How Did the Man in the Well Reach Parma?

Yesterday, 8 September, was the feast of St Josaphat of India in the Greek Orthodox calendar — 26 August by the Julian reckoning. The Roman Martyrology kept him on 27 November until the 1960s; the Slavic churches still keep 19 November Julian. Josaphat is the Buddha. Sanskrit bodhisattva became Middle Persian Bōdisav, then Arabic Būdhasaf, then Yūdhasaf when a copyist put one extra dot on a bāʾ, then Georgian Iodasaph, Greek Ἰωάσαφ, Latin Iosaphat. His father took the same beating: Arabic Junaysar, misplaced iʿjām, Habeneser, Georgian Abeneser, Greek Ἀβεννήρ. That is the famous part, usually told as an accident nobody caught until Laboulaye and Liebrecht worked it out in 1859–60. But the accident is the least interesting property of the text. Trace the lineage border by border and you can watch which component was actually travelling — and it was never the biography.

The frame gets rebuilt at every crossing. Manichaean missionaries picked up the Buddha legend before 800 and replaced its doctrine wholesale; fragments survive at Turfan in Parthian, Middle Persian and Uyghur, and the Manichaean spelling bwdysdf is the direct ancestor of the Arabic form (Encyclopædia Iranica). Baghdad made it the Kitāb Bilawhar wa-Būdhāsaf; Georgia made that the Christian Balavariani; around 985 the Georgian monk Euthymius the Hagiorite put that into Greek. Payload swapped twice. The parable, meanwhile, does not change. Apologue IV — the man in the well — comes out of Mahābhārata XI.5, where Vidura tells it to Dhṛtarāṣṭra, and it is carved at Amaravati and Nagarjunakonda in the 2nd–3rd century CE (Zin 2011). The Greek at Loeb ch. XII keeps the full inventory: the pursuing beast, the pit, the tree, “two mice, the one white, the other black,” the fire-breathing dragon below, four asps in the wall, and the drop of honey (Woodward & Mattingly 1914). Around 1270 that same inventory is carved onto the south portal of the Parma Baptistery. Two stone carvings of one story, a millennium and six thousand kilometres apart, on monuments belonging to religions with nothing doctrinal in common.

The scribes themselves knew which part was portable. Robert Volk’s critical edition — 160-plus manuscripts, all personally inspected, the oldest from 1021, thirty-five years after composition — closes with appendices of Greek epitomai and excerpt-collections consisting especially of the parables (BMCR 2009.06.04). The detachment is documented inside the tradition, not merely downstream of it. Romania gives the clean chronometer. The Teachings of Neagoe Basarab to his son Teodosie, written in Slavonic around 1518–21, contains six of the parables — the unicorn, the trumpet of doom, the four caskets, the nightingale, the three friends, the king for a year: apologues I–VI, in order — and no Josaphat at all; the novel itself does not reach Romanian until 1648 (Stanciu, RRL 2013). The parables beat the story into the language by roughly 130 years, and the same article notes a later Life of the Saints version circulating with the parables stripped back out. The two halves come apart cleanly in both directions — and the parables keep going after the frame is dead. Barlaam’s Apologue II has four caskets: two gilded and full of bones, two smeared with pitch and full of pearls. The Gesta Romanorum cuts it to three, adds the inscriptions, and hands the choice to a woman; from there, Gower, Boccaccio, and Portia’s suitors. Tolstoy took the other one — chapter 4 of A Confession retells the man in the well with the black and white mice and the honey intact, calls it “une fable orientale… bien vieille,” and then refuses the frame outright: “Ceci ce n’est pas une fable, mais c’est la pure, l’incontestable vérité.” Zero doctrine, zero Buddha, zero Josaphat. Pure form, still load-bearing.

Run the test in reverse and the frame turns out to be a cargo hold. In 1889 Rendel Harris found a Syriac Apology of Aristides at St Catherine’s on Sinai; while his edition was in press, Armitage Robinson realised the Greek had never been lost. It is Nachor’s speech, Loeb ch. XXVII — and the plot around it is that Nachor is a pagan astrologer hired to impersonate Barlaam and lose the debate on purpose, who panics and wins it. The earliest surviving Greek Christian apology, addressed to Hadrian or Antoninus Pius, is extant because a Georgian monk needed a speech for a fraud who changes his mind (Apology of Aristides). So the claim is not that Christendom mistook the Buddha for a saint. It is that the durable unit of transmission across this whole lineage is the freestanding parable of two to four hundred words, and everything above that scale — biography, doctrine, names, proper nouns — is rebuilt at every border it crosses. If that generalises, it is measurable. My bet: for any medieval compendium assembled out of embedded exempla — Kalīla wa-Dimna, the Seven Sages, the Disciplina Clericalis — the individual exempla are attested earlier, in more languages, and in more manuscripts than the frames carrying them. Nobody has built that stemma, because philology catalogues complete texts and treats excerpt-manuscripts as damaged goods. Volk found the parable-excerpts and put them in an appendix. They belong on the main stemma, as the fast-moving line.