Online Education and Jazz

A common responses to my article, Why Online Education Works, is that there is something special, magical, and “almost sacred” about the live teaching experience. I agree that this is true for teaching at its best but it’s also irrelevant. It’s even more true that there is something special, magical and almost sacred about the live musical experience. The time I saw Otis Clay in a small Toronto bar, my first Springsteen concert, the Teenage Head riot at Ontario Place these are some of my favorite and most memorable cultural experiences and yet by orders of magnitude most of the music that I listen to is recorded music.

In The Trouble With Online Education Mark Edmundson makes the analogy between teaching and music explicit:

Every memorable class is a bit like a jazz composition.

Quite right but every non-memorable class is also a bit like a jazz composition, namely one that was expensive, took an hour to drive to (15 minutes just to find parking) and at the end of the day wasn’t very memorable. The correct conclusion to draw from the analogy between live teaching and live music is that at their best both are great but both are also costly and inefficient ways of delivering most teaching and most musical experiences.

Edmundson also says this about online courses:

You can get knowledge from an Internet course if you’re highly motivated to learn. But in real courses the students and teachers come together and create an immediate and vital community of learning. A real course creates intellectual joy, at least in some. I don’t think an Internet course ever will.

Edmundson reminds me of composer John Philip Sousa who in 1906 wrote The Menace of Mechanical Music, an attack on the phonograph that sounds very similar to the attack on online education today.

It is the living, breathing example alone that is valuable to the student and can set into motion his creative and performing abilities. The ingenuity of a phonograph’s mechanism may incite the inventive genius to its improvement, but I could not imagine that a performance by it would ever inspire embryotic Mendelssohns, Beethovens, Mozarts, and Wagners to the acquirement of technical skill, or to the grasp of human possibilities in the art.

Sousa could not imagine it, but needless to say recorded music has inspired many inventive geniuses. Edmundson’s failure of imagination is even worse than Sousa’s, online courses are already creating intellectual joy (scroll down).

(Sousa was right about a few things. Recorded music has reduced the number of musical amateurs and the playing of music in the home. Far fewer pianos are sold today, for example, than in 1906 when Sousa wrote and that is true even before adjusting for today’s much larger population. Online education will similarly change teaching and I don’t claim that every change will be beneficial even if the net is good.)

Sousa and Edmundson also underestimate how much recording can add to the pursuit of artistic excellence. Many musical works, for example, cannot be well understood or fully appreciated with just a few listens. Recording allows for repeated listening and study. Indeed, one might say that only with recording, can one truly hear.

Recording also let musicians truly hear and thus compare, contrast and improve. Most teachers will also benefit from hearing and seeing themselves teach. With recording, teaching will become more like writing and less like improv. How many people write perfect first drafts? Good writing is editing, editing, editing. Live teaching suffers from too much improv and not enough editing. Sometimes I improv in class–also called winging it–but like most people I am usually better when I am better prepared. (Tyler, in contrast, is the Charlie Parker of live teaching.)

Sousa and the modern critics of online education also miss how new technologies bring new possibilities. For Sousa then, as for Edmundson today, the new technologies are simply about recording the live experience. But recorded music brought the creation of new kinds of music. Indeed, a lot of today’s music can’t be played live.

In his excellent 1966 disquisition, The Prospects for Recording (highly recommended, fyi), pianist Glenn Gould said that using the technology of the studio “one can very often transcend the limitations that performance imposes upon the imagination.” The same will be true for online education.

Addendum: Andrew Gelman comments.


Nailed it.

Nailed it exactly!

Doublecheck the statistics on amateur musicanship. In 1906 comparatively few people could afford even inexpensive musical instruments. If fewer pianos are sold today, it likely reflects changing tastes in music and the availability of inexpensive, portable synthesizers. Certainly there are millions of guitars sold each year.

I'd like to call attention to a recent piece of technology that has completely changed the way I learn music. A few months ago I picked up the banjo, after a lifetime of failing to achieve adequacy with a variety of instruments and a variety of teachers. Today there is a wealth of online instructional material on the web (lesson videos, tablatures, books) but the really revolutionary tool is a piece of software that will take any recording and _slow it down_ while preserving pitch. Instead of plodding through boring beginner songs with a metronome, I've pretty much started from the beginning playing "real" songs along with musical giants like Earl Scruggs. Not only is it waaaaay more fun, but it's suddenly easy to pick out the timing nuances that made the greats great.

