Easy Come, Easy Go: The Surprising Paradox of Task-Switching

What do pit orchestra and my dissertation have in common? Well, aside from the fact that I maybe read some relevant articles during the occasional rehearsal and/or show (or, more often the case, discovered that articles weren’t as relevant as I thought they might be), almost nothing…aside from one thing: task-switching. As a reed player, I have doubled for most of the shows I’ve done, meaning I play multiple instruments, usually in the clarinet and/or saxophone families, though I have done some flute (not my forte—not my piano either, for that matter). Though still playing music, when switching between instruments, I’m essentially switching between tasks (at least as psychologists would define task-switching), because I’m alternatingly applying different sets of rules (i.e., fingerings). When doing this, rules will inevitably be misapplied (i.e., I play the wrong fingering). Or at least it’s thus far been consistent for me. 😉

But now let’s add another dynamic into the mix. (Maybe a nice mezzo forte? Man, the musical puns are strong with me!). Let’s consider a related—and more challenging—type of task-switching: transposing. For those of you who aren’t instrumental musicians, this may be uninteresting and/or hard to follow, so I’ll try to keep it quick and fun, because it does set up the example and the whole point of this post. Not all musical instruments are in the same key. For example, when C instruments (e.g., piano, flute, oboe, violin, tuba, and others) play a C, you get what’s called a concert C; that’s why they’re called C instruments. However, when a B-flat instrument, such as the clarinet, plays the same written C, you actually get a concert B-flat, which sounds a whole step lower than when a C instrument plays the same written note. (Or, in other words, a clarinet has to play a D to produce the same absolute pitch as a flute playing a C.) When an alto saxophone plays a C, you actually get a concert E-flat, which sounds a sixth lower that if played on a piano. I’ve included a diagram below that might help to visualize this.


Do you find this different keys business totally confusing and seemingly unnecessary? You’re in good company. However, I believe the advantage of this system (and Wikipedia seems to agree with me, so it must be right. 😉 ) is that it preserves fingerings within and across similar families, at least within woodwinds. Thus, when you see a top-of-the-staff G, you will always finger that the same on saxophone, whether it’s a soprano, an alto, a tenor, or a bari. It will produce a different pitch on the different saxophones (it will be an octave lower on tenor than soprano and an on bari than alto; it will be a fourth lower on tenor than on alto; etc.), but at least you finger it the same way. And, bonus, the fingering for that G is the same on clarinet and on flute. So while it might sound different on all of those instruments (save the clarinet and soprano sax, which are in the same key and octave, which has helped me on more than one occasion), at least there is a pretty consistent set of fingerings, which is immensely helpful for people who double, because reed books for pit orchestras are often written for multiple instruments. Because the other option would be having the same fingering be a B-flat on soprano sax, tenor sax, and clarinet, while being an E-flat on alto sax and bari sax, a C on flute. With that, you would have to relearn your note-to-finger mappings for each instrument, and I can’t imagine that being in any way easy.

I recently finished playing in the orchestra for a community theatre production of “The Little Mermaid”. I played the Reed 3 book, which had saxophone, clarinet, and bass clarinet parts. But because there was no double reed player, I also interspersed key oboe and English horn parts into my music and played them on clarinet. (I quite literally copied, cut, and pasted them so that I didn’t have to flip between two books. I did that once. It was a hassle. And I have no desire to do that again.) I only picked up these new parts a few days before opening night, so, not having had much time to do anything with them, I was transposing them in my head at first. For the oboe parts, that means playing a step up. I’ve had experience with this, so it’s not too hard—as long as the music is easy enough and the key signatures aren’t too complex. But this is pit orchestra, meaning that the music is, at times, quite challenging, and the key signatures get to be heinously complex. To transpose the English horn, I have to play a fourth lower than what’s written on the page. No matter what the music’s like, that one’s an ouch.

Basically, what it came down to was this: for a few small parts of the music, I was not playing what was written on the page. As I was first fumbling through this new music, I avoided any horribly and grotesquely obviously misplayed notes by playing very quietly (or not at all) as soon as I began to lack confidence in my transpositional abilities. But I knew I couldn’t keep doing that, because these were some important and exposed solo-like (or actual solo) parts. So they needed playing. I began to fear the inevitable: a glaring, ugly mistake on some really exposed part where it would be painfully obvious. I feared that it could happen on one of the parts I was transposing. But I also feared—probably to a greater degree—that it would happen on a clarinet part as I came back to that from having just been transposing an oboe or English horn part. Because of course that’s how it would happen to me: I mess up the easier playing on what’s my main instrument. It would be so laughably and pathetically ironic.

But, reflecting upon this one night, I actually realized that it wouldn’t be ironic. Even though playing the clarinet parts and not having to transpose is far easier for me (what, with it being my primary instrument and all, and with reading music as-is being immeasurably easier than transposing in my head), that’s exactly why it might be harder to come back to a play-as-written-on-the-clarinet part. Why? Task-switching.

(Aside: The first nasty mistake actually came as I was playing what should have been a quite lovely oboe part. It was one of those aforementioned heinous key signatures. After that, I realized I shouldn’t be leaving the audience’s aural pleasure in the hands of my less-than-consistent transpositional abilities, so I wrote out most of the transpositions. Much improved.)

