The Goths began their famous foray into the history books in the third century AD, when they launched the first attack of what would become in the eyes of contemporaries a barbarian ravaging and pillaging machine. Their first entry into the literary records comes with their attacks on the Roman empire. At the time, the most exposed province was Transylvanian Dacia; but this region was protected by the Carpathian mountains, hence attacks generally came from south of the range across the lower reaches of the Danube or through a stretch of land connecting the so-called Dacian Salient to the Roman-occupied regions of the Balkans. The first known attack was the sack of Histria at the mouth of the Danube in 238. This was followed some 10 years later by other land attacks: Marcianople, a city inland from the westernmost extent of the Black Sea, was sacked in 249 by a Gothic group led by Argaith and Guntheric; in 250 Cniva crossed the Danube at Oescus and captured Philippopolis, wintered his army, and then in 251 defeated Roman forces and killed the emperor Decius at Abrittus.
The Goths then changed tactics and led sea-borne attacks via the Black Sea, the first series coming sometime around 255-257. The first attack fell on Pityus, on the eastern coast of the Black Sea, but was unsuccessful. In the next attack a year later, the 'Boranoi' -- a group possibly including the Goths -- ravaged Pityus, Trapezus, and a large part of the Pontus, the region of Asia Minor bordering the Black Sea in the southeasternmost reaches. A year later the Goths explicitly attacked the cities Apamea, Chalcedon, Nicaea, Nicomedia, Prusa, as well as the surrounding areas of Bithynia and Propontis, lying near the southwesternmost reaches of the Black Sea.
The next attacks came another 10 years later. In 268 the Goths and other tribes formed a large fleet which headed across the Black Sea to the south, unsuccessfully attacking Tomi and Marcianople, and subsequently (also unsuccessfully) Cyzicus and Byzantium. They pushed through the Dardanelles into the Aegean and dispersed in three forces: (1) composed of Heruli attacking the northern Balkans near Thessalonica -- subsequently defeated by Emperor Gallienus in 268; (2) composed of Goths and Heruli, attacking Attica -- after pushing north over land they were defeated by Claudius at Naissus in 270; (3) probably led by the Gothic chieftains Respa, Thuruar, and Veduc, and attacking Asia Minor, then Rhodes and Cyprus, then Side and Ilium and Ephesus, destroying the temple of Diana. This third group was pushed back to the Black Sea in 269. After this foray into the Mediterranean, there were no other attacks through the Dardanelles.
The Goths returned to land battles in an attack across the Danube in 270 against Anchialus and Nicopolis. This was followed by a Roman attack across Danube in 271, defeating the Gothic king Cannabaudes. The Goths attacked the Pontus again in 276-277, pushing farther inland to Galatia and Cilicia.
The impact of these waves of Gothic attacks is clearly felt in the Canonical Letter of Gregory Thaumaturgus, bishop of Neocaesarea (modern Niksar) in the Pontus region of Asia Minor. His episcopate fell during the emperor Decius's persecution of Christians, which started c. 250 AD. He may have lived until the reign of Aurelian (270-275 AD). The letter is a response to a neighboring bishop's questions concerning the conduct of Roman soldiers during and after the most recent period of Gothic raids in the Pontus (translated in Heather and Matthews, 1991):
| Canon 5 | Others delude themselves by keeping the property of others which they have found, in place of their own which they have lost, in order that, since the Boranoi and Goths worked on them deeds of war, so they may become Boranoi and Goths to others. I have therefore sent my brother... [that he may] advise you whose accusations you should accept.... | |
|---|---|---|
| Canon 6 | Concerning those who forcibly detain captives (who have escaped) from barbarians.... Send men out into the countryside, lest divine thunderbolts descend upon those who perpetrate such wickedness! | |
| Canon 7 | ... As for those who have been enrolled among the barbarians and followed after them as prisoners, forgetting that they were men of Pontus, and Christians, and have become so thoroughly barbarised as even to put to death men of their own race by the gibbet or noose, and to point out roads and houses to the barbarians, who were ignorant of them; you must debar them even from the ranks of Hearers, until a common decision is reached about them by the assembly of saints, with the guidance of the Holy Spirit. | |
| Canon 8 | ... As for those who brought themselves to attack the houses of others, if they are convicted after accusation, let them not be fit even to be Hearers. If however they confess their own guilt and make restitution, they are to prostrate themselves among the ranks of the penitent. |
It thus seems clear that the raiding parties, though in origin homogeneous groups of Goths or other individual tribes, were soon augmented by other enterprizing -- or coerced -- individuals within the lands they attacked. The attacks themselves thus only left a transitory material calamity. More long-term problems were felt when the raids had ceased, and there arose the issues of readmitting offenders and presumed offenders into the societies to which they formerly belonged.
