|The orthography of machine-readable Neolatin texts: A plaidoyer for minimal intervention|
Automated 'correction' can create new problems: A test case
The automated correction of cygn- to cycn- has created problems in several passages in the Mannheim corpus.
Cygnea, Latin name for Schwanenstadt, Zwickau. Matthiae, Theatr. hist., p.1050b cum ... Cygneam et Torgam ... occupavit. CAMENA changes the name to Cycn-; this seems to be without basis, I have not been able to find a single instance where the place name Cygnea is originally spelt Cycnea.
Confusion is introduced into the Lemma CYGNUS in Hofmann's Lexicon universale, where the spelling -g- is retained in the Lemma (unavoidable because of the alphabetical position between CYDRUS and CYITES), but changed to -c- in the text.
Completely mystifying is Nolte, Antibarbarus, p.128 "in veteribus libris legi ... cignus, pro ... cycnus l. cycnus", until one realizes that the first cycnus is not Nolte's, who wrote "cygnus l. cycnus", a reference to his previous discussion of this word (p.47). The initial statement in these pages in CAMENA refers the reader, if in doubt, to the page image. But in many cases the reader will be unsuspecting initially, or even assume a word does not occur, if he cannot find it, because its spelling has been changed.
cygnaeus/cycnaeus/cycneus: Sometimes, cygnaeus is 'emended' to cycneus (Widl, Carmina, p.268), but cycnaeus remains unchanged. A further drawback: With the change from cygnaeus to cycneus the long vowel in the penultimate syllable is no longer visible (cycneus can also be measured with short -e-, e.g. in Boiardo, Pastoralia, G 5.63).
There is the further possibility cicn-, e.g. Taubmann, Postuma schediasmata, p.217 Cicnos (registered as a variant by CAMENA). The spelling is rare, but not unique, e.g. Verino, Carlias, 6.619 leporem aut cicnum (Byzantine pronunciation ?). A reader might expect also these (and further ?) cases to be normalized.--
Even a seemingly harmless or even apparently useful 'improvement' like the change of u before a vowel to v can have side-effects; thus in GOTTFRIED Avity arch p.334 In coronatione regum suorum vel cum bellum immineret solebant Pervani ducentos pueros trucidare, in the CAMENA-transcription the Peruvians (Peruani) have become 'very vain' (pervani).