbatim quotation from al-Fatāwā al-Hindiyya [The Indian Fatwās], which supplies the bridge discussed in section 6: “He dedicated his estate as waqf for the poor of his relatives and his village, and they either cannot be enumerated, can be enumerated, or one group can be enumerated while the other cannot. In the first case, the administrator may assign half the yield to his poor relatives and half to the poor of the village, then give to whichever members of each group he chooses and favour some as he wishes, because his intention is charity, and such is the rule in charity. In the second case, the yield is distributed to the two groups according to their numbers, and he may not favour some over others, because his intention is a bequest, and such is the rule in bequests. In the third case, he first divides the yield between the two groups, paying those who can be enumerated according to their numbers and assigning a single share to those who cannot be enumerated. He then gives this share to those who cannot be enumerated whenever he wishes and favours some within this share, as we have explained”. Al-Fatāwā al-Hindiyya was not examined directly; the quotation is transmitted through the fatwā text. The heading supplies no Gregorian date, so only the Hijri date is given. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “al-Iqrār bi-l-istiḥqāq fī al-waqf” [“Acknowledgement of Entitlement in a Waqf”], fatwā of ʿAbd al-Majīd Salīm, Dhū al-Qaʿda 1358 AH / 3 January 1940, location 3033 (F432-06063): “This account book among the case papers contains only an enumeration of the waqf’s receipts and expenditure and the distribution of the net amount among the entitlement-holders”. ↩
-
The identifiers of the seventeen records appear in Appendix B, paragraph 9. The category’s definition and lexical indicators are stated in the main text. Every passage was read in context. Figures expressing a quota stipulated in a deed or a denominator for distribution per head were not included in this category. One example is “the nineteen persons for whose benefit the waqf was established” in F432-05689, where the number is stipulated in the deed. That record was included because of another passage restricting entitlement to a single person after the others had died. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Waqf khayrī” [“Charitable Waqf”], fatwā of ʿAbd al-Raḥmān Qarāʿa, Rajab 1344 AH / 30 January 1926, location 2808 (F432-05613). The question came in a letter from the Office of the Shaykh of al-Azhar dated 14 January 1926, no. 2790. It concerns the waqf of Zaynab bint Saʿīd for students from the island of Djerba residing at al-Azhar, and the sum held in al-Azhar’s treasury for 1924 and 1925. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Waqf bi-sharṭ al-suknā” [“Waqf with a Residence Stipulation”], fatwā of Ḥassūna al-Nawāwī, Dhū al-Qaʿda 1315 AH, location 2874 (F432-05745). The heading supplies no Gregorian date. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Tanāzul ʿan al-naẓar” [“Relinquishment of Waqf Administration”], fatwā of Muḥammad Bakhīt, Shaʿbān 1334 AH / 8 June 1916, location 2951 (F432-05899). ↩
6 · The Threshold of Enumerability: A Source-Based Examination of the Juristic Question
The two tests show that the waqf system did not lack a concept of enumeration; it made enumeration an explicit condition. In Ḥanafī jurisprudence, this appears in the question of whether “waqf recipients can or cannot be enumerated”. The question must be examined here because it supplies the rule that makes enumerability part of the legal characterisation of the right itself.
6.1. Three Legal Consequences of Exhaustive Enumeration in the School’s Texts
Three doctrinal consequences of enumerability and the capacity for exhaustive enumeration are established in Ḥanafī texts
First consequence (the validity of the waqf):
Al-Sarakhsī states in “Sharḥ Kitāb al-waqf” [Commentary on the Book of Waqf], quoted and critically discussed by Ibn Māza in al-Muḥīṭ al-Burhānī, that a waqf is valid for a disbursement purpose expressly stipulating poverty and need, whether its recipients can be enumerated or not. Where rich and poor are treated alike, the ruling divides:
-
Valid in respect of their individual persons, by way of conferring ownership: (if they can be enumerated)
-
Invalid: (if they cannot be enumerated)
Two independent sources examined directly support this ruling: the wording of al-Mabsūṭ itself and that of Ibn Nujaym in al-Baḥr al-rāʾiq18. Enumerability is thus a condition of validity in a particular configuration.
Second consequence (the rule of financial distribution):
Al-Isʿāf states that if orphaned relatives can be enumerated on the day the waqf is created, every one of them existing on that day is entitled, rich or poor. If they cannot be enumerated, the yield goes to the poor among them rather than the rich, and the administrator may give it to whichever of them he chooses19. (Exhaustive enumeration thus moves the right from an entitlement individuated for every person to an allocation made by the administrator at his discretion.)
Third consequence (a change in the right when the instrument of enumeration changes):
Exhaustive enumerability is a variable condition rather than a fixed attribute. The same passage in al-Isʿāf states: “Then, whenever they become enumerable, the rich share in it with them”. The text therefore entails that a group’s transition from non-enumerability to enumerability changes the rule governing distribution to it. This establishes that the unit of counting is a legal variable with operative effect, rather than an external description of reality.
6.2. Defining the “Threshold of Enumerability” and the Numerical Threshold (One Hundred)
The threshold of enumerability itself is the subject of a specified disagreement. In al-Bināya, al-ʿAynī reports three opinions from Fatāwā Qāḍī Khān. For Abū Yūsuf, the threshold is that they cannot be enumerated by a written record or calculation; another opinion uses a pebble criterion. Muḥammad says that if they number more than a hundred, they cannot be enumerated. Others say that the matter is left to the judge’s judgement, which is the position adopted for fatwā, while “what Muḥammad said is the easier”20.
Attribution of the numerical position to Muḥammad ibn al-Ḥasan requires investigation, since al-Bināya does not clearly separate the pebble criterion from the hundred-person criterion. A more explicit parallel account in the chapter on bequests resolves this. Al-Nihāya by al-Sighnāqī is a commentary on al-Hidāya predating al-Bināya and drawing on the same Qāḍī Khān. It introduces the pebble criterion with “it is said”, without attribution, and explicitly attributes the numerical criterion to Muḥammad21. This is supported by the answer in the 1331 AH fatwā cited in this paper, which gives the same attribution and refers its transmission to al-Durr and its glosses22. Sources in the waqf chapter, however, give a different attribution for the numerical thresholds, recorded here without preferring it. Al-Isʿāf and al-Baḥr al-rāʾiq report them in the opposite arrangement: ten from Muḥammad and a hundred from Abū Yūsuf23. The attribution is therefore resolved only for the account in the bequests chapter and cannot be extended to the waqf chapter. The accounts are set out in their wording in Appendix D.
The location of this material in the two books must be stated explicitly. Al-Bināya and al-Nihāya discuss the threshold of enumerability in the chapter on bequests to a group of people, rather than in the waqf chapter. Transferring it to waqf consequently requires an evidentiary bridge. That bridge is explicit in the sources, rather than inferred by this paper. Al-Isʿāf and al-Sarakhsī’s wording establish the rule of exhaustive enumeration specifically for waqf. Moreover, al-Fatāwā al-Hindiyya, as quoted within the answer to the 1331 AH fatwā, grounds the distinction between the two chapters in the founder’s intention. For those who can be enumerated, “the yield is distributed to the two groups according to their numbers, and he may not favour some over others, because his intention is a bequest”. Those who cannot be enumerated receive “a single share”, which the administrator gives to whichever of them he chooses, because the founder’s intention is charity24. Where counting is possible, it supplies the denominator for distribution; where it is not, the allocation becomes a collective share. Through this text, the muftīs establish the bridge between the two chapters.
The accounts in al-Bināya and al-Nihāya contain three criteria of different kinds. One is technical, making exhaustive enumeration depend on two counting instruments: written records and calculation. Another is numerical, setting a threshold of a hundred. As noted above, the two accounts differ over the attribution of these criteria, but their respective character is consistent in both. The criterion adopted for fatwā is procedural, assigning the assessment to the judge. The first criterion’s relevance to this paper is evident: making the threshold of enumerability depend on the existence of a register and the possibility of calculation poses precisely the question addressed by the history of quantification in the nineteenth century, albeit in a different language and for a different purpose.
The rule for deriving instances of application from the corpus is explicit. Enumeration expressions matched eleven records, of which six instances were included and five excluded for recorded reasons. The first excluded record concerns the enumeration of receipts and expenditure in a waqf account book, rather than of waqf recipients. The remaining four contain “maḥṣūr” in the sense of restriction or limitation, rather than enumerability. The eleven identifiers and the disposition of each record appear in Appendix (B).
6.3. Instances of Application in Dār al-Iftāʾ Fatwās (1314–1337 AH)
- First instance: (Shaykh Muḥammad ʿAbduh, Dhū al-Ḥijja 1314 AH / May 1897)
A waqf for students at al-Aḥmadī Mosque, with a question from Ibrāhīm Bey on interpreting the founder’s stipulation. The muftī ruled that “distribution among them all shall follow the judgement of the waqf’s nāẓir and the shaykh of al-Aḥmadī Mosque, without unfairness and on the basis of the founder’s stipulation. This applies if the aforementioned students can be enumerated. If they cannot be enumerated, it is to be paid to some of them; the determination of what cannot be enumerated is entrusted to the judge’s judgement, according to the position adopted for fatwā”25.
The connection between this instance and the paper’s introduction is deliberate. Al-Aḥmadī Mosque in Ṭanṭā supplied the figure for 1875 in the statistical digest. Approximately two decades later, a question about its students receives an answer allowing for the possibility that they cannot be enumerated. The two answers do not contradict one another. Enumeration in the statistical apparatus is an observational procedure producing a figure; enumeration in jurisprudence is a condition with consequences for a right. The same group may be countable in one system and non-enumerable in another because the two systems ask different questions. Here the paper’s thesis is presented in a single instance.
- Second instance (Shaykh Bakrī al-Ṣadafī, Jumādā al-Thāniya 1331 AH / May 1913)
The waqf of Aḥmad al-Ḥalabī in the al-Sanṭa area, one-third of which had devolved to twenty-four reciters, twenty-five students and twenty-five scholars. The shaykh of al-Aḥmadī Mosque asked about the fact that the number actually present exceeded the number entitled under the waqf deed. The muftī ruled that scholars and students do not become individually designated through designation because they cannot be enumerated, whereas the reciters of the Qurʾān reading circle do become individually designated because they can be enumerated26. Exhaustive enumerability is therefore not simply an attribute of a group, but of its relationship to an authority that keeps track of it: the two categories differ within a single waqf. At the end of the answer, the muftī also favours prohibiting payment to one person in two capacities, so as to spread the benefit more widely. This is an explicit provision against double counting, to be read alongside section 11. The instance’s four elements are detailed in Appendix (D).
- Third instance (Shaykh Muḥammad Bakhīt, Muḥarram 1337 AH / 30 October 1918)
The waqf of Muṣṭafā Bey ʿUmar for the sharia sciences at the Religious Institute in Asyūṭ. The question came from the governor of Asyūṭ and concerned the distribution of two years’ rent. The muftī ruled that the first disbursement purpose includes all scholars and students: “If they can be enumerated, meaning that they number fewer than a hundred, their portion is given to them according to the number of their heads. If they are among those who cannot be enumerated, meaning that they number a hundred or more, the nāẓir has a choice: he may give to all equally or unequally, or he may give to some and not to others”27.
This is the clearest evidence found in the records examined of the effect of the unit of counting on the structure of the right. The numerical threshold is explicit: a hundred. Below it, each individual has a share calculated “according to the number of their heads”, making the right individual and calculable. Above it, the matter passes to the nāẓir’s discretion: he may give to some and not to others, and even to scholars rather than students. Crossing a single numerical threshold transforms the entitlement-holder from the holder of a calculated share into a candidate for discretionary allocation. No claim is made to a comprehensive survey of the entire corpus, since its complete manual coding remains pending.
A qualification of the inference is necessary here. The threshold is a transmitted juristic standard, rather than the result of a count conducted among students at the Asyūṭ institute. The fatwā states the rule under two possible conditions without reporting which one actually obtained. This is an instance of stating a counting rule, rather than an instance of counting. The distinction must remain explicit.
The remaining three instances complete the picture. The first is Bakrī al-Ṣadafī’s fatwā of Dhū al-Ḥijja 1324 AH on a waqf for poor relatives. The second is his fatwā of Jumādā al-Thāniya 1327 AH on a charitable waqf for the staff of a mosque and a zāwiya (small religious lodge). The third is Muḥammad Bakhīt’s fatwā of Shawwāl 1335 AH / 16 August 1917, on an entitlement waqf28. The rule operates in the same way throughout: if they can be enumerated, each has an individual share; otherwise, the matter passes to allocation by the administrator. Their texts appear in Appendix (D).
6.4. Four Implications Established by This Section
This section yields four implications. First, Ḥanafī jurisprudence contains a concept of enumeration predating the statistical apparatus by centuries. Second, it is an operative rather than descriptive concept, affecting the validity of a waqf and its rule of distribution. Third, its threshold is disputed, and one definition makes it depend on the two instruments of record-keeping and calculation. Fourth, Egyptian muftīs applied it to particular cases between 1314 AH and 1337 AH, making number a condition of a right rather than merely a description of a population. The thesis of a transition from the uncounted to the counted therefore fails here.
