Airstreams, Voicing, and Phonation
Chapter Objectives
By the end of this chapter, you will be able to:
- Distinguish pulmonic, glottalic, and lingual airstream mechanisms at an overview level.
- Distinguish egressive and ingressive airflow.
- Distinguish voiced and voiceless production using articulatory evidence.
- Describe modal, breathy, and creaky phonation at an introductory level.
- Use safe self-observation to detect vocal-fold vibration.
- Explain why voicing is not a simple on-off property in every context.
- Classify a set of supplied sounds by airstream mechanism and laryngeal setting and explain the evidence used.
An Observable Problem
Place a fingertip lightly on the front of your throat, just below your jaw, and say a long [zzzz], then switch to a long [ssss] without stopping. Most speakers can feel a buzzing vibration during [zzzz] that disappears during [ssss], even though your tongue and lips barely move between the two sounds. Something is happening lower in your vocal tract, at the larynx, that has nothing to do with where in the mouth the sound is made.
Now try something different. Say a normal [a] vowel, then let your voice drop into a low, creaky register — the kind of sound some speakers use at the end of a long, relaxed sentence, sometimes called "vocal fry." The articulators in your mouth have not changed position at all, but the quality of the sound has changed noticeably. This chapter gives you the tools to describe both of these observations precisely: how airflow is set in motion, and what the larynx does to it once it is moving.
- Before reading further, write one sentence describing, in your own words, what you felt or heard change between [zzzz] and [ssss].
Airstream Mechanisms
Purpose: identify what sets air in motion for speech before describing what happens to that airflow.
Every speech sound requires a moving column of air. An airstream mechanism is the part of the vocal tract that initiates that movement, and the body part responsible for the movement is called the initiator. The vast majority of the world's speech sounds, in every language examined so far, use one mechanism: the lungs.
Pulmonic airstream
A pulmonic airstream mechanism uses the lungs as the initiator, pushing air up through the trachea and into the vocal tract. Nearly every sound you have studied so far in this book, and the large majority of consonants and all vowels in most languages, are pulmonic.
Glottalic and lingual airstreams (overview)
Two other mechanisms exist and are used productively, though less frequently, in many of the world's languages. A glottalic airstream mechanism uses the larynx itself as the initiator, moving a closed or nearly closed glottis upward or downward to compress or rarefy air trapped above it. A lingual airstream mechanism uses the tongue as the initiator, trapping air between two points of closure in the mouth and moving the air by tongue motion alone, independent of the lungs.
At this stage in the course, you only need to recognize that these two additional mechanisms exist and that they are not exotic curiosities confined to a handful of languages — they are regular, systematic parts of the sound systems of many languages around the world. Chapter 13 returns to glottalic sounds (ejectives and implosives) and lingual sounds (clicks) in full detail, including their own dedicated IPA symbols and safe guidance for attempting them. For now, this chapter focuses on the pulmonic mechanism, since it underlies almost everything you will transcribe before Chapter 13.
An airstream mechanism is not the same thing as voicing or place of articulation. It answers a single, narrower question: what part of the body is moving the air? For the rest of this chapter, assume a pulmonic airstream unless stated otherwise.
Egressive and Ingressive Airflow
Purpose: distinguish the direction of airflow from the mechanism that initiates it.
Any airstream mechanism can, in principle, move air outward or inward. Airflow moving out of the vocal tract is egressive; airflow moving into the vocal tract is ingressive. These two terms describe direction only, and they apply independently of which mechanism is doing the initiating.
Virtually all speech in virtually all languages is produced with a pulmonic egressive airstream: lung air moving outward. This is the default combination assumed throughout this book unless a chapter states otherwise. Some languages use a glottalic ingressive airstream for implosives, and lingual airstreams used for clicks are also typically ingressive. You will study these combinations directly in Chapter 13.
Pulmonic ingressive speech does occur, most often as a marginal or stylistic phenomenon, such as gasped speech used for brief interjections in some communities, rather than as a productive part of a regular phoneme inventory. Do not assume that every attested combination of mechanism and direction plays an equally large role in the world's languages.
Voiced and Voiceless Production
Purpose: connect the auditory and tactile observation from this chapter's opening to a precise articulatory description.
Once air is moving through the vocal tract, the larynx, introduced in Chapter 4, can shape that airflow before it reaches the mouth or nose. The vocal folds sit inside the larynx and can be held in different configurations. When the vocal folds are brought close enough together that the moving air causes them to vibrate rapidly, the resulting sound is voiced. When the vocal folds are held apart and do not vibrate, the resulting sound is voiceless.
