A pre-Mongol New Persian legal document from Islamic Khurāsān dated ah 608/1212 ce

Abstract Since the 1990s several caches of New Persian documents have come to light in Afghanistan. These documents, written on paper, are now the most significant sources for understanding how New Persian in Arabic script was used as an administrative and legal language in the eastern Islamic lands between the eleventh and early thirteenth centuries before the Mongol conquest of Khurāsān. After a brief survey of the three main collections in which these New Persian paper documents are preserved today, this article presents a preliminary edition, translation and commentary on one of the New Persian documents held in the Nasser D. Khalili Collection of Islamic Art. The document, dated ah 608/1212 ce, is a record of court proceedings and the decision of a judge (qāḍī) in a lawsuit over water rights initiated by a woman.


Introduction
Since the 1990s more than 200 New Persian paper documents from Afghanistan have come to light in separate caches. 1 These documents are of outstanding historical significance as they provide us with a glimpse of everyday life in medieval Islamic Khurāsān (in the region of present-day Afghanistan) betweenthe eleventh and early thirteenth centuries. 2They also have the potential to transform our understanding of the emergence and use of New Persian in Arabic script following the Arab conquests. 3Until now, such research has mostly relied on the earliest known inscriptions and manuscript codices, from the ninth century onwards, which use New Persian in Arabic script. 4The New Persian documents from Afghanistan are of particular significance for understanding how New Persian in Arabic script was used to write legal and administrative documents in the Islamic east prior to the Mongol conquest of Khurāsān in the early thirteenth century.Until now there were no known pre-Mongol administrative documents in New Persian in Arabic script, and only a handful of pre-Mongol legal documents in New Persian in Arabic script dating from the eleventh to the early thirteenth centuries from Khotan,5 Khurāsān 6 and Ardabil 7 had been studied.
The exact circumstances in which the pre-Mongol New Persian documents from Afghanistan were first discovered, then appeared on the market and, in at least two cases, were acquired from Afghanistan through the intermediary of dealers are unclear.This poses an ethical dilemma for scholars wishing to work with this material given its outstanding research potential.Today these documents are preserved in three separate collections.The first is the Afghan Geniza collection acquired between 2013-16 by the National Library of Israel (NLI) in Jerusalem. 8 The pre-Mongol New Persian documents of the Afghan Geniza collection can be divided chronologically into two separate groups.An earlier group consists mainly of legal documents and letters dating from the beginning of the eleventh century during the period of Ghaznavid rule in Khurāsān.Most of the documents in this group belonged to the private archive of a Jewish family living in Bāmiyān in central Afghanistan. 9The second group has legal and administrative documents dating from the second half of the twelfth to the early thirteenth centuries. 10here is some internal evidence to suggest that the documents from this group are also from Bāmiyān and its region. 11The administrative documentsdecrees, receipts, letters and listsof the second group are of particular significance for research on archival practices as they appear to be linked to various state officials and local archives (dīwāns). 12esides the documents in the Afghan Geniza collection, a second collection of pre-Mongol New Persian documents was discovered by treasure hunters in 1370 sh./ 1991 inside a cave near the village of Shahr-i Kharu, in Ghalmīn, 30 kilometres north of Chaghcharān (Fīrūzkūh), the capital of Ghūr province in central Afghanistan. 13Until recently these New Persian documents (hereafter the Ghūr New Persian documents) were held in the private collection of an inhabitant of Ghūr, a local calligrapher named Mīrzā Khwāja Muḥammad.In 1388 sh./2009, in collaboration with Nabī Sāqī, Mīrzā Khwāja Muḥammad published an edition of 84 documents with facsimiles. 14In 1399 sh./ 2020, Mīrzā Khwāja Muḥammad entrusted the Ghūr documents to the National Archives of Afghanistan where they are presently held. 15The edited Ghūr material consists of different types of legal and administrative documents which closely resemble the second group of the Afghan Geniza New Persian documents.The Ghūr documents are also dated between the second half of the twelfth to the early thirteenth centuries and mention villages and places mainly in Ghūr province itself.There appears to be no identifiable connection to Bāmiyān and its region.
