WEBVTT 00:00:00.000 --> 00:00:02.470 align:middle line:90% [SQUEAKING] 00:00:02.470 --> 00:00:04.446 align:middle line:90% [RUSTLING] 00:00:04.446 --> 00:00:05.928 align:middle line:90% [CLICKING] 00:00:05.928 --> 00:00:11.698 align:middle line:90% 00:00:11.698 --> 00:00:13.990 align:middle line:84% NORVIN RICHARDS: Here, I'm going to try to answer that. 00:00:13.990 --> 00:00:15.240 align:middle line:90% That's a really good question. 00:00:15.240 --> 00:00:18.400 align:middle line:90% 00:00:18.400 --> 00:00:21.650 align:middle line:84% And in some cases, we just kind have to guess. 00:00:21.650 --> 00:00:25.633 align:middle line:90% But in some cases, we have-- 00:00:25.633 --> 00:00:27.550 align:middle line:84% especially for languages like Greek and Latin, 00:00:27.550 --> 00:00:30.520 align:middle line:84% we have descriptions by people who attempted 00:00:30.520 --> 00:00:32.200 align:middle line:90% to describe their own sounds. 00:00:32.200 --> 00:00:35.210 align:middle line:84% That's a kind of thing that happens. 00:00:35.210 --> 00:00:38.140 align:middle line:84% We can infer things from mistakes 00:00:38.140 --> 00:00:41.180 align:middle line:84% that people made at various points. 00:00:41.180 --> 00:00:46.360 align:middle line:84% So people who were learning how to spell, or say children, 00:00:46.360 --> 00:00:49.510 align:middle line:84% or people who weren't highly educated, 00:00:49.510 --> 00:00:54.520 align:middle line:84% when they misspelled words in, say, Greek or Latin, 00:00:54.520 --> 00:00:55.700 align:middle line:90% you can infer things. 00:00:55.700 --> 00:00:56.830 align:middle line:90% So if you're making-- 00:00:56.830 --> 00:00:59.000 align:middle line:84% if a certain kind of misspelling is common, 00:00:59.000 --> 00:01:02.030 align:middle line:84% you get to infer that those sounds were kind of common. 00:01:02.030 --> 00:01:05.462 align:middle line:84% So in the case of ancient Greek, we think that there was an "Y" 00:01:05.462 --> 00:01:07.420 align:middle line:84% [high front rounded vowel] partly because there 00:01:07.420 --> 00:01:10.690 align:middle line:84% was a period where people were-- when people misspelled it. 00:01:10.690 --> 00:01:14.080 align:middle line:84% They either misspelled it as "u" or as "i". 00:01:14.080 --> 00:01:17.020 align:middle line:84% And then later, there was a period where people reliably 00:01:17.020 --> 00:01:19.720 align:middle line:84% misspelled it as "i," leading us to think 00:01:19.720 --> 00:01:22.240 align:middle line:84% that it had shifted to what it is in modern Greek, which 00:01:22.240 --> 00:01:24.220 align:middle line:90% is an "i." 00:01:24.220 --> 00:01:26.740 align:middle line:84% That's one of the kinds of things people do. 00:01:26.740 --> 00:01:29.380 align:middle line:84% You also get to look at what happens to these words 00:01:29.380 --> 00:01:33.170 align:middle line:84% when they get borrowed into other languages, 00:01:33.170 --> 00:01:37.480 align:middle line:84% so languages that have richer or poorer systems. 00:01:37.480 --> 00:01:40.870 align:middle line:84% There's one classic example from Greek and from Latin. 00:01:40.870 --> 00:01:44.758 align:middle line:90% We know that Greek-- 00:01:44.758 --> 00:01:46.570 align:middle line:84% I see, I'm sorry, Raquel, you were right. 00:01:46.570 --> 00:01:48.362 align:middle line:84% This is going to be a long-- a long answer. 00:01:48.362 --> 00:01:51.940 align:middle line:84% We know that the Greek letter, this Greek letter, 00:01:51.940 --> 00:01:56.510 align:middle line:84% used to be pronounced as a "p" with an extra puff of air. 00:01:56.510 --> 00:01:59.260 align:middle line:84% It used to be pi or pea [pronounced with aspiration], 00:01:59.260 --> 00:02:02.020 align:middle line:84% so just an especially emphatic "p." 00:02:02.020 --> 00:02:04.930 align:middle line:84% They also had another letter, which 00:02:04.930 --> 00:02:07.960 align:middle line:84% was a "p" without the extra puff of air, "p." 00:02:07.960 --> 00:02:10.780 align:middle line:90% And then this one shifted to-- 00:02:10.780 --> 00:02:13.390 align:middle line:84% well, actually, the modern-- this is the symbol for it 00:02:13.390 --> 00:02:18.970 align:middle line:84% in IPA, the bilabial fricative, a "fuh," a "fuh." 00:02:18.970 --> 00:02:21.190 align:middle line:84% And the way that we know-- we can actually 00:02:21.190 --> 00:02:24.650 align:middle line:84% date pretty precisely when this shift happened. 00:02:24.650 --> 00:02:28.570 align:middle line:84% And that's because Greek was loaning lots of words 00:02:28.570 --> 00:02:29.300 align:middle line:90% into Latin. 00:02:29.300 --> 00:02:31.180 align:middle line:84% Latin was borrowing words from Greek, 00:02:31.180 --> 00:02:34.465 align:middle line:84% including a word that they spelled like this. 00:02:34.465 --> 00:02:38.550 align:middle line:90% 00:02:38.550 --> 00:02:42.800 align:middle line:84% And the fact that they spelled it like that, first of all, 00:02:42.800 --> 00:02:44.872 align:middle line:84% suggests that it was a "p" with an-- 00:02:44.872 --> 00:02:46.580 align:middle line:84% a "p" with an "h" sound and not something 00:02:46.580 --> 00:02:49.940 align:middle line:84% more like an "f" because Latin had an "F." 00:02:49.940 --> 00:02:52.550 align:middle line:84% And they would have used an F if they had 00:02:52.550 --> 00:02:53.930 align:middle line:90% thought that this was an "f." 00:02:53.930 --> 00:02:57.260 align:middle line:84% Also, because people who were bad at spelling-- children, 00:02:57.260 --> 00:03:00.980 align:middle line:84% and people writing graffiti, and stuff like that-- 00:03:00.980 --> 00:03:03.740 align:middle line:84% when they misspelled this word, when they were writing graffiti 00:03:03.740 --> 00:03:06.620 align:middle line:84% about philosophy, they would misspell it 00:03:06.620 --> 00:03:10.070 align:middle line:84% by leaving out the H. So they would spell it just with a P. 00:03:10.070 --> 00:03:11.810 align:middle line:84% And then later, there was a period where 00:03:11.810 --> 00:03:13.352 align:middle line:84% people who were misspelling this word 00:03:13.352 --> 00:03:15.800 align:middle line:84% would misspell it by writing it with an F. 00:03:15.800 --> 00:03:18.770 align:middle line:84% And that's how you know that the word had changed 00:03:18.770 --> 00:03:21.410 align:middle line:84% its pronunciation right then at that point 00:03:21.410 --> 00:03:26.570 align:middle line:84% from an aspirated "p," a "p" with an "h" sound, to an "f." 00:03:26.570 --> 00:03:28.070 align:middle line:84% AUDIENCE: So when I misspell things, 00:03:28.070 --> 00:03:29.660 align:middle line:84% that's actually benefiting the future. 00:03:29.660 --> 00:03:30.952 align:middle line:90% NORVIN RICHARDS: You're being-- 00:03:30.952 --> 00:03:34.610 align:middle line:84% providing valuable data for future historical linguists, 00:03:34.610 --> 00:03:35.450 align:middle line:90% absolutely. 00:03:35.450 --> 00:03:36.450 align:middle line:90% That's absolutely true. 00:03:36.450 --> 00:03:38.720 align:middle line:84% Now, please engrave your misspellings in stone. 00:03:38.720 --> 00:03:41.690 align:middle line:84% That will make them a lot easier for people to handle. 00:03:41.690 --> 00:03:43.996 align:middle line:90% Any other questions about this? 00:03:43.996 --> 00:03:48.170 align:middle line:84% OK, where was I going with all that? 00:03:48.170 --> 00:03:50.870 align:middle line:90% Oh, right, so I was saying-- 00:03:50.870 --> 00:03:53.120 align:middle line:84% I was telling you some things about typology. 00:03:53.120 --> 00:03:55.950 align:middle line:84% There are languages that have poorer and richer vowel 00:03:55.950 --> 00:03:56.450 align:middle line:90% systems. 00:03:56.450 --> 00:03:58.940 align:middle line:84% English has a particularly rich system of vowels. 00:03:58.940 --> 00:04:00.895 align:middle line:84% We have a fairly large number of vowels. 00:04:00.895 --> 00:04:03.020 align:middle line:84% There are plenty of languages out there that have-- 00:04:03.020 --> 00:04:06.500 align:middle line:84% languages that have fewer, that have five. 00:04:06.500 --> 00:04:08.990 align:middle line:84% There are plenty of languages that have three. 00:04:08.990 --> 00:04:12.140 align:middle line:84% If you have three vowels, your three vowels 00:04:12.140 --> 00:04:17.000 align:middle line:84% are pretty invariably "a," "i," and "u" 00:04:17.000 --> 00:04:19.040 align:middle line:84% That is, they're the vowels that are kind of 00:04:19.040 --> 00:04:21.290 align:middle line:90% in the corners of this system. 00:04:21.290 --> 00:04:24.230 align:middle line:84% They're at the bottom right, and the top right, 00:04:24.230 --> 00:04:25.958 align:middle line:90% and the top left, yeah. 00:04:25.958 --> 00:04:28.250 align:middle line:84% There aren't languages out there that have three vowels 00:04:28.250 --> 00:04:31.520 align:middle line:84% and the three vowels are "eh," "ih," and "ah." 00:04:31.520 --> 00:04:35.000 align:middle line:84% That's not a kind of language that we find. 00:04:35.000 --> 00:04:37.790 align:middle line:84% And there's some work on this, including 00:04:37.790 --> 00:04:41.540 align:middle line:84% classic work by my colleague Edward Fleming, 00:04:41.540 --> 00:04:44.270 align:middle line:84% talking about optimal distributions 00:04:44.270 --> 00:04:46.970 align:middle line:84% of vowels and vowel space, where the idea is 00:04:46.970 --> 00:04:51.200 align:middle line:90% have this space in your mouth. 00:04:51.200 --> 00:04:54.860 align:middle line:84% You want your vowels to be maximally contrastive. 00:04:54.860 --> 00:04:57.320 align:middle line:84% So you want people to always be able to hear the difference 00:04:57.320 --> 00:04:58.520 align:middle line:90% between different vowels. 00:04:58.520 --> 00:05:00.603 align:middle line:84% And the best way to do that is to have your vowels 00:05:00.603 --> 00:05:04.280 align:middle line:84% as far apart from each other as possible, as a theory about why 00:05:04.280 --> 00:05:06.860 align:middle line:84% we get the kinds of vowel systems that we get and not 00:05:06.860 --> 00:05:09.390 align:middle line:90% others. 00:05:09.390 --> 00:05:12.113 align:middle line:90% OK. 00:05:12.113 --> 00:05:13.780 align:middle line:84% Oh, right, and then there are languages, 00:05:13.780 --> 00:05:16.513 align:middle line:84% like French, that have nasalized vowels. 00:05:16.513 --> 00:05:19.180 align:middle line:84% So French has words like "main," [with nasalized vowel] which is 00:05:19.180 --> 00:05:21.790 align:middle line:90% hand, and "met," which is dish. 00:05:21.790 --> 00:05:24.490 align:middle line:84% So that's another option that some languages take. 00:05:24.490 --> 00:05:26.290 align:middle line:84% You can lower your velum and allow some air 00:05:26.290 --> 00:05:29.770 align:middle line:84% to flow through your nasal cavity as you're making sounds. 00:05:29.770 --> 00:05:32.110 align:middle line:84% French, Portuguese, lots of languages do this. 00:05:32.110 --> 00:05:35.060 align:middle line:90% 00:05:35.060 --> 00:05:38.150 align:middle line:84% OK, so I said we're classifying vowels. 00:05:38.150 --> 00:05:39.440 align:middle line:90% Why are we classifying vowels? 00:05:39.440 --> 00:05:40.880 align:middle line:84% Partly because it's fun, but also 00:05:40.880 --> 00:05:44.810 align:middle line:84% because it enables us to realize that we wonder 00:05:44.810 --> 00:05:47.450 align:middle line:90% about kinds of vowel spaces. 00:05:47.450 --> 00:05:50.420 align:middle line:84% Why are all the English vowels in the upper right-hand corner? 00:05:50.420 --> 00:05:51.565 align:middle line:90% Do they have to be? 00:05:51.565 --> 00:05:53.690 align:middle line:84% If you had a rounded vowel that was somewhere else, 00:05:53.690 --> 00:05:54.732 align:middle line:90% what would it sound like? 00:05:54.732 --> 00:05:56.310 align:middle line:90% All that good stuff. 00:05:56.310 --> 00:05:58.340 align:middle line:84% Another thing that it does is help 00:05:58.340 --> 00:06:01.700 align:middle line:84% us to develop theories of phonologically natural sound 00:06:01.700 --> 00:06:02.370 align:middle line:90% changes. 00:06:02.370 --> 00:06:04.760 align:middle line:90% So here are some Turkish nouns. 00:06:04.760 --> 00:06:08.330 align:middle line:84% I'm sorry, I'm going to use your native language. 00:06:08.330 --> 00:06:10.070 align:middle line:84% And I'll just ask you to be quiet 00:06:10.070 --> 00:06:11.450 align:middle line:90% for the next little while. 00:06:11.450 --> 00:06:13.670 align:middle line:90% Here are some Turkish nouns. 00:06:13.670 --> 00:06:16.160 align:middle line:84% Some of you, if you've ever read CS Lewis, 00:06:16.160 --> 00:06:20.180 align:middle line:84% you might recognize the first one. 00:06:20.180 --> 00:06:25.070 align:middle line:84% And here are the plurals of these nouns. 00:06:25.070 --> 00:06:27.140 align:middle line:84% What's the plural suffix in Turkish? 00:06:27.140 --> 00:06:31.320 align:middle line:90% 00:06:31.320 --> 00:06:31.890 align:middle line:90% Yeah? 00:06:31.890 --> 00:06:32.850 align:middle line:90% AUDIENCE: L-A-R. 00:06:32.850 --> 00:06:36.810 align:middle line:84% NORVIN RICHARDS: So it's L-A-R, [TURKISH] 00:06:36.810 --> 00:06:40.073 align:middle line:84% like in the word for lions, and the word for arms, 00:06:40.073 --> 00:06:42.240 align:middle line:84% and the word for slaves, and the word for daughters. 00:06:42.240 --> 00:06:46.710 align:middle line:84% Oh, but not in the words for winds, teeth, and roses, 00:06:46.710 --> 00:06:50.580 align:middle line:84% where we seem to get a [TURKISH].. 00:06:50.580 --> 00:06:55.500 align:middle line:90% So we get L-E-R for those three. 00:06:55.500 --> 00:06:56.760 align:middle line:90% I'll write them down-- 00:06:56.760 --> 00:06:57.510 align:middle line:90% [TURKISH] 00:06:57.510 --> 00:07:01.770 align:middle line:90% 00:07:01.770 --> 00:07:06.150 align:middle line:90% And we get [TURKISH] for these-- 00:07:06.150 --> 00:07:06.870 align:middle line:90% [TURKISH] 00:07:06.870 --> 00:07:17.340 align:middle line:90% 00:07:17.340 --> 00:07:20.470 align:middle line:84% Having done that, I'm now going to go back to the last slide. 