alphabet

Etymology

The English word alphabet came into Middle English from the Late Latin word alphabetum, which in turn originated in the Greek ἀλφάβητος alphábētos; it was made from the first two letters of the Greek alphabet, alpha (α) and beta (β). The names for the Greek letters, in turn, came from the first two letters of the Phoenician alphabet: aleph, the word for ox, and bet, the word for house.

History

Alphabets related to Phoenician

The Ancient Egyptian writing system had a set of some 24 hieroglyphs that are called uniliterals, which are glyphs that provide one sound. These glyphs were used as pronunciation guides for logograms, to write grammatical inflections, and, later, to transcribe loan words and foreign names. The script was used a fair amount in the 4th century AD. However, after pagan temples were closed down, it was forgotten in the 5th century until the discovery of the Rosetta Stone. There was also cuneiform, primarily used to write several ancient languages, including Sumerian. The last known use of cuneiform was in 75 AD, after which the script fell out of use.

In the Middle Bronze Age, an apparently alphabetic system known as the Proto-Sinaitic script appeared in Egyptian turquoise mines in the Sinai Peninsula c. 1840 BC, apparently left by Canaanite workers. Orly Goldwasser has connected the illiterate turquoise miner graffiti theory to the origin of the alphabet. In 1999, American Egyptologists John and Deborah Darnell discovered an earlier version of this first alphabet at the Wadi el-Hol valley. The script dated to c. 1800 BC and shows evidence of having been adapted from specific forms of Egyptian hieroglyphs that could be dated to c. 2000 BC, strongly suggesting that the first alphabet had developed about that time. The script was based on letter appearances and names, believed to be based on Egyptian hieroglyphs. This script had no characters representing vowels. Originally, it probably was a syllabary—a script where syllables are represented with characters—with symbols that were not needed being removed. The best-attested Bronze Age alphabet is Ugaritic, invented in Ugarit before the 15th century BC. This was an alphabetic cuneiform script with 30 signs, including three that indicate the following vowel. This script was not used after the destruction of Ugarit in 1178 BC.

The Proto-Sinaitic script eventually developed into the Phoenician alphabet, conventionally called Proto-Canaanite, before c. 1050 BC. The oldest text in Phoenician script is an inscription on the sarcophagus of King Ahiram c. 1000 BC. This script is the parent script of all western alphabets. By the 10th century BC, two other forms distinguish themselves, Canaanite and Aramaic. The Aramaic gave rise to the Hebrew alphabet.

The South Arabian alphabet, a sister script to the Phoenician alphabet, is the script from which the Geʽez script was descended. Abugidas are writing systems with characters comprising consonant–vowel sequences. Alphabets without obligatory vowels are called abjads, with examples being Arabic, Hebrew, and Syriac. The omission of vowels was not always a satisfactory solution due to the need of preserving sacred texts. "Weak" consonants are used to indicate vowels. These letters have a dual function since they can also be used as pure consonants.

Types

The term "alphabet" is used by linguists and paleographers in both a wide and a narrow sense. In its broader sense, an alphabet is a segmental script at the phoneme level—that is, it has separate glyphs for individual sounds and not for larger units such as syllables or words. In its narrower sense, some scholars distinguish "true" alphabets from two other types of segmental script: abjads and abugidas. These three differ in how they treat vowels. Abjads have letters for consonants and leave most vowels unexpressed. Abugidas are also consonant-based but indicate vowels with diacritics, a systematic graphic modification of the consonants. The earliest known alphabet using this sense is the Wadi el-Hol script, believed to be an abjad. Its successor, Phoenician, is the ancestor of modern alphabets, including Arabic, Greek, Latin (via the Old Italic alphabet), Cyrillic (via the Greek alphabet), and Hebrew (via Aramaic).

Examples of present-day abjads are the Arabic and Hebrew scripts; true alphabets include Latin, Cyrillic, and Korean hangul; and abugidas, used to write Tigrinya, Amharic, Hindi, and Thai. The Canadian Aboriginal syllabics are also an abugida, rather than a syllabary, as their name would imply, because each glyph stands for a consonant and is modified by rotation to represent the following vowel. In a true syllabary, each consonant-vowel combination gets represented by a separate glyph.

