The role of numerical and nonnumerical magnitudes in discriminative behavior

Humans and nonhuman animals can quickly identify the number of items in a set, which is vital for adaptation and survival as it enhances the organisms’ ability to collect food, navigate their environment, and assess social interactions, among other situations. This ability relies on nonsymbolic representations of number, which do not require language or symbolic representations of number. In nature, the number of items in a set is correlated with nonnumerical magnitudes like the total area, volume, or density of the set, among others. This means that it may be possible to solve approximate numerical discriminations by relying on differences on nonnumerical magnitudes among the sets rather than using numerical information. Prominent theories of numerical processing differ on the importance that nonnumerical magnitudes have in numerical discrimination responses in humans and nonhuman animals. This dissertation assessed the role that numerical and nonnumerical magnitudes have on discrimination responses made by pigeons and humans. In a series of experiments both species were trained and tested in identical two-alternative forced-choice discrimination tasks that manipulated the numerical and nonnumerical magnitudes of the training stimuli. Results demonstrate that the extent to which both species’ discrimination responses rely on numerical and nonnumerical magnitudes is flexible and depends on the relative amount of information each stimulus dimension provides in any given task. Moreover, both species can identify and adapt to changes in tasks requirements, prioritizing their use of either numerical or nonnumerical information to maximize choice accuracy.