The following passage is Luke 4.1-13, in which Jesus is tempted by Satan. In the first verse, we find the collocation gawandida sik 'returned', encountered in several places in other readings. This phrase is more literally 'brought himself back'. Though Gothic has a functioning morphological mediopassive which might, on Indo-European historical grounds, serve to connote exactly the meaning reached here, Gothic is already very close to other Germanic languages in the frequent use of such periphrastic reflexive collocations to render transitive verbs intransitive.
In Luke 4.3 we find a nominative form suna'us instead of the proper nominative sunus 'son'; a similar form diabula'us replaces the proper nominative diabulus in Luke 4.5. It is likely that, shortly after the time of Wulfila (if not before), the diphthong a'u was monophthongized and subsequently indistinguishable from u, leading to occasional scribal confusion. Similar phonetically derived scribal confusion occurs in Luke 4.4: we find the forms hla'ib and liba'id, which do not show the expected devoicing of final -b and -d to -f and -th, respectively.
Note also the construction in Luke 4.3: qith thamma sta'ina ei wai'rtha'i hla'ibs 'command this stone that it be made bread'. This is an example of the particle ei used to introduce an indirect command. The particle ei was originally a general relative marker, setting off an entire phrase as dependent on surrounding material. Only later did this particle become attached to demonstrative pronouns to form relative pronouns, such as saei. Some of these subsequently became subordinate conjunctions, e.g. thatei. Gothic may also use prepositional constructions for subordinate clauses, such as Luke 4.10: du gafastan '(in order) to support' shows the frequent use in Gothic of the preposition du with the infinitive in purpose constructions.
The phrase in Luke 4.7, in andwai'rthja meinamma 'in my presence', translates the Greek eno:pion emou, which is not explicit in the King James Version. The English phrase 'Get thee behind me, Satan' in Luke 4.8 does not appear in the Gothic, an indication that the Gothic translation was made from a different manuscript than the King James Version.
4:1 - Ith Iesus, ahmins weihis fulls, gawandida sik fram Iaurdanau jah tauhans was in ahmin in authidai
2 - dage fidwor tiguns, fraisans fram diabulau. jah ni matida waiht in dagam jainaim, jah at ustauhanaim thaim dagam, bithe gredags warth.
3 - jah qath du imma diabulus: jabai sunaus sijais gudis, qith thamma staina ei wairthai hlaibs.
4 - jah andhof Iesus withra ina qithands: gamelid ist thatei ni bi hlaib ainana libaid manna, ak bi all waurde gudis.
5 - jah ustiuhands ina diabulaus ana fairguni hauhata, ataugida imma allans thiudinassuns this midjungardis in stika melis.
6 - jah qath du imma sa diabulus: thus giba thata waldufni thize allata jah wulthu ize, unte mis atgiban ist, jah thishwammeh thei wiljau, giba thata.
7 - thu nu jabai inweitis mik in andwairthja meinamma, wairthith thein all.
8 - jah andhafjands imma Iesus qath: gamelid ist, fraujan guth theinana inweitais jah imma ainamma fullafahjais.
