WEBVTT - From the Vault: The Whistling, Part 2 0:00:06.160 --> 0:00:07.960 Hey, you welcome to Stuff to Blow your Mind. My 0:00:08.039 --> 0:00:09.320 name is Robert Lamb. 0:00:09.360 --> 0:00:12.000 And I'm Joe McCormick, and it's Saturday. Time to go 0:00:12.039 --> 0:00:14.160 into the vault for an older episode of the show. 0:00:14.440 --> 0:00:18.160 This one originally aired August fourth, twenty twenty two, and 0:00:18.239 --> 0:00:20.840 it is part two of our series on whistling. 0:00:21.400 --> 0:00:27.240 Enjoy Welcome to Stuff to Blow Your Mind, a production 0:00:27.280 --> 0:00:33.040 of iHeartRadio. 0:00:34.920 --> 0:00:36.960 Hey you welcome to Stuff to Blow your Mind. My 0:00:37.040 --> 0:00:38.360 name is Robert Lamb. 0:00:38.280 --> 0:00:40.920 And I'm Joe McCormick, and we're back with part two 0:00:41.479 --> 0:00:46.240 of our series on whistling. In the last episode, we 0:00:46.360 --> 0:00:50.000 talked a bit about how whistling works physically, what happens 0:00:50.000 --> 0:00:53.360 when you're creating a sort of resonator cavity within the mouth. 0:00:54.280 --> 0:00:59.240 We also talked about the whistle speech of the Mastaco 0:00:59.360 --> 0:01:03.240 languages and in Mexico, and I wanted to start off 0:01:03.240 --> 0:01:07.800 today's episode by talking about some other examples of whistled 0:01:07.920 --> 0:01:11.160 languages and some of the common characteristics between them, because, 0:01:11.200 --> 0:01:14.559 of course, the Mezteko whistle speech is not the only 0:01:14.720 --> 0:01:20.039 example of a whistled language that carries information. In fact, 0:01:20.080 --> 0:01:22.400 I was looking at a paper by an author named 0:01:22.520 --> 0:01:27.560 Julian Meyer called Environmental and Linguistic Typology of Whistled Languages 0:01:28.120 --> 0:01:31.039 in the Annual Review of Linguistics twenty twenty one. So 0:01:31.080 --> 0:01:35.160 it's a very recent paper. And according to Meyer, there 0:01:35.200 --> 0:01:39.200 are reports of more than eighty languages around the world 0:01:39.600 --> 0:01:43.800 that contain a whistled lexicon, and about half of those 0:01:43.880 --> 0:01:47.920 have been confirmed by formal studies and published recordings. So 0:01:48.080 --> 0:01:51.680 really solid documentation of at least forty or so whistled 0:01:51.760 --> 0:01:54.960 languages around the world. And so I think it's worth 0:01:55.040 --> 0:01:58.040 mentioning a few more examples of these in describing how 0:01:58.080 --> 0:02:00.800 they work and seeing what we can compare in contrast 0:02:00.840 --> 0:02:04.840 with them. So one example I was reading about was 0:02:04.880 --> 0:02:09.079 in a really interesting twenty seventeen article in BBC Travel 0:02:09.280 --> 0:02:13.640 by Elliott Stein, and the story here goes like this. 0:02:14.320 --> 0:02:19.040 In Greece, there is a remote mountain village called Antia, 0:02:19.720 --> 0:02:23.799 which is found on the southern eastern coast of the 0:02:23.919 --> 0:02:28.280 Greek island of Evia in the Aegean Sea. And within 0:02:28.400 --> 0:02:32.640 this village there has long been a whistle based language 0:02:32.760 --> 0:02:38.279 called Spheria, which allows speakers to communicate across great distances, 0:02:38.480 --> 0:02:41.400 and it seems to have been passed down from parents 0:02:41.400 --> 0:02:43.920 to children among the shepherds and the farmers of the 0:02:44.000 --> 0:02:47.400 village for literally thousands of years, for more than two 0:02:47.480 --> 0:02:50.960 thousand years to read from Stein here quote. But in 0:02:51.000 --> 0:02:54.679 the last few decades, Antia's population has dwindled from two 0:02:54.760 --> 0:02:58.600 hundred and fifty to thirty seven, and as older whistlers 0:02:58.680 --> 0:03:03.280 lose their teeth, many can no longer sound Spheria's sharp notes. 0:03:03.840 --> 0:03:07.000 Today there are only six people left on the planet 0:03:07.080 --> 0:03:10.760 who can still speak this unspoken language. Now this was 0:03:11.000 --> 0:03:13.640 five years ago as of this recording, so I don't 0:03:13.639 --> 0:03:16.120 know how that number six has changed since then. There 0:03:16.560 --> 0:03:19.560 are descriptions of some efforts to try to teach it 0:03:19.600 --> 0:03:23.160 to more people, but of course, whenever you're talking about 0:03:23.160 --> 0:03:28.160 a language with that few speakers, it's certainly extremely endangered. 