Contra previous uniform approaches for wh-phrases, the current paper argues that there is a clear asymmetry between in-situ argument and adjunct wh-phrases with respect to Intervention Effects (IEs) in Korean and Japanese. Based on the categorical (nominal vs. adverbial) dichotomy evidenced by structural case attachment tests and formation of complex wh-expressions, different base locations for wh-arguments (inside vP) and wh-adjuncts (outside vP) are suggested in these languages. Finally, I propose that IE asymmetries be attributed to the inherently different properties of argument and adjunct wh-phrases under scrambling operation.