00:07:20.470 --> 00:07:24.260 align:middle line:90% 00:07:24.260 --> 00:07:25.060 align:middle line:90% What do you think? 00:07:25.060 --> 00:07:27.610 align:middle line:84% Why are we getting [TURKISH] with these and [TURKISH] 00:07:27.610 --> 00:07:28.180 align:middle line:90% with those? 00:07:28.180 --> 00:07:31.040 align:middle line:90% 00:07:31.040 --> 00:07:32.740 align:middle line:84% Joseph, what do you think is happening? 00:07:32.740 --> 00:07:34.990 align:middle line:84% AUDIENCE: I think this is an example of vowel harmony. 00:07:34.990 --> 00:07:37.198 align:middle line:84% NORVIN RICHARDS: I think this is an example of what's 00:07:37.198 --> 00:07:38.450 align:middle line:90% called vowel harmony. 00:07:38.450 --> 00:07:40.550 align:middle line:90% What do you mean? 00:07:40.550 --> 00:07:42.260 align:middle line:84% AUDIENCE: So a lot of languages will 00:07:42.260 --> 00:07:45.920 align:middle line:84% have a certain suffix or a prefix, affix 00:07:45.920 --> 00:07:47.310 align:middle line:90% without a morpheme. 00:07:47.310 --> 00:07:50.780 align:middle line:84% And when you attach to another word, the vowel, 00:07:50.780 --> 00:07:55.400 align:middle line:84% the primary vowel in that found morpheme will change to-- 00:07:55.400 --> 00:07:57.860 align:middle line:90% will harmonize with whatever-- 00:07:57.860 --> 00:08:02.770 align:middle line:84% with some vowel in the other [? word ?] [INAUDIBLE].. 00:08:02.770 --> 00:08:05.320 align:middle line:84% NORVIN RICHARDS: So that's well said. 00:08:05.320 --> 00:08:12.900 align:middle line:84% So what does this vowel have in common with these vowels? 00:08:12.900 --> 00:08:16.860 align:middle line:84% And what does this vowel have in common with these vowels? 00:08:16.860 --> 00:08:26.700 align:middle line:84% Bear in mind that "oo" is a front high grounded vowel 00:08:26.700 --> 00:08:35.610 align:middle line:84% and "uh" is a back high unrounded vowel. 00:08:35.610 --> 00:08:40.530 align:middle line:84% What do these vowels, "ay," "ee," and "oo" all have 00:08:40.530 --> 00:08:41.500 align:middle line:90% in common? 00:08:41.500 --> 00:08:42.120 align:middle line:90% Yes? 00:08:42.120 --> 00:08:43.120 align:middle line:90% AUDIENCE: They're front. 00:08:43.120 --> 00:08:45.650 align:middle line:84% NORVIN RICHARDS: They're all front. 00:08:45.650 --> 00:08:46.790 align:middle line:90% And what do these vowels-- 00:08:46.790 --> 00:08:50.090 align:middle line:90% "ah," "oh," "oo," and "uh"-- 00:08:50.090 --> 00:08:50.910 align:middle line:90% have in common? 00:08:50.910 --> 00:08:54.350 align:middle line:90% Well, they're all back, right? 00:08:54.350 --> 00:09:00.650 align:middle line:84% And the vowel in "lar" is back and the vowel in "ler" 00:09:00.650 --> 00:09:02.240 align:middle line:90% is front. 00:09:02.240 --> 00:09:05.450 align:middle line:84% So just as Joseph said, this is vowel harmony. 00:09:05.450 --> 00:09:07.820 align:middle line:84% The vowel of this suffix is having 00:09:07.820 --> 00:09:12.475 align:middle line:84% its properties determined by the vowel of the thing it's 00:09:12.475 --> 00:09:14.600 align:middle line:84% being attached to, the noun it's being attached to. 00:09:14.600 --> 00:09:18.170 align:middle line:84% Whether it's front or back is determined 00:09:18.170 --> 00:09:20.820 align:middle line:84% by the vowel of the noun it's attached to. 00:09:20.820 --> 00:09:23.880 align:middle line:90% So this is front-back harmony. 00:09:23.880 --> 00:09:25.550 align:middle line:90% There's also rounding harmony-- 00:09:25.550 --> 00:09:28.640 align:middle line:84% so cases where you have a suffix whose roundedness 00:09:28.640 --> 00:09:31.070 align:middle line:84% is determined by the roundedness of the vowel 00:09:31.070 --> 00:09:32.305 align:middle line:90% that it's attaching to. 00:09:32.305 --> 00:09:33.680 align:middle line:84% There are other kinds of harmony. 00:09:33.680 --> 00:09:35.270 align:middle line:84% But those are things that happen. 00:09:35.270 --> 00:09:38.180 align:middle line:84% And categorizing vowels, looking at vowels and thinking, 00:09:38.180 --> 00:09:40.520 align:middle line:84% ah yes, some of these are front and others are back, 00:09:40.520 --> 00:09:42.557 align:middle line:84% it's not just a fun thing to do that allows 00:09:42.557 --> 00:09:44.390 align:middle line:84% us to think about the insides of our mouths. 00:09:44.390 --> 00:09:47.018 align:middle line:84% It allows us to understand Turkish better, right? 00:09:47.018 --> 00:09:48.560 align:middle line:84% What the heck is going on in Turkish? 00:09:48.560 --> 00:09:51.080 align:middle line:84% Well, it's harmony for front versus back. 00:09:51.080 --> 00:09:56.440 align:middle line:90% 00:09:56.440 --> 00:10:02.230 align:middle line:84% OK, I think when I was first introducing you to consonants, 00:10:02.230 --> 00:10:09.370 align:middle line:84% I said, consonants, they involve disturbing the flow of air. 00:10:09.370 --> 00:10:14.660 align:middle line:84% And the flow of air is usually going outward from your lungs. 00:10:14.660 --> 00:10:16.150 align:middle line:84% But there are some other options, 00:10:16.150 --> 00:10:18.910 align:middle line:84% I said, and chuckled mysteriously. 00:10:18.910 --> 00:10:23.410 align:middle line:84% And so I want to just show you why I was doing that. 00:10:23.410 --> 00:10:27.690 align:middle line:84% There are other kinds of sounds out there. 00:10:27.690 --> 00:10:30.230 align:middle line:84% So the first kind of sound that I wanted to talk about 00:10:30.230 --> 00:10:32.270 align:middle line:90% are what are called ejectives. 00:10:32.270 --> 00:10:35.240 align:middle line:84% And I have some sound files that I'll play for you. 00:10:35.240 --> 00:10:37.640 align:middle line:84% These were all compiled and put online 00:10:37.640 --> 00:10:43.310 align:middle line:84% by the great phoneticist Peter Ladefoged, who 00:10:43.310 --> 00:10:44.570 align:middle line:90% used to teach at UCLA. 00:10:44.570 --> 00:10:47.330 align:middle line:90% He's no longer with us. 00:10:47.330 --> 00:10:52.550 align:middle line:84% I'll put a link on the website both to the particular classes 00:10:52.550 --> 00:10:54.050 align:middle line:84% of sounds that I'm going to show you 00:10:54.050 --> 00:10:56.120 align:middle line:84% and to the larger UCLA website, which 00:10:56.120 --> 00:10:57.590 align:middle line:84% has all kinds of cool sound files 00:10:57.590 --> 00:10:59.870 align:middle line:84% on it that you can listen to and amaze and annoy 00:10:59.870 --> 00:11:01.470 align:middle line:90% your friends with. 00:11:01.470 --> 00:11:03.770 align:middle line:84% So the first type of sound that I want to talk about 00:11:03.770 --> 00:11:05.750 align:middle line:90% are what are called ejectives. 00:11:05.750 --> 00:11:08.660 align:middle line:84% There aren't languages of Europe that have ejectives. 00:11:08.660 --> 00:11:11.210 align:middle line:84% But they're quite popular in the languages of North and South 00:11:11.210 --> 00:11:12.870 align:middle line:90% America. 00:11:12.870 --> 00:11:16.560 align:middle line:84% And so it's worth learning how to do them. 00:11:16.560 --> 00:11:19.970 align:middle line:84% Let's start by making an ejective "t." 00:11:19.970 --> 00:11:23.330 align:middle line:84% This is how you make an ejective "t." 00:11:23.330 --> 00:11:26.733 align:middle line:84% So first, we'll practice like this, you'll hold your breath. 00:11:26.733 --> 00:11:29.150 align:middle line:84% Don't do it now because I have other things to talk about. 00:11:29.150 --> 00:11:30.320 align:middle line:84% But what you're going to do is you're 00:11:30.320 --> 00:11:31.880 align:middle line:84% going to hold your breath and try 00:11:31.880 --> 00:11:34.050 align:middle line:90% to make an audible "t." sound. 00:11:34.050 --> 00:11:37.075 align:middle line:90% So go t, t, t. 00:11:37.075 --> 00:11:38.357 align:middle line:90% AUDIENCE: t, t, t. 00:11:38.357 --> 00:11:39.440 align:middle line:90% NORVIN RICHARDS: OK, good. 00:11:39.440 --> 00:11:42.400 align:middle line:84% And then if you were actually doing this in speech, 00:11:42.400 --> 00:11:45.160 align:middle line:84% you're not allowed to stop and hold your breath. 00:11:45.160 --> 00:11:48.880 align:middle line:90% The idea is to get-- 00:11:48.880 --> 00:11:51.070 align:middle line:84% well, so hold your breath, make your ejective "t," 00:11:51.070 --> 00:11:55.347 align:middle line:84% and then release into a vowel, so go ah, ta, ah, ta. 00:11:55.347 --> 00:11:58.340 align:middle line:90% AUDIENCE: Ah, ta, ah, ta. 00:11:58.340 --> 00:12:00.830 align:middle line:84% NORVIN RICHARDS: Cool, you guys sound very ejective. 00:12:00.830 --> 00:12:02.720 align:middle line:84% And similarly, you can do this with bilabial. 00:12:02.720 --> 00:12:07.010 align:middle line:84% There are bilabial ejective-- ah, pa, ah, pa-- 00:12:07.010 --> 00:12:12.860 align:middle line:84% or velar ejective stops like ah, ca, ah, ca. 00:12:12.860 --> 00:12:15.480 align:middle line:90% Cool, excellent. 00:12:15.480 --> 00:12:19.890 align:middle line:84% So we're all amateurs at this, but the Lakota 00:12:19.890 --> 00:12:21.270 align:middle line:90% are professionals. 00:12:21.270 --> 00:12:25.073 align:middle line:84% So let's listen to a Lakota speaker do this. 00:12:25.073 --> 00:12:25.740 align:middle line:90% [AUDIO PLAYBACK] 00:12:25.740 --> 00:12:26.335 align:middle line:90% 00:12:26.335 --> 00:12:28.085 align:middle line:84% NORVIN RICHARDS: That's an ejective velar. 00:12:28.085 --> 00:12:30.767 align:middle line:90% 00:12:30.767 --> 00:12:31.350 align:middle line:90% [END PLAYBACK] 00:12:31.350 --> 00:12:34.470 align:middle line:84% NORVIN RICHARDS: As opposed to, here's her regular velar stop. 00:12:34.470 --> 00:12:35.220 align:middle line:90% [AUDIO PLAYBACK] 00:12:35.220 --> 00:12:36.887 align:middle line:90% 00:12:36.887 --> 00:12:37.470 align:middle line:90% [END PLAYBACK] 00:12:37.470 --> 00:12:39.143 align:middle line:84% NORVIN RICHARDS: Here's a bilabial one. 00:12:39.143 --> 00:12:39.810 align:middle line:90% [AUDIO PLAYBACK] 00:12:39.810 --> 00:12:40.637 align:middle line:90% 00:12:40.637 --> 00:12:41.220 align:middle line:90% [END PLAYBACK] 00:12:41.220 --> 00:12:42.660 align:middle line:84% NORVIN RICHARDS: So that's what ejectives 00:12:42.660 --> 00:12:43.618 align:middle line:90% sound like in the wild. 00:12:43.618 --> 00:12:44.250 align:middle line:90% This is Lakota. 00:12:44.250 --> 00:12:47.850 align:middle line:84% It's a language still spoken, well, 00:12:47.850 --> 00:12:49.200 align:middle line:90% near North and South Dakota. 00:12:49.200 --> 00:12:52.245 align:middle line:84% So there are the Lakotas, and the Nakotas, and the Dakotas. 00:12:52.245 --> 00:12:53.370 align:middle line:90% They all live in that area. 00:12:53.370 --> 00:12:54.630 align:middle line:90% They're Siouan languages. 00:12:54.630 --> 00:12:55.130 align:middle line:90% Yeah? 00:12:55.130 --> 00:12:55.782 align:middle line:90% Oh, Raquel? 00:12:55.782 --> 00:12:57.990 align:middle line:84% AUDIENCE: What's the difference between that and just 00:12:57.990 --> 00:13:01.440 align:middle line:84% the normal "k" sound followed by the glottal stop? 00:13:01.440 --> 00:13:04.690 align:middle line:84% NORVIN RICHARDS: So it's very similar. 00:13:04.690 --> 00:13:07.530 align:middle line:84% You're engaging your-- so what you're 00:13:07.530 --> 00:13:11.130 align:middle line:84% doing when you make these sounds is 00:13:11.130 --> 00:13:16.920 align:middle line:90% you are making a velar closure. 00:13:16.920 --> 00:13:21.600 align:middle line:84% And then, you are also making a glottal stop. 00:13:21.600 --> 00:13:24.450 align:middle line:84% And then, you are pushing with your glottis 00:13:24.450 --> 00:13:27.840 align:middle line:90% to make the velar closure open. 00:13:27.840 --> 00:13:30.600 align:middle line:84% So you make your-- you move your larynx in order 00:13:30.600 --> 00:13:32.310 align:middle line:90% to get the closure to open. 00:13:32.310 --> 00:13:34.553 align:middle line:84% And that's what gives it this unique burst. 00:13:34.553 --> 00:13:35.220 align:middle line:90% So you're right. 00:13:35.220 --> 00:13:37.925 align:middle line:84% There is a glottal stop involved when you're doing that. 00:13:37.925 --> 00:13:41.610 align:middle line:84% We're at-- yeah, so when I said that the flow of air 00:13:41.610 --> 00:13:44.070 align:middle line:84% is usually from your lungs, here's 00:13:44.070 --> 00:13:46.470 align:middle line:84% the place where it isn't exactly. 00:13:46.470 --> 00:13:48.840 align:middle line:84% The air originally came from your lungs. 00:13:48.840 --> 00:13:51.660 align:middle line:84% Well, it originally came from the world. 00:13:51.660 --> 00:13:53.070 align:middle line:84% It spent some time in your lungs. 00:13:53.070 --> 00:13:56.240 align:middle line:84% And then, it got trapped between this glottal stop 00:13:56.240 --> 00:13:57.990 align:middle line:84% at your larynx and the closure that you're 00:13:57.990 --> 00:13:59.130 align:middle line:90% making in your mouth. 00:13:59.130 --> 00:14:01.080 align:middle line:84% And the flow of air that causes-- 00:14:01.080 --> 00:14:03.710 align:middle line:84% that's present as you're opening the stop-- 00:14:03.710 --> 00:14:06.060 align:middle line:90% is actually from your larynx. 00:14:06.060 --> 00:14:09.480 align:middle line:90% It's not from your lungs. 00:14:09.480 --> 00:14:12.790 align:middle line:84% It's not originally from your lungs, if that makes sense. 00:14:12.790 --> 00:14:14.300 align:middle line:90% So those are ejectives. 00:14:14.300 --> 00:14:17.460 align:middle line:90% 00:14:17.460 --> 00:14:19.770 align:middle line:90% Clicks. 00:14:19.770 --> 00:14:22.495 align:middle line:84% I have some recordings of clicks in here somewhere. 