CC BY
32
S. D. Kleyner
THE GOTHIC FUTURE: A TENSE THAT DOESN'T EXIST*
В статье рассматриваются три готских глагола, которые часто называют вспомогательными при образовании готского будущего времени - duginnan, haban и skulan. При внимательном рассмотрении оказывается, что только у глагола haban в сочетании с инфинитивом есть коннотации будущности.
Ключевые слова: готский язык, будущее время, skulan, duginnan, haban.
0. The problem
The problem with the Gothic future started with Streitberg, who, in the end of the 19th century, stated in his Gotisches Elementarbuch that while Gothic, as well as other old Germanic languages, uses present instead of future tense, there is indeed a duratives Futurum in Gothic, and it can be expressed in several ways:
1. duginnan + infinitive - Ph. 1:18 jah in pamma fagino, akei jah faginon duginna (Gk Kai ¿v тоотф xaip®: aAM Kai xap^aopai) and Lk 6:25 wai izwis, jus hlahjandans nu, unte gaunon jah gretan duginnid (Gk ovai, oi ye^&vxe« vuv, oxi nevB^aexe Kai K^avaexe).
2. haban + infinitive - Jn 12:26 jah parei im ik, paruh sa andbahts meins wisan habaip (Gk Kai onou elpi sy® ¿Kei Kai о SiaKovo« о eaxai), 2 Cor 11:12 ip patei tauja jah taujan haba (Gk о Se noi® Kai noi^a®), and 2 Thes 3:4 jah taujip jah taujan habaip (Gk [Kai] noieixe Kai noi^aexe).
3. skulan + infinitive in Lk 1:66 hva skuli pata barn wairpan (Gk xi apa to naiSiov touto eaxai)
4. Optative: Jn 5:47 pande nu jainis melam ni galaubeip, hvaiwa meinaim waurdam galaubjaip (Gk el Se xoi< ¿Keivou ypa^^aaiv ov niaxevexe, xoi< p^^aaiv niaxevaexe;), Mk 4:13 ni witup po gajukon, jah hvaiwa allos pos gajukons kunneip (Gk ovk o'iSaxe x^v napapoX^v xavx^v, Kai naaa< xa« napapo^a« yv®aea0e;), Lk 1:34 hvaiwa sijai pata, pandei aban ni kann (Gk eaxai xovxo, enei avSpa ov yiv®aK®;)
* This research has been carried out with the financial support of PrH® (Project № 14-04-00394). I am thankful to S. Ivanov for his valuable comments and interest.
Streitberg contrasts the duratives Futurum against the perfektives Futurum, expressed with the prefix ga- or with the help of wairpan that translates Gk eaopai on no less than 48 accounts (Streitberg 1910: 200-202).
This vision of the Gothic future has travelled from textbook to textbook ever since: it's given as a fact in the Russian textbook by Gukhman (ryxMaH 1958: 172-173); it's quoted by Wackernagel in his lectures - and left uncommented by Langslow (Wackernagel 2009: 254); it's a starting point for Leiss's working hypothesis on perfective and imperfective infinitives (Leiss 2012: 189) in Gothic.
On the other hand, there are authors who summarily dismiss the existence of the Gothic future altogether, like A. Lloyd who simply writes, 'In Gothic, where no formal future tense has developed...' (Lloyd 1979: 145), or like Hewson, who states that Germanic languages adopted an Ascending time paradigm that allows to use the non-past tense in reference to future, and 'the need to create a contrastive future is less imperative' (Hewson 1997: 221).
This is by no means the only description of the Gothic future; Streitberg now stands corrected in many respects. But the real issue remains: are those expressions really a Gothic way to speak about the future?
This paper will only address the compound forms of the duratives Futurum; the Optative, and all the discussion concerning wairpan or the prefixes and their connection to the Gothic Aspect will remain out of this framework.
1. Other Germanic languages versus Gothic: a brief couple of words
The ground work for the Germanic future in general was also done over a century ago; F. A. Blackburn (Blackburn 1892: 10-14) meticulously examined translations of the Latin futurum, and stated that different languages had developed different approaches to the issue:
Old High German and Old English seem to prefer the present indicative approach with rare insertions of the present Optative. Now we know that in OHG, many of these indicatives included the verb werdan (especially Latin future passives), while OE preferred beon1.
Old Norse and Old Saxon, on the other hand, seem to prefer the periphrasis, mainly with skulu/sculan to the present indicative, with
1 For the numbers, see Diewald, Wischer 2013.
present Optative being virtually non-existent. Old Norse actually uses another auxiliary, munu, which 'seems to be closest to our idea of future time' (Bauschatz 1982: 201)2.
While there are many similarities with Gothic, such as the OHG werdan clearly coming from the same verb as the Gothic wairpan, and also the use of skulan and the optative, the differences are rather striking. First, there is only one Gothic example where skulan is used to express the future - Lk 1:66, where it translates Gk £oxai. This is not the case for OHG and OE, where skulan is used much more frequently.
Second, while there may be hints to other ways of expressing the future, for example, willan in OE (Diewald, Wischer 2013: 211), nowhere else in the Germanic languages do verbs like 'to have' and 'to begin' play any kind of part.
Here, the obvious should be noted, that is, that Gothic is much older than all its known relatives, and, therefore, closer to the time where no grammatical expression of future existed in Germanic. Another obvious statement is the fact that dealing with Gothic we inevitably deal with a translation, and the source language (or, indeed, languages) of this translation has two future tenses, and this has to come into play at some point.
