A syllable (σ) is a phonological unit of sonority. Sonority can be described by the degree of airflow obstruction and voicing that occurs during phonation. Sonority is inversely correlated with constriction of the articulators in the oral cavity. Sonorous sounds have a more ‘sing-able’ quality, that is they are more prominent in amplitude and length than less sonorous sounds. Sonority shows the resonance of one sound segment in relation to another.
The structure of a syllable represents sonority peaks and optional edges, and is made up of three elements: the onset, the nucleus, and the coda. This can be seen in (1).
(1) σ
/ \
onset rime
/ \
nucleus coda
The Sonority Sequencing Principle and the Sonority Hierarchy
In an ‘optimal syllable,’ sonority increases towards the nucleus, forming a peak in sonority, and then decreases away from the nucleus towards the coda. This is known as the Sonority Sequencing Principle (SSP).
In binary terms, sounds can be classified as either sonorants [+son] or obstruents [-son].
(2)
However, within these binary groupings, there exists a hierarchy of sonority. Take obstruents for example. Both stops and fricatives are [-son], however, stops are less sonorous than fricatives, as we see in (3). Remember that sonority is inversely correlated with constriction of the articulators in the oral cavity. Therefore, stops are less sonorous than the latter because airflow is completely occluded during production of the former, whereas there is less occlusion in the production of latter, where air has more space to flow between articulators.
(3) ‘The Sonority Hierarchy’ ranks sounds from most to least sonorous.
Most sonorous Least sonorous
low Vs – mid Vs – hi Vs and glides– liquids – nasals – fricatives – stops/affricates
Note that:
1. Voiced obstruents rank higher in sonority than their voiceless counterparts.
2. Round vowels are ranked higher in sonority than their unrounded counterparts.
Syllable Structure
When we represent syllable structure as in (1), the nucleus and coda are right-branching forming the ‘rime.’ This type of branching is the most common across languages. However, there are some languages, such as Japanese and Korean, for which a case has been made for a left-branching structure (4) in which the onset and nucleus would form a ‘body.’
(4) Alternate view
σ
/ \
body coda
/ \
onset nucleus
Despite the differences in branching, both of these syllable structures account for onsets, nuclei, and codas. Not all of these elements are required in every syllable.
All languages require syllable nuclei. In fact, the nucleus is the only universally obligatory component of a syllable. In most languages onsets are preferred yet optional – although they are required in some languages. No language requires codas. In most languages codas are optional, and they are restricted or even prohibited in others.
According to the SSP, the nucleus is the peak of sonority. The nucleus is usually filled with a vowel (V), because vowels are the most sonorous sounds. To then allow the nucleus to be the peak of sonority, Consonants (Cs), being less sonorant, flank the nucleus in the onset and coda positions. There are exceptions, which will be discussed later. However, this is why we represent syllables as V, CV, CVC, etc.
The most common structure across languages consists of an onset and a nucleus (CV). CV is known as canonical structure because it is universally the most prevalent structure, and the first to be acquired by children. Although syllabic structure is language-specific in terms of form and the types of phonemes that are permitted in any of the three positions, the canonical structure exists in all languages and is preferred in most.
Onset
CVC
The onset is the beginning of a syllable boundary, and is the strongest consonantal position. The onset is required in many languages and is optional or restricted in others. Onsets are almost always preferred to codas; when a C is found intervocalically it will be parsed to the subsequent onset rather than the previous coda. For instance, a CVCV structure would be syllabified as CV.CV, and not as CVC.V.
Nucleus
CVC
Nuclei are generally filled by vowels.
A vowel cannot fill any position of a syllable besides the nucleus. If a vowel were to be in either the onset or coda position, the nucleus would be required to contain a sound more sonorous than a vowel in order to not violate the SSP. However, there are no sounds more sonorous than vowels. If a vowel is present, it must be in the nucleus of the syllable.
