爆料王

(music plays)

Chris Neff:听You're listening to听Open for Discussion,听a 爆料王 podcast that discusses research through a personal and critical lens. I'm your host Chris Neff.

In 2015 my guest won accolades for his research on how the word 鈥榟uh鈥� is essential to human communication. They built on that research to show that no five minutes of conversation go by in any language without someone attempting to fix a misunderstanding.

So, from studying how we strive to understand each other with precision, I'm intrigued to know how our guest is now a member of a research initiative examining how fake news, alternative facts, lies and propaganda are spread and how we detect them.听

Professor Nick Enfield is Chair of Linguistics at the 爆料王 and Director of the Sydney Social Science and Humanities Advanced Research Centre. Thank you for joining us today Professor Enfield.

Professor Enfield:听Thank you for having me.

Chris Neff:听Nick it seems a long way from field work in Laos to examining fake news. How did studying linguistics get you here?

Professor Enfield:听(laughs) Well you'd be surprised. I think it's not actually that far away. My interest in linguistics really sort of鈥�was you know kindled through seeing my professors go out into the field and do research on endangered languages. And so that was exciting to me - the idea that you could travel across the world and go to a very different type of society and culture and see how people speak there.

听

And once you're doing that, well you see language from the point of view of its social utility. Not just to convey information but to create alliances and to get things done and the problem of truth is right at the core of that really or at least it's a crucial element of that.

Political language and the language you know of sort of media and that kind of thing operates on much higher scales but deep down it's all to me the same stuff.

Chris Neff:听Listen I love this stuff. Before we talk about deliberate misunderstandings and language, I'd like to discuss the work you've done Nick that points to just how much subtle negotiation or adjustments take place to understand each other in one-on-one conversations. This 90 second threshold. Do you mind explaining that a bit?

Professor Enfield:听I think actually the precise figure is 84 seconds but you know it's much of a muchness. About every 90 seconds on average across the languages of the world in free-flowing conversations, someone will say something like 鈥榟uh鈥�, 鈥榳hat鈥�, 鈥榳ho鈥�...this kind of a question that gets the other person to back up and either repeat or clarify or rephrase what they just said.听

You know this is a fairly remarkable statistic in a number of ways. In one it shows that language is far from perfect. That actually a lot of the time we don't quite succeed. What it also shows us though is that people are willing to help the other person by backing up and doing the required repetition and the required sort of rephrasing and so on.

And this points to a strong theme in a lot of the work that I鈥檝e done which is that language usage is a form of cooperation and there鈥檚 a true sort of core of cooperation in language usage and it鈥檚 essential to human cognition and to the way that humans interact. So, it鈥檚 not the case that in some languages people ask for clarification you know鈥�all the time, every 10 seconds and then in other languages they do it maybe every 10 minutes. You just don鈥檛 get that. You get across the languages which we鈥檝e looked at, which is from quite a broad sample from around the world, a very similar sort of rate and set of strategies for doing these kinds of requests for clarification.

Chris Neff: I love the idea of conversational repair. That there is this dance that鈥檚 going on when people are having a conversation and you know every 84 seconds or so they鈥檙e figuring out what the other person鈥檚 saying and that that鈥檚 a universal principal that鈥檚 applied across languages. This is really profound information and I saw that you won the Ig Nobel prize for this at Harvard University. How was that experience?

Professor Enfield: That experience was really interesting in a whole range of ways. So, the Ig Nobel prize it鈥檚 for research that first makes you laugh and then makes you think and the prize was for a finding from a large project that all languages that we have evidence for have the same word 鈥榟uh鈥� for you know鈥�you have no idea what the person just said in any way at all.

And we found that in every language we looked at, the word鈥�one word that people could use to do that is 鈥榟uh鈥�. The same sound occurring in all these languages which is remarkable because languages normally have very different sounding words for the same types of objects or the same types of functions.

