(PDF) Hittite Clause Architecture

HITTITE CLAUSE ARCHITECTURE Andrej V. Sideltsev Presses Universitaires de France | « Revue d'assyriologie et d'archéologie orientale » 2015/1 Vol. 109 | pages 79 à 112 ISSN 0373-6032 ISBN 9782130651246 DOI 10.3917/assy.109.0079 Article disponible en ligne à l'adresse : -------------------------------------------------------------------------------------------------------------------- https://www.cairn.info/revue-d-assyriologie-2015-1-page-79.htm © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) -------------------------------------------------------------------------------------------------------------------- Distribution électronique Cairn.info pour Presses Universitaires de France. © Presses Universitaires de France. Tous droits réservés pour tous pays. La reproduction ou représentation de cet article, notamment par photocopie, n'est autorisée que dans les limites des conditions générales d'utilisation du site ou, le cas échéant, des conditions générales de la licence souscrite par votre établissement. Toute autre reproduction ou représentation, en tout ou partie, sous quelque forme et de quelque manière que ce soit, est interdite sauf accord préalable et écrit de l'éditeur, en dehors des cas prévus par la législation en vigueur en France. Il est précisé que son stockage dans une base de données est également interdit. Powered by TCPDF (www.tcpdf.org) – © PUF – [RA 109-2015] 79 HITTITE CLAUSE ARCHITECTURE1 BY Andrej V. SIDELTSEV 1. INTRODUCTION. BASIC HITTITE CLAUSE ARCHITECTURE: STATUS QUAESTIONIS Although Garrett (1990); Luraghi (1990); Hoffner, Melchert (2008), building upon earlier research, established basic Hittite syntactic features, there still remains a lot to be done, both purely descriptively and in the context of advances in typological understanding of how syntax works. This paper focuses on the seemingly well- studied area of the structure of the Hittite clause. As described in (Hoffner, Melchert 2008), Hittite attests a very rigid SOV word order: (1) OH/OS (CTH 547.II) KUB 37.223 c. 5 LÚ KÚR-aš URU-an ḫulāli-zzi # enemy-NOM.SG.C city-ACC.SG.C strike-3SG.PRS “The enemy strikes the city”.2 Some non-canonical word orders, especially involving a constituent in the derived position3 in the left periphery (clause leftmost position) are described as being conditioned by the information structure, but the details remain unclear. Some classes of pronouns (e.g., indefinite and negative pronouns) are described as having aberrant syntactic behavior, different from nouns, namely they are positioned closer to the verb. Subordinators and relative pronouns are positioned clause initially, some of them turn up later in the clause. Negations are normally immediately in front of the verb (Hoffner, Melchert 2008; Luraghi 1990). In the meantime there has been a lot of cross-linguistic work done on clause architecture, mostly on © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) focus, wh-words, relative pronouns and subordinator positions in a clause.4 Considering the research is revealing for a better understanding of the structure of Hittite clause. The first steps in this direction have been done for Hittite by (Goedegebuure 2009; 2014; Huggard 2011; 2013; 2014; Sideltsev 2014b; 2014c). Goedegebuure (2009; 2014) established the preverbal position of replacing/counterexpectant focus and wh- words. The same was done for relative pronouns and the conjunction kuit “because” in (Huggard 2011; 2013). Goedegebuure (2014) provides, among many other things, a reliable framework for the analysis of Hittite information structure, which I follow in the present paper. Sideltsev (2014b) attempted to keep apart (mostly on information structure grounds) two non-canonical verb positions in the clause and to define the preverbal position. Huggard (2014) and Sideltsev (2014c) assessed the position of wh-words, relative and indefinite pronouns in the Hittite clause. 1. Previous versions of the paper were delivered as lectures at a meeting of Sodalizio Glottologico Milanese in Milano on 9 December 2013, and at a meeting of Celtic-Anatolian Seminar at the Department of Anatolian and Celtic Studies, Institute of Linguistics on 24 April 2014. I am grateful to the audience, and particularly to A. Rizza, Ya. Testelets and M. Vai for the feedback. Drafts of the paper were read by O. Belyaev, D. Erschler and E. Ljutikova who provided invaluable criticism. Naturally, all errors are mine. 2. Following An De Vos (2013: 124). 3. = ex situ. 4. (Büring 2009; Hyman, Polinsky 2009; Lyutikova, Tatevosov 2009; Samek-Lodovici 2005; 2006; 2009; Sturgeon 2008; Travis 2005; van der Wal 2012; Zubizarreta, Vergnaud 2005; Zubizarreta 2010), to name just a few. Revue d’Assyriologie, volume 109 (2015), p. 79-112 assyriologie_109.indd 79 16/11/15 14:12 – © PUF – 80 ANDREJ V. SIDELTSEV [RA 109 The present paper aims to consolidate previous research and provide a unified account of Hittite clause architecture, in the way that I term “cross-linguistically informed”, but without going into technical details of any formalism. 2. PREVERBAL POSITION IN THE HITTITE CLAUSE: FOCUS I will start by exploring the preverbal position in a clause. As recent research of P. Goedegebuure (2003; 2009; 2014) showed, despite the claim for rigid SOV word order above, the relative placement of S and O to each other is frequently determined not by their grammatical function, but by their information structure: e.g., the canonical SO word order is determined by the dominant topical status of subject and focal status of object. If subject is contrastive focus and object is topic, the order is reversed –it becomes OS, as in: (2) MH/MS (CTH 186) HKM 13 rev. 13-14 (This capitulation (to the enemy) by Marruwa, the ruler of Ḫimmuwa, about which you wrote me, (adding): “I have dispatched him (to you).” On a tablet you wrote to me about him: “I have dispatched him (to you),” but as of now he has not come. Now put him in the charge of an officer, and have him conduct him quickly before My Majesty. Otherwise,) nu=za apēl waštul zik dā-tti # CONN=REFL his sin.ACC.SG.N you.NOM.SG.C take-2SG.PRS “you take upon yourself his ‘sin’”.5 Following (Goedegebuure 2014: 401), in this example the actual offender is replaced with another person, who might take his sin upon himself. Thus zik “you” which is preverbal in the non-canonical OSV word order is contrastive focus.6 2.1. Preverbal position in the Hittite clause: other constituents As illustrated above, contrastive focus is preverbal. However, the linearly preverbal position is targeted by many other constituents.7 The following constituents can be preverbal: 8 SOME SUBORDINATORS: (3) eNH/NS (CTH 49.II) KBo 10.12+ obv. i 5-6 nu=tta tuēl [ZI=KA SAG.DU=KA DAMMEŠ=KA DUMUMEŠ]=KA **[] Ù CONN=you your soul=your person=your wives=your sons=your and KUR.URU=KA GIM-a[n nakki-ēš] # © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) land=your as dear-NOM.PL.C “As your [soul, your person, your wives], your [sons], and your land are dear to you…”;9 5. Following (Hoffner 2009: 118). 6. Replacing focus in P. Goedegebuure’s terminology (Goedegebuure 2014). Alongside preverbal position, P. Goedegebuure posits another, clause-initial, position of narrow informational foci, such as additive focus, etc (Goedegebuure 2014). 7. In traditional Hittite grammars (Hoffner, Melchert 2008) the preverbal position of many of the constituents was not well understood. See (Goedegebuure 2014; Luraghi forthcoming; Huggard 2011, 2013, 2014; Sideltsev 2014b) for independently formulating the rules for the preverbal position. 8. Commonly kuit “as” (Huggard 2013), kuwapi “when” (Hoffner, Melchert 2008: 417) and occasionally maḫḫan “when, as” (ibid: 417). 9. Following (del Monte 1986: 128-9; Beckman 1996: 33). assyriologie_109.indd 80 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 81 10 RELATIVE PRONOUNS OR RELATIVE PHRASES: (4) NH/lNS (CTH 106.A.1) Bo 86/299 obv. ii 22-24 (They shall not take those revenues and supplies for ceremonies for the gods of Tarhuntassa) 1. ANA mDLAMMA=ma LUGAL KUR URU DU-tašša ABU=YA kui-t pe-šta # to Kurunta=but king land Tarhuntassa father=my which-ACC.SG.N give-3SG.PST D = 2. UTUŠ I=ya=šši kui-t pi-ḫḫun # Majesty=My=him which-ACC.SG.N give-1SG.PST “(1) from that which my father gave to Kurunta, king of the land of Tarhuntassa, (2) and that which I, My Majesty, have given him”.11 12 WH-PHRASES: (5) NH/NS (CTH 89.A) KUB 21.29(+) rev. iv 13-14 (But if you men of the city tolerate/condone (?) someone/something,) šummeš=kan kui-t ney-ari # you.DAT!.PL=LOC what-NOM.SG.N happen-3SG.PRS.MED “what will happen to you?”.13 14 NEGATION MARKERS AND NEGATIVE PRONOUNS (6) NH/NS (CTH 204) Msk. 73.1097 17-19 kinuna=šši=kan apāt É-er GIŠKIRI₆.GEŠTIN=ya ar[ḫa] lē kuitki ta-t[t]i # now=him=LOC that house vineyard=and away PROHIB something take-2SG.PRS “Now, that house(hold) and vineyard you should in no way take from him!”15. 16 INDEFINITE PRONOUNS: (7) NH/NS (CTH 85.2) KUB 21.37 obv. 48’ [mān ANA] DUTUŠ=I ḪUL–lu kuiški peda-i # if to Majesty=My evil.ACC.SG.N somebody.NOM.SG.C bring-3SG.PRS “[If] somebody brings evil to My Majesty…”.17 18 PREVERBS: (8) NH/lNS (CTH 106.A.1) Bo 86/299 rev. iii 3-5 zilati!=wa=kan LUGAL-UTTA ŠA KUR URU DU-tašša ANA NUMUN in.the.future=QUOT=LOC kingship of land Tarhuntassa to progeny m NIR.GÁL lē kuiški arḫa dāi # © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) Muwattalli PROHIB anybody.NOM.SG.C away take.3SG.PRS “For all time no one shall take the kingship of the land of Tarhuntassa away from the progeny of Muwattalli”.19 10. A relative phrase is relative pronoun + NP. Cf. (Huggard 2011). Both relative pronouns in determinate and indeterminate relative clauses can be preverbal. Relative pronouns functioning as indefinite in conditional clauses behave in the same way: (a) MH/MS (CTH 147) KUB 14.1+ rev. 45 nu=wa=mu mān idālu-n memia-n kui-š [mema-i] # CONN=QUOT=me if evil-ACC.SG.C word-ACC.SG some-NOM.SG.C tell-3SG.PRS “If anybody tells me a bad word” following (Beckman 1996: 150). 11. Following (Otten 1988: 16-17; Beckman 1996: 111-2). 12. See (Hoffner 1995; Goedegebuure 2009; 2014). 13. Following (Neu 1968: 115; CHD L-N: 215, 363). Cf. (González Salazar 1994: 165). 14. See (Hoffner, Melchert 2008). 15. Following (Hoffner 2009: 370). 16. See (Sideltsev 2014b; Huggard 2014; Goedegebuure 2014; Luraghi forthcoming). 17. Following Ünal (1974: 124). 18. See (Tjerkstra 1999: 173; Luraghi 1990: 32, 35; Francia 2002a; Salisbury 2005: 216). Preverbs are separable from the verb, they do not form one morphosyntactic or phonological complex with the verb. 19. Following (Otten 1988: 20-21; Beckman 1996: 114). assyriologie_109.indd 81 16/11/15 14:12 – © PUF – 82 ANDREJ V. SIDELTSEV [RA 109 20 LOW MANNER ADVERBS AND ADVERBIALS like SÍG₅-in “well”, kiššan “this way”, apeniššan “that 21 way”, namma “then, again”, mekki “much”: (9) MH/MS (CTH 138.1) KUB 13.27+ obv!. 22’-23’ [n]aš?ma=wa ḫāl[i?] ŪL SÍG₅-in u-ška-nzi # or=QUOT watch.ACC.SG.N NEG properly see-IMPF-3PL.PRS “Or they do not keep watch properly”. Some of these constituents, primarily negation markers, preverbs and low adverbs, are preverbal in the canonical word order, but they can move to the clause initial position for information structure and discourse reasons.22 Negation can also move into any position if its scope is immediately over the constituent it moves in front of. The preverbal position of the constituents in the section is statistically different –for some it clearly dominates (negation markers, negative pronouns, low adverbs), for some it is one of the two options (indefinite, relative and interrogative pronouns, subordinator kuit).23 For some it is a statistically minor position (subordinator maḫḫan). 2.2.1. The inner structure of the preverbal position If several of these constituents are preverbal simultaneously, they are linearly ordered as follows, from left to right:24 (a) focus, wh-phrases; (b) subordinators, relative pronouns, and indefinite pronouns; (c) preverbs;25 (d) negation markers, negative pronouns;26 (e) low manner adverbs and adverbials.27 Evidence for the ordering: WH-WORD – NEGATION: (10) NH/NS (CTH 177.3) KUB 23.101 obv. ii 5 LÚ nu tu-el ṬEMU kuwat ŪL punuš-ta # CONN you-GEN.SG messenger why NEG ask-3SG.PST “Why did you not ask your messenger”.28 WH-WORD – MANNER ADVERB: (11) NH/NS (CTH 171) KUB 23.102 obv. 5′-6′ ŠEŠUTTA=ma Ù ŠA ḪUR.SAGAmmana uwauwar kuit namma meme-ške-ši # brotherhood=but and of Mt. Ammana coming why then speak-IMPF-2SG.PRS “But why do you continue to speak about “brotherhood” and about coming to Mt. Ammana?”.29 © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) 20. See (Tjerkstra 1999; Sideltsev 2014b; Goedegebuure 2014). Namma is in fact floating with predominantly clause-initial position –out of 57 entries in M. Molina’s MH/MS letter corpus only 14 are not clause-initial. What matters, however, is that one of the positions namma can occupy in the clause is preverbal. 21. Just like Georgian low adverbs (Skopeteas, Fanselow 2010). 22. E.g., apenišš(uw)an is overwhelmingly preverbal –in the MH/MS letter corpus of M. Molina out of 17 entries apenišš(uw)an is 15 times preverbal (once it is postverbal out of 15). In two cases it is not preverbal because it is topicalized. So we might want to change the taxonomy –low adverbs are preverbal when they are focus and not preverbal when they are topical, but see below (section 2.2.3) for the reasons to keep them apart from the rest of foci. 23. See below for the other option. 24. See (Sideltsev 2014b; Goedegebuure 2014). 25. Preverb position is more complex and will be dealt with in a separate paper. 26. I.e. negation marker + indefinite pronouns. 27. I know of only one case where the linear sequence is reversed, it quite expectably involves adverbs: LOW ADVERB – SUBORDINATOR: (b) MH/MS (CTH 186) HKM 17 rev. 28-29 URU Kapapaḫšuwa-š mekki kuit [paḫḫašn]uw-anza # Kapapaḫšuwa-NOM.SG.C much since protect-PRTC.NOM.SG.C “Since Kapapaḫšuwa is well protected” following (Hoffner 2009: 124). In a number of cases where the subordinator is between the negation marker and the verb, it may just be in the second position. 28. Following (CHD Š: 61; de Roos 2005: 52; Hoffner, Melchert 2008: 428). Ср. (Hagenbuchner 1989: 278-279). assyriologie_109.indd 82 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 83 FOCUS – SUBORDINATOR: (12) NH/NS (CTH 81.A) KUB 1.1+ rev. iv 7-8 ammuk=ma LUGAL-UTTA DIŠTAR GAŠAN=YA annišan=pat kuit memi-ške-t # me.DAT.SG=but kingship Istar lady=my previously=EMPF because say-IMPR-3SG.PST “Because to me my lady Istar had previously promised the kingship, (at that time my lady Istar appeared to my wife)”.30 FOCUS – PREVERB – MANNER ADVERB: (13) NH/NS (CTH 407) KBo 15.1 obv. i 12-14 1. nu=kan ANA LÚ LUGAL–u-š anda kišan memai #… CONN=LOC to man king-NOM.SG.C into in.this.way speak.3SG.PRS 2’. [nu=kan AN]A LÚ apāš anda kišan [mema]i # CONN=LOC to man he.NOM.SG.C into in.this.way speak.3SG.PRS “(1) And it is the king who speaks to the man in this way. (But if it does [not] please the king (to do so), 3 he will send someone else, 4 and that one will assist at the ritual instead (of the king).) (2’) He will [spea]k [to] the man in this way instead”.31 RELATIVE PRONOUN32 – LOW ADVERB: (14) NH/lNS (CTH 106.A.1) Bo 86/299 rev. iii 2-3 našma ANA AWAT KUR DU-tašša kui-t kiššan or to matter land Tarhuntassa which-NOM.SG.N as.follows EGIR-an iya-n # then do-PRTC.NOM.SG.N “Or concerning the problem of the land of Tarhuntassa something is stipulated subsequently (to my father’s treaty tablet) as follows”.33 FOCUS – INDEFINITE PRONOUN: (15) NH/NS (CTH 106) Bo 86/299 rev. iii 3-8 (In the future let no one take away the kingship of Tarhuntassa from the offspring of Muwatalli). nu apāt kuiški iya-zi # CONN that.ACC.SG.N someone.NOM.SG.C do-3SG.PRS “(If) someone does the following instead, (that is, he gives it to another descendant of Muwatalli while taking it away from the offspring of Kuruntiya, anyone who performs that action instead, may the Stormgod of Hatti and the Sungoddess of Arinna destroy him)”.34 © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) 2.2.2. The position of indefinite pronouns It was observed in passing above that indefinite pronouns are in the preverbal position as different from all other verbal arguments: (16) NH/NS (CTH 42.A) KBo 5.3+ obv. ii 27 našma kūrur KI.BAL kuiški ē[p-z]i # or hostile.ACC.SG.N revolt someone.NOM.SG.C take-3SG.PRS “Or (if) anybody starts a hostile revolt”.35 If a clause simultaneously attests verbal arguments which are instantiated by noun phrases and by indefinite pronouns, including noun phrases modified by indefinite pronouns, all the rest of verbal arguments, 29. Following (Hoffner 2009: 323). 