In form, most letters resemble letters of theGreek alphabet. The origin of the alphabet is disputed: it is debated whether (or how) theLatin andRunic alphabets were used as a source. The set of letters, and the way that they are used, show some innovations to expressGothic phonology.
The origin of the Gothic alphabet is controversial.[2][3] In addition to the Gothic language,Ulfilas knewGreek andLatin.[4] When developing the Gothic alphabet, he may have drawn on theGreek alphabet,Latin alphabet, andRunic alphabet.[5] According to Cercignani, it is generally agreed that Ulfilas had knowledge of each of these three alphabets.[6] However, this has been debated. It is not known that Ulfilas was familiar with runes;[7][8] on the other hand, there is also lack of evidence showing that Ulfilas was unfamiliar with runes.[9] There are a few extant runic inscriptions thought to be East Germanic, possibly Gothic[10] (seeGothic runic inscriptions). The influence of Latin and Runic letters on the Gothic alphabet is disputed,[11] and some scholars have argued that they were not used as inputs.[12]
Cercignani (1988:172–178) supposes that Ulfilas used a twenty-four letter Runicfuþark alphabet as his starting point and assigned most of its letters to corresponding letters of the twenty-seven letter Greek alphabet, taking the value of the Greek letters as numerals and their alphabetical position, and using additional symbols (some taken from the Latin alphabet) to take the place of remaining unmatched Greek letters. After the correspondences were established, Cercignani supposes the form of the Runic letters were either modified or outright replaced by the corresponding Latin or Greek letters.Miller (2019:25) calls Cercignani's hypothesis of an adapted runic alphabet "not implausible".
Snædal (2015:96–97, 104–105) argues that Ulfilas's alphabet was initially and primarily adapted from the Greek alphabet (cursive and perhaps alsouncial), with secondary influence from Latin and only minor, uncertain influence from Runic.
Miller (2019:25) cites a number of authors that consider the Gothic alphabet to be derived from the Greek alphabet, influenced by Latin and Runic.
Ulfilas is thought to have consciously chosen to avoid the use of the olderRunic alphabet for this purpose, as it was heavily connected withpagan beliefs and customs.[13] Also, the Greek-based script probably helped to integrate the Gothic nation into the dominantGreco-Roman culture around theBlack Sea.[14]
Below is a table of the Gothic alphabet. Two letters used in itstransliteration are not used in current English:thorn⟨þ⟩ (representing/θ/), andhwair⟨ƕ⟩ (representing/hʷ/).
As with the Greek alphabet, Gothic letters were also assigned numerical values. When used as numerals, letters were written either between two dots (•𐌹𐌱• = 12) or with an overline (𐌹𐌱 = 12). Two letters,𐍁 (90) and𐍊 (900), have no phonetic value.
Names for the letters are recorded inCodex Vindobonensis 795,[15] a 9th-century manuscript ofAlcuin. Most of them seem to be Gothic forms of names also appearing in therune poems. The antiquity of these names is not certain:Cercignani (1988:172, 178–180) argues that the names recorded in this manuscript may be influenced by Old English and Old High German, and concludes that the original rune names are unknown. Most of the names follow the principle ofacrophony (starting with the sound that the letter represents), which Cercignani argues must have been a feature of the original names used by Ulfilas.[16] In the following table, the names are given in their attested forms followed by reconstructed Gothic forms and their meanings.[b]
𐍇 (x) is only used in proper names and loanwords containing GreekΧ (xristus "Christ",galiugaxristus "Pseudo-Christ",zaxarias "Zacharias",aiwxaristia "eucharist").[37]
The letter/numeral𐍊 (900) is not attested in the Gothic bible, but only in the Salzburg-Vienna manuscript[38] (Codex Vindobonensis 795).[39]
Most of the letters clearly bear a one-to-one correspondence to matching letters of theGreek alphabet,[40] having similar forms and sounds and sharing the same alphabetical order andvalue as numerals.
However, a few letters have uncertain or disputed origins, and may have been taken fromLatin or possibly (more controversially[41])Runic letters. These are:
𐌵 (q). Its numeric value of 6 corresponds to Greekdigamma orstigma (), which may also be the source of its form.[41] Other possible sources are a cursive variant ofkappa (ϰ), which could strongly resemble au,[42] or Latin minusculeq.[43]
𐌷 (h). Its numeric value of 8 corresponds to Greeketa (η). Its form and phonetic value may be borrowed from Latin[44] uncialh;[45] however,Snædal (2015:98) argues that there is no need to use Latin to explain the form or sound value of the Gothic letter.
