The Encoding of Information in the Brain

The representation of objects and faces by neurons in the temporal lobe visual cortical areas of primates has the property that the neurons encode relatively independent information in their firing rates. This means that the number of stimuli that can be encoded increases exponentially with the number of neurons in an ensemble. Moreover, the information can be read by receiving neurons that perform just a synaptically weighted sum of the firing rates being received. Some ways in which these representations become grounded in the world are described. The issue of syntactic binding in representations, and of its value for a higher order thought system, is discussed. 1. THE ENCODING OF INFORMATION IN THE BRAIN How is information encoded in the cerebral cortex? Can we read the code being used by the cortex? What are the advantages of the encoding scheme used for the computations being performed by neurons in different areas of the cortex? These are some of the key issues considered here. Because information is exchanged between the computing elements of the cortex, the neurons, by their spiking activity, the appropriate level of analysis is how single neurons, and populations of single neurons, encode information in their firing. More global measures which reflect the averaged activity of large numbers of neurons (for example PET (positron emission tomography) and fMRI (functional magnetic resonance imaging), EEG (electroencephalographic recording), and ERPs (event-related potentials)) cannot reveal how the information is represented, or how the computation is being performed. Some of the types of representation that might be found at the neuronal level are as follows (see Rolls and Treves 1998). A local representation is one in which all the information that a particular stimulus or event occurred is provided by the activity of one of the neurons. This is sometimes called a grandmother cell representation, because in a famous example, a single neuron might be active only if one’s grandmother was being seen (see Barlow 1995). A fully distributed representation is one in which all the information that a particular stimulus or event occurred is provided by the activity of the full set of neurons. If the neurons are binary (for Synthese 129: 153–171, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.