Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Linguistics. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 14 April 2021

Gothiclocked

  • D. Gary MillerD. Gary MillerLinguistics, University of Florida

Summary

This is an advance summary of a forthcoming article in the Oxford Research Encyclopedia of Linguistics. Please check back later for the full article.

Apart from runic inscriptions, Gothic is the earliest attested language of the Germanic family, dating to the 4th century. Along with Crimean Gothic, it belongs to the branch known as East Germanic. The bulk of the extant Gothic corpus is a translation of the Bible, of which only a portion remains. The translation is traditionally ascribed to Wulfila, who is credited with inventing the Gothic alphabet. The many Greek conventions both help and hinder interpretation of the Gothic phonological system. As in Greek, letters of the alphabet functioned as numerals, but the late letter names were from runic.

Gothic inflectional categories include nouns, adjectives, and verbs. Nouns are inflected for three genders, two numbers, and four cases. Various stem types inherited from Indo-European constitute different form classes in Gothic. Adjectives have the same properties and are also inflected according to so-called weak and strong forms, as are Gothic verbs. Verbs are inflected for three persons and numbers, an indicative and a nonindicative mood (here called “optative”), past and nonpast tense, and voice. The mediopassive survives in Gothic morphologically as a synthetic passive and syntactically in innovated periphrastic formations; middle and anticausative functions were taken over by reflexive-type structures. Nonfinite forms are the infinitive, the imperative, and two participles.

In syntax, Gothic had null subjects as an option, mostly in the third person singular. Aspect was effected primarily by prefixes, which have many other functions, and aspect is not consistently indicated. Absolute constructions with a participle occurred in various cases with functional differences. Relativization was effected primarily by relative pronouns built on demonstratives plus a complementizer. Complementizers could be used with subordinate clause verbs in the indicative or optative. The switch to the optative was triggered by irrealis, matrix verbs that do not permit a full range of subordinate tenses, expression of a hope or wish, potentiality, and several other conditions. Many of these are also relevant to matrix clauses (independent optatives).

Essentials of linearization include prepositional phrases, default postposed genitives and possessive adjectives, and preposed demonstratives. Verb-object order predominates, but there is much Greek influence. Verb-auxiliary order is native Gothic.

A Brief History of the Goths

Most of what is known about the Goths is from Jordanes, a Romanized Goth.1 Born ?ca. 480 ce on the lower Danube, he served in Moesia (north of Thrace, northern Bulgaria today) as a notarius (secretary) to the otherwise unknown eastern military commander Gunthigis or Baza. Jordanes’s Getica or ‘The Getae/Goths’ (confused in late antiquity), composed in Constantinople before April 1, 551, is written in Moesian administrative Latin and departs considerably from what little is known about the lost twelve-book Historia Gothica (Gothic History), written ca. 533 by Cassiodorus (ca. 490–ca. 583) (see Barnish, 1984; Croke, 1987). Jordanes, who was writing a world history, was asked by his friend Castalius to summarize that work, but without access to it he had to rely on memory from a single reading plus other sources (Croke, 1987).

According to Jordanes, the Goths moved from Scandza (Scandia) to Gothiscandza near the delta of the Vistula,2 then southeast in the 2nd century, splitting around the Black Sea. The Ostrogoths occupied the area north of the Black Sea and in the Crimea, while the Visigoths settled west of the Black Sea and the Dniepr and north of the Danube, in the Roman province of Dacia. In 376 the Visigoths crossed the Danube from Dacia to Moesia, then to Thrace, where they defeated and killed Emperor Valens in 378.

Around this time the Ostrogothic inscription was made on the golden ring of Pietroassa (Pietroasele, Romania). As of August 14, 2015, the Kiel Rune Project lists fifteen interpretations, but gutaniowihailag, defended by Bammesberger (1994: 5–6) and MacLeod and Mees (2006: 173), is confirmed by a republished photo (Mees, 2004: 78–79). Whether this indicates Gutanio or ‘Gothic female’ (MacLeod & Mees, 2006: 173) or gutani-o remains in dispute. If the latter, the first word can be Gutani (Goth. *Gutanē ‘of the Goths’); hence, the old interpretation gutanī ō(þal) wī(h) hailag (‘possession of the Goths, sacred, holy’) defended by Nedoma (2010: 29ff., 44–45). The last word is likely hailag ‘holy’ (not in Wulfila’s text), and wi(h) may be wīh ‘sanctuary’ or ‘sacred’, comparable to Wulfila’s weihs ‘holy’ or weiha ‘priest.’

The Visigoths sacked Rome in 410 during the 395–410 reign of Alaric (ca. 370–410; Goth. *Alareiks ‘king/ruler of all’)(). Theoderic (ca. 454/5–526); Goth. *Þiudareiks ‘people-king/ruler’, the Ostrogothic king of Italy [493–526]) grew up in Roman Constantinople. Before 475, he led his people down the Danube from Pannonia to Lower Moesia. Theoderic entered Italy in 489 and by 490 controlled most of mainland Italy and Sicily. In 493 he captured Ravenna, established an Ostrogothic empire, and reigned thirty-three years. By 497, his rule of Italy was recognized by the eastern Roman government.

Homoianism (cf. Arianism: Brown, 2007) may be reflected in some Gothic passages (Quinlin, 2007; Pakis, 2008) but not others (Stutz, 1966: 12ff.; Falluomini, 2015: 15). Of the seven buildings for Arian worship in Ravenna, three survive, including Theoderic’s church dedicated to St. Martin, now Basilica di Sant’ Apollinare Nuovo (see Falluomini, 2015: 28–29, w. rich lit.). During Theoderic’s reign, manuscripts of the Gothic Bible were recopied, and Gothic documents from Ravenna date to this period.

In 552/3, the Ostrogoths were driven from Italy. Some remained in the Crimea through the 16th century, and possibly into the 18th on the evidence of influence on the Greek dialect spoken there.

Crimean Gothic

Crimean Gothic has 86 entries (101 lexical items), elicited in 1560/1562 by Flemish diplomat Ogier Ghiselin de Busbecq (1522–1592) from two Crimeans—one a Greek, and the other possibly a Crimean Goth but more competent in Greek. De Busbecq’s report was written in Latin and published in France in 1589 from a pirated copy.3

Nothing definitive can be said about the dialectology of Crimean except that it is not a direct descendent of Wulfila’s Gothic. It is probably a variety of East Germanic parallel to Gothic (Zadorožnyj, 1960; Stearns, 1978; Ganina, 2011). A West Germanic dialect influenced by Gothic (Grønvik, 1983) would entail many direct loans from Gothic. For instance, those two alone have a /d/ in *fedwōr ‘four’ (Goth. fidwor, Crim. fyder) and /z/ where the rest of Germanic has rhotacism (cf. Ringe, 2012: 34; Stiles, 2013: 15).

Moreover, Crimean ada ‘egg’ has Verschärfung of the Gothic kind (Ganina, 2011: 108–109), and forms with -d- do not exist in North Germanic (ON egg) or West Germanic (OHG ei ‘egg’) (cf. Stiles, 2013: 7). For verschärfung, or Holtzmann’s law (Holtzmann, 1835: 862f.), compare Gothic twaddje and Old Norse tveggja versus Old High German (Isidor) zweiio (‘of two’). A succinct overview and references can be found in Kroonen, 2013: xxxviii ff.

Differences between biblical Gothic and recorded Crimean are not surprising, given that (1) at least 10 centuries separate the two, (2) the informants may not have been native Crimean Gothic speakers, (3) Flemish misperceptions are rampant (e.g., tria vs. Goth. þriu ‘three’), and (4) transcription errors abound (e.g., goltz for gulþ ‘gold’).

Wulfila and Gothic Documents

Probably of Anatolian parents enslaved by Goths in Western Cappadocia, Wulfila lived ca. 307/311–ca. 381/383 (Stutz, 1972: 388; Metzger, 1977: 384–385). What little is known about him is from his student and later bishop of Moesia, Auxentius, and the 5th-century church historians Philostorgios, Socrates, and Sozomen. His name was variously rendered as Ulfila (by Auxentius), Vulphilas (by Cassiodorus), and Vulfila (by Jordanes), etc., which point to Gothic Wulfila, ‘little wolf’ (Ebbinghaus, 1991a; Gothic Etymological Dictionary 375).

Constantine sanctioned Christianity in 325. Between 337 and 341 (Sivan, 1996: 381ff.), Wulfila was consecrated bishop of the Visigoths for Gothia (in eastern Dacia: (Vasiliev, 1936: 12ff.; Kokowski, 2007). He preached for forty years in Greek, Latin, and Gothic (cf. Burton, 2002). During that time, he began his translation of the Gothic Bible and, persecuted by Athanaric, led his people across the Danube in 347/8. Around 369 (traditional date) he finished many religious texts for the Goths of Moesia, on which see Velkov, 1989. Auxentius mentions treatises and commentaries by Wulfila but not a Bible translation (Griepentrog, 1990: 33ff., w. lit.), implying that others were involved (Gryson, 1990: 13).

The Gothic Corpus

The Bible translation is the bulk of the extant Gothic corpus, fewer than 70,000 words (Snædal, 2009a). Most of it is in codex Argenteus (now in Uppsala, Sweden), produced ca. 520 in Ravenna, Italy (Munkhammar, 2011; cf. Ebbinghaus, 1997; Burton, 2002). Written in two hands (visible in von Friesen & Grape, 1927), with differences between Matthew–John (Scribe 1) and Luke–Mark (Scribe 2), Argenteus contains 188 of the original 336 parchment leaves, one found in 1970 (Szemerényi, 1972).

Copies of the Pauline Epistles (less Hebrews), especially codices Ambrosiani A and B, attest some textual modifications but share 19 errors that point to a common ancestor (Friedrichsen, 1939: 62–127). Ambrosianus D has fragments of Nehemiah. There were two Gothic-Latin bilinguals: the destroyed codex Gissensis, with a double parchment leaf of Luke 23–24 (reconstructed in Snædal, 2003), and codex Carolinus (beginning of the 6th century), with four leaves of Romans 11–15 (Falluomini, 1999). The Gothic text on the left suggests a Latin-speaking audience (ibid.; Falluomini, 2015: 29–30).

Skeireins (aíwaggeljons þaírh Iohannen), which translates as ‘Explication (of the Gospel according to John),’ was so-named by Maßmann (1834), its first editor. Of the eight parchment leaves (sixteen pages), leaves 1, 2, and 5–7 are in codex Ambrosianus E, and leaves 3, 4, and 8 are in codex Vaticanus 5750, all by the same scribe, with corrections by a second (Bennett, 1960: 26–27). The original length is unknown. A Greek source is likely because of its classical stylistic features (McKnight, 1897; Bennett, 1960: 41–42; Friedrichsen, 1961a). Friedrichsen (1961b, 1962, 1963, 1964, 1970) attempts to reconstruct the original by Wulfila’s contemporary, Theodorus, bishop of Heraclea (cf. Snædal 2015a), who wrote commentaries on John and Matthew. Schäferdiek (1981) argues that theologically, Skeireins must date to the second quarter of the 4th century. Griepentrog (1990) takes the next leap and claims that Wulfila translated Skeireins. Though usually considered to be later, a contemporaneous work is plausible with team involvement. Stylistic differences from the Gothic Bible can be due to different translators and/or the different text type and linguistic content (Bennett, 1959b).

To summarize, Gothic is attested in (portions of) the Bible translation; eight leaves of the Skeireins ‘commentary’; a fragment of a calendar of martyrs (codex Ambrosianus A, leaf 197); Gotica Veronensia (end of the 5th century or beginning of the 6th), consisting of at least 27 marginal notes (about 13 legible) in a collection of Latin homilies (Marchand, 1973b; Snædal, 2002b); a ‘Vandal’ epigram in a Latin anthology (Snædal, 2009b); a land transfer title deed (ca. 550/551) from Arezzo, Italy (only a copy from 1731 is extant); four debt settlement deeds from Naples (“Naples Deed,” Germanica: The Germanic Encyclopedia and Resource Library)—all ultimately from Ravenna (Penzl, 1977); a few runic inscriptions (Ebbinghaus, 1990), but see Nielsen, 2011 and Snædal, 2011a; two partial abecedaria in the Salzburg-Vienna Alcuin manuscript (9th or 10th century; Sickel, 1875: 471ff.) with letter names like the Old English and Old Norse rune names; and several other fragments, borrowings, names, and 60 margin glosses that reveal a tradition of Gothic textual exegesis (Stutz, 1972: 381, w. lit.). See the complete list and texts in the Wulfila Project website and Snædal (2013a: Vol. 1). For descriptions, cf. Stutz, 1966 and Gryson, 1990.

Discovered in 2010, codex Bononiensis is a bifolium (two pages recto and verso) on a palimpsest (first quarter of the 6th century) from the Basilica of San Petronio in Bologna, Italy. The Gothic text (Finazzi & Tornaghi, 2013; cf. Fragmenta Bononiensia; Falluomini, 2014) was scraped off but was mostly visible behind the letters superimposed to reuse the manuscript for Augustine’s De civitate Dei. The Gothic is an eclectic composition of Old and New Testament quotes, likely part of a sermon or liturgical prayer, with some passages and words previously unattested. Verses parallel to those in codex Argenteus do not differ in substance.

The Bible Translation

The Gothic version contains many foreign words (with variant spellings) and calques (Schulze, 1905; Gaebeler, 1911; Lühr, 1985; Davis, 2002; Casaretto, 2010, 2014; Snædal, 2015a). Divergences in the rendering of the foreign words and constructions are partly stylistic or interpretive and partly due to Wulfila’s several sources and team of translators (Friedrichsen, 1939: 259, 1961a: 103–104; Gryson, 1990: 13; Falluomini, 2015: 147).

Different word and form densities may signal multiple translators. For instance, ‘devil’ is usually diabulus in the Gospels and unhulþa in the Epistles (Jellinek, 1926: 191). In the dative-accusative plural of ‘us’, unsis predominates in the Gospels and uns in the Epistles, especially 2 Corinthians (Snædal, 2010). The neuter nominative-accusative singular of the strong adjective has no suffix (e.g., þein ‘your’, all ‘all’) or -ata (þeinata, allata). On the use of -ata, Scribe 1 (21 -ata) and the Epistles (19 -ata versus 239 bare formations) pattern together, different from Scribe 2 with 36 -ata (Ratkus, in press).

The Gothic Bible is not uniform for a variety of reasons. Ignoring copy errors, these include scribal preferences (Friedrichsen, 1926, 1930), revisions in Ravenna (cf. Stutz, 1972), variations among Wulfila’s translators (Friedrichsen, 1961a: 103–111; Griepentrog, 1990: 33ff.; Falluomini, 2009: 312; cf. Jellinek, 1926: 10–11). These must not be confused with stylistic features (Falluomini, 2015: 82–88), such as avoidance of the same Gothic word in close succession (cf. Stutz, 1972: 380), or interpretive variations in Aktionsart, viewpoint, theological factors, and so on (e.g., Götti, 1974; Lloyd, 1979).

The Greek Vorlage

The primary source or model (‘Vorlage’) for the Gothic translation was the Greek New Testament, but 5,400 manuscripts with 200,000 to 300,000 differences are extant (Ehrman, 2000: 443), and the Gothic version does not reflect any one of them entirely.

The Gothic Gospels are sequenced Matthew–John–Luke–Mark in the misleadingly named ‘Western’ order (Burton, 1996b: 82). This is the order followed by the Greek-Latin codex Bezae (ca. 400) (Parker, 1992), with only Luke complete, and some 10 other sources, including Peshitta Syriac manuscripts (Metzger & Ehrman, 2005: 276–277).

It is generally agreed (e.g., Hug, 1821: 462–489; Kauffmann, 1911a; Friedrichsen, 1961a; Metzger, 1977: 384–385; Ratkus, 2011; Falluomini, 2013a, 2015) that Wulfila used a Byzantine text (e.g., Robinson & Pierpont, 2005)4 more than the Alexandrian manuscripts (e.g., Nestle et al., 2012), the main source of the Latin Vulgate of Jerome (347–420).

To illustrate this issue, one difference involves the ending of the Lord’s Prayer:

(1)

About a dozen manuscripts containing Matthew have the doxology (Falluomini, 2015: 149). The very early Didache (end of the 1st century or beginning of the 2nd; see Didache (Teaching of the Twelve), which bears similarities to Matthew, also has a variant: hóti soũ estin hē dúnamis kaì hē dóxa eis toùs aiõnas (Didache 8.2), ‘because yours is the power and the glory into the eons’. The main Byzantine text has the complete doxology (Robinson & Pierpont, 2005: 11), as does codex Brixianus and several other pre-Vulgate manuscripts (cf. Jülicher (1972): 31; cited under Jülicher, 1963–1976). Both the Didache and the doxology are ignored by Jerome’s Vulgate. See the Gospel According to St. Matthew, Parallel Greek New Testament.

Burkitt (1899) argued that the Gothic translation influenced northern Italian manuscripts of the Vetus Latina, or (misnamed) ‘Old Latin’ (Bible), a pre-Vulgate Latin translation of a scantily preserved Greek text (see Vetus Latina). One of those is codex Brixianus (6th century), a purple manuscript with silver ink, like Argenteus. Some Brixian readings differ from other Vetus Latina versions and the Vulgate but agree with the Gothic text (Burkitt, 1899; Metzger, 1977: 386, w. lit.). See the extensive literature in Pakis, 2010.

In the Gospels (excluding Matthew), the historical present is prompted only 10 times by the same construction in Greek, while deviations from the Greek agree 138 times with the Vetus Latina (Pakis, 2010). But this may be a stylistic feature of Gothic and Latin.

For agreement of codex Brixianus exclusively with the Gothic, cf. ustáuh (Mk 1:12) ‘led out’, not a match to Vetus Latina or Vulgate expulit ‘drove out’, Vetus Latina tulit ‘led’, or dūxit ‘id.’, for example (cf. Jülicher, 1970: 3; cited under Jülicher, 1963–1976), but only to Brixian ēdūxit ‘led out’. However, Brixianus can pattern with the Greek against the Vulgate and Gothic text (Stutz, 1972: 389, w. lit.), and some other pre-Vulgate manuscripts show similarities to the Gothic (Falluomini, 2015: §5.2).

Especially in Luke and the Epistles, the Gothic sometimes agrees with Latin or Alexandrian texts, but non-Byzantine readings in many different manuscript traditions imply their presence in the Byzantine area and Wulfila’s Vorlage prior to stabilization of the proferred Byzantine readings (Friedrichsen, 1959; Gryson, 1990; Falluomini, 2013a, 2015).

Unlike the reason(s) for them, relationships between the Latin versions and the Gothic Bible are manifest (Hunter, 1969). Due to codicological and text-critical similarities to codex Brixianus (Gryson, 1990; Falluomini, 2013a, 2013b, 2015: 33), Argenteus was once hypothesized to have been Latinized (Friedrichsen, 1961a: 68; Metzger, 1977: 386). Nevertheless, the “Gothic and Latin may represent independent renderings of the same Greek readings” (Falluomini, 2013b: 146).

To conclude this section, “Wulfila probably used, beside a Greek Vorlage which transmitted an early Byzantine text, a Latin translation, in order to better render difficult passages of the Greek. This would justify some similar renderings in the Gothic and Latin versions” (Falluomini, 2015: 147).

Gothic Alphabet and Pronunciation

Wulfila invented the Gothic alphabet (see Lendinara, 1992). Most of the letters were adapted from a Greek script, as shown in Table 1 (variant forms in Marchand, 1973a: 18–22). The second row contains the numerical value; the third, the Greek letter; the fourth, the English letter equivalent; and the last row, the late Gothic letter name.

Table 1: Gothic alphabet

𐌰

𐌱

𐌲

𐌳

𐌴

𐌵

𐌶

𐌷

𐌸

1

2

3

4

5

6

7

8

9

Α

Β

Γ

Δ

ε

(υ‎)

Ζ

(η‎)

(Θ‎)

a

b

g

d

e

q[u]

z

h

th

aza

bercna

geuua

daaz

eyz

quertra

ezec

haal

thyth

𐌹Ï

𐌺

𐌻

𐌼

𐌽

𐌾

𐌿

𐍀

𐍁

10

20

30

40

50

60

70

80

90

ι

κ

λ

Μ

Ν

(υ‎)

Π

ϙ

i

k

l

m

n

j

u

p

iiz

chozma

laaz

manna

noicz

gaar

uraz

pertra

𐍂

𐍃

𐍄

𐍅

𐍆

𐍇

𐍈

𐍉

𐍊

100

200

300

400

500

600

700

800

900

(Ρ‎)

(S)

Τ

Υ

ϝ

χ

Ω

(ϡ‎)

r

s

t

w

f

wh

o

reda

sugil

tyz

uuinne

fe

enguz

uuaer

utal

The letter names (see Ganina, 2007; Seebold, 2010) and some forms (e.g., for /f/, /þ/, /j/) suggest runic input. Viehmeyer (1971) derives them all from runic. Most Gothic letters have a Greek shape, alphabet order, numerical value, and sound (Granberg, 2010, 2013), but runic input is possible (Mees, 2004; Raschellà, 2011; pace Marchand, 1955, 1973a; Ebbinghaus, 1996). Snædal (2015b) derives the entire Gothic alphabet from the Greek, with j and q influenced by the Latin alphabet.

The Latinate 𐍃s has a sigmatic variant in codex Ambrosianus B and D and in some of the smaller fragments. If 𐍈‎/hw/ is a wheel by the acrophonic principle for Gmc. *hwehwlaz/*hwegwulaz ‘wheel’ (Wagner, 1986), its runic name is slightly irregular. Snædal (2015b) claims that Θ‎/θtheta was used for 𐍈 /hw/. By that account, it is almost as if the decision to use a Latin / runic F for /f/ left the perceptually close þ‎ open for [θ‎], which in turn left Θ‎/θ‎ available for /hw/. It is just as plausible that ‘classical’ theta was used for /hw/ and cursive theta or runic thorn for /þ‎/.

The numbers 90 and 900 have no (known) sound value, and the latter occurs only in the Salzburg-Vienna manuscript (cod. Vindobonensis 795). The Gothic Bible spells out niun hunda (Neh 7:39) ‘nine hundred’.

The letters 𐌹‎ and ï were positional variants: ï was syllable-initial (e.g., ïzei ‘(he) who’, saiïþ ‘sows,’ sauïl ‘sun’) (Rauch, 2011: 5; cf. Grammatik der gothischen Sprache 25). ï also occurs in some foreign names, like Gaïus (Rom 16:23A) (Gk. Γάϊος‎, Lat. Gaius). Both i and ï are transcribed i.

