Phonetics and Phonology

Introduction

Phonetics and Phonology examines how human languages use sound. In this university-level aiMOOC, you study both the physical side of speech and the abstract systems that organize speech sounds into meaningful patterns. Phonetics asks how speech is produced, transmitted, and perceived. Phonology asks how languages organize sounds into contrastive units, sequences, alternations, syllables, and prosodic patterns.

The two fields are closely connected. A phonetic difference can become phonologically contrastive, and phonological patterns are constrained by what speakers can articulate and perceive. At the same time, similar phonetic material may be organized differently in different languages. This course therefore moves repeatedly between physical evidence and grammatical analysis.

You will work with the International Phonetic Alphabet, articulatory descriptions, acoustic displays, phonemic and phonetic transcription, distinctive features, distributional evidence, syllable structure, phonological processes, stress, tone, and intonation. You will also use research-style reasoning and introductory speech-analysis methods.

Datei:IPA unitipa (2026).pdf

Learning Objectives

By the end of the course, you should be able to explain the difference between phonetics and phonology, describe speech sounds using standard articulatory terminology, read and produce basic IPA transcriptions, interpret simple acoustic evidence, distinguish phonemes from allophones, use minimal pairs and distributional evidence, formulate basic phonological generalizations, analyze syllable and prosodic structure, and design a small empirical investigation of speech.

Phonetics and Phonology as Complementary Sciences

What Phonetics Studies

Phonetics investigates speech as a physical event. Articulatory phonetics studies how the vocal tract creates speech sounds. Acoustic phonetics studies measurable properties of the speech signal. Auditory and perceptual phonetics studies how listeners receive and interpret that signal. These perspectives can be combined: for example, a change in tongue position alters the resonances of the vocal tract, which alters the acoustic spectrum and may alter what a listener perceives.

What Phonology Studies

Phonology studies systematic sound organization in a language or language variety. It asks which differences create contrasts, where sounds can occur, how sounds alternate across contexts, which segments form natural classes, how syllables are structured, and how stress, tone, and intonation are represented. Phonological analysis aims to state generalizations that explain patterns rather than merely list pronunciations.

A useful distinction is between a phone, an observed speech sound, and a phoneme, a contrastive unit in a phonological system. A phoneme can have multiple context-dependent phonetic realizations called allophones. This distinction is analytical rather than visual: the same IPA symbol may appear in a broad phonemic transcription or a narrow phonetic transcription depending on what the analyst is representing.

Representation Conventions

Linguists conventionally use slashes for phonemic representations, such as /p/, and square brackets for phonetic realizations, such as [pʰ]. Angle brackets may be used for spelling, such as

. This notation helps you keep distinct the orthographic form, the phonological representation, and the observed pronunciation. A good analysis always states what level of representation is intended. Writing [pʰ] claims a phonetic detail, while writing /p/ abstracts away from many predictable details. Neither level is inherently more correct; each answers a different question.

The Speech Production System

The Vocal Tract

Most spoken-language sounds depend on airflow from the lungs, activity at the larynx, and shaping by the supralaryngeal vocal tract. Important articulators include the lips, teeth, alveolar ridge, hard palate, soft palate or velum, tongue, and glottis. The nasal cavity becomes acoustically important when the velum is lowered and airflow can pass through the nose.

When you describe an articulation, distinguish the active articulator, such as the tongue tip or lower lip, from the passive articulator, such as the alveolar ridge or upper teeth. This makes labels such as alveolar, dental, labiodental, palatal, and velar easier to interpret as descriptions of contact or constriction.

Phonation and Voicing

At the larynx, the vocal folds can be configured in different ways. In many voiced sounds, vocal-fold vibration creates a quasi-periodic source. In many voiceless sounds, such vibration is absent for at least part of the relevant interval. Real languages also use more complex phonation types, and the timing of laryngeal gestures can be phonologically important.