00:14:22.495 --> 00:14:23.370 align:middle line:90% Yeah, there they are. 00:14:23.370 --> 00:14:25.260 align:middle line:90% These are Zulu clicks. 00:14:25.260 --> 00:14:28.710 align:middle line:84% There are three main kinds of clicks to think about. 00:14:28.710 --> 00:14:32.910 align:middle line:84% And clicks, again, involve a strange and cross-- 00:14:32.910 --> 00:14:35.250 align:middle line:84% typologically unusual, kind of, airflow. 00:14:35.250 --> 00:14:39.930 align:middle line:90% So clicks are found in-- 00:14:39.930 --> 00:14:43.320 align:middle line:84% so the languages with the richest click inventories 00:14:43.320 --> 00:14:47.850 align:middle line:84% are the Khoisan languages of places like Namibia. 00:14:47.850 --> 00:14:51.090 align:middle line:84% And then, there are a bunch of Bantu languages 00:14:51.090 --> 00:14:56.460 align:middle line:84% that also have clicks, also in Africa, in theory 00:14:56.460 --> 00:14:59.100 align:middle line:84% because they've been in contact with the Khoisan languages. 00:14:59.100 --> 00:15:01.320 align:middle line:84% So clicks are really cross-linguistically 00:15:01.320 --> 00:15:02.610 align:middle line:90% not terribly popular. 00:15:02.610 --> 00:15:09.420 align:middle line:84% They're only found in that area and in one 00:15:09.420 --> 00:15:12.420 align:middle line:84% culturally restricted language spoken by the Lardil. 00:15:12.420 --> 00:15:15.880 align:middle line:84% So the Lardil have-- it's almost a language game. 00:15:15.880 --> 00:15:20.700 align:middle line:84% It's called Damin, which is used after a certain initiation 00:15:20.700 --> 00:15:21.270 align:middle line:90% ceremony. 00:15:21.270 --> 00:15:22.560 align:middle line:84% So there's a period where you're not supposed 00:15:22.560 --> 00:15:23.850 align:middle line:90% to speak normal Lardil. 00:15:23.850 --> 00:15:26.007 align:middle line:84% And so you speak this other thing instead, 00:15:26.007 --> 00:15:28.590 align:middle line:84% which has all kinds of strange sounds in it, including clicks. 00:15:28.590 --> 00:15:29.340 align:middle line:90% It's called Damin. 00:15:29.340 --> 00:15:31.720 align:middle line:84% We may get a chance to talk more about it later. 00:15:31.720 --> 00:15:33.330 align:middle line:84% But for languages that are not spoken 00:15:33.330 --> 00:15:36.180 align:middle line:84% under special cultural circumstances, 00:15:36.180 --> 00:15:38.850 align:middle line:84% these African languages, the Khoisan languages and the Bantu 00:15:38.850 --> 00:15:40.767 align:middle line:84% languages that have been in contact with them, 00:15:40.767 --> 00:15:42.840 align:middle line:84% are the only ones that have clicks. 00:15:42.840 --> 00:15:44.910 align:middle line:84% And I'm talking about clicks now because they 00:15:44.910 --> 00:15:47.370 align:middle line:90% do involve an unusual airflow. 00:15:47.370 --> 00:15:50.580 align:middle line:84% They involve making a stop, a velar stop, 00:15:50.580 --> 00:15:54.750 align:middle line:84% and then using your tongue to sharply draw air 00:15:54.750 --> 00:15:58.450 align:middle line:90% into your mouth for a moment. 00:15:58.450 --> 00:15:59.790 align:middle line:90% So what's a click? 00:15:59.790 --> 00:16:02.070 align:middle line:90% One click is the dental click. 00:16:02.070 --> 00:16:07.770 align:middle line:84% It's the, kind of, sound you make to criticize someone, 00:16:07.770 --> 00:16:10.540 align:middle line:84% to [TSK-ING] thing that you sometimes-- 00:16:10.540 --> 00:16:11.040 align:middle line:90% [LAUGHTER] 00:16:11.040 --> 00:16:11.868 align:middle line:90% --see spelled tsk. 00:16:11.868 --> 00:16:18.870 align:middle line:84% T-S-K. T-S-K. Yeah, there is a lateral click. 00:16:18.870 --> 00:16:22.980 align:middle line:84% Whenever I read about lateral clicks, the description of them 00:16:22.980 --> 00:16:26.730 align:middle line:84% is always that it's the sound you make to encourage a horse. 00:16:26.730 --> 00:16:28.605 align:middle line:84% I have never had to encourage a horse-- 00:16:28.605 --> 00:16:29.820 align:middle line:90% [LAUGHTER] 00:16:29.820 --> 00:16:30.755 align:middle line:90% --or discourage a-- 00:16:30.755 --> 00:16:32.880 align:middle line:84% I've never-- I haven't interacted much with horses. 00:16:32.880 --> 00:16:36.940 align:middle line:84% So I'm not-- that description doesn't help me a whole lot. 00:16:36.940 --> 00:16:39.960 align:middle line:84% I gather it involves sucking the sides of your tongue 00:16:39.960 --> 00:16:41.996 align:middle line:90% in from the sides of your mouth. 00:16:41.996 --> 00:16:43.121 align:middle line:90% So you're going [CLICKING]. 00:16:43.121 --> 00:16:45.285 align:middle line:90% AUDIENCE: [CLICKING] 00:16:45.285 --> 00:16:47.660 align:middle line:84% NORVIN RICHARDS: Yes, I think I hear some lateral clicks. 00:16:47.660 --> 00:16:48.560 align:middle line:90% [LAUGHTER] 00:16:48.560 --> 00:16:49.460 align:middle line:90% 00:16:49.460 --> 00:16:52.800 align:middle line:84% And then, there's another click, which 00:16:52.800 --> 00:16:55.767 align:middle line:84% is sometimes called a palatal click, which goes [CLUCKING].. 00:16:55.767 --> 00:16:56.600 align:middle line:90% AUDIENCE: [CLUCKING] 00:16:56.600 --> 00:16:58.072 align:middle line:90% NORVIN RICHARDS: So you're-- 00:16:58.072 --> 00:16:59.780 align:middle line:84% oh, you guys are really good at that one. 00:16:59.780 --> 00:17:00.280 align:middle line:90% [LAUGHING] 00:17:00.280 --> 00:17:03.830 align:middle line:84% Yeah, so all three of these clicks 00:17:03.830 --> 00:17:05.420 align:middle line:84% are present, for example, in Zulu 00:17:05.420 --> 00:17:08.660 align:middle line:84% and in Xhosa, which is the native language of Nelson 00:17:08.660 --> 00:17:11.119 align:middle line:90% Mandela. 00:17:11.119 --> 00:17:14.359 align:middle line:84% And here are some Zulu ones that you can listen to. 00:17:14.359 --> 00:17:16.723 align:middle line:84% So we've got the dental one first. 00:17:16.723 --> 00:17:17.390 align:middle line:90% [AUDIO PLAYBACK] 00:17:17.390 --> 00:17:18.817 align:middle line:90% 00:17:18.817 --> 00:17:19.400 align:middle line:90% [END PLAYBACK] 00:17:19.400 --> 00:17:21.991 align:middle line:84% NORVIN RICHARDS: And here's the lateral one. 00:17:21.991 --> 00:17:22.658 align:middle line:90% [AUDIO PLAYBACK] 00:17:22.658 --> 00:17:23.962 align:middle line:90% 00:17:23.962 --> 00:17:24.545 align:middle line:90% [END PLAYBACK] 00:17:24.545 --> 00:17:26.180 align:middle line:84% NORVIN RICHARDS: And the alveopalatal one. 00:17:26.180 --> 00:17:26.847 align:middle line:90% [AUDIO PLAYBACK] 00:17:26.847 --> 00:17:28.520 align:middle line:90% 00:17:28.520 --> 00:17:29.183 align:middle line:90% [END PLAYBACK] 00:17:29.183 --> 00:17:31.100 align:middle line:84% NORVIN RICHARDS: And as you can see from this, 00:17:31.100 --> 00:17:36.440 align:middle line:84% Zulu draws distinctions depending on, basically, 00:17:36.440 --> 00:17:38.540 align:middle line:90% how you release the stave. 00:17:38.540 --> 00:17:41.510 align:middle line:90% So there are clicks-- 00:17:41.510 --> 00:17:43.910 align:middle line:84% those clicks, the ones I just played for you, 00:17:43.910 --> 00:17:47.902 align:middle line:84% involve a voiceless stop that's not aspirated, 00:17:47.902 --> 00:17:50.360 align:middle line:84% that doesn't have this puff of air that I was talking about 00:17:50.360 --> 00:17:51.867 align:middle line:84% and that we will talk about more. 00:17:51.867 --> 00:17:53.450 align:middle line:84% I was talking about when I was talking 00:17:53.450 --> 00:17:55.940 align:middle line:90% about "philosophy" and Greek. 00:17:55.940 --> 00:17:58.550 align:middle line:84% But you can also build them on aspirated stops, 00:17:58.550 --> 00:18:02.187 align:middle line:84% or on voiced stops, or even on nasals. 00:18:02.187 --> 00:18:04.520 align:middle line:84% So maybe I'll just run through all of the dental clicks. 00:18:04.520 --> 00:18:05.720 align:middle line:90% Here's the voiceless one again. 00:18:05.720 --> 00:18:06.387 align:middle line:90% [AUDIO PLAYBACK] 00:18:06.387 --> 00:18:07.807 align:middle line:90% 00:18:07.807 --> 00:18:08.390 align:middle line:90% [END PLAYBACK] 00:18:08.390 --> 00:18:09.350 align:middle line:84% NORVIN RICHARDS: But here, it is aspirated. 00:18:09.350 --> 00:18:10.200 align:middle line:90% [AUDIO PLAYBACK] 00:18:10.200 --> 00:18:10.797 align:middle line:90% 00:18:10.797 --> 00:18:11.380 align:middle line:90% [END PLAYBACK] 00:18:11.380 --> 00:18:12.338 align:middle line:90% NORVIN RICHARDS: Ta ga. 00:18:12.338 --> 00:18:15.080 align:middle line:84% So there's air coming out as you do this. 00:18:15.080 --> 00:18:16.040 align:middle line:90% And here it is voiced. 00:18:16.040 --> 00:18:17.340 align:middle line:90% [AUDIO PLAYBACK] 00:18:17.340 --> 00:18:18.050 align:middle line:90% 00:18:18.050 --> 00:18:19.440 align:middle line:90% [END PLAYBACK] 00:18:19.440 --> 00:18:21.065 align:middle line:84% NORVIN RICHARDS: And here, it is nasal. 00:18:21.065 --> 00:18:21.732 align:middle line:90% [AUDIO PLAYBACK] 00:18:21.732 --> 00:18:23.107 align:middle line:90% 00:18:23.107 --> 00:18:23.690 align:middle line:90% [END PLAYBACK] 00:18:23.690 --> 00:18:25.250 align:middle line:84% NORVIN RICHARDS: Whoa, that hardly sounded like a click 00:18:25.250 --> 00:18:25.750 align:middle line:90% at all. 00:18:25.750 --> 00:18:27.736 align:middle line:84% Here, let me try to find another nasal one. 00:18:27.736 --> 00:18:28.450 align:middle line:90% [AUDIO PLAYBACK] 00:18:28.450 --> 00:18:29.387 align:middle line:90% 00:18:29.387 --> 00:18:29.970 align:middle line:90% [END PLAYBACK] 00:18:29.970 --> 00:18:31.553 align:middle line:84% NORVIN RICHARDS: Yep, that was clicky. 00:18:31.553 --> 00:18:32.250 align:middle line:90% [LAUGHTER] 00:18:32.250 --> 00:18:34.080 align:middle line:90% And here it is as a lateral. 00:18:34.080 --> 00:18:35.220 align:middle line:90% [AUDIO PLAYBACK] 00:18:35.220 --> 00:18:36.343 align:middle line:90% 00:18:36.343 --> 00:18:37.510 align:middle line:90% NORVIN RICHARDS: [CHUCKLING] 00:18:37.510 --> 00:18:38.200 align:middle line:90% 00:18:38.200 --> 00:18:38.940 align:middle line:90% [END PLAYBACK] 00:18:38.940 --> 00:18:41.230 align:middle line:84% NORVIN RICHARDS: Zulu really does sound like that. 00:18:41.230 --> 00:18:41.730 align:middle line:90% [LAUGHING] 00:18:41.730 --> 00:18:43.280 align:middle line:84% We did a field methods class on Zulu. 00:18:43.280 --> 00:18:45.300 align:middle line:84% And it has this beautiful-- because it's tonal. 00:18:45.300 --> 00:18:47.460 align:middle line:84% So it has all of this-- the intonation is just 00:18:47.460 --> 00:18:50.190 align:middle line:90% really pretty to listen to. 00:18:50.190 --> 00:18:51.967 align:middle line:84% So clicks is another kind of thing, not 00:18:51.967 --> 00:18:53.550 align:middle line:84% all that popular cross-linguistically. 00:18:53.550 --> 00:18:54.270 align:middle line:90% But it happens. 00:18:54.270 --> 00:18:55.800 align:middle line:90% And they're magnificent. 00:18:55.800 --> 00:18:59.340 align:middle line:84% The Khoisan languages have even richer arrays 00:18:59.340 --> 00:19:01.380 align:middle line:90% of click inventories than this. 00:19:01.380 --> 00:19:05.510 align:middle line:84% In particular, they also have bilabial clicks. 00:19:05.510 --> 00:19:07.340 align:middle line:84% So they have [CLICKING] as a sound. 00:19:07.340 --> 00:19:09.807 align:middle line:84% [CLICKING],, that kind of a sound. 00:19:09.807 --> 00:19:12.530 align:middle line:84% And they make all of these distinctions 00:19:12.530 --> 00:19:13.940 align:middle line:90% among their clicks. 00:19:13.940 --> 00:19:16.400 align:middle line:84% The result ends up being that the Khoisan languages, if you 00:19:16.400 --> 00:19:19.940 align:middle line:84% count all of these clicks as different consonants, 00:19:19.940 --> 00:19:22.775 align:middle line:84% the Khoisan languages have just huge inventories of consonants 00:19:22.775 --> 00:19:24.650 align:middle line:84% because they have all these different clicks. 00:19:24.650 --> 00:19:25.670 align:middle line:90% Yes? 00:19:25.670 --> 00:19:28.964 align:middle line:84% AUDIENCE: Are the accents above the vowels indicative 00:19:28.964 --> 00:19:30.757 align:middle line:90% of what direction the tone is? 00:19:30.757 --> 00:19:32.590 align:middle line:84% NORVIN RICHARDS: Yeah, those are tone marks. 00:19:32.590 --> 00:19:37.690 align:middle line:84% Yeah, so the accents that points down to the right 00:19:37.690 --> 00:19:38.500 align:middle line:90% are low-tones. 00:19:38.500 --> 00:19:40.875 align:middle line:84% And the accents the point up to the right are high-tones. 00:19:40.875 --> 00:19:42.580 align:middle line:90% And that's all Zulu's got. 00:19:42.580 --> 00:19:44.470 align:middle line:84% It's got high-tones and low-tones. 00:19:44.470 --> 00:19:47.440 align:middle line:84% And then, you can have long vowels that transition 00:19:47.440 --> 00:19:50.950 align:middle line:84% from a high-tone to a low-tone or vice versa. 00:19:50.950 --> 00:19:54.460 align:middle line:90% Yeah, yep. 00:19:54.460 --> 00:19:56.940 align:middle line:90% Clicks. 00:19:56.940 --> 00:19:57.900 align:middle line:90% Implosives. 00:19:57.900 --> 00:19:59.640 align:middle line:90% Let me find you some implosives. 00:19:59.640 --> 00:20:01.260 align:middle line:90% They're around here somewhere. 00:20:01.260 --> 00:20:02.240 align:middle line:90% Where were they? 00:20:02.240 --> 00:20:04.410 align:middle line:90% These are implosives. 00:20:04.410 --> 00:20:10.800 align:middle line:84% Implosives, what implosives are, are sounds where-- 00:20:10.800 --> 00:20:13.130 align:middle line:84% so what you should do if you want to make an implosive, 00:20:13.130 --> 00:20:14.550 align:middle line:90% ah... 00:20:14.550 --> 00:20:18.090 align:middle line:84% "b," for example, is you should make your bilabial closure. 00:20:18.090 --> 00:20:20.250 align:middle line:84% And then, you should try to inhale. 