The earliest literary references to the Goths are found in the works of Pliny, Strabo, and Tacitus. In his Natural History, Pliny the Elder mentions a certain Pytheas of Messalia, a Greek historian who records his observations on a journey through the 'parts near Ocean', sometime around the time of Alexander the Great. Pliny states that Pytheas believed a tribe called Gutones inhabited regions of Germania. Pliny's statement of Pytheas' findings concerning the Goths, however, presents scholars with two difficulties: (1) that the statement in question actually contains Guiones, which must be emended to Gutones to bring it in line with other presumed references to the Goths in Pliny; (2) the reference to Germania is clearly Pliny's own, since no such province was in existence in the time of Pytheas. Hence we cannot be sure what Pytheas himself said about their location. Pliny later mentions the Gutones as one of five tribes of the Germani.
Strabo, in his Geography, mentions the Gutones in a discussion of the Hercynian Forest. Once again, however, such an association rests on textual emendment: the manuscript reads Boutones, which scholars emend to Goutones. The location is not specified, which is not unexpected, since few authors could claim to know anything certain about regions beyond the Danube in this period.
Tacitus, in his work the Germania, written sometime around 98 AD, says in chapter 43: "beyond the Lugii, the Gotones are ruled by kings..., and next, close to the Ocean, the Rugii" and others. According to his account, the Suebi are in northern Europe, the Lugii beyond them, and the Gotones beyond them; but the latter must not quite be on the Baltic coast, since the Rugii and others are closer to the Baltic than the Gotones. Tacitus also mentions in his later work the Annales, chapters 2.62-63, that a certain Catualda was a noble among the Gotones.
In his Geography, Ptolemy locates the Guthones near the Vistula river. He elsewhere lists the Goutai as one of the seven tribes inhabiting Skandiai, presumably Sweden. It is not clear if both of these terms refer to the same tribe. If so, these are perhaps reflexes of strong and weak forms of the name. If not, one is not sure which ones are 'the' Goths. Some link the Goutai to the Geats of Beowulf, whose history thereafter is know from other medieval sources. But it is not clear that these are the Goths of Scandinavia.
The late 4th century, non-Christian author Ammianus Marcellinus is an important source for our understanding of the early movements of the Gothic tribes and their interaction with imperial forces. But he mentions nothing of Gothic origins, even though he mentions origins of others, such as the Alans, who descended from the Massagetae, and the Persians from the Scythians. Ammianus focuses on the movements of individual Gothic groups, most importantly the Tervingi and the Greuthungi.
Bishop Ambrose of Milan, in composing his work De Fide sometime around 380 AD for the emperor Gratian, links the Gothi with the Biblical Gog, ruler of the land Magog, which is perhaps set to the north, and maybe connected with islands. Ambrose seems to have taken the occasion to place Gratian's struggle with the Goths in a more divine setting, since in Revelations 20.7-10, Gog is destined to compass 'the camp of the saints'. The genealogist Josephus, earlier writing the Antiquitates in 93-94 AD, links Magogites with Scythians; Josephus is directly quoted by the later Gothic historian Jordanes.
Jerome, writing sometime c. 390, challenges the identification of the Goths with Gog and his people. He identifies Getae and Gothi. Orosius, writing the Historia adversum paganos in 417 AD and seeking to play down the prophetic overtones of a link between Goths and Gog, follows Jerome's association. The association was a simple one, since the Getae had lived along the lower Danube, and this was the origin of the Gothi in their attack on Rome. Augustine, however, writing De civitate Dei between 413 and 427 AD, denies the equation of Goths and Scythians, as well as that of Goths and Getae.
One of our most important sources of Gothic history is Jordanes, who wrote the Origins and Acts of the Goths or Getica in 550 AD in Constantinople. Though he wrote in Latin, Jordanes is unique among our sources because he is the only one who is himself a Goth. He states in his work that he relies on Gothic oral tradition, but nevertheless claims some personal acquaintance with the material he treats. He also mentions that he closely follows the written work of another historian, the Gothic History written by Cassiodorus, a Roman Senator in the 520s in the court of Theodoric the Great, the Ostrogothic king of Italy. He thrice mentions another historian, Ablabius, who perhaps wrote in the court of a Visigothic king.
The Getica gives an account of Gothic history from its inception, i.e. from the origin of the Gothic people to the time of writing, providing several concepts central to modern attempts to reconstruct Gothic history:
Under the assumption that there are no new peoples, just the same peoples with new names and new locations (a typical motif of ancient history writing), Cassiodorus, and hence Jordanes, were able to equate the Goths with Scythians, Amazons, Getes, and Dacians: the Gothic kingdom was founded before Rome, and the Goths fought in the Trojan war. Under Berig the Goths crossed the Baltic in 1490 BC, and under Filimer they moved to the Black Sea only five generations later, i.e before any of the earliest mentions of the Goths.