Finally, a third, much smaller, set of ten pre-Mongol New Persian documents, also from Afghanistan, now forms part of the Nasser D. Khalili Collection of Islamic Art in London.These New Persian documents were acquired, along with the Arabic and Bactrian documents from Afghanistan which became known in the 1990s.The Arabic and Bactrian documents have since been edited and published. 16Based on internal evidence, the Arabic and Bactrian documents originate from northeastern Afghanistan, in an area lying between Balkh and Bāmiyān.The Khalili New Persian documents have not yet been examined.Seven documents contain the text of eight complete deeds of acknowledgement (iqrārs). 17These iqrārs acknowledge: the sale or transfer of agricultural land (593/ 1197, 18 597/1201, 19 610/1214 20 and 617/1220 21 ), marriage (594/1198, 22 598/1202 23 ), debt (605/1209 24 ) and marital relations (undated). 25In addition, there is an undated list of items, 26 an undated legal fragment 27 and a court record of proceedings in a lawsuit over water rights that was held before a judge (qāḍī) dated 608/1212. 28It should be mentioned here that the same group of documents also includes an iqrār in Arabic (600/ 1204) 29 and three talismanic rolls. 30he spatial setting of the Khalili pre-Mongol New Persian documents in relation to the pre-Mongol New Persian Afghan Geniza and Ghūr documents is not clear as the toponyms mentioned in the Khalili New Persian documents have not yet been identified.Some tentative identifications, however, can already be made at this stage.One of the Khalili New Persian iqrārs mentions an individual who is described as an inhabitant of Fīrūz province (wilāyat-i fīrūz), which presumably refers to the area of Fīrūzkūh, the summer capital of the Ghurid Dynasty in Ghūr province. 31Three of the Khalili New Persian iqrār documents mention Bāmiyān, and one mentions Nīshāpūr [var.Nīsābūr] when specifying where currency was minted. 32Far more difficult to identify are the names of the villages.One of the iqrārs refers to the sale of agricultural lands in the mountainous area around the village or small town (qaṣaba) of ‫ﺍ‬ ‫ﺳ‬ ‫ﺘ‬ ‫ﺎ‬ ‫ﻕ‬ . 33This could possibly be identified as present-day Istak/Estak, situated around 200 kilometres southwest of Bāmiyān.This seems to be confirmed by the fact that the mountainous lands referred to in the document are located among the mountains of ‫ﺍ‬ ‫ﺳ‬ ‫ﺘ‬ ‫ﺎ‬ ‫ﻕ‬ at a place on the outskirts of ‫ﺣ‬ ‫ﺎ‬ ‫ﺑ‬ ‫ﻚ‬ . 34This could read as Chabak, a mountain located midway between Bāmiyān and Istak.According to the document, the agricultural lands in question were known locally as Ist.ʾ/n.wā.bīk . 35 The reading and vocalization of this toponym beginning with Ist is uncertain.This is perhaps also the same place referred to in the Khalili New Persian qāḍī court record.
In what follows, I propose to examine this court record dated 608/1212 in more detail.After some general remarks on its significance, I provide an edition, translation and commentary on the document.Arabic vocalization marks, shadda, madda, final dotted yāʾ and initial hamza are only indicated if they appear in the original.The Arabic tāʾ marbūṭa is indicated in the edition when it is not in a Persian iḍāfa construction.In the commentary, I compare the document with two twelfth-century court records in Arabic from the Yārkand oasis in present-day Xinjiang, China, and the examples of such documents found in Ḥanafī model legal formularies (shurūṭ) from twelfth-to thirteenth-century Transoxiana.I have chosen these sources for comparison as they use similar legal formulae and are thus crucial for deciphering the New Persian Khalili qāḍī court record.