00:20:20.250 --> 00:20:23.370 align:middle line:84% So it's like, "bah, bah, bah, bah" 00:20:23.370 --> 00:20:25.110 align:middle line:90% [pronounced with implosive "b"]. 00:20:25.110 --> 00:20:26.980 align:middle line:84% And similarly, so don't hurt yourself. 00:20:26.980 --> 00:20:28.480 align:middle line:84% But that's what you're trying to do. 00:20:28.480 --> 00:20:29.260 align:middle line:90% [LAUGHTER] 00:20:29.260 --> 00:20:31.170 align:middle line:84% And there are also velar implosives, 00:20:31.170 --> 00:20:32.130 align:middle line:84% which you should really be careful with. 00:20:32.130 --> 00:20:33.167 align:middle line:84% They're "ag ba, ag ba" [pronounced with implosive 00:20:33.167 --> 00:20:33.667 align:middle line:90% "g"]. 00:20:33.667 --> 00:20:34.600 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:20:34.600 --> 00:20:35.730 align:middle line:84% NORVIN RICHARDS: Or "ah dah, ah dah" 00:20:35.730 --> 00:20:36.104 align:middle line:90% [pronounced with implosive "d" 00:20:36.104 --> 00:20:37.040 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:20:37.040 --> 00:20:39.060 align:middle line:84% NORVIN RICHARDS: Yeah, [INAUDIBLE].. 00:20:39.060 --> 00:20:41.950 align:middle line:90% So here's a native speaker. 00:20:41.950 --> 00:20:45.960 align:middle line:84% So these are not all that uncommon in languages of India. 00:20:45.960 --> 00:20:49.230 align:middle line:84% There are a bunch of them in Indonesia. 00:20:49.230 --> 00:20:49.890 align:middle line:90% Here are some. 00:20:49.890 --> 00:20:50.610 align:middle line:90% [AUDIO PLAYBACK] 00:20:50.610 --> 00:20:51.197 align:middle line:90% 00:20:51.197 --> 00:20:51.780 align:middle line:90% [END PLAYBACK] 00:20:51.780 --> 00:20:54.450 align:middle line:84% NORVIN RICHARDS: That's a bilabial one. 00:20:54.450 --> 00:20:55.658 align:middle line:90% Here's a retroflex one. 00:20:55.658 --> 00:20:56.325 align:middle line:90% [AUDIO PLAYBACK] 00:20:56.325 --> 00:20:56.612 align:middle line:90% 00:20:56.612 --> 00:20:57.195 align:middle line:90% [END PLAYBACK] 00:20:57.195 --> 00:20:57.536 align:middle line:84% NORVIN RICHARDS: "Dinu" [pronounced with voiced 00:20:57.536 --> 00:20:58.065 align:middle line:90% retroflex implosive "d"]. 00:20:58.065 --> 00:20:58.565 align:middle line:90% All right. 00:20:58.565 --> 00:21:00.600 align:middle line:84% So we can compare his implosive one 00:21:00.600 --> 00:21:01.793 align:middle line:90% with the regular voiced one. 00:21:01.793 --> 00:21:03.210 align:middle line:84% So here's the implosive one again. 00:21:03.210 --> 00:21:03.877 align:middle line:90% [AUDIO PLAYBACK] 00:21:03.877 --> 00:21:04.030 align:middle line:90% 00:21:04.030 --> 00:21:04.220 align:middle line:90% [END PLAYBACK] 00:21:04.220 --> 00:21:06.637 align:middle line:84% NORVIN RICHARDS: Yeah, and then, here it is not implosive. 00:21:06.637 --> 00:21:07.303 align:middle line:90% [AUDIO PLAYBACK] 00:21:07.303 --> 00:21:07.470 align:middle line:90% 00:21:07.470 --> 00:21:08.160 align:middle line:90% [END PLAYBACK] 00:21:08.160 --> 00:21:10.150 align:middle line:84% NORVIN RICHARDS: Yeah, and here's 00:21:10.150 --> 00:21:11.650 align:middle line:84% what it sounds like when it's velar. 00:21:11.650 --> 00:21:12.317 align:middle line:90% [AUDIO PLAYBACK] 00:21:12.317 --> 00:21:12.357 align:middle line:90% 00:21:12.357 --> 00:21:12.940 align:middle line:90% [END PLAYBACK] 00:21:12.940 --> 00:21:14.530 align:middle line:84% NORVIN RICHARDS: As opposed to the regular one. 00:21:14.530 --> 00:21:15.197 align:middle line:90% [AUDIO PLAYBACK] 00:21:15.197 --> 00:21:15.507 align:middle line:90% 00:21:15.507 --> 00:21:16.090 align:middle line:90% [END PLAYBACK] 00:21:16.090 --> 00:21:19.750 align:middle line:84% NORVIN RICHARDS: Yeah, so again, I'll put these files up. 00:21:19.750 --> 00:21:25.098 align:middle line:84% You can play with them at home, amaze and annoy your friends. 00:21:25.098 --> 00:21:26.390 align:middle line:90% No questions about any of that? 00:21:26.390 --> 00:21:30.200 align:middle line:84% So again, this has all been about interesting air flows. 00:21:30.200 --> 00:21:33.110 align:middle line:84% So here's one where the flow of air is not outward at all. 00:21:33.110 --> 00:21:36.070 align:middle line:84% It's going inward for just a millisecond, yeah, trying 00:21:36.070 --> 00:21:38.860 align:middle line:84% to suck air into your body instead of getting it out 00:21:38.860 --> 00:21:40.660 align:middle line:90% as you speak. 00:21:40.660 --> 00:21:44.860 align:middle line:84% There are-- so I don't know of any languages that 00:21:44.860 --> 00:21:48.070 align:middle line:84% have, for example, a class of vowels 00:21:48.070 --> 00:21:50.920 align:middle line:84% that you must make while inhaling. 00:21:50.920 --> 00:21:53.110 align:middle line:84% So these are consonants that you make 00:21:53.110 --> 00:21:54.428 align:middle line:90% while attempting to inhale. 00:21:54.428 --> 00:21:56.720 align:middle line:84% That's the sense in which the flow of air is different. 00:21:56.720 --> 00:21:59.320 align:middle line:84% You could imagine a language that had the vowel "a" 00:21:59.320 --> 00:22:02.622 align:middle line:90% and also the vowel [GASP]. 00:22:02.622 --> 00:22:04.330 align:middle line:84% I've never heard of a language like that. 00:22:04.330 --> 00:22:08.500 align:middle line:84% There are languages that will inhale for particular words 00:22:08.500 --> 00:22:09.340 align:middle line:90% or classes of words. 00:22:09.340 --> 00:22:15.190 align:middle line:84% So I'm told that in Finland, there's 00:22:15.190 --> 00:22:18.400 align:middle line:84% an expression that basically means "yes." 00:22:18.400 --> 00:22:22.480 align:middle line:84% But it's used only by middle-aged women. 00:22:22.480 --> 00:22:24.010 align:middle line:90% And it involves inhaling. 00:22:24.010 --> 00:22:26.530 align:middle line:90% You say, [GASP]. 00:22:26.530 --> 00:22:29.180 align:middle line:84% And that marks you as a middle-aged Finnish woman. 00:22:29.180 --> 00:22:29.680 align:middle line:90% [LAUGHTER] 00:22:29.680 --> 00:22:31.245 align:middle line:84% So if you're ever attempting to sound 00:22:31.245 --> 00:22:32.620 align:middle line:84% like a middle-aged Finnish woman, 00:22:32.620 --> 00:22:34.535 align:middle line:90% you should use that a lot, yeah. 00:22:34.535 --> 00:22:36.160 align:middle line:84% But I've never heard of a language that 00:22:36.160 --> 00:22:38.830 align:middle line:84% had-- there are languages that have a series of nasal vowels, 00:22:38.830 --> 00:22:39.470 align:middle line:90% for example. 00:22:39.470 --> 00:22:40.360 align:middle line:84% But there aren't languages that have 00:22:40.360 --> 00:22:41.590 align:middle line:90% a series of inhaled vowels. 00:22:41.590 --> 00:22:44.080 align:middle line:84% That doesn't happen, no, as far as I know. 00:22:44.080 --> 00:22:48.198 align:middle line:90% 00:22:48.198 --> 00:22:50.490 align:middle line:84% And then trills, these don't involve a special airflow. 00:22:50.490 --> 00:22:53.150 align:middle line:84% But I wanted to talk about them partly because they came up 00:22:53.150 --> 00:22:53.650 align:middle line:90% in class. 00:22:53.650 --> 00:22:53.730 align:middle line:90% AUDIENCE: [TRILLING] 00:22:53.730 --> 00:22:55.260 align:middle line:84% NORVIN RICHARDS: Yeah, like that. 00:22:55.260 --> 00:23:00.210 align:middle line:84% So trills involve getting some type of-- some part 00:23:00.210 --> 00:23:04.200 align:middle line:90% of your vocal tract to flutter. 00:23:04.200 --> 00:23:07.890 align:middle line:84% And there are ones that are comparatively famous, 00:23:07.890 --> 00:23:09.615 align:middle line:90% like the so-called rolled "r." 00:23:09.615 --> 00:23:12.833 align:middle line:90% 00:23:12.833 --> 00:23:15.250 align:middle line:84% These are some examples that are going to sound like that. 00:23:15.250 --> 00:23:16.042 align:middle line:90% This is a language. 00:23:16.042 --> 00:23:17.700 align:middle line:84% This is a Dravidian language called 00:23:17.700 --> 00:23:22.360 align:middle line:84% Toda that has dental trills and also postalveolar trills. 00:23:22.360 --> 00:23:24.650 align:middle line:90% So you can listen to those. 00:23:24.650 --> 00:23:25.610 align:middle line:90% Sure we can. 00:23:25.610 --> 00:23:27.080 align:middle line:90% Come on. 00:23:27.080 --> 00:23:29.755 align:middle line:90% Wake up, website. 00:23:29.755 --> 00:23:30.380 align:middle line:90% I take it back. 00:23:30.380 --> 00:23:31.992 align:middle line:90% We can't listen to those. 00:23:31.992 --> 00:23:33.200 align:middle line:90% Why can't we listen to those? 00:23:33.200 --> 00:23:35.270 align:middle line:84% I was listening to them just now. 00:23:35.270 --> 00:23:36.250 align:middle line:90% What is your deal? 00:23:36.250 --> 00:23:36.750 align:middle line:90% Never mind. 00:23:36.750 --> 00:23:37.875 align:middle line:90% We'll go to this other one. 00:23:37.875 --> 00:23:40.680 align:middle line:90% 00:23:40.680 --> 00:23:43.700 align:middle line:84% So the rolled "r" the [PURRING] sound, 00:23:43.700 --> 00:23:46.550 align:middle line:84% that's the alveolar trill, where you're 00:23:46.550 --> 00:23:48.920 align:middle line:84% holding your tongue fairly close to your alveolar ridge, 00:23:48.920 --> 00:23:51.950 align:middle line:84% and getting the air to flow, and holding your tongue in such 00:23:51.950 --> 00:23:55.610 align:middle line:84% a way that your tongue flutters in the breeze as it goes by. 00:23:55.610 --> 00:23:56.730 align:middle line:90% English doesn't have that. 00:23:56.730 --> 00:24:00.080 align:middle line:84% But if you want to learn to speak Italian, or Spanish, 00:24:00.080 --> 00:24:02.870 align:middle line:84% or a variety of languages of Australia, 00:24:02.870 --> 00:24:06.440 align:middle line:84% you need to learn how to roll your "r"s. 00:24:06.440 --> 00:24:10.610 align:middle line:84% It's possible to be physically incapable of doing that. 00:24:10.610 --> 00:24:13.340 align:middle line:84% In fact, so for example, there are Italians who cannot roll 00:24:13.340 --> 00:24:15.320 align:middle line:90% their "r"s. 00:24:15.320 --> 00:24:16.880 align:middle line:90% But it's not all that common. 00:24:16.880 --> 00:24:19.550 align:middle line:84% Yeah, it's the kind of thing that is a speech impediment 00:24:19.550 --> 00:24:21.980 align:middle line:84% if your native language has rolled "r"s. 00:24:21.980 --> 00:24:23.660 align:middle line:84% If your native language is English then 00:24:23.660 --> 00:24:25.820 align:middle line:90% you're off the hook. 00:24:25.820 --> 00:24:28.760 align:middle line:84% There are also languages that have uvular trills. 00:24:28.760 --> 00:24:30.440 align:middle line:84% That's one possible pronunciation 00:24:30.440 --> 00:24:33.012 align:middle line:84% of the French "r'' or the German "r." 00:24:33.012 --> 00:24:35.150 align:middle line:90% [PURRING IN BACK OF MOUTH] 00:24:35.150 --> 00:24:37.220 align:middle line:84% I can only do a uvular trill with my head 00:24:37.220 --> 00:24:40.550 align:middle line:90% tilted back slightly. 00:24:40.550 --> 00:24:43.070 align:middle line:84% So yeah, that's, again, getting your uvula 00:24:43.070 --> 00:24:44.063 align:middle line:90% to move in the breeze. 00:24:44.063 --> 00:24:44.855 align:middle line:90% AUDIENCE: Gargling. 00:24:44.855 --> 00:24:45.500 align:middle line:90% NORVIN RICHARDS: I'm sorry. 00:24:45.500 --> 00:24:46.500 align:middle line:90% AUDIENCE: Like gargling. 00:24:46.500 --> 00:24:47.803 align:middle line:90% NORVIN RICHARDS: Like gargling. 00:24:47.803 --> 00:24:49.220 align:middle line:84% And then, there are languages that 00:24:49.220 --> 00:24:51.890 align:middle line:84% have bilabial trills, which was what Raquel was doing for us. 00:24:51.890 --> 00:24:55.540 align:middle line:84% Let me see if I can get these to come out. 00:24:55.540 --> 00:24:58.490 align:middle line:90% Ah, come on. 00:24:58.490 --> 00:24:58.990 align:middle line:90% All right. 00:24:58.990 --> 00:24:59.440 align:middle line:90% The UCLA website. 00:24:59.440 --> 00:25:00.107 align:middle line:90% [AUDIO PLAYBACK] 00:25:00.107 --> 00:25:00.190 align:middle line:90% 00:25:00.190 --> 00:25:00.400 align:middle line:90% [END PLAYBACK] 00:25:00.400 --> 00:25:00.850 align:middle line:90% NORVIN RICHARDS: There we go. 00:25:00.850 --> 00:25:01.517 align:middle line:90% [AUDIO PLAYBACK] 00:25:01.517 --> 00:25:01.537 align:middle line:90% 00:25:01.537 --> 00:25:02.120 align:middle line:90% [END PLAYBACK] 00:25:02.120 --> 00:25:02.950 align:middle line:84% NORVIN RICHARDS: So that's the sound that 00:25:02.950 --> 00:25:04.150 align:middle line:90% begins with a bilabial trill. 00:25:04.150 --> 00:25:04.633 align:middle line:90% [BLOWING] 00:25:04.633 --> 00:25:05.116 align:middle line:90% AUDIENCE: [BLOWING] 00:25:05.116 --> 00:25:06.760 align:middle line:84% NORVIN RICHARDS: [BLOWING] Yeah, you're better at it than me. 00:25:06.760 --> 00:25:07.427 align:middle line:90% [AUDIO PLAYBACK] 00:25:07.427 --> 00:25:07.927 align:middle line:90% 00:25:07.927 --> 00:25:08.573 align:middle line:90% [END PLAYBACK] 00:25:08.573 --> 00:25:10.240 align:middle line:84% NORVIN RICHARDS: Oh, here's another one. 00:25:10.240 --> 00:25:10.907 align:middle line:90% [AUDIO PLAYBACK] 00:25:10.907 --> 00:25:13.947 align:middle line:90% 00:25:13.947 --> 00:25:14.530 align:middle line:90% [END PLAYBACK] 00:25:14.530 --> 00:25:16.510 align:middle line:84% NORVIN RICHARDS: Yeah, so this is Kele, 00:25:16.510 --> 00:25:21.520 align:middle line:84% it's a language spoken in Gabon, and Titan, which is, I believe, 00:25:21.520 --> 00:25:22.903 align:middle line:90% a Polynesian language. 00:25:22.903 --> 00:25:24.820 align:middle line:84% These are languages that have bilabial trills. 00:25:24.820 --> 00:25:25.800 align:middle line:90% Is that a question? 00:25:25.800 --> 00:25:26.950 align:middle line:90% AUDIENCE: Question. 00:25:26.950 --> 00:25:31.960 align:middle line:84% Remind me why rolled "r"s are called "rolled 'r's" again. 