November 25, 2005
Musical Training Might Improve Brain Processing Of Speech
Trained musicians are better at detecting differences in syllable sounds.
In two Stanford studies, researchers demonstrated that people with musical experience found it easier than non-musicians to detect small differences in word syllables. They also discovered that musical training helps the brain work more efficiently in distinguishing split-second differences between rapidly changing sounds that are essential to processing language.
Nadine Gaab, a former Stanford postdoctoral fellow, will present the findings at 9:30 a.m. Nov. 16 at the Society for Neuroscience's annual meeting in Washington, D.C. "These results have important potential implications for improving speech processing in children struggling with language and reading skills," she said. They also could help "seniors experiencing a decline in their ability to pick up rapid changes in the pitch and timing of sounds, as well as speech perception and verbal memory skills, and even for people learning a second language."
But does musical training really help language skills development all that much? The problem with this study (and with most social science for that matter) is that it compares people at a snapshot point. All sorts of selection effects could be responsible for this result.
Adults with and without musical training were compared in their ability to recognize syllable sequences.
In the first study, researchers took 28 adults, divided into musicians and non-musicians, who were matched for age, gender, intelligence and general language ability. Musicians in the study were required to have started playing an instrument before the age of 7, to have never stopped playing and to have continued to play several hours a week. When musicians play, Gaab said, they must actively distinguish between sounds and their order, and adjust as necessary.
Suppose a large cohort of 7 year olds were tested for general intelligence and language skills and then some proceeded to learn music an others did not. Would the music learners be better at understanding and speaking language 20 years later? I'm skeptical.
Non-musicians in the study had to be native English speakers with minimal experience studying non-tonal foreign languages such as Spanish. People who had studied a tonal language such as Mandarin were not included.
During the experiment, participants listened to pairs of syllables such as ba-da, ba-wa and ga-ka, and noted if each syllable in the pair sounded the same or different. Depending on how they performed, the scientists made the task increasingly difficult by using syllables that sounded more and more alike. Musicians outperformed their non-musician peers in how quickly and accurately they perceived these rapid changes, Gaab said.
In the second experiment, researchers used functional magnetic resonance imaging (fMRI) to find out whether musical training changes the way the brain processes sound. The fMRI scanning machines, which look like beds that slide into tubes, normally are used to check for brain injuries or tumors. With slightly different software they can be used to measure which regions of the brain are active by looking for changes in blood oxygenation, a process that occurs in parts of the brain where the neurons are active.
Forty people, evenly divided into musicians and non-musicians, listened to three-tone sequences made from different combinations of low and high pitches. Participants had to reproduce the order of the tones they heard by manually pressing buttons on a panel.
Musicians once again beat the non-musicians with this task. "We were surprised that musicians could do it almost perfectly," Gaab said. Musicians got the fastest tone sequences right at least 85 percent of the time, compared to non-musicians who hit a 50-percent average. They also could replicate the sequences a lot faster. "Non-musicians needed to make a lot more effort—their brains were not as finely attuned."
According to Gaab, musical training appears to alter the ability of the brain's language areas to process pitch and timing changes that are common to perceiving both words and music. "The brain becomes more efficient and can process more subtle auditory cues that occur simultaneously," she said.
I see these results as analogous to how a doctor or nurse has a fingertip that is very good at measuring a pulse. The ability measured is fairly narrow. A person who learns music might be better able to distinguish syllables and words which are mumbled or spoken quietly or spoken with an acccent or under noisy conditions. The effect would be equivalent to simply having better hearing. But the benefit might be limited to just an enhanced ability to signal process sounds and may not carry over at all into enhanced ability to parse sentences and understand, say, written language.
I've long wondered how much the learning of particular skills enhance or interfere with the ability to learn other skills. For example, learning math enables the learning of physics, engineering, andd sciences because it provides a conceptual toolbox needed to understand many other subjects. But does, say, learning foreign languages use up neuons that then are unavailable for learning other subjects? Or if a person is born blind does that person have more mental resources available to learn, say, spoken languages because their mind is not cluttered by visual image information? Do parts of their brain which otherwise would have been recruited for processing images instead get recruited for better processing of sounds?
I believe this is possible, but I didn't notice the bit where they demonstrate that musical training enhances any of the things they are talking about. What about the obvious alternative hypothesis: that people who already have these abilities to an unusually high degree are therefore likely to become musicians? A snapshot study without a time element just tells us that musicians have unusually acute ability to distinguish and remember sounds - well, big wow - but nothing about whether they had that ability innately, or developed it over time. Probably both, but it's clearly a question that a study like this should address.
