Definite archaeological remains of the early Goths are even more elusive than clear literary references. Part of the problem concerns dating: the are many sites which offer possible remains of Gothic culture, but the relevant timeframes are difficult to establish. In general closed archaeological finds (e.g. burial finds) with Roman coins and pottery provide the best sources for dating, but the time lag between production and placement leaves a certain amount of uncertainty. It is therefore easier to establish relative chronology by looking at the development of certain specific types of objects, e.g. brooches, buckles, pots, combs. The simpler forms are considered earlier, the more complex later. Such dating, however, does not apply well to individual objects, rather only to groups of objects. The more objects in a closed area, the more secure the relative chronology.
This still leaves open a crucial issue, namely that a material culture is not the same as an ethnic culture, or even political, social, or linguistic culture. Material items, and technology in general, have the ability to move across socio-ethnic boundaries much quicker than linguistic or ethnic traits. When however archaeologists discover not only continuity of material items, but also of ritual practice, such as burial rites, then this strengthens the argument that the material in question is associated with a somewhat homogenous culture. Keeping these caveats in mind, then, we may discuss the two cultures on which archaeologists have focused in their attempts to find physical traces of the Goths.
The Wielbark culture is named after an area in the north of present-day Poland in which many characteristic remains were discovered. This material culture formed in the middle of 1st century AD in Pomerania on both sides of the Vistula, which is roughly the area in which Tacitus, Ptolemy, and Strabo place the Goths, and at roughly the same time. An early phase of development lasted for roughly a century. Then followed a second phase which spread over a wider area, first encompassing the northernmost regions of Poland and Mazovia east of the Vistula (c. 160-210), and then extending farther south along the Vistula, San, and Bug rivers into Byelorussia, Volhynia, and northern Ukraine (c. 180-230). This area is where the Chernjakhov culture, discussed below, later developed.
The Wielbark culture is characterized by stone circles found in cemeteries. In trying to align the movement of this culture with the picture presented in the literary record, especially as pertains to the purported Scandinavian origin of the Goths, it is important to find traits common to material cultures both in Scandinavia and on the southern shore of the Baltic, and furthermore among which the Scandinavian finds are the clear antecedents. These stone circles in cemeteries are the only practice found to be earlier in Scandinavia than on the European mainland. They do not, however, appear in the earliest Wielbark cemeteries.
Another trait of the culture is that inhumation and cremation graves are found side by side in Wielbark cemeteries. Surrounding burial sites do not display this two-fold practice, but instead display only cremation burials. A rather peculiar trait is that the members of the Wielbark culture did not bury iron objects, most importantly weapons, with any male dead. By contrast, the surrounding sites, as well as earlier inhabitants of the same areas, did bury iron weapons with the dead. Thus, if this culture does in fact represent the archaeological remains of the Goths, they appear to have broken with the typical Germanic tradition of burying the dead with their weapons. In addition, women's dress, at least in burials, was characterized by a double brooch, one on each shoulder.
The period of the second phase of the Wielbark culture coincides with the Marcomannic wars, c. 150 AD, which caused dramatic changes in the material cultures of present-day Poland. This phase of the Wielbark culture spread into the area of the Przeworsk culture, the area south of Pomerania, between the Notec and Warta rivers, and to Masovia in the southeast. The period 180-300 AD finds not only the expansion of Wielbark culture, but the incorporation of Wielbark traits into other cultures, particularly the Chernjakhov.
The Chernjakhov culture began in the middle of the 3rd century. It reached its fullest extent in the 4th century, covering a large area between the Danube and Don, to the north and west of the Black Sea, and to the south and east of the Carpathian mountains. The temporal overlap with the Wielbark culture is also physically accompanied by shared features of material culture, particularly handmade pottery, some types of brooch, and the style of women's dress.
More intriguing, perhaps, are the shared cultural practices. As with the Wielbark culture, the Chernjakhov culture left behind cemeteries with mixed inhumation and cremation burials. The majority of Chernjakhov inhumation graves are placed along a north-south line, with the head of the deceased to the north. Also like the Wielbark culture, the Chernjakhov culture did not bury weapons with their male dead. There are however a few cemeteries in Cozia-Iasi, Todireni, and Braniste where the dead were buried with weapons. These weapons may have originated outside the Wielbark culture, though, perhaps in Przeworsk; the other equipment is consonant with the idea of Germanic intruders from the north. They also buried some wheel-made pottery with the dead, as well as some bone combs and iron implements.