A New Persian qa ̄ḍ ı ̄court record on water rights in the Khalili collection dated 608/1212 The New Persian qāḍī court record in the Khalili collection dated 608/1212 (see Figure 1) is a rare example of this type of legal document from pre-Mongol Khurāsān.The remaining legal documents in the pre-Mongol New Persian Khurāsān corpus are either iqrārs, fatwās, deeds of sale or settlement. 36The only known equivalent so far is a qāḍī court record of a dispute over custody and maintenance payment (nafaqa) dated 26 Ramaḍān 5[.]4/1169-98 among the Ghūr documents. 37The Ghūr New Persian qāḍī court record has survived only partially as the top fragment is missing.It is not clear therefore if it contained the qāḍī's authenticating signature (tawqīʿ) in Arabic at the top of the document, as in the case of the Khalili qāḍī court record (see below). 32Khalili doc.48, line 12: sīm-i rasmī-yi ḍarb-i bāmiyān; Khalili doc.49, lines 11-12: haftād dīnār sīm-i ḍarb-i bāmiyān; Khalili doc.39: az sīm-i rasmī-yi naqd-i waqt-i haḍrat-i bāmiyān; Khalili doc.37, lines 7-8: sīm-i zarʿīn-i maḍrūb bi-nīsābūrī.
36 On the New Persian fatwās from medieval Khurāsān, see Bhalloo and Ishkawari (forthcoming, 2024c).On two unedited deeds of sale dated 400?/1009?and 405/1015, see Haim (2019c).For the earliest settlement deed, dated 473/1080-81, which settles an inheritance dispute between a sister and her brother, see Fīrūzbakhsh (1400sh./2022).The Ghūr documents also include a deed of settlement of grain dated 607/1211 between the male heirs of a deceased man; see P. Ghur 18.  37 P.Ghur 14.The facsimile of the document does not include the last segment, which has the date and the names of the witnesses.According to the edition, the number between 5 and 4 is illegible.Based on the dates of the other Ghūr legal documents, it is likely the record was produced in the second half of the twelfth century.
The Ghūr qāḍī court record is more informal in its style and structure compared to the Khalili qāḍī court record.Unlike the latter, it does not contain a detailed description of the proceedings in the lawsuit with a protocol of claims made by both parties, a record of witness testimonies and the decision and note of certification of the judge.Moreover, the entire record is narrated by the qāḍī himself.This contrasts with the Khalili qāḍī court record where the text shifts, depending on the stage of the proceedings, between the voice of the qāḍī, the parties involved in the lawsuit and the witnesses.The Ghūr qāḍī court record is also written entirely in New Persian with only the witness clauses in Arabic.The Khalili qāḍī court record, however, in addition to using New Persian, has a significant amount of Arabic and uses Perso-Arabic clauses for both the proceedings and the witness clauses.The distinctive formulae of the Khalili qāḍī court record is therefore of considerable interest.It is also the only known example we have so far from medieval Islamic Khurāsān of a woman initiating legal proceedings before a qāḍī to claim the restitution of her rights. 38In this case, the judge dismissed her claim.
The proceedings described in the Khalili New Persian qāḍī court record took place on 15 Dhū l-Ḥijja 608/26 May 1212 before the qāḍī Muḥammad b.Ismāʿīl b.Ibrāhīm b.ʿUmar.The claimant was a woman named Fāṭima bt.Luqmān b. al-Ḥasan, and the defendant, a man named Mīr Khwāja Muḥammad b.Ibrāhīm b. ʿAlī.According to the claimant, the defendant was in illegal possession of one tīr out of nine tīr of the waters of a certain place, which rightfully belonged to her. 39 She claimed she had received this measure of water from her husband Ḥusayn b.Ibrāhīm (presumably deceased at the time of the lawsuit) in lieu of her dowry (mahr).In his reply to the claimant's claim, the defendant said he had bought the measure of water from the claimant's husband in her presence for 70 silver dīnārs.The defendant brought two male witnesses to court to testify.The latter both