I tried a traditional banjo teacher. It wasn't as productive as working with online material - or at least, it would be prohibitively expensive to have him spend several hours a day one-on-one with me while I practice. I still occasionally consult with my teacher when I have specific questions or problems I want to work through, but technology has radically changed the way we interact, and reduced my demand for his services. And the value of his service comes from one-on-one interaction; a large classroom environment would be pointless.

It's not just me; this phenomenon is becoming common enough that banjo teachers lament it openly. Here one thread:

This one anecdote does not make good data, but I'm nevertheless quite convinced that technology has made it far easier to become an amateur musician than ever before.

Wasted hours in a classroom >> Great moments in a classroom
Learning from casual conversation with professor >> Lecture from professor
Learning in lab working with peers >> All most other learning experiences

The moments that stick out to me is when the professors taught things that they were interested in, not what was on the syllabus, and the moments when it was a shared interest. This only happened for me with physics, and I would have preferred to have just had online courses for the remainder of the classes, due to the fact that for me youtube's explanations >> the average professor's.


Spacing sucks on this editor.

It's really a good posting. I like it. It's pretty much impressive to me. We know that “hard work always pays off”, after a long struggle with sincere effort it’s done

You have been reading David Byrne's "How Music Works", these references are in there.

Gave a quick re-read to your last piece. I guess there's a word that caused all the controversy: "better". Online ed is better than traditional ed. It is true for universities, they got big profits from online ed. But, students?

The problem arises when people that believes traditional education is better, fears the change. Let people decide what they prefer, traditional or online. People too busy to attend a classroom? online ed is better. Developing world? Online ed. But, for some people still traditional education is going to be the best. Let students decide.

'With recording, teaching will become more like writing'

Or writing was already like teaching - after all, Plato supposedly simply wrote down the teachings of his teacher. Leading to an immense number of teachers introducing those teachings to students over several thousand years. Perhaps not too surprising, there don't seem to be many few records of those teachers, yet Plato's works continue to be read, in an ever changing framework.

A framework in which writing is more timelessly universal - think about the language and format of a book, and then think how many streaming media formats existed 30 years ago.

The correct conclusion to draw from the analogy between live teaching and live music is that at their best both are great but both are also costly and inefficient ways of delivering most teaching and most musical experiences.

Teaching itself is an inefficient way of delivering education. Wouldn't books and articles do much better? If we are going online anyways, is there still any great utility to delivered lectures rather than written books, notes etc.?

I think that, while a degree of improvisation can go along well with the recording process, a major benefit comes in touching up composition, production and editing that comes with the recording process.

The raw experience is something else altogether, but it's surely not an efficient way of diffusing music.

I have mixed feelings about this subject. I love listening to people who have mastered their subjects and can provide insights I have missed. It may take me years to develop the proper wording for an argument, and a skilled teacher can bring it out in moments.

My online education is much more practical. Youtube is a wealth of information. Do you want to know how to make a world class marinara? How about fix a DLP television? How about diagnose the problem with your stove and the parts you should replace? I love online education for its "How to..." simplicity and the low cost barrier to entry.

I am seeing the some variant of the phrase "No UoP, Kaplan, Ashford type degrees..." more and more listed on job postings.

I guess employers are finding out on-line students are of a different quality.

Would TC / AT hire a post-doc / grad-student with an online degree in Econ? I think that'd be a good signal.

Perhaps, but if not it would not be because of the on-line degree per se. For a prototypical TC to hire a post doc he really needs to believe that that student has Nobel prize potential.

So, it is not the on-line degree's fault that academia is batshit crazy.

Or that on-line educational vendors are of diverse quality...I'd be willing to bet that the 200-odd students who got a perfect score in Sebastian Thrun's online Artificial Intelligence course at Stanford (none of whom were actually Stanford students) received a much different reception from potential employers.

I'll wager that a large fraction of those 200 were either students at a brick-and-mortar university elsewhere or already professionals in AI or related sectors.

That's because it is not a fair test. If you can do that, why wouldn't you go to Harvard, even if you might not in the future equilibrium?

'Would TC / AT hire a post-doc / grad-student with an online degree in Econ?'