Voice onset time is one useful temporal measure in stop consonants. It is commonly defined as the interval between stop release and the onset of periodic voicing. Languages can use different ranges of this timing to realize laryngeal contrasts, so a phonetic measurement should not be equated automatically with one universal phonological category.

Consonants

Describing Consonants

A basic articulatory description of a consonant uses three dimensions: voicing, place of articulation, and manner of articulation. For example, [t] is commonly described as a voiceless alveolar plosive, while [v] is a voiced labiodental fricative. Additional properties may be needed for languages that contrast aspiration, ejectives, implosives, clicks, secondary articulations, or other laryngeal and oral gestures.

Place describes where the main constriction occurs. Manner describes how airflow is shaped, as in plosives, nasals, fricatives, approximants, trills, taps or flaps, and lateral sounds. These labels are not simply vocabulary to memorize; they define dimensions on which phonological patterns can generalize.

Natural Classes of Consonants

Phonological rules often target groups of sounds that share properties. If a process affects /p t k/ but not /m n ŋ/, it may be sensitive to obstruency, continuancy, or laryngeal features rather than to three unrelated segment names. Such sets are called natural classes when they can be characterized by a coherent feature description.

This is why detailed phonetic classification matters for phonology. A feature such as [voice], [nasal], [continuant], or a place feature can express a generalization more economically than a list of individual sounds.

Vowels

Articulatory Dimensions of Vowels

Vowels are commonly described using tongue-body height, tongue-body backness, and lip rounding, with additional dimensions such as length, nasalization, phonation, and tongue-root position where relevant. The IPA vowel chart is an idealized map of these articulatory and auditory relationships, not a literal anatomical picture of the mouth.

A vowel transcribed [i] is high and front with unrounded lips, while [u] is high and back with rounded lips. Central vowels occupy a region between front and back. Actual vowel spaces vary across speakers and languages, and a phonological vowel category may include considerable phonetic variation.

Acoustic Correlates of Vowels

Vowel quality can be investigated acoustically using formants, frequency regions strengthened by vocal-tract resonances. The first two formants, F1 and F2, are especially useful for many vowel analyses. As a broad tendency, F1 relates inversely to vowel height, while F2 often relates to frontness and backness, but interpretation depends on the speaker, vocal-tract dimensions, and the analytical context.

Formant plots often reverse the numerical axes so that the plot resembles the traditional vowel quadrilateral. You should therefore read axis labels carefully rather than assuming that visual position has a simple geometric meaning.

Acoustic Phonetics

Waveforms and Spectrograms

A waveform displays changes in air pressure or an electrical representation of those changes over time. A spectrogram displays how acoustic energy is distributed across frequency and time. Darker regions typically indicate greater energy in a frequency band. Spectrograms make it possible to inspect periodicity, formant structure, stop closures and releases, frication, transitions, and other cues.

No acoustic cue should be interpreted in isolation. Listeners integrate multiple cues, and a cue that is important for one contrast may be less important for another. Acoustic analysis is strongest when measurements are motivated by an explicit hypothesis about production, perception, or phonological structure.

Fundamental Frequency, Duration, and Intensity

Fundamental frequency or F0 is the repetition rate associated with periodic vocal-fold vibration and is a major acoustic correlate of perceived pitch. Duration is crucial for timing contrasts, rhythm, stress, and many segmental distinctions. Intensity relates to signal amplitude and can contribute to prominence, although perceived loudness also depends on frequency and listening conditions.

In prosodic analysis, you should avoid treating any single acoustic measure as equivalent to a category such as stress. Stress can be cued by combinations of duration, F0, intensity, vowel quality, and language-specific patterns.

Auditory and Perceptual Phonetics

Speech perception is not a simple decoding of one acoustic property at a time. Listeners use multiple acoustic cues, linguistic expectations, context, and prior experience. Some experiments test identification, in which listeners label stimuli, while others test discrimination, in which listeners judge whether stimuli differ.

