Rosenblatt's Contributions

In contrast to Wikipedia's characterization of Frank Rosenblatt as a handsome bachelor driving a classic sports car around the Cornell campus, those who knew him would consider him a rather shy genius and more of a Renaissance man because he excelled in a wide variety of subjects, including psychology (his original field), computing, mathematics, neurophysiology, astronomy, and music.

The Mark I Perceptron
Rosenblatt's early work on perceptrons at the Cornell Aeronautical Laboratory (1957-1959) culminated in the development and hardware construction of the Mark I Perceptron. Mark I, a visual pattern classifier, had an input (sensory) layer of 400 photosensitive units in a 20x20 grid modeling a small retina, an association layer of 512 units (stepping motors) each of which could take several excitatory and inhibitory inputs, and an output (response) layer of 8 units. The connections from the input to the association layer could be altered through plug-board wiring, but once wired they were fixed for the duration of an experiment. The connections from the association to the output layer were variable weights (motor-driven potentiometers) adjusted through the perceptron error-propagating training process. Mark I consisted of 6 racks (approximately 36 square feet) of electronic equipment and numerous experiments were conducted on this machine (can we describe some?). When Dr. Rosenblatt moved his research to Cornell University in 1959, Mark I was shipped to the Office of Naval Research (the funding agency) in Washington D.C. for demonstration purposes but was returned to Cornell rather than to the Aeronautical Lab. The Mark I Perceptron currently resides in the Smithsonian Institute.

The Tobermory Perceptron
The Tobermory Perceptron was designed to recognize speech and was named after a talking cat, Tobermory, in a story by H.H. Monroe (aka Saki). It was a large machine consisting of 45 band-pass filters and 80 difference detectors as sensory units, 1600 A1-units (20 time samples per detector), 1000 A2-units, and 12 R-units, with 12,000 adaptive weights between the A2-units and R-units.

Perceptron Computer Simulations
Whereas hardware implementations were good for demonstrations and for proving feasibility, software simulations were far more flexible. In the late 1950s and early 1960s, however, computer simulations of perceptrons required machine language coding for reasonable speed and memory usage. Nevertheless, Rosenblatt initiated work on a large perceptron simulation software package in which the user could specify the number of layers, the number of units in each layer, the types of connections between layers, etc. In addition to computer time at Cornell, Rosenblatt was provided free time on the computer at NYU and trips were made into NYC to use the time. The results of these simulations were presented at conferences and summarized in Rosenblatt's book.

Rosenblatt's Book
The following is from the Preface of Frank Rosenblatt's book, Principles of Neurodynamics: Perceptrons and the Theory of Brain Mechanisms, Spartan Books, 1962.

It is only after much hesitation that the writer has reconciled himself to the addition of the term "neurodynamics" to the list of such recent linguistic artifacts as "cybernetics, "bionics", "autonomics", "biomimesis", "synnoetics", "intelectronics", and "robotics". It is hoped that by selecting a term which more clearly delimits our realm of interest and indicates its relationship to traditional academic disciplines, the underlying motivation of the perceptron program may be more successfully communicated. The term "perceptron", originally intended as a generic name for a variety of theoretical nerve nets, has an unfortunate tendency to suggest a specific piece of hardware, and it is only with difficulty that its well-meaning popularizers can be persuaded to suppress their natural urge to capitalize the initial "P". On being asked, "How is Perceptron performing today?" I am often tempted to respond, "Very well, thank you, and how are Neuron and Electron behaving?"
For this writer, the perceptron program is not primarily concerned with the invention of devices for "artificial intelligence", but rather with investigating the physical structures and neurodynamic principles which underlie "natural intelligence". A perceptron is first and foremost a brain model, not an invention for pattern recognition. As a brain model, its utility is in enabling us to determine the physical conditions for the emergence of various psychological properties.