0:03:28.200 --> 0:03:31.000 In fact, this is considered one of the most endangered 0:03:31.080 --> 0:03:35.120 languages in the world. Now. Apparently the existence of this 0:03:35.360 --> 0:03:39.920 language Spheria here was not documented anywhere in the outside 0:03:40.000 --> 0:03:43.840 world until the year nineteen sixty nine, when a plane 0:03:43.960 --> 0:03:47.240 crashed in the mountains nearby, and there was a rescue 0:03:47.280 --> 0:03:50.280 team that was attempting to locate the pilot and they 0:03:50.320 --> 0:03:55.560 reported hearing strange, melodious whistling echoing through the hillsides, and 0:03:55.760 --> 0:03:58.960 this led to investigation and brought the language to the 0:03:59.000 --> 0:04:03.280 attention of the media and two academics. So a big 0:04:03.360 --> 0:04:06.560 question here is where does a language like this come from. 0:04:06.760 --> 0:04:10.360 Linguists do seem to agree that it dates back to 0:04:10.440 --> 0:04:12.640 ancient times. It's been around for a long time, but 0:04:12.720 --> 0:04:17.279 exactly how it was created is less certain, and apparently 0:04:17.400 --> 0:04:21.919 local legends abound. So one story I came across this 0:04:22.080 --> 0:04:26.080 was described in some detail in a documentary piece on 0:04:26.160 --> 0:04:30.480 PBS News Hour that was about Spherea, and it claimed 0:04:30.560 --> 0:04:35.279 that the language was invented about two five hundred years ago, 0:04:35.480 --> 0:04:39.919 not by Greeks, but by Persians after they were defeated 0:04:40.120 --> 0:04:44.000 at the Battle of Salamis. So Salamis was a battle 0:04:44.200 --> 0:04:48.640 in four ADBCE during the Persian invasion of Greece under 0:04:48.720 --> 0:04:52.160 zerk Sees the Great, and so Salamis was It was 0:04:52.200 --> 0:04:55.239 a naval battle where the coalition of Greek city states 0:04:55.320 --> 0:04:58.800 was able to fight off and defeat the larger Persian 0:04:58.839 --> 0:05:02.640 allied fleet. And I think this is widely considered the battle, 0:05:02.760 --> 0:05:05.080 or one of the battles that turned the tide of 0:05:05.120 --> 0:05:08.120 the war in favor of the Greek defenders and pushed 0:05:08.160 --> 0:05:11.880 back the Persian invasion. But anyway, the legend about the 0:05:11.880 --> 0:05:16.039 whistle speech goes that the Persian survivors of the battle 0:05:16.200 --> 0:05:18.960 I guess they were, you know, their ship was sank 0:05:19.240 --> 0:05:21.520 or defeated in some way, and they managed to swim 0:05:21.560 --> 0:05:24.279 to shore on the island of Evia, where they had 0:05:24.279 --> 0:05:29.440 to survive hiding in the mountains inhabited by hostile native Greeks. 0:05:29.960 --> 0:05:33.080 And one way they avoided detection was by coming up 0:05:33.160 --> 0:05:36.560 with a way of speaking in whistles that would sound 0:05:36.720 --> 0:05:40.000 just like the birds, so they could communicate with each other, 0:05:40.160 --> 0:05:43.880 but their speech would not be intelligible and in many 0:05:43.880 --> 0:05:46.919 ways would probably not even be detected. Oh wow, but 0:05:47.080 --> 0:05:49.560 that story has a kind of legendary quality. I'm not 0:05:49.560 --> 0:05:52.280 sure how much there is behind that, but it's a 0:05:52.279 --> 0:05:56.600 great story nonetheless, and steinsit some other local legends as well. 0:05:56.640 --> 0:06:00.719 Some residents believe it was invented during the Busante Empire 0:06:00.760 --> 0:06:03.400 by locals who wanted a secret way to communicate that 0:06:03.440 --> 0:06:06.800 would elude the understanding of pirates and people from hostile 0:06:06.880 --> 0:06:10.200 nearby villages. And so a common theme here seems to 0:06:10.240 --> 0:06:12.880 be the idea that somehow this language was created to 0:06:12.880 --> 0:06:16.080 be a secret way of communicating, to allow the locals 0:06:16.480 --> 0:06:20.360 to communicate