confirmed having witnessed the sale transaction.The qāḍī made both witnesses take an oath on the veracity of their statements as a precaution and then issued a decision in favour of the defendant's ownership of the measure of water.The proceedings and issuance of the qāḍī's decision was witnessed by nine witnesses.Structurally, the text of the court record is arranged as follows: 7. Mīr Khwāja Muḥammad b.Ibrāhīm b. ʿAlī.This person (Fāṭima) who was present in court made a claim against this person (Mīr Khwāja Muḥammad) also present in court 8. saying: "One tīr out of nine tīr of water at the begining of the [S/sh.n.j/ḥ.ūna?] valley 9. along with all its boundaries is my right and property as the measure (qadr) which (my husband) Ḥusayn b.Ibrāhīm 10. gave to me in lieu of my dowry.This measure, which is my right, is in the possession of the defendant 11. illegally.The (purchase) amount (of this measure) must be surrendered (to me) by him."After the hearing of the claim, 12. upon her (i.e. the claimant's) word, the defendant was questioned.He (Mīr Khwāja Muḥammad) replied: "It (i.e. this measure) is my right 13. and property which I bought from the claimant in the presence of her husband Ḥusayn b.Ibrāhīm 14. for 70 silver dīnārs in current use minted in [J/Ḥ.l.b/t/th a/l.ḥ.f?], may God exalt it, through a lawful sale, 15. and I gave the (purchase) amount to her husband in her presence."The claimant denied the said sale 16.and surrender of the said sum.Witnesses were requested from the defendant.17 "These are my witnesses, question them."19.After asking for their testimony, they (i.e. the two witnesses) gave valid testimony (saying): "One tīr 20.out of nine tīr of water which is disputed was sold for 70 dīnārs currently in use by this claimant 21. to this defendant in a valid sale.The (purchase) amount was received in full 22. by the claimant.This is our witness testimony and it makes the claim void 23. and for this reason this water is the right and property of this defendant."They (the two witnesses) pointed 24. to the appropriate positions and their testimonies were consistent in their aspect and manner.25.I heard their testimonies and made them each swear an oath successively on the veracity of their testimony in order to confirm it 26.and out of precaution, and I recorded it (i.e. the testimonies).After taking refuge in God, 27. the exalted, from error, oversight and what brings down retribution, I gave a judgement in favour of the ownership of the defendant and 28. the validity of the sale and transfer of the sum to her (i.e. the claimant).This defendant requested this record from me 29.so that it could be a proof in his hands.I agreed to this and wrote this on the date written at the beginning of the document.30.Muḥammad b.Ismāʿīl b.Ibrāhīm says: "It (i.e. this document) contains my decision and judgment.31.The signature at the begnining and this (record) from the beginning 32. to the end is in my own handwriting and these four lines 33. after the date and its mention is in my own handwriting and I called upon those present before me to witness to it."

Textual notes
1.A small ornamental calligraphic ḥāʾ is visible under the ḥāʾ of al-ḥukm.
3. The letter bāʾ is used instead of pāʾ for the Persian number pānzdah.The Persian iḍāfa is indicated with a hamza or small yāʾ after the silent final hāʾ of dhū l-ḥijja.This hamza or small yāʾ also appears after qaṣaba (line 4), fāṭima (line 6) and darra (line 8).It is also visible in P.Ghur 10, line 5 after zawja and after fāṭima and paywasta in Khalili doc. is uncertain.The honorific term ḥaḍrat that precedes it occurs in Khalili doc.39 in relation to the place where coins were minted: az sīm-i rasmī-yi naqd-i waqt-i haḍrat-i bāmiyān.This suggests the term is a toponym, in particular in relation to the clause ajallahā l-lāh that follows it.19.The reading of the adjective mazāj after ṣaḥīḥ is uncertain.It is possibly a lapsus calami for mujāz, meaning permitted.