2. skulan + infinitive
There is only one instance where skulan translates a Greek future, and that is Lk 1:66 hva skuli pata barn wairpan (Gk ti ара то naiSiov touto eatai, 'What will this child be?'. There is much to consider here: first of all, skuli is an Optative. Secondly, the verb in infinitive is none other than wairpan - the verb that translates Gk ёаорш on no less than 48 accounts (Streitberg 1910: 200-202). This instance is unparalleled anywhere in the Gothic text we have: another example of ара used together with eatai is seen in Mt 19:27, but this part has been lost for us in Gothic. However, other instances of ара + future in Greek, such as Mk 11:13 and Lk 18:8 yield Gothic Optatives without any trace of modality; and ара is rendered as aufto and swepauh respectively.
2.1 Other examples
There are several Greek sources for Gothic skulan.
2 In modern Icelandic, mun is a standard auxiliary to form the Future tense.
a) ^eXXro etc. + infinitive. In all, the existing parts of the Gothic New Testament feature 9 instances where ^eXXro + infinitive is translated by skulan + infinitive3. In 3 of these cases, the tense is past, and in 2 cases skulan translates the Greek participle. There also are 2 cases where ^eXXro + infinitive is translated in Gothic as skulds ist4.
^eXXro + infinitive is translated into Gothic in a variety of ways:
In Jn 6:6, Jn 6:71 and Mk 10:32 ^eXXro is rendered as haban.
In Lk 10:1, Lk 19:4, Jn 6:15 and Jn 14:22, ^eUro is translated with munan, which is interesting in the light of a similar future construction in Old Norse, but even more so because in 3 of those instances, the Greek infinitive is epxea9ai5.
In 1 Thess 3:14 and 1 Tim 1:16, it is rendered with the construction anawairp + wesan + du + infinitive, which means 'the future was to'.
On one occasion, is turned into izei skaftida sik du + inf
'prepared himself to' (Jn 12:4), and on another occasion, in Lk 7:2, "^eAAev xeXeuxav is replaced with one Gothic word, swultawairj 'someone who's dying'.
b) Set etc + infinitive. In Gothic, there are 23 cases of skulan translating Set6, and 7 more where Set is translated by skulds ist1.
3 2 Tim 4:1 saei skal stojan, Gk tov peXXovto<; Kpivsiv; Jn 7:35 kadre sa skuli gaggan, Gk nov ovxo^ peXXei rcopEVEGBai and nibai in distahein ^iudo skuli gaggan jah laisjan ^iudos? Gk p^ eI<; t^v SiaGrcopav Tröv 'EXX^vrov peXXEi rcopEVEG&ai Kai 5i5&gkeiv tov^ "EXX^va^; Mt 11:14 saei skulda qiman, Gk o psXXrov spxEGBai; Jn 7:39 skuldedun niman, Gk e'peXXov XapßavEiv; Jn 12:33 dau^au skulda gadau^nan, Gk BavaTro %eXXev ano^v^GKEiv; Jn 18:32 dau^au skulda gaswiltan, Gk BavaTro %eXXev anoBv^GKEiv; Lk 9:31 skulda usfulljan, Gk %eXXev rcX^povv.
4 Lk 9:44 skulds ist atgiban, Gk peXXEi rcapa5i5oGBai; Lk 19:11 skulda wesi ^iudangardi gudis gaswikunjn, Gk rcapaxp^pa peXXei ^ ßaGiXEia tov BEOV ava^aivEGBai.
5 Jn 14:22 has ep^aviZsiv, translated as gabairhtjan.
6 Present: Lk 4:43 EvayyEXiGaGBai pE 5eT t^v ßaGiXEiav tov Beov wailamerjan ik skal bi ^iudangardja gudis, Lk 9:22 AeT tov uiov Tov avBprorcou noXXa naBETv skal sunus mans manag winnan, Lk 17:25 8eT avTov noXXa naBETv skal manag ga^ulan, Lk 19:5 5eT pE pETvai skal ik wisan, Mk 8:31 5eT tov uiov tov avBprorcou noXXa naBETv skal sunus mans filu winnan, Jn 9:4 ^pa<; 5eT spy&ZEGBai ik skal waurkjan, Jn 10:16 KaKETva 5eT pE ayayETv jah ^o skal briggan, 1 Cor 15:25 5eT yap avrov ßaGiXEVEiv skal auk is ^iudanon, 1 Tim 3:2 5eT ow tov sniGKonov avEniXnpnTov Eivai skal nu aipiskaupus ungafairinonds wisan, 1 Tim 3:7 5eT 5e Kai papxupiav KaX^v exeiv skal auk is weitwodi^a goda haban, 2 Tim 2:6 tov KonirövTa yEropydv 5eT npröTov Tröv Kapnröv pETaXapßavEiv arbaidjands air^os
Curiously, the only other way of translating Seî into Gothic is present in 2 Cor 12:1: KauxàtfQai ôeî hvopan binah 'it is allowed to boast', although the same Kau%fc9ai ggt is rendered as hvopan skuld sijai 'if there is need to boast' in 2 Cor 11.30.
c) o^eiX® etc + infinitive. There are 7 such cases8, and one more for skulds ist9, and there are no other translations for this Greek construction in Gothic.
d) e'x® + infinitive. There are 3 cases10; all of them are in the 1st person, present tense, and all the infinitives are verbs of speech.
waurstwja skal frumist akrane andniman, 2 Tim 2:24 SovXov 8e Kupíou ov 5eí páxEaBai i^ skalks fraujins ni skal sakan, Tit 1:7 5eí yap xóv éníoKonov ávéyKiniov EÍvai skalu^an aipiskaupus ungafairino^s wisan, Tit 1:11 ow; 5eí érciaxopíZEiv ^anzei skal gasakan, 1 Thess 4:1 5eí vpa; rcEpircaxEÍv Kai ápéoKEiv be® uns kaiwa skulu^ gaggan jah galeikan guda; Optative: Eph 6:20 5eí pE XaX^oai swe skuljau rodjan, Mk 9:11'HXíav 5eí ¿xbeív rcproxov Helias skuli qiman faur^is, Rom 12:3 p^ vrcEp^povEÍv nap' o 5eí ^poveív ni mais franjan ^au skuli franjan, Col 4:4 8eí pE XaX^oai swaswe skuljau rodjan, Col 4:6 5eí vpa; évi ÉKáaxro árcoKpívEaBai skulei^ ainkarjammeh andhafjan; Past: Lk 2:49 év xoí; xoü rcaxpó; pou 5eí Eívaí pE ^aim attins meinis skulda wisan, 2 Cor 2:3 skulda faginon; and one case of Lk 18:1 "exeyev 5e rcapaPoX^v avxoí; npó; xó 5eív rcávxoxE rcpooEvxEaBai avxov; Kai p^ éyKaKEÍv Qa^u^-^an jah gajukon im du ^ammei sinteino skulun bidjan jah ni wair^an usgrudjans.
