Freakonomics Radio - 301. What Would Be the Best Universal Language? (Earth 2.0 Series)

Episode Date: September 21, 2017

We explore votes for English, Indonesian, and … Esperanto! The search for a common language goes back millennia, but so much still gets lost in translation. Will technology finally solve that?

In our previous episode, we talked about living under the ancient curse of the Tower of Babel. So the curse of Babel is an existential condition in which we live every day. We use language to communicate, but we cannot rely on it to make ourselves understood. We can't always rely on it because... Well, we have 7,000 languages. 7,000 languages? We learned about the many costs associated with this linguistic diversity, financial costs, psychic costs, even war. So many people died in the war, which in fact easily could have been avoided. And we learned that linguistic diversity has plenty of benefits, too.

There are certainly claims about types of thinking that become very hard without language or become unlikely without language. Those are some of the things we know about language here on Earth 1.0. But today's episode is part of our Earth 2.0 series, in which we imagine we could reboot the planet and do some optimizing, or at least some tidying up. So, if we were starting over, what would be the best way for everyone on Earth to be able to communicate with one another? From WNYC Studios, this is Freakonomics Radio, the podcast that explores the hidden side of everything. In the future, human-to-human communication may be so different that it'll render our mission today moot. Between auto-translation and artificial intelligence,

maybe even mind-melding, will anything ever get lost in translation? Maybe. Maybe not. But that's the future. Let's talk about language on Earth 2.0 using the tools and knowledge at our disposal today. If we could start from scratch, what would that look like?

I somehow think if we did this from scratch, it would be a very surprising outcome. It's Michael Gordon. He's a historian of science at Princeton. And who knows how it would work without the path dependency of previous empires, current economic structures, our current modes of transportation, and media and communication. It would be very interesting to see how that would shake out. Okay, let's start with a couple basic questions. Number one, should we consider, please don't throw things at me, should we consider having one common language? I would be wary of thinking of common language as the solution to perfect communication. Lara Boroditsky is a cognitive scientist at the University of California, San Diego.

Because we already have common language and that doesn't lead to perfect communication. You would need, oddly, a language that had a lot less in it than many people would expect. John McWhorter is a linguist at Columbia. He's also an author and host of the Lexicon Valley podcast. You want it to be something that's maximally easy for all of the world's language speakers to use. And so you could have a universal language where tense was largely left to context, as it is in a great many of the world's normal languages. You certainly wouldn't have anything like grammatical gender. The vocabulary could be quite rich. That would be fun. But the grammar would be something where you could pick it up in a week. I'm curious to know the degree to which language generally is

things like utilitarian, like I want to pick up that thing, or transactional, I want that thing from you, or romantic or relationship or gossip or lying and so on. And I'm just curious how a linguist might think about that. So what language is, is more than questions, commands, and certainly more than just naked statements. Real language is about communication and charting feelings, telling people new things. And that means that a language is a whole lot more than just nouns, verbs, and adjectives. So if somebody says, oh, she's totally going to call you, that totally means you and I both know that other people think she isn't going to call, but we have reason to think that she is. We are full of things like that.

Okay, this leads to question number two. If there were a universal language, should it be a pre-existing one or an invented one? English, while hardly universal, has of course become a very powerful language. And what makes this regrettable to many, and quite understandably, is that English was the vehicle of a rapaciously imperial power, and now America is the main driver. So any pre-existing language will come with baggage, with lots of votes for and against. Does this mean we'd be better off inventing a new one? Apparently, some Facebook bots recently gave it a try.

According to several reports, Facebook's artificial intelligence researchers had to shut down two chatbots after they developed a strange English shorthand. The bots were supposed to learn how to trade virtually. A shorthand that its human creators couldn't understand. As it happens, the dream of inventing a universal language has long been pursued by scholars, priests, even, as you'll hear, by an ophthalmologist. So the history of language invention, which goes back millennia, has to do with reversing the curse of Babel. Esther Shore is a professor of English at Princeton.

