The ligaturesÆ, Œ, and the symbol ß, when used in English, French, or German, are normally not counted as separate alphabetic letters but as variants of AE, OE, and ss, respectively. Letters bearing diacritics are also not counted as separate letters in these languages. This is often not the case for Æ and Œ and some letters bearing diacritics in other variations of the Latin alphabet. For example, å, ä, and ö all count as separate letters in Swedish.
The letters [[Þ|]], Ðð, Ææ and Ƿƿ are no longer a part of the Latin alphabet as used in English, but they were considered Latin letters in the past, and except for the last, are still used in Icelandic. For a short time in Roman history, the three Claudian letters were added to the alphabet, but the innovations didn't stick.
Later the Greek zeta (Z) was dropped and a new letter G was placed in its position. After the conquest of Greece in the first century BC the letters Y and Z were, respectively, adopted and readopted from the Greek alphabet and placed at the end. Now the new Latin alphabet contained 23 letters. It was not until the Middle Ages that the letter J (representing non-syllabic I) and the letters U and W (to distinguish them from V) were added.
The alphabet used by the Romans consisted only of capital (upper case or majuscule) letters. The lower case (minuscule) letters developed in the Middle Ages from cursive writing, first as the uncial script, and later as minuscule script. The old Roman letters were retained for formal inscriptions and for emphasis in written documents. The languages that use the Latin alphabet generally use capital letters to begin paragraphs and sentences and for proper nouns. The rules for capitalization have changed over time, and different languages vary in their rules for capitalization. Old English, for example, used to capitalise all nouns, in the same way that Modern German does today.
As late as 1492, the Latin alphabet was limited primarily to the nations of western and central Europe. The Orthodox Christian Slavs of eastern and southern Europe mostly used the Cyrillic alphabet, and the Greek alphabet was still in use by Greek-speakers around the eastern Mediterranean. The Arabic alphabet was widespread within Islam, both among Arabs and non-Arab nations like the Turks and Iranians. Most of the rest of Asia used a variety of Brahmic alphabets or the Chinese script.
In the course of its history, the Latin alphabet was adapted for use for new languages, some of which had phonemes which were not used in languages previously written with this alphabet, and therefore new letters and diacritics were created as needed, for example:
the cedilla in ç, originally a small z written below the c (once symbolized /ts/ in Romance languages, now gives c a 'soft' sound before a, o, and u, e.g. /s/ in French façade and /θ/ in Catalan Barça)
the háček in č š ž (used in Slavonic languages to mark palatalized versions of the base letter), and in ǔ (used in Belarusian language to mark a consonant similar to w)
the tilde in Spanish ñ, Portuguese ã and õ, and Estonianõ, originally a small n written above the letter (once used to mark the elision of a former n, now marks nasalization of the base letter)
the acute accent in á é í ó ú in Spanish and other languages
the grave accent in à è ù in French, Italian, and other languages
the circumflex in the vowels â ê î ô û in French, Romanian, and other languages, and in the consonants ĉ ĝ ĥ ĵ ŝ in Esperanto.
the umlaut in ä ö ü in German and other languages (changes quality of the vowel)
the diaeresis (same visual appearance as the umlaut above) in ä ë ï ö ü in several languages (to indicate that two successive vowels do not form a diphthong)
the dot above in ė ż in Polish, Latvian, and other languages
the ogonek in ą ę į ų in Polish, Latvian, and other languages
the macron in ā ē ī ū in Latvian and other languages
W is a letter made up from two V's or U's. It was added in late Roman times to represent a Germanic sound. U and J were originally not distinguished from V and I respectively.
In Old English, ash æ, eth ð and the Runic letters thorn þ, and wynn ƿ were added. Eth and thorn were replaced with 'th', and wynn with the new letter 'w'. In modern Icelandic, thorn and eth are still used.
The additional letters added in German are special presentations of earlier ligature forms (ae → ä, ue → ü or ſs → ß). French adds the circumflex to record elided consonants that were present in earlier forms and are often still present in the modern English cognate forms (Old French hostel → French hôtel = English hotel or Late Latin pasta → Middle French paste → English paste. Note Modern French divergence to pâte, and preservation of the original pasta in Italian, and now borrowed into English).
West Slavic and most South Slavic languages use the Latin alphabet rather than the Cyrillic. Among these, Polish uses a variety of diacritics and digraphs to represent special phonetic values, as well as the l with stroke - ł - for a sound similar to w. Czech uses diacritics as in Dvořák — the term háček (caron) originates from Czech. Croatian and the Latin version of Serbian use carons in č, š, ž, an acute in ć and a bar in đ. The languages of Eastern Orthodox Slavs generally use Cyrillic instead which is much closer to the Greek alphabet. Serbian language uses two alphabets.
The African language Hausa uses three additional consonant letters: ɓ, ɗ and ƙ, which are variants of b, d and g employed by linguists to represent certain sounds similar to them.
In French and English, characters with diaeresis (ä, ë, ï, ö, ü, ÿ) are usually treated just like their un-accented versions. If two words differ only by an accent in French, the one with the accent is greater. (However, the Unicode 3.0 book specifies a more complex traditional French sorting rule for accented letters.)
In German letters with umlaut (Ä, Ö, Ü) are treated generally just like their non-umlauted versions; ß is always sorted as ss. This makes the alphabetic order Arg, Ärgerlich, Arm, Assistent, Aßlar, Assoziation. For phone directories and similar lists of names, the umlauts are to be collated like the letter combinations "ae", "oe", "ue". This makes the alphabetic order Udet, Übelacker, Uell, Ülle, Ueve, Üxküll, Uffenbach.
