Three witnesses and a voyage episode
You are editing a 16th-century Gujarati narrative poem preserved in three manuscripts. A (dated 1554) is water-damaged and lacks 40 lines in canto 3; it is otherwise archaic and internally consistent. B (1621) supplies the 40 lines; it shares nine unique errors with C, none of them found in A (for example, all three of B's instances of 'samudra' for A's 'sāmudra', and a repeated half-line). C (1698) contains the same 40 lines plus an additional 60-line episode describing a sea voyage and a merchant's dealings with a 'kumpaṇī'. The episode uses 'jāhāj' for ship and 'kumpaṇī' for company. Your publisher wants a readable text and asks whether the voyage episode should go in, since it is 'the most exciting part'.
1. Describe the stemma the evidence supports and say what it implies about the status of the 40-line passage.
Read the worked answer
Stemma: A stands apart; B and C share nine errors absent from A, so they descend from a common ancestor β that is not A. C additionally contains material absent from B, so C is not a copy of B; it descends from β via its own line, or from a now-lost sibling of B.
A → (independent branch) β → B β → (branch leading to C)
The nine shared errors are the key: they are conjunctive errors, proving B and C are related, and since A lacks them, A is not their parent.
Implication for the 40 lines: they are not automatically suspect. They are present in B, which is a β-descendant, and B's text there is not marked by the errors that show β's corruption elsewhere. So the 40 lines may well be genuine, and A's damage, not A's authority, explains their absence in A. They should be printed as text, with A's lacuna recorded.
However: the voyage episode is a different matter. It is unique to C, the latest witness, and its vocabulary is diagnostic (see next answer). Being 'exciting' is not a textual criterion.
2. Should the 60-line voyage episode be printed as part of the poem? Argue from the vocabulary and from the stemma.
Read the worked answer
No - print it as an appendix or in a separately marked section, and say why.
- Stemmatic argument: the episode is singular to C. A singular reading in the latest witness, absent from both A and B, requires positive evidence before it can be accepted; none is offered. The burden of proof lies on the inclusion, not the exclusion.
- Lexical argument: 'jāhāj' (ship) and 'kumpaṇī' (company) belong to the commercial vocabulary of European trade in Gujarat, which postdates the 16th-century horizon of the poem. They set a terminus post quem for the episode, not for the poem. A 16th-century poet could describe a sea voyage, but not a 'kumpaṇī' in this sense.
- Literary argument: the episode's length (60 lines) and its self-contained narrative suggest a separate composition attracted into C's exemplar - a common pattern with merchant and voyage material.
Presentation: print the episode after the text, headed 'Appendix: passage unique to C', with a short note on the loanwords. Do not bracket it inside the canto and do not let the publisher's preference for excitement override the stemma.