 Linguists love to do experiments, but how do these experiments actually work and what do they reveal? In this video I want to go over three experiments that yield fundamental insights into how language works in the human mind. The experiments also show that linguists often have a pretty weird sense of humor. My name is Martin Hilpert. I'm a professor of linguistics at the University of Neuchâtel in Switzerland. So what are these three experiments all about? Experiment one is about joy and happiness. The design of the experiment is extremely simple. The participants were shown a picture of a person and all they had to do was answer one simple question, namely what emotion is this person experiencing? Is it joy or is it happiness? Now joy and happiness are two very similar emotions, but in English we talk about them in different ways. With joy I can say something like I'm full of joy or she was overflowing with joy, brimming over with joy or I could even say that joy welled up inside me. In other words I'm talking about joy as if it were a liquid inside my body and indeed it has been argued that this is how we understand what joy actually is. Now happiness by contrast is something that you are searching for. It can be very hard to find, but you have to keep looking for it. You have to pursue it and in the end you may actually find it. Now with this in mind the experimenters showed the picture of the person to three different groups of people. Group 8 were people in libraries or grocery stores. They were looking for the right book or the right food item that they wanted to buy. So in essence they were people who were searching for something and the hypothesis was that these people would be more likely to respond to the picture by saying that person is experiencing happiness because the idea of searching for something was already active in their minds. Now by contrast group B were people in street cafes who were drinking coffee or cold drinks. So there the hypothesis was that these people would be more likely to respond with joy because the idea of a liquid inside their bodies was immediately present. There was of course a control group, a neutral control group of students who were sitting in the classroom neither drinking anything nor searching anything. Okay what came up? The results show that the searching group were actually more likely to say happiness and the drinking group were more likely to say joy in line with the predictions of the researchers. The control group were also more likely to say joy but not quite as much as the drinking group. Now what does this show? It shows that we're thinking metaphorically more specifically we're thinking in terms of metaphors that our language provides. When we talk about happiness as something we're searching for we're not just talking about it that way we actually understand it as something that we're looking for and that we hope to find someday. Let's move on to experiment two which involves a dog, a tiger and a polar bear. As a participant of this experiment you see a table with three animals in a row. You see a dog, a tiger and a polar bear. The experimenter shows you the animals and asks you to memorize the sequence. You got it? Okay they mix it up and you have to put the sequence back together. Easy right? Okay here's the twist. The experimenter asks you to switch around to another table. They give you the animals and they say put the animals in a row just like before. What do you do? If you're like me you do this. You start on the right with the dog followed by the tiger and the polar bear. The experiment has been carried out with speakers of Dutch and with speakers of Japanese and they do exactly the same thing and that seems only logical. Why would anyone line up the animals in a different way? Well speakers of the language of Zeltal, Longu and Erandik do it differently. When they are asked to put the animals in a row they have the dog on the left followed by the tiger and by the polar bear. Here we see a graph with the results. Speakers of Zeltal, Longu and Erandik overwhelmingly arrange the animals in the way you see here. Whereas speakers of Dutch and Japanese tend to arrange them like this. Now why is there this difference? The explanation has to do with the fact that languages such as Zeltal and Longu have a way of expressing spatial relations in a way that differs a lot from the way Dutch and Japanese do it. Dutch, Japanese and also French, Spanish, English and German have words for left and right that are relative to the speaker. So what's left to me is actually on your right if we're opposite one another. Languages with words such as left and right are said to use the relative frame of reference. And you will be wondering if Zeltal and Longu don't have words for left and right, then how do they actually indicate where something is in space? The answer would be that they use absolute directions such as north and south. And they would do that not only for let's say the expression of the village being to the north of the city, but rather they would also say that hey we have to move the couch a little bit further to the west. Now for speakers of languages that use left and right, it is a major mystery how someone can point to north and south in every situation they're in. The intriguing observation is that speakers of Zeltal and actually many more absolute frame of reference languages are perfectly capable of doing that even in the dark, even when they're not in their everyday surroundings. The animals in a row experiment yields