Bela Julesz

Bela Julesz

He addressed this question, as he did everything in his life, uniquely. While the mid-20th century still had its share of philosophers and talkers on such questions, Bela investigated the mental representations of external reality by fully implementing the experimental program first devised by Gestalt philosophers and—this is especially crucial—by informing it with the most current neuroscience.

Bela Julesz died on 31 December 2003 in New Jersey—47 years to the day after he began at Bell Laboratories, 47 years of providing guidance and questions to the burgeoning field of neuroscience, 47 years as a physicist of the mind.

Bela was born in Budapest, Hungary, on 19 February 1928. It is not clear how his early years formed the foundation of an interdisciplinary scientist. From his doctorate at the Hungarian Academy of Sciences, he certainly understood physics and complex engineering. When the Soviet tanks rolled into Hungary in 1956, he and his wife Margit escaped, swimming the Danube to the West.

The first of the two positions he held in the US were at AT&T Bell Laboratories. Manfred Schroeder hired Bela to continue his doctoral studies on television signals. Yet Bela's unique approach to science rapidly became apparent. From studies by John W. Tukey, Harry Nyquist, and Claude Shannon on random number generators, Bela was given the task of assessing binary sequences for randomness. Rather than use numerical measures, he reached across domains and made images of the sequences using the unique pattern-recognizing ability of our visual system to evaluate randomness. One will never know how long that approach had incubated, or from whence it came. I prefer to think it was Bela's innate ability to rapidly assess problems from conventional viewpoints and then let his mind roam afield, bringing in his own experiences and unusual knowledge for resolution.

From that simple exploitation of the visual system to solve a mathematical problem, Bela addressed the question of mind and brain. He selected visual illusions in which the internal perceptions were not identical to the visual inputs on the retina. His first great success came with the invention of the random dot stereogram.

The random dot stereogram resolved a fundamental conflict between Charles Wheatstone and David Brewster from the mid-19th century. Bela computed two identical pairs of random dot images that lacked any identifiable cues; a square patch of dots in one image was displaced slightly. When viewed monocularly, each image lacked global structure. But viewed through a stereoscope, so that each eye saw one image, a central square emerged in depth. Those displays—eventually known as Julesz random dot stereograms, or RDSs—and subsequent explorations described in his first monograph, Foundations of Cyclopean Perception (U. of Chicago Press, 1971), established the primacy of stereoscopic vision to shape and form vision. Bela saw stereoscopic fusion as analogous to alignment of magnetic dipoles—one of his many physical analogies for perception.

Importantly for the field of vision neuroscience, in its toddler stage at that time, RDS provided a paradigm to study higher cognitive function. A behavioral test is created for which the internal representation differs from the external world; the mental processes are then explored using physical approaches embedded in the brain's anatomy and physiology. Bela termed this approach "psychoanatomy." Although the term never stuck, the approach is evident throughout contemporary neuroscience. In experiments ranging from single neuron recordings in monkeys to functional magnetic resonance imaging in humans, the modus operandi is to locate a physical neural correlate in response to an external physical event that is not a simulacrum. The "ghost in the machine," the dualist nature of mind and brain, is finally excised from any scientific thinking about the mind. This is Bela's legacy.

In his later years, through his second book, Dialogues on Perception (MIT Press, 1995), one got a picture of how his particular mind worked. It was as if there were two people, constantly trying to best each other. Two opposing viewpoints could then be held in his head and addressed. Bela's science, in which he appreciated a reductionist physical approach to the function of the most complex mechanisms of brain, also could endorse a synthetic model of brain with emergent behavior.

Through the 1980s at Bell, Bela developed a powerful research group and worked exclusively with postdoctoral trainees and faculty. Before retiring from Bell in 1989, he chose to move to Rutgers University, where he was awarded the highest academic position, the New Jersey State University Professorship, to establish the Laboratory of Vision Research. There, he constructed an international team to continue examining stereovision, texture, form, color, attention, and motion.

Bela received many awards, including a MacArthur fellowship from the John T. and Catherine D. MacArthur Foundation in 1983. He used the MacArthur money to travel and spent significant time at Caltech, where he lectured to students and faculty.

An approachable man, Bela had an inexhaustible collection of anecdotes. In his last years, his health deteriorated. He would rally for visitors, and his wit and questing intelligence continued to be evident. In 2003, he had begun a second edition of Cyclopean Perception. Knowing Bela, that text certainly would have raised more questions about the physical nature of brain and mind than it answered.