𐍇 is Greek χ‎ (chi). It occurs mainly in religious words, for example (Ambros. B ) for Xristus ‘Christ’; pasxa ‘Passover; paschal feast’ (Jn 6:4, 18:28, 18:39) beside paska (Mk 14:12 [two times], 14:14, etc. [six times in all]). 𐌺k is normal for Greek χ‎; cf. Twkeikus (Eph 6:21B, Col 4:7A), Twkekus (Col 4:7B) for Τυχικός‎ (Gaebeler, 1911: 19–20, 48–50; cf. Grammatik der gothischen Sprache 29, Marchand 1973a: 26, Gotische Grammatik 67).

Phonological System 1: Consonants

Table 2 contains the inventory of Gothic consonantal segments (labvel = labiovelar).

Table 2: Gothic consonantal system

labial

coronal

palatal

velar

labvel

low

stop

vcl

p

t

k

kw

vcd

b

d

g

(gw)

continuant

vcl

f

ɵ

s

[x]

hw

h/χ

vcd

[β‎]

[ð]

z

[ɣ]

sonorant

nasal

m

n

[ŋ]

liquid

l

r

glide

j

w

Alternations with voiceless fricatives suggest that b, d, and g were voiced continuants (β‎, ð‎, ɣ‎) after vowels and possibly after liquids, and stops (b, d, g) after consonants and when geminated (see Hench, 1897; Marchand, 1973a: 64–8, 76; Harbert, 2007: 50; Rauch, 2011: 47–48; Kotin, 2012: 64–65).

Gothic-specific word-final devoicing affected only continuants, for example ƕ‎as ‘who’: ƕ‎azuh ‘each’; acc goþ : gen godis ‘good’; nom twalif : gen twalibe ‘12’. Contrast lamb ‘sheep’ (nom/acc pl lamba). 𐌲 g did not alternate (magan* ‘can’: pret mag), possibly because of the absence of a letter for velar [x], if indeed h h was [h] word-initially and uvular [χ‎] elsewhere (Vennemann, 1972: 878–879). For Roberge (1983), the roughly 226 final -d, -b, -z (in decreasing order of frequency) suggest post-Wulfilian devoicing, which may also account for word-final -g, if it is just [g] (Roberge, 1984: 337).

Rather than not fully developed (e.g., Bernharð‎sson, 2001), verner’s law was likely lost early in Gothic (see Verner, 1875; d’Alquen, 1988; Suzuki, 1994; Schaffner, 2001; Liberman, 2010; Kiparsky, 2010). Alternations are isolated, for example (ga-)filhan ‘conceal; bury’ : PPP (ga-)fulgins; áih : áigum (~ áihum) ‘possess’ (1/3 sg : 1 pl); -saízlep(-) [2x] ~ (-)saíslep(-) [2x] ~ -saisleip [1x] ‘slept’. Note the lack of alternation in wisan ‘to be’, wesum ‘we were’ vs. OE wesan : wǣ‎ron, or Gothic saian : saíso (for *sezō-) ‘sow : sowed’ vs. ON : sera ‘id.’

Unique to Gothic is thurneysen’s law (Thurneysen, 1898), by which the first and second continuants “received a converse specification in terms of voice” (Suzuki, 1992: 41), for example waldufni ‘power’ (< *walðufni) but fastubni ‘fasting’ (< *fastuβ‎ni). This rule was generalized to continuants after noncontinuants, as in witubni ‘knowledge’, but affected only certain derivational suffixes and was mostly leveled (e.g., -iþ‎a occurs exclusively as -ida only in áuþ‎ida ‘desert’ and waírþ‎ida ‘worthiness’). See Woodhouse, 2000.

Variant spellings like bringiþ (Lk 15:22) for briggiþ ‘brings’ confirm the [ŋ‎] value of g(g/k) as in Greek (Snæ‎dal, 2011b); cf. 𐌰𐌲𐌲𐌹𐌻𐌿𐍃aggilus [aŋ‎gilus] ‘angel’ (Gk. ἄγγελος‎ [áŋ‎gelos] ‘messenger’; ‘angel’). Prefixes like in- and un- keep [n] (ibid.; cf. Penzl, 1950).

Gothic 𐌵 q is always voiceless and represents /kw/, as in qrammiþ‎a ‘moistness’ (discussion in Kotin, 2012: 63); 𐍈 ƕ‎ represents /hw/ (and not a sequence [hw]) because it reduplicates as a single consonant (ƕ‎aíƕ‎ot ‘boasted’) (pace Voyles, 1968: 721f.; sC- behaves differently cross-linguistically), counts as one consonant for class 5 verbs, like saíƕ‎-an ‘to see’, whose roots end in a single consonant, and is always voiceless (Bennett, 1959a, 1967); /gw/ occurs 98 times, but there is no internal evidence for the etymological contrast between, for example, 𐍃𐌰𐌲𐌲𐍅𐍃saggws [saŋ‎gws] ‘song’ and 𐍄𐍂𐌹𐌲𐌲𐍅𐍃‎ [triggws] ‘true’ (Snæ‎dal, 2011b).

The glide /j/ in Gothic (𐌾), as in all of Germanic, is conventionally transcribed j. That it was indeed a glide, and not a fricative (Vennemann, 1985: 206–217), is maintained by Barrack (1997: 5), Gotische Grammatik 57, Pierce (2007), and Kotin (2012: 62).

Phonological System 2: Vowels

In terms of frequency, the order of the vowel letters is 𐌰 a > 𐌹 ï i > 𐌿 u > 𐍉 o > 𐌴 e (Grammatik der gothischen Sprache 25–26, 28). The customary interpretation is in Table 3.

Table 3: Gothic vowel system

short

long

front

back

front

back

high

i

(y)

u

ī

ū

mid

high mid

ē

ō

low mid

ɛ

ɔ

ɛ̄

ɔ̄

low

a

ā

Another hypothesis is that the relevant contrast was between tense 𐌴/e/, 𐍉 /o/ and lax 𐌰𐌹 /ɛ‎/, 𐌰𐌿 /ɔ‎/ (see Gotische Grammatik 27–28, 48–49). But some alternations suggest that Gothic still had the long/short contrast (cf. Voyles, 1968: 727; Vennemann, 1971; Beck, 1973).

One alternation is antevocalic lowering: /ē ō‎/ > [ɛ̄‎ ɔ̄‎]/__V (Paul, 1880, 1882):

(2)

This alternation occurred in stressed syllables (contrast waíwoun ‘they blew’: Fullerton, 1991:12) and is consistent with the hypothesis that vowels remained long at least in root syllables, for example áins̄‎ns] ‘one’ < *ó‎inos (Voyles & Barrack, 2009:53; Rauch, 2011:51–60). For unstressed syllables, see Boutkan, 1995.

By breaking, radical/stressed /i/, /u/ were lowered to ai [ɛ‎], au [ɔ‎] before r, h, ƕ‎ (all [+low] if h was [χ‎] except word-initially), for example faíhu ‘chattels’, haú‎rn ‘horn’, and saíƕan ‘to see’; /ī/ and /ū/ were unaffected: skeirs ‘clear’, leiƕ‎an ‘to loan’, þū‎hta ‘it seemed’.

Breaking in pre-Gothic may have blocked raising: *beran- > Gothic /bɛran/ <bairan> ‘to bear’ (Bennett, 1952; Cercignani, 1979b; Voyles & Barrack, 2009:54), rather than *beran- > *biran- > baíran.

Breaking was opaque. Factors include (1) analogy, for example nih ‘and not’ (neg ni), hiri ‘come here’ (hi- ‘this’, but see Heffner, 1929), nuh ‘then’ (nu ‘now’) (see Voyles, 1968: 740; Cercignani, 1979b: 277–278; Kotin, 2012: 431)—all often claimed to be unstressed like -uh ‘and’ (e.g., Nielsen, 2010: 434); (2) composition, as in duƕ‎e [to/for what] ‘why?’; (3) assimilation, for example urrists* ‘resurrection”’< *us#ris-ti- (Nominale Wortbildung der gotischen Sprache 505); (4) generalization at the boundary (Cercignani, 1979a) or proclisis (Ebbinghaus, 1991b) of reduplicating [ɛ‎], for example laí-lot ‘let’; and (5) occasional aí, aú in unstressed position before /r/: dat sg paúrprái (Lk 16:19) beside paúrpurái (Mk 15:17, 15:20) ‘in/with purple’; contrast spaíkulatur (Mk 6:27) ‘bodyguard, executioner’, and the very strange Nabukaúdaúnaúsaúr (Bn.2v.22f.) ‘Nebuchadnezzar’.

A third alternation is the sievers’ law realization of *-je- as -ji- after a light syllable and -ei- ([ī‎]) after heavy, for example sitjiþ ‘sits’ but sokeiþ ‘seeks’ (Beade, 1972; Suzuki, 1995; Barrack, 1989, 1998; Kiparsky, 2000; Riad, 2004; Pierce, 2006, 2013b).

Since phonological alternations can be residues of an earlier phonological system, spelling variations are a safer criterion. By Snædal’s count (2013b: 287), in nonborrowed vocabulary there are 74 (possibly lexically conditioned; see Bethge, 1900: 33–34; Marchand, 1973a: 50–51) occurrences of ei for /ē‎/ and 44 of e for /ī‎/, but only three ei for /ĭ‎/ and nine i for /ī‎/. The greater confusion among the long vowels suggests a vowel shift in progress (cf. Nielsen, 2010: 431), implying that Gothic still had the long/short contrast.

Additional evidence is from line-end word breaks, like ak-ran ‘fruit’ but hlei-þ‎rái ‘hut’ (dat sg). If due to prokosch’s law (Prokosch, 1939: 140), by which the main-stressed syllable is preferentially bimoric, a syllable break [ak.ran] is licensed but not *hleiþ.rai because ei is long (Riad, 1992: 87, 2004: 188–189; Pierce, 2013b). But there are breaks for which Prokosch’s law is irrelevant, for example iupaþ-ro ‘from above’ (Frey, 1989), and the Gothic syllabification -VC.RV- (Riad, 1992: chapter 2, 2004; Suzuki, 1995; Pierce, 2013b) was inherited from Indo-European (Miller, 1994; Byrd, 2010). Together, these two (nonexhaustive) conditions for line-end word division support the hypothesis that Gothic retained long vowels and heavy syllables.

It is generally thought that 𐌰𐌹 ai and 𐌰𐌿 au can be short /ɛ/, /ɔ/ or long /ɛ̄/, /ɔ̄/. Another hypothesis is that the digraphs can represent diphthongs, distinguished by a diacritic not in the Gothic script: fhu ‘chattels’, fr ‘before’ with short vowel, versus máizo ‘more’, sunáus ‘son’s’ with diphthong.5 The counterevidence is laid out in Grammatik der gothischen Sprache 30–3; Bennett, 1949; Jones, 1958; and Marchand, 1973a: 74ff. Beck (1973) argues convincingly that final -ai and -au were monophthongized. D’Alquen (1974) and Kotin (2012: 45–46), contrarily, assume post-Wulfilian monophthongization, but alleged ái/áu and aí/aú seem parallel, including the reflex of antevocalic lowering and that before r, h, and ƕ (Snædal, 2013b: 287), which would require a number of late orthographic practices, as would aw for the diphthong (e.g., Pawlus /pau̯lus/), in contrast to au for the short vowel (e.g., Paúntius for Pontius) (cf. Bennett, 1949).

In summary, the letters 𐌴e and 𐍉o are long /ē‎/, /ō‎/ and by convention need not be so indicated. The vowel /ī‎/ is written 𐌴𐌹 ei. The other vowel letters can indicate long or short vowels but are not distinguished in the orthography; cf. rum /rū‎m/ ‘room’ versus sunus /sunus/ ‘son’. Long /ā‎/ (from *-anχ‎-) is rare (Vennemann, 1971: 104), for example fahan /fāχ‎an/ ‘to grasp, seize’, 1/3sg pret brahta /brāχ‎ta/ ‘brought’; 𐌰a in the suffix *-ā‎r(i)ja- (from Latin -ā‎rius: Miller, 2012: 140ff.), for example bokareis ‘scribe’ (not *bokarjis < *-ărja-) is ambiguous (Nominale Wortbildung der gotischen Sprache 423).

The Greek letter Υ‎ was borrowed as Gothic 𐍅, usually /w/ but was also used for υ‎ and οι‎ in Greek loanwords, suggesting that both υ‎ and οι‎ were pronounced /y/ at that time, for example 𐌻𐍅𐍃𐍄𐍂𐍅𐍃 Lwstrws (2 Tim 3:11) /lý‎strys/ (probably not so pronounced by monolingual Goths: Collitz, 1925; Bennett, 1959a; Jones, 1960) = Gk. ΛύστροιςLústrois ‘in Lustra’, a city in Asia Minor. Those who propose /ȳ‎/ for iu [iu̯‎] (Riad, 1992: 56) or [iu] (Voyles, 1968: 722f., 1981: 11f.), as in tiuhan ‘to lead’, do not explain why it is spelled iu and never 𐍅 w (Grammatik der gothischen Sprache 35).

Numerals

The numbers from ‘one’ to ‘three’ are declined in all genders and cases. ‘Five’ to ‘eight’ and ‘ten’ are indeclinable. ‘Four’ to ‘nineteen’ are normally undeclined, but some dative and genitive forms occur, for example gen niune ‘9’ (Lk 15:7), twalibe ‘12’ (5 times), dat ainlibim ‘eleven’ (Mk 16:14S, 1 Cor 15:5A), fimftaihunim ‘fifteen’ (Jn 11:18). Hund* (neuter plural hunda) ‘hundred’ and þūsundi ‘thousand’ are nouns.

As in Greek, alphabet letters also functioned as numerals (Table 4). A sample follows:

Table 4: Gothic numerals

𐌰

1

áins

𐌹𐌰

11

áinlif*

𐌻

30

þrins tiguns (acc)

𐌱

2

twái

𐌹𐌱

12

twalif

𐌼

40

fidwor tigjus*

𐌲

3

þrija (nom/acc.n)

þrins (acc.m)

𐌹𐌲

13

𐌽

50

fimf tigjus*

𐌳

4

fidwor

𐌹𐌳

14

fidwortaíhun

𐌾

60

saíhs tigjus*

𐌴

5

fimf

𐌹𐌴

15

fimftaíhun*

𐌿

70

sibuntehund

𐌵

6

saíhs

𐍀

80

ahtáutehund

𐌶

7

sibun

𐍁

90

niuntehund

𐌷

8

ahtáu

𐍂

100

taíhuntehund

𐌸

9

niun

𐍃

200

twa hunda

𐌹

10

taíhun

𐌺

20

twái tigjus*

𐍆

500

fimf hunda

Numerical letter symbols are set off by a horizontal stroke under or over, the latter with or without raised dots (Gotische Grammatik 22). Combinations include acc = þrins tiguns [three tens] ‘30’, = taíhuntehund [ten tens] ‘100’, and = taíhuntehund jah þrins tiguns [and 3 tens] jah þrins [and 3] ‘133’. A complete list appears in Snædal (2013a: ii. 626–628).

For -lif in twalif from PIE *likw-, compare Lith. -lika (e.g., dvýlika ‘twelve’) (Grienberger, 1900: 14f.; Ringe, 2006: 116; Kotin, 2012: 170). On the origin of -tehund, see Ringe, 2006: 206. The variant spelling -taíhund is remodeled after taíhun ‘ten.’

The Nominal System

Gothic nouns are inflected for gender (masculine, feminine, neuter), number (singular and plural), and case (nominative, vocative, accusative, genitive, dative). The vocative has the form of the accusative and/or is syncretized with the nominative. Demonstratives and pronominals have a residual instrumental (e.g., þe ‘by this’, biþe ‘while’) and ablative (e.g., jáinþro ‘from there’).

The two largest noun classes were thematic (-a-) stems (Goth. dags ‘day’) and feminine -ō- stems (Goth. giba ‘gift’), plus similar formations (-ja-, -jō-, etc.). Other stem types include -i- (runic -gastiz, Goth. gasts ‘guest’), -u- (sunus ‘son’), -n- (guma ‘male human being’, qino‘woman’), -nd- (nasjands ‘savior’), -r- (broþar ‘brother’), consonant (baúrgs ‘castle, citadel’), mixed -C- and -n- (manna ‘man[kind], human being, person’ sometimes overlapping with waír ‘adult male, man’).

Table 5 contains a synopsis of the nouns above with unattested forms (indicated by an asterisk * after the form) supplied from many other nouns.

Table 5: Gothic noun classes

-a- stem

-ja- stem

-wa-

-Ō-

-jŌ-

masc

neut

masc

masc

neut

neut

fem

fem

‘day’

‘word’

‘shepherd’

‘army’

‘clan’

‘tree’

‘gift’

‘girl’

sg

NOM

dags

waúrd

haírdeis

harjis

kuni

triu*

giba

mawi

VOC

dag*

haírdi*

hari*

(mawi)

ACC

dag

waúrd

haírdi*

hari*

kuni

triu*

giba

máuja

GEN

dagis

waúrdis

haírdeis

harjis

kunjis

triwis*

gibos

máujos

DAT

daga

waúrda

haírdja*

harja*

kunja

triwa*

gibái

máujái

pl

N/VOC

dagos

waúrda

haírdjos

harjos*

kunja

triwa*

gibos

máujos*

ACC

dagans

waúrda

haírdjans

harjans*

kunja*

triwa*

gibos

máujos

GEN

dage

waúrde

haírdje*

harje*

kunje*

triwe*

gibo*

máujo*

DAT

dagam

waúrdam

haírdjam

harjam*

kunjam*

triwam

gibom*

máujom*

-i-stem

-u-stem

-n- stem

-nd-stem

-r-stem

-C-stem

mixed

masc

masc

masc

fem

masc

masc

fem

masc

‘guest’

‘son’

‘male’

‘woman’

‘savior’

‘brother’

‘citadel’

‘man’

sg

NOM

gasts

sunus

guma

qino

nasjands

broþar

baúrgs

manna

VOC

gast*

sun(á)u

nasjand*

(broþar)

ACC

gast

sunu

guman*

qinon

nasjand

broþar

baúrg

mannan

GEN

gastis*

sunáus

gumins*

qinons

nasjandis

broþrs

baúrgs

mans

DAT

gasta*

sunáu

gumin

qinon

nasjand

broþr

baúrg

mann

pl

N/VOC

gasteis

sunjus

gumans*

qinons

nasjands*

broþrjus

baúrgs*

ma(nna)ns

ACC

gastins

sununs

gumans*

qinons

nasjand*

broþruns

baúrgs

ma(nna)ns

GEN

gaste*

suniwe

gumane

qinono

nasjande*

broþre

baúrge

manne

DAT

gastim

sunum

gumam*

qinom

nasjandam*

broþrum

baúrgim

mannam

The paradigm of sunus, like all –u- stems, has much leveling, for example nom sun(á)us, gen sun(á)us, dat sun(á)u, and npl sunjus, sunjos.

For the missing forms of guma, cf. atta ‘father; God’ attested in all cases: sg nom atta, gen attins, dat attin, acc attan, pl nom/acc attans, gen attane, dat attam.

Mawi is syntactically vocative at Luke 8:54, but Snædal classifies it as nominative. Likewise broþar is syntactically vocative at Luke 6:42 and Philemon 1:20 (Gotische Grammatik 107) but listed as nominative, even though the only occurrence of the morphologically identical fadar ‘father’ (Gal 4:6A) is labeled vocative.

A neuter -n- stem is haírto ‘heart’, gen haírtins, dat haírtin, plN/A haírtona, gen haírtane, dat haírtam.

The Demonstrative

The main Gothic demonstrative was sa, so, þata ‘this, that; the; he, she, it’ (neutral deixis); see Table 6. All the forms are attested (Snædal, 2009a: 159). For the functions, see Vilutis, 1972, 1982. Substantival uses include deictic, correlative, proleptic, anaphoric, and relative, among others. Instrumental þe occurs in biþe ‘while’ and duþe ‘for this (reason)’, for instance, and as a free form only in ni þe haldis ‘by no means’ (Sk 4.4.4).

Table 6: Neutral deixis demonstrative

masc

neut

fem

sg

nom

sa

þata

so

acc

þana

þata

þo

gen

þis

þis

þizos

dat

þamma

þamma

þizái

inst

(þe)

pl

nom

þái

þo

þos

acc

þans

þo

þos

gen

þize(i)

þize(i)

þizo

dat

þáim

þáim

þáim

Demonstratives agree with their noun in gender, number, and case. They do not translate Greek articles. In (3), sunus ‘son’, mans ‘(of) man’, háubiþ ‘head’ have no determiner in contrast to the Greek text (fourth line):

(3)

Weak and Strong Adjectives

Gothic adjectives were inflected for strong and weak forms in several stem classes in all genders and cases and in two numbers. The weak adjective is an -n- stem with forms identical to those of guma ‘man’, haírto ‘heart’, and tuggo ‘tongue’. Table 7 contains the strong and weak paradigms of -a- stem ‘blind’, reconstructed from many adjectives.6

Table 7: Strong and weak adjectives

(1)

Strong

masc

neuter

fem

sg

nom

blinds

blind

blindata

blinda

acc

blindana

blind

blindata

blinda

gen

blindis*

blindis

blindáizos

dat

blindamma

blindamma

blindái

pl

nom

blindái

blinda

blindos

acc

blindans

blinda

blindos

gen

blindáize*

blindáize

blindáizo

dat

blindáim

blindáim

blindáim

(2)

Weak

masc

neut

fem

sg

nom

blinda

blindo

blindo

acc

blindan

blindo

blindon

gen

blindins

blindins

blindons

dat

blindin

blindin

blindon

pl

nom

blindans

blindona

blindons

acc

blindans*

blindona

blindons

gen

blindane*

blindane

blindono

dat

blindam*

blindam

blindom

The extended neuter blindata* is built on þata ‘that’. There are 76 -ata forms in Gothic (Ratkus, 2011: 111–115, in press), for example juggata (4 times) ‘young’, allata (38 times) and ‘all’, þeinata (8 times) ‘your’. Only weihata ‘holy’ (Rom 7:12) is unequivocally predicative. There was competition between -ata and the bare-stem neuter (cf. Meyer 1863: 3):

(4)

The fuller form -ata could also serve as a reverential in God contexts, for example þeinata namo, waúrd þeinata (Jn 17:6) ‘thy name, thy word’, beside namo þein (Mk 5:9, Lk 8:30) ‘your name’ addressed to demons (Ratkus, in press).

Of the 2,056 adjectives and quantifiers in the main corpus of Gothic, 1,587 are strong and only 469 are weak (Ratkus, 2011: 136).