I think the Mercatus Center hires whoever the hell they want - it isn't as if the center's contributors are that interested in degrees, after all.

"Edmundson also says this about online courses: 'You can get knowledge from an Internet course if you’re highly motivated to learn. But in real courses the students and teachers come together and create an immediate and vital community of learning.'"

He is right about the first. The second? Well, as my grandfather used to say about bad beer, "They should have left it in the horse." Most students in most courses are trying to memorize (and then forget) enough to pass the course and move on. Classmates who are in the same situation, and the requirement of getting out and seeing a live teacher every few days, helps keep many students going. I'm pretty sure the pass rate will be higher with live courses. But that doesn't say much about which modality involves more learning.

I have earned a full set of academic credentials, and I've just searched my brain for a memorable teaching experience. I don't think there is any, though it's not hard to see why teachers would like to think otherwise.

I like the analogy. Live music has a much higher retention rate (I can name all the concerts I've attended, but I obviously can't do the same with recorded music). Moreover, my investment in listening to recorded music is much lower. However, I've listened to a lot more recorded music than live *because* that investment (money, time, and energy) is much lower.

I'm highly skeptical that online education will produce results that are even close to face-to-face *on a per person taught* basis.

But if it can reach a factor of 100x more students, then it could easily be more educationally relevant than face-to-face education in all.

Great post Alex. I especially like the point about teaching becoming more like writing because I think in retrospect, this point will become so obvious. Especially for the more basic classes, the idea that these concepts are currently taught with varying degrees of competency, enthusiasm etc. is pretty ridiculous. These courses could be much better taught if a professor or even a group of professors sat down and really drummed up a coherent. well-paced video series with what amounts to a script. Of course as you and Tyler have both brought up before, the incentives for this seem in some ways unfortunate. How many instructors would be willing to "put themselves out of business" by effectively mechanizing some of their labor?

I do have one quibble with your post though. You claim that recorded music has reduced the number of musical amateurs. While I can't speak to the effect of recorded music on rates of amateur musicians around the beginning of the 20th century, I find it very unlikely that the rate of musical amateurs declined throughout the century. As you say, piano sales have declined significantly, but surely the rise of rock music and the accompanying holy grail of the modern amateur musician, the guitar, have more than made up for the decline in "classical" instruments. According to your link, in the US piano sales have decreased from 360,000 to 60,000 and over that same time period population has tripled. So are there 1.02 million "replacement" musical amateurs to make up for the scaled "lost" piano amateurs? I think probably.

I actually think that this trend is sort of relevant to your general point. There surely will be short-term tradeoffs with the rise of online universities and courses, but in the long-term really cool things are likely to happen. The rise of rock was surely accelerated by the ability to record music and its exploding popularity opened up the door to amateur musicianship to many many new people. Lectures as carefully composed works of pedagogic art may have similar positive effects.

there were very few electric guitars sold at the beginning of the century

If knowledge and processing information is gained by listening to someone, by doesn't it work with textbooks? Or any book for that matter? Teaching isn't about passing on information to passive individuals who sit on the back end of a screen. Successful teaching is engaging students by having them think critically about things and how the world works around them. Unfortunately, only about 8% of the population has a true passion for learning, and it's the successful teacher that increases that number.

Homework, exams, and evaluation result in more critical thinking, engagement, and learning than most lectures.

Without the, "now do" part, the classroom lecutre is as weak a forum for learning as the Sunday sermon is a forum for producing moral behavior.

Cyrus says:
"[T]he classroom lecture is as weak a forum for learning as the Sunday sermon is a forum for producing moral behavior."

I don't disagree with that point at all, but that's my point. The vast majority of academics don't know how to teach. Think of the best professors you had in college. (And I don't mean the nicest or easiest, but those from whom you learned the most.) I believe I can count all (including grad school) on one hand.

Taking inept academics and putting them in front of thousands of people doesn't solve the problem.

I inadvertently hit submit prematurely.

I meant to end with:

I'm not certain that you can improve on that number by passively talking to students via the web. It may work, I'm just skeptical.

If you're trying to create a better (or equal) pure learning environment, the arguments make sense.

I would point out however that college, for example, is as much about networking as it is learning. For better or worse.

