Reproducibility and uncertainty

Different uncertainties

Missing surface, an extant but illegible trace, uncertain sign identification, disputed word division, ambiguous morphology and doubtful translation are separate problems. Name the earliest uncertain step. Later layers inherit that uncertainty unless independent evidence resolves it. Oracc’s ATF Primer and papyri.info provide explicit but different symbols for several of these states; read each project’s legend before transferring a line.

Codex Sinaiticus shows another kind of plurality: corrections by later hands can be inspected in the project’s transcription view. A corrected reading is not merely a modern editor’s repair. Record the hand or stage when the project identifies it, rather than collapsing all readings into one unqualified text.

A replicable path

A second reader should be able to open the same source, locate the line, apply the editorial key, repeat dictionary or corpus searches, and see where your judgement enters. Keep query strings, dates, edition versions and original notation. If an image was unavailable, say so. If you change a reading, preserve the earlier note with a dated reason.

For each conclusion, write what could overturn it: a clearer image, another manuscript, a newly published join, a closer parallel or a grammatical counterexample. This turns “uncertain” into a useful research question.

Practice

Compare the Oracc Primer’s marks for damage, loss and queried reading with papyri.info’s marks for restored loss and visible uncertainty. Make three rows headed “observed state,” “editorial notation,” and “what new evidence would help.” Fill them from the documented examples, keeping the two systems identified by name.