across distance and understand each other without other 0:06:20.400 --> 0:06:24.120 people detecting or understanding what they were saying. Now, the 0:06:24.160 --> 0:06:27.800 author of this BBC travel piece describes visiting the village 0:06:27.800 --> 0:06:30.839 and spending time with the handful of people there who 0:06:30.880 --> 0:06:35.000 still use the whistle language, and they apparently use it 0:06:35.040 --> 0:06:38.440 in many of the same scenarios described in that paper 0:06:38.520 --> 0:06:41.000 on the Mastako whistle speech that I've talked about in 0:06:41.000 --> 0:06:44.880 the last episode. A big scenario of use seems to 0:06:44.880 --> 0:06:50.000 be communicating across great distance on the mountain side and 0:06:50.040 --> 0:06:54.640 sort of greeting or summoning people from far away. And 0:06:54.920 --> 0:06:59.719 Stein cites a Greek linguist named Dimitra Hengen who studies 0:06:59.720 --> 0:07:03.320 Sphere and she says that Spheria is in some sense 0:07:03.400 --> 0:07:09.120 a whistled version of spoken Greek, where specific whistled tones 0:07:09.720 --> 0:07:14.160 correspond to specific phonetic syllables or letters, and you can 0:07:14.280 --> 0:07:18.080 build words out of them. Now, again, this is another 0:07:18.120 --> 0:07:20.760 way that it's similar to the Mezateeko example, because in 0:07:20.800 --> 0:07:25.880 both cases the whistled language is not like an totally independent, 0:07:26.120 --> 0:07:29.640 unique language. Instead, it is in some way adapting and 0:07:29.760 --> 0:07:32.920 existing spoken language to whistles. 0:07:32.720 --> 0:07:36.200 And of course that would lean us more towards the 0:07:36.560 --> 0:07:38.960 Greek origin story as opposed to the Persian one. 0:07:39.280 --> 0:07:41.480 Yeah, I thought about that, but I don't know if 0:07:41.520 --> 0:07:44.760 it actually informs that one way or another, But yeah, 0:07:44.920 --> 0:07:48.000 I had the same intuition at least. So one of 0:07:48.000 --> 0:07:51.400 the most remarkable things about Spheria, again similar to the 0:07:51.440 --> 0:07:56.040 Mezteko whistle speech example, is that it is intelligible at 0:07:56.040 --> 0:07:59.960 a great distance. You can understand messages in Spheria update 0:08:00.080 --> 0:08:05.320 to about four kilometers away on this mountainous terrain, which 0:08:05.600 --> 0:08:09.520 Hingen says is about ten times farther than you can 0:08:09.600 --> 0:08:13.480 usually understand speech, loud speech or shouting. And I saw 0:08:13.480 --> 0:08:16.640 that number of the ten times distance multiplier mentioned in 0:08:16.680 --> 0:08:20.280 other sources, such as a Cambridge University press paper that 0:08:20.360 --> 0:08:22.960 I looked at. But there's a great part in this 0:08:23.120 --> 0:08:27.160 article where Stein quotes a local shop owner named Maria 0:08:27.560 --> 0:08:32.160 Cathalis who tells a story about some of the social 0:08:32.200 --> 0:08:35.800 opportunities offered by the whistle speech. And so her story 0:08:35.840 --> 0:08:38.920 goes like this quote. One night, a man was in 0:08:39.000 --> 0:08:42.520 the mountains with his sheep when it started snowing, he 0:08:42.559 --> 0:08:45.040 knew that somewhere deep in the mountains there was a 0:08:45.080 --> 0:08:48.720 beautiful girl from Antia with her goats. So he found 0:08:48.760 --> 0:08:51.440 a cave, built a fire, and whistled to her to 0:08:51.480 --> 0:08:55.120 come keep warm. She did, and that's how my parents 0:08:55.160 --> 0:08:55.720 fell in love. 0:08:56.480 --> 0:08:59.880 Well, that is a better ending than something like this. 0:09:00.080 --> 0:09:04.400 And the descendants of Persian soldiers slaughtered him in the woods, 0:09:05.880 --> 0:09:07.160 so I will say that much. 0:09:07.520 --> 0:09:10.400 Yeah, and then the Greeks finally came for revenge. 0:09:11.200 --> 0:09:13.000 Yeah. I continue to just tell a lot of questions 0:09:13.040 --> 0:09:16.960 about the Persian origin theory. It just seems like it 0:09:17.040 --> 0:09:20.440 seems I could be missing something major here, but it 0:09:20.440 --> 0:09:22.760 seems like it begs more questions than it answers. 