Grzegorz Jagodzinski wrote:
> Borrowings do not form, as a rule, rich word families, especially with
> miscellaneous meanings. Slavic xopiti / xapati does, e.g. Old Polish chopic'
> (catch, adhere, cling), chapac', ochapiac', pochopic', pochopiac',
> wschopic', ochopien', ochopny, pochop, pochopny (survived till today,
> 'hasty, inconsiderate, eager, ready'). So, this word does not look like a
> loanword.
Come on, this "rich" word family consists of two members, whose
relationship to each other is dubious, plus a small group of productive
derivatives for each. Your "rule" concerning borrowings is easy to
falsify with counterexamples. Take any well-known early loan from
Germanic, e.g. *kuningaz --> *kUneN3I and count the derivatives in
Polish: <ksia,dz> 'priest', <ksia,z.e> 'duke, prince', <ksie,z.yc>
'moon' (I hope the meanings are miscellaneous enough) plus a good many
derived adjectives and secondary nouns. Once a loan becomes sufficiently
naturalised, its behaviour is indistinguishable from that of inherited
words.
Piotr