As modern scholarship sifts through the ancient sources it becomes clear that, by the time of the fourth century, some twelve or thirteen groups of Goths are known from the records. Five coalesce in the fifth century to form the well-known Visigoths and Ostrogoths, while the others remain distinct (after Heather, 1996):
|Visigoths:||1||the greater part of the Tervingi|
|2||the Greuthungi under Ermenaric|
|3||Goths led by Radagaisus|
|5||Goths under Theodoric Strabo|
|Others:||6||the remaining Tervingi, perhaps the same as those led by Arimer|
|7||Greuthungi led by Farnobius|
|8||Greuthungi led by Odotheus|
|9||Goths under Bigelis|
|10||Goths under Dengizich|
|12||Goths near the Sea of Azov|
Ammanianus mentions that the Tervingi, groups 1 and 6, formed one unit; the Greuthungi under Ermenaric were another important political group of the 4th century. If the other groups eventually came under the control of Ermenaric, this might fall in line with the history of Jordanes; but the latter's account of Ermenaric is believed by some scholars to be more an embellished version of Ammianus' history rather than a previously attested tradition, and so Ermenaric's conquests have been exaggerated by Jordanes to put the Gothic leader on par with the later Attila.
In 399 the Roman poet Claudian wrote (In Eutropium, 2.152-153):
|Ostrogothis colitur mixtisque Gruthungis Phryx ager|
|The Phrygian plain is inhabited by Ostrogoths and mixed Greuthungi.|
If not merely a rhetorical device, this would make a further distinction between Ostrogoths and Greuthungi, both separate from the Tervingi, and thus adding to the number of Gothic tribes known in the 4th century.
The following passage is the nativity scene found in Luke 2:1-14. The Gothic text has somewhat redundant material in Luke 2.2: at [wisandin kindina Swriais] raginondin Saurim Kwreinaiau, leading scholars to believe that a marginal gloss has crept into the text during its transmission. This particular phrase shows the common construction at + substantive + participle, an absolute construction in Gothic similar to the genitive absolute in Greek, or the ablative absolute in Latin.
The text also contains other notable grammatical features. Luke 2.3 shows the occasional use in Gothic of ei + subjunctive for purpose clauses: iddje:dun alla'i, ei melida'i we:seina. The demonstrative tho: appears in Luke 2.6 to refer to Joseph and Mary; Gothic uses the neuter plural to refer to individuals of different genders. This however is not exclusive: compare ins in Luke 2.9. Luke 2.7 gives an example of the use of the genitive in negated clauses: ni was im rumis, literally 'there was not for them of room'.
In Luke 2.14 we find an instance of the Gothic translation remaining more faithful to the Greek than the English of the King James Version. Where the English is 'and on earth peace, good will toward men', with 'peace' and 'good will' in apposition, Gothic in fact preserves the Greek genitive in godis wiljins 'of good will', qualifying the phrase 'among men': 'among men of good will'. Compare the Vulgate in hominibus bonae voluntatis, which is elaborated in the Spanish en la tierra paz, a los hombres que aman el Senor 'on earth peace to those men who love the Lord.'
2:1 - Warth than in dagans jainans, urrann gagrefts fram kaisara Agustau, gameljan allana midjungard.
2 - soh than gilstrameleins frumista warth at [wisandin kindina Swriais] raginondin Saurim Kwreinaiau.
3 - jah iddjedun allai, ei melidai weseina, hwarjizuh in seinai baurg.
4 - Urrann than jah Iosef us Galeilaia, us baurg Nazaraith, in Iudaian, in baurg Daweidis sei haitada Bethlahaim, duthe ei was us garda fadreinais Daweidis,
5 - anameljan mith Mariin sei in fragiftim was imma qeins, wisandein inkilthon.
6 - warth than, miththanei tho wesun jainar, usfullnodedun dagos du bairan izai.
7 - jah gabar sunu seinana thana frumabaur jah biwand ina jah galagida ina in uzetin, unte ni was im rumis in stada thamma.
8 - jah hairdjos wesun in thamma samin landa, thairhwakandans jah witandans wahtwom nahts ufaro hairdai seinai.
9 - ith aggilus fraujins anaqam ins jah wulthus fraujins biskain ins, jah ohtedun agisa mikilamma.
10 - jah qath du im sa aggilus: ni ogeith, unte sai, spillo izwis faheid mikila, sei wairthith allai managein,
11 - thatei gabaurans ist izwis himma daga nasjands, saei ist Xristus frauja, in baurg Daweidis.
12 - jah thata izwis taikns: bigitid barn biwundan jah galagid in uzetin.
13 - jah anaks warth mith thamma aggilau managei harjis himinakundis hazjandane guth jah qithandane:
14 - wulthus in hauhistjam guda
jah ana airthai gawairthi in mannam godis wiljins.
