Charter of Pucangan (1041-11-6)

⟨A1⟩ nr̥ṇāv vidhāne • One expects nr̥ṇāṁ vidhāne. Of this idiosyncratic sandhi, of which several more examples will follow below, we have seen one other instance in the Campā inscription C. 167, st. IV (§1.3). See also Kern’s remark (1917:87). ALSO BAN NAXONE LAOS

⟨A2⟩ [A](p)[i] ⬦ (A)pi K • In fact only a part of the p is actually visible, but the shape of the lacuna certainly allows conjecturing Api. — ⟨A2⟩ yas sthāṇur ⬦ ya sthāṇur K. — ⟨A2⟩ atitarāy • Norm. atitarāṁ. For this idiosyncratic sandhi see our note on nr̥ṇāv vidhāne in pāda b of stanza I. — ⟨A2⟩ ya¿v?⟨th⟩epsitārtha- ⬦ yathepsitārtha- K • The akṣara in question quite clearly seems to be v and not th, but the two akṣaras are very similar and the latter was almost certainly intended, so we translate yathepsita-.

⟨A3⟩ °tayā ⬦ dhiyā K • The reading included in Crawfurd 1816 seems to be kayā at this place.

⟨A5⟩ lakṣmīś ca no gatvarī ⬦ lakṣmīn dadhau gatvarī K • We initially read dorgatvarī, because something that resembles a repha/layar seems discernible above the ga, but we are unable to come up with any justification for why the author would have invented a unique expression ‘going from the arms’ or ‘going from arm to arm’ to mean ‘fickle’. It is more likely that he used no, a metrically conditioned equivalent of the negation na (cf Raghuvaṁśa 11.60: diśo no babhūvur avalokanakṣamāḥ). For gatvarī as an attribute of śrī (meaning ‘riches’ in this case), see Kirātārjunīya 11.21: abhidroheṇa bhūtānām arjayan gatvarīḥ śriyaḥ ‘The man who amasses transient wealth by injuring creatures…’ (tr. Peterson). — ⟨A5⟩ sadāvavaddhā ⬦ sahābhavadyā K. — ⟨A5⟩ mudam K • The reading sudam in Kern 1917 is a misrepresentation of the original Devanāgarī edition.

⟨A7⟩ -vibhavo ⬦ -vi(śa)do K. — ⟨A7⟩ -dhiyām bhām bhāvayan K ⬦ -dhiyā kṣmām bhāvayan K. — ⟨A7⟩ Ābhiś K ⬦ śauriś K.

⟨A8⟩ śa(t)r(ū)n ⬦ śubhram K. — ⟨A8⟩ -rupā K ⬦ -rūpā K • It seems impossible to read here the metrically required long vowel ū. See another case of rūpa spelt rupa in line 5 of face B.

⟨A9⟩ rāja- ⬦ bhāva (?) K.

⟨A10⟩ mahendradattāv • Regarding the idiosyncratic sandhi, see our note on nr̥ṇāv vidhāne in pāda b of stanza I.

⟨A12⟩ -rāṣṭram iva – dya – ⏑ – ⬦ -rāṣṭram iva mudyutaṁ ciraṁ (?) K; -rāṣṭram avasādya taṁ ciram· P • We have considered several possibilities to complete this line but none of them seemed satisfactory, e.g. -m iva madyavādhitaṁ, ‘afflicted by intoxicating drinks’ (for another instance of writing v instead of b see stanza V. -valānvito, and stanza VI. sadāvavaddhā), -m iva māndyasāditaṁ, ‘destroyed by indolence’, -mavasādya pārthivaḥ, ‘having ruined it, the king’, etc. Arjunavivāha 1.3 might possible be related to this passage: (daitya Nivātakavaca) sumyūhaṅ indrālaya ‘(the demon Nivātakavaca) was about to destroy Indra’s abode’. — ⟨A12⟩ kalinā khalena ⬦ talinā ⏑ lena K; kalinā jhalena P • See Damais (1952: 90–91, n.1): “Nous dirons dans une étude consacrée à la stèle de Pucangan pourquoi il nous est impossible de lire avec Purbatjaraka un jha”. Alas, Damais never published this study. — ⟨A12⟩ vinā P ⬦ viś¡ā!⟨āṁ⟩ K. — ⟨A12⟩ narottamena sahito P ⬦ narottamair upahito K.

⟨A13⟩ śaśalāñchanā[bdhi]vadane ⬦ tha ⏑ locanāgnivadane K • The first syllable of line 13 is almost illegible in all sources available to us. For the middle chronogram word, the choice to read bdhi (4 oceans, abdhi) instead of gni (3 fires,agni) is made on the ground that the specified thirteenth tithi of the waxing fortnight of Māgha fell on a Monday (vāre śaśini) in 941 Śaka, while in 931 it did not; moreover, in line 15 on face B, the same episode is dated to 941. See on this verse and Damais’ reading of the chronogram n. 4 (p. 66) in Boechari’s article on the Mūla-Maluruṅ inscription = Boechari 2012: 430 n.4.

⟨A14⟩ (pā)hīty apātāṅ ⬦ pāhītyutāntāṅ K • Other possibilities we have taken into consideration include: pāhīty anantāṅ / adāntāṅ / aśāntāṅ kṣitim, pāyāt samantāṅ kṣitim, pāhi tvam asmatkṣitim, pāty adya śāntāṅ / dāntāṅ kṣitim. None seemed satisfactory. — ⟨A14⟩ samrājya- • the intended word is sāmrājya-. But it is imaginable that for our author, samrājya- was an acceptable form. — ⟨A14⟩ śaśvad ⬦ bibhyad K. — ⟨A14⟩ abhyasyatīha K ⬦ abhyasyatīva K. — ⟨A14⟩ (capa)latvam K ⬦ mukhalatvam K • Other conceivable readings would be kuśalatvam, mr̥dulatvam and musalavam. If we read capalatvam, it might be an allusion to the term hastacapala, found frequently in OJ inscriptions and sometimes in other textual sources. CHECK Boechari. Pāṇicapala is also found in dharmaśāstra texts, e.g. Manusmṛti 4.177, where it refers to “someone conducting himself in a fickle manner with his hands” (cf. O livelle’s translation p. CHECK). Here in the inscription, however, the meaning of this expression is probably more positive.

