Musical Skill and Cognition
John Jonides, Ph.D., University of Michigan
We explored the effects that training in music and in acting have on memory skills. Our studies to date suggest that the benefits of enhanced memory skills conferred by training in these two art forms result from strategic changes on the cognitive system used to maintain and store retrievable memories.
Our behavioral studies on memory were undertaken in a group of college-aged participants who had been extensively trained in music, and in a similar group (in terms of gender, age, and education) of non-musician controls. We tested the two groups’ long-term verbal memory. We found that the musically trained participants had better scores compared to the controls, but this difference disappeared when the musicians were prevented from rehearsing.
These results suggest that people trained in music apply strategies of rehearsal to maintain information in memory more effectively.
We used functional magnetic resonance imaging (fMRI) to visualize brain activity involved in long-term verbal memory, and to determine whether the musicians’ better scores, compared to non-musicians, were accompanied by more extensive activation in brain regions that have been implicated in rehearsal processes. While undergoing fMRI imaging, 11 trained musicians and 11 non-musician demographically similar control participants were instructed to rehearse and remember groups of three words. All participants first were instructed to use simple repetition to rehearse the words, and then they were instructed to use “elaborative” rehearsal (creating an emotionally aroused context) for remembering the words.
The musicians showed more brain activation using rote rehearsal, while the non-musicians showed more brain activation using elaborative rehearsal. The results suggest that the musicians, through rote rehearsal, are able to bind the remembered items to the context in which they occur.
Our studies in 21 actors trained in theater performance, compared to 24 demographically similar control participants, revealed that actors do not have better verbatim verbal memory. Instead, they are better able to extract the general idea from the verbal material, and this skill is transferable to other verbal cognitive skills.
In aggregate, our results to date indicate that the process of rehearsal, used in music and acting training, implements a strategy for focusing attention that enhances memory, and that this skill transfers to other cognitive functions involving memory. Rehearsal is an implementation of an attentional strategy, or “executive attention” as defined by Michael Posner.
Introduction and Conclusions
Most final reports about research programs begin with a summary of the empirical studies and the resulting findings. We begin, by contrast, with a summary of our conclusions and then move on to the findings that support these conclusions. We do this as a way of giving the reader a sense of the destination that this research has reached before reviewing the journey that led to this destination.
What we seem to be learning from the study of the effects of two arts (music and acting) on cognition is that the benefits of each are not a result of structural changes that they produce in the cognitive system; rather, the benefits are the result of strategic changes on the cognitive system. In an important way, this is consistent with Michael Posner’s work on the effect of arts training on “executive attention.”
He summarizes his work by arguing that the mediational path between arts training and cognitive performance goes through changes in motivation, which lead to changes in attentional focus. It is this change in attentional focus, he argues, that heightens performance in a variety of cognitive tasks, because most cognitive tasks require the kind of cognitive control that is needed to overcome prepotent responses when there is conflict present in the environment.
Our work focuses on this general theme within the context of memory processes. What we have found for musicians is that it is not that their memory is better; rather, it is that they apply strategies of rehearsal to maintain information in memory more effectively. Rehearsal is a way to implement an attentional strategy, in the sense defined by Posner as “executive attention.”
Similarly, what we have found for actors is not that their memory for verbal material is better; it is that they apply strategies for extracting semantic themes from verbal material, and these strategies result in better memory for the material in question. Again, this is a matter of strategy, and a matter of applying attention to semantic extraction rather than verbatim memorization.
Overall, then, what is emerging is a picture in which there is a mediational path that governs the now amply documented effects of arts training on memory. The path goes from skill in the arts, to heightened use of effective strategies for memorization, to better memory. We do not yet understand the mechanisms at work in learning in the various arts domains that lead to this path, but the tentative conclusion that the mediational path governs the effects of arts training on memory opens up the field for studying how training in the arts, and in other domains, can have such beneficial effects.
The Studies and their Results
The empirical work that leads us to this overall view comes from our studies of music and memory on the one hand, and from the study of acting and memory on the other hand. Let us summarize our findings.
Music and Memory
Comparing Groups on Verbal Memory
We have completed a program of behavioral research that examined the relationship of musical skill and memory. The results of this work are reported in Franklin (in press).
In brief, we selected a cohort of college-aged participants who were well-matched demographically, but differed substantially in musical experience. Specifically, one-half of the study participants had at least 10 years of musical experience, and practiced at least 10 hours of music per week. The other half had played an instrument for less than a year in their entire lives and did not play music currently. The two groups were similar in terms of demographics, including gender, age, education, grade point average, SAT scores, and performance on the Raven’s Test of Advanced Progressive Matrices.
We found that musicians scored better compared to the non-musicians on a test of long-term verbal memory, but this advantage disappeared when we prevented the musicians from rehearsing the material. We also found evidence that the musicians had a greater span of verbal working memory compared to the non-musicians. We attribute both of these effects to the enhanced use of rehearsal skills in musicians, rather than to a “hard-wired” difference in verbal memory capacity. This result suggests that musical training has the added benefit of training another cognitive skill, rehearsal, which has spillover effects onto cognitive tasks that engage verbal memory.
