The Stylistic Uses of Gothic Passive Constructions

This paper explores the variation between non-past (present and future) synthetic and periphrastic passive verb forms in the Gothic Gospels in an effort to evaluate the possibility that the availability of functionally identical forms of the passive was exploited by the translators of the Gothic Bible as a way of manipulating the stylistic composition of the Gothic text. Based on the evidence of the Gothic translation of the Gospels, although the Gothic synthetic passive constructions do mostly occur in stylistically special environments, the existence of other clearly verifiable competing motivations makes the stylistic motivations difficult to verify. It is concluded that the distribution of forms is largely determined by factors such as literalism as the main translation technique as well as contrasts between the synthetic and periphrastic ‘be’ passives in terms of the actionality of the former and stativity of the latter.


1.
In the older Germanic languages, the passive voice is realised periphrastically by means of the auxiliary 'be' or 'become' and the past participle of the notional verb. Gothic is the only older Germanic language, which, in addition to realising the passive periphrastically, has retained an older synthetic passive form, utilised in the non-past (present or future) side by side with a periphrastic non-past passive. In the past tense, passive constructions could only be periphrastic.
'You were born in utter sin, and you teach us?' (John 9:34) (5) akei þan sa us þiujai bi leika gabaurans was 'But the one from the handmaid was born according to the flesh' (Galatians 4:23B) Examples (4) and (5) illustrate past passives, with the past participle of the verb gabairan 'give birth' combining with preterite forms of the auxiliaries wairþan 'become' and wisan 'be'. In the present tense as well as the past, the difference between the 'become' and 'be' passives is that the former are actional passives that convey a sense of inchoativity or change of state, while the latter are stative passives (Harbert 2007: 317f. ;Miller 2019: 216ff.). In the past tense, the 'be' passives can also have an actional sense: Luke 1:65 merida wesun 'were (being) proclaimed/noised', Mark 1:5 daupidai 1 The Gothic examples are cited from Streitberg (2000). The Greek citations refer to the "Majority Text" of Robinson and Pierpont (2005) as the primary comparator and the "Critical Text" of Nestle et al. (2012) as the secondary comparator. For pre-Vulgate Latin translations I consulted Jülicher (1963)(1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976) as well as the online Vetus Latina Database (VLD). Where a pre-Vulgate witness was lacking, the Vulgate (Weber and Gryson 2007) was consulted instead. Line references to the Skeireins follow Snaedal (2013 a, b). wesun 'were (being) baptised', both of which translate Greek imperfect passives. The synthetic forms of the non-past are semantically ambiguous, as they can be either stative or actional.

2.
As is well established, the Gothic Bible represents a close literal translation of a Greek original (Friedrichsen 1926: 15;Falluomini 2015: 66ff.). 2 It is no surprise, therefore, that Gothic passive constructions typically translate Greek passive constructions. Occasionally, however, Gothic passive constructions correspond to Greek actives. For example, in Matthew 25:42 the Greek active aorist ἐδίψησα '(I) thirsted' is rendered in Gothic with the passive afþaursiþs was 'was thirsted'. The opposite is also possible. For example, in Luke 2:21 the Greek passive ἐπλήσθησαν 'were fulfilled/accomplished' is rendered with the Gothic active (class 4 intransitive) usfulnodedun lit. 'fulfilled, accomplished'. Sometimes, the Greek passive is not rendered with a voice construction, with a lexical periphrasis utilised instead. For example, in Matthew 8:3 the Greek aorist passive ἐκαθαρίσθη 'was cleansed' is translated with an adjective and a copula in Gothic: hrain warþ 'became clean'. These variations in translation technique are accounted for by a range of factors including the native linguistic constraints of Gothic, translational convenience dictated by the goal of expository transparency, lexical insufficiency and the resulting need to paraphrase, etc. 3 In individual instances, the influence of pre-Vulgate Latin translations of the Bible can be suspected.
The Gothic passives, irrespective of their realisation, are mostly matched by monolectic Greek forms. For example, the past passive laþodai wesuþ '(you) were called' in Colossians 3:15 translates the Greek aorist passive ἐκλήθητε '(you) were called'. Likewise, the non-past passives in examples (2) and (3) translate, respectively, the Greek aorist subjunctive passive ἀποκαλυφθῇ 'may be revealed' and aorist indicative passive ἀπεκαλύφθη 'was revealed', with both of these Greek passives being realised monolectically. However, while in the past tense the only available passive forms in Gothic were periphrastic (because a synthetic past passive had never existed-see Ringe 2017: 194;Fulk 2018: 283), the availability of choice between the inherited synthetic and new periphrastic passives forms in the non-past raises the question of how the choice was motivated in translating monolectic Greek forms.
In a study of variation between uns 'us' and unsis 'us'-competing forms of the first-person personal pronoun weis 'we' in the dative and accusative plural- Snaedal (2010: 313) finds that the choice of forms was motivated stylistically: "It can be as-2 In Ratkus (2020), I offer an overview of the principal problems related to the authorship and properties of the Gothic translation, with further references. 3 For a detailed discussion of the motivations behind Greek active constructions being rendered as passives in Gothic, see Ratkus (2020). sumed that unsis was the more formal and ceremonious (not to say pompous) variant of the two. The use of uns in the Lord's Prayer may be due to the wish to make it more colloquial and intimate." In my own study of the competition between the bare-stem and -ata forms of singular nominative and accusative neuter possessives pronouns, adjectives and other modifiers (Ratkus 2015), I conclude that modifiers augmented with -ata were stylistically charged forms invoked in contexts of second-person address or contexts that warranted an elevated or reverential tone. The availability of these variant forms enabled the Gothic translators to imbue the Gothic text with contextual stylistic nuances or contrasts that were unavailable in the original Greek (and Latin). With the periphrastic passive encroaching upon the synthetic passive of the non-past, the variation between the competing forms was, in all likelihood, correlated with a stylistic contrast along similar lines. Examples (6) -(10) illustrate a selection of present-tense periphrastic passive forms. In 1 Corinthians 7:14 (6) both the stative passive weihaida ist 'is sanctified' and the resultative passive gaweihaids ist 'is sanctified; has been sanctified' refer to the Greek perfect (medio)passive ἡγίασται 'is sanctified'. The different Gothic aspectual realisations attest to the translator's effort in drawing subtle semantic distinctions. In 2 Cor-inthians 7:4 usfulliþs im 'am filled' and ufarfulliþs im 'am overfilled; am exceedingly full' translate, respectively, Greek perfect (medio)passive πεπλήρωμαι '(I) have been filled' and present (medio)passive ὑπερπερισσεύομαι '(I) overabound'. In 1 Thessalonians 5:7 (8) the passives drugkanai wairþand 'get/become drunk' translate different Greek forms: in the first instance of the Gothic passives, the Greek substantivised passive participle οἱ μεθυσκόμενοι 'the (ones) being drunk' is rendered with a finite passive clause; in the second instance, the Gothic passive clause translates the Greek active μεθύουσιν '(they) drink to intoxication; get drunk'. In Galatians 3:27 (9) the passive gahamodai sijuþ '(you) are clothed' translates the Greek aorist middle ἐνεδύσασθε '(you) put on, clothed'. Finally, in 2 Timothy 2:9 (10) the Gothic passive nist gabundan 'is not bound' corresponds to the perfect (medio)passive οὐ δέδεται 'is not bound; has not been bound' in the Greek.