𐌸 (þ). Its numeric value of 9 corresponds to Greektheta (θ)/θ/. It is used for theta in the Gothic transcription of Greek names, such as𐍄𐌴𐌹𐌼𐌰𐌿𐌸𐌰𐌹𐌿𐍃 (teimauþaius) forΤιμόθεος. Its form may be derived from theta via the 4th-century cursive formϑ.[46] Alternatively, its form has been argued to derive from Greekphi (Φ)[47]/f/ orpsi (Ψ)/ps/ with phonetic reassignment, or from Runicᚦ.[48]
𐌾 (j). Its numeric value of 60 corresponds to Greekxi (ξ)/ks/. Its form may be derived from LatinG /ɡ/,[49] Greekξ, the Greek epsilon-iotaligature, or Runicᛃ[50] (the last of which is itself speculated to be derived from Greek epsilon-iota ligature).
𐌿 (u). Its numeric value of 70 corresponds to Greekomicron (ο) (originally namedοὖ/uː/[51][52]). Its form may be derived from Greekο or from Runicᚢ/u/.[53]
𐍈 (ƕ). Its numeric value of 700 corresponds to Greekpsi (ψ)/ps/. Its form is potentially derived from GreekΘ/θ/ with phonetic reassignment; or from GreekΟ/o/; possibly the letterform was switched with𐌸;[41] also possibly pictographic in nature, with O for rounding and • for aspiration.[54]
𐍉 (o). Its numeric value of 800 corresponds to Greekomega (ω). Its form may be derived from Greekω or from Runicᛟ.[55][14] An alternative proposal[56] derives it from a cursive form of Greekomicron (ο):Snædal (2015:103) argues that this better matches the shape of the Gothic letter (though the variant of Greek omicron is often vertically mirrored compared to the Gothic letter), and that the use of Greek omicron to represent close-mid/o/ would be symmetrical to the use of Greek epsilon to represent close-mid/e/. On the other hand, Miller argues that the idea that𐍉 was derived from omicron does not explain why it has the same numerical value and alphabetical position as omega, with the (pronounced) Gothic alphabet being ordered from𐌰 to𐍉 like the Greek alphabet is ordered fromΑ to Ω.[56]
𐍂 (r),𐍃 (s) and𐍆 (f) appear to be derived from their Latin equivalents rather than from the Greek, although the equivalent Runic letters (ᚱ,ᛋ andᚠ), assumed to have been part of the Gothic futhark, possibly played some role in this choice.[57][58] However, Snædal claims that "Wulfila's knowledge of runes was questionable to say the least", as the paucity of inscriptions attests that knowledge and use of runes was rare among the East Germanic peoples.[41] Miller refutes this claim, stating that it is "not implausible" that Wulfila used a runic script in his creation of the Gothic alphabet, noting six other authors—Wimmer, Mensel, Hermann, d'Alquen, Rousseau, and Falluomini—who support the idea of the Gothic alphabet having runic contributions.[56] Some variants of𐍃 (s) are shaped like a sigma and more obviously derive from the GreekΣ.[41]
Diacritics and punctuation used in theCodex Argenteus include atrema placed on𐌹i, transliterated asï (used at the start of a word or syllable[59]); theinterpunct (·) andcolon (:) as well asoverlines to indicatesigla (such asxaus forxristaus) and numerals.
The Gothic alphabet was added to theUnicode Standard in March 2001 with the release of version 3.1.
The Unicode block for Gothic isU+10330–U+1034F in theSupplementary Multilingual Plane. As older software that usesUCS-2 (the predecessor ofUTF-16) assumes that all Unicode codepoints can be expressed as 16bit numbers (U+FFFF or lower, theBasic Multilingual Plane), problems may be encountered using the Gothic alphabet Unicode range and others outside of theBasic Multilingual Plane.
^The forms which are not attested in the Gothic corpus are marked with an asterisk. For a detailed discussion of the reconstructed forms,cf.[17] For a survey of the relevant literature, cf.[18]
^Zacher arrives at *iuya, *iwja or *ius, cognate to ONȳr, OEīw, ēow, OHGīwa "yew tree", though he admits having no ready explanation for the formezec; cf.[24]
^conventionally called *algiz or *elhaz, but unclear original name
^Zacher assents to a hypothesis of the Brothers Grimm that the Alcuin name "thyth" is not Germanic, but instead is a phonetic adaptation of the name of Greek theta. Based on comparison with Old English, Zacher reconstructs𐌸𐌰𐌿𐍂𐌽𐌿𐍃 (þaúrnus) "thorn" as the original Gothic rune name (this name was also previously postulated by Munch).[26]
Snædal, Magnús (2015). "Gothic Contact with Latin: Gotica Parisina and Wulfila's Alphabet". In Askedal, John Ole; Nielsen, Hans Frede (eds.).Early Germanic Languages in Contact. NOWELE Supplement Series. Vol. 27. pp. 91–107.doi:10.1075/nss.27.05sna.ISBN978-90-272-4073-6.