⟨A15⟩ pr̥thvīv • Regarding the idiosyncratic sandhi, see our note on nr̥ṇāv vidhāne in pāda b of stanza I. — ⟨A15⟩ na K ⬦ nu (?) K. — ⟨A15⟩ bubhujus ⬦ bubhujas Kern 1917 K • Damais writes an u above the a — ⟨A15⟩ taj(jā) K ⬦ ta – r K • This reading is proposed by K in the note to his translation, where he observes that an appropriate reading would be tuṣṭiṁ, but that it cannot be adopted because no u is visible. It indeed seems hard to read a tu; for the following akṣara, something with j or ṣṭ would be possible, but these are not the only possibilities judging by what little remains identifiable of the akṣara. — ⟨A15⟩ kintu K ⬦ tiktaṁ K. — ⟨A15⟩ dhirājāgraṇīr ⬦ – irājāgraṇīr K; dhi(nāthā)graṇīr K. — ⟨A15⟩ bhobhuṅkte ⬦ bho caṅkte / bhaṅktve K; bho aṅge K.

⟨16⟩ -digdha- ⬦ – kta- K; -(sa)kta K • Cf. K. 263 C, st. XV = K. 669 A, st. VI, pāda a: āsīd bhūpālamaulisphuritamaṇiśikhārāgadigdhāṅghrijaśrīr. For a slightly different expression of the same idea in Old Javanese, see face B lines 29–30.

⟨A16⟩ -laucana- ⬦ -locana- Kern 1917 K • The word -locana- is surely intended, but the stone shows lau instead of lo.

⟨A18⟩ to hy ājñāvidheyas ⬦ tāmpyājñavidheyas K; –ājñāvidheyas K. — ⟨A18⟩ -hr̥daya(ḥ stai)nyā ⬦ -hr̥dayastenyā Kern 1917 K • One of the problems with our reading is that this inscription elsewhere prefers the sandhi -s st-. — ⟨A18⟩ rājāpy upālabhyate ⬦ rājampyupālanyate K; – – ⏑ yā lapyate K • The possibility of reading as we do was proposed to Arlo Griffiths by the late Abhijit Ghosh, when they discussed the stanza in January 2011. The UI estampage seems indeed to show these akṣaras.

⟨A19⟩ – vāptaye dhanamalāni K ⬦ – – ptapo vanasamāni K. — ⟨A19⟩ tuṅgāsau ⬦ tuṅgāho K; tuṅgā K.

⟨A20⟩ sya maha – – – ⏑ – – ⏑so • Maybe it is possible to read ... jijñā ⏑ so? — ⟨A20⟩ cikīrṣayā kṣa • Maybe it is possible to read kṣi or kr̥? Or cikīrṣayārkṣi ...? — ⟨A20⟩ rabhasuyā • K states in a note that the reading is unmistakable to him, but that what is expected is rabhasayā.

⟨A21⟩ śuddhikarīndradā ⬦ śuddhikarī ⏑ dā K. — ⟨A21⟩ [na]reṣu ⬦ mareṣu K • The first syllable of this word is completely illegible on the UI estampage . — ⟨A21⟩ [mati]vāk- • K states in a note that he guesses the stone has pr̥thuvāk. Our conjecture is based on the assumption that we must have here an expression of the triple ’thoughts, words and deeds’. An alternative would be to conjecture [mata]. — ⟨A21⟩ [dha]rmmo ⬦ dhr̥ṣṭo K. — ⟨A21⟩ (pāśa)bhr̥d ⬦ (bhā)gakr̥d K. — ⟨A21⟩ hanta [nr̥pa]r(ā)d ⬦ hanta ⏑ rarāḍiti K • Kern notes that the syllable after hanta clearly bears a u, but supposes this has been misplaced, and believes the intended reading is hantu mararāḍ iti. To us, the akṣara supposedly bearing a u rather seems to be the one before r(ā). Instead of hanta, we could also read hasta. Understand -rāḍ.

⟨A22⟩ nr̥po ⏑ ⏑ purīpralayaṁ sa – – ⬦ nr̥ – ⏑ ⏑ ⏑ – prala – ⏑ – – K. — ⟨A22⟩ viṣṇu- ⬦ bhīṣma- K.

⟨A23⟩ tañ candrabhūtavadane śakarājavarṣa • this entire pāda, except the last three syllables, was unreadable for Kern. — ⟨A23⟩ ekādaśī ⏑ daka – cu ⏑ phalguṇe (v)ā(n)· ⬦ – – ⏑ – ⏑ ⏑ ⏑ – ⏑ ⏑ – ⏑ lokāt K. — ⟨A23⟩ adhamaḥ parudābh- ⬦ adhamāpanudābh- K • Would parujābh° (including OJ word parujar) somehow be possible? — ⟨A23⟩ Ivāvyathayaj jaganti ⬦ vādhyadhamāṅgatantiḥ K.

⟨A24⟩ narendro ⬦ nagendre K. — ⟨A24⟩ (rā)mātviramyaca(ri)to nyavadhīt tam āśu • K only read from nyavadhīt onwards. Emend rāmāti-. Or is it possible that what looks like subscript -v- is actually an accidental scratch in the stone? — ⟨A24⟩ jigīṣu(r ga)tas ⬦ jigīṣur bhuvās K.