Irrespective of the grammatical details of the Greek forms that the Gothic passives in (6)-(10) translate, all of the above Gothic periphrastic present forms, whether stative (6), (7), (9), (10) or change-of-state (8), refer to monolectic items in the Greek. Also, and significantly, the contexts in which the Gothic periphrastic passives occur are neutral with regard to the contextual properties that necessitate the choice of stylistically charged forms as outlined above with reference to variation between uns 'us' vs. unsis 'us' as well as the bare-stem vs. -ata. The examples neither occur in a context of direct address, nor are they solemn declarations or exhortations.
(11) bi usbeisnai jah wenai meinaim, unte ni in waihtai gaaiwiskoþs wairþa, ak in allai trauainai, swe sinteino jah nu, mikiljada Xristus in leika meinamma, jaþþe þairh libain jaþþe þairh dauþu 'According to my long-suffering and hope that in nothing I will be(come) ashamed, but that will all confidence, as always so now also, Christ shall be magnified in my body, whether by life or by death' (Philippians 1:20) Philippians 1:20 in (11) contains two passive constructions, which translate, respectively, the Greek future passive αἰσχυνθήσομαι '(I) will be shamed/put to shame' and future passive μεγαλυνθήσεται 'will be magnified'. Although, similar to the examples in 6-10, both Greek forms are monolectic, the difference in the Gothic translation is that the first one is rendered with a periphrastic passive, while the second one is given a synthetic form. Since, in this case, there is no grammatical difference between the Greek forms that might account for the different Gothic forms, the choice of form is entirely within the discretion of the Gothic translator. An examination of the context of Philippians 1:20 shows the verse to be composed of two conceptual halves. In the first half ('According to my long-suffering and hope that in nothing I will be(come) ashamed'), the apostle alludes to the circumstances of his situation (as a prisoner in Rome), the yearning of a positive outcome from his affliction and the possible risk that he might be put to shame by denying the Christian faith or his work for God. Set against the background of the first half, the second half of Philippians 1:20 issues a proclamation that the apostle is eager and determined to glorify God, whether through continuing his ministry in life or by remaining faithful even when put to death. Thus, the first half of the verse builds towards the noble message of the second half. The elevated tone of the exhortation mikiljada Xristus 'Christ shall be glorified' is what justifies the use of the older (inherited) synthetic passive form. By contrast, the periphrastic passive in the first half of the verse has reference to the apostle rather than Jesus and is therefore given a neutral periphrastic rendering. Thus it is evident that the availability of different passive forms in Gothic enabled the Gothic translator to contribute a degree of stylistic expressiveness to a context linguistically undifferentiated in the original Greek.
Below, I examine the variation between (non-past) synthetic and periphrastic passive forms in the Gothic translation of the Gospels in the order they are printed in Streitberg's (2000) edition of the Gothic texts (Matthew, John, Luke, Mark).

3.
The Gothic Gospel of Matthew begins with an incomplete Chapter 5, which contains nine passive constructions. Of these, six verses (21,27,31,33,38,43) illustrate the formula qiþan ist 'is said', attested mostly in Matthew (including encliticised qiþanuh in verse 31), as well as twice in Luke (2:24; 4:12) and once in Romans 9:12. This "timeless formula" (Miller 2019(Miller : 2017, as well as the formula gameliþ ist (or ist gameliþ) 'is written' (Romans 14:11, Luke 2:23, John 6:45, etc.), are used to introduce what is perceived as an established or doctrinal truth, or a precept to follow. These formulae convey states and do not, as a rule, occur in the synthetic form. 4 The remaining three passives in Chapter 5 are synthetic forms. Matthew 5:19 attests two passive forms of haitan 'call': minnista haitada in þiudangardjai himine '(he) shall be called the least in the kingdom of heaven' and sah mikils haitada in þiudangardjai himine 'This (man) shall be called great in the kingdom of heaven'. In both instances, the passive haitada is not a naming verb, but an actional passive form: 'shall be referred to as'-see Ratkus (2020) for a more detailed discussion 4 The only synthetic passive attestation of qiþan 'say, tell, speak' occurs in Romans 9:26: jah wairþiþ in þamma stada þarei qiþada im: ni managei meina jus, þai haitanda sunjus gudis libandins 'And it shall happen in the place where it is declared to them: You are not my nation; there they shall be called the sons of the living God.' In this verse qiþada 'is said' is not stative, but clearly actional, as an act of communication follows it in the form of direct address. (Interestingly, in translating the Greek aorist passive ἐρρέθη 'it was declared' the Gothic translator chooses a synthetic present form, rather than a periphrastic past form. The present-tense periphrastic formula qiþan ist 'is said' likewise translates the aorist passive ἐρρέθη.) The passives qiþada 'is declared' and haitanda 'shall be called' occur in a context of solemn appeal, which is likely the motive behind the choice of the synthetic forms. of haitan 'name, call, hail' as a synthetic passive. Finally, Matthew 5:25 has jah sa staua þuk atgibai andbahta, jah in karkara galagjaza 'And (lest) the judge should deliver you to the officer, and you should be cast in prison'. It is noteworthy that both verses occur in a context of Jesus teaching his disciples or preaching.