9 - thathroh gatauh ina in Iairusalem jah gasatida ina ana giblin alhs jah qath du imma: jabai sunus sijais gudis, wairp thuk thathro dalath;
10 - gamelid ist auk thatei aggilum seinaim anabiudith bi thuk du gafastan thuk,
11 - jah thatei ana handum thuk ufhaband, ei hwan ni gastagqjais bi staina fotu theinana.
12 - jah andhafjands qath imma Iesus thatei qithan ist: ni fraisais fraujan guth theinana.
13 - jah ustiuhands all fraistobnjo diabulus, afstoth fairra imma und mel.
4:1 Ith Iesus, ahmins weihis fulls, gawandida sik fram Iaurdanau jah tauhans was in ahmin in authidai 2 dage fidwor tiguns, fraisans fram diabulau. jah ni matida waiht in dagam jainaim, jah at ustauhanaim thaim dagam, bithe gredags warth. 3 jah qath du imma diabulus: jabai sunaus sijais gudis, qith thamma staina ei wairthai hlaibs. 4 jah andhof Iesus withra ina qithands: gamelid ist thatei ni bi hlaib ainana libaid manna, ak bi all waurde gudis.
5 jah ustiuhands ina diabulaus ana fairguni hauhata, ataugida imma allans thiudinassuns this midjungardis in stika melis. 6 jah qath du imma sa diabulus: thus giba thata waldufni thize allata jah wulthu ize, unte mis atgiban ist, jah thishwammeh thei wiljau, giba thata. 7 thu nu jabai inweitis mik in andwairthja meinamma, wairthith thein all. 8 jah andhafjands imma Iesus qath: gamelid ist, fraujan guth theinana inweitais jah imma ainamma fullafahjais.
9 thathroh gatauh ina in Iairusalem jah gasatida ina ana giblin alhs jah qath du imma: jabai sunus sijais gudis, wairp thuk thathro dalath; 10 gamelid ist auk thatei aggilum seinaim anabiudith bi thuk du gafastan thuk, 11 jah thatei ana handum thuk ufhaband, ei hwan ni gastagqjais bi staina fotu theinana. 12 jah andhafjands qath imma Iesus thatei qithan ist: ni fraisais fraujan guth theinana.
13 jah ustiuhands all fraistobnjo diabulus, afstoth fairra imma und mel.
From the King James version:
4:1 And Jesus being full of the Holy Ghost returned from Jordan, and was led by the Spirit into the wilderness, 2 Being forty days tempted of the devil. And in those days he did eat nothing: and when they were ended, he afterward hungered. 3 And the devil said unto him, If thou be the Son of God, command this stone that it be made bread. 4 And Jesus answered him, saying, It is written, That man shall not live by bread alone, but by every word of God.
5 And the devil, taking him up into an high mountain, shewed unto him all the kingdoms of the world in a moment of time. 6 And the devil said unto him, All this power will I give thee, and the glory of them: for that is delivered unto me; and to whomsoever I will I give it. 7 If thou therefore wilt worship me, all shall be thine. 8 And Jesus answered and said unto him, Get thee behind me, Satan: for it is written, Thou shalt worship the Lord thy God, and him only shalt thou serve.
9 And he brought him to Jerusalem, and set him on a pinnacle of the temple, and said unto him, If thou be the Son of God, cast thyself down from hence: 10 For it is written, He shall give his angels charge over thee, to keep thee: 11 And in their hands they shall bear thee up, lest at any time thou dash thy foot against a stone. 12 And Jesus answering said unto him, It is said, Thou shalt not tempt the Lord thy God.
13 And when the devil had ended all the temptation, he departed from him for a season.
The weak declension of nouns is nothing more than n-stem nominal formation. In principle it is no different from the formation of r-stems or nd-stems: the affix -Vn- intercedes between nominal root and endings, where V is some vowel. Just as bro:thar 'brother' is inherently an r-stem noun (employing no other formations), likewise weak nouns are inherently weak, i.e. only n-stem. While adjectives may employ strong or weak declension (see Sections 13 and 17) according to contextual demands, a given noun by contrast employs only one declension at all times: a weak noun always declines weak, a strong noun always strong.