In other words, to return the world to a single language of perfect understanding. And for some language inventors, this was imagined to be God's own language, the language of divine truth. In the 13th century, for instance, Ramon Llull, a Mallorcan philosopher with Franciscan ties, sought to create the perfect language for channeling the truth and converting people to Christianity. He created a formula for generating propositions from letters and words. And he felt that some of them would be propositions to which an infidel would of necessity have to consent. But his truth was not the truth, or at least it didn't seem like the truth,

to the Saracens who eventually murdered him. A few centuries later, the German philosopher Leibniz, an admirer of Juhu's, by the way, tried to build a language based on logic. So Leibniz's idea was to represent propositions by numbers. And he would reason by getting the ratio of one proposition to another and just sort of calculate an answer. And again, we have the idea of a language of logic without words.

And in the 19th century, a Jewish ophthalmologist named Ludwig Lazarus Zamenhof created a language both idealistic and pragmatic. It's called Esperanto because that was his pseudonym, Doctoro Esperanto,

which means the hopeful one. And he brags in this initial pamphlet that you can learn it in an afternoon, you know, and that it's fun. So it was supposed to be easy to learn and easy to pronounce. Esperanto was derived from various European roots. Zamenhof's idea was not to have Esperanto displace other languages. He called it a helping language or an auxiliary language. It would stand next to national languages

and be a helping language to make bonds among people who were not like one another. Zamenhof was a universalist. But he was also a universalist who understood what it meant to have warm feelings for one's people. And Esperanto was to somehow reconcile those two things. It was to try to breed in us these feelings of attachment for other people who were really quite unlike us. The larger goal of Esperanto was nothing less than world peace. He knew that language could be a wall between ethnicities, but that it could also be a bridge.

And that was his motivation, to build a language that would be a bridge among ethnicities. And he modeled it on the teaching of Hillel, do not unto others as you would not have them do unto you. Hillel was a first century rabbi, so it had a very Jewish cast to it. that Zamenhof shear away its religious ideology. Hitler and Stalin would also reject Esperanto. Regardless, if you remove its religious and utopian components, what's left, Esther Schor says, is a language with some substantial benefits over many other languages, whether existing or invented. What he wanted was maximal flexibility and simplicity. So for one thing, the verbs are all regular in Esperanto. He wanted a language that was egalitarian and neutral. He didn't want people to be disadvantaged because they weren't a native speaker. And he speaks very movingly about what it's like to try to speak a language that's not your own.

And he talks about his pulse racing and his palms sweating. It's an experience I've had. Perhaps you have had it also. So Esperanto is a lot easier to learn than other languages because it has very regular rules and very regular grammar. That's Ruth Cohen. She helped develop an online Esperanto course for the language site Duolingo. And that's where you find that it's taking you a lot less time than you thought to learn the language.

So here's a sentence in Esperanto. Mi estas knabo. I am a boy. There is no a in Esperanto. Mi estas knabo. I am a boy. There is no a in Esperanto. Knabo, you can see, is a noun because it has an o at the end. Every noun ends in the letter o. Every adjective ends in the letter a. Every verb in the present ends in a-s, as. So you already know that. Estas is is, am, are. It's the same. There's no conjugation of that. We spoke with Cohen at this year's

Esperanto USA National Congress, or Landa Congreso, as you say it, in Esperanto. Our producer Stephanie Tam spent a couple days there. You will hear about that in an upcoming special episode. You may be surprised to learn that Esperanto is still spoken. Esther Shore again. These days, the most informed estimates I hear are several hundred thousand people speak Esperanto. And the strength of Esperanto is in its continuity over 130 years in 62 countries from generation to generation without being passed down from generation to generation.

