Abstract
Gambling machines have been referred to as the “crack cocaine” of gambling, due to the rapid speed at which people develop a pathological addiction. The American Psychological Association (2000) defines pathological gambling as a disorder of impulse control with continual and recurring maladaptive gambling behaviour. Extensive neurological research has demonstrated that pathological gamblers display changes in prefrontal cortex (PFC) activity. Given the importance of the PFC in decision-making and reward we recorded from the nidopallium caudolaterale (NCL), the avian equivalent of the PFC, employing single-unit electrophysiology. Specifically, four gambling-related neuron types (reward-proximity, I-Lost, I-Won, and near-miss effect (NME)) were investigated for changes in overall prevalence and changes in activity as a function of either short or long experience playing a slot machine. Four pigeons (Columba livia) served as the subjects, two subjects were trained for a short period (short-trained), and two subjects were trained for a long period (long-trained) in order to mimic pathological gamblers length of gambling experience. The subjects played a computer-generated slot machine whereby the animal pecked on the virtual arm to initiate four tumblers to begin rolling. The pigeon then proceeded to peck each of the four tumblers in a consecutive order to stop each from rolling. If four of a kind appears on the tumblers the subjects received a food reward. The task was very similar to real slot machines that humans use in gaming rooms and casinos.
It was hypothesised that the prevalence of reward-proximity, I-Lost and I-Won neurons would be higher in long-trained subjects. In contrast, the prevalence of NME neurons was predicted to be higher in the short-trained subjects. Moreover, with regard to activity rate (or magnitude effects) the long-trained subjects were hypothesised to demonstrate higher levels with reward-proximity, I-Lost and I-Won neurons. In contrast, the short-trained subjects were hypothesised to demonstrate a higher magnitude effect. The results illustrated higher prevalence’s in the long-trained group with reward proximity, I-Lost and I-Won neurons, and a higher prevalence of NME neurons in the short-trained group, thus, the first hypothesis were supported. Higher magnitude effects were found with reward-proximity, I-Won and NME in the long-trained group. The short-trained group, however, demonstrated a higher magnitude effect with I-Lost neurons. The second hypothesis, therefore, was partially supported. In conclusion, the present experiment demonstrates that with extensive gambling experience changes in neural activity is observed in the avian NCL. Due to the similarities of the NCL to the PFC, this experiment has significant implications with regard to developing effective treatments and interventions for pathological gamblers. The findings, here, provide a framework for further research to investigate gambling related neural changes.