- [z]voiced — vocal folds vibrating, as felt at the throat during the opening observation
- [s]voiceless — vocal folds apart, no vibration
Voicing is independent of place and manner of articulation, the two systems you will study beginning in Chapter 7. The same place and manner of constriction can typically be produced with or without voicing, which is why languages commonly organize consonants into voiced-voiceless pairs, such as English [s] and [z], or [t] and [d].
Two safe, simple checks can help you detect voicing in your own speech:
- Place a fingertip lightly on the front of your throat and sustain a consonant such as [v] or [f]. A buzzing vibration indicates voicing; its absence indicates voicelessness.
- Cover both ears lightly with your palms and sustain the same sounds. Voiced sounds typically produce a noticeably louder internal buzz than voiceless sounds.
Try both checks on the pair [f] and [v], then on [ʃ] and [ʒ] (the consonants in the middle of pressure and pleasure). Stop immediately if either technique causes any discomfort; ordinary self-observation of your own speech should never strain your throat.
Modal, Breathy, and Creaky Phonation
Purpose: show that the larynx can do more than switch voicing on or off, returning to the creaky-voice observation from the opening of the chapter.
Phonation describes the overall manner in which the vocal folds vibrate, or fail to vibrate, during speech. Voiced and voiceless are the two most basic phonation categories, but the vocal folds can also vibrate in other characteristic ways. At an introductory level, this book distinguishes three phonation types:
- Modal voicethe ordinary, default voiced phonation used for most voiced sounds in most speech; relatively regular vocal-fold vibration
- Breathy voicevocal folds vibrate but remain relatively far apart, allowing audible airflow alongside the voicing; sometimes described as a whispery or airy voice quality
- Creaky voicevocal folds vibrate slowly and irregularly, often described informally as "vocal fry"; common at the ends of phrases in some varieties of English and used contrastively as part of the consonant or vowel system in a number of other languages
The creaky-voice demonstration from the opening of the chapter illustrates a phonation contrast within a single vowel: the articulators did not move, but the laryngeal setting changed from modal to creaky. In English, this kind of phonation shift is typically a matter of style or phrase position rather than a contrast between two different words. In some other languages, however, modal, breathy, and creaky phonation can each define a separate phoneme.
In Jalapa Mazatec, an Otomanguean language of Mexico, modal, breathy, and creaky phonation distinguish separate vowel phonemes rather than functioning only as a stylistic effect, so that otherwise identical vowels with different phonation types form separate words (Silverman et al. 1995). This is a useful reminder that a laryngeal setting English speakers may associate mainly with casual style can carry full phonemic weight in another language's sound system.
Why Voicing Is Not Simply On or Off
Purpose: prevent an oversimplified picture of voicing before later chapters introduce timing-sensitive cases such as aspiration.
It is tempting to treat voicing as a simple switch: either the vocal folds vibrate, or they do not. In practice, voicing interacts with timing, position in the word, and neighboring sounds in ways that this chapter only previews.
Consider aspiration, briefly introduced here for recognition and developed fully in Chapter 9: a period of voicelessness, heard as a small puff of air, that can follow the release of an otherwise voiceless stop. The transcription symbol ʰ marks this period. You do not need to produce or transcribe aspiration accurately yet; you only need to recognize that the timing of voicing relative to a consonant's release, not merely its presence or absence, can matter.
- [pʰ]for recognition only — a voiceless stop released with audible aspiration; developed fully in Chapter 9
A second complication is that voiced consonants are not always voiced for their entire duration in connected speech, and voicing can be partially devoiced near the edges of an utterance or next to a voiceless neighbor. You are not asked to transcribe this kind of partial devoicing yet; the relevant diacritics are introduced for recognition only and developed productively in Chapter 25. For now, the important conceptual point is simply that "voiced" and "voiceless" describe a typical pattern, not an absolute and unchanging property of a sound in every context.
Error: Assuming that every sound spelled with a "voiced letter," such as b, d, or g, is voiced for its entire duration in every context.
Correction: Voicing is a property of articulation, established through observation, not a property guaranteed by spelling. Whether a given token of a voiced consonant is fully voiced throughout depends on its position in the word and utterance, a topic this book returns to directly in Chapters 9 and 23.
Guided Self-Observation
Complete this self-observation task using only your own speech. No audio recording is required, and none of these steps should ever cause discomfort. Stop any step immediately if it feels strained.
- Place a fingertip lightly on the front of your throat. Sustain [f] for two seconds, then switch to [v] for two seconds without stopping. Note which one produces a vibration you can feel.