0:09:23.080 --> 0:09:26.440 Well. Yeah, as I said, it sounds more like legend 0:09:26.520 --> 0:09:30.119 to me than like an strongly evidence based explanation. 0:09:30.240 --> 0:09:33.040 Yeah, when I agree, away from being a full blown 0:09:33.040 --> 0:09:33.720 ghost story. 0:09:34.960 --> 0:09:38.800 But so I've described two examples here of whistled language 0:09:38.840 --> 0:09:41.880 in detail, and as I mentioned before, there are many 0:09:41.920 --> 0:09:45.319 others around the globe. There are something like eighty ish 0:09:45.559 --> 0:09:48.719 that have been reported somewhere around forty of them are 0:09:48.840 --> 0:09:52.680 very well documented. So an obvious question to ask is 0:09:52.800 --> 0:09:56.040 what do these languages have in common? What causes a 0:09:56.120 --> 0:09:59.840 whistled language to arise? So I was looking at a 0:09:59.840 --> 0:10:02.560 few sources here. One of them is that paper in 0:10:03.000 --> 0:10:06.440 the Annual Review of Linguistics by Julian Meyer already mentioned. 0:10:06.800 --> 0:10:11.959 Another is an article in Smithsonian Magazine from twenty twenty 0:10:11.960 --> 0:10:15.599 one by Bob Holmes which cites that paper and summarizes 0:10:15.679 --> 0:10:19.040 some other research in this area, for example, focusing on 0:10:19.760 --> 0:10:23.000 a whistled variant of Spanish that is used in the 0:10:23.000 --> 0:10:28.440 Canary Islands, on the mountainous islands of Lagomera and eliero 0:10:28.720 --> 0:10:32.240 In that are both in the Canary Islands. And that 0:10:32.360 --> 0:10:36.600 paper by Julian Meyer tries to gather together all of 0:10:36.640 --> 0:10:41.120 these languages and say, okay, are there common topographical or 0:10:41.160 --> 0:10:45.720 sort of geographical features that these languages tend to have 0:10:45.760 --> 0:10:49.800 in common? And he finds yes, indeed there are. Almost 0:10:49.960 --> 0:10:53.480 all of the whistled languages occur in two different types 0:10:53.520 --> 0:10:58.760 of environments, either in mountainous terrains or rugged mountains, or 0:10:59.320 --> 0:11:03.880 in dense vegetation like dense forest or dense savannah. So 0:11:03.880 --> 0:11:06.440 why would it be those two places mountains or in 0:11:06.520 --> 0:11:10.319 dense vegetation. Well, to focus on mountains first, Maya writes, 0:11:10.360 --> 0:11:14.240 that in mountainous terrain, settlements and the people living in 0:11:14.400 --> 0:11:18.040 the mountainous terrain tend to be much more scattered across 0:11:18.280 --> 0:11:23.520 larger distances that are more difficult to traverse quickly than 0:11:23.640 --> 0:11:29.800 people in other types of topographical settings. So Maya writes 0:11:29.880 --> 0:11:33.640 quote in Eliero and Lakomera in the Canary Islands, in 0:11:33.679 --> 0:11:37.720 the region around Kushkoi in Turkey, in the High Altus, 0:11:37.840 --> 0:11:41.160 or in the Pyrenees near the village of Os two points, 0:11:41.200 --> 0:11:44.960 only five hundred meters apart can easily represent an hour 0:11:45.160 --> 0:11:49.640 in walking time. Thus, whistled forms of languages serve as 0:11:49.679 --> 0:11:53.640 soon as the spoken forms become ineffective. Between forty and 0:11:53.720 --> 0:11:57.520 one hundred meters. Depending on terrain, whistles can be heard 0:11:57.640 --> 0:12:01.680 up to seven kilometers away in some vas okay. So 0:12:01.840 --> 0:12:05.000 the idea here is that in mountainous terrain you have 0:12:05.480 --> 0:12:09.040 the problem of people are often situated farther apart from 0:12:09.080 --> 0:12:13.120 each other, and those distances to cross are difficult to cross. 0:12:13.160 --> 0:12:15.520 They take a long time. So if you need to 0:12:15.559 --> 0:12:18.760 communicate actually coming to be close enough together that you 0:12:18.800 --> 0:12:23.360 could understand each other by shouting, that's a long that's 0:12:23.360 --> 0:12:26.360 a big time investment. So it's actually worth your time 0:12:26.559 --> 0:12:29.840 to learn a whistle speech that will carry better across 0:12:29.920 --> 0:12:33.160 longer distances and save you all of that climbing and 0:12:33.240 --> 0:12:37.559 walking time. Now, what about the forests or the dense vegetation. Well, here, 0:12:38.200 --> 0:12:41.920 Meyer writes quote the vegetation in dense tropical forests and 0:12:41.920 --> 0:12:47.200 savannahs restricts visual contact and limits the propagation of sound. 