7 1 Cor 15:53 5eí yap xó ^Bapxóv xoüxo évSúaaaBai á^Bapaíav (s)kuld ist auk ^ata riurjo ga(ham)on unriurein, 2 Cor 5:10 xov; yap rcávxa; ^pa; ^avEpro-Bnvai 8eí unte allai weis ataugjan skuldai sijum, 2 Cor 11:30 Eí KauxaoBai 5eí jabai kopan skuld sijai, 1 Tim 3:15 'iva EÍSfl; rcro; 5eí év o'ixro beoü ávaaxpé^EaBai ei witeis kaiwa skuld ist in garda gudis usmitan, 2 Thess 3:7 rcro; 5eí pipEÍaBai ^pa; kaiwa skuld ist galeikon unsis, Jn 12:34 5xi 5eí vyroB-qvai xóv uióv xoü ávBprórcou ^atei skulds ist ushauhjan sa sunus mans Lk 15:32 EÜ^pavBnvai 5e Kai xap^vai e5ei waila wisan jah faginon skuld was.
8 (present) Jn 13:14 Kai vpEÍ; ó^eíxexe áXX^Xrov vínxEiv xov; rcóSa; jah jus skulu^ izwis misso ^wahan fotuns, Jn 19:7 Kai Kaxa xóv vópov ó^eíxei ánoBavEÍv jah bi ^amma witoda unsaramma skal gaswiltan, 2 Cor 12:14 ov yap ó^eíxei xa xÉKva xoí; yovEÜaiv Bnoaupí^Eiv ni auk skulun barna fadreinam huzdjan, Eph 5:28 ovxro; ó^eíxouoiv Kai oi avSpEc; áyarcav swa jah wairos skulun frijon, 2 Thess 1:3 EvxapioxEÍv ó^EíXopEV x® be® awiliudon skulum guda, (past) Lk 17:10 o é^EíXopEV noi^oai nEnoi^KapEV unte ^atei skuldedum taujan gatawidedum, 1 Cor 5:9 énEi ó^eíxexe apa éK xoü Kóopou é^E^BEÍv unte skuldedei^ ^an us ^amma fairkau usgaggan,
9 2 Cor 12:11 éyé yap ro^Eilov v^' vprov ouvíoxaoBai a^an ik skulds was fram izwis gakannjan.
10 Lk 7:40 E'x® ooí xi eítceív skal ^us ka qi^an, Jn 8:26 rcoAM ex® nEpi vprov Xa^EÍv Kai KpívEiv manag skal bi izwis rodjan jah stojan, and Jn 16:12 "Exi rcoXXa ex® vpív XéyEiv Nauh ganoh skal qi^an izwis
These are, incidentally, all the cases of e%ro + a verb of speech in the Greek original.
e) finally, there is one instance where skulan does not translate a Greek verb used with an infinitive: the Gothic version of 1 Cor 15:2 xivi Xoy® evnYYe^iaa^nv v^tv ei Kaxexexe ('if you keep in memory what I preached unto you') is in ho saupo wailamerida izwis, skulup gamunan. Here, skulup clearly has a purely modal function11.
Skereins contains two sentences with the same construction:
Sk 1:23 swa auk skulda du galeikon seinai frodein jah mans aftra galapon waurdamjah waurstwam jah spilla wairpan aiwaggeljons usmete 'For thus He would accomodate His Wisdom and invite man again by word and deed and become a proclaimer of the way of life of the Gospel'12.
Sk 5:17 skulum nu allai weis at swaleikai jah swa bairhtai insahtai guda unbauranamma andsaljan sweripa jah ainabaura sunau gudis gup wisan anakunnan 'Now we all should, at such and so clear a declaration, render honor to the Unborn God, and recognize the Only Begotten Son of God to be God'
In the latter case, the sense is 'we ought to'. The former case is slightly more complex; Bennet13 understands it differently: 'that for us He should become a teacher of the righteousness in accord with God'; it seems that if we had a Greek original, we would half expect to find ^eXXro in the text.
2.2 Explanation
While it is quite understandable that so many different Greek phrases were again and again translated into Gothic with the help of skulan, perhaps several of these translations could shed some light on the processes that were happening to skulan.
2.2.1 Mt 11:14
An interesting case is Mt 11:14 Kai ei SeXexe Se^aaSai, avxo« eaxiv 'HXia< o ep%ea$ai.14 The Vulgate has et si vultis
recipere ipse est Helias qui venturus est - 'who is about to come',
11 In two other places where KaxéYsxs is present, it is once translated as gahabaip (1 Thess 5:21), and once as gafastaip (1 Cor 11:2). 2 In not indicated otherwise, translations of Skereins are by James W.
Marchland, available at http://www.gotica.de/skeireins/translations/english/ marchand.html
13 http://www.gotica.de/skeireins/translations/english/bennett.html
14 And if you are willing to accept it, he is the Elijah who is to come.
but in Gothic, the line is jah jabai wildedeip mipniman, sa ist Helias, saei skulda qiman - 'who had to come'. This past tense is also absent in the Anglo-Saxon Bible (he ys Helias pe to cumenne ys), nor in Tatian (her ist Helias thie thar zuouuert ist), but is curiously present in the Geneva Bible, the Bishop's Bible15 - and in most major English versions since (the source of such a translation is found in French: Calvin's Bible reads c'est Elie qui deuoit venir16).
