A mora (plural morae or moras; often symbolized μ) is a unit in phonology that determines syllable weight, which in some languages determines stress or timing. The definition of a mora varies. In 1968, American linguist James D. McCawley defined it as "something of which a long syllable consists of two and a short syllable consists of one". The term comes from the Latin word for "linger, delay", which was also used to translate the Greek word chronos (time) in its metrical sense.
Monomoraic syllables have one mora, bimoraic syllables have two, and trimoraic syllables have three, although this last type is relatively rare.
In general, morae are formed as follows:
In general, monomoraic syllables are called "light syllables", bimoraic syllables are called "heavy syllables", and trimoraic syllables (in languages that have them) are called "superheavy syllables". Some languages, such as Old English and present-day English, can have syllables with up to four morae.
For the purpose of determining accent in Ancient Greek, short vowels have one mora, and long vowels and diphthongs have two morae. Thus long ē (eta: η) can be understood as a sequence of two short vowels: ee.
Ancient Greek pitch accent is placed on only one mora in a word. An acute (έ, ή) represents high pitch on the only mora of a short vowel or the last mora of a long vowel (é, eé). A circumflex (ῆ) represents high pitch on the first mora of a long vowel (ée).
In Old English, short diphthongs and monophthongs were monomoraic, long diphthongs and monophthongs were bimoraic, consonants ending a syllable were each a mora, and geminate consonants added a mora to the preceding syllable. In Modern English, the rules are similar, except that all diphthongs are bimoraic. In English, and likely also in Old English, syllables cannot have more than 4 morae, with loss of sounds occurring if a syllable would have more than 4 otherwise. From the Old English period through to today, all content words must be at least 2 morae long.
In Luganda, a short vowel constitutes one mora while a long vowel constitutes two morae. A simple consonant has no morae, and a doubled or prenasalised consonant has one. No syllable may contain more than three morae. The tone system in Luganda is based on morae. (See Luganda tones.)
Gilbertese, an Austronesian language spoken mainly in Kiribati, is a trimoraic language. The typical foot in Gilbertese contains three morae. These trimoraic constituents are units of stress in Gilbertese. These "ternary metrical constituents of the sort found in Gilbertese are quite rare cross-linguistically, and as far as we know, Gilbertese is the only language in the world reported to have a ternary constraint on prosodic word size."
In Hawaiian, both syllables and morae are important. Stress falls on the penultimate mora, though in words long enough to have two stresses, only the final stress is predictable. However, although a diphthong, such as oi, consists of two morae, stress may fall only on the first, a restriction not found with other vowel sequences such as io. That is, there is a distinction between oi, a bimoraic syllable, and io, which is two syllables.
Most dialects of Japanese, including the standard, use morae, known in Japanese as haku (拍) or mōra (モーラ), rather than syllables, as the basis of the sound system. Writing Japanese in kana (hiragana and katakana) is said by those scholars who use the term mora to demonstrate a moraic system of writing. For example, in the two-syllable word mōra, the ō is a long vowel and counts as two morae. The word is written in three symbols, モーラ, corresponding here to mo/o/ra, each containing one mora. Such scholars also argue that the 5/7/5 pattern of the haiku in modern Japanese is of morae rather than syllables.
The Japanese syllable-final n is also said to be moraic, as is the first part of a geminate consonant. For example, the Japanese name for "Japan", 日本, has two different pronunciations, one with three morae (Nihon) and one with four (Nippon). In the hiragana spelling, the three morae of Ni-ho-n are represented by three characters (にほん), and the four morae of Ni-p-po-n need four characters to be written out as にっぽん.
Similarly, the names Tōkyō (to-u-kyo-u とうきょう), Ōsaka (o-o-sa-ka おおさか), and Nagasaki (na-ga-sa-ki ながさき) all have four morae, even though, on this analysis, they can be said to have two, three and four syllables, respectively. The number of morae in a word is not always equal to the number of graphemes in its written form; for example, even though it has four morae, the Japanese name for Tōkyō (とうきょう) is written with five graphemes, because one of these graphemes (ょ) represents a Yōon, a feature on the Japanese writing system that indicates that the preceding mora is palatalized.
In India, the mora was an acknowledged phenomenon well over two millennia ago in ancient Indian linguistics schools studying the dominant scholarly and religious lingua franca of Sanskrit. The mora was first expressed in India as the mātrā.
For example, the short vowel "a" (pronounced like a schwa) is assigned a value of one mātrā, the long vowel "ā" is assigned a value of two mātrās, and the compound vowel (diphthong) "ai" (which has either two simple short vowels, "a"+"i", or one long and one short vowel, "ā"+"i") is assigned a value of two mātrās.
Sanskrit prosody and metrics have a deep history of taking into account moraic weight, as it were, rather than straight syllables, divided into "laghu" (लघु, "light") and "dīrgha" / "guru" (दीर्घ / गुरु, "heavy") feet based on how many morae can be isolated in each word.
Thus, for example, the word kartŗ, meaning "agent" or "doer", does not contain, contrary to intuitive English prosodic principles, simply two syllabic units, but contains rather, in order, a "dīrgha" / "guru "/ "heavy" foot and a "laghu" / "light" foot. The reason is that the conjoined consonants 'rt' render the normally light 'ka' syllable heavy.