Frequency, memory formation, memory retrieval or access

Once we know that the brain works differently when making semantic judgments about words encountered with different frequencies, we can ask how the brain processes these words so that information about them can be accessed later. So, with regards to frequency, why is it that when asked to recall word lists, we tend to recall more frequent items, whereas when asked to recognize word lists, we do it better for less frequent items. Clearly, something about the previous history of these words of different frequency of encounters has an affect on the way information about a specific encounter of these words is retrieved later.

First, we can look at what the brain does differently for words of high and low frequency that leads to different memory access. In sum, the brain seems to be more active during semantic judgments for words that are later recognized. Furthermore, these words that are more active tend to be the low frequency words. Perhaps, because we have to work harder at making semantic distinctions for such words, we end up enhancing their recent representation. This makes it easier for them to be retrieved later for recogntion. Abstract: Word frequency and subsequent memory effects studied using event-related fMRI.

Next, we look at how the brain handles these words during recognition itself. Perhaps during this stage, other processes additively or interactively affect the resulting recognition response. The mechanism might operate in this manner: the top-down criteria is made active (words that appeared at a particular point in time), probe words are encountered and identified, information for these words are matched to the active set, a decision is made about the recognition memory for the word. In sum, word frequency of the test words seen during encoding did not affect brain activity differentially, however, word frequency of new words not seen at encoding did make a difference to brain activity. That is, high frequency novel words activated the parietal region more than low frequency words when subjects recognized them as such. Abstract: Recognition memory for studied words is determined by cortical activation differences at encoding but not during retrieval.

So once we encode the word recently, the word frequency does not matter anymore, at least to the recognition judgment. What matters in the recognition is trace of recency or specific time information associated with the probe word. This, presumably, is not dependent on the frequency history of the word. However, if the word is not encountered before, word frequency seems to make a difference in the parietal area, an area that is involved in attention modulation and visualization. Perhaps, the higher activity for high frequency words in this case marks the more connected representation of high frequency words compared to low frequency words. Sort of an automatic engagement of the relatively wider spread of activation for high frequency words.