Still, for all its thoughtfulness and pragmatism, Esperanto never got anywhere close to its intended universal status, what Esperantists refer to as La Fina Venco, the final victory. Why not? Well, I can answer that by looking at what does look like a universal language in our world, which is English. What looked like a universal language in Zamenhof's day was French. And both French and English were propelled into the world by commerce and armies. And Desperanto had neither of those. In order to keep a language constant enough so that it can function as a global, universal language, the way English is functioning now, you need to have a global communications infrastructure that standardizes dialects and pronunciations. Michael Gordon again.

You need to have a global entertainment industry that produces books with standard spelling and a pattern of accents that are considered acceptable or that mark different classes or regional identities. And that kind of constant reinforcement requires an infrastructure. It's something we don't think about. At least I'd never thought about it. But there's a lot of upkeep associated with language. When classical Chinese was being used as a lingua franca for a very broad region, it was used in Japan and Korea and Vietnam as the language of written communication, a very strict civil service exam system, privileged learning the language to precision. And that stabilized that language. And to a certain extent, the anglophone

entertainment publication and media industry, as well as the scientific institutions, stabilize a certain kind of global English now. Gordon points to another factor that would make it hard to install a universal language, the nature of language itself. The reason why I think you can't just blanket install and say, okay, everybody's going to learn Esperanto is because people will experiment and mess with the language. They'll change it. Which, by the swell, but because as you use a language over time and you pass it on to new generations, brains tend to start hearing things slightly differently than they were produced. And after a while, you start producing them that way. And that is as inherent to language as it is inherent for clouds to change their shapes. And so it isn't that that happens to

some languages and not others. That's how human speech goes. All right. So imagine in our thought experiment now that we've got Earth 2.0. So you've got, you know, seven, eight billion people. And let's say we want to give everybody the most prosperity and opportunity and equity it's possible. And let's say that we make you the chief, let's say, communications advisor of Earth 2.0. And we give you the task of writing the plan, the blueprint for creating from scratch our new language systems and institutions. So what would that blueprint look like? I would say that an ideal in the future is that everybody in the world can communicate in one language, and that then people have another language that they use with their in-group, and that we have as many of those languages as possible. I don't think that it's going to be

another 6,999 ever. But there does need to be one language that everybody uses so that as many people in the world as possible can take advantage of economic benefits such as they are. I would go with global language on some higher level. That's Shlomo Weber, an economist who studies language. But still keeping the local language for everybody because, you know, sensibilities of the people are just very important things.

Let's say this Earth 2.0 experiment, just to be a little more realistic, that we're still working with the resources we've got. In other words, the languages that exist now would still exist. So English obviously has a big, big, big head start, but it obviously also comes with a lot of baggage, right? People learn English because it's useful, but English has a history of, you know, colonialism and domination and so on. So would picking a language like

English just doom it to failure? I don't know. Most of the languages, maybe except the Chinese, have the history of domination too. That's true. Does that mean you're nominating Chinese because they took the Middle Kingdom route and they never really tried? No, they definitely would be one of the leading languages,

absolutely. But, you know, I think we could have chosen six or seven. To choose one, it's really a very difficult thing. And of course, the colonial legacy of English is questionable, but it's true for many others. So the history of Russian language, of Japanese, of French, of German, obviously. Turkish empire had also its ups and downs. But really, given our circumstances, English, again, reluctant word for English. I almost wish that there was some reason that everybody had to learn colloquial Indonesian because it's the only language I've ever encountered where you can learn a whole bunch of words.

And even though you're going to sound like an idiot, you can get an awful lot done. You don't sound nearly as much like an idiot stringing together your lonely planet words. In many parts of Indonesia, there's no such thing as the moon being a girl and a boat being a boy. None of those things that make languages hard to learn. Really, almost none. And I thought this should be the world's universal language. Indonesian is one of those languages like English, which has been learned by so many different people speaking so many different languages that it's relatively user-friendly as languages go.

