"Data-Rich, but Theory-Poor"
It’s often said neuroscience is “data-rich, but theory-poor.” Francis Crick, famous for his part in discovering DNA, went even further, stating neuroscience lacked a theoretical framework entirely. In the words of Thomas Kuhn, neuroscience remains in its pre-paradigm stage. Despite the tremendous progress in the study of cognition, the question has to be asked: “Why do we not have a good brain theory yet?”
While every neuroscientist will attest to the lack of a theoretical framework in neuroscience, the reasons as to why vary. Some argue we still need more research and more data in order to develop an adequate brain theory. This reasoning suggests there is still not quite enough information about the anatomy and physiology of the brain in order to propose a broad theoretical framework. However, we just noted the vast amount of data about the brain available to man, so it’s unlikely continued research will suddenly result in neuroscience’s first paradigm. It’s improbable to assume another 20 years of data collection will eventually result in a theoretical framework that helps to interpret and integrate a myriad of observational and experimental data into a comprehensive theory when the past 20 years of research have failed to suggest any sort of theory.
Jeff Hawkins, founder of the Redwood Institute for Theoretical Neuroscience, suggests there is another reason as to why we have yet to develop a theoretical framework to understand existing data. Hawkins claims an “intuitive, strongly-held, but incorrect assumption” has blinded neuroscientists in their efforts to develop a working theory of the brain. Hawkins states this intuitive yet incorrect assumption is that intelligence is defined by behavior. Instead, Hawkins suggests intelligence is defined by prediction. As Hawkins points out, IQ testing relies heavily on prediction, testing one’s ability to predict the next number in a sequence and recognize and recall relationships. This simple misunderstanding of how intelligence is judged, Hawkins argues, has blinded neuroscientists and researchers in their efforts to propose such a framework.
Now that we have an understanding as to why we don’t have a good brain theory yet, we have yet to answer the question: “Why is having a good brain theory necessary?” First and foremost, developing a strong theoretical framework is necessary to interpret much of the still unexplained data in neuroscience, similar to what Copernicus’ heliocentric theory did for our understanding of the solar system or what Darwin’s theory of natural selection did for our understanding of evolution. It is important to note Copernicus’ and Darwin’s respective theories were not the first paradigms in their respective fields. Both the study of the solar system and evolution experienced numerous paradigm shifts and it was only after a series of successive iterations were we able to discover the truth about each Conversely, neuroscience has yet to have its first paradigm. Interestingly enough, the framework of ideas Copernicus and Darwin proposed were rooted in mathematical models and empirical evidence, methods neuroscience has only recently embraced.
Perhaps even more important than helping to interpret existing data, science can sometimes tell us something about ourselves. In this case, it can tell us about who we are. It can tell us about how we perceive our surroundings and about how we learn. It can offer an explanation as to why we think and behave the way we do. And every so often science contributes to the advancement of the human race. It helps society move forward and further the range of possibilities for humanity. Artificial intelligence, the primary application of the research in theoretical and computational neuroscience, is capable of such scale and impact. The use of intelligent machines and supercomputers has the potential to not only transform our lives but broaden our intellectual horizons and help create a more intelligent tomorrow.