0:12:47.960 --> 0:12:52.280 In such contexts, Whistled speech frequencies are also well shielded 0:12:52.360 --> 0:12:57.239 against acoustic energy loss due to reverberation, which is particularly 0:12:57.280 --> 0:13:02.120 important in densely vegetated environments because the whistled frequencies belong 0:13:02.200 --> 0:13:06.360 to the most favorable frequency window, ranging from one to 0:13:06.480 --> 0:13:12.000 three killihertz, within which reverberation in forests varies less with distance. 0:13:12.679 --> 0:13:17.280 In dense vegetation, whistled language facilitates the coordination of individuals 0:13:17.360 --> 0:13:22.160 during group movements, especially during hunting and fishing. Whistling also 0:13:22.240 --> 0:13:26.280 allows human dialogue to go undetected by animals, blending in 0:13:26.320 --> 0:13:30.199 with natural sounds, since many animal species also use whistling. 0:13:30.960 --> 0:13:34.040 Other advantages are that whistles are easy to locate and 0:13:34.120 --> 0:13:38.760 difficult for strangers to recognize, especially other tribes, even those 0:13:38.840 --> 0:13:43.080 that speak different dialects of the same language. Whistled communications 0:13:43.120 --> 0:13:46.440 are used for distances from about ten meters up to 0:13:46.480 --> 0:13:51.400 five hundred meters, depending on the density of vegetation. Okay, 0:13:51.440 --> 0:13:54.000 so there are a lot of advantages in the forest 0:13:54.080 --> 0:13:57.640 or thick savanna. So the idea is that, of course 0:13:57.640 --> 0:14:01.280 whistling speech allows you to communicate kate without being able 0:14:01.320 --> 0:14:05.960 to see each other. Sitelines are limited by the vegetation itself, 0:14:06.400 --> 0:14:10.640 but also whistling just carries better in the forest. It 0:14:10.800 --> 0:14:14.040 propagates better through the forest without being drowned out by 0:14:14.080 --> 0:14:17.560 the sort of the reverberation effects of having all that 0:14:17.600 --> 0:14:22.200 foliage there. And it also seems to pierce through ambient 0:14:22.320 --> 0:14:26.840 sound much better. And Holmes also summarizes some of these 0:14:26.880 --> 0:14:32.200 advantages of whistling in the Smithsonian paper, saying that if 0:14:32.200 --> 0:14:34.960 you're good at whistling, and you've been practicing this all 0:14:34.960 --> 0:14:37.400 your life, sometimes you can reach one hundred and twenty 0:14:37.440 --> 0:14:41.360 decibels with a whistle, which is loud. That's like he 0:14:41.480 --> 0:14:43.920 compares it to a car horn. It says it's actually 0:14:43.960 --> 0:14:48.040 louder than a car horn, and that whistles pack almost 0:14:48.080 --> 0:14:52.640 all of that energy into the perfect frequency range. The 0:14:52.680 --> 0:14:56.400 most piercing frequency range, which Holmes says is between one 0:14:56.440 --> 0:14:59.240 to four killer hertz. Meyer said between one to three 0:14:59.320 --> 0:15:03.640 killer hertz, roughly the same space which Holmes says is 0:15:03.720 --> 0:15:07.760 above the pitch of most ambient noise. And this is 0:15:07.800 --> 0:15:11.960 interesting because I was thinking about, why do we keep 0:15:12.320 --> 0:15:16.920 noticing that whistling sounds, the ones made by humans are 0:15:16.960 --> 0:15:20.640 similar to bird song. Well, one thing that occurs to 0:15:20.640 --> 0:15:24.480 me here is that bird song is probably shaped by 0:15:24.560 --> 0:15:28.560 natural selection to propagate through vegetation and to cut through 0:15:28.640 --> 0:15:32.280 ambient noise from the environment so as to be clear, 0:15:32.320 --> 0:15:34.240 you know, to be clear and audible at a distance 0:15:34.280 --> 0:15:37.760 where maybe a potential mate could hear it. So whistle 0:15:37.760 --> 0:15:42.120 speech probably sounds like bird song, having similar frequency ranges 0:15:42.160 --> 0:15:45.840 because similar forces are shaping