Chapter 6 of Matthew contains six (optative) passives: 6:2 hauhjaindau 'may be glorified', 6:5 gaumjaindau 'may be seen/observed', 6:7 andhausjaindau 'shall be heard', 6:16 gasaiƕaindau 'may be seen/perceived (as)', 6:18 ni gasaiƕazau 'would not be seen/perceived (as)'. Chapter 7 contains three synthetic passive constructions, including two passives in 7:19 usmaitada jah in fon atlagjada 'is/shall be hewn down and shall be cast into fire', followed by one synthetic passive in 7:26 galeikoda mann dwalamma 'shall be likened unto a foolish man'. All of these examples occur in the language of Jesus teaching or preaching, including contexts of direct address. The first significant instance of variation occurs in Matthew 8.
(12) iþ þai sunjus þiudangardjos uswairpanda in riqis þata hindumisto; jainar wairþiþ grets jah krusts tunþiwe 'But the children of the kingdom shall be cast out into outer darkness; there shall be weeping and gnashing of teeth' (Matthew 8:12) (13) jah sai, wegs mikils warþ in marein, swaswe þata skip gahuliþ wairþan fram wegim; iþ is saislep 'And, behold, there arose a great storm in the sea, so that the ship was covered with the waves; however, he was asleep' (Matthew 8:24) Matthew 8:12 in (12) illustrates a synthetic passive uswairpanda 'shall be cast out' in a context of Jesus addressing his disciples. In contrast, the periphrastic passive gahuliþ wairþan 'become/get covered' in (13) occurs in a context spoken by the narrator (Matthew 8:24). Thus, by manipulating the synthetic and periphrastic passive forms the Gothic translator is able to weave stylistic contrasts into the text, with the elevated language of Jesus being brought out against the neutrality of the narrative.
Matthew 10 contains five non-past passives (including both synthetic and periphrastic), of which three occur in the same verse.
(14) ni nunu ogeiþ izwis ins; ni waiht auk ist gahuliþ, þatei ni andhuljaidau, jah fulgin, þatei ni ufkunnaidau 'So, let them not terrify you; for there is nothing covered that shall not be revealed, and hidden that shall not be made known' (Matthew 10:26) Matthew 10:26 (14) opens with the periphrastic passive ist gahuliþ 'is covered', followed by two synthetic passives: ni andhuljaidau 'shall not be revealed' and ni ufkunnaidau 'shall not be recognised/made known'. Since the verse illustrates Jesus directly addressing his disciples and followers, the use of the periphrastic form is unexpected. However, a look at the Greek makes it clear that the choice of form in Gothic need not be confined to stylistic motivations. In particular, because the Greek also has a periphrastic form ἐστιν κεκαλυμμένον 'is covered' followed by monolectic passives ἀποκαλυφθήσεται 'will be revealed' and γνωσθήσεται 'will be known', literalism as the principal translation technique overrides any stylistic considerations, and the Greek periphrastic structure is given a periphrastic translation in Gothic. In Matthew 10:29 niu twai sparwans assarjau bugjanda 'Are not two sparrows bought for a penny?' the Gothic synthetic passive translates a monolectic Greek form and is stylistically in line with the context of preaching and direct address. The final -periphrastic -passive in this chapter, Matthew 10:30 aþþan izwara jah tagla haubidis alla garaþana sind 'But the very hairs of your head are all counted', again strays from the stylistic criterion of direct interaction, as it translates the Greek periphrastic form ἠριθμημέναι εἰσίν lit. 'numbered are'.
Matthew 11 has five (non-past) passive forms. The synthetic forms wailamerjanda 'are (having the Gospel) preached' in 11:5 and ni gamarzjada 'shall not be offended/ made to stumble' in 11:6, uttered by Jesus in his reply to John's disciples, are stylistically justified. Two verses later, in 11:8, the translator switches to a periphrastic passive. Although Matthew 11:8 clearly continues the context of Jesus addressing John's disciples and, after they have left, the crowd, the invocation of the periphrastic passive wasidai sind 'are clothed' seems out of line with the tone of the address, additionally marked by the elevated use of an -ata participle in 11:7 (raus fram winda wagidata 'A reed shaken by the wind?'). This periphrastic form is even more surprising in light of the existence of gawasjan* 'clothe, dress' in the synthetic passive form. This sole attestation of the passive gawasjada 'is clothed' in 1 Corinthians 15:54 A shows that there was nothing wrong in principle with synthetic passivisation of gawasjan*. A side by side comparison of the passives wasidai sind 'are clothed' (translating Greek active present participle φοροῦντες 'wearing, carrying') and gawasjada 'is clothed' (translating Greek subjunctive aorist middle ἐνδύσηται 'shall have put on') highlights a difference discussed above with reference to the well attested stative passive qiþan ist 'is said' and the actional passive qiþada 'is said/declared', attested only in Romans 9:26. In particular, the periphrastic form wasidai sind 'are clothed' is a stative passive, while the synthetic gawasjada 'is clothed' is clearly actional. This difference points to an additional motive behind the distribution of Gothic passive forms, namely that the synthetic passives were not only stylistically distinctive, but at the same time preferred in the actional sense. If this was indeed the case, then the stylistically neutral 'be' passives were a convenient means of disambiguation invoked where the stative sense required an unequivocal realisation.
The remaining four passives in the Gothic Matthew include the formula gameliþ ist 'is written' in 11:10, and three synthetic forms: anamahtjada 'is taken by force' (lit. 'is overpowered') in 11:12, atgibada 'is/will be given away/betrayed' in 26:2 and haitada 'is called' in 27:17. The former two occur in the language of Jesus, while the latter is uttered by Pontius Pilate in a direct address to a crowd. Similar to haitada in Matthew 5:19 (see above), this passive in Matthew 27:17 is not a static naming construction with a sense of 'his name is X', but an actional form that means 'is hailed/addressed as'.