Footnotes
-
A necessary attribution clarification, added after inspecting the original. The passage is in Burhān al-Dīn Ibn Māza al-Bukhārī, al-Muḥīṭ al-Burhānī fī al-fiqh al-Nuʿmānī [The Burhānī Compendium of Nuʿmānī Jurisprudence], p. 116. It is explicitly quoted from al-Sarakhsī, rather than composed by Ibn Māza, and reads: “The eminent shaykh and imam, Shams al-Aʾimma al-Sarakhsī, may God have mercy on him, in Sharḥ Kitāb al-waqf [Commentary on the Book of Waqf]: The conclusion for this class of questions is that whenever he specifies a disbursement purpose expressly stipulating poverty and need, the waqf is valid whether they can be enumerated or not. His words ‘whether they can be enumerated or not’ indicate that perpetuity is not a condition, whereas we have previously stated the contrary. Whenever he specifies one in which rich and poor are treated alike … if they can be enumerated, it is valid in respect of their individual persons, meaning that it is valid by way of conferring ownership; if they cannot be enumerated, it is invalid”. The intervening sentence is Ibn Māza’s critical comment on the implication of al-Sarakhsī’s wording; it is recorded here rather than omitted. “Meaning that it is valid by way of conferring ownership” is Ibn Māza’s explanation, rather than the wording of the author quoted. The page number comes from the indexed version examined; its volume number was not established. Two independent supporting sources were examined directly. The first is Abū Bakr al-Sarakhsī, al-Mabsūṭ [The Expanded Treatise], p. 34: “The conclusion is that whenever he specifies a disbursement purpose expressly stipulating poverty and need, it is valid whether they can be enumerated or not”. The second is Zayn al-Dīn ibn Ibrāhīm ibn Muḥammad, known as Ibn Nujaym, al-Baḥr al-rāʾiq sharḥ Kanz al-daqāʾiq [The Clear Sea: A Commentary on the Treasure of Subtleties], second edition, eight volumes, p. 215. Its original was consulted for this version after it had previously been quoted through an intermediary. It reads: “The conclusion is that whenever he specifies a disbursement purpose expressly referring to the poor and to need, the waqf is valid whether they can be enumerated or not. His words ‘can be enumerated’ indicate that perpetuity is not a condition. Whenever he specifies a disbursement purpose in which rich and poor are treated alike, if they can be enumerated it is valid by way of conferring ownership; if they cannot be enumerated, it is invalid”. Comparing the original with the intermediary yields two points worth recording. First, Abū Zahra’s wording in Muḥāḍarāt fī al-waqf [Lectures on Waqf], p. 96, from which the quotation had been taken, omits without ellipsis the sentence “His words ‘can be enumerated’ indicate that perpetuity is not a condition”—the very sentence recorded by Ibn Māza in al-Muḥīṭ. Second, Ibn Nujaym introduces the passage with a quotation from Shams al-Aʾimma: “Accordingly, if he establishes a waqf for the students of a particular town, it is permissible, because poverty predominates among them, so the designation indicates need”. This is the basis of the ruling in the waqf cases concerning students presented below. A correction to attribution in an earlier version: the statement “the asset passes to them by way of conferring ownership, and the disposition is not restricted to the yield alone” had been attributed to Abū Zahra as his own doctrinal derivation. In its context, it is his inference from Ibn Nujaym’s preceding statement, introduced in the third person (“for he says”). Abū Zahra then qualifies it in the same context: “This is undoubtedly so where the wording of the waqf expressly indicates that the disposition confers ownership … If the wording does not indicate conferral of ownership, the formulation is invalid from its inception”. See also al-Muḥīṭ al-Burhānī, p. 115, on the requirement of perpetuity. ↩
-
Ibrāhīm ibn Mūsā ibn Abī Bakr al-Ṭarābulusī al-Ḥanafī, al-Isʿāf fī aḥkām al-awqāf [Assistance in the Rulings Governing Awqāf], Egypt: Maṭbaʿat Hindiyya, al-Mahdī Street, al-Azbakiyya, second edition, 1320 AH / 1902, p. 139, in the “Chapter on Waqf for Orphans, Widows, Unmarried Women, Previously Married Women and Virgins”. The two passages read: “If they can be enumerated on the day the waqf is created, every one of them existing on that day is entitled to it, rich or poor, and any orphans subsequently arising among them share with them … If they cannot be enumerated on the day the waqf is created, nor can those subsequently arising among them be enumerated, the yield belongs to the poor among them rather than the rich, and the administrator may give it to whichever of them he chooses. Then, whenever they become enumerable, the rich share in it with them”; and “If he says, ‘This land of mine is a charitable waqf for God Almighty in perpetuity for the unmarried women among my relatives’ … if they can be enumerated, the waqf is valid and its yield is paid to them; if they cannot be enumerated, it is not valid for them, because we do not know to whom the yield should be given”. ↩
-
Badr al-Dīn al-ʿAynī, al-Bināya sharḥ al-Hidāya [The Edifice: A Commentary on the Guidance], edited by Ayman Ṣāliḥ Shaʿbān, Beirut: Dār al-Kutub al-ʿIlmiyya, first edition, 1420 AH / 2000, thirteen volumes, p. 473. The indexed version examined reads: “Then the threshold of enumerability according to Abū Yūsuf is that they cannot be enumerated by a written record or calculation, so they cannot be enumerated. It is also said: such that pebbles are counted with them until a child is born among them and someone among them dies, and this is Muḥammad’s opinion; if they number more than a hundred, they cannot be enumerated. Others say that the matter is left to the judge’s judgement, which is the position adopted for fatwā; what Muḥammad said is the easier. Thus in Fatāwā Qāḍī Khān [The Fatwās of Qāḍī Khān]”. The volume number was not established in the copy examined. The editor of al-Nihāya refers to al-Bināya 13/472 at the parallel passage, which offers an indication rather than establishing the volume. This wording does not clearly separate the pebble criterion from the hundred-person criterion; attribution must therefore be sought in the parallel account quoted in the following footnote. The text’s location was established after inspecting the entire page: it belongs to the chapter on bequests to a group of people, rather than the waqf chapter. Its immediate context before and after concerns bequests to the orphans, widows, unmarried women, blind persons and chronically disabled persons of a particular clan. It includes: “According to the three imams, the bequest is for all of them, whether they can be enumerated or not”, and a quotation from Mukhtaṣar al-Karkhī [Al-Karkhī’s Compendium] concerning a person who “bequeathed a third of his property to the orphans of a particular clan”. Transferring the threshold to waqf rests on the explicit bridge described in section 6, rather than on analogy. ↩
-
Ḥusām al-Dīn Ḥusayn ibn ʿAlī al-Sighnāqī, al-Nihāya fī sharḥ al-Hidāya [The Culmination in Explaining the Guidance], p. 170: “If they are a group that can be enumerated, both their poor and their rich are included in the bequest; if they cannot be enumerated, the bequest is for the poor among them. The threshold of enumerability reported from Abū Yūsuf is that they not be enumerated … It is also said: if the enumerator cannot enumerate them before a child is born among them and one of them dies, they cannot be enumerated. Muḥammad said: if they number more than a hundred, they cannot be enumerated. Others say that the matter is left to the judge’s judgement, which is the position adopted for fatwā. What Muḥammad said is the easier: thus in Fatāwā Qāḍī Khān [The Fatwās of Qāḍī Khān]”. The pebble criterion is introduced by “it is said”, without attribution, whereas the numerical criterion is explicitly attributed to Muḥammad; both commentaries quote the same source. The volume number was not established in the copy examined. The passage in al-Nihāya also occurs in the bequests chapter. Its immediately preceding wording is: “If they are a group that can be enumerated, both their poor and their rich are included in the bequest; if they cannot be enumerated, the bequest is for the poor among them”. The ellipsis in the wording of Abū Yūsuf’s criterion replaces a corrupted passage in the indexed version, which reads “that they not be enumerated, absorption and calculation for them, they cannot be enumerated”. Attribution of this criterion is therefore restricted to the sound wording in al-Bināya, without correcting the corrupted passage by conjecture. The different account in the waqf chapter was recorded after direct inspection in two separate passages. The first is Ibrāhīm ibn Mūsā al-Ṭarābulusī, al-Isʿāf fī aḥkām al-awqāf [Assistance in the Rulings Governing Awqāf], p. 14: “It is related from Muḥammad, may God have mercy on him, that what cannot be enumerated is ten, and from Abū Yūsuf, may God have mercy on him, that it is a hundred, which is the position adopted by some. Others say forty, and others eighty. The position adopted for fatwā is that the matter is left to the judge’s judgement”. The second is Ibn Nujaym, al-Baḥr al-rāʾiq [The Clear Sea], p. 215, quoting al-Isʿāf in corresponding wording: “In al-Isʿāf, it is related from Muḥammad that what cannot be enumerated is ten, and from Abū Yūsuf a hundred, which is the position adopted by some. Others say forty, and others eighty. The position adopted for fatwā is that the matter is left to the judge’s judgement”. Two implications follow. First, the numerical threshold in the waqf chapter is attributed to Abū Yūsuf rather than Muḥammad, while Muḥammad has a different threshold there: ten. Second, the waqf chapter contains five opinions rather than three, with forty and eighty added. This paper does not prefer either account, since resolving attribution is not its subject. It records only that arguments in the main text relying on attribution of the threshold to a particular imam are confined to the account in al-Bināya and al-Nihāya. The ruling on which the paper relies does not depend on attribution at all: it turns on a numerical threshold separating a right calculated per head from discretionary allocation. This is established in both accounts and in the wording of Muḥammad Bakhīt’s fatwā of Muḥarram 1337 AH, without naming the proponent. ↩
-
See note 12. ↩
-
See note 21. ↩
-
See note 12. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Waqf ʿalā ṭalabat al-ʿilm bi-l-Jāmiʿ al-Aḥmadī” [“Waqf for Students at al-Aḥmadī Mosque”], fatwā of Muḥammad ʿAbduh, Dhū al-Ḥijja 1314 AH, location 2982 (F432-05961). It concerns a waqf of agricultural land in Kafr al-ʿArab, Daqahliyya. The founder stipulated that eight shares out of twenty-four be paid to students at al-Aḥmadī Mosque “according to the judgement of the original nāẓir and whoever is shaykh over the scholars … without unfairness”. The heading supplies no Gregorian date, so only the Hijri date is given; Dhū al-Ḥijja 1314 AH falls in 1897, not 1895. ↩
-
See note 12. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Waqf khayrī” [“Charitable Waqf”], fatwā of Muḥammad Bakhīt, Muḥarram 1337 AH / 30 October 1918, location 2797 (F432-05591). It concerns the waqf of Muṣṭafā Bey ʿUmar for the Islamic Institute of Religious Learning in the city of Asyūṭ. The question came from the governor of Asyūṭ under no. 2950. The passage quoted in the main text comes from the answer block. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ]: “Waqf ʿalā al-qarāba al-fuqarāʾ” [“Waqf for Poor Relatives”], fatwā of Bakrī al-Ṣadafī, Dhū al-Ḥijja 1324 AH, location 2745 (F432-05487); “Waqf khayrī” [“Charitable Waqf”], his fatwā of Jumādā al-Thāniya 1327 AH, location 2785 (F432-05567); and “Waqf istiḥqāq” [“Entitlement Waqf”], fatwā of Muḥammad Bakhīt, Shawwāl 1335 AH / 16 August 1917, location 3006 (F432-06009), which refers to Ibn ʿĀbidīn, Radd al-muḥtār [The Reply of the Perplexed], vol. 3, p. 629, Amīriyya edition of 1286, and to al-Anqarawiyya [The Anqarawī Fatwās], vol. 1, p. 220. Neither edition was examined for this paper; the references are transmitted through the fatwā text. ↩
6.A · The Replacement Test: The German Purpose Foundation as Comparator
The result is mixed. The purpose foundation under the German Civil Code (BGB) (Stiftung bürgerlichen Rechts) was put in the place of waqf in this paper’s argument: one boundary held, while another collapsed. As regards the counting apparatus and the absence of an aggregate figure, waqf is not distinguished in this respect from its comparator within the family of purpose institutions. But treating the possibility of enumeration as a variable in the legal characterisation of the right fails under replacement.
The comparator is neutral and symmetrical: neutral because it is not the other party to the comparison undertaken in this paper, which is the Egyptian statistical and planning apparatus; symmetrical because it is a regular form within its system, not a marginal exception. It is regulated within the Civil Code alongside the association, and such foundations number in the tens of thousands.
The common ground is this: the German Civil Code does not confer supervisory powers over the foundation on recipients of its benefits. This was established in the judgement of the Federal Court of Justice of 15 December 2016. ⚠️ The judgement predates the 2021 reform, which entered into force on 1 July 2023 and changed both the numbering and the substance of the sections. It therefore supports the absence of supervisory powers, but cannot support a claim that the provisions currently in force contain no provision governing the recipient’s relationship with the foundation, as established in Appendix E. The negative claim must remain within its limits: what is absent is any supervisory power, not regulation of the relationship as a whole. §§ 86 no. 4 and 86a no. 3 protect the claims of those to whom the charter grants benefits, while § 86e(2) requires them to be heard before the authority’s decision. These claims derive from the charter, not the statute. The system also draws a distinction resembling the Ḥanafī distinction. If the charter lays down objective qualifying descriptions through which status is acquired without any discretionary selection, the beneficiary has a justiciable claim. If a governing body is authorised to select from a defined circle, there is no claim, only a “legally undefined opportunity” (nicht rechtlich definierte Chance). Its counterpart in the juristic texts is that those who can be enumerated become individually designated, whereas, if they cannot be enumerated, the administrator may give to whomever he wishes. Nor does the German system require an aggregate count of recipients. The absence of an aggregate figure from the corpus therefore establishes no distinction.
Replacement fails at the criterion governing this division. In German law, the criterion lies solely in how the founder has framed the instrument; number has no bearing on it. The circle may comprise a thousand and its qualifying descriptions establish a claim, or it may comprise three and the delegation of selection defeats a claim. In Ḥanafī jurisprudence, the criterion is an attribute of the group itself: whether it can be enumerated. This is determined by writing and calculation according to Abū Yūsuf, by the threshold of one hundred according to Muḥammad, or by the judge’s judgement. ⚠️ These attributions follow the account in the chapter on bequests in al-Bināya and al-Nihāya. The chapter on waqf in al-Isʿāf and al-Baḥr al-rāʾiq reverses them, assigning one hundred to Abū Yūsuf and ten to Muḥammad, as established in section 6 and footnote 20. This section’s conclusion does not depend on the attribution: it turns on the criterion being an assessed number, which holds in both accounts. Its effect extends to the validity of the waqf from its inception, not merely to claimability. It changes as circumstances change, without an amendment to the instrument or a decision by a governing body: “Then, whenever they become enumerable, the rich share in it with them.” There is no counterpart to this in the comparator’s provisions. The conditions for recognition in § 82 concern form and finances, accompanied by a single negative substantive condition: the foundation must not endanger the public good. None concerns the circle of beneficiaries or their number. The charter may be amended only under the conditions of § 85 and with the approval of the federal state’s authority (§ 85a(1) sentence 2).
The limits of this test are explicit. It was conducted on the comparator’s provisions reproduced in Appendix E and on the 2016 judgement. Specialist German legal scholarship was not examined, an acknowledged gap. The negative finding concerns the absence of a provision from the sections examined, not its absence from the system as a whole. The passages cited were checked against the text in force, and the sections were identified by number. A citation concerning the register of foundations was removed because its entry into force had been deferred. No conclusion is based on a provision whose operation has been deferred. The mixed result remained unchanged.
Re-examination using a comparator from a third family narrows the distinctive component without eliminating it. Within this programme, the result of replacement was tested again against the Hindu religious and charitable endowments system in Tamil Nadu and a French control, reversing the assessment of six elements. One bears on the central argument of this section: the gap-filling authority and the source of the defining attribute—whether it lies outside the founding instrument or within it.
The challenge arises because the third comparator has a normative gap-filling authority external to the instrument, arising from circumstances rather than from the act of the person allocating the property. Section 6 paragraph 20 of the Tamil Nadu law establishes temple status through several routes, including long-standing use as of right, without any act of dedication. Section 28 paragraph 1 binds the trustee to the usage of the institution alongside the terms of the trust.29 The implication is that the claim fails in its general form: a criterion external to the instrument that changes with circumstances is not a distinctive feature of Islamic jurisprudence. It is shared with a system from a third family that recognises waqf and does not belong to the Romano-Germanic family.
What remains after this narrowing is more limited and precise: the criterion must be a number assessed through writing and calculation or a specified threshold, and its effect must extend to the validity of creation, not merely to claimability. No counterpart combining these two features was reached in the third comparator. Narrowing the claim therefore gains precision rather than losing the result: the distinctive feature becomes a testable quantitative attribute, replacing a general claim about the source of the attribute that the first comparator outside the family falsifies.
The description of the recognition condition must also be qualified. Our statement that § 82 contains a single negative substantive condition is a judgement about the German text, not its family. French law belongs to the same family and requires a positive attribute of the purpose, as does Indian law. This qualification does not affect this section’s result, since neither positive attribute involves counting the circle of beneficiaries or a condition concerning whether they can be enumerated.
