Two for two.
Not only was the day a complete success, but we had accomplished all of this in two hours—half the time of the previous session.
It was still exhausting. When you’re locked face-to-face inside the magnet with jackhammers all around you, the level of concentration, for both dogs and humans, is intense. When Callie and I finally got home, we crashed together on the couch. We looked at each other once and then closed our eyes.
19
Eureka!
ANDREW DIDN’T WASTE ANY TIME. The next day, he had already begun the analysis of Callie’s and McKenzie’s data. Just like the peas and hot dogs experiment, the first and trickiest part of the analysis would be the motion correction. We had to carefully identify which scans contained brains and discard the ones in which the dogs moved too much. Animating the sequence of images in rapid speed helped make the task easier.
Andrew showed me the animation.
“Check this out,” he said. A pixelated image of a dog’s brain danced on the computer screen. For stretches of several frames, which were actually tens of seconds in real time, the image didn’t move. Except the eyeballs, which darted left and right.
“This is Callie,” Andrew continued. “She did really well. If we throw out the scans with movement artifacts, we still have 62 percent left for analysis.” My heart swelled in pride at my beloved feist.
“That’s amazing,” I said. “That is five times better than the previous session. How about McKenzie?”
“Almost as good. We can keep 58 percent. She had sixteen hot dog trials and eleven no hot dog trials.”
“Melissa was really making her hold still for a long time,” I said.
“Yes,” Andrew said, “but that means we’ll have a lot of scans for each repetition.”
We spent the next two days checking and rechecking each step of the analysis. To make sure that we didn’t mistakenly confuse brain activation with movement artifacts, we kept ratcheting up our criteria for whether to keep a scan in the analysis. Andrew and I would stare at the animations, looking for even the slightest twitch of the head. Most of the head motion occurred when we gave the dogs hot dogs. This was no surprise. But we weren’t interested in the brain response to hot dogs. We were interested in the response to the hand signals. When we were satisfied that we had identified and discarded all the scans with motion, the remaining scans showed that the dogs had held their heads with less than one millimeter of movement during the critical period of the hand signals. That was as good as humans do in the scanner. We were ready for the final step: comparing the activation between the two hand signals.
All fMRI experiments measure relative changes in brain activity between different conditions. With only two conditions—the signal for “hot dog” and the signal for “no hot dog”—all we had to do was subtract the brain activity in one condition from the other. The difference would show us which parts of the dogs’ brains processed the meaning of the signals.
To do this, we usually calculate the difference in activity at every location in the brain and perform a statistical test to determine whether the results are real or simply random fluctuations in the fMRI signal. We then create a map from this analysis and overlay it on the structural image. By convention, neuroscientists use a color scheme that ranges from yellow for weak activations to bright red for strong ones. Andrew did the subtraction for Callie.
Everyone in the lab had been waiting for this moment and gathered around the computer screen.
There, overlaid on Callie’s pyramidal-shaped brain, were several hot spots of yellow and red. We still didn’t know what most of the brain was doing. It was important to stay focused on the one region that we knew a lot about.
“Zoom in on the caudate,” I said.
Sure enough, an orangish blob sat squarely on top of the right caudate. There was no doubt. The lab stared in amazement and let out a collective gasp.
McKenzie’s activation map was even stronger. Both dogs showed unmistakable proof of caudate activation to the signal for “hot dogs” but not the signal for “no hot dogs.”
Had only one of the dogs shown caudate activation, it would be easy to dismiss as a fluke. But we were looking at caudate activation in both dogs. The odds of that happening by chance we calculated to be 1 in 100.
“Caudate activation in both dogs?” I said. “That it is no accident. That is real.”
At dinner that night, I broke the good news to the girls.
“The Dog Project worked,” I announced.
“What do you mean?” Kat asked.
“We found reward-system activation in both dogs.”
“So,” Kat said skeptically, “you discovered that dogs like hot dogs?”
“No,” I replied. “We discovered that they understand the meaning of hand signals.”
This was a crucial distinction. In fact, Andrew and I did observe caudate activation to the hot dogs. But because the delivery of the hot dogs also caused Callie and McKenzie to move their heads as they swallowed and licked their lips, we had to discard a high proportion of those scans. Even so, the caudate activation was still plainly evident. But for the reasons Kat implied, such a finding would not be very surprising. Everyone knows that dogs like food.
No, the big result was caudate activation to the hand signal for “hot dog” but not “no hot dog.”
The Pavlovian behaviorists would say, “Ah, the dogs learned the association between a neutral stimulus—the hand signal—and an unconditioned response from the food. Nothing in the brain implies an understanding of meaning.” Had we done the experiment like Pavlov, using the ringing of a bell, for example, or the turning on of a light in place of the hand signal, this would certainly be true. But we used hand gestures. Humans take it for granted that hand gestures convey a great deal of information, almost as much as the eyes. Is it possible that dogs place as much importance on hand movements as we do?
A growing body of evidence suggests that they do.
Brian Hare, an anthropologist at Duke University, has pioneered the study of social cognition in dogs, especially the extent to which they understand human social signals. In his initial experiments, Hare hid food in one of several possible locations in a room. A human would stand in the room and point to the correct location. When a dog entered the room, it was able to use the pointing cue to more quickly find the food. Often, the dogs did this on the first try, indicating that simple associative learning, like the behaviorists believed, could not explain dogs’ ability to intuit the meaning of human social signals. Dogs seem to be particularly skilled at reading human signals. Hare later tested wolves and chimpanzees, and neither did as well as dogs.
Even at the dinner table, I could see that Callie was exquisitely attuned to our human social interaction. Lyra—not so much. But Callie sat in relaxed attention. Her head would swivel to whoever was speaking. Although she couldn’t understand all the words, if someone said one of the words she did know, like walk, she would run to that person and start wagging her tail vigorously. More than speech, I knew she understood hand signals, because all I had to do was point at the MRI tube, and she would go in.
Now we were faced with the conundrum of reverse inference.
If we had been studying humans, the interpretation of the caudate activation would be pretty simple. In fact my colleagues and I had done exactly this kind of experiment ten years earlier. Instead of hot dogs, we used Kool-Aid. In that experiment, our human subjects lay in the scanner with a tube snaking into their mouths. When a green light appeared on a computer screen, the subjects would have to press a button, and then, a few seconds later, they would get a squirt of Kool-Aid on their tongues. Just like Callie and McKenzie, the humans’ caudates activated to the signal indicating impending Kool-Aid. Since then, this result has been replicated dozens of times by us and other researchers. The advantage with humans, of course, is you can ask them what they thought and felt in response to the signals.