The Gothic
Gospel of John opens with four (non-past) periphrastic passives in Chapter 6. John 6:31 attests the formula ist gameliþ 'is written', prompted by Greek ἐστιν γεγραμμένον 'is (having been) written', as is evident from the ordering of the auxiliary and the participle. 5 John 6:45 attests two periphrastic passives: ist gameliþ ana praufetum: jah wairþand allai laisidai gudis 'It is written in the prophets: and they shall all be taught by God'. In both instances, these passives mimic the corresponding Greek structures ἔστιν γεγραμμένον 'is (having been) written' and καὶ ἔσονται πάντες διδακτοὶ 'shall be all taught'. What is clear, however, is that the Gothic translation of these Greek passives draws a distinction between stativity and actionality (change of state), with the first form rendered with a 'be' passive and the second with a 'become' passive. The final passive in this chapter, John 6:65 ist atgiban 'is given' is likewise Greek-prompted (ᾖ δεδομένον 'may be (having been) given'). 5 In natural Gothic, auxiliaries and copulas appear clause-finally (Miller 2019: 502, 504, 508ff.).
John 7:8 unte meinata mel ni nauh usfulliþ ist 'For my time is not yet fulfilled', uttered by Jesus in reply to his brethren, has an elevated tone, as suggested by the use of -ata. 6 This notwithstanding, the Gothic periphrastic passive (which translates the Greek perfect passive πεπλήρωται 'has been fulfilled') is not given a synthetic realisation. The translator's insistence on the stativity of the passive evidently overrides any stylistic considerations. The verb usfulljan 'fulfil, complete', well attested as a synthetic 3rd person passive usfulljada (Galatians 5:14, 2 Corinthians 10:6, Luke 3:5, Romans 13:9), not only occurs in contexts of prophetic declaration, solemn or authoritative tone, but also conveys an actional sense or one of change of state. Unsurprisingly, therefore, in John 7:23 ei ni gatairaidau witoþ þata Mosezis 'So that the law of Moses would not be broken', with Jesus authoritatively replying to the multitude, the passive is actional and therefore synthetic. In John 7:47 ibai jah jus afairzidai sijuþ 'Are you also deceived?', asked by the Pharisees, the periphrastic passive translates Greek perfect passive πεπλάνησθε 'have been deceived' and is stative. The final passive construction, fraqiþanai sind 'are cursed' in 7:49 is not only stative, but also an accurate rendering of the Greek ἐπικατάρατοί εἰσιν lit. 'accursed are', with an adjective and a copula.
The formulaic stative gameliþ ist 'is written' in John 8:17 is followed by a stative passive in 8:41 weis us horinassau ni sijum gabauranai 'We are not born of adultery', with the Gothic periphrastic passive predictably translating a Greek perfect passive οὐ γεγεννήμεθα lit. 'not were born'. In John 9:7 gagg þwahan in swumsl Siloamis, þatei gaskeirjada insandiþs 'Go and wash in the pond of Siloam, which is clarified/explained as Sent' (intended meaning: 'Which means/stands for Sent') Jesus addresses a blind man. The Gothic synthetic form translates the Greek present mediopassive ἑρμηνεύεται 'is translated/explained'. The putative actional value of the Gothic synthetic form is, however, not apparent. Gaskeirjan* 'clarify, interpret' is attested in a stative (periphrastic) passive construction ist gaskeiriþ 'is explained' three more times (Mark 5:41;15:22;15:34), which translates the Greek periphrastic passive ἐστιν μεθερμηνευόμενον 'is (being) translated/interpreted'. The periphrastic passive uswahsans ist 'is grown' (intended meaning: 'is of age') in John 9:21 and 23 translates the Greek transitive construction ἡλικίαν ἔχει lit. 'maturity/age has', with an accusative noun and a transitive verb. The Gothic construction is therefore not merely a stative periphrastic passive, but also a lexical periphrasis of the Greek construction. The formula ist gameliþ 'is written' in John 10:34 is a verbatim translation of the Greek ἔστιν γεγραμμένον 'is written'.
John 11 contains two synthetic passives. In 11:4 ei hauhjaidau sunus gudis þairh þata 'So that the Son of God might be glorified through it' the synthetic passive form 6 Compare John 7:6 mel mein ni nauh ist 'My time is not yet (come)', only two verses before 7:8, with the stylistically neutral bare stem possessive pronoun mein 'my'.
is uttered by Jesus in his reply to the sisters of Lazarus. In contrast John 11:16 þanuh qaþ Þomas saei haitada Didimus þaim gahlaibam seinaim 'Then spoke Thomas, who is called Didymus, to his companions'. In this verse the Gothic relative clause with a passive verb translates the Greek substantivised passive participle ὁ λεγόμενος 'the (one) being said'. Importantly, Didymus is not merely a name, but a Greek translation of the Hellenized Aramaic name Thomas (Ridderbos 1997: 392, 646), both Thomas and Didymus meaning 'twin'. Thus, the sense of the verse in Greek is 'Thomas, who is called the twin, said to his co-disciples' (Brodie 1993: 101). In the Gothic version, the translator makes it clear that Thomas was alternatively referred to as Didymus. Thus, similar to the examples of the passive haitada discussed above, haitada in John 11:16 is not a naming verb, but an actional passive with the sense of 'was hailed/referred to as'. Conveyance of the actional meaning of the verb overrides any stylistic considerations, justifying the synthetic form in a neutral context spoken by the narrator.