2:1 Warth than in dagans jainans, urrann gagrefts fram kaisara Agustau, gameljan allana midjungard. 2 soh than gilstrameleins frumista warth at [wisandin kindina Swriais] raginondin Saurim Kwreinaiau. 3 jah iddjedun allai, ei melidai weseina, hwarjizuh in seinai baurg. 4 Urrann than jah Iosef us Galeilaia, us baurg Nazaraith, in Iudaian, in baurg Daweidis sei haitada Bethlahaim, duthe ei was us garda fadreinais Daweidis, 5 anameljan mith Mariin sei in fragiftim was imma qeins, wisandein inkilthon. 6 warth than, miththanei tho wesun jainar, usfullnodedun dagos du bairan izai. 7 jah gabar sunu seinana thana frumabaur jah biwand ina jah galagida ina in uzetin, unte ni was im rumis in stada thamma.
8 jah hairdjos wesun in thamma samin landa, thairhwakandans jah witandans wahtwom nahts ufaro hairdai seinai. 9 ith aggilus fraujins anaqam ins jah wulthus fraujins biskain ins, jah ohtedun agisa mikilamma. 10 jah qath du im sa aggilus: ni ogeith, unte sai, spillo izwis faheid mikila, sei wairthith allai managein, 11 thatei gabaurans ist izwis himma daga nasjands, saei ist Xristus frauja, in baurg Daweidis. 12 jah thata izwis taikns: bigitid barn biwundan jah galagid in uzetin. 13 jah anaks warth mith thamma aggilau managei harjis himinakundis hazjandane guth jah qithandane:
14 wulthus in hauhistjam guda
jah ana airthai gawairthi in mannam godis wiljins.
From the King James version:
2:1 And it came to pass in those days, that there went out a decree from Caesar Augustus, that all the world should be taxed. 2 (And this taxing was first made when Cyrenius was governor of Syria.) 3 And all went to be taxed, every one into his own city. 4 And Joseph also went up from Galilee, out of the city of Nazareth, into Judaea, unto the city of David, which is called Bethlehem; (because he was of the house and lineage of David:) 5 To be taxed with Mary his espoused wife, being great with child. 6 And so it was, that, while they were there, the days were accomplished that she should be delivered. 7 And she brought forth her firstborn son, and wrapped him in swaddling clothes, and laid him in a manger; because there was no room for them in the inn.
8 And there were in the same country shepherds abiding in the field, keeping watch over their flock by night. 9 And, lo, the angel of the Lord came upon them, and the glory of the Lord shone round about them: and they were sore afraid. 10 And the angel said unto them, Fear not: for, behold, I bring you good tidings of great joy, which shall be to all people. 11 For unto you is born this day in the city of David a Saviour, which is Christ the Lord. 12 And this shall be a sign unto you; Ye shall find the babe wrapped in swaddling clothes, lying in a manger. 13 And suddenly there was with the angel a multitude of the heavenly host praising God, and saying,
|14 Glory to God in the highest,|
|and on earth peace, good will toward men.|
The Gothic alphabet closely resembles the Greek uncial alphabet of the fourth century AD. Where the Greek uncials proved insufficient for rendering some of the sounds, Roman or runic letters were borrowed. These alphabetic characters are also used to represent numerals. Rather than work with the Gothic alphabet itself, scholars generally work with a transliteration using the Roman alphabet, augmented with two additional characters and with the acute accent mark. The following chart lists the transliterated letters, their corresponding numerical values, and a rough guide to pronunciation.
|a||1||[a], o as in 'cot'|
|[a:], a as in 'father'|
|b||2||[v], v as in 'have'||medially after vowel or diphthong|
|[b], b as in 'bob'||otherwise|
|g||3||[ng], n as in 'sing'||before k, g, q|
|[x], ch as in 'Bach'||finally, or before s, t|
|[gh], g as in North Ger. 'sagen'||otherwise|
|d||4||[dh], th as in 'father'||medially after vowel or diphthong|
|[d], d as in 'did'||otherwise|
|e||5||[e:], a as in 'gate'|
|q||6||[kw], qu as in 'queen'|
|z||7||[z], z as in 'buzz'|
|h||8||[x], ch as in 'Bach'|
|th||9||[th], th as in 'with'|
|i||10||[i], i as in 'with'|
|k||20||[k], k as in 'kick'|
|l||30||[l], l as in 'lazy'|
|m||40||[m], m as in 'mouth'|
|n||50||[n], n as in 'nose'|
|j||60||[j], y as in 'you'|
|u||70||[u], o as in 'do it'|
|[u:], oo as in 'boot'|
|p||80||[p], p as in 'pin'|
|r||100||[r], trilled r as in Sp. 'rueda'|
|s||200||[s], s as in 'hiss'|
|t||300||[t], t as in 'tin'|
|w||400||[u], oo as in 'boot'||between consonants, finally after consonant|
|[w], w as in 'with'||otherwise|
|f||500||[f], f as in 'fife'|
|x||600||[k], k as in 'kick'|
|hw||700||[xw], ch w as in 'Bach was'|
|o||800||[o:], o as in 'phone'|
We use the numerical values to establish the order of the alphabet. The numerical value assigned to each letter corresponds closely to the Greek system employed at the time, supporting the assertion initially based on visual similarity that the Gothic alphabet was in fact modeled on the Greek. The symbols used to represent 90 and 900 occur only in their numeral function, never representing sounds of the Gothic language. They consequently have no transliterations.
