Serendip is an independent site partnering with faculty at multiple colleges and universities around the world. Happy exploring!

The Future of Neural Imaging and the Ethics of Mind Reading

Jeremy Posner's picture

      The greatest limitation in the field of human biology has traditionally been the inability for research to be conducted upon the body while it is still functioning.  This has been a particularly large barrier to overcome in the study of the nervous system.  Non-living tissue may suffice to study the anatomy of the brain and of the peripheral nervous system, but the activity of nerve cells is difficult to measure accurately even in living tissue, and so for many years the function of the nervous system was understood through the study of animal models and via observation of the effects of neurological disorders and injuries.  The earliest attempt to obtain images of living nerve tissue employed the X-Ray, and accurate but inanimate images of the brain were available in the late 1920s.  The CAT scan was developed in the 1970s, allowing for very precise three dimensional representations of the brain to be assembled from a series of two dimensional X-ray images.  The MRI and PET scan were then introduced in the 1980s and eventually the principles of both techniques were combined to create the fMRI.  Most fMRI work using blood-oxygen level dependent imaging, which can be used to view changes in bloodflow patterns within the brain at a high temporal resolution.  The amount of blood being used by a particular area within the brain has been shown to correlate strongly with point neural activity within the brain.

  Recent work with fMRI imaging has led to the proposal of some new and radical uses for neural imaging.  These novel applications for neural imaging revolve around the premise that, if it is possible with imaging methods such as fMRI to identify how the brain activates in response to a certain stimulus, and if it is also possible to identify within the brain the particular functions of certain areas and pathways that the combination of the two would allow researchers to determine what the brain is doing at a particular time, to read minds.  fMRI as it exists currently is unlikely to be up to the task, imprecise spatial resolution prevents the activity of individual neurons or even neural pathways to be examined and individual differences in neuroanatomy and neural function either require that imaging be extensively calibrated to each individual or prevent meaningful results of the level of specificity needed for “mind reading” to be possible.  While these uses of fMRI are still in their infancy it is unclear whether neural imaging as it currently exists will ever be able to accurately provide an image of the brain’s function, even as improvements in spatial and temporal resolution continue.  It is possible that, in order to accurately capture the neural processes that underlie complex thoughts the method by which neural imaging occurs will have to dramatically change. 

The potential for future imaging techniques to accurately capture the activity of the brain and to identify neural processes brings into play a very different issue: the extent to which an individual’s thoughts should be private.  The Wired article discussing the use of fMRI as a form of lie detection and introducing brain imaging as a source of evidence in court provides both immediate moral issues related to the use of potentially unreliable data provided by fMRI in a trial but also introduces the long term dilemma of the fairness of being convicted based upon one’s thoughts.  In the immediate there is no doubt that the use of fMRI or other current imaging techniques to identify so complex a neurological process as lying reliably and to depend upon the result in making an important decision such as whether to convict an accused man or woman or to hire that man or woman for a job is both unfair and unwise.  Traditional lie detectors are similarly unreliable though less expensive and time consuming to employ, and should be viewed similarly, though some states allow for the use of lie detectors in the hiring process.

Recent studies that have achieved results with neural imaging that were previously inconceivable have pushed some to consider the implications of the opposite, the availability of a technique that can accurately determine if a person is lying, or whatever else they are thinking.  A group at the Max Planck institute (Haynes et al, 2007) found that it was possible to distinguish whether subjects intended to engage in subtraction or addition based upon fMRI scans.  A joint venture between the Carnegie Mellon psychology and computer science departments managed to create an algorithm that interprets fMRI and is able to, with some reliability (though still with vast limitations) determine what a test subject is thinking of, without even calibration to the brain of that particular individual.  While even these new advances in the use of imaging are a far cry from mind reading the question is being broached: should the mind be off limits?    

There is something to be said for both sides of that debate.  Some in the scientific community, such as Forbes’ Paul Wolpe (Ord, 2009) argue that the potential abuses of a mind reading technology are so extensive that such an application of neuroimaging technology should not even continue to be explored.  Many, such as Oxford’s Toby Ord, disagree with this position, citing potential uses of neuroimaging based mind reading that would provide considerable benefit to society.  Ord specifically anticipates using neuroimaging to help prevent false convictions of accused criminals (Ord, 2009).  That specific example is meant to serve as a counterpoint to fears of abuse of “mind reading” by the government; Ord also points out that such a technology could be enormously important for national security. 

