Podcast: “This Idea Must Die Full” from Freakanomics (Transcript)
I enjoyed the conversation put forth in this podcast. The premise of the show is really interesting: what are some ideas that are commonly accepted but are no longer providing value—even hindering progress—so they need to die?
One idea that was put forth as needing to die was that statistics can entirely answer our questions and solve our problems. What’s interesting is that this is coming from a guy named Emanuel Derman, now a professor at Columbia University, who worked on Wall Street for 20 years as a quantitative analyst. Here’s what he said (emphasis mine):
The scientific idea that I believe is ready for retirement is one that’s very fashionable now, and that’s the use and the power of statistics. It’s a subject that’s become increasingly popular with increasing power of computers, computer science, information technology…which have all come together in some nexus to make people think that just looking at data is going to be enough to tell you truths about the world…
I can give examples from physics. I worked as a particle theoretical physicist for a long time, and all the really great discoveries in physics have come from a burst of intuition, which people tend to look down on these days. Johannes Kepler was an astronomer about 50 years or so before Newton, and he actually spent a lot of time studying Tycho Brahe, who was a Danish observational astronomer who collected tons of very detailed data, the position of the planets. Kepler got access to them and over 30 or 40 years analyzed them. Actually it was an astonishing feat. If you think about what you see when you see the trajectories of lights in the sky, which are planets, you see their motion relative to the Earth. But what Kepler was interested in [was] their motion relative to the sun. The Earth is moving around the sun, so God knows how he did it, but he had to extract out the motion of the earth from the whole picture and describe how the planet moves relative to the sun. How he did this without computers is quite beyond me. But in the end, his second law says that the line between the planet and the sun sweeps out equal areas in equal times. It’s an astonishing thing to say, because he’s describing an invisible line between the sun and the planet. There is no line between the sun and the planet, and yet he’s come up with this burst of intuition, which lets him talk about something you can’t see and that isn’t in the data. That’s a good instance of the bursts of insight that people have when they use their intuition to make great discoveries. There’s no understanding how he came to look at things in that way.
But what’s fashionable these days is simply doing statistics, and correlations. I don’t believe you can really find deep truths like Kepler’s Laws that are trying to describe something below the surface, simply by looking at data. It’s what’s wrong with a lot of financial modeling too, the idea that somewhere there’s a formula that will tell you how to manage risks, tell you how to price things, and absolve you of the responsibility of the struggle to actually understand the world in a deeper way.
I find this especially interesting in the web design / software world. “If there’s no data behind it, you’ve got nothing” the reasoning goes. Often there’s intuition behind it — intuition backed by decades of experience. And I think there’s value in intuition that we’re often overlooking.
I think we all know (but don’t always want to admit to ourselves) that you can use data to prove anything. I think, like Derman said, data can be the easy way out. It’s a way to “absolve you of the responsibility of the struggle to actually understand the world in a deeper way”. I think vision and intuition can be much more valuable than data (imagine what you could do with vision and intuition coupled with data). As Derman pointed out in his Kepler example, vision and intuition is what allowed Kepler to discover something that couldn’t actually be seen and that wasn’t in the data.
Now, that’s not to say that data cannot be useful. Of course it can. Which leads me to the next portion of this podcast I really enjoyed. This interview featured Alan Alda who is famously known for his role as Captain Hawkeye Pierce in the TV series MASH. He also loves science and has been working at the Center for Communicating Science at Stony Brook University for the last 25 years trying to teach scientists to communicate about their work in a way that ordinary people can understand. During the podcast, he questioned the premise of the show that ideas need to die:
It’s eye-catching to say this idea must die, and I’m not sure that [many of the ideas here] need to be retired … or just rethought. Therefore, I would say that asking for these ideas to be retired is really a way of saying, “This is the received wisdom. Do we need to reexamine it?” That’s a good approach to take.
Which is the point I’m trying to make about data: maybe we’ve gone overboard in thinking data can solve all our problems. Maybe the idea that data can solve problems is true, but only in some contexts and not in others. It’s a true statement — some of the time. Alda expounds on this (emphasis mine):
The idea that maybe is due for a rest — you notice I said, “Needs to take a rest,” I didn’t say it needed to “die” — is the idea that things are either true or false. I know that’s an impertinent thing to say and it sounds stupid. But what I mean by it is the idea that something is either true or false for all time and in all respects. I think about this because when I was being taught to think in school, I was taught that the first rule of logic is that a thing cannot both be and not be at the same time and in the same respect. And that last part, ‘in the same respect,’ really has a lot to do with it because something is determined to be true through research. Then further research finds out that it’s only true under certain conditions or that there are other factors that are involved.
Here’s a very interesting example: a lot of people were interested, I know I was interested, when I read that red wine was good for you. At first, we might have even thought, “The more red wine the better. Look at all that antioxidant stuff going into it.” But it was a terrible disappointment sometime later when some other scientists said, “Under certain conditions red wine could be not so good for you.” Again, there’s this other thing that it might be really great for mice and less good for us. But what really disturbs me is when the public decides that that means that science can’t make up its mind, or that scientists are just making things up. Some people actually do think that. Some people think the findings in science are hogwash because if one day they say one thing and the next day they say another thing it sounds like they just are taking wild guesses at things, when in fact, the progress of science is just that.
You go deeper and deeper. You open up one door and you find another hundred doors that have locks on them that you have to figure out the combination for. I personally find it exciting to see what we thought we understood to be contradicted, but I don’t think the public has enough of a grasp of how science is done, how it’s based on evidence. When you say this is true, in the mind of the person receiving that information, they’re going to accept it as true for all time, under all circumstances, unless you warn them that things might change in the future. We might learn more about this. That shift in the frame of reference is something that ought to be allowed for. I want to see science prosper. I want to see evidential thinking be the norm for the public as it is for scientists. So my suggestion that we alter the way we talk about things being true or false is really to help in the communication of science so that people don’t get confused.
That last line feels extraordinarily relevant to design and technology. We need a way of altering the way we talk about things being right or wrong, true or false, because they are perceived as “this is the right thing to do at all times in all places in all circumstances”. That’s everything from “how to create an artboard in sketch” to “how to iterate over an array” to “what framework you should use in your app”. I think we (and I) need to be better at relating to our audience. Because so much of our conversation on these types of things happens online, we miss the opportunity to observe the person we’re listening to. We miss reading their face and tone of voice. We overlook trying to read their mind, their intent, and that’s the basis of good communication. As Alda says, “You’ve got to know what’s going on in the mind of the person listening to you to know if you’re getting through to them or not.”
Even simply writing this post, I can see I need to practice my own form of communication and what I mean when I say “I liked this podcast”.