The Missing Link: There is no Copula in Akkadian

The Missing Link There is no Copula in Akkadian1 ØYVIND BJØRU (University of Texas, Austin) & NA'AMA PAT-EL (University of Texas, Austin) Abstract: Unlike many West-Semitic languages, for most of its history Akkadian did not develop a copula. Some scholars (GAG §196c; Hämeen-Anttilla 2000: 107; de Ridder 2018 §676) argue that at least in Neo-Assyrian, and possibly in Middle Assyrian, a copula, on the basis of the 3rd person personal pronoun, is used. In this paper, we will argue that the 3rd person personal pronoun in Neo-Assyrian has not, in fact, grammaticalized to become a copula. We will show that the pronoun in Akkadian shows none of the expected behaviors of a copula and therefore should not be considered one. Keywords: syntax, copula, Neo-Assyrian, historical linguistics Contact: [email protected]; [email protected] While many Semitic languages developed a copula, either nominal, verbal, or both, for most of its attested history, Akkadian shows very few indications that it possesses such a feature. The two main candidates for a copula in Akkadian are the enclitic particle =mā and the independent pronoun šū/šī (GAG §126c). The function of =mā has been discussed extensively and has been shown to possibly be a marker of the predicate rather than a copula (e.g., Cohen 2000). However, at least for Neo-Assyrian (NA), prominent scholars have argued that a pronominal copula is available. In this paper we aim to reevaluate the evidence in order to determine whether Akkadian has indeed developed a copula, and if so, at which stage and how. For clarity, we will follow Pustet’s definition of copula: “A copula is a linguistic element which co-occurs with certain lexemes… when they function as predicate nucleus. A copula does not add any semantic content to the predicate phrase it is contained in.” (Pustet 2003: 5) In Akkadian, adjectival predicates assume a predicative form whose subject is marked with personal suffixes (“stative”). This strategy was not available, at least 1 We use the following abbreviations: Cop(ula); FAOS 19: Kienast & Volk 1995; GAG: von Soden 1995; KAV: Schroeder 1920; MA: Middle Assyrian; NA: Neo-Assyrian; NP: noun phrase; o(bverse); OA: Old Assyrian; OAkk: Old Akkadian; Prag 1: Hecker & Matouš 1998; Pred(icate); Pron(oun); r(everse); SAA: State Archives of Assyria; Subj(ect); Top(ic); VAT: Sigla for tablets in Vorderasiatisches Museum, Berlin; X/k: excavation seasons at Kültepe. Copyright Institut für Orientalistik Wien 2022 24 Ø. Bjøru & N. Pat-El originally, for nouns, prepositional phrases, or adverbs (Huehnergard 1986); nominal sentences with such predicates remain unmarked for most of the history of Akkadian. OA am-tum a-ma-sú amtum amas su ‘The slavegirl is his slavegirl.’ (Prag 1, 605: 12) However, in some cases a personal pronoun is also used even when an overt nominal subject is expressed in the construction, and such clauses are known as tripartite nominal sentences. NA ú-ma-a mulSAG.ME.GAR d30 šu-u ūmâ Sagmegar Sîn šū ‘Sagmegar is Sîn, i.e., Jupiter is the moon.’ (SAA 10, 43, r 4–5) This type of clause is more common in NA than in earlier dialects, and it is therefore no surprise that in grammatical treatments of this dialect a question arose as to what function the pronoun performs in such sentences (GAG 224 §126f).2 In the only full grammar of NA, Hämeen-Anttila remarks that it is preferable to treat the pronoun as a copula rather than a subject (2000: 107) but does not provide a rationale for this position. Based on an exhaustive list of tripartite sentences,3 Vinnichenko offers a syntactic argument for analyzing the pronoun as a copula in sentences with a nominal predicate (2016: 81–83). Since most of the nominal sentences on her list show Subject-Predicate (S-P) word order, and the 3rd person pronoun is in final position, she identifies the pronoun as non-subject, and therefore as a copula. Vinnichenko does not believe that the copula in NeoAssyrian is a result of Aramaic influence, because the Neo-Assyrian copula is, according to Vinnichenko, at a much more advanced stage of grammaticalization, while