The houses of the Chernjakhov culture are of two types. The most numerous are sunken huts. These are usually rectangular, though there are some with a more oval shape. These huts are cut into the ground, some so deep that only roofing would need to be added. The earthen floors are generally between 5 and 16 square meters. The walls were wattle and daub, and each house had a hearth. The other type of houses were surface dwellings. These are often found in the same settlements as the sunken huts. The smaller ones are usually between 6 and 8 square meters, the larger between 11 and 16 square meters. They were divided in two parts, one providing quarters for people, the other for animals.
If the Chernjakhov culture is in fact the continuation of the Wielbark culture, then the shared burial practices, as well as the shared mode of women's dress and the style of other implements, show not only a carryover of material culture, but also a carryover of social customs and beliefs. The Wielbark culture did not nevertheless cease to exist when the Chernjakhov culture began. The latter is therefore unlikely to be the result of a near total migration of the former.
The following passage, Luke 2:41-52, gives an account of an incident in Jesus's boyhood, the only boyhood incident reported in the New Testament. Luke says that Jesus was twelve years old: the Gothic translates twalibwintrus, literally 'twelve winters' (Luke 2.42). Ancient Germanic cultures often used 'winter' as an equivalent of 'year' when reckoning spans of time. Similar constructions occur in both the Old English Beowulf and the Old Saxon Heliand: xii wintra tīd 'a span of twelve winters' (B.147); gebad wintra worn 'he endured countless winters' (B.264); Huand wit habdun aldres ēr efno tuēntig uuintro an uncro uueroldi, ēr than quāmi that uuīb ti mi 'the two of us had an age of about twenty winters in our world when that woman came to me' (H.144-145). Even the Old English translation of this biblical passage has and ða he wæs twelf wintre, hy foron to Hierusalem 'and when he was twelve winters (old), they went to Jerusalem'.
This phrase is followed by an example of the Gothic dative absolute: jah biþē warþ twalibwintrus, usgaggandam þan im in Iaírusaúlwma bi biūhtja dulþáis jah ustiuhandam þans dagans, literally 'when he became twelve-years-old, (with) them then going out to Jerusalem according to the custom of the feast, and (with them) fulfilling the days...' (Luke 2.42-43). These absolute constructions constitute a substantive and associated participle to give what would be in English a subordinate clause.
Luke 2.48 provides a notable instance in which Gothic employs the plural where one might expect the dual: ƕa gatawides uns swa? sai, sa atta þeins jah ik winnandona sokidedum þuk 'why hast thou thus dealt with us? behold, thy father and I have sought thee sorrowing.' Here uns and sokidedum clearly refer to Mary and Joseph, and so the reader might expect the dual ugkis for uns and the dual sokidedu for sokidedum. Evidently the dual in Gothic, as in many other Indo-European languages, was a category in decline (by the time of the New Testament it had completely fallen out of the Greek language). In this particular instance, as some scholars suggest, the difference in gender (as evidenced by the neuter form of the adjective winnandona) might have triggered the use of the plural in place of the dual. But in other passages we find the plural replacing the dual with no obvious trigger.
Luke 2.49 also begins with an interesting collocation: ƕa þatei sōkidēduþ mik? The phrase ƕa þatei is a compressed phrase 'what is this?', the þata then pointing to what follows, hence the relative marker ei. This phrase has, by the time of the Gothic text, become frozen as a way of saying 'why'. A similar development happened within Latin during the Middle Ages, where quid est quod -- literally 'what is (this, the fact) that...' -- came simply to denote 'why'.
2:41 - jah wratodedun þai birusjos is jera ƕammeh in Iairusalem at dulþ paska.
42 - jah biþe warþ twalibwintrus, usgaggandam þan im in Iairusaulwma bi biuhtja dulþais,
43 - jah ustiuhandam þans dagans, miþþane gawandidedun sik aftra, gastoþ Iesus sa magus in Iairusalem, jah ni wissedun Iosef jah aiþei is.