Commentary 1.The qāḍī's signature (tawqīʿ).In line 31, the qāḍī refers to the pious formula in Arabic at the beginning of the documental-ḥukm li-llāh al-ʿaliyy al-kabīr ( judgement belongs to God, the exalted and great)as his tawqīʿ (signature).We know from Abbasid literary sources that the term tawqīʿ was used to refer to pious formulae that functioned as a personal signature. 42This usage of the term tawqīʿ survived in the Islamic east for the pious formula used by the qāḍī as his signature.In Egypt, however, al-Asyūṭī (d.800/1475) refers to the pious formula used by the qāḍī as his personal signature as his ʿalāma (sign). 43In the Khurāsān court record examined here, the qāḍī's tawqīʿ is written vertically at the top left-hand corner of the document perpendicular to the basmala.This spatial orientation of the tawqīʿ in relation to the basmala is so far the only known example of its kind.The qāḍī's tawqīʿ in two comparable Arabic court records from the Yārkand oasis is written parallel to the basmala on the top-left hand corner of the document. 44The first Yārkand document is a court decision in a land ownership dispute dated 474/1082 or 494/1101 (P.GronkeYarkand 1).The tawqīʿ used by the qāḍī is aḥmadu l-lāh waḥdahu (I praise God alone).The second (P.GronkeYarkand 2), an order of court concerning an intestacy dated 503/1110, has the tawqīʿ: iʿtaṣamtu bi-llāh (I seek assistance from God).The qāḍīs in both these Yārkand documents, as in our document, refer to these pious formulae at the beginning (ṣadr) of the document as their tawqīʿ.In addition to the tawqīʿ at the start of the document, the qāḍī's note of certification at the end of the court record and the accompanying witness clauses all had an authenticating function.In case of later infringement of rights, the court record could not serve as an argument or proof (ḥujja) in court without them (see 29-32 below).467-8 (facsimile).This parallel orientation is also the case for the ʿalāma, alḥamdu li-llāhi rabbi l-ʿālamīn of the qāḍī Sharaf al-Dīn ʿIsā in the Ḥaram documents.See, for example Ḥaram document no.39, facsimile in Müller (2013): 634. 45The Arabic equivalent majlis al-ḥukm for the qāḍī's court is well attested in other documents; see, for example, T-S 28.8; P.TillierRancon.al-ḥukm. 46After the opening formula, the place where the court was located is described.According to the Ḥanafī jurist al-Ṭaḥāwī (d.321/933), mentioning the place was only necessary when the court record being produced was a sijill not a maḥḍar. 47From a Ḥanafī perspective, therefore, the Khurāsān court record was a sijill.According to Ḥanafī shurūṭ, the sijill was the final court record containing the qāḍī's decision, while the maḥḍar was an initial record of court proceedings upon which the sijill was based. 48Precisely what constituted a maḥḍar and sijill in the Islamic world, however, varied depending on the place, the period and on the school of law in question. 49A distinction should also be made between the model maḥḍars and sijills presented in shurūṭ works and actual surviving court records.Though closely intertwined, as Hallaq has argued, actual documents show how the legalese of shurūṭ works was used in practice and, moreover, provide important local perspectives missing in the shurūṭ literature. 50After the date of the proceedings and location of the court, the name of the qāḍī is mentioned in a clause concerning his appointment.This appointment clause confirms that Muḥammad b.Ismāʿīl b.Ibrāhīm b.ʿUmar was the qāḍī in the place mentioned on behalf of the holder of authority according to Islam's sacred law (min qibal man lahu l-wilāya sharʿan).This is most likely a reference to the qāḍī's appointment by the political ruler in the region (whose name is not mentioned).The sijill examples in the twelfth-to thirteenth-century Ḥanafī shurūṭ works of Ẓahīr al-Dīn al-Marghīnānī (fl.c. 600/1203) (ZM) 51 and Ibn Māza al-Bukhārī (d.616/1219) (MB-K) 52 use the clause min qibal al-sulṭān fulān and min qibal al-khāqān fulān. 53As this min qibal clause is not mentioned by al-Ṭaḥāwī, it is likely that the addition of this clause was a later Ḥanafī development. 54It is also not found in the Ḥaram court records of the Shāfiʿī judge Sharaf al-Dīn (d.797/1395) from Mamluk Jerusalem. 55.GronkeYarkand 1 mentions the name of the ruler after the min qibal clause. 56n P.GronkeYarkand 2, however, the qāḍī mentions deriving his appointment from a higher ranking qāḍī. 576-15.Record of the claim made by the claimant and the reply of the defendant.The record of the proceedings itself begins when the claimant, Fāṭima bt.Luqmān b. al-Ḥasan, brings the defendant, Mīr Khwāja Muḥammad b.Ibrāhīm b. ʿAlī, to court.The Persian verbal construction used here is a direct translation of the Arabic formula ḥaḍara…wa aḥḍara maʿa nafsihi found in shurūṭ works. 