All three types may be augmented with syllabic glyphs. Ugaritic, for example, is essentially an abjad but has syllabic letters for /ʔa, ʔi, ʔu/ These are the only times that vowels are indicated. Coptic has a letter for /ti/. Devanagari is typically an abugida augmented with dedicated letters for initial vowels, though some traditions use अ as a zero consonant as the graphic base for such vowels.

The boundaries between the three types of segmental scripts are not always clear-cut. For example, Sorani Kurdish is written in the Arabic script, which, when used for other languages, is an abjad. In Kurdish, writing the vowels is mandatory, and whole letters are used, so the script is a true alphabet. Other languages may use a Semitic abjad with forced vowel diacritics, effectively making them abugidas. On the other hand, the ʼPhags-pa script of the Mongol Empire was based closely on the Tibetan abugida, but vowel marks are written after the preceding consonant rather than as diacritic marks. The short a is not written in the Geʽez script now used for Amharic and Tigrinya, unlike the Indic abugidas; the source of the term "abugida" assimilating the short a vowel into its consonant modifications made the abugida no longer systematic, and by definition, a syllabary rather than a segmental script. Even more extreme, the Pahlavi abjad eventually became logographic.

Alphabetical order

Alphabets often come to be associated with a standard ordering of their letters; this is for collation—namely, for listing words and other items in "alphabetical order."

Latin alphabets

The ordering of the Latin alphabet (A B C D E F G H I J K L M N O P Q R S T U V W X Y Z), which derives from the Northwest Semitic "Abgad" order, is already well established. Although, languages using this alphabet have different conventions for their treatment of modified letters (such as the French é, à, and ô) and certain combinations of letters (multigraphs). In French, these are not considered to be additional letters for collation. However, in Icelandic, the accented letters such as á, í, and ö are considered distinct letters representing different vowel sounds from sounds represented by their unaccented counterparts. In Spanish, ñ is considered a separate letter, but accented vowels such as á and é are not. The ll and ch were also formerly considered single letters and sorted separately after l and c, but in 1994, the tenth congress of the Association of Spanish Language Academies changed the collating order so that ll came to be sorted between lk and lm in the dictionary and ch came to be sorted between cg and ci; those digraphs were still formally designated as letters, but in 2010 the Real Academia Española changed it, so they are no longer considered letters at all.

In German, words starting with sch- (which spells the German phoneme /ʃ/) are inserted between words with initial sca- and sci- (all incidentally loanwords) instead of appearing after the initial sz, as though it were a single letter, which contrasts several languages such as Albanian, in which dh-, ë-, gj-, ll-, rr-, th-, xh-, and zh-, which all represent phonemes and considered separate single letters, would follow the letters ⟨d, e, g, l, n, r, t, x, z⟩ respectively, as well as Hungarian and Welsh. Further, German words with an umlaut get collated ignoring the umlaut as—contrary to Turkish, which adopted the graphemes ö and ü, and where a word like tüfek would come after tuz, in the dictionary. An exception is the German telephone directory, where umlauts are sorted like ä=ae since names such as Jäger also appear with the spelling Jaeger and are not distinguished in the spoken language.

The Danish and Norwegian alphabets end with ⟨æ, ø, å⟩, whereas the Swedish conventionally put ⟨å, ä, ö⟩ at the end. However, ⟨æ⟩ phonetically corresponds with ⟨ä⟩, as does ⟨ø⟩ and ⟨ö⟩.

Early alphabets

It is unknown whether the earliest alphabets had a defined sequence. Some alphabets today, such as the Hanuno'o script, are learned one letter at a time, in no particular order, and are not used for collation where a definite order is required. However, a dozen Ugaritic tablets from the 14th century BC preserve the alphabet in two sequences. One, the ABCDE order later used in Phoenician, has continued with minor changes in Hebrew, Greek, Armenian, Gothic, Cyrillic, and Latin; the other, HLĦM, was used in southern Arabia and is preserved today in Geʽez. Both orders have therefore been stable for at least 3000 years.