⟨A25⟩ (varo) ⬦ varo K • It would also be possible to read gate or gato.

⟨A26⟩ e[tān tatkṣa]ṇaṅ ⬦ etal lakṣaṇaṅ K • As is clear from the Leiden estampage, Kern’s reading must already have been partly conjectural. His interpretation of lakṣaṇaṅ as meaning ‘prize’ seems impossible, for one would then have expected lakṣyaṅ or lakṣaṅ. — ⟨A26⟩ a(ja)yyān ⬦ anāryyān K • Neither ā-mātra on the second syllable nor repha above y are discernible.

⟨A27⟩ labdhvā ⬦ luṇṭaṁ K. — ⟨A27⟩ -saṅghe ⬦ -madhye K. — ⟨A27⟩ bhadre • Kern notes that bhadre is to be corrected to bhādre, but this is probably unnecessary. Damais reads caitre, but the estampages seem to favor Kern’s reading; neither reading yields a Wednesday.

⟨A28⟩ -titho ⬦ -tithau K • Of course one must understand -tithau.

⟨A29⟩ [A]tha mu(kha)- ⬦ (atha) muni- K • What has been read as mu looks more like pr̥ or possibly pu. Virtually no trace of the last syllable in this line is actually preserved. — ⟨A29⟩ vijayavarm(m)ā ⬦ vijayanāmā K.

⟨A30⟩ m(ukha)- ⬦ maka- K • See Kern 1917: 96 n. 4 for details. As Kern observed, the passage in line 29–30 of face B makes clear that the year in question was 959, and hence that the word we are dealing with here had to have the value 9.Judging by the visual material at our disposal, we doubt that the reading maka- was ever as “perfectly clear (volkomen duidelijk)” as Kern claimed. We find it likely that the same word with value 9 was used in stanza XXIX and XXX.

⟨A31⟩ nihitapādo • it seems that the e-mātrā of do stands at the end of line 30, but if this is correct, we probably have to assume that it was repeated by the stone carver, or that some other error was made at the beginning of l. 31, for there are traces of more signs than dā before ratna-. The hypothesis of redundant e-mātrā is to be preferred, for there is an analogous case in face B, lines 14–15. — ⟨A31⟩ -pīḍaṅ • We cannot distinguish the ḍ with certainty from a d. In all likelihood, the stone carver spelled -pīdaṅ.

⟨A32⟩ bratitayā K ⬦ vratitayā K. — ⟨A32⟩ devatārādhanaiḥ rantuñ ⬦ devatārādhanairantru/añ K; devatā rādhanai rantuñ K • A pausa form is expected at the end of pāda b. The Leiden estampage seems to show the expected visarga. — ⟨A32⟩ -mahā(kra/śra)mas ⬦ -mahā¿d?⟨t⟩vam¿u?⟨a⟩s K; -mahānr̥pas K • The second ma seems relatively clear.

⟨A34⟩ -netr¿ā?⟨āḥ⟩ • This was no doubt intended by Kern’s netrā(ḥ). — ⟨A34⟩ mālābhiprītikārās ⬦ [6+]rās K; mālādiprītikārās K. — ⟨A34⟩ enan K ⬦ etan K.

⟨A35⟩ na me K ⬦ nr̥pe K.

⟨A36⟩ jīvyād • the supposed ī-sign is not clearly distinguishable from a short i.

⟨B1⟩ va, śu K ⬦ pa, bu B K; va, śa P. — ⟨B1⟩ caragraha • This term is anomalous. Elsewhere, one finds grahacāra. See Gomperts 2001: 107–110. — ⟨B1⟩ -Uttarabhadravā⟨da⟩-nakṣatra K ⬦ -rabhadravānākṣatra K; -ra( )dratanakṣatra B; -rabhadrava(da) nakṣatra P. — ⟨B1⟩ -budhna- Kern 1917 K ⬦ -budha- B P.

⟨B2⟩ gara(ḍ)i-: K • We tentatively follow D’s suggestion for the name of the karaṇa; dadhi- P. B and K leave the syllables in question open. — ⟨B2⟩ rakai P K ⬦ rake B K.

⟨B3⟩ dhamasuparṇa(vā)haṇa tguh uttuṅga- ⬦ dhamma suparṇa – hana tahutuṅgadeva B; dhamma suparṇa – haṇata hutuṅgadeva K; dharmmasuparṇacaraṇa tguh uttuṅga- P; dh(ā)ma suparṇavāhaṇa tguh uttuṅga- K • On the reading of the first element of the name, see Damais 1955: 66–7 n. 3. We reject Damais’ conclusion that the reading must be dhāma, because we see no trace in the RTI of the -ā suspected by Damais at the end of line 2. Although we thus read dhama instead of dhāma, on the grounds that the intended meaning of the name may have been ‘Viṣṇu (suparṇavāhana) in heaven (dhāma)’, we suppose that dhāma may indeed have been intended. — ⟨B3⟩ dānaśūra K ⬦ janaśūra P • B and K leave these syllables blank. — ⟨B3⟩ Iṁ barahəm· • After dharmmamūrtti B leaves the syllables blank until niṅ barahəm·. K has dharmamūrti [11+] kə [5+] maṅan, iṅ barahəm. — ⟨B3⟩ bapuri K ⬦ śapuri P; basuri B K.

⟨B5⟩ 939 P ⬦ 938 (928) B; 928 K. — ⟨B5⟩ ri pra(hāra) ⬦ ri prahara P; mra?– – K • B leaves these syllables blank. — ⟨B5⟩ An· vijil· B K ⬦ masə: mijil P. — ⟨B5⟩ -rŭpa- ⬦ -rūpa- B K P.

