This commentary reviews some of the main findings in speech sound perception using the brain imaging techniques and comments briefly on the recent findings by the session contributors. The main emphasis is on the experimental settings used in these studies. The aim is to demonstrate how the search for the neural correlates for abstract linguistic units has resulted in various types of experimental designs and how the stimulus selection may play a crucial role in the findings. It seems that the experimental settings are becoming more and more elaborate thus offering an access to the abstract levels of representation.