In recent years, neural-inspired architectures have gained significant popularity for Multiple-Choice Question Answering (MCQA) tasks. However, these approaches often need help with interpretability, resource-intensive training requirements, and limited transfer learning capabilities. In this study, we deviate from this mainstream trend and propose novel methods based on fuzzy logic to address these issues. We aim to achieve comparable performance to neural-based solutions while offering advantages such as improved interpretability, reduced training resource demands, and enhanced knowledge transferability. Through our experiments, we demonstrate the potential of fuzzy logic approaches in MCQA, showcasing their ability to obtain results on par with state-of-the-art neural solutions while providing intuitive and cost-effective training and knowledge transfer solutions.