The an/jan/wan-stem nouns are generally masculine or neuter. Nouns with the j- or w-augment take the same endings as those without. The nouns atta 'father', arbja 'heir', and gawau'rstwa 'fellow-worker' -- respectively an-, jan-, and wan-stems -- serve to illustrate the masculine forms. The nouns hai'rto: 'heart', a'ugo: 'eye', and sigljo: 'seal' -- respectively an-, an-, and jan-stems -- serve to illustrate the neuter forms.
| Weak an-Stem | Masculine | Neuter | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Stem | attan- | arbjan- | gawau'rstwan- | hai'rtan- | a'ugan- | sigljan- | ||||||
| N Sg. | atta | arbja | gawau'rstwa | hai'rto: | a'ugo: | sigljo: | ||||||
| A | attan | arbjan | gawau'rstwan | hai'rto: | a'ugo: | sigljo: | ||||||
| G | attins | arbjins | gawau'rstwins | hai'rtins | a'ugins | *sigljins | ||||||
| D | attin | arbjin | gawau'rstwin | hai'rtin | a'ugin | *sigljin | ||||||
| N Pl. | attans | arbjans | gawau'rstwans | hai'rto:na | a'ugo:na | *sigljo:na | ||||||
| A | attans | arbjans | gawau'rstwans | hai'rto:na | a'ugo:na | *sigljo:na | ||||||
| G | attane: | arbjane: | gawau'rstwane: | hai'rtane: | a'ugane: | *sigljane: | ||||||
| D | attam | arbjam | gawau'rstwam | hai'rtam | a'ugam | *sigljam | ||||||
Note that the dative plural is built by analogy to a-stem nouns, without the intervening Vn-affix. The nouns aba (masc.) 'man, husband'; au'hsa (masc.) 'ox'; namo: (neut.) 'name'; and wato: (neut.) 'water' have different forms in the plural. These are boldfaced in the chart below. The masculine noun manna 'man' generalized the zero-grade of the n-stem formation (-n- rather than -Vn-), yielding a peculiar declension.
| Weak an-Stem | Masculine | Neuter | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Stem | ab(a)n- | au'hs(a)n- | man(a)n- | nam(a)n- | wat(a)n- | |||||
| N Sg. | aba | au'hsa | manna | namo: | wato: | |||||
| A | aban | au'hsan | mannan | namo: | wato: | |||||
| G | abins | au'hsins | mans | namins | watins | |||||
| D | abin | au'hsin | mann | namin | watin | |||||
| N Pl. | abans | au'hsans | mans, mannans | namna | wato:na | |||||
| A | abans | au'hsans | mans, mannans | namna | wato:na | |||||
| G | abne: | au'hsne: | manne: | namne: | watane: | |||||
| D | abnam | au'hsam | mannam | namnam | watnam | |||||
For the zero-grade forms of manna, compare Latin declension: nom. sg. caro: 'flesh' with acc. carnem, as opposed to nom. sg. homo: 'man' with acc. hominem.
The o:n/jo:n/wo:n-stem nouns are generally feminine. Nouns with the j- or w-augment take the same endings as those without. The nouns tuggo: 'tongue', arbjo: 'heiress', and u:htwo: 'early morning' -- respectively o:n-, jo:n-, and wo:n-stems -- serve to illustrate the forms.
| Weak o:n-Stem | Feminine | |||||
|---|---|---|---|---|---|---|
| Stem | tuggo:n- | arbjo:n- | u:htwo:n- | |||
| N Sg. | tuggo: | arbjo: | u:htwo: | |||
| A | tuggo:n | arbjo:n | u:htwo:n | |||
| G | tuggo:ns | arbjo:ns | u:htwo:ns | |||
| D | tuggo:n | arbjo:n | u:htwo:n | |||
| N Pl. | tuggo:ns | arbjo:ns | u:htwo:ns | |||
| A | tuggo:ns | arbjo:ns | u:htwo:ns | |||
| G | tuggo:no: | arbjo:no: | u:htwo:no: | |||
| D | tuggo:m | arbjo:m | u:htwo:m | |||
The vowel o: has generalized throughout the declension, so that the dative, though characteristically lacking any sign of the n-stem, still maintains the vowel.