It is fascinating that we see 'should' in the English Tyndale and Miles Coverdale versions; of course, there is no way the translators in the 16th century even knew about the existence of the Gothic version; but the translating device they used seems similar, if not the same. But that would mean that the past tense of Gothic skulan can function as a modal verb. Unfortunately, the list of examples containing the past tense of skulan is too short17 to give any parallels, so we have to turn to the present tense.
Gisela Ferraresi has put forward the idea that magan and skulan, when used with infinitives, 'should be re-analysed as modal verbs' (Ferraresi 1998: 46). As an example, she gives Lk 9:22 skal sunus mans manag winnan jah uskusans fram sinistam wairpan ('The son of man must suffer many things and be rejected by the elders'). There is a problem with some of her examples, though. In this specific case, in the Greek original we see Aei xdv uidv xou av$p©nou noXXa naSeiv Kai anoSoKip.aaS'qvai and x®v npeapuxep©v. We already know that Sei, in all cases but one, is translated into Gothic with skulan, so the choice of skulan here is not really indicative of anything.
Another example is Lk 9:44 unte sunus mans skulds ist atgiban in handuns manne ('Because the son of man is due to be delivered
15 Wycliffe Bible (1395): he is Elie that is to come; Tyndale NT (1525): this is Helyas which shuld come; Miles Coverdale Bible (1535): this is Helias, which shulde come; The Bishop's Bible (1568): this is Elias which was for to come; the Geneva Bible (1587) this is that Elias, which was to come, King James Bible (1611) this is Elias which was for to come, and so forth, up to the New International Version, although some revised editions, like the New Revised Standard, have 'who is to come'. The source of this translation was probably the first version of the Geneva Bible that appeared in 1560 outside England.
16 La bible d'Olivetan (1535): iceluy est Elias qui doibt venir; La bible de Lefevre d'Etaples (1530): icilui est Elie qui est ad venir.
17 Skulan is used in past tense only in 9 places; one of them (Lk 2:49) indeed translates a Greek present, but this is due to the simple sequence of tenses.
into the hands of men') which she interprets as 'passivization of skulan' and calls it a 'langes Passiv'. In the original, we see the passive infinitive: o yap uiö< tov avOprnnou napaSiSoa&ai ei<
%dpa<; av$p©n©v, and such usage in Gothic is indeed interesting and has parallels in the modal verb usage in German (ibid. 47).
I found two more examples like this in Gothic. 2 Cor 12:11 ey® yap ©^eiXov v^' auviaxaaöai appan ik skulds was fram izwis gakannjan ('For I was due to be commended by you'), and Lk 19:11 napaxp^^a piXXei ^ ßarnXeia tov &eov ava^aivea&ai is suns skulda wesi piudangardi gudis gaswikunpjan ('the Kingdom of God was due to appear') where the present tense is rendered as the past
Optative wesi because of the sequence of tenses. It has to be noted that the first example, with o^eiXro, is always translated as skulan or skulds ist, and that there seems to be no strict correlation between the choice of skulan and skulds ist, and the form of the infinitive (cf Lk 9:22 above, where dnoSoKi^aaO^vai is translated with the proper passive, uskusans wairpan, or Jn 7:35 nopeveaöai is
translated as skuli gaggan). Thus, in the first example, only its syntax is interesting; but in the second, the choice of skulan was not the only choice that could be made.
Ferraresi also cites our original example, Mt 11:14, and calls it an 'epistemic verb in a futuristic paraphrase' (ibid. 48), but it is hard to agree with her here. It is obvious from other Germanic languages that if this were indeed futuristic, this 'paraphrase' could be done differently; it is also clear that this switch to the past tense could simply be done by mistake, so this is not necessary a paraphrase at all. But if we admit that skulan in Gothic is indeed a real modal verb, there is a distinct possibility that this usage is intentional and is meant in pretty much the same way as the English should, with the only difference that in Gothic, skulan may be epistemic, and not just
deontic18.
This would mean a lot for an interpretation of Lk 1:66 hva skuli pata barn wairpan.
2.2.2 Odd cases and Lk 1:66
One rare example of skulan used without any particular reason is 1 Cor 15:2. In Greek, we see two conditional clauses: 8t' ov Kai
18 It also means that the «working hypothesis» by Leiss that skulan + perfective infinitive = deontic, and skulan + imperfective infinitive = imperfective future (Leiss 2012: 190) is not working: skal sunus mans manag winnan in Lk 9:22 is not an imperfective future.
aéÇea&e, xivi Xoy® eùnyyeXiaâMnv ù^îv el Kaxéxexe, éKxôq el etefi éniaxeuaaxe ('By this gospel you are saved, if you hold firmly to the word I preached to you. Otherwise, you have believed in vain'). The clause is repeated in Latin: per quod et salvamini qua ratione praedicaverim vobis si tenetis nisi si frustra credidistis. In Gothic, the second clause remains, but the first one is replaced with skulan+infinitive: frairh fratei _jah ganisifr: in hvo saufro wailamerida izwis, skulufr gamunan, niba sware galaubidedufr, and the literal translation would be as follows: 'By that you are saved: in the way I preached to you, you should remember; if not, you have believed in vain'. This is a strange replacement that can be explained only if we read skulup as an epistemic modal verb.
In Lk 1:66, we see several issues at once. The obvious one is the fact that it is a question, and about 1/6 of questions in the future tense are translated with Optatives in Gothic 9. The use of wairpan to translate âoxai is well attested (Streitberg 1910: 202). The only remaining issue is the verb skulan. Considering all above and noting that the Optative is used with skulan and not with wairpan, the logical conclusion seems that skulan + infinitive is not a standard way to designate a future - and not such a way at all; it is a situative use of an epistemic modal verb in a question.