Of the ten passive constructions in John 12, four are non-past. Verse 12:14 attests the formula ist gameliþ 'is written', in imitation of the Greek ἐστιν γεγραμμένον 'is written'. Verses 12:23, 31 and 32 attest three synthetic passives: 23 sweraidau 'should be glorified' (Greek subjunctive aorist passive δοξασθῇ 'should be glorified'), 31 uswairpada 'shall be driven out/overthrown' (Greek future passive ἐκβληθήσεται 'shall be cast out'), 32 ushauhjada 'am lifted up' (Greek subjunctive aorist passive ὑψωθῶ '(I) may be lifted/raised up'). This sequence of three synthetic forms is uttered by Jesus in an episode of Jesus discussing his death and addressing God and the crowd. The first non-past passive in John 13 occurs in verse 31 (17), side by side with a 'become' past passive. This example departs from the Greek in two ways. Because both passive constructions refer to Greek aorist passive ἐδοξάσθη 'is esteemed/glorified' (repeated twice in the Greek), Gothic engages two lexically different periphrases. 7 More 7 It is possible, though not conclusively verifiable, that the Gothic lexical variants refer to some of the variants found in pre-Vulgate Latin. The Latin translations (Jülicher 1963: 155) attest four variants (clārificātus est 'is made illustrious', honōrificātus est 'is honoured', h[onōrā]tus est 'is honoured/respected', glōrificātus est 'is glorified') in rendering the first ἐδοξάσθη, and three variants (clārificātus est 'is made illustrious', honōrificātus est 'is honoured', glōrificātus est 'is glorified') that render the second ἐδοξάσθη. With the only exception of Manuscript a (Vercellensis s. IV/V), which attests honōrātus est … honōrificātus est, the individual Latin manuscripts engage lexically identical passives. It is therefore not very likely that the Gothic variants are due to Latin. It is more likely that the Gothic variants are a result of avoiding lexical repetition (i.e. lexical diversification) in close contexts-for instance, see Stutz (1966: 78), Falluomini (2015: 82). importantly, Gothic engages different tenses in translating the same Greek form, with a perfective past passive gasweraids warþ 'got/became honoured' followed by a nonperfective present passive hauhiþs ist 'is exalted/lifted up'. This stative passive, as well as the same passive in the following verse (13:32 hauhiþs ist 'is exalted') is in contrast to the synthetic form in John 14:13 hauhjaidau 'may be exalted/glorified'.
(18) jah þatei ƕa bidjiþ in namin meinamma, þata tauja, ei hauhjaidau atta in sunau 'And whatever you ask in my name, I will do that, so that the Father may be exalted in the Son' (John 14:13) In John 14:13 (18) the synthetic optative passive hauhjaidau 'may be exalted/glorified' translates the Greek subjunctive aorist passive δοξασθῇ 'should be glorified'. The difference between the periphrastic forms in John 13:31-32 and the synthetic form in 14:13 (with all three uttered by Jesus) is not merely that the former are stative passives and the latter is actional, but that the periphrastic forms are stylistically neutral declarative statements, while clause with the synthetic form in 14:13 has a sense of solemn promise. The same solemn tone seems to define John 14:21 Saei habaid anabusnins meinos jah fastaiþ þos, sa ist saei frijoþ mik: jah þan saei frijoþ mik, frijoda fram attin meinamma 'He who has my commandments and keeps them, he is one who loves me: and he who loves me shall be loved by my father'. In this verse, the passive clause functions as a solemn edict. In view of the elevated tone that defines John 14 as a whole, it is unsurprising that this chapter attests the stylistically distinctive pronoun unsis 'us' (in John 14:8-9, 22) as well as -ata (John 14:26 allata 'all').
John 15 attests five non-past passives, including a string of three in 15:6: uswairpada 'is cast out', galisada 'is gathered', inbrannjada 'is burned'. In this verse Jesus lays out the plight of a man who does not abide in Christ by comparing him with a withered vine that is cast out, gathered and burned. The passive galisada 'is gathered' departs from the Greek in that it translates an active transitive Greek verb συνάγουσιν αὐτὰ '(they/men) gather them', with the object 'them' eliminated in the Gothic passive construction. 8 As a whole, the string of three passives conveys a clear actional sequence. In contrast, John 15:8 hauhiþs ist 'is exalted/glorified' is a stative that translates the Greek aorist passive ἐδοξάσθη 'is esteemed/glorified'. In John 15:11 þata rodida izwis, ei faheþs meina in izwis sijai, jah faheds izwara usfulljaidau 'I have spoken this unto you, so that my joy would be/remain in you, and your joy would be fulfilled/completed'. The synthetic form in this verse is not only an actional passive in a context of direct address, but also one of a series of precepts and pledges Jesus makes in John 15. 8 At the same time, the Greek aorist passive ἐξηράνθη 'is dried/withered' is rendered with an active in Gothic: gaþaursniþ 'dries up/withers away'.
John 16 opens with a synthetic passive in 16:1, spoken in direct address: Þata rodida izwis, ei ni afmarzjaindau 'I have spoken this unto you, so that you would not be offended/hindered'. In 16:21 the periphrastic passive gabauran ist 'is born' in iþ biþe gabauran ist barn, ni þanaseiþs ni gaman þizos aglons faura fahedai 'But when the child is born, (she) no longer remembers the anguish due to joy' translates the Greek active subjunctive aorist γεννήσῃ 'brings forth, bears' (see Ratkus 2020 for a more detailed discussion). Not only is the Gothic passive stative, but it occurs in stylistically neutral narration. Three verses later, in 16:24 bidjaiþ jah nimiþ, ei faheþs izwara sijai usfullida 'Ask, and you shall receive, that your joy may be fulfilled' a periphrasic passive is attested. The context of exhortation and direct address that defines this verse would warrant the use of a stylistically distinctive form (usfulljaidau 'may/would be fulfilled'). This is exactly the form attested in John 15:11 jah faheds izwara usfulljaidau 'and your joy would be fulfilled/completed', discussed above. Contrary to expectations, however, the attested passive form in John 16:24 is periphrastic. The motivation for the choice of the periphrastic form, no doubt, derives from the original Greek, which attests a periphrastic construction: ᾖ πεπληρωμένη 'may be filled'. Since in John 15:11 the Greek form is monolectic (subjunctive aorist passive πληρωθῇ 'may be filled'), the Gothic translator is not in violation of literalism (i.e. the principal translation technique) in choosing a synthetic passive, justified by the stylistic context of the verse. The final passive in this chapter occurs in 16:32: sai, qimiþ ƕeila jah nu qam, ei distahjada ƕarjizuh du seinaim, jah mik ainana bileiþiþ 'Behold, the hour is coming, and has now come, that each man shall be scattered to his own, and shall leave me alone'. Distahjada 'shall be scattered', translating Greek subjunctive aorist passive σκορπισθῆτε 'should be scattered', is clearly an actional passive; attested in a context of Jesus addressing his disciples, it is also contextually justified.
