Modern Greek depends on Ancient Greek language?

[Terasgr]

Hello people. 

When I tried to OCR a Greek text using tesseract I found that the modern Greek data (ell) uses the ancient Greek language (grc) as a sublanguage.

Since in terms of glyphs, this is the other way around (i.e. all modern glyphs are found inside the ancient glyphs --  the modern glyphs is a subset of ancient glyphs), I am thinking what is the reasoning for this choice. Is this a bug? Is it solving something that I can't see right now?

Thank you