00:25:31.960 --> 00:25:34.210 align:middle line:84% NORVIN RICHARDS: That's the-- so in linguistics, 00:25:34.210 --> 00:25:35.543 align:middle line:90% they aren't usually called that. 00:25:35.543 --> 00:25:37.510 align:middle line:90% They're called alveolar trills. 00:25:37.510 --> 00:25:39.040 align:middle line:84% We call them rolled "r"s, I think, 00:25:39.040 --> 00:25:41.890 align:middle line:84% because they sound a little bit like a drum roll. 00:25:41.890 --> 00:25:43.675 align:middle line:84% It's like there's a repetitive sound. 00:25:43.675 --> 00:25:46.540 align:middle line:84% [PURRING] You're repeatedly striking 00:25:46.540 --> 00:25:48.585 align:middle line:90% the roof of your mouth. 00:25:48.585 --> 00:25:50.710 align:middle line:84% That's what I always thought that term referred to. 00:25:50.710 --> 00:25:51.370 align:middle line:90% But as I say. 00:25:51.370 --> 00:25:52.972 align:middle line:90% It's not a linguistic term. 00:25:52.972 --> 00:25:55.930 align:middle line:84% AUDIENCE: Remind me why drums rolls [INAUDIBLE].. 00:25:55.930 --> 00:25:58.890 align:middle line:84% NORVIN RICHARDS: Oh, that's a good question. 00:25:58.890 --> 00:26:00.640 align:middle line:84% Well, because they sound like rolled "r"s. 00:26:00.640 --> 00:26:03.340 align:middle line:90% [LAUGHTER] 00:26:03.340 --> 00:26:05.560 align:middle line:84% I guess they sound like something is rolling 00:26:05.560 --> 00:26:09.465 align:middle line:84% and repeatedly-- so they involve a repeated strike of something. 00:26:09.465 --> 00:26:10.840 align:middle line:84% So if you have something rolling, 00:26:10.840 --> 00:26:14.020 align:middle line:84% bumping down a staircase or something, 00:26:14.020 --> 00:26:15.920 align:middle line:84% then it might sound a little bit like that. 00:26:15.920 --> 00:26:16.462 align:middle line:90% I don't know. 00:26:16.462 --> 00:26:17.860 align:middle line:90% I am now making things up. 00:26:17.860 --> 00:26:21.290 align:middle line:84% I will try to flag it when I'm making things up. 00:26:21.290 --> 00:26:23.262 align:middle line:90% But yes, I don't know. 00:26:23.262 --> 00:26:24.220 align:middle line:90% That's a good question. 00:26:24.220 --> 00:26:26.262 align:middle line:84% You should go look it up in the OED or something. 00:26:26.262 --> 00:26:27.960 align:middle line:90% Maybe they'll have a theory. 00:26:27.960 --> 00:26:29.620 align:middle line:90% Yeah, other questions? 00:26:29.620 --> 00:26:33.340 align:middle line:90% 00:26:33.340 --> 00:26:34.340 align:middle line:90% All right. 00:26:34.340 --> 00:26:37.850 align:middle line:84% So enough of the bilabial trills. 00:26:37.850 --> 00:26:41.180 align:middle line:90% 00:26:41.180 --> 00:26:45.950 align:middle line:84% And so yeah, as I say, I'll put the UCLA sound files up 00:26:45.950 --> 00:26:51.350 align:middle line:84% there so that you can play around with them. 00:26:51.350 --> 00:26:52.340 align:middle line:90% Oh, sorry. 00:26:52.340 --> 00:26:55.040 align:middle line:90% I spoke too soon. 00:26:55.040 --> 00:26:59.000 align:middle line:84% So we've been talking about sounds for a little while. 00:26:59.000 --> 00:27:04.310 align:middle line:84% And I've been encouraging you to try to make the strange sounds. 00:27:04.310 --> 00:27:06.890 align:middle line:84% You've all been clicking at me and trilling at me. 00:27:06.890 --> 00:27:08.430 align:middle line:90% It's been great. 00:27:08.430 --> 00:27:12.770 align:middle line:84% But there is another way to study speech sounds. 00:27:12.770 --> 00:27:15.860 align:middle line:84% And I just wanted to show you how it works. 00:27:15.860 --> 00:27:16.610 align:middle line:90% I hope this works. 00:27:16.610 --> 00:27:18.350 align:middle line:90% We'll see. 00:27:18.350 --> 00:27:22.220 align:middle line:84% It's possible to make what are called spectrograms. 00:27:22.220 --> 00:27:24.260 align:middle line:84% And maybe what I'll do is just show you one. 00:27:24.260 --> 00:27:27.185 align:middle line:84% And that'll make it easier to talk about what they are. 00:27:27.185 --> 00:27:28.685 align:middle line:84% Let's see if I can get this to work. 00:27:28.685 --> 00:27:39.236 align:middle line:90% 00:27:39.236 --> 00:27:39.736 align:middle line:90% You. 00:27:39.736 --> 00:27:43.480 align:middle line:90% 00:27:43.480 --> 00:27:46.720 align:middle line:90% Let's see if we can see that. 00:27:46.720 --> 00:27:47.320 align:middle line:90% There. 00:27:47.320 --> 00:27:50.680 align:middle line:90% That's a spectrogram. 00:27:50.680 --> 00:27:52.820 align:middle line:90% Let me Zoom in on it. 00:27:52.820 --> 00:27:55.260 align:middle line:90% Yeah, there we go. 00:27:55.260 --> 00:27:57.230 align:middle line:84% And I can zoom in on it a little further. 00:27:57.230 --> 00:27:58.550 align:middle line:90% It isn't all vowels, is it? 00:27:58.550 --> 00:28:01.090 align:middle line:90% 00:28:01.090 --> 00:28:04.330 align:middle line:84% So this is a program called Praat. 00:28:04.330 --> 00:28:05.410 align:middle line:90% It's a freeware program. 00:28:05.410 --> 00:28:07.300 align:middle line:90% I'll put it up online. 00:28:07.300 --> 00:28:09.730 align:middle line:84% It's something that allows you to make spectrograms. 00:28:09.730 --> 00:28:11.710 align:middle line:90% As you can see, it's very easy. 00:28:11.710 --> 00:28:13.660 align:middle line:84% It's a pretty easy program to use. 00:28:13.660 --> 00:28:16.000 align:middle line:84% I'll put up-- it's a program that's 00:28:16.000 --> 00:28:17.270 align:middle line:90% been around for a long time. 00:28:17.270 --> 00:28:20.110 align:middle line:84% It was developed by linguists and not by software engineers. 00:28:20.110 --> 00:28:21.190 align:middle line:90% And you can tell. 00:28:21.190 --> 00:28:23.608 align:middle line:90% So it's a little bit-- 00:28:23.608 --> 00:28:25.150 align:middle line:84% the interface is a little bit clunky. 00:28:25.150 --> 00:28:28.492 align:middle line:84% But I'll put up some suggestions about how to use it. 00:28:28.492 --> 00:28:30.950 align:middle line:84% And I encourage you to download it and play around with it. 00:28:30.950 --> 00:28:32.770 align:middle line:90% It's fun. 00:28:32.770 --> 00:28:35.920 align:middle line:84% What's going on here, the thing on the bottom of this window 00:28:35.920 --> 00:28:38.950 align:middle line:90% is the spectrogram. 00:28:38.950 --> 00:28:42.250 align:middle line:84% And it's a spectrogram showing me making those vowel sounds, 00:28:42.250 --> 00:28:46.180 align:middle line:84% "a," "i," and "u," transitioning between them. 00:28:46.180 --> 00:28:48.778 align:middle line:84% I think I might be able to play what we're looking at here. 00:28:48.778 --> 00:28:49.445 align:middle line:90% [AUDIO PLAYBACK] 00:28:49.445 --> 00:28:49.497 align:middle line:90% 00:28:49.497 --> 00:28:50.080 align:middle line:90% [END PLAYBACK] 00:28:50.080 --> 00:28:52.000 align:middle line:84% NORVIN RICHARDS: Yeah, so I didn't 00:28:52.000 --> 00:28:53.770 align:middle line:90% keep very much of the "u." 00:28:53.770 --> 00:28:57.400 align:middle line:84% That we can think just about the "a" and the "i." 00:28:57.400 --> 00:29:01.120 align:middle line:84% What you're doing when you are making vowels especially, 00:29:01.120 --> 00:29:03.190 align:middle line:84% but actually when you're speaking generally, 00:29:03.190 --> 00:29:05.870 align:middle line:84% is your vocal cords, in this case, 00:29:05.870 --> 00:29:07.930 align:middle line:90% my vocal chords were vibrating. 00:29:07.930 --> 00:29:10.720 align:middle line:84% They were making the air vibrate. 00:29:10.720 --> 00:29:12.490 align:middle line:84% There's this column of air that comes up 00:29:12.490 --> 00:29:14.710 align:middle line:90% through my vocal tract. 00:29:14.710 --> 00:29:16.160 align:middle line:90% They're making that air vibrate. 00:29:16.160 --> 00:29:19.150 align:middle line:84% And then, those vibrations are interacting with other things 00:29:19.150 --> 00:29:21.160 align:middle line:90% I'm doing with my vocal tract. 00:29:21.160 --> 00:29:24.230 align:middle line:84% They're bouncing off the sides of my vocal tract. 00:29:24.230 --> 00:29:26.350 align:middle line:84% And as I change the shape of my vocal tract, 00:29:26.350 --> 00:29:29.530 align:middle line:84% I change what those vibrations are doing. 00:29:29.530 --> 00:29:32.560 align:middle line:84% Yeah, it's a complicated phenomenon. 00:29:32.560 --> 00:29:36.490 align:middle line:84% You end up creating this complicated waveform, which 00:29:36.490 --> 00:29:40.570 align:middle line:84% you can do Fourier analysis of and analyze it 00:29:40.570 --> 00:29:44.320 align:middle line:84% as a waveform that has a bunch of regular sine 00:29:44.320 --> 00:29:48.220 align:middle line:84% waves of various frequencies and amplitudes 00:29:48.220 --> 00:29:49.480 align:middle line:90% that are contributing to it. 00:29:49.480 --> 00:29:52.240 align:middle line:84% Yeah, and this is what you do when you do a Fourier analysis, 00:29:52.240 --> 00:29:53.410 align:middle line:90% yeah. 00:29:53.410 --> 00:29:57.130 align:middle line:84% And what the spectrogram does is mark for you 00:29:57.130 --> 00:30:00.310 align:middle line:84% the frequencies that are making the greatest contribution 00:30:00.310 --> 00:30:04.840 align:middle line:84% to the general complicated air waveform. 00:30:04.840 --> 00:30:07.400 align:middle line:90% There are the darker parts. 00:30:07.400 --> 00:30:10.570 align:middle line:84% So these dark bands that you're seeing 00:30:10.570 --> 00:30:13.000 align:middle line:90% are what are called formants. 00:30:13.000 --> 00:30:16.000 align:middle line:84% And they are the things that are contributing 00:30:16.000 --> 00:30:18.670 align:middle line:90% the most to the sound. 00:30:18.670 --> 00:30:23.020 align:middle line:84% As you can see, they're shifting as I shift from "ah" to "ee." 00:30:23.020 --> 00:30:26.500 align:middle line:84% So this formant down-- no, let's see. 00:30:26.500 --> 00:30:31.240 align:middle line:90% Can we see my cursor on this? 00:30:31.240 --> 00:30:34.390 align:middle line:84% Yes, this formant right here, you 00:30:34.390 --> 00:30:36.710 align:middle line:84% can see as I shift from "ah" to "ee," it goes up. 00:30:36.710 --> 00:30:37.210 align:middle line:90% Yeah? 00:30:37.210 --> 00:30:39.745 align:middle line:84% And this formant up here, as I shift from "ah" to "ee," 00:30:39.745 --> 00:30:41.950 align:middle line:90% it dips. 00:30:41.950 --> 00:30:45.880 align:middle line:84% So how far apart the formants are from each other 00:30:45.880 --> 00:30:49.060 align:middle line:84% is what distinguishes the vowels from each other. 00:30:49.060 --> 00:30:51.850 align:middle line:84% You're hearing this difference in formants. 00:30:51.850 --> 00:30:53.710 align:middle line:90% You can do this for yourself. 00:30:53.710 --> 00:30:55.180 align:middle line:84% Here, I'm going to do another one. 00:30:55.180 --> 00:30:58.924 align:middle line:90% 00:30:58.924 --> 00:30:59.730 align:middle line:90% Hang on. 00:30:59.730 --> 00:31:05.486 align:middle line:90% 00:31:05.486 --> 00:31:05.986 align:middle line:90% "Wah." 00:31:05.986 --> 00:31:10.490 align:middle line:90% 00:31:10.490 --> 00:31:11.840 align:middle line:90% Yeah, so here's another one. 00:31:11.840 --> 00:31:14.860 align:middle line:90% 00:31:14.860 --> 00:31:17.800 align:middle line:84% I didn't get a very good recording of that one. 00:31:17.800 --> 00:31:18.580 align:middle line:90% Try that again. 00:31:18.580 --> 00:31:24.040 align:middle line:90% 00:31:24.040 --> 00:31:24.540 align:middle line:90% "Wah." 00:31:24.540 --> 00:31:33.870 align:middle line:90% 00:31:33.870 --> 00:31:34.370 align:middle line:90% Well. 00:31:34.370 --> 00:31:35.037 align:middle line:90% [AUDIO PLAYBACK] 00:31:35.037 --> 00:31:36.039 align:middle line:90% 00:31:36.039 --> 00:31:36.622 align:middle line:90% [END PLAYBACK] 00:31:36.622 --> 00:31:37.796 align:middle line:90% NORVIN RICHARDS: Hmm. 00:31:37.796 --> 00:31:38.296 align:middle line:90% Hmm. 00:31:38.296 --> 00:31:40.046 align:middle line:84% That's not doing what I'm expecting it to. 00:31:40.046 --> 00:31:40.570 align:middle line:90% Oh. 00:31:40.570 --> 00:31:44.668 align:middle line:84% Oh, I think I see what it's doing. 00:31:44.668 --> 00:31:47.210 align:middle line:84% It has to do with the fact that the microphone in my computer 00:31:47.210 --> 00:31:48.830 align:middle line:90% is lousy. 00:31:48.830 --> 00:31:51.523 align:middle line:84% You can see the formant shifting for this one. 00:31:51.523 --> 00:31:53.690 align:middle line:84% Actually, you can't see them shifting all that much. 00:31:53.690 --> 00:31:55.760 align:middle line:84% But you can hear a formant shifting. 00:31:55.760 --> 00:32:01.030 align:middle line:90% If you do the shift-- 00:32:01.030 --> 00:32:01.740 align:middle line:90% oh, no, wait. 00:32:01.740 --> 00:32:04.240 align:middle line:84% I think I can think of something that will show this better. 00:32:04.240 --> 00:32:05.032 align:middle line:90% Hang on just a sec. 00:32:05.032 --> 00:32:05.680 align:middle line:90% One more thing. 00:32:05.680 --> 00:32:13.980 align:middle line:90% 00:32:13.980 --> 00:32:14.490 align:middle line:90% Stop. 00:32:14.490 --> 00:32:14.790 align:middle line:90% Stop. 00:32:14.790 --> 00:32:15.290 align:middle line:90% Stop. 00:32:15.290 --> 00:32:17.310 align:middle line:90% Let's try "ee-oo." 00:32:17.310 --> 00:32:22.050 align:middle line:90% 00:32:22.050 --> 00:32:22.550 align:middle line:90% There we go. 00:32:22.550 --> 00:32:23.690 align:middle line:90% That's better. 00:32:23.690 --> 00:32:26.110 align:middle line:84% So here, you can clearly see a formant 00:32:26.110 --> 00:32:29.470 align:middle line:84% going from up here to down here as I transition 00:32:29.470 --> 00:32:30.558 align:middle line:90% from "ee" to "oo." 00:32:30.558 --> 00:32:31.600 align:middle line:90% Is everybody seeing that? 00:32:31.600 --> 00:32:32.720 align:middle line:90% I'll play it again. 