Dialectical language model evaluation: An initial appraisal of the commonsense spatial reasoning abilities of LLMs