One worry that I suspect a lot of people have currently is that it might be harmful to job prospects not to take the traditional education route. There may be a generation of experimenters who get screwed for their part in experimentation. This worry has the potential to adversely affect experimentation.

There is also a feeling among young people that the best path to job security is through the government, which leans even harder toward credential seeking.

A lot of great teachers may also hold off from producing online content if they cannot easily attain accreditation for their classes. In the end, a lot of the content that does attain accreditation might be far less interesting than the best jazz. Imagine what the music world would be like if great young bands had to first seek accreditation before producing their music. We might be missing nearly all the best of rock and roll and its derivatives.

"The work of art is a stuffed crocodile."

-Alfred Jarry

While I agree with the substance of this piece, "fewer pianos" is a pretty spurious indication of fewer amateur musicians. There are fewer pianos because people don't play paino music anymore. You might have heard of these things called rock and hip-hop that people seem to do in their free time, often without pianos!

Good post, Alex. I don't know how we can figure out whether online education will work until it is tried for a while. Clearly, there are classes of students who have different motives for education and different responses to varieties of teaching. I doubt online ed will work for everybody, but I feel confident that it will work for many (how many - 80%? 50% ? 25% ?)

The Teaching Company already works. It just operates tangential to the signaling/credentialing system. The only question is how on-line will assimilate into the signaling/credentialing component.

Maybe take the analogy of teaching and music a step further. Many online classes to work on the material individually - as in music practice. Then a few live classes of practical demonstrations and Q&A - as in a music performance (and improv) both for students and teachers.
This potentially uses the best of both methods.   

I'm still concerned with online education in certain situations. I don't feel that "real classes" can be effectively taught in an online format for the majority of students. Specifically, I don't feel that STEM courses (and some econ & finance) can be done justly. There are some individuals - typically of higher intelligence - who would do fine with an online format of these courses, but the average individual?

Engineering which, with all due respect, is the tip of the spear, does the vast majority of its teaching when you are alone (or in a group) doing the homework and the professor is just a vague threat in the back of your mind.

Don't underestimate the efficiency of that vague threat. Can MOOC instructors provide the same sense of impending doom?

I'd say STEM could be some of the easiest to do well online. A basic, well put together intro to the material followed by an exhaustive amount of walked-through sample problems that can be worked through until the student feels they've grasped the concept. All the questions in any of my engineering and math classes were of the "but what if-" sort. Some people were easily annoyed by these questions (they grasped the concept quickly) and some really needed the practice through examples.

Also, it's generally some of the smart kids who ask questions in class. The average student is more likely to stay quiet and ride off the questions of the smart kids. So I guess that could be an argument for in person education, but a motivated average student can work through all the examples without having to put his neck out too far.

>>"Quite right but every non-memorable class is also a bit like a jazz composition"

That's brilliant. Very well done.

I would only add that most of those extolling paleo-education techniques are just selfish louts trying to save their own hide. Just another version of Krugman, really.

This is the best analogy of the benefits of online education I've ever heard : )

On a similar note, music is an excellent way to motivate under-privileged children to find something productive that doesn't involve, say, selling drugs or their bodies for money.

So they go hand-in-hand pretty well.

For someone who is such an advocate of classroom teaching, Edmundson seems not to have witnessed in a large number of actual classes. Sure, for a small English class, where guided discussion is most of the benefit, an online class will never match even a mediocre teacher (with motivated students), but one need only walk through the halls of a university classroom building to see the failings of many lecturers when it comes to other subjects. Regurgitated Powerpoints, students on their phones, and a droning lecturer are the order of the day.

I agree with Professor Tabarrok that online education can improve the lecture (which often just rehashes what was in the textbook anyhow) and make it less expensive in the long run by eliminating the lecturer.

So there is more upfront cost to develop the lecture...a good lecture at least...but fewer costs to maintain it. Though I wonder how pricing might ultimately play out. The current assumption is that anyone can post the material online and thus the cost is trivial to create an online course. But is this true? A good online course will probably have a team working with the professor developing questions, testing tools etc. As these tools get more sophisticated the barriers to entry (to create a really good course with animations, high quality graphics, etc) may actually rise. How long did it take Tabarok and Cowen to write their own textbooks? How many people helped them do it? In the end, creating an online course will probably resemble writing a textbook more than it does preparing a lecture.

