Broken forms in C9006 ⲉⲡⲓⲕⲁⲗⲉⲓ

[amir-zeldes]

There is a form ⲉ (S) in C9006 ⲉⲡⲓⲕⲁⲗⲉⲓ and a form ⲡⲓⲕⲁⲗⲓ (M) - are these two broken parts of the same form? If so, does this need to be fixed on the DDGLC side or just CCL? @KaJohn-DDGLC @dwerning @simondschweitzer

[KaJohn-DDGLC]

They are not two parts of one broken form.

CF22181 (S) is very likely a processed result from our variant spelling ID 5063 ⲉ(ⲡⲓⲕⲁⲗⲓ) occurring twice in Sahidic in the form of a symbol and treated as an abbreviated form in our data.
CF22194 (M) appears to be a processed result from our variant spelling ID 8615 <ⲉ>ⲡⲓⲕⲁⲗⲓ (M), which occurs once in dialect M with a resolved haplography.
It doesn't need fixing in the DDGLC data.

It appears that angle brackets, parentheses, and their contents have been removed during processing at the TLA, at least for those two cases. However, this also raises the question of whether these two forms are an oversight, with other cases possible or systematic. An initial quick look for variant spellings with the same patterns (ⲁⲅⲁⲑⲟⲛ, ⲁⲅⲁⲡⲏ, ⲁⲅⲉⲛⲛⲏⲧⲟⲥ, ...) does not suggest that it is systematic.

As far as resulting strings are concerned, CF22194 can be seen as a duplicate of CF22192 and is very likely to be deleted. CF22181 might be considered a duplicate of CF22190.

[amir-zeldes]

I see - I understand the need to document all orthographic variants in the comprehensive database. But in these particular cases I think I would prefer not to include the abbreviated forms in those dictionary entries, since they seem less like orthographic variants and more like ad hoc and more unusual situations that could apply to almost any word (letter missing/damaged and a kind of acronym usage). Would it be OK to remove these entries from CDO? And if so, @dwerning / @simondschweitzer would you consider removing them from CCL too?

[KaJohn-DDGLC]

Just a reply to missing/damaged characters and abbreviations: I agree, from an entry perspective, some spellings (i.e., datasets in our database) are not meant to be shown in a dictionary environment. The forms of CF22194 and CF22192 will be represented as one spelling in our dictionary. The same applies to missing/damaged spellings ([]; with resolutions), which are traced and will be listed in our third volume in a separate section.
However, this differs from a type of spelling status in our data (abbreviations), to which CF22181 also belongs. Those are not cases of damaged or 'missing' characters, which indeed can happen to almost every word. However, the latter are abbreviations of a type similar, although not as systematic or functionally 'loaded' as Nomia Sacra, but they still represent their own spellings. Mainly but not exclusively occurring in documentary texts, which can be spelled with or without an abbreviation indicator (/) e.g. ⲇⲓⲁⲕ/ or ⲇⲕ, constituting a spelling on their own, something that can't happen to any word.

[amir-zeldes]

OK, I take your point that some of these may be more conventional than others, but indeed, in a dictionary environment I think these over-generate for searches that are not likely to be intended to find such strings (and for print resources like Crum's dictionary I don't think such spellings are included)