Relationships Between Working Memory & Language

We remember some things for our entire lives, but other memories (like memory of a security code we just entered on a website) are gone in seconds. Researchers have often explained this difference between long-lasting and brief memories by suggesting that there are two different memory systems in the brain—long term memory for the permanent stuff, and short term or working memory for information that doesn’t last.

An alternative view, dating from the 1960’s, is that there’s no dedicated memory system for temporary information. Instead, the parts of the brain that are acting on this information take care of holding onto it temporarily. This is called the “Procedural” or “Emergent” view of working memory. I’ve long been an advocate of this approach, and my colleagues and I have written many theory papers (marked with an *asterisk below) explaining this view, specifically for memory for language, aka verbal working memory. The most recent of these papers, Schwering and MacDonald (2020), lays out these ideas in more detail.

The question of whether verbal working memory is separate from our language skills isn’t just some narrow academic exercise. It has real implications for how we can help children and adults who have impaired language abilities. For example, children who have been diagnosed with developmental language disorder (DLD) have poor comprehension, language production, and they will likely do very poorly on tests of verbal working memory. Some researchers have concluded that such children have an impairment in their (separate) working memory system, and they need memory training. If their memory gets better, the thinking goes, then their language will improve. 

The other view is that these children really have a language delay, and their poor memory test performance is a symptom, not a cause, of their language problems. On this view, these children need help with language, not memory training. See the Schwering and MacDonald (2020) paper for more information, including why memory training doesn’t work. Chapter 6 of my book makes similar points about memory and language in the elderly.

* Indicates that the article offers review, new theory, or response to other researchers.