The syllable in Spanish

To speak Spanish more naturally and to understand spoken Spanish more easily, it will help to have an idea of how the syllable works in Spanish. Syllables are the "rhythmic units" that we subconciously divide our speech into. (Usually, when you clap a word or phrase, you clap once per syllable.) In general:

• every vowel forms the centre of a syllable;
• consonants generally sit at the edges of syllables;
• languages differ in their phonotactics: that is, the combinations or choices of sounds allowed in different positions of the syllable;
• the pronunciation of sounds typically varies depending (among other things) on where it is in the syllable (for example, in English, voiceless stops such as t are accompanied by an aspiration or "puff or air" at the beginning of a syllable, but not at the end).

Note that strictly speaking, syllabification is not a spelling issue. (Some grammars and style guides appear to confuse syllabification with hyphenation.)

Syllable structure

Syllables are often analysed as having an internal structure, as shown in Figure 1. The beginning of the syllable, called the onset, can hold a consonant or a cluster of consonants. The middle of the syllable, called the nucleus, holds a vowel¹. And the end of the syllable or coda holds another consonant.

1. A vowel sound is generally one where the airway is not impeded, whereas a consonant "blocks" or "impedes" the airway. But there are certain consonants, such as l, n, that are actually quite "vowel-like" in that they only block the airway to a limited extent. In English, but not Spanish, these "vowel-like consonants" can potentially fill the syllable nucleus, as in the final l sound in the English word little, usually syllabified lit.l, or possibly the final s in the word strengths.

There are some similarities and some differences in how this structure can be filled in English and Spanish (phonotactics):

• In both Spanish and English, onset clusters can be a stop or f plus a liquid (l or r), giving groups such as pl-, br-, cl-, fr- etc.
• In English, the onset can also be formed by s plus a stop or cluster with a stop (e.g. spray, scanner etc). Such clusters are unstable in Spanish; we'll look at this issue in more detail below.
• English readily allows clusters in the coda (e.g. film, meetings; possibly even [kst] in mixed); in Spanish, only a very limited range of single consonants tend to occur in the coda (d, s, n, r and the fricative z¹ in Castillian Spanish), although other codas are possible in loanwords.
• Unlike English, some varieties of Spanish allow tl in the onset (essentially Latin America and Western Spain), e.g. in the Mexican words tlacoyo, tlayuda etc. In such varieties, a word such as atlas would be syllabified a.tlas, as opposed to at.las in other varieties (principally Central and Eastern Spain).
• Both languages allow diphthongs and triphthongs: vowels that are glides between two or three targets (cf English height where the vowel glides between an [a] and an [i] sound). But the inventory of diphthongs and thripthongs allowed in the two languages differs, as do the circumstances when they occur. In Spanish, diphthongs and triphthongs tend to occur as more of an automatic process when certain vowels occur in sequence, rather than being simply dictated by specific words.

1. This is the voiceless interdental fricative, similar to US English th in think, where the tip of the tongue sits between the backs of the two rows of teeth, at the right position to cause friction.

How utterances are grouped into syllables

The start of a word or prefix doesn't necessarily coincide with the start of a syllable. In general, words and phrases are grouped into syllables in Spanish as follows:

• In general, the vowels i and u form a diphthong with the neighbouring vowel, so that the i and u become a "glide into" or "glide out of" the other vowel. So for example, the word fuerte is two syllables, and fue- occupies a single syllable, with ue a dipthong.
• In rapid speech, any sequence of two vowels can become a diphthong and fill a single syllable. So for example, tea.tro is often pronounced as two syllables, with ea forming a diphthong.
• Consonants generally fill syllable onsets in preference to syllable codas. So given a sequence of consonants, they will form a cluster in a syllable onset if Spanish allows that combination in an onset (see above), else they will be split into the coda of one syllable and the onset of the next. For example:
  • the word hablar is syllabified ha.blar (and not *hab.lar), because Spanish allows the cluster bl- in a syllable onset (put another way, this sequence can come at the beginning of a word)¹;
  • the word es.pe.rar is syllabified with s.p in two separate syllables, because sp- can't generally form an onset cluster (or stably come at the start of a word) in Spanish;
  • the word en.trar is syllabified with tr- at the start of the second syllable, rather than *ent-rar, because tr- is a valid syllable onset (and beginning of a word) in Spanish, and syllabification maximises onsets at the expense of codas. However, we of course can't syllabify *e.ntrar, because ntr- isn't a valid syllable onset in Spanish (or indeed English).
• These syllabification patterns span over words. So for example, come y toma is generally pronounced as four syllables co.mei.to.ma, with "(com)e y" forming a diphthong (although written with a letter y, the word y is actually just the vowel [i], so the usual diphthongisation rule applies).
• When the same vowel accurs twice across a word boundary (i.e. at the end of a word then at the beginning of the following word), then one of the vowels is normally elided (so in effect it's "pronounced once"): la alfombra becomes three syllables: lal.fom.bra, as does le escucho: les.cu.cho.

---

1. Some slightly more complex situations actually arise. In the case of a word such as hablar, speakers appear to have a strong intuition that the division is ha-blar. But in the case of a word such as subliminal, speakers appear less sure intuitively as to whether the division is su.bli.mi.nal (as expected by the maximum onset principle) or sub.li.mi.nal (not expected, but possibly influenced by the fact that sub- is a morpheme).