58According to Ḥanafī shurūṭ, if the claimant and defendant were known to the judge, their names and genealogies could be recorded directly.If not, their names and genealogies were to be mentioned as follows: rajulun dhukira annahu yusammā fulān b. fulān (a man who it is said is named so-and-so, son of so-and-so). 59Ḥanafī shurūṭ also recommend that a description of the physical features of the person be provided if they were not known to the judge. 60The use of musammāt (so-called) before the name of the claimant in our record might suggest, therefore, that she was not known to the qāḍī.This is by no means certain.In P.GronkeYarkand 1, the qāḍī knew both parties in question by face, genealogy and name (bi-l-wajh wa-l-nasab wa-l-ism), yet a physical description is nevertheless provided of the claimant.Moreover, the names and genealogies of both claimant and defendant appear after al-musammā. 61fter the appearance of the parties in court, the claimant, Fāṭima, states her claim against the defendant, Mīr Khwāja Muḥammad.According to Ḥanafī shūrūṭ, this had to be expressed as follows: iddaʿa hādha l-ladhī ḥaḍara ʿalā hādha l-ladhī aḥḍara maʿahu anna (the one who appeared in court made the following claim against the one who he brought to court).Our record has a similar (but not identical) Perso-Arabic clause which retains the use of the Arabic demonstrative pronouns.Ḥanafī jurists considered it necessary for Arabic demonstrative pronouns to be used after the names of the claimant and defendant throughout the court record, irrespective of whether it was a maḥdar or a sijill. 62This was to prevent any future confusion regarding who the claimant and the defendant were.The court record was considered invalid without it.After describing their appearance in court, each subsequent mention refers to the claimant and defendant as mud-daʿiyya hādhihi (this female claimant) and muddaʿā ʿalayh hādha (this male defendant) respectively.
Once the claimant and defendant are present in court, the text switches to the first person, as the claimant states her claim against the defendant.After mentioning the disputed object (the measure of water), the claimant claims it as her property.This is followed by the demand for the restitution of the disputed object.The precise formula of the clause where the claimant claimed his right and demanded the restitution of the disputed object from the defendant was debated by Ḥanafī jurists.In general, this had to be as explicit as possible to prevent subsequent interpolation or misinterpretation. 63he transition to the defendant's reply ( jawāb) occurs after a question clause where the defendant is asked to respond to the claim of the claimant.In Yārkand I, this is concisely expressed in Arabic as wa-saʾla fa-suʾila: he (the claimant) asked (for the defendant to be questioned), so he was questioned. 64 questioned, the defendant rejects (inkār) the claim of the claimant, saying he legally bought the disputed object from the claimant's husband in the claimant's presence.The claimant, however, denies the truth of this counter claim made by the defendant.As a result, the defendant is asked to present witnesses in support of his counterclaim.15-22.Witness testimonies by the defendant's witnesses.The defendant brings two male witnesses to court.The clause where the defendant introduces his witnesses is recorded in the first person.It is not clear, however, whether these witnesses were professional court witnesses (ʿudūl) whose reliability (lit."justness") (ʿadāla) was already established or ordinary witnesses whose reliability had to be examined by the qāḍī.Ḥanafī shurūṭ recommended providing a detailed description of the witnesses, including their physical features, place of residence and mosque, presumably if they were not known to the qāḍī. 65Since there is no such description of the witnesses here, nor an account of the examination of their reliability, it is likely that their good character had already been established for the qāḍī. 66One of the witness's Arabic title raʾīs suggests, in the rural setting of the document, that he was a local landowner. 67The record now switches back to the third person, stating that after testimony was requested from the witnesses (baʿd az istishhād), both witnesses gave valid testimony (guwāhī-yi ṣaḥīḥ).In accordance with Ḥanafī shurūṭ stipulations, the text of this oral testimony (alfāẓ al-shahāda) is included in the court record. 