The duration of doubled consonants is roughly twice that of their single counterparts. For example, inn 'within' has the prolonged n sound in English 'penknife', while in 'into' has the short n of 'cannon'; fulla 'full' (strong adj., fem. nom. sg.) has the prolonged l of 'call later', while fula 'foal' (noun, fem. nom. sg.) has the short l of 'caller'. Similarly atta 'father' has the prolonged t of 'Fat Tuesday', and likewise for other consonants. The exception to this practice is gg. As in Greek, gg is used to represent the the sound of ng in English 'finger', cf. Goth. figgrs. This practice extends to all velars, so that g before any velar represents the same nasal sound before that velar. For example, gk in drigkan represents the nasal plus unvoiced velar plosive as in the corresponding English 'drink'; gq in sigqan 'sink' represents roughly the sound of nkw in English 'inkwell'. Some words -- e.g. bliggw- 'scourge', glaggw- 'accurate', skuggw- 'mirror', triggw- 'faithful' -- may have contained a true prolonged g as in (a slow pronunciation of) English 'doggone', but this has probably given way to the sound [ng] by the time of Wulfila's translation.
The letter w is often found in words borrowed from Greek, where the Greek contains upsilon. As can be seen from the alphabetic order and numerical value, the Gothic symbol for w actually is the Greek upsilon. For this reason some editors of Gothic texts instead use y in such loanwords, breaking from a more uniform transliteration with w. Thus Lwstrws, the dative plural of 'Lystra', is transcribed by some editors as Lystrys.
Other sounds of the Gothic language are represented by digraphs. Specifically, the long-i sound [i:] is represented by ei, mimicking the contemporaneous Greek pronunciation of epsilon followed by iota. The digraph ai has a threefold distinction. In some instances ai represents the short-e vowel [e] found in Modern English 'bet' (or perhaps a slightly more open sound, as in the a of 'hat'). In other instances ai represents the long version of the same sound. And in the last instance ai represents the diphthong formed by its two constituents, namely the sound of i in Modern English 'white'. In transcription, these three values are distinguished by placement of an acute accent mark: ai' is [e], ai is [e:], and a'i is [ai]. A similar threefold distinction holds for the digraph au: au' is the vowel sound in Modern English 'bought', au is a long version of the same sound, and a'u is the diphthong represented by ou in Modern English 'about'. The digraph iu represents a falling diphthong (i.e. a diphthong accented on its first element) much like the eu of Modern English 'reuse' when the re- carries the stress. The situation is summarized in the following chart.
|ei||[i:], ee as in 'meet'|
|ai'||[e], e as in 'bet'|
|ai||[e:], same as above, but prolonged|
|a'i||[ai], i as in 'white'|
|au'||[o], ou as in 'bought'|
|au||[o:], same as above, but prolonged|
|a'u||[au], ou as in 'about'|
|iu||[i'u], eu as in 'reuse'|
It appears that these sounds were in fact all distinct in the period leading up to the emergence of Gothic and in its earliest stages. But the three values of ai may have merged by the time of Wulfila's translation, and likewise the three values of au may also have merged.
Being that of the earliest documented Germanic language, the sound system of Gothic is of great importance for historical studies. Some sound changes have occurred, however, in the span of time leading up to Wulfila's translation, so that Wulfila's own pronunciation is not necessarily the nearest approximation to the original sound system which Gothic inherited from Proto-Germanic. From a synchronic point of view, it is clear that o [o:] is already colored with some of the qualities of u, since we find spelling mistakes confusing the two, e.g. supu:da for supo:da. Likewise e [e:], though open, was close enough to be confused with ei [i:], e.g. qeins for qe:ns. It is also likely that h is in Wulfila's time closer to the h of Modern English 'he' than it is to the ch of 'Bach', and similarly with hw. Nevertheless, on etymological grounds and because of the archaic nature of the morphology, it is common in scholarship to ascribe values to Gothic letters which preserve the distinctions between, say, ei and e:, or between a'i, ai, and ai', though they may be prior to Wulfila's time, and not in accordance with Wulfila's own pronunciation. By the same token, given the fact that the same spelling mistakes are made in several languages of the other branches of Germanic, it is possible that the distinctions were never actually as clean as the historical linguist would like. In this scenario, the marking of these distinctions is merely a theoretical construct, but one to which we shall nevertheless adhere.
We may group the archaic pronunciation of the Gothic consonants according to points of articulation. This is done in the following chart.
|Voiceless:||p||t, tt||k (x), kk||q|
|Voiced:||b||d, dd||g, gg [gg]|
|Voiced:||b [v]||d [dh]||g [gh]|
|Nasals||m, mm||n, nn||gg [ng]|
Note that gg is listed among the nasals, reducing to g when marking a nasal before k or q. The letters b, d, and g appear both as stops and fricatives. The above chart is a phonetic, rather than phonemic, description. For example the difference in pronunciations of d is purely conditioned by environment (allophonic), never serving as the sole distinction of a word's change in meaning.
The vowels may similarly be organized according to articulation. This is done in the following chart.
