Olds, James (1922-1973)
OLDS, JAMES (1922-1973)
James Olds became a prominent figure in physiological psychology when he discovered, in 1953, that rats could be trained to perform a variety of experimental tasks, some at very high rates, in order to obtain a pleasurable electrical stimulus applied to a discrete central nervous system site. This procedure, often called self-stimulation or intracranial self-stimulation, is intimately related to the brain pathways that mediate positive reinforcement. At the time of his sudden death, Olds was Bing Professor of Behavioral Biology at the California Institute of Technology in Pasadena and a member of the National Academy of Sciences. The events leading up to the serendipitous discovery of pleasure centers in the brain were described both by Olds (1973b, 1975a, 1977) and others (Miller, 1980; Milner, 1989).
Olds was born in Northbrook, Illinois, where his father, Leland, was industrial editor for the Federated Press. The family moved in 1931 to Nyack, New York, where Olds's father headed the New York State Power Commission. In 1939 the family moved to Washington, DC, where the elder Olds served as commissioner and then chairman of the Federal Power Commission until 1949. Olds began his college career as an undergraduate at the University of Wisconsin in Madison. He transferred to St. John's College in Annapolis, Maryland, and then worked briefly for the International News Service before being drafted into the army during World War II. Trained in Arabic, he spent most of his war service in Cairo, Egypt, with the Persian Gulf Command. After returning to the United States at the end of the war, Olds transferred to Amherst College, of which his grandfather had been president, to finish his undergraduate studies.
Olds received his doctoral training in the department of social relations at Harvard University. A relatively new department at that time, it included social, experimental, and clinical psychology and sociology. Olds's mentor was Richard Solomon, who gave him thorough training in experimental psychology and exposed him to the current literature on physiological psychology. Olds had a part-time job as editor of a book by Talcott Parsons, chairman of the department, to supplement his graduate fellowship. Olds's contributions were so substantial that Parsons made him a coauthor. This was an unusual intellectual apprenticeship for someone who later made major contributions to the understanding the neural substrates of reward learning. His contact with Olds during subsequent years provided the biological perspective for Parsons's sociological theorizing.
Olds's thesis dealt with motivation and was influenced by Donald Hebb's landmark book, The Organization of Behavior (1949). He planned to train under Hebb as a first step toward developing a neural realization of a model of Edward Tolman's sign-gestalt theory that ideas determine behavior (Olds, 1954). The basic conviction underlying Olds's career plans at this stage was that behavior had to be explained in terms of underlying brain activity. Olds felt that the two principal problems of physiological psychology were motivation and learning, that these two problems were closely intertwined, and that their solution depended on a detailed knowledge of the central nervous system. These basic ideas motivated Olds's entire professional career.
After completing his doctoral thesis, Olds spent a year as a lecturer at Harvard and then moved on to McGill's psychology department to spend two years as a postdoctoral trainee under Hebb. During this period, Hebb's laboratory interacted extensively with the Wilder Penfield and Herbert Jasper groups at the Montreal Neurological Institute. When Olds arrived at McGill to begin his training, he was placed under the guidance of Peter Milner, who was then finishing his doctoral thesis. Milner taught him how to implant stimulating electrodes in the brain, and Olds prepared a rat with stimulating electrodes in what he assumed was the reticular formation, an area of considerable interest at that time. The hypothesis to be tested was that the animal would avoid stimulation in the reticular formation (assumed to be the source of neural activity to be reduced by behavior). But when the rat received stimulation through the implanted electrode, quite the opposite effect occurred. The rat actually was attracted to those locations that were associated with the stimulation. This observation in the very first rat he studied turned Olds's initial hypothesis on its head and also suggested that the brain contains reward or pleasure centers that mammals seek to activate during goal-seeking behaviors.
Some unsuccessful attempts to replicate the original electrode placement led to the determination that the hypothalamus and septal regions supported self-stimulation. Olds and Milner described their landmark discovery in a paper that appeared in 1954. It
[Image not available for copyright reasons]
is interesting to note Milner's description of the energy and organizational ability that Olds showed in fleshing out their original observation (Milner, 1989). Olds went on to perform many systematic studies using a standard and sensitive technique to determine which portions of the brain supported self-stimulation. He also developed new technical approaches and took advantage of new developments in the brain/behavioral sciences. This approach also characterized the contributions he made to the study of learning and memory.
Mapping the Brain for Learning
Olds's laboratory at Caltech in the early 1970s was the scene of a unique series of studies designed to "map the brain for learning" (Disterhoft and Buchwald, 1980; Olds, 1973a). These studies used a technique he had perfected for recording small groups of neurons in the freely moving rat (Olds et al., 1972; Olds, 1975b). The basic objective was to determine how neurons in various brain regions in animals learning the same associative response compared over time. The regions that changed earliest, by definition, were especially notable in the formation and readout of the learned response.
