The study of learning has a rich tradition, going back at least to the days of Aristotle, who proposed that the formation of associations between coincident events is the way humans learn. More famously associated with learning, especially in the popular mind, is Pavlov who, in the 1920s, studied what is now known as classical conditioning or Pavlovian conditioning. In these well-known experiments, an unconditioned stimulus (food), which naturally causes an unconditioned reflex (salivation), was presented along with a neutral stimulus (a bell) with enough repetition that, eventually, the bell began to evoke the salivation even without the presence of food (conditioned response). At that point the bell had become a conditioned stimulus, i.e., a stimulus that, after learning a new association, evokes the conditioned response. Classical conditioning refers to an environmental stimulus that can elicit a response, but it does not address the ways in which an animal might learn how its own behavior could cause an environmental response. Work by Thorndike in the late 1890s began to address this by studying what he termed instrumental conditioning. Thorndike observed that a hungry cat could learn through trial and error that rubbing up against the side of its cage would open a latch allowing it access to food. From this he proposed his law of effect, in which he argued that the tendency to repeat a behavior is dependent on the consequences that behavior evokes. In a series of studies from the 1930s through the 1950s, Skinner followed up on Thorndike’s work, studying primarily pigeons and rats in a wide variety of conditions, including positive reinforcement (in which a behavior is more likely to recur if it is followed by a reward such as a piece of food), negative reinforcement (a behavior becoming more likely if it is followed by the removal of an aversive stimulus such as an electric shock). Skinner renamed this paradigm operant conditioning because a spontaneously emitted behavior (or operant) is what elicits the response. Although the rich psychological history of studying classical conditioning, operant conditioning, and other forms of learning has led to greater understanding of the phenomenology of learning, less is known about the neurophysiological mechanisms of these forms of learning. This is because most studies of cortical and subcortical function have involved either functional imaging, in which case the spatial resolution is too gross to permit the study of precise mechanisms of neuronal learning, or acute single-electrode recording methods, which are inherently limited in their ability to examine functional interrelationships between various cortical areas or to study any changes in neuronal firing that might occur over days or weeks. This inability to track changes across cortical areas has meant that most studies have used highly trained animals who have reached a stable level of performance on a previously learned task, rather than animals learning a new task. Multielectrode ensemble recording, however, provides the ability to record from a large number of cells simultaneously. The mammalian brain contains many millions of neurons that work in an interconnected manner to produce complex behaviors and thoughts. Understanding the interrelations between many neurons that make up a functional circuit, therefore, requires simultaneous recording from many more than one at a time. Furthermore, because the brain’s encoding of a given event (be it sensory, motor, or cognitive) relies on complex interactions between neurons, our ability to understand fundamental neural-circuit mechanisms is greatly improved when one simultaneously records the firing of many neurons, rather than just a single one at a time. Thus, multielectrode recording brings us a step closer to understanding normal brain function. Another important advantage of multielectrode recording is that it provides a more random sample of the neurons in the implanted area, obviating a priori decisions about the cell types of interest and permitting comparison of the contributions of different neurons to the encoding of, for example, a given motor action. Finally, and most important for the study of learning, chronic implantation of multielectrode arrays allows us to study ongoing processes that take more than a single session to complete. Thus, multielectrode arrays may be implanted in a naïve animal and recordings made throughout the course of learning. This allows the sampling of many neurons each day of the study, even if it lasts weeks, months, or years.
Copyright © 2008, Taylor & Francis Group, LLC.
Visualizing Adult Cortical Plasticity Using Intrinsic Signal Optical Imaging.In: Frostig RD, editor. In Vivo Optical Imaging of Brain Function. 2nd edition. Boca Raton (FL): CRC Press/Taylor & Francis; 2009. Chapter 9. In Vivo Optical Imaging of Brain Function. 2nd edition. 2009. PMID: 26844321 Free Books & Documents. Review.
Multielectrode Recordings in the Somatosensory System.In: Nicolelis MAL, editor. Methods for Neural Ensemble Recordings. 2nd edition. Boca Raton (FL): CRC Press/Taylor & Francis; 2008. Chapter 6. Methods for Neural Ensemble Recordings. 2nd edition. 2008. PMID: 21204443 Free Books & Documents. Review.
In Vivo Observations of Rapid Scattered Light Changes Associated with Neurophysiological Activity.In: Frostig RD, editor. In Vivo Optical Imaging of Brain Function. 2nd edition. Boca Raton (FL): CRC Press/Taylor & Francis; 2009. Chapter 5. In Vivo Optical Imaging of Brain Function. 2nd edition. 2009. PMID: 26844322 Free Books & Documents. Review.
Recording human electrocorticographic (ECoG) signals for neuroscientific research and real-time functional cortical mapping.J Vis Exp. 2012 Jun 26;(64):3993. doi: 10.3791/3993. J Vis Exp. 2012. PMID: 22782131 Free PMC article.
A comparison of neuronal reactions during classical and instrumental conditioning under similar conditions.Neurobiol Learn Mem. 2004 Jan;81(1):82-95. doi: 10.1016/j.nlm.2003.08.002. Neurobiol Learn Mem. 2004. PMID: 14670362