You've argued that isolation in a language breeds complexity. So considering that English is the least isolated language there is these days, it's everywhere, does that necessarily mean that it will or is becoming less complex to make it accessible to newer users all over the world? It doesn't mean that, but only because this business of languages being more complex when they're isolated and becoming simpler when they're spoken by a lot of adults is largely something that happens before widespread literacy. And so English didn't become relatively user-friendly because of the Bosnian cab driver in New York.

It happened when Scandinavian Vikings flooded Britain and learned bad Old English but were dominant enough that generations started speaking the way they did. That became the language. And so you and I right now are speaking really crappy old English, and we feel fine about it. Speak for yourself. I feel I've been pretty literate today. So, see, I didn't even use the right word for literate. Literate is written, right? I can't even think of the right word for what I'm trying to say. What do you call it when I'm being... Well, articulate, I suppose. Articulate, yeah. I couldn't even come up with that. That's how bad I know you're right. I just proved your point. You know what that was? That

was Mupri's Law. Do you know Mupri's Law? No, what's that? Mupri's Law is whenever you try to correct someone's mistake, you make an additional mistake. I didn't know there was a name for that. Of course there is, because our language is so rich, of course. It is exactly that. As rich as our language may be, there's still plenty of room for improvement. Coming up after the break, let's say we bit the bullet and went with English as our universal language. How could it be made more accessible and equitable? Oh, easy. Magic wand.

Something that we must get rid of is linguistic prescriptivism. And let's not overlook how much technology is already changing our communication. It's not going to be a babblefish that you stick in your ear and will translate everything immediately. But it does improve the possibilities of translating roughly between language groups. That's coming up next. On Freakonomics. On Freakonomics Radio. On Earth 2.0, it might be nice if all 7 plus billion of us spoke one shared language, and then, as John McWhorter suggested, and that then people have another language that they use with their in-group,

and that we have as many of those languages as possible. This, McWhorter says, is pretty close to the way a lot of people already communicate. If you think about the typical person who speaks Arabic, for example, they almost certainly speak two different languages. There's the Arabic that we would learn in a book, and then there's either Moroccan Arabic or Iraqi Arabic or Sudanese Arabic, Libyan Arabic. Those are completely different languages from standard Arabic. Different basic words, different grammatical constructions.

And so you grow up speaking your Libyan Arabic. That's, you know, mommy's language. And then when you go to school, you learn something that often I've heard people from these countries also call Arabic. And that's this other language. That happened because of history, because of cultural history. In the case of Arabic, the Koran and the religious unity of the nations has a lot to do with it. But ideally, nobody would have to go to school to quote unquote, learn Arabic. And that sort of thing is going on in many South Asian countries.

It's what a typical African often has to go through. Or you know, if you're Sicilian, you speak Sicilian. You go to school and you learn Italian. Okay, fine. But then there's the task of selecting the universal language. Michael Gordon of Princeton. Even if we picked a universal language that was neutral, politics being what it is, and I doubt this could be engineered away, we'd find ways to particularize the previously general. But it'd be interesting if there was some sort of academy that were designed to keep people from making it more complicated. I love that the linguist is coming up with the

academy to keep language from becoming more complicated because you guys are the ones that have contributed, obviously, to the way we think about languages so complicated. See, we contain multitudes. It might be helpful to look at some of the countries that already use formulas calling for two or three languages. The Indians are actually in some other countries, in Nigeria, Kazakhstan, they try to implement this formula. The economist Shlomo Weber. They try to combine all these things. So every child has to study his own language, English, and the language of the other part of the country.

So everything beautiful, you bring national cohesiveness, you bring efficiency through English, and you still sustain your individual languages, your individual attachment, your identification. But it didn't work because the people didn't accept this formula. And why didn't accept it? Because the attachment to own language was much stronger than doing anything else. I thought that Kazakhstan worked better than,

let's say, India or Nigeria. So what did Kazakhstan do or what happened there that made it work better? Well, they have a strong government there. But, you see, in the case of Kazakhstan, I think the people were convinced that this is the right way to go. Because in Kazakhstan, with its oil and gas resources, English is very important to be a part of the international community. Of course, Kazakh language is important as their own language, but they also recognize that for cohesiveness of the country, Russian is an important language.