Studies of how the auditory system might change its processing of the tone-conditioned stimulus demonstrated that neurons from the inferior colliculus up to the auditory cortex showed alterations in firing rate during differential auditory conditioning. The research focused intensively on the posterior nucleus of the thalamus, a region that receives multisensory afferent drive and showed large firing-rate changes very early in the learning process. The involvement of the hippocampus, a region that Olds sometimes referred to as the "Rosetta Stone of the brain," also received considerable attention. Olds's previous mapping of the brain for self-stimulation had demonstrated that widely spread brain regions supported this phenomenon. Studies of neurons in many other regions of the brain, including the nonspecific thalamus, the basal ganglia, the reticular formation, and the hypothalamus showed altered firing rates during an important behavioral event such as auditory learning.
The series of studies done by Olds's group yielded some additional general contributions to the study of learning and memory. First, they delineated the advantage of using an apparently simple task as a "model system" to study mechanisms of learning in the mammalian brain. Second, they clearly demonstrated that many neurons in many brain regions do change during acquisition of even a simple learning task. These data firmly supported the idea that learning is truly a distributed process in the brain. Third, they demonstrated the advantages of formulating temporal maps of alterations in the brain during learning as a way to determine where important alterations were occurring (Olds et al., 1972). The temporal maps were drawn in the two dimensions: from conditioned-stimulus onset through conditioned-response performance and from the first training trial through acquisition and overlearning of the conditioned response. This approach allowed a clear ranking of the relative importance of the many alterations in single-neuron activity observed during learning in widely scattered brain regions. Finally, the studies clearly emphasized the importance and strength of an approach that dissected individual subsystems for detailed analysis during learning of a common task. Subsystem analysis allows a more precise delineation of the total system interactions.
The approach Olds designed and that his group used to such obvious advantage was adapted by many laboratories and is still being profitably used in the continuing search for mechanisms of learning and memory. For example, Berger and Thompson (1978) used it in their analysis of hippocampal-system activity during learning of the nictitating membrane response in the rabbit. Woody (1974) had begun studies using temporal analysis of the motor side of the eye-blink pathways in overtrained cats at about the time the Olds group's mapping studies began. More recent approaches have combined in vitro analyses of hippocampal brain slices with in vivo recording to ensure precise study of localized cellular and subcellular alterations in the hippocampus during eyeblink conditioning (Disterhoft and McEchron, 2000).
Considerable progress has been made in the analysis of the mechanisms of fear conditioning over the past two decades. The most successful studies of this phenomenon have used and extended the approach pioneered by Olds and his colleagues in their mapping studies (Davis and Whalen, 2001; Schafe et al., 2001). These studies of emotional learning have resonated among nonneuroscientists interested in brain mechanisms of learning and behavior (Damasio, 1994; LeDoux, 1996), imparting renewed impetus to Olds's position that learning and motivation are pivotal problems in physiological psychology (now subsumed under the more encompassing term neuroscience).
Setting the Stage for Future Research
Olds spent considerable time and effort setting the stage for subsequent laboratory work; his colleagues were intimately involved in the daily sequence of events that led to the important observations that ensued. Olds's approach to postdoctoral training was modeled after that of Donald Hebb—learning by doing. His experience in Hebb's laboratory, where he discovered self-stimulation as a postdoctoral fellow, clearly impressed upon him the value of serendipity, personal effort, and scientific tinkering. The series of brain-mapping studies used that general approach and remain landmarks in the study of associative learning in mammalian brain.
Olds was passionately dedicated to studying the functioning of the brain. His one frustration was that he felt he had not devoted enough time to studying the brain. He felt that the more facts he had at his disposal, the more likely it was that he could assemble them into compelling insights. He spent a good deal of time thinking, talking, and writing about how the brain works (Olds, 1975a, 1977, 1980).
Olds was also fascinated with computers and electronics. Many of his ideas about brain function, such as his speculations about memory-storage function in the hippocampus, used computers and their memories as analogies (Olds, 1969). He expended much effort on developing and testing software and hardware for the abundant computer equipment in his laboratory. His studies of associative learning used a combined hardware-software system simultaneously to study a large number of brain regions in animals learning the same relatively simple associative task.
The technical problem of developing better ways to study single neurons in conscious animals was one to which Olds devoted onsiderable energy. One of the reasons he came to Caltech was to take advantage of the possibilities for developing electronic gadgetry for his experiments. He collaborated with an electrical engineer from the Jet Propulsion Laboratory in the design and building of what must have been one of the earliest telemetry systems for multiple single-neuron recording. The idea was to transmit signals from ten microwire electrodes simultaneously without incurring the danger of cable artifacts. The rat looked a little ungainly with the miniature transmitter on its head (considerably less miniature in 1971 than it would be today), but the system worked pretty well. Olds was always trying to come up with a better operational amplifier than the ones he was using, although the old standbys often worked better. He also was involved in troubleshooting things like electronic waveform identifiers; he wanted his to work more simply and efficiently. He would hook up a rat, sit down in front of an oscilloscope, and run a new waveform identifier through its paces by himself. He knew from experience that often the product did not work as precisely as the engineerintended.