00:32:32.720 --> 00:32:33.394 align:middle line:90% [AUDIO PLAYBACK] 00:32:33.394 --> 00:32:33.894 align:middle line:90% 00:32:33.894 --> 00:32:35.307 align:middle line:90% [END PLAYBACK] 00:32:35.307 --> 00:32:37.515 align:middle line:84% NORVIN RICHARDS: I sound like a dolphin or something. 00:32:37.515 --> 00:32:39.000 align:middle line:90% [AUDIO PLAYBACK] 00:32:39.000 --> 00:32:39.467 align:middle line:90% 00:32:39.467 --> 00:32:40.050 align:middle line:90% [END PLAYBACK] 00:32:40.050 --> 00:32:42.090 align:middle line:84% NORVIN RICHARDS: Yeah, but as I switch 00:32:42.090 --> 00:32:46.005 align:middle line:84% from the "ee" to the "oo," you can see a formant descending, 00:32:46.005 --> 00:32:48.030 align:middle line:90% yeah. 00:32:48.030 --> 00:32:49.650 align:middle line:90% And you can hear that. 00:32:49.650 --> 00:32:51.177 align:middle line:90% If you do the-- 00:32:51.177 --> 00:32:53.010 align:middle line:84% so this is something that we were doing back 00:32:53.010 --> 00:32:54.385 align:middle line:84% when I was trying to convince you 00:32:54.385 --> 00:32:56.880 align:middle line:84% that vowels were worth categorizing, that the switch 00:32:56.880 --> 00:33:00.180 align:middle line:84% from "ee" to "oo" involves your tongue going from being front 00:33:00.180 --> 00:33:03.330 align:middle line:84% to being back, and also your lips going from being not 00:33:03.330 --> 00:33:05.190 align:middle line:90% rounded to being rounded, yeah. 00:33:05.190 --> 00:33:15.820 align:middle line:84% And if you whisper that to yourself, so go [SLURPING],, 00:33:15.820 --> 00:33:19.730 align:middle line:84% you might be hearing something descending in pitch. 00:33:19.730 --> 00:33:20.750 align:middle line:90% Are you hearing that? 00:33:20.750 --> 00:33:22.710 align:middle line:90% Just go [SLURPING]. 00:33:22.710 --> 00:33:27.068 align:middle line:90% 00:33:27.068 --> 00:33:28.360 align:middle line:90% And that's what you're hearing. 00:33:28.360 --> 00:33:29.560 align:middle line:90% It's that formant. 00:33:29.560 --> 00:33:32.590 align:middle line:84% Yeah, by whispering, you're getting 00:33:32.590 --> 00:33:33.610 align:middle line:90% that to leap out at you. 00:33:33.610 --> 00:33:36.220 align:middle line:84% And you can hear it descending in pitch. 00:33:36.220 --> 00:33:37.960 align:middle line:90% Why is it descending in pitch? 00:33:37.960 --> 00:33:39.530 align:middle line:84% If you think about what you're doing, 00:33:39.530 --> 00:33:42.160 align:middle line:84% think about the inside of your vocal tract. 00:33:42.160 --> 00:33:48.820 align:middle line:84% As you go from "ee" to "oo," let's talk first about "oo." 00:33:48.820 --> 00:33:54.280 align:middle line:84% So "ooh" ["u"] is a high, back rounded vowel. 00:33:54.280 --> 00:33:55.535 align:middle line:90% Was any of that surprising? 00:33:55.535 --> 00:33:56.035 align:middle line:90% Yeah? 00:33:56.035 --> 00:33:57.220 align:middle line:90% "Oo." 00:33:57.220 --> 00:33:59.980 align:middle line:84% So your tongue is high and bunched up 00:33:59.980 --> 00:34:01.300 align:middle line:90% in the back of your mouth. 00:34:01.300 --> 00:34:02.710 align:middle line:90% And your lips are rounded. 00:34:02.710 --> 00:34:03.790 align:middle line:90% Ooh. 00:34:03.790 --> 00:34:06.400 align:middle line:84% Yeah, that means that so what you're 00:34:06.400 --> 00:34:08.170 align:middle line:84% doing when you're making vowels is 00:34:08.170 --> 00:34:11.920 align:middle line:84% you're taking this column there that's in your vocal tract. 00:34:11.920 --> 00:34:14.380 align:middle line:84% And you're, depending on the vowel, 00:34:14.380 --> 00:34:18.639 align:middle line:84% making a partition somewhere that isn't complete. 00:34:18.639 --> 00:34:20.500 align:middle line:84% You're not blocking off part of the tract. 00:34:20.500 --> 00:34:22.630 align:middle line:84% But you're putting your tongue in the way. 00:34:22.630 --> 00:34:26.650 align:middle line:84% So you're making part of that column of air narrower. 00:34:26.650 --> 00:34:28.310 align:middle line:90% Does that make sense? 00:34:28.310 --> 00:34:30.699 align:middle line:84% So for a high back rounded vowel, 00:34:30.699 --> 00:34:33.790 align:middle line:84% by bunching your tongue up toward the back of your mouth, 00:34:33.790 --> 00:34:37.840 align:middle line:84% you're making not a complete obstruction of the flow of air. 00:34:37.840 --> 00:34:40.120 align:middle line:84% There's still space for the air to flow past it. 00:34:40.120 --> 00:34:43.100 align:middle line:84% But you're making it narrower back there, yeah. 00:34:43.100 --> 00:34:46.900 align:middle line:84% And when you make an "oo" by making a high back gesture 00:34:46.900 --> 00:34:49.449 align:middle line:84% with your tongue and by rounding your lips, 00:34:49.449 --> 00:34:55.960 align:middle line:84% you're creating a compartment for your vocal tract, which 00:34:55.960 --> 00:34:58.330 align:middle line:90% is as large as you can make it. 00:34:58.330 --> 00:35:00.760 align:middle line:84% Your lips are out there in front. 00:35:00.760 --> 00:35:03.340 align:middle line:84% And your tongue is as far back as it goes. 00:35:03.340 --> 00:35:05.710 align:middle line:90% Yeah, does that make sense? 00:35:05.710 --> 00:35:09.370 align:middle line:84% You're compartmentalizing off that top space 00:35:09.370 --> 00:35:13.240 align:middle line:90% of your vocal tract. 00:35:13.240 --> 00:35:16.180 align:middle line:84% And you're making that space as large as it can be. 00:35:16.180 --> 00:35:19.630 align:middle line:84% As opposed to "ee" ["i"], where your lips are not rounded 00:35:19.630 --> 00:35:21.880 align:middle line:84% and your tongue is toward the front of your mouth. 00:35:21.880 --> 00:35:24.700 align:middle line:84% So that makes that space that you created for ooh. 00:35:24.700 --> 00:35:26.530 align:middle line:90% That makes that space collapse. 00:35:26.530 --> 00:35:29.110 align:middle line:84% It makes that space as small as it can possibly 00:35:29.110 --> 00:35:30.968 align:middle line:84% be because now your tongue is shoved up 00:35:30.968 --> 00:35:32.260 align:middle line:90% toward the front of your mouth. 00:35:32.260 --> 00:35:33.760 align:middle line:90% And your lips are not rounded. 00:35:33.760 --> 00:35:35.620 align:middle line:84% The space between your tongue and your lips 00:35:35.620 --> 00:35:37.075 align:middle line:90% is now as small as it gets. 00:35:37.075 --> 00:35:38.200 align:middle line:90% Yeah? 00:35:38.200 --> 00:35:42.460 align:middle line:84% And that's why the transition from "ee" to "ooh" 00:35:42.460 --> 00:35:45.040 align:middle line:84% involves something descending in pitch. 00:35:45.040 --> 00:35:47.710 align:middle line:84% It's because you're going from a very small compartment 00:35:47.710 --> 00:35:49.810 align:middle line:84% with a little bit of vibration inside it 00:35:49.810 --> 00:35:53.590 align:middle line:84% to a large compartment with the same amount of vibration 00:35:53.590 --> 00:35:55.270 align:middle line:84% inside it because your vocal cords are 00:35:55.270 --> 00:35:57.160 align:middle line:90% vibrating at a constant rate. 00:35:57.160 --> 00:35:59.230 align:middle line:84% We're not-- you're not fooling around with pitch. 00:35:59.230 --> 00:36:00.320 align:middle line:90% Does that make sense? 00:36:00.320 --> 00:36:04.320 align:middle line:84% So as you switch from a small compartment to a large one, 00:36:04.320 --> 00:36:07.390 align:middle line:84% one component of the very complicated things that make up 00:36:07.390 --> 00:36:10.780 align:middle line:84% this complicated waveform is descending in pitch 00:36:10.780 --> 00:36:13.480 align:middle line:84% because you're creating this little space that's 00:36:13.480 --> 00:36:16.210 align:middle line:84% getting larger, and by putting the same amount of energy 00:36:16.210 --> 00:36:20.020 align:middle line:84% to it, the molecules of air are not moving as fast. 00:36:20.020 --> 00:36:22.150 align:middle line:84% That's what we hear as a lower pitch. 00:36:22.150 --> 00:36:24.050 align:middle line:90% Does that make sense? 00:36:24.050 --> 00:36:26.440 align:middle line:84% It's a way of talking about this, anyway. 00:36:26.440 --> 00:36:30.700 align:middle line:84% So encourage you to fool around with Praat. 00:36:30.700 --> 00:36:33.100 align:middle line:84% Let me just show you some consonants 00:36:33.100 --> 00:36:38.770 align:middle line:84% so that you can see what else you can do with Praat. 00:36:38.770 --> 00:36:40.015 align:middle line:90% We'll try another one here. 00:36:40.015 --> 00:36:46.800 align:middle line:90% 00:36:46.800 --> 00:36:47.460 align:middle line:90% "Bab. 00:36:47.460 --> 00:36:48.190 align:middle line:90% Dad. 00:36:48.190 --> 00:36:48.690 align:middle line:90% Gag." 00:36:48.690 --> 00:36:55.370 align:middle line:90% 00:36:55.370 --> 00:36:56.270 align:middle line:90% View that one. 00:36:56.270 --> 00:36:59.120 align:middle line:90% [AUDIO PLAYBACK] 00:36:59.120 --> 00:37:03.590 align:middle line:90% 00:37:03.590 --> 00:37:04.610 align:middle line:90% [END PLAYBACK] 00:37:04.610 --> 00:37:07.235 align:middle line:84% NORVIN RICHARDS: So here, I got the same vowel all three times. 00:37:07.235 --> 00:37:10.652 align:middle line:84% And you can see these nice clear bars in the middle. 00:37:10.652 --> 00:37:12.860 align:middle line:84% The difference has to do with what the consonants are 00:37:12.860 --> 00:37:15.770 align:middle line:90% on either side of the vowel. 00:37:15.770 --> 00:37:21.030 align:middle line:84% One of the things that this representation, 00:37:21.030 --> 00:37:23.780 align:middle line:84% this spectrogram, allows us to concentrate on, 00:37:23.780 --> 00:37:27.050 align:middle line:84% if you think about it, how do you 00:37:27.050 --> 00:37:32.120 align:middle line:84% tell the difference between "bab" and "gag"? 00:37:32.120 --> 00:37:36.740 align:middle line:84% So we know articulatorially what the difference is. 00:37:36.740 --> 00:37:38.900 align:middle line:84% The difference is that, for "bab," you're 00:37:38.900 --> 00:37:40.550 align:middle line:90% making a bilabial stop. 00:37:40.550 --> 00:37:43.970 align:middle line:84% And for "gag," you're making a velar stop. 00:37:43.970 --> 00:37:45.500 align:middle line:84% But when you're hearing people make 00:37:45.500 --> 00:37:48.710 align:middle line:84% those sounds, how are you hearing what the difference is 00:37:48.710 --> 00:37:50.090 align:middle line:90% between a B and a G? 00:37:50.090 --> 00:37:52.850 align:middle line:84% Bear in mind that B and G are both stops. 00:37:52.850 --> 00:37:55.250 align:middle line:84% The flow of air is stopped, yeah, 00:37:55.250 --> 00:37:57.480 align:middle line:90% while you're making the sound. 00:37:57.480 --> 00:37:59.450 align:middle line:84% So while you're making the sound, 00:37:59.450 --> 00:38:01.820 align:middle line:90% you're not hearing things. 00:38:01.820 --> 00:38:04.050 align:middle line:90% Yeah? 00:38:04.050 --> 00:38:07.050 align:middle line:84% The way you distinguish "bab" from "gag" 00:38:07.050 --> 00:38:11.850 align:middle line:84% is by looking at what it does to the vowel that's in between. 00:38:11.850 --> 00:38:15.660 align:middle line:84% You can see that these formants for the "ah" vowel, 00:38:15.660 --> 00:38:18.343 align:middle line:84% and remember that the vowel was the same in all three cases, 00:38:18.343 --> 00:38:20.010 align:middle line:84% it's especially clear for this last one. 00:38:20.010 --> 00:38:21.210 align:middle line:90% Here's "gag." 00:38:21.210 --> 00:38:24.630 align:middle line:84% And you can see that the formants, this format 00:38:24.630 --> 00:38:26.910 align:middle line:90% is smile-shaped. 00:38:26.910 --> 00:38:28.620 align:middle line:84% And this formant is frown-shaped. 00:38:28.620 --> 00:38:33.180 align:middle line:84% This one goes down as it goes close to the stop. 00:38:33.180 --> 00:38:35.610 align:middle line:84% And this one goes up as it goes close to the top. 00:38:35.610 --> 00:38:39.030 align:middle line:84% As opposed to the bilabial, which is, if anything, 00:38:39.030 --> 00:38:42.750 align:middle line:90% slightly down on both sides. 00:38:42.750 --> 00:38:45.860 align:middle line:84% This is one of the things that makes speech perception so hard 00:38:45.860 --> 00:38:50.330 align:middle line:84% and that makes it so hard to get machines to understand speech. 00:38:50.330 --> 00:38:52.760 align:middle line:84% It's a problem that's more or less solved. 00:38:52.760 --> 00:38:57.770 align:middle line:84% But it's a hard problem because your speech is not just 00:38:57.770 --> 00:39:00.410 align:middle line:90% a sequence of sounds. 00:39:00.410 --> 00:39:04.580 align:middle line:84% It's a sequence of sounds that all influence each other. 00:39:04.580 --> 00:39:07.580 align:middle line:84% It's a mess in other words, yeah? 00:39:07.580 --> 00:39:11.720 align:middle line:84% And the best cues for what sound you made at a particular point 00:39:11.720 --> 00:39:13.610 align:middle line:84% might not be at that point at all. 00:39:13.610 --> 00:39:16.020 align:middle line:84% They might be on neighboring points. 00:39:16.020 --> 00:39:19.290 align:middle line:84% That's how you find out what sound you're looking at. 00:39:19.290 --> 00:39:21.470 align:middle line:84% And that's what makes speech perception hard. 00:39:21.470 --> 00:39:25.430 align:middle line:90% It's also part of why-- 00:39:25.430 --> 00:39:29.510 align:middle line:84% so there are generalizations about the kinds of syllables 00:39:29.510 --> 00:39:32.570 align:middle line:90% that we find in the world. 00:39:32.570 --> 00:39:35.300 align:middle line:84% There are plenty of languages out there 00:39:35.300 --> 00:39:40.230 align:middle line:84% in which it's OK for a syllable to start with a stop. 00:39:40.230 --> 00:39:43.250 align:middle line:84% Let's say you can have a syllable like "pa." 00:39:43.250 --> 00:39:44.750 align:middle line:84% That's an English word more or less. 00:39:44.750 --> 00:39:48.500 align:middle line:84% It's a word for father, "pa," but in which 00:39:48.500 --> 00:39:53.780 align:middle line:84% it's not OK to have a P followed by a fricative and then 00:39:53.780 --> 00:39:54.770 align:middle line:90% a vowel. 00:39:54.770 --> 00:39:57.950 align:middle line:84% So you don't have syllables like "psah," in English, 00:39:57.950 --> 00:40:00.873 align:middle line:90% or "pfah," yeah, or "pkah." 00:40:00.873 --> 00:40:02.540 align:middle line:84% You can imagine that there are languages 00:40:02.540 --> 00:40:03.930 align:middle line:90% that allow syllables like that. 00:40:03.930 --> 00:40:06.980 align:middle line:84% But if a language is going to restrict its syllables, 00:40:06.980 --> 00:40:08.960 align:middle line:84% if it's only going to have one of these kinds, 00:40:08.960 --> 00:40:10.670 align:middle line:84% it's only going to have this kind. 00:40:10.670 --> 00:40:12.620 align:middle line:84% Or to put it another way, there are languages 00:40:12.620 --> 00:40:13.700 align:middle line:90% that have both of these. 00:40:13.700 --> 00:40:15.830 align:middle line:84% And there are languages that only have this. 00:40:15.830 --> 00:40:19.370 align:middle line:84% There aren't any languages that only have this, yeah? 00:40:19.370 --> 00:40:21.080 align:middle line:90% And we can see why. 00:40:21.080 --> 00:40:24.890 align:middle line:84% The crucial cue to what, kind of, stop you're hearing, 00:40:24.890 --> 00:40:27.420 align:middle line:90% it comes from the vowel. 00:40:27.420 --> 00:40:31.160 align:middle line:84% So if you put a fricative in between the stop and the vowel, 00:40:31.160 --> 00:40:34.250 align:middle line:84% you deprive yourself of one of the big cues for figuring out 00:40:34.250 --> 00:40:36.000 align:middle line:84% what, kind of, stop you're listening to. 00:40:36.000 --> 00:40:39.800 align:middle line:84% And so it makes sense that we get this type 00:40:39.800 --> 00:40:40.765 align:middle line:90% of syllable everywhere. 00:40:40.765 --> 00:40:42.140 align:middle line:84% And the other type of syllable is 00:40:42.140 --> 00:40:43.940 align:middle line:84% only-- has a higher degree of difficulty, 00:40:43.940 --> 00:40:45.810 align:middle line:84% basically, as far as perception goes. 00:40:45.810 --> 00:40:47.210 align:middle line:90% Yeah, Joseph. 00:40:47.210 --> 00:40:51.003 align:middle line:84% AUDIENCE: What effect does that have on [INAUDIBLE] especially. 00:40:51.003 --> 00:40:51.920 align:middle line:90% NORVIN RICHARDS: Yeah. 00:40:51.920 --> 00:40:55.130 align:middle line:84% AUDIENCE: For one example, where it's perfectly OK 00:40:55.130 --> 00:40:56.422 align:middle line:90% to string three stops in a row. 00:40:56.422 --> 00:40:57.297 align:middle line:90% NORVIN RICHARDS: Yep. 00:40:57.297 --> 00:40:58.490 align:middle line:90% AUDIENCE: Or at least two. 00:40:58.490 --> 00:40:59.407 align:middle line:90% NORVIN RICHARDS: Yeah. 00:40:59.407 --> 00:41:01.522 align:middle line:84% AUDIENCE: [INAUDIBLE] then you can just-- 00:41:01.522 --> 00:41:02.480 align:middle line:90% NORVIN RICHARDS: Right. 00:41:02.480 --> 00:41:05.450 align:middle line:84% So I've shown you the easiest way to figure out 00:41:05.450 --> 00:41:06.980 align:middle line:84% what a stop is, which is by looking 00:41:06.980 --> 00:41:08.450 align:middle line:84% at its effect on the nearby vowel. 00:41:08.450 --> 00:41:09.470 align:middle line:90% But you're absolutely right. 00:41:09.470 --> 00:41:10.887 align:middle line:84% There are languages out there that 00:41:10.887 --> 00:41:14.360 align:middle line:84% have strings of stops, strings of stops and fricatives. 00:41:14.360 --> 00:41:15.620 align:middle line:90% It absolutely happens. 00:41:15.620 --> 00:41:18.815 align:middle line:84% Russian has words like [RUSSIAN] it's 00:41:18.815 --> 00:41:20.990 align:middle line:84% got of a whole string of stops and fricatives 00:41:20.990 --> 00:41:25.490 align:middle line:84% at the beginning of a word, a beginning of a syllable. 00:41:25.490 --> 00:41:29.318 align:middle line:84% The most famous examples of this are languages-- 00:41:29.318 --> 00:41:30.860 align:middle line:84% there are two kinds of languages that 00:41:30.860 --> 00:41:32.300 align:middle line:90% are especially famous for this. 00:41:32.300 --> 00:41:37.160 align:middle line:84% There's a Salish language called Bella Coola. 00:41:37.160 --> 00:41:40.130 align:middle line:84% The Salish languages are spoken in the Pacific Northwest. 00:41:40.130 --> 00:41:42.290 align:middle line:84% Bella Coola, there's only one word of Bella Coola 00:41:42.290 --> 00:41:43.340 align:middle line:90% that I know. 00:41:43.340 --> 00:41:44.360 align:middle line:90% And it's this. 00:41:44.360 --> 00:41:48.680 align:middle line:84% So this is a uvular fricative, a voiceless uvular fricative. 00:41:48.680 --> 00:41:52.070 align:middle line:84% Anybody want to try to pronounce that? 00:41:52.070 --> 00:41:53.834 align:middle line:90% It's something like sauce. 00:41:53.834 --> 00:41:54.758 align:middle line:90% [LAUGHTER] 00:41:54.758 --> 00:41:57.050 align:middle line:84% It's the only thing I know how to say in Bella Coola. 00:41:57.050 --> 00:41:58.640 align:middle line:84% If I ever meet a Bella Coola, we'll 00:41:58.640 --> 00:42:01.310 align:middle line:90% have a very simple conversation. 00:42:01.310 --> 00:42:03.680 align:middle line:90% It means seal fat. 00:42:03.680 --> 00:42:07.670 align:middle line:84% And so here's a place where you hope 00:42:07.670 --> 00:42:10.220 align:middle line:84% there are other cues besides effects on vowels because there 00:42:10.220 --> 00:42:12.920 align:middle line:90% are no vowels in this word. 00:42:12.920 --> 00:42:16.220 align:middle line:84% Or the Berber languages of Northern Africa 00:42:16.220 --> 00:42:18.330 align:middle line:84% are famous for having sounds like this. 00:42:18.330 --> 00:42:21.860 align:middle line:84% Here's the only word I know, the only word 00:42:21.860 --> 00:42:25.400 align:middle line:84% I know in [INAUDIBLE] Berber, which is spoken in Morocco. 00:42:25.400 --> 00:42:29.300 align:middle line:84% It means "you (feminine) sprained it." 00:42:29.300 --> 00:42:31.305 align:middle line:84% And it's something like [CLICKING] 00:42:31.305 --> 00:42:32.930 align:middle line:84% Phonologists just get really interested 00:42:32.930 --> 00:42:34.620 align:middle line:84% in these languages for lots of reasons. 00:42:34.620 --> 00:42:36.110 align:middle line:84% One is the kind of thing Joseph is asking about, 00:42:36.110 --> 00:42:38.690 align:middle line:84% how do these people figure out what sounds they are hearing? 00:42:38.690 --> 00:42:41.930 align:middle line:84% And the answer has to be something like, yeah, there are 00:42:41.930 --> 00:42:43.550 align:middle line:90% lots of good cues on vowels. 00:42:43.550 --> 00:42:46.010 align:middle line:84% But there are also cues on other kinds of things. 00:42:46.010 --> 00:42:48.230 align:middle line:84% And we have to figure out what they are, how-- 00:42:48.230 --> 00:42:51.650 align:middle line:90% what people are cueing in on. 00:42:51.650 --> 00:42:54.392 align:middle line:84% There are other questions too, like I was just talking about. 00:42:54.392 --> 00:42:55.850 align:middle line:84% Syllables, which is something we'll 00:42:55.850 --> 00:42:57.800 align:middle line:84% talk about as we get further into phonology. 00:42:57.800 --> 00:43:00.320 align:middle line:84% And the question is, are there any syllables in this word? 00:43:00.320 --> 00:43:00.820 align:middle line:90% 00:43:00.820 --> 00:43:03.590 align:middle line:84% To which, the answer turns out to be yes. 00:43:03.590 --> 00:43:06.290 align:middle line:84% So if you look carefully at these languages, 00:43:06.290 --> 00:43:10.370 align:middle line:84% you can convince yourself that they have syllables, yeah. 00:43:10.370 --> 00:43:12.290 align:middle line:84% Sorry, very long answer to Joseph's question, 00:43:12.290 --> 00:43:13.957 align:middle line:84% to which the short version of the answer 00:43:13.957 --> 00:43:15.767 align:middle line:84% would be the best cues are on vowels. 00:43:15.767 --> 00:43:18.350 align:middle line:84% But there must be cues on other kinds of things as well, which 00:43:18.350 --> 00:43:19.865 align:middle line:90% are not as clear. 00:43:19.865 --> 00:43:21.890 align:middle line:90% All right? 00:43:21.890 --> 00:43:24.890 align:middle line:84% And it's a question of your level of tolerance 00:43:24.890 --> 00:43:27.860 align:middle line:84% for unclarity, yeah, or how good you are 00:43:27.860 --> 00:43:31.290 align:middle line:90% at picking up on those cues. 00:43:31.290 --> 00:43:31.790 align:middle line:90% Cool. 00:43:31.790 --> 00:43:35.618 align:middle line:84% So Praat, was i going to do anything else with Praat? 00:43:35.618 --> 00:43:37.910 align:middle line:84% No, those were the things I was going to do with Praat. 00:43:37.910 --> 00:43:39.670 align:middle line:90% I will leave Praat up. 00:43:39.670 --> 00:43:41.420 align:middle line:84% I think-- actually, we may be seeing Praat 00:43:41.420 --> 00:43:42.780 align:middle line:90% again pretty shortly. 00:43:42.780 --> 00:43:44.960 align:middle line:84% But I'll leave Praat on the website 00:43:44.960 --> 00:43:48.594 align:middle line:84% so you can play around with it at home. 00:43:48.594 --> 00:43:51.810 align:middle line:90% Going to full screen mode. 00:43:51.810 --> 00:43:52.500 align:middle line:90% Oh, sorry. 00:43:52.500 --> 00:43:54.330 align:middle line:90% Go back out of screen mode. 00:43:54.330 --> 00:43:56.316 align:middle line:84% So we've just been talking about how 00:43:56.316 --> 00:44:01.200 align:middle line:84% you know what sounds it is that you are hearing. 00:44:01.200 --> 00:44:03.315 align:middle line:84% And some of the cues come from what's 00:44:03.315 --> 00:44:04.440 align:middle line:90% happening during the sound. 00:44:04.440 --> 00:44:06.900 align:middle line:84% And some of the cues come from neighboring sounds. 00:44:06.900 --> 00:44:10.440 align:middle line:84% And the study of this is one of the central studies 00:44:10.440 --> 00:44:12.390 align:middle line:90% of phonology and of phonetics. 00:44:12.390 --> 00:44:14.640 align:middle line:84% I want to show you about another, kind of, information 00:44:14.640 --> 00:44:15.630 align:middle line:90% that we have, though. 00:44:15.630 --> 00:44:18.670 align:middle line:84% It's something called the McGurk effect. 00:44:18.670 --> 00:44:20.170 align:middle line:84% And I'll just demonstrate it to you. 00:44:20.170 --> 00:44:21.420 align:middle line:90% And then, we'll talk about it. 00:44:21.420 --> 00:44:25.960 align:middle line:84% This guy is going to repeat a syllable a couple of times. 00:44:25.960 --> 00:44:28.780 align:middle line:84% And then, when he's done, I'm going to ask you 00:44:28.780 --> 00:44:32.200 align:middle line:90% what syllable you heard. 00:44:32.200 --> 00:44:33.370 align:middle line:90% So here's the guy. 00:44:33.370 --> 00:44:35.617 align:middle line:90% [AUDIO PLAYBACK] 00:44:35.617 --> 00:44:36.117 align:middle line:90% - Bah. 00:44:36.117 --> 00:44:37.108 align:middle line:90% Bah. 00:44:37.108 --> 00:44:37.608 align:middle line:90% Bah. 00:44:37.608 --> 00:44:38.600 align:middle line:90% Bah. 00:44:38.600 --> 00:44:39.100 align:middle line:90% Bah. 00:44:39.100 --> 00:44:40.647 align:middle line:90% Bah. 00:44:40.647 --> 00:44:41.230 align:middle line:90% [END PLAYBACK] 00:44:41.230 --> 00:44:43.230 align:middle line:84% NORVIN RICHARDS: I'm going to turn up the volume 00:44:43.230 --> 00:44:45.292 align:middle line:84% and do that again so you can hear it better. 00:44:45.292 --> 00:44:47.500 align:middle line:90% [AUDIO PLAYBACK] 00:44:47.500 --> 00:44:48.000 align:middle line:90% - Bah. 00:44:48.000 --> 00:44:49.000 align:middle line:90% Bah. 00:44:49.000 --> 00:44:49.500 align:middle line:90% Bah. 00:44:49.500 --> 00:44:50.000 align:middle line:90% Bah. 00:44:50.000 --> 00:44:50.500 align:middle line:90% Bah. 00:44:50.500 --> 00:44:52.471 align:middle line:90% Bah. 00:44:52.471 --> 00:44:52.971 align:middle line:90% Bah. 00:44:52.971 --> 00:44:54.355 align:middle line:90% Bah. 00:44:54.355 --> 00:44:54.855 align:middle line:90% Bah. 00:44:54.855 --> 00:44:55.768 align:middle line:90% Bah. 00:44:55.768 --> 00:44:56.268 align:middle line:90% Bah. 00:44:56.268 --> 00:44:57.567 align:middle line:90% Bah. 00:44:57.567 --> 00:44:58.150 align:middle line:90% [END PLAYBACK] 00:44:58.150 --> 00:45:03.190 align:middle line:84% NORVIN RICHARDS: How many of you heard a "bah"? 00:45:03.190 --> 00:45:06.610 align:middle line:90% How many of you heard a "da"? 00:45:06.610 --> 00:45:09.695 align:middle line:90% How many of you heard a "gah"? 00:45:09.695 --> 00:45:11.820 align:middle line:84% Several of you are raising your hand several times. 00:45:11.820 --> 00:45:12.320 align:middle line:90% [LAUGHTER] 00:45:12.320 --> 00:45:13.820 align:middle line:84% You're just not sure what you heard? 00:45:13.820 --> 00:45:15.840 align:middle line:84% Or you thought you heard several things at once? 00:45:15.840 --> 00:45:19.150 align:middle line:84% Yeah, you thought you heard several things at once. 00:45:19.150 --> 00:45:21.330 align:middle line:84% So there's something very confusing going on. 00:45:21.330 --> 00:45:22.840 align:middle line:90% I'm going to play it again. 00:45:22.840 --> 00:45:24.420 align:middle line:84% And I want you to close your eyes. 00:45:24.420 --> 00:45:26.242 align:middle line:90% AUDIENCE: [INTERPOSING VOICES] 00:45:26.242 --> 00:45:28.242 align:middle line:84% NORVIN RICHARDS: Everybody got your eyes closed? 00:45:28.242 --> 00:45:31.116 align:middle line:90% [AUDIO PLAYBACK] 00:45:31.116 --> 00:45:31.616 align:middle line:90% - Bah. 00:45:31.616 --> 00:45:32.544 align:middle line:90% Bah. 00:45:32.544 --> 00:45:33.044 align:middle line:90% Bah. 00:45:33.044 --> 00:45:33.972 align:middle line:90% Bah. 00:45:33.972 --> 00:45:34.472 align:middle line:90% Bah. 00:45:34.472 --> 00:45:35.682 align:middle line:90% Bah. 00:45:35.682 --> 00:45:37.173 align:middle line:90% [END PLAYBACK] 00:45:37.173 --> 00:45:39.090 align:middle line:84% NORVIN RICHARDS: I'll put the link so that you 00:45:39.090 --> 00:45:40.590 align:middle line:90% can play with this at home. 00:45:40.590 --> 00:45:42.070 align:middle line:90% But here's what's happening. 00:45:42.070 --> 00:45:44.340 align:middle line:84% You're hearing a recording of "bah." 00:45:44.340 --> 00:45:45.360 align:middle line:90% That's what he's doing. 00:45:45.360 --> 00:45:46.680 align:middle line:90% He's saying "bah." 00:45:46.680 --> 00:45:49.380 align:middle line:84% But the "bah" is being carefully timed together 00:45:49.380 --> 00:45:51.065 align:middle line:84% with a video of him saying "gah". 00:45:51.065 --> 00:45:52.560 align:middle line:90% [LAUGHTER] 00:45:52.560 --> 00:45:55.410 align:middle line:84% And so you can only tell that it's "bah" 00:45:55.410 --> 00:45:56.655 align:middle line:90% if you don't look at him. 00:45:56.655 --> 00:45:57.720 align:middle line:90% [LAUGHTER] 00:45:57.720 --> 00:46:01.140 align:middle line:84% So if your eyes are closed, you hear it as "bah," yeah? 00:46:01.140 --> 00:46:03.870 align:middle line:84% But if your eyes are not closed, if you look at him-- 00:46:03.870 --> 00:46:05.760 align:middle line:84% you should try this at home maybe if you're 00:46:05.760 --> 00:46:07.020 align:middle line:90% on a better quality screen. 00:46:07.020 --> 00:46:09.420 align:middle line:84% The illusion becomes fairly clear. 00:46:09.420 --> 00:46:11.070 align:middle line:84% People generally report that they're 00:46:11.070 --> 00:46:15.150 align:middle line:84% hearing "dah," which is neither what they are seeing 00:46:15.150 --> 00:46:18.100 align:middle line:90% or what they are hearing, yeah? 00:46:18.100 --> 00:46:20.400 align:middle line:84% So what this is teaching us is that although there's 00:46:20.400 --> 00:46:22.500 align:middle line:84% all this complicated phonology going on, 00:46:22.500 --> 00:46:25.140 align:middle line:84% all of these things that you do to try to figure out the speech 00:46:25.140 --> 00:46:28.470 align:middle line:84% signal, you are also paying attention to what you see. 00:46:28.470 --> 00:46:30.010 align:middle line:84% So apparently, what's going on here, 00:46:30.010 --> 00:46:32.250 align:middle line:84% again, the recording is of "bah." 00:46:32.250 --> 00:46:33.900 align:middle line:90% But you're looking at this guy. 00:46:33.900 --> 00:46:37.200 align:middle line:84% And he's clearly not saying "bah" because his lips are not 00:46:37.200 --> 00:46:38.010 align:middle line:90% closing. 00:46:38.010 --> 00:46:41.070 align:middle line:84% And apparently, the evidence of your eyes 00:46:41.070 --> 00:46:44.883 align:middle line:84% is enough to convince your brain to ignore your ears when 00:46:44.883 --> 00:46:47.050 align:middle line:84% it's trying to figure out what you're hearing, yeah. 00:46:47.050 --> 00:46:50.310 align:middle line:84% So the cause of the illusion is that because you 00:46:50.310 --> 00:46:52.830 align:middle line:84% can see that he's not making a bilabial closure, 00:46:52.830 --> 00:46:55.500 align:middle line:84% your brain is rejecting the hypothesis that it's 00:46:55.500 --> 00:46:58.560 align:middle line:84% a bilabial closure and trying to come up with the closest sound 00:46:58.560 --> 00:46:59.310 align:middle line:90% that it can. 00:46:59.310 --> 00:47:00.930 align:middle line:90% And you end up with confusion. 00:47:00.930 --> 00:47:01.710 align:middle line:90% Yeah? 00:47:01.710 --> 00:47:03.240 align:middle line:84% AUDIENCE: So then, what would language acquisition 00:47:03.240 --> 00:47:05.100 align:middle line:84% look like in individuals who are born blind? 00:47:05.100 --> 00:47:07.810 align:middle line:84% NORVIN RICHARDS: Yeah, so that's a really good question. 00:47:07.810 --> 00:47:10.910 align:middle line:84% I don't know whether there's any work on that, 00:47:10.910 --> 00:47:14.310 align:middle line:84% that question of whether the McGurk effect has anything 00:47:14.310 --> 00:47:15.060 align:middle line:90% to say about that. 00:47:15.060 --> 00:47:18.013 align:middle line:84% We obviously can get along just fine without seeing the people 00:47:18.013 --> 00:47:18.930 align:middle line:90% that we're talking to. 00:47:18.930 --> 00:47:21.960 align:middle line:84% We can talk to people on the phone and all kinds of things. 00:47:21.960 --> 00:47:25.080 align:middle line:84% But we are apparently geared to pay so much attention 00:47:25.080 --> 00:47:28.020 align:middle line:84% to visual input that it actually overrides 00:47:28.020 --> 00:47:29.640 align:middle line:90% the evidence of our ears. 00:47:29.640 --> 00:47:31.110 align:middle line:90% And it's a very powerful effect. 00:47:31.110 --> 00:47:32.440 align:middle line:90% I know what's going on. 00:47:32.440 --> 00:47:33.630 align:middle line:90% And it happens to me. 00:47:33.630 --> 00:47:34.890 align:middle line:90% You know what's going on now. 00:47:34.890 --> 00:47:37.233 align:middle line:84% Then it will happen to if you watch this again. 00:47:37.233 --> 00:47:38.400 align:middle line:90% Again, I'll put the link up. 00:47:38.400 --> 00:47:42.010 align:middle line:84% And you can amaze your friends with it. 00:47:42.010 --> 00:47:43.410 align:middle line:90% It's a very robust phenomenon. 00:47:43.410 --> 00:47:45.420 align:middle line:84% It's been demonstrated for a bunch of different sounds 00:47:45.420 --> 00:47:47.003 align:middle line:84% and in a bunch of different languages. 00:47:47.003 --> 00:47:49.110 align:middle line:90% It's not just an English thing. 00:47:49.110 --> 00:47:52.260 align:middle line:90% It's all over the place. 00:47:52.260 --> 00:47:55.710 align:middle line:84% Now, am I done with other things? 00:47:55.710 --> 00:47:58.740 align:middle line:84% Well, now I am about to start talking about something 00:47:58.740 --> 00:48:00.300 align:middle line:90% which is pretty complicated. 00:48:00.300 --> 00:48:04.470 align:middle line:90% And so I'll just give you-- 00:48:04.470 --> 00:48:07.030 align:middle line:84% yeah, so I'll tell you briefly what I'm 00:48:07.030 --> 00:48:08.280 align:middle line:90% going to talk about next time. 00:48:08.280 --> 00:48:09.720 align:middle line:84% And we'll just start with this next time 00:48:09.720 --> 00:48:11.470 align:middle line:84% because I don't think we have time to talk 00:48:11.470 --> 00:48:14.162 align:middle line:90% about this in any detail. 00:48:14.162 --> 00:48:16.620 align:middle line:84% We've been talking-- well, here, let me show you something. 00:48:16.620 --> 00:48:20.700 align:middle line:84% And then we'll talk more about it next time. 00:48:20.700 --> 00:48:21.395 align:middle line:90% Back to Praat. 00:48:21.395 --> 00:48:25.730 align:middle line:90% 00:48:25.730 --> 00:48:26.390 align:middle line:90% "Pie. 00:48:26.390 --> 00:48:27.170 align:middle line:90% Bye. 00:48:27.170 --> 00:48:29.366 align:middle line:90% Spy." 00:48:29.366 --> 00:48:30.290 align:middle line:90% Ah, darn. 00:48:30.290 --> 00:48:31.683 align:middle line:90% I failed to save that. 00:48:31.683 --> 00:48:32.600 align:middle line:90% Hang on just a second. 00:48:32.600 --> 00:48:35.350 align:middle line:90% 00:48:35.350 --> 00:48:36.010 align:middle line:90% Pi. 00:48:36.010 --> 00:48:36.695 align:middle line:90% Bye. 00:48:36.695 --> 00:48:37.195 align:middle line:90% Spy. 00:48:37.195 --> 00:48:42.110 align:middle line:90% 00:48:42.110 --> 00:48:46.775 align:middle line:84% So that's me saying "pie, bye, spy." 00:48:46.775 --> 00:48:47.506 align:middle line:90% [AUDIO PLAYBACK] 00:48:47.506 --> 00:48:48.877 align:middle line:90% 00:48:48.877 --> 00:48:49.460 align:middle line:90% [END PLAYBACK] 00:48:49.460 --> 00:48:52.220 align:middle line:84% NORVIN RICHARDS: I'm sorry, the-- 00:48:52.220 --> 00:48:53.540 align:middle line:90% everything is up again. 00:48:53.540 --> 00:48:57.992 align:middle line:84% So here's-- let's just listen to "pie." 00:48:57.992 --> 00:48:59.450 align:middle line:84% Actually, let's just look at "pie." 00:48:59.450 --> 00:49:01.990 align:middle line:90% 00:49:01.990 --> 00:49:02.770 align:middle line:90% Can I do that? 00:49:02.770 --> 00:49:05.428 align:middle line:90% Come on, you. 00:49:05.428 --> 00:49:06.289 align:middle line:90% [AUDIO PLAYBACK] 00:49:06.289 --> 00:49:07.310 align:middle line:90% 00:49:07.310 --> 00:49:07.970 align:middle line:90% [END PLAYBACK] 00:49:07.970 --> 00:49:10.553 align:middle line:84% NORVIN RICHARDS: So that's the spectrogram of me saying "pie." 00:49:10.553 --> 00:49:13.730 align:middle line:84% And you can see-- here, I'll use my cursor instead. 00:49:13.730 --> 00:49:16.670 align:middle line:84% What's going on here, my-- here's the stop. 00:49:16.670 --> 00:49:20.930 align:middle line:84% And then, this little burst is my lips opening. 00:49:20.930 --> 00:49:24.020 align:middle line:84% And then, there's a little burst of white noise. 00:49:24.020 --> 00:49:25.960 align:middle line:84% And then, we start in on the vowel. 00:49:25.960 --> 00:49:30.440 align:middle line:84% And you can see the formants, the black bars that indicate 00:49:30.440 --> 00:49:31.970 align:middle line:90% that we're doing a vowel. 00:49:31.970 --> 00:49:34.430 align:middle line:84% This little burst of white noise is the period 00:49:34.430 --> 00:49:39.980 align:middle line:84% that elapses in between my lips opening and my vocal chords 00:49:39.980 --> 00:49:41.690 align:middle line:90% beginning to engage. 00:49:41.690 --> 00:49:43.910 align:middle line:84% So there's this period where there's nothing 00:49:43.910 --> 00:49:45.940 align:middle line:90% coming out of my lungs but air. 00:49:45.940 --> 00:49:47.690 align:middle line:84% There's always air coming out of my lungs. 00:49:47.690 --> 00:49:48.607 align:middle line:90% Let me try that again. 00:49:48.607 --> 00:49:50.900 align:middle line:84% There's nothing coming out of my vocal tract but air. 00:49:50.900 --> 00:49:51.830 align:middle line:90% Yeah? 00:49:51.830 --> 00:49:53.720 align:middle line:84% My vocal chords aren't vibrating yet. 00:49:53.720 --> 00:49:54.950 align:middle line:90% And it's not long. 00:49:54.950 --> 00:49:57.320 align:middle line:84% It lasts some fraction of a second. 00:49:57.320 --> 00:49:58.820 align:middle line:90% How long does it last? 00:49:58.820 --> 00:50:02.960 align:middle line:84% It lasts-- who cares how long it lasts? 00:50:02.960 --> 00:50:03.643 align:middle line:90% Do I care? 00:50:03.643 --> 00:50:05.060 align:middle line:84% Do I care long enough to find out? 00:50:05.060 --> 00:50:06.620 align:middle line:90% It lasts less time than that. 00:50:06.620 --> 00:50:08.780 align:middle line:90% It lasts 0.06 seconds. 00:50:08.780 --> 00:50:11.040 align:middle line:90% Yeah, there we go. 00:50:11.040 --> 00:50:13.680 align:middle line:84% So that's what's going on in "pie." 00:50:13.680 --> 00:50:15.530 align:middle line:90% Now let's look at "spy." 00:50:15.530 --> 00:50:18.450 align:middle line:90% 00:50:18.450 --> 00:50:19.276 align:middle line:90% [AUDIO PLAYBACK] 00:50:19.276 --> 00:50:20.285 align:middle line:90% 00:50:20.285 --> 00:50:21.120 align:middle line:90% [END PLAYBACK] 00:50:21.120 --> 00:50:23.120 align:middle line:84% NORVIN RICHARDS: And you can see-- and actually, 00:50:23.120 --> 00:50:24.560 align:middle line:90% allow me to let you hear. 00:50:24.560 --> 00:50:25.913 align:middle line:90% So here's the "s." 00:50:25.913 --> 00:50:27.330 align:middle line:84% This is what fricatives look like, 00:50:27.330 --> 00:50:29.000 align:middle line:90% these bursts of light and noise. 00:50:29.000 --> 00:50:32.180 align:middle line:84% Lots more energy up here for the "s." 00:50:32.180 --> 00:50:33.175 align:middle line:90% Here's the stop. 00:50:33.175 --> 00:50:34.550 align:middle line:84% You can see what stops look like. 00:50:34.550 --> 00:50:35.450 align:middle line:90% It's blank. 00:50:35.450 --> 00:50:37.718 align:middle line:84% Here are my lips opening for the end of the stop. 00:50:37.718 --> 00:50:39.635 align:middle line:84% And let's just play it starting with the stop. 00:50:39.635 --> 00:50:40.302 align:middle line:90% [AUDIO PLAYBACK] 00:50:40.302 --> 00:50:42.848 align:middle line:90% 00:50:42.848 --> 00:50:44.480 align:middle line:90% [END PLAYBACK] 00:50:44.480 --> 00:50:47.190 align:middle line:84% NORVIN RICHARDS: How many of you hear that as "pie"? 00:50:47.190 --> 00:50:48.840 align:middle line:84% How many of you here it as "bye"? 00:50:48.840 --> 00:50:53.350 align:middle line:84% Yeah, so this is a non-aspirated voiceless stop. 00:50:53.350 --> 00:50:54.780 align:middle line:90% So my lips are opening. 00:50:54.780 --> 00:50:56.760 align:middle line:84% And then, my vocal cords get busy right away. 00:50:56.760 --> 00:51:00.000 align:middle line:84% You can see there isn't a big gap right in between the "p' 00:51:00.000 --> 00:51:02.310 align:middle line:90% and what's happening after it. 00:51:02.310 --> 00:51:04.230 align:middle line:84% So we will talk more about this next time. 00:51:04.230 --> 00:51:06.630 align:middle line:84% But this is what's called aspiration. 00:51:06.630 --> 00:51:09.422 align:middle line:84% English has aspiration in some places, but not in others. 00:51:09.422 --> 00:51:11.380 align:middle line:84% So we'll do this demonstration again next time. 00:51:11.380 --> 00:51:13.750 align:middle line:84% And then, we'll talk about what it means. 00:51:13.750 --> 00:51:15.660 align:middle line:84% So for now, I think we're done unless anybody 00:51:15.660 --> 00:51:18.360 align:middle line:84% has any questions about any of this. 00:51:18.360 --> 00:51:19.350 align:middle line:90% All right. 00:51:19.350 --> 00:51:21.900 align:middle line:84% See you not on Tuesday, but on Thursday. 00:51:21.900 --> 00:51:24.320 align:middle line:84% So we don't have class on Tuesday. 00:51:24.320 --> 00:51:32.000 align:middle line:90%