Note that the Gothic letters e and o always denote long vowels, e: and o: respectively. On the other hand, i always denotes a short vowel.
The above system is complemented by the three diphthongs a'i, a'u, iu, which are all stressed on the initial vowel. The resonants l, m, n, r may also function as vowels. For example: fugls 'bird', ma'ithms 'treasure', ta'ikns 'token', ligrs 'bed'. The semivowel w may also form the nucleus of a syllable. For example, wau'rstw 'work'.
Conventionally scholars divide syllables in the Gothic language so that non-initial syllables begin with a consonant. Thus hai'r-to: 'heart', sle:-pith 'sleeps', af-le:t 'forgive thou'. Syllables ending in a short vowel are short, all others are considered long.
In general initial stress is the norm, though there are some exceptions. We may say there are three types of stress: primary, secondary, and weak stress. If we use bold italics to represent primary stress, italics for secondary stress, and no marking for weak stress, then these correspond to the stresses in the Modern English word 'incubate'. Primary stress falls on the first syllable, secondary stress on the last, and weak stress on the intervening syllable. The stress system of Gothic is similar. More specifically, root syllables bear primary stress when initial, secondary when non-initial. Consider the following examples:
In matters of stress, it helps to realize that not all prefixes are equal. Adverbial prefixes to nouns, and the reduplicated syllables of verbs, follow the same rule as above. For example:
|un- 'un-, not'||mahts||unmahts|
|sai'- (redupl. syll.)||sle:-pith||sai'-sle:p|
However, adverbial prefixes to verbs (i.e. preverbs) bear secondary stress. For example:
Thus stress alone may in some instances serve to distinguish verbs and nouns. Consider the following pair:
|af 'from'||le:tan 'let'||af-le:t 'forgive thou'||af-le:t 'forgiveness' (acc.)|
The contrast is similar to that found in Modern English 'project' (noun) vs. 'project' (verb). An exception to the secondary stress of preverbs is ga-, which carries weak stress: ga-sahw 'saw'.
Suffix syllables (but not endings), when following a weakly stressed syllable, follow the accentuation rules of root syllables. Compare sal-bo:nd 'they anoint' vs. salbo:-de:deina 'they might anoint'; mi-kils 'great' vs. mikil-du:ths 'greatness'.
Weak stress falls on syllables between those with primary or secondary stress. Compare -ra- in figgra-gulth above, and also -na- in ana-sai'-sle:p. In general, the prefix ga-, the interrogative particle -u, and the conjunction -uh 'and' carry weak stress. For example: ga-leiks 'like'; ga-u-laubjats 'do ye two believe'; ub-uh-wo:pida 'and he cried out'.
Nouns in Gothic are inflected for case, number, and grammatical gender. There are three grammatical genders: masculine, feminine, neuter. For nouns with clear sexual gender, the grammatical gender generally agrees with the sexual gender. For example, qe:ns 'woman' is feminine, so that natural gender and grammatical gender agree; but graba 'ditch' is also feminine, though the referent has no natural gender. There are two numbers: singular and plural (though personal pronouns and verbs also have a dual number). There are four cases: nominative (N), accusative (A), genitive (G), dative (D).
Case inflection is essentially a means of marking by suffixes grammatical functions which would otherwise be signalled by prepositions in Modern English. The most obvious remnant of the older case system in English is 's (apostrophe-s), which at the end of a noun fills the same role as the preposition of before a noun. For example, the bark of a dog is the same as a dog's bark. In much the same way, the nouns of Gothic use suffixes in order to denote grammatical function within a clause. Below is a short description of the case system of Gothic.
We may get a jump-start on nominal syntax if we step back for a moment and discuss what we might term logical cases. That is to say, before we pin down the meanings of the specific cases found in Gothic, we may first discuss a number of possible cases. We will take as our starting point the case system of Sanskrit, as being reasonably robust among the Indo-European languages. The chart below gives the eight cases found in the Sanskrit language, along with the basic meanings associated with those cases. (The fact that Sanskrit's case system is being used is immaterial -- the only purpose is to make certain logical distinctions among types of grammatical occurrences; the fact that these distinctions also happen to be made explicit in the suffixal system of a particular language is merely an added bonus.)
