GUEST ESSAY: Further Response to the Psychedelic Neuroimaging Researchers

By Prof. Edward F. Kelly

(This is a guest essay by Prof. Edward F. Kelly, University of Virginia School of Medicine, Department of Psychiatry and Neurobehavioral Sciences. It follows up on an earlier essay and disclosure.)

Bernardo has already provided a swift and trenchant response to the critique of our recent Scientific American blog on psychedelic neuroimaging by six leading figures in that rapidly bourgeoning area of research. Here I wish only to add a few further comments, focusing primarily on issues of particular interest or concern to me.

It is noteworthy first that the authors do very little to address directly the intriguing topic announced in their title – “What Psychedelic Research Can and Cannot Tell Us About Consciousness” – and devote most of their attention instead to the theme of their subtitle – “A recent Scientific American blog misconstrues and oversimplifies the research”.

Let me speak immediately to our allegedly oversimplifying “the details” of the studies we discussed. These studies are all enormously complex, and within the constraints of the Scientific American blog forum (strict word limits, technical discussions discouraged) there is no realistic possibility of getting into the details of any one of them, let alone the whole lot. The best we could hope to accomplish was to convey a faithful overall impression of what has been accomplished to date and to indicate where the research seems to us to be heading. I believe we succeeded in that, but our critics do not.

They go on to give two primary examples of how we supposedly misconstrued the available findings. One, which Bernardo has already dealt with at some length, concerns the observed sharp reductions by various agents of activity in brain networks associated with ordinary conscious awareness – in effect, the neural embodiment of the everyday self or ego. Amazingly to us, the researchers now want to back away from their own early findings of this sort, which as Bernardo shows again in his rejoinder are anything but subtle. The observed reductions, that is, are big effects, easily detected using conventional and straightforward analysis techniques. They satisfy what statisticians sometimes jokingly refer to as the IOT test (InterOcular Trauma – it socks you right between the eyes). Those results are certainly not going away.

Even relatively simple neuroimaging methods can easily distinguish between wakeful and drowsy states and other commonplace conditions, yet here we seem to be having difficulty finding anything in patterns of brain activity that is even remotely commensurate with the extraordinary phenomenology of psychedelic states.

Now consider by contrast their other example, which involves possible increases in other aspects of brain activity that they think might help us understand the extraordinary phenomenology of psychedelic states. They accuse us of misrepresenting the observed increases in multichannel MEG signal “diversity” produced by several agents as “small” effects. What really matters, they now declare, are the statistical significance and effect sizes (using Cohen’s d) of the observed increases. Those results, however, depend strongly upon the very large numbers of 2-second MEG data segments that enter into the analyses. The exact Ns are not explicitly stated, but given information provided elsewhere in the report they likely range from something like several hundred for the psilocybin data to several thousand for the ketamine data. These large numbers are what drive successful statistical discrimination of mean differences between drug and placebo conditions that in absolute magnitude are in fact appropriately described as “tiny” rather than merely “small”. Specifically, on a diversity scale of 0 to 100 they average on the order of .005 across all drug/placebo pairs, and occupy the interval between .98 and .99 (see Fig. 2 in Schartner et al., 2017). For several subjects the observed mean differences between drug and placebo conditions even go in the wrong direction.

Now, an enormous amount of unusually sophisticated computation went into the generation of these results, and to both Bernardo and me they seem puzzlingly meagre. After all, even relatively simple neuroimaging methods can easily distinguish between wakeful and drowsy states and other commonplace conditions, yet here we seem to be having difficulty finding anything in patterns of brain activity that is even remotely commensurate with the extraordinary phenomenology of psychedelic states. Our critics are not troubled by this; instead they go on to describe some of the relevant experiential properties and simply urge us to believe that they will all somehow, someday, be “accounted for” or “underpinned”, or “explained by” physical properties such as the “diversity” of the accompanying brain states.

Let me remark here parenthetically that I don’t think “diversity” in the sense of unpredictability or randomness in multichannel neuroelectric signals is by itself a very good candidate for such a property. To me the facile equation of brain signal diversity with experiential richness and complexity seems at best premature, and reminiscent of the many instances since publication of Shannon and Weaver’s original treatise on information theory in which cognitive scientists and neuroscientists have glibly conflated technical and everyday meanings of the word “information”.

Be that as it may, however, Bernardo and I can certainly agree with our critics that psychedelic neuroimaging research is at a very early stage in its development, and we certainly don’t profess to know how far it may ultimately get. Indeed, in the original blog we even offered a concrete suggestion of our own as to how to carry it further.

But our critics don’t know either, and here we arrive at the bottom-level disagreement. All of them are apparently conventional physicalists, and for persons of that persuasion it is simply axiomatic, a given, that everything in mind and consciousness must be manufactured by, or in some mysterious way identical to, neurophysiological events and processes in the brain. In our original blog we had provided indications that not only popular interpreters of psychedelic neuroimaging research but the researchers themselves were showing signs of bias in favor of this physicalist picture, and Bernardo has already demonstrated that this 6-person response to our blog amounts to more of the same.

Everything comes to a head in their final paragraph, where they now start by accusing us not merely of misrepresenting but of deliberately misrepresenting their findings in order to motivate an alternative non-physicalist view. That accusation is false and offensive. To set the record straight, Bernardo and I began our very different scientific careers as conventional physicalists ourselves, but we have long since changed our minds in light of arguments and evidence presented in voluminous detail in earlier publications. We approached psychedelic neuroimaging research as genuinely interested and admiring outside observers, viewing the subject through that independently justified lens, and we could not help noticing the evident tensions between physicalist expectations and the experimental results that had started coming in. As a scientist with long-time involvement in the field of psychical research, I also noticed that the most prominent reported physiological effect of psychedelics, the reduction or disruption of activity in networks that stabilize ordinary everyday consciousness, has parallels in recent neuroimaging research with advanced practitioners of various kinds of meditation, and that both of those situations could be viewed as steps on the way toward one of the main empirical challenges to physicalist neuroscience – namely, mystical-type near-death experiences (NDEs) occurring under extreme circumstances such as deep general anesthesia and/or cardiac arrest in which physiological conditions regarded by virtually all contemporary neuroscientists as necessary for consciousness have been abolished.

Our critics end by openly revealing themselves as guardians of the physicalist faith, intent upon defending it against potentially dangerous infidels. It seems to them that merely by entertaining any possibility of an alternative conceptual framework or worldview we are somehow putting at risk the legitimacy of the emerging neuroscience of consciousness. That in my opinion is patent nonsense. Their own efforts in psychedelic neuroimaging are certainly not at risk. Indeed, one of the most exciting developments in contemporary neuroscience is surely the rapid intensification of efforts to relate improved measures of brain activity to improved measures of the phenomenological character of associated states of consciousness, and psychedelic neuroimaging provides an ideal arena in which to pursue and deepen that inquiry precisely because of the extreme character of the states involved. Friendly criticism ought not be perceived as a threat. The research will certainly go on unabated, and all of us – we and our critics alike – are enthusiastically supportive of that. Let the empirical chips fall where they may.

Bernardo and I have evidently had no impact whatsoever on these six critics, at least not yet, but perhaps other readers will prove more open-minded.

Copyright © 2018 by Edward F. Kelly. Published with permission.