The finding concerning the beneficiary’s status remains unchanged and gains further support: the three cited provisions—§ 86 no. 4, § 86a no. 3 and § 86e(2)—are reproduced in Appendix E. These are precisely the provisions that changed the description of beneficiary status from absolute absence to variation. This had already been established here before that reversal, so it requires no correction to this section.
The limits of this addition are explicit: no replacement test using the third comparator was carried out on this paper’s texts. The provisions cited from it were read in the comparator file adopted by this programme, and its law is the law of a single state, which was not compared with the laws of other states. What has been established is a narrowing of the claim, not a second test.
Footnotes
-
Tamil Nadu Hindu Religious and Charitable Endowments Act, 1959, section 6 paragraph 20 and section 28 paragraph 1. The first reads: ““Temple” means a place … dedicated to, or for the benefit of, or used as of right by, the Hindu community … as a place of public religious worship”; the second reads: “the trustee of every religious institution is bound to administer its affairs and to apply its funds and properties in accordance with the terms of the trust, the usage of the institution and all lawful directions which a competent authority may issue in respect thereof”. Scope of the evidentiary record: quoted through the comparator file adopted by this programme; the state gazette was not examined, and this is the law of a single state whose provisions cannot be generalised. ↩
7 · Units of Counting in the Material: Four Coexisting Operations
This paper distinguishes four documented counting operations, not four successive stages. These accounting systems may coexist in the same year and institution: an al-Azhar riwāq keeps a register of names, distributes regular provisions in daily units, has its members counted in an educational table, and has a decision concerning its waqf entered in a register. This coexistence cannot be converted into a linear chronological succession.
- The first operation: counting entitlement
Its object is the person as the holder of a share recognised by sharia; its accounting unit is the registered name or stipulated person-slot; its inclusion rule is status under the sharia deed or position in the register. It preserves identity to the highest degree, since the name is linked to the entry standard, rank and share. Its scope of aggregation is as narrow as possible, stopping at a single waqf or riwāq. The precision of this operation therefore produces no knowledge of the total number of beneficiaries in the country: it was designed for distribution within a particular patrimony.
- The second operation: counting services or benefits
Its object is the portion disbursed, and its unit is the loaf, garment or monetary amount. The clearest evidence in the corpus is an account statement included in a case dated (Muḥarram 1372 AH / 19 October 1952), with entries including: “8 pounds and 500 milliemes for cloth for the Qurʾān teacher and pupils of the Qurʾān school at Ḥawsh al-Mujāwirīn, comprising 85 metres of dablān cloth; 6 pounds and 600 milliemes for fezzes and shoes for the Qurʾān teacher and pupils of the Qurʾān school at Ḥawḍ al-Mujāwirīn, comprising 11 fezzes and the same number of pairs of shoes; 40 pounds for bread for distribution; 20 pounds for beans for distribution”30.
This single statement reveals the entire problem of conversion. The entry for fezzes gives an explicit count of eleven (11), together with a corresponding number of pairs of shoes, allowing the number of beneficiaries to be derived with reasonable certainty: (eleven people, including the Qurʾān teacher). The bread and bean entries, however, give a value in pounds without a count, since they were disbursed “for distribution”. The units within this single statement are not uniform: one item is counted in units that indicate persons, while another is valued in money and does not indicate persons. If this waqf’s expenditures were aggregated in a central table, all would appear as a single row under charitable expenditure, and the distinction between what can and cannot be counted would disappear.
The effect of a change in the form of benefit appears in two cases separated in time. The first is Shaykh Muḥammad ʿAbduh’s fatwā of Dhū al-Ḥijja 1314 AH, which gives students at al-Aḥmadī Mosque “the choice between receiving the amount allocated to them in cash or receiving regular provisions of equivalent value”31. The second is Shaykh Muḥammad Bakhīt’s fatwā of Rabīʿ al-Awwal 1334 AH / 1 February 1916 on the regular provisions of the Disūq Institute, responding to an enquiry from the Secretariat of the Supreme Council of al-Azhar: “The entitlement-holders have the choice: if they wish, they may take the bread itself as stipulated by the founder, and if they wish, they may take its value in cash”32.
The difference between the two cases over these twenty years lies not in the ruling but in the requesting party. In the first, an enquirer asks about a founder’s stipulation; in the second, an administrative council requests a fatwā before issuing a general regulatory decision for an institute. The juristic ruling remains constant, while its position within the decision-making cycle changes. Replacing the unit of counting matters because a loaf is a daily unit in kind that cannot be aggregated with other units, whereas cash is a periodic value that can enter a budget and be compared. Yet the possibility of financial aggregation does not establish a constant social value: purchasing power changes while the nominal figure remains the same.
- The third operation: counting institutional affiliation
Its object is the individual as a member of a school or institute, and its unit is the pupil or resident student. The most precise evidence in the corpus is Shaykh Bakrī al-Ṣadafī’s fatwā of Jumādā al-Ūlā 1325 AH, responding to an enquiry from the Ministry of Public Instruction. It concerns Aḥmad Pasha al-Minshāwī’s allocation of two hundred and fifty pounds annually “for the admission of a number of pupils who are children of poor Muslims” to the Ṭanṭā Government School. Its principles include three consecutive statements reproduced in Appendix (D)33. ⚠️ The “Principles” block was composed by the collection editor and is not the muftī’s wording, as established in section 3’s separation of the layers. What follows from it rests on the editor’s summary of the fatwā’s effect and is not attributed to Bakrī al-Ṣadafī.
These three principles provide a complete operational definition of the unit of counting. The inclusion rule is enrolment following authorisation, and a vacant place generates no right. Leaving the institution extinguishes the status. The unit here is therefore an occupied place, not capacity, and counting follows enrolment rather than personal entitlement. This contrasts with the previously discussed hospital text, where the number of patients is “as many as these funds can accommodate”. There, capacity is calculated from money; here, occupancy is calculated from the register. They have no common denominator.
A fourth restriction on the reference population appears in the fatwā of Shaʿbān 1333 AH. The muftī ruled that distribution of the sum dedicated as waqf for scholars of al-Aḥmadī Mosque was to follow the number of shares held by those present when the waqf was created, excluding those who obtained the ʿālim qualification afterwards34. The reference population is thus frozen at the date of the deed, rather than open to new entrants. This is a form of fixing the denominator that differs from the logic of the statistical table: the table counts those present at the date of observation, while the deed counts those present at the date of creation.
- The fourth operation: counting the apparatus
Its object is the document, waqf or office, and its unit is the decision or register (discussed in detail in section 9).
Footnotes
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Waqf ahlī wa-khayrī” [Family and Charitable Waqf], fatwā of Ḥasanayn Muḥammad Makhlūf, Muḥarram 1372 AH / 19 October 1952, location 2840 (F432-05677). Its heading carries the notice: “From this point, implementation pursuant to Law 180 of 1952 begins.” It concerns the waqf of Shaykh Muḥammad al-Inbābī, and the account statement appears in the main text of the enquiry. ↩
-
See note 25. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Istibdāl al-jirāya bi-l-nuqūd” [Conversion of Regular Provision into Cash], fatwā of Muḥammad Bakhīt, Rabīʿ al-Awwal 1334 AH / 1 February 1916, location 3003 (F432-06003). The enquiry came in a communication from the Secretariat of the Supreme Council of al-Azhar dated 31 January 1916, no. 1255. It concerns the regular provisions of the Disūq Institute and encloses a communication from the Office of the Shaykh concerning the stipulations in the waqf deeds of Aḥmad Bey al-Sharīf and al-Minshāwī Pasha. Article 115 of Qānūn al-ʿadl wa-l-inṣāf [Code of Justice and Equity] and Radd al-muḥtār [The Reply of the Perplexed] are two references cited within the fatwā. Neither was examined in this paper; the references are transmitted through the fatwā’s text, following the practice of footnote 21. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Waqf” [Waqf], fatwā of Bakrī al-Ṣadafī, Jumādā al-Ūlā 1325 AH, location 2782 (F432-05561). The enquiry came in a communication from the Ministry of Public Instruction, no. 2353, and concerns the waqf of Aḥmad Pasha al-Minshāwī, registered at the Ṭanṭā Sharia Court on 5 Dhū al-Qaʿda 1322 AH. The three principles quoted in the main text come from the “Principles” block, composed by the collection editor. They must be read as the editor’s summary of the fatwā’s effect, not the muftī’s wording. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “al-Waqf ʿalā al-mudarrisīn bi-l-Jāmiʿ al-Aḥmadī” [Waqf for Teachers at al-Aḥmadī Mosque], fatwā of Muḥammad Bakhīt, Shaʿbān 1333 AH, location 2884 (F432-05765): “The sum dedicated as waqf in shares for the scholars of al-Aḥmadī Mosque in the manner described is based solely on the number of shares of the scholars present when this waqf was created, excluding those who obtained the ʿālim qualification after its creation.” ↩
8 · Non-Equivalence of Units: What Prevents the Construction of a Single Series
8.1. The Structural Impossibility of Converting Between Statistical Units
The inability to convert statistical units or aggregate them into a unified time series is a structural result in this material, not a technical defect in the sources, because each unit of counting was designed to answer a different question. “Stock” counts those registered or present at a particular moment; “flow” counts those entering during a period; “exposure” counts instances of service receipt; the “unique person” removes repetition; the “portion” counts units disbursed; “capacity” counts possible places; and the “institution” counts bodies rather than persons.
Appendix (D) contains a table of these units, what they count, their problems of repetition and the limits of conversion between them. A single series cannot be constructed before identifying the possible conversions and the information lost in each. The waqf counting apparatus therefore uses multiple units that cannot be internally converted, and the “stipulated person-slot” is its unit in the context of the deed alone, not in every context.
8.2. Application to the Two Documented Cases
The effects of non-equivalence between statistical units appear in two documented cases in this paper’s material:
- The first case: (al-Aḥmadī Mosque in Ṭanṭā between 1875 and the beginning of the twentieth century)
The two figures are three thousand eight hundred and seventy-four resident students, followed by four hundred students. They occur in the same source, one in the main text and the other in its footnote, with no conversion linking them35.
Three reasons together prevent their inclusion in a single series. The first is the difference in terminology: a “resident student” resides in a riwāq and holds an entitlement, while a “student” is enrolled in lessons. The second is the difference in the counting body: the first figure is quoted from a general educational digest, whereas the available text attributes the second to no body. The third is that the text in the copy examined breaks off at “and in that year the number of resident students of”. The difference between the figures may therefore reflect a change in circumstances or an effect of classification. This source does not distinguish between them.
- The second case: (the al-Azhar statistical table of 1875)
The results of its arithmetic audit appear in Appendix (D); section 1 has already established that the discrepancy is confined to a single row. In addition, after the two figures the text mentions an increase “towards the end of eighteen hundred and seventy-five of about five hundred and sixty-four students”36. The same source thus contains both a stock and a flow. The text does not establish whether the increase is included in the stated stock or is additional to it. Adding them produces a figure the source does not state, while disregarding them discards information the source contains.
8.3. Rules Governing the Reading of Claims About Growth in Beneficiary Numbers
This section yields a rule governing the reading of any claim about growth in beneficiary numbers. Moving from counting names to counting disbursements raises the figure by counting the same person repeatedly; moving from a daily stock to an annual flow makes growth appear greater than the change in the number of individuals; and expanding the definition of an institution or the boundaries of coverage makes the difference an effect of classification. These are not technical reservations appended after the analysis: they are the substance and object of the analysis itself.
Footnotes
9 · Counting the Apparatus: Establishing the Unit Behind a Large Figure
9.1. Clarifying the Terminology and Identifying the Figure’s Administrative Meaning
A description of Ministry of Awqāf registers gives a total of thirty thousand six hundred and eleven (30,611) nāẓir reports. The historical source divides them into three categories according to the body that appointed the nāẓir (the sharia court), and their sum exactly equals the total37. The categories are detailed in Appendix (D).
Establishing the unit behind this figure, however, matters more than checking its arithmetic total. The same source states that each report is in fact “a decision issued by a sharia court appointing a nāẓir to a waqf, or appointing a new nāẓir alongside the existing nāẓir”38. The statistical unit is therefore an instance of judicial appointment, not a report submitted by a nāẓir about his waqf. Here, the word is an Egyptian administrative term whose meaning differs from its apparent meaning in general usage.
9.2. The Three Implications of This Clarification
Clarifying the unit behind this figure has three implications. First, the figure cannot belong to a beneficiary series: its objects are the waqf and the nāẓir, not the entitled person. Second, it cannot be converted arithmetically into the number of awqāf existing in Egypt. A single waqf generates multiple reports whenever its nāẓir changes or another nāẓir is appointed alongside him. The figure measures turnover in the office of nāẓir more than the size of the sector. Third, an increase in this number does not indicate greater visibility of the individual within the apparatus; it indicates the frequency of one type of administrative occurrence.
9.3. The Structural Paradox and the Relocation of Archival Detail
The central paradox is that an institution’s greater capacity to produce documents does not guarantee a greater presence of the beneficiary within them. The ministry may become better able to ascertain the number of awqāf, their revenues and their expenditures, while entitlement-holders remain within an aggregate entry. Yet the absence of names from a central table cannot support an inference that identity has disappeared. The numerical survey in section 5 established that the name remains the unit of counting even in observational cases. The formulation the material can sustain is that aggregation changes the location of detail: what appeared in the main body of the instrument of distribution moves to a file that must be requested, making access conditional on a dispute or request.
This is a testable hypothesis, and its test is specified: comparing the central report with the registers or files feeding into it for a particular waqf in a particular year. This test was not conducted in this paper, so the item remains subject to the conditions for upgrading claims.
A limitation concerning the source must be recorded here, not concealed. The table of registers in the copy examined presents four rows of nāẓir reports, reproduced in Appendix (D)39, but not all its rows were read. This is an acknowledged gap in examination, not an objection to the source.
Footnotes
-
Ibrāhīm al-Bayyūmī Ghānim, al-Awqāf wa-l-siyāsa fī Miṣr [Awqāf and Politics in Egypt], “E — Nāẓir Reports” section, visible page number 41 in the copy examined. It reads: “Their total number recorded in Ministry of Awqāf registers is 30,611 reports: of these, 1,471 appoint Egypt’s rulers—from Khedive Ismāʿīl to King Fārūq—as nāẓirs over certain awqāf… and 1,333 appoint the Minister of Awqāf, in his official capacity, as temporary nāẓir over the awqāf administered by the ministry until 1952. The remaining 27,807 reports concern nāẓirs drawn from private individuals.” It also states: “Each nāẓir report is a decision issued by a sharia court appointing a nāẓir to a waqf—or, in some cases, appointing a new nāẓir alongside the existing nāẓir.” The three categories sum to 30,611, exactly matching the total. ↩
-
See note 37. ↩
-
Ghānim, al-Awqāf wa-l-siyāsa fī Miṣr [Awqāf and Politics in Egypt], “A. Registers” section. Four rows concerning nāẓir reports were read in the copy examined: “Registers of reports concerning private individuals 74 / 24483 / reports appointing nāẓirs over awqāf”; “Registers of reports appointing the ministry 9 / 1333”; “Registers of reports concerning director nāẓirs 7 / 701”; “Registers of reports concerning royal nāẓirs 4 / 2168”. Not all rows of the table were read, so the rows read cannot be used to reconcile the total. ↩
10 · From Administrative Statistics to the Planning Category: Dating an Approach, Not Claiming Completion
10.1. Distinguishing Completed Bureaucratic Transformation from Completed Integration into Planning
The available fatwā and legislative material establishes changes in “body, scope and purpose”. Counting moves from a nāẓir, court and muftī to a specialised government agency and ministry; its scope moves from a waqf and riwāq to an institution and province; and its purpose moves from distribution and accounting to reporting. The internal material does not, however, establish a transition to the planning category in its full sense: defining a target population, constructing a population denominator, and linking the figure to an allocation decision and an outcome indicator.