Categorical patterns in identification can be important evidence, but they do not prove that perception is perfectly categorical. Modern experimental work often examines graded responses, cue weighting, adaptation, and variation across listeners.

The International Phonetic Alphabet

The International Phonetic Alphabet provides a standardized system for representing speech sounds. Its consonant chart organizes many pulmonic consonants by place and manner of articulation, while the vowel chart organizes vowel qualities. Additional symbols and diacritics represent non-pulmonic consonants, secondary articulations, phonation, nasalization, length, tone, stress, and other properties.

Use a broad transcription when you want to represent contrastive or major phonetic categories without every predictable detail. Use a narrow transcription when fine phonetic detail is relevant to the analytical question. A narrow transcription is not automatically better; it is better only when the added detail serves the purpose of the analysis.

For the current official chart and chart resources, consult the International Phonetic Association: https://www.internationalphoneticassociation.org/content/ipa-chart

Phonemes, Allophones, and Contrast

Minimal Pairs and Near-Minimal Pairs

A minimal pair consists of two forms that differ in one sound position and have different meanings. Such evidence strongly supports a phonemic contrast in the relevant language or variety. English pairs such as sip and zip support a contrast involving /s/ and /z/ for speakers who distinguish those words.

Near-minimal pairs can also contribute evidence when strict minimal pairs are unavailable, but a good analysis considers distribution, morphology, phonetic similarity, and the possibility that other differences are involved.

Complementary Distribution and Allophony

Two phonetically similar sounds that occur in mutually exclusive environments may be analyzed as allophones of one phoneme if a coherent rule or constraint predicts their distribution. English provides a classic illustration with aspirated and unaspirated voiceless stops in many varieties: [pʰ] is common at the beginning of a stressed syllable, whereas [p] commonly occurs after /s/ in the same syllable. The exact phonetic timing varies by speaker and context.

The decision chart is a useful heuristic, not a substitute for analysis. Complementary distribution alone is not sufficient; phonetic similarity, morphological evidence, historical structure, and the broader phonological system can matter.

Free Variation

Two pronunciations are in free variation when more than one realization can occur in the same environment without a change in lexical meaning. In real speech, apparent free variation may still correlate with speaking style, region, social identity, speech rate, or discourse context. Therefore, "free" does not mean "random" or "socially meaningless."

Distinctive Features and Phonological Rules

Features and Natural Classes

Distinctive-feature systems represent segments as bundles or structures of phonological properties. Different theories propose different feature inventories and organizations, but the analytical goal is similar: to capture contrasts and recurring natural classes.

For example, if a process applies to voiced obstruents, a feature-based statement can target a class defined by properties rather than naming every relevant segment. A good feature analysis should explain why the process groups those segments together.

Rule Notation

A traditional phonological rule states an input, an output, and an environment. A simplified schema is:

A → B / X __ Y

This reads: A becomes B when it occurs after X and before Y. The notation is compact, but the important intellectual work is deciding what A, B, X, and Y must contain so that the rule captures all and only the relevant data.

Rules can describe assimilation, deletion, epenthesis, neutralization, stress assignment, and many other patterns. Contemporary phonological theories may instead use constraints, representations, or probabilistic models, but the basic task remains to explain systematic distribution and alternation.

Phonotactics and Syllable Structure

Phonotactics

Phonotactics concerns restrictions on possible sound sequences and positions. A sequence that is well formed in one language may be disallowed in another. Speakers often adapt unfamiliar sequences by deleting sounds, inserting vowels, changing segments, or reorganizing syllable boundaries. These repairs reveal the structure of the receiving phonological system.

Phonotactic judgments are gradient in some contexts and can be influenced by lexical frequency and experience, so formal restrictions and usage-based evidence should be kept analytically distinct.

Syllables

A common syllable model distinguishes an onset from a rhyme; the rhyme contains a nucleus and, where present, a coda. The nucleus is typically vocalic, but some languages permit syllabic consonants. Languages differ in which onset and coda shapes they allow and in how syllable weight interacts with stress or other processes.

