Review of Assigned Readings

Pinker:

1) Chapter 6: The Sounds of Silence (phonology and phonetics)

1) File 3: Phonetics (Files 3.1-3.7)

 Phonetics

Phonetics is the study of the sounds we use when we speak. The technical term that linguists often use for speech sounds is the phone, so we can say that phonetics is about the study of phones. It turns out that there are a number of perspectives from which we can consider the properties of speech sounds, and these can be broken down into three fairly tidy categories: articulatory phonetics, acoustic phonetics, and auditory phonetics. Let's define each:

• articulatory phonetics: the study of how sounds are produced
• acoustic phonetics: the study of the physical properties of sounds
• auditory phonetics: the study of how sounds are perceived

Actually, we'll start by clearing up a common confusion that arises whenever people first start to learn about phonetics. This problem has to do with the difference between sounds and letters. Ready...tatoo this on your heart...
 
 

Why is this the case? Well, in simple terms, letters are symbols used in a writing system (aka orthography), such as our Roman alphabet, but sounds refer to coordinated sequences of gestures that speakers make in order to produce the words and sentences of their languages. Basically, there are lots of spoken languages that don't have orthographies (i.e. they are oral languages only), but there are NO spoken languages that don't have sounds. In fact, if you assume that sounds and letters are the same, you'll get into some immediate trouble with English.

"Why?," you ask. Well, consider the following problems. In English, as your textbook points out (and it would be a very good idea to commit these problems to memory and to be able to give examples (hint)):

• the same sound can be represented by different letters: sea, ski, amoeba OR leaf, cough
• the same letters can sometimes stand for different sounds: sigh, please OR character, chip
• a single sound can sometimes be expressed with more than one letter: lock, took, boat, apple
• a single letter can sometimes represent more than one sound: exit, use
• and even sometimes there are letters that don't represent any sounds: dough, knight, island, rhubarb

The reason to talk about this is not actually to attack English spelling, but rather, to make an important distinction between the sounds a word has and the way that it might be spelled in a particular language. Linguists don't make much use of spelling because of the problems above. (As an aside, spelling in English IS interesting to historical linguists because it gives us insight into how words used to be pronounced, but we'll set this issue aside.) However, linguists are very interested in being able to represent the sounds that a word has. So, we have developed a code for writing down the sounds of a word (or of an utterance of any length).

Writing down the actual sounds that a word has is called transcription or phonetic transcription. The obvious benefit of phonetic transcription is that if properly done, it provides us with a way of consistently and unambiguously writing down the sounds of a word so that anyone (even a non-speaker of the language) can gain a fairly good idea of how the word is pronounced. This type of accuracy allows us to record data for descriptive purposes and for later use by other investigators.

Imagine trying to do math if the number "1" sometimes meant "1" and sometimes meant "0", but you could never be sure when it was one or the other. If this seems like a problem to you, then you can understand why linguists would want to have a consistent means of transcribing sounds. The symbols that we use in our transcriptions are referred to as the phonetic alphabet. The phonetic alphabet that your textbook presents to you is one that is typically used in the United States, although there is another phonetic alphabet that is slightly different from what's given to you in the text and that is more widely used internationally. It's called the IPA, which is short for International Phonetic Alphabet.

Much of the phonetic alphabet is based on the Roman alphabet that we are all familiar with, because we use it in writing languages such as English, French, German, and so forth (but not Arabic, Chinese, Russian, or Sanskrit). Anyway, the basic idea is that each sound should have its own unique symbol (or combination of symbols), thus allowing us to transcribe anything we hear without ever having to use the same symbol for different sounds. The symbols on pp. 37-39 of Language Files are the symbols necessary for transcribing the consonants and vowels of English.

Okay, how can we see an example of transcription vs. spelling in action? Well, take the following two English words : "laugh" and "staff". In our orthography, they look pretty different. But, they are actually quite similar. The first word, "laugh" consists of three sounds, and the second word "staff" has four sounds. Both words end with the same sound, an "f", and are preceded by the same vowel. (That's why they rhyme!). They are different because "laugh" starts with a single sound (an [l]), while "staff" starts with a sequence of two sounds ([s] + [t]).

Having said all this, on the exam you'll be expected:

a) to be able to identify how many sounds a word has;

b) to be able to accurately transcribe English words;

c) to be able to provide English spellings for words that are given to you in transcription.

As I note above, articulatory phonetics is the study of how sounds are produced. In understanding how sounds are produced, we're going to divide this discussion into two general parts. The first will discuss the production of consonant sounds, while the second will discuss the production of vowel sounds. Each part will detail the phonetic features that we need to begin to understand how consonants and vowels are produced and how they can be grouped together into larger classes of sounds based on their shared features.

Consonants

In discussing how we make consonant sounds, we're going to look at three crucial phonetic dimensions or aspects of their production. These dimensions are listed here:

• VOICING (ALSO CALLED PHONATION or referred to as STATE OF THE GLOTTIS)
• PLACE
• MANNER

The first thing we want to know about a consonant is whether or not it is voiced. How can you tell whether a sound is voiced or not? One trick is to take your thumb and index finger and rest them on your throat on the thyroid, on either side of your Adam's apple. Now, make a long [sssssssssssssssssss] sound. You'll feel that there is no vibration. Having done this, now make a long [zzzzzzzzzzzzzzzz] sound. You should feel a change. Now you should feel a fairly clear buzzy vibration. That's voicing. Now make an long [sssssssssszzzzzzzz] sequence so that you can feel the change over from voicing to voicelessness.

How does voicing work? Essentially, voicing involves bringing your vocal cords (also called vocal folds) together and then driving them apart by blowing air up through them from the lungs. You can't simply look down a partner's throat and see this happening. However, you can get a pretty good sense of how things work if you do the following. Grab your upper and lower lips lightly between your thumb and index finger at both sides of your mouth. Stretch your lips apart so that there is some tension on them along their length. You'll now be in a position where your lips are together and stretched a bit. Now, blow air through them and you'll see that they make quite a bit of noise. This noise is produced because they are blown apart by the stream of air and then their own elasticity pulls them back together. Of course, they are then blown apart again by the air, and then they come back together.

This is basically what happens at your larynx (the voicebox). When a sound is voiced, the vocal cords are brought together with a certain amount of tension applied across their length. Then, air from the lungs blows them apart, and they snap back together. As a result, puffs of air are produced in very rapid succession. These puffs of air are what we hear as voicing. You saw this happening, in fact, on the video that I played for you all in class. You can also see voicing on the web at the U. of Washington web site that I mentioned in your discussion forum.

If a sound is produced this way. We call it a voiced sound. If by contrast, a sound is produced with the vocal cords held apart from one another so that they don't vibrate then the sound is voiceless. You can try this idea out for yourself by opening your mouth wide and just breathing out. Note that your lips don't vibrate now. This is because you've drawn them apart by opening your mouth.

Summing up:

• Voiced: vocal cords together and vibrating
• Voiceless: vocal cords spread apart and NOT vibrating

Voiced: b, d, z, g

Voicelss: p, t, s, k

Note in these examples that I have given you pairs of sounds: [p,b], [t,d], [s,z], [k,g]. The members of each pair are identical, except for the fact that one is voiced and one is voiceless. In English, the so-called obstruents sounds (oral stops, fricatives, and affricates) come in voiced and voiceless pairs.

The next phonetic dimension of interest to us is where in the vocal tract our consonant is made. Linguists refer to this as the place of articulation of the consonant, or place for short. Perhaps the simplest way to begin to understand this concept is to continue to think about the air that is flowing up from the lungs through the vocal cords. Before it can get out into the word, it is going to get obstructed somewhere in the vocal tract. Think about how you make the sound [b]. Before the air can get out into the world, it gets obstructed by your two lips. So, we say that this is a labial, or more precisely, a bilabial sound. Labial sounds are visible to the eye, of course. The place of articulation of other consonants isn't, but the issue is the same. Somewhere, our articulators come together to form a constriction that obstructs the airflow. That's what place is. Here's a list of the places of articulation of English sounds, as presented in Language Files.

Bilabial: These are sounds produced by bringing both lips together. Examples in English are [p], [b], [m].

Labio-dental: These are sounds produced by placing the lower lip against the upper front teeth. Examples are [f] and [v].

Interdental: Interdental sounds are made by placing the tongue BETWEEN the upper and lower teeth. The two sounds that are spelled with a [th] in English are interdental sounds. (Remember, I don't have phonetics fonts here.) The voiced one is the first sound in the word [thy], while the voiceless one is the first sound in the word [thigh].

Alveolar: These are sounds that are made at or near the alveolar ridge. The alveolar ridge is a small ridge just behind your upper front teeth. There are a bunch of English sounds made here: [t], [d], [s], [z], [n], [l], [r]. English is not special in this regard. The alveolar ridge is the most common place of articulation across languages.

Palatal: Behind the alveolar ridge, you'll note if you check with your tongue, that the mouth slopes up steeply into a kind of hard boned dome. This is the hard palate, and sounds articulated here are called palatal sounds. Your book lists five English sounds produced here. [sh] as in 'shore', [zh] as in 'measure', [ch] as in 'chip', [j] as in 'Joe', and [y] as in 'yes'. (Again, NOTE THAT THESE ARE NOT THE APPROPRIATE PHONETICS SYMBOLS. YOU NEED TO MEMORIZE THE ONES GIVEN TO YOU IN LANGUAGE FILES).

I'm not actually delighted with your book here. More precisely, only [y] is truly a palatal sound here. The others are more appropriately described as palato-alveolar sounds. If you make the first sound in my name, for example, you'll see why. It's actually realized just behind the alveolar ridge, rather than completely up on the hard palate. At any rate, I won't penalize you for calling these palatals, because that's what your book says.

Velar: Behind the hard palate,we get to the velum. This is soft tissue, so you can feel the change from hard bone to mushy tissue if you run your tongue back behind the hard palate. English has three sounds made here: [k], [g], and the velar nasal [ng] (not its appropriate phonetic symbol) as in the word "sing".

Glottal: Technically, the glottis is defined as the space between your vocal cords. English has two sounds that are made here. When we say made here, what we mean is that they have no place of articulation above the vocal cords in the way that sounds like [b] and [d] and the rest of the sounds that we've mentioned above do.

One of these is the sound [h] which is simply produced by blowing air through a wide open glottis (i.e. when your vocal cords are held widely apart), as in the word 'hi'.

The other is a glottal stop. This is just the opposite. Here, your vocal cords are pressed so tightly together that the air coming up from the lungs can't push them apart. Glottal stop is transcribed like a question mark [?], only without the dot under it. Glottal stop in English is produced in words such as [uh-oh]. It's what appears between the two vowels. (There is no [h] in this word. That's just how we spell it in our orthography.) Cockney English has lots of glottal stops. That's one of its most salient properties.

Uvular: There's one more place of articulation that I mentioned in class. The uvula is that funny looking thing that hangs down in the back of your throat. Many languages make sounds at the uvul, though English does not. A famous case is the French [R] sound. This is a uvular sound. Learn this one for the exam, even though English doesn't have uvulars.

We've just reviewed two of our three general dimensions: voicing and place. Let's keep thinking for a moment about our airstream coming up from the lungs. We've seen that it gets obstructed at different points in the vocal tract and that this is the phonetic dimension of place of articulation. If we think about it, we'll also notice that the airsteam gets obstructed in different ways. Specifically, there are different degrees of constriction when our articulators come together to make a consonant sound. These different degrees comprise what we refer to as the set of manner features. That is, manner refers to the degree to which the airflow is obstructed when we make a consonant.

I'm going to review the manner features that are covered in your book and that we discussed in class. I am going, however, to modify the presentation from Language Files slightly in order to better organize the discussion of manner. So, pay attention to what you see here!

Stops

Stops are sounds that are produced by completely stopping the flow of air in the oral cavity (the mouth). Since you can actually see them, [p] and [b] provide nice examples. Note that when we say a word such as [bye], we close our lips and completely block the flow of air for a brief time. Then, we release the air (this is called the release of the stop) and go on to say the following vowel. The stops in English are [p], [b], [t], [d], [k], [g]. Actually, these are more precisely called oral stops. There are three more stop sounds that are nasal sounds. I'll address these here.

Nasal Stops. I just said that a stop involves totally blocking the flow of air in the oral cavity (the mouth). Think about how you make an [m]. What happens is that just as you would for a [b], you press your lips together so that air can't escape from the mouth. What's different about an [m] is that air is flowing out of your nose. This is because your velum is lowered and so some of the air coming up from the lungs is diverted out through the nasal cavity. So, [m] is a stop because the flow of air is completely blocked in the mouth, but it's a special kind of stop because air is flowing out of the nose!

Lowering and raising the velum is something you do all the time without realizing it. In order to breathe through your nose or to produce sounds like [n] and [m], you need to lower the velum. And in order to produce sounds such as [p], [t], [s], and most other English consonants, you need to raise the velum and seal it tight. For those of you who are curious about such things, this is what is at the core of the problem in people who have cleft palates. They can't simply raise the velum and seal off the nasal cavity. As a consequence, their speech sounds very nasalized and they have a tough time producing the non-nasal consonants of their language.

There are three nasal stops in English: the bilabial nasal stop [m], the alveolar nasal stop [n], and the velar nasal stop [ng] (<--remember that this isn't the appropriate symbol for the velar nasal). When we are talking about consonants, all the nasal consonants in English are also stops. So, nasal stops can be said to be a subset of the larger set of stops. A picture:

STOPS (in English)
 
 

Oral Nasal

p, b m (bilabial)

t, d n (alveolar)

k, g ng (velar)

Generally speaking, linguists refer to oral stops simply as stops, and they refer to nasal stops as either nasal stops or simply as nasals. For your exam, you want to be sure, for example, that you can identify [p] as a voiceless bilabial stop and that you can identify [m] as a voiced biliabial nasal or as a voiced bilabial nasal stop. Either answer would be fine for the nasal.

Fricatives are different from stops in that they are made by forming a constriction that does not completely obstruct the flow of air out of the mouth. They do involve making a fairly small constriction however, through which the air rushes noisily out. The technical term that linguists use to refer to the production of fricatives is "turbulence" in that fricatives result in the generation of turbulent airflow.

An easy way to appreciate the difference between stops and fricatives is to think about a garden hose and your thumb. If you have a big enough thumb or a small enough hose, you can for a brief moment totally block the flow of water from your hose. This is the garden hose experiment equivalent of making an oral stop. Now, move your thumb just slightly so that there is only a tiny opening through which the water can escape. This is the garden hose equivalent of a fricative. Notice how the water, which might flow quite smoothly if your weren't blocking the end of the hose, jets out when you only leave a tiny opening for its escape. Well, this is essentially what happens to the air streaming up from your lungs when you make a fricative constriction somewhere in your vocal tract.

English has 8 fricatives made at four different places of articulation. Again, with the warning that I don't have access to phonetics fonts because I'm writing this in html, here are examples:

• [f] as in 'fat' and [v] as in 'vat' are both labiodental fricatives
• theta as in 'thigh' and eth as in 'thy' (the two "th" sounds) are both interdental fricatives
• [s] as in 'sue' and [z] as in 'zoo' are both alveolar fricatives
• [sh] as in 'mesher' and [zh] as in 'measure' (note that these are not the right symbols) are both palatal (or palato-alveolar) fricatives

Affricates.

These sounds are hybrids. Affricates are produced by completely obstructing airflow for a brief period and then releasing the flow in such a way as to produce friction (or turbulence). Sound familiar? If so, good. What affricates are, in effect, are sequences of a stop + fricative!

English has only two affricates: the palatal sounds [ch] as in my name 'Chip', and [j] as in the first and last sounds of the word 'judge' (remember, [ch] and [j] are not the correct phonetic symbols for accurate transcription.)

Liquids.

Okay, back to our garden hose example. If, after we've produced our noisy jet of water (the fricative), we gradually move our thumb away a bit more so that the constriction is not as tight, we'll see that the flow of water at a certain point will become smooth. This is comparable to what happens in the mouth when a liquid sound is produced. There is still a constriction formed that obstructs the flow of air, but it isn't narrow enough to generate the turbulent airflow that is characteristic of fricative sounds.

The class of liquid sounds can be broken down into two groups: [l] sounds and [r] sounds. The fancy terms for these are lateral (remember "l" for lateral) and rhotic (remember "r" for rhotic).

Lateral sounds are special because they are made such that the airflow coming up from the lungs is channeled down one or both sides of the tongue. As I pointed out in class, you can check how you make your [l] by making an [l] sound, stopping without changing anything about the position your tongue position in the mouth, and then breathing in. You'll feel a sensation of coldness on the side or sides of the tongue along which the air flows. Lateral sounds (the family of [l] sounds) are different from all other sounds in that they are the only ones that channel the air in this fashion. All other sounds channel air down the center of the tongue.

Rhotic sounds are [r] sounds. There is a great variety of ways in which [r] is produced in various dialects of English, as well as across languages. We're not going to delve into the complications of the world of rhotics here, but it is useful to know that one of the common realizations of [r] in English is what we call a retroflex [r]. This is what we do in American English. We make a retroflex [r] by curling the tongue tip up and pointing it just behind the alveolar ridge (though some people get the same acoustic effect by just bunching the tongue back).

Give a phonetic description of the following sounds

[l]:

[r]:

[l]: voiced alveolar lateral-liquid

[r]: voiced alveolar retroflex-liquid

• liquids are produced by a constriction that is more open than a fricative constriction, i.e. that doesn't generate turbulent airflow
• liquids can be divided into two classes: [l] sounds (laterals) and [r] sounds (rhotics)

Glides are consonants that are realized with only a slight amount of closure or constriction. Thus, like liquids, they are realized with a smooth rather than turbulent airflow. In fact, glides have such a slight constriction that they are actually vowel-like, and that is why they are often referred to as semi-vowels.

We have two glides in English: the palatal glide [y] and the voiced bilabial glide [w]. These are the terms that Language Files uses. I would prefer to call [w] a voiced labio-velar glide. It's true that it's labial because the lips are rounded, but it's also velar because the back of the tongue is raised towards the velum. Say the word "we" and try to concentrate on how your tongue moves back and up during the [w] before sliding forward for the [i] sound that follows. At any rate, note also how similar [y] is to the vowel [i] and [w] is to the vowel [u]. Again, that's why they are called semi-vowels.

Finally, note that in some dialects of English, we also have a voiceless labio-velar glide. So, some people make a distinction between 'which' and 'witch' such that the first sound in 'which' is voiceless while the first sound in 'witch' is a voicelss labio-velar glide.

Glottal Stop.

I've separated glottal stop from the other stop sounds for a particular reason. Unlike those sounds, which involve totally blocking the flow of air by bringing two articulators together (such as the lips for [p] or the tongue tip and the alveolar ridge for [t]) glottal stops are made by bringing the vocal folds together tightly for a brief period. This produces a kind of momentary hesitation or catch in the throat. These are different from other stops because other stops can be either voiced or voiceless. Glottal stops, on the other hand, must be voiceless in that the vocal cords must be held tightly together in order to produce a glottal stop. And if they are being held tightly together to completely block the flow of air at the glottis, then they are not together and vibrating. (In truth, this is actually an oversimplification, but it will do for our purposes in this class. If you want to know more about glottal stops, come and become a lab rat in my phonetics lab!)

• VOICING:

VOICED

VOICELESS

• PLACE:

BILABIAL

LABIO-DENTAL

ALVEOLAR

PALATAL

VELAR

UVULAR

GLOTTAL

• MANNER: STOP

oral

nasal

FRICATIVE

AFFRICATE

LIQUID

lateral

rhotic

GLIDE

As with consonants, we have a group of phonetic features that we use to describe vowels. I've focused on four of these with you all in class. These are:

• vowel height
• vowel backness/frontness
• vowel roundness
• vowel tenseness/laxness

One of the things that makes reviewing all of this here difficult is my lack of phonetics fonts in html. So, I'll say things like "the vowel in 'beat'", etc... With this in mind, here's a quick sketch of the vowel features.

Vowel height

The dimension of vowel height refers to the relative degree to which the tongue body is raised or lowered in the mouth during the production of a given vowel. We divide the vowel space up into three general heights: high, mid, and low. Here are examples of English words with high, mid, and low vowels:

• high: beat, bit, boot, good
• mid: bait, bet, but, oat, bought
• low: bat, pot

Besides differing in the height of the tongue body, different vowels with relatively more or less retraction of the tongue body in the mouth. Again, we have three general areas or divisions: front, central, and back vowels. Here are examples from English:

• front: beat, bit, bait, bet, bat
• central: but (i.e. the schwa)
• back: boot, good, boat, bought, pot

Some vowels are made with the lips rounded, while others are made without rounding the lips. Vowels that are made with rounded lips are called, not surprisingly, round vowels. Here, we only make two distinctions: round vs. unrounded. Examples:

• round: boot, good, boat, bought
• unrounded: beat, bit, bait, bet, bat, but, pot,

The final feature for vowels that we'll cover here is what we call the tense vs. lax distinction. Essentially, the idea is that some vowels are made with more extreme tongue gestures than others are. You can feel this if you practice making the tense vs. lax pairs of vowels. Here's a list of examples from English:

• tense: beat, bait, boot, boat
• lax: bit, bet, good, bought

NOTE: a very nice and useful chart is given to you on p. 49 of Language Files. Learn it. Each of the phonetic features for vowels that we've reviewed here is listed there, along with the appropriate phonetic vowel symbols. (Remember, you only need to know one symbol for the schwa vowel--the upside down "e". Don't worry about the wedge symbol.) Also, two more caveats: the chart lists the low vowels [ae] and [a] as lax, along with schwa. That's fine. I focused in class on the ones that come in pairs, e.g. i/I, e/E and so forth. Note that for the ones I focused on, the tense/lax distinction is needed to teases these vowels apart, while for the other vowel, there aren't tense/lax pairs. They're just kind of redundantly lax. So, if I say, give me the low front vowel in English, there's only one and it happens to be lax. But if I say, give me the high front vowel in English, there are actually two, and you need tenseness v. laxness to differentiate between the two of them.

Natural Classes

Once we've learned our phonetic features, it is important to be able to see how individual sounds can be grouped into more general classes. A natural class is simply a group of sounds that share a common phonetic feature or set of features. So, if I list the following sounds {p, t, k}, you should be able to recognize immediately that these constitute the natural class of voiceless stops in English. Here's a couple of natural classes of English sounds. Can you tell me what the natural class for each is?

1. {i, e, o, u}

2. {s, z}

3. {p, b, m}

This should be easy come exam time. Here are the answers: (1) tense vowels, (2) alveolar fricatives, (3) bilabial stops. Notice also that if I had only listed {p, b}, we wouldn't have the natural class of bilabial stops in English, because we wouldn't have included all of its members. We would have the class of bilabial oral (non-nasal) stops, but NOT the whole class of bilabial stops. Make sure you understand the exercises in File 3.6.

Suprasegmentals

The last phonetic features that I want you to make sure to review are the features that make up what we call the suprasegmentals. Remember, we use the term suprasegmental because these features are thought of as "riding on top of" sounds (often vowel sounds). There are four that we discussed in class:

• length
• intonation
• tone
• stress

Length refers to how long the duration of a sound is. Many times length differences are important. Your book cites the example of Finnish, where both the length of consonants and vowels can make a meaning difference, as seen in the following triplet of words:

[muta] 'mud'

[mu:ta] 'some other'

[mut:a] 'but'

We mark long vowels and consonants by writing a colon after the sound (or by doubling the consonant or vowel, eg. muuta, mutta). What's important in these Finnish examples is that vowel length is crucial in distinguishing between the words meaning 'mud' and 'some other', while consonant length is what's important in distinguishing between the words meaning 'but' and 'mud'.

In English, there are noticeable differences in vowel length, depending on where a vowel falls in a word. Say the following pair of words to yourself: 'bead'/'beat'. You can hear that the [i] vowel is much longer in 'bead' than in 'beat'. We can transcribe these this way: [bi:d] vs. [bit]. This length difference is totally predictable. Vowels are always longer before voiced consonants than before voiceless consonants in English. As we know from phonology, we can say that in Finnish, long vowels are contrastive, while in English they are not. Nevertheless, from a purely phonetic point of view, English does have two flavors of vowels: relatively long vowels and relatively shorter vowels. (Of course, if you plunge ahead later in life and take phonetics, you'll see that this is too simple, but you should get the main point anyway!)

Intonation

Intonation refers to the way that some languages like English use changes in pitch. We use changes in pitch in English to signal differences in the meanings of whole sentences. So, a very common example is that we use a rising pitch when we are asking a question but a flatter pitch when we are making a declarative statement. Say these two things and listen carefully to yourself to see what I mean:

Tone also involves the way that a language uses pitch. Only in a tone language, pitch changes signal changes in meaning from one word to another. Chinese is a language that is a tone language, and so are Thai and Vietnamese. On page 57 of Language Files, you are given a very famous paradigm from Mandarin Chinese which shows how the same two sounds [ma] can have FOUR different meanings because of FOUR different tones. So, in Chinese, changes in pitch mark changes in meaning for individual words.

The difference between intonation and tone is simply this: in intonation languages like English, pitch is used to express sentence level meanings such as question, disbelief, declarative statements, surprise, doubt, etc... By contrast, in tone languages, the very same mechanism, pitch change, is used instead to signal differences in word meaning. So, in Chinese, [ma] can mean 'mother', 'horse', 'hemp', or 'scold' depending on what pitch it is realized on. Pitch is the vehicle in both cases, but its linguistic use is different!

Stress

Of all of the suprasegmentals, stress is perhaps the most abstract. Why? Because while length can be measured in milliseconds, and pitch can be measured in Hertz, stress doesn't have a single, straightforward correlate in the physical world. Stress is about prominence. Stress is about some syllables standing out as more salient than others in a word. They may stand out for various reasons. They may be louder, longer, of a higher pitch, or a combination some or all of these. But the key is that no one of these physical features always correlates with stress.

Nevertheless, stress is REAL. Think about how stress can make a meaningful difference in English. (I will use capital letters to mark stress.) If you say "PERvert" you are not saying the same thing as if you say "perVERT". The former is a noun, while the latter is a verb. And all that is making the difference is where the stress falls--on the first syllable of the noun and on the second syllable of the verb.

A final note on phonetic transcription

What used to be File 3.8 in Language Files (but no longer exists in the new versions) presented a very nice run down of the difference between broad and narrow transcription. This is still a nice point to end our discussion of phonetics here, because it serves as a useful bridge into phonology. What is important to note is that there are two different modes of transcription: a broad way of transcribing that doesn't record all of the aspects of how we pronounce a word, and a narrow mode which more strictly transcribes what is actually produced.

The big point is that some aspects of how we pronounce a word are *entirely predictable* from context. Thus, we see that vowels are predictably nasalized when they appear next to a nasal consonant. And we see that vowels are longer when they precede a voiced sound. And we see that liquids are devoiced when they follow an aspirated stop (as in 'play' or 'prey').

These predictable aspects of pronunciation are what will lead us into a discussion of phonology--of the difference between the level of the phoneme and the level of the allophone. I've gone over a number of the predictable aspects of English pronunciation in class. For example, I've shown you that voiceless stops are predictably aspirated when they are the first sound of a stressed syllable, as in the word /pIt/. I've shown you that vowels are predictably lengthened before voiced consonants. I've shown you that vowels are nasalized both before and after nasal consonants in English. I've shown you that stops can be realized as unreleased as the end of English words like /tap/.  I'll tell you now that I'm going to give you a question like this on the exam:

"Provide a broad and narrow transcription of the word "toad". Make sure your narrow transcription is different in at least two ways from the broad transcription."

How would you answer this? Here's how:

broad transcription: /tod/

narrow transcription: add aspiration to the t, lengthen the vowel, make the final stop unreleased!

(you'd actually do the transcription this way, but, as you know, I don't have the phonetics fonts to do so here!)

Okay, onward to phonology.