68What is missing, however, is a clause confirming that the recorded testimony in Persian was also read out to and confirmed by the witnesses themselves. 69his procedure is described in some detail by Ibn Māza in his discussion of the maḥḍar.The qāḍī had to ensure the witness testimony was first recorded on a piece of paper (qiṭʿa al-qirṭās). 70A court official (ṣāḥib al-majlis) would then read the witness testimony in Persian to the witnesses.After this the qāḍī would ask the witnesses to confirm if they had heard the testimony that was read out to them from beginning to end by the court official and that they were witnesses to it.The witnesses would then respond in Persian testifying that they heard what was read to them from beginning to end by the court official and that they were witnesses to it.23-27.The qāḍīs assessment of evidence in the case and his judgement.As soon as the record of the witness testimony in Persian ends, the text of the court record returns to the third person, with a clause in Arabic on the consistency of the witness testimony and the fact that the witnesses pointed to the correct places in their testimony.This clause is almost identical to that which is found in the Ḥanafī shurūṭ literature. 71The meaning of pointing to the correct places was that the witnesses identified the claimant and the defendant correctly when referring to them in their testimony and, when referring to the disputed object, to the court record tawqīʿ at the beginning (of the sijill) and these four linesor five lines depending on what fitsin my own handwriting." 76The left-hand placement of the yaqūlu clause, as prescribed by Ibn Māza, is visible in the Khurāsān court record.This is not the case, however, in P.YarkandGronke 1 and P.YarkandGronke 2 where the yaqūlu clause appears directly below the last sentence of the text of the proceedings.In the yaqūlu clause in P. YarkandGronke 1 77 and P.YarkandGronke 2, 78 and in the example of Ibn Māza, the qāḍī confirms that the tawqīʿ and the yaqūlū clause are in his own handwriting, but the rest of the text was written by a scribe upon his order.In contrast, in our record the qāḍī confirms that the text of the proceedings itself was also in his handwriting.

Witness clauses
The Khurāsān court record has nine witness clauses: five appear on the recto and four on the verso of the document.The first part of each clause which refers to the witnessing is in Persian, while the second part which concerns the writing of the clause onto the document is in Arabic.Witness 1 uses the formula bi-ḥudūr-i man būd wa kataba fulān bi-khaṭṭihi: it (occurred) in my presence written by so and so in his own hand.The remaining witnesses use the formula ham bar-īn jumla būd (it was like this) followed by the name of the witness and either wa kataba [var.katabahu] bi-khaṭṭihi or bi-yaddihi (in his own hand) or bi-amrihi (upon his order by a scribe).An exception is Witness 4 who begins the clause with the formula ham bar-īn jumla guwāh-am (I am a witness to this).The ham bar-īn jumla būd witness clause is attested in pre-Mongol New Persian iqrār documents from Khurāsān and appears to be a twelfth-century development since the eleventh-century New Persian iqrārs from Bāmiyān use witness clauses entirely in Arabic, with the verb shahida. 79The type of witness clause beginning with ham is also not found in the New Persian legal documents from al-Ḥaram al-Sharīf or Ardabil, which suggests regional differences. 80As the handwriting of the clauses with bi-khattihi and bi-yaddihi appears to be different from those with bi-amrihi in the Khurāsān court record, it is likely that they are autograph witness clauses, while the remaining clauses were probably recorded by the qāḍī himself or by a different scribe.

Conclusion
The pre-Mongol thirteenth-century Khalili New Persian qāḍī court record from Khurāsān studied here uses a combination of New Persian and Arabic for different parts of the document.The qāḍī's signature (tawqīʿ), his decision (ḥukm) and his yaqūlu note of certification 76 MB-K, 118. 77P.GronkeYarkand 1, 480.The text of this yaqūlu clause is effaced in places and difficult to decipher.Gronke's reading of the yaqūlu clause is that the qāḍī wrote the record (wa l-sijill kutiba bi-aydayy).The earlier reading by Barthold, which she cites, suggests the qāḍī ordered a scribe to write the record (wa l-sijill kutiba bi-amrī).Barthold's reading is closer to the text of the proceedings, which suggests the qāḍī ordered a scribe to write the record (…wa amara bi-kitbati hādha l-dhikr…).