10.2. Specifying the Limitation Concerning Studies of Quantification and Preventing a Mistaken Analogy
The literature places transformations in Egyptian quantification in the first half of the twentieth century. This paper does not, however, attribute dating to a particular decade to works whose editions were not examined40. Nor can the transfer of this language to charitable benefit be inferred by analogy with labour and population. The fields differ in their bodies, purposes and legal effects. Such an analogy confuses the design of the apparatus with its operation.
10.3. Two Established Findings and a Third Unknown
Two findings can be drawn from this paper’s material with direct evidentiary support, while a third remains unknown. First, the corpus continues to 1400 AH without the word “beneficiary” (mustafīd) appearing. This is a claim about the release as a whole, not a particular period: the denominator after 1959 is twenty-two records, which have been excluded from the inferential denominator as established in section 12. Second, the final item in the corpus is a fatwā on the authority for abolishing family waqf, issued by Shaykh Jād al-Ḥaqq ʿAlī Jād al-Ḥaqq in (Shaʿbān 1400 AH / 29 June 1980), in response to an enquiry from the University of Malaya in Malaysia41. This fatwā marks the endpoint of the fatwā source whose history the paper traces. What remains unknown is whether beneficiary numbers themselves were used to determine programme priorities or transfer funds from one category to another. This field is therefore recorded as “unknown”, not “absent”.
Footnotes
-
See note 7. ↩
-
Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], “Ilghāʾ al-waqf al-ahlī wa-sanaduhu” [Abolition of Family Waqf and Its Authority], fatwā of Jād al-Ḥaqq ʿAlī Jād al-Ḥaqq, Shaʿbān 1400 AH / 29 June 1980, location 3266 (F432-06529). The enquiry came from the University of Malaya in Malaysia in a letter dated 9 Jumādā al-Ākhira 1400 AH / 24 April 1980, registered at Dār al-Iftāʾ under no. 175 of 1980. ↩
11 · From Input to Impact: What Was Being Measured?
11.1. Distributing the Waqf Material Across the Contemporary Measurement Sequence
Contemporary measurement terminology requires a distinction between: (inputs, activities, outputs, reach, outcomes and impact).
Along this sequence, the entitlement material falls into six layers: evidence on inputs is rich, activities are available descriptively, and outputs are relatively well documented. Evidence on reach is scattered and confined to individual awqāf; evidence on outcomes is scattered and unsystematic; and impact is undocumented in the internal material. Appendix (D) contains a table of these layers with examples from the paper’s material, together with the boundary between counting persons, measuring outcomes and establishing impact, and the position of the “reach” row following the numerical survey.
This does not mean that historical actors neglected outcomes: they described the school’s benefit, the continuity of feeding and the students’ well-being. They expressed these through narrative and testimony, however, rather than a standardised indicator. The distinction between an absence of concern and an absence of an indicator is decisive here. The paper traces the history of the measurement instrument, not the full range of purposes held by the actors.
11.2. Structural Convergence: The Fatwā Preventing Duplicate Receipt and the “Unique Person” Indicator
The material contains one point where the “language of jurisprudence” approaches the contemporary “language of reach”: Shaykh Bakrī al-Ṣadafī’s ruling in Jumādā al-Thāniya 1331 AH: “What more closely accords with the founder’s purpose is to ensure that benefit is widely shared; it is therefore more appropriate not to give to the same person in two capacities”42. The rationale is to spread benefit, and the means is to prevent double counting. In its logical structure, this is precisely what the unique person indicator does in contemporary measurement. The difference lies in the aim: jurisprudence prevents duplication to spread benefit more widely, whereas the indicator prevents it to make the figure accurate (this comparison establishes only a structural resemblance; methodologically, a classical scholar cannot be credited with a contemporary bureaucratic theory he did not articulate).
Footnotes
-
See note 12. ↩
12 · Competing Explanations: What Was Tested and What Remains Viable
Five competing explanations for this paper’s claim are presented here, each with a different status:
12.1. The First Explanation (the Strongest Alternative): The Effect of Documentary Genre
The argument is that the narrow scope of observed counting in the corpus merely reflects the fatwā’s character as an answer to an individual waqf case and is not evidence of the system’s structure.
This explanation has been narrowed by three pieces of evidence from within the corpus and two from outside it, but has not been ruled out. First, the corpus contains detailed account statements counting items by unit; if documentary genre precluded numerical detail, it could not contain these. Second, documentary genre did not prevent counting to verify actual presence from appearing in seventeen records, yet none extended beyond a single waqf. Third, when disputes concern numbers, they concern the interpretation of the stipulation rather than the accuracy of observation.
The two external pieces of evidence are not fatwās. The first is a waqf statistical account submitted by the Director-General of Awqāf and published on 8 April 1899. It counts the buildings renovated by the Dīwān in three categories, followed by “comprehensive general statistics on waqf revenues and expenditures”, divided between charitable and family awqāf year by year. Its units are the building and the pound; it contains no count of even a single person43. The second is a Tunisian article from 1964 whose vocabulary includes “entitlement-holder”, “nāẓirs”, “regular provision” and “entitlement-holders of aḥbās” (the Maghribi term for awqāf), but not “beneficiary” (mustafīd)44. These two sources imply that neither the narrow scope of observed counting nor the absence of the word is unique to the fatwā genre. Their limits are set out in section 13. Even after this narrowing, a decisive determination requires an operational register for a particular riwāq or waqf. This explanation therefore remains a viable alternative that the paper’s design does not rule out. Its specified future test is a disbursement register for an al-Azhar riwāq in a particular year, examined for a count of those present independent of the list of those registered.
12.2. The Second Explanation: Difference of Field Rather Than Time
The argument is that units of counting differ because the fields of waqf service—education, feeding and entitlement—are inherently different, not because time changed the system. The material narrows this explanation: the six fatwās on enumeration concern a single field, waqf-supported religious education. Their counting rules differ according to the group’s relationship with a body that keeps track of it, rather than according to field. This narrowing is strongest in the fatwā of 1331 AH, where the two categories diverge within a single waqf: the reciters of the Qurʾān reading circle can be exhaustively enumerated, while the students cannot. The explanation nevertheless retains a limited role: the difference between the unit of regular provision and the unit of enrolment concerns the nature of the benefit, not time alone.
12.3. The Third Explanation: Actual Expansion of Services
The argument is that larger figures reflect actual service expansion rather than merely bureaucratic reclassification. This remains a viable explanation in the case of al-Aḥmadī Mosque in Ṭanṭā, and the paper’s design does not rule it out: the decline from three thousand eight hundred and seventy-four to four hundred may represent an actual contraction following the emergence of modern schools. The paper establishes that the source does not distinguish between the two possibilities, not that the second is less likely.
12.4. The Fourth Explanation: Later Editing and the Interval Before Printing
The argument is that some statistical series were recompiled in later publications, so the date of the table differs from that of the occurrence. The material contains evidence supporting this: the source quoting the digest from which the 1875 figure is taken describes it as dating from 1847 and as having been printed in 194245. Thus, about a century separates the date of observation from the date of printing, and ʿAlī Mubārak gives it a different title46. Whether these are one work under two titles or two works remains unresolved. Neither can support a calculation before the original is consulted. A more immediate editorial layer belongs under this explanation: the collection editor’s layer within the corpus itself. This paper has separated it out and republished the count without it, as described in sections 3 and 4.
12.5. The Fifth Explanation: Bias in the Corpus Rather Than Structural Transformation
The argument is that what this programme’s series measure across periods reflects changes in what the collection published, not changes in what occurred within the system. This explanation was tested collectively along four axes: main-text length, the composition of issue types, the enquirer’s identity, and publication density and period representation. The results appear in the Common Appendix, and this paper relates to them in three respects.
-
First, the central result is protected from this explanation: the word “beneficiary” (mustafīd) is entirely absent across all nine decades. Its density is zero per ten thousand cleaned characters in each. Absolute absence cannot be generated by length or size bias, because there is no quantity to contract or expand as the denominator changes. This indicator therefore remains independent of the four axes. Its only limit is the genre of the source.
-
Second, no claim of increasing lexical prevalence is possible, and none is made. The density of “mustaḥiqq” fluctuates between 4.22 and 8.76 per ten thousand cleaned characters across the nine decades, without a trend. An increase in the number of records containing it in a period therefore cannot be read as the spread of the word. The denominator changes, while density remains constant within the range of fluctuation. This paper makes no claim of increasing lexical prevalence. It establishes only the dominance of the two terms and the absence of the third.
-
Third, and this is where the claim is narrowed, the competing explanation remains viable for Table (3): the increased share of the category of counting to verify actual presence after 1946 is a proportion calculated against the period’s denominator. This is precisely the form of result targeted by the explanation. Here, the competing explanation has become measured rather than hypothetical. The share of cases concerning a “designated or descendant” entitlement-holder rises from thirty-five point seven per cent (35.7%) in the 1930s to seventy-five point three (75.3%) in the 1950s. The share of issues concerning devolution and transfer rises from forty-one (41%) to seventy-nine (79%). Counting to verify actual presence concerns which waqf recipients are present, making it more closely connected to waqf for designated descendants than to other forms. The explanation based on the composition of cases thus has numerical support; an explanation based on the 1952 law cannot be preferred without an evidentiary bridge. The association is recorded, while its explanation awaits the specified test.
This test imposes a governing restriction on the final period. The ratios of the final two decades’ shares in the corpus to their shares in all published fatwās are thirty-seven out of one hundred and then nineteen out of one hundred. Their samples comprise seventeen texts and then four. Material after 1959 therefore describes a sample rather than providing observation. None of it supports inference, whether the zero for “beneficiary” (mustafīd) or the zero for counting to verify actual presence. This paper’s central result rests on the corpus as a whole, not on a particular period, so this restriction does not affect it. Chronological inference ends here at 1959, while the sixth period remains displayed as a description of its sample.
Footnotes
-
Majallat al-Manār [The Beacon], issue 5, 8 April 1899, pp. 75–76, “News and Historical Events” section, “Iḥṣāʾ al-awqāf” [Statistics on Awqāf], followed by “Iḥṣāʾ al-awqāf al-Miṣriyya” [Statistics on Egyptian Awqāf]. The item has no attributed author, and the article-author field in the copy examined is empty; it is therefore attributed neither to the journal’s proprietor nor to anyone else. Its occasion was the celebration by the “Dīwān of General Awqāf” of the opening of its new premises in Bāb al-Lūq, “belonging to the waqf of the late ʿAbbās Pasha I”, and the submission by “His Excellency Fayḍī Pasha, Director-General of Awqāf, of a summary containing statistics on the buildings renewed and renovated by the Dīwān”. The starting date of the period is corrupted in the copy examined to “1793 of the Christian era”. It is preserved as it appears rather than corrected by conjecture; the only qualification is that the text identifies it as “the period when he assumed the administration of the Dīwān following the accession of His Exalted Highness the Khedive”. The three counting categories and their arithmetic audit appear in Appendix D. The revenue section contains three columns—total, charitable awqāf and family awqāf—for the years 1892 to 1896 and thereafter, with the first row reading “190735 153775 36960 year 1892”. The passage examined contains no count of persons, whether entitlement-holders, employees or beneficiaries. The text was extracted from the Alsharekh magazine collection; its location in the original is: archive.alsharekh.org/Articles/33/3352/93323. ↩
-
Majallat al-Fikr [Thought] (Tunis), issue 6, 1 March 1964, pp. 73–80, Muṣṭafā Āghā, “al-Nāẓir wa-l-mustaḥiqq” [The Nāẓir and the Entitlement-Holder]. Attribution is to the article’s author, not the journal’s editor-in-chief. It takes the form of an argument in rhymed prose voiced by two parties. Its genre is therefore literary satire, neither administrative nor juristic, and this limit on its evidentiary use is recorded. Its expressions include: “I see others among the entitlement-holders of aḥbās”; “it is a word that delights the nāẓirs and terrifies the weak entitlement-holders”; “this is no difficult matter for the nāẓirs”; and “for you provide me with enough and continue the regular provision”. It contains a satirical enumeration of account-statement entries corresponding to the discussion in section 7: “By God, how much renovation, how many fees and valuations, how much timber and how many nails, how much plaster and how many stones, how many keys and locks, ladders and baskets… the cost of books and the expenses of journeys… a tax on inzāl leases, workers’ wages and a cleaning charge.” The statement’s units are things and amounts, not persons, as established in section 8. The word “beneficiary” (mustafīd) does not occur in the article. The text was extracted from the Alsharekh magazine collection; its location in the original is: archive.alsharekh.org/Articles/97/9839/197852. ↩
-
See note 2. ↩
-
See note 4. ↩
13 · Declared Limitations
The governing restrictions and source limitations of this third paper are recorded here through direct and comprehensive disclosure, to protect the scholarly results and prevent unwarranted upgrading of inferences:
Type of contribution: a test carried out on one component of the thesis—the vocabulary of the receiving side and the structure of counting within the entitlement corpus. Beyond that lie a concept and a hypothesis.
Falsifier: its formulation is the presence of an observed count of recipients in the corpus. It was applied to the entire corpus through two procedures and was realised in a restricted form: seventeen records whose unit is the designated name and whose scope is the individual waqf. The claim was accordingly narrowed to the absence of an aggregate figure rather than the absence of counting. Its limit is that it falsifies a claim about counting only within this genre of sources.
Layer: the paper concerns the design of counting instruments, not the outcomes of benefit. Rates of reach and actual receipt await operational registers that have not been consulted.
First source: a corpus of a single genre, the waqf fatwā, in the frozen release WEC-432/1.0, comprising three hundred and ninety-six texts. It does not represent the corpus comprising the press, associations and social reports. Full manual coding of the fields in Table M1 remains pending.
Limit of the frozen release: freezing preserves a state; it does not constitute final approval. The script that tightened the selection and separated four hundred and five from one hundred and thirty-five has not been preserved. The boundary between candidates and accepted items is therefore described, not reproduced. The judgements for the sixty-item precision-test sample have not been preserved, so its rate cannot be recalculated. The release contains one flag attached to a record in which the muftī and date conflict. All the figures are in Appendix B.
Miss rate: its initial measurement of “zero missed items” cannot be reviewed because the sample identifiers were not preserved, and reading a new sample at the time of freezing refuted it. The known minimum is two missed cases out of one hundred and thirty-five, with four likely. This is a lower bound, not an estimate, and none was added to the release.
Fewer cases than texts: the denominator of the results published here comprises texts, not distinct cases. The release contains two documented pairs, each published under two titles. What is known with certainty is therefore three hundred and ninety-four distinct cases in three hundred and ninety-six texts. Any estimate of duplication beyond this lacks documentation by identifiers.