44 - hugjandona in gasinþjam ina wisan qemun dagis wig jah sokidedun ina in ganiþjam jah in kunþam.
45 - jah ni bigitandona ina gawandidedun sik in Iairusalem sokjandona ina.
46 - jah warþ afar dagans þrins, bigetun ina in alh sitandan in midjaim laisarjam jah hausjandan im jah fraihnandan ins.
47 - usgeisnodedun þan allai þai hausjandans is ana frodein jah andawaurdjam is.
48 - jah gasaiƕandans ina sildaleikidedun, jah qaþ du imma so aiþei is: magau, ƕa gatawides uns swa? sai, sa atta þeins jah ik winnandona sokidedum þuk.
49 - jah qaþ du im: ƕa þatei sokideduþ mik? niu wisseduþ þatei in þaim attins meinis skulda wisan?
50 - jah ija ni froþun þamma waurda þatei rodida du im.
51 - jah iddja miþ im jah qam in Nazaraiþ, jah was ufhausjands im; jah aiþei is gafastaida þo waurda alla in hairtin seinamma.
52 - jah Iesus þaih frodein jah wahstau jah anstai at guda jah mannam.
2:41 jah wratodedun þai birusjos is jera ƕammeh in Iairusalem at dulþ paska. 42 jah biþe warþ twalibwintrus, usgaggandam þan im in Iairusaulwma bi biuhtja dulþais, 43 jah ustiuhandam þans dagans, miþþane gawandidedun sik aftra, gastoþ Iesus sa magus in Iairusalem, jah ni wissedun Iosef jah aiþei is. 44 hugjandona in gasinþjam ina wisan qemun dagis wig jah sokidedun ina in ganiþjam jah in kunþam. 45 jah ni bigitandona ina gawandidedun sik in Iairusalem sokjandona ina. 46 jah warþ afar dagans þrins, bigetun ina in alh sitandan in midjaim laisarjam jah hausjandan im jah fraihnandan ins. 47 usgeisnodedun þan allai þai hausjandans is ana frodein jah andawaurdjam is. 48 jah gasaiƕandans ina sildaleikidedun, jah qaþ du imma so aiþei is: magau, ƕa gatawides uns swa? sai, sa atta þeins jah ik winnandona sokidedum þuk. 49 jah qaþ du im: ƕa þatei sokideduþ mik? niu wisseduþ þatei in þaim attins meinis skulda wisan? 50 jah ija ni froþun þamma waurda þatei rodida du im. 51 jah iddja miþ im jah qam in Nazaraiþ, jah was ufhausjands im; jah aiþei is gafastaida þo waurda alla in hairtin seinamma. 52 jah Iesus þaih frodein jah wahstau jah anstai at guda jah mannam.
From the King James version:
2:41 Now his parents went to Jerusalem every year at the feast of the passover. 42 And when he was twelve years old, they went up to Jerusalem after the custom of the feast. 43 And when they had fulfilled the days, as they returned, the child Jesus tarried behind in Jerusalem; and Joseph and his mother knew not of it. 44 But they, supposing him to have been in the company, went a day's journey; and they sought him among their kinsfolk and acquaintance. 45 And when they found him not, they turned back again to Jerusalem, seeking him. 46 And it came to pass, that after three days they found him in the temple, sitting in the midst of the doctors, both hearing them, and asking them questions. 47 And all that heard him were astonished at his understanding and answers. 48 And when they saw him, they were amazed: and his mother said unto him, Son, why hast thou thus dealt with us? behold, thy father and I have sought thee sorrowing. 49 And he said unto them, How is it that ye sought me? wist ye not that I must be about my Father's business? 50 And they understood not the saying which he spake unto them. 51 And he went down with them, and came to Nazareth, and was subject unto them: but his mother kept all these sayings in her heart. 52 And Jesus increased in wisdom and stature, and in favour with God and man.
