The Dancer and the Trunk
In his Essay Concerning Human Understanding, the English philosopher John Locke tells the story of a man who learned how to dance “to great perfection”. He practiced consistently every day in the same room, which was furnished distinctively with a large, ornate trunk at one end.
In that room, the man could dance like a true professional, but he couldn’t perform as well in any other place. It was as though the intricate steps and movement patterns of the dances he’d perfected were so indelibly linked to that space that they couldn’t be replicated anywhere else.
What this man was experiencing was a form of context dependency, the ability to recall knowledge or skills in the specific environment in which they were learned but not in others.
We’ve all experienced this before, whether it’s feeling that we’ve perfected our speech when practicing to ourselves in our living room only to fall flat when we deliver it at work; or believing we’ve gained a solid grasp of French in a weekly evening class but feel hopeless in our very first conversation of a trip to Paris.
Scuba Diving, Music and the Theory of Reinstatement
The anecdote of Locke’s dancing man raises an important question. If we can revisit or recreate the same external environment in which we initially learned something, will our learning and recall be stronger?
Psychologists Baddeley and Godden from Stirling University in Scotland set out to answer this question in 1975. In an experiment in the diving destination of Oban facing the Southern Hebrides, they divided 18 scuba divers into two groups, and gave them a list of 36 words to study while submerged 20 feet under water.
An hour later the first group was tested on the words on dry land, while the other group took the test underwater, using a waterproof microphone to communicate with the researchers.
Surprisingly the difference was significant – the divers who took the test underwater remembered 30% more words than those who did it on land. This led Baddeley and Godden to propose the existence of a reinstatement effect, suggesting that “recall is better if the environment of the original learning is reinstated.”
This work was built on by other cognitive scientists, who began to examine what other environmental cues influenced recall. In a 1985 study by Steven Smith at Texas A&M University, the effect of music on recall was examined using 54 psychology undergraduates.
The students were divided into 3 groups and given a list of 40 words to study for 10 minutes; the only difference was the background music they studied to. The first group studied in silence, the second with jazz by Milt Jackson and the third with classical music – Mozart’s Piano Concerto Number 24.
Two days later they returned to the study room and were given a test without warning to see how many of the words they could remember. This time Smith subdivided the three study groups and changed the music for many of them – some of the students who’d studied to jazz took the test with jazz; others took it in silence and others with classical music; this procedure was repeated with the other subgroups to produce 9 variations.
He found that the students who studied and tested in the same conditions performed better than those who did so in different ones, providing more evidence for reinstatement theory. One possible explanation for this particular result is that the background music became entwined with the stored memory of words and acted as a cue for recall when the students heard it again.
Silence, Contextual Cues and The Spy at the Dinner Party
Another interesting finding from Smith’s study was that the silence-silence group did the worst of the 3 groups who had the same conditions, remembering about half as many words as the jazz-jazz and classical-classical groups. Smith speculated that this was due to an “impoverished” study environment with fewer cues to latch on to than the more enriched ones with music in the background.
Although it’s rarely possible for us to stream our favourite playlist during an exam or to cue it up when it’s time to perform, this finding suggests that the advice we’ve always been given to find a quiet place to focus when studying might not be applicable to all of us, all of the time.
Our brains are naturally attracted to external stimuli which signal changes in environment, as this would often have been a sign of danger for our ancestors. While these stimuli are rarely a matter of life and death anymore, we can take advantage of these contextual cues to help us encode memories better.
When thinking about the impact of context on memory, it helps to imagine our learning brain as a spy at a dinner party. While listening to the main conversation at the table, it’s also scanning the room periodically, taking in the associated sights and sounds and noticing any anomalies.
This observational process can help us associate different parts of the conversation with distinct images, like a painting on the wall, or sounds, like the clinking of glasses, which will make them more memorable. This analogy highlights the subtle but important impact that contextual cues can have on our memory, which can be especially important when we’re learning something new.
Performing in Different Environments and Becoming Antifragile
What about evidence that contradicts reinstatement theory? Is it possible that changing our environment can have a positive impact on memory and learning?
Smith (the same one), Glenberg and Bjork ran an experiment to see what would happen if people studied the same material twice, in different places. They presented a group of students with forty four-letter words to study in two ten-minute sessions and divided them into two groups.
The first studied the material in the same location twice – either a small cluttered basement room, or a more spacious room with a window looking out onto a courtyard. The second group did one ten-minute session in both rooms.
Three hours later the students were tested on the material in a neutral classroom, which they hadn’t been in before. The difference in scores was huge – the one-room group remembered an average of 16 words, while the two-room group remembered an average of 24 words, meaning that a simple change in location led to a 40% improvement in recall.
One explanation for these results is that the brain encodes one subset of the words in one room and a slightly different one in another, which overlap and improve recall. Another possible explanation is that rehearsing in the two rooms doubles the number of cues linked to the studied words but the truth is, we can’t identify a causal effect with certainty.
So what do we do with all of these results? And doesn’t the evidence seem contradictory?
If you’re certain of the environment in which you’ll have to recall knowledge or a skill, then applying reinstatement theory by trying to recreate some of those conditions is a good idea.
The problem is, this is rarely possible in real life – we almost never know when we’re going to need to apply our knowledge, which means the ability to improvise and adapt is crucial. Given this need to perform spontaneously, it’s a good idea to vary some environmental factors during your learning.
In his book Antifragile, Naseem Taleb discusses this idea at length, emphasising that avoiding variations in external conditions be very harmful. While you may be able to perform in one context, the chances are you’ll struggle in a different one if you get too comfortable in a closed system.
The concept of antifragility has been applied in risk analysis, molecular biology, and computer science but we can also apply it to learning. Being antifragile in a learning context is being able to respond to changes in our environment and perform well, even thriving and growing in response to shocks and uncertainty.
If I’m playing a tennis tournament in two weeks time on clay courts, it’s probably a good idea to practice in context and use the same surface in my preparation.
But if I want to work on a particular aspect of my game in the off season, like my forehand, it might be a good idea to mix up my practice on different surfaces – some clay, grass and hard courts. This will make me more likely to perfect the shot and be able to use it effectively in different contexts, on different surfaces.
There appears to be evidence for the existence of some reinstatement effect where recreating the same external conditions in our environment can have a positive effect on learning and recall.
If we could predict our target performance environment, we could try to produce a carbon copy of these conditions but this is rarely possible in real life. In most cases we want to apply what we learn in response to challenges as they arise and adapt on the fly. Ultimately, effective learning should be antifragile.
1) Start Consistent then Mix it Up
When you start learning something new, consistency is key to establish the habit and build some early confidence. But once you get past the initial learning phase, vary a few elements of your environment one at a time.
So if you’re learning to play the guitar and your initial practice sessions have been at home try going to the park next time. If you’ve been practicing alone, do it in front of a small group of friends. What’s most important is not what you vary but that you do it in the first place.
2) Scenario Plan
You may not be able to control your environment but you can think about the likely changes you might face and prepare an action plan to deal with them. This is something that’s used extensively by military strategists, CEOs and sports coaches so that they can respond rapidly to the most probable environmental changes if they arise. Even if those changes don’t arise, the fact they’ve taken the time to identify key strategic issues gives them added confidence in uncertainty.
3) Make the Most of Strong Cues
While contextual cues do play a role in memory and learning, evidence suggests that strong hints are more reliable for recall. If you’re learning a language, consider using some form of mental elaboration to make the words you use more familiar. Check out this tip on the Gruneberg method here and my post on elaboration for more on this.