them. In the case of birds, 0:15:45.840 --> 0:15:48.560 it would be evolution, and in the case of humans, 0:15:48.840 --> 0:15:52.400 it would be people intentionally selecting whatever noise they are 0:15:52.440 --> 0:15:56.000 able to make with their bodies that is the clearest 0:15:56.160 --> 0:16:00.400 at the longest distance, cutting through ambient noise, and losing 0:16:00.440 --> 0:16:04.000 the least energy to reverberation in the forest. And that 0:16:04.120 --> 0:16:06.520 just happens to be the whistle that sounds like a bird. 0:16:07.120 --> 0:16:10.200 Yeah, yeah, this is interesting to think think about on 0:16:10.200 --> 0:16:14.400 one hand, I'm a quiet whistler. My whistler, my whistle 0:16:14.440 --> 0:16:17.120 is not very loud, and therefore it can be a 0:16:17.120 --> 0:16:19.880 little surprising when I encounter someone who has a very 0:16:19.920 --> 0:16:23.080 loud whistle, and you're reminded just how loud a whistle 0:16:23.160 --> 0:16:26.360 can be. So that's important to factor into all of this. 0:16:26.600 --> 0:16:29.920 And another connection that came up in some of the 0:16:29.920 --> 0:16:34.120 research I was doing was that you end up encountering 0:16:34.120 --> 0:16:39.680 this whole realm of of non linguistic sounds that humans 0:16:39.800 --> 0:16:43.360 can make that can be used to communicate ideas or 0:16:43.400 --> 0:16:47.240 to gain attention, et cetera. And you also see things 0:16:47.240 --> 0:16:53.080 like yodling thrown in there. Yodling also an art form, 0:16:53.120 --> 0:16:55.960 if you will, or a performance or to a sound 0:16:56.240 --> 0:16:59.560 that developed that also had to do with communicating or 0:16:59.600 --> 0:17:03.840 calling animals, or communicating with other herdsmen across across long 0:17:03.880 --> 0:17:05.000 distances in the wild. 0:17:05.359 --> 0:17:08.320 When you're trying to speak normal phonemes like we're using 0:17:08.320 --> 0:17:11.600 in words here, I think a lot of that information 0:17:11.680 --> 0:17:14.280 probably easily gets lost at a distance, Like you might 0:17:14.320 --> 0:17:17.080 be able to hear that somebody is shouting, but you 0:17:17.080 --> 0:17:19.879 can't hear the difference between consonants or they making a 0:17:19.960 --> 0:17:22.600 T sound or a K sound like I don't know. 0:17:22.640 --> 0:17:26.480 At a distance, that kind of all disappears. But if 0:17:26.560 --> 0:17:31.560 you're judging more on sequences of pitches, yeah, then suddenly 0:17:31.760 --> 0:17:34.440 the confusion created by distance is reduced. 0:17:34.920 --> 0:17:38.720 Yeah. Yeah, just just yelling doesn't necessarily cut it, right, 0:17:38.800 --> 0:17:41.760 because if you can't, if your particular words are not 0:17:41.800 --> 0:17:43.920 going to be overheard, then you might end up having 0:17:43.960 --> 0:17:46.600 to do something like just some sort of rhythmic barking. 0:17:46.640 --> 0:17:48.960 And if you're doing some sort of rhythmic barking, well 0:17:49.000 --> 0:17:52.920 why not further develop that and get somewhere, get to 0:17:52.960 --> 0:17:56.120 somewhere where it is yodling, or you shift over into 0:17:56.119 --> 0:17:59.040 a whistling, and that develops into some sort of a 0:17:59.080 --> 0:18:02.239 whistling language. So yeah, it's just the more you look 0:18:02.280 --> 0:18:03.920 at it, the more sense it makes for this kind 0:18:03.960 --> 0:18:04.920 of purpose. 0:18:11.400 --> 0:18:14.040 Now, another thing going on with whistled languages is that 0:18:14.920 --> 0:18:17.360 most of them, perhaps all of them, but I'm not 0:18:17.480 --> 0:18:19.280 sure about that. So I'm going to say at least 0:18:19.320 --> 0:18:23.239 the vast majority of whistled languages appear to be not 0:18:23.680 --> 0:18:28.639 wholly independent languages of their own, but whistled versions of 0:18:28.880 --> 0:18:32.800 spoken languages. So this was true of all the examples 0:18:32.800 --> 0:18:36.000 I've talked about before. You know, the Mazteko whistle speech 0:18:36.200 --> 0:18:40.520 was a whistled variant of the tonal Mesteko language. Spherea 0:18:40.560 --> 0:18:43.639 appears to be a whistled system for encoding spoken Greek. 