The most conspicuous sound shift affecting the Germanic languages is Grimm's Law. According to this rule, the following sound correspondences obtain between Proto-Indo-European and Proto-Germanic:
|p, t, k remain unchanged after s, and|
|t remains unchanged after p, k.|
Certain discrepancies in the correspondences appear upon closer inspection of the data. In particular, one frequently finds that the Germanic voiceless spirants (f, þ, x, xw) and s become voiced:
|f, þ, x, xw, s > v, ð, ḡ, ḡw, z.|
Many of these counterexamples are explained by Verner's Law. This states that the voiceless spirants remain when initial, or when immediately preceded by the PIE accent. For example, *t > þ in PIE *bhréH-ter > Gothic broþar, but *t > ð in PIE *pH-tér > fadar [faðar]. Note in this last example that initial *p > f, with no voicing.
Consider the following examples:
|p-||podós (Gk.)||f-||fōtáus 'of the foot'||p > f||Grimm|
|-p-||kléptēs (Gk.)||-f-||hliftus 'thief'||p > f||Grimm|
|-p-||kapálam (Skt.)||-v-||háubiþ 'head'||p > v||Verner|
|t-||tu (Lat.)||þ-||þu 'thou'||t > þ||Grimm|
|-t-||várte (Skt.)||-þ-||waírþa 'I become'||t > þ||Grimm|
|-t-||patér (Gk.)||-ð-||fadar 'father'||t > ð||Verner|
|k-||cordis (Lat.)||x-||haírtins 'of the heart'||k > x||Grimm|
|-k-||déka (Gk.)||-x-||taíhun 'ten'||k > x||Grimm|
|k-||com-mūnis (Lat.)||-ḡ-||ga-máins 'common'||k > ḡ||Verner|
|s-||sá (Skt.)||s-||sa 'that'||s > s||Grimm|
|-s-||geú(s)ō (Gk.)||-s-||kiusa 'I choose'||s > s||Grimm|
|-s-||bhárase (Skt.)||-z-||baíraza 'art borne'||s > z||Verner|
Note in the instance of Latin commūnis and Gothic gamáins that Verner's Law applies to proclitics. The example of Sanskrit bhárase and Gothic baíraza illustrates that the accent must be on the vowel immediately preceding the consonant for Grimm's Law to apply.
Among the ancient Germanic languages Gothic is uniquely conservative in terms of phonology. As will be seen in the section on the weak preterite, the dental suffix retains a fuller expression than in languages such as Old English or Old Norse. Gothic also preserves a situation which precedes regular umlaut due to i in a following syllable. For example, compare Gothic alþeis 'old' and alþiza 'older' to Old English eald and ieldra, respectively; similarly compare PGmc. *gastiz > Goth. gasts to Old Norse gestr.
Gothic does, however, show some important sound changes. Some of the more notable examples appear below.
6.2.1 Initial *fl-
Initial *fl- > þl- in syllables ending in h /x/. Compare Old Saxon and Old High German fliohan 'flee' to Gothic þliuhan. The same may also occur in stems ending in q /kw/: for example, Gothic þlaqus 'putting out leaves, tender' with possible relation to Old High German flah 'flat', Latin placidus ('flat, even' and hence) 'gentle, quiet'. The change *fl- > þl- does not occur in stems with other final consonants. For example, compare Gothic flōdus to Old English and Old Saxon flōd 'flood, stream'.
6.2.2 Final -s
As mentioned in Section 3.1, final -s > zero before (short vowel) + (consonantal r). For example, waír + -s > waír 'man'; similarly baúr 'son', anþar 'second', unsar 'our'. Compare dags 'day', gasts 'guest', akrs 'field', swērs 'honored', skeirs 'clear', G brōþrs 'of a brother'.
Additionally, -s > zero before stem-final s. For example, runs + -s > runs 'a running'. Compare accusative runs, showing the s is part of the stem, not the nominative ending.
Several sound changes occur frequently when consonants become final in the past tense or combine with the second person singular past tense ending -t. These are collected here for reference.
Several nouns have stems ending in i. As mentioned above, this does not lead to regular umlaut of the root vowel as it does in other Germanic languages such as Old Norse and Old English. The nouns gards 'court', staþs 'place', and gasts 'guest' illustrate the masculine forms of the i-declension; ansts 'grace', fahēþs 'joy', and qēns 'woman' illustrate the feminine forms. Recall that ei is the Gothic spelling of [ī].
