Combined reinforcement: Poisoned cue or a panacea for modern equine training? A narrative review

Equine industry stakeholders have identified that inappropriate training practices are of significant concern for the welfare of domestic horses. Common approaches to horse training are largely based upon traditional practices, which frequently do not align with equine telos, ethology or cognition, and consequently, the techniques used to facilitate human-horse interactions warrants this concern. Given the current pressure on equestrianism’s social licence, it is timely to review what is currently known about equine behaviour modification approaches and identify areas where there is a deficit of knowledge. Uniquely in horse training, negative reinforcement is the fundamental basis for the majority of human-horse interactions, compared to the wider reliance on positive reinforcement in other animal species. While positive reinforcement offers many advantages for animal training, the characteristics of the ridden horse interaction necessarily relies on the use of aversive stimuli during training, such as pressure cues, even if of low intensity. Therefore, the question of whether a combined reinforcement model, incorporating both positive and negative reinforcement, can provide the benefits of positive reinforcement while accommodating pressure cues from negative reinforcement is worth investigating. This review seeks to outline current knowledge and application of combined reinforcement in the equine training literature for comparison with positive and negative reinforcement modalities. A survey of the literature highlighted a large number of inconsistencies in the description and classification of different forms of reinforcement used in training. Many studies putatively exploring positive reinforcement incorporate unacknowledged aspects of negative reinforcement, thereby confounding experimental outcomes and the interpretation of the results. Consequently the lack of specific focus on how these approaches actually consist of combined reinforcement has meant opportunities to advance scientific understanding in this area have been missed. As a result, knowledge of combined reinforcement remains poor, whilst training approaches and outcomes arising from positive reinforcement are inconsistent or incorrectly classified. These inconsistencies highlight the need for targeted research on the use of combined reinforcement, to better understand how to incorporate rewards-based training programs in horse training, including during riding.