Sounds and Tone: How Yorùbá Means

Yorùbá is a tone language. Pitch is not decoration on top of a word, it is part of the word, in the same way that a consonant is. The syllables ra, rá, and rà are made of the same consonant and the same vowel, and they are three different words: to rub, to disappear, and to buy 1. Getting the tone wrong is not an accent, it is saying a different word, which is why writing Yorùbá without tone marks is closer to writing English without vowels than to writing English without italics.

This file lays out the sound inventory and then the tone system in detail, because tone is the single thing that most often defeats a heritage speaker returning to the written language.

The sound inventory

Vowels

Standard Yorùbá has seven oral vowels and five nasal vowels 23. The oral vowels are written i, e, ẹ, a, ọ, o, u. Two of these carry a subdot, and the subdot is a letter-distinguishing mark, not an accent:

Written	IPA	Approximate English guide
i	[i]	as in machine
e	[e]	close to the vowel in they, without the glide
ẹ	[ɛ]	as in bet
a	[a]	as in Spanish casa
ọ	[ɔ]	as in thought (British)
o	[o]	close to the vowel in go, without the glide
u	[u]	as in rule

The pairs e / ẹ and o / ọ are separate vowels distinguishing separate words. Dropping the subdot merges distinct words, which is the same category of error as dropping tone marks.

The five nasal vowels are conventionally written with a following n: in, ẹn, an, ọn, un 2. Here the n is not a consonant to be pronounced separately, it marks that the vowel itself is nasalised, air passing through the nose. This is why ìwọ̀n (measure) is two syllables and not three. Note that dialects differ in their vowel systems: several Central and Southeast varieties have nine oral vowels where Northwest Yorùbá and the standard have seven 3.

Consonants

Yorùbá has eighteen consonants: b, d, f, g, gb, h, j, k, l, m, n, p, r, s, ṣ, t, w, y 2. Three deserve comment because they trip up readers who assume English values.

p is not the English p. It represents [k͡p], a doubly articulated labial-velar stop made by closing at the lips and at the soft palate simultaneously and releasing them together 1. There is no plain [p] in native Yorùbá words.

gb is its voiced counterpart [ɡ͡b], made the same way with voicing. It is one sound and one letter of the alphabet, not a g followed by a b. English has nothing like it, and it is worth spending time on.

ṣ is [ʃ], the sh sound, distinguished from plain s [s]. The subdot again marks a separate letter.

The alphabet as conventionally taught has 25 letters, and it has no c, q, v, x, z, and no plain p in the English sense.

Syllable structure

Yorùbá syllables come in three shapes: consonant plus vowel (CV), a vowel alone (V), and a syllabic nasal (N) 2. There are no consonant clusters and no syllable-final consonants in native words. This is why English loanwords are restructured on entry: school becomes sùkúù, bread becomes búrẹ́dì, each consonant getting its own vowel.

The syllabic nasal is a nasal consonant that is a syllable on its own, carrying its own tone 1. It appears in the continuous aspect marker ń and in the negative kò/ò, and in words like ǹkan (thing). Because it bears tone independently, it is written with a tone mark.

A phonological point that becomes important in etymology: nearly all native Yorùbá nouns begin with a vowel 1. Consonant-initial nouns behave as though they have an underlying vowel prefix that surfaces in genitive constructions 1. This is a structural fact about how Yorùbá builds nouns, and it explains the vowel prefixes discussed in yoruba-etymology-and-word-formation.

The tone system

Yorùbá has three phonemically distinctive tones: High (H), Mid (M), and Low (L) 1. In the standard orthography, High is written with an acute accent, Low with a grave accent, and Mid is left unmarked 4.

• High: á é ẹ́ í ó ọ́ ú
• Mid: a e ẹ i o ọ u (unmarked)
• Low: à è ẹ̀ ì ò ọ̀ ù

Tones are relative, not absolute. A High tone is high relative to that speaker's pitch range in that utterance, not a fixed frequency. A child and an adult both produce the same three-way contrast in different absolute registers.

Minimal pairs, worked

The following set is from Akinlabi and Liberman's survey of Yorùbá tonal phonology and shows the contrast systematically, first on monosyllables, then across the nine possible tone patterns of a disyllable 1.

Monosyllabic, all three tones on one segmental shape:

Word	Tone	Meaning
rá	H	to disappear
ra	M	to rub
rà	L	to buy

Three unrelated verbs. Nothing but pitch separates them.

Disyllabic, showing that tone patterns combine freely 1:

Word	Tone pattern	Meaning
ọkọ́	MH	hoe
ọkọ	MM	husband
ọkọ̀	ML	vehicle
ìlú	LH	town
ìlu	LM	opener
ìlù	LL	drum
pákó	HH	plank
pákò	HL	chewing stick
kése	HM	a mythological place name

The ọkọ series is the one to sit with. Ọkọ is husband, ọkọ́ is hoe, ọkọ̀ is vehicle, and if you also count the subdot distinction, oko is farm. A sentence about a husband, a hoe, a vehicle, and a farm is four different sentences distinguished by marks that untoned writing throws away.