|Logical Case||Description of Use||Basic Preposition||Example of Use|
|Nominative||case of the subject||(none)||I killed him.|
|case of something predicated to the subject||(none)||The sky is blue.|
|Accusative||case of the direct object||(none)||I killed him.|
|case of the terminus of directed motion||(none), to(wards)||I ran (to the) east.|
|case of an expression involving extent in time or space||(none), for||The event lasted (for) five days.|
|Instrumental||case of the instrument of an action||with||I killed him with a knife.|
|case of accompaniment||with||I travelled with my friend.|
|Dative||case of the indirect object||(none)||He gave me a book.|
|to||He gave a book to me.|
|for||I wrote a recommendation letter for my student.|
|Ablative||case of origin, source, or separation||from||I went from New York to Austin.|
|Genitive||case of possession||of||The shoes of the man are dirty.|
|'s||The man's shoes are dirty.|
|case of the sphere of relation||of||I shed tears of joy.|
|This soup needs a pinch of salt.|
|The canyon is a day's journey from here.|
|case of the subject or object of nominalized action||of, 's (s')||Man's killing of man speaks to human nature.|
|Locative||case of location in space or time||in, on, at, within||I stood on the corner for an hour.|
|Vocative||case of direct address||(none), o!||(O) Luck, be a lady tonight!|
As one can see from the chart, the logical meanings of the cases may be expressed in a language, even if such a case system is not present. English retains overt marking only of the genitive, so that prepositions take over the role of the case system. Gothic declension, however, is more robust than that of Modern English, though more sparse than that of Sanskrit. One may then envision the syntax associated with the Gothic case system in terms of the following question: how do the eight logical cases above fit into the four extant cases of Gothic? The following chart gives the cases of Gothic, along with the logical cases whose role each has subsumed.
|Gothic Case||Logical Case||Example||Translation|
|Nominative||Nominative||frija ist this wito:dis||'she is free of that law'|
|Vocative||atta unsar thu in himinam||'our Father, thou in heaven'|
|Accusative||Accusative||gasai'hwith thana sunu mans||'ye shall see the son of man'|
|Vocative||hails thiudan Ju:daie:||'hail, king of the Jews'|
|Genitive||Genitive||gasai'hwith thana sunu mans||'ye shall see the son of man'|
|Ablative||frija ist this wito:dis||'she is free of (from) that law'|
|Dative||Dative||gif mis sei undrinna'i mik da'il a'iginis||'give me the portion of property which is coming to me'|
|Instrumental||wo:pida Ie:sus stibna'i mikila'i||'Jesus cried with a loud voice'|
|Locative||swe:gnida ahmin Ie:sus||'Jesus rejoiced in spirit'|
|Ablative||sa afar mis gagganda swintho:za mis ist||'he who comes after me is mightier than me'|
The above identifications are not iron-clad, nor should they be taken for actual historical evolution. But they do go a long way to explaining the syntactic descriptions of Gothic case usage found in the standard handbooks.
Like the other Germanic languages, Gothic has strong and weak nominal declensions. These are terms originally applied by J. Grimm to distinguish two types of declension within Germanic languages. Among nouns, the property of being strong or weak is inherent, each noun being either strong (only) or weak (only). Adjectives, by contrast, can be strong or weak depending on the situation: adjectives are declined strong when indefinite, weak when definite.
The a/ja/wa-stem nouns historically derive from o/jo/wo-stem nouns, respectively, and some grammars use the historical terminology. These nouns are generally masculine or neuter. Among masculine nouns, dags 'day' and hla'ifs 'loaf, bread' are a-stems; harjis 'army' and hai'rdeis 'herdsman' are ja-stems; thius 'servant' is a wa-stem. Their declensions are as follows.
When the nominative singular ends in -s, the -s is lost for the vocative, and so the vocative and accusative fall together. In the plural, nominative and vocative are the same. Note the substitution of b for f between vowels in the paradigm of hla'ifs; that is, intervocalic [f] becomes the voiced allophone [v].
Final -s drops when it immediately follows the combination (short vowel) + (consonantal r). For example one finds nominative singular wai'r + s > wai'r 'man', bau'r + s > bau'r 'son', both nouns following the declension of dags.
Among neuter nouns, wau'rd 'word' and wito:th 'law' are a-stems; kuni 'race' and reiki 'kingdom' are ja-stems; kniu 'knee' is a wa-stem. Their declensions are as follows.
The nominative and accusative singular forms of neuter nouns are always identical, as are the plural forms. The vocative is identical with these. As with f and b in hla'ifs, the th of wito:th alternates with d between vowels: intervocalic [th] becomes the voiced allophone [dh].
The o:/jo:/wo:-stem nouns historically derive from a:/ja:/wa:-stem nouns, respectively. These nouns are exclusively feminine. The noun giba 'gift' is an o:-stem; sunja 'truth', bandi 'band, bond', and mawi 'maiden' are jo:-stems; triggwa 'covenant' is a wo:-stem. Their declensions are as follows.
The declension of wo:-stems follows that of the o:-stems. The jo:-stems fall into two types, depending on whether or not the nominative and accusative singular forms are the same.
Verbs in Gothic, as in the other Germanic languages, fall into two categories: strong and weak. These terms have no relation to the same names applied to nouns and adjectives.
There are two tenses in Gothic, present and preterite. As with other Indo-European languages exhibiting this type of two-tense system, the distinction between preterite and present is the distinction between past and non-past, since the present forms are used for both present and future. This is similar to Modern English 'I am going on vacation next week', where the present tense has future meaning, equivalent to 'will go'. Likewise, the preterite forms subsume the roles of several different tenses in Modern English, such as the simple past 'did', perfect 'has done', and pluperfect 'had done'. There are three moods: indicative, subjunctive, imperative. The moods are formed with either the preterite or present stems, except for the imperative, which only employs the present stem. Generally the past subjunctive forms denote potential completed actions, whereas the present subjunctive has no such implication of completion. This parallels somewhat Modern English 'might have done' vs. 'might do'. There are also two voices in Gothic: active and (medio)passive.
