Sounds

Chapter 1 showed that human language is built on a small set of meaningless sounds that combine into words. This chapter looks at those sounds as physical events: what the mouth, tongue, lips and throat do to make them. The study of speech sounds is called phonetics. It matters because the spelling of a language is a poor guide to its sounds: English ship, nation and sure all begin or contain the same sound spelled three different ways. To describe sounds precisely, linguists use the International Phonetic Alphabet (IPA), which this chapter introduces.

How Speech Is Made

Every spoken sound needs three things:

1. An airstream. For almost all speech sounds this is air pushed out from the lungs.
2. A source of sound. The vocal folds in the larynx (voice box) either vibrate, producing a buzz, or stay apart, producing none. This is voicing.
3. Shaping. The tongue, lips, jaw and the soft palate (velum) change the shape of the space above the larynx and so shape the sound.

You can feel voicing yourself. Put your fingertips on your throat and say a long sssss, then a long zzzzz. The z buzzes against your fingers; the s does not. Everything else about the two sounds is the same: the difference is voicing alone. S is voiceless and z is voiced.

Consonants: Voicing, Place and Manner

A consonant is described by answering three questions: is it voiced or voiceless; where in the mouth is the airflow obstructed (place of articulation); and how is it obstructed (manner of articulation)?

Place of Articulation

Manner of Articulation

Putting the three together gives a precise label for any consonant. The English s in sip is a voiceless alveolar fricative; the b in bat is a voiced bilabial plosive; the m in mat is a voiced bilabial nasal. Each label picks out exactly one sound.

The IPA: One Symbol, One Sound

The IPA assigns one symbol to each sound, and one sound to each symbol, so that a transcription is unambiguous. Many symbols are ordinary Latin letters with their usual values; the extras fill the gaps that spelling leaves. Here is the full English consonant set arranged by place and manner. Where a cell has two symbols, the voiceless one is first.

The symbols that are not ordinary letters, with the sound they stand for:

Notice the mismatches with spelling. Sing has four letters but only three sounds, because ng is a single sound: [s], [ɪ], [ŋ]. Thin and this are spelled with the same th but begin with different sounds, one voiceless and one voiced. The IPA j is the sound of y in yes, which is how German, Hungarian and many other languages spell it too.

Vowels: Height, Backness and Rounding

Vowels have no obstruction, so place and manner do not apply. Instead a vowel is described by where the tongue sits and what the lips do:

• Height: how high the tongue is in the mouth (close, mid or open).
• Backness: whether the tongue is pushed forward (front), central or pulled back.
• Rounding: whether the lips are rounded or spread.

Say ee (as in see) and then oo (as in too): your tongue moves from front to back and your lips from spread to rounded. Now say ee and then ah: your jaw drops. Those are the three dimensions at work.

The central vowel ə, called schwa, is the most common vowel in English: it is the unstressed vowel in about, sofa and the first syllable of again. A : after a vowel (as in [iː]) marks a long vowel. In a Scottish accent several of these distinctions are absent: many speakers do not distinguish put from too, or cot from caught, which is an example of how accents differ in the number of vowel sounds, not in how "correct" they are.

Sounds That English Doesn't Have

Each language uses only some of the sounds the human vocal tract can make. A sound that is ordinary in one language may be missing from another, which is a major reason adult learners have a "foreign accent": they substitute the nearest sound of their own language.

The English th sounds are themselves unusual across the world's languages, and many English learners find them hard for the same reason. A Japanese speaker is not "bad at" the English r and l: Japanese has one liquid sound where English has two, and the contrast between them is simply not part of the Japanese sound system. Chapter 3 explains why this makes some differences impossible to hear until you learn them.

Esperanto's Sounds

Esperanto was designed so that each letter has a single, fixed sound, which the IPA makes explicit. The six accented letters from Esperanto Grammar 1, Chapter 1 map to symbols you now know:

Why Spelling Is Not Enough

English spelling is a historical record, not a transcription. The letters ough stand for at least six different sounds in bough, cough, through, though, thorough and enough, and the same sound [ʃ] is spelled sh, ti, ci, ch, s, ss in ship, nation, special, machine, sure, tissue. Hungarian and Esperanto have far more regular spelling, and the IPA treats all three the same way: it records the sound, not the spelling. How writing systems relate to speech is the subject of the course Writing Systems of the World; for now, remember that a transcription in square brackets shows what is pronounced.

Hands-On Exercises