Primal Information

Over at Not Exactly Rocket Science, Ed Yong has a great summary of a new paper trying to figure out why information (at least in primates) can be just as rewarding as primal, biological rewards, such as calories and sex.

Ethan Bromberg-Martin and Okihide Hikosaka trained two thirsty rhesus monkeys to choose between two targets on a screen with a flick of their eyes; in return, they randomly received either a large drink or a small one after a few seconds. Their choice of target didn't affect which drink they received, but it did affect whether they got prior information about the size of their reward. One target brought up another symbol that told them how much water they would get, while the other brought up a random symbol.

After a few days of training, the monkeys almost always looked at the target that would give them advance intel, even though it never actually affected how much water they were given. They wanted knowledge for its own sake. What's more, even though the gap between picking a target and sipping some water was very small, the monkeys still wanted to know what was in store for them mere seconds later. To them, ignorance is far from bliss.

[SNIP]

This preference for knowledge about the future was intimately linked to the monkeys' desire for water. The same neurons in the middle of their brains signalled their expectations of both rewards - the watery prizes and knowledge about them.

All the neurons in question release the signalling chemical dopamine. While the monkeys were making their choices, Bromberg-Martin and Hikosaka recorded the activity of 47 dopamine neurons in their midbrains. These neurons became very excited when the monkeys saw a symbol that predicted a large amount of water, while the symbol that cued a smaller drink inhibited the neurons. The same dopamine neurons were excited during trials where the monkey only saw the symbol that heralded forthcoming information, and they were inhibited if they monkey only saw the other non-informative symbol.

So the same population of midbrain neurons signal changes in both the thirst for water and for knowledge. The more active they are, the stronger that thirst is. One monkey had a stronger preference for early information than the other and indeed, its dopamine neurons were more active when it saw the informative symbol. Even for each individual monkey, the neurons were more active on specific trials where they showed a preference for advanced knowledge.

These experiments elegantly demonstrate an essential feature of the human mind, which is how evolution bootstrapped our penchant for ideas to the same reward circuits that govern our animal appetites. In other words, the political activist on hunger strike might still be relying on his reward circuity, even though he's actually denying himself caloric treats: the cause is simply more important than food. That's what makes ideas so powerful: No matter how esoteric or ethereal or abstract they get, they are ultimately plugged back into the same system that makes us want sex and sugar. The end result is that we can crave knowledge and facts just like a thirsty person craves water.

UPDATE: And doesn't this also help explain the allure of suspenseful narratives? A good story, after all, is simply the artful denial of information - Will Elizabeth marry Darcy? Will Jason Bourne survive Moscow? - so that the audience craves resolution, which arrives in the form of information. The happy ending is a universal human reward.