- Repeat the same comparison with [s] and [z], then with [ʃ] and [ʒ] (the middle consonants in pressure and pleasure).
- Say a relaxed, sustained [a] vowel in your ordinary modal voice. Then let your pitch drop and allow the vowel to become low and creaky, as at the end of a long, relaxed sentence. Note what changed and what stayed the same.
- Write two or three sentences describing, in your own words and without technical jargon you have not yet learned, what you noticed in steps 1 through 3.
Classification Lab
Purpose: apply airstream, direction, and laryngeal classification to a small, verified dataset.
The sounds below are drawn from descriptions of English and Jalapa Mazatec in standard phonetic reference sources. Use only the information in the table and in this chapter to answer the questions that follow.
| Symbol | Language | Brief description |
|---|---|---|
| [z] | English (General American) | Alveolar fricative, vocal folds vibrating throughout |
| [s] | English (General American) | Alveolar fricative, vocal folds held apart, no vibration |
| [pʰ] | English (General American) | Bilabial stop released with audible aspiration, word-initial in a stressed syllable |
| a̤ | Jalapa Mazatec | Low central vowel with breathy phonation, contrasts with modal and creaky counterparts (Silverman et al. 1995) |
| a̰ | Jalapa Mazatec | Low central vowel with creaky phonation, contrasts with modal and breathy counterparts (Silverman et al. 1995) |
Source: Jalapa Mazatec phonation data adapted from Silverman, Daniel, Barbara Blankenship, Paul Kirk, and Peter Ladefoged. 1995. "Phonetic Structures in Jalapa Mazatec." Anthropological Linguistics 37(1): 70–88. English forms reflect standard General American citation pronunciation.
Step 1 — Airstream and direction
All five sounds in this dataset share the same airstream mechanism and the same direction of airflow. Name both, and explain how you know without relying on a textbook definition alone.
Step 2 — Voicing
Classify [z], [s], and [pʰ] as voiced or voiceless. For [pʰ], identify which part of the sound is voiceless even though the surrounding vowel is voiced.
Step 3 — Phonation beyond voicing
Compare a̤ and a̰. Both are voiced. What additional phonation distinction separates them, and why does the dataset description note that they "contrast," rather than simply describing them as two possible pronunciations of the same vowel?
Step 4 — Write the analysis
In two to three sentences, explain why this dataset shows that phonation involves more than a single voiced-versus-voiceless choice. Support your claim using at least one specific form from the table.
Common Mistakes
Error: Treating "airstream mechanism" and "place of articulation" as the same kind of information, for example describing a sound as having a "pulmonic place."
Correction: Airstream mechanism answers what initiates the airflow; place of articulation, covered beginning in Chapter 7, answers where in the vocal tract a constriction occurs. A single sound is described using both systems together, but they are independent questions.
Error: Assuming that creaky or breathy voice is always a sign of vocal strain, illness, or incorrect pronunciation.
Correction: Creaky and breathy phonation are normal, systematic features of human speech. They function stylistically in some languages and varieties and contrastively as part of the phoneme system in others, such as Jalapa Mazatec. Neither pattern is a speech disorder or a mispronunciation.
Error: Concluding that the glottalic and lingual airstream mechanisms are rare or marginal because this chapter spends limited time on them.
Correction: Glottalic and lingual airstreams are regular, systematic parts of the sound systems of many languages. This chapter previews them briefly because their full productive treatment, including safe attempts at production, is scheduled for Chapter 13.
Chapter Review
Part A — Definitions
1. Define airstream mechanism and name the three types introduced in this chapter.
2. Define egressive and ingressive, and state which combination of mechanism and direction is most common across the world's languages.
3. Define voiced and voiceless in articulatory terms, without referring to spelling.
Part B — Phonation
1. Name and briefly describe the three phonation types introduced in this chapter.
2. Using the Jalapa Mazatec example from this chapter, explain why phonation can be a phonemic property of vowels in some languages even though it is mainly stylistic in English.
Part C — Applying the Concepts
1. Using the self-observation techniques from this chapter, predict whether English [θ] (as in thin) and [ð] (as in this) are voiced or voiceless, and describe the evidence that would let you check your prediction safely.
2. Explain, in your own words, why aspiration shows that voicing involves timing as well as presence or absence.
Cumulative Review
1. Using the anatomy vocabulary from Chapter 4, identify which structure is primarily responsible for the vibration you felt during the self-observation task in this chapter.