In the absence of an overt determiner, strong forms occur in most of the 608 attributive constructions (Ratkus, 2011: 108) (e.g., bagms ubils [tree bad] ‘bad tree’ and mahtins mikilos ‘great miracles’). Weak forms occur primarily with a determiner (Ratkus, 2011: 86ff.), including appositional contexts, for example Lazarus sa dáuþa (Jn 12:1) ‘Lazarus the dead’ (cf. Falluomini, 2013b: 157). The appositional structure was mostly reanalyzed as a postnominal syntactic attributive (cf. Ratkus, 2011: 233f.), for example haírdeis sa goda (Jn 10:11) (shepherd the good [one]) ‘the good shepherd’ (Gk. ho poimḕn ho kalós [the shepherd the good] ‘id.’).

Since Germanic vocatives were determined, they occur with weak adjectives when modified (19 times), for example goda skalk (Lk 19:17) ‘good servant’ (Ratkus, 2011: 141).

Gothic has 701 predicative adjectives, many with null copula (Ratkus, 2011: 117ff.).7 In (5), the Greek text, as frequently, has no overt copula either.

(5)

Strong forms also occur in predicate adjuncts:8

(6)

Words with exclusively strong inflection include jáins ‘that’ (distal), áins ‘one’, alls ‘all’, anþar ‘other, second’, halbs ‘half’, fáus ‘few’, fulls ‘full’, ganohs ‘enough’, and several others, plus possessive adjectives. Exclusively weak-inflected are áinaha ‘sole, only’, taíhswa ‘right’, and some others (Trutmann, 1972; Dvuxžilov, 1980; Ratkus, 2011).

The Present (Incompletive) Participle

Except for the nom sg masc gibands, which some consider strong but the -s may be just an old consonantal -nd- stem form, the so-called present participle (PrP) has only weak adjectival forms. The feminine is an -ī‎n- stem like managei, gen manageins ‘multitude’) (Gotische Grammatik 122–123).

The -s nominative can in principle contrast with the weak form: sa saiands ‘the sower’: sa saianda* ‘the one sowing’, but there is no minimal contrast. The choice has been claimed to be stylistic (Trutmann, 1972: 161ff.; cf. Grammatik der gothischen Sprache 130–131, Gotische Grammatik 123), but lexical aspect plays a role (Götti, 1974: 10).

Table 8 contains the paradigm of gibands ‘giving’, reconstructed from many participles (Gotische Grammatik 122–123). Only forms of the masculine (including gibands) are attested. In other words, gibando, gibandei, and other such examples do not occur.

Table 8: The present participle

masc

neut

fem

sg

nom

gibanda*

gibando

gibandei

acc

gibandan

gibando

gibandein

gen

gibandins*

gibandins

gibandeins

dat

gibandin

gibandin

gibandein

pl

nom

gibandans

gibandona

gibandeins

acc

gibandans*

gibandona

gibandeins

gen

gibandane*

gibandane

gibandeino

dat

gibandam*

gibandam

gibandeim

Among other functions, -s nominatives signal consonant-stem agentives, for example nasjands ‘savior’, dáupjands ‘baptist’, talzjands* ‘teacher’ (voc talzjand; participial talzjands ‘teaching’ is attested), frijonds ‘friend’, and fijands ‘enemy’. In all of Germanic, at least the last two are ordinary nouns declined as consonant stems: nasjands, gen nasjandis, and so on (Gotisches Elementarbuch 113; Mossé, 1956: 99; Casaretto, 2004: 437–444; Gotische Grammatik 122–123); cf. nom/acc pl bisitands ‘neighbors’, frijonds ‘friends’ versus participial nom/acc pl frijondans ‘loving’, and so on.

Since most Gothic PrPs translate a Greek participle (Metlen, 1932)—but see Bennett, 1959b for Skeireins—it is not surprising that many are calques.9

One Greek calque, of Hebrew origin (Rubio, 2009: 218), features the -s form with ‘be’ to insist on durativity in a past timeframe (Mossé, 1938: i. 21–30, 1956: 179f., 273):

(7)

‘Reflexive’ warmjands sik translates a Greek middle thermainómenos. In the Gothic corpus available to Metlen, he counted 97 examples of the past durative, 21 differing from the Greek (Metlen, 1932: 30ff.), plus an additional 26 PrPs with waírþan ‘get to be’ (34–35).

Adjunct clauses in various cases occur with the participle (Metlen, 1938; Morgaleva, 2008a, 2008b). Consider the accusative absolute in (8).

(8)

Greek soũ dè poioũntos eleēmosúnēn ‘but you doing almsgiving’ is a genitive absolute. Absolute constructions were supposedly able to be calqued because the cases could express, for example, temporality or attendant circumstances (Lühr, 2005; cf. Krause, 1994), but an Indo-European inheritance is possible (Keydana, 1997; Meier-Brügger, 2010: 379; Kotin, 2012: 324–328), even if most, if not all, absolutes are Greek prompted (Metlen, 1938: 643).

According to Dewey and Syed (2009), the prepositionless dative absolute is the default for subordination and the nominative for main clause features (on based on the thematic role of the subject). Other cases are semantically determined. Accusative interacts with the tense and (lexical) aspect of the participle and is linked to durativity or iterativity.

Personal and Interrogative Pronouns

First- and second-person pronouns (‘I/we’, ‘thou/ye2019) are inflected for all three numbers but are defective in gender and have much suppletion. See Table 9. For the history, see Katz, 2003 and Ringe, 2006: 57–58.

Table 9: Gothic personal pronouns

sg

nom

ik

þu

acc

mik

þuk

gen

meina

þeina

dat

mis

þus

du

nom

wit

jut*

acc

ug(g)kis

igqis/inqis

gen

ugkara*

iggqara

dat

ug(g)kis

ig(g)qis

pl

nom

weis

jūs

acc

uns(is)

izwis

gen

unsara

izwara

dat

uns(is)

izwis

The corresponding possessive adjectives are meins ‘my’, þeins ‘your’, unsar ‘our’, izwar ‘your’ (plus the dual form igqar ‘your’), and anaphoric seins* ‘his, hers, its (own)’.

Little is known about functional differences between uns and unsis (with -is after mis ‘to me’, etc.). Like -ata, the fuller form unsis could be more formal or ceremonious (Snædal, 2010: 313). This development may have been in progress. For instance, in the Gospels uns occurs only 16 times, four of which are in the Lord’s prayer (Mt 6:11–12): gif uns himma daga. (12) jah aflet uns þatei skulans sijáima ‘give us on this day. And forgive us that we be debtors’—with agentive skula ‘debtor’ substituted for the Greek result noun opheleímata ‘debts’ (cf. Mittermüller, 1983: 55). The short form may have been used either (1) because the Lord’s prayer was translated into Gothic before the longer forms became productive (Jellinek, 1926: 193), or (2) to make the prayer more colloquial or intimate (Snædal, 2010: 313). Since unsis outnumbers uns an average of 2 :1 in the Gospels, unsis was gaining in frequency as the more characterized alternant but had the potential for greater formality.

The third-person pronoun is, si, ita ‘he, she, it’ is inflected for all three genders and singular/plural number. See Table 10.

Table 10: Third-person pronoun

masc

neut

fem

sg

nom

is

ita

si

acc

ina

ita

ija

gen

is

is

izos

dat

imma

imma

izái

pl

nom

eis

ija

ijos*

acc

ins

ija*

ijos

gen

ize(i)

ize

izo

dat

im

im

im

The interrogative and indefinite pronoun ‘who, what’ and “any(one), any(thing)” is inflected only in the singular, as shown in Table 11. Instrumental ƕe remains a functional case attested 10 times and in the prepositional constructs biƕe “whereby, how?” and duƕe “for what, why?”

Table 11: The Gothic interrogative and indefinite pronoun

masc

neut

fem

sg

nom

ƕas

ƕa

ƕo

acc

ƕana

ƕa

ƕo

gen

ƕis

ƕis

ƕizos*

dat

ƕamma

ƕamma

ƕizái

inst

ƕe

The same forms are used for negative polarity items, for instance, ni ƕashun ‘not any(one), no one’. Gothic has a number of different negative polarity formations (Klein, 2011: 136).

The Verbal System

Verbs in Gothic are inflected for person (first, second, third), number (singular, dual, plural), tense (past/preterite and nonpast; for expressions of futurity, see Cuendet, 1924; Davis, 1929; Coleman, 1996), mood (indicative, optative), and voice (active and passive). The Indo-European mediopassive is best preserved in Gothic as a synthetic passive in the nonpast indicative and optative. The past system features a periphrastic passive. Middle functions are mostly represented by simple reflexive structures and -nan verbs. Nonfinite categories include one (mostly active) infinitive, a present and past participle (the former active, the latter passive on transitive bases), and a present active imperative. The third-person imperative is normally expressed by an optative. Verbs follow three main classes: thematic, athematic, and preterite-present. Verbs are also classified as strong or weak.

The Strong Verb

Strong verbs (ablauting type sing, sang, sung) have seven form classes, cited by four principal parts: infinitive, preterite 1/3sg, preterite 1pl (or 3pl -un), and the past participle (Table 12).

Table 12: Strong verb classes

infinitive

1/3sg pret

1pl pret

ppp

gloss

Class 1

-steigan

-stáig

-stigum

-stigans*

‘ascend’

Class 2

-biudan*

-báuþ

-budum

-budans*

‘command’

Class 3

-bindan

-band

(-bundun)

bundans

‘bind’

Class 4

niman

nam

nemum

-numans

‘take’

Class 5

giban

gaf

(gebun)

-gibans

‘give’

Class 6

faran*

for*

forum*

farans*

‘travel’

Class 7

-maítan

-maímáit

(maímáitun)

-máitans

‘cut’

Classes 1–4 have a subclass with breaking, for example 1b leiƕan, láiƕ*, laíƕum*, laíƕans* ‘loan’, 2b tiuhan, -táuh, (-taúhun), taúhans ‘lead’, 3b waírþan, warþ, waúrþum, waúrþans ‘get to be,’ 4b baíran, bar, berum, -baúrans ‘bear.’

Alternations like the above are found in other Indo-European languages (e.g., Gk. leípō ‘I leave’, léloipa ‘I have left’, élipon ‘I left’). Only Germanic productively regularized ablaut to indicate tense: e-grade present, o-grade preterite singular, zero-grade preterite plural (plus a lengthened-grade subsystem), and past passive participle with an e-grade subsystem (Ringe, 2006; Mailhammer, 2007).

Gothic alone transparently retained reduplication in 21 class 7 verbs (Bennett, 1967), including a subclass with ablaut: for example -letan, laílot ‘let’, and -tekan, taítok ‘touch’. The rest of Germanic has only relics, like Old English (Anglian dialect) heht (Goth. haíháit) versus WS hēt ‘named, called’, and leolc (Goth. laíláik) versus WS lēk ‘played’ (Jasanoff, 2007).

A sample paradigm can be found under the Thematic Verb section.

The Weak Verb

The weak verbs had four main form classes: (-j-) causative, deadjectival; (-ō-) iterative, intensive, denominal; (-ái-) stative, durative; and (-na-/-no-) anticausative, unaccusative11 (See Table 13). In the fourth class, -no- is found in the preterite (e.g., 3sg-fullnoda). A potential fifth class (ignored here) is based on the preterite with a bare consonant stem: 1/3sg þāhta (þagkjan ‘reflect on’). Many forms are unattested and are supplied from other verbs.

Table 13: Weak verb paradigms

class 1

2

3

4

nasjan

sokjan

salbon

haban

fullnan*

‘save’

‘seek’

‘anoint’

‘have’

‘be filled’

Indicative:

sg

1

nasja*

sokja

salbo*

haba

fullna*

2

-nasjis

sokeis*

salbos*

habáis

fullnis*

3

-nasjiþ

sokeiþ

salboþ*

habáiþ/d

fullniþ*

du

1

nasjos

sokjos*

*salbos

habos

fullnos*

2

*nasjats

*sokjats

*salbots

*habats

*fullnats

pl

1

nasjam*

sokjam*

salbom*

habam

fullnam*

2

nasjiþ*

sokeiþ

salboþ*

habáiþ/d

fullniþ*

3

nasjand*

sokjand

salbond*

haband

fullnand*

Imperative:

sg

2

nasei

sokei

salbo

habai

fulln*

3

*nasjadáu

*sokjadáu

*salbodáu

*habadáu

*fullnadáu

du

2

*nasjats

*sokjats

*salbots

*habats

*fullnats

pl

1

nasjam*

sokjam*

*salbom

*habam

*fullnam

2

nasjiþ*

sokeiþ*

salboþ*

habaiþ*

fullniþ*

3

nasjandáu*

sokjandáu*

salbondáu*

habandáu*

fullnandáu*

Optative:

sg

1

-nasjáu

-sokjáu

salbo*

habáu

-fullnáu

2

-nasjáis

sokjáis*

salbos*

habáis*

fullnáis*

3

nasjái*

sokjái

salbo*

habái

-fulnái

du

1

*nasjáiwa

*sokjáiwa

*salbowa

*habáiwa

*fullnáiwa

2

*nasjáits

*sokjáits

*salbots

*habáits

*fullnáits

pl

1

nasjáima*

sokjáima*

salboma*

habáima

fullnáima*

2

nasjáiþ*

sokjáiþ*

salboþ*

habáiþ

ful(l)náiþ

3

nasjáina*

sokjáina*

salbona*

habáina

fullnáina*

Passive indicative:

sg

1

nasjada*

sokjada*

salboda*

habada*

2

*nasjaza

*sokjaza

*salboza

*habaza

3

-nasjada

-sokjada

salboda*

habada*

pl

1–3

nasjanda*

sokjanda*

salbonda*

habanda

Passive optative:

sg

1

nasjáidáu*

-sokjáidáu

salbodáu*

habáidáu*

2

nasjáizáu*

sokjáizáu*

salbozáu*

habáizáu*

3

nasjáidáu*

sokjáidáu*

salbodáu*

habáidáu*

pl

1–3

nasjáindáu*

sokjáindáu*

salbondáu*

habáindáu*

Preterite:

sg

1

nasida*

sokida

salboda*

habáida

fullnoda*

2

nasides*

sokides*

salbodes

habáides*

fullnodes*

3

-nasida

sokida

salboda

habáida

-fullnoda

pl

1

nasidedum*

sokidedum

salbodedum*

habáidedum

fullnodedum*

2

nasideduþ*

sokideduþ

salbodeduþ*

habáideduþ*

fullnodeduþ*

3

nasidedun*

sokidedun

-salbodedun

habáidedun

-ful(l)nodedun

Preterite optative:

sg

1

-nasidedjáu

sokidedjáu*

salbodedjáu*

habáidedjáu*

fullnodedjau*

2

nasidedeis*

sokidedeis*

salbodedeis*

habáidedeis*

fullnodedeis*

3

-nasidedi

sokidedi*

salbodedi*

habáidedi*

-fullnodedi

pl

1

nasidedeima*

sokidedeima*

salbodedeima*

habáidedeima

fullnodedeima*

2

nasidedeiþ*

sokidedeiþ*

salbodedeiþ*

habáidedeiþ

fullnodedeiþ*

3

nasidedeina*

sokidedeina*

-salbodedeina

habáidedeina

-fullnodedeina

The weak verb is unique to Germanic. Its most distinctive feature is the dental preterite (type Eng. -ed), for which there are many different accounts (e.g., Kiparsky, 2009; Hill, 2010; Kim, 2010; Stiles, 2010; Ringe, 2012). Most likely, a form like Gothic waúrhtedun ‘they wrought’ goes back to *wurhta dedun [worked made.3pl] ‘they made (something) wrought’; cf. runic dedun Northumbrian OE dedun. and examples in other early Germanic languages. For the linearization, see the Linearization section.

The Thematic Verb

The thematic verb was a form class with a stem vowel characterized by an alternation between -a- and -e- (-i-). Table 14 contains a composite paradigm of niman ‘take’ (strong 4).12

Table 14: A Gothic thematic verb

Nonpast:

ind act

ind pass

opt act

opt pass

imp

sg

1

-nima

nimada*

nimáu

nimáidáu*

2

nimis

nimaza*

-nimáis

nimáizáu*

nim

3

nimiþ

-nimada

nimái

nimáidáu*

nimadáu*

du

1

nimos*

nimáiwa*

2

nimats*

nimats*

nimats*

pl

1

nimam*

nimanda*

-nimáima

nimáindáu*

nimam*

2

nimiþ

nimanda*

nimáiþ

nimáindáu*

nimiþ*

3

nimand

nimanda*

nimáina

nimáindáu*

nimandáu*

Preterite:

ind act

opt act

sg

1

nam

-nemjáu

2

namt

nemeis

3

nam

nemi

du

1

*nemu

*nemeiwa

2

nemuts*

*nemeits

pl

1

nemum

nemeima*

2

nemuþ

nemeiþ*

3

nemun

nemeina

The Verb ‘Be’

The athematic verb was characterized by a different set of endings attached to the root. Of these, at least from the Indo-European point of view, wisan ‘to be’ is isolated in Germanic. For the paradigm in Table 15, cf. Snædal, 2009a: 161.

Table 15: The Gothic copula

nonpst

opt

pret

pret opt

sg

1

im

si(j)áu

was

we(i)sjáu

2

is

si(j)áis

wast

we(i)seis

3

ist

si(j)ái

was

wesi

du

1

siju

sijáiwa*

wesu*

weseiwa*

2

sijuts*

sijáits*

wesuts*

weseits*

pl

1

si(j)um

sijáima

we/isum

weseima

2

si(j)uþ, siud

sijáiþ/d

wesuþ

weseiþ

3

sind

sijáina

we(i)sun

weseina

The forms without -j- occur in Luke and the Epistles and are rare even there, but argue in favor of j being a glide (Barrack, 1997: 5; Pierce, 2007: 241).

For the missing imperative, the optative forms sijáis, sijái, and sijáiþ are substituted. This verb has only one participle, wisands ‘being’, with 16 case forms attested.

The Dual

The dual is an archaic category in Gothic (on the history, see Fritz, 2011), preferentially used with a dual pronoun (Mossé, 1956: 178) but otherwise variably: 16 plurals in dual contexts versus 48 duals (Seppänen, 1985). Consider the duals (underscored) in (9) and (10), which are native Gothic because New Testament Greek lost the dual.

(9)

(10)

Periphrastic Passive Formations

Apart from relics like OE hātte ‘am/is called’, Gothic alone preserves the inherited mediopassive. Beside the synthetic nonpast passive of the type gibada ‘is (being) given’, Gothic created a periphrastic past passive with the preterite participle plus ‘be’: gibans was* ‘was given’. The participle agrees with its subject in gender, number, and case (but see Pagliarulo, 2011a). To insist on inchoativity or change of state, waírþan ‘become’ was used: -gibans warþ ‘came to be given, got given’. By analogy, these formations were extended to the nonpast: gibans* (giban) ist ‘is given’ and gibans waírþiþ* ‘gets given’.14

The contrast in the past is illustrated in (11) (Schröder, 1957: 10; Abraham, 1992b: 3; Ferraresi, 2005: 122).

(11)

Fralusans was is stative (‘was in a lost state’) and bigitans warþ is inchoative (‘got to be in a found state’). Compare German ‘mein Sohn . . . war verloren und wurde gefunden’. However, the ‘be’ passive is stative in German but is general in Gothic (Schröder, 1957: 14; Abraham, 1992b; Ferraresi, 2005: 121–124; Pagliarulo, 2008). See (12).

(12)

Daupidai (= dáupiþái) is nom pl masc in agreement with allái. That German requires a werden passive here illustrates the difference. The Gothic waírþan passive is restricted to inchoative or change-of-state contexts, while the ‘be’ passive is more general.

The Preterite-Presents

This class has preterite forms but a present-tense meaning (Table 16), due to the origin of some as Indo-European perfects (traditional account). Wáit (= Gk. oĩda ‘I know’, oĩde ‘knows’) was a perfecto-present in Indo-European (but see Jasanoff, 2003: 228–233). A few important preterite-presents are Gothic wáit ‘know’, þarf ‘need’, ga-dars ‘dare’, kann ‘be acquainted, know (how)’, skal ‘owe’, mag ‘can’, and áih ‘possess’ (Comparative Germanic Grammar 187–193; Mossé, 1956: 141–143; Gotische Grammatik 167–170).

Table 16: Some Gothic preterite-presents

witan

þaúrban*

kunnan

magan*

skulan*

[-aihan]

sg

1

wáit

þarf

kann

mag

skal

áih

2

wáist

þarft

kan(n)t

magt

skalt

áiht*

3

wáit

þarf*

kann

mag

skal

áih

pl

1

witum

þaúrbum

kunnum

magum

skulum

áigum

2

wituþ

þaúrbuþ

kunnuþ

maguþ/d

skuluþ

áihuþ

3

witun

þaúrbun

kunnun

magun

skulun

áigun

preterite

sg

wissa

þaúrfta

kunþa

mahta

skulda

áihta

pl

wissed-

kunþedum

mahted-

skulded-

áihtedun

optative

nonpst

witjáu

þaúrbeiþ

kunnjáu25

magjáu

skuljáu

áigi

pret

wissedjáu

kunþedeiþ

mahtedi

skuldedi

áihtedeis

participles

nonpst

witands

þaúrbands*

kunnands

magands

áigands

pret

þaúrfts

kunþs

mahts

skulds

Witan has a 2du wituts (spelled wituþs), and magan has magu and maguts. In the preterite and optative, an attested form is cited (usually third person) instead of one singular and one plural. For witan and skulan, the optative has the preterite plural stem. Forms of áih exhibit some leveling (áigum/áihum, áigands/nom pl masc áihandans); one factor is the lack of an apophonic alternation, in contrast to, for example, þarf/þaúrbum (Sturtevant, 1931).

The preterite participles sometimes have special meanings (e.g., kunþs ‘known’, þaúrfts ‘necessary; useful’, skulds ‘obliged, obligated; guilty’).

The Germanic modal verbs had participles but no imperatives (Cuendet, 1924) and no infinitives (Coupé & van Kemenade, 2009). The functions of these verbs are discussed by Ferraresi (1998) and Rousseau (2003).

The Verb ‘Will’

Wiljan ‘will, be willing, wish, want’ is by origin an optative identical to, for example, Lat. velim ‘I’d wish’, velī‎s ‘you’d like’ and therefore has only optative forms in the nonpast tense system The preterite is a standard weak type and has a separate optative (Table 17).

Table 17: The verb ‘will’

nonpst

pret

pret opt

sg

1

wiljáu

wilda

wildedjáu*

2

wileis/z

wildes*

wildedeis*

3

wili

wilda

wildedi

pl

1

wileima

wildedum

wildedeima*

2

wileiþ/d

wildeduþ

wildedeiþ

3

wileina

wildedun

wildedeina*

The only participle is wiljands ‘willing’. A 2du wileits is attested at Mark 10:36.

