Postscript

When frustrated, the inordinate expectations urged on us by neuroscience enthusiasts often induce us to undervalue the contributions and the rate of progress of functional neuroimaging. But were we to adopt the dispassionate perspective of the historian of science, we would marvel at the tremendous accomplishments of that discipline that has yet to reach the age of maturity. Only four short decades ago we had no real access to the normal living brain and we were trying, most often in vain, to estimate what was transpiring and where, inside the skull, with surface electrophysiological measurements: electroencephalogram (EEG) and evoked and event-related potentials. Now we can not only discern with exquisite spatial resolution minute details in brain structures and assess the integrity of their connections, but we can also localize with remarkable accuracy nodes or hubs of the primary sensory and motor networks and, with increasing certainty, those of language and other cognitive functions. So much so, in fact, that such estimates can now be and often are used for the evaluation of brain surgery candidates. Yet we are not satisfied, and we shouldn’t be until we can specify the mechanisms of all neurologically valid cognitive operations to the degree of precision that our models of the same operations allow and prescribe. For that, however, much time is needed. But what is needed mostly are some corrections of the trajectory of our research activities, the necessity of which has been alluded to in various sections in this Handbook and will be summarized here in a moment.

We became aware of the need for such corrections when confronted with the vast literature on intriguing topics only to realize how very limited was the set of securely established facts that could be extracted from them to be included in this Handbook. The first correction that it is within our means to implement is to strive for reproducibility of findings and to moderate the tendency to always study new issues and publish what “has never been done before” when much more fundamental issues remain unresolved, and they are many. We have, for example, the discrepancy between the phenomenon of the “hyperfrontal activation pattern” and that of the “default mode network.” We still do not know whether the hippocampus is or is not always involved in encoding and consolidation operations when all other evidence besides that of neuroimaging urge that it always is. This tendency to always opt for the new and the intriguing can easily be moderated since the responsibility for it rests squarely with us, the investigators, in our capacity as editors, consultants, and reviewers for journals or funding institutions or as members of promotion and tenure committees.

A second correction we are capable of implementing is to attend as carefully to the theoretical propositions we wish to test empirically as we do to the technical aspects of recording brain activity and activation and to the algorithms for estimating from these recordings the nature and topography of the intracranial events they reflect, although both the recording techniques and the algorithms also call for improvements. Unless we outgrow the tendency to call and treat as “models” sets of often unconnected conjectures and strive to construct cohesive and testable models instead, no real progress can be made no matter how sophisticated the neuroimaging technology becomes. Unless vague terms (which we also have used with abandon in this Handbook) such as “processing” this or that or “involvement” of this region or that in this or that process are replaced with operationally specified concepts; unless obvious distinctions such as that between activation due to an operation and activation corresponding to the product of the operation (e.g., the experience of the meaning of a word) are appreciated so as to inform the research designs, progress in reaching the ambitious goals we are setting will be unnecessarily retarded.

The consumers of the neuroimaging literature to whom this Handbook is addressed can also aid in correcting the course of the research in this area if they would only demand the same rigor from it as they do of psychophysics, of interventional physiology, and of all other better-established and productive scientific endeavors. And it is also with this purpose in mind and not only with the purpose of offering the consumer of this literature the means for evaluating it critically that this Handbook was designed and composed.

Andrew C. Papanicolaou