As in English, ablaut, or vowel gradation, characterizes the strong verbs of Gothic. This system employs vowel alternation within a root to signify change in meaning or function. Take, for example, the English forms: sing-sang-sung-song. Within the base s-ng, an i gives present forms, an a past forms, a u the past participle, and o a derived noun. Other verbs may follow the same ablaut pattern in full or in part, e.g. ring-rang-rung (with no o-grade form). Still other verbs follow an entirely different ablaut pattern, e.g. hold-held-held.
There are seven classes of strong verbs. Six of these are characterized solely by ablaut. The seventh is characterized by reduplication, or by reduplication coupled with ablaut. In order to distinguish, then, the different ablaut classes, specific forms are listed illustrating the gradation sequence. This can be accomplished by listing four principal parts, from which all forms of a given verb may be derived:
The forms chosen as principal parts are, respectively, (1) the infinitive, (2) the first (or third) person singular preterite, (3) the first person plural preterite, (4) the nominative singular masculine preterite participle. The different strong verb classes are listed below with verbs illustrating the vowel gradation.
|Class||Meaning||(1) Infinitive||(2) 1st Sg. Pret.||(3) 1st Pl. Pret.||(4) Past Ptcple.|
As can be seen, the singular and plural preterite forms of class VII are derived from the same stem. The seventh class functions somewhat differently from the rest, and this will be treated in more detail in a later lesson.
The class IV verb bai'ran 'bear, carry' illustrates the active forms of the strong verb.
Note how all present forms are built from the first principal part, and all the preterite forms except the singular are built from the third principal part. The second principal part supplies the singular preterite forms, and the fourth principal part the preterite participle.
The notion of voice concerns the way in which logical action is manifested in a grammatical statement. By 'logical action' is meant action in the abstract, or the underlying process being referred to. Any action may be referred to in a number of ways, and the morphology of the language dictates whether different expressions of the same action may be rendered concisely or through circumlocution. Within the arena of logical action, one may distinguish agent and patient. The agent is the logical actor, the one doing the logical action; the patient, by contrast, is the one undergoing the logical action, the logical recipient. Within the arena of the grammatical action, one distinguishes the (grammatical) subject and the (direct) object. The grammatical subject denotes the one performing the action expressed by the verb in the statement; the direct object denotes the recipient of that verbal action, when different from the grammatical subject. A statement is active when agent and subject are the same. For example, 'I ate the cookie' is active; the logical action is that of 'eating', and 'I' is the agent of this logical action. 'I' is also the subject of the verb expressed: 'ate'. Here the patient, 'the cookie', is the direct object. On the other hand, a statement is passive when patient and subject are the same. For example, 'The cookie was eaten (by me)'. Here again the logical action is 'eating', and 'I' is the agent, while 'the cookie' is still the patient. But now the patient is the subject of the verb expressed: 'was eaten'. The agent, 'I', need not even be expressed, though it is possible with the phrase 'by me'.
A third voice is distinguished, called the middle voice. The middle voice is somewhere between the active and passive voices, where the distinction between agent and patient is blurred. In many of the ancient Indo-European languages, this voice denotes action which is reflexive (e.g. 'you'll get (yourself) killed'), for the personal benefit of the subject (e.g. 'I had a house built'), or representing an internal process (e.g. 'I wondered at its beauty'). In these languages, the morphology denoting the middle voice is often the same as that denoting the passive. Such uses of the morphological passive in Gothic are not very common, and the term mediopassive, rather than simply passive, is employed based largely on historical and comparative grounds.
Gothic has a morphological mediopassive only in the present. The forms of bai'ran 'carry' illustrate the conjugation.
There are no forms for the dual.
Any mention of Gothic word order and syntax must begin by saying that the Gothic translation of the New Testament follows the Greek extremely closely. The case system of Gothic is as robust as that of Greek, so that one to one correspondence of constructions is possible. Nevertheless, Gothic use often departs from Greek use, perhaps most conspicuously in the occurrence of Greek genitive absolute constructions, which are rendered in Gothic by dative constructions. (Ironically, the same dative rendering occurs in Old Church Slavonic, which possesses an even richer case system.)