In theory, putting material online also reduces overhead costs, since you don’t need a lecture hall or fancy projectors etc. Though some of these costs are transferred to students who must pay for internet access, own a computer, etc. (I am not sure these costs are as trivial as might be assumed.)

But even assuming lower overhead and improved lectures, teaching, even without the lecturer, will remain labor intensive in most cases: You still need someone to grade papers and problem sets, someone to answer student questions, and someone to run tech support. In advanced courses this grading will be extremely hard to automate. (So perhaps most savings will be in intro courses.)

I would like to hear Professor Tabarrok comment on this (often overlooked) side of the online revolution: how it will restructure or reinforce existing labor practices within the university.

Most grading today is done with cheap graduate student labor (as I am sure is the case at GMU). Furthermore, most courses that are labor intensive (like intro writing courses) are taught by adjuncts and lecturers.

It’s not clear to me that online education will really revolutionize anything in this regard. In other words, the cost savings in labor that online courses claim they will realize are really just an acceleration of existing hiring practices at Universities. It may even worsen those practices because those teachers will be less visible and because working at a distance less likely to unionize, organize, or communicate with one another.

Yes, you can get students to grade their own work. But, of course, you can do this in a normal classroom too. We don’t need technology to do that.

This also raises the question of who creates value in a lecture course? Is it the lecturer or the TAs? Why not experiment with traditional classes without lecturers and instead TAs working with a textbook? The graduate student assigns readings and problem sets, then grades them and goes over them in section. Isn’t this very similar to what online education is promising?


Why do you think grading will be done by a graduate student at GMU?

Think India.

Because grading might be done by the grad student as part of his training. Or, it COULD be outsourced to India if that makes sense.

Academia is about performing a magic act so that people don't realize you are just doing Montessori.

I haven't seen any discussion of online education here begun by first defining the market. I think there are several markets for online eduation, each with their own opportunities, problems and cost/benefits.

1. 18-21 Year Old Market

The problem I see here is that students "stay in school", rather than quit after the first bad exam, because of "peer pressure" and the commitment of being in a dorm where you would have to leave and say good-bye.

But, what is the peer pressure to continue an online course? You just unenroll with a click.

So, for the student who is easily discouraged, needs some hand holding, responds to peer pressure to stick it out--online might not be the best for them.

On the other hand, if you had a crappy high school math teacher, and you are struggling with calc, you can supplement in college with Khan Academy to refresh yourself on that thing in math you never thought you would see again...but you did in today's calc lecture.

2. Pre-College Market

Prediction: colleges will begin screening applicants by their ability to master their online courses before they admit them to college. If you were Stanford, and some high school kid in the hinterlands scored high in the artificial intelligence class and had good SATs, would you admit that person over a person who had the same SAT score but had not taken the Stanford online course?

3. Creditialing and Testing Market

This is the real issue and problem. If you go to MIT or CalTech and took the calc exams, your employer knows you know or should know calc. The tests are hard. I am waiting for some testing organization (SAT, GMAT, College Boards, etc.) to begin offering course specific exams you can take to show (or prove) you mastered the material. This will be easy for some of the sciences, but how do you do this for English Lit?

4. Post College or Special Interest Market

This is a market also. A friend of mine, Ph.D. University of Chicago, teaches a graduate marketing class. He never had a class in graph theory, networks and network analysis, social network analysis, programming Gephi or NetLogo. Today, he is at a disadvantage, and his students are not learning some things they should be learning. So, the answer is that is taking some Coursera courses (my suggestion). No risk to him. They don't know who he is, and he likes it that way, but he is learning enough so that he can nod intelligently.

I think you're exactly right Bill.

Most people suggest that online education will spread education to the far reaches of the globe. But of course that ignores the fact that you still need the technological infrastructure to support online courses. (Does subSaharan Africa have that infrastructure? And who pays for it?)

Students who have reliable broadband access probably already live in countries (or counties) with decent educational systems.

But the real expansion will be in non-traditional students who are in "traditional" places.

"Tyler, in contrast, is the Charlie Parker of live teaching". Can anyone explain to me what it means to be the Charlie Parker of live teaching? All I know is that Parker was an innovative jazz musician.

Tyler is black and used heroin?

Good joke Bill! But seriously, what did Alex mean by calling Tyler Charlie Parker of live teaching?

Comments for this post are closed