⟨B7⟩ vālaka P ⬦ vālajaka B; vāla, ka K. — ⟨B7⟩ dahat· K P ⬦ dahat iṁ B. — ⟨B7⟩ enāk B ⬦ enak· K P. — ⟨B7⟩ baṅətni P ⬦ baṅəṅgi B K. — ⟨B7⟩ rumə⟨B8⟩gəp· B K P • One detects what seems to be a character to the right of mə, which could theoretically be a vestige of an akṣara rə or R̥, or else be a space filler such as one often finds in later inscriptions but never in original stone inscriptions from Airlangga’s time. Although a form rumərəgəp could theoretically be justified, it is unattested. We hence believe that we must be dealing with an accidental scratch.

⟨B8⟩ pasarikəpany ⬦ pasariṅkəpany B P; sapariṅkəpany K. — ⟨B8⟩ sākṣāt(n)iran ⬦ sākṣāt iran B K; sakṣātiran P. — ⟨B8⟩ Inahākən B K ⬦ Innahakən P. — ⟨B8⟩ kavaśa K P ⬦ livaśa B. — ⟨B8⟩ paṅava(śaniṁ) P ⬦ paṅavananiṁ B; paṅavara niṁ K. — ⟨B8⟩ maṅantīri ⬦ maṅanti ri B K P. — ⟨B8⟩ -niṁ vanagiri B ⬦ -ni vanagiri K; -iṅ vanagiri P.

⟨B9⟩ lāvan· K P ⬦ lavan B. — ⟨B9⟩ saṅ e(kā)ntapratipatti ⬦ [4+]nta pradipatti B; [4+]nta pratipatti K; samekāntapratipatti P. — ⟨B9⟩ (dāśabhū)ta-tanŭpacāran· bhaktiprahva ri ⬦ [3+] tanu – – n – – pra – ri B; – tanu [4+] n – – prahva ri K; Umaṅga tatan upakaran bhakti sraddha ri P • For -tanupacāran, one must either understand -tanūpacāran, or emend -tanupracāran. We tentatively opt for the former. The whole passage from ekāntapratipatti to pāduka śrī mahārāja is reminiscent of the Adulengen charter, where we read in 2r1–3: tan· kapālaṅalaṁ suṣṭubhakti I pāduka śrī mahārāja, lot kahudanan·, kapyayān·, An paṅekānta I pāduka śrī mahārāja “they were undisturbed in their extreme loyalty to his Majesty the Great King, steadfast [despite being] exposed to rain and sun, as they had but one aim, [namely] his Majesty the Great King”. — ⟨B9⟩ pu narottama K ⬦ saṁ narottama B; mpu narottama P.

⟨B10⟩ sirādini K ⬦ sirādiniṁ B P. — ⟨B10⟩ Ati¡sayeṁ!⟨śayeṅ⟩ K ⬦ Atiśayeṁ B P. — ⟨B10⟩ tumūt P ⬦ hambak B; humbat (?) K. — ⟨B10⟩ tan sāḥ ⬦ tansaḥ B K; tan sah P. — ⟨B10⟩ saparān· ⬦ Ayunan B K; saparan P. — ⟨B10⟩ makāhāra P ⬦ I[.]ā hāra B; maṅāhāra K.

⟨B11⟩ vanaprastha B P ⬦ vanapraṣṭha K. — ⟨B11⟩ tātan· vismr̥¿d?⟨t⟩i ⬦ yataniviṣadi B K; tatan vismr̥ti P. — ⟨B11⟩ ri kārādhanan· ⬦ rikā bhavanan· B K; ri kāradhanan· P • The Sanskrit text on face A also uses the word ārādhana: see st. XXXII. — ⟨B11⟩ ahorātra K ⬦ ahoratra B P.

⟨B12⟩ (ka)lpapādapa Umə:bana ⬦ Alpapādi padamə:lana B; –lpapādi padamə:lana K; kalpapādapa ṅahə:bana P. — ⟨B12⟩ kumaliliranaṁ P ⬦ kumalilirana B K. — ⟨B12⟩ makadrabyaṁ P ⬦ makadrabya B K. — ⟨B12⟩ muvahakna sāśvatā(niṁ) ⬦ muvaḥakna haṣa nikanaṅ B; muvahakna harṣanikanaṅ K P • The RTI clearly shows sā instead of ha. Cf. the parallel passage in line 33: de śrī mahārāja munarjīvākən· sāśvatāniṁ sayavadvīpa. The word recorded in OJED as śāśvata is intended in both passages. — ⟨B12⟩ munarjīvāknaṁ ⬦ munarjīvākna P; [5×] B K.

⟨B13⟩ maṁkanābhamataniṁ ⬦ maṁkanābhimataniṁ B K P • The reading of the previous editors suggests that an -i was visible to them in the reproductions they used. It seems rather that none was ever engraved atop bh, and that it needs to be inserted by emendation. Another instance of the same scribal error occurs below in l. 16. — ⟨B13⟩ -devata K P ⬦ -devatā B. — ⟨B13⟩ kapadasthāniṁ pituṁ ⬦ ka [2+] gə:ṅniṅ pitu B; kapūja(?) [2+] pitu K; kapadasthaniṁ pitu P.

⟨B14⟩ haji devatā K P ⬦ haridevatā B. — ⟨B14⟩ lumāḥ ⬦ lumāh B K; lumah P. — ⟨B14⟩ īśānabajra K P ⬦ iśānabajra B. — ⟨B14⟩ pinaka(ka)padasthān· ⬦ pinakakapratisthān· B K; pinakakapratiṣṭhān· P.

⟨B15⟩ -vikramottuṅgadeva • the e-element (taling) of mo is written first at the end of line 14, and then redundantly repeated at the beginning of line 15. See identical case on face A, in st. XXX. — ⟨B15⟩ mahārāja, de B K ⬦ mahārāja de P. — ⟨B15⟩ Irikāṁ ⬦ irikaṅ B K P. — ⟨B15⟩ (tā)tan· pahī(ṅan de) śrī ⬦ tatan pahīṅan śrī P; I tan pahiṅan· śrī B; – tan pahīṅan· śrī K • All previous editors ignore the presence of de.