Syntax

One major problem with Gothic syntax is distinguishing native constructions from Greek calques. Scholarship is divided on how much Gothic is genuine (see Cuendet, 1929; Greiner, 1992; Klein, 1992a, 1992b; Berard, 1993; Burton, 1996b; Ferraresi, 2005; Harbert, 2007; Goetting, 2007; Walkden, 2012; Kotin, 2012: 317–374; Ratkus, 2011, 2016).

Subject Pronouns

Gothic was the most null subject of any Germanic language (Friedrichs, 1891; Schulze, 1924; Abraham, 1992a; Harbert, 2007: 221–222). In many instances there is no difference from the Greek text, which provides no information (Meillet, 1908–1909: 87ff.).

In Matthew, there is no statistical difference in the use of null or overt pronominal subjects to express a referential pronoun (Walkden, 2012: 174). The highest number of referential null subjects is in the 3sg, with 62 examples (93.9%). Of the 229 instances of a null referential subject in Matthew, there is no instance of an overt pronoun in the Greek text (Walkden, 2012: 175). When there is a difference between the Greek and the Gothic text, it is generally because Gothic has inserted an overt subject pronoun with change of subject (Lenk, 1910: 243; Ferraresi, 2005: 42, 48–49; Walkden, 2012: 174–175).

As shown in (13), Gothic, like Greek and Latin, did not use overt expletive subjects (Grimm, 1819–1837: 252; Fertig, 2000: 5; Ferraresi, 2005: 59; Walkden, 2012: 174).

(13)

Note that is (not seins*) is used for ‘his’ because swe introduces an adjunct clause ‘as his teacher (is)’ (Peeters, 1978), like the Greek and Latin versions (cf. Jülicher, 1972: 60; cited under Jülicher, 1963–1976).

Reflexive and Pseudo-Reflexive Structures

Gothic has a system of anaphors consisting of reflexive sik ‘himself, herself, itself’ (gen seina Lk 7:32, dat sis) and the possessive adjective sein- ‘his, her, its (own)’. Sein- can be reflexive or a discourse anaphor; that is, context dependent rather than syntactically bound (Kiparsky, 2011).15 Only the third person has a distinct anaphor. Mik ‘me’, mein- ‘my’, þuk ‘you’, þein- ‘your’, and the plural forms are both anaphoric and pronominal. This is true of the oblique case forms such as misand þus, in contrast to anaphoric sis, for example.

Reflexive sik often occurs with a form of the intensive predicate of identity silb- ‘self’:

(14)

Silban most often translates Greek heautón ‘himself”’ (reflexive he ‘himself’ plus autón ‘self’), especially in contexts where the action is directed by the subject toward itself (Rose, 1976: 47; Harbert, 2007: 210).16

Unlike English, the anaphor can occur in a prepositional phrase contained in a clause with a more local potential binder, such as the object of the matrix verb:

(15)

The object of the matrix verb is the more local potential binder, ignored by the anaphor whose binder is the null subject of the participle. Such constructions are genuine Gothic because the Greek text has a pronoun, not an anaphor (Ferraresi, 2005: 93; Harbert, 2007: 198), and the Vetus Latina manuscripts, with one exception (featuring a pronoun, not an anaphor), have a totally different construction (cf. Jülicher, 1970: 26; cited under Jülicher, 1963–1976).

That binding is anaphoric and not logophoric is suggested by that- clauses in which a pronoun rather than an anaphor refers to the matrix subject (Harbert, 2007: 197). See (16).

(16)

Gothic was in the process of replacing the Indo-European mediopassive with a nonargument reflexive formation (Ferraresi, 2005: 109–123), illustrated in (17).

(17)

Galesun sik corresponds more closely to Vetus Latina congregāta est ‘gathered’ (cf. Jülicher, 1970: 27; cited under Jülicher, 1963–1976) than to Gk. sunágetai ‘gathers’ (Alexandrian) or Byzantine sunḗkhthē ‘was gathered’. Generally speaking, for 28 verbs, a simple reflexive translates a Greek mediopassive (Ferraresi, 2005: 89–90).

In (18), the simple reflexive translates the Greek intransitive hupéstrepsen ‘returned’; compare the Vulgate and Vetus Latina deponent reversa est ‘id.’ (cf. Rose, 1976: 48).

(18)

See also García García, 2005: passim; p. 152 for this verb.

The 10 Greek infinitives (counting tense and voice) versus one in Gothic created translation difficulties (Greiner, 1992). Some Gothic infinitives involve raising (Joseph, 1981), but some are just functionally passive (Gotisches Elementarbuch 209–210; Klein 1992a: 342, 360); cf. (19).

(19)

Dáupjan translates Greek baptisthẽnai ‘to be baptized’. Passivity is indicated by fram + dat, the most frequent Gothic expression of the agent ‘by’-phrase (Corazza, 1982: 92; Klein, 1992b: 7, 12, 73). This residual construction is normally replaced by a reflexive (Gotisches Elementarbuch 210), as in þaiei qemun . . . háiljan sik (Lk 6:18): ‘(they) who came to be healed’. The preterite system had a periphrastic passive infinitive formed with wisan or waírþan.

Aspect and Aktionsart

Not all Gothic verbal prefixes were relevant to lexical or grammatical aspect (pace Streitberg, 1891; see the refutation in Mirowicz, 1935).17 Bucsko (2011: 61f.) classifies 63 as idiomatic (71 with some idiomaticity), for example Gothic us-qiman [come out] ‘kill’; 92 as metaphorical, for example dis-tahjan [tear apart] ‘destroy, waste’ (ibid. 23f.); and many as polysemous, for example us-baíran (1) compositional ‘carry out’, (2) metaphorical ‘produce’, and (3) idiomatic ‘answer’ (ibid. 52, 124). There were also stylistic choices (Götti, 1974).

There is some competition between us- and ut ‘out’, but in prefixed forms they are not identical; cf. utgaggan (two times) ‘go out, exit’ : usgaggan (102 times) ‘come/go (out, up)’, and only the former is resultative and contrasted with inngaggan ‘enter’ (Goetting, 2007: 325–326).

The main Gothic perfectivizing and telic prefix was ga-, which has many functions, including idiomatic and metaphorical (Bucsko, 2011: 176–179, 254–267). Two important functions are (20a) collective/sociative and (20b) telic (bounded in space or time).18

(20)

With (20a), cf. Gothic ga-gaggan ‘gather, assemble’ and ga-ga-haftjan ‘join together’.

Miþ ‘with’ is productive as sociative, as in miþ-qiman ‘come along with’, miþ-rodjan ‘speak with’, and miþ-ga-dáuþnan ‘perish (ga-dáuþnan) along with’.

Other functions of ga- were temporal completion in the past and definiteness (Krause, 1987: 209; Leiss, 2000, 2007; Kotin, 2012: 214–220).

The ga- forms in Gothic often translate a Greek verb that is noniterative/durative, nonprogressive, ingressive or egressive, instantaneous, or punctual. Contrasts include háusjan ‘to hear’ : gaháusjan ‘take heed’; táujan ‘do’ : gatáujan ‘complete, accomplish, produce’; waúrkjan ‘work, effect’ : (telic) gawaúrkjan ‘bring about, achieve, produce’ (Lloyd, 1979: 294–302); and meljan ‘write’ : (telic) gameljan ‘write down’:

(21)

Prefixes other than ga- could indicate telicity. In jah in fon atlagjada (Mt 7:19) ‘and is laid upon the fire’, at ‘to, towards’ makes the verb telic.

Complementizers and Relativizers

The main Gothic complementizers were ei and þatei ‘that’. Þatei derives from the pronoun þata ‘that’ plus relativizer/complementizer ei.19 For ei as a relativizing particle, cf. faúr-þiz-ei (before this that) ‘before’ (always + optative: Diekhoff, 1912) and ik-ei ‘I who’, for instance. Especially frequent is sa-ei ‘he that/who’, acc sg masc þan-ei ‘him that/whom’, for example.

(22)

(23)

The complementizer þatei has a dative counterpart þammei ‘(in/for) that’ from dative pronoun þamma + ei.20 In a relative clause þammei gets lexical case, for example stáins þammei uswaúrpun þái timrjans (Mk 12:10) ‘the stone which the builders threw out’.

Mood

Complement clauses can be finite or nonfinite. Infinitival complements are frequent under identity of the matrix and infinitival subject. Switch reference is normally effected by a finite ‘that’ clause. Purposives are often expressed by du ‘to’ plus infinitive (Gotisches Elementarbuch 213), as in urrann saiands du saian (Lk 8:5) ‘the sower ran out to sow’.

Whether the optative occurs in a finite subordinate clause or not depends to some extent on the semantics of the matrix verb. Factive, epistemic (including pseudo-perception verbs), and declarative predicates (combined as reflective verbs in Miller, 2002, w. lit.), normally take a complement clause with þatei or ei plus the indicative. Verbs that do not allow a full range of independent tenses in the complement clause, or whose complements are not realized, or are only potentially realizable, trigger a shift to the optative. Of course, Greek influence and other factors play a role (cf. Lindberg, 2010).

The Optative

The term “optative” is used here because formally it corresponds to the Indo-European (IE) optative. Since it conflates the functions of the IE subjunctive and optative, either term can be used and many scholars prefer “subjunctive.”

The Dependent Optative

The dependent optative accompanies matrix verbs that do not allow a full range of independent tenses in the complement clause, or whose complements are not realized, or are only potentially realized.

Volitional complements are unrealized (inherently irrealis), and the time frame of the complement is at least partly dependent on the matrix time frame; cf. (24).

(24)

Verbs of the asking class admit several complement clause types (e.g., ‘whether’, ‘that’). The latter do not license a full range of tenses in the subordinate clause independent of those in the matrix clause (cf. Mourek, 1893: 168). Note, however, the tense harmony in both of the examples in (25), where matrix and subordinate verbs are preterite but the latter have no temporal value (cf. Gotisches Elementarbuch 240, 244).

(25)

In (25b), the embedded clause patterns with Latin versions (which observe sequence of tenses), for example Vulgate sī‎ quid vidēret ‘if / whether he saw anything’. The Greek has epērṓtā autòn eí ti blépei ‘he was asking him if he sees anything’ (v.l. blépeis ‘you see’, not in the Byzantine main text). Note Gothic interrogative u (Klein & Condon, 1993: 34–46; Pagliarulo, 2011b) and ƕa ‘what / anything’ supposedly preserving the Greek order eí ti (Danielsen, 1968: 123, w. lit.), intercalated in the prefixed verb ga-saíƕan ‘catch sight of’, but the construct is native Gothic (cf. Ivanov, 1999).21

Any epistemic verb can have its complement clause verb in the optative when unrealized or potential. In (26), the Greek text has ind ménei ‘remains’, and the Vulgate and Vetus Latina manuscripts have manet ‘id.’ (Jülicher, 1963: 142; cited under Jülicher, 1963–1976).

(26)

A negated matrix verb sets up a question as to the actuality of the realization; hence, the optative (cf. Mossé 1956: 195). Nevertheless, even with a negated matrix verb a direct observation can take the indicative (cf. Guxman, 1958: 236), although (27) is an exact translation (Falluomini, 2013b: 158); cf. Vetus Latina loquitur/dī‎cit ‘says’ beside rare subjunctive loquā‎tur/dī‎cat (Jülicher, 1963: 176; cited under Jülicher, 1963–1976).

(27)

The Independent Optative

Functions include deontic or agent-oriented modalities (e.g., obligation); functional illocution markers of reinforcing mode, such as admonitive or prohibitive (ni + opt: Gotisches Elementarbuch 205); and proposition markers of subjective mode, such as doubt/uncertainty, possibility/potentiality, and boulomaic modalities—that is, wishing (desiderative/volitive) and hoping (cf. Mourek, 1893: 155–284). Less fulfillable or unfulfillable wishes take the preterite optative (Gotisches Elementarbuch 204). Conditionals mostly take the indicative; the nonpast optative expresses possibility or an implied command, and the preterite optative expresses impossibility or counterfactuality (Weisker, 1880).

The optatives in (28) translate Greek aorist 3sg imperatives.22

(28)

To the category of reinforcing mode belong (rare) imperative substitutes; cf. (29).

(29)

The optative replaces the imperative for attenuation (Gotisches Elementarbuch 204f.), when no immediate action is required, or when a prescription for the future or all time is involved (Gotisches Elementarbuch 205; Cuendet, 1924: 59; Klein, 1992a: 364–367), as in (29).

One of the most frequent functions of the optative is to express potentiality, especially in questions (Gotisches Elementarbuch 207), or eventuality (cf. Mossé, 1956: 194). (30) contains two optatives, one boulomaic or reinforcing mode, the second potential:

(30)

Linearization

Studies of linearization include Cuendet, 1929; Fourquet, 1938: chapter 6; Longobardi, 1978; Ferraresi, 1991, 2005; Harbert, 2007; Kotin, 2012: chapter 4; and Ratkus, 2011, 2016.

Pronouns

Subject pronouns preferentially precede the verb directly. Ik (less often jūs) allows an intercalated phrase (Ferraresi, 2005: 38). When a subject and an object pronoun are present, the order is SO (ibid. 50). Reflexives precede nonreflexives (ibid. 52). For complete data, see Friedrichs, 1891.

With double object verbs, the order is determined lexically or semantically. For instance, atáugjan ‘show’ and nine other verbs take dat-acc order; (at-, fra-, us-)giban ‘give’ and six other verbs require acc-dat; bidjan and fraíhnan ‘ask’ and (ga-)fulljan ‘fill’ take acc-gen; and láisjan ‘teach' takes accperson accthing(Ferraresi, 2005: 64ff.).

Determiners

Where Gothic determiners occur, they precede the noun directly unless accompanied by an adjective, in which case the determiner-adjective complex can precede or follow the noun. In general, determiners in Gothic are rare except in the context of the weak adjective, where they may have some demonstrative feature.23 In Matthew there are “some 475 Greek articles which have no equivalents in the Gothic text . . . [versus] some 165 which have been duly rendered” (Metlen, 1933: 534). Other text samples yield similar results (Berard, 1993: 119). Overall Kovari (1984: 35) counts 1,417 Greek articles but only 371 Gothic examples of det + N versus 992 plain nouns. Gothic determiners correspond to Greek articles in emphatic, demonstrative, and relative contexts (Vilutis, 1976). One frequent use refers to old information (Behaghel, 1923–1928: 39; Kotin, 2012: 213–216), for example ufar þans fimf hláibans jah twans fiskans (Sk 7.2.15ff.) ‘over those five loaves and two fish’, referring to .e. hláibans barizeinans jah twans fiskans ‘(the) five barley loaves and two fish’ mentioned in 7.1.10ff. (Lenk, 1910: 244).

Prepositional Phrases

Gothic has no (even Greek-prompted) postpositions (Cuendet, 1929: 153). Prepositional phrases (PPs) rarely admit determiners. Articleless PPs still occur in present-day Germanic languages (Vilutis, 1976: 155). Kovari (1984: 37) counts 573 examples of P + N versus only 141 with det, in contrast to Greek, where articles were normal in PPs. Gothic PPs have a det with previously mentioned material, strong demonstratives (e.g., bi þamma razna jáinamma [Mt 7:25] [against that house yon] ‘against that house’), and a few other contexts (Vilutis, 1976; Kotin, 2012: 475f.).

Attributive and Predicative Adjectives

Attributive adjectives occur after the noun 333 times and precede only 275 (Ratkus, 2011: 145). With quantifiers the figures are reversed: 55 QN beside 31 NQ (ibid.). In Skeireins, likewise, there are seven QN but only three NQ (ibid. 154). Adjectives, however, prefer prenominal position in Skeireins (40 AN : 9 NA), which may have been the default position for attributive adjectives (Ratkus, 2011: 165ff.).

In predicative constructions, adjectives precede the verb 285 times and follow 184 (Ratkus, 2011: 145). This is also true of Skeireins, with 15 AV versus 2 VA (ibid. 151).

Possessive adjectives preferentially follow the noun (cf. Cuendet, 1929: 42f.). Ratkus (2016) counts 1,564 matches with the Greek linearization versus only 28 nonprobative deviations. Where the Gothic text is nothing like any Greek or Latin version, there are 14 examples of postnominal possessives versus one Poss-N (ni sokja izwaros áihtins, ak izwis [2 Cor 12:14A/B]: ‘I do not seek your property, but you’). Skeireins has 18 N-Poss versus eight Poss-N, and the deeds have 12 N-Poss and zero Poss-N—nothing unequivocal. But even in calques, native Gothic order can emerge, for example so armahaírtiþa þeina (Mt 6:4) ‘your charitable deed’, rendering sou hē [~ hē sou] eleēmosúnē, literally ‘your the [~ the your] charitable deed’. N-Poss prevails in the new Gothic fragments, for example naseins meina jah wulþus meins (Bn.1v.18) ‘my salvation and my glory’, and ana managein þeinái | þiuþeins þeina (Bn.1v.19f.) ‘among your people [is] your blessing’. Possible orders are N-Poss-Det, Det-Poss-N, and Det-N-Poss. *Poss-Det-N occurs only when a weak adjective follows.

Genitives

The default position for genitives is postnominal, but exceptions do not require Greek prompting (Koppitz, 1900: 435–438; Lenk, 1910: 271; Cuendet, 1929: 12). Genitives are consistently right branching in the new fragments, for example in alláim waíladede is (Bn.1r.23) ‘in all of his good deeds’, and nasjands alláize manne þishun þize ga[l]áubjandane (Bn.1r.24ff.) ‘savior of all people, especially of the believers’ (= 1Tim 4:10B minus þize). Predicative genitives are different, for example f(rauj)ins ist naseins (Bn.1v.19) ‘the lord’s is salvation’.

Numerals and Quantifiers

In the Greek gospels a numeral precedes a noun 212 times and follows 85 times. Gothic normally follows suit (Cuendet, 1929: 142ff.). Some (especially preposed) numerals behave as quantifiers; cf. fidwor unkjane ‘four (of) uncias’ in the landsale deed from Arezzo beside skilliggans .j. ‘sixty shillings’ in the debt-settlement deed from Naples (Signature 2). For the opposite order, note .e. hláibans ‘five loaves’ (Sk 7.1.10), dage fidwor tiguns [of days four tens] ‘forty days’ (Lk 4:2).

All- ‘all’ can modify a noun directly (Koppitz, 1900: 455f.), but the partitive structure is frequent, for example all bagme ‘all of trees’ (gen pl), different from Gk. pãn déndron (Mt 7:19) ‘every tree’ (nom sg) or omnis (. . .) arbor ‘id.’ in the Latin versions (cf. Jülicher,1972: 38; cited under Jülicher, 1963–1976). All dagis (Rom 8:36A) [all of day] ‘all day long’ renders Greek hólēn tḕn hēmérān ‘the entire day’ (cf. Vulgate tōtā diē ‘id.’).

Verbs and Auxiliaries

Auxiliaries follow verbs in native Gothic (Fourquet, 1938: 252ff.). Pagliarulo (2006) finds that translation of a Greek periphrastic structure prompts the same linearization, for example estin gegramménon (in Jn 6:31) ist gameliþ ‘is written’ and ẽn gnōstós (Jn 18:15) was kunþs ‘was known’. By sharp contrast, with only two exceptions out of 62 instances, Gothic has final verbs or auxiliaries to render a Greek synthetic construct, as in qiþan ist (Mt 5:27) ‘is said’ (Gk. erréthē) and gasuliþ was (Mt 7:25) ‘was founded’ (Gk. tethemelíōto). This extends to hráin warþ (Mt 8:3) ‘got clean’ (Gk. ekatharísthē), and even to blinds gabaúrans warþ (Jn 9:20) ‘was born blind’ (Gk. tuphlòs egennḗthē) (ibid.; cf. Eythórsson, 1995: 20, 1996: 109).

In most biblical Gothic documents the verb precedes its direct object. Object-before-verb (OV) order is more rare, especially in main clauses. Relative clauses that are not translated from Greek tend to be verb final (Eythórsson, 1996: 109).

V1 and V2

Certain functional elements in the left periphery force verb movement (Eythórsson, 1995, 1996; Ferraresi, 1991, 2005; Fuß, 2003; Harbert, 2007: 410–415; Buzzoni, 2009).

Particles like emotive interrogative -u force initial position, for example of a verb: maguts-u driggkan (Mk 10:38) ‘can you two drink?’ (Ferraresi, 2005: 148ff.; Pagliarulo, 2011b); compare also -uh ‘and’: iddjedun-uh ufar marein (Jn 6:17) ‘and they went over the sea’ (Fuß, 2003: 201, w. lit.).

Imperatives and their optative substitutes are normally sentence-initial, as in (31).

(31)

Only a pronoun, clitic, or adverb can separate neg ni from the verb (Koppitz, 1900: 15–23; Mourek, 1903: 15; Coombs, 1976: 44ff.); cf. ni naúhþanuh galagiþs was in karkarái ioha(n)nes (Sk 3.1.4–7) [not yet cast was in prison John] ‘John had not yet been cast into prison’. The most complete study of negation in Gothic is Danielsen, 1968.

With wh-questions, violations of V2 are Greek prompted (Fuß, 2003: 200–205).

The Position of Gothic within Germanic

A single but stratally differentiated language called Proto-Germanic made a number of innovations after splitting off from ‘central’ Indo-European. Subsequently, Proto-Germanic speakers migrated in different directions, and the separation facilitated evolution of different dialects, which became Gothic, Old Norse, and the other languages.24

Innovations common to all the Germanic languages include Grimm’s and Verner’s laws (the latter preserved only in traces in Gothic), collapse of the subjunctive and optative into a single mood, innovation of a class of preterite-presents, the strong and weak verb, and the strong and weak adjective. The system of strong versus weak adjectives is best preserved in German, alive in Scandinavian, lost predicatively in Dutch, and gone in English.

Some 70 derivational suffixes have specialized uses in Germanic. The nominal, pronominal, and verbal systems are remarkably consistent throughout the Germanic world. Consider the pronoun ‘I/me’: Gothic ik/mik, OE ic/mec, Old Norse ek/mik, and Old High German ih/mih.

The differences between Gothic and the rest of Germanic are equally significant. These are due in part to the early attestation of Gothic and in part to the fact that East Germanic apparently split off the earliest from Proto-Germanic.

In the realm of phonology, Gothic has no evidence of ever having had umlaut, which is so characteristic of North and West Germanic; Gothic alone had Thurneysen’s law of spirant dissimilation, and preserved -u- in final syllables. Verner’s law in the verbal system is restricted to a few forms, and Gothic probably lost all diphthongs.

Since different demonstratives grammaticalized into definite articles in different areas, Proto-Germanic had no articles. Gothic determiners had only a few of the features associated with articles and, for this reason, rarely translate Greek articles.