Syllabification is not always recoverable from spelling. You should derive syllable structure from phonological evidence such as phonotactic patterns, allophony, stress, and morphological behavior.

Common Phonological Processes

Assimilation

In assimilation, one sound becomes more similar to another in some phonological property. Place assimilation in nasal consonants is common cross-linguistically because articulatory gestures can overlap. Assimilation may be regressive, influenced by a following segment, or progressive, influenced by a preceding segment.

The best analysis identifies the changing feature, the triggering context, and the domain in which the process applies.

Dissimilation, Epenthesis, and Deletion

Dissimilation makes nearby sounds less similar in a relevant property. Epenthesis inserts a segment, often to repair a disfavored sequence. Deletion removes a segment in a particular environment. These labels describe outcomes; a full analysis must also explain the structural conditions and whether the process is productive in present-day grammar.

Lenition, Fortition, and Neutralization

Lenition is a cover term for changes often described as weakening, such as increased sonority or reduced constriction in certain contexts. Fortition describes changes in the opposite direction. Because these terms cover different phonetic pathways, you should specify the actual change rather than relying on the label alone.

Neutralization occurs when a contrast found in one environment is absent in another. To establish neutralization, compare the relevant distribution and alternations rather than assuming that surface similarity proves an underlying identity.

Prosody: Stress, Tone, and Intonation

Stress

Stress refers to language-specific patterns of prominence. It may distinguish words, organize rhythmic structure, or interact with vowel reduction and morphology. Acoustic correlates can include duration, F0, intensity, and spectral properties, but no single correlate defines stress in every language.

Tone

In a tone language, pitch patterns can distinguish lexical or grammatical meaning. Tone may be represented using level targets, contours, features, or autosegmental associations depending on the analysis. Tone must be distinguished from the broader phenomenon of intonation, even though both involve F0.

Intonation

Intonation concerns phrase- and discourse-level pitch patterns. It can signal boundaries, information structure, pragmatic meaning, and speaker stance. Analyses often distinguish pitch accents from boundary tones, but theoretical conventions vary. You should state the framework you are using and support labels with acoustic and distributional evidence.

From Data to Analysis

A Distributional Workflow

When analyzing an unfamiliar sound pattern, begin by transcribing the data consistently. Identify potentially contrasting sounds, search for minimal or near-minimal pairs, list the environments of each sound, and test whether a natural-class generalization predicts the distribution. Then examine morphological alternations and look for counterexamples.

A strong analysis makes predictions. If you propose that one allophone occurs only before a particular class of segments, test forms outside that environment. If you propose a phonotactic restriction, ask how borrowed words or newly coined forms are adapted.

Avoiding Common Analytical Errors

Do not assume that spelling reveals phonemic structure. Do not treat English categories as universal. Do not infer a phonological contrast from acoustic difference alone. Do not assume complementary distribution automatically implies allophony. Do not write a rule that simply restates a list when a natural class is available.

Your analysis should separate observation from explanation. First state what occurs. Then propose a representation or generalization. Finally test that proposal against the full dataset.

Practical Acoustic Analysis with Praat

Praat is widely used for phonetic analysis. With it, you can inspect waveforms and spectrograms, annotate recordings, estimate F0 and formants, measure duration and intensity, and create figures for research. Measurements require methodological choices such as time windows, frequency ranges, segmentation criteria, and speaker normalization.

A basic student workflow is to record or open a sound, inspect the waveform and spectrogram, label relevant intervals, make a small number of motivated measurements, save those measurements systematically, and report how they were obtained. If you record people for a project, follow your institution's requirements for consent, privacy, storage, and research ethics.

Variation, Social Meaning, and Language Change

Phonetic and phonological systems vary across speakers, regions, social groups, styles, and historical periods. Variation is not noise to be discarded automatically; it may be the phenomenon you need to explain. A variable pronunciation can carry social meaning, while repeated phonetic biases can contribute to sound change over generations.