Later on, we found these other much more nuanced findings about things like the 90 second threshold. My understanding of that prior to before we won it was that it is kind of jokey, right? It鈥檚 a kind of parody prize in a way. You know鈥�on the one hand you feel like yeah you don鈥檛 really want to win that right because it鈥檚 sort of make fun of your work and this type of thing. It鈥檚 actually become prestigious so yeah it was really exciting to win that and it was obviously great for our research. It鈥檚 given this project a lot more of a profile and I think it was very good for the subsequent work that came out.

Chris Neff: I love this stuff. Can I go down the rabbit hole with you for a moment here?

Professor Enfield: Sure thing.

Chris Neff: Because if, from what I read, when you looked at languages this also included sign language.

Professor Enfield: That鈥檚 right.

Chris Neff: I don鈥檛 know why my brain went here but it went to like Coco the ape鈥�like Michael Jackson鈥檚 ape that learned sign language. If humans teach sign language to animals then would that 鈥榟uh鈥� or the conversational repair translate into their vocabulary as well?

Professor Enfield: Yeah that鈥檚 a good question. It鈥檚 one that would seem natural to ask and the answer that I give is a definite no. That animals will not use a signal like 鈥榟uh鈥� and there鈥檚 a very particular reason for it. So, you know鈥�animal communication systems are very complex in a whole range of ways and you know鈥�there鈥檚 visual signals and there鈥檚 vocal signals and there鈥檚 all manner of signals. But most of the time those signals are sort of broadcast. Maybe they were directed at another individual or maybe they were just directed at whoever happened to be looking or listening. As in a bird鈥檚 sort of territorial call or what have you.

But what you never see in animal forms of communication is a truly cooperative type of exchange. You know鈥�this is exactly what鈥檚 exemplified by the 鈥榟uh鈥� type phenomenon. So, when I say 鈥榟uh鈥�, you don鈥檛 just keep going on talking right? You will cooperate with me by backing up and saying what you just said again. So, there鈥檚 this real sort of high level cooperation that takes place there.

And so even in animals that have something that looks a little bit like turn taking, so there are various primate species and birds also have something that looks like a turn taking system, you never get this kind of possibility of one saying something like 鈥榟uh鈥�. So, turning to the sign language question that you asked, teaching a primate鈥�an ape to use signs is not the same as teaching them to use sign language.

Chris Neff: Mmm hmm.

Professor Enfield: So, they鈥檙e able to use signs and sometimes in interesting ways they might combine those in similar ways to a child of between one and two years of age. But there鈥檚 nothing like the sophistication in those cases of an ape using signs. Nothing like the sophistication that you find in human sign language.

Chris Neff: So, Nick, can I ask just a really basic question? Do you have a favorite word? Could you tell us what your favorite word is and why?

Professor Enfield: That鈥檚 a kind of typical question that a linguist would be asked, what鈥檚 your favorite word.

Chris Neff: Aww.

Professor Enfield: No, I like it though. I mean we can talk about it. So, one answer would be you know鈥�I love the word 鈥榟uh鈥� for the reasons we鈥檝e been talking about but of course there are plenty of other sorts of ones. So, I was recently on a panel discussing correct grammar and correct punctuation and proper words and this kind of thing. And one of the members of the audience was complaining about the lack of distinctions that were being made by you know鈥�young people today.

And the example that came to mind to sort of counter that was to say well let鈥檚 take another example where in English, a second person pronoun, 鈥榶ou鈥�, doesn鈥檛 distinguish between singular and plural. So, if I say 鈥榶ou鈥�, you don鈥檛 know if I mean one person or multiple people and 鈥榶ouse鈥欌��you know in American dialect you鈥檝e got you all or y鈥檃ll. 鈥榊ouse鈥� actually is a form that makes an important meaning distinction and so you know鈥�as I was talking about this I thought well, I like that word. I think that word does something really nice.

So that would be one kind of word that might be my favorite one but I think that on any given day you know鈥�there could be a million words so鈥�

Chris Neff: Well since you picked a semi-American word, I鈥檓 going to pick a semi-Australian word. I was walking down King Street in Newtown and there was one guy walking toward me with a bike with a flat tyre and his friend was walking next to me and he goes, 鈥榳hat happened to your bike?鈥� And the guy goes, 鈥榝lat tyre鈥�. And the other guy goes, 鈥榙evo鈥�. And the other guy goes, 鈥榙evo鈥�. And I鈥檝e got to say 鈥榙evo鈥� is one of my favorite words and鈥�

Professor Enfield: (laughs)

Chris Neff:鈥�I know it鈥檚 an Australian abbreviation but I鈥檓 just going to go there that 鈥榙evo鈥� won my heart sort of immediately.