Every vowel will fill a nucleus, but not every nucleus will be filled by a vowel. Some languages do not require a vowel in the nucleus and instead permit certain consonants (C). Languages have different rules for determining what is allowed to occupy the nucleus position. In English, in addition to vowels, syllabic liquids and nasals are permitted to fill the nucleus. Other languages that permit non-vocalic content in their nuclei have far more complex syllabification processes. In languages such as Nuxálk (Bella Coola) (5a) and Berber (5b) there exist entire words and phrases without vowels. They require much more complex algorithms to become syllabified.
(5a) Nuxálk
[sxs] ‘seal blubber’
[xɬpʼχʷɬtʰɬpʰɬːskʷʰt͡sʼ] ‘He had in his possession a bunchberry plant’
(5b) Berber
[tftktst tfktstt] ‘you sprained it and then gave it‘
[rkkm] ‘rot’ (imperf.)
Nevertheless, when we are discussing syllable structure, we will assume that the nucleus of the syllable is a vowel unless otherwise indicated.
Coda
CVC
The coda is optional in most languages. In some languages, it is restricted or even prohibited. Old Bulgarian does not permit codas, and will therefore syllabify (parse) a structure such as CVCCCV to CV.CCCV. On the other hand, some languages such as Persian do not permit complex onsets but do allow complex codas. Persian speakers would most likely syllabify CVCCCV as CVCC.CV.
Syllables without codas are referred to as open syllables, and syllables with codas are referred to as closed syllables. Many languages permit both of these, and in those that do, a difference in syllabification can sometimes change the meaning. This phenomenon is found in English (6).
(6) English
[aj.skɹim] ‘I scream’
VG.CCVC
σ σ
/ \ / \
onset rime onset rime
/ / \
nucleus nucleus coda
[ aj skɹ i m ]
[ajs.kɹim] ‘ice cream’
VCC.CCVC
σ σ
/ \ / \
onset rime onset rime
/ \ / \
nucleus coda nucleus coda
[ aj s kɹ i m ]
Ambisyllabicity
Sometimes breaking a consonant cluster will result in lexical contrast, but other times it won’t. In the instances that lexical contrast is not created, it can be difficult to decide where to parse a syllable. In cases where parsing is not clear, medial Cs are referred to as ambisyllabic. In (7), the [s]s would be considered ambisyllabic.
(7) English
[æs.pɹIn] vs. [æ.spɹIn] ‘aspirin’
[sIs.təɹn] vs. [sI.stəɹn] ‘cistern’
These different syllabifications and parsings can be considered correct as they meet the constraints of their language, but the differences in syllabification will affect the weights of the syllables.
Syllable Weight
Syllable weight is measured in ‘moras.’ A mora is a phonological unit used to distinguish various syllable structures from one another. Syllable weight is important because it can clarify issues pertaining to language-specific constraints on syllable structure.
Languages differ in terms of how moras are measured. In most languages, each element of the rime (the nucleus and coda) counts as a single mora. However, as always, there are exceptions. For example, Ponapean (the major language of the Federated States of Micronesia) does not permit final consonants in coda position to be moraic.
All syllables have at least one mora, which comes from the nucleus of the syllable. In English all of the sound segments in the rime contribute toward syllable weight. Therefore, CV and a CCCV syllables have the same weight as well as number of moras because the complexity of the onset has no bearing on how the weight of the syllable is measured. When syllables are identified by their weight, they are classified as light, heavy, or superheavy.
Light syllables have only one mora (i.e. a short vowel and no coda). One example would be ‘me’ [mi] or CV.
Heavy syllables have two moras. The two moras could both be found in the nucleus in the form of a long vowel or diphthong as in ‘my’ [maj], CVV or one mora could be found in the nucleus and the other in the coda as in ‘mean’ [min], CVC.
Superheavy syllables have three or more moras, which can consist of a short vowel with a complex coda as in ‘meets’ [mits], CVCC, or a long vowel or diphthong with either a simple or complex coda as in ‘mine’ [majn], CVVC, or ‘mines’ [majnz], CVVCC.
For introductory-level practice on syllable structures, check out our Syllable Structure Exercise under Phonology.
T. Molnar