30. Following (Otten 1981: 24-25; Huggard 2013: 4). 31. Following (Goedegebuure 2014: 395). The example is from ritual description. The nearest diplomatic parallels are P. Goedegebuure’s exx. 7.13 (apiya kuiški anda), 7.40 (apiya anda). 32. Functioning as indefinite in the conditional clause. 33. Following (Otten 1988: 20-21). Cf. (Beckman 1996: 114): “Or concerning the problem of the land of Tarhuntassa, because it is hereby stipulated subsequently (to my father’s treaty tablet) as follows”. 34. Following (Otten 1988: 20f; Goedegebuure 2014: 393 = ex. 7.31). 35. Following (G. Wilhelm (ed.), hethiter.net/: CTH 42 (TX 26.07.2013, TRde 19.02.2014), Beckman 1996: 26). assyriologie_109.indd 83 16/11/15 14:12 – © PUF – 84 ANDREJ V. SIDELTSEV [RA 109 both definite/specific and indefinite/non-specific,36 both topics and foci, are always in front (to the left) of indefinite pronouns, irrespective of their syntactic function. E.g., in (16) the direct object kūrur KI.BAL “hostile revolt” is in front of the subject instantiated by the indefinite pronoun, which brings about the non- canonical OSV word order. This behavior of indefinite pronouns is not available in well-studied SOV languages where indefinite pronouns behave like other verbal arguments.37 This is true even for Hungarian which otherwise attests a dedicated quantifier position (Kiss 2004): Hungarian existential quantifiers like valaki “somebody” are either in the topic position or postverbal (Kiss 2004: 106-7). 2.2.3. How many preverbal positions are there? Thus wh-words, relative and indefinite pronouns as well as subordinators are preverbal. Now two questions arise: (a) whether the preverbal position is the same for all these constituents and (b) whether the position is original or derived.38 2.2.3.1. One vs. several preverbal positions The evidence for one vs. several positions is extremely scanty and unclear. It looks like there are data that indefinite pronouns and wh-words occupy different preverbal positions. Wh-words are likely to occupy the same position as preverbal focus: (a) both preverbal contrastive focus (17) and wh-words (18) are in front of preverbs:39 (17) NH/NS (CTH 407) KBo 15.1 obv. i 12-13 nu=kan ANA LÚ LUGAL–u-š anda kišan memai # CONN=LOC to man king-NOM.SG.C into in.this.way speak.3SG.PRS “And it is the king who speaks to the man in this way”.40 (18) MH/MS (CTH 186?) HKM 43 obv. 1’-5’ m n=[aš]ta Tarul[i?]y[a]š? tuzzi-n m Zilapiyašš=a CONN=LOC Taruliya.GEN.SG army-ACC.SG.C Zilapiya.GEN.SG=and ÉRINMEŠ GIBIL maḫḫan šarā uwat-er # troops new how up bring-3PL.PST “How could they have brought up the army of Taruli(ya) and the new troops of Zilapiya?”.41 (b) Both focus and wh-words stay higher than the non-canonical clause-internal position of the verb (19):42 (19) NH/NS (CTH 63.A) KUB 19.31+ rev. iii 27”-31” © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) nu k[ū]n memiya-n kuwat iya-tten QATAMMA # CONN this.ACC.SG.C matter-ACC.SG.C why do-2PL.PST in.this.way “So, why have you handled this matter in this way: (you keep taking those civilian captives away from Tuppi-Teššub)?”.43 Naturally, as QATAMMA in (19) stands for kiššan, not apeniššan, it is not replacing focus. 36. Cf. wrongly (Huggard 2014). 37. Cross-linguistically, this is highly unusual. Indefinite/non-specific arguments often occupy a different structural position than definite/specific ones (see, among others, (Gračanin-Yüksek, İşsever 2011; Kiss 2004; Kahnemuyipour, Megerdoomian 2008: appendix A; Kahnemuyipour, Megerdoomian 2011 with ref.; Travis 2005: 209; Vikner 1995), or raise to a lower position, so indefinite/non-specific arguments are one of the few constituents which can intervene between focus and verb in SOV languages (Kim 1988). However, I know of no language where only indefinite/non-specific arguments which are indefinite pronouns behave in a different way from all noun phrases, both definite/specific and indefinite/non-specific. Thus the different syntactic behavior of indefinite pronouns cannot be attributed to their non- specificity/indefiniteness contra (Huggard 2014). 38. In the paradigms which operate with movement the original position is the position where a constituent is base- generated (in situ) whereas the derived position is the position where the constituent moves (ex situ). 39. See (Goedegebuure 2014). 40. Following (Goedegebuure 2014: 395). 41. Following (Hoffner 2009: 169). 42. See below in more detail. 43. Following (Miller 2007: 126-127, 129-130). assyriologie_109.indd 84 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 85 (c) Finally, the information structure of both focus and wh-words is similar.44 Actually, there are no cases in my corpus where both indefinite pronouns and wh-words are attested in the same clause.45 But if wh-words and focus are in the same position, there are ample data that both indefinite pronouns and focus are attested in the same clause, as in (20) where focused apiya “then” precedes indefinite pronoun kuiški “someone”: (20) NH/NS (CTH 379) KUB 31.121 obv. ii 11'-13' 1. karū kui-ēš LUGALMEŠ eš-ir # before who-NOM.PL.C kings be-3PL.PST 2. nu=kan mā[n] apiya kuiški anda dai[-š] # CONN=LOC whether then any.NOM.SG.C into put-3SG.PST “(I do not know) (1) whether any (2) of the kings that ruled (lit. were) in the past (1) added (any word) then”.46 So, albeit admittedly indirectly, it can be assumed that indefinite pronouns and wh-words occupy two distinct preverbal positions. Several different positions are not so clear in case of indefinite pronouns, relative pronouns and subordinators. At first sight it looks like they all occupy the same position. All of them are different from wh- words in that they are not attested to the left of the clause-internal verb as in (19), they rather stay lower (21- 23):47 (21) NH/lNS (CTH 383) KUB 21.19+ obv. ii 9 apāt=ma ḪUL–lu uttar iya-t kui-š # that.ACC.SG.N=but evil.ACC.SG.N thing.ACC.SG.N do-3SG.PST who-NOM.SG.C “The one who did that evil thing…”.48 (22) NH/NS (CTH 61.II.7.A) KBo 2.5+ rev. iii 34-35 m KUR nu Aparru-š LÚ [Kal]ašma kūruriaḫ-ta kuit # CONN Aparru-NOM.SG.C man Kalasma get.hostile-3SG.PST as “As Aparru, the man of Kalasma, started hostilities, (he mobilized 3000 troops)”.49 (23) OH/OS (CTH 291.I.b.A) KBo 22.61+ obv. i 4 (§ 3) [takku LÚ-a(n našma MUNUS-an ELLAM walaḫ-zi k)]uiški50 # if man-ACC.SG. or woman-ACC.SG.C free strike-3SG.PRS somebody.NOM.SG.C “[If] anyone strikes a free [man] or woman…”.51 Besides, they all attest wide-going similarities in the clause distribution –they can all be preverbal, second position and clause initial/first.52 The parallelism is at first sight contradicted by the fact that © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) subordinators and indefinite/relative pronouns are simultaneously attested preverbally53 in the same clause (24): 44. There is enormous literature on the topic. Cf. (Goedegebuure 2009) who I believe overinterpreted the difference between information structure of preverbal and clause-initial/first wh-words. 45. The only example I am aware of is restored in the relevant point and thus cannot serve as independent evidence: (c) MH/MS (CTH 190) HKM 84 obv. 12’-14’ nu=tta n[am]ma kui-š [kuitki iya-zi] # CONN=you then who-NOM.SG.C anything.ACC.SG.N make-3SG.PRS “(But if there is no grain,) then who ever again will make anything for you?” following (Hoffner 2009: 247). 46. Following (Goedegebuure 2014: 385). 47. The verbs are either topics, as in (19, 21-22) or part of broad informational focus, as in (23). Cf. wrongly (Sideltsev 2014c). 48. Following (Singer 2002b: 742-743). 49. Following (Goetze 1967: 188-189). 50. Restored from OH/NS copy KBo 6.3+ obv. i 6. 51. Following (Hoffner 1997: 18). 52. Admittedly, with considerable statistical differences between the frequency of each position. 53. Naturally, subordinators and indefinite/relative pronouns are attested many times in the same clause if the subordinator is clause-first/initial and the relative/indefinite pronoun/subordinator functioning as indefinite pronoun is preverbal. assyriologie_109.indd 85 16/11/15 14:12 – © PUF – 86 ANDREJ V. SIDELTSEV [RA 109 (24) NH/NS (CTH 255.2.A) KUB 26.1+ rev. iii 50-51 [(š)]ummaš=ma kuit GIM-an išdamaš-ten # you=but which-ACC.SG.N when hear-2PL.PST “But when you have heard something…”.54 (25) NH/NS (CTH 255.2.A) KBo 26.1+ rev. iii 16 nu=za kui-t GIM-an kiš-ari # CONN=REFL which-ACC.SG.N when hear-2PL.PST “And should something happen …”.55 (26) NH/NS (CTH 255.2.B) KBo 26.8 obv. ii 5' […] x=ma kuit GIM-an *u-š*ke-tteni # x=but which-ACC.SG.N when see-IMPF-2PL.PRS “But when you observe something…”.56 Exx. (24–26) would certainly be a strong argument against subordinators and indefinite/relative pronouns occupying the same position.57 However, it is conspicuous that the three examples are the only ones from my corpus of diplomatic texts, they are extremely late and come from two very closely related texts. Moreover they are lexically very similar. Besides, in one of the three examples, the relative pronoun functioning as an indefinite one is clause-first. In two other examples, the position of the relative pronoun is ambiguous between clause second and preverbal. In view of the scantiness and ambiguity of evidence it is an open question whether subordinators and indefinite/relative pronouns occupy the same position. What is potentially relevant for the discussion is the fact58 that in relation to the verb’s position in the clause, including non-canonical verb’s position, the position of indefinite pronouns, relative pronouns and subordinators is the same. I also expressly reject the claim of Huggard (2011) that relative pronouns occupy the same position as wh-words and preverbal focus as the information structure status of relative pronouns has nothing to do with contrastive focus or even with the narrow informational focus of wh-words. 2.2.3.2. Original vs. derived position? The next question is whether the preverbal positions of wh-words, relative pronouns, indefinite pronouns and subordinators are original or derived. The arguments in favor of the position being the original one are the following: if wh-words, relative pronouns, and indefinite pronouns are verbal arguments, only objects are preverbal whereas subjects are not.59 Thus preverbal wh-words, relative pronouns, and indefinite pronouns which are objects simply conform to the canonical SOV Hittite word order and are in the original position. As different from this, subjects are not preverbal60 and thus again conform to the canonical SOV word order. © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) However, this is not really so. Both for relative pronouns, wh-words and especially indefinite pronouns, there are attestations where subjects instantiated by relative pronouns (27), wh-words61 and indefinite pronouns (28) are preverbal, thus bringing about non-canonical OSV word order: (27) NH/OS (CTH 292.II.a.B) KBo 6.26 rev. iii 16 (§ 186) ŠA 2 GU₄ iugaššaš UZU=ŠUNU [k]ui-š wāš-i # of 2 cattle yearling.GEN.PL? meat=their who-NOM.SG.C buy-3SG.PRS “Whoever buys the meat of 2 yearling cattle”.62 54. Following (Miller 2013: 302-303). 55. Following (Miller 2013: 300-301). 56. Following (Miller 2013: 298-289). 57. The position of relative and indefinite pronouns is different vis-à-vis negation markers: whereas the former normally precede it even when they are preverbal, the latter (as part of negative pronouns) follow it when they are in the preverbal position. 58. Which I will demonstrate further. 59. See (Huggard 2011). 60. Thus (Huggard 2011). 61. See (Goedegebuure 2009) for the examples. 62. Following (Hoffner 1997: 147). assyriologie_109.indd 86 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 87 (28) NH/NS (CTH 584) KUB 15.1+ obv. ii 6-8 URU ŠÀ Ù=kan GIM-an MUNUS.LUGAL INA Iyamma inside dream=LOC when queen in Iyamma É tarnuw-i EGIR-an LÚMEŠ GURUŠ kuiēšqa ḫatkiššanu-šk-ir # bathhouse-LOC.SG behind men young some.NOM.PL.C oppress-IMPF-3PL.PST “When in a dream some young men behind/at the back of the bathhouse in Iyamma (intended to) oppress the queen”.63 This means that the position of relative pronouns, wh-words and indefinite pronouns is not original, but rather derived –just like the preverbal position of contrastive focus which brings about the same non- canonical OSV word order is derived. Naturally, however, it might be supposed that all the rest of verbal arguments move in these very particular cases, leaving the constituents which surface preverbally in their original position. This, however, is refuted by the placement of relative pronouns, wh-words, subordinators and indefinite pronouns vis-à-vis preverbs. As is obvious from the corpus of data from diplomatic texts, both the position of relative pronouns, wh-words, subordinators and indefinite pronouns in front of the preverb, and their position between the preverb and the verb are attested. See for preverbal indefinite pronouns the following two examples where (29) attests the indefinite pronoun between the preverb and the verb and (30) –the indefinite pronoun in front of the preverb: (29) OH/NS (CTH 292.II.a.B) KBo 6.26 obv. i 18 (§ 162a) takku PA₅-an EGIR-an arḫa kuiški nāi if canal-ACC.SG.C subsequently away somebody.NOM.SG.C turn.3SG.PRS “If anyone (totally?) diverts an irrigation ditch…”.64 (30) NH/NS (CTH 62.II.A) KBo 5.9+ obv. ii 25'–26' m mān tuk=ma TUPPI–DU–up-an uddān-aza kuiški if you.ACC.SG=but Tuppi-Tessup-ACC.SG.C word-ABL some.NOM.SG.C anda damaš-ti # into oppress-3SG.PRS “But if somebody oppresses you, Tuppi-Tessup, with a word…”.65 But statistically the two positions are very differently represented in the texts. For indefinite pronouns, the counts are as follows: the position between the preverb and the verb is attested 7 times (19%) whereas the position in front of the preverb is attested 30 times (81%). Three of the seven attestations occur in the lexically identical phrase which occurs in three NH/NS treaties.66 Besides, subject or direct object © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) indefinite pronouns occur only 1 time between the preverb and the verb whereas they occur 19 times in front of the preverb. Thus the statistics is very different if syntactic functions are considered. It is an impressive 5% to 95%! For relative pronouns, only the position in front of the preverb is available in my corpus: (31) MH/MS (CTH 190) HKM 64 obv. 22 nu uddan-aš arkuwar kuit EGIR-pa iēr # CONN word.GEN.SG reply.ACC.SG.N which.ACC.SG.N back make.3PL.PST “And the reply to the word which they have made …”.67 (32) MH/MS (CTH 190) HKM 71 rev. 24-26 kāša=kan kī tuppi kuedani UD–t-i parā ne-ḫḫun # PRF=LOC this tablet which day-LOC.SG out send-1SG.PST “On the day that I have dispatched this tablet”.68 63. = obv. ii 38-40. Following (de Roos 2007: 91, 99; Mouton 2007: 261-263). 