The i:n-stem nouns (recall [i:] is spelled ei in Gothic) are generally feminine. These nouns derive for the most part from adjectives, forming the associated abstract noun. The nouns managei 'multitude', a'ithei 'mother', and fro:dei 'understanding' serve to illustrate the forms.
| Weak i:n-Stem | Feminine | |||||
|---|---|---|---|---|---|---|
| Stem | managein- | a'ithein- | fro:dein- | |||
| N Sg. | managei | a'ithei | fro:dei | |||
| A | managein | a'ithein | fro:dein | |||
| G | manageins | a'itheins | fro:deins | |||
| D | managein | a'ithein | fro:dein | |||
| N Pl. | manageins | a'itheins | fro:deins | |||
| A | manageins | a'itheins | fro:deins | |||
| G | manageino: | a'itheino: | fro:deino: | |||
| D | manageim | a'itheim | fro:deim | |||
The formation is similar to the declension of o:n-stems. The vowel i: has generalized throughout the declension, so that the dative maintains the vowel while losing the n.
Adjectives decline according to weak or strong paradigms based on the requirements of context (cf. Section 13 on strong adjective declension). The choice is one of specificity: Gothic employs weak adjective forms to modify a definite noun, and strong forms to modify an indefinite noun. For example, ahma sa weiha 'the holy ghost' and tha'i ana ai'rtha'i thiza'i go:do:n saianans 'they that are sown on the good ground' (Mark 4.20). For comparison, weihs ahma would be 'a holy spirit', ana go:da'i ai'rtha'i 'on (some) good ground', saiana'i 'sown'. Weak endings are generally used for nominalized adjectives: unkarjans 'careless (ones)'. Compare weihs 'holy' (strong) to weiha 'priest' (weak), literally '(the) holy one'. From a morphological point of view, weak adjective endings are simply n-stem endings. In this way, their use for definite reference parallels proper names in Latin, e.g. catus 'sly' vs. Cato: (G. Cato:nis) 'the Sly One', and in Greek, e.g. platu's 'broad' vs. Pla'to:n 'the Broad (Shouldered) One'.
As mentioned above, the weak declension of adjectives is actually simply n-stem declension. The weak masculine endings of a-stem adjectives exactly parallel those of the an-stem noun atta, and the neuter a-stem weak endings those of the an-stem hai'rto: (Section 16.1); the feminine weak endings of a-stem adjectives parallel those of the o:n-stem noun tuggo: (Section 16.2). Again the adjective blinds 'blind' illustrates the declension.
| Weak a-Stem | Masculine | Neuter | Feminine | |||
|---|---|---|---|---|---|---|
| N Sg. | blinda | blindo: | blindo: | |||
| A | blindan | blindo: | blindo:n | |||
| G | blindins | blindins | blindo:ns | |||
| D | blindin | blindin | blindo:n | |||
| N Pl. | blindans | blindo:na | blindo:ns | |||
| A | blindans | blindo:na | blindo:ns | |||
| G | blindane: | blindane: | blindo:no: | |||
| D | blindam | blindam | blindo:m | |||
The ja-stem adjectives decline analogously, the masculine forms following the jan-stem noun arbja, the neuter following sigljo:, the feminine following arbjo:. The distinctions of Group (1) and Group (2) ja-stems do not play a role in weak declension. Likewise, i-stem and u-stem adjectives follow the same weak declension as ja-stems, exhibiting the same j-augment in all forms. The adjectives niujis 'new' and wiltheis 'wild' illustrate the weak declension of ja-stems; hra'ins 'clean' illustrates the i-stems; hardus 'hard' illustrates the u-stems. The forms are as follows.