In John 17:10 jah meina alla þeina sind jah þeina meina, jah hauhiþs im in þaim 'And all mine are yours and yours (are) mine, and I am exalted/glorified in them' the periphrastic passive translates the Greek perfect passive δεδόξασμαι 'I have been glorified'. The resultative sense of the Greek form is predictably captured by the Gothic stative. In 17:23 ik in im jah þu in mis, ei sijaina ustauhanai du ainamma 'I in them and you in me, that they may be drawn/perfected into one' attest a periphrastic passive. Although a synthetic form (ustauhaindau* 'would be drawn/completed/perfected') would be justified in the context of the verse-with Jesus issuing this plea while praying for his followers-the Gothic translator stays faithful to the Greek, which attests a periphrastic passive ὦσιν τετελειωμένοι 'may be accomplished/perfected'. The final non-past passive in the Gothic John, attested in John 18:37 ik du þamma gabaurans im 'To this (end) I am born' translates the Greek perfect (medio)passive γεγέννημαι '(I) have been born'. As is to be expected, here the Gothic periphrastic 'be' passive captures the resultative sense of the Greek perfect.

5.
The Gothic Gospel of Luke attests by far the richest set of passive forms of the four Gospels, at least in part because Luke is preserved better than the other Gothic Gospel texts. Since many of the examples in Luke repeat the types discussed above, in what follows I address only the more important examples of variation attested in Luke.
The Gothic Luke opens with a resultative periphrastic form Luke 1:4 galaisiþs is '(you) are/have been instructed', which translates the Greek aorist passive κατηχήθης '(you) were instructed'. In 1:13 andhausida ist bida þeina 'Your prayer is/has been heard' the periphrastic form is likewise resultative, translating Greek aorist passive εἰσηκούσθη 'is heard/heeded'. This passive, uttered by an angel of the Lord, is in contrast to the passive jah ahmins weihis gafulljada 'and (he) shall be filled with the Holy Spirit' in 1:15, also uttered by the angel. The synthetic form is not only part of a series of prophetic statements, and is stylistically justified as such, but it is also actional. The same explanation applies to the contrast between the periphrastic form insandiþs im 'am/have been sent' in 1:19 and the synthetic passive usfulljanda 'shall be fulfilled' in 1:20, both uttered by the angel Gabriel.
(18) ahma weihs atgaggiþ ana þuk, jah mahts hauhistins ufarskadweid þus, duþe ei <jah> saei gabairada weihs haitada sunus gudis 'The Holy Spirit shall come upon you, and the might of the Highest shall overshadow you, to the end that the holy (one) that shall be born of you shall be called the Son of God' (Luke 1:35) Luke 1:35 (18) illustrates a sequence of two actional synthetic forms-gabairada 'shall be born' and haitada 'shall be called'-in a proclamation by the angel Gabriel. As noted above (section 3) with reference to haitada 'shall be called' in Matthew 5:19, the synthetic passive is not a static naming construction with the sense of 'X's name is', but an actional form with the senses of 'is hailed as', 'is referred to as' and the like (for more discussion see Ratkus 2020). These senses are identifiable in the sequence of synthetic passives of haitan 'call' in this chapter of Luke: Luke 1:26 haitada 'is referred to as', 32 haitada 'shall be called/addressed as', 36 haitada 'is called/said to be', 60 haitadau 'shall be named', 76 haitaza '(you) shall be called/hailed as' (as well as haitada in Luke 2:4, 23; haitazau 'you are called' 14:10, etc. The contrast between the periphrastic and synthetic passives of haitan 'call' is best illustrated by 1 Corinthians 7:22 (19), where both passive constructions refer to the same Greek aorist passive participle (repeated twice): κληθεὶς 'being called'. The Gothic renderings of this Greek form illustrate a significant difference in the grammatical properties of the Gothic periphrastic and synthetic forms, as well as the translator's tactical judgment behind using them. In particular, the periphrastic passive haitans ist 'is/has been called/summoned' refers to the result of the action of summoning, while the synthetic passive haitada 'is called/summoned' is an actional passive with a sense of a regular or repetitive present action. It should also be mentioned that the awkwardness of the Greek clauses (ἐν κυρίῳ κληθεὶς δοῦλος ἀπελεύθερος κυρίου ἐστίν lit. 'in Lord (being) called slave freeman of Lord is'; ἐλεύθερος κληθεὶς δοῦλός ἐστιν χριστοῦ lit. 'the (one) free (being) called slave is Christ's') may have been an additional motive behind the different passives, aimed at disambiguating the Greek.
Luke 3:4 opens with the formula gamelid ist 'is written', which introduces the words of the prophet Esaias (Isaiah), detailed in the following verse. Unsurprisingly, in 3:5 the actional passives usfulljada 'shall be filled' and gahnaiwjada 'shall be lowered' are synthetic. Verse 3:9 attests one more pair of actional passives in a context of direct address (to the crowd) by John the Baptist: all dalei usfulljada jah all fairgunje jah hlaine gahnaiwjada 'All valleys shall be filled, and all mountains and hills shall be brought low'. Luke 4:6 also attests a direct address, with the devil appealing to Jesus, a context whose elevated tone is marked by -ata. However, the passive unte mis atgiban ist 'For it is given unto me' attested in this verse is periphrastic, as it has a resultative sense, translating the Greek perfect (medio)passive παραδέδοται 'has been given/handed over'.
