I know that polarity in a EEG signal is arbitrary, and a matter of convention, as clearly explained in this past question.
However, assuming said conventions are well-followed, I wonder whether is there any intrinsic meaning at all to a particular ERP/component being described (as they often are in the literature) as either a positivity or a negativity?
It seems indeed that this aspect (voltage polarity) is a salient-enough feature of EEG components, that they are metonymically referred to by this very feature, e.g. "this positivity is associated with cognitive process X", as if the component's sign was just as important as its latency and absolute-value amplitude (which are of course not arbitrary, and clearly meaningful).
Some papers (e.g. Koelsch et al 2005) seem to suggest that some cognitive tasks (e.g. perceiving some deviance in a stimulus) is only associated with one specific polarity, in this case negative:
(..)it is also possible that any kind of deviance-related negativity has an effect on linguistic syntax processing.
To sum up: is there any intrinsic meaning at all to be found in a ERP's polarity?