3. duginnan + infinitive
There are only two instances of this usage:
Ph. 1:18 jah in pamma fagino, akei jah faginon duginna ('I rejoice in that, and will rejoice'; Gk Kai év xo"ux® %aip®: àXXà Kai Xap^aopai)
Lk 6:25 wai izwis, jus hlahjandans nu, unte gaunon jah gretan duginnid ('Woe upon you, laughing now, for you will mourn and weep' ; Gk oùai, oi yeXôvxeç; wv, oxi nevB^aexe Kai K^auaexe)
3.1 Other examples
In Gothic, there are three more instances of duginnan + infinitive, and all three translate Gk àp%o^ai:
Lk 3:8 waurkjaip nu akran wairpata idreigos jah ni duginnaifr qifran in izwis: attan aigum Abraham ('Make therefore fruits worthy of repentance, and begin not to say within yourselves'; Gk noi^aaxe oûv Kapnoùç à^iouç x^ç pexavoiaç: Kai ap^naBe Xéyeiv év èauxoîç, naxépa ë%opev xôv àppaâp)
19 For the statistics, see Blackburn 1892: 7.
Lk 14:29 allai pai gasaihvandans duginnaina bilaikan ina ('All seeing it begin to mock him'; Gk navxe« oi 0e©pouvxec< ap^®vxai avx® ¿pnaiZeiv)
II Cor. 3:1 Duginnam aftra uns silbans anafilhan? ('Do we begin again to commend ourselves?'; Gk ap%op,e0a naXiv eauxo"uc< auviaxaveiv;)
The construction 'ap%op,ai+ infinitive' had been in use since classical times (Goodwin 1890: 358), and its meaning is rather direct: to begin to do something. The question is, why is it used in the two instances where Greek does not offer ap%op,ai?
The only other examples where Gothic has duginnan + infinitive, and Greek does not, are Mk 14:72 and Lk 1:1; still, both Greek verses have a verb that can be translated directly as duginnan: noXkoi ¿nexeipnaav avaxa^aa0ai 'many have undertaken to write' managai dugunnun meljan in Lk 1:1, and Kai ¿nipaX©v eKXaiev 'and having broken down he wept' jah dugann greitan in Mk 14:72.
One obvious reason would be to suppose an attempt to translate a distinction between 'now' and 'later' present in the Greek text in both original instances: joy now and joy later in Ph 1:18, and joy now and woe later in Lk 6:25.
3.2 Contrast between the Present and the Future?
There is a problem with such an explanation, and it is twofold.
Firstly, other Germanic translators did not think it necessary to underline the contrast: for Lk 6:25 in OE, we have Wa eow pe gefyllede synt: forpam pe ge hingriad; Wa eow pe nu hlihad. forpam pe ge heofad and wepad, and Tatian writes, Uue iu thie thar gisatote birut, bithiu uuanta ir hungeret. Uue iu thie nu lahhet, bithiu uuanta ir vvuofet inti riozet. The other instance is, unfortunately, absent in both languages, but it is not the only case where Gothic invents something whereas other Germanic languages do not.
The other, more difficult problem, is the Gothic Bible itself: just four lines above Lk 6:25 we have Lk 6:21 audagai jus gredagans nu, unte sadai wairpip. audagai jus gretandans nu, unte ufhlohjanda ('Blessed are you now hungering: for you shall be filled. Blessed are you now weeping: for ye shall laugh'; Gk ^aKapioi oi neiv®vxe< vuv, oxi xopxaa0^aea0e. ^aKapioi oi K^aiovre« vuv, oxi yeXaaexe), that is, two more ways of expressing the Greek future: one with wairpan, and the other with a simple passive present. Why is that so?
The first translation - sadai wairpip - is explained by the fact there is no verb in Gothic to translate %opxaa0^aeG0e; it is not an
exception - there is no such verb in modern English, or in modern German, or in their old forms, so the translation 'become full' is quite self-evident, and we are dealing with a simple case of translation the Greek future with a Gothic present. The other verb, ufhlohjanda, is also a simple present20.
Presumably, gaunon and gretan do not take prefixes (why not?), and thus, duginnan was needed in our instances. But why duginnan? The second case in Greek clearly means 'I rejoice and will continue to rejoice', so why is duginnan used, and not a verb that means 'to continue', or indeed haban like in 2 Cor 11:12 ip patei tauja jah taujan haba - 'I do so and will continue to do'?
This situation is further complicated by the fact that there are other instances when present and future are contrasted, and nothing is done in Gothic at all. For example, Jn 14:19 nauh leitil, jah so manaseips mik ni panaseips saihvip; ip, jus saihvip mik, patei ik liba, jah _jus libaip ('Before long, the world will not see me anymore, but you will see me; because I live, you also will live'; Gk exi paKpdv Kai o Koap,oc< oukgxi 0e®pei, up,eic< Se 0e®peixe p,e, oxi ey® Z® Kai up,eic< Z^aexe) is full of Greek future tenses, but there are no prefixes in Gothic, not even the ever-present ga-, and the contrast between the present and the future is not reflected in any way.
3.3 Explanation
All the facts - that there are no other instances of duginnan + infinitive having any Greek original meaning anything else but 'to begin doing something' and that in other, very similar instances different translations were made - point to the conclusion that faginon duginna or gaunon jah gretan duginnid do not have specific durative connotations or semantics21.
But the key to the answer of why haban was not chosen for Ph. 1:18, like it was for Cor 11:12, may lie in Mk 14:72. Kai enipaX©v eK^aiev does not literally mean 'and began to weep', but it should be noted that the Gothic greitan is a verb of emotion, just like gaunon and faginon.