one very important insight, namely that the language that you speak affects how you think. And in this particular case, the way your language lets you talk about space affects how you arrange toy animals when you're supposed to put them in a row just as before. For speakers of languages with absolute frames of reference, that means that if the animals were facing south before, you're going to put them facing south again. If your language lets you talk about left and right, then well if they were facing left before, you're going to make to face them left again. Some of you may know that this fundamental insight has a name. It's called the Sapir War of Hypothesis and it comes in different versions. The strongest version of it is that language is something like a prison for your thoughts. If you don't have words for something, then you can't express it and you can't even think about it. Now in this very strong version, the hypothesis is wrong. However, the version of the Sapir War of Hypothesis that has a lot of support for it is that your thoughts are often guided by your language. Your language makes it easier to think in one way than in another. Let me turn to experiment three, which has to do with the sounds of language. Linguists call these sounds phonemes and the experiment investigates how speakers hear these phonemes. The study is from the year 1957, so Elvis Presley was doing the jailhouse rock, the first satellite was launched into space, Dr. Seuss wrote the cat in the hat and Noam Chomsky wrote syntactic structures. So arguably, this is well the second most exciting event in linguistics in 1957. Let's take a look at how the experiment works. The experiment is a so-called ABX task. That is, the participant hears three sounds in a row. Sounds A and B are different from one another and sound X is identical either to A or to B. As a participant, you have to decide whether X is the same as A or whether it is the same as B. That sounds easy, right? In the actual experiments, we have our participant in front of a computer screen and the participant has to press either the A key or the B key depending on whether the X sound is identical to either A or B. Let's check out how this works in practice. I'm sure you're wondering what this is all about, so let me explain. Let's listen to the sounds first that were used in this experiment. There are 13 of them and they have been created with state-of-the-art synthesizer technology from 1957. If you have a set of good headphones, I suggest that you put them on now so that you can hear the differences more clearly. Let's give it a listen. What most speakers of English here is a continuum from sounds that resemble a bear to a dare and then a gaire. Three different phonemes of English, B, D and G. When you play the continuum, you start with bear up to a certain point and then it suddenly sounds like dare and the last ones sound like gaire. What I've indicated on the slide here with the red lines are the boundaries between phonemes. So step five still sounds like bear and step six already sounds like dare. The sounds in this experiment are combinations that are exactly three steps apart on the continuum. So if step A is two, then B is five and so on and so forth. So what this means is that there are some AB combinations that are within the boundaries of one phoneme and other AB combinations reach across a phoneme boundary. For example, if we have steps four and seven, A sounds like bear and seven sounds like dare so it should actually be easy to distinguish the two. By contrast, step six and step nine both sound like dare so there it should be quite hard to say if X sounds identical to A or identical to B. Moving on, step eight sounds like dare and step 11 sounds like gaire so that should again be relatively easy. What this means is that there are some combinations that are easy and some combinations that are more difficult and this is shown in this graph from the study where the authors contrast their predictions and their actual observations. So the y-axis shows how accurately the participants identified the X sound as either A or B. Points high up mean that most answers were correct and points that are lower they indicate that participants were often unable to distinguish between A and B. The observations are not as drastic as the predictions but they are actually in line with them so combinations are easier when they reach across phoneme boundaries and they're harder to distinguish when the participants have to hear variants of the same phoneme. The fundamental insight here is that speech sounds are heard as categories. When we perceive two sounds as different that is because in our language those sounds instantiate different phonemes. Two sounds can be objectively different but if they fall into the same category that means that we'll find it hard to hear a difference and among other things this actually explains why we find it so hard to learn the sounds of a second language. So speakers of French for instance differentiate between sounds that speakers of English would categorize as the same and vice versa. So I hope to have given you a little taste of linguistic experiments and how they work. If you want to find out more about language and linguistics there are more than 100 videos on this channel and if there are any linguistic topics that you would like to know more about but can't find anything on this channel let me know in the comments and I'll try if I can make something happen. That's it for now thanks for watching and see you soon. Bye.