0:18:44.200 --> 0:18:47.160 The whistle speech system of the Canary Islands, called Silbo, 0:18:47.400 --> 0:18:51.159 is a whistled version of Spanish, and so forth, and 0:18:51.440 --> 0:18:54.920 for this reason, one of the main differences in whistled 0:18:54.960 --> 0:18:59.160 languages appears to be whether they are encoding a tonal 0:18:59.320 --> 0:19:03.560 language or a non tonal language, and based on that distinction, 0:19:03.800 --> 0:19:07.760 the encoding process is different. Tonal languages tend to be 0:19:07.880 --> 0:19:10.560 whistled in a way that preserves the tones of the 0:19:10.600 --> 0:19:13.520 spoken words, and in the last episode we talked about 0:19:14.119 --> 0:19:17.679 tonal languages. Tonal languages where you know the syllables of 0:19:17.720 --> 0:19:21.479 the words also carry information based on the tone you 0:19:21.640 --> 0:19:25.439 use when speaking them, So say, like a high pitched 0:19:25.600 --> 0:19:29.119 version of the syllable ma means something different than a 0:19:29.160 --> 0:19:32.560 lower pitched version of the syllable MA, or an upgliding 0:19:32.640 --> 0:19:35.159 tone on that syllable, and so forth, like the tone 0:19:35.160 --> 0:19:38.600 of the syllable actually makes a difference. Non tonal languages 0:19:38.600 --> 0:19:42.120 are not like so in English, we don't encode much 0:19:42.160 --> 0:19:45.240 information into the tones of syllables. It's just like, what 0:19:45.400 --> 0:19:46.840 are the vowels and consonants. 0:19:47.359 --> 0:19:52.880 Yeah, the tone can contain some information, but not nearly 0:19:52.960 --> 0:19:56.480 to the extent that you find in true tonal languages. Right. 0:19:56.560 --> 0:20:00.679 No, it's not lexical information, more like maybe sort of 0:20:01.400 --> 0:20:04.520 contextual mood information or inflection. 0:20:04.880 --> 0:20:07.919 Right, Like the difference between saying I would like you 0:20:08.000 --> 0:20:10.280 to walk the dog and I would like you to 0:20:10.400 --> 0:20:13.800 walk the dog. Well, that implies that maybe something the 0:20:13.840 --> 0:20:17.080 last time the dog was walked it was not it 0:20:17.119 --> 0:20:19.960 was not good enough it was or maybe you ran 0:20:20.080 --> 0:20:22.119 the dog, you know, like that sort of thing. But 0:20:22.440 --> 0:20:27.080 it doesn't change the actual yeah information contained in the 0:20:27.119 --> 0:20:27.960 word no. 0:20:28.000 --> 0:20:31.440 It's more like about the implied information about the attitude 0:20:31.480 --> 0:20:32.320 of the speaker or. 0:20:32.280 --> 0:20:34.480 Something, yeah, or I need you to walk the dog. 0:20:34.880 --> 0:20:37.440 It applies maybe you didn't walk the actual dog, maybe 0:20:37.480 --> 0:20:40.280 you took some other creature or item from the house 0:20:40.359 --> 0:20:41.680 with you on the walk instead. 0:20:42.080 --> 0:20:44.640 Okay, so you got a tonal language and you want 0:20:44.680 --> 0:20:48.159 to make a whistled version of that. In most cases, 0:20:48.160 --> 0:20:51.640 it seems like you preserve the tones of the spoken words. 0:20:52.240 --> 0:20:56.399 Non tonal languages that have whistled speech tend to involve 0:20:56.440 --> 0:21:00.960 a sort of approximation of consonants and vowels, and the 0:21:01.280 --> 0:21:04.960 Holmes article I mentioned quotes this scholar, Julian Meyer, explaining 0:21:05.000 --> 0:21:10.560 that we already use subtle differences in frequencies to distinguish 0:21:10.720 --> 0:21:17.040 between spoken phonemes, like the differences between certain vowels and consonants. 