⟨B16⟩ [luma]risakən· ⬦ [4+]kən B K; manarīrākən P. — ⟨B16⟩ sābhamataniṁ • K and P have proposed the emendation sābhimataniṁ which underlies our translation. See the emendation maṁkanābhamata to maṁkanābhimata in l. 13. It is very curious that the same error is made twice; normally, this would impose taking the reading seriously as it is, but we see no other way than to assume that the word abhimata was really intended in both cases. — ⟨B16⟩ kapva kakal(imbaṅ ikaṁ) kriyā vāhyāntara ⬦ kapva kalimba – – kri phala – hyantara B; kapva kakal imban (?) iṁ kriyā vāhyāntara K; kapva kakalimbaṅ ikaṅ kriya vahyāntara P • Several factors conspire to make the reading rather uncertain here. First, there is interference with the descender (itself rather unclear) of ṅku in the preceding line; what previous scholars have read, and what we read after them, as mb could possibly be ṅk or jb; the squiggle to its top left, which might be taken as -ā on mb pushed to the left by the descender from line 15, seems indeed to be an -i on the preceding akṣara, as previous editors assumed, and as we find confirmed by an apparently parallel passage in the Kamalagyan inscription, lines 21–22: yāvat kavaṅunaniṅ yaśa donanya, an kapva kinalimbaṅ juga denira, sahana saṅ hyaṅ sarvvadharmma kabeḥ ‘As many constructions of foundation as were his aim, they were all simply carried out by him, absolutely all holy foundations’. The li seems to be endowed with descender to make -u (lu), but this must be an accidental scratch. — ⟨B16⟩ denira, tan· kasalimur i kadamlaniṁ pūjā P ⬦ denira tan· kapalimura I dharmma dhi – jā B; denira tansalimur i kadharmmaniṁ prajā K.

⟨B17⟩ [3+] prāyaścitta [6+] (ma)hābhāraniṁ kasiṅhe(śva)ranira ⬦ [9+] jro mrā nikaṅ siṅho [2+]nira B; [2+] mahārāja [1+] nika siṅho [2+] nira K; riṁ bhaṭara, deni kasiṅheśvaranira P. — ⟨B17⟩ marava(śā)k(ə)n· P ⬦ mara [3+] ni B; mara [5+] ni K. — ⟨B17⟩ sakveḥnik(āṁ ka)la(ṅka) satumuvuh i ⬦ sakveḥnikāṅ [2+] maṅhyaṅgadrabya B; sakveḥ nikāṅ [2+] K; sakveḥnikaṅ kalaṅka satumuvuh P • P’s reading breaks off after this point. — ⟨B17⟩ [2+] hyaṁ • for the gap of two akṣaras, satva or mata would seem to be possible readings judging from the UI estampage and the RTI, but the resulting meaning in either case is not sufficiently persuasive to allow us to make a choice. — ⟨B17⟩ sap(i)nakahani(t)uniṁ ⬦ sapinakahanituni B K. — ⟨B17⟩ pra(bheda) [4+] dru(ha) [6+] ⬦ prabheda [12+] B; prabheda [5+] K.

⟨B18⟩ ṅ inaranan· B ⬦ ṅinarani K • A possible restitution of the illegible akṣaras at the end of line 17 would be salvirani, to give the words salviraniṅ inaranan. — ⟨B18⟩ si (śuvuka)l·, (man)tu (ha)j(ī)nadgakn (iṁ) ka(ṭ)ula ⬦ [10+] nakekale B; [4+] kekale K. — ⟨B18⟩ si la[ḍ]iṁ ⬦ si baḍiṁ B K • Brandes’ note indicates his hesitation between sibaḍing and silaḍing. — ⟨B18⟩ nirava(ś)eṣa K ⬦ naravaśeṣa B. — ⟨B18⟩ I [1+] (tta) s(ama) [2+] (n)ya ⬦ ikā [6+] B; ikā [2+] K. — ⟨B18⟩ de śrī mahārāja, ṅuni-ṅuni haji tṅaḥ [7+] sapuluḥ ⬦ de śrī mahā [16+] B; de śrī mahārāja ṅuni ṅuni haji [3+] K.

⟨B19⟩ paṁ(di)ri, pinakarovaṁ mad(v)andva yuddha ca(kra)[10+](vv)at· pinja(han iṁ) ⬦ [15+]rika [5+]nasikanaṅ sarat pinjahaniṅ B; [2+]diri (?) [5+]rika [6+] ikanaṅ sarat pinjahan, iṅ K. — ⟨B19⟩ vaṅuntur· mvaṁ saṅ ibu Irikāṁ ⬦ [3+] mvaṅ – – – irikāṅ B; [3+] mvaṅ irikāṅ K. — ⟨B19⟩ 95(1) [5+] ⬦ 951 ri kanaṁ pitu B K • Kern translates “in the seventh month”. This corresponds to nothing in stanza XXIII (Phalguṇa is not the 7th month by any reckoning). The akṣaras in question can but by no means need to be read as B and K did.