Another interesting feature of both cases is the word order, where duginnan stands in the final position. The normative order in Gothic
20 The discussion on the Gothic 'perfective' prefixes is rather large and old; here it will suffice to say that Scherer (1954), West (1981), and Coleman (1996) have proven that the situation is very different from Streitberg's perfektives Futurum.
1 Coleman (Coleman 1996: 14) calls these instances 'minor experiments'.
seems to be verb + bare infinitive, and not the other way around22, so this word order may be marked. If that is so, then presumably, with words of emotion, the infinitive + duginnan may be used in some sort of an inchoative construction.
4. haban + infinitive
These are the three instances:
Jn 12:26 jah parei im ik, paruh sa andbahts meins wisan habaip ('and where I am, my servant will be'; Gk Kai onou eipi éy® éKeî Kai o ôiâKovoç o épôç; eaxai)
2 Cor 11:12 ip patei tauja jah taujan haba ('I do that and will do'; Gk o ôè noi® Kai noi^a®)
2 Thes 3:4 jah taujip jah taujan habaip ('and you do and will do'; Gk [Kai] noieîxe Kai noi^aexe).
The OE Bible has min pen bid pœr pœr ic eom for Jn 12:26; in Tatian, the passage reads: thar ih bin thar ist min ambaht. In the other two instances, the context is identical, as is the verb taujan; and the meaning is clearly 'I do and will continue to do', and 'you do and will continue to do'.
4.1 Other examples
In the Gothic Bible, there are 6 situations of haban + infinitive, but in three of them, the meaning is simply 'to have [money] to recompense' (Lk. 14:14), 'to obtain [means] to share with those in need' (Eph 4:28), and 'to have power to take' (Jn 10:18).
The remaining three instances, however, all translate the Greek construction with piAA®: Jn 6:6 patuh pan qap fraisands ina; ip silba wissa patei habaida taujan ('And he said this to prove him, for he himself knew what he would do'; Gk xoùxo ôè è'Xeyev neipâÇœv aùxov, aùxôç yàp flôei xi epeAAev noieîv); Jn 6:71 qapuh pan pana ludan Seimonis, Iskariotu, sa auk habaida ina galewjan, ains wisands pize twalibe ('He spoke of Judas Iscariot the son of Simon: for he it was that would betray him, being one of the twelve' ; Gk è'Xeyev ôè xôv 'Iovôav Ei^œvoç 'IaKapiéxou ovxoç yàp epeAAev napaôiôovai aùxov, elç [®v] éK t®v ôéôeKa); Mk 10:32 jah andnimands aftra pans twalif dugann im qipan poei habaidedun ina gadaban ('And he took again the twelve, and began to tell them what
22 There is, strangely, very little research on that particular issue; in the latest available work, Th. Eythorsson (1995) treats bare infinitives as regular objects, which does not seem to be a correct approach.
things would happen to him'; Gk Kai napaAap®v naAiv xou< S®SeKa ^p^axo auxoi« Aeyeiv xa piAAovxa aux® auppaiveiv). They are, however, all in the past tense, and that distinguishes them from the original instances.
Interestingly, two more examples (also in past tense) are found in Skeireins: Sk 1:3 pata nu gasaihands iohannes, po sei ustauhana habaida wairpan fram fraujin 'John now seeing this, the plan which was to be carried out by the Lord...', and Sk 2:7 ip nasjands pana anawairpan dom is gasaihands jah patei in galaubeinai pei(han) habaida, gaskeirjands imma, swe mippan unkunnandin, qipands 'But the Savior, seeing his future judgment, and that he was to grow in faith, explaining to him as to one still ignorant, saying...'.
The debate on whether Skeireins is a translation is not closed, but so far, these parts of the text have not been pointed out as direct translations. In his seminal paper on the subject, Knut Schaferdiek (Schaferdiek 1981: 187) states that Skeireins seems to be a 'Katenenuberlieferung' of Theodore from Heraclea's commentary to John's Gospel, but it is unclear if this means that both examples have a real underlying p,eAA®-construction. If they do not, that might mean that haban + infinitive developed at least into a kind of predetermined future-in-the-past.
4.2 Explanation
The examples from Skeireins and translations of Greek +
infinitive are exactly what they are supposed to be: expressions of a planned future chain of events. The only question is why haban was chosen in these very cases: there are plenty of + infinitive in
the Greek New Testament (see below), and only three of them were translated with haban.
In the cases where haban + infinitive translates Greek future, the logic may be very similar: tauja jah taujan haba, along with taujip jah taujan habaip means not only 'do and will continue to do', but 'do and plan to continue to do', which would call for a construction with this semantics. Still, Jn 12:26 calls for an explanation; the context is, rather, that of the imperative dKoAou0eix© earlier in the same sentence. Another difficulty is presented by the three future-in-the-past-like instances; note that all three are preceded by qipan (Gk Aey®), and all three concern prophetic knowledge. The Greek infinitive is seen in this position (after Aey®, and with a pre-knowledge of the future) in several other instances, such as Jn 7:39, and especially Jn 12:33, Jn 18:32 and Lk 9:31 - and all of them were
translated with the help of Gothic skulan, not haban. There is no clear difference between these two groups; one could conjecture that the Greek tense of p,eXX® is different (ep,eXXev vs ^p,eXXev in three of the latter examples), but Jn 7:39 contains ep,eXXov. There is no syntactic difference between them; nor is there anything special in the infinitives that follow.
4.2.1 The case of p.kXXw.
In Koine, ^eXXro + infinitive is a phrase undergoing a rapid grammaticalisation23 on its way to the analytical verb used to form the immediate future in modern Greek. This unclear - and, perhaps, more analytical - status of ^eXXro could be reflected in the attempts to translate it into Gothic. Note that we are now talking not about Streitberg's duratives Futurum and its instances, but about a broader context that may shed some light on the choices made by the translator (Wulfila?) and allow for a better understanding of those choices.
