Lets assume that a person grows up thinking that religion X is true religion. So he builds up a profile in ones head and when notion of religion X appears the neurons that is related are fired, giving sensation of God and related stuff.
Lets us also assume that later in life he learns that religion X was not correct, then does "realization" of this fact happens by modifying old neural networks or a new profile is created for religion X and old profile gets deleted?
In other words do neurons associated with sensation that religion X was true religion "modified" or a new network is created giving a new sensation that religion X was not true religion along!
In my general understanding modification of old network could be challenging assuming that they are surrounded by other neurons associated with other stuff, a certain modification by having more or less neurons would result in borrowing neighboring neurons and wiring them to create a modification or loosing some connections to create a modification, which doesn't seem efficient.