⟨B20⟩ ikāṁ parad[āra]putra hyaṁ ma(ṅ)adg (i v)uru tum(u)ṁgal (i) maddhyadeśaniṁ ⬦ -ikaṅ parada [7+] guru tumaṅgal caddhya deśani B; -ikāṅ para [5+] guru tumaṅgal – ddhya deśa ni K • We tentatively take paradāraputra as a tatpuruṣa compound, although a dvandva or sequence of two coordinated words is also a possibility (cf. tanayadāra in l. 28). We consider that a toponym is required after hyaṁ, and since we are reminded of the well-attested Vuru Tuṅgal (Damais 1970: 584–4), we read (v)uru tum(u)ṁgal where both B and K read guru tumaṅgal: the first consonant is extremely damaged, and can easily be read v or g; a faint trace of -u seems visible on the RTI below the m, but it must be admitted that the prima facie reading is tumaṁgal. — ⟨B20⟩ vurahan· ⬦ –ratan· B; vu?ratan· K • See Dharma Pātañjala 314:4 (ed. Acri 2017), for an occurrence of vurahan, taken as ‘confusion’ by the editor but potentially a toponym, indicating the opposite of a deśa rahayu. — ⟨B20⟩ -pa(rākrama) sumorakən haji vəṅkər· • B and K left these syllables blank.

⟨B21⟩ saṅa ta ha(jy a)n· ⬦ sa[4+]n B; samāhvayan K • A problem with our tentative reading is that we don’t really expect sandhi to be applied to haji An. — ⟨B21⟩ vaṅk(ə)r· ⬦ vaṅkir· B; vəṅkər· K • B’s note “vaṅkir-vəṅkər” suggests the emendation he had in mind, which was adopted by K into his text as such. It is possible to read i or ə on top of the ṅk, but there really is no vocalization sign on the v. In a damaged part of line 23, it seems the text may have had the spelling vaṅkar. — ⟨B21⟩ kavnaṅ ata ⬦ kavada ta B; kavaśa ta K. — ⟨B21⟩ mvaṁ tinuluy (hani)hanitu[n]i [8+] ⬦ maṁkinakuyanahani [10+]ḥ B; maṅkin[2+] hanitu [2+]ḥ K • The ostensible conjunct yh is hard to recognize and perhaps not so likely. If, as seems nevertheless quite likely, the passage is parallel to the one in line 26, we can fill in the first part of the gap with (hani)hanitu[n]iṁ yadadvīpa, leaving about 4 akṣaras unaccounted for.

⟨B22⟩ (m)uvah akaḍatvan i L̥ca, ⬦ ṅu – ha kaḍatvaniL̥va B; mulih(e) kaḍatvan i – vā K. — ⟨B22⟩ tinūtnikāṁ ⬦ bunutnikāṁ B; tinūt? nikāń K. — ⟨B22⟩ sināhasan ⬦ sināhsan B; sināhāsan K.

⟨B23⟩ [d]e (śrī) mahārāja ⬦ [2+] śrī mahārāja B; [3+] mahārāja K. — ⟨B23⟩ muvaḥ hana ta putra [haji] vaṅ[ka]r· maṅadg i vur(u rt)u p(j)aḥ ⬦ [19+]ḥ B; [4+]ḥ K. — ⟨B23⟩ vargga la(ṁ)ḍiṁ rāk· ⬦ va[1+] laṁ[3+] B; vargga la[3+] K. — ⟨B23⟩ hiR̥ṁhiR̥ṅan(·), Iṁ ⬦ haR̥pi[3+], bu B; haR̥p[5+], [1+] K.

⟨B24⟩ maṁdo(n)· lāvan· ⬦ b[5+]n· B; – – – – lāvan· K. — ⟨B24⟩ sahananikāṁ makira(k)i(rāpā)yāmutra hyaṅ amana(ḥnya) [8+] ⬦ sahananikaṅ – – – – – – – – – – – – – – – – – – – – – – B; sahana nikaṅ [8+] K • Two epigraphic passages, both contemporary with our inscription, contain the sequence amutra hyaṅ. See Turun Hyang, face A, line 6: ... nikanaṅ amutra hyaṅ kāvaḥ dūdu kika.... maṅabhayā manambu sarāt nakavāmaṇa rara bilṣani pāduka śrī mahā ...; plate MNI E.91, verso, lines 6–7: an tan dumaṇāta sira sumaṅgaha ikaṅ vaṅ sahayaniṅ amutra hyaṅ mvaṅ katibān tulis kaṅāśvāsa. Alas, the published readings of these two inscriptions are insecure, and we are unable to translate them.

⟨B25⟩ makapuruṣa(kā)ra ⬦ makapuruṣa – – B; mako (?) puruṣa – – K. — ⟨B25⟩ (U)saṁ[ṅ-u]saṁ ⬦ [3+]saṅ B; [3+] K • The initial akṣara can easily be read as U, but could also be a vocalization -e. The entirely lost akṣara, indicated here as ṅu, may equally well have been another U. Whatever its precise spelling, we are unsure that the intended word here was usaṅ-usaṅ to begin with. If it was, then we guess that it is a so far unattested form of the word recorded as usəṅ in OJED. See our note on p.XX below.

⟨B26⟩ hane magəṁ ha – – ⬦ haneṁ matiha – – B; haneṁ magəhan (?) – – K • The syllables tentatively represented here as magəṁ seem rather clear, but they are somewhat difficult to make sense of, even if we presume that they are the first element of a toponym. — ⟨B26⟩ h(ī)ṅa(nyan·) śrī mahārāja [sa](ṅka)ni ⬦ hiṅanya – – mahārāja maṅkana B; hīṅan ya – mahārāja, saṅkani K • We feel quite uncertain about the readings hīṅanyan· and saṅkani, but believe the punctuation sign read by Kern cannot have been present if the text really contained both these words. However, there does seem to be enough space for such a punctuation sign or even for a whole extra akṣara between the ja and the akṣara restored as sa. — ⟨B26⟩ hilaṁniṁ sahanani B ⬦ hilaṁ sahanani K. — ⟨B26⟩ hanihanituniṁ ⬦ hanihanituni B K. — ⟨B26⟩ kramani kahilaṁ K ⬦ kramanika hilaṁ B. — ⟨B26⟩ [mū]la kaḍatvani(ra) ⬦ Iṅakaḍatvanira B; – sva?kaḍatvanira K • Another possible reconstruction for the first word, whose initial syllable is lost, would be [kā]la. But since kāla is not used as conjunction elsewhere in this text, nor even recorded in that function in OJED, we tentatively prefer mūla.

