Neuroscience, Mindreading, and the Courts: The Example of Pain


Publish Date:
May 22, 2015
Publication Title:
Journal of Health Care Law & Policy
Journal Article Volume 18 Issue 2 Pages 171
  • Henry T. Greely, Neuroscience, Mindreading, and the Courts: The Example of Pain, 18 Journal of Health Care Law & Policy 171 (2015).
Related Organization(s):


From the first 4 pages of the article article (footnotes omitted):

Our brains hold about 100 billion neurons. At the synapses where neurons connect, the neurons are constantly giving off and picking up chemicals called neurotransmitters, which in turn can cause those neurons to “fire”—to run cascading ions down the neurons’ “wires” or axons. And that process creates the Universe we live in.

Not quite, literally. I do believe, though I cannot rigorously prove, that you exist, the Earth exists, and the Universe exists outside of our own brains, but our only interaction with that reality is through our brains and the physical events that happen there. Those objective physical events create a subjective and non-physical “thing” we call the mind. If you remember tomorrow that you read this Article (or this much of the Article), it is because this Article (and I) will have made physical changes to your brain.

As we get better at looking at those physical changes in the brain through various new technologies, we can begin to correlate those objective physical brain states with subjective mental states. We can begin to say “any time you move the big toe on your left foot, these neurons fire,” or “every time you see a face, those neurons fire.”

Consider, for example, some spectacular work by Professor Jack Gallant at the University of California at Berkeley. Gallant’s group showed thousands of hours of YouTube videos to some volunteers while they were in a magnetic resonance imaging (“MRI”) scanner. The MRI noted the changes at different times in the relative amounts of oxygenated and deoxygenated hemoglobin in different parts of the volunteers’ brains, in a process called functional magnetic resonance imaging (“fMRI”). Computers analyzed the resulting data and found correlations between what the volunteers were seeing at any given time and the patterns of these hemodynamic changes. Gallant then took different volunteers, put them in the MRI scanner, and showed them trailers from movies. His team took the resulting brain scans and, using the correlations from the original work, “re-created” the scenes from the trailers as the volunteers saw them. The results are far from perfect—but still close to amazing. When, in a trailer, an elephant walks across a plain from left to right, the recreation of what the viewer sees from the viewer’s brain scan shows something that looks like an elephant-shaped haystack walking from left to right across a plain. The results come from correlating perceived physical states of the brain with subjective mental states. It comes from, in some small way, reading minds.

This Article is about mindreading and its applications to the law. We are beginning to be able to use neuroimaging and other techniques to read minds. Most of the attention in the burgeoning field of law and neuroscience has focused on issues of free will and criminal responsibility, but the most important contribution that neuroscience will make to the law will be through neuroscience-based mindreading. And I suspect its first important use will be in the area of detecting “pain,” on which this Article will focus.

This Article makes that argument in four parts. First, it looks at what kind of evidence the law currently uses to read minds, and how neuroscience-based evidence would and would not be different. Second, it discusses some of the possible ways the law could use neuroscience-based mindreading. Third, in its most novel contribution, it analyzes what kind of proof the law should demand of the accuracy of such mindreading techniques—and what we would have to invest in developing these technologies to be confident in their use. Finally, it touches on one of the deepest problems that might be raised by the use of accurate mindreading evidence in the law.

We are near the edge of neuroscience-based mindreading in the law. As two of my former Stanford post-docs said in an article, we currently see “through a scanner darkly.” We don’t see clearly, but we see a little, and the resolution of the scanner, or at least our understanding of what it means, is getting better and better. As our resolution and understanding of the scanners gets better, it will become more important in the law, in some ways discussed here and in others still unforeseen.