In matters of word order, then, the concern is not solely to what degree the extant Gothic matches its Greek source, but also to what degree the extant Gothic matches patterns expected by comparison with other Germanic languages. Because of the antiquity of the Gothic documents and the general conservatism Gothic displays in morphology, the most pertinent comparanda are the early runic inscriptions. As the Gallehus inscription, c. 400 AD, shows,
|ek hlewagastiz holtijaz horna tawido:|
|I, Hlewagastiz, son of Holtagastiz, the horn made|
the unemphatic word order of the earliest Germanic documents was predominantly
|(Subject) + Object + Verb.|
This word order apparently lasted well into the time period of the Gothic documents. The same word order is frequently found in the Old English poem Beowulf, as in the opening lines:
|Hwaet, we: Ga:r-dena in ge:ardagum|
|the:odcyninga thrym gefru:non,|
|hu: dha: aethelingas ellen fremedon.|
|Listen! We have heard the glory of the Spear-danes,|
|of the people's kings, in days past,|
|how the heroes performed courageous deeds.|
In both clauses, the verbs (gefru:non, fremedon) occupy the last position, and the object directly precedes them. Such word order is even found in comparative constructions in Old English, e.g. sta:ne heardran 'than-stone harder', and in Old Norse, e.g. so'lo fegra 'than-(the)-sun fairer'. It is then quite likely that typical Gothic word order -- that is, word order emphasizing no particular element of the utterance -- during the time of Wulfila's translation was also SOV. The Gothic translation of Mark 8.23,
|frah ina ga-u-hwa-sehwi|
|asked him if he saw anything|
is, as often, a word-for-word translation of the Greek. But the second clause ga-u-hwa-sehwi suggests that the tendency for object to precede verb was strong enough that the object could even intervene between verb and prefix.
Wulfila's translation of the New Testament, however, frequently departs from SOV word order, and does so more often than one would expect if such departure were merely for stylistic reasons. For example, simple declarative sentences often have the structure
|Subject + Verb (+ Adverb) (+ Object),|
as in manne: sums a'ihta twans sununs 'a certain one among men had two sons' (Luke 15.11). This generally agrees with the Greek word order. When the Adverb is placed first, the Verb often follows directly, and the Subject is moved to the position following the Verb:
|Adverb + Verb + Subject (+ Object),|
e.g. suns qimith Satans 'immediately Satan comes' (Mark 4.15). This also agrees with the Greek. The common conjunction jah 'and' is frequently followed by the verb of the second clause. Thus,
|S + V (+ Adv) (+ O) + jah + V + S (+ O).|
For example, tharuh is qath du imma thatei bro:thar theins qam, jah afsna'ith atta theins stiur thana alidan 'then he said to him (that) your brother came, and your father killed a fattened calf' (Luke 15.27). Again this construction agrees with the Greek. When Gothic breaks with Greek word order, it frequently reverts back to verb-final word order: jah gai'rnida sad itan hau'rne: tho:ei matide:dun sweina, jah manna imma ni gaf 'and he yearned to eat his fill of the husks which the swine were eating, and the man did not give him (any)' (Luke 15.16). Here the Greek has kai oudeis edi'dou auto:i, literally 'and no one gave to him'.
As the verb-final structure of Modern German subordinate clauses and Modern English indirect questions shows, one might expect subordinate clauses in Gothic to preserve SOV word order. But even here the word order tends to follow the same Greek patterns found in main clauses:
|Relative (+ Subject) + Verb (+ Adverb) (+ Object), or|
|Relative (+ Adverb) + Verb (+ Subject) (+ Object).|
For example, und thatei usleithith himins jah ai'rtha, jo:ta a'ins ai'ththa'u a'ins striks ni usleithith af wito:da 'up to the point when heaven and earth pass, not one iota or one bit shall pass from the law' (Matthew 5.18). If the subject is simply a single relative pronoun, then the dependent word order may be
|Rel.Pron. + Verb (+ Adverb) (+ Object), or|
|Rel.Pron. + Adverb + Verb (+ Object).|
For example, ith than sa sunus theins, saei fre:t thein swe:s mith kalkjo:m, qam 'but then your son, who squandered your fortune on harlots, came' (Luke 15.30).
An adjective may either precede or follow its referent. They agree in gender, case, and number, with some exceptions. Feminine substantives are occasionally modified by masculine adjectives, or even neuter if the feminine noun denotes a thing. For example, ei kannith we:si... handugei guths 'that the wisdom of God... might be known' (Ephesians 3.10), where the neuter adjective kannith 'known' modifies the feminine abstract noun handugei 'wisdom'. A plural adjective or pronoun referring to two nouns of different gender is put in the neuter, e.g. ba (Zakarias jah Aileisabaith) framaldra wesun 'both (Zachary and Elizabeth) were very old' (Luke 1.7). A possessive pronoun generally follows the noun it governs, e.g. ahman izo:s 'her spirit', and likewise for possessive adjectives: atta theins 'your father', mith frijo:ndam meina'im 'with my friends'. Though there is a definite article in Gothic, or rather a demonstrative adjective which frequently assumes the role of an article, it often does not modify a noun governed by a genitive, e.g. in thiudangardja'i himine: 'in the kingdom of heaven'. Either noun, however, may also appear with an article: thana attan thizo:s ma'ujo:s 'the father of the(se) maidens'.
The general conservatism displayed by Gothic in terms of morphology leads scholars to expect the unemphatic word order was typically SOV in accordance with the earliest Germanic inscriptions. The Gothic New Testament however generally looks to be a mirror image of its Greek model. Given the rich morphology of the Gothic language, such word order would not have posed much difficulty for the intended audience, be it a native speaker's choice of word order or not.