⟨B27⟩ kapaṁ sira naṁ ⬦ tapa sira B K. — ⟨B27⟩ pratiniyata ⬦ pratinayaka B K. — ⟨B27⟩ hinārohara deśanirāṅkən· K ⬦ hina – ro – – deśani raṅkən B. — ⟨B27⟩ [vvay a]ta, • the punctuation sign after ta was ignored by B and K; in the light of its undeniable presence, and of the likely presence of ta in the next sentence, B and K’s reading vvaya ta in an area where the surface of the stone is now obliterated, is very suspicious. For in all occurrences known to us (see OJED s.v. wwaya), the sequence vvaya ta stands at the beginning of a sentence. For our tentative interpretation see note 257.

⟨B28⟩ (ta) sirar kaparājaya ⬦ ta sira – – rājaya B; ta sira [3+] (pa)rājaya K. — ⟨B28⟩ kapaṁ ⬦ tapa B K. — ⟨B28⟩ deśa K ⬦ de– B. — ⟨B28⟩ rājavāhana B ⬦ rājavahana K.

⟨B29⟩ vaR̥gg anusup· ⬦ – rəp manusup B K. — ⟨B29⟩ kavnaṁ ta sira B ⬦ kavnaṁ sira K.

⟨B30⟩ ha – – – pinakatapakan ⬦ ha – – s panaka ta pa kan B; – – – pinaka ta pakan K. — ⟨B30⟩ (t)aṇḍasnira ⬦ kaṇḍa sira B K. — ⟨B30⟩ Ar paliṅ(g)iḥ ⬦ Apaliṅgiḥ B K • The (g) in (ṅgi) has wrongly been carved as a full circle. — ⟨B30⟩ saṁkṣipta ta Ikāṁ ⬦ saṁkṣipta Ikāṁ B K • Both previous editors have omitted the particle ta.

⟨B31⟩ tan (hana) ⬦ nhana B; (ta)n hana K. — ⟨B31⟩ rāt· An· (p)aṅəb· makapārāyana (c)chā[yā]ni ⬦ [9+] pā – yanacchāyāni B; [5+] maka – – yan acchāyā ni K. — ⟨B31⟩ mataṅ yar siddh[ākən][·] ⬦ mataṅya siddhākən B K. — ⟨B31⟩ [pratijñānira] K ⬦ prājīnānira B.

⟨B32⟩ Inusa[n·-U]san· raghu, yāṅkən· ⬦ [4+]san· rake yāṅkən· B; Ista – – s an· rake yāṅkən· K • Both previous editors ignore the punctuation sign after raghu, which they misread as rake. Our restoration Inusa[n·-U]san (there seems to be too much space for restorting Inusa[n·-U]san) is based on the assumption that we have a typical Old Javanese-Sanskrit dyad and that usan-usan is a variant spelling for OJED’s usən-usən. — ⟨B32⟩ mantras(t)avanamaskāra K ⬦ mantrasta – namastāra B • Brandes’ note “( )namastāra-vana sangskāra” reveals that he had really not understood the present passage. — ⟨B32⟩ śrī mahārāja ri bhaṭāra B ⬦ śrī bhaṭāra K. — ⟨B32⟩ sārisāri ⬦ hari sāri B K. — ⟨B32⟩ paṁ[liṅgānanikāṁ] ⬦ paliṅgānanikāṁ B; paliṅgānanikaṁ K • Both previous editors ignored the first anusvāra, which is crucial to get the correct sense. We are dealing with an irrealis of the word paṅliṅgan. Cf. Sumanasāntaka 37.3c: paṅliṅgāna savaṅsavaṅni hayunīdəmanira səmuniṅ varāpsarī ‘This was clear because the young girls, whose beauty was on the point of blossoming, resembled divine nymphs’ (ed. and transl. Worsley et al. 2013).

⟨B33⟩ anāgataprabhu B ⬦ aṅāgataprabhu K. — ⟨B33⟩ hlam i ⬦ tkā ri B K. — ⟨B33⟩ kramani de śrī K ⬦ kramani – śrī B. — ⟨B33⟩ munarjīvākən· ⬦ munajikakən· B K • Cf. line 12.

⟨B34⟩ yavadvīpa, ka [3+] sa(dbhāva) [13+] han denira ⬦ yavadvīpa [23+]nenira B; yavadvīpa [8+]nira K. — ⟨B34⟩ t(kā)maṅgiḥ B ⬦ tkā, maṅgiḥ K. — ⟨B34⟩ hayunira • B leaves open an amount of space between hayu and nira, implying that Brandes thought there were some illegible akṣaras in the gap. But K indicates no such gap.

⟨B35⟩ [abhimata] B K • the reading has become unverifiable, because the surface of the stone has been lost since their time. In view of the fact that the text twice showed abhamata instead of expected abhimata (lines 13 and 16), one may wonder if that is what was actually engraved here too.

⟨B36⟩ (th)ā(n)i [vat]ə[k] ⬦ nāni – ti – B; – – ni – – – K • Our reconstruction is based on the occurrence of the sequence thāni vatək in other inscriptions of Airlaṅga, e.g. Terep I XX. — ⟨B36⟩ [6+] (maṅaran i) h(i)no(,) – (dr̥bya) ⬦ [12+]ti hino – madrabya B; [5+] hino madrabya K. — ⟨B36⟩ ni • This akṣara, rather clear in the RTI, was not read by B and K, whose reading resumes with nikāṁ at the end of the line, where the surface of the stone has by now entirely peeled off.