78 P.GronkeYarkand 2, 489.The text of the proceedings, however, suggests that the qāḍī wrote the record (…wa katabtu hādha l-dhikr li-yakūna ḥujjatan lahu…). 79See, for example, Ms. Heb. 8333.217=4. 80 Many of the fourteenth-century Ḥaram New Persian legal documents use witness clauses in Persian only: bi-guwāhī-yi fulān (witnessed by so and so): see, for example, document no.863 in Little (1984), facsimile 16.The Ardabil New Persian legal documents use both Arabic and Persian witness clauses in the same deed.See, for example, the sale deed (517/1123): Urkunde I, in Gronke (1982): 94-112.are in Arabic.The text of the proceedings is recorded in New Persian with Perso-Arabic clauses.The text shifts, however, to Arabic when recording the qāḍī's interventions in the proceedings.The witness clauses combine Persian and Arabic.The Perso-Arabic formulae used in the document was clearly based on earlier and contemporary Arabic recording norms. 81As we have seen, the text bears a close resemblance to the Arabic formulae and structure of the Yarkand court records produced further east in the twelfth century.In addition, the qāḍī-scribe of the Khalili New Persian qāḍī court record was familiar with the legal genre of shurūṭ.He carefully follows the prescriptions of Transoxanian Ḥanafī shurūṭ works of the twelfth to thirteenth centuries for writing such sijill certificates containing the qāḍī's decision.As shurūṭ stipulations differed in different parts of the Islamic world depending on the school of law -Shāfiʿī, Ḥanafī, Ḥanbalī, Māliki, Zaydī, Imāmī, etc.and the period in question, the Khalili New Persian qāḍī court record is significant as it shows us how local recording practice, in a rural mountainous area of Khurāsān, interacted with the theories of a particular school, here with Ḥanafī law, on the eve of the Mongol conquest of these lands.As comparable documents in New Persian from the fourteenth century onwards are studied from the Ardabil and Ḥaram al-Sharīf collections, future research might be able to shed light on continuities or ruptures in the Mongol period with the pre-Mongol Khurasanian practice of our document. 82
. He brought the noble raʾīs Khaṭīr al-Dīn Muḥammad b.Muḥammad b.Ḥusayn and Khwāja 18. Muḥammad b.Muḥammad well known as Rashīd.He (i.e. the defendant) said: occurred in my presence.Written by Muḥammad al-Ḥasan ʿAbd al-Warrāq in his own hand.2. It was like this.Written by Ḥusayn ʿAlī b. ʿAbd al-Khāliq upon his order.3. It was like this.Written by Bakhtiyār b.Ismāʿīl b.Muḥammad upon his order.4. I am a witness to this.Written by ʿAlī b.Aḥmad b.Muḥammad in his own hand. 5. Muḥammad b.Masʿūd b. ʿAlī wrote it in his own hand.Verso 6.It was like this.Written by Muḥammad Dihqān al-Ḥusayn upon his order.7. It was like this.Written by Aḥmad Ḥakīm Abū l-Faḍl upon his order.8.It was like this.Written by Ḥusayn Dihqān al-Ḥusayn upon his order.9.It was like this.Written by Rukn al-Dīn Rukn upon his order.
Wardak is the name a well-known province southeast of Bāmiyān and it might suggest the document originates from this area.If this toponym beginning with Ist is the same one mentioned in Khalili doc.37, line 3 and Khalili doc.38, line 6, then it is possibly situated in the region southwest of Bāmiyān near Estak/Istak.Another possible reading after the toponym is az dāng-i.The second alif of Ibrāhīm is generally omitted.Both forms of the relative pronoun kay and ki (line 8) are used in the document.5.The word directly preceding min qibal is uncertain.It is probably an honorific title related to the appointment of Muḥammad b.Ismāʿīl b.Ibrāhīm b. ʿUmar as qāḍī by The unit of measurement mentioned here is uncertain.Possible readings are tīr and sitīr.The same measure is mentioned in line 20 in Khalili doc.50 recto: dawāzdah-bāra az sih si/tīr yak si/tīr-wār az naṣīb-i amyāh.The reading ān qadr also appears in line 18 in Khalili doc.50 recto.11.Instead of samāʿat, tamāmat (meaning completion) is also plausible.
50, lines 2, 26 and 27.In the eleventh-century New Persian manuscript Codex Vindobonensis (447/1055-56), small yāʾs and hamzas already begin to indicate the New Persian iḍāfa.41Thealif madda is not indicated above the alif of āmad.The dagger alif next to the ḥāʾ of ḥāḍir is probably related to āmad and is used in place of the alif madda.