Text this: Evaluation of event plausibility recognition in Large (Vision)-Language Models