0:21:17.240 --> 0:21:20.720 So think about the vowels E and O. A long 0:21:20.920 --> 0:21:24.240 E vowel sound has a higher pitch than a long 0:21:24.359 --> 0:21:26.600 oh vowel sound, and if you say them back to back, 0:21:26.640 --> 0:21:30.040 you can listen to the descending melody of those vowel 0:21:30.040 --> 0:21:34.359 sounds EO, EO, And in fact, though it's harder to 0:21:34.359 --> 0:21:37.360 hear it first, the same is sort of true of consonants, 0:21:37.400 --> 0:21:41.399 Like a T sound contains more high frequencies than a 0:21:41.520 --> 0:21:45.320 K sound, and these differences can to some extent be 0:21:45.520 --> 0:21:50.560 reproduced in whistles. And so the discussion of this in 0:21:50.600 --> 0:21:54.960 the article got me thinking about, even without having an 0:21:55.040 --> 0:21:58.320 established version of a language like this, and without any training, 0:21:59.040 --> 0:22:02.800 can you sort of attempt to whistle English phrases and 0:22:02.880 --> 0:22:06.880 have people understand what you're saying. In some cases you can, 0:22:06.960 --> 0:22:09.159 and I actually tried this out with my wife Rachel 0:22:09.200 --> 0:22:12.800 before we recorded here. This was a kind of weird exercise, 0:22:12.840 --> 0:22:15.000 but I was like, hey, can you tell what I'm 0:22:15.040 --> 0:22:21.800 saying here? And so I tried things like which she 0:22:22.119 --> 0:22:24.520 took a minute on but decided I was saying hello, 0:22:24.640 --> 0:22:26.840 nice to meet you, which is what I was trying 0:22:26.880 --> 0:22:31.480 to say, so that one worked a few other phrases 0:22:31.520 --> 0:22:34.720 I tried did not work as well, but the ones 0:22:34.760 --> 0:22:39.120 that really seemed to work immediately were the ones where 0:22:39.119 --> 0:22:43.200 it was phrases she had heard me say before, especially 0:22:43.200 --> 0:22:46.760 when I tried to whistle, common phrases that we use 0:22:46.880 --> 0:22:53.880 with our dog, so immediately she heard as all buddy. 0:22:55.320 --> 0:22:57.760 And I think this ties into something we've talked about 0:22:57.760 --> 0:23:02.040 on the show before, the exactg CD musicality that humans 0:23:02.119 --> 0:23:05.240 tend to use when speaking to babies and pets. For 0:23:05.280 --> 0:23:09.280 some reason, there may be evolutionary reasons for this, that 0:23:10.080 --> 0:23:13.200 when we speak to cute things that need our care 0:23:13.240 --> 0:23:16.960 and attention, you know, babies or pets, says kind of 0:23:17.400 --> 0:23:19.400 it might be creepy to think about them this way, 0:23:19.480 --> 0:23:24.080 but to some extent, kind of psychologically surrogate babies that 0:23:24.160 --> 0:23:29.159 we speak with an exaggerated musicality, or tonal variation that 0:23:29.200 --> 0:23:32.400 we don't use when speaking to adults, and that stereotyped 0:23:32.400 --> 0:23:37.080 phrases within this kind of highly musical speech are much 0:23:37.119 --> 0:23:40.000 easier to recognize when you try to whistle them instead 0:23:40.040 --> 0:23:41.159 of say them phonetically. 0:23:41.680 --> 0:23:45.000 So are you going to keep whistling? Was it a 0:23:45.040 --> 0:23:45.880 big enough success? 0:23:46.200 --> 0:23:50.960 Oh? No, I think that would be a horrible idea. Also, strangely, 0:23:51.240 --> 0:23:53.920 the dog did not seem to get it. So when 0:23:53.920 --> 0:23:57.000 I whistled all buddy Rachel could tell what I was saying, 0:23:57.119 --> 0:23:59.560 but Charlie did not seem affected. 0:24:00.160 --> 0:24:05.000 I tried whistling to my cat whilst researching information for 0:24:05.080 --> 0:24:08.600 these episodes, and yes, you didn't care. And my wife 0:24:08.640 --> 0:24:11.639 was like, like, you can't speak to a cat in whistles. 0:24:11.720 --> 0:24:14.240 You have to use the kissie sound. That's what they understand. 0:24:14.280 --> 0:24:16.520 That's what that's the language they speak. 0:24:16.760 --> 0:24:17.400 It is known. 0:24:17.920 --> 0:24:21.040 But the kissie sound clicks, I mean, these are all 0:24:21.160 --> 0:24:24.240 these are not too far removed from whistling. Some of 0:24:23.920 --> 0:24:26.320 these type of sounds will come up again later on. 0:24:26.840 --> 0:24:29.680 Indeed. Now, one of the most interesting lines of thought 0:24:29.720 --> 0:24:33.920 emerging from all this is that some experts think that 0:24:34.240 --> 0:24:38.639 studying whistled languages might help us understand the origin of 0:24:38.880 --> 0:24:43.240 human language as a whole, because again, some linguists think 0:24:43.720 --> 0:24:47.760 that these whistled languages could be similar to the first 0:24:47.880 --> 0:24:51.320 languages that probably emerged in human evolution. Now why on 0:24:51.359 --> 0:24:54.400