If the cables are close enough to influence each other thermally, then I'd say yes, de-rate as if they were two circuits.

I can understand the confusion, the wording in the regs of "number of circuits of multi-core cables" only covers common situations adequately (i.e. not singles in parallel). The physics seem to be clear though - the cables don't care if they're feeding two separate circuits or are paralleled into a single circuit - they carry the same current, have the same resistance and produce the same heat regardless. 

I suppose the wording should be more like 'how many sets of loaded conductors you have divided by the number in a set described in the ratings table (usually 2 for single phase or 3 for 3-phase) but descriptions like that don't trip off the tongue quite as easily (or fit into table headings).

   - Andy.

seconded - its all about the spacing - widely separated cables may be thermally independent, so no derating, but if they are close enough that they warm each other up - usually to be independent needs a few cable diameters ,  then some de-rating should be applied.

Mike.

Well noted. Thank you so much. 

Well noted. Thank you so much. 

Well noted. Thank you so much.