Is there a valid comparison to be made between phonemic (underlying) rep and Saussure's langue?
If you set up the framing of the comparison so that the language/parole is specified as being the same as Chomsky's competence/performance distinction, then you could use the example of someone with some sort of articulatory difficulty, having an underlying phonemic knowledge that he isn't able to execute phonetically, then if you link performance to parole, then the counterpart to that would be an underlying phonemic relationship, representative of competence, or in Saussurean terms, langue. That way you could say that an underlying property isn't able to surface properly, for whatever reason, and in that case the underlying phonemic relationship would be equivalent to an inability to show true linguistic competence, because it would affect the performance.
With a few links of separation, you can connect almost anything though. But, there is something to that comparison, I have to admit. It might be broad, but it makes more sense than the other comparison.
