lemmatization of ὄσσε as ὄσσα

[bcrowell]

The inflected form ὄσσε is tagged as a dual ("eyes") 56 times. Of these instances, about half are lemmatized as ὄσσε and half as ὄσσα ("rumor"). As always, I could just be wrong or confused, in which case I apologize, but it seems to me that the latter must all be mistakes, since it seems unlikely that Homer is frequently using the dual number of "rumor." As an example, Iliad 3.427 seems to be clearly a reference to Helen's eyes. Possibly some of the hard-working folks who did the tagging of Homer felt that it would be a good idea to lemmatize this as a fictitious singular form, but I don't think it should be half one way and half the other. IMO it should be lemmatized as ὄσσε (which is what is done in both Cunliffe and the Cambridge Greek Lexicon), not as a fictitious singular. (Beekes seems to think that the dual was inherited from pre-Greek.) If so, then the instances that should be fixed are the following:

iliad 3.427
iliad 4.461
iliad 4.503
iliad 4.526
iliad 5.82
iliad 5.310
iliad 6.11
iliad 11.356
iliad 11.453
iliad 12.466
iliad 13.3
iliad 13.7
iliad 16.316
iliad 16.325
iliad 16.333
iliad 16.645
iliad 16.792
iliad 21.181
odyssey 4.186
odyssey 4.758
odyssey 10.247
odyssey 12.232
odyssey 13.401
odyssey 13.433
odyssey 20.204
odyssey 20.348

If you would like this sort of thing submitted through github as a patch to the two xml files, please let me know. I've been submitting issues on github because I thought it would be of some value to record such things as I came across them, but I'm just an amateur, so I'm never 100% sure if the things I think are errors are just mistakes on my part.