At the same time, you should avoid stereotyping speakers. Linguistic variation should be investigated with explicit data, careful sampling, and respect for communities. Describing a variety as systematic is more accurate than ranking it against an imagined single "correct" form.

Phonetics-Phonology Interface

The phonetics-phonology interface asks how categorical linguistic patterns relate to continuous physical signals. For example, a phonological feature may have gradient phonetic realizations, and a phonetic cue may support more than one phonological contrast. Laboratory phonology studies such questions experimentally by connecting speech production, perception, and phonological theory.

University-level work in this area often combines corpus data, acoustic measurements, perception experiments, statistical analysis, and formal modeling. The central question is not whether phonetics or phonology is more fundamental, but how evidence at different levels constrains a coherent explanation.

Interactive Tasks

Quiz: Test Your Knowledge

Which statement best distinguishes phonetics from phonology? (Phonetics studies physical speech while phonology studies systematic sound organization) (!Phonetics studies spelling while phonology studies vocabulary) (!Phonetics studies syntax while phonology studies semantics) (!Phonetics studies writing while phonology studies punctuation)

What does a minimal pair primarily demonstrate? (A phonemic contrast in the relevant language or variety) (!A universal spelling rule) (!A difference in sentence structure) (!A historical relationship between two languages)

Which notation normally represents a phonetic realization? (Square brackets) (!Curly brackets) (!Quotation marks) (!Angle brackets)

Which three dimensions are basic to a consonant description? (Voicing place and manner) (!Pitch rhythm and spelling) (!Height stress and tone) (!Meaning syntax and morphology)

What is a formant? (A resonance-related concentration of acoustic energy) (!A unit of spelling) (!A type of phonological rule) (!A punctuation mark)

What is an allophone? (A context-dependent realization of a phoneme) (!A synonym with different spelling) (!A syllable containing two nuclei) (!A word borrowed from another language)

What does complementary distribution mean? (Two forms occur in nonoverlapping environments) (!Two words have opposite meanings) (!Two sounds always create a minimal pair) (!Two syllables receive equal stress)

Which syllable constituent normally contains the nucleus and coda? (Rhyme) (!Onset) (!Morpheme) (!Foot)

What is assimilation? (A process that makes one sound more similar to another) (!A process that always deletes a vowel) (!A rule that changes word meaning) (!A method for measuring formants)

Which acoustic measure is most directly associated with repetition rate in periodic voicing? (Fundamental frequency) (!Orthographic depth) (!Syllable count) (!Lexical frequency)

Memory Game

Phoneme	Contrastive sound unit in a phonological system
Allophone	Contextual realization of a phoneme
Formant	Resonance-related frequency concentration in speech
Onset	Syllable material before the nucleus
Epenthesis	Insertion of a segment into a phonological form
Intonation	Phrase-level organization of pitch patterns

Drag and Drop

Match the correct terms.	Topic
Minimal pair	Evidence for contrast through a one-sound difference in meaning-bearing forms
Complementary distribution	Nonoverlapping environments for two sounds
Spectrogram	Frequency-by-time display of acoustic energy
Natural class	Group of segments characterized by shared phonological properties
Coda	Syllable material following the nucleus

...

Crossword Puzzle

Phoneme	What contrastive sound unit can distinguish words in a language?
Allophone	What term names a contextual realization of a phoneme?
Formant	What resonance-related acoustic feature is important in vowel analysis?
Spectrogram	What display shows acoustic energy across frequency and time?
Assimilation	What process makes one sound more similar to a neighboring sound?
Intonation	What term describes phrase-level pitch organization?