⟨B37⟩ mantən· ta B ⬦ mantən· tā K. — ⟨B37⟩ [bapuri] ⬦ basuri B K • Even though these akṣaras have peeled off by now, we emend the reading of our predecessor by comparison with our reading of the same toponym in line 3.

⟨B38⟩ ya [10+]· saṁ • B and K were unable to read any akṣaras in the damaged stretch before saṁ. — ⟨B38⟩ [[tan katamāna deniṁ]] • The restoration was made by Kern (1917: 106: n. 1), while Brandes left open a gap for the illegibly damaged akṣaras.

⟨B39⟩ mvaṁ (saṁ nāyaka)pratya[ya] ⬦ mvaṁ – nāyaka, partyaya B; mvaṁ nāyaka, partyaya K • The two words nāyaka and pratyaya are normally not separated by a punctuation sign (although see n. XX below), and are quite often preceded by saṁ, which we can faintly make out in the gap left open by Brandes that was ignored by Kern.

⟨B40⟩ m – – – – [sukhaduḥkha,] ⬦ – khaduḥkha, B; sukhaduḥkha, K • In B’s reading, line 39 ends with a substantial lacuna after haji m, while 40 starts with khaduḥkha; in K’s, there is a lacuna of five akṣaras after haji m at the end of 39, while line 40 starts with sukhaduḥkha. In cases where it could be verified, we have seen that K’s edition is not very precise in dealing with the distinction between preserved and conjectured text; moreover, the amount of space before the first legible akṣaras on line 40 suggests that B’s reading of the beginning of this line has a more plausible number of akṣaras. We therefore conjecture that the necessary akṣara su stood at the end of line 39. If we may otherwise rely on K’s edition, it means the lacuna between m and su extends over four akṣaras. If this estimation is exaggerated, then we could conjecture a reading like haji mvaṁ sukhaduḥkha, inspired by OJO XLIX 2b.

⟨B41⟩ […] [[saṁ maṅilāla] drabya haji vuluvulu mvaṁ] […] ⬦ […] [drabya haji vuluvulu mvaṁ] B; [8+] [[saṁ maṅilāla] drabya haji vuluvulu mvaṁ] [3+] K • We are unable to confirm any of the sequences of akṣaras that our predecessors have discerned on this line. We see a fairly clear ṅi around the 12th position in the line, but what remains of the surrounding akṣaras cannot be matched with Kern’s conjecture maṅilāla; three positions to the left of this ṅi stands a clear na which may also have been a ne. The number of illegible akṣaras indicated by Kern implies that there were only 25 akṣaras on this line which must be far off the mark as completely preserved lines higher up show around 55 akṣaras per line. But we are unable to propose an alternative estimate because we are unable to situate the string of akṣaras read by B and K anywhere on the line. Note that B’s text suggests mvaṁ stood at the end of the line, whereas K’s suggests a short lacuna stood between this word and the end of the line.

⟨B42⟩ [siṅ ata lviranya] ⬦ siṅa – lviranya B; siṅa ta lviranya K • K’s translation suggests that he was thinking of a word like asiṅ, an idea that receives some support from the Cane inscription, faces C and d, line 30: kita kamuṅ hyaṅ yāvat ya hana umulahulaḥ ikeṅ sīma ri cane asiṅ lvirani kavvaṅanya jaḥ tasmat kabvat kārmmāknanya ... But if this passage is indeed a relevant comparison, it also implies that Kern’s word division is unlikely to be correct. This is why we tentatively divide the words as asiṅ ata. — ⟨B42⟩ [luməbu][ra] ⬦ luməbu B K • It is tempting to conjecture further luməburāta or luməbura ta, and to assume no further loss of akṣaras before saṁ hyaṁ, but this issue depends on whether K’s estimate of 3 lost akṣaras between luməbu and saṁ hyaṁ is reliable or not. — ⟨B42⟩ [nāyaka, partyaya,] B K • We are unable to verify any part of this sequence, but expect that the stone rather showed pratyaya before it got damaged, and that there was no punctuation sign separating this from nāyaka. See n. XX above.

⟨B43⟩ va[huta rāma] [14+] [umulahu]laha • Almost all of this line is now lost. While B, as usual, only leaves open a large gap, K estimates the lacuna between rāma and umulahulaha to extend over 7 akṣaras. Taking the expected entire number of akṣaras for this line into account, and estimating the distance between the akṣaras that are still preserved today, we estimate that the gap which was already present in our predecessors’ time extended over at least 14 akṣaras.

⟨B44⟩ –nanira ⬦ [3+]nira B K • It might be possible to read the first akṣaras as sthā, conjecturing sthānanira. But ṅhananira or ṅkānanira also seem to be possible readings.

⟨B45⟩ [bhukt]inira [pañca]mahāpātaka ⬦ caṇḍinira [2+]mahāpātaka B K • The word caṇḍi in our predecessors’ reading is so unlikely that we feel forced to reject it, and the word pañca cries out so hard to be read before mahāpātaka that we feel obliged to conjecture it. Although we have found no exact parallel for the wording we propose here, assuming that nira was correctly read by B and K, our conjecture is based among numerous other comparable passages on Baru (Cd, lines 39–40), Cane (C, 26) and Patakan (C, 26), where we read: salvirniṅ pañcamahāpātaka bhuktinya. — ⟨B45⟩ [mahārāja] B ⬦ mahāraja K. — ⟨B45⟩ […] [citralekha I pāduka śrī mahārāja] […] [sira] […] • K estimates a gap of 17 akṣaras between pātaka and citralekha, then gaps of 3 akṣaras each after mahārāja and sira. The total number of akṣaras on the line would thus be 41 akṣaras, which is far too few. We are unable to verify any akṣaras after pātaka, and so are unable to propose our own estimates for the three gaps.