Psychology

Share this article:

Psychology

  • Join our comunity:

Computational Neuropsychology, Studying Emergence

By: , Posted on: January 16, 2015

Warren Tryon Collecting WoodIt is not enough to call for the study of emergence as I have done in my appeal for a paradigm shift in my book, Cognitive Neuroscience and Psychotherapy: Network Principles for a Unified Theory and in some of my previous blogs. One must also provide tools and some direction for using them to get the ball rolling.

Every parallel-distributed processing connectionist neural network (PDP-CNN) model that I have encountered has focused on the properties of the fully trained “adult” model rather than the process by which these properties emerged. This is because the authors of these simulations have presented their models as demonstration proofs that artificial neural networks are capable of performing certain functions. I agree that this is a necessary first step. It would be premature to study the emergent process unless, or until, one first demonstrated that the network in question is capable of generating the desirable psychological properties. But now that so many psychological and behavioral phenomena have been effectively simulated using PDP-CNN models, it is time to ask how these properties emerge. This line of inquiry is needed to generate full scientific explanations of these psychological and behavioral phenomena.

The following proposed method for studying emergence was inspired by Gazanniga’s (2011, p. 125) position that emergence is a type of phase change that entails reorganization. The physical phase changes that occur when ice melts into water and/or when water vaporizes into steam can be explained by molecular reorganization that changes the physical structure of water and gives rise to the emergent properties associated with each phase state. I suspect that the functional properties that artificial and real neural networks develop as a result of training also entail reorganization. I refer to synaptic reorganization of the connectome. We need to study how this reorganization occurs and the principles that govern it to advance our scientific explanations of psychology and behavior. Next, I propose a methodology for such study.

Networks are initialized by assigning small random values to the connection weights that simulate the state of synapses that connect the neurons to one another. Positive values represent, simulate, excitation. Negative values represent, simulate, inhibition. One can color the excitatory connections green and the inhibitory connections red. The structure of the network can be visualized by creating a 3D image where the distance among the simulated neurons is directly proportional to the strength of these positive and negative connections. A simple three layered system could be represented by placing simulated neurons along three orthogonal axes. More complex neural architectures could be represented by adding another dimension for each additional layer. The problem here is that it would soon become difficult to visualize the resulting structure. Alternatively, all simulated neurons could be randomly assigned X, Y, and Z coordinates normalized to create a sphere. Software currently exists that can rotate objects around each of the three axes thereby enabling inspection of the initial network structure from any angle. This initial network structure is associated with the pre-emergent state of not being able to properly perform the designated functions.

Training changes the connection weights. This constitutes a reorganization of the connectome. This reorganization can be visualized. The positive weights that become more positive can be represented by longer green lines. The negative weights that become more negative can be represented by longer red lines. In both cases the distance between processing nodes, simulated neurons, increases and thereby changes the 3D structural representation of the network’s connectome. The positive weights that become less positive, smaller, can be represented by shorter green lines. The negative weights that become less negative, smaller, can be represented by shorter red lines. In both cases, the distance between processing nodes decreases. Two processing nodes overlap as their connection weights approach zero. In these cases the decreased distance between processing nodes, simulated neurons, also changes the 3D structure of the network’s connectome. In sum, the connectome reorganization that results from training is revealed as structural changes of this sphere or image.

I propose the following method to track and study this reorganization process. I recommend that one color image, as described above, be constructed after each processing cycle, or set of processing cycles, and displayed as a movie to better visualize the dynamic structural changes, reorganization, that occur as the network learns and the desired functionality emerges from the artificial neural network under study. This movie can be sped up or slowed down to any desired degree in order to better understand how emergence occurs.  Each frame can be rotated along any or all three spatial axes in order to better understand how emergence occurs. One could select a viewing orientation prior to playing this “movie” as the result may depend upon the viewer’s perspective. In sum, this method may reveal how connectome reorganization enables psychology and behavior to emerge from neural networks.