First, let’s revisit what task-switching is, at least as defined in the many psychological experiments that have tested it. Despite what its name suggests, task-switching rarely (if ever?) deals with switching between widely disparate tasks, such as verbal reasoning and mental arithmetic. Rather, the tasks have a high degree of overlap, generally to the point of applying different rules and/or responses to the same set of stimuli. For example, it could be asking you to sort the same set of images by color in one task and by shape in another task. The stimuli are the same, but what you’re being asked to do with them is different. Hopefully now the connection to pit orchestra makes sense: I’m looking at the same notes, but having to play them differently, either transposed or not, or with clarinet fingerings or saxophone fingerings.

Looking at the task-switching literature, one of the key findings is entirely unsurprising: performance is easier on the easier task—or, rather, I should say, on the dominant task; that distinction will be crucial when we discuss the third key finding. For example, I miss far fewer notes when not transposing than when transposing. The second key finding is about as unsurprising: when switching between tasks, performance on the first few trials after the switch is usually worse. This is called a switch cost. There are different theories as to what accounts for this (perhaps many of them having truth), but I don’t need to go into that for y’all to understand why that effect is consistently observed. Basically, you’re warming up to the new task, getting acquainted with how it works. It’s going to take you a few trials to get your sea legs (or should I say C fingers?), as it were. That’s to be expected. So let’s move on to the third key finding, which I find to be the most interesting—and the most surprising (at least until you understand the underlying psychology): when switching to the easier task, relatively more mistakes are made than when switching to the harder task. This is called a switch cost asymmetry. Why does it occur? The central executive. (Yep, we’re discussing that again. Third time’s a charm. We’ll see more if its inner workings this time!)

This is where the distinction between easy and dominant becomes critical. Yes, a dominant task will probably be easier (and exhibit accordant behavioral measures, such as faster responding and fewer errors), but that could have just as much to do with its being so well-learned and well-practiced that it feels rather natural as it could have to do with the fact that it is simply and objectively easier. (Though, can easiness ever be objective? Perhaps that will be a post for another time.) So, yes, a dominant task will likely be performed more effortlessly and efficiently and automatically—and, dare I say, easily?—but, because it is so dominant, the rules for this task are also more dominant than the rules for the other task (e.g., when I read music, my fingers automatically know the fingerings for those notes; it’s much less dominant and more effortful for me to see a C and instead finger a D or a G). Why does that matter? It means that you’ll have to work harder to inhibit these dominant rules when you don’t want them. This inhibition is a key function of the central executive. It’s also a key process in task-switching and why we see the switch cost asymmetry. (The figure below might help you to visualize this: both tasks encounter the switch cost when switching from the previous task, hence accuracy being on trials where the task is repeated [maroon bars], but that switch cost is larger for the easy task [on the left].)


Now the picture hopefully starts to make sense: yes, the dominant task is “easier”, but it’s also harder to inhibit. (Okay, I’m going to be done using the scare quotes on “easy” and “difficult” [and their morphological derivations] but just know that when I refer to a task as easy or difficult, I’m really referring to how dominant the task rules are.) So, while, you’re engaged in the dominant task, you’re easily inhibiting the rules for the less dominant task as you cruise through the “easy” task. But, when prompted to switch tasks, you have to inhibit these dominant rules. This requires more inhibitory processes and resources—but it’s doable. You also have to activate the less dominant rules. Since it didn’t take much to inhibit them, it’s rather easy to activate them. Still, the task is less dominant (“harder”), so performance isn’t quite what it was on the dominant task, especially for the first few trials (switch cost) as you get used to doing the new task. Now, you’re asked to go back to the “easy” task. Remember, you were just inhibiting these rules, which, given how dominant they are, took a lot of inhibitory effort and power. So, to reactivate those rules takes a lot of effort and control. You’re going to struggle through the first few trials (switch cost) as you recover these previously deeply inhibited rules. You will likely struggle more here than you did as you began the less dominant task. That is, the switch cost will be greater for the dominant task than for the less dominant task. Thus, while you still experience a switch cost when starting either task, it is harder to get yourself back in the frame of mind for the easy task, because it took so much effort to get yourself out of that frame of mind. That’s why I had such a great fear of messing up a clarinet part: I was afraid (with good reason) that I was going to inadvertently transpose a part that ought not to have been transposed.

Simply put, whether it’s task-switching, instrument-switching, or switching between transposing and playing as written, task performance (as measured by something like accuracy, response time, etc.) in switching paradigms comes down to dominance and inhibition: the more dominant the task, the more effort required to inhibit it, meaning the more challenging it will be to reactivate it, making what was once easy, easy no longer—at least for a brief transitional period. So what does this all mean? What do we do with this? Truth be told, I don’t know. Task-switching has been extensively studied, and we understand it much better than we used to, but I’m unaware of any studies that have been able to eliminate switch costs. So I guess the best that we can do for now is to be aware of the phenomenon and have a little grace with ourselves and others when we fall prey the inevitable. Also, be mindful of how you plan your tasks and, when possible, avoid switching between tasks so that you can keep performance up. (Also, it’s generally best to avoid multi-tasking. That’s just not as effective as we like to think it is. Perhaps I’ll have more to say on that in the future…)

Yours truly,
D. R. Meriwether
Renaissance Man


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s