LearningApps

Cloze Text

Open-Ended Tasks

Easy

IPA Sound Inventory: Choose ten speech sounds from a language you know, provide IPA symbols, and describe the articulation of each sound in your own words.
Minimal Pair Collection: Find five minimal or near-minimal pairs in one language variety and explain what phonological contrasts they support.
Vocal Tract Observation: Use a mirror or safe self-observation to compare how your lips and jaw move for three vowels and three consonants, then write a short articulatory report.
Spectrogram Reading: Select a publicly available spectrogram and annotate where you expect periodic voicing, strong formant structure, and any visible transitions.

Standard

Praat Mini Lab: Record a short word list with informed consent where required, measure one motivated variable such as vowel duration or F0, and present the results in a small table.
Allophone Analysis: Analyze a dataset containing two phonetically similar sounds, map their environments, and argue whether they are separate phonemes, allophones, or a case of variation.
Syllable Structure Project: Compare permissible word-initial and word-final clusters in two languages or varieties and propose phonotactic generalizations.
Pronunciation Interview: Interview a speaker about pronunciation variation in their community, avoid prescriptive judgments, and connect at least two observations to phonetic or phonological concepts.

Advanced

Acoustic Vowel Study: Design a small vowel study using F1 and F2 measurements, explain your segmentation and measurement choices, and interpret variation across tokens or speakers.
Phonological Rule Analysis: Work from an unfamiliar dataset to formulate a feature-based rule or constraint, test it against all forms, and discuss at least one possible alternative analysis.
Perception Experiment Design: Propose an identification or discrimination experiment for a sound contrast, define stimuli and dependent measures, and explain how the results would test a phonetic or phonological hypothesis.
Research Video Explanation: Produce a short scholarly video that teaches one phonetics-phonology interface problem using your own examples, diagrams, and evidence, and include a transparent source list.

Learning Assessment

Contrast and Distribution Analysis: Given a novel dataset, determine whether selected sounds are contrastive or allophonic and justify your conclusion with minimal-pair and distributional evidence.
Articulation to Acoustics Transfer: Explain how one articulatory change should affect an acoustic measurement, then identify what additional evidence you would need before making a phonological claim.
Feature Generalization: Replace a list-based rule with a natural-class description and argue why the feature-based analysis is more explanatory.
Syllable and Phonotactic Reasoning: Predict how a language might repair an illegal sound sequence, then compare your prediction with observed adaptation data.
Prosodic Evidence Evaluation: Evaluate a proposed stress or intonation analysis using duration, F0, intensity, and distributional evidence without treating any single measure as definitive.
Mini Research Interpretation: Interpret a small set of acoustic or transcription data, distinguish observation from inference, identify one limitation, and propose a follow-up study.

Evidence of Learning

Evidence of learning should show both conceptual understanding and analytical practice.

Knowledge: You can accurately explain core concepts such as phone, phoneme, allophone, feature, syllable, formant, F0, stress, tone, and intonation.
Transcription skill: You can use IPA symbols and appropriate bracket conventions consistently for the level of analysis required.
Analytical reasoning: You can use minimal pairs, complementary distribution, natural classes, and alternations to support a phonological argument.
Acoustic skill: You can interpret basic waveforms and spectrograms and make simple, methodologically transparent measurements.
Research product: You can produce a short report, annotated dataset, figure, poster, presentation, or video that connects evidence to a phonetic or phonological claim.
Transfer: You can apply the same analytical principles to an unfamiliar language, variety, or dataset rather than relying only on memorized English examples.
Methodological awareness: You can identify uncertainty, variation, ethical requirements, and limitations in a speech study.

OERs on the Topic

The following open resources support deeper study and practical work:

International Phonetic Association IPA Chart: Official chart resources and symbol information.
MIT OpenCourseWare Introduction to Linguistics: Open lectures and notes with university-level phonetics and phonology material.
MIT OpenCourseWare Introduction to Phonology: A dedicated course on phonological grammars, representations, and analyses.
University of Iowa Sounds of Speech: Articulatory animations, videos, and audio examples.
UBC eNunciate: Visual and ultrasound-supported resources for articulatory phonetics.
Praat: Free software and documentation for speech analysis.