64. Following (Hoffner 1997: 129). 65. Following G. Wilhelm-F. Fuscagni (ed.), hethiter.net/: CTH 62 (TX 16.10.2013, TRde 15.10.2013). Cf. (del Monte 1986: 168-169; Beckman 1996: 57). 66. Two of them of Mursili II ((CTH 67) KBo 5.4 obv. 28’-29’ and (CTH 68.D) KBo 19.69+ obv. iii 17-18) and one in that of Muwatalli II (NH/lNS (CTH 76.A) KBo 19.73+ rev. iii 17-18). 67. Following (Hoffner 2009: 216). 68. Following (Hoffner 2009: 228). assyriologie_109.indd 87 16/11/15 14:12 – © PUF – 88 ANDREJ V. SIDELTSEV [RA 109 If one follows Huggard (2014) that the position between the preverb and the verb is the original position for all verbal arguments, then the data can only be interpreted in one way: the preverbal position in reality is heterogeneous. All of the preverbal categories are located originally between the preverb and the verb, but they stay there only in very rare cases. In the majority of cases the preverbal position is derived –it is in front of the preverb. 2.2.3. low adverbs and wh-words It follows from section 2.2.1 that low adverbs follow all other preverbal constituents and immediately precede the verb. However, it happens only when low adverbs are part of the broad VP focus, as in exx. (11, 13, 14). When low adverbs are topicalized or narrowly focussed, they target the same position as other contrastive foci as seen in (34), or other topics, as seen in (33). (33) MH/MS (CTH 186) HKM 8 obv. 12-14 LÚ n=ašta KÚR QATAMMA kuit KUR-e anda lammar CONN=LOC enemy thus because land-LOC.SG into instantly lammar ia-ttar[i] # instantly march-3SG.PRS.MED “Because the enemy thus marches into the land at a moment’s notice”.69 QATAMMA/apeneššuwan in ex. (33) refers back to the previous situation and is part of the old information, thus it receives a topic reading. (34) NH/NS (CTH 40) KBo 5.6 obv. iii 52 kuwat=wa apeniššan TAQBI # why=QUOT in.that.way speak “Why have you spoken in that way?”.70 In (34) the wh-word is in front of the contrastively focused adverb apeneššuwan “in that way”.71 Following (Goedegebuure 2014: 251) “…the adverb apeniššan is used to express a type of counter-expectant focus: “Why did you write in this way (focus) ‘…’, (instead of believing me immediately and send a son”)? The fact that we find eniššan in iv 4, referring backward to the same piece of text “Do they deceive me?”, shows that apeniššan replaces non-focal kiššan”. In this case it can be supposed that the wh-word is in the clause initial position, see for such position section 2.2.4, whereas the focussed adverb is in the same position as other preverbal foci. © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) However, the explanation is not applicable to (35) which similarly attests both a contrastive focus and a wh-word in the clause. But, as different from (34), in (35) the position of the wh-word is preverbal: (35) NH/NS (CTH 127) Bo 2810 obv. 8-9 nu=mu DUMU=YA kuwat iya-t apeneššuwan # CONN=me son=my why do-3SG.PST that.way “Why has my son acted that way towards me?”72 The only way to assess (35) is to modify the explanation for (34), i.e. to suppose that the wh-phrases in (34-35) are topical, and thus are positioned in the topic position, i.e. in front of replacing focus which is regularly preverbal. The preverbal position of wh-word in (35) follows from the non-canonical position of the verb.73 An important conclusion is that wh-words are sensitive to the information structure.74 The same applies to relative pronouns.75 Other preverbal constituents –indefinite pronouns and subordinators– are not. 69. Following (Hoffner 2009: 109). 70. Following (Goedegebuure 2014: 251). 71. See for the information structure (Goedegebuure 2014: 251). 72. Cf. (Hoffner 2009: 363). 73. Which will be assessed below. 74. Cf. in the same spirit, but along very different lines (Goedegebuure 2009). 75. See (Huggard 2011, 2014; Becker 2014). assyriologie_109.indd 88 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 89 However, the sensitivity very seldom brings about clause positions which are different from preverbal or, as we will shortly see, other positions in the clause. 2.2.4. Clause initial vs. preverbal positions in a Hittite clause I have already mentioned that relative pronouns can be both preverbal and clause initial. The same is true for the majority of the constituents which are preverbal –they can also be at the left edge of the clause, as is the case with the subordinator maḫḫan “as”, which was preverbal in (3)76 and is clause initial in (36): (36) NH/lNS (CTH 106.A.1) Bo 86/299 obv. ii 101-102 URU nu maḫḫan ANA LUGAL KUR Ḫatti ZI-anza # CONN as to king land Hatti soul.NOM.SG.C “(He shall be treated) as the King of Hatti pleases”.77 As is seen from the example, constituents in the clause initial position can follow sentence connectives like nu, especially in Middle and New Hittite texts.78 The clause initial/first and preverbal positions in Hittite are identical only to a very limited degree – only subordinators, relative pronouns and wh-words can occur in either position.79 This does not, naturally, mean that I equate the positions themselves. The majority of subordinators are clause initial,80 only a few are preverbal (most frequently kuit “because”81 and much rarer maḫḫan “as”), and all of the preverbal ones can occur in the first/initial position. Relative pronouns are normally clause second in determinate relative clauses and clause initial in indeterminate relative clauses,82 they are quite seldom unambiguously preverbal.83 Wh- words occur in either position,84 although preverbal position clearly dominates: out of 145 questions in (Hoffner 1995), only 12 wh-words are unambiguously clause initial and not preverbal.85 Contrastive foci are only preverbal, contrastive topics are only clause initial86 whereas non-contrastive topics are not necessarily clause initial. Indefinite pronouns are attested once in the first position in my corpus. Both initial and preverbal positions can be filled in a clause simultaneously: the clause initial position in such cases normally hosts subordinators, whereas the preverbal position may contain focus, relative pronouns, indefinite pronouns, subordinators functioning as indefinite pronouns. 2.2.4.1. Clause second position As it was observed in section 2.2.2.5, not all the left edge constituents are clause initial. Some of them occupy the second position. The main second position constituents in Hittite are indefinite pronouns, relative pronouns87 and second position subordinators88 like kuit “because”, kuwapi “where”, more seldom maḫḫan © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) 76. Spelled with the Sumerogram GIM-an. 77. Following (Otten 1988: 20-21; Beckman 1996: 113). 78. Thus it would be more accurate to label the position first, and not initial, as is done in (Luraghi 1990). 79. There is just a handful of cases in my corpus where the subordinator is neither clause second, nor preverbal. I leave them out of consideration for the moment for the future research because of their rarity. 80. See for a descriptive overview (Hoffner, Melchert 2008). 81. See (Huggard 2013). 82. Or, following the recent reassessment of relative pronouns’ distribution by (Becker 2014), the information structure is different and more complex. 83. (Huggard 2011) assesses the preverbal position as the original one and treats first and second positions as topicalizations or focusing, but see above. 84. (Hoffner 1995; Goedegebuure 2009). 85. (nu) wh-word-V clauses are ambiguous for me. 86. (Goedegebuure 2014) discusses some not completely clear examples which can be preverbal contrastive topics. 87. In what was traditionally termed determinate relative clauses (Held 1957, Garrett 1994), but cf. now (Becker 2014). 88. See for subordinators (Hoffner, Melchert 2008), for indefinite pronouns (Huggard 2014, Sideltsev forthcoming). assyriologie_109.indd 89 16/11/15 14:12 – © PUF – 90 ANDREJ V. SIDELTSEV [RA 109 “as”. As different from clitics in the Wackernagel position, which can be second after connectives like nu,89 relative pronouns and subordinators do not count a number of constituents as the first position:90 (37) NH/NS (CTH 61.II.7.A) KBo 5.8 rev. iii 24-5 nu uni kui-n 9 LIM ÉRINMEŠ mPitaggatalli-š uwate-t # CONN this who-ACC.SG.C 9000 troops Pitagattalli-NOM.SG.C bring-3SG.PST “That 9,000-man army which Pitagattalli led (joined battle with me)”.91 (38) OH-MH/MS (CTH 262) IBoT 1.36 rev. iv 30 LÚ.MEŠ MEŠEDI kuwapi *du*[nnake]šn-i ti-ššak-anzi # Bodyguards where inner.chamber-LOC.SG step-IMPF-3PL.PRS “Where the bodyguards stand by the inn[er cham]ber…”.92 (39) NH/lNS (CTH 106.A.1) Bo 86/299 rev. iv 8-9 našma ANA DUTUŠ=I kuiški našma ANA NUMUN DUTUŠ=I or to majesty=my someone or to progeny majesty=my ANA LUGAL-UTTI ŠA KUR URUḪatti uwāi pēda-i # for kingship of land Hatti difficulty bring-3SG.PRS “Or (if) someone brings difficulties upon My Majesty or upon the progeny of My Majesty concerning the kingship of Hatti, […]”.93 2.2.5. Postverbal position in a Hittite clause After the detailed treatment of the preverbal position I will pass to the main topic of the paper, the postverbal position. The logic behind the structure of the paper is that the constituents which land postverbally are identical to the constituents which are preverbal. I will remind that the preverbal constituents are (a) focus, wh- phrases; (b) some subordinators, relative and indefinite pronouns; (c) preverbs; (d) negation markers, negative pronouns; (e) low manner adverbs and adverbials. Now part of these can be postverbal, namely (b) subordinators, relative and indefinite pronouns; (d) negation markers, negative pronouns; (e) low manner adverbs and adverbials. It is curious that focus and wh-words are very sporadically postverbal. There is only one example of wh-words being postverbal in my corpus, see ex. (41), vs., e.g., 15 indefinite pronouns in the postverbal position, as in (42). It is also curious that contrastive focus is represented only by contrastively/scalarly focused low adverbs, as in (35) above:94 POSTVERBAL SUBORDINATORS: (40) NH/NS (CTH 61.II.7.A) KBo 2.5+ KBo 16.17+ rev. iii 34-35 m nu Aparru-š LÚ KUR[Kal]ašma kūruriaḫ-ta kuit # © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) CONN Aparru-NOM.SG.C man Kalasma get.hostile-3SG.PST as “As Aparru, the man of Kalasma, started hostilities, (he mobilized 3000 troops)”.95 POSTVERBAL WH-WORDS: (41) OH/NS (CTH 19.II.A) KBo 3.1+ obv. i 40 [k]-ī=wa iya-nun kuit # this-ACC.SG.N=QUOT do-1SG.PST why “Why did I do this?”.96 89. (Garrett 1996; Hoffner, Melchert 2008; Luraghi forthcoming). It is likely that nu is a proclitic. 90. I.e., not after nu, mān, našma, namma, etc., see (Sideltsev 2002, 2014b). It implies that there are two clause second positions in Hittite, one for clitics, and one for relative/indefinite pronouns/some subordinators. 91. Following (Goetze 1967: 158-159; Held 1957: 18; Hoffner, Melchert 2008: 158). 92. Following (Miller 2013: 118-119). 93. Following (Otten 1988: 26-27; Beckman 1996: 116). 94. See in more detail (Sideltsev 2014b). It is significant that the only postverbal wh-word is also adverbial. Cf. (Luraghi forthcoming) who misses the real distribution. 95. Following (Goetze 1967: 148-149). 96. Following (Hoffmann 1984: 18-21; HED K: 219). assyriologie_109.indd 90 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 91 POSTVERBAL INDEFINITE PRONOUNS: (42) NH/lNS (CTH 106.A.1) Bo 86/299 rev. iii 28 mD mān=ma ANA NUMUN Tutḫaliya nakkiēš-zi kuitki if=but to progeny Tudhaliya weigh-3SG.PRS anything “If something becomes difficult for a descendant of Tadhaliya…”.97 POSTVERBAL RELATIVE PRONOUNS: (43) NH/lNS (CTH 383) KUB 21.19+ obv. ii 998 apa-t=ma ḪUL–lu uttar iya-t kui-š # that-ACC.SG.N evil.ACC.SG.N thing.ACC.SG.N do-3SG.PST who-NOM.SG.C “The one who did that evil thing…”.99 2.2.5.1. The last point –that focus can be postverbal– was discovered by (Bauer 2011).100 However, she attempted to describe all non-canonical V-X clauses as VTOPIC-XCONTRASTIVE FOCUS. The point is very important and I will dwell on it in some detail. First of all, it is obvious that some postverbal foci are really contrastive, as (44), (45) or (46): (44) OH/NS (CTH 8.A) KBo 3.34 obv. ii 8-12 (Askaliya was the lord in Hurma, he too was a man in every respect. They defamed him to my father; so he transferred him, brought him to Ankuwa, and made him an LÚAGRIG in Ankuwa) 1. šarkuš LÚ-eš₁₇ ēšta # 2. aki-š=ma=a-š101 tepšauwann-i # die-3SG.PST=but=he-NOM.SG.C in.diminished.circumstances-LOC.SG “ (1) He was an eminent/powerful/prominent man, (2) but he died in diminished circumstances”.102 Here the verb akiš “died” can be understood as topical only if the topic is inferable from the 103 context. There is no obvious mention of “dying” in the previous context. The post-verbal tepšauwanni “in diminished circumstances” is replacing focus: in the spirit of (Goedegebuure 2014) it replaces the šarkuš LÚ-eš₁₇ “an eminent/powerful man” in cl. 1. The particle -ma here is contrastive and its scope is over all cl. 2. The following example attests yet another subtype of contrastive focus, scalar focus: (45) NH/NS (CTH 293) KUB 13.35+ obv. ii 39-45 1. ANŠEḪI.A=wa kuiēš ḫarkun # 2. nu=wa=za ŪL kuitki daḫḫun # …104 da-ḫḫun=ma=wa=za © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) 3’. ŪL kuitki # take-1SG.PST=but=QUOT=REFL NEG something.ACC.SG.N “(1) Of the asses which I had (charge of) (2) I took for myself nothing. (Five asses died, and I replaced them from (my own) house. Five asses died from abuse. They will drive back here five jackasses (as replacements). Admittedly they haven't yet driven them here. Mr. AMAR.MUSEN the animal-driver worked them to death, and he hasn't yet replaced them) (3’) But I took nothing for myself”.105 In (45) cl. 2 and 3’ are identical but for the word order. Both negative pronoun and the verb are informational focus in cl. 2, in cl. 3’ the same proposition is repeated, this time with scalar focus on the negative pronoun (= traditional emphasis) and the verb being topical. 97. Following (Otten 1988: 22-23; Beckman 1996: 114). 98. The same word order is attested ibid. obv. ii 9, i 31. 99. Following (Singer 2002b: 742-743). 100. The first observations to this effect go as far back as (Hoffner 1977), but we owe the systematic research to (Bauer 2011). 101. It is noteworthy that in KBo 3.36+ ex. C the adversative particle is not employed: 18’ akiš=šan. 102. Following (Dardano 1997: 46-47, 97-100, 169; CHD L-N: 121, P: 58, Š: 269-270). 103. Topical information structure status is not normally acknowledged for verbs, but see for discourse linked verbs, e.g., (Szendröi 2003: 72-73) in Hungarian, and (Yanko 2011) for Russian topical verbs. 104. Presented only in translation. 