Montréal, 1954.
Fortunately, Peter Milner and James Olds didn’t have
perfect aim with their electrodes.
While postdoctoral fellows at McGill University under the direction of
the renowned psychologist Donald Hebb, Olds and Milner were conducting
experiments which involved implanting electrodes deep into the brains of
rats. The implanting surgery was
done under anesthesia and the electrodes, two of them, half a millimeter apart,
were then cemented to the skull.
After a few days to recover from the surgery, the rats were fine. Long, flexible wires were attached to
the electrodes at one end and to a electrical stimulator at the other, to allow
for discrete activation of the specific brain region where the tips of the
electrodes had come to rest. One
particular Fall day, Olds and Milner were testing a rat in which they had
attempted to target a structure called the midbrain reticular system. Located at the midline of the brain, at
the point where the base of the brain tapers to form the brainstem, this region
had previously been shown by another lab to control sleeping and waking
cycles. But in this surgery, the
electrode had gone astray and it wound up, still at the midline, but in
somewhat more forward position in the brain, a region called the medial
forebrain bundle.
The rat was
placed in a large box with corners labeled A, B, C, and D and was allowed to
explore freely. But, whenever the
rat went to corner A, the experimenter pressed a button to deliver a brief,
mild electrical shock through the implanted electrodes. After a few jolts, the rat kept
returning to corner A and finally fell asleep in a different location. The next day, however, the rat seemed
even more interested in corner A.
Olds and Milner were excited: they believed that they has found a brain
region that, when stimulated, provoked curiosity. However, further experiments on this same rat soon proved
that not to be the case. By this
time, the rat had a acquired a habit of returning often to corner A to be
stimulated. The experimenters then
tried to coax the rat away from corner A: they would give a shock every time
the rat made a step in the direction of corner B. This worked all too well—within 5 minutes, the rat was in
corner B. Further investigation
revealed that this rat could be directed to any location within the box with
well-timed brain shocks—brief ones to guide the rat to the target location and
more sustained ones once there.
Many years
earlier, the psychologist B.F. Skinner had devised the operant conditioning
chamber or “Skinner Box” in which a lever press by an animal triggered either a
reinforcing stimulus such as delivery of food or water, or a punishing stimulus
like a painful footshock. Olds and
Milner soon adapted the chamber so that a lever press would deliver direct
brain stimulation through the implanted electrodes. What resulted was perhaps the most dramatic experiment in
the history of behavioral neuroscience—rats would press the lever as many as
7,000 times per hour to stimulate their brains. They weren’t stimulating a “curiosity center” at all-- this
was a reward circuit, the activation of which was much more powerful than any
natural stimulus. A series of
amazing experiments revealed that rats preferred reward circuit stimulation to
food (even when they were hungry) and water (even when they were thirsty). Self-stimulating male rats would
ignore a female in heat and would repeatedly cross footshock-delivering flood
grids to reach the lever. Female
rats would abandon their newborn nursing pups to continually press the lever. Some rats would self-stimulate 2,000
times per hour for 24 hours, to the exclusion of all other activities. They had to be unhooked from the
apparatus to prevent starvation!
Further work
was done to systematically vary the placement of the electrode tips and thereby
map the reward circuits of the brain.
These experiments revealed that stimulation of the upper surface of the
brain, the neocortex, where sensory and motor processing reside, produced no
reward—the rats continued to press the lever at chance levels. However, deep in the brain, there was
not just a single discrete location underlying reward. Rather, a group of interconnected
structures, all located at the base of the brain and distributed along the
midline comprised the reward circuit.
These included a variety of locations with names like the ventral
tegmental area, amygdala, medial forebrain bundle and septum as well as
portions of the thalamus and hypothalamus. Not all of these areas were equally rewarding. Stimulation in some parts of this
“medial forebrain reward circuit” could support self stimulation rates of 7,000
times/hour while others only elicited 200 times/hour.
It’s hard to
imagine now, but at the time, the notion that motivational or reward mechanisms
could be localized to certain brain regions or circuits was highly
controversial. The dominant
theory, which had held sway for many years, was that excitation of the brain
was always punishing and that learning and the development of behavior could be
explained solely by punishment avoidance. This was called the “drive-reduction hypothesis.” In Olds’ characterization of this
theory, “…pain supplies the push and learning based on pain reduction supplies
the direction.” There was no need
for reward: it was all stick, no carrot.
The pioneering experiments of Olds and Milner clearly demolished the
punishment-only model in favor of a more comprehensive, hedonistic view that
“behavior is pulled forward by pleasure as well as pushed forward by pain”
(Olds, 1958). In this way, brain
pleasure/reward circuits were revealed as important determinants of behavior.