Elder Fuþark runes had unrelated sequences beginning in "Fuþark", which got simplified later on in Younger Fuþark and expanded in Anglo-Frisian Fuþorc. Arabic usually uses its sequence, although Arabic retains the traditional abjadi order, which is used for numbers.

The Brahmic family of alphabets used in India uses a unique order based on phonology: the letters are arranged according to how and where the sounds are produced in the mouth. This organization is present in Southeast Asia, Tibet, Korean hangul, and even Japanese kana, which is not an alphabet.

Acrophony

In Phoenician, each letter got associated with a word that begins with that sound. This is called acrophony and is still used to varying degrees in Samaritan, Aramaic, Syriac, Hebrew, Greek, and Arabic.

Acrophony was abandoned in Latin. It referred to the letters by adding a vowel—usually ⟨e⟩, sometimes ⟨a⟩ or ⟨u⟩—before or after the consonant. Two exceptions were Y and Z, which were borrowed from the Greek alphabet rather than Etruscan. They were known as Y Graeca "Greek Y" and zeta (from Greek)—this discrepancy was inherited by many European languages, as in the term zed for Z in all forms of English, other than American English. Over time names sometimes shifted or were added, as in double U for W, or "double V" in French, the English name for Y, and the American zee for Z. Comparing them in English and French gives a clear reflection of the Great Vowel Shift: A, B, C, and D are pronounced /eɪ, biː, siː, diː/ in today's English, but in contemporary French they are /a, be, se, de/. The French names (from which the English names got derived) preserve the qualities of the English vowels before the Great Vowel Shift. By contrast, the names of F, L, M, N, and S (/ɛf, ɛl, ɛm, ɛn, ɛs/) remain the same in both languages because "short" vowels were largely unaffected by the Shift.

In Cyrillic, originally, acrophony was present using Slavic words. The first three words going, azŭ, buky, vědě, with the Cyrillic collation order being, А, Б, В. However, this was later abandoned in favour of a system similar to Latin.

Orthography and pronunciation

When an alphabet is adopted or developed to represent a given language, an orthography generally comes into being, providing rules for spelling words, following the principle on which alphabets get based. These rules will map letters of the alphabet to the phonemes of the spoken language. In a perfectly phonemic orthography, there would be a consistent one-to-one correspondence between the letters and the phonemes so that a writer could predict the spelling of a word given its pronunciation, and a speaker would always know the pronunciation of a word given its spelling, and vice versa. However, this ideal is usually never achieved in practice. Languages can come close to it, such as Spanish and Finnish. Others, such as English, deviate from it to a much larger degree.

The pronunciation of a language often evolves independently of its writing system. Writing systems have been borrowed for languages the orthography was not initially made to use. The degree to which letters of an alphabet correspond to phonemes of a language varies.

Languages may fail to achieve a one-to-one correspondence between letters and sounds in any of several ways:

A language may represent a given phoneme by combinations of letters rather than just a single letter. Two-letter combinations are called digraphs, and three-letter groups are called trigraphs. German uses the tetragraphs (four letters) "tsch" for the phoneme German pronunciation: [tʃ] and (in a few borrowed words) "dsch" for [dʒ]. Kabardian also uses a tetragraph for one of its phonemes, namely "кхъу." Two letters representing one sound occur in several instances in Hungarian as well (where, for instance, cs stands for [tʃ], sz for [s], zs for [ʒ], dzs for [dʒ]).

A language may represent the same phoneme with two or more different letters or combinations of letters. An example is modern Greek which may write the phoneme Greek pronunciation: [i] in six different ways: ⟨ι⟩, ⟨η⟩, ⟨υ⟩, ⟨ει⟩, ⟨οι⟩, and ⟨υι⟩.

A language may spell some words with unpronounced letters that exist for historical or other reasons. For example, the spelling of the Thai word for 'beer' เบียร์ retains a letter for the final consonant /r/ present in the English word it borrows, but silences it.