105. Following (Hoffner 2003: 58; Werner 1967: 8-9). assyriologie_109.indd 91 16/11/15 14:12 – © PUF – 92 ANDREJ V. SIDELTSEV [RA 109 The following context attests a counterexpectant focus in the postverbal position: (46) NH/NS (CTH 89.A) KUB 21.29+ obv. i 11-19 1. URUTiliuraš URU-aš IŠTU UDKAM mḪantili dannattiš ēšta # 2. n=an ABI=YA mMuršiliš EGIR-pa wetet # 3. ašešanu-t=ma=an nāwi5 SIG5-in # resettle-3SG.PST=but=it not.yet well GIŠ 4. ašešanu-t=ma=an ap-ēl IŠTU NAM.RA TUKUL taraḫḫ-ant-it # resettle-3SG.PST=but=it he-GEN.SG with civil.prisoners weapon conquer-PRTC-INSTR “(1) From the time of Hantili the city of Tilura was empty, (2) my father, Mursili, rebuilt it. (3) But he had not yet resettled it in an appropriate way.106 (4) He resettled it with his civil prisoners conquered by arms. (But he took away those who were the ancient inhabitants of the city of Tilura. I, My Majesty, brought them back. I settled them again in the city of Tilura)”.107 The identical verbs in cl. 3 and 4 are inferable topics –resettling logically follows rebuilding in the previous clause. The other constituents (negative adjunct in cl. 3 and adjunct in cl. 4) are informational foci. The question is why non-canonical word order is used in both clauses. Despite presence of -ma108 I do not see any proper lexical contrast in either clause, I rather suppose it is unexpectedness that brought about the non- canonical word order. Unexpectedness is coded by adjuncts, thus they are counterexpectational foci: the way resettling was done is not expected and runs counter to what is considered to be the normal state of affairs by the author of the text. There is no lack of discourse continuity109 that might have conditioned verb fronting – resettling logically follows rebuilding. Thus the fact that postverbal constituents are contrastive or counterexpectational foci appears to be firmly established. However, as P. Goedegebuure (2014) showed, both contrastive and counterexpectational foci are regularly preverbal, see ex. (2) above. Moreover, in other examples the postverbal focus is not contrastive, as in (47): (47) MH/NS (CTH 264.A) KUB 13.4 rev. iii 77-79 (If someone sleeps with a woman,110 and his superior (or) his supervisor presses him (to work), he must speak! If he, however, does not dare to tell (his superior), let him speak to his colleague) 1. nu=za war(a)ptu=pat * * # 2. mān šekkantit=ma ZI-it parā dāi # 3. war(a)p-zi=ma=za nāwi # bathe-3SG.PRS=but=REFL not.yet “(1) But he must by all means bathe.111 (2) If, however, he intentionally postpones112 (the bath), © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) (3) (and) has not yet bathed”,113 (yet approaches the sacrificial loaves (and) libation vessels of the gods (while) unclean, […] it is a capital crime)”.114 A. Bauer (2011) assesses the verb in cl. 3 of this context as topical, whereas the negation is contrastive focus for her. She is undoubtedly right about the information structure status of the verb. However, due to the inherent polarity value of the negation marker, it is hard to unambiguously interpret the information structure function of the negation as contrastive and not simply as inherently polar. I suppose that the 106. Cf. “But he had not yet resettled it fully” (CHD L-N: 422); “Pero aun no la hecho habitar convenientemente” (González Salazar 1994: 160). 107. Following (González Salazar 1994: 160; CHD L-N: 422). 108. -ma has scope over all the clause and has nothing to do with verb fronting or non-canonical word order in this example. Cf. (Bauer 2011). 109. For which see below in fn. 130. 110. Following (Miller 2013: 261; Taggar-Cohen 2006: 81); (CHD L-N: 422; Š: 331): “his wife”. 111. (McMahon 2003: 220): “He still must bathe.” (Miller 2013: 261): “He shall bathe in any case.” 112. Following (CHD L-N: 422; McMahon 2003: 220). Cf. (Taggar-Cohen 2006: 81): “if he puts (it) forth knowingly”. 113. Following (CHD L-N: 422). Cf. (Taggar-Cohen 2006: 81): “while not yet washed”; (McMahon 2003: 220): “without bathing”. 114. Following (Miller 2013: 260-1; Taggar-Cohen 2006: 54-5, 81; McMahon 2003: 220; CHD L-N: 422; Š: 48, 331). assyriologie_109.indd 92 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 93 contrastive effect mostly comes from the fact that in (47) the verb is topical, whereas lack of contrast in (48) results from the verbs being informational foci: (48) MH/MS (CTH 268) KUB 23.82+ obv!. 23'-27' (Even if someone with evil intent seeks to stir up a rebellion against me, and I write to you, if you do not instantly come to my aid; or (if) you hear about this, if you do not tell it to My Majesty right away –even if he is not an enemy for you) n=an laḫ[ḫiy]a-tteni ŪL # CONN=him make.war-2PL.PRS NEG “(so that) you do not make war upon him (that shall be pu[t] under oath)”.115 Ex. (48) does not even display contrast on the clausal level, which is demonstrated by absence of -ma. The information structure of (48) is identical to that of the preceding clauses, from the same context, reproduced here in translation, some of which are also negative. However, there is no trace of non-canonical word order in any clause in the previous context. Naturally, one can always assume that the negation marker is scalar focus116 (= emphatic in the traditional terms) when it is postverbal, but this assumption remains totally ad hoc and is not supported by the only available analysis –contextual one. So I tend not to interpret the examples involving negation markers as containing postverbal replacing focus. The negation marker in such cases is simply informational focus for me. Still, for (47) the contrastive interpretation of the postverbal negation is not totally excluded: one can still consider the negation as contrastive if it replaces the positive polarity expressed in warptu=pat in cl. 1. A further proof that the postverbal focus need not always be contrastive, but can be just informational, comes from: (49) NH/NS (CTH 61.II.10) KBo 2.5+ rev. iii 39-42 (He defeated Aparru with three thousand troops (and) chariotry of his, and slew them). 1. nu ēpp-irr=a mekki # CONN capture-3PL.PST=and much 2. kuenn-ir[r=]a mekki # kill-3PL.PST=and much “(1) They captured in large numbers, (2) and they killed in large numbers”. (But Aparru escaped).117 Pace (Luraghi 1990: 101) there is absolutely no verb contrast in the example. The information structure of the example should probably be better interpreted as informational focus on mekki, which is an © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) adverb (CHD L-N: 247). The verbs of cl. 3 and 4 are inferable topics –they are inferred from the military © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) context. Still, both verbs host particles -ya, which, as (Melchert 2009a) showed, commonly trigger fronting of the constituent it cliticizes to. Double -ya means “both… and”, “neither… nor” under negation. The non-contrastiveness of postverbal constituents is even more obvious in case of indefinite and relative pronouns, subordinators and some adverbs: their information structure status is identical in the preverbal and postverbal positions. Generally speaking, in my corpus as well as in P. Goedegebuure’s corpus (Goedegebuure 2014) there are no examples where a constituent is informational focus or topic preverbally and replacing focus postverbally. Rather, contrastive focusing takes place in the preverbal position. Then, as I will argue shortly, the verb sporadically moves to the clause-internal position, past both the constituents which are focused preverbally and the constituents which are in the preverbal position, irrespective of their information structure status.118 This is arguably best seen in the following pair of examples involving preverbal (51) and postverbal (50) wh-words: (50) OH/NS (CTH 19.II.A) KBo 3.1+ obv. i 33-35, 39-40 1. nu ḪUL-lu utt[ar iē]r # 2. nu=kan mMuršilin kue[(nnir)] # 3. [(nu)] ēšḫar iēr # § 115. Following (Goedegebuure 2003: 261), cf. (CHD L-N: 9). 116. Presumably assessed so by (Goedegebuure 2003: 261): “(so) that you do not march against him”. 117. Following (CHD L-N: 248; HED H: 365; Goetze 1967: 188-189; Luraghi 1990: 101). 118. As indefinite and relative pronouns, negation markers and negative pronouns, subordinators. assyriologie_109.indd 93 16/11/15 14:12 – © PUF – 94 ANDREJ V. SIDELTSEV [RA 109 4. [nu mḪ]antiliš naḫšariyatati # … § 5’. [mān mḪ]antilišš=a URUTagarama ā[raš] # 6’. [nu memiski]wan [dāiš] # 7’. [k]-ī=wa iya-nun kuit # this-ACC.SG.N=QUOT do-1SG.PST why “(1) They made a bad deed: (2) they killed Mursili, (3) they made blood(shed). (4) And Hantili became afraid… (5’) When Hantili came to Tagarama, (6’) he started to ask: (7’) ‘Why did I do this?’”.119 It is obvious that the postverbal wh-word in cl. 7’ of this example is not just informational focus, it also codes surprise/unexpectedness: “why on earth did I do this?”. However, as P. Goedegebuure has shown (2009), this discourse function is normally coded by wh-words in the preverbal position, which is illustrated in (51): (51) MH/NS (CTH 42.A) KBo 5.3+ rev. iii 56’ zik=wa=kan apūn anda kuwat auš-ta # you=QUOT=LOC that.ACC.SG.C into why look-3SG.PST “Why did you look at that (woman)?”.120 Thus the postverbal placement of the wh-word in (50) is totally identical to the preverbal one in (51) as for the information structure and discourse functions. It is not conditioned by some special information structure status or discourse function of the constituent in question. But it is the verb in (50) which is not identical to the verb in the canonical clause-final position: all the constituents save the wh-word in clause 7’ are anaphoric to the previous situation, and thus they are topical. This pair of examples (50-51) shows beyond all reasonable doubt that the natural temptation to ascribe some emphatic function to all postverbal constituents as different from preverbal ones is not applicable to such cases. In all other (admittedly less clear) cases it is also completely ad hoc and not at all required by the context. The argument up to this point does not imply that postverbal foci cannot be contrastive foci. It implies that they need not be contrastive foci, prototypical contrastive foci are preverbal –i.e. preverbal and postverbal arguments are identical as for their information structure status. 2.2.5.2. Postverbal position: right dislocation? It is suggested in (Luraghi forthcoming) that postverbal constituents should rather be described not via verb movement to the left, but via right, postverbal, position of the constituents. In the cross-linguistic perspective it © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) can only be interpreted as right dislocations,121 see, e.g., right dislocated wh-words in Japanese: (52) (one) wh-phrase out of multiple wh-question: Dare-ga nomiya-de noN-da no, NANI-O? wh-NOM bar-LOC drink-TNS Qwh WH-ACC “Who drank at the bar, WHAT?” (Nakagawa, Asao, Nagaya 2008); (53) reduplicated wh-phrase out of wh-question: Mari-ga nani-oᵢ nomiya-de noN-da no, NANI-Oᵢ? M.-NOM wh-ACC bar-LOC drink-TNS Qwh WH-ACC “Whatᵢ did Mari drink at the bar, WHATᵢ?” (Yamashita 2010: 4.2). Right dislocations are constituents outside the matrix clause, so it would be highly surprising that functional constituents whose scope is over the clause should be placed outside the clause. This is not so for Japanese. The Japanese right dislocated wh-words are one of two wh-words. Hittite examples like this exist, see ex. (50) in (Hoffner 1995: 94), but they are clearly different from postverbal constituents: postverbal wh- words in Hittite are not just one of the two wh-words, as in Japanese. This is the main argument against assessing postverbal constituents as right dislocated. The same refers to subordinators, negation markers, 119. Following (Hoffmann 1984: 18-21; HED K: 219; CHD L-N: 345). 120 . Following (Hoffner, Melchert 2008: 352; hethiter.net/: CTH 42 (TX 17.11.2011, TRde 17.11.2011)). Cf. (Beckman 1996: 22). 121. (Skopeteas, Fanselow 2010; Hyman, Polinsky 2009 with ref.). assyriologie_109.indd 94 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 95 negative and relative pronouns: they have to stay in the same clause which they have scope over for the clause to be grammatical. Possibly, indefinite pronouns can in principle be described as right dislocations, but this will destroy the deep parallelism between the classes of constituents I advocated above. 3. VERB POSITIONS Thus the postverbal position of the constituents in 2.2.5 can only be explained by verb movement. Before I tackle the position of the verb in the examples immediately above, I will outline verb fronting strategies in Hittite. 3.1. Hittite attests not only verb movement to a clause-internal position, which is necessary to account for exx. from 2.2.5, but also verb movement to the clause initial position, i.e. proper verb fronting,122 see, e.g., (54) MH/MS (CTH 200) ABoT 1.60 obv. 5`-8` (As soon as I dispatched those tablets to Your Majesty, my lord,) m URU šalik-aš=ma=mu karuwariwar Niriqqaili-š LÚ Tabḫa[llu] # awake-3SG.PST=but=me following.morning Nerikkaili-NOM.SG.C man Taphallu “early the following morning Nerikkaili, the man from Taphallu, awoke me (and brought me the message)”.123 This verb fronting is triggered by discourse functions: the verb clause initially is either head-tail linking device or it marks unexpectedness,124 as in (54) above. From the functional point of view it is obvious that there is no contrast in (54) despite the presence of -ma –the clause with the non-canonical word order is just new information, i.e. it is thetic,125 thus the verb codes part of the broad informational focus. However, there is also a discourse function of the verb which conditions its movement to the edge of the clause: (54) comes from a letter where it introduces new information which appeared (immediately?) after the author communicated the previous state of affairs to the addressee. It is very likely that this new information made it necessary to send another letter before there was a reply to the previous one –the state of affairs beyond the customary one. Thus discourse discontinuity is marked.126 Some clause initial verbs are due to second position phenomena.127 In two constituent clauses, i.e. clauses where the only other constituent was the verb, the subordinator/relative/indefinite pronoun which has to © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) be in the second position could only be clause second behind the verb: (55) NH/NS (CTH 176) KUB 21.38 obv. 40’ (But my brother has not accepted in his own mind my status as a sister and my dignity, saying: “…, and do what should not/cannot be done!”… would I not? write…,) waḫnu-mi=an=kan kuwapi # turn-1SG.PRS=it=LOC when “when I change it?”.128 122. See (Luraghi 1990; Rizza 2011; Sideltsev 2014b). Cf. (Bauer 2011). 123. Following (Hoffner 2009: 177). 124. See for detail (Sideltsev 2014b). Cf. (Rizza 2011). In (Sideltsev 2014b) I supposed that contrastive foci are also clause-initial. Now in the light of (Goedegebuure 2014) I prefer to assess the examples like (59), see below, as contrastive focusing in the clause-internal position. 125. Only karuwariwar “early the following morning” can be assessed as a kind of topic-coding setting. 126. This discourse function, often realized as unexpectedness or surprise, is not introduced ad hoc to explain this example. It is assumed to be operative in many cases not involving verb fronting analyzed by (Goedegebuure 2003, 2009, 2014; Meacham 2000: 148-151, 203-204). Some of them are marked by the particle -ma (ibid). See along general lines for the function (Hopper 1979; Givón 1983). For an exact cross-linguistic parallel in Biblical Hebrew see (Baayen 1997) where loose clause linkage is textually realized as unexpectedness, surprise, contra-expectation, mostly when it operates on the events in the foreground. 127. See above for a general outline. 128. Following (Hoffner 2009: 285). Cf. (Beckman 1996: 127). assyriologie_109.indd 95 16/11/15 14:12 – © PUF – 96 ANDREJ V. SIDELTSEV [RA 109 So the verb is fronted to the left edge. However, as different from the movement in the examples in 2.2.5, the verb moves solely to provide the first position for the second position constituent. 