| Masculine | ja-Stem (1) | ja-Stem (2) | i-Stem | u-Stem | ||||
|---|---|---|---|---|---|---|---|---|
| N Sg. | niuja | wilthja | hra'inja | hardja | ||||
| A | niujan | wilthjan | hra'injan | hardjan | ||||
| G | niujins | wilthjins | hra'injins | hardjins | ||||
| D | niujin | wilthjin | hra'injin | hardjin | ||||
| N Pl. | niujans | wilthjans | hra'injans | hardjans | ||||
| A | niujans | wilthjans | hra'injans | hardjans | ||||
| G | niujane: | wilthjane: | hra'injane: | hardjane: | ||||
| D | niujam | wilthjam | hra'injam | hardjam | ||||
| Neuter | ||||||||
| N Sg. | niujo: | wilthjo: | hra'injo: | hardjo: | ||||
| A | niujo: | wilthjo: | hra'injo: | hardjo: | ||||
| G | niujins | wilthjins | hra'injins | hardjins | ||||
| D | niujin | wilthjin | hra'injin | hardjin | ||||
| N Pl. | niujo:na | wilthjo:na | hra'injo:na | hardjo:na | ||||
| A | niujo:na | wilthjo:na | hra'injo:na | hardjo:na | ||||
| G | niujane: | wilthjane: | hra'injane: | hardjane: | ||||
| D | niujam | wilthjam | hra'injam | hardjam | ||||
| Feminine | ||||||||
| N Sg. | niujo: | wilthjo: | hra'injo: | hardjo: | ||||
| A | niujo:n | wilthjo:n | hra'injo:n | hardjo:n | ||||
| G | niujo:ns | wilthjo:ns | hra'injo:ns | hardjo:ns | ||||
| D | niujo:n | wilthjo:n | hra'injo:n | hardjo:n | ||||
| N Pl. | niujo:ns | wilthjo:ns | hra'injo:ns | hardjo:ns | ||||
| A | niujo:ns | wilthjo:ns | hra'injo:ns | hardjo:ns | ||||
| G | niujo:no: | wilthjo:no: | hra'injo:no: | hardjo:no: | ||||
| D | niujo:m | wilthjo:m | hra'injo:m | hardjo:m | ||||
The wa-stem adjectives maintain the w-augment. Few weak forms are extant. The adjective triggws 'true' exhibits only the weak N sg. triggwa and D sg. triggwin. The wa-stem adjectives lasiws 'weak', *qius 'alive', *fa'us 'little', *usska'us 'vigilant' show no weak forms.
The comparative of adjectives is formed by addition of the suffix -iz- or -o:z-. The superlative is derived by addition of the suffix -st- or -o:st-. Consider the following examples.
| Meaning | Positive | Comparative | Superlative | |||
|---|---|---|---|---|---|---|
| 'much, many' | manags | managiza | managists | |||
| 'wise' | fro:ths | fro:do:za | ||||
| 'poor' | arms | armo:sts | ||||
| 'strong' | swinths | swintho:za | ||||
| 'old' | altheis | althiza | ||||
| 'sweet' | suts | sutiza | sutists | |||
| 'hard' | hardus | hardiza | ||||
| 'high' | *ha'uhs | ha'uhists | ||||
As with the positive degree, the superlative degree declines either strong or weak according to the requirements of context. The superlative, however, does not take the alternate pronominal ending -ata in the neuter singular nominative or accusative. The comparative, by contrast, only assumes weak adjectival endings, regardless of context. The only difference between comparative endings and general weak adjectival endings occurs in the feminine: the feminine declines like the i:n-stem managei, rather than tuggo:. The adjective ju:hiza 'younger', from juggs 'young', serves to illustrate comparative declension.
| Comparative | Masculine | Neuter | Feminine | |||
|---|---|---|---|---|---|---|
| N Sg. | ju:hiza | ju:hizo: | ju:hizei | |||
| A | ju:hizan | ju:hizo: | ju:hizein | |||
| G | ju:hizins | ju:hizins | ju:hizeins | |||
| D | ju:hizin | ju:hizin | ju:hizein | |||
| N Pl. | ju:hizans | ju:hizo:na | ju:hizeins | |||