But you're also suggesting that authoritarianism is handy if you want to get everybody to speak the three languages, yeah? Because democracy is a little sloppier. Well, a little sloppy in this regard, right? Some other advantages, but not that, yeah. To be fair, there are a lot of differences between Kazakhstan and India. India is much larger, much more diverse. Even so, says Michael Gordon,

you have to give people a reason to want to engage with the language. The energy required to learn a language is high enough that you really have to work on the motivation. And the constructed languages and the natural languages provide lots of examples of the importance of that. Okay, so how do you get people to engage with a language? As we've seen on Earth 1.0, most of the big legacy languages come with a lot of baggage, cultural baggage at least, more likely colonialist baggage. So what would happen if we chose English as the new universal language? I mean, with one and a half billion speakers, it's already 20% of the way there. What would you do to make English truly accessible to everyone, especially non-native speakers?

Something that we must get rid of is linguistic prescriptivism. And by that, I mean that we live within an idea that some ways of speaking a language are bad, broken, and some ways aren't. It's all based on myths. And that's not to say that in a formal situation, you can get up and say, Billy and me went to the store. In the 19th century, the standard by which people had to know a language, a foreign language that wasn't their own. So let's for the moment pretend like everybody in the world speaks either French, English or German. You had to be really fluent in one of those three but only pretty competent in the others, a much weaker level of fluency. So the French person didn't have to know a lot of English but they had to be able with a dictionary to puzzle their way through a scientific article.

So you could relax the assumption that everything has to be perfect grammar book English and just allow the publication of rougher English in a variety of forms without this obsessive copy editing. That would be fairer. I mean there are some kinds of English that would be so difficult for anybody else to understand that maybe there would have to be some adjustment. But schematically, the idea that most people in most nations have to learn a form of what they speak that requires effort to master, that's crummy. You could imagine subsidizing global English education. Another fair option is to say, no, we actually language and the arts together, for bringing language and engineering together. And this would have to come from some kind of organization or donors, of course. But that's as much of an institution as I would like to imagine

negotiating language in Earth 2.0. I would like to have a piece on this planet and then to approach those things. What do you think would be a better way for everyone in the world to learn English? I'm especially curious to know, as an economist, what you think is the ROI on an education dollar versus like an entertainment dollar. In other words, would it be better

just to have all Hollywood movies distributed globally for free? Would that be the best way for people to learn English? It could be the case. You know, once again, example of India, where the Bollywood, actually the Bollywood movies contributed to tremendous development of India. You know, the language was not spoken very widely in India before the development of Bollywood. I mean, maybe like even five years from now,

a technology like movies will seem very old-fashioned because, you know, there may be technology that's essentially instant and perfect translation from any language to any language, right? Of course, of course, the technology will play a part. Machine translation, I think, will never be perfect. It's not going to be a babblefish that you stick in your ear and will translate everything immediately. But it does improve the possibilities of translating roughly between language groups. It's not going to be a babblefish.

The babblefish is from Hitchhiker's Guide to the Galaxy, by the way, that you stick in your ear and will translate everything immediately. Maybe not, but maybe. A New York startup called Waverly Labs has been working on a babblefish-like earbud that will do live translation. They say they've already taken in $5 million in pre-orders. There's also the rapidly developing Google Translate and Skype Translator. And it's not just major languages that benefit from the digital revolution.

Oh, I don't think there's any doubt that technology has been a great boon to Esperanto. Esther Shore again. And I know many Esperantists, especially in the United States, who essentially live their Esperantic lives online. Some of them Skype, some of them do it on Facebook. Lernu.net has several hundred thousand registered users. And there's also Duolingo, which in the past two years since its inception, it has signed on about a million people into the Esperanto course, which is really amazing and marvelous. But overall, the Internet is dominated by what John McWhorter calls the big dude languages, especially English. Google searches in English return roughly four times more results than Arabic searches.