Professor Enfield: That鈥檚 great. If words win your heart then you know鈥�that鈥檚 good for the linguist in all of us.

Chris Neff: So where do you look? When a linguist is looking to find out about a language where do they look?

Professor Enfield: It all depends on what aspect of the language you want to look at and which language it is. So, when you come to languages that are not actually written down and that are spoken by much smaller communities and that haven鈥檛 really been studied before, what they have to do is to collect the language data themselves. Now that鈥檚 not a very easy thing to do and so what鈥檚 traditionally been taught is to go into the field, get to know the people that you鈥檙e working with and ask them to tell some narratives and then you record those narratives.

So, for example a person might say tell you the story of their life and you get this monologue and then you transcribe it and then you sit down with people who speak the language and you find out ok, what does this word mean? What does this word mean? What does this sort of structure, how does it work? And that鈥檚 the sort of typical way of working in doing sort of linguistic field work.

听

And there鈥檚 a really major blind spot in that. I mean very few people have gone out and collected recordings of conversation. That is, real dialogue back and forth. You know鈥�not staged but just everyday conversation which in fact is far and away the most common sort of mode of using language.

It鈥檚 actually when you look at free flowing conversation that you see these things that animals just don鈥檛 do and secondly because that order actually does tell us something quite important about what is essential to the human capacity for language. And that is after all what linguists are looking for.

Chris Neff: So, can we go from negotiating and talking about individual communication to mass communication? Cause I鈥檓 interested in research that you鈥檙e doing and I鈥檓 wondering what the aims are of the post truth initiative?

Professor Enfield: The post truth initiative is a Sydney research excellence initiative joined by a number of colleagues from areas you know鈥�from physics, to media and communications, to philosophy, to government and international relations and the Sydney Environment Institute. There鈥檚 a whole range of people involved.

For us as academics and as scientists, as people who are trying to find out how the world is and to discover new things and to come to new knowledge, you know鈥�after we鈥檝e learnt new things about the world through our work then we want to make impact from that right. We want that to affect policy or practice or we want that to lead to further discoveries so it matters to us that the facts that we arrive at through our work, are actually treated as facts.

As well as that many of my colleagues in鈥�for example in politics and in media, they鈥檙e interested in the concepts of truth themselves and in the sort of power dynamics surrounding truth and all that kind of thing. So, what we鈥檙e really doing is studying the problem that seems to be rearing its head right now globally in public discourse, the truth and empirical findings and evidence have less weight than they used to. But we鈥檙e also through studying that, trying to understand well how could we then you know counteract it because鈥�

Chris Neff: Well and it sounds like this goes back to the point you were making about the way individual communication works in that we鈥檙e each accountable to each other for the words that we say and the way that we communicate in making sure that the other person understands. Just a quick language warning. That you鈥檙e trying to build a bullshit detector? Is that correct?

Professor Enfield: That鈥檚 right, we are. We鈥檙e trying to build a bullshit detector. It is a鈥�well essentially what we鈥檙e interested in doing is exploring ways in which new information technology and AI in the area of sort of language processing, speech recognition, that kind of thing. I mean this is a field that is moving very, very quickly.

One sort of thing that we鈥檙e focusing on right now is looking for inconsistencies over time in how certain individuals have spoken about certain topics. So, for example politicians are speaking about coal and what they鈥檝e said about the [Great Barrier] Reef itself and what they鈥檝e said about climate change and these kinds of things, then we can see well here鈥檚 a person that鈥檚 been quite consistent or here鈥檚 a person that flip flops back and forth like crazy. And that in a way is a kind of measure of well what we鈥檙e regarding as bullshit kind of technically defined.