Proto-Indo-European (PIE) had a system of past/nonpast tense preserved in Anatolian (e.g., Hittite) and Gothic. In the rest of Germanic, a future was created with an infinitive plus an auxiliary (e.g., Eng. I will go). The most frequent future auxiliary in the other Germanic languages corresponds to ‘shall’, the oldest form of which is Gothic skulan ‘owe, be obligated’.

The PIE mediopassive voice became passive in Germanic, vestigial everywhere but in Gothic (e.g., gibada ‘is given’). A new middle was attempted with reflexive *sik (e.g., Goth. gawandida sik ‘returned’). This construction became most productive in North Germanic, for example Old Norse felr sik ‘hides himself’ and fel-a-sk ‘to hide (oneself)’. Note also þyngjask ‘grow heavy’ versus þyngja ‘make heavy’. By sharp contrast, English and more generally North Sea Germanic (Old Saxon in part) lost *sik in the prehistoric period.

In the past tense a new periphrastic passive was created with ‘be’ or ‘become’, as in Gothic gibans was ‘was / had been given’ and gibans warþ ‘came to be given’. This formation was then generalized to the nonpast system; cf. Gothic gibans ist ‘is/was given’, gibans waírþiþ ‘gets given’. Though distributed differently in the later Germanic languages (e.g., modern German), in early Germanic the ‘become’ passive was typical of changes of state, like the 18 verða passives in the Edda, restricted to being born and dying, and the 14 weorðan passives in Beowulf tied to fate and similar contexts (see Schröder, 1957–1958: 102ff.).

In the aspectual system, the inherited Indo-European perfect became the strong preterite in Germanic. North and West Germanic created a new perfect system most often with the auxiliary ‘have’ (type I have done it), as in Old English (32a). By contrast, Gothic had only the antecedent predicative adjunct resultative construction in (32b).

(32)

In terms of linearization typology, Germanic was in the process of a shift from Verb-final to non-V-final and regularizing the position of modifiers, specifiers, heads, and complements. Prepositions and sentence-initial complementizers were already fully stabilized and exceptionless in Gothic. These two features overwhelmingly correlate with non-Verb-final linearization (Miller, 2010: ii. 21, 48, w. lit.).

Another Germanic innovation was V2 order, triggered by certain functional forms in the left periphery. This was violable in Gothic and never became fully regular in Old or Middle English, as it did in North and Continental West Germanic.

All of the early Germanic languages, including Gothic, were standardly V1 with imperatives, direct questions, and several other structures (Eythórsson, 1996; Harbert, 2007: 410–415; Miller, 2010: ii. 50).

On the conservative side, the default postpositioning of auxiliaries in Gothic was shared with early runic inscriptions, for example haitinaz was ‘was commanded / called’ (Kalleby stone [?ca. 400]). This is a residue of PIE V-final order (Miller, 2010: ii. 47ff.).

To some extent, early Germanic might qualify as a free word order (FWO) language, a slight misnomer because completely free orders do not exist (Miller, 2010: ii. 13, w. lit.). FWO languages have proportionately the highest number of mixed adjective-noun / noun-adjective orders, but a slight preference for genitive-noun (Miller, 2010: ii. 26f., w. lit.). In Gothic, however, the default orders (by a slight preference) were adjective-noun but noun-genitive and noun-possessive, illustrating the problem of the mixed orders resulting from a series of microparametric changes. Ultimately, the Germanic target was to have all modifiers and complements predictable with respect to all heads, both lexical and functional (Miller, 2010: ii, chapters 1–2; cf. Kiparsky, 1996).

Acknowledgements

I am indebted to Artūras Ratkus for extensive comments and many references, as well as for extensive help with works from the Eastern Bloc; to Rob Howell and Sara Pons-Sanz for valuable discussion and suggestions; and to Wayne Harbert for input on the syntax section.

Further Reading

  • Falluomini, C. (2015). The Gothic version of the Gospels and Pauline epistles: Cultural background, transmission and character. Arbeiten zur neutestamentlichen Textforschung 46. Berlin: De Gruyter.
  • Ferraresi, G. (2005). Word order and phrase structure in Gothic. Orbis Supplementa 25. Leuven, Belgium: Peeters.
  • Harbert, W. (2007). The Germanic languages. Cambridge Language Surveys. Cambridge, U.K.: Cambridge University Press.
  • Howell, R. B., Roberge, P. T., & Salmons, J. C. (2016). The Cambridge history of the Germanic languages (Vols. 1–2). Cambridge, U.K.: Cambridge University Press.
  • Kaliff, A., & Munkhammar, L. (Eds.). (2013). Wulfila 311–2011: International symposium, Uppsala University June 15–18, 2011. Occasional Papers in Archaeology 57. Uppsala, Sweden: Edita Västra Aros.
  • Marchand, J. W. (1973). The sounds and phonemes of Wulfila’s Gothic. Janua Linguarum 25. The Hague: Mouton.
  • Nielsen, H. F., & Stubkjær, F. T. (Eds.). (2010). The Gothic language: A symposium. NOWELE 58–59. Odense: University Press of Southern Denmark.
  • Ratkus, A. (2011). The adjective inflection in Gothic and early Germanic: Structure and development. Ph.D. dissertation, University of Cambridge, Cambridge, U.K.
  • Ratkus, A. (ed). (2016). Studies in Gothic. Oxford: Oxford University Press.
  • Vennemann, T. (1971). The phonology of Gothic vowels. Language, 47, 90–132.

