 Good morning. Are you all right, comfortable? Okay, now in this part of the course, we are going to begin the second unit of the course. Until yesterday, we were looking at the design features of language. What are those features of language that make it any language, Telugu or English, a tribal language or Latin or Sanskrit, any language, any natural language has all of these features and we call them design features. They are part of the design of the language. If even one of them is not there, it's not a natural language and these features, all of them may not be unique to human language but some of them are and these some features make human language very versatile. Say for example, we saw that human languages have a structure, have a system. Okay, they can be learnt, they can be taught, they can be ambiguous. Meaning is not only the property of sound. Meaning is also the property of the way sounds are arranged. L-I-T and T-I-L we saw may have the same sounds but they do not mean the same thing. John kicked Mary and Mary kicked John have the same words but they don't mean the same thing. Meaning also depends upon the way they are arranged, the way they are structured and we also saw that languages have dual structure. Not just one, not just linear but you also vertical structure. Every time you speak, every time I speak, every time we speak we make choices. Should I take X or Y and if you take X then you have to take A. If you take Y then you have to take B. There are these constraints on the structure of language. Beginning today we will talk about the second unit of the course which is please write speech sounds. Okay, not noise, not non-speech sounds. Say this is also sound. Tapping the table is also sound. Clapping, it is also sound. Okay, or snapping your fingers. Can you do it? These are all sounds. We are not talking about them. We are talking about speech sounds. The sounds that we produce when we speak. We will talk about the structure of speech sounds between now and next two weeks Not more than next two weeks, quite some time. About a dozen hours about how speech sounds are produced, how organs of articulation are organized and how we get different kinds of sounds from within the same organs of articulation. Are you okay? Are you with me? Say yes or no please. Are we together? You have to take notes. Please take out your pen and paper or pencil. Better beginning tomorrow for this class. Bring a pencil and eraser. You will have to draw lots of diagrams and that sort of thing to be able to understand the way speech sounds flow. The branch of subject which studies speech sounds is called, please write, phonetics. In Greek, phon means sound or speech sound and tix is the science, phonetics. The science or the study of speech sounds. We can study speech sounds in any one of the three ways. We can study the way it is produced, production of speech sounds. Or the way it is transmitted. It goes from me to you. I produce the speech sounds, it reaches you. How does it travel? What happens? Are there microphones? Are there wires between you and me? What is the medium? How do speech sounds travel? And third, reception of speech sounds. How are speech sounds received? We know that we will hear through our ears. But just as we produce different sounds in different ways. When you say, both your lips come together. But when you say, your lips are far apart. Please say it, both your lips are together, right? But when you say, your lips are far apart. Say, are they not far apart? Yes or no? Yes. So, you know different speech sounds. Different speech sounds are produced differently. Are they also received differently? Is it that your ear bends when you speak per? And it rises when you say, how? How do we receive them? Primarily, there may be many more. Science, as somebody said, has no limits. Because science is God. Science is nature. It has no beginning. It has no end. It is all of it. So, from the little that we have understood, we try to understand. We say that the entire speech sounds can be studied from one of the three angles, or all the three angles. Namely, what are they? Can you please repeat? Production of speech sounds. Everybody together, please. Production, transmitted. Transmission of speech sounds. And reception of speech sounds. Actually, I made a mistake. So, production of speech sounds, reception of speech sounds. Each of these branches has a name. For example, you can see at the PPD screen, the production of speech sounds is studied under a branch of phonetics called articulatory phonetics. Please note. Articulatory phonetics. Production of speech sounds. What is it called? Articulatory phonetics. What is it called? Articulatory phonetics. Everybody, please. Repeat after me. Articulatory phonetics. Articulatory phonetics. Yeah, it is not phonetics. It is phonetics. What is it? Phonetics. Correct. It is articulatory phonetics. The science that studies production of speech sounds. Or you can look at the transmission of speech sounds. And what is it called? This branch of science? Acoustic. What is it called? Acoustic phonetics. Study of the transmission of speech sounds. It is known as acoustic phonetics. What is it known as? Acoustic phonetics. And then you have what is called the reception of speech sounds. That branch of science is called auditory phonetics. Audition, listening. Auditory phonetics. We will not be looking at in any detail at the first two, you know. Here on this course if you want to know more about some of these things. More about, sorry, the second two. More about acoustic phonetics and auditory phonetics you will have to do. You should do higher courses here or elsewhere at this institute or some other place. Here on this course we will be talking mostly about articulatory phonetics. Not about acoustic phonetics, neither about auditory phonetics. What are the problems? You know, I will just give you an overview of those two subjects and then come back to it. I will give you an overview of those two sub-branches or branches of the subject and before I come back to. But let us first see what is primary to speech? What is the most fundamental requirement for the production of speech sounds? You know, you can close your eyes and answer. Think about it for ten seconds and then answer that question. Open your eyes and then answer the question. The question is, can dead people speak? Yes or no? Can dead people speak? You do not seem to be sure. Can dead people speak? Vishal, Anurag, where are you? Can dead people speak? No. No. Say that formally. Say no. Can dead people speak? No. That is what we call a clear answer. Otherwise it looks you are not sure. Maybe my grandfather could speak but my wife's grandfather could not. No dead people can speak. It's a universal rule. Why can't they speak? Close your eyes. Think for ten seconds and tell me. In the meanwhile, I'll write something else on the board. Open your eyes now. Why can't they speak? They have the tongue. They have the lips. They have the lungs. Why can't they speak? What is it they don't have and living people have? Living people have breath. Thank you very much. Who said that? Please, my friends kindly capture my friend. Remove your hand from your face. Please don't clap. You are embarrassing me. Excellency is routine here, right? Dead people cannot speak because dead people do not have breath. Living people cannot speak because living people have breath. Dead people do not have breath. Breath or air is primary to speech. You may have lungs. You may have tongue. You may have teeth. You may have gums. You may have X. You may have Y. No matter what you have, if you don't have air going into your mouth or coming out of your mouth, you cannot speak. Every time you speak something, it is either air coming in or going out. Keep your hand here and start speaking. Now sing it here this way. Don't close your mouth. Keep your hand before your mouth and speak it now. Or any Sanskrit mantra or Shloka or keep saying IIT is best, IIT is best. Okay? Do you feel some air comes out? Every time you speak, actually from some people's mouth, bad air comes out. But air does come out. If there is no air, there is no speech. Don't let any air come out. Now speak, but don't let any air come out. Just keep your hand here, but don't let air come out or go in. No. You cannot speak. Air, whether getting in, you can also speak while getting in. Say for example when you say, or for when you say, do it. You are sucking air in. You are taking air in. Okay? Or when you throw air out, as you say, or you say, Abba, Abha, Akka, Agha, whatever you say. Okay? Air stream is important to the production of speech sounds. Dead people cannot speak because their lungs cannot take air in, cannot throw air out. Okay? They have everything else. They have tongue. They have teeth. They have gums. They have heart pallet. They have vocal cords. Everything is there yet. No speech sound can come out because there is no air. Speech air is primary to speech. Right? Obviously, you know, we can have two kinds of air stream. We can have air stream. We can have air stream that is outgoing. Or we can have air stream that is incoming. When you say, you are taking air in. When you say, you are taking air in. There are some languages like there is among Indian languages. People say Sindhi. Do you know where it is spoken? Sindhi? Yeah, in Sindh. Now in Pakistan. And there are Sindhi speakers all over India. Okay? But primarily it is spoken in Karachi and adjoining areas in the Sindh province of Pakistan. What is Pakistan today? But you know what used to be all of India. Sindhi has some sounds where you take air in. For example, they don't say ba. They say swelior cheek. Swelior cheek. Puff it. And now suck air in. Ba. Say it. That is air going in. So some sounds are also produced by incoming air. Air goes in. A large number of speech sounds, nearly 99% of speech sounds are produced through air going out. We call it, please write, egressive. What is it called? Egressive. Airstream mechanism. Airstream mechanism. And because this air comes from the lungs, so we call it, we call it pulmonic. What is it called? Pulmonic. Pulmus in Greek. Pulmus in Latin is, I'm not sure, check please whether it is Latin or Greek. Pulmus is called, lungs are known as pulmus in Latin or perhaps I should also check. Greek. And because this air comes out of pulmus, it is known as pulmonic. Please give me the complete name. Pulmonic. Airstream. Airstream mechanism. Everybody together once again please. Pulmonic. Agressive. Airstream. It looks like this. Together. One, two, three. Pulmonic. Egressive. Airstream mechanism. Slowly please. We are in such a great hurry. Come again. Pulmonic. Egressive. Airstream. Mechanism. Mechanism. Mechanism. Not mechanism. Please, at least if you don't know mechanical engineering it doesn't matter. But pronounce English correctly. It is not mechanism. It's mechanism. What's it? Mechanism. Mechanism. Not mechanism. It is mechanism. What is it? Mechanism. Okay. Pulmonic. Egressive. Airstream. Mechanism. Therefore incoming will be known as what is it now? What is it called? Pulmonic. Egressive. Egressive. Airstream. Mechanism. Mechanism. What is the incoming air known as? Once again please. Pulmonic. Egressive. Egressive. Airstream. Mechanism. All speech sounds are produced no matter what language you speak. No matter you are a Lata Mangeshkar or a Siddhi Chaudhary. How sweet or how sour. How noisy or how melodious. All speech sounds are produced through either of these two mechanisms. Either it is through pulmonic, aggressive, airstream mechanism or it is through pulmonic, ingressive, airstream mechanism. Because this is mostly and this is common we don't need to remember it. We remember this. Egressive airstream or ingressive airstream. Again there are very few sounds which are produced through ingressive airstream. Largely it is egressive airstream which produces most speech sounds in any language. Not just English, not just Telugu, not just Hindi, any language. You will find that most speech sounds are produced through egressive airstream mechanism. Look at the drawing, the diagram of the mouth. The speech air goes in this way through the oral passage, through the pharynx into the lungs or it arises in the lungs. The stomach muscles press, press lungs. That is why you will find many teachers going to canteen after the class because their lungs, their vocal cords are tired. They want a cup of tea. Students go to the class because they are tired of listening. But it is a muscular effort that produces speech. Stomach muscles press lungs, lungs press air out and air comes through the mouth, this passage. We will talk about that in some detail, the airstream mechanism, the production of speech sounds in some detail a little later. Let us look at something else. Let us look at, because we will not be talking about acoustic phonetics at any length I am just trying to give you an overview. What is an overview? Can somebody tell me what is an overview? Say for example, you want to see the entire IIT campus and you have only two minutes. Now what can you do? You can take a helicopter, go up above and from there you see the entire campus. On your left is Sindhu, on your right is Sharavati etc., etc. Or go to the ad block on the top of a tall building. Stand there, look around and you see. Similarly from a greater distance if you go to the moon you can see the entire half of the earth hemisphere. So I am giving you an overview, not details. You will just see peaks and valleys. How does a speech air travel? Acoustic phonetics talks about transmission of speech sounds. How do speech sounds travel? Speech sounds travel in waves. Why do they travel in waves? Because lungs have to contract and to expand. It is like this, contract and expand. So air comes out, the wave is stopped, comes out again, wave is stopped again. So it goes out in waves. And because it goes out in waves, you remember your basic physics, it has a crest. All waves have a crest. Have you been to the beach? Go to the beach again. I mean the holidays that are coming, Independence Day and other things. Go there and check again. You will see that all waves have a trough. They go as low as they can. And all waves have a, what is the opposite of crest? Sorry, trough? Crest. Speech air travels in waves. Because it travels in waves, waves necessarily have troughs. What is a trough? The lowest, the valley. This is the, you see, crest, trough, crest, trough, crest, trough, crest, trough. It always goes this way. That is how speech air flows in waves. And waves have crests and troughs. Waves also have other characteristics. They are called frequency characteristics. What was the pressure with which you threw air out into the lungs? Anything, you know, anything thrown is thrown with a pressure. You can throw the ball such that it reaches only the batsman in cricket. You can throw the ball such that it reaches the fielder far away. The fielder returns the ball, good fielders return the ball such that it hits the wicket without the wicket keeper having to bother about it. So, you know, you throw anything with a particular amount of force. So, waves are also thrown with force. So, waves can have, sorry, I should have written crests here. So, waves are also produced with a kind of force. There can be a force which says maximum it will go there. And there will be a force which says it will start here. Let's write a starting force of the wave, you know. You can start it at 2 kilograms per inch. Or you can start it at the level of 2 tons per inch. Now, 2 kilograms per inch will go maybe 20 meters. Or 2 tons per inch will perhaps go to America. So, all waves have two kinds of frequencies. Number one, please write, fundamental frequency. The frequency at which it started, fundamental frequency. The frequency at which the wave started. And the terminal, it cannot go beyond. It has reached its crest. Now, it has to decline. It has to fall. So, let's call it terminating force. Or in the language of physics, call it formant frequency. Can you take it? So, you know, all waves are produced through, you know, both. Fundamental frequency. In book, in literature, you will find it abbreviated as F0. And formant frequency, FF. Fundamental frequency and formant frequency. Or initial and terminal frequencies. Because we do not produce each sound with the same kind of initial and terminal frequencies, their waves differ. And that is why our ears catch these waves differently. Look at the picture I have given you. This is called a spectrogram. This spectrogram gives you the pictures of some sounds. This is E, the vowel E as in cheese in English. Or this is, this is one kind of vowel, another kind of vowel. Yet another kind of, one kind of sound, another kind of sound, another kind, three different kinds of sounds you will see have three different kinds of waves. Okay? In the case of one, we have lot more air coming out. Okay? We have different kinds of fundamental and formant frequency. We have different kinds of band structures. We have different kinds of perception. We have different kinds of reception. Because we have different kinds of production. So in acoustic phonetics, we study transmission of speech sounds. Where we study the following, how speech sounds are formed. Fundamental frequency, formant frequency, transmission characteristics, relative differences. What's the difference between pa and ba that your ear can, you know, you and I otherwise require when somebody says, write the phone number and write the name. Then you say, we, we agree. And you ask, we as in Bangalore? No, we as in victory. Okay? You and I need clues. But routinely, face to face, your ear catches. All of these frequencies very appropriately so that you understand one another without having to ask your speaker to repeat himself or herself. How does that happen? That happens because each speech sound has different fundamental and formant characteristics. Please write their fundamental and formant characteristics. Their fundamental frequency and formant frequencies differ. That is why we are able to perceive different sounds. You can do a lot of engineering in a speech, you know, processes. You can design a machine which can listen like human beings, which can answer like human beings. But if you design that machine, your machine will have to have the ability to recognize the different wave characteristics. Okay? Only if you can recognize the difference between wave characteristic of a and wave characteristics of u, can your machine perceive that now we are talking about car or we are talking about cut. Okay? These wave characteristics differ for each sound. Actually, some people say that no two sounds are produced alike. Some people go further and say even the same sound is not produced alike on two occasions. Okay? That is the dynamics of nature. You know? We are always changing, constantly changing, constantly influx. That is why there has been a great philosopher, a Greek philosopher, senior to Plato by about 300 years, somebody called Heraclitus. Please write. You know, it may not be important now, but before you are 40, before you think of buying another car or buying a plane, okay, think of reading it. Okay? And people like, you know, this is also there in Sanskrit, the Indian tradition, that, you know, Heraclitus said, we swim and do not swim in the same river twice. Because by the time you approach the river, yesterday's water has gone away. And yet you go to the same spot. So, you know, it's the same thing. On no two occasions, any speech sound is produced the same way. If you take the spectrogram and today's computers have it by default, almost. Or if you like, you can go to the speech laboratory of the computer science and engineering department, ask colleagues there. I want to see the photograph of my own voice, and they will let you speak into the microphone, and the computer will produce a picture of your voice, and you will see, speak the same thing twice, and you will find it is not exactly the same. Okay? Are you with me? Are we together? All I'm trying to get you to understand is, not these details just now, but the fact that production, transmission, reception of a speech is a highly complex engineering. Okay? But nature has gifted us with apparatus, with mechanism, which helps us understand things easily, almost effortlessly. We don't even say that, look, look, I am going to now switch on my turning, my listening machine. It happens automatically. It happens in the middle of noise. It happens in the middle of quarrel. You can take a phone, talk to somebody else. You can also pretend to listen to a third person. Actually, there are people who can take two dictations at the same time, just as there are people who can do two paintings at the same time. Okay? They are the freaks of nature, which speaking is a natural ability. Similarly, listening is also a natural ability. How does that happen? That happens because all speech sounds come with different formant and fundamental frequencies, different wave characteristics. They strike our ears differently and our ears know how to recognize them and that is how they are recognized. Okay? Any problem? Are you with me? No, only three people say yes. Are you with me? Yes, sir. Lovely. Great. Okay? How do we listen? Look at the picture of your ear. Touch your ear, please. If it is dirty, touch it with a handkerchief. Okay? Why do we have grooves in the ear? Why do our ears, are the bones or flesh or what? Cartilage. Why cartilage? Why not bones? You know? Or some people say, I will have ears of diamond. You know? Rich people could have had ears of diamond. Slum dwellers could have had ears of paper. Disposable. Throw it tomorrow. Okay? Why do we have ears like this? Any guesses? Look at your two ears. You know, go today and take a tape. Ask your friend to measure it. They are not exactly the same height. Okay? They are not from the crest of your head. Okay? There is a minuscule difference. They are not exactly symmetrically aligned. There is, there is a, you know, there is a slight difference. One is a little higher than the other. Why do we have two ears? Why are the ears designed this way? And then why do ears have these grooves? They should have in one something flat. Keep it clean, you know? Maybe they were designed so that we could have jewelry. Okay? The reason is we need two ears because it is the mean of that angle. No matter where the sound comes from. Okay? It hits right ear and left ear and they make a certain kind of angle. And your, your brain, the processing mechanism finds the mean of that angle and knows this is the source. Those who go weak in one ear have difficulty recognizing the source of the sound. It happens when you get aged. It happens when people age. The ability, physical ability starts weakening. Their eyes become weak. Their ears become weak. Then, you know, they have difficulty seeing who called from where. You may be calling them from Ganga hostel, but they go answering to Sharavati hostel. Okay? They have difficulty listening. They have difficulty answering. But ideally nature has given us two eyes, two ears, so that we can locate exactly taking the average where the sounds come from. These groups because it is with, to these groups that speech, air, wave strikes. They catch those waves and then, you know, waves are directed inside through the inner air and inner air is a very complex thing. Look at the drawing of the inner air, inner ear. Who can do the drawing of the inner ear in 30 seconds? Can you? At least the outlines? Try. I will give you 30 seconds. Let us see. Or one minute, full one minute. Can you do the drawing of the inner ear? First do the outer ear. Doesn't matter. Is it an Indian ear or an American ear? Don't know. Can't say. I took it from Google. Okay? Come on, quick. One minute only, please. Do the outer ear first. Just do the outer ear. Don't worry about inner ear. Inner ear is very, very difficult, very complex phenomenon. And you can find it on the net. Okay? It is anyway there, you know. I am only trying to draw your attention to the groups. We know hardly enough about it. We don't know much. We only know that there is a tunnel inside this tunnel. Okay? This speech here comes here. And then here, this thing, you know, the cochlea, the tiny hair, more than or close to 40,000 tiny hair cells here. So this speech sounds strike these hair cells. And at the strike of the hair, at the strike of these waves, and they regain their normal position. They regain their normal position. Okay? So depending upon, you see, longer hair makes a bigger arch. It goes down in a long distance, comes back in a long distance. So the distance traveled, the time taken, the entire thing, all these mechanisms, perhaps, this is a very big word, an important word, perhaps. We know nothing about how speech sounds are produced, how speech sounds are processed between cochlea, the inner ear, hair cell, and the brain. How those signals are transmitted, we don't know. We are human beings today. People say, know as much only about these things as they know about. They knew about the solar system in the time of Galileo, which means we know very, very little. Right. Let's continue another. So we know hardly enough about how speech sounds are processed. We only know that speech sounds strike the hair cells in the inner ear with a certain force, and they produced electrical impulses which traveled to brain, and brain interprets them in various ways. We do not have any algorithm which tells us that A is produced in, perceived in this way, and B is produced in this way. We don't know enough. But finally, the most important thing we'll be talking about on this course is articulatory phonetics, how speech sounds are produced. Human beings have a very large capacity. Human beings are versatile. They have the capacity to produce all kinds of sounds. They can speak like bees. They can speak like flies. They can speak like birds. They can speak like other animals, other mammals, other four-footed animals. Actually, in animal world, only four kinds of sounds are produced. People say we can either produce hissing sounds like reptiles, you know, like snakes and crocodiles. They produce sounds like the hissing sound. Or we can produce buzzing sounds like flies. Or we can talk like birds, you know. Birds can whistle. They can say, can you do it? Yeah, right. So birds can do that kind of sound. They can produce, you know, gliding sounds, whistles. Or like dogs or lions, you can produce roaring sounds. You can say, wow, wow, wow, okay? So, you know, if you look at the entire animal world, only four kinds of sounds are possible. People say you can either produce hisses or buzzes or banks or glides. Human beings can produce all the four different kinds of sounds. They can produce hissing sounds such as when we say sss, as in sit, or as in sing, or as in soap, or as in, you know, all different kinds of sss, as in shop, or yeah, or as in chivalrous, okay? Or they can produce buzzing sounds. Sss, as in is, was, zee, zee TV, okay? Or, you know, all kinds of other buzzing sounds that we have. Or they can produce sounds like glides, you know. Say for example, when you say, when you say, when you say, or, yeah, or sounds that are produced with great explosion like pa, ba, a, a, o. So, you know, human beings can produce all the four or five different kinds of sounds. That is because how is that possible? What is it that flies cannot do? Bees cannot do? Lions and elephants and dogs cannot do? Birds cannot do? Humans beings can. The idea is, the reason is, human beings have a very versatile apparatus of speech, organs of articulation. They have a tongue, they have vocal cords, they have teeth, they have jaws, they have lips which manipulate sounds in many different ways. And then we produce all kinds of different sounds. Beginning tomorrow we will talk about organs of articulation and we will talk about production of speech sounds. You may refer to the books I have already listed in the course outline. All of these books are known as some kind of phonetics, you know, either articulatory phonetics or introduction to phonetics, a number of things. Look up any book on phonetics and you will have a very easy time understanding the subject. If you have difficulties or doubts, please ask me. Please look at the references I have given and you will sure enjoy and understand. This is where the next frontier of challenge for engineering lies. Can you design a machine which can speak and listen like human beings? Which can ask and answer questions like human beings can? If you want to do that, your machine will have to understand some of these mechanisms. We will talk about the production of speech sounds for the next 2 to 3 weeks. Thank you. Do you have any questions? Thank you. Have a good day.