The classic teaching example is the igba set, and it is worth showing carefully because popular sources disagree about which pattern means what. The distinct words include igbá (calabash), ìgbà (time, season), igba (two hundred), ìgbá (garden egg), and igbá in the sense of a climbing rope in some varieties. The forms and their glosses vary by dialect and by source, so treat any single tidy list of four with caution and check it against a tone-marked dictionary. That the contrast exists is certain; the exact inventory in a given variety is what differs.

Tone carries grammar, not just vocabulary

Tone in Yorùbá is not only lexical. It does grammatical work that English does with word order or auxiliaries.

There is a purely tonal morpheme in Yorùbá, the subject-marking High tone, which marks the end of most subject noun phrases and consists of nothing but a floating High tone with no segmental content of its own 1. A grammatical marker made entirely of pitch has no English analogue at all.

Tone also alternates under syntactic conditions. A Low-toned verb becomes Mid when followed by a noun object: rà ìwé (buy book) surfaces as ra ìwé 5. The genitive linker between two nouns is a Mid-toned vowel, obligatory when the second noun begins with a consonant: ilé Kúnlé is realised as ilé e Kúnlé, Kunle's house 51.

Mid tone is the difficult one

There is a well-supported analysis, going back to Akinlabi (1985) and Pulleyblank (1986), that the Yorùbá Mid tone is underlyingly no tone at all rather than a third specified pitch 1. The evidence is tonal stability. When a verb and a following noun combine, Yorùbá deletes one of the adjoining vowels, and High or Low tones of a deleted vowel survive and reattach. Mid tones do not survive: they behave in combination as though they were simply not there, producing no contour and no downstep 1.

This has a practical consequence for learners. Ọlanikẹ́ Ọla Orie's study of American learners of Yorùbá found error rates of 12 percent for High tones and 26.5 percent for Low, but over 97 percent for Mid tones 5. Learners tend to hear a word-initial Mid as High and a word-final Mid as Low, so owó is misproduced as ówó and Ṣẹ́gun as Ṣẹ́gùn 5. Orie records learners' own explanation, that High resembles English stressed-syllable pitch and Low resembles unstressed pitch, but Mid has no English equivalent to hang onto 5.

The lesson for a returning heritage speaker: the unmarked tone is not the easy default, it is the hardest one, and the fact that the orthography leaves it unwritten makes it easier to overlook rather than easier to produce.

Contour tones and downstep

Beyond the three level tones, two contours arise in derived contexts 5:

• A High after a Low becomes rising: ọ̀rẹ́ (friend) is realised with a rising contour on the second syllable.
• A Low after a High becomes falling: Báyọ̀ is realised with a falling contour.

These are predictable consequences of tonal transitions and are not written separately in the orthography. Orie found that even learners who master the three level tones typically fail to acquire the contours 5.

Downstep is the phenomenon where a High tone later in an utterance is realised at a lower absolute pitch than an earlier High, without ceasing to be a High tone. The pitch register steps down, and subsequent Highs are measured against the new lowered ceiling. In Yorùbá this arises characteristically where a Low tone has been removed or is present only as a floating tone between two Highs; Bamgbóṣé described the effect in terms of an assimilated Low tone 6. Yorùbá downstep can also be lexicalised, meaning some words carry it as a fixed property rather than deriving it from context 7. Standard orthography does not mark downstep, which is one reason skilled reading aloud requires knowing the word and not only decoding the marks.

Vowel harmony

Yorùbá has vowel harmony, meaning the vowels within a word are constrained to be compatible with each other rather than freely combinable. The relevant feature is Advanced Tongue Root (ATR), the position of the tongue root that distinguishes e and o from ẹ and ọ 89.

The core generalisation in Standard Yorùbá concerns the mid vowels: in a two-syllable word, the mid vowels must agree in ATR, so e and o occur with each other and with the high vowels, while ẹ and ọ occur with each other. Forms such as ọ̀wọ̀ and ewé are well formed; a word mixing e and ọ across the relevant positions is not. The vowel a and the high vowels i and u behave as neutral or transparent in Standard Yorùbá rather than triggering harmony 8.

Dialects differ in how far harmony extends. Àkúrẹ́ Yorùbá extends ATR harmony to the high vowels as well as the mid vowels, while Standard Yorùbá and Mọ̀bà do not 9. This is a genuine structural difference between varieties, not a matter of careful versus careless speech.

Harmony matters for etymology, because it constrains what a proposed compound can plausibly reduce to. An etymology that requires a vowel sequence Yorùbá phonotactics do not permit is weakened by that fact, a test applied in yoruba-etymology-and-word-formation.

Why this is not optional detail

Two practical conclusions follow from everything above.

First, Yorùbá written without tone marks and subdots is systematically ambiguous, and the ambiguity is not marginal. Ọkọ alone covers husband, hoe, and vehicle. Readers resolve this from context most of the time, which is why untoned Yorùbá is usable, but "usable with effort by someone who already knows the language" is a much weaker property than "readable," and it fails completely for names, for poetry, for anything unfamiliar, and for computational processing.

Second, the tone system is a reason to distrust colonial-era characterisations of the language. A system with three contrastive level tones, two derived contours, downstep, a floating tonal grammatical morpheme, and ATR vowel harmony is not a simple system. It was described as such by observers who could not hear it, and the inability was theirs.