Can we measure working memory without contamination from knowledge held in long-term memory?