As was already listed above, there are only 20 instances of ^eXXro + infinitive translations in the surviving parts of the Gothic text. 9 of them were translated with skulan or skulds ist, 4 - with munan, 3 -with haban, 2 - with anawairp + wesan, 1 - with sik skaftjan and 1 -with a single noun, swultawairpja. This is very different from other similar constructions like the ones with Set and o^eiX® that are all but one translated with skulan/skulds ist, and it certainly means that both Greek and Gothic allowed for a wide variation and interpretation of this construction. The 'future--in-the-past' -like constructions with haban as well as the weird past tense of skulan in Mt 11:14 all have ^eXXro + infinitive as their source. This is scarce; but this allows for some speculation.
It is quite clear that in case of obvious words with obvious meaning, the obvious solution would be to find an analogous lexeme and use it throughout the text with few exceptions (perhaps, when the language does not permit a certain usage or there is a much better
23 Even in Classical times, piXXro + infinitive was used as a 'voluntative future' (see Schwyzer, Delbrunner 1950: 293-294). Basset goes even further and states that piXXro generally behaves as an auxiliary verb and is never autonomous even in Homer (Basset 1979: 93), and that piXXro + inf. praes. can always be translated as 'it's probable that...' (ibid., 104). In Hellenistic-Roman Greek, up to the 4th cent. AD, piXXro + infinitive 'made small steps towards a more grammatical status' (Markopoulos 2009: 59), such as developing a deontic meaning in the 1st cent. AD (ibid., 48).
expression at hand). We see that in the cases of Sei and o^eiA®. But when the meaning of a word cannot be simply translated - and especially when its function and even the grammar category it is used to express do not exist in the target language, a good option would be to look for synonymous constructions in each separate case. But this would mean two possibilities: a) bending the words of the target language into something that looks similar and creating a calque; and b) looking for phrases that exist in the target language and have a clear meaning. Naturally, both possibilities could be realized in the course of one translation, so the difficulty would be to see which is which.
It would be simple to translate ^peAAev xeAeuxav with a clause, but the translator chose swultawairpja; here he found something that already existed in Gothic. If we are to assume that haban + infinitive means something specific in Gothic, we must find evidence that this is not a calque.
Typologically, there would be nothing strange in such a construction (though Harbert 2006: 299 calls it 'striking'): it exists, most notably, in Romance languages, where je ferai derives from facere habeo24, and, though with a more specific meaning, is present in Slavic languages (Russ. vot chto ja imeju skazat - 'I'm going to say something now', literally, 'here's what I have to say'), and in Albanian25. Fascinatingly, Greek's own e%® eineiv etc. is never translated with haban, but only with skulan.
Haban + infinitive in Gothic exists in two versions, like in Latin: with the infinitive in ante- and postposition. In the Bible text, all present tenses have the infinitive in anteposition, and all the past tenses have it in the postposition. But in Skeireins 2:7 we read jah patei in galaubeinai pei(han) habaida. This may be brought about by the object in galaubeinai in preposition; we have no way to verify this hypothesis. The consensus on the Gothic word order is that it was shifting from OV to VO (Lehmann 1994: 34-36; further corroborated by Tamasauskaite 2013), so it could be relatively free.
In Latin, it seems that the order habeo + infinitive was normative and usually had deontic or epistemic meaning, while the marked order infinitive + habeo acquired a secondary future meaning, and later became the new future of the Romance languages (Adams
24 And, as shown by Adams 1991, infinitive + habeo had a deontic modal meaning.
25 Sources like Neumark 1982: 50 list kam + infinitive as an exlusively Gheg form.
1991: 161). It is hard to establish if the same were true for Gothic, because the examples are scarce, and there is little research on this specific word order, but it may also be marked.
It is unlikely if the future semantics of the infinitive + haban was drawn on Latin: the Latin meaning is well attested in Tertullian and other post-classical writers26; if it was, then the Gothic translation has borrowed much more from Latin than we could ever know: in none of the instances where the Gothic Bible has haban + infinitive, the Vulgate has anything like habeo + infinitive27, so the borrowing would have come through oral medium. There are several instances, though, when Greek ^eXXro + infinitive is translated by habeo + infinitive in the Latin Bible, but in those cases Gothic uses other means in all those cases (such as munan), so the meaning of habeo + infinitive must be different from that of haban + infinitive. Even if the construction indeed came to Gothic via language contacts, it must have acquired a very different meaning and started functioning on its own.
Overall, it seems possible that we are dealing with an authentic Gothic expression 'infinitive + haban'28, which may or may not be distinct from 'haban + infinitive' and has a modal, but also clearly a futuristic meaning of a 'planned future'.
26 This was described by Coleman in 1971, and he does not think this to be an areal (African) phenomenon (Coleman 1971: 226), but finds roots for this construction in Classical Latin.
27 Jn 12:26 erit, 2 Cor 11:12 facio et faciam; 2 Thess 3:4 facitis et facietis, Jn 6:6 esset facturus, Jn 6:71 erat traditurus, Mk 10:32 essent ventura; the Vulgate does have such examples, cf. Mk 10:38 baprizari habeo, or Jn 16:12 habeo dicere etc. Curiously, though the Latin version of Jn 16:12 is a literal translation from Greek, in Gothic it is translated as skal qi^an, not haba qi^an - and none of those instances were translated as haban + infinitive. Ronsch (Ronsch 1875: 447-449) gives a long list of habeo + infinitive in the Vulgate; many of these are literal translations of Greek e'xro + infinitive, but some (like Jn 8:22) can be seen as the basis for the Romance future.
28 Coleman writes, 'For unlike skulan and munan + infinitive, which had other functions also, haban + inf. had none and was therefore available to repair this deficiency in the tense system' (Coleman 1996: 22), and this may be true, though it seems too rare for something that 'repairs a deficiency'.
