Insights from a new study on word learning
Kapnoula, E. C., & Samuel, A. G. (2023) Wait long and prosper! Delaying production alleviates its detrimental effect on word learning. Language, Cognition and Neuroscience, 38(5), 724-744. DOI: 10.1080/23273798.2022.2144917
I am convinced that the study of language can teach us a lot about clinical practice – but also about ourselves as human beings. And sometimes, the study of language actually delivers on both. Like in this case.
So these folks – out of Spain – tested how word learning is affected by the way the stimulus is processed. What’s best for retention: only hearing the word; saying the word right after you hear it; or saying it after a delay of a few seconds? So they went to work, having their participants listen to novel words. They either only heard the words or repeated them: right away, after 2 seconds, and after 4 seconds. Four conditions for a nice comparison re: which one facilitates learning most. And guess what they found.
One of our most common practices in speech-language pathology is to give people input – like, a word – and have them repeat it. This practice is based on our intuition and on a formidable body of science. So, based on that, we’d guess they’d retain the word better if they repeated it, rather than just hearing it. Right?
Well, not quite. Turns out it depends on the amount of time you let elapse. Just hearing the word actually makes for significantly better learning than repeating it right away! However – and this is crucial – if you give your learners four seconds and then have them repeat the word, that effect goes away. Check out the figures below, straight from the article.
© Kapnoula & Samuel (2023)
What you see here is the mean difference between correct responses by condition. As you can tell, accuracy of responses is highest in the "perception only" condition, where participants just heard the word. It plummets in the condition where learners repeated the word immediately, creeps back up in the condition where they repeated it after 2 seconds, and reaches almost the same level as "perception only" in the condition where they repeat it after 4 seconds. Surprising, right?
Now, in this dataset the "only hearing the word" condition wins. Does that mean we shouldn't have our clients repeat words at all?
Nope – at least not if you want generalization. Turns out retention over time is best in the "4 seconds" condition!
© Kapnoula & Samuel (2023)
This figure shows the same data but comparing the first to the second day of the experiment. See what's happening? If you only hear a new word, you remember it well right away but you've already started forgetting it a day later. If you repeat it right after hearing it, it's almost like your immediate mental image of it is smothered, but it bobs back up on the second day. If you repeat it after 2 seconds, it kind of just hovers at the same level. But if you repeat it after 4 seconds, you not only get Day 1 learning that's basically as good as in the "only hearing" condition but – ta-dah! – you get your best learning outcome on Day 2.
Now, notice many of those differences aren't ginormous – in fact, the ones between Day 1 and Day 2 within each condition are all non-significant. But the Day 2 ones between conditions sure are substantial. And that's where we want our patients to go, of course: into good learning over time.
One more thing. This effect doesn't only happen in learning new lexical items but also in learning phonemic contrasts. The authors report on a previous study of theirs and, helpfully, put a pertinent figure in their article:
© Kapnoula & Samuel (2023)
What you see here is Day 1 data from a study in the same format, where participants learned a phonemic contrast not present in their native tongue. As you can see, the "4 seconds" condition wins outright this time, beating the "only hearing" condition even on the first day of the study. And both leave the "say it immediately" condition far behind.
So what's all this mean for clinical practice?
It's very simple, actually. Two words. Slow. Down.
What Kapnoula and Samuel are telling us here is that our therapy works best when we give our patients time. Whether you want to teach a new word or a new sound, let them hear it, let the new percept really sink in and start to form a mental image, and only then have them say it after you.
How often does speech and language therapy become a rapid-fire, endless sequence of RREs (requests, responses, evaluations), as my mentor Jack Damico called them? Turns out that kind of mindless drill is detrimental to learning. Slow it down. Pause. Really let them hear the target. Really let it sink in. Then let them produce it with awareness. That's how you get results.
And what does that tell us about our human condition? Well, what I thought of immediately when I read this article wasn't only how it could improve therapy, but also just how much we are all in need of slowing down. Just. Slowing. Down. Stopping the endless rush to getting more and more done that means less and less. Who says this 4-second effect is only there in word and phoneme learning? Mental processing of novel stimuli – of any stimuli, really – requires the same basic preconditions across contexts and types of stimuli. One of them being time.
So, after you complete this article – or that assessment, or a round of therapy, or simply while enjoying a cup of coffee – why not try and pause for a moment, and let the experience sink in? Allow it to truly become part of your system. Quiet time is the (often unspoken) foundation for reflective practice, after all.
Photo by Jan Kahánek on Unsplash
There is no time for that, you say? You are correct. The way we work these days, everything is bent on squeezing meaningless activity out of every second. The only way to fight back* is to demonstrate that speeding up often means accomplishing less. And that's what this study does. A tiny contribution, by the science of language, to making the world a better place.
* No, I'm not naïve. I know it takes more than just dangling one study in front of one supervisor. And my lab is on it. We have a study coming out that not only shows why SLPs are burned out and overwhelmed but also gives some directions to combat that. Stay tuned!
Comments