The portion of the laboratory used for the mapping of learning was set up with four training stations. Olds used a recording station of his own and carried on a series of experiments separate from those of the postdoctoral fellows and graduate students. He traveled extensively, but when he was in town, he came in every morning to check the rat-in-training in his station and to check the setting of the waveform discriminators on the unit channels. Olds was very demanding about the quality of the data he and his group gathered. He was a firm believer that high-quality findings came from high-quality data. His system had numerous checks for electronic noise and other artifacts. He also took an intense interest in the experiments as they were being run. Those with freely moving rats ran from evening until early morning, the peak of the rats' diurnal cycle. A collaborator coming in during the evening to check experimental progress was likely to discover that Olds had been there shortly before. Olds was almost always in the laboratory early on Sunday to make notes on the printout or adjustments to the computer.
Olds and his collaborators met in his office every afternoon to discuss the data and their interpretation. These meetings often included theoretical discussions that ranged far from the data at hand and discussions of appropriate strategies to use in current or future experiments. Descriptions of Olds's scientific activities by Neal Miller (1980) and Peter Milner (1989) confirm Old's lifelong eagerness to share ideas with colleagues and students and to approach brain function with novel perspectives.
Olds was a gracious, urbane man with a good sense of humor. He was also a family man. His wife, Marianne, worked closely with him in the lab. He also lavished considerable attention on his son, James, who was in high school during the time Olds was on the Caltech faculty.
Berger, T. W., and Thompson, R. F. (1978). Neuronal plasticity in the limbic system during classical conditioning of the rabbit nictitating membrane response. I. The hippocampus. Brain Research 145, 323-346.
Davis, M., and Whalen, P. J. (2001). The amygdala: Vigilance and emotion. Molecular Psychiatry 6, 13-34.
Damasio, A. R. (1996). Descartes' error: Emotion, reason and the human brain. New York: G. P. Putnam.
Disterhoft, J. F., and Buchwald, J. S. (1980). Mapping learning in the brain. In A. Routtenberg, ed., Biology of reinforcement. New York: Academic Press.
Disterhoft, J. F., and McEchron, M. D. (2000). Cellular alterations in hippocampus during acquisition and consolidation of hippocampus-dependent trace eyeblink conditioning. In D. Woodruff-Pak and J. E. Steinmetz, eds., Eyeblink classical conditioning, Vol. 2: Animal models. Norwell, MA: Kluwer.
Hebb, D. O. (1949). The organization of behavior. New York: John Wiley.
LeDoux, J. (1998). The emotional brain: The mysterious underpinnings of emotional life. New York: Simon and Schuster.
Miller, N. E. (1980). Introduction: Brain stimulation reward and theories of reinforcement. In A. Routtenberg, ed., Biology of reinforcement. New York: Academic Press.
Milner, P. M. (1989). The discovery of self-stimulation and other stories. Neuroscience and Biobehavioral Reviews 13, 61-67.
Olds, J. (1954). A neural model for sign-gestalt theory. Psychological Review 61, 59-72.
—— (1969). The central nervous system and the reinforcement of behavior. American Psychologist 24, 114-132.
—— (1973a). Brain mechanisms of reinforcement learning. In D. E. Berlyne and N. B. Madsen, eds., Pleasure, reward, preference. New York: Academic Press.
—— (1973b). Commentary on Olds and Milner's "Positive reinforcement produced by electrical stimulation of septal area and other regions of rat brain." In E. S. Valenstein, ed., Brain stimulation and motivation: Research and commentary. Glenview, IL: Scott, Foresman.
—— (1975a). Mapping the mind onto the brain. In F. G. Worden, J. P. Swazey, and G. Adelman, eds., The neurosciences: Paths of discovery. Cambridge, MA: MIT Press.
—— (1975b). Unit recordings during Pavlovian conditioning. In N. A. Buchwald and M. A. B. Brazier, eds., Brain mechanisms in mental retardation. New York: Academic Press.
—— (1977). Drives and reinforcements: Behavioral studies of hypothalamic functions. New York: Raven Press.
—— (1980). Thoughts on cerebral functions: The cortex as an action system. In A. Routtenberg, ed., Biology of Reinforcement. New York: Academic Press.
Olds, J., Disterhoft, J. F., Segal, M., Kornblith, C. L., and Hirsh, R. (1972). Learning centers of rat brain mapped by measuring latencies of conditioned unit responses. Journal of Neurophysiology 35, 202-219.
Olds, J., and Milner, P. M. (1954). Positive reinforcement produced by electrical stimulation of septal area and other regions of rat brain. Journal of Comparative and Physiological Psychology 47, 419-427.
Schafe, G. E., Nader, K., Blair, H. T., and LeDoux, J. E. (2001). Memory consolidation of Pavlovian fear conditioning: A cellular and molecular perspective. Trends in Neuroscience 24, 540-546.
Woody, C. D. (1974). Aspects of the electrophysiology of cortical processes related to the development and performance of learned motor responses. The Physiologist 17, 49-69.