5. Conclusion
Of all three constructions, duginnan + infinitive seems to be a situational choice made to describe a presupposed state of emotions that does not reflect any specific Gothic way of describing the future.
Skulan + infinitive is a modal construction undergoing some grammaticalization as such, but not a future-related construction, at least, at the time when the Bible was written.
Only haban + infinitive, or, perhaps, its reverse version, infinitive + haban, looks like a genuine construction with a futuristic meaning. It was rare enough not to be used constantly throughout the text, but established enough to emerge in a situation that required a non-straightforward translation.
References
Adams 1991 - Adams J. N. Some neglected evidence for Latin habeo with infinitive: the order of the constituents // Transactions of the Philological Society. 1991.Vol. 89. No.2. P. 131-196. Basset 1979 - Basset L. Les emplois périphrastiques du verbe grec pé^siv. Étude de linguistique grecque et essai de linguistique générale. Lyon: Maison de l'Orient et de la Méditerranée Jean Pouilloux, 1979. Bauschatz 1982 - Bauschatz P. C. The Well and the Tree: World and Time
in Early Germanic Culture. The Univ. of Massachusetts Press, 1982. Blackburn 1892 - Blackburn F. A. The English future; its origin and
development. Leipzig-Reudnitz, 1892. Coleman 1971 - Coleman R. The Origin and Development of Latin Habeo+Infinitive // The Classical Quarterly New Series. 1971. Vol. 21, No. 1. P. 215-232.
Coleman 1996 - Coleman R. Exponents of Futurity in Gothic // Transactions of the Philological Society. 1996. Vol. 94, No. 1, P. 1-29. Diewald, Wischer 2013 - Diewald G., Wischer I. Markers of Futurity in Old High German and Old English: A Comparative Corpus-Based Study // Comparative Studies in Early Germanic Languages. John Benjamins Publishing Co, 2013. P. 195-216. Eythorsson 1995 - Eythorsson Th. Verbal syntax in the early Germanic
languages. Diss. Cornell University, 1995. Ferraresi 1998 - Ferraresi G. Die Syntax des Infinitivs im Gotischen: die Modalverben skulan und magan, in J. Strässler (Hrsg.), Tendenzen europäischer Linguistik. Akten des XXXI. Linguistischen Kolloquiums, Bern, Niemeyer, Tübingen, 1998. P. 46-49. Goodwin 1890 - Goodwin W. W. Syntax of the Mood and Tenses of the
Greek Verb. Boston, 1890. Gukhman 1958 - Гухман М. М. Готский язык. М., 1958. Harbert 2006 - Harbert W. The Germanic Languages. Cambridge University Press, 2006.
Hewson 1997 - Hewson J. with Bubenik V. Tense and Aspect in Indo-European Languages. Theory, Typology, Diachrony. John Benjamins Publishing Co, 1997. Leiss 2012 - Leiss E. Aspectual patterns of cover modality in Gothic and Old High German // Covert patterns of modality. Abraham W. and Leiss E. ed. Cambridge Scholars Publishing, 2012. Lehmann 1994 - Lehmann W. P. Gothic and the Reconstruction of Proto-Germanic // The Germanic Languages. König E. and van der Auwera J. ed. Routledge, 1994. Lloyd 1979 - Lloyd A. L. Anatomy of the Verb: the Gothic Verb as a Model for a Unified Theory of Aspect, Actional Types and Verbal Velocity. John Benjamins Publishing Co, 1979. Markopoulos 2009 - Markopoulos T. The Future in Greek from Ancient to
Medieval. Oxford University Press, 2009. Neumark 1982 - Neumark L., Hubbard P., Prifti P. Standard Albanian.
A reference grammar for students. Stanford University Press, 1982. Rönsch 1875 - Rönsch H. Itala und Vulgata, das Sprachidiom der urchristlichen Itala und der katholischen Vulgata unter Berücksichtigung der römischen Volkssprache. Marburg, 1875. Schäferdiek 1981 - Schäferdiek K. Die Fragmente der "Skeireins" und der Johanneskommentar des Theodor von Herakleia // Zeitschrift für deutsches Altertum und deutsche Literatur. 1981. Bd. 110. P. 175-193. Scherer 1954 - Scherer P. Aspect in Gothic // Language. 1954. Vol. 30, No. 2. P. 211-223.
Schwyzer, Debrunner 1950 - Schwyzer E., Debrunner A. Griechische Grammatik. Auf Grundlage von K. Brugmanns griechischer Grammatik. Band II. Syntax und syntaxische Stylistik. Munich, 1950. Streitberg 1910 - Streitberg W. Gotisches Elementarbuch. Dritte und vierte neubearbeitete Auflage. Carl Winter's Universitätsbuchhandlung, Heidelberg, 1910.
Tamasauskaite 2013 - Tamasauskaite Z. Gothic Word Order Patterns as Attested in the Gothic Gospel of Luke // Sustainable Multilingualism. 2013. Vol. 3. P. 77-95. Wackernagel 2009 - Wackernagel J. Lectures on syntax with special reference to Greek, Latin and Germanic / Ed. by Langslow D. Oxford University Press, 2009. West 1981 - West J. Proklitische Verbalpartikel und ihr Gebrauch in Bezug auf das verbale Aspek tsystem des Gotischen // Zeitschrift für Deutsche Philologie. 1981. Bd. 100, 3. P. 331-338.
S. D. Kleyner. The Gothic Future: a Tense That Doesn't Exist
The paper focuses on the three Gothic verbs that are commonly said to help form the Gothic future: duginnan, haban and skulan. While all three sometimes help translate the Greek future, only haban + infinitive turns out to have a future-related meaning.
Keywords: Gothic, future tense, skulan, duginnan, haban