References

  • Abraham, W. (1992a). Null subjects: From Gothic, Old High German and Middle High German to Modern German; From pro-drop to semi-pro-drop. Groninger Arbeiten zur germanistischen Linguistik, 34, 81–107.
  • Abraham, W. (1992b). The emergence of the periphrastic passive in Gothic. Leuvense Bijdragen, 81, 1–15.
  • Abraham, W., & Leiss, E. (Eds.). (2012). Covert patterns of modality. Newcastle upon Tyne, U.K.: Cambridge Scholars.
  • Adams, J. N., Janse, M., & Swain, S. (Eds.). (2002). Bilingualism in ancient society: Language contact and the written text. Oxford: Oxford University Press.
  • Afros, E. (2006). Gothic relative and explicative clauses introduced by enclitic ei. Amsterdamer Beiträge zur älteren Germanistik, 61, 5–15.
  • Afros, E. (2010). Gothic relative clauses introduced by izei and sei revisited. Amsterdamer Beiträge zur älteren Germanistik, 66, 5–20.
  • Aland, K. (Ed.). (1972). Die alten Übersetzungen des Neuen Testaments, die Kirchenväterzitate und Lektionare: Der gegenwärtige Stand ihrer Erforschung und ihre Bedeutung für die griechische Textgeschichte. Arbeiten zur Neutestamentlichen Textforschung 5. Berlin: Walter de Gruyter.
  • Andrason, A. (2010). Resultative predicative adjunct constructions in the Gothic Bible. Interlingüística, 20, 1–12.
  • Askedal, J. O. (2009). Some general evolutionary and typological characteristics of the Germanic languages. In J. O. Askedal, I. Roberts, T. Matsushita, & H. Hasegawa (Eds.), Germanic languages and linguistic universals (pp. 7–56). Development of the Anglo-Saxon Language and Linguistic Universals 1. Amsterdam: John Benjamins.
  • Askedal, John Ole, & Nielsen, Hans Frede (eds). (2015). Early Germanic Languages in Contact. Amsterdam: John Benjamins.
  • Askedal, J. O., Roberts, I., Matsushita, T., & Hasegawa, H. (Eds.). (2009). Germanic languages and linguistic universals. Development of the Anglo-Saxon Language and Linguistic Universals 1. Amsterdam: John Benjamins.
  • Bahder, K. von. (1880). Die Verbalabstracta in den germanischen Sprachen: Ihrer Bildung nach dargestellt. Halle, Germany: M. Niemeyer.
  • Baldi, P., & Cuzzolin, P. (Eds.). (2009–2010). New perspectives on historical Latin syntax (Vols. 1–4). Trends in Linguistics. Berlin: Mouton de Gruyter.
  • Bammesberger, A. (1990). Die Morphologie des urgermanischen Nomens. Heidelberg, Germany: Winter.
  • Bammesberger, A. (1994). The development of runic script and its relationship to Germanic phonological history. In T. Swan, E. Mørck, & O. J. Westvik (Eds.), Language change and language structure: Older Germanic languages in a comparative context (pp. 1–25). Trends in Linguistics. Berlin: De Gruyter.
  • Bandle, O., Braunmüller, K., Jahr, E. H., Karker, A., Naumann, H.—P., & Teleman, U. (Eds.). (2002). The Nordic languages: An international handbook of the history of the North Germanic languages (Vol. 1). Berlin: Walter de Gruyter.
  • Barðdal, J., & Chelliah, S. L. (Eds.). (2009). The role of semantic, pragmatic, and discourse factors in the development of case. Studies in Language Companion Series 108. Amsterdam: John Benjamins.
  • Barnish, S. J. B. (1984). The genesis and completion of Cassiodorus’ Gothic history. Latomus, 43, 336–361.
  • Barnish, S. J. B., & Marazzi, F. (Eds.). (2007). The Ostrogoths from the migration period to the sixth century: An ethnographic perspective. Studies in Historical Archaeoethnology 7. Woodbridge, U.K.: Boydell.
  • Barrack, C. M. (1989). Keyser, Kiparsky, O’Neil and Postal versus Sievers. Lingua, 77, 223–296.
  • Barrack, C. M. (1997). The putative strengthening of glides in Gothic. In I. Rauch & G. F. Carr (Eds.), Insights in Germanic linguistics II: Classic and contemporary (pp. 1–8). Trends in Linguistics 94. Berlin: Mouton de Gruyter.
  • Barrack, C. M. (1998). Sievers’ law in Germanic. Berkeley Insights in Linguistics and Semiotics 22. New York: Peter Lang.
  • Beade, P. (1972). Sievers’ law in Gothic and other related matters. Lingua, 30, 449–459.
  • Beck, R. (1973). Length and monophthongization in Gothic. Indogermanische Forschungen, 78, 113–140.
  • Behaghel, O. (1923–1928). Deutsche Syntax: Eine geschichtliche Darstellung (Vols. 1–3). Heidelberg, Germany: Winter. Vol. 1, Die wortklassen und wortformen: A. Nomen; Pronomen (1923); Vol. 2, Die wortklassen und wortformen: B. Adverbium; C. Verbum (1924); Vol. 3, Die satzgebilde (1928).
  • Bennett, W. H. (1949). The monophthongization of Gothic ái, áu. Language, 25, 15–21.
  • Bennett, W. H. (1952). The earliest Germanic umlauts and the Gothic migrations. Language, 28, 339–342.
  • Bennett, W. H. (1959a). The phonemic status of Gothic w h q. Language, 35, 427–432.
  • Bennett, W. H. (1959b). The function of present participial constructions in the Skeireins. In F. Mossé, Mélanges de linguistique et de philologie: Fernand Mossé in memoriam (pp. 32–36). Paris: Didier.
  • Bennett, W. H. (1960). The Gothic commentary on the gospel of John: Skeireins aiwaggeljons þairh iohannen; A decipherment, edition, and translation. Modern Language Association of America Monograph 21. New York: Modern Language Association of America.
  • Bennett, W. H. (1967). Some phonological effects of pre-Gothic juncture. Language, 43, 661–665.
  • Bennett, W. H. (1970). The stress patterns of Gothic. Transactions and Proceedings of the Modern Language Association, 85, 463–472.
  • Bennett, W. H. (1980). An introduction to the Gothic language. Introductions to the Older Languages of Europe 2. New York: Modern Language Association of America.
  • Berard, S. A. (1993). Biblical Gothic and the configurationality parameter. American Journal of Germanic Linguistics and Literatures, 5, 111–162.
  • Berard, S. A. (1995). Infinitival subject sentences in Gothic. In I. Rauch & G. F. Carr (Eds.), Insights in Germanic linguistics II: Classic and contemporary (pp. 5–46). Trends in Linguistics 94. Berlin: Mouton de Gruyter.
  • Bernharðsson, H. (2001). Verners law in Gothic. Ph.D. dissertation, Cornell University, Ithaca, NY.
  • Bethge, R. (1898–1900). Gotisch. In F. Dieter (Ed.), Laut- und Formenlehre der altgermanischen Dialekte: Zum Gebrauch für Studierende dargestellt von R. Bethge, O. Bremer, F. Dieter, F. Hartmann und W. Schlüter (pp. 21–35, 193–214, 391–408, 568–602). Leipzig: O. R. Reisland. The work is scattered throughout: Vokalismus des Gotischen (pp. 21–35). Konsonantismus des Gotischen (pp. 193–214), Gotische Konjugation (pp. 391–408), and Gotische Deklination (pp. 568–602).
  • Boutkan, D. (1995). The Germanic ‘Auslautgesetze’. Amsterdam: Rodopi.
  • Braune, W., & Ebbinghaus, E. A. (1981). Gotische Grammatik (19th ed.). Tübingen, Germany: Niemeyer.
  • Braune, W., & Heidermanns, F. (2004). Gotische Grammatik: Mit Lesestücken und Wörterverzeichnes (20th ed.). Tübingen, Germany: Niemeyer.
  • Brown, T. (2007). The role of Arianism in Ostrogothic Italy: The evidence from Ravenna. In S. J. B. Barnish & F. Marazzi (Eds.), The Ostrogoths from the migration period to the sixth century: An ethnographic perspective (pp. 417–441). Studies in Historical Archaeoethnology 7. Woodbridge, U.K.: Boydell.
  • Bucsko, J. M. (2011). Preverbs and idiomatization in Gothic. New York: Peter Lang.
  • Budanova, V. P. (1999). Goty v èpoxu velikogo pereselenija narodov (2d ed.). St. Petersburg, Russia: Aletejja.
  • Burkitt, F. C. (1899). The Vulgate Gospels and the Codex Brixianus. Journal of Theological Studies, 1, 129–134.
  • Burton, P. (1996a). Fragmentum Vindobonense 563: Another Gothic-Latin bilingual? Journal of Theological Studies n.s., 47, 141–156.
  • Burton, P. (1996b). Using the Gothic Bible: Notes on Jared S. Klein ‘On the independence of Gothic syntax’. Journal of Indoeuropean Studies, 24, 81–98.
  • Burton, P. (2002). Assessing Latin-Gothic interaction. In J. N. Adams, M. Janse, & S. Swain (Eds.), Bilingualism in ancient society: Language contact and the written text (pp. 393–418). Oxford: Oxford University Press.
  • Buzzoni, M. (2009). Ibai mag blinds blindana tiuhan? Pragmatic functions and syntactic categories in the Gothic left sentence periphery. Filologia Germanica / Germanic Philology, 1, 29–62.
  • Byrd, A. M. (2010). Reconstructing Indo-European syllabification. Ph.D. dissertation, University of California at Los Angeles.
  • Carr, C. T. (1939). Nominal compounds in Germanic. London: Oxford University Press.
  • Casaretto, A. (2004). Nominale Wortbildung der gotischen Sprache: Die Derivation der Substantive. Heidelberg, Germany: Winter.
  • Casaretto, A. (2010). Evidence for language contact in Gothic. NOWELE: North-Western European Language Evolution, 58–59, 217–237.
  • Casaretto, A. (2014). Zur Sprachkontaktsituation im Gotischen: Die Verbalkomposita. Linguarum Varietas: An International Journal, 3, 43–54.
  • Cercignani, F. (1979a). The reduplicating syllable and internal open juncture in Gothic. Zeitschrift für vergleichende Sprachforschung, 93, 126–132.
  • Cercignani, F. (1979b). The development of the Gothic short/lax subsystem. Zeitschrift für vergleichende Sprachforschung, 93, 272–278.
  • Coleman, R. (1996). Exponents of futurity in Gothic. Transactions of the Philological Society, 94, 1–29.
  • Collitz, H. (1925). Gothic siponeis, a loan word from Greek. American Journal of Philology, 46, 213–221.
  • Coombs, V. M. (1976). A semantic syntax of grammatical negation in the older Germanic dialects. Göppingen, Germany: Kümmerle.
  • Corazza, V. D. (1982). Le prepositioni gotiche fram, us, af, þairh e la loro funzione agentive. Aevum, 56, 92–106.
  • Coupé, G., & van Kemenade, A. (2009). Grammaticalization of modals in Dutch: Uncontingent change. In P. Crisma & G. Longobardi (Eds.), Historical syntax and linguistic theory (250–270). Oxford: Oxford University Press.
  • Crisma, P., & Longobardi, G. (Eds.). (2009). Historical syntax and linguistic theory. Oxford: Oxford University Press.
  • Croke, B. (1987). Cassiodorus and the Getica of Jordanes. Classical Philology, 82, 117–134.
  • Cuendet, G. (1924). L’impératif dans le texte grec et dans les versions gotique, arménienne et vieux slave des évangiles. Paris: Paul Geuthner.
  • Cuendet, G. (1929). L’ordre des mots dans le texte grec et dans les versions gotique, arménienne et vieux slave des évangiles: Vol. 1. Les groupes nominaux. Paris: Champion.
  • d’Alquen, R. J. E. (1974). Gothic AI and AU: A possible solution. The Hague: Mouton.
  • d’Alquen, R. J. E. (1988). Germanic accent, grammatical change and the laws of unaccented syllables. New York: Peter Lang.
  • Daniels, P., & Bright, W. (Eds.). (1996). The world’s writing systems. Oxford: Oxford University Press.
  • Danielsen, N. (1968). Status und Polarität im Gotischen im Lichte des Kymrischen dargestellt. Odense, Denmark: Odense University Press.
  • Davis, E. P. (1929). The injunctive in Gothic. Modern Philology, 26, 427–432.
  • Davis, G. (2002). Codex argenteus: Lingua gotorum aut lingua gotica? Journal of Language and Linguistics, 1–3, 308–313.
  • Desportes, Y. (Ed.). (1994). Philologische Forschungen: Festschrift für Philippe Marcq. Heidelberg, Germany: Winter.
  • Dewey, T. K., & Syed, Y. (2009). Case variation in Gothic absolute constructions. In J. Barðdal & S. L. Chelliah (Eds.), The role of semantic, pragmatic, and discourse factors in the development of case (pp. 3–21). Studies in Language Companion Series 108. Amsterdam: John Benjamins.
  • Diekhoff, T. J. C. (1912). The so-called prospective or anticipatory subjunctive in Gothic. Journal of English and Germanic Philology, 11, 173–179.
  • Dieter, F. (Ed.). (1898–1900). Laut- und Formenlehre der altgermanischen Dialekte: Zum Gebrauch für Studierende dargestellt von R. Bethge, O. Bremer, F. Dieter, F. Hartmann und W. Schlüter. Leipzig: O. R. Reisland.
  • Dolcetti Corazza, V., & Gendre, R. (Eds.). (2008). Intorno alla Bibbia gotica. Alessandria, Italy: Edizioni dell’Orso.
  • Dvuxžilov, A. V. (1980). Semantiko-sintaksičeskaja priroda protivopostavlenija ‘slabaja/sil’naja forma’ prilagatel’nogo v gotskom jazyke. Ph.D. dissertation, University of Kiev.
  • Ebbinghaus, E. A. (1990). The question of Visigothic runic inscriptions re-examined. General Linguistics, 30, 207–214.
  • Ebbinghaus, E. A. (1991a). Ulfila(s) or Wulfila? Historische Sprachforschung, 104, 236–238.
  • Ebbinghaus, E. A. (1991b). The vowel of the Gothic reduplicating syllable. General Linguistics, 31, 177–179.
  • Ebbinghaus, E. A. (1992). Some remarks on the life of Bishop Wulfila. General Linguistics, 32 (2-3), 95–104. Reprinted in Ebbinghaus, 2003, 192–203.
  • Ebbinghaus, E. A. (1996). The Gothic alphabet. In P. Daniels & W. Bright (Eds.), The world’s writing systems (pp. 290–296). Oxford: Oxford University Press.
  • Ebbinghaus, E. A. (1997). The Gothic documents: Their provenance and age. NOWELE: North-Western European Language Evolution, 31–32, 101–103.
  • Ebbinghaus, E. A. (2003). Gotica: Kleine Schriften zur gotischen Philologie (P. Scardigli & W. Meid, Eds.). Innsbruck, Austria: Institut für Sprachen und Literaturen der Universität Innesbrück. Contains 37 articles.
  • Ehrman, B. D. (2000). The New Testament: A historical introduction to the early Christian writings (2d ed.). Oxford: Oxford University Press.
  • Ehrman, B. D., & Holmes, M. W. (Eds.). (2013). The text of the New Testament in contemporary research: Essays on the status quaestionis (2d ed.). Leiden, The Netherlands: Brill.
  • Eythórsson, T. (1995). Verbal syntax in the early Germanic languages. Ph.D. dissertation, Cornell University, Ithaca, NY.
  • Eythórsson, T. (1996). Functional categories, cliticization, and verb movement in the early Germanic languages. In H. Thráinsson, S. D. Epstein, & S. Peter (Eds.), Studies in comparative Germanic syntax II (pp. 109–139). Dordrecht, The Netherlands: Kluwer.
  • Falluomini, C. (1999). Der sogenannte Codex Carolinus von Wolfenbüttel (Codex Guelferbytanus 64 Weissenburgensis): Mit besonderer Berücksichtigung der gotisch-lateinischen Blätter (255, 256, 277, 280). Wolfenbütteler Mittelalter-Studien 13. Wiesbaden, Germany: Harrassowitz.
  • Falluomini, C. (2009). Textkritische Anmerkungen zur gotischen Bibel. Annali della Facoltà di Lingue e Letterature Straniere dell’Università di Sassari, 5, 311–320.
  • Falluomini, C. (2013a). The Gothic version of the New Testament. In B. D. Ehrman & M. W. Holmes (Eds.), The text of the New Testament in contemporary research: Essays on the status quaestionis (2d ed., pp. 329–350). Leiden, The Netherlands: Brill.
  • Falluomini, C. (2013b). The Gothic Gospel of John and its text-critical character. In A. Kaliff & L. Munkhammar (Eds.), Wulfila 311–2011: International Symposium, Uppsala University June 15–18, 2011 (pp. 145–164). Uppsala, Sweden: Edita Västra Aros.
  • Falluomini, C. (2014). Zum gotischen Fragment aus Bologna. Zeitschrift für deutsches Altertum und deutsche Literatur, 143, 281–305.
  • Falluomini, C. (2015). The Gothic version of the Gospels and Pauline epistles: Cultural background, transmission and character. Berlin: De Gruyter.
  • Ferraresi, G. (1991). Die Stellung des gotischen Verbs im Lichte eines Vergleichs mit dem Althochdeutschen. M.A. thesis, University of Venice.
  • Ferraresi, G. (1998). Die Syntax des Infinitivs im Gotischen: Die Modalverben skulan und magan. In J. Strässler (Ed.), Tendenzen europäischer Linguistik: Akten des 31. Linguistischen Kolloquiums, Bern 1996 (pp. 46–49). Tübingen, Germany: Max Niemeyer.
  • Ferraresi, G. (2005). Word order and phrase structure in Gothic. Leuven, Belgium: Peeters.
  • Ferreiro, A. (2011). The Visigoths in Gaul and Iberia (update): A supplemental bibliography, 2007–2009. Medieval and Early Modern Iberian World 45. Leiden, The Netherlands: Brill. Also see Ferreiro’s The Visigoths in Gaul and Iberia (Update): A supplemental bibliography, 2004–2006 (Leiden, The Netherlands: Brill, 2008); he has earlier bibliographies as well.
  • Ferreiro, A. (2014). The Visigoths in Gaul and Iberia (update): A supplemental bibliography, 2010–2012. Leiden, The Netherlands: Brill.
  • Fertig, D. (2000). Null subjects in Gothic. American Journal of Germanic Linguistics and Literatures, 12, 3–21.
  • Finazzi, R. B., & Tornaghi, P. (2013). Gothica Bononiensia: Analisi linguistica e filologica di un nuovo documento. Aevum, 87(1), 113–155.
  • Fourquet, J. (1938). L’ordre des éléments de la phrase en germanique ancien. Paris: Les Belles Lettres.
  • Frey, E. (1989). Worttrennung und Silbenstruktur des Gotischen mit besonderer Berücksichtigung der Skeireins. Indogermanische Forschungen, 94, 272–293.
  • Friedrichs, E. (1891). Die Stellung des Pronomen personale im Gotischen. Jena, Germany: Hermann Pohle.
  • Friedrichsen, G. W. S. (1926). The Gothic version of the Gospels: A study of its style and textual history. London: Oxford University Press.
  • Friedrichsen, G. W. S. (1930). The silver ink of the Codex Argenteus. Journal of Theological Studies, 31, 189–192.
  • Friedrichsen, G. W. S. (1939). The Gothic version of the Epistles: A study of its style and textual history. London: Oxford University Press.
  • Friedrichsen, G. W. S. (1959). The Greek text underlying the Gothic version of the New Testament: The Gospel of St. Luke. In F. Mossé, Mélanges de linguistique et de philologie Fernand Mossé in memoriam (pp. 161–184). Paris: Didier. A critical discussion of selections made by Streitberg for the Vorlage.
  • Friedrichsen, G. W. S. (1961a). Gothic studies. Oxford: Blackwell.
  • Friedrichsen, G. W. S. (1961b). The Gothic ‘Skeireins’ in the Greek original. New Testament Studies, 8, 43–56.
  • Friedrichsen, George Washington Salisbury. (1962). The Gothic “Skeireins”: A new approach to interpretation. New Testament Studies, 9, 53–55.
  • Friedrichsen, George Washington Salisbury. (1963). The Gothic “Skeireins”, Leaf VI. New Testament Studies, 10(3), 368–373.
  • Friedrichsen, George Washington Salisbury. (1964). The Gothic commentary on St John “Skeireins”, Leaf VIII. New Testament Studies, 10(4), 499–504.
  • Friedrichsen, George Washington Salisbury. (1970). The Gothic “Skeireins” in the Greek original, Leaves V and VII. New Testament Studies, 16(3), 277–283.
  • Friedrichsen, G. W. S. (1977). Limitations of Gothic in representing Greek. In B. M. Metzger (Ed.), The early versions of the New Testament: Their origin, transmission, and limitations (pp. 388–393). Oxford: Oxford University Press.
  • Friesen, O. von, & Grape, A. (1927). Codex argenteus Upsaliensis jussu senatus universitatis phototypice editus. Uppsala, Sweden: Almqvist & Wiksell.
  • Fritz, M. (2011). Der Dual im Indogermanischen: Genealogischer und typologischer Vergleich einer grammatischen Kategorie im Wandel. Heidelberg, Germany: Winter.
  • Fritz, M., & Gippert, J. (Eds.). (2009). International conference on morphology and digitisation. Prague: Charles University Press.
  • Fullerton, G. L. (1991). Reduplication and the prosody of ancient Germanic. Beiträge zur Geschichte der deutschen Sprache und Literatur, 113, 1–21.
  • Fuß, E. (2003). On the historical core of V2 in Germanic. Nordic Journal of Linguistics, 26, 195–231.
  • Gabelentz, H. C. von der, & Löbe, J. (1846). Ulfilas: Veteris et Novi Testamenti versionis gothicae fragmenta quae supersunt ad fidem codd. Castigata, latinitate donata, adnotatione critica instructa, cum glossario et grammatica linguae gothicae coniuntis curis: Vol. 2. Grammatik der gothischen Sprache. Leipzig: Brockhaus.
  • Gaebeler, K. (1911). Die griechischen Bestandteile der gotischen Bibel. Zeitschrift für deutsche Philologie, 43, 1–118.
  • Ganina, N. A. (2007). ‘Alkuinova rukopis’: K ètimologičeskoj i istorikokul’turnoj traktovke. Voprosy jazykoznanija, 6, 60–72.
  • Ganina, N. A. (2011). Krymsko-gotskij jazyk. St. Petersburg, Russia: Aletejja/Aletheia.
  • García García, L. (2005). Germanische Kausativbildung: Die deverbalen jan-Verben im Gotischen. Göttingen, Germany: Vandenhoeck & Ruprecht.
  • Gast, V. (2006). The grammar of identity: Intensifiers and reflexives in Germanic languages. London: Routledge.
  • Glaser, E., Seiler, A., & Waldispühl, M (Eds.). (2011). LautSchriftSprache: Beiträge zur vergleichenden historischen Graphematik. Zurich, Switzerland: Chronos.
  • Goetting, L. (2007). Greek textual influence on Gothic complex verbs with pleonastic prepositions. Journal of Germanic Linguistics, 19, 309–347.
  • Götti, E. (1974). Die gotischen Bewegungsverben: Ein Beitrag zur Erforschung des gotischen Wortschatzes mit einem Ausblick auf Wulfilas Übersetzungstechnik. Berlin: Walter de Gruyter.
  • Granberg, A. (2010). Wulfila’s alphabet in the light of neighbouring scripts. NOWELE: North-West European Language Evolution, 58–59, 169–193.
  • Granberg, A. (2013). Establishing new alphabets (300–900 AD) and the relation between the structure of the alphabet and the shape of its letters. In A. Kaliff & L. Munkhammar (Eds.), Wulfila 311–2011: International Symposium, Uppsala University June 15–18, 2011 (pp. 165–177). Uppsala, Sweden: Edita Västra Aros.
  • Green, D. H. (1998). Language and history in the early Germanic world. Cambridge, U.K.: Cambridge University Press.
  • Greiner, P. (1992). Tempted by original syntax: Luther, Wulfila, and the Greek New Testament. In I. Rauch, G. F. Carr, & R. L. Kyes (Eds.), On Germanic linguistics: Issues and methods (pp. 97–107). Berlin: Mouton de Gruyter.
  • Grienberger, T. von. (1900). Untersuchungen zur gotischen Wortkunde. Vienna: Carl Gerolds Sohn.
  • Griepentrog, W. (1988). Synopse der gotischen Evangelientexte. Munich: Kitzinger. A study of parallel passages.
  • Griepentrog, W. (1990). Zur Text- und Überlieferungsgeschichte der gotischen Evangelientexte. Innsbrucker Beiträge zur Sprachwissenschaft 49. Innsbruck, Austria: Institut für Sprachwissenschaft der Universität Innsbrück.
  • Grimm, J. L. C. (1819–1837). Deutsche Grammatik (Vols. 1–4). Göttingen, Germany: Dieterich. The four volumes are Erster Theil (1819; 2d ed. 1822), Zweiter Theil (1826), Dritter Theil (1831), and Vierter Theil (1837).
  • Grønvik, O. N. (1983). Die dialektgeographische Stellung des Krimgotischen und die krimgotische cantilena. Oslo, Norway: University of Oslo Press.
  • Gryson, R. (1990). La version gotique des évangiles: Essai de réévaluation. Revue théologique de Louvain, 21, 3–31.
  • Gussmann, E. (Ed.). (1985). Phono-morphology: Studies in the intersection of phonology and morphology. Lublin, Poland: Catholic University of Lublin Press.
  • Guxman (Gukhman), Mirra Moiseevna. (1940). Proisxoždenije stroja gotskogo glagola [Origin of the system of the Gothic verb]. Moscow: Nauka.
  • Guxman (Gukhman), M. M. (1958). Gotskij jazyk. Moscow: Vysšaja škola. Excellent summaries of Gothic syntax (pp. 119–133, 214–240).
  • Guxman (Gukhman), M. M. (1964). Razvitie zalogovyx protivopostavlenij v germanskix jazykax. Moscow: Nauka.
  • Guxman (Gukhman), M. M., Makaev, È. A., & Jarceva, V. N. (1977–1978). Istoriko-tipologičeskaja morfologija germanskix jazykov (Vols. 1–3). Moscow: Nauka.
  • Hansen, B. S. S. (2015). Review article of Kotin (2012). NOWELE: North-West European Language Evolution, 68, 121–128.
  • Harbert, W. (1992). Gothic relative clauses and syntactic theory. In I. Rauch, G. F. Carr, & R. L. Kyes (Eds.), On Germanic linguistics: Issues and methods (pp. 109–146). Berlin: Mouton de Gruyter.
  • Harbert, W. (2007). The Germanic languages. Cambridge, U.K.: Cambridge University Press.
  • Harbert, W. (2012). Contrastive linguistics and language change: Reanalysis in Germanic relative clauses. Languages in Contrast, 12, 27–46.
  • Heather, P. (1996). The Goths. Oxford: Blackwell.
  • Heather, P. (2010). Empires and barbarians: The fall of Rome and the birth of Europe. Oxford: Oxford University Press.
  • Heather, P., & Matthews, J. (1991). The Goths in the fourth century. Liverpool, U.K.: Liverpool University Press. Translated texts for historians are in Vol. 11. Reprinted in 2004.
  • Heffner, R. M. S. (1929). Gothic hiri. Journal of English and Germanic Philology, 28, 343–359.
  • Hempel, H. (1962). Gotisches Elementarbuch (3d ed.). Sammlung Göschen 79–79a. Berlin: Walter de Gruyter.
  • Hench, G. A. (1897). The voiced spirants in Gothic. Journal of Germanic Philology, 1, 45–58.
  • Hill, E. (2010). A case study in grammaticalized inflectional morphology: Origin and development of the Germanic weak preterite. Diachronica, 27(3), 411–458.
  • Holtzmann, A. (1835). Review of Hans Ferdinand Maßmann, Skeireins . . . (Munich: George Jaquet, 1834). Heidelberger Jahrbücher der Literatur, 28, 854–863.
  • Holtzmann, A. (1863). Das gothische Adjectivum. Germania, 8, 257–268.
  • Horn, W. (Ed.). (1924). Beiträge zur germanischen Sprachwissenschaft: Festschrift für Otto Behaghel. Heidelberg, Germany: Winter.
  • Hug, J. L. (1821). Einleitung in die Schriften des Neuen Testaments I (2d ed.). Stuttgart: Cotta.
  • Hunter, M. J. (1969). The Gothic Bible. In G. W. H. Lampe (Ed.), Cambridge History of the Bible: Vol. 2. The West from the Fathers to the Reformation (pp. 338–362). Cambridge, U.K.: Cambridge University Press.
  • Ioanesjan, E. R. (Ed.). (2005). Problemy i metody sovremennoj lingvistiki: Sbornik naučnyx trudov; Vypusk 1. Moscow: Institut Jazykoznanija RAN.
  • Ivanov, V. V. (1999). Indo-European syntactic rules and Gothic morphology. UCLA Indo-European Studies, 1, 103–120.
  • Jamison, S. W., Melchert, H. C., & Vine, B. (Eds.). (2010). Proceedings of the 21st Annual UCLA Indo-European Conference: Los Angeles, October 30th and 31st, 2009. Bremen, Germany: Hempen.
  • Jasanoff, J. H. (2003). Hittite and the Indo-European verb. Oxford: Oxford University Press.
  • Jasanoff, J. H. (2004). Gothic. In R. D. Woodard (Ed.), The Cambridge encyclopedia of the world’s ancient languages (pp. 881–906). Cambridge, U.K.: Cambridge University Press.
  • Jasanoff, J. H. (2007). From reduplication to ablaut: The class VII strong verbs of Northwest Germanic. Historische Sprachforschung, 120, 241–284.
  • Jellinek, M. H. (1926). Geschichte der gotischen Sprache. Berlin: Walter de Gruyter.
  • Jones, O. F. (1958). Gothic au in inflectional syllables. Language, 34, 33–39.
  • Jones, O. F. (1960). Nonsyllabic allophones of Gothic /w/. Language, 36, 508–515.
  • Joseph, B. (1981). On the so-called ‘passive’ use of the Gothic active infinitive. Journal of English and Germanic Philology, 80, 369–379.
  • Jülicher, A. (1963–1976). Itala: Das Neue Testament in altlateinischer Überlieferung nach den Handschriften (2d ed., Vols. 1– 4). Berlin: Walter de Gruyter. Vol. 1, Matthäus-Evangelium (1972); Vol. 2, Marcus-Evangelium (1970); Vol. 3, Lucas-Evangelium (1976); Vol. 4, Johannes-Evangelium (1963).
  • Kaliff, A., & Munkhammar, L. (Eds.). (2013). Wulfila 311–2011: International Symposium, Uppsala University June 15–18, 2011. Uppsala, Sweden: Edita Västra Aros.
  • Karpov, V. I. (2005a). Struktura, semantika i funkcii složnyx slov i slovosočetanij v gotskom jazyke (sopostavitelnyj aspekt). Ph.D. dissertation, Institut jazykoznanija Rossijskaja akademija nauk.
  • Karpov, V. I. (2005b). K voprosu ob otnošenii složnyx slov i slovosočetanij v gotskom jazyke. In E. R. Ioanesjan (Ed.), Problemy i metody sovremennoj lingvistiki: Sbornik naučnyx trudov; Vypusk 1 (pp. 199–214). Moscow: Institut Jazykoznanija RAN.
  • Katz, J. T. (2003). Studies in Indo-European personal pronouns. Göttingen, Germany: Vandenhoeck & Ruprecht.
  • Kauffmann, F. (1911a). Zur Textgeschichte der gotischen Bibel. Zeitschrift für deutsche Philologie, 43, 118–132.
  • Kauffmann, F. (1911b). Beiträge zur Quellenkritik der gotischen Bibelübersetzung. Zeitschrift für deutsche Philologie, 43, 401–428.
  • Kenyon, F. G. (1937). The text of the Greek Bible. London: Duckworth.
  • Keydana, G. (1997). Absolute Konstruktionen in altindogermanischen Sprachen. Göttingen, Germany: Vandenhoeck & Ruprecht.
  • Kieckers, E. (1960). Handbuch der vergleichenden gotischen Grammatik. Munich: Max Hueber. First published in 1927.
  • Kim, R. (2010). On the prehistory of Old English dyde. Medieval English Mirror, 6, 9–22.
  • Kiparsky, P. (1996). The shift to head-initial VP in Germanic. In H. Thráinsson, S. D. Epstein, & S. Peter (Eds.), Studies in comparative Germanic syntax II (pp. 140–179). Dordrecht, The Netherlands: Kluwer.
  • Kiparsky, P. (2000). Analogy as optimization: ‘Exceptions’ to Sievers’ law in Gothic. In A. Lahiri (Ed.), Analogy, levelling, markedness: Principles of change in phonology and morphology (pp. 15–46). Berlin: Mouton de Gruyter.
  • Kiparsky, P. (2009). The Germanic weak preterite. In P. O. Steinkrüger & M. Krifka (Eds.), On inflection (pp. 107–124). Berlin: Mouton de Gruyter.
  • Kiparsky, P. (2010). Compositional vs. paradigmatic approaches to accent and ablaut. In S. W. Jamison, H. C. Melchert, & B. Vine (Eds.), Proceedings of the 21st Annual UCLA Indo-European Conference: Los Angeles, October 30th and 31st, 2009 (pp. 137–181). Bremen, Germany: Hempen.
  • Kiparsky, P. (2011). Greek anaphora in cross-linguistic perspective. Manuscript, Stanford University.
  • Klein, J. S. (1992a). On the independence of Gothic syntax, I: Interrogativity, complex sentence types, tense, mood, and diathesis. Journal of Indo-European Studies, 20, 339–379.
  • Klein, J. S. (1992b). On the idiomatic nature of the Gothic New Testament: A comparative study of prepositional usage in Gothic and New Testament Greek. Transactions of the Philological Society, 90, 1–80.
  • Klein, Jared Stephen. (2011). Negation and polarity in the Greek, Gothic, Classical Armenian, and Old Church Slavic Gospels: A preliminary study. In Welo (ed.) (pp. 131–154).
  • Klein, J. S. (1994). Gothic þaruh, þanuh, and -(u)h þan. Indogermanische Forschungen, 99, 253–276.
  • Klein, J. S., & Condon, N. L. (1993). Gothic (u)h: A synchronic and comparative study. Transactions of the Philological Society, 91, 1–62.
  • Kluge, F. (1926). Nominale Stammbildungslehre der altgermanischen Dialekte (3d ed.). Halle, Germany: Niemeyer.
  • Köbler, G. (1989). Gotisches Wörterbuch (2d ed.). Leiden, The Netherlands: Brill.
  • Kokowski, A. (2007). The agriculture of the Goths between the first and fifth centuries AD (central and eastern Europe—the Roman period and the early migration period). In S. J. B. Barnish & F. Marazzi (Eds.), The Ostrogoths from the migration period to the sixth century: An ethnographic perspective (pp. 221–236). Studies in Historical Archaeoethnology 7. Woodbridge, U.K.: Boydell.