Ninety-five percent of Wikipedia concepts are represented in fewer than six languages. There is, of course, no guarantee that this march toward English hegemony continues. History shows us that language is inherently mutable. So what can we assume about the future of language? Since we're not changing the biology of humans, we can assume a couple of things. Michael Gordon, the historian of science from Princeton. That people will learn languages, that they'll learn them pretty well when they're kids, and that languages won't stay stable. And if you want a more broadly communicative, more inclusive infrastructure, you should focus on training children while they're young and still able to

learn multiple different languages and keeping them straight. In the 19th century in Bohemia, the Czech region of the Habsburg Empire, it was quite common for neighboring peasant villages, one of which was predominantly German-speaking and one of which was predominantly Czech-speaking, to send kids to be educated in the other town. So that way the kid would know both languages. That kind of leveraging the way children can soak up languages almost effortlessly in a way to create a kind of more dense web of people who understand each other's languages would improve some aspects of the system. But here's the thing. However judiciously we might draw up the best course

of language for Earth 2.0, the original blueprint is unlikely to hold. Language evolves. It diverges. It constantly sparks its own offshoots. Consider a recent group of languages that were created from scratch. Computer languages are very definitely created. And so somebody sits down and says, this is the way we want to have our language work. Brian Kernaghan is a computer science professor at Princeton. He used to work at Bell Labs, the famous incubator of various operating systems and coding languages. Kernaghan himself worked on the Unix OS and the languages AUK and AMPLE. The first major programming languages were invented in the late 1950s.

The first high-level languages, I would say, would fundamentally be Fortran, COBOL, BASIC, and a language called ALGOL, which was in some sense more an academic exercise. These languages were built for different tasks. Like scientific and engineering computation, which was Fortran, or business computation, which was COBOL, or even educational computation, if you like, which was BASIC. They're definitely created for a purpose as opposed to being a natural process. On the other hand, once they're created, then there's a pressure for them to evolve. Just a few years later, in 1961... In 1961, a professional journal called Communications of the ACM, in their January issue, had a cover piece of art which showed a schematic version of the Tower of Babel. And it listed on that probably 200 programming languages.

The message was, boy, there's a lot of programming languages. Today, there are at least 1,500 programming languages. Do we need that many languages? Of course not. Do we use that many languages? Actually, no. The repertoire of sort of most journeyman programmers is probably half a dozen to a dozen or something like that. The parallel between programming languages and natural languages is not perfect, but still striking. A new language costs time, effort, and money to create, to learn, to maintain. Why, then, has there been so much growth?

People are trying to write bigger programs, and they're trying often to address programming problems that is taking on tasks that were not part of the original. And therefore the language evolves because the environment in which it lives is changing, the resources that are available for programmers, that is hardware resources, are changing. And the desires of the people who write programs

change as well. Or an optimist would say, developing into varieties of pronunciations and accents that display the diversity of who we are. Michael Gordon speaking now about natural languages. And that process we've seen over world history many times as things fragment and then they coalesce again and they fragment and they coalesce again. Part of that has to do with tribal tendencies. Part of it has to do with just love of experimentation, regional loyalty,

something that sounds aesthetically interesting. And you can end up with something like a guy writing a poem in the late medieval period in the Tuscan dialect, Dante producing a standard for a language by the act of his particularity. This kind of change can create chaos, but it's also a hallmark of being human, a dissatisfaction with the status quo, a desire to experiment, to build, to adapt to changing circumstances. We're champions in the animal world at creating our own niches, right? So taking the environment that we're given and then radically transforming it to suit our needs. That's the cognitive scientist, Lara Boroditsky. And we do this with language as well. And what is Boroditsky's vision for language on Earth 2.0?