Functional Neuroimaging of Verbal Memory
Our results from studies of the effects of musical training on verbal memory were sufficiently promising to prompt us to conduct a study using functional MRI (fMRI, functional magnetic resonance imaging) of the brain in 11 trained musicians and 11 matched “control” (non-musician) participants, to delve into the neural mechanisms involved. We searched for a long-term verbal memory model test that would yield robust activation in the brain’s medial temporal region, and in other regions of the brain’s cortex. Our search led to a model developed by Davachi and Wagner (2002).
These authors were kind enough to provide us with the stimuli they used for their experiment, and we used those stimuli to construct our study of the musicians and their “matched controls” (demongraphically similar non-musicians). The study design involved having participants undergo fMRI imaging while they encoded (stored in memory in a retrievable way) triplets of words in one of two instruction conditions. One instruction condition involved simple repetition rehearsal, and the other involved “elaborative” rehearsal (constructing a plausible relationship among the words) in order to make a quality judgment on the stimuli.
The interesting feature of this paradigm is that it compared these two rehearsal conditions, since the results of our first study (described above) implicate rehearsal skill as the critical outcome of musical training. After participants encoded all the stimuli while being scanned, they later were tested on their memory of all the individual words, in a recognition test.
We then took a retrospective review of the data (“back-sorted” the data), comparing participants’ performance on the word-recognition test with an examination of brain activations, seen on imaging, that had resulted in successful versus unsuccessful encoding of the words. This allowed us to compare brain activation profiles for musicians versus control participants, to see whether the advantage shown in our first study for musicians behaviorally, is accompanied by a more extensive activation profile in brain regions that have been implicated in rehearsal.
Although all of our data have been collected and we are just at the beginning stages of analysis, our first pass at analyzing the data shows some interesting features. First of all, as with the findings of Davachi and Wagner, we also find that engagement of the brain’s medial temporal lobe occurs with both elaborative and rote encoding instructions. This demonstrates the robustness of this brain activation effect, and that the effect is replicable.
Even more interesting, however, is the pattern of brain activations for rote versus elaborative encoding in musicians compared to non-musicians. The musicians showed more engagement of the brain’s medial temporal lobe for rote than for elaborative encoding, while the non-musicians showed exactly the reverse.
We need to explore this finding in greater detail, and we certainly need to conduct analyses to determine whether these brain activation patterns were associated with whether or not the participants later correctly recognized each word. Nonetheless, this preliminary imaging finding seems to be consistent with our behavioral finding of enhanced verbal memory in musicians as a function of a strategic allocation of rehearsal. The musicians appear to be more able to rehearse the verbal memory, and this is what causes greater engagement of the brain’s medial temporal lobe structures, as a mechanism to bind the items to the context in which they occur. As we noted, it is yet premature to draw any firm conclusions from these data, and we shall continue our analyses, including a potential collaboration with Mark D’Esposito to conduct some multivariate analyses of the data.
Acting Skill and Memory
Our results with musicians led us to ponder the effects of training in other art forms on memory skill. One seemingly obvious domain in which to address this has to do with training in acting. On the face of it, training in acting would seem a natural place to examine its benefits on verbal memory. Surprisingly, our review of the literature unearthed remarkably little insight about this issue.
Study Design and Results
Consequently, we devised a behavioral study in which we tested 21 actors skilled in live theater, and compared them to 24 demographically matched control participants who had not had acting training. We tested for both long-term and short-term verbal memory skills. We found, contrary to what popular lore may suggest, that actors do not differ from control participants in their verbatim memory for verbal material. It is verbatim memory that one might suppose would be a skill that would be enhanced by training in acting, because actors (at least for live theater) need to memorize long corpora of lines.
Instead, what does seem to differentiate the actor and non-actor groups is how well they can extract gist from verbal material. Perhaps the most interesting implementation of this result comes from the research design used by Roediger and McDermott (1995). In this task, participants receive a set of lists to memorize. Each list is built around a theme. For example, subjects may see the words, “nap, bed, pillow, rest, night, snooze,” etc. They then later have to recall the list. The paradigm distinguishes itself because the theme word around which each list is constructed (in this example, “sleep”) is not actually included in the list. Yet in many cases, study participants mistakenly “recall” that this word was part of the list.
We have found that this false recall is much more pronounced among actors than among non-actors. This result suggests to us that what acting confers is a generally enhanced skill to extract gist from verbal material, a skill that is transferrable to other verbal cognitive tasks. It is this skill, we also suggest, that leads to actors being able to assume the roles of their parts. We are currently preparing a manuscript based on these results to submit for publication.
Franklin, M.S., Rattray, K., Sledge-Moore, K., Moher, J., Yip, C-Y., and Jonides, J. (In press). The effects of musical training on verbal memory. Psychology of Music.
Davachi, L. and Wagner, A.D. (2002). Hippocampal contributions to episodic encoding: Insights from relational and item-based learning. Journal of Neurophysiology, 88, 982-990.
Roediger, H.L., and McDermott, K.B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803-814.