Any of the many published neural network models could be used to study emergence. They all involve memory. We now understand learning as memory formation (Carlson, Miller, Heth, Donahoe, & Martin, 2010, p. 440). Learning does not occur if memory does not form because our measures of learning all require some degree of retention. This better understanding of how emergence provides a gateway to all psychological phenomena because memory and learning provide the foundation, basis, for everything psychological.

Cognitive Neuroscience and Psychotherapy coverRead more from Warren Tryon:

Warren’s book, Cognitive Neuroscience and Psychotherapy: Network Principles for a Unified Theory is available for purchase on the Elsevier Store. Use discount code “STC215” at checkout and save up to 30% on your very own copy.

About the Author

Warren ComputerWarren W. Tryon received his undergraduate degree from Ohio Northern University in 1966. He was enrolled in the APA approved Doctoral Program in Clinical Psychology at Kent State University from 1966 – 1970. Upon graduation from Kent State, Dr. Tryon joined the Psychology Department faculty at Fordham University in 1970 as an Assistant Professor. He was promoted to Associate Professor in 1977 and to Full Professor in 1983. Licensed as a psychologist in New York State in 1973, he joined the National Register of Health Service Providers in Psychology in 1976, became a Diplomate in Clinical Psychology from the American Board of Professional Psychology (ABPP) in 1984, was promoted to Fellow of Division 12 (Clinical) of the American Psychological Association in 1994 and a fellow of the American Association of Applied and Preventive Psychology in 1996. Also in 1996 he became a Founder of the Assembly of Behavior Analysis and Therapy.

In 2003 he joined The Academy of Clinical Psychology. He was Director of Clinical Psychology Training from 1997 to 2003, and presently is in the third and final year of phased retirement. He will become Emeritus Professor of Psychology in May 2015 after 45 years of service to Fordham University. Dr. Tryon has published 179 titles, including 3 books, 22 chapters, and 140 articles in peer reviewed journals covering statistics, neuropsychology, and clinical psychology. He has reviewed manuscripts for 45 journals and book publishers and has authored 145 papers/posters that were presented at major scientific meetings. Dr. Tryon has mentored 87 doctoral dissertations to completion. This is a record number of completed dissertations at the Fordham University Graduate School of Arts and Sciences and likely elsewhere.

His academic lineage is as follows. His mentor was V. Edwin Bixenstein who studied with O. Hobart Mowrer at the University of Illinois who studied with Knight Dunlap at Johns Hopkins University who studied with Hugo Munsterberg at Harvard University who studied with Wilhelm Wundt at the University of Leipzig.

Cognitive Neuroscience and Psychotherapy: Network Principles for a Unified Theory is Dr. Tryon’s capstone publication. It is the product of more than a quarter of a century of scholarship. Additional material added after this book was printed is available at www.fordham.edu/psychology/tryon. This includes chapter supplements, a color version of Figure 5.6, and a thirteenth “Final Evaluation” chapter. He is on LinkedIn and Facebook. His email address is wtryon@fordham.edu.

This blog and all others by Dr. Warren Tryon can be found on his Fordham faculty webpage.

References

  • Carlson, N. R., Miller, H., Heth, C. D., Donahoe, J. W., & Martin, G. N. (2010). Psychology: The science of behavior (7th ed.), Boston: Allyn & Bacon.
  • Gazzaniga, M. S. (2011). Who’s in charge? Free will and the science of the brain. New York: HarperCollins.

Connect with us on social media and stay up to date on new articles

2 thoughts on “Computational Neuropsychology, Studying Emergence

Comments are closed.

Psychology

Researchers and clinicians in psychology work across a vast array of sub-disciplines, including applied psychology, addictions, cognitive psychology, developmental and educational psychology, experimental physiological psychology, forensic psychology, neuropsychology, and behavioral and cognitive therapy. For these professionals, and students as well, cross-disciplinary study is a given. For more than 75 years, Elsevier has cultivated portfolios of psychology books, eBooks, and journals covering current and critical issues in all of these areas. This vital content provides a sound basis of understanding for all those involved in this multi-faceted field.