In the legal arena neuroimaging might be viewed in a similar light to testing for DNA evidence, though it is possible that, like DNA evidence, the scientific character of neuroimaging might lend undue weight to neuroimaging evidence. As a part of legal proceedings suspects can be required to submit to DNA testing within the U.S. and in that regard the required participation in an fMRI session is not out of line with the normal operation of our legal system.  For that reason if a “mind reading session” can be considered akin to DNA testing it could be applied with impunity, but if it is considered to be an actual statement by the accused there may be a potential problem when considering the application of neuroimaging in the legal setting with regard to the right against self-incrimination and how that right might be applied to the use of neuroimaging as well as issues of Miranda rights concerning neuroimaging (though of course traditional council would do very little to help a person deal with an evaluation of activity in their brain). 

As neuroimaging technology advances another potential issue may also become more prominent; if it is possible to determine via fMRI not just if someone is lying or not but to examine their thoughts, or to examine memory; are those uses of neuroimaging to be regarded in the same way?  Lie detection would be useful in the examination of statements made by witnesses and suspects, and would not necessarily reveal anything invasive beyond the veracity of what they provide relevant to the trial.  If a suspect refuses to disclose information, or is not asked the right questions, or perhaps even tells a half truth or lies by omission it is possible that even an effective lie detector would be of little help.  As was discussed during the seminar on memory it may not be possible to actually access a person’s memory via a scan of their brain, because there may not exist within the brain a physical memory to examine.  It may, however, be possible to use a combination of interview, suggestion and even potentially hypnosis to make particular memories salient to an individual and through that process make them accessible via neural imaging.  That particular use of neural imaging does extend beyond the limits of ethically acceptable practices.  This practice would take the power to control expression out of the hands of the individual and thus invalidate their Fifth Amendment rights in a legal context and even outside of a legal context the freedom of expression to which U.S. citizens are guaranteed and which I believe all should be guaranteed extends to the freedom to withhold expression.     

Of course not every justification for mind reading can be justified as necessary to insure national security; research into improving fMRI scanning techniques for a “mind reading” purpose are currently underway with the intention of marketing the result commercially.  Some states do allow for the use of conventional lie detection devices as a part of the hiring process.  The less than stellar efficacy of these tests, if nothing else, makes this practice ethically unacceptable to me.  Brandon Keim also evokes another potential realm of commercial use, using neural imaging to conduct market research in order to create products and advertising strategies to more perfectly entice customers to spend their money.  Keim is strongly opposed to this use of neural imaging, but I am less sure that even especially precise neural imaging would result in particularly unethical commercial practices.  It would be completely impractical for individual customers to receive fMRI and tailoring marketing to neural imaging findings is not very different than the psychology research around which advertising is often based currently. 

A final, interesting future direction for the neural imaging of thought is in the marriage of therapeutic and medical treatments for diseases and mental disorders.  While it is currently understood that similar effects can be achieved neurologically via very different treatments of the same disorder, such as CBT and drug treatment of major depressive disorder, the exact overlap between these types of treatments and what mental the mental processes that may be beneficial in each are is based largely upon observation and self-report measures at this point.  Each treatment type has serious flaws, most prominently the risk of serious side effects in drug treatments and the cost and demands placed upon the time of the patient in talk therapy.  If the key, efficacious components of each treatment could be identified the downsides of each might eventually be minimized and their effectiveness maximized.  Sophisticated neural imaging techniques might also help to minimize the trial and error that is currently a large part of psychiatry.  There is a large amount of variation in how effective a particular treatment may be from patient to patient or even potentially how detrimental a particular treatment may be.  Through the examination of the common neural characteristics of groups of patients within a population suffering from a particular disorder and the characteristics of positive and negative reactions to a particular treatment those reactions may eventually become better understood and anticipated.      



Haynes, J. D., Sakai, K., Rees, G., Gilbert, S., Frith, C., & Passingham, R. E. (2007). Reading hidden intentions in the human brain. Current Biology, 17(4), 323–328.


Ord, t. (2009, october 12). The Ethics of mind reading. Practical Ethics, Retrieved from


Paul Grobstein's picture

brain imaging: "should the mind be off limits?"

Interesting question indeed, if we get rid of the Decartesian dichotomy and presume that mind and brain are the same thing.  Along these lines, its useful to be reminded, as in the example of the fifth amendment, that some presumptions about mind privacy have long been part of the American legal system.  And to equally note that there are some inconsistencies about this.  While one need not testify about one's own behaviors/motivations in some cases, other evidence intended to establish "intent" is not only permitted but essential in legal proceedings in the case, for example, of first degree murder.