So, you asked about the question of accountability and the kind of relationship between you know鈥�my area of work on the kind of micro level of interaction and the kinds of things that I was just talking about in relation to how politicians have spoken over time and that kind of thing.

And you鈥檙e absolutely right to point to accountability as being the key kind of factor there. We expect that when people say things to us that they鈥檙e telling the truth. That鈥檚 what makes lying a possible thing that might work because it exploits that fact. It exploits the fact that people assume that we鈥檙e telling the truth. Just telling falsehoods like that in your everyday life, you will find yourself being held accountable to them in a pretty major way and you know鈥�whole relationships can break down around things like that very quickly.

So, the issue about kind of scaling up to political discourse and to the media and all that kind of thing, that doesn鈥檛 really change in terms of you know鈥�how our brains are oriented and that鈥檚 kind of the big part of the problem here and that is that we have the same levels of trust in the individuals that we鈥檙e kind of listening to, like a politician talking or you know鈥�an author that鈥檚 written a blog post that we鈥檙e reading or whatever. We apply that same level of trust as we would sort of in everyday interaction which is where all of these capacities kind of evolve and as we all know, with a system like the one that we have with the information revolution being what it is then we can abuse that trust very, very easily and there鈥檚 very little the rest of us can do about it.

So, you know鈥�I think at the core of it for me that鈥檚 the problem. And I think that kind of lack of trust and cynicism and so forth about politicians and so on, which is surprising it hasn鈥檛 happened a lot earlier. You know鈥�it really does make sense and I think people are tapping into the fact that those people who are making decisions that affect all of us, are just completely distant from us. We have no way of knowing what they really mean and so forth.

So that in a way destroys the currency of truth and trust and so I suspect there鈥檚 a kind of collapse of that currency going on that鈥檚 a little bit like the collapse of a monetary currency in a society where people have lost their faith in that.

Chris Neff: Oh god that was a depressing answer to the question.

Professor Enfield: Sorry.

Chris Neff: Umm鈥�no that鈥檚 fine. It鈥檚 politics as usual and that brings me to my final question which is the way language boundaries stabilise ethnic boundaries. Could you give us a quick outline of that research and its applications because I鈥檝e heard that this is another area that you鈥檙e working on.

Professor Enfield: I work in Laos and it鈥檚 like many countries that are not far from the equator, it鈥檚 a place that has a lot of linguistic diversity which is to say there are many different languages spoken in a small area of land. And these languages are spoken by people of different ethnic identities and so you know鈥�usually we think of a language as sort of being attached to a certain ethnic identity and what I鈥檓 interested in looking at here is how鈥�is what happens when you have actually quite small languages being spoken together in communities where there鈥檚 a lot of social interaction between those communities.

So, in the area that we鈥檙e going to be working in in a sort of upland area of Central Laos, you鈥檝e got quite small language communities of about let鈥檚 say 300 speakers of one language and maybe 1000 speakers of another language and maybe 2000 speakers of another language. Villages that are sort of just 1 hours walk from each other. So, you have groups of people who are meeting each other every day on the paths and in the fields and so forth and they are visiting each other鈥檚 houses. Actually, the languages that they speak in Laos have over time they鈥檝e come to resemble each other very, very closely in terms of their structure. Even though they鈥檙e not historically related to each other.

And so, the question is how do you maintain your ethnic distinction while at the same time sort of letting your language gravitate more and more closely to the language of the other group. So, we鈥檙e really sort of looking at the dynamics of that. If you check back in with us in about three years鈥� time I鈥檒l tell you how the project went.

(music plays)

Chris Neff: I would love to. Well I learned a lot today and I really appreciate our conversation so thank you again Professor Nick Enfield from The 爆料王.

Professor Enfield: Thank you very much, it鈥檚 been a pleasure.

Chris Neff: Thank you for joining us on Open for Discussion. If you鈥檇 like to know more about our research be sure to visit our website sydney.edu.au/news.

Voice: Enjoyed this podcast? We think you鈥檇 love another from the 爆料王 鈥� where 爆料王 academics gave talks in small bars across Sydney. Available now on or

听

听