Last week two of our manuscripts were accepted by the peer-reviewed journal Learning & Memory. They will appear back to back in the same issue. Those are my first back to back publications and I'm happy it turned out that way.
The articles are about how fruit flies ( Drosophila melanogaster) learn what is important in their environment and how they use this information to form flexible memories. For instance, the flies learn that certain visual patterns in their environment can predict dangerous heat. If this predictive learning takes place in a colored background (say, green), the flies are flexible enough to treat the "dangerous" patterns still as dangerous in a new background color (one they can distinguish from green, say, blue-green). If however, these same colors indicate that different sets of patterns are dangerous in different colors, the flies can learn such "if-then" relationships as well. For example, the animals can learn: "if the background is green, then one of the two patterns is dangrous. If the background is bluegreen, then the other one is dangerous". In other words, even flies can learn to generalize or discriminate flexibly between different situations (in this case background colors in the environment), depending on the information content of the situation. Our studies suggest that this flexibility is partly dependent on a prominent insect brain area, the mushroom bodies. Only generalization requires the mushroom-bodies, discrimination does not. This finding is corroborated by the result that different parameters in the background colors are used for generalization and discrimination, respectively.
In conclusion, we hypothesize that generalization and discrimination are separate, competing, simultaneous processes in a sort of push/pull relationship. Which of the two comes to dominate the behavior is determined by both a set of innate rules as well as the predictive value of either situation.
I'll present more about this on brembs.net as soon as the articles are published.