Note that, because of the lack of umlaut, the singular forms of masculine i-stems parallel those of the a-stems. Feminine abstract nouns in -ōns and -áins, derived from verbs of the second and third weak conjugation, decline like ansts. For example, laþōns 'invitation' from laþōn 'to invite', mitōns 'a thought' from mitōn 'to think over', bauáins 'a dwelling' from bauan 'to inhabit', libáins 'life' from liban 'to live'.
Some nouns which belong to the i-declension in Gothic belong to the a-declension in sister languages. For example, compare Gothic gards to the Old Norse a-stem garðr, but Goth. gasts and ON gestr are both i-declension. By contrast, Gothic qēns appears both as i-stem in the Eddic form kvæn and as the ōn-stem kona (G pl. kvenna) in Old Norse.
The masculine noun náus 'corpse' has plural forms N. naweis and A. nawins. The feminine noun háims 'village' follows the declension of ansts in the singular, but follows the ō-stem giba in the plural (see Section 3.2). Feminine abstract nouns in -eins follow the declension of ansts, except for N pl. -ōs and G pl. -ō. The forms of the masculine náus 'corpse' and of the feminine háims 'village' and láiseins 'doctrine' are given below.
The u-stem nouns appear in all genders, though there are few remnants of neuter forms. The masculine sunus 'son', feminine handus 'hand', and neuter faíhu 'cattle' illustrate the declension.
Some scribes write N sg. sunáus beside sunus, D sg. sunu beside sunáu, and V sg. sunáu beside sunu. The neuter noun filu 'much' falls under this declension, with G sg. filáus used adverbially with comparatives in the sense 'very'.
Because of the rich morphology of the Gothic verb, subject pronouns are generally unnecessary. They are used only for emphasis. In addition to singular and plural, the first and second person pronouns also distinguish a dual number, e.g. wit 'we two', igqara 'of you two'.
The forms of the Gothic first and second person pronouns are as follows.
|1st Person||2nd Person|
Note that the first person plural A uns and D unsis often interchange with one another. The oblique forms fill the role of first and second person reflexive pronouns, so that e.g. Modern English 'I hit myself' would be more literally in Gothic 'I hit me'.
The third person pronoun is built from a stem i-. Unlike the first and second person pronouns, these do not serve as reflexives. Instead the forms sik, seina, sis serve as reflexives. Though the forms are singular, they serve as reflexives for all genders and numbers. Thus the reflexive of is in the accusative is sik 'himself', and likewise the reflexive of neut. N pl. ija in the accusative is sik 'themselves'. The forms of the third person pronoun and the reflexive are given below.
Weak verbs form a category separate from strong verbs. Whereas vowel gradation (ablaut) characterizes strong verbs (cf. Modern English sing-sang-sung-song), this is not so for weak verbs. Rather the addition of a dental suffix -d- in the past tense characterizes weak verbs (cf. Modern English arrive-arrived). This dental suffix is appended to the verbal stem, before the addition of personal endings. The dental suffix is found not only in finite verbal forms, but also in the past participle (cf. Modern English 'That problem, addressed by Einstein, was the beginning of modern quantum theory').
Gothic has four classes of weak verbs. These classes are distinguished by the vowel which precedes the dental suffix, and the presence or absence of a nasal appended to the stem: -i-, -ō-, -ái-, -nō-. These correlate with different forms of the infinitive. The following chart lists examples of the Gothic weak verb classes.
|Class||Preterite (1/3 Sg.)||Infinitive||Meaning|
The class i verb nasjan 'save' illustrates the active forms of the weak verb.
All present forms derive from the infinitive, minus the -an ending. All past forms contain the dental suffix. Note that all finite preterite forms, except for the singular indicative, contain the suffix -dēd- rather than simply -d-.
As with the strong verbs, the morphological mediopassive has only present forms. The forms of nasjan 'save' illustrate the paradigm.
There are no forms for the dual. Be careful to note that the -d- of the mediopassive forms is found in all verbs, strong and weak; it is not to be confused with the -d- of the weak dental preterite. The stem of these forms derives from the infinitive, as with the present active forms.