My emphasis would be on preserving diversity and preserving flexibility. So making things really easy to learn and really adaptable to environment rather than focusing on making something that is exactly the same and common across everyone. I don't know that we can judge that we now have the best solution and we should just build it right in. I'd still want people to learn lots of things through cultural transmission and adjust to their environment the way that we do so well as humans. And so in some ways, becoming more aware of the relationship that we have with language is the thing that helps communication

more than simply trying to build one system. It probably hasn't escaped your attention that just about everyone we've heard from in the series on language has been an academic. They, like all tribes, have their own dialects and sub-languages, which is often not all that decipherable to the rest of us. I asked Shlomo Weber about this. He's an economist. Yes, at the moment I'm the director of the New Economics School in Moscow. So I have to tell you, I love academia. I love academics. I love the research you do. But my one big complaint is this. The way that you academics communicate to the rest of us,

to the non-academics, is terrible. Now, I understand these are areas of technical expertise, but this strikes me as its own little Tower of Babel where, you know, there are academic researchers all over the world doing this amazing and valuable research, which, by the way, is often funded by us, the taxpayers, and yet we can't really participate in it because of the way that you all communicate. So I'm curious to know if we can't solve the language or communication problem globally, if we could at least address this problem. Believe me, Stephen, I agree with you. You know, I am doing my small part. I try to write newspapers. I go on television to talk about general things and not using the language. But I think it comes back to economics.

There are incentives, and the incentives are not to go to tell you about this research. There is nothing in my incentive mechanism what university or community offers me, go to talk to people who are interested in some kind of simplified version of this research. So, for this, you really need to grow as an individual and to understand that, indeed, your research is supported by your dollars. Well, I will say this. Honestly, as much as I complain about the gap, I'm grateful for it because I wouldn't have a job if you guys communicated directly to people, because basically I am the translator, so keep doing what you're doing, keep doing what you're doing

Shlomo, okay? Thank you, Steven, you keep doing what you're doing Thanks, as always, for listening to Freakonomics Radio. Coming up next time... Larry Summers is a Harvard economics professor, but he's also a former president of Harvard, a former secretary of the Treasury,

and he was the chief White House economist under Obama when the Great Recession hit. What was that like? It was a very tense time. House economist under Obama when the Great Recession hit. What was that like? It was a very tense time. We would meet with the president each morning and talk about what was happening. Summers gives himself and his team a crisis grade. While battlefield medicine's never perfect, I think you'd have to say that the approach we chose was effective.

Summers also sort of admits a past policy mistake. Perhaps given what happened, you can say it was a mistake. And Summers also reveals, big surprise, that he is not a fan of the current White House. It's the disregard for ascertainable fact and disregard for analysis of the consequences of policy actions. That's next time.

That will come in the next episode. Episode 7 of Radio Freakonomics. Coming soon on Freakonomics Radio. Also, keep your ears open for our upcoming special episode with producer Stephanie Tam about modern day Esperanto. Freakonomics Radio is produced by WNYC Studios and Dubner Productions. This episode is produced by Stephanie Tam. Our staff also includes Allison Hockenberry, Merritt Jacob, Greg Rosalski, Eliza Lamber, Emma Morgenstern, Harry Huggins, and Brian Gutierrez. We had help this week from Sam Baer. Special thanks to our intern, Kent McDonald, and to the many listeners who contributed their voices and their languages to this episode. The music you hear throughout the episode is composed by Luis Guerra.

You can subscribe to Freakonomics Radio on Apple Podcasts, Stitcher, or wherever you get your podcasts. You can also find us on Twitter, Facebook, or via email at radio at freakonomics.com.

Merci beaucoup. Spasibo bolshoia. Dankan trei molta. Toda raba. Domo arigato. Senka anjanji. Toda raba. Terima kasih. Shukran jazila. Muito obrigada. Goodbye. Thank you. Goodbye. Goodbye. Thank you.

Thank you. Thank you. Thank you.