In the Swedish alphabet, "W" is seen as a variant of "V" and not a separate letter. It is however recognised and maintained in names, like in "William". The alphabet also has three extra vowels placed at its end (..., X, Y, Z, Å, Ä, Ö). The same alphabet and collating rules are used for Finnish.
The same extra vowels as in Swedish are also present in the Danish and Norwegian alphabets but in a different order and with different glyphs (..., X, Y, Z, Æ, Ø, Å). Also, "Aa" collates as an equivalent to "Å". The Danish alphabet has traditionally seen "W" as a variant of "V", but nowadays "W" is considered a separate letter.
The Faroese alphabet also has some of these extra letters, namely Æ and Ø. Furthermore, the Faroese alphabet uses the eth, which follows the D. Five of the six vowels A, I, O, U and Y can get accents and are after that considered separate letters. The consonants C, Q, X, W and Z are not found. Therefore the first five letters are A, Á, B, D and Ð, and the last five are V, Y, Ý, Æ, Ø
Some languages have more complex rules: for example, Spanish treated (til 1997) "CH" and "LL" as single letters, giving an ordering of CINCO, CREDO, CHISPA and LOMO, LUZ, LLAMA. This is not true anymore since in 1997 RAE adopted the more conventional usage, and now LL is collated between LK and LM, and CH between CG and CI. The only Spanish specific collating question is Ñ (eñe) as a different letter collated after N.
Welsh also has complex rules: the combinations CH, DD, FF, NG, LL, PH and TH are all considered single letters, and each is listed after the letter which is the first character in the combination, with the exception of NG which is listed after G. However, the situation is further complicated by these combinations not always being single letters. An example ordering is LAWR, LWCUS, LLONG, LLOM, LLONGYFARCH: the last of these words is a juxtaposition of LLON and GYFARCH, and, unlike LLONG, does not contain the letter NG.
In Dutch the combination IJ (representing Ĳ (letter IJ)) was formerly to be collated as Y (or sometimes, as a separate letter Y < IJ < Z), but is currently mostly collated as 2 letters (II < IJ < IK). Exceptions are phone directories; IJ is always collated as Y here because in many Dutch family names Y is used where modern spelling would require IJ. Note that a word starting with ij that is written with a capital I is also written with a capital J, e.g. the town IJmuiden (mun. Velsen) and the river IJssel.
The Hungarian language has accents, umlauts, and double accents. The accent is ignored in collating, and the double accent, which indicates a long umlaut vowel, is treated as equal to the umlaut.
Both letters were also used by Anglo-Saxon scribes who also used the Runic letter Wynn to represent /w/.
Þ (called thorn; lowercase þ) is also a Runic letter.
Ð (called eth; lowercase ð) is the letter D with an added stroke.
In Polish, specifically Polish letters derived from the Latin alphabet are collated after their originals: A, Ą, B, C, Ć, D, E, Ę, ..., L, Ł, M, N, Ń, O, Ó, P, ..., S, Ś, T, ..., Z, Ź, Ż.
In Czech (and Slovak), accented vowels have secondary collating weight - compared to other letters, they are treated as their unaccented forms (A-Á, E-É-Ě, I-Í, O-Ó U-Ú-Ů, Y-Ý), but then they are sorted after the unaccented letters (e.g. the correct lexicographic order is baa, baá, báa, bab, báb, bac, bác, bač, báč). Accented consonants (the ones with hacek) have primary collating weight and are collocated immediately after their unaccented counterparts, with exception of Ď, Ň and Ť, which have again secondary weight. CH is considered to be a separate letter and goes between H and I.
In Esperanto, consonants with circumflex accents (ĉ, ĝ, ĥ, ĵ, ŝ), as well as ŭ (u with breve), are counted as separate letters and collated separately (c, ĉ, d, e, f, g, ĝ, h, ĥ, i, j, ĵ ... s, ŝ, t, u, ŭ, v, z).
In Romanian, special characters derived from the Latin alphabet are collated after their orginals: A, Ă, Â, ..., I, Î, ..., S, Ş, T, Ţ, ..., Z.
In Tatar, there are 9 additional letters. 5 of them are vowels, paired with main alphabet vowels as hard-smooth: a-ä, o-ö, u-ü, í-i, ı-e. The four remaining are consonants: ş is sh, ç is ch, ñ is ng and ğ is gh.
In Croatian and Serbian and related South Slavic languages, the five accented characters and two conjoined characters are sorted after the originals: ..., C, Č, Ć, D, DŽ, Đ, E, ..., L, LJ, M, N, NJ, O, ..., S, Š, T, ..., Z, Ž.
In Filipino and other Philippine languages, the letter Ng is treated as a separate letter. Also, letter deriviatives (such as Ñ) immediately follow the base letter. Filipino also is written with accents and other marks , but the marks are not in very wide use (except the tilde).
Jensen, Hans. 1970. Sign Symbol and Script. London: George Allen and Unwin Ltd. Transl. of Die Schrift in Vergangenheit und Gegenwart. VEB Deutscher Verlag der Wissenschaften. 1958, as revised by the author.
Rix, Helmut. 1993. "La scrittura e la lingua" In: Cristofani, Mauro (hrsg.) 1993. Gli etruschi - Una nuova immagine. Firenze: Giunti. S.199-227.
Sampson, Geoffrey. 1985. Writing systems. London (etc.): Hutchinson.
Wachter, Rudolf. 1987. Altlateinische Inschriften: sprachliche und epigraphische Untersuchungen zu den Dokumenten bis etwa 150 v.Chr. Bern (etc.): Peter Lang.