  • Koppitz, A. (1900). Gotische Wortstellung. Zeitschrift für deutsche Philologie, 32, 433–463. Continued in 1901 in Zeitschrift für deutsche Philologie, 33, 7–45.
  • Kotin, M. L. (2012). Gotisch: Im (diachronischen und typologischen) Vergleich. Heidelberg, Germany: Winter.
  • For discussion, see Pierce, 2013a; Seebold, 2013; Robinson, 2014; Quak, 2014; Hansen, 2015; and Roland Schuhmann, Anmerkungen zu M. Kotin, Gotisch, Heidelberg, 2012. On derivation there is little difference between Kotin and Guxman, 1958, as a brief comparison of Kotin (pp. 386–392) with Guxman (pp. 100–205) reveals.
  • Kovari, G. (1984). Studien zum germanischen Artikel: Entstehung und Verwendung des Artikels im Gotischen. Vienna: Halosar. With valuable statistics.
  • Krahe, H., & Meid, W. (1967). Germanische Sprachwissenschaft: Vol. 3. Wortbildungslehre. Sammlung Göschen 1218b. Berlin: Walter de Gruyter.
  • Krahe, H., & Seebold, E. (1967). Historische Laut- und Formenlehre des Gotischen (2d ed.). Heidelberg, Germany: Winter.
  • Krause, M. (1987). Sémantique et syntaxe des préverbes en gotique (Vols. 1–4). Ph.D. dissertation, University of Paris IV, Sorbonne, France. Available in 1988 through Atelier National de Reproduction des Thèses (Lille, France).
  • Krause, M. (1994). Zeitangaben im Gotischen. In Y. Desportes (Ed.), Philologische Forschungen: Festschrift für Philippe Marcq (pp. 18–42). Heidelberg, Germany: Winter.
  • Krause, M. (1995). Das System der spatialen Präpositionen im Gotischen. Sprachwissenschaft, 20, 1–31.
  • Krause, W. (1968). Handbuch des Gotischen (3d ed.). Munich: Beck.
  • Kroonen, G. (2013). Etymological dictionary of Proto-Germanic. Leiden, The Netherlands: Brill.
  • Lahiri, A. (Ed.). (2000). Analogy, levelling, markedness: Principles of change in phonology and morphology. Berlin: Mouton de Gruyter.
  • Lambdin, T. O. (2006). An introduction to the Gothic language. Eugene, OR: Wipf & Stock.
  • Lampe, G. W. H. (Ed.). (1969). Cambridge history of the Bible: Vol. 2. The West from the Fathers to the Reformation. Cambridge, U.K.: Cambridge University Press.
  • Lehmann, W. P. (Ed.). (1986). A Gothic etymological dictionary. Leiden, The Netherlands: Brill.
  • Leiss, E. (2000). Artikel und Aspekt: Die grammatischen Muster von Definitheit. Berlin: Walter de Gruyter.
  • Leiss, E. (2007). Covert patterns of definiteness/indefiniteness and aspectuality in Old Icelandic, Gothic, and Old High German. In E. Stark, E. Leiss, & W. Abraham (Eds.), Nominal determination: Typology, context constraints, and historical emergence (pp. 73–102). Amsterdam: John Benjamins.
  • Leiss, E. (2012). Aspectual patterns of covert coding of modality in Gothic and Old High German. In W. Abraham & E. Leiss (Eds.), Covert patterns of modality (pp. 175–200). Newcastle upon Tyne, U.K.: Cambridge Scholars.
  • Lendinara, P. (1992). Wulfila as the inventor of the Gothic alphabet: The tradition in late antiquity and the Middle Ages. General Linguistics, 32, 217–225.
  • Lendinara, P., Raschellà, F. D., & Dallapiazza, M. (Eds.). (2011). Saggi in onore di Piergiuseppe Scardigli. Bern, Switzerland: Peter Lang.
  • Lenk, R. (1910). Die Syntax der Skeireins. Beiträge zur Geschichte der deutschen Sprache und Literatur, 36, 237–306.
  • Liberman, A. (2010). Verner’s law. NOWELE: North-West European Language Evolution, 58–59, 381–425.
  • Lindberg, C. E. (2010). Die Subjunktion þatei nach verba dicendi und sentiendi im Codex Argenteus. NOWELE: North-West European Language Evolution, 58–59, 255–268.
  • Lloyd, A. L. (1979). Anatomy of the verb: The Gothic verb as a model for a unified theory of aspect, actional types, and verbal velocity. Amsterdam: John Benjamins.
  • Longobardi, G. (1978). Problemi di sintassi gotica: Aspetti teorici e descrittivi. M.A. thesis, University of Pisa, Pisa, Italy.
  • Lühr, R. (1985). Die Deklination griechischer und lateinischer Wörter in Wulfilas gotischer Bibelübersetzung. Münchener Studien zur Sprachwissenschaft, 46, 139–155.
  • Lühr, R. (2005). Der Einfluß der klassischen Sprachen auf die Germanische Grammatik. In G. Meiser & O. Hackstein (Eds.), Sprachkontakt und Sprachwandel: Akten der XI. Fachtagung der Indogermanischen Gesellschaft, 17.–23. September, Halle an der Saale (pp. 341–362). Wiesbaden, Germany: Reichert.
  • MacLeod, M., & Mees, B. (2006). Runic amulets and magic objects. Woodbridge, U.K.: Boydell.
  • Mailhammer, R. (2007). The Germanic strong verbs: Foundations and development of a new system. Berlin: Mouton de Gruyter.
  • Makovskij, M. M. (2011). K probleme vida v gotskom jazyke. Voprosy Jazykoznanija, 60(6), 90–104.
  • Marchand, J. W. (1955). Der Ursprung des gotischen þorn-Zeichens. Beiträge zur Geschichte der deutschen Sprache und Literatur, 77, 490–494.
  • Marchand, J. W. (1973a). The sounds and phonemes of Wulfila’s Gothic. The Hague: Mouton.
  • Marchand, J. W. (1973b). On the Gotica Veronensia. New Testament Studies, 19, 465–468.
  • Marold, E., & Zimmermann, C. (Eds.). (1995). Nordwestgermanisch. Berlin: Walter de Gruyter.
  • Maslov, Y. S. (1959). Kategorija predel’nosti/nepredel’nosti glagol’nogo dejstvija v gotskom jazyke. Voprosy jazykoznanija, 5, 69–80. Reprinted with minor changes and updates in Maslov’s Izbrannye trudy: Aspektologija; Obščee jazykoznanie, edited by A. V. Bondarko, T. A. Majsak, and V. A. Plungjan, pp. 249–266 (Moscow: Jazyki slavjanskoj kul’tury, 2004).
  • Maßmann, H. F. (1834). Skeireins aiwaggelions þairh Iohannen: Auslegung des Evangelii Johannis in Gothischer Sprache; Aus römischen und mayländischen Handschriften nebst lateinischer Uebersetzung, belegenden Anmerkungen, geschichtlicher Untersuchung, gothisch-lateinischem Wörterbuche und Schriftproben. Munich: George Jaquet.
  • McKnight, G. H. (1897). The language of the Skeireins. Modern Language Notes, 12, 205–209.
  • McLynn, N. (2007). Little Wolf in the big city: Ulfila and his interpreters. Bulletin of the Institute of Classical Studies [Special issue, Suppl. 591], 50, 125–135.
  • Mees, B. (2004). Runo-Gothica: The runes and the origin of Wulfila’s script. Die Sprache, 43, 55–79.
  • Meier-Brügger, M. (2010). Indogermanische Sprachwissenschaft (9th ed.). Berlin: De Gruyter.
  • Meillet, A. (1908–1909). Notes sur quelques faits gotiques. Mémoire de la Société de Linguistique de Paris, 15, 73–103.
  • Meillet, A. (1949). Caractères généraux des langues germaniques (7th ed.). Paris: Hachette.
  • Originally published in 1917; reprinted as recently as 2009 (Cambridge, U.K.: Cambridge University Press).
  • Meiser, G., & Hackstein, O. (Eds.). (2005). Sprachkontakt und Sprachwandel: Akten der XI. Fachtagung der Indogermanischen Gesellschaft, 17.–23. September, Halle an der Saale. Wiesbaden, Germany: Reichert.
  • Metlen, M. (1932). Does the Gothic Bible represent idiomatic Gothic?—an investigation based primarily on the use of the present participle in the Gothic Bible: With some corroborating facts drawn from other materials. Ph.D. dissertation, Northwestern University, Evanston, IL.
  • Metlen, M. (1933). What a Greek interlinear of the Gothic Bible can teach us. Journal of English and Germanic Philology, 32, 530–548.
  • Metlen, M. (1938). Absolute constructions in the Gothic Bible. Publications of the Modern Language Association of America, 53, 631–644.
  • Metzger, B. M. (1977). The early versions of the New Testament: Their origin, transmission, and limitations. Oxford: Oxford University Press.
  • Metzger, B. M., & Ehrman, B. D. (2005). The text of the New Testament: Its transmission, corruption, and restoration. Oxford: Oxford University Press.
  • Meyer, L. K. H. (1863). Über die Flexion der Adjectiva im Deutschen: Eine sprachwissenschaftliche Abhandlung. Berlin: Weidmann.
  • Meyer, L. K. H. (1869). Die gothische Sprache: Ihre Lautgestaltung, insbesondere im Verhältniss [sic] zum Altindischen, Griechischen und Lateinischen. Berlin: Weidmann.
  • Miller, D. G. (1975). The Gothic complementizers þammei and ei. Indogermanische Forschungen, 80, 110–117.
  • Miller, D. G. (1994). Ancient scripts and phonological knowledge. Amsterdam: John Benjamins.
  • Miller, D. G. (2002). Nonfinite structures in theory and change. Oxford: Oxford University Press.
  • Miller, D. G. (2010). Language change and linguistic theory (Vols. 1–2). Oxford: Oxford University Press.
  • Miller, D. G. (2012). Latin suffixal derivatives in English and their Indo-European ancestry (2d ed.) Oxford: Oxford University Press.
  • Mirowicz, A. (1935). Die Aspektfrage im Gotischen. Vilnius, Lithuania: Nakładem Towarzystwa Przyjaciół Nauk.
  • Mittermüller, K. (1983). Zur Struktur der verbalen Personalindikation im Gotischen. Kirchzarten, Germany: Burg-Verlag.
  • Morgaleva, V. V. (2008a). Absoljutnye pričastnye oboroty v gotskom i drevneanglijskom. Vestnik TGPU, 2(76), 73–77.
  • Morgaleva, V. V. (2008b). Absoljutnyj datelnyj pričastnyj oborot v gotskom jazyke. Vestnik Tomskogo Gosudarstvennogo Universiteta, 316, 21–24.
  • Mossé, F. (1938). Histoire de la forme périphrastique être + participe présent en germanique (Vols. 1–2). Paris: Klincksieck.
  • Mossé, F. (1956). Manuel de la langue gotique: Grammaire, textes, notes, glossaire (2d ed.). Paris: Montaigne.
  • Mourek, V. E. (1890). Syntaxis gotských předložek. Prague: Nauka.
  • Mourek, V. E. (1893). Syntaxis složených vět v gotštině. Prague: České Akademie Císaře Františka Josefa. In Czech with a German summary, Syntax der mehrfachen Sätze im Gotischen, pp. 287–334.
  • Mourek, V. E. (1895). Nochmals über den Einfluss des Hauptsatzes auf den Modus des Nebensatzes im Gotischen. Sitzungsberichte der Königl. Böhmischen Gesellschaft der Wissenschaften: Classe für Philosophie, Geschichte und Philologie, 17, 1–21.
  • Mourek, V. E. (1903). Zur Negation im Altgermanischen. Prague: Verlag der Königl. Böhmischen Gesellschaft der Wissenschaften.
  • Munkhammar, L. (2011). The Silver Bible: Origins and history of the Codex Argenteus. Uppsala, Sweden: Selenas.
  • Nedoma, R. (2010). Schrift und Sprache in den ostgermanischen Inschriften. NOWELE: North-West European Language Evolution, 58–59, 1–70.
  • Nestle, E., Nestle, E., Aland, B., & Aland, K. (Eds.). (2012). Novum Testamentum Graece (28th ed.). Stuttgart: Deutsche Bibelgesellschaft.
  • Nielsen, H. F. (1989). The Germanic languages: Origins and early dialectal interrelations. Tuscaloosa: University of Alabama Press.
  • Nielsen, H. F. (1995). Methodological problems in Germanic dialect grouping. In E. Marold & C. Zimmermann (Eds.), Nordwestgermanisch (pp. 115–123). Berlin: Walter de Gruyter.
  • Nielsen, H. F. (2002). Nordic–West Germanic relations. In O. Bandle, K. Braunmüller, E. H. Jahr, A. Karker, H.‑P. Naumann, & U. Teleman (Eds.), The Nordic languages: An international handbook of the history of the North Germanic languages (Vol. 1) (pp. 558–568). Berlin: Walter de Gruyter.
  • Nielsen, H. F. (2010). Gothic and early runic: Two sound systems compared. NOWELE: North-West European Language Evolution, 58–59, 427–442.
  • Nielsen, H. F. (2011). Gothic runic inscriptions in Scandinavia? Futhark, 2, 51–61.
  • Pagliarulo, G. (2006). On the alleged functions of word-order in Wulfilian Gothic. Journal of Indo-European Studies, 34, 437–450.
  • Pagliarulo, G. (2008). Innovazione e conservazione nel passivo gotico. In V. Dolcetti Corazza & R. Gendre (Eds.), Intorno alla Bibbia gotica (pp. 329–339). Alessandria, Italy: Edizioni dell’Orso.
  • Pagliarulo, G. (2011a). Su alcuni casi di cattivo accordo in perifrasi passive gotiche. Alessandria, 4, 95–101.
  • Pagliarulo, G. (2011b). Notes on the function of Gothic -u. Journal of Indo-European Studies, 39, 395–413.
  • Pakis, V. A. (2008). Homoian vestiges in the Gothic translation of Luke 3,23–28. Zeitschrift für deutsches Altertum und deutsche Literatur, 137, 277–304.
  • Pakis, V. A. (2010). Praesens historicum and the question of Old Latin influence on the Gothic Bible. NOWELE: North-West European Language Evolution, 58–59, 239–254.
  • Parker, D. C. (1992). Codex Bezae: An early Christian manuscript and its text. Cambridge, U.K.: Cambridge University Press.
  • Paul, H. (1880). Gotisch ai und au vor Vocal. Beiträge zur Geschichte der deutschen Sprache und Literatur, 7, 152–160.
  • Paul, H. (1882). Noch einmal Gotisch ai und au vor Vocal. Beiträge zur Geschichte der deutschen Sprache und Literatur, 8, 210–222.
  • Peeters, C. (1978). Reflexive and non-reflexive possessives in Gothic (a particular use). Zeitschrift für vergleichende Sprachforschung, 92, 233–234.
  • Penzl, H. (1950). Orthography and phonemes in Wulfila’s Gothic. Journal of English and Germanic Philology, 49, 217–230.
  • Penzl, H. (1977). Names and historical Germanic phonology: The bilingual sixth century Ravenna deeds. Names, 25, 8–14.
  • Petersen, C. T. (Ed.). (2005). Bibliographia gotica amplificata. Gotica Minora 5. Darmstadt: Syllabus.
  • Pierce, M. (2006). Syllable structure and Sievers’ law in Gothic. Journal of Germanic Linguistics, 18, 275–319.
  • Pierce, M. (2007). Die Rolle linguistischer Theorien in der historischen Sprachwissenschaft: Eine Fallstudie aus dem Gotischen. Indogermanische Forschungen, 112, 236–243.
  • Pierce, M. (2013a). Review of Kotin (2012). Folia Linguistica Historica, 34, 276–281.
  • Pierce, M. (2013b). Syllable weight in Gothic. Indogermanische Forschungen, 118, 213–225.
  • Piras, A. (2007). Manuale di gotico: Avviamento alla lettura della versione gotica del Nuovo Testamento. Rome: Herder.
  • Poulter, A. (2007). Invisible Goths within and beyond the Roman Empire. Bulletin of the Institute of Classical Studies [Special issue, Suppl. 591], 50, 169–182.
  • Probert, P., & Willi, A. (Eds.). (2012). Laws and rules in Indo-European. Oxford: Clarendon.
  • Prokosch, E. (1939). A comparative Germanic grammar. Baltimore: Linguistic Society of America. Reprinted in 1966.
  • Quak, A. (2014). Review of Kotin (2012). Amsterdamer Beiträge zur älteren Germanistik, 71, 268–271.
  • Quinlin, D. P. (2007). Wulfila’s (mis)translation of Philippians 2:6. Indogermanische Forschungen, 112, 208–214.
  • Raschellà, F. D. (2011). Models and principles of Wulfila’s Gothic alphabet: Some methodological remarks. In E. Glaser, A. Seiler, & M. Waldispühl (Eds.), LautSchriftSprache: Beiträge zur vergleichenden historischen Graphematik (pp. 109–124). Zurich, Switzerland: Chronos.
  • Ratkus, A. (2011). The adjective inflection in Gothic and early Germanic: Structure and development. Ph.D. dissertation, University of Cambridge, Cambridge, U.K.
  • Ratkus, A. (in press). Gothic possessives, adjectives and other modifiers in -ata. Journal of Germanic Linguistics.
  • Ratkus, A. (2016). Linearisation of adnominal possessives in Gothic. Manuscript, University of Vilnius, Vilnius, Poland.
  • Rauch, I. (2011). The Gothic language: Grammar, genetic provenance and typology, readings (2d ed.). New York: Peter Lang.
  • Rauch, I., & Carr, G. F. (Eds.). (1995). Insights in Germanic linguistics I: Methodology in transition. Berlin: Mouton de Gruyter.
  • Rauch, I., & Carr, G. F. (Eds.). (1997). Insights in Germanic linguistics II: Classic and contemporary. Berlin: Mouton de Gruyter.
  • Rauch, I., Carr, G. F., & Kyes, R. L. (Eds.). (1992). On Germanic linguistics: Issues and methods. Berlin: Mouton de Gruyter.
  • Riad, T. (1992). Structures in Germanic prosody: A diachronic study with special reference to the Nordic languages. Ph.D. dissertation, Stockholm University, Stockholm.
  • Riad, T. (2004). Syllabification and word division in Gothic. Journal of Germanic Linguistics, 16, 173–202.
  • Rice, A. L. (1932). Gothic prepositional compounds in their relation to their Greek originals. Language Dissertations 11. Philadelphia: Linguistic Society of America.
  • Ringe, D. R. (2006). From Proto-Indo-European to Proto-Germanic: A linguistic history of English (Vol. 1). Oxford: Oxford University Press.
  • Ringe, D. R. (2012). Cladistic principles and linguistic reality: The case of West Germanic. In P. Probert & A. Willi (Eds.), Laws and rules in Indo-European (pp. 33–42). Oxford: Clarendon.
  • Roberge, P. T. (1983). Those Gothic spirants again. Indogermanische Forschungen, 88, 109–155.
  • Roberge, Paul T. (1984). Remarks on Gothic final -g. Beiträge zur Geschichte der deutschen Sprache und Literatur 106:325–343.
  • Robinson, O. W. (2014). Review of Kotin (2012). Diachronica, 31, 286–290.
  • Robinson, M. A., & Pierpont, W. G. (Eds.). (2005). The New Testament in the original Greek: Byzantine textform. Southborough, MA: Chilton.
  • Rose, M. L. (1976). The syntax of the reflexive pronoun in Gothic and Old Norse. Ph.D. dissertation, University of Wisconsin, Madison.
  • Rousseau, A. (2003). Nouveau regard sur les modalités du gotique. Revue belge de philologie et d’histoire, 81, 729–748.
  • Rousseau, A. (2012). Grammaire explicative du gotique. Paris: L’Harmattan.
  • Rubio, G. (2009). Semitic influence in the history of Latin syntax. In P. Baldi & P. Cuzzolin (Eds.), New perspectives on historical Latin syntax (Vol. 1, pp. 195–239). Trends in Linguistics. Berlin: Mouton de Gruyter.
  • Sauvageot, A. (1929). L’emploi de l’article en gotique. Paris: Champion.
  • Scardigli, P. (1973). Die Goten: Sprache und Kultur. Munich: Beck.
  • Scardigli, P. (2002). Nordic-Gothic linguistic relations. In O. Bandle, K. Braunmüller, E. H. Jahr, A. Karker, H.‑P. Naumann, & U. Teleman (Eds.), The Nordic languages: An international handbook of the history of the North Germanic languages (Vol. 1, pp. 553–558). Berlin: Walter de Gruyter.
  • Schäferdiek, K. (1981). Die Fragmente der Skeireins und der Johanneskommentar des Theodor von Herakleia. Zeitschrift für deutsches Altertum und deutsche Literatur, 110, 175–193.
  • Schaffner, S. (2001). Das Vernersche Gesetz und der innerparadigmatische grammatische Wechsel des Urgermanischen im Nominalbereich. Innsbrucker Beiträge zur Sprachwissenschaft 103. Innsbruck, Austria: Institut für Sprachen und Literaturen der Universität Innsbrück.
  • Schröder, E. (1910). Busbecqs krimgotisches Vokabular. Nachrichten der königlichen Gesellschaft der Wissenschaften zu Göttingen: Philologisch-historische Klasse, 1910, 1–16.
  • Schröder, W. (1957). Die Gliederung des gotischen Passivs. Beiträge zur Geschichte der deutschen Sprache und Literatur, 79, 1–24.
  • Schröder, W. (1957–1958). ‘Germanisches’ werden-Passiv und ‘christliches’ sein-Passiv bei Wulfila. Zeitschrift für deutsches Altertum und Literatur, 88, 101–115.
  • Schubert, H.‑J. (1968). Die Erweiterung des bibelgotischen Wortschatzes mit Hilfe der Methoden der Wortbildungslehre. Munich: Hueber.
  • Schulze, W. (1905). Griechische Lehnworte im Gotischen. Sitzungsberichte der königlichen Preußischen Akademie der Wissenschaften, 36, 726–759.
  • Schulze, W. (1924). Personalpronomen und Subjektausdruck im Gotischen. In W. Horn (Ed.), Beiträge zur germanischen Sprachwissenschaft: Festschrift für Otto Behaghel (pp. 92–109). Heidelberg, Germany: Winter.
  • Schuhmann, R. (2015). Gotische Grammatik. Wiesbaden, Germany: Reichert.
  • Seebold, E. (2010). Die gotischen Buchstabennamen: Mit einem Exkurs über die englischen Manuskriptrunen. NOWELE: North-West European Language Evolution, 58–59, 71–168.
  • Seebold, E. (2013). Review of Kotin (2012). Beiträge zur Namenforschung, 48, 124–126.
  • Seppänen, A. (1985). On the use of the dual in Gothic. Zeitschrift für deutsches Altertum und deutsche Literatur, 114, 1–41.
  • Shimomiya, T. (2009). Characteristics of Germanic languages. In J. O. Askedal, I. Roberts, T. Matsushita, & H. Hasegawa (Eds.), Germanic languages and linguistic universals (pp. 57–68). Development of the Anglo-Saxon Language and Linguistic Universals 1. Amsterdam: John Benjamins.
  • Sickel, F. A. T. von. (1875). Alcuinstudien. Sitzungsberichte der Wiener Akademie der Wissenschaften, 79, 461–550.
  • Sivan, H. (1996). Ulfila’s own conversion. Harvard Theological Review, 89, 373–386.
  • Sizova, I. A. (1978). Stanovlenie germanskogo glagol’nogo slovoobrazovanija: Na materiale gotskogo jazyka. Moscow: Nauka.
  • Skeat, W. W. (1868). Mœso-Gothic glossary, with an introduction, an outline of Mœso-Gothic grammar, and a list of Anglo-Saxon and old and modern English words etymologically connected with Mœso-Gothic. London: Asher.
  • Snædal, M. H. (2002a). Naples, Arezzo, Verona. In C. T. Petersen (Ed.), Gotica Minora: Miscellanea de linguae Ulfilae collecta (Vol. 1, pp. 1–2). Hanau, Germany: Syllabus.
  • Snædal, M. H. (2002b). Gotica Veronensia. In C. T. Petersen (Ed.), Gotica Minora: Miscellanea de linguae Ulfilae collecta (Vol. 1, pp. 3–6). Hanau, Germany: Syllabus.
  • Snædal, M. H. (2003). The Gothic text of codex Gissensis. Gotica Minora, 2, 1–20.
  • Snædal, M. H. (2006). Wulfila and Oddur Gottskálksson. Gothica Minora, 6, 1–4.
  • Snædal, M. H. (2009a). Ostgermanische Morphologie / East-Germanic morphology. In M. Fritz & J. Gippert (Eds.), International conference on morphology and digitisation (pp. 147–167). Prague: Charles University Press.
  • Snædal, M. H. (2009b). The ‘Vandal’ epigram. Filologia Germanica / Germanic Philology, 1 [Special issue, Lingua e Cultura dei Goti / Language and Culture of the Gods], 181–215.
  • Snædal, M. H. (2010). Uns/Unsis and colloquial Gothic. NOWELE: North-West European Language Evolution, 58–59, 301–322
  • Snædal, M. H. (2011a). The Runic inscriptions from Kovel and Pietroassa. In P. Lendinara, F. D. Raschellà, & M. Dallapiazza (Eds.), Saggi in onore di Piergiuseppe Scardigli (pp. 233–243). Bern, Switzerland: Peter Lang.
  • Snædal, M. H. (2011b). Gothic «ggw»*. Studia Linguistica, 128, 145–154.
  • Snædal, M. H. (2013a). A concordance to biblical Gothic (3d ed., Vols. 1–2). Reykjavik: University of Iceland Press. Vol. 1, Introduction, texts; Vol. 2, Concordance.
  • Snædal, M. H. (2013b). Gothic letter (and phoneme) statistics. Studia Linguistica Universitatis Iagellonicae Cracoviensis, 130, 277–295.
  • Snædal, Magnús Hreinn. (2015a). Gothic contact with Greek: Loan translations and a translation problem. In Askedal & Nielsen (ed.) (pp. 75–90).
  • Snædal, Magnús Hreinn. (2015b). Gothic contact with Latin: Gotica Parisina and Wulfila’s alphabet. In Askedal & Nielsen (ed.) (pp. 91–107).
  • Sroka, K. A. (Ed.). (1996). Kognitive Aspekte der Sprache: Akten des 30. Linguistischen Kolloquiums, Gdańsk, 1995. Linguistische Arbeiten 360. Tübingen, Germany: Max Niemeyer.
  • Stark, E., Leiss, E., & Abraham, W. (Eds.). (2007). Nominal determination: Typology, context constraints, and historical emergence. Amsterdam: John Benjamins.
  • Stearns, M., Jr. (1978). Crimean Gothic: Analysis and etymology of the corpus. Saratoga, CA: Anma Libri.
  • Stearns, M., Jr. (1989). Das Krimgotische. In H. Beck (Ed.), Germanische Rest- und Trümmersprachen (pp. 173–191). Berlin: Walter de Gruyter.
  • Steinkrüger, P. O., and Krifka, M. (Eds.). (2009). On inflection. Berlin: Mouton de Gruyter.
  • Stiles, P. V. (2010). The Gothic extended forms of the dental preterit endings. NOWELE: North-West European Language Evolution, 58–59, 343–366.
  • Stiles, P. V. (2013). The pan-West Germanic isoglosses and the sub-relationships of West Germanic to other branches. NOWELE: North-West European Language Evolution, 66, 5–38.
  • Strässler, J. (Ed.). (1998). Tendenzen europäischer Linguistik: Akten des 31. Linguistischen Kolloquiums, Bern 1996. Linguistische Arbeiten 381. Tübingen, Germany: Max Niemeyer.
  • Streitberg, W. A. (1891). Perfective und imperfective Actionsart im Germanischen. Beiträge zur Geschichte der deutschen Sprache und Literatur, 15, 70–177.
  • Streitberg, W. A. (1920). Gotisches Elementarbuch (5th & 6th ed.). Heidelberg, Germany: Winter. The syntax portion (§234–) has been edited by Hugo Stopp as a separate book, Gotische Syntax (Heidelberg, Germany: Winter, 1981).
  • Streitberg, W. A. (2000). Die gotische Bibel (7th ed.). Revised by Piergiuseppe Scardigli. Heidelberg, Germany: Winter.
  • Strid, J. P. (2010). The origin of the Goths from a topolinguistic perspective: A short proposal. NOWELE: North-West European Language Evolution, 58–59, 443–452.
  • Strid, J. P. (2013). Retracing the Goths. In A. Kaliff & L. Munkhammar (Eds.), Wulfila 311–2011: International Symposium, Uppsala University June 15–18, 2011 (pp. 41–54). Uppsala, Sweden: Edita Västra Aros.
  • Sturtevant, A. M. (1931). Aih (aig), aigum (aihum). Germanic Review, 6, 57.
  • Stutz, E. (1966). Gotische Literaturdenkmäler. Stuttgart: Metzler. A rich resource with references on every topic.
  • Stutz, E. (1972). Das Neue Testament in gotischer Sprache. In K. Aland (Ed.), Die alten Übersetzungen des Neuen Testaments, die Kirchenväterzitate und Lektionare: Der gegenwärtige Stand ihrer Erforschung und ihre Bedeutung für die griechische Textgeschichte (pp. 375–402). Arbeiten zur Neutestamentlichen Textforschung 5. Berlin: Walter de Gruyter.
  • Suzuki, S. (1989). The morphosyntax of detransitive suffixes -þ- and -n- in Gothic: A synchronic and diachronic study. New York: Peter Lang.
  • Suzuki, S. (1992). Toward an explanatory account of Thurneysen’s law in Gothic. Beiträge zur Geschichte der deutschen Sprache und Literatur, 114, 28–46.
  • Suzuki, S. (1994). Final devoicing and elimination of the effects of Verner’s law in Gothic. Indogermanische Forschungen, 99, 217–251.
  • Suzuki, S. (1995). The decline of the foot as a supersyllabic mora-counting unit in early Germanic. Transactions of the Philological Society, 93, 227–272.
  • Swan, T., Mørck, E., & Westvik, O. J. (Eds.). (1994). Language change and language structure: Older Germanic languages in a comparative perspective. Berlin: Mouton de Gruyter.
  • Szemerényi, O. J. L. (1960). Gothic auhuma and the so-called comparatives in -uma. Beiträge zur Geschichte der deutschen Sprache und Literatur, 82, 1–30.
  • Szemerényi, O. J. L. (1972). A new leaf of the Gothic Bible. Language, 48, 1–10.
  • Thöny, L. (2013). Flexionsklassenübertritte: Zum morphologischen Wandel in der altgermanischen Substantivflexion. Innsbrucker Beiträge zur Sprachwissenschaft 146.Innsbruck, Austria: Institut für Sprachen und Literaturen der Universität Innsbrück.
  • Thráinsson, H., Epstein, S. D, & Peter, S. (Eds.). (1996). Studies in Comparative Germanic Syntax II. Dordrecht, The Netherlands: Kluwer.
  • Thurneysen, R. (1898). Spirantenwechsel im Gotischen. Indogermanische Forschungen, 8, 208–214.
  • Tollenaere, F. de, & Jones, R. L., in cooperation with F. van Coetsem, P. H. Smith Jr., & H. T Wong. (1976). Word-indices and word-lists to the Gothic Bible and minor fragments. Leiden, The Netherlands: Brill.
  • Trutmann, A. (1972). Studien zum Adjektiv im Gotischen. Berlin: Walter de Gruyter.
  • Van de Velde, F. (2009). De nominale constituent: Structuur en geschiedenis. Leuven, Belgium: Universitaire Pers Leuven.
  • Vasiliev, Alexander Alexandrovich. (1936). The Goths in the Crimea. Cambridge, MA: The Medieval Academy of America.
  • Velkov, V. (1989). Wulfila und die Gothi minores in Moesien. Klio, 71, 525–527.
  • Vennemann, T. (1971). The phonology of Gothic vowels. Language, 47, 90–132.
  • Vennemann, T. (1972). Phonetic detail in assimilation: Problems in Germanic phonology. Language, 48, 863–892.
  • Vennemann, T. (1985). Phonologically conditioned morphological change. In E. Gussmann (Ed.), Phono-morphology: Studies in the intersection of phonology and morphology (pp. 193–219). Lublin, Poland: Catholic University of Lublin Press.
  • Verner, Karl. (1875). Eine Ausnahme der ersten Lautverschiebung. KZ 23: 97–130 (1877).
  • Verner, K. (1877). Eine Ausnahme der ersten Lautverschiebung. Zeitschrift für vergleichende Sprachforschung auf dem Gebiete der indogermanischen Sprachen, 23, 97–130.
  • Viehmeyer, L. A. (1971). The Gothic alphabet: A study and derivation. Ph.D. dissertation, University of Illinois, Urbana-Champaign.
  • Vilutis, J. (1972). K voprosu ob upotreblenii substantivnogo ‘sa, þata, so’ v gotskom jazyke. Kalbotyra, 23, 155–162.
  • Vilutis, J. (1973). Die zusammengesetzten Demonstrativa im Gotischen. Kalbotyra, 25, 91–98.
  • Vilutis, J. (1976). Zum Problem des (bestimmten) Artikels im Gotischen. Kalbotyra, 27, 153–159. Continued in Kalbotyra, 28, 50–56 (1977) and 30, 43–49 (1979).
  • Vilutis, J. (1982). Zum Problem der syntaktischen Funktionen des adjektivischen sa, þata, so im Gotischen. Kalbotyra, 33, 103–107.
  • Vilutis, J. (1986). Zum Entstehungsproblem einiger unbestimmter Pronomina der gotischen Sprache. Kalbotyra, 36, 122–129.
  • Viteau, J. (1893). Étude sur le grec du nouveau testament: Le verbe; Syntaxe des propositions. Paris: Émile Bouillon.
  • Voyles, J. B. (1968). Gothic and Germanic. Language, 44, 720–746.
  • Voyles, J. B. (1981). Gothic, Germanic, and Northwest Germanic. Wiesbaden, Germany: Steiner.
  • Voyles, J. B., & Barrack, C. M. (2009). An introduction to Proto-Indo-European and the early Indo-European languages. Bloomington, IN: Slavica.
  • Wagner, N. (1986). Das ƕ im gotischen Alphabet. Zeitschrift für deutsches Altertum und deutsche Literatur, 115, 143–150.
  • Walkden, G. L. (2012). Syntactic reconstruction and Proto-Germanic. Ph.D. dissertation, University of Cambridge, Cambridge, U.K. Published in 2014 (Oxford: Oxford University Press).
  • Watts, S., West, J., & Solms, H.‑J. (Eds.). (2001). Zur Verbmorphologie germanischer Sprachen. Tübingen, Germany: Max Niemeyer.
  • Weisker, E. (1880). Über die Bedingungssätze im Gotischen. Jahresbericht . . . Höhere Bürgerschule, 6, 3–14.
  • Welo, Eirik, (ed.), (2011). Indo-European syntax and pragmatics: Contrastive approaches. Oslo Studies in Language 3/3.
  • Wolfram, H. (2005a). Die Goten und ihre Geschichte (2d ed.). Munich: Beck.
  • Wolfram, H. (2005b). Gotische Studien: Volk und Herrschaft im frühen Mittelalter. Munich: Beck.
  • Woodard, R. D. (Ed.). (2004). The Cambridge encyclopedia of the world’s ancient languages. Cambridge, U.K.: Cambridge University Press.
  • Woodhouse, R. (2000). The origin of Thurneysen’s law: A detailed analysis of the evidence. Beiträge zur Geschichte der deutschen Sprache und Literatur, 122, 187–229.
  • Wright, J. (1954). Grammar of the Gothic language: And the gospel of St. Mark, selections from the other gospels, and the second epistle to Timothy; With notes and glossary (2d ed.). With a supplement to the grammar by Olive Lenore Sayce. Oxford: Clarendon.
  • Zadorožnyj, B. M. (1960). Porivnjalna fonetyka i morfolohija hotskoj movy. Lviv, Ukraine: Vydavnystvo lvivskoho Universytetu.
  • Zucha, I. (1989). Zum schwachen Adjektiv in prädikativer Stellung. Indogermanische Forschungen, 94, 301–305.