Luke 5 attests four non-past passives, including 5:20 afleitanda 'are forgiven' and 5:23 afletanda 'are forgiven', 5:35 afnimada 'shall be taken away' and 5:38 gafastanda 'will be preserved'. All of these synthetic forms are actional and occur in contexts of direct address or prophesying. In Luke 6:21, the synthetic passive ufhlohjanda '(you) will be made to laugh' translates the Greek active future form γελάσετε '(you) will laugh'. It occurs in a context of Jesus addressing his disciples and granting blessings. The verb is passivised in the Gothic translation in the interests of semantic symmetry within the verse (Ratkus 2020). The sequence of synthetic passives in Luke 6:37 ni stojaindau '(you) shall not be judged', ni afdomjanda '(you) shall not be condemned', fraletanda '(you) shall be liberated/forgiven' and 6:38 gibada 'shall be given' (2x), mitada 'shall be measured' all occur in a context of direct address and prophesying. All are actional passives, including the forms ni afdomjanda '(you) shall not be condemned', fraletanda '(you) shall be liberated/forgiven', whose future reference presents them as changes of state rather than states.
Aside from the formula gamelid ist 'is written' in Luke 7:27, all non-past passives in Luke 7 are actional synthetic forms: 7:22 gahrainjanda 'are cleansed' and wailamerjanda 'are (having the Gospel) preached', 7:23 ni gamarzjada 'shall not be offended', 7:47 af-letanda 'are forgiven' and fraletada 'is forgiven', 7:48 afletanda 'are forgiven'. All occur either in direct address (by Jesus) or as solemn proclamations. Interestingly, the synthetic forms in 7:47 afletanda frawaurhteis izos þos managons 'Her sins, which are many, are forgiven' and iþ þammei leitil fraletada, leitil frijod 'But (the one) to whom little is forgiven loves little' refer to different Greek forms, namely the perfect (medio)passive ἀφέωνται 'have been forgiven' and present (medio)passive ἀφίεται 'is (being) forgiven. In spite of the expectation that the perfect (medio)passive ἀφέωνται 'have been forgiven/pardoned' in 7:47 and 7:48 (also in Luke 5:20 and 23) should be rendered with a stative periphrastic form, it is translated with a synthetic form. It is impossible to establish with certainty whether the synthetic form has been chosen because it is stylistically more suited to the context of the verse, or because the translator prefers to see two actional forms in the interests of semantic symmetry between the clauses, or simply because literalism in translating a Greek word with a Gothic word overrides any other considerations-or perhaps all of the above. Luke 7:48 qaþuh þan du izai: afletanda þus frawaurhteis þeinos 'And he said to her: your sins are forgiven to you' is a short verse with a single passive and no semantic parallelism. Since in rendering it (as well as afletanda in Luke 5:20 and 23) the translator insists on the same synthetic form in spite of the resultative sense of the Greek perfect, it is conceivable that the synthetic passive is primarily motivated stylistically by the elevated tone of pledging forgiveness in direct address.
In contrast to the above, the Greek perfect (medio)passive δέδοται 'has been given' in Luke 8:10 is rendered with a periphrastic passive in izwis atgiban ist kunnan runos þiudinassaus gudis '(It) is given unto you to know the secrets of the kingdom of God'. It is non-actional, as it conveys a resultative sense. In 10:22 all mis atgiban ist fram attin meinamma 'All is given to me by my father' the periphrastic form atgiban ist 'is/has been given/handed over' is likewise construed as a resultative. However, in this case it translates the Greek aorist passive παρεδόθη 'was given'.
The synthetic forms 8:17 ni gakunnaidau 'shall not be revealed/(made) known', 8:18 gibada 'shall be given' and afnimada 'shall be taken away', 8:50 ganasjada 'shall be saved' are all actional and occur in contexts of prophetic declaration or direct address. In Luke 9:22 skal sunus mans manag winnan jah uskusans fram sinistam wairþan 'The son of man must suffer many things and become rejected by the elders' the periphrastic form uskusans wairþan 'get rejected' translates the Greek aorist passive infinitive ἀποδοκιμασθῆναι 'be rejected'. Since the verse occurs in a context of Jesus addressing his disciples, stylistically it would warrant the (actional) synthetic form. However, the passive infinitive refers to a morphological gap in the Gothic passive paradigm. As a result, there is no possibility of variation in this instance, with the only available passive form being periphrastic. In contrast, in Luke 17:25 aþþan faurþis skal manag gaþulan jah uskiusada fram þamma kunja 'But first (he) must endure much and is rejected by this generation' the same Greek aorist passive infinitive ἀποδοκιμασθῆναι 'be rejected' is rendered with the finite (3rd p.) synthetic passive uskiusada 'is rejected'. While it is possible that in this case the motivation for the choice of the synthetic form is confined to mechanical literalism in translating a Greek word with a Gothic word (rather than the grammatical literalism of translating a Greek infinitive with a Gothic infinitive, as in 9:22 above), it is likewise possible that the synthetic form was in part motivated stylistically.
6. The first passive in the Gothic Gospel of Mark is the formula gameliþ ist 'is written' in Mark 1:2. It is followed by three synthetic forms in Mark 2, including afletanda 'are forgiven' in 2:5 and 2:9 and afnimada 'shall be taken away' in 2:20. The verses occur in a context of direct address or prophesying by Jesus, whose elevated tone is additionally marked by -ata in 2:9. The difference between these verbs is that afletanda 'are forgiven' translates the Greek perfect (medio)passive ἀφέωνταί 'have been forgiven/pardoned' (see discussion with reference to Luke 7:47 and 7:48 above), while afnimada 'shall be taken away' translates the Greek subjunctive aorist passive ἀπαρθῇ 'may be taken away'. If indeed the Greek perfect was the source of Gothic afletanda then, similar to afletanda in Luke 7:47 and 48, it is possible that the synthetic passive was primarily motivated on stylistic grounds, as well as being syntactically convenient. (Otherwise, one would expect a Greek perfect to be rendered with a Gothic stative/ resultative periphrastic passive.) It must be noted, however, that specifically in Mark 2:5 and 2:9 Robinson and Pierpont (2005: 72) and Nestle et al. (2012: 108) give the present (medio)passive ἀφίενταί 'are forgiven' as a variant form, which is a straightforward match for the actional Gothic synthetic form. 9 9 It should also be mentioned in this connection that in Matthew 9:2 and 9:5 afletanda 'are forgiven' translates the Greek present (medio)passive ἀφίενταί 'are forgiven'. It is possible, therefore that in Luke 5:20, 5:23, 7:47 and 7:48 the Gothic synthetic form also refers to this Greek present, even if indirectly.
