In her interesting commentary, Valerie Gray Hardcastle argues that a P300 can be obtained even with "completely" novel stimuli which do not access or index LTM codes (1995). My argument is that, in order to get recognized, even a novel stimulus requires some sort of LTM information. Thus, even in this case some sort of "indexing of a memory code" would be necessary. I agree that the P300 reflects much more than just this type of process. However, this should be expected if the P300 reflects processes of a complex Working Memory System as I have proposed.
2. Let me briefly state the critical aspects of the WMS. It was proposed that the interaction between LTM and the WMS plays an important role for encoding, searching, retrieving, and recognizing information. Processes underlying this close interaction between the WMS and LTM were termed "monitoring processes". One of the most obvious tasks of these monitoring processes was seen in the identification of a familiar object: Top down processes which are guided by expectancy and selective attention are capable of directing the recognition process towards a certain outcome by preactivating or preselecting appropriate templates or prototypes in LTM (see also, Grossberg, 1980; Carpenter & Grossberg, 1993 for a similar conception).
3. My argument is that even a novel stimulus would undergo a similar process and that even in this case some sort of LTM will be accessed. If no LTM information can be accessed, then no stimulus will be recognized. However, the geometric Figures Hardcastle used in her experiment were certainly recognized as geometric figures by the subjects.
4. I do not argue that the P300 just reflects access to LTM or just some matching process between LTM and the WMS. My hypothesis is that the P300 reflects all of the heterogeneous aspects of monitoring processes including attention, contextual information, expectancy, emotion, and certain autobiographic aspects.
5. I think that it will hardly be possible to unconfound these different aspects of monitoring processes by using ERPs. From my point of view, the hotly debated issue concerning the functional meaning of the P300 is an example which well documents this difficulty. However, spectral analyses focusing on power changes in different frequency bands might offer a promising way to unconfound different cognitive processes, because our data indicate that different EEG-frequencies reflect different types of cognitive processes.
Grossberg, S. (1980) How Does a Brain Build a Cognitive Code? Psychological Review, 87, 1-51.
Carpenter, G.A. & Grossberg, S. (1993) Normal and amnesic learning, recognition and memory by a neuronal model of cortico-hippocampal interactions. Tins, 16, 131-137.
Hardcastle, V.G. (1995) An Expanded Role for the P300: An Addendum to Klimesch. PSYCOLOQUY 6(23) memory-brain.2.hardcastle.
Klimesch, W. (1995). Memory Processes Described as Brain Oscillations in the EEG-Alpha and Theta Bands. PSYCOLOQUY 6(6) memory-brain.1.klimesch.