3.2. Clause internal verbs Now after briefly reviewing verb fronting to the clause initial position, I will go back to verb movement to the clause-internal position in section 2.2.5. Part of the examples are clearly conditioned by the information structure status of the verb which is not identical to that in the canonical clause-final position, part of them are not information structure conditioned and appear to be identical to those in the canonical clause-final position. What is extremely curious, however, is that in both cases the verb moves to the clause-internal position past only the preverbal position, i.e. adverbial wh-phrases, relative, indefinite and negative pronouns, negation markers, low adverbs and adverbials, subordinators.129 If there is fully stressed subject and/or object in the clause, either topical or focal, it does not raise past them: (56) NH/lNS (CTH 106.A.1) Bo 86/299 obv. ii 87 apūn=ma=za MUNUS-an mDLAMMA-aš ANA PĀNI ABI=YA that.ACC.SG.C=but=REFL woman Kurunta.NOM.SG.C in to father=my da-tta=pat nawi # take-3SG.PST=EMPH not.yet “Kurunta during the reign of my father had not yet even taken that woman”.130 (57) NH/NS (CTH 68.F) KUB 19.54 rev. iv! 8’ m mān tuk=ma Kupanta-DLAMMA–a-n waggariya-zzi kuiški # if you.ACC.SG=but Kupanta-LAMMA-ACC.SG.C revolt-3SG.PRS anyone.NOM.SG.C “If anyone revolts against you, Kupanta-Kurunta…”.131 The only subjects or objects the verb moves past are instantiated by indefinite pronouns which are in the preverbal position, see above section 2.2.2. The only foci verbs that move past to the clause-internal position are instantiated by adverbs and adverbials. This would imply two focus positions for Hittite –the high one for subjects and objects and the low one for adverbs and adverbials. 3.3. Clause internal verbs: information status different from that in the original clause-final position In a number of cases the information structure status of the verb clause-internally is different from that clause- finally. © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) In the following context the verb is topical as it refers back to [nu=m]u kūrurriaḫta “he became hostile to me” in l. 29.132 129. As there are at least two distinct positions within the preverbal position (see section 2.2.1), the verb can move past only one of preverbal positions, as in (e) below, or in: (d) NH/NS (CTH 63.A) KUB 19.31+ rev. iii 27”-31” 1. [ … (tezzi)] # 2. […]xx=mu=kan arḫa daškanz[i] # 3. [nu=war=aš=za IN]A URU=ŠU EGIR-pa ašišanušk[anzi] # 4. nu k[ū]n memiya-n kuwat iya-tten QATAMMA # CONN this.ACC.SG.C matter-ACC.SG.C why do-2PL.PST in.this.way 5. nu=šši=kan apūš NAM.RAMEŠ ANA mTuppi-D10 arḫa dašketteni # “(1) [But Tuppi-Teššub] says, (2) “they are [still] taking [them] away from me, (3) they keep resettling [them i]n their city.” (4) So, why have you handled this matter in this way: (5) you keep taking those civilian captives away from Tuppi-Teššub?” following (Miller 2007: 126-7, 129-130). 130. Following (Otten 1988: 16; Garrett 1990: 79; CHD P: 226; Goedegebuure 2014: 260-261). 131. Following (Beckman 1996: 75, Hoffner, Melchert 2008: 287). 132. See Melchert (2005: 92) for the context. assyriologie_109.indd 96 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 97 (58) NH/NS (CTH 61.II.7.A) KBo 2.5+ rev. iii 34-35 m nu Aparru-š LÚ KUR[Kal]ašma kūruriaḫ-ta kuit CONN Aparru-NOM.SG.C man Kalasma get.hostile-3SG.PST as “As Aparru, the man of Kalasma, started hostilities, (he mobilized 3000 troops)”.133 Actually, corpus studies demonstrate that the canonical clause-final position of topical verbs heavily dominates: M. Molina (pers.comm.) observes that in the corpus of MH/MS letters (1422 clauses) there is not a single example of non-canonical clause-internal topical verbs, all the topical verbs are clause-final, both in two and more than two constituent clauses. Thus verb movement to the clause-internal position, as in (58), is optional, i.e. it occurs only in part of the cases where it could have occured.134 The fact that movement driven by information structure is optional, i.e. occurs only in some cases, is not surprising. For (Miyagawa 2006) “altering the focus potential of a sentence” is one of the few triggers for optional movement. Optional movement is commonly observed in the information structure sphere. Focusing can be done in situ, i.e. in the original position: narrowly focused subject and object are normally postverbal in Ossetic, whereas contrastive subject and object are preverbal (Lyutikova, Tatevosov 2009). In Georgian contrastive foci can be both preverbal (ex situ) and postverbal (in situ) (Skopeteas, Fanselow 2010). In Hungarian out of several foci only the first one is preverbal, the rest are postverbal, i.e. in situ (Szendröi 2003: 49-52; Kiss 2004: 91).135 Hungarian verb is focused in situ whereas noun phrases are focused ex situ (Szendröi 2003: 52-3). Hungarian topics can be both in situ (postverbally) and ex situ, preverbally (Szendröi 2003). See also (Melchert 2009a) for optional coding of information structure and discourse functions in Hittite by word order change. The following case also attests the verb in the clause-internal position with the information structure status different from that in the original clause-final position, although in this case the information structure status is different from the previous example: (59) MH/NS (CTH 261.B) KUB 13.2+ rev. iii 25–28 (Let him not, however, decide it (the case) for (his) superior. Let him not decide it for his brother, his wife or his friend. Let no one take a bribe.) 1. DINAM šarazzi katteraḫḫ-i lē # case upper.ACC.SG.N down-3SG.PRS PROHIB 2. katterr-a136 šaraz<zi>yaḫ-i lē # downy-ACC.PL.N up-3SG.PRS PROHIB “(1) (He) shall not make winning cases lose, © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) (2) (he) shall not make losing ones win, (you do what is right)”.137 In clauses 4 and 5 of this example, all the VPs are contrastive to each other: both direct object DINAM šarazzi and the verb katteraḫḫi in cl. 4 are contrasted with katterra and šaraz<zi>yaḫi in cl. 5. The direct objects are contrastive topics: DINAM “case” is established topic for the context, it was coded by the anaphoric pronoun -at “it” in the previous clauses. The verbs are contrastive foci. The negation is very obviously the only constituent which is not contrastive focus in either of these clauses.138 133. Following (Goetze 1967: 188-189). 134. See section 2.2.5.1. above for the discussion of the information structure status of postverbal constituents. 135. Cf. (Kiss 2004: 89; 2007). 136. Cf. (Pecchioli Daddi 2003: 154). 137. Following (CHD Š: 250). Cf. (Pecchioli Daddi 2003: 154-155; Hoffner, Melchert 2008: 342; CHD Š: 152). 138. There are two more analogous examples, known at least since (Garrett 1990: 79; CHD P: 226), where the verb is in the clause-internal position and hosts the particle -pat which is traditionally assessed as emphatic (CHD P: 224-6): (e) NH/lNS (CTH 106.A.1) Bo 86/299 obv. ii 31 m ūk=ma Tudhaliya-š LUGAL.GAL kuitman LUGAL-izziahhat=pat nawi # I=but Tudhaliya-NOM.SG Great King when become.king.1SG.PST.MED=EMPH not.yet “When I, Tudhaliya, Great King, had not yet even become king” following (Otten 1988: 16; Garrett 1990: 79; CHD P: 226). assyriologie_109.indd 97 16/11/15 14:12 – © PUF – 98 ANDREJ V. SIDELTSEV [RA 109 The following example can in my opinion also be classified as a variety of contrastive focus in the clause-internal position,139 although it is ambiguous due to the lack of subject or object in front of the verb: (60) NH/NS (CTH 85.1.A) KBo 6.29+ obv. ii 14-17 (I held up my hand to my lady Istar of Samuha. My lady Istar of Samuha helped me. Upper and lower regions darkened;140 she shook heaven and earth. And My lady Istar of Samuha stood to my side). 1. nu ḫatrā-nun kuedaš KUR.KU]R–e-aš # CONN write-1SG.PST which.DAT.PL land-DAT.PL 2. EGIR-an=wa=mu tiyatten # 3. n=at=mu EGIR-an t[ī]ēr # 4. ŪL=ya kuedaš KUR.KUR–e-aš ḫatrā-nun # NEG=and which.DAT.PL land-DAT.PL write-1SG.PST 5. nu ḫūman=pat ammētaz tiyat # “(1) The lands to which I wrote (2) “Stand by me”, (3) they stood by me. (4) And the lands to which I did not write, (5) all of them stood by me”.141 The key to the analysis of this context is the fact that cl. 1 and 4 are correlated: the countries that Hattusili wrote to are confronted with the countries to which he did not write to. There is no full-scale clausal contrast, rather some kind of what is termed in the Russian tradition as сопоставительное выделение (Testelets 2001), parallel focus for (Krifka 2007). The fact that the clauses are not contrasted, but rather parallel follows from the fact that there is a coordinative conjunction -ya, not the adversative -ma in the second clause (cl. 2). What is crucial for the analysis is the fact that both cl. 1 and cl. 4 attest a constituent in the non- canonical position –in cl. 1 it is the verb, in cl. 4 it is the negation. The conditioning for the non-canonical position is contrast: negation marker in cl. 4 is contrastive to the affirmative verb in cl. 1, which is verum focus:142 “the countries to which I did write” – “the countries to which I did not write”. Here it will be helpful to remember that the dedicated contrastive focus position for noun phrases is preverbal in Hittite, i.e. clause-internal.143 So, both verb and nominal phrases focusing occurs in a similar position. Other languages with preverbal focus also attest contrastive focusing of the verb clause-internally, e.g., Georgian: (61a) A: {Peter ate apples.} B: ara. P’et’er-ma GA-TAL-A vašl-eb-i. NEG Peter-ERG PR-peel-AOR.S.3.SG apple-PL-NOM ‘No, Peter peeled apples.’ © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) (61b) A: {Peter ate quickly.} B: ara. P’ et’ er-ma GA-TAL-A čkara. NEG Peter-ERG PR-peel-AOR.S.3.SG quickly ‘No, Peter peeled quickly.’ (Skopeteas, Fanselow 2010: ex. 22). (f) NH/lNS (CTH 106.A.1) Bo 86/299 obv. ii 87 mD apūn=ma=za MUNUS-an LAMMA-aš ANA PĀNI ABI=YA that.ACC.SG.C=but=REFL woman-ACC.SG.C Kurunta.NOM.SG.C in before father=my da-tta=pat nawi # take-3SG.PST=EMPH not.yet “Kurunta during the reign of my father had not yet even taken that woman” following (Otten 1988: 16; Garrett 1990: 79; CHD P: 226; Goedegebuure 2014: 260-261). More specifically, following Hoffner, CHD describes -pat in such clauses as counterexpectational with the meaning “even, even though” (CHD P: 224-6). Thus the verbs are likely to be counterexpectational foci. 139. Cf. (Sideltsev 2014b). 140. Following (HED M: 89). Or “And she m.-ed above and below” (CHD L-N: 202, Š: 249). 141. Following (Goetze 1925: 48-49; Held 1957: 19; Ünal 1974: 125; Garrett 1994: 37; CHD L-N: 202; HED M: 89). Cf. (HED K: 222): “I wrote to some lands”. (Parker 1990: 281): “whateve[r la]nds I wrote, ‘follow me!’, they followed me; those lands also, to which I wrote not –they, all of them, took their place by my side”. 142. I.e. focus on the truth value of a sentence (Féry 2007; Krifka 2007). 143. (Goedegebuure 2014). assyriologie_109.indd 98 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 99 Ex. (62) is at first sight destructive for the suggested analysis as it appears to attest contrastive focusing clause initially, i.e. to the left of subject: (62) OH/NS (CTH 291.II.a.B) KBo 6.26 obv. i 56-59 (§ 187/*73) 1. takku LÚ-iš GU₄-aš katta [wašt]ai # 2. ḫūrkil # 3. aki=aš # 4. LUGAL-an āški uwat[ezz]i # 5. kuen-zi=ma=an LUGAL–u-š # kill-3SG.PRS=but=him king-NOM.SG.C 6. ḫuiš[n]u-zi=y[a=an144 LUGAL–u]-š # let.live-3SG.PRS=and=him king-NOM.SG.C 7. LUGAL=i=ma=aš ŪL tiyaizzi # “(1) If a man sins (sexually) with a cow, (2) it is an unpermitted sexual pairing: (3) he will be put to death. (4) They shall conduct him to the king's court (lit. gate). (5) The king orders him killed, (6) or the king spares his life, (7) but he (the man) shall not appear (personally) before the king (lest he defile the royal person)”.145 According to S. Luraghi, the particle -ma here codes “unexpected or in some ways exceptional events”, the two fronted verbs also bear contrastive focus (Luraghi 1990: 98). I cannot fully accept this position: the fact the offender must die follows from the previous situation. -ma might rather be ascribed here a somewhat prospective function –it refers forward to cl. 7 where the real contrast is expressed. However, cl. 7 is already marked with the particle -ma in the contrastive meaning. The most realistic position in this case is to consider -ma to code discourse discontinuity: cl. 1-4 and 7 are narration, the main line of the story, whereas cl. 5-6 are a clear digression, the general state of affairs, and not part of the narration. The verbs are contrastively focused, but they are not contrastive to the previous situation, they rather form a kind of coordinated structure where all the contrast is between the two verbs. It does not extend outside of the coordinated structure. What I believe triggers the verbs’ fronting to the clause initial position in this case is not contrast, it is discourse discontinuity: the narrative line is broken by the two clauses (5 and 6) and then continued by cl. 7 with the canonical word order. Thus (62) is not analogous to verb movement to the clause- internal position in exx. (59-60) above, triggered by contrastive focusing, it is analogous to verb fronting to the clause initial position in ex. (54) above, triggered by discourse functions. Just as contrastive focusing of noun phrases occurs in the preverbal position whereas additive noun phrase focusing is clause initial (Goedegebuure 2014), contrastive focusing of verbs is clause-internal while © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) additive focusing of verbs is clause initial.146 I listed above exx. (59-60) with contrastive verb focusing clause- internally, now I will illustrate additive focusing of the verb clause initially by the two following examples: (63) NH/NS (CTH 42.A) KBo 5.3+ obv. i 8-16 1. nu=za zik mḪuqqanaš DUTUŠI=pat AŠŠUM BELUTIM šāk # 2. DUMU=YA=ya kuin DUTUŠI temi # 3. kūn=wa=za ḫūmanza šākdu # 4. n=an=kan ištarna tekkuššami # 5. nu=za ziqq=a mḪuqqanāš apūn šā[k] # § 6. namma=ma kuiēš ammel DUMUMEŠ=YA ŠEŠMEŠ=ŠU ammell=a ŠEŠMEŠ=[YA] # 7. n=aš=za aššuli AŠŠUM ŠEŠUTTIM Ù AŠŠUM LÚTAPPU[TIM] šāk # 8. namma=ma=za damain BELAM # 9.147 # kuiēš=aš kuiš [UN-aš] # 8a. ANA DUTUŠI EGIR-an arḫa lē kuinki šākti # 10. DUTUŠI-i[n=za=pat] šāk # D 11. paḫši=ya=an UTUŠ=I # protect=and=him, Majesty=My 144. Cf. ḫuiš[n]uziy[=an] (Luraghi 1990: 98). 145. Cf. (Luraghi 1990: 98; Hoffner 1997: 148). 146. This information structure status has to be added to the list of discourse functions clause-initial verbs attest in section 2.3.1. 147. Additional clause inside clause 8 –between 8 and 8а, which are actually one clause. assyriologie_109.indd 99 16/11/15 14:12 – © PUF – 100 ANDREJ V. SIDELTSEV [RA 109 “(1) You, Huqqana, recognize only My Majesty as overlord. (5) And recognize my son (2) of whom I, My Majesty, will say: (3) “Everyone shall recognize this one”, (4) and whom I will present among <his brothers(?)>. § (7) Furthermore, benevolently recognize (6) those who are my sons –his brothers– and [my] brothers in brotherhood and comradeship. (8) But beyond that you shall not recognize behind the back of My Majesty any other nobleman, (9) whoever he might be. (10) Recognize only My Majesty (11) and protect him, My Majesty!”