the copula in Aramaic is only nascent (ibid. 179–80).4 She suggests, rather, that the copula was already established in Neo-Assyrian and that it was likely inherited from Middle Assyrian; namely this is an internal development, not a result of external influence. There are two basic arguments in support of the analysis of certain uses of the pronoun as a copula: First, as was argued in previous literature, the pronoun is found in a position that is otherwise occupied by the predicate. Second, in sentences with an overt nominal subject and a predicate, the pronoun is redundant and should not be analyzed as the subject. If the pronoun is indeed a copula we 2 3 4 The form of the singular 3rd person pronouns in NA is variable: šū/šūtu and šī/šīti, both occur multiple times in similar contexts in our corpus (Luukko 2004: 133). For an attempt to distinguish differences between them, see Hämeen-Anttila (2000: 45–47). The list contains less than 20 examples and excludes, without explanation, any example with nominal predicate and a demonstrative subject. Note that Zielenbach (2021) lists copula function already in Imperial Aramaic. Copyright Institut für Orientalistik Wien 2022 The Missing Link 25 should be able to identify the rules governing this new category, for example, under which conditions is it obligatory, does it have allomorphs, etc. The first argument, that a pronoun is in final position and therefore must be part of the predicate, should be dismissed. Although in Akkadian the predicate is typically in final position, in verbless clauses with a personal (but not demonstrative) pronoun as the subject, it is in fact the pronominal subject that is in the final position. The order Predicate-Pronominal Subject is attested from the earliest stages of Akkadian (Huehnergard 1986: 223; Kouwenberg 2017: 714), and is still the typical word order in NA. O akk ù-la a-bí at-tá ulā abī atta ‘Are you not my father?’ (FAOS 19, Ki1: o 5) OA5 ší-ir-kà da-ma-kà a-na-ku šīrka damāka anāku ‘I am your flesh (and) blood.’ (b/k 95: 6–7) MA ARAD ša EN-ia a-na-ku urdu ša bēlīya anāku ‘I am my lord’s servant.’ (KAV 159: o 6) a-bi-ia at-ta EN-⸢ia at-ta⸣ abī atta bēlīya atta ‘You are my father; you are my lord.’ (VAT 8851: 19‒20) NA6 SAG.DU ITI šu-ú kaqqad urḫe šū ‘It is the first of the month.’ (SAA 10, 43: r 2) lú ARAD ra-ʾi-mu ša ENmeš-šú a-na-ku urdu rāʾimu ša bēlīšu anāku ‘I am a servant who loves his lords.’ (SAA 16, 36: r 5) Furthermore, this is likely the inherited word order from Proto-Semitic, as it is not only the normal order in West Semitic, but also the order underlying the stative construction in Akkadian itself. The evidence is, therefore, indicative that the position of the pronoun in the right-most periphery is original, while the occurrence of the predicate in the same position is a result of language contact and, therefore, secondary (Edzard 2003: 174). Consequently, one cannot conclude from the final position of the pronoun that the pronoun is a copula in Akkadian. 5 6 In OA, the order Subject-Predicate is also fairly common. Kouwenberg speculates that it may be contrastive (2017: 714). In the following, all examples without language or dialect specification are Neo-Assyrian. Copyright Institut für Orientalistik Wien 2022 26 Ø. Bjøru & N. Pat-El The second argument – that the pronoun in tripartite sentences is not the subject – is more complicated. In tripartite sentences, namely where there are overt subjects and predicates in addition to the 3rd person pronoun, it may be tempting to analyze the pronoun as a copula,7 but tripartite sentences with an overt nominal subject, a predicate, and a pronoun could be interpreted as either containing copula or as topic-comment constructions; for example: an-nu-rig UNmeš an-na-ka šu-nu annurig nišū annāka šunu ‘Now the people are here / the people, they are here.’ (SAA 16, 49: r 3) DUMU mdEN-SUM-na EN ḫi-i-ṭu šu-ú mār Bel-iddina bēl ḫīṭu šū ‘The son of Bel-iddina is a criminal / the son of Bel-iddina, he is a criminal.’ (SAA 5, 210; o 15–16) lú The fact that the same sentence can be interpreted in two distinct ways is not accidental. