Speech, Sound, and Writing
Learning Objectives
By the end of this chapter, you will be able to:
- Distinguish speech as a physical, acoustic event from writing as a conventional symbol system.
- Explain why writing systems cannot perfectly represent pronunciation.
- Identify multiple types of mismatch between letters and sounds.
- Describe what different writing systems represent — sounds, syllables, or meanings.
- Use angle brackets ⟨ ⟩, slashes / /, and square brackets [ ] correctly to distinguish orthographic, phonemic, and phonetic representation.
- Explain why linguists developed the International Phonetic Alphabet.
- Identify situations where spelling provides misleading information about pronunciation.
Opening Observation
You already know how to read. That skill is so automatic that it can actually get in the way of hearing speech clearly. Consider the word knight. When you read it, you probably hear one syllable and one initial consonant — not a k, not a gh. Now consider why a beginning student of English, looking at the letters k-n-i-g-h-t, might try to pronounce six distinct sounds. The spelling is not the word. The word is what speakers say.
- Write down five English words where at least one letter is silent. How did you know which letter to ignore?
- Try to write the sound of a sneeze, a click of disapproval, or the sound made by a dripping faucet, using only the letters of the English alphabet. What problems did you encounter?
- If you know another language, think of one word whose spelling a native English speaker would almost certainly mispronounce. What makes it difficult?
Speech Came Before Writing
Human beings have been speaking for at least 100,000 years, and perhaps much longer. Writing, by contrast, is a recent invention — the oldest known writing systems appeared only about 5,000 years ago in Mesopotamia and Egypt. For at least 95,000 years, language existed entirely as speech and gesture: no alphabet, no characters, no page.
This historical fact has a direct consequence for phonetics. Writing did not create language, and it did not create the sounds of language. Writing is a technology that attempts to represent language — one that different civilizations invented independently, in different ways, with different degrees of accuracy, and for different purposes.
Anatomically modern humans
The vocal tract capable of producing a wide range of speech sounds is in place. Language, in some form, almost certainly predates writing by hundreds of millennia.
Earliest writing systems
Cuneiform in Mesopotamia and hieroglyphics in Egypt emerge as record-keeping and administrative tools — not as transcriptions of speech designed for phonetic accuracy.
Early alphabets
Phoenician develops a system in which symbols correspond to consonant sounds rather than syllables or whole words — a new approach to representing speech.
The International Phonetic Association
Phoneticians establish a systematic, language-independent alphabet designed to represent any sound in any human language. For the first time, a writing system is explicitly designed to represent phonetics, not just to serve one language community.
Speech is primary; writing is secondary. In linguistics, this is not a value judgment — it is a historical and developmental fact. Children learn to speak before they learn to write. Languages can exist for millennia without any writing system. No writing system, however carefully designed, can capture everything speakers do. Phonetics is the study of speech, not the study of spelling.
Languages Without Writing
Linguists estimate there are between 6,000 and 7,000 languages spoken in the world today. Of these, only a few hundred have a substantial writing tradition, and only a few dozen have large bodies of written literature extending back more than a century or two. Most of the world's languages exist primarily or entirely as spoken systems.
A language without a writing system is not a primitive, incomplete, or deficient language. The Pirahã language of the Brazilian Amazon, the ǀXam language of the Kalahari, and thousands of others are fully complex, expressive, and rule-governed systems capable of conveying any idea a speaker wishes to communicate. They simply do not use writing — or did not until linguists and community members developed writing systems for documentation and education.
Sign languages offer a powerful demonstration that language is fundamentally independent of any external medium. American Sign Language (ASL), British Sign Language (BSL), and the hundreds of other sign languages worldwide are complete, grammatically sophisticated languages in the visual-gestural modality. They are not invented or artificial — they emerge naturally in communities of deaf signers. No writing system captures sign language in a fully adequate way, yet no linguist would describe ASL as an "incomplete" language. Language is the system; writing — and signed notation systems — are secondary representations of it.
The implication for this course is fundamental: when you study the sounds of Spanish, Mandarin, Arabic, or any other language, you are studying something that existed long before any writing system tried to represent it. The phonetics of a language is not contained in the alphabet. You have to listen.
What Writing Systems Actually Do
Writing systems differ in what they choose to represent. Once you understand this, many puzzles about spelling — across many languages — become much clearer.
Alphabetic
Symbols represent individual consonant and vowel sounds (or attempt to). English, Greek, Russian, and most European scripts work this way. The match between symbols and sounds varies widely in accuracy.
Syllabic (Syllabary)
Symbols represent whole syllables rather than individual sounds. Japanese kana (hiragana and katakana) work this way. Each symbol covers one consonant-vowel combination rather than the consonant or vowel alone.
Abjad (Consonantal)
Symbols represent consonants; vowels are omitted or written as optional marks. Arabic, Hebrew, and Aramaic work this way. Readers infer vowels from context and familiarity with the language.
Logographic / Morphosyllabic
Symbols represent morphemes or words, often with phonetic components embedded. Chinese characters work this way. A single character can represent a syllable and a meaning simultaneously.
Alphabetic writing systems differ in how consistently they match letters to sounds. Linguists describe this variation as orthographic depth. A shallow orthography (such as Finnish or Spanish) maps letters to sounds quite consistently, with few exceptions. A deep orthography (such as English or French) has accumulated many irregularities through historical change, borrowing, and the failure of spelling reform. The depth of an orthography affects how difficult it is for children to learn to read — and how much a phonetician must distrust the spelling when analyzing speech.
The key point: no writing system was designed by a committee of phoneticians sitting down to represent every acoustic detail of human speech. Writing systems evolve historically, get borrowed across languages and cultures, and are constantly playing catch-up with the actual sounds of the language.
The Many-to-Many Problem
One of the most important things a phonetician learns to recognize is that the relationship between letters and sounds is not one-to-one. It is systematically many-to-many: one letter can represent many sounds, and one sound can have many spellings. The following examples are all from English, a notoriously deep orthography — but every alphabetic language shows some version of this pattern.
One Letter, Many Sounds
The letter ⟨c⟩ in English can represent at least three distinct sounds
The letter ⟨a⟩ appears in strikingly different pronunciations
One Sound, Many Spellings
The sound [iː] (as in "feet") is spelled in at least ten ways in English
The sound [ʃ] (as in "shoe") appears with many different spellings
Misconception: "This word has six letters, so it must have six sounds."
Reality: The number of letters in a word tells you almost nothing about the number of sounds it contains — especially in English. Strength has eight letters but six sounds. Ax has two letters but three sounds [æks]. Queue has five letters but two sounds [kjuː]. A phonetician counts sounds, not letters.
The many-to-many relationship is not unique to English. It appears in every alphabetic language to some degree, because all alphabetic writing systems were developed for one historical state of the language and then carried forward as the sounds changed over centuries. The notation we need for phonetics — for accurately representing speech — must be free from this accumulated inconsistency.
Orthography versus Pronunciation
Looking at words across different languages reveals quickly that the same letter can mean very different things in different orthographic systems. The letter ⟨j⟩ is a useful example:
| Language | Example word | What ⟨j⟩ sounds like | IPA symbol |
|---|---|---|---|
| English | jump | like the first sound in "judge" | [dʒ] |
| Spanish | jardín (garden) | like a strongly breathed ⟨h⟩ — velar or uvular fricative | [x] or [χ] |
| German | ja (yes) | like English ⟨y⟩ in "yes" | [j] |
| French | je (I) | like the middle sound in "measure" | [ʒ] |
| Portuguese | janela (window) | similar to French ⟨j⟩ | [ʒ] |
When you read a word in a foreign language using your native spelling intuitions, you will almost always mispronounce it. This is not a failure of intelligence or effort. It is the predictable consequence of treating one orthographic system as if it were universal — which it is not.
English spelling was largely standardized in the fifteenth and sixteenth centuries, largely through the influence of early printers. By that point, many sound changes had already occurred that the spelling did not follow. The word knight was actually pronounced with an initial [k] and a final [xt] (a velar fricative) in Middle English — the letters reflected real sounds. The sounds changed; the spelling did not. This historical "freezing" of spelling while pronunciation continues to evolve is one of the main reasons English orthography is so irregular today.
The Three Levels of Representation
Phoneticians and linguists use a precise notation system to distinguish among three fundamentally different things that people often confuse: the way a word is spelled, the sound categories that make the word meaningful in a language, and the detailed physical realization of those sounds when an actual speaker produces them.
These three levels require three different types of notation. Learning to use them correctly is one of the most important foundational skills in this course.
Level 1 — Orthography
⟨ ⟩ Angle BracketsUsed to represent written letters and spellings — what appears on the page. Angle brackets signal that you are discussing a symbol in a writing system, not a sound.
Level 2 — Phonemic
/ / SlashesUsed for phonemic representations — the abstract sound categories of a particular language that distinguish meaning. Also called "broad" transcription.
Level 3 — Phonetic
[ ] Square BracketsUsed for phonetic transcription — the actual, detailed physical realization of sounds by a particular speaker in a particular context. Also called "narrow" transcription.
Note that ⟨ough⟩ is spelled the same way across many words — but represents completely different sounds in each one: though, through, thought, thorough, cough, rough, hiccough.
Using the Notation Correctly
The three delimiter types are not interchangeable. Using the wrong one signals a claim you did not intend to make.
| If you want to discuss… | Use… | Example |
|---|---|---|
| A letter or spelling | Angle brackets ⟨ ⟩ | The word is spelled ⟨knight⟩. |
| A phoneme — a sound category that distinguishes meaning | Slashes / / | The first phoneme is /n/. |
| A phone — an actual, physically realized sound event | Square brackets [ ] | The final consonant may be realized as [tʰ] or [t̚]. |
Error: Writing ordinary English spelling inside slashes or square brackets — for example, /knight/ or [knight].
Correction: Slashes and square brackets contain IPA symbols, not ordinary letters. Putting standard spelling inside phonetic delimiters implies that the letters directly represent sounds — exactly the confusion this notation system is designed to prevent. The spelling goes in angle brackets: ⟨knight⟩. The phonemic representation goes in slashes using IPA: /naɪt/.
Why Linguists Need the IPA
By the 1880s, phoneticians and language teachers — especially in France and Britain — were confronting the same problem from different directions. Teachers wanted to teach accurate pronunciation without confusing students through spelling. Linguists wanted to describe the sounds of languages around the world without being constrained by any particular alphabet. Both groups needed the same thing: a notation system that could represent any sound in any language consistently, systematically, and without ambiguity.
The solution was the International Phonetic Association, founded in Paris in 1886, and its International Phonetic Alphabet, first published in 1888. The IPA was designed on several explicit principles:
- One symbol, one sound. Each symbol represents one specific phonetic value, and each phonetic value has one dedicated symbol. This eliminates the one-to-many and many-to-one problems of ordinary spelling.
- Symbols are based on familiar Latin and Greek letters where possible, with new shapes created for sounds that have no familiar alphabetic representation. This makes the system learnable without requiring entirely new visual vocabulary.
- The system is language-neutral. No one language's orthographic conventions are privileged. The same symbol means the same sound whether you are transcribing Swahili, Finnish, or Japanese.
- The system is expandable. Diacritics can add phonetic detail beyond what the base symbols represent. This allows transcribers to be as precise as their purposes require.
One of the most powerful demonstrations of why the IPA is necessary: look at the spelling sequence ⟨ough⟩ in English. This single four-letter sequence represents at least seven different pronunciations:
| Word | Pronunciation | IPA |
|---|---|---|
| though | rhymes with "go" | [ðoʊ] |
| through | rhymes with "blue" | [θruː] |
| thought | rhymes with "caught" | [θɔːt] |
| cough | rhymes with "off" | [kɔːf] |
| rough | rhymes with "stuff" | [rʌf] |
| bough | rhymes with "cow" | [baʊ] |
| hiccough | rhymes with "pick-up" | [ˈhɪkʌp] |
Each IPA transcription is unambiguous. The spelling is not. This is exactly the problem the IPA was designed to solve.
The IPA is not a pronunciation guide in the style of a tourist phrasebook. It is not a decorative respelling system. It is not a spelling reform proposal for any particular language. And it is not simply a list of symbols to memorize. The IPA is an analytical system — a set of tools for making precise, language-independent claims about the physical properties of speech. Learning the IPA means learning to think about sounds differently, not just memorizing new symbols for sounds you already know.
Practice Activities
Speech or Writing?
For each item below, decide whether it is a property of speech, a property of writing, or potentially both. Be prepared to explain your reasoning.
- The word cat has three letters.
- Two words rhyme when they share the same vowel and final consonant sounds.
- The word through contains the sequence ⟨ough⟩.
- A baby babbles before producing words.
- Capitalization signals the beginning of a sentence.
- Some speakers pronounce the final consonant in sing and the one in sin differently.
- A comma creates a pause in reading.
- A speaker's voice rises at the end of a question.
- The letters ⟨gh⟩ are silent in night.
- A word can be stressed on different syllables to change its meaning or grammatical function.
Using Angle Brackets, Slashes, and Square Brackets
Rewrite each sentence replacing the underlined segment with the correct notation. Use angle brackets ⟨ ⟩ for spellings, slashes / / for phonemic representations (when IPA is provided), and square brackets [ ] for phonetic representations (when IPA is provided).
- The word phone begins with the letter p.
- In many English accents, the sound ʃ appears in the word shoe.
- The spelling tion often represents the sound ʃən.
- The phoneme k can be spelled as c, k, ck, or ch in English.
- The phonetic symbol ɑ looks similar to but is not the same as the letter a.
Counting Sounds, Not Letters
For each word below, write the number of letters and estimate the number of distinct sounds. (Do not worry about IPA yet — just think about what you hear when you say the word.)
| Word | Number of letters | Approximate number of sounds | Where letters and sounds differ |
|---|---|---|---|
| through | 7 | ____ | ____ |
| mix | 3 | ____ | ____ |
| strength | 8 | ____ | ____ |
| she | 3 | ____ | ____ |
| exit | 4 | ____ | ____ |
| queen | 5 | ____ | ____ |
The Letter ⟨j⟩ Across Languages
Based on the table in this chapter, answer the following questions:
- A native English speaker encounters the Spanish word jardín. What pronunciation error are they most likely to make, and why?
- A German speaker sees the English word jump. If they apply German spelling conventions, what would they produce instead of the correct English pronunciation?
- What is the only practical solution to these cross-language spelling traps? Explain in one or two sentences.
Writing Systems and Speech
Based on the information in this chapter, respond to each question in two to four sentences.
- A student argues that Japanese kana is a "more accurate" writing system than English spelling because each symbol corresponds to a fixed syllable. Do you agree? What phonetic information does kana still fail to represent?
- Someone tells you that a language "has no grammar" because it has never been written down. How would you respond, drawing on this chapter?
- Why might a linguist find the Arabic abjad system particularly interesting as a demonstration of the relationship between writing and spoken language?
Look at the following group of words. All of them contain the letter sequence ⟨ea⟩. Say each one aloud and listen carefully. How many different sounds does ⟨ea⟩ represent across this list?
head · heat · heart · heard · great · idea · ocean · real · thread · break
Write down each distinct vowel sound you hear. Then consider: if you were a foreign speaker reading these words for the first time, how would you know which pronunciation to use? What does this exercise demonstrate about the need for phonetic notation?
Observation Portfolio — Entry 2
Observation Portfolio
Documenting the Speech-Writing Gap
The Observation Portfolio begun in Chapter 1 continues here. This entry builds your growing record of phonetic observations with material that you can reference throughout the course.
- Spelling-trap collection. Find five words in any language you know where spelling and pronunciation diverge in an interesting way. For each word, record: (a) the spelling in angle brackets, (b) a brief description of what the spelling might suggest to an outsider, and (c) a description in plain words of what it actually sounds like. You do not need IPA yet.
- Writing system encounter. Find a written word or short phrase in a writing system unfamiliar to you — an Arabic word, a Chinese character, a Japanese hiragana sequence, a Devanagari syllable. Note: (a) what type of writing system it is, according to this chapter, (b) what information about pronunciation you can and cannot recover from the symbols, and (c) what additional information you would need to know how to say it.
- Notation practice. Select any five English words and write the following for each: the spelling in angle brackets ⟨ ⟩, a brief description of how many sounds you think it contains, and one observation about a letter that does not correspond straightforwardly to a sound. Do not use IPA yet — just describe what you notice.
- Cross-language reflection. If you know any language other than English, write two or three sentences describing the relationship between its spelling and pronunciation. Is it more or less consistent than English? Are there sounds that the writing system cannot represent? This reflection can be informal but should use the vocabulary of this chapter.
Chapter Review
Part A — Core Concepts
1. In your own words, explain why linguists say that speech is "primary" and writing is "secondary." What does this mean, and what evidence supports it?
2. What is orthographic depth? Name one language with a shallow orthography and one with a deep orthography, and explain what makes them different.
3. Explain the difference between an alphabetic writing system, a syllabary, an abjad, and a logographic system. Give one example of each.
4. What is the "many-to-many problem" in the relationship between letters and sounds? Provide one example of one letter representing multiple sounds, and one example of one sound being spelled multiple ways.
5. What principles did the founders of the IPA use when designing the alphabet? Which of these principles would be impossible for a natural language orthography to satisfy, and why?
Part B — Notation
1. What does each of the following notation types represent? Give an example of each in use.
- Angle brackets ⟨ ⟩
- Slashes / /
- Square brackets [ ]
2. Identify the error in each of the following and correct it:
- (a) The word is spelled /phone/.
- (b) The first sound in city is spelled ⟨s⟩.
- (c) The letter ⟨ʃ⟩ appears at the start of shoe.
3. What is the difference between a phoneme and a phone? Which level of representation corresponds to each notation type?
Part C — Application
1. The word rough is spelled ⟨rough⟩. Explain, using the ideas from this chapter, why that spelling gives an unreliable guide to pronunciation for a foreign learner of English.
2. A student studying Spanish says: "I already know the alphabet, so learning to read Spanish should be easy." Based on the concepts in this chapter, what is valid about this claim, and what important caveat would you add?
3. Why is the claim "Language X has no grammar because it has never been written down" incorrect? Use evidence from this chapter to explain your answer.
Self-Assessment
Before moving to Chapter 3, check that you can do each of the following. If any item feels uncertain, review the relevant section before continuing.
- I can explain, in plain English, why speech is considered primary and writing secondary in linguistics.
- I can describe at least two ways in which writing systems fail to accurately represent speech.
- I can correctly use angle brackets ⟨ ⟩ when referring to spellings and letters.
- I can correctly use slashes / / when referring to phonemic representations.
- I can correctly use square brackets [ ] when referring to phonetic transcriptions.
- I understand that the same letter can represent different sounds in different languages.
- I can name the four main types of writing system and give one example of each.
- I can explain what the International Phonetic Alphabet is and why it was invented.
- I have completed my Observation Portfolio entry for this chapter.
Answer Key
Exercise A — Speech or Writing?
- Writing. The number of letters is a property of orthography, not of the spoken word.
- Speech. Rhyme is defined by shared sounds, not shared spellings (cf. through and blue do not share spelling but share vowel sound).
- Writing. The letter sequence ⟨ough⟩ is a property of the orthographic representation.
- Speech. Babbling occurs in the oral-motor domain before any exposure to written language.
- Writing. Capitalization is a convention of written systems with no direct spoken correlate.
- Speech. The distinction between final consonant sounds is phonetic.
- Both — with important qualifications. A comma signals a potential pause in reading aloud, but readers do not always pause at commas, and spoken language can create pauses through prosody where there is no comma in a written version.
- Speech. Rising intonation is a prosodic feature of spoken language. Writing may represent it with a question mark, but only approximately.
- Writing. Silence is a property of the letters — the sounds those letters once represented are no longer produced.
- Speech. Stress is a prosodic feature realized in sound through pitch, duration, and loudness.
Exercise B — Notation
- The word ⟨phone⟩ begins with the letter ⟨p⟩.
- In many English accents, the sound [ʃ] appears in the word shoe.
- The spelling ⟨tion⟩ often represents the sound [ʃən].
- The phoneme /k/ can be spelled as ⟨c⟩, ⟨k⟩, ⟨ck⟩, or ⟨ch⟩ in English.
- The phonetic symbol [ɑ] looks similar to but is not the same as the letter ⟨a⟩.
Exercise C — Counting Sounds
Approximate answers; exact counts vary by accent and analysis.
- through: 7 letters, 3 sounds — ⟨th⟩=1 sound, ⟨r⟩=1, ⟨ough⟩=1 (the long [uː] vowel)
- mix: 3 letters, 4 sounds — ⟨x⟩ represents two sounds [ks]
- strength: 8 letters, 6 sounds — the cluster at the end is complex but counts as approximately 6 phonemes: /s.t.r.ɛ.ŋ.θ/
- she: 3 letters, 2 sounds — ⟨sh⟩ is one sound
- exit: 4 letters, 5 sounds — ⟨x⟩ represents [gz] in this word
- queen: 5 letters, 4 sounds — ⟨qu⟩ represents [kw], a sequence of two sounds
Exercise D — Cross-Language Observation
- A native English speaker would likely pronounce the ⟨j⟩ in jardín as [dʒ] — the English value of the letter. The correct Spanish pronunciation is [x] or [χ], a velar or uvular fricative. The error arises from transferring English spelling conventions to Spanish.
- A German speaker applying German conventions would pronounce ⟨j⟩ as [j] — like English ⟨y⟩ in yes. So jump would begin with [j] rather than [dʒ].
- The only reliable solution is to learn the specific sound value of letters in each language's orthography — or better, to use a language-neutral phonetic notation (like the IPA) that makes no assumptions drawn from any single orthographic tradition.
Chapter Review — Part B (Notation Errors)
- (a) Error: ordinary spelling in slashes. Correction: The word is spelled ⟨phone⟩.
- (b) Error: IPA symbol in angle brackets. Correction: The first sound in city corresponds to the letter ⟨c⟩ but is realized as [s].
- (c) Error: An IPA symbol (ʃ) should never be placed inside angle brackets — angle brackets are for orthographic symbols. Correction: The letter ⟨sh⟩ appears at the start of shoe, representing the sound [ʃ].
Key Terms
Looking Ahead
What You Now Understand
You have established something essential: spelling is not sound. The letter is not the phoneme. The phoneme is not the phone. Every time you encounter a word in any language, you now know to ask: "What does this spelling actually tell me about pronunciation, and what does it not tell me?" That skepticism toward orthography is the beginning of phonetic thinking.