Gothic makes use of a number of conjunctions. The simple joining of two clauses in a copulative fashion is most often accomplished with jah 'and'. This conjunction stands as the first element in its clause: warþ hūhrus abrs and gawi jáinata, jah is dugann alaþarba waírþan 'a great famine came over that people, and he started to become very poor' (Luke 15.14). jah is also found sentence-initially, continuing a previous sentence in a style parallel to Greek kaí in the New Testament. The enclitic -uh 'and' affixes to the first word of its clause: Galáiþ in praitaúria aftra Peilātus jah wōpida Iēsu qaþuh imma... 'Pilate came into the pretorium again and called Jesus and said to him...' (John 18.33). Following iþ 'but', the conjunction -uh generally attaches to the verb of the clause: iþ Iesus iddjuh miþ im 'and then Jesus went with them' (Luke 7.6). The negative copulative is nih 'and not', a combination of ni 'not' and -uh. nih generally stands at the beginning of its clause: ni maúrnáiþ sáiwalái izwarái ƕa matjáiþ jah ƕa drigkáiþ nih leika izwaramma ƕē wasjáiþ 'have no thought for your life, what you shall eat and what you shall drink; nor for your body, how you shall dress' (Matthew 6.25).
The general disjunctive particle is aíþþáu 'or'. The sequence 'either... or' is typically jabái... aíþþáu or andizuh... aíþþáu. For example, untē jabái fijáiþ áinana jah anþarana frijōþ, aíþþáu áinamma ufháuseiþ iþ anþaramma frakann 'for either he will hate the one, and love the other; or else he will hold to the one, and despise the other' (Matthew 6.24). In direct questions of a disjunctive nature one finds -u... þáu '(either)... or', where -u is appended to the first word of the first question: abu þus silbin þu þata qiþis, þáu anþarái þus qēþun bi mik? 'Sayest thou this thing of thyself, or did others tell it thee of me?' (John 18.34) Indirect disjunctive questions usually omit -u, though þáu remains to mark the second part.
There are several adversative particles: iþ, aþþan, ak, akei. These generally stand at the head of their clause: Háusidēduþ þatei qiþan ist þáim áirizam: ni maúrþrjáis; iþ saei maúrþreiþ skula waírþiþ stauái. aþþan ik qiþa izwis... 'Ye have heard that it was said of them of old time, Thou shalt not kill; and whosoever shall kill shall be in danger of the judgment: But I say unto you...' (Matthew 5.21-22). The particle ak generally follows a negative introduction, while akei follows a positive. For example, ni hugei hauhaba, ak ogs 'Be not highminded, but fear' (Romans 11.20); compare all binah, akei ni all daug 'All things are lawful for me, but all things are not expedient' (1 Cor. 10.23).
The main causal conjunctions are áuk, allis, raíhtis, untē. The particles áuk, allis, raíhtis generally occupy second position in their clause, though áuk and raíhtis may also occupy third position. For example: mahteigs áuk ist 'for [God] is able' (Romans 14.4); maht wēsi áuk 'For it might have been' (Mark 14.5). áuk may combine with raíhtis or allis: sa áuk raíhtis 'for he' (Mark 6.17); saei áuk allis 'for whoever' (Mark 9.41). By contrast, untē 'for, because, until' occupies the first position in its clause: áudagái þái hráinjahaírtans, untē þái guþ gasaiƕand 'Blessed are the pure in heart: for they shall see God' (Matthew 5.8).
For ease of reference, the most common connective particles are listed in the following chart, along with their essential roles and basic translations.
|andizuh... aíþþáu||disjunctive||'either... or'|
|final||'to the end that, because'|
|duþþē ei||final||'to the end that, because'|
|du þamma ei||final||'to the end that, because'|
|ei||final||'that, so that'|
|ibái (iba)||final||'lest, that... not'|
|jah... jah||copulative||'both... and'|
|jaþþē... jaþþē||disjunctive||'whether... or'|
|ni þatáinei... ak jah||copulative||'not only... but also'|
|ni (nih)... ni (nih)||disjunctive||'neither... nor'|
|nibái (niba)||conditional||'unless, if... not'|
|nih... ak jah||copulative||'not only... but also'|
|sunsei||temporal||'as soon as'|
|temporal||'when, as long as'|
|temporal||'when, as long as'|
|'until, until that, as long as'|
|þáu||concessive||'in that case'|
|und þatei||temporal||'until, until that, as long as'|
|temporal||'until, until that, as long as'|