2. Using the phone-phoneme-allophone framework from Chapter 5, explain whether the modal-versus-creaky distinction in English [a] from this chapter's opening scenario is best treated, at an introductory level, as a phonemic contrast or as a difference between two phones of the same phoneme. Justify your answer.
Chapter Performance Task
Performance Task
A Voicing and Phonation Field Note
Choose three English consonant pairs that differ mainly in voicing (for example, [f]/[v], [s]/[z], and [ʃ]/[ʒ]). Working independently:
- For each pair, describe the self-observation evidence — tactile or auditory — that lets you classify each member as voiced or voiceless.
- Identify one moment in your own ordinary speech, such as the end of a relaxed sentence, where you notice modal voice shifting toward creaky voice, and describe what changed.
- State, in your own words, why this kind of stylistic phonation shift in English is different from the phonemic phonation contrasts found in a language such as Jalapa Mazatec.
Answer Key (click to expand)
Classification Lab
Step 1: All five sounds are pulmonic egressive: each is produced with lung air moving outward through the vocal tract. This can be confirmed by the fact that all five are ordinary sounds produced during normal exhalation, with no special closure or movement of the larynx or tongue acting as an independent initiator.
Step 2: [z] is voiced; [s] is voiceless. [pʰ] is voiceless at the moment of closure and release, including during the aspiration itself; voicing for the following vowel begins only after the aspiration ends.
Step 3: Both a̤ and a̰ are voiced, but they differ in phonation type: a̤ is breathy, with audible airflow accompanying relatively open vocal-fold vibration, while a̰ is creaky, with slow, irregular vocal-fold vibration. The dataset describes them as contrasting because, in Jalapa Mazatec, substituting one phonation type for another on an otherwise identical vowel can produce a different word, meaning the distinction is phonemic rather than a matter of free stylistic variation.
Step 4 (model answer): This dataset shows that voicing alone cannot capture every laryngeal distinction relevant to language, because a̤ and a̰ are both voiced yet are phonemically distinct in Jalapa Mazatec, differing only in phonation type. A complete phonetic description therefore needs a more detailed phonation category, not just a voiced versus voiceless label. Other well-reasoned answers that correctly use the dataset are acceptable.
Chapter Review — Part A (model answers)
1. An airstream mechanism is the part of the vocal tract that initiates the movement of air for speech; the three types introduced are pulmonic, glottalic, and lingual. 2. Egressive airflow moves outward from the vocal tract; ingressive airflow moves inward. Pulmonic egressive airflow is by far the most common combination across the world's languages. 3. Voiced sounds are produced with the vocal folds vibrating; voiceless sounds are produced with the vocal folds held apart and not vibrating. This distinction is established through articulatory observation, not spelling.
Chapter Review — Part B
1. Modal voice is the ordinary, default voiced phonation with relatively regular vocal-fold vibration; breathy voice involves vibration with the folds held relatively far apart, allowing audible airflow; creaky voice involves slow, irregular vibration, often called vocal fry. 2. In Jalapa Mazatec, modal, breathy, and creaky phonation can each define a separate vowel phoneme, so changing phonation alone can change a word's meaning. In English, a comparable phonation shift, such as modal voice becoming creaky at the end of a sentence, is a matter of style or phrase position and does not change which word is being said.
Chapter Review — Part C
1. Both [θ] and [ð] should be checked, not assumed: a learner can place a fingertip on the throat or cover the ears while sustaining each sound. For most speakers, [ð] (as in this) is voiced and [θ] (as in thin) is voiceless, but the chapter's point is that this should be confirmed by observation rather than assumed from spelling. 2. Aspiration shows that a stop can be articulated with the vocal folds apart for a short period after release, even though the same stop is not described simply as "voiceless" in every sense throughout its entire duration; the relevant question becomes when voicing begins relative to the release, not only whether voicing eventually occurs.
Cumulative Review
1. The larynx, and specifically the vocal folds within it, introduced in Chapter 4, is primarily responsible for the vibration felt during the self-observation task. 2. At an introductory level, the modal-versus-creaky distinction in English [a] is best treated as a difference between two phones of the same phoneme, since the shift is conditioned by phrase position and style rather than by a contrast that changes word meaning; English does not appear to use this distinction to distinguish separate words the way Jalapa Mazatec does.
Looking Ahead
What You Can Now Do
You can now distinguish how airflow is initiated from what the larynx does to that airflow, use safe self-observation to detect voicing, and recognize that phonation involves more than a simple voiced versus voiceless choice. These distinctions will be active throughout the consonant chapters that follow.