Fix typos in the metadata for genre_narrow

[bambooforest]

|written_traditions |
|writtten tradition |
|written tradition

• should also write some tests to catch any new data that doesn't adhere to some set of values for the metadata headers:

https://github.com/uzling/100LC/blob/master/Database/clc/models.py#L28

Ah, already left myself a note to do this:

https://github.com/uzling/100LC/blob/master/Database/clc/models.py#L108-L109

[bambooforest]

Note also:

corpus.genre_narrow --> mode

[bambooforest]

9 Bagirmi spoken grammar spoken
16 Burushaski spoken grammar spoken
64 Kayardild spoken grammar spoken
98 Rama spoken grammar spoken

I think these should just be NAs, as with the case of Otomi:

91 Otomí (Mezquital) written grammar NA