The aMEG approach made it possible to estimate spatio-temporal stages of the event-related total theta power to words and pseudowords. The observed pattern is consistent with the left-lateralized posterior-to-anterior fronto-temporal processing stream as reported for different language tasks with time-domain aMEG analysis (Dhond et al., 2001, 2007; Halgren et al., 2002; Marinkovic et al., 2003; Marinkovic, 2004). Figure 2 depicts group average event-related theta source power estimates for the three stimulus types under placebo as they change across time. The estimates in the form of maps and timecourses for the relevant ROIs are shown for the left hemisphere only since they were much weaker on the right overall. Wavelet-based estimates suffer from temporal smearing due to the width of the wavelet especially at lower frequencies. Nonetheless, we have observed differential modulation across time in different brain areas in agreement with the activity progression from the sensory-specific to supramodal regions during language processing (Marinkovic, 2004). The earliest activity was seen in the occipital area at ∼130 ms and it did not differ among the conditions. Lateral middle temporal area (LT) was