Mark 3 contains four non-past passives, including one periphrastic and three synthetic forms.
(20) jah hausjandans fram imma bokarjos jah anþarai usiddjedun gahaban ina; qeþun auk þatei usgaisiþs ist 'And having heard of him the scribes and others came out to lay hold on him; for (they) said that (he) is stunned' 10 (Mark 3:21) (21) jah jabai þiudangardi wiþra sik gadailjada, ni mag standan so þiudangardi jaina 'And if a kingdom is divided against itself, that kingdom cannot stand' 11 (Mark 3:24) In Mark 3:21(20) the Gothic periphrastic (resultative) passive usgaisiþs ist 'is stunned/ stupefied' translates an active aorist of the Greek ἐξίστημι 'stand out/aside from' (< ἐκ 'from, out of ' + ἵστημι '(make to) stand'). Some of the secondary senses of the Greek verb include 'astonish, amaze' (in intransitive use), 'be out of one's mind', etc. In translating the aorist ἐξέστη 'was beside himself ' the Gothic translator was aiming at capturing a state perceived as arising from an event or set of circumstances, which he achieved by means of the periphrastic resultative passive usgaisiþs ist. This clause is spoken by the 'scribes and others' who label Jesus as being out of his mind and is in contrast to three non-past passives in the following few verses-spoken by Jesus-all of which are synthetic. The synthetic form gadailjada 'is divided' in 3:24(21) (as well as the same form in 3:25) translates the Greek subjunctive aorist passive μερισθῇ 'should be divided' and is clearly actional (i.e. 'if a kingdom divides against itself '). This is suggested by the contrast of statal (resultative) past passive gadailiþs warþ 'was/became divided' in 3:26, which predictably translates the Greek perfect passive μεμέρισται 'has been divided'. In 3:28 þatei allata afletada þata frawaurhte sunum manne 'That all sins shall be forgiven to the sons of men', a prophetic declaration made by Jesus, the synthetic passive is actional, translating the Greek future passive ἀφεθήσεται 'shall be forgiven'. The use of -ata (allata 'all') is an additional indication of the elevated tone of the verse.
Mark 4 opens with the passive atgiban ist 'is given' in 4:11, which captures the resultative sense of the Greek perfect (medio)passive δέδοται 'has been given' (in spite of -ata in this verse). This resultative is in contrast to the synthetic passive atgibada 'is/ shall be delivered' in 9:31 and 10:33. In 4:29, the synthetic form atgibada 'is deliv- 10 The final clause in Mark 3:21 is traditionally translated or interpreted as "he was beside himself " or "he was out of his mind". In the Gothic translation, however, the passive construction usgaisiþs ist means 'is stunned'. For a detailed discussion of this construction and the semantics of usgaisjan* 'stupefy, stun, strike aghast' see Ratkus (2020). condemned' in 16:16, translating the Greek future passive κατακριθήσεται 'shall be condemned'. Similar to most synthetic passives, afdomjada is uttered by Jesus in a prophetic declaration.

7.
To sum up, in this article I set out to test the hypothesis that, similar to some instances of morphological variation (such as the bare stem vs. -ata in Gothic modifiers), the variation between Gothic inherited synthetic passives and periphrastic forms is motivated stylistically, with the synthetic passive being a stylistically specialised form used in circumstances that require the use of an elevated tone (for instance, direct address, noble proclamation, exhortation, prophetic declaration, pledge, etc.), while the nonpast periphrastic passive is the regular form used in stylistically neutral circumstances.
I begin with presenting some evidence in section 2 that the choice of the synthetic vs. periphrastic non-past passives may be stylistically motivated. In order to be verifiable, this hypothesis requires clear evidence of contrasts, where the Gothic synthetic forms, observed only in specialised stylistic environments, would be contrasted by functionally (semantically) identical periphrastic forms used only in neutral environments. On the evidence of passive constructions attested in the Gothic Gospel texts, it does indeed look like the synthetic forms mostly occur in stylistically specialised contexts. When examination of the evidence is conducted with a stylistic focus exclusive of other criteria, it does indeed seem that stylistics may be an important factor in the choice of passive constructions, which allows the translator to vary the different types of passives as a way of making the (generally very literal) Gothic translation of the Bible more expressive. However, in reality the existence of other clear motives that explain the variation between the synthetic and periphrastic passive forms makes the stylistic variable difficult to isolate.
The discussion of the Gothic examples has shown that, although Greek monolectic forms can be rendered with Gothic periphrastic forms, this typically happens when either the Greek source form or the context of the verb is conceived as a state ('be' passives) or inchoative or change of state ('become' passives). A Greek monolectic form can also be rendered with a Gothic periphrastic construction where the Greek form refers to a gap in the Gothic synthetic passive paradigm. In virtually all other instances, the principal factor in the choice of a Gothic synthetic form is that it corresponds to a monolectic Greek form (i.e. mechanical literalism). At the same time, the Greek periphrastic forms almost invariably correspond to Gothic periphrastic forms. These clearly verifiable motivations in the distribution of the non-past synthetic and periphrastic forms obscure the assumed stylistic motivations. Needless to say, this does not mean that the stylistic motivations did not exist; what this means is that the nature of the Gothic evidence makes the stylistic hypothesis virtually impossible to verify.
As suggested by virtually all examples of contrasts, the most clearly demonstrable difference between the Gothic non-past synthetic and periphrastic 'be' passives is that the synthetic forms are actional, while the periphrastic 'be' forms are (in the majority of the examples) stative. As a matter of fact, I have not been able to find a single synthetic passive that would be unambiguously stative. If this finding can be substantiated on the full evidence of Gothic passive constructions (i.e. including the Epistles and smaller fragments), it would call for a revision of the assumption that the Gothic synthetic forms were ambiguous (either actional or stative), providing an important insight into the decline of the inherited synthetic passive and the development of the periphrastic passive in Gothic and, more broadly, in early Germanic.