>I've long wondered how much the learning of particular skills enhance or interfere with the ability to learn other skills.
Well, the opportunity cost of becoming proficient at a musical instrument is huge. I've seen 10,000 hours of practice quoted as a reasonable estimate for getting really (professional-level good) - that's ten years of three hours a day that you could otherwise be learning other things. In that sense there's a direct tradeoff in what you choose to do with a limited time budget. On the other hand, learning anything to a high level requires diligence and an ability to focus the mind for sustained periods of time even when things get tedious & frustrating - that at least must be a transferable skill.
About the musicians, even assuming causation rather than selection effect, discriminating different syllable sounds isn't really useful in "processing speech" unless you're talking about learning a foreign language (L2) post-puberty or if you're a kid w/ a pronounced speech pathology (e.g., an English-speaker who can't say both /p/ and /b/). By puberty you've already learned all the distinctions you'll need to know for your native language (L1), whether you played an instrument growing up or not. And most kids who have pathologies fall into the "can't say it" camp rather than the "can't even recognize the difference" camp. It's not unlike toddlers who may say both /b/ and /p/ as a /b/ sound, but who can still tell the difference -- if an adult tries to mimic their error by pronouncing "pie" as the word "bye," they'll get angry and protest "that's now how you say it!" So the real question for the speech pathology kids is, does music help you *produce* two different sounds, beyond merely recognizing their distinction?
I can see this helping more for learning L2 -- in that case you really do have to learn distinctions de novo, say the unaspirated and aspirated consonants (written as, e.g., "k" vs "kh"), which are the bane of the beginning Hindi student's existence. The study didn't look at learning altogether new sounds (all of the ga-ka, ba-wa sounds are in English), so the next place to look would be: does music putatively help you learn to discriminate, and more importantly *produce*, novel sounds like the French and German /r/?
One remaining puzzle: the blurb says the participants were matched for "intelligence" -- does that mean that they were administered an IQ test w/ an auditory component? Otherwise they might not have actually matched for intelligence, and the musicians could have had a higher auditory IQ component score.
BTW, not that it matters but good god, Nadine Gaab is smokin' hottt!
About the parts of the brain being recruited for other purposes. For learning foreign languages, certainly learning them as a normal part of growing up bilingual doesn't hurt -- look at how great is the scientific output of the typically trilingual Dutch. Would be interesting to look at those who learn L2 post-puberty -- but of course here that's typically restricted to high-IQ people who want to be able to travel throughout Italy or something. You'd have to look at, say, Russia where you could sample both high-IQ university students as well as lowly cab drivers who are trying to learn English / German / etc.
As for blind kids, I believe most of their visual processing area gets recruited primarily for tactile use. Think about it: vision serves the purpose of exploring your spatial environment, so how else are they going to explore it? So, if you ask a blind kid "can you see the chair," they'll sort of tap it once to show that they've verified it's existence. If you ask them to "look at the chair," they'll start to explore it using their hands, since "look" implies a goal-oriented exploration of an object's properties. And Braille, the use of a walking stick, etc. are also tactile. They do develop a sharpened sense of our weak echolocution (using echos of footsteps, etc. to sense if there's an obstruction coming up), but I don't know about phoneme-discrimination.
I read in Pinker something to the effect that everybody is born with the ability to make, and distinguish, the full range of sounds the human vocal system is capable of, but as part of learning one's mother tongue, one loses the ability to distinguish/make sounds that aren't in it. Hence why hardly any adults can ever learn to speak a foreign language with no accent. My wife is Russian, studied German at university and has lived in Germany for ten years. Her German is 100% fluent, but she still has a slight but detectable Russian accent. Another example I noticed recently on TV is Peter Maffay, who is a big pop star in Germany. Rumanian born, has lived in Germany since his teens, now in his 50s. Still speaks German with an accent that I personally couldn't place as Rumanian, but clearly not a native speaker.
Flip side of that: I've lived in Germany for seven years and speak good but not really fully fluent German. But I do notice I sometimes seem to have less difficulty understanding really heavy Swiss, Bavarian or Austrian accents than native Hochdeutsch speakers do. As a non-native speaker perhaps you're less keyed in to relatively subtle differences that might seem big and jarring to a native speaker?
Re Hindi: dental vs retroflex consonants were the bane of my sanskrit studies. Understand the difference in theory, and can just about hear it if somebody goes out of their way to enunciate it really clearly for me.
I've heard that playing an instrument can improve your skills in mathematics. I've been playing the guitar for about 4 years now and have become very good (not to sound full of myself or anything) and I'm still in Algebra 1! And I'm a sophomrore!