Notes

  • 1. For documentation on the Goths and their history, see Heather & Matthews, 1991. See also Scardigli, 1973; Budanova, 1999; Wolfram, 2005a, 2005b; McLynn, 2007; Barnish & Marazzi, 2007; and the papers in Kaliff & Munkhammar, 2013. Many unknowns remain about the Goths, Wulfila, and the Bible translation (Ebbinghaus, 1992; Poulter, 2007). Useful aids are the bibliographies by Petersen (2005) and Ferreiro (2008, 2011, 2014); the references in Falluomini, 2013a, 2015; the glossary in Skeat, 1868; Lehmann, 1986; the word index in Tollenaere & Jones, 1976; Wörterbuch (Köbler, 1989), with every Gothic form plus English gloss; the concordance in Snædal, 2013a; Snædal’s Academia.edu profile; and the searchable Wulfila Project, with Snædal’s corrections to Streitberg’s text and valuable links. The PROIEL parallel parsed corpus of early New Testament translations, including Gothic, requires caution. Gothic is cited from Snædal, 2013a: Vol. 1.

    The classic grammar is von der Gabelentz & Löbe, 1846. Early historical treatments include Meyer, 1869 and Jellinek, 1926; more-recent treatments include Krahe & Seebold, 1967; Jasanoff, 2004; and Ringe, 2006. For derivation, see Schubert, 1968; Casaretto, 2004; Thöny, 2013 (which also contains inflection classes). For compounds, see Carr, 1939 and Karpov, 2005a, 2005b; for gradation, see Szemerényi, 1960. Unreferenced material can be found in all handbooks: Streitberg, 1920; Jellinek, 1926; Kieckers, 1960; Wright, 1954; Mossé, 1956; Guxman, 1958; Hempel, 1962; Krause, 1968; Braune-Ebbinghaus, 1981; Braune-Heidermanns, 2004; Piras, 2007; Rauch, 2011; Rousseau, 2012; Kotin, 2012; and Schuhmann, 2015. Useful textbooks include Bennett, 1980 and Lambdin, 2006.

  • 2. Another interpretation of Jordanes’s Gothiscandza is *Gutisk andja ‘Gothic end/coast’, possibly Gdańsk (CGG 29; Green, 1998: 166f.). This is based in part on the identification of the Wielbark culture (between the Oder and the Vistula) with the Goths (Heather, 2010: 104f.). There is no secure evidence for a Scandinavian origin of the Goths (Heather, 1996: 25–30), nor is there runic evidence (Nielsen, 2011), and linguistic parallels between Gothic and Old Norse are inconclusive (CGG 30; Nielsen, 1989: 80–103, 1995, 2002; cf. Scardigli, 2002: 555) in light of the Northwest Germanic innovations and derivational differences between Northwest Germanic and Gothic (Bahder, 1880; Kluge, 1926; Krahe-Meid, 1967; Schubert, 1968; Bammesberger, 1990; Casaretto, 2004; Kotin, 2012). The immigrating East Germanic tribes could have split, some continuing west to Poland and points north; others, south to the Black Sea. If something like this occurred, it is compatible with the topographic evidence of the Goths’ relation to the Gautoi (Procopius) in Scandinavia, the Swedish Östgötar (cf. Ostrogothae), Gutland/Gotland, etc. (Strid, 2010, 2013). Vol. 6 (2006) of Gotica Minora (ed. Christian T. Petersen) is devoted to historical writings on the name of the Goths.

  • 3. The main discussions are Schröder (1910), Stearns (1978, 1989), Grønvik (1983), and Ganina (2011), with a new photograph of the Busbecq manuscript, discussion of every word in the corpus, and early-21st-century archaeological finds. It is especially useful for words that Busbecq did not consider Germanic. Henceforth, Crimean Gothic and Wulfila’s Gothic are distinguished as ‘Crimean’ and ‘Gothic’, respectively. For historical writings on Crimean Gothic, see Vol. 4 (2005) of Gotica Minora (ed. Christian T. Petersen).

  • 4. The Byzantine text developed slowly (Kenyon, 1937: 199). It was only partially standardized by the time of the Gothic translation. These issues, their evolution, and the Greek Vorlage are discussed most extensively by Falluomini (2013a, 2013b, 2015). For editions of many Greek and Latin biblical manuscripts, see Falluomini, 2014: 286–287. Some comparisons of the different Greek versions can be found at the Greek New Testament.

  • 5. The convention of the diacritics is observed here as a heuristic for the beginner.

  • 6. Only forms of the masculine are attested. In other words, blind, blinda, and so on do not occur. The best-attested adjectives have gaps. For instance, alls ‘all’ lacks gen sg masc; allis is attested only as gen sg neut 6x and as adv/conj 27x (Snædal, 2009a: 159).

  • 7. The number should be 700 (and 609 attributive adjectives) by Ratkus’s revised count (p.c.) of 24 rather than 25 predicative weak adjectives (PWAs). Of these, some, like usfaírina ‘blameless’ and áinaha ‘only (begotten)’, may be weak only (Dvuxžilov, 1980: 112, 121). The latter, like usháista ‘financially impoverished (person)’, may be a noun (Snædal, 2013a). Two PWAs are bahuvrihis, which are weak elsewhere in Germanic (Zucha, 1989). Eleven of the 24 end in -ans, which may not be weak: weak acc pl –ans : strong acc pl -ans = weak nom pl –ans : x (→ strong nom pl -ans). This is especially cogent when an -ans form is surrounded by strong adjectives, for example ungaƕaírbai … unaírknans (2 Tim 3:2A) ‘disobedient, unthankful, corrupted’, and manuscript B has unambiguously strong unaírknái (Artūras Ratkus, p.c.). For Trutmann (1972: 50), more of these forms are substantives, and prefixed forms are treated like compounds.

  • 8. Gothic diverges from the Greek and Vetus Latina texts (Jülicher (1970); cited under Jülicher, 1963–1976: 86) where ‘lame’ is accusative, despite dative Gk. soì (v.l. acc se, not in the Byzantine main text) and Lat. tibi ‘to you’. The position of þus suggests a matrix dative unless it is cliticized to goþ, but usually object pronouns are cliticized to a verb (Ferraresi, 2005). Syntactically, haltamma is difficult to justify unless þus is infinitival subject.

  • 9. In the Gothic corpus available to him, Metlen (1932: 9–14) counted 2,067 PrPs. Of the 246 Greek participles in the Gospels that are not translated with a PrP in Gothic, 100 are rendered by a relative clause (ibid. 15), for example saei saíƕiþ (Mt 5:28 and other verses) ‘he who sees’, as is typical of the Latin versions (cf. Jülicher (1972); cited under Jülicher, 1963–1976: 24), for Greek ho blépōn ‘the seeing one’. Some are rendered by adjectives, others by miscellaneous clause types, and so on. With lexical gaps a Gothic PrP can render a Greek word, for example leprós ‘leper’ (Mt 8:2), translated manna þrūtsfill habands ‘a man having leprosy’ (ibid. 17). Þ‎rū‎tsfill is an endocentric compound ‘distended skin’ (Carr, 1939: 66; Karpov 2005b: 202).

  • 11. Despite the various descriptions of -nan verbs, they were detransitive (anticausative, nonagentive, unaccusative, some inchoative or stative) (Suzuki, 1989); cf. gaháiljan ‘heal’: gaháilnan ‘become healed’, gadáuþjan ‘kill’: gadáuþnan ‘die’, mikiljan ‘make great’: mikilnan ‘become great’, etc. All 61 Gothic -nan verbs are discussed by Suzuki (1989) and Guxman (1964: 71–94), some 20 of which have no attested transitive counterpart. See also Ferraresi 2005: 111–117 and Ringe, 2006: 176–178, 259–260

  • 12. The best-attested verb, qiþan ‘say’, has only 37% of its possible forms (Snædal, 2009a: 162–163). The preterite passive is periphrastic.

  • 14. The type gibans ist* could also translate a past tense; cf. unte galáubida ist weitwodei unsara du izwis (2Thess 1:10A) ‘because our testimony among you was believed’, in which galáubida ist translates a Greek aorist passive episteúthē; cf. crēditum est ‘was believed’ in most Latin versions (see Schröder, 1957–1958). For details of the Gothic verb, including the periphrastic formations, see Jarceva in Guxman et al., 1977: Vol. 2.

  • 25. A form kunnjai ‘(that) I might know’ is sometimes cited (from Streitberg), but it is a misreading for kunnjá‎u at Colossians 4:8 (some early editors got it right), itself a mistake based on a misinterpretation of the ambiguous Greek manuscript form ΓΝΩ‎ as 1sg γνῶ‎ rather than 3sg γνῷ‎ (Snæ‎dal, 2006, w. lit.).

  • 15. Context dependency does not include coordinated structures on the evidence of the syntactic minimal pair in John 18:1: Iesus usiddja miþ siponjam seinam . . . aúrtigards in þanei galáiþ Iesus jah siponjos is ‘Jesus went out with his disciples . . . a garden into which Jesus went, and (also) his disciples’. Greek has a pronominal genitive autoũ ‘his’ in both places, but the Vulgate and Vetus Latina manuscripts with the same structure (cf. Jülicher (1963); cited under Jülicher, 1963–1976: 188) have the same distribution as the Gothic, possibly independently for a similar syntactic reason.

  • 16. The concept of ‘typically self-directed’ and ‘typically other-directed’ predicates is elaborated for the Germanic languages by Gast (2006). Unfortunately, his four Gothic examples illustrate only silba ‘self’.

  • 17. Of the studies of Gothic preverbs and prepositions (Mourek, 1890; Rice, 1932; Sizova, 1978; Lloyd, 1979; Krause, 1987, 1995), Bucsko, 2011 confirms considerable overlap between the two but little overlap with adverbs (p. 47ff.).

  • 18. The Verner’s law reflex (ga- < PIE *ko(m)-) is due to its pretonic position (Bennett, 1970; Schaffner, 2001: 60, 440). Aspectual ga- originates from the uniting of partial actions into a complete action (Behaghel, 1923–1928: 100; Lloyd, 1979: 110). Makovskij (2011) follows the Slavic tradition (cf. Lloyd, 1979: 7f.) in denying the perfectivizing use of Gothic prefixes because aspect is not consistently indicated. Also, aspect and Aktionsart have been confused (Mirowicz, 1935). Maslov (1959/2004) argues that ga- was telic, not perfective. Kotin (2012: 294f., 312f., 397f., 492) claims it could be both. Still, aspectual prefixes were not grammaticalized in Gothic, although Lloyd’s ‘complexive’ (prefixed forms, esp. with ga-) is restricted to completed actions (Lloyd 1979: 141–143). The aspect of Greek imperatives is almost never captured (Cuendet, 1924), and Guxman (1940: 121) points out that the single Gothic infinitive (with or without ga-) translates all of the aspectually different Greek infinitives.Nevertheless, Leiss (2012) demonstrates parallels in the selection of infinitives with or without ga- and Slavic perfective or imperfective infinitives after affirmative and negated modal verbs.

  • 19. Discussion of Gothic relative clauses and complementizers can be found in Longobardi, 1978; Ferraresi, 1991: 30–35; 2005; Miller, 1975, 2010: Vol. 2, 235ff.; Afros, 2006, 2010; and Harbert, 1992, 2012.

  • 20. Verbs that take complement clauses with þammei include fraþjan ‘realize’ (Mk 7:18), gáumjan ‘see’ (Lk 17:15, Mk 16:4, Jn 6:5), gatrauan ‘trust’ (Phil 2:24B; 2 Tim 1:12A/B), faginon ‘rejoice’ (Lk 15:6). Verbs of this class can also take þatei: gatrauan (2 Cor 2:3A/B, Rom 8:38A), fraþjan (Jn 8:27), gáumjan (Sk 7.4.5f.). It is impossible to prove that the meaning of þammei is always different from þatei, as Afros (2006) claims. Also, in all cases, the Greek has hóti ‘that’, and the Vetus Latina has quod, quia, quoniam ‘id.’, suggesting that this was the translation target.

  • 21. In most cases the Greek order is not kept (Jellinek, 1926: 50; Fourquet, 1938: 248), for example ga-þ-þan-miþ-sandidedum imma broþar (2 Cor 8:18B) ‘and (þ = uh) we then (þan) sent with (miþ) him (imma) brother,” very different from the Gk. sun-epémpsamen dè metautoũ tòn adelphón, literally ‘with-sent.1pl and with him the brother’. For the syntax of clitics in Gothic, see Eythórsson, 1995: 103–142. Adjunction of clitics and prefixes alike suggests that the prefixes were independent particles by origin (Goetting, 2007: 313, w. lit.).

  • 22. Gothic attests only three third-person imperatives: atsteigadau ‘let him climb down’ (Mt 27:42, Mk 15:32), lausjadau ‘let him free/rescue’ (Mt 27:43), liugandau ‘let them marry’ (1 Cor 7:9A) (cf. GE 206).

    Qimái is not aspectually equivalent to Greek elthétō (Davis, 2002), but neither is the Vulgate present subjunctive adveniat ‘let (it) come’ or the Vetus Latina (ad)veniat ‘id.’ (cf. Jülicher (1972); cited under Jülicher, 1963#x2013;1976: 31). On the overall irrelevance of Greek aspect to that in Gothic, see Lloyd, 1979: 143ff., in which it is argued that style and interpretive theological factors play a role.

  • 23. Most scholars recognize articles in Gothic (e.g., Sauvageot, 1929; Vilutis, 1976; Kovari, 1984; Kotin, 2012: 23, 211–224). Meillet (1949: 191) denied that Gothic had an article, and Van de Velde (2009) claims that Germanic had no determiners before Old Dutch. This ignores the many features an article can have.

  • 24. For additional discussion, see Meillet, 1949; Harbert, 2007: 6f.; Askedal, 2009; and Shimomiya, 2009.