.148 Cl. 11 here displays verb in the clause initial position149 accompanied by -ya “and”. The work on noun phrases (Melchert 2009a: 194; Goedegebuure 2014) showed that clause initial noun phrase hosting -ya “and” is additive focus. This suits perfectly well the information structure status of the clause initial verb in cl. 11: pahši “protect” is an additional new action to šāk “recognize” in previous cl. 10. Its role runs parallel to šāk, see (Melchert 2009a: 194). The following example is analogous: (64) MH/MS (CTH 286.6T) KUB 29.50 obv. i 23'-24' 1. [namm]a=aš katta pa[l]ahšiyanzi # 2. n=at 1⁄2 DA[NNA zallaz] uwanzi # 3. par[ḫ-a]nzi=ya=aš 4 ME gipeššar # gallop-3PL.PRS=and-they 400 ell “(1) [The]n they c[o]ver? them. (2) They [tr]ot half a mi[le] (3) and gallop 400 ells”.150 Here too the action of par[ha]nzi “make horses gallop” is seen as additional, whose role runs parallel to [zallaz] uwanzi “trot”, which is demonstrated by the use of the particle -ya. It is highly intriguing that the same particle cliticizes to the verb in the clause-internal position, when it marks scalar additive focus best translated as “even”: (65) NH/lNS (CTH 106.A.1) Bo 86/299 obv. ii 99-100 mān=ma DUMU=KA DUMU.DUMU=KA katta151 wašta-i=ya kuiški # if=but son=your grandson=your later sin-3SG.PRS=and someone.NOM.SG.C “But even if any son or grandson of yours later commits an offense…”.152 In the following case the verb is anaphoric to the previous situation, but at the same time it is scalar additive focus which again hosts -ya “even”: (66) MH/MS (CTH 147) KUB 14.1+ rev. 23 (‘I will either smite the country Hapalla or I will remove it together with civilian captives, cattle (and) sheep, and will give [it to] My Majesty’.) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) URU namma=ma=kan KUR Ḫapālla kuen-ta=ya ŪL # then=but=LOC country Hapalla smite-2/3SG.PST=and NEG “But subsequently you/he did not even smite the country Hapalla”.153 148. Following (Beckman 1996: 23-24, CHD Š: 29; Garrett 1990: 254; Hoffner, Melchert 2008: 409). 149. Besides linear clause-initial position, this placement follows from the fact that the verb is to the left of direct object. Clause internal verbs can be to the left of only canonically preverbal constituents (wh-phrases, relative pronouns, negative and indefinite pronouns, negation markers, low adverbs), whether their linear position is clause-initial or clause- internal. 150. Following (CHD P: 61) w. lit. the inverted clause is identical to ibid. obv. i 9'-10' (frgm.), rev. iv 13'-14'; 4T. KUB 29.49+ obv. i 21'-22'; with different distances: (6 ME): 6.T KUB 29.50 obv. i 11'-12'; 26'-27'; rev. iv 16'-17'; (2 ME): 3T. KUB 29.46+ obv. i 10'-11'; 1T. KUB 29.45(+) obv. i 13'; the numeral is in the lacuna: 7T. 165/q(+) 99 obv. i 42''. See (Kammenhuber 1961: 216 f.). 151. Pace (Garrett 1990: 79; Salisbury 2005: 85), katta is not here a preverb, rather an adverb meaning “later”, as follows from its meaning, which is identical to the unambiguous adverb, see Salisbury (2005: 83-85). It is true that katta is not normally located clause-internally in this meaning, but it is also very obvious that the aberrant syntactic behaviour should tip the balance in favour of katta being an adverb. A parallel for the clause-internal placement of an otherwise clause-initial/first adverb comes from (g) OH-MH/MS (CTH 262) IBoT 1.36 rev. iii 23 mān DUMU É.GAL=ma EGIR-anda me[mi]an uda-i # if son palace=but afterwards word.ACC.SG.C bring-3SG.PRS “After that, though, if a palace servant brings a message…” following (Miller 2013: 114-5). 152. Following (Otten 1988: 20-21; Beckman 1996: 113). assyriologie_109.indd 100 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 101 As it was observed by A. Bauer (Bauer 2011), negation is focus in clause 4, and all the rest of constituents are topical. However, verbs with the identical topical information structure status can also occur in the canonical position, i.e. clause-finally. Thus, once again, we see different information structure clause-internally and clause initially. It is conspicuous that -ya after mān in the meaning “even” does not show at first sight this distribution with other constituents –it is clause-first in the data collected in (CHD L-N: 155) and (Hoffner, Melchert 2008: 419). However, after a closer examination of data, it is obvious that in all the cases but one collected by CHD and Hoffner-Melchert, the enclitic is either on the only noun phrase in the clause, or on the verb which is the only constituent. In the only case where it is on the nominal part of the predicate and where other noun phrases are attested in the clause, there are no preverbal positions filled in the clause, so it is impossible to say whether the verb is clause-internal or not. Thus the examples like (65-66) are simply the only unambiguous ones and indicate clause-internal contrastive focusing of verbs. 3.4. Clause internal verbs: information status identical to that in the original clause-final position Finally, I will list the contexts where it is impossible to ascribe to the verb any information structure status different from that in the canonical position, i.e. clause-finally. To make the analysis more convincing I have limited it to the unambiguously clause-internal cases where the verb does not host the enclitics -pat, -ma or -ya and where it unambiguously follows from the context that the verb is part of the broad informational focus. The cases are quite numerous, especially in case of indefinite pronouns. Out of 15 cases with clause- internal verbs and indefinite pronouns, only two attest verbs accompanied by either focus particles -pat or -ya. One of them is (65) above. The other is: (67) NH/lNS (CTH 383) KUB 21.19+ rev. iv 14'-15' x x x x x ANA DINGIRMEŠ piran apē waškuwan-a to gods before that.NOM.PL.N sin-NOM.PL.N ēš-zi=pat kuitki nūwa # be-3SG.PRS=EMPH some.NOM/ACC.SG.N still “[And if] those sins somehow still exist before the gods…”.154 In 13 cases, the verb hosts no particles and is part of broad informational focus, thus its information structure status is identical to that clause-finally: (68) OH/OS (CTH 291.I.b.A) KBo 22.61+ obv. i 4 (§ 3) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) [takku LÚ-a(n našma MUNUS-an ELLAM if man-ACC.SG. or woman-ACC.SG.C free walaḫ-zi k)]uiški155 # strike-3SG.PRS somebody.NOM.SG.C “[If] anyone strikes a free [man] or woman…”.156 (69) MH/MS (CTH 41.II.2) KUB 36.127 rev. 16`-17` mān=wa kel ŠA ÌR=YA ēš-zi kuitki # if=QUOT this.GEN.SG of slave=my be-3SG.PRS anything.NOM.SG.N “If anything belongs to this slave of mine…”.157 (70) MH/MS (CTH 199) ABoT 1.65 obv. 8 mGIŠ GIDRU-DINGIRLIM–i-n tapaššīē-t kuitki # Hattušili-ACC.SG.C get.a.fever-3SG.PST something.ACC.SG.N “Hattušili contracted a bit of a fever”.158 153. Following (Luraghi 1990: 102; Beckman 1996: 149). 154. Following (Sürenhagen 1981: 98-99, Singer 2002a: 100S). 155. Restored from OH/NS copy KBo 6.3+ obv. i 6. 156. Following (Hoffner 1997: 18). 157. Following F. Fuscagni (ed.), hethiter.net/: CTH 41.II.2 (INTR 2011-08-24). 158. Following (Hoffner 2009: 243, 245). assyriologie_109.indd 101 16/11/15 14:12 – © PUF – 102 ANDREJ V. SIDELTSEV [RA 109 (71) NH/NS (CTH 176) KUB 21.38 obv. 16' ŠEŠ=YA=ma ammēd-aza NÍG.TUKU-ti kuitki # brother=my=but I-ABL enrich-2SG.PRS something.ACC.SG.N “Yet, my brother, you want to enrich yourself at my expense!”.159 (72) NH/lNS (CTH 106.A.1) Bo 86/299 obv. ii 75–76 mānn=a ANA NUMUN mDLAMMA nakkiš-zi kuitki # if=and upon progeny Kurunta weigh-3SG.PRS anything “And if anything weighs upon the progeny of Kurunta…”.160 (73) NH/lNS (CTH 106.A.1) Bo 86/299 rev. iii 28 mān=ma ANA NUMUN mDTutḫaliya nakkiēš-zi kuitki # if=but to progeny Tudhaliya weigh-3SG.PRS anything “If something becomes difficult for a descendant of Tadhaliya…”.161 (74) NH/NS (CTH 68.F) KUB 19.54 rev. iv! 8’ m [mān tuk=ma Kupanta-D]LAMMA–a-n waggariya-zzi kuiški # if you.ACC.SG=but Kupanta-Kurunta-ACC.SG revolt-3SG.PRS anyone.NOM.SG.C “[If] anyone revolts against [you, Kupanta-]Kurunta…”.162 (75) NH/NS (CTH 89.A) KUB 21.29(+) rev. iv 13 mā[n] šumeš=ma LÚMEŠ URULIM mazzallaša-duwari ku[it]ki163 # ? if you=but men city tolerate -2PL.PRS.MED something “But if you men of the city tolerate/condone (?) someone/something, (what will happen to you?)”.164 (76) NH/NS (CTH 255.1.A) KUB 21.42 + obv. i 33'-35' [n]ašma=šmaš EGIR-ziaz ištamaš-zi kuiški [k]uitki # or=you subsequently hear-3SG.PRS anyone.NOM.SG.C anything.ACC.SG.N “Or (if) anyone of you subsequently hears anything…”.165 (77) NH/NS (CTH 293) KUB 13.35+ obv. i 30 E[GI]R-zi=man=wa=za da-ḫḫi kuitki # later=IRR=QUOT=REFL take-1SG.PRS something.ACC.SG.N “Would I afterwards take something for myself?”.166 (78) NH/NS (CTH 70.1.A) KUB 14.4+ rev. iii 21 DAM=YA MUNUS.LUGAL idalawaḫ-ta kuitki # wife=my queen harm-3SG.PST somehow “Did my wife harm the queen in some way?”.167 (79) NH/NS (CTH 69.B) KUB 19.50+ rev. iii 12 m ziqq=a=an Manapa-DU-aš išt[amaš-ti ku]watqa # you.NOM.SG=and=it Manapa-Tarhunta-NOM.SG.C hear-2SG.PRS somehow “And you, Manapa-Tarhunta, somehow h[ear] about him…”.168 © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) (80) NH/lNS (CTH 383) KUB 21.19+ obv. ii 10-11 nu=kan mān DUTU URU Arinna GAŠAN=YA ANA INIM CONN=LOC if Sungoddess Arinna lady=My to word f Danuḫepa šer TUKU.TUKU–iš-ta kuitki # Danuhepa because.of become.angry-3SG.PST somehow “If somehow the Sungoddess of Arinna, my lady, became angry over the matter of Danuhepa…”.169 Similar cases are attested with relative pronouns. The clearest context is attested in several variants in the same group of texts. The first one is: 159. Following (Hoffner 2009: 283). 160. Following (Otten 1988: 18-19; Beckman 1996: 113; CHD L-N: 372). 161. Following (Otten 1988: 22-23; Beckman 1996: 114). 162. Following (Friedrich 1926: 140-141; Beckman 1996: 75). Cf. (Hoffner, Melchert 2008: 287). 163. González Salazar (1994: 165 fn. 35) thinks it also possible to restore ku[in]ki. 164. Following (Neu 1968: 115; CHD L-N: 215). Cf. (González Salazar 1994: 165). 165. Following (Miller 2013: 284-285). 166. Following (Hoffner 2003: 58; Werner 1967: 4-5). 167. Following (Hoffner 1995: 101; Miller 2014). 168. Following (G.Wilhelm-F.Fuscagni (ed.), hethiter.net/: CTH 69 (TX 17.02.2014, TRde 17.02.2014, Beckman 1996: 79). 169. Following (Singer 2002b: 743). assyriologie_109.indd 102 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 103 (81a) OH/OS (CTH 627.3.a) KBo 20.33+ obv. 12 [LÚKA]Š₄.E tarḫ-zi kui-š runner win-3SG.PRS who-NOM.SG.C 1 MA.NA KÙ.BABBAR Ù 2 NINDAwagadaš pianzi # “They give one mina of silver and two w.-breads to the runner who wins”.170 The second one is virtually identical: (81b) OH/OS (CTH 627.1.h.A) KBo 25.12+ ii 10–2 [LÚK]AŠ₄.E [(tarruḫ-zi kui)]š runner win-3SG.PRS who-NOM.SG.C 2 NINDAwagataš 1 MA[.NA KÙ.BABBA]R LUGAL-waš [(kiššarraz=š)]et dā[i] # “The runner who wins takes two w.-breads and one mina of silver from the hand of the king”.171 The most extensive context in the group of texts172 is provided by the late text in which, however, the verb provides the second position for the relative pronoun and thus is by itself of no relevance for the analysis: (81c) OH/NS (CTH 627.1.j.D) IBoT 1.13+ rev. v? 14’-18’ 1. EGIR-ŠU=ma 10 LÚMEŠ KAŠ₄.E uwanzi # 2. nu taraḫ-zi kui-š # CONN win-3SG.PRS who-NOM.SG.C 3. dān pēdāšša=a kuiš # 4. nu=šmaš 2 TÚGHI.A ÉRINMEŠ pianzi # “(1) Ten runners come next; (2) to the one who wins (3) and to the one who is in the second place (4) they give two ‘tunics’ ”.173 The relative NP in (81a-b) is topical whereas the verb is informational focus. Thus it is identical to the verb clause-finally. Analogous cases occur when low adverbs are involved: (82) NH/NS (CTH 40.II.2.E) KUB 19.10+ obv. i 7’ [a]mmel=ma ABI ABI=YA hattuliš-ta namma # my=but grandfather=my recover-3SG.PST again “But my grandfather once more became well”.174 (83) NH/lNS (CTH 40.II.3.F) KBo 14.3+ rev. iii 22' ABI ABI=YA=ma hattuliš-ta namma # grandfather=my=but recover-3SG.PST again © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) “But my grandfather once more became well”.175 The recovering of “my grandfather” occurs several times in the text,176 but each time it involves a new situation and is not an indication of topical status of the verb. Finally, clauses with subordinators are also attested: (84) MH/MS (CTH 186) HKM 7 rev. 23-25 m 1. tug=a=az Kaššū-n IDI maḫḫan you.ACC.SG=but=REFL Kaššū-ACC.SG know as 2. n=ašta ANA LÚMEŠ URUGašga kattan arḫa anku ŠUPUR “(1) And since he knows you, Kaššū, (2) write by all means secretly to the Kaška men”.177 Here the action of knowing is neither anaphoric, additive nor contrastive in the context of the letter, thus it is again identical to that in the canonical clause-final position. 170. Following (Neu 1980: 52; Singer 1984: 89; Probert 2006: 39). 171. Following (Neu 1980: 32; Singer 1984: 34; Probert 2006: 39). 172. See for the discussion of the contexts (Singer 1983: 103-104). 173. Following (Neu 1980: 52; Singer 1983: 103; 1984: 89; Probert 2006: 39). 174. Following (Güterbock 1956: 65; del Monte 2008: 16, 42-43). 175. Following (Güterbock 1956: 67; del Monte 2008: 18, 28-29). 176. (del Monte 2008: 28-29 fn. 34). Also in the fragmentary KUB 19.11+ rev. iv 17. 177. Following (CHD Š: 107). Cf. less convincingly (Alp 1991: 131; Hoffner 2009: 107). assyriologie_109.indd 103 16/11/15 14:12 – © PUF – 104 ANDREJ V. SIDELTSEV [RA 109 I believe the examples unambiguously demonstrate that part of clause-internal verbs are not information structure conditioned in their non-canonical position. This, naturally, raises the question why the verb is forced to move to a non-canonical position. An answer was attempted in (Sideltsev 2014b): Cf. (Huggard 2014). 4. CLAUSE STRUCTURE CONTRAINTS ON VERB MOVEMENT As follows from the previous sections, Hittite verbs are canonically clause-final, but they can be in several non-canonical position. Non-canonical verbs are extremely rare –they are attested once or twice per an average Hittite text, some texts do not attest them at all.178 Curiously, though, there are two constructions where verb ex situ is much more regular. The first structure is two-constituent clauses179 where the only other constituent besides the verb is preverbal in the canonical word order. The second structure is verb movement with second position phenomena. The constraint is not unique to Hittite. Similar clause structure constraints on movement are attested, e.g., according to (Hyman, Polinski 2009), in Aghem (a Western Grassfields Bantu language). It allows verb fronting with transitive verbs only if the object is externalised outside the verb phrase, i.e. right-dislocated (Hyman, Polinski 2009). If the object is not right-dislocated, verb fronting is impossible with two-places predicates. Another dependency of movement on clause architecture is attested in Hungarian where preverb movement out of dependent clause into the main one is acceptable only if the section of the embedded clause crossed by the raised prefix contains nothing but a complementizer –and perhaps a topic (Kiss 2004: 58 fn. 13). The only possible explanation for the predominance of verb movement in two-constituent clauses is some kind of on-going but not completed grammaticalization: in the majority of cases, just as seen by A. Bauer,180 the verb raises if the preverbal constituent is contrastive/scalar/counterexpectant focus and the verb is topical. This happens much less frequently in more-than-two-constituent clauses. The fact that the grammaticalization is not yet finished is demonstrated by the data attesting: (a) preverbal contrastive focus in more than two constituent clauses, (b) preverbal contrastive focus in two-constituent clauses, (c) postverbal informational focus constituents, (d) informational focus verbs in the non-canonical position.181 The fact that grammaticalization is at work is proved by the fact that the previous four types of data are much less frequently attested than VTOPIC-XCONTRASTIVE FOCUS clauses, although it is curious that a corpus study of MH/MS letters by M. Molina (pers. comm.) did not reveal any statistical correlation between topical verbs and non-canonical word order. The beginning grammaticalization is responsible for the greater frequency of verb movement in © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) two-constituent clauses. The reason for the grammaticalization might have been the realization by some speakers of Hittite that preverbal focus is the most deeply embedded argument or adjunct if there are several arguments/adjuncts.182 In the wish to recreate the preverbal contrastive focus position as the most deeply embedded in two-constituent clauses, some speakers might have reanalyzed X-XCONTR FOCUS-V clauses as V- XCONTR FOCUS along the lines of clause second constituents which were postverbal only in two-constituent clauses.183 The linguistic reality might have been the wish to make the contrastive focus the most deeply embedded constituent even if the only other overt constituent in the clause was the verb. 178. See (Luraghi 2012) who evaluates V-initial clauses at ~1% and (in her termninology) postverbal focused negations at < 1%. See now on a limited sample corpus (MH/MS letters) M. Molina (pers. comm.): in the corpus comprising 1422 clauses there are 15 clauses with non-canonical verb positions. Due to the rarity I suppose that no chronologically-based study breaking the data down into OH, MH, NH layers is possible. 179. (Hoffner 1977; Bauer 2011). 180. Although clearly not in all cases contra A. Bauer, see above. 181. See above for the data supporting each claim. 182. That the fact is not limited to Hittite speakers reanalysis is demonstrated by (Hyman, Polinsky 2009), where a description of information structure/syntax is based on embeddedness. 183. Cf. the explanation in (Sideltsev 2014b). assyriologie_109.indd 104 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 105 5. HITTITE CLAUSE ARCHITECTURE: A SUMMARY 5.1. Verb position Hittite verbs are canonically clause-final, but they can be in the non-canonical position. Non-canonical verbs are extremely rare –they are attested once or twice per an average Hittite text, some texts do not attest them at all. If the verb is clause-initial, either verbal arguments (both topics and foci) or constituents which are preverbal in the canonical word order can be postverbal. Preverbal constituents and verbal arguments are not simultaneously postverbal in any clause of my corpus.184 If the verb is clause-internal, verbal arguments (subject, object) are to the left of the verb, whereas what was the preverbal constituents in the canonical word order are to the right of the verb. In this case only low focus (adverbial wh-words, low adverbs or nominal part of the predicate) can be postverbal, high focus is always preverbal. If the verb is to the left of subject and object, its information structure and discourse functions are not identical to those in the original clause-final position: the verb is either in additive focus or it possesses discourse functions, such as unexpectedness or head-tail linking. If the verb is clause-internal (i.e., if subject and object are to the left of the verb, whereas the originally preverbal constituents are to the right of the verb), its information structure is either identical to that in the original clause-final position or different from it. In the latter case the verb is either contrastive focus or topic. Schematically, all available word order options can be represented as follows: (85) Canonical verb position: wh/Rel S-O V S-O wh/Rel/Neg V O-S wh/Rel/Neg V wh/Rel S-O Neg V wh/Rel/Neg V Non-canonical verb position: V S-O S-O V wh/Rel/Neg V wh/Rel/Neg V wh/Rel/Neg wh/comp V indef.pron./Neg185 It is significant that the following word orders are not attested: © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) *V S-O wh/Rel/Neg *V wh/Rel/Neg S-O *wh/Rel/Neg V S-O 5.2. Preverbal/postverbal position Hittite attests two focus positions186 –high (subject and object) and low (adverbs and adverbials, including adverbial wh-words). High focus is always preverbal, low focus can be preverbal or postverbal. If the verb is clause-internal, there is a very small number of constituents that can be postverbal. They are limited to the constituents which occupy the preverbal position in the canonical word order: adverbial wh-phrases, relative pronouns, some subordinators, indefinite pronouns, negation markers, negative pronouns, low adverbs and adverbials. Only the low focus can be postverbal. 184. Except clause second constituents. 185. All the examples above which attest S/O-Comp-V-wh/Rel/Neg involve second position subordinators. 186. In addition to the clause-initial one, see (Goedegebuure 2014). assyriologie_109.indd 105 16/11/15 14:12 – © PUF – 106 ANDREJ V. SIDELTSEV [RA 109 APPENDIX HITTITE DIALECTS? The above study is based on the corpus of diplomatic texts (treaties, letters, annals), laws and other legal documents, instructions, oracles, dreams, vows and prayers. The corpus of rituals and myths displays divergent clause architecture, with three main points of difference. It was argued above that in diplomatic texts, clause-internal verbs are to the right of preverbal constituents, but to the left of verbal arguments (subject and object), i.e. S-O-V-wh/rel/indef.pron/neg. However, there is a considerable number of counterexamples in rituals and myths, especially in the direct speech which attest clause-internal verbs to the left of subject and/or object (O-V-S/S-V-O/V-S-O, S-Prv-O-V, Prv-S-V).187 Second, E. Rieken (2011) argued that Hittite proper texts do not attest simultaneous fronting of preverb and verb, but, e.g., in the Mastigga ritual the pattern is very common. Third, the verb cannot raise past preverb in diplomatic texts. However, myths attest some counterexamples, as in (86) MH/MS (CTH 789) KBo 32.16 obv. ii 1, 3 mema-i=šši kui-š menaḫḫanda # speak-3SG.PRS=him who-NOM.SG.C against “Who speaks against him…”.188 How does one assess the distribution? There are several options: (a) question the conclusions drawn above on the basis of diplomatic texts only; (b) assess diplomatic and ritual usage as reflecting different dialects of Hittite; with the further option to assess only diplomatic usage to be genuine Hittite usage. Ritual and myth usage will then be construed as either directly borrowed or as reflecting stylistic reanalysis of original calquing from Hattian.189 I suppose the first option is untenable as there is very consistent usage in both diplomatic and ritual texts, different from each other and clearly reflecting some linguistic reality. As for the evaluation of ritual usage as calqued,190 it remains totally a matter of interpretation and is of no direct bearing on the object of the paper. An argument in favor of original calquing and later stylistic reanalysis might be the fact that non- © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) canonical word orders involving verbal arguments heavily dominate in the texts which are either clearly translated from Hattian or Hurrian or which display such aberrant usage as to suggest composition by a non- native speaker of Hittite.191 This is particularly likely in case of verb raising past preverb, unambiguously attested only in translations from Hurrian and in the “Egyptian” letter MH/MS (CTH 151) VBoT 1 obv. 18.192 In any case, the fact that Hittite rituals and myths have experienced extremely strong interference, both culturally and linguistically, is beyond any doubt.193 The distinction between diplomatic and ritual usage is clear-cut and sharp, but it is not absolute. There are sporadic examples in diplomatic texts which clearly attest ‘ritual’ usage. In this section I listed three characteristics which occur in ritual texts, but do not occur in diplomatic ones. Now I will provide sporadic examples from my diplomatic corpus which go with the ritual usage. 187. See (Sideltsev 2002; Rizza 2007, 2008, 2009; Rieken 2011; Sideltsev 2014a). 188. Following (Neu 1996). 189. (Bauer 2011; Rieken 2011). 190. Or originally calqued and later reassessed as stylistic marker. 191. See (Rizza 2007; 2008; 2009; Sideltsev 2002; 2010; Rieken 2011). 192. Which is normally considered to be a translation or to be composed by a non-native speaker of Hittite, see (Sideltsev 2002; 2010; Francia 2002b), cf. (Hoffner 2009: 274 with ref.). See also (Tjerkstra 1999: 172) for an example from a ritual. 193. See the discussion in (Rizza 2007; 2008; 2009; Sideltsev 2002; 2010; 2014a; Rieken 2011). assyriologie_109.indd 106 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 107 The position of the verb in the following example can be interpreted only as clause-internal, because the verb is to the right of the preverbal relative pronoun. But the postverbal position is occupied by the subject, a feature typical of ritual texts: (87) MH/MS (CTH 186) HKM 7 obv. 12-13 D kašma=tta karū=ya kui-t ḫatra-nun UTUŠ=I # hereby=you previously=and what-ACC.SG.N write-1SG.PST Majesty=My “Concerning what (I,) My Majesty, have written to you previously…”.194 The following example is also likely to attest clause-internal verb, due to the fact that the verb does not possess any discourse function, which typically triggers verb movement to the clause-initial position,195 but the verb is to the left of subject, which is not otherwise attested in diplomatic texts: (88) NH/NS (CTH 81.A) KUB 1.1+ rev. iv 9-13 (I will run before your husband. All Hattusa will defect to the side of your husband.) šallanu-nun=war=an kuit ammuk # raise-1SG.PST=QUOT=him since I.NOM.SG “Since I raised him, (I have never subjected him to an evil doom, to an evil deity)”.196 The following example from a prayer attests simultaneous preverb and verb movement,197 which is frequently attested only in the Mastigga ritual: (89) NH/NS (CTH 381.A) KUB 6.45+ rev. iii 13 D šarā=kan uw[(a-š)]i nepiš-aš UTU–u-š arun-az # up=LOC come-2SG.PRS heaven-GEN.SG Sungod-NOM.SG.C sea-ABL “You, Sungod of Heaven, arise from the sea”.198 Finally, there is even an example which may attest verb raising past preverb:199 (90) MH/MS (CTH 190) HKM 71 obv. 4 BELU # man=wa ūnna-tti kattan ## lord OPT=QUOT drive-2SG.PRS down “Lord, if only you would drive down here!”.200 Naturally, it is possible to assume that kattan “down” is not a preverb, but rather an adverb. As is well known,201 preverbs and local adverbs are a class of lexically identical constituents which are extremely difficult to break down into separate taxonomical classes, but the assumption would in this particular case be © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) entirely ad hoc. Besides, locatival adverbs, which kattan definitely is, behave in syntactic terms identically to preverbs.202 How do we evaluate such examples? They most likely reflect some marginal but quite expected interference between the two dialects. 194. Following (Hoffner 2009: 107). 195. The subordinator kuit “since” is in the second position, see above. 196. Following (Otten 1981: 24-25; Luraghi 1990: 99; CHD Š: 87). 197. Following (Salisbury 2005: 224). 198. Following (Singer 1996: 20, 39; Salisbury 2005: 224). 199. As is supposed by (Tjerkstra 1999: 172). 200. Following (Hoffner 2009: 227; Klinger 2001: 68; Alp 1991: 255; Tjerkstra 1999: 172). 201. (Melchert 2009b: 613) with ref., esp. to (Tjerkstra 1999: 172-173). 202. (Francia 2002a). assyriologie_109.indd 107 16/11/15 14:12 – © PUF – 108 ANDREJ V. SIDELTSEV [RA 109 REFERENCES Alp 1991 – Alp S. Hethitische Briefe aus Mașat-Höyük, Ankara, 1991. An De Vos 2013 – An De Vos. Die Lebermodelle aus Boğazköy. Wiesbaden, 2013. StBoTBh 5. Baayen 1997 – Baayen R.H. “The pragmatics of the ‘tenses’ in Biblical Hebrew”//Studies in Language. 21/2: 245-285. Bauer 2011 – Bauer A. “Verberststellung im Hethitischen”//Indogermanistik und Linguistik im Dialog – Akten der XIII. Fachtagung der Indogermanischen Gesellschaft vom 21. bis 27. September 2008 in Salzburg/hrsg. von Th. Krisch et al. Wiesbaden: 39-48. Becker 2014 – Becker K. Zur Semantik der hethitischen Relativsätze. Hamburg, 2014. Studien zur historisch- vergleichenden Sprachwissenschaft 5. Beckman 1996 – Beckman G. Hittite Diplomatic texts. Atlanta, Georgia, 1996. Büring 2009 – Büring D. “Towards a Typology of Focus Realization”//Information Structure/ed. by M. Zimmermann, C. Féry. Oxford, 2009: 177-205. CHD – The Hittite Dictionary of the Oriental Institute of the University of Chicago/ed. by H. Güterbock, H. Hoffner, and T. van den Hout. Chicago, 1980-. Dardano 1997 – Dardano Paola. L’Anedotto e il racconto in età antico-hittita: la cosiddetta “cronaca di palazzo”. Biblioteca di Ricerche Linguistiche e Filologiche 43. Roma: Il Calamo. del Monte 1986 – Monte G.F. del. Il trattato fra Muršili Il di Ḫattuša e Niqmepa' di Ugarit. Roma, 1986. Orientis Antiqui Collectio 18. del Monte 2008 – Monte G.F. del. Le gesta di Suppiluliuma. Traslitterazione, traduzione e comment, L’opera storiografica di Mursili II re di Hattusa, vol. I, Pisa, 2008. Féry 2007 – Féry C. “Information Structural Notions and the Fallacy of Invariant Correlates”//The Notions of Information Structure, Interdisciplinary Studies on Information Structure/ed. by C. Féry, G. Fanselow, M. Krifka. Potsdam, 2007. Working Papers of the SFB 632: 161-184. Francia 2002a – Francia R. Le funzioni sintattiche degli elementi avverbiali di luogo ittiti. anda(n), āppa(n), katta(n), katti-, peran, parā, šer, šarā. Roma, 2002. StudAs 1. Francia 2002b – Francia R. “Scelte di linguaggio e anomalie nell’ufficio dello scriba ittita”//Ufficio Documento: 349-359 Friedrich 1926 – Friedrich J. Staatsverträge des Ḫatti-Reiches in hethitischer Sprache 1. Teil: Die Verträge Muršiliš II. mit Duppi-Tešup von Amurru, Targašnalliš von Ḫapalla. Leipzig, 1926. MVAeg 31. Garrett 1990 – Garrett A. J. The Syntax of Anatolian Pronominal Clitics. Ph.D. Diss., Harvard University, 1990. Garrett 1994 – Garrett A. J. “Relative Clause Syntax in Lycian and Hittite”//Die Sprache. 36: 29-69. © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) Garrett 1996 – Garrett A. “Wackernagel’s Law and Unaccusativity in Hittite”//Approaching Second: Second Position Clitics and Related Phenomena/ed. by A.L. Halpern and A.M. Zwicky. Stanford, 1996: CSLI-LN 61: 85-134. Givón 1983 – Givón T. “Topic continuity in discourse: The functional domain of switch reference”//Switch reference and Universal Grammar/ed. by J. Haiman, P. Munro. Amsterdam, 1983. Typological Studies in Language 2: 51-82. Goedegebuure 2003 – Goedegebuure P. Reference, Deixis and Focus in Hittite. The demonstratives ka- “this”, apa- “that” and asi “yon”. PhD Diss. Amsterdam, 2003. Goedegebuure 2009 – Goedegebuure P. “Focus structure and Q-words questions in Hittite”//Linguistics. 47: 945-969. Goedegebuure 2014 – Goedegebuure P. The use of demonstratives in Hittite: deixis, reference and focus. Wiesbaden, 2014. StBoT 55. Goetze 1925 – Goetze A. Hattušiliš. Der Bericht fiber seine Thronbesteigung nebst den Paralleltexten. Leipzig, 1925. MVAG 29.3. Goetze 1967 – Goetze A. Die Annalen des Mursilis. Darmstadt, 1967. González Salazar 1994 – González Salazar J.M. “Tiliura, un ejemplo de la política fronteriza durante el imperio hitita (CTH 89)”//AuOr. 12: 159-176. Gračanin-Yüksek, İşsever 2011 – Gračanin-Yüksek M., İşsever S. “Movement of Bare Objects in Turkish”//Dilbilim Araştırmaları, 2011/I: 33-49. Güterbock 1956 – Güterbock H., “The Deeds of Suppiluliuma as Told by His Son, Mursili II”//JCS. 10/2-4: 41-68, 75-98, 107-130. assyriologie_109.indd 108 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 109 Hagenbuchner 1989 – Hagenbuchner A. Die Korrespondenz der Hethiter. 2. Die Briefe mit Transkription, Übersetzung und Kommentar. Heidelberg, 1989. THeth 16. HED – Puhvel J. Hittite Etymological Dictionary. Berlin – New York, 1984- Held 1957 – Held W. H. Jr. The Hittite Relative Sentence. 1957. Language 33, 4/2. Hoffmann 1984 – Hoffmann I. Der Erlaβ Telipinus. Heidelberg, 1984. THeth 11. Hoffner 1977 – Hoffner H. A. “Studies in Hittite Vocabulary, Syntax, and Style”//JCS 29 (1977): 151-156. Hoffner 1995 – Hoffner H. A. Jr. “About Questions”//Studio Historiae Ardens, Fs. Houwink ten Cate/ed. by Th. van den Hout, J. de Roos. Istanbul, 1995: 87-104. Hoffner 1997 – Hoffner H. A. Jr. The Laws of the Hittites: A Critical Edition. Leiden. DMOA 23. Hoffner 2003 – Hoffner H. A. Jr. “The case against Ura-Tarhunta and his father Ukkura”//The Context of Scripture, vol. 3 Archival Documents from the Biblical World/ed. by W. Hallo, Leiden-Boston, 2003: 57-60. Hoffner 2009 – Hoffner H. A. Jr. Letters from the Hittite Kingdom. Atlanta, 2009. Hoffner, Melchert 2008 – Hoffner H. A. Jr., Melchert C. A Grammar of the Hittite Language, Part 1: Reference Grammar. Winona Lake, Indiana, 2008. Hopper 1979 – Hopper P. “Aspect and foregrounding in discourse”//Syntax and Semantics, Vol. 12, Discourse and Syntax/ed. by T. Givón. New York, 1979: 213-241. Huggard 2011 – Huggard M. “On Wh-(Non)-Movement and Internal Structures of the Hittite Preposed Relative Clause”//Proceedings of the 22nd Annual UCLA Indo-European Conference/ed. by S. W. Jamison, H. Craig Melchert, and B. Vine. Bremen, 2011: 83–104. Huggard 2013 – Huggard M. “More on kuit: Causal clauses in Hittite”//ECIEC XXXII Poznan, June 21-24 2013 (handout). Huggard 2014 – Huggard M. “On Semantics, Syntax and Prosody” // ECIEC XXXIII, June 6-8 2014 (handout). Hyman, Polinsky 2009 – Hyman L., Polinsky M. “Focus in Aghem”//Information Structure: Theoretical, Typological, and Experimental Perspectives/ed. by M. Zimmermann, C. Féry. Oxford, 2009: 206-233. Kahnemuyipour, Megerdoomian 2008 – Kahnemuyipour A., Megerdoomian K., “Second Position, Focus and Multiple Spell-Out: Toward an Analysis of the Armenian Auxiliary”//Meeting Clitics. Barcelona, August 23, 2008. Kahnemuyipour, Megerdoomian 2011 – Kahnemuyipour A., Megerdoomian K. “Second-Position Clitics in the vP Phase: The Case of the Armenian Auxiliary”//Linguistic Inquiry. 42/1: 152-162. Kammenhuber 1961 – Kammenhuber A. Hippologica Hethitica. Wiesbaden, 1961. Kim 1988 – Kim Alan Hyun-Oak. “Preverbal focusing and type XXIII languages”//Studies in Syntactic Typology/ed. by M. Hammond, E. Moravcsik and J. Wirth. Amsterdam, 1988: 147-169. © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) Kiss 2004 – Kiss K. É. The Syntax of Hungarian. Cambridge, 2004. Krifka 2007 – Krifka M. “Basic notions of information structure”//Interdisciplinary Studies on Information Structure, Interdisciplinary Studies on Information Structure/ed. by C. Féry, G. Fanselow and M. Krifka. Potsdam, 2007. Working Papers of the SFB 632: 13-56 Klinger 2001 – Klinger J. “Briefe aus hethitischen Archiven”//TUAT Erg., 2001: 65-69. Luraghi 1990 – Luraghi S. Old Hittite Sentence Structure. London – New York, 1990. Luraghi 2012 – Luraghi S. “Opposite tendencies of placement rules for P2 clitics and the position of the finite verb”//SWL V, 1-4 October 2012 Dubrovnik (handout). Luraghi forthcoming – Luraghi S. “Anatolian syntax: The simple sentence”//Comparative Indo-European Linguistics/ed. by J. Klein and M. Fritz. Berlin, New York. Lyutikova, Tatevosov 2009 – Lyutikova E., Tatevosov S. “The clause internal left edge: Exploring the preverbal position in Ossetian” // International Conference on Iranian Linguistics 3, University of Paris III: Sorbonne Nouvelle, Paris, 11-13 Sept. 2009, URL: http://otipl.philol.msu. ru/staff/people/tatevosov/ clause_internal_left_edge_4.pdf. Meacham 2000 – Meacham, M.D. A Synchronic and Diacronic Functional Analysis of Hittite -ma, Ph.D. Diss., Berkeley, 2000. McMahon 2003 – McMahon G. “Hittite Canonical Compositions - Instructions: Instructions to Priests and Temple Officials”//The Context of Scripture, vol. 1/ed. by W. Hallo, Leiden-Boston, 2003: 217-221. Melchert 2005 – Melchert H. Craig “Latin īnsolēscō, Hittite šulle(šš)- and PIE Statives in -ē-”//Hŗdā mánasā. Studies presented to Professor Leonard G. Herzenberg on the occasion of his 70-th birthday/ed. by N.N. Kazansky. Saint Petersburg, 2005: 90-98. assyriologie_109.indd 109 16/11/15 14:12 – © PUF – 110 ANDREJ V. SIDELTSEV [RA 109 Melchert 2009a – Melchert H. Craig “Discourse Conditioned Use of Hittite -ma”//Pragmatische Kategorien. Form, Funktion und Diachronie, Akten der Arbeitsteigung der Indogermanischen Gesellschaft vom 24. bis 26. September 2007 in Marburg/hrsg. von E. Rieken, P. Widmer. Wiesbaden: 187-195. Melchert 2009b – Melchert H. Craig. “Local Adverbs in Hittite: Synchrony and Diachrony”//Language and Linguistics Compass. 3/2: 607–620. Miller 2007 – Miller J. “Muršili II’s dictate to Tuppi-Teššub’s Syrian antagonists”//KASKAL. 4: 121-152. Miller 2013 – Miller, J. Royal Hittite Instructions. Atlanta, 2013. SBL Writings from the Ancient World 31. Miller 2014 – Miller, J. “Mursili II’s Prayer Concerning the Misdeeds and the Outstanding of Tawananna”//Procedings of the VIII International Congress of Hittitology/ed. by P. Taracha. Warsaw, 2014: 516-557. Miyagawa 2006 – Miyagawa Sh. “Moving to the Edge”//Proceedings of the 2006 KALS-KASELL International Conference on English and Linguistics, Busan: 3-18. Mouton 2007 – Mouton A. Rêves hittites. Leiden, Boston, 2007. CHANE 28. Nakagawa, Asao, Nagaya 2008 – Nakagawa N., Asao Y., Nagaya N. “Information Structure and Intonation of Right- Dislocation Sentences in Japanese”//Kyoto University Linguistic Research 27, 2008. Neu 1968 – Neu E. Interpretation der hethitischen mediopassiven Verbalformen. Wiesbaden, 1968. StBoT 5. Neu 1980 – Neu E. Althethitische Rituale in Umschrift. Wiesbaden, 1980. StBoT 25. Neu 1996 – Neu E. Das hurritische Epos der Freilassung Bd. 1: Untersuchungen zu einem hurritisch-hethitischen Textensemble aus Ḫattuša. Wiesbaden, 1996. StBoT 32. Otten 1981 – Otten H. Die Apologie Hattusilis IlI. Das Bild der Überlieferung. Wiesbaden, 1981. StBoT 24. Otten 1988 – Otten H. Die Bronzetafel aus Bogazkoy. Wiesbaden, 1988. StBoT Bh. 1. Parker 1990 – Parker V. “Reflexions on the Career of Hattušili III until the Time of his Coup d’État”//AoF. 26/2: 269-290. Pecchioli Daddi 2003 – Pecchioli Daddi F. Il vincolo per i governatori di provincia. Pavia, 2003. SM 14; Series Hethitica 3. Probert 2006 – Probert, Philomen, 2006. Probert Ph. “Clause Boundaries in Old Hittite Relative Sentences” Transactions of the Philological Society vol. 104: 1 (2006): 17-83. Rieken 2011 – Rieken E. “Verbanfangsstellung in den hethitischen Übersetzungstexten”//Indogermanistik und Linguistik im Dialog – Akten der XIII. Fachtagung der Indogermanischen Gesellschaft vom 21. bis 27. September 2008 in Salzburg/hrsg. von Th. Krisch et al. Wiesbaden: 498-507. Rizza 2007 – Rizza A., I pronomi enclitici nei testi etei di traduzione dal hattico, Pavia, 2007. SM 20. Rizza 2008 – Rizza A. “"Cercato in traduzione": su di un passo problematico della bilingue «SÌR para tarnum(m)aš»”//Traduzione di traduzioni e tradizioni di traduzione. Atti del IV incontro «Orientalista – Associazioni Orientalisti», Pavia, 19-21 aprile 2007/ed. by B. Bellucci, E. Jucci, A. Rizza et B. M. © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) Tomassini Pieri. Milano, 2008: 53-79. Rizza 2009 – Rizza A. “Left and Right Periphery in Hittite. The case of the translations from Hattic”//Pragmatische Kategorien. Form, Funktion und Diachronie, Akten der Arbeitsteigung der Indogermanischen Gesellschaft vom 24. bis 26. September 2007 in Marburg/hrsg. von E. Rieken, P. Widmer. Wiesbaden: 275-286. Rizza 2011 – Rizza A. “The ritual for the Royal Couple CTH 416. Syntax of non verbal predicates and numerals”//Anatolistica, indoeuropeistica e oltre nelle memorie dei seminari offerti da Onofrio Carruba (anni 1997-2002) al Medesimo presentate. Milano, 2011. Antiqui Aevi Grammaticae Artis Studiorum Consensus. Series maior I: 13-37. Roos 2005 – Roos J. de. “Die Hethiter und das Ausland”//Motivation und Mechanismen des Kulturkontaktes in der späten Bronzezeit/hrsg. von D. Prechel. Firenze, 2005. Eothen 13: 39-58. De Roos 2007 - Roos J. de. Hittite Votive Texts (PIHANS CIX), Leiden. Salisbury 2005 – Salisbury D. Local adverbs in Neo-Hittite. ProQuest company, 2005. Samek-Lodovici 2005 – Samek-Lodovici Vieri. “Prosody-Syntax Interactions in the Expression of Focus”//Natural Language and Linguistic Theory. 2005: 687-755. Samek-Lodovici 2006 – Samek-Lodovici Vieri. “When Right Dislocation Meets the Left-Periphery”//Lingua. 116: 836- 873. Samek-Lodovici 2009 – Samek-Lodovici Vieri. “Topic, Focus, and Background in Italian Clauses”//Focus and Background in Romance Languages/ed. by A. Dufter and D. Jakob. Amsterdam, 2009: 333-358. Sideltsev 2002 – Sideltsev A. “Inverted Word Order in Middle Hittite”//Anatolian Languages/ed. by V.V. Shevoroshkin, P.J. Sidwell. Canberra, 2002. Association for the History of Language Studies in the Science & History of Language 6: 137-188. assyriologie_109.indd 110 16/11/15 14:12 – © PUF – 2015] HITTITE CLAUSE ARCHITECTURE 111 Sideltsev 2010 - Sideltsev A. “Proleptic Pronouns in Middle Hittite”//Babel und Bibel 4. Proceedings of the 53e Rencontre Assyriologique Internationale. Vol. I. Part 1. Language in the Ancient Near East. Ed. by L. Kogan, N. Koslova, S. Loesov, and S. Tishchenko. Eisenbrauns, Winona Lake, Indiana, 2010: 53-79. Sideltsev 2014a – Sideltsev A. “The origin of right dislocations”//Procedings of the VIII International Congress of Hittitology/ed. by P. Taracha. Warsaw, 2014: 827-872. Sideltsev 2014b – Sideltsev A. “Clause internal and clause leftmost verbs in Hittite”//AoF. 41/1: 80-111. Sideltsev 2014c – Sideltsev A. “Wh-in situ in Hittite”//Tipologija morfosintaksicheskix parametrov. Proceedings of the Conference “Typlogy of Morphosyntactic Parameters 2014”. Moscow: MGGU: 198-222. Sideltsev forthcoming - Sideltsev A. “Clause Internal Left Periphery in Hittite: a Formal Description”. Singer 1983 – Singer I. The Hittite KI.LAM Festival. Part 1. Wiesbaden: Harrasowitz. StBoT 27. Singer 1984 – Singer I. The Hittite KI.LAM Festival. Part 2. Wiesbaden: Harrasowitz. StBoT 28. Singer 1996 – Singer I. Muwatalli’s Prayer to the Assembly of Gods Through the Storm-God of Lightning (CTH 381). Atlanta, Georgia, 1996. Singer 2002a – Singer I. Hittite Prayers. Atlanta, 2002. SBL Writings from the Ancient World 11. Singer 2002b – Singer I. “Danuhepa and Kurunta”//Anatolia antica. Studi in memoria di F. Imparati. T. II, Firenze, 2002: 739-751. Skopeteas, Fanselow 2010 – Skopeteas S., Fanselow G. “Focus in Georgian and the expression of contrast”//Lingua. 120 (6): 1370–1391. Sturgeon 2008 – Sturgeon A. The left periphery: the interaction of syntax, pragmatics and prosody in Czech. Amsterdam- Philadelphia, 2008. Linguistik Aktuell 129. Sürenhagen 1981 – Sürenhagen D. “Zwei Gebete Hattušilis und der Puduhepa”//AoF. 8, 1981: 83-168. Szendröi 2003 – Szendröi K. “A stress-based approach to the syntax of Hungarian focus”//The Linguistic Review. 20: 37– 78. Taggar-Cohen 2006 – Taggar-Cohen A. Hittite Priesthood. Heidelberg, 2006. THeth 26. Testelets 2001 – Testelets Ya. G. Vvedenije v obschij sintaksis. Moscow, 2001. Tjerkstra 1999 – Tjerkstra, F.A. 1999, Principles of the Relation between Local Adverb, Verb and Sentence Particle in Hittite. Groningen, 1999. Cuneiform Monographs 15. Travis 2005 – Travis L. “VP-internal structure in a VOS language”//Verb First. On the syntax of verb initial languages/ed. by A. Carnie, H. Harley, S. Dooley. Amsterdam-Philadelphia, 2005. Linguistik Aktuell 129: 203-226. Ünal 1974 – Ünal A. Hattušili III. Teil 1. Hattušili bis zu seiner Thronbesteigung. Bd. 2: Quellen und Indexes. Heidelberg, 1974. THeth 4. © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) van der Wal 2012 – van der Wal J. “Why does focus want to be adjacent to the verb?”//Workshop ‘Parametric variation in discourse configurationality’, 28-29 August 2012, Meeting Societas Linguistica Europaea, Stockholm. Vikner 1995 – Vikner S. Verb movement and expletive subjects in the Germanic languages. Oxford, 1995. Werner 1967 – Werner R. Hethitische Gerichtsprotokolle. Wiesbaden, 1967. StBoT 4. Yamashita 2010 – Yamashita H. “What Makes Right Dislocation of Wh-phrases in (Tokyo) Japanese (Im)Possible?”//WCCFL 28 Online Proceedings. 2011. Yanko 2011 – Yanko T.E. Intonatsionnyje strategii russkoj rechi v sopostaviteljnom aspekte. Moscow, 2011. Zubizarreta 2010 – Zubizarreta M.L. “The Syntax and Prosody of Focus: the Bantu-Syntax Connection”//Iberia: An International Journal of Theoretical Linguistics. Special Issue on Information Structure, Vol. 2.1: 1-39. Zubizarreta, Vergnaud 2005 – Zubizarreta M.L., Vergnaud J.-R. “Phrasal Stress, Focus, and Syntax”//The Syntax Companion/ed. by M. Everaert and H. van Riemsdijk. Blackwell, 2005: 522-568. KEY WORDS: Hittite, syntax, clause architecture, word order, focus, verb position, preverbal position. assyriologie_109.indd 111 16/11/15 14:12 – © PUF – 112 ANDREJ V. SIDELTSEV [RA 109-2015] ABSTRACT The paper aims to provide a comprehensive description of Hittite clause structure.203 The picture that emerges is quite different from both the view of Hittite clause architecture as codified in (Hoffner, Melchert 2008) and as documented in the parallel line of research (Luraghi 1990; 2012; forthcoming). The paper focuses on two key features of Hittite clause architecture: (a) preverbal vs. clause initial vs. clause second positions; (b) verb’s positions in the clause, although in-depth study of these aspects involves examination of virtually every significant feature of Hittite syntax. Preverbal position is constituted by wh-words, subordinators, negation markers, negative, indefinite and relative pronouns204 as well as some adverbs, only part of these constituents can alternatively be clause initial or second. Contrastive focus is normally preverbal, contrastive topic is clause initial. Two focus positions are distinguished in a Hittite clause205 –high (subjects and objects) and low (adverbs, adverbials). Wh-words, subordinators and relative/indefinite pronouns can also be optionally postverbal. It is significant that only lower focus can be postverbal, never high focus, even though in the canonical word order both high and low focus is preverbal. No information structure difference is detectible between the preverbal and postverbal positions. It is shown that non-canonical positions of the verb can be described by two movements to the left from the canonical clause-final position: (a) to the clause-internal position which follows subject and object, both topical and focal, on the one hand, and precedes what is in the canonical word order the preverbal position,206 on the other, producing V-wh/Neg/Rel, S- O-V-wh/Neg/Rel word orders; (b) to the clause leftmost position, producing V-S-O word order. *V-wh/Neg/Rel-S-O or *wh/Neg/Rel-V-S-O word orders are not attested in my ‘diplomatic’ corpus. The last point raises an important question of sociolinguistics of the Hittite language, namely evidence for (idio)lects. Dr. Andrei Sideltsev - Head of the Department of Anatolian and Celtic Languages, Institute of Linguistics, Russian Academy of Sciences, 125 009 B. Kislovskij per 1/1, Moscow Russia

© Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) © Presses Universitaires de France | Téléchargé le 23/01/2022 sur www.cairn.info (IP: 107.21.170.183) 203. Building upon recent work on Hittite syntax (Bauer 2011; Rizza 2011; Huggard 2011, 2013, 2014; Goedegebuure 2009; 2014; Sideltsev 2014b, 2014c). 204. Negative pronouns are negation marker + indefinite pronoun. 205. Apart from clause-initial one, for which see (Goedegebuure 2014). 206. Thus the position is actually postverbal. assyriologie_109.indd 112 16/11/15 14:12