Rosenblatt's book summarized his work on perceptrons at the time. The book is divided into four parts. The first gives an historical review of alternative approaches to brain modeling, the physiological and psychological considerations, and the basic definitions and concepts of the perceptron approach. The second covers three-layer series-coupled perceptrons: the mathematical underpinnings, performance results in psychological experiments, and a variety of perceptron variations. The third covers multi-layer and cross-coupled perceptrons, and the fourth back-coupled perceptrons and problems for future study. Rosenblatt used the book to teach an interdisciplinary course entitled "Theory of Brain Mechanisms" that drew students from Cornell's Engineering and Liberal Arts colleges.

Rosenblatt-Minsky Debates and Minsky-Papert Book
During the late 1950s and early 1960s, much to the enjoyment of those in the audience, Rosenblatt and Minsky debated on the floors of scientific conferences the value of biologically inspired computation, Rosenblatt arguing that his neural networks could do almost anything and Minsky countering that they could do little. Minsky, wanting to decide the matter once and for all, began collaborating with Seymour Papert in the mid 1960s, and after much delay they published a book in 1969, Perceptrons: An Introduction to Computational Geometry, where they asserted about perceptrons (page 4), "Most of this writing ... is without scientific value..." Although Minsky was well aware that the powerful perceptrons have multiple layers and that Rosenblatt's basic feed-forward perceptrons have three layers, he defined a perceptron as a two-layer machine that can handle only linearly separable problems and, for example, cannot solve the exclusive-OR problem. Although considered a political maneuver (hatchet job) for contract funding by some knowledgeable scientists, this strong criticism of perceptrons essentially halted work on neural-based devices for a decade. [Note: Minsky and Rosenblatt knew of each other from their days at the Bronx High School of Science, graduating respectively in 1945 and 1946.]

Rosenblatt built a modest observatory on a hilltop behind his house in Brooktondale about 6 miles east of Ithaca. Work began in the summer of 1961 and he bought a Fecker 12" cassegrain telescope. About the time that construction began, Frank had considerable interest in SETI (Search for Extraterrestrial Intelligence) and wrote a grant proposal touting a "Stellar Coherometer" that he had designed. Through his connections, probably at NASA, he was awarded $75K for the project if Cornell would administer the grant. Because Cornell was not interested in being associated with such a project, Rosenblatt went to Ithaca College and they agreed to administer the project. However, by the time he secured the required additional appointment as Associate Professor at Ithaca College the grant period for that fiscal year elapsed and the $75K reverted to the funding agency. The observatory was finally completed about 1966. It is a circular cinderblock structure with a dome mounted on top to house the telescope. The structure was built mostly by himself with some help from his students and friends.

Rosenblatt was an accomplished pianist and bought a grand piano for his house in Brooktondale. Although he played the well-known classical pieces of Mozart, Beethoven, etc., he had a penchant of improvising endlessly on "Three Blind Mice".

Practical Joker
Frank did both his undergraduate and graduate studies at Cornell which was not a common practice at the time. He was a psychology major and Professor James Gibson was a well-known member of the department. As the story goes, Frank and some other graduate students drove to the town of Gibson one night and stole the town's "Gibson" signs, which they then mounted at the door of Professor Gibson's office in Morrill Hall. When Department Chair Professor Robert MacLeod saw them, he remarked to the department secretary, "Don't you think Gibby's getting a little ostentatious?"

Rat Brain Experiments
In the late 1960s Rosenblatt began experiments within the Cornell Department of Entomology on the transfer of learned behavior via rat brain extracts. Rats were taught discrimination tasks such as Y-maze and two-lever Skinner box, their brains extracted and injected into untrained rats that were then tested in the discrimination tasks to determine whether or not there was behavior transfer from the trained to the untrained rats. Rosenblatt spent his last several years on this problem and showed convincingly that the initial reports of larger effects were wrong and that any memory transfer was at most very small.

Rosenblatt's Untimely Death
Frank Rosenblatt died in a boating accident on his 43rd birthday. He "was a most gifted human being ... had made his entire life a contribution to mankind" [Tribute to Dr. Frank Rosenblatt, U.S. Congressional Record].

Rosenblatt's Publications
See partial list of Rosenblatt's publications.