Text length: the length field recorded in the file measures the text together with additions from the publishing website. These comprise approximately four thousand eight hundred and ninety-three characters and are nearly constant. Every length figure in this paper is therefore calculated solely on the cleaned text, represented by the main-text word-count column in Table D2 of Appendix D.
Dating key: constructed from the heading in this version, not from the automated field. The automated field is known to be defective, with its figures for the frozen release recorded in Appendix B. The period distribution was reconstructed using a second, exclusively Hijri key on the third layer. The central result held, while the size of the final period changed.
Lexical matching: automated, with declared word boundaries and separated layers. It measures the presence of the form, not every expression of the meaning. Forms outside the survey may still occur.
Numerical survey: restricted to proximity to names for the receiving side drawn from a closed list. A figure counting recipients under a description outside that list escapes it.
Final period: its denominator is twenty-two records, and it was excluded entirely from the inferential denominator following the collective test of the competing explanation. It is read as a description of a sample, and chronological inference ends at 1959.
Statistical digest: the original was not consulted. The Aḥmadī Mosque figure and the figure for the beginning of the twentieth century are both at the grade of material drawn from a direct source. They do not rise to the grade of directly examined evidence. Its title differs between the two sources, and its observation date differs from its printing date.
Al-Azhar table: the difference between the total number of shaykhs and its breakdown is thirty-six, and no explanation has been preferred. The total was not used in any calculation in this paper.
Ministry of Awqāf registers: not all rows of their table were read in the copy examined, preventing reconciliation of the table’s total with the total in the prose.
Juristic references: al-Bināya and al-Muḥīṭ al-Burhānī were printed in volumes. The indexed version used here numbers pages without identifying the volume. No volume number was established or invented. This gap in the references is recorded pending comparison with the printed edition.
Location of the threshold of enumerability: the threshold and its three positions occur in the chapter on bequests, not the chapter on waqf, in al-Bināya and al-Nihāya. Its transfer to waqf rests on a bridge explicitly stated in the sources, not an analogy made by the paper. The evidence is the muftīs’ application of it to waqf cases. Two references were corrected in this version without affecting any conclusion. Full details appear in Appendix D, items D18 and D19.
Two external pieces of evidence: the statistics of 1899 and the article of 1964 are two passages, not a comparable corpus. ⚠️ The 1964 article is literary satire, neither administrative nor juristic, so the absence of “beneficiary” (mustafīd) from it cannot be taken as evidence of its absence from administrative writing. They narrow the first competing explanation without ruling it out.
First competing explanation: the effect of documentary genre remains a viable alternative that this design does not rule out.
Fifth competing explanation: bias in the corpus is ruled out for the central result, remains viable for the distribution across periods of counting to verify actual presence, and governs the treatment of material after 1959.
Evidentiary tiers of the claims, aligned with the shared inference protocol. The first tier, comprising what the fatwās establish through their express wording, contains five claims: the absence of “beneficiary” (mustafīd) from the frozen release; the occurrence of “one who derives benefit” (muntafiʿ) in two records without denoting a category; the division of counting terms into six uses; the seventeen instances of counting to verify actual presence, with their texts and identifiers; and the muftīs’ application of the enumeration rule to particular cases. The second tier, comprising what the corpus makes likely after controlling the denominator and applying the falsifier, contains one claim: the share of cases involving observed counting across periods. In this version, it has been reduced to a described association for which no explanation is preferred. The third tier, comprising what remains a hypothesis requiring the administrative archive, contains one claim: the transition of the unit of counting into a programme category within the planning apparatus. This paper does not describe that claim as a finding.
Overlap with the first paper of this series: the source is shared, while the unit and instrument differ, so agreement between the two papers cannot be read as independent corroboration.
What was not investigated where it would be expected to be found: the original statistical digest; operational registers of the riwāqs; Ministry of Social Affairs reports after 1952; minutes of family waqf liquidation committees; and registration forms in charitable institutions. What was not found here was not sought in its proper location. It does not count as a negative result.
14 · Conclusion
The statistical figure was not the endpoint of the waqf beneficiary’s story, but one way of defining that beneficiary. In the entitlement register, the beneficiary appeared as a name, share and rank; in the founder’s deed, as a stipulated person-slot; in the register of regular provisions, as a portion disbursed; in the Public Instruction table, as a member of an institution; and in the rows examined from the ministry’s register, not at all.
This paper establishes three things through testing rather than inference of likelihood. First, “beneficiary” (mustafīd) does not occur in the Waqf Entitlement Corpus, either in the main texts of the fatwās or in the collection editor’s layer. Its frozen release comprises three hundred and ninety-six texts between 1313 and 1400 AH. Nor does “one who derives benefit” (muntafiʿ) occur as the name of a category. Second, observed counting is confined to one category: verifying the presence of waqf recipients within an individual waqf. No figure was found aggregating recipients within a denominator extending beyond a single waqf. Third, Ḥanafī jurisprudence contained an operative concept of enumeration whose threshold was subject to explicitly defined disagreement, and Egyptian muftīs applied it to particular cases, making the numerical threshold a dividing point in the structure of the right.
These three findings imply that the proposition of a transition from the uncounted to the counted mistakes the starting point. The waqf fatwā records an existing practice of counting, but one that distributes before it observes. When it observes, it observes names within an individual waqf, not heads within a population. It makes the possibility of counting a condition of the right, not a description of the population. The proper question is therefore not when the beneficiary became counted, but when the meaning of counting that beneficiary changed. All this must be read within the limits of its genre. The competing explanation based on the effect of documentary genre remains viable, as established in section 12. The conclusion cannot be transferred from fatwā to system without an evidentiary bridge.
The replacement test’s result must be stated within its limits. The purpose foundation under the German Civil Code was put in the place of waqf, and the counting apparatus did not sustain a distinction: waqf is not distinguished in this respect from its comparator within the family of purpose institutions. Replacement does fail, however, at the threshold of enumerability. Ḥanafī jurisprudence makes the validity of waqf and its rule of distribution depend on an attribute of the group that changes with circumstances. The comparator makes the beneficiary’s claim depend solely on the wording of the charter, not on the number of beneficiaries. The test and its limitation are detailed in section 6.A. This component must itself be narrowed: a criterion external to the instrument that changes with circumstances is not a distinctive feature of Islamic jurisprudence, since it has a counterpart in a system from a third family. What remains is that the criterion must be a number, and its effect must extend to the validity of creation.
The transition to “measuring outcomes and impact”, and to the “beneficiary” (mustafīd) as a stable programme category, is not a conclusion established by this paper. That requires consulting the original digest and taking an inventory of reports after 1952. It also requires tracing administrative vocabulary in a corpus other than fatwās and following a decision in which recipient numbers were actually used. Only then can the research move from the history of counting to the history of what numbers did in distributing benefit and judging it.
The argument passes to the paper on contestation: if the beneficiary has been redefined as a unit in a table, through what voice can that beneficiary contest the table itself?
Sources and References
First: The Main Source
— Waqf Entitlement Corpus (1313–1400 AH): Fatāwā Dār al-Iftāʾ al-Miṣriyya [Fatwās of the Egyptian Dār al-Iftāʾ], compiled by al-Maktaba al-Shāmila from the collection of Dār al-Iftāʾ al-Miṣriyya (the Egyptian fatwā authority), book no. 432 in al-Shāmila’s numbering, in ten volumes. The release used here is the frozen release WEC-432/1.0, containing 396 texts with stable identifiers in the form F432-NNNNN. This paper does not refer to the two intermediate layers waqf_corpus.jsonl and waqf_corpus_v2.jsonl. Every citation gives the fatwā’s title, muftī, date and location in the numbering of the collected version. The bibliographic details of the collection’s reference print edition were not examined; establishing them is an item under the conditions for upgrading claims. The full protocol appears in the first paper of the programme and in Appendix B below.
— Source-selection funnel for the corpus of entitlement fatwās and release statement for the frozen release WEC-432/1.0 (28 July 2026): a record of the five layers, inclusion rules, hashes and exclusion table. Every claim in this paper concerning the size and scope of the corpus, or the layer on which a figure was calculated, refers to this record.
Second: Primary and Descriptive Sources Examined
— ʿAlī Pasha Mubārak, al-Khiṭaṭ al-Tawfīqiyya al-jadīda li-Miṣr wa-l-Qāhira wa-mudunihā wa-bilādihā al-qadīma wa-l-shahīra [The New Tawfīqī Topography of Egypt, Cairo, and Their Ancient and Famous Cities and Towns], second revised edition, Cairo: Dār al-Kutub wa-l-Wathāʾiq al-Qawmiyya, 1425 AH / 2004, p. 36 (description of al-Azhar and its statistics for 1875).
— Ibrāhīm ibn Mūsā al-Ṭarābulusī al-Ḥanafī, al-Isʿāf fī aḥkām al-awqāf [Assistance in the Rulings Governing Awqāf], second edition, Egypt: Maṭbaʿat Hindiyya, al-Mahdī Street, al-Azbakiyya, 1320 AH / 1902, p. 139.
— Badr al-Dīn al-ʿAynī, al-Bināya sharḥ al-Hidāya [The Edifice: A Commentary on the Guidance], edited by Ayman Ṣāliḥ Shaʿbān, Beirut: Dār al-Kutub al-ʿIlmiyya, first edition, 1420 AH / 2000, p. 473 (according to the numbering of the indexed version examined, which did not identify the volume number).
— Ḥusām al-Dīn Ḥusayn ibn ʿAlī al-Sighnāqī, al-Nihāya fī sharḥ al-Hidāya (Sharḥ Bidāyat al-mubtadiʾ) [The Culmination in Explaining The Guidance (A Commentary on The Beginner’s Beginning)], p. 170 (according to the numbering of the indexed version examined, which did not identify the volume number or edition details).
— Burhān al-Dīn Ibn Māza al-Bukhārī, al-Muḥīṭ al-Burhānī fī al-fiqh al-Nuʿmānī [The Burhānī Compendium of Nuʿmānī Jurisprudence], edited by ʿAbd al-Karīm Sāmī al-Jundī, pp. 115–116 (according to the numbering of the indexed version examined, which did not identify the volume number or publication details). The passage cited explicitly quotes al-Sarakhsī and includes Ibn Māza’s critical comment on it.
— Abū Bakr Muḥammad ibn Aḥmad al-Sarakhsī, al-Mabsūṭ [The Expanded Treatise], edited by a group of distinguished scholars, Egypt: Maṭbaʿat al-Saʿāda, p. 34 (according to the numbering of the indexed version examined, which did not identify the volume number).
— Muḥammad Abū Zahra, Muḥāḍarāt fī al-waqf [Lectures on Waqf], Cairo: Dār al-Fikr al-ʿArabī, pp. 94–96. This paper draws on Ibn Nujaym’s passage from “al-Baḥr”, quoted verbatim within it, together with Abū Zahra’s qualification.
— Zayn al-Dīn ibn Ibrāhīm ibn Muḥammad, known as Ibn Nujaym, al-Baḥr al-rāʾiq sharḥ Kanz al-daqāʾiq [The Clear Sea: A Commentary on the Treasure of Subtleties], second edition, eight volumes, p. 215. 🔴 The original was consulted for this version and its wording reproduced, so it is a directly examined source rather than a quotation through an intermediary, as established in footnotes 17 and 20 and items D18 and D19. (An earlier version quoted it through Abū Zahra’s lectures, p. 96, under the title “al-Baḥr” without further specification; comparing the original with the intermediary revealed that a sentence had been omitted from the intermediary’s quotation without ellipsis marks.)
Third: References
— Ibrāhīm al-Bayyūmī Ghānim, al-Awqāf wa-l-siyāsa fī Miṣr [Awqāf and Politics in Egypt]. (A digital copy examined in the research library; the two sections used here are “Analysis of Waqf Deed Data”, with visible page number 230, and “E — Nāẓir Reports”, with visible page number 41. The edition to which these numbers belong has not been established; verifying it is an item under the conditions for upgrading claims.)
— al-Fadhlaka al-iḥṣāʾiyya al-ʿumūmiyya li-l-madāris wa-l-katātīb bi-l-quṭr al-Miṣrī min sanat 1847 [General Statistical Digest of Schools and Qurʾān Schools in Egypt from 1847], printed at Maṭbaʿat al-Madāris al-Malakiyya, 1942. (Quoted through Ghānim; the original was not examined, and the title contains an evident transcription error in the digital copy.)
— Kitāb al-Natīja al-iḥṣāʾiyya li-l-madāris wa-l-makātib bi-l-quṭr al-Miṣrī [Book of Statistical Results for Schools and Qurʾān Schools in Egypt]. (As named by ʿAlī Mubārak; not examined, and it cannot be stated with certainty that it is the same work as the preceding entry.)
— al-Fatāwā al-ʿĀlamgīriyya al-maʿrūfa bi-l-Fatāwā al-Hindiyya [The ʿĀlamgīrī Fatwās, Known as The Indian Fatwās]. (Not examined in this paper; its text appears only as a quotation within the answer to the fatwā of Jumādā al-Thāniya 1331 AH.)
— Majallat al-Manār [The Beacon], issue 5, 8 April 1899, pp. 75–76, “News and Historical Events” section, “Iḥṣāʾ al-awqāf” [Statistics on Awqāf]. (The item is unattributed in the copy examined; extracted from the Alsharekh magazine collection.)
— Muṣṭafā Āghā, “al-Nāẓir wa-l-mustaḥiqq” [The Nāẓir and the Entitlement-Holder], Majallat al-Fikr [Thought] (Tunis), issue 6, 1 March 1964, pp. 73–80. (Extracted from the Alsharekh magazine collection; its genre is literary satire, which limits its evidentiary use.)
— Alain Desrosières, The Politics of Large Numbers: A History of Statistical Reasoning, trans. Camille Naish (Cambridge, MA: Harvard University Press, 1998).
— Wendy Nelson Espeland and Mitchell L. Stevens, “Commensuration as a Social Process,” Annual Review of Sociology 24 (1998): 313–343.
— Wendy Nelson Espeland and Mitchell L. Stevens, “A Sociology of Quantification,” European Journal of Sociology 49, no. 3 (2008): 401–436.
— Malak Labib, Recenser l'Égypte: Dette publique et politiques de quantification à l'ère impériale (1875–1922) (Cairo: Institut français d'archéologie orientale, RAPH 47, 2024).
— Malak Labib, “An Unlimited Supply of Labor? Counting Workers and Planning Development in Mid-Twentieth-Century Egypt (1939–1960),” International Labor and Working-Class History 108 (2025): 1–17.
— Theodore M. Porter, Trust in Numbers: The Pursuit of Objectivity in Science and Public Life (Princeton: Princeton University Press, 1995).
— James C. Scott, Seeing Like a State (New Haven: Yale University Press, 1998).
The seven foreign-language works above are cited as a whole for their general arguments. Their editions were not examined as part of this paper’s material, so no specific page references are given.
Fourth: Papers from the Two Programmes (Unpublished)
— Bandar, “When the Reference Point of the Answer Changed” (2026). Draft.
— Bandar, “The Entry Standard: Measured Content, a Hypothesised Source” (2026). Draft.
— Bandar, “Did Charity Become a Right?”, the fourth paper of the “Architecture of Entitlement to Charitable Benefit” programme (2026). Draft.
— Bandar, “When the Waqf Became Legible”, in the “Architecture of Charitable Benefit Production” programme (2026). Draft.
— Bandar, “Common Appendix to the Architecture of Entitlement to Charitable Benefit Programme: Inference Protocol, Variable Matrix and Declaration of Dependence” (2026).
— Bandar, “The Collective Test of the Competing Explanation: A Biased Corpus or a Changing Structure?” (28 July 2026).
Appendix A: Coding Instrument
| Field | Coding question |
|---|---|
| Object of counting | Person, portion, payment, institution, waqf, document |
| Unit of counting | Registered name, stipulated person-slot, head, loaf, share, pupil, seat |
| Reference population | Who can enter the denominator of the count? |
| Inclusion rule | Registration, presence, receipt, legal entitlement, affiliation, status |
| Temporal dimension | A snapshot on a day, a stock at a date, a flow over a year |
| Repetition | Is the person counted once, or with every portion or visit? |
| Counting authority | Nāẓir, shaykh of a riwāq, court, muftī, ministry, statistical agency |
| Purpose of counting | Distribution, accounting, supervision, funding, planning |
| Scope of aggregation | Waqf, riwāq, school, province, sector |
| Identity retention | Name and full status, register identifier, or anonymised total |
| Connection to a decision | Description alone, or use in allocating funds or changing an organisational arrangement |
| Indicator type | Input, activity, output, reach, outcome, impact |
Table M1: The twelve fields of an “instance of counting the receiving side”, with their coding questions.
Appendix B: Counting Protocol
1 · Layers of the corpus selection funnel, the inclusion rule and its error rate: the funnel has five layers: crawling 4,930 pages; segmenting 10,035 fatwās in book 432; lexical screening yielding 540 waqf-related candidates; tightening the criteria after a precision test, retaining 405; and complete manual coding, excluding nine for recorded reasons, leaving the frozen release WEC-432/1.0 with 396 texts.
The automated rule in the second layer was the occurrence of “waqf”, “wāqif” (founder) or “mawqūf” (dedicated as waqf) at the opening of the text, together with an entitlement term from a closed lexicon of sixteen words (waqf, al-waqf, istiḥqāq [entitlement], mustaḥiqq [entitlement-holder], al-mustaḥiqq, al-mawqūf, al-wāqif, nāẓir [trustee-administrator], al-nuẓẓār [nāẓirs], ṭabaqa [tier], naṣīb [share], dhurrī [family-related], ḥikr [long-term ground lease], al-ghalla [yield], rayʿ [revenue], jirāya [regular provision]), or a waqf-related title. This was a purely lexical rule, not a semantic one, so it admitted material unrelated to waqf, such as “a marriage pending approval”, “the unborn child’s share is held in abeyance” and “Islam’s position on such-and-such”.
The boundary of the third layer is described but has not been reproduced: the script that implemented the tighter criteria has not been preserved, and the stated rule does not automatically reproduce the split into 405 and one hundred and thirty-five records. Reapplying it retained 378 of the 405 and reinstated 25 of the 135. The reasons for excluding the one hundred and thirty-five have not been recorded individually. The judgements on the sixty-record precision-test sample have not been preserved; what survives is its identifiers and the fact that 44 remained in the third layer and 16 were excluded.
Both directions of error are reported together, and a lower bound has now been measured for the second. Over-inclusion can be detected through review: 135 such records were detected in the third layer and nine in the fourth. For under-inclusion, however, the initial finding of “zero misses” cannot be reviewed, because the identifiers of its fifteen-record sample were not preserved. Reading a sample of 22 records when the release was frozen overturned that finding, identifying two confirmed misses, F432-05731 on the substitution of waqf assets to widen Al-Azhar Street and F432-05987 on the sale of entitlement, and two likely misses, F432-05939 and F432-05537. The known lower bound for misses is therefore 2 out of 135, and the likely figure is 4 out of 135. This is a lower bound, not an estimate, and none of these records was added to the release. It follows that the denominator underlying the negative finding is a lower bound rather than a complete count, and that the claim to cover “the complete corpus” must be read within the limits of what this rule admitted.
There are fewer instances than texts. The release contains two pairs, documented by their identifiers, in each of which the same instance was published under two titles: F432-06347 with F432-05671, and F432-06495 with F432-05673. What is known with certainty is therefore 394 distinct instances in 396 texts; any estimate of further duplication is not documented by identifiers. The release also contains one flag, attached to F432-05973, for an inconsistency between its muftī and its date.
2 · Text cleaning: the publishing website’s technical header is removed from each record by cutting everything after the first occurrence of “Load the next page”, which appears in all 396 records of the frozen release. Diacritics and the elongation character are then stripped from the text, and hamza-bearing alif and alif maqṣūra are normalised to a single form.
3 · Separating the layers: each record is divided at the first occurrence of “Question” into two layers: the editorial layer, comprising the title, the muftī’s name, the date and the “Principles” block; and the fatwā text, comprising the question and answer. This division is possible in 395 of the 396 records; as a precaution, the remaining record is assigned entirely to the editorial layer.
4 · Matching pattern: the stem is matched with one optional prefix, “w” or “f”, a second optional prefix, “al”, “b”, “l” or “k”, and no more than four trailing letters, with a word boundary preventing attachment to anything before the prefix or after the suffix.
The matched stems are: “mustafīd” (beneficiary); “muntafiʿ” (one who derives benefit); “mustaḥiqq”; “mawqūf ʿalay” (stem of “waqf recipient”); “faqīr” and “fuqarā” (poor person and poor people); “mujāwir” (resident student); “nafar” and “anfār” (person and persons); “ʿadad” (number); “iḥṣā” (enumeration); “yuḥṣawn”, “yuḥṣayn”, “yuḥṣī” and “iḥṣāʾihim” (enumeration-related forms); and “maḥṣūr” (bounded or enumerable).
5 · Counts in the two layers of the frozen release: figures are given as records/occurrences. The first column covers the complete corpus without separating the layers, the second the fatwā text, and the third the editorial layer:
| Term | Complete text | Fatwā text | Editorial layer |
|---|---|---|---|
| mustafīd | 0 / 0 | 0 / 0 | 0 / 0 |
| muntafiʿ | 2 / 2 | 2 / 2 | 0 / 0 |
| mustaḥiqq | 174 / 692 | 166 / 563 | 79 / 131 |
| mawqūf ʿalayhi | 144 / 437 | 130 / 367 | 51 / 72 |
| faqīr | 84 / 307 | 81 / 254 | 33 / 54 |
| mujāwir | 20 / 51 | 19 / 48 | 3 / 3 |
| nafar and anfār | 7 / 13 | 7 / 13 | 0 / 0 |
| ʿadad | 27 / 49 | 27 / 46 | 3 / 3 |
| iḥṣāʾ | 2 / 2 | 1 / 1 | 1 / 1 |
| yuḥṣawn and related forms | 6 / 30 | 6 / 24 | 3 / 6 |
| maḥṣūr | 5 / 6 | 5 / 6 | 0 / 0 |
The effect of freezing the release on these counts is confined to two of the nine excluded records: F432-00323, which contains one occurrence of “mustaḥiqq” and seven occurrences of “faqīr” in the fatwā text and four in the editorial layer; and F432-07844, which contains one occurrence of “faqīr” and another of “ʿadad” in the fatwā text. None of the other seven contains any term in the table. The zero count for “mustafīd” holds across both layers.
6 · Identifiers of records containing enumeration expressions, and the treatment of each: the pattern “yuḥṣawn” and related forms matched six records. F432-05961, F432-05763, F432-05591, F432-05487 and F432-05567 were all included among the instances of application; F432-04937 was excluded because “yuḥṣī” there refers to recording receipts and expenditure in the waqf ledger, not enumerating waqf recipients. The pattern “maḥṣūr” matched five records. F432-06009 was included among the instances of application; F432-05733, F432-05851, F432-06113 and F432-06133 were excluded because the word denotes restriction or limitation rather than the possibility of counting. Their wording is “a bounded plot of land”, “the founder’s relatives, restricted to his brother’s children”, “the disagreement … is confined to a few questions” and “the excuses are limited to the four matters mentioned”. There are therefore six included instances and five exclusions, with reasons recorded.
7 · Dating key: the key used in this version is constructed from the record heading. It uses the Gregorian year explicitly stated there where available, as in 271 records of the frozen release; otherwise, the year is converted from the Hijri month and year in the heading using the tabular calendar, as in 120 records. Five records have no key because their headings contain no date. For comparison, the status of the automated field in the file is recorded for the same release: accurate in 273 records; calculated using a fixed offset in 117 records (581 years in 105, 580 in ten and 579 in two); and disregarding an explicit Gregorian date in the heading in 15 records.
The counts of included instances under the corrected key are: 101 between 1895 and 1913; 124 between 1914 and 1933; 52 between 1934 and 1945; 42 between 1946 and 1953; 50 between 1954 and 1959; and 22 between 1960 and 1980; a total of 391. The effect of freezing the release is confined to the sixth period, from which four records were excluded: F432-00769, F432-00323, F432-01121 and F432-01463. Their fatwā texts contain a total of 4,335 words, reducing the word-count column for that period from 17,048 to 12,713. The five records that the automated key had placed in 1997 were excluded from the release through manual coding and therefore no longer require subtraction.
The exclusively Hijri key used for verification was applied to the third layer before the release was frozen, and its figures must therefore be attributed to that layer: 107, 118, 50, 43, 48 and 15, totalling 381. It was not rerun on the frozen release. The central finding does not depend on it, since the zero count for “mustafīd” applies to the corpus as a whole and cannot be changed by distributing its records across periods.
The Gregorian boundary of the corpus must be stated accurately: the frozen release’s manifest gives its earliest Gregorian year as 1894, whereas the heading-based key adopted here places the earliest instance in 1895. The difference arises from matching the Hijri month to a Gregorian year; it is recorded rather than elided. This paper adopts the Hijri range 1313–1400 AH because that is what the headings state.
8 · The independent numerical survey and the layer on which its counts were calculated: all numerals written in Arabic-Indic or Western digits are captured in the fatwā texts, together with all explicit number words from a closed lexicon comprising units, tens, hundred and thousand. Processing is then restricted to matches within a narrow window around a noun from a closed list of nouns referring to the receiving side: eighteen characters after the number or twelve before it. A predetermined rule then excludes matches within a date, amount, area, share, article-number or law-number window, and the remainder are read in context. A targeted survey for indications of actual presence was conducted alongside it.
The results of its stages were calculated on the third layer (405) and were not rerun on the frozen release, so they must be attributed to that layer: 5,325 matches, comprising 3,213 written numerals in 294 records and 2,112 number words in 304 records; then 299 matches in 102 records; then 103 matches in 52 records, all read in context. The targeted survey identified 39 passages in 32 records, all of which were read; 17 records met the definition.
Only the reproducible component was rerun on the frozen release: capturing written numerals in the fatwā texts. It yielded 2,994 numerals in 285 records. The substantive result of the survey does not change between the two layers, since all seventeen record identifiers are included in the frozen release, and none of the nine excluded records contains a passage belonging to the sixth category.
9 · Identifiers of the seventeen records in the category of verifying actual presence, ordered by Hijri year: F432-05745 and F432-05963 (Dhū al-Qaʿda 1315 AH); F432-05545 (Jumādā al-Ākhira 1333 AH); F432-05583 (Ṣafar 1334 AH); F432-05899 (Shaʿbān 1334 AH); F432-05991 (Rajab 1337 AH); F432-05613 (Rajab 1344 AH); F432-06075 (Rabīʿ al-Thānī 1371 AH); F432-05677 (Muḥarram 1372 AH), F432-05685 (Jumādā al-Ākhira 1372 AH), F432-06501 (Dhū al-Qaʿda 1372 AH); F432-05497 (Rabīʿ al-Thānī 1373 AH); F432-05689 (Ramaḍān 1374 AH); F432-05705 and F432-06113 (Shawwāl 1375 AH), F432-05839 (Dhū al-Qaʿda 1375 AH); F432-05835 (Shawwāl 1387 AH).
10 · Classifying the falsifier: every occurrence of “ʿadad”, “nafar” and “iḥṣāʾ” was examined by reading it within a window of one hundred and forty surrounding characters and assigning it to one of five uses specified before reading: counting documents, a denominator for distribution, a stipulated quota, capacity derived from funding, and a hypothetical example used in reasoning. None fell outside these five. The independent numerical survey, however, identified a sixth category absent from the prior classification: “counting to verify actual presence”, defined in section 5. It was added after reading, not before; this is recorded rather than elided. It therefore remains an exploratory classification until tested on a second corpus with the same definition declared before reading.
Appendix C: Conditions for Upgrading Claims
— Items transferred here from the conditions for falsification: the following items were removed from their roles as direct conditions for falsification because they had not been carried out and cannot serve as declared conditions for falsification while remaining unimplemented: dating the first occurrence of “mustafīd” in Egyptian ministry reports; comparing a central report with the underlying ledger for a particular waqf in a particular year; and finding a decision in which a recipient count was used to allocate funds.
— Items concerning the source material: consulting the original statistical digest and examining whether its series was annual, how it defined pupils and institutions, and when each table was produced; resolving the relationship between its title as given by Ghānim and its title as given by ʿAlī Mubārak; comparing the al-Azhar table with its original to assess the likely reason for the difference of thirty-six; reading the Ministry of Awqāf registers table in full; consulting printed editions of al-Bināya, al-Nihāya and al-Muḥīṭ al-Burhānī to establish volume numbers; and manually coding every instance in the corpus against the fields in Table M1.
— Testing the sixth category: declaring the definition of “counting to verify actual presence” before reading a second corpus, then comparing its proportion of instances in the fatwā corpus with its proportion in the minutes of family waqf liquidation committees between 1952 and 1957, to examine the hypothesis that Law 180 of 1952 prompted observed enumeration.
— Items concerning the frozen release: fulfilling the two outstanding finalisation requirements—a stratified sample for coding reliability, reported field by field, and collation of a sample of texts against printed editions to distinguish the questioner’s words from the muftī’s answer and the editor’s title; then reading all one hundred and thirty-five excluded records to estimate misses rather than establish a lower bound; then rerunning every stage of the numerical survey on the adopted release rather than the third layer. Completing these tasks requires a release 1.1 using the same identifiers.
— Required research protocol: compiling an inventory of registration forms and performance reports in the Ministries of Awqāf and Social Affairs between 1946 and 1980; then coding every number in them against the fields in Table M1 under a rule declared before reading; then restricting the claim to the years and domains actually covered by the results.
Appendix D: Details Transferred from the Main Text
This appendix contains procedures, tables of raw evidence, case details and source clarifications transferred from the main text: material intended for consultation rather than continuous reading. The points referring readers here are marked in the main text. Footnote markers remain in their positions in the main text and are not repeated here.
D1 · Arithmetic audit of the al-Azhar statistics table for 1875. The breakdown of shaykhs by the four legal schools totals three hundred and twenty-five: one hundred and forty-seven Shāfiʿīs, ninety-nine Mālikīs, seventy-six Ḥanafīs and three Ḥanbalīs. There is therefore a difference of thirty-six shaykhs between the total and the breakdown, with no corresponding explanation. The breakdown of resident students, however, accounts for their total exactly: five thousand six hundred and fifty-one Shāfiʿīs, three thousand eight hundred and twenty-six Mālikīs, one thousand two hundred and seventy-eight Ḥanafīs and twenty-five Ḥanbalīs, totalling ten thousand seven hundred and eighty. This audit confines the discrepancy to a single row of the table. It cannot therefore be attributed to a convention of deliberately providing incomplete breakdowns: if that were the table’s convention, it would apply to its second row as it does to its first. Two possibilities remain: some shaykhs were not assigned to legal schools, or something was lost in copying or printing. The paper does not choose between them on intuition. Nor is the total used in any calculation before comparison with the original.
D2 · The lexical matching protocol and the “nafar” incident. Diacritics and the elongation character are stripped from the text, and hamza-bearing alif and yāʾ maqṣūra are normalised. Everything after the first occurrence of the publishing website’s navigation phrase is cut from each record. Stems are then matched with word boundaries that prevent the inclusion of words sharing the same letters. This final constraint is more than a procedural detail. Matching “nafar” without a word boundary also captures “infarada” (acted alone), “munfaridan” (alone) and “al-irāda al-munfarida” (unilateral will), raising the count from thirteen occurrences to more than a hundred. This happened in an initial trial before the pattern was refined, and it was corrected before any inference was based on it. The difference between the two counts is therefore not a difference in precision, but the difference between measuring a phenomenon and measuring an illusion.
D3 · Table D1: The vocabulary of the receiving side in the entitlement corpus.
| Term | Fatwā text: records/occurrences | Editorial layer: records/occurrences | Role in the corpus |
|---|---|---|---|
| mustaḥiqq | 166 / 563 | 79 / 131 | A legal position defined by status and share |
| mawqūf ʿalayhi | 130 / 367 | 51 / 72 | The relationship of a person or entity to property dedicated as waqf |
| faqīr | 81 / 254 | 33 / 54 | A description of need and an entry standard |
| mujāwir | 19 / 48 | 3 / 3 | Educational membership tied to a place |
| muntafiʿ | 2 / 2 | 0 / 0 | Use of an asset or a share in a partnership |
| mustafīd | 0 / 0 | 0 / 0 | Absent |
Table D1: The vocabulary of the receiving side in the Waqf Entitlement Corpus, frozen release WEC-432/1.0, using lexical matching across all three hundred and ninety-six texts after removing the publishing website’s header and separating the title and principles layer from the question and answer. The overall total for each term is smaller than the sum of the two columns, since the term occurs in both layers in many records. Eight records contain “mustaḥiqq” only in the editorial layer, and fourteen contain “mawqūf ʿalayhi” only there. These are counts of records and occurrences, not people.
D4 · Table D2: Presence of terms referring to the receiving side by period.
| Period | Instances | Words in the fatwā text | mustaḥiqq | mawqūf ʿalayhi | muntafiʿ | mustafīd |
|---|---|---|---|---|---|---|
| 1895–1913 | 101 | 38,969 | 33 (33%) | 19 (19%) | 1 | 0 |
| 1914–1933 | 124 | 68,420 | 53 (43%) | 40 (32%) | 0 | 0 |
| 1934–1945 | 52 | 25,299 | 23 (44%) | 12 (23%) | 0 | 0 |
| 1946–1953 | 42 | 26,282 | 18 (43%) | 19 (45%) | 0 | 0 |
| 1954–1959 | 50 | 31,685 | 27 (54%) | 29 (58%) | 0 | 0 |
| 1960–1980 | 22 | 12,713 | 8 (36%) | 10 (45%) | 1 | 0 |
Table D2: Presence of terms referring to the receiving side by period, in the fatwā texts excluding the editorial layer, counted by records rather than occurrences, with percentages relative to each period’s denominator. The dating key is constructed from the heading as explained in section 3. The periods total three hundred and ninety-one instances, with five records outside the key because their headings contain no date. Word counts for the fatwā texts are calculated from the cleaned text rather than the recorded length field. Column totals differ from the overall totals for the terms because some records fall outside the dating key: the “mustaḥiqq” column totals 162, against 166 across the fatwā texts, and “mawqūf ʿalayhi” totals 129, against 130 overall. The upper bound compatible with zero occurrences varies with the period’s denominator: three per cent in the first period (3%), two and a half in the second (2.5%), six in the third (6%), seven in the fourth (7%), six in the fifth (6%), and thirteen and a half in the sixth (13.5%). Across the corpus as a whole, it is approximately three-quarters of one per cent (0.75%), certainly below one per cent (1%).
D5 · Uses of “ʿadad”: the second, third and fourth.
The second is the denominator for distribution by head. In several instances, revenue is divided “according to the number of their heads, with a male receiving the share of two females”, and the waqf is described as “multiple awqāf, corresponding to the number of their heads”. Here, a head is a computational denominator in a distribution, not a unit of observation for a population.
The third is the stipulated quota in the sharia deed. In an instance from Ṣafar 1340 AH / 25 October 1921, the muftī rules that “there is no specified number of residents at the said zāwiya; it depends on what the nāẓir considers appropriate”, whereas for collectors the founder “stipulated a specific number, so his stipulation must be followed”. In an instance from Ṣafar 1347 AH / 31 July 1927, he rules that “where the founder designated neither a known group nor a specific number, but left the provision unrestricted”, those possessing the stated characteristic were included in the waqf.
The fourth is capacity derived from funding. A fatwā from Rabīʿ al-Thānī 1327 AH concerning al-Minshāwī Hospital cites the founder’s stipulation that “the number of patients shall be as many as the aforementioned funds can accommodate”. Here, the number is a calculated output of the budget, not an observation of those who fell ill.
D6 · The seven instances of “nafar”, enumerated. Ten reciters who had memorised the Book of God, in a waqf established by a deed from the al-Bāb al-ʿĀlī Court in 1217 AH. Five reciters in a stipend schedule from Rabīʿ al-Awwal 1316 AH. Eleven Sufis whose disbursement is divided into twelve parts: two for the shaykh and ten for the ten Sufis. Ten reciters who complete a recitation of the Qurʾān every Thursday eve at al-Aḥmadī Shrine. Eighteen bearers of the Qurʾān from among the Jabartī residents of Medina. Twenty Muslim children taught in an elementary school above a public drinking fountain. Five persons forming an accounts-review board for a waqf from Rajab 1351 AH / 31 October 1932.
D7 · Clarifying the attribution of the numerical position to Muḥammad ibn al-Ḥasan. The wording of al-Bināya is: “Then the threshold of enumerability according to Abū Yūsuf is that they cannot be enumerated by a written record or calculation, so they cannot be enumerated. It is also said: such that pebbles are counted with them until a child is born among them and someone among them dies, and this is Muḥammad’s opinion; if they number more than a hundred, they cannot be enumerated. Others say that the matter is left to the judge’s judgement, which is the position adopted for fatwā; what Muḥammad said is the easier. Thus in Fatāwā Qāḍī Khān.” The phrase “and this is Muḥammad’s opinion” may be read as referring only to the pebble criterion, with the sentence about a hundred beginning a new statement. It may also be read as referring to both sentences. A parallel account with more explicit wording resolves this. In al-Sighnāqī’s al-Nihāya, a commentary on al-Hidāya predating al-Bināya and likewise quoting Qāḍī Khān, the two positions are explicitly separated: “The threshold of enumerability reported from Abū Yūsuf is that they not be enumerated … It is also said: if the enumerator cannot enumerate them before a child is born among them and one of them dies, they cannot be enumerated. Muḥammad said: if they number more than a hundred, they cannot be enumerated. Others say that the matter is left to the judge’s judgement, which is the position adopted for fatwā. What Muḥammad said is the easier: thus in Fatāwā Qāḍī Khān.” The pebble criterion is thus introduced by an unattributed “it is said”, while the numerical criterion is explicitly attributed to Muḥammad. Supporting this, the answer in the 1331 AH fatwā cited in this paper makes the same attribution and identifies its source: “Muḥammad’s position is that if they number more than a hundred, they cannot be enumerated … Thus in al-Durr and its glosses.”
D8 · The second instance of applying the enumeration rule: the answer and its four elements.
The second instance is Bakrī al-Ṣadafī’s fatwā of Jumādā al-Thāniya 1331 AH. Aḥmad al-Ḥalabī’s waqf comprised approximately fifty-two feddans in the al-Sanṭa area, a third of which had devolved to twenty-four reciters, twenty-five students and twenty-five scholars. The shaykh of al-Aḥmadī Mosque asked whether, since those present outnumbered those entitled under the waqf deed, the nāẓir could designate recipients in rotation from year to year. The muftī ruled that scholars and students “do not become individually designated through designation, because these scholars and students cannot be enumerated … a waqf for them is a form of charity, for which the rule is that they are not individually designated”, whereas “the reciters of the aforementioned Qurʾān reading circle become individually designated through designation because they can be enumerated”.
This instance contains four elements of the counting system in its complete form. First, the stipulated quota is smaller than the eligible group, creating an excess for whom no seats are available. Second, Islamic jurisprudence resolves this excess through the criterion of enumerability: those who can be enumerated become individually designated, while giving rotates among those who cannot. Third, the muftī distinguished between two categories within a single waqf: the reciters of the Qurʾān reading circle were enumerable because of their connection to a known place and function, whereas the students were not. Enumerability is therefore not a property of the group alone, but of its relationship to an authority that keeps track of it. Fourth is the muftī’s ruling at the end of the answer: “Moreover, what more closely accords with the founder’s purpose is to ensure that benefit is widely shared; it is therefore more appropriate not to give to the same person in two capacities, as mentioned.” This explicitly addresses the prevention of double counting. The questioner asks directly whether a reciter may receive in his capacity as a reciter and in his capacity as a scholar or student, thereby receiving two shares. The muftī answers by favouring prohibition. The problem of the same person appearing under two descriptions—what contemporary measurement calls deduplication and identifying the unique person—was thus raised here in Jumādā al-Thāniya 1331 AH in the language of description and share.
D9 · The remaining three instances of applying the enumeration rule. In Bakrī al-Ṣadafī’s fatwā of Dhū al-Ḥijja 1324 AH concerning a waqf for poor relatives, he quotes Nuṣayr and Dāwūd as holding that the distribution is divided equally between the poor and relatives. Some later jurists held that this applies “if the relatives cannot be exhaustively enumerated; if they can, each of them receives one share and the poor receive one share”. In his fatwā of Jumādā al-Thāniya 1327 AH concerning a charitable waqf, he rules that the sum stipulated for employees of the mosque and zāwiya “is given to each of them, who can be exhaustively enumerated, in an amount appropriate to his circumstances, as required by custom and the founder’s purpose”. In Muḥammad Bakhīt’s fatwā of Shawwāl 1335 AH / 16 August 1917 concerning an entitlement waqf, he interprets the passage quoted from Ibn ʿĀbidīn as applying “where the revenue is not due to specific, enumerable persons”, referring to a particular passage in Radd al-muḥtār and to al-Anqarawiyya.
D10 · The three principles of the al-Minshāwī waqf fatwā. The most precise evidence for this in the corpus is Bakrī al-Ṣadafī’s fatwā dated Jumādā al-Ūlā 1325 AH, responding to a question from the Ministry of Public Instruction. It concerns Aḥmad Pasha al-Minshāwī’s allocation, in his waqf deed, of two hundred and fifty pounds annually “for the admission of a number of pupils who are children of poor Muslims” to Ṭanṭā Government School. Three consecutive principles appear in the fatwā: consulting the founder and obtaining his approval means seeking his opinion on their admission, “and once the founder gives permission, procedures are undertaken to register them there”; “the nāẓir may not be asked to pay sums for the past period, even if there were no pupils at the school, since that would contravene the founder’s stipulation”; and “the sum designated for disbursement to pupils of this school may not be disbursed to them after they leave it and enrol in schools of a higher level”.
D11 · Breakdown of the nāẓir reports figure and the rows read from the registers table. The source divides the figure into three categories: one thousand four hundred and seventy-one reports appointing Egypt’s rulers as nāẓirs over certain awqāf; one thousand three hundred and thirty-three appointing the Minister of Awqāf in his official capacity as nāẓir up to 1952; and twenty-seven thousand eight hundred and seven appointing private individuals as nāẓirs. The three categories sum to the total exactly. The source defines the report unit as follows: “Each nāẓir report is a decision issued by a sharia court appointing a nāẓir to a waqf, or, in some cases, appointing a new nāẓir alongside the existing nāẓir.” The four rows read from the registers table are: seventy-four registers containing twenty-four thousand four hundred and eighty-three reports appointing nāẓirs; nine registers containing one thousand three hundred and thirty-three appointing the ministry; seven containing seven hundred and one; and four containing two thousand one hundred and sixty-eight.
D12 · Table of measurement levels and the availability of source material for them.
| Level | Example in the paper’s material | Availability |
|---|---|---|
| Input | Funds, endowed principal, employee, building | Abundant |
| Activity | Teaching, feeding, disbursing regular provision | Available descriptively |
| Output | Registered pupil, fez, share received | Relatively abundant |
| Reach | A unique person who received the service | Scattered and confined to individual awqāf |
| Outcome | Learning, health, improved circumstances | Scattered and irregular |
| Impact | Change causally attributable to the benefit | Undocumented in the internal material |
Table D3: Measurement levels and the availability of source material for them. Counting people or portions is not equivalent to measuring outcomes, and measuring outcomes is not equivalent to establishing impact. The assessment of the “reach” row was raised from its initial level after the numerical survey revealed instances of verifying actual presence, although these remained confined to individual awqāf.
D13 · The first use of “ʿadad”: counting papers and documents. In an instance from Ramaḍān 1338 AH, documents are submitted to the muftī “numbering 23”; in another from Rabīʿ al-Awwal 1314 AH, papers are sent “numbering 4”; and in a third from Dhū al-Ḥijja 1317 AH, papers are sent “numbering 3”. The clearest use of counting terminology in this corpus thus concerns documents rather than persons. The fifth use is a hypothetical example used in reasoning within the answer: in a fatwā of Shaʿbān 1333 AH, the muftī says, “If we suppose that the number of scholars when the waqf was established was forty, of whom ten were each resident in Ṭanṭā.” The number is an assumption in an illustrative example, not an instance of counting.
D14 · The third piece of evidence in the category of verifying actual presence. It dates from Shaʿbān 1334 AH / 8 June 1916 and reads: “Since none of the freed persons remains except the aforementioned Sitt Wasīla alone, and the children of the freed persons who are entitlement-holders and are present now in the second tier are …”, followed by their names. The enumeration is by name and exhaustive, and it is tied to a particular moment: “now”.
D15 · Table D4: Units of counting in the material and the limits of conversion between them.
| Unit | What it counts | Repetition problem | Identity retention | Limit of conversion |
|---|---|---|---|---|
| Registered name | A person with a position in a register | May remain despite absence | High | Does not equal a person present or a person receiving payment |
| Stipulated person-slot | A fixed seat in the sharia deed | Filled through replacement | Medium | Does not equal those eligible under the description |
| Portion or loaf | A unit of benefit disbursed | Repeated with every disbursement | Low | Cannot be converted into a person without a distribution rule |
| Resident student or pupil | A person affiliated with an institution | May recur across years | Medium | Stock does not equal the flow of entrants |
| Seat or bed | Capacity | Does not represent actual receipt | None | Does not equal a user |
| Document or decision | An administrative instance | Recurs for the same waqf | None | Outside the beneficiary denominator |
Table D4: Units of counting in the material and the limits of conversion between them. A single series cannot be constructed before identifying the possible conversions and the information lost in each. The implication is that the waqf counting apparatus uses multiple units that cannot be internally converted into one another, and that the stipulated person-slot is its unit only in the context of the sharia deed, not in every context.
D16 · The evidentiary scale. Each piece of evidence is tagged with one of four grades: directly examined; drawn from a direct source through an accessible reference; contemporary descriptive; and exploratory.
D17 · Arithmetic audit of the Dīwān (bureau) of Awqāf statistics table for 1899. The text contains three categories of counting, each with a construction row, a repair row and a total, followed by a grand total. Its figures are reproduced as written, preserving their leading zeros: “030 mosques and charitable shelters constructed, 1000 mosques and charitable shelters repaired, 1030 in total; 090 revenue-producing buildings constructed, 2750 revenue-producing buildings repaired, 2740 in total; 060 agricultural buildings constructed, 0300 agricultural buildings repaired, 0360 in total; 4230 total buildings on which work was undertaken.” The audit finds that the first and third categories reconcile, whereas the second does not, and that the grand total resolves the discrepancy. The mosque row gives 30 and 1000, summing to 1030, exactly the stated figure. The agricultural-buildings row gives 60 and 300, summing to 360, again exactly the stated figure. For revenue-producing buildings, however, 90 and 2750 sum to 2840, whereas the stated figure is 2740. The grand total supports 2840 as the likely figure, since 1030, 2840 and 360 sum to exactly 4230, the stated grand total; using 2740 instead yields four thousand one hundred and thirty. It is therefore likely that “2740” is a transcription error for “2840”. The evidence is arithmetic internal to the table, not a guess about the appearance of the digits. This differs from the al-Azhar table in D1, where no evidence resolved the discrepancy. Neither table is used as the basis for a calculation in this paper; the first is used only as evidence of the unit of counting, not its magnitude.
D18 · The place of the threshold of enumerability within the chapters of Islamic jurisprudence, and the layers of its transmission. This version distinguishes three layers of the question. The first is the enumeration rule for a particular waqf, supported by three independent sources: al-Sarakhsī’s wording in “Sharḥ Kitāb al-Waqf”, quoted in al-Muḥīṭ al-Burhānī; its counterpart in al-Mabsūṭ; and Ibn Nujaym’s wording in al-Baḥr al-rāʾiq, p. 215, examined in the original. Alongside these is al-Isʿāf on a waqf for orphaned relatives. The second is the threshold of enumerability itself and the positions on it, for which there are two accounts rather than one. The account in al-Bināya and al-Nihāya presents three positions and appears in the chapter on bequests to a group, not the chapter on waqf; both works quote Fatāwā Qāḍī Khān. The account in al-Isʿāf and al-Baḥr al-rāʾiq presents five positions and appears in the waqf chapter itself. It attributes the numerical threshold to Abū Yūsuf and the figure of ten to Muḥammad, reversing the first account. This is clarified in footnote 20. The third layer is the textual link between the two chapters, found in al-Fatāwā al-Hindiyya and quoted within the answer to the 1331 AH fatwā: an intention to give charity entails differentiation at the discretion of the qayyim (waqf administrator), whereas an intention to make a bequest entails disbursement “according to their number”. This distinction means that the paper does not transfer a ruling from one chapter to another by analogy; it presents a link explicitly stated in the sources, then shows how muftīs applied it to particular instances. What has not been examined in the original among these layers is al-Fatāwā al-Hindiyya, which remains at the grade of material drawn from a direct source. This version corrects an earlier description of these layers which stated that Ḥanafī jurists had not initially specified a numerical threshold in the waqf chapter. Inspection of al-Isʿāf and al-Baḥr al-rāʾiq showed that the numerical threshold, with its five positions, is explicitly stated in the waqf chapter itself. The evidence used in section 6 nevertheless remains unchanged: Egyptian muftīs’ application of the threshold to particular waqf instances between 1314 AH and 1337 AH.
D19 · Four corrected attributions and their effect on the conclusion. The first concerns a passage previously attributed to Ibn Māza in al-Muḥīṭ al-Burhānī. In its original location it is explicitly quoted from al-Sarakhsī’s “Sharḥ Kitāb al-Waqf”, followed by a critical comment from Ibn Māza on the implications of its wording that had been omitted through ellipses. The second concerns a statement previously attributed to Abū Zahra as his own juristic derivation. In its original location, it is his distillation of Ibn Nujaym’s preceding words, expressed in the third person; Abū Zahra then qualifies it with a restriction stated afterwards. Both passages are clarified in footnote 17. The correction changes none of the paper’s conclusions, since the same ruling is established by three independent sources examined directly: al-Muḥīṭ al-Burhānī, al-Mabsūṭ and Ibn Nujaym’s wording. It strengthens the evidentiary standard: the ruling now has two independent sources in addition to the first, whereas the first source and the parallel statement attributed to Abū Zahra had in fact derived from a single layer of transmission. The third correction concerns Ibn Nujaym’s wording, which had been presented as an independently examined source although it had been quoted indirectly through Abū Zahra. Its original in al-Baḥr al-rāʾiq, p. 215, was therefore consulted and its exact wording recorded. Comparing the original with the intermediary revealed that a sentence had been omitted from the latter without ellipses; this is clarified in footnote 17. The fourth, and most consequential, is that the waqf chapter contains an account of the numerical threshold that differs from the account in the bequests chapter in its initial attribution of the two positions. Its exact wording from al-Isʿāf, p. 14, and al-Baḥr al-rāʾiq, p. 215, is set out in footnote 20. The fourth correction affects attribution alone, not substance: the paper’s argument is that a numerical threshold separates a quantified right from discretionary allocation, and both accounts establish this. Determining the more likely attribution between the two accounts remains outside this paper’s scope and is submitted for the author’s review.
Appendix E: Texts of the Comparator’s Provisions in Force (German Civil Code (BGB))
The provisions of the German Civil Code (Bürgerliches Gesetzbuch) cited in this paper are reproduced here in their German wording, so that readers can compare the main text’s account with the provisions themselves without consulting an external source. Appendices are excluded from the main-text word count.
These provisions support the statements in section 6.A: the conditions for recognising a foundation (§ 82), the conditions for amending its charter (§ 85), approval of the amendment by the federal state’s authority (§ 85a(1) sentence 2), and protection of recipients’ claims and their right to be heard during absorption and merger (§ 86 no. 4, § 86a no. 3 and § 86e(2)). Their numbers have now been supplied in the main text, where they had previously been described without numbered references, enabling readers to compare each statement with its statutory text.
Scope of the evidentiary record: the texts come from a consolidated mirror of federal legislation (buzer.de), which gives each provision’s authority in the official gazette (BGBl. 2021 I S. 2947, effective 1 July 2023). The gazette itself has not been examined.
The authority for entry into force is the same for all these provisions and reads: Text in der Fassung des Artikels 1 Gesetz zur Vereinheitlichung des Stiftungsrechts und zur Änderung des Infektionsschutzgesetzes G. v. 16. Juli 2021 BGBl. I S. 2947; zuletzt geändert durch Artikel 33 G. v. 08.12.2025 BGBl. 2025 I Nr. 319 m.W.v. 1. Juli 2023.
A note on numbering: the 2021 reform changed both numbering and content, so any reference in literature published before 1 July 2023 to §§ 80–88 refers to a text other than the one reproduced here.
§ 82 BGB · Anerkennung der Stiftung
1 Die Stiftung ist anzuerkennen, wenn das Stiftungsgeschäft den Anforderungen des § 81 Absatz 1 bis 3 genügt und die dauernde und nachhaltige Erfüllung des Stiftungszwecks gesichert erscheint, es sei denn, die Stiftung würde das Gemeinwohl gefährden. 2 Bei einer Verbrauchsstiftung erscheint die dauernde Erfüllung des Stiftungszwecks gesichert, wenn die in der Satzung für die Stiftung bestimmte Zeit mindestens zehn Jahre umfasst.
§ 85 BGB · Voraussetzungen für Satzungsänderungen
(1) 1 Durch Satzungsänderung kann der Stiftung ein anderer Zweck gegeben oder der Zweck der Stiftung kann erheblich beschränkt werden, wenn 1. der Stiftungszweck nicht mehr dauernd und nachhaltig erfüllt werden kann oder 2. der Stiftungszweck das Gemeinwohl gefährdet. 2 Die Voraussetzungen des Satzes 1 Nummer 1 liegen insbesondere vor, wenn eine Stiftung keine ausreichenden Mittel für die nachhaltige Erfüllung des Stiftungszwecks hat und solche Mittel in absehbarer Zeit auch nicht erwerben kann. 3 Der Stiftungszweck kann nach Satz 1 nur geändert werden, wenn gesichert erscheint, dass die Stiftung den beabsichtigten neuen oder beschränkten Stiftungszweck dauernd und nachhaltig erfüllen kann. 4 Liegen die Voraussetzungen nach Satz 1 Nummer 1 und Satz 3 vor, kann eine auf unbestimmte Zeit errichtete Stiftung auch abweichend von § 83c durch Satzungsänderung in eine Verbrauchsstiftung umgestaltet werden, indem die Satzung um Bestimmungen nach § 81 Absatz 2 ergänzt wird. (2) 1 Durch Satzungsänderung kann der Stiftungszweck in anderer Weise als nach Absatz 1 Satz 1 oder es können andere prägende Bestimmungen der Stiftungsverfassung geändert werden, wenn sich die Verhältnisse nach Errichtung der Stiftung wesentlich verändert haben und eine solche Änderung erforderlich ist, um die Stiftung an die veränderten Verhältnisse anzupassen. 2 Als prägend für eine Stiftung sind regelmäßig die Bestimmungen über den Namen, den Sitz, die Art und Weise der Zweckerfüllung und über die Verwaltung des Grundstockvermögens anzusehen. (3) Durch Satzungsänderung können Bestimmungen der Satzung, die nicht unter Absatz 1 oder Absatz 2 Satz 1 fallen, geändert werden, wenn dies der Erfüllung des Stiftungszwecks dient. (4) 1 Im Stiftungsgeschäft kann der Stifter Satzungsänderungen nach den Absätzen 1 bis 3 ausschließen oder beschränken. 2 Satzungsänderungen durch Organe der Stiftung kann der Stifter im Stiftungsgeschäft auch abweichend von den Absätzen 1 bis 3 zulassen. 3 Satzungsbestimmungen nach Satz 2 sind nur wirksam, wenn der Stifter Inhalt und Ausmaß der Änderungsermächtigung hinreichend bestimmt festlegt.
§ 85a BGB · Verfahren bei Satzungsänderungen
(1) 1 Die Satzung kann durch den Vorstand oder ein anderes durch die Satzung dazu bestimmtes Stiftungsorgan geändert werden. 2 Die Satzungsänderung bedarf der Genehmigung der nach Landesrecht zuständigen Behörde. (2) Die Behörde kann die Satzung nach § 85 ändern, wenn die Satzungsänderung notwendig ist und das zuständige Stiftungsorgan sie nicht rechtzeitig beschließt. (3) Wenn durch die Satzungsänderung der Sitz der Stiftung in den Zuständigkeitsbereich einer anderen Behörde verlegt werden soll, bedarf die nach Absatz 1 Satz 2 erforderliche Genehmigung der Satzungsänderung der Zustimmung der Behörde, in deren Zuständigkeitsbereich der neue Sitz begründet werden soll.
§ 86 BGB · Voraussetzungen für die Zulegung
Durch Übertragung ihres Stiftungsvermögens als Ganzes kann die übertragende Stiftung einer übernehmenden Stiftung zugelegt werden, wenn 1. sich die Verhältnisse nach Errichtung der übertragenden Stiftung wesentlich verändert haben und eine Satzungsänderung nach § 85 Absatz 2 bis 4 nicht ausreicht, um die übertragende Stiftung an die veränderten Verhältnisse anzupassen, oder wenn schon seit Errichtung der Stiftung die Voraussetzungen für eine Auflösung nach § 87 Absatz 1 Satz 1 vorlagen, 2. der Zweck der übertragenden Stiftung im Wesentlichen mit einem Zweck der übernehmenden Stiftung übereinstimmt, 3. gesichert erscheint, dass die übernehmende Stiftung ihren Zweck auch nach der Zulegung im Wesentlichen in gleicher Weise dauernd und nachhaltig erfüllen kann, und 4. die Rechte von Personen gewahrt werden, für die in der Satzung der übertragenden Stiftung Ansprüche auf Stiftungsleistungen begründet sind.
§ 86a BGB · Voraussetzungen für die Zusammenlegung
Mindestens zwei übertragende Stiftungen können durch Errichtung einer neuen Stiftung und Übertragung ihres jeweiligen Stiftungsvermögens als Ganzes auf die neue übernehmende Stiftung zusammengelegt werden, wenn 1. sich die Verhältnisse nach Errichtung der übertragenden Stiftungen wesentlich verändert haben und eine Satzungsänderung nach § 85 Absatz 2 bis 4 nicht ausreicht, um die übertragenden Stiftungen an die veränderten Verhältnisse anzupassen, oder wenn schon seit Errichtung der Stiftung die Voraussetzungen für eine Auflösung nach § 87 Absatz 1 Satz 1 vorlagen, 2. gesichert erscheint, dass die neue übernehmende Stiftung die Zwecke der übertragenden Stiftungen im Wesentlichen in gleicher Weise dauernd und nachhaltig erfüllen kann, und 3. die Rechte von Personen gewahrt werden, für die in den Satzungen der übertragenden Stiftungen Ansprüche auf Stiftungsleistungen begründet sind.
§ 86e BGB · Behördliche Zulegungsentscheidung und Zusammenlegungsentscheidung
(1) Auf den Inhalt der Entscheidungen über die Zulegung oder Zusammenlegung von Stiftungen durch die nach Landesrecht zuständige Behörde ist § 86c Absatz 1 und 2 entsprechend anzuwenden. (2) Die Behörde hat Personen nach § 86c Absatz 1 Satz 2 mindestens einen Monat vor der Entscheidung über die Zulegung oder Zusammenlegung anzuhören und auf die möglichen Folgen der Zulegung oder Zusammenlegung für deren Ansprüche gegen eine übertragende Stiftung hinzuweisen.
What this appendix establishes is limited to the wording of these provisions. The basis for denying the beneficiary supervisory powers is the Federal Court of Justice judgement of 15 December 2016, not any of these provisions. The 2016 judgement predates the 2021 reform that entered into force in 2023. It is therefore cited to support the absence of supervisory powers, not to deny the existence of a provision governing the recipient’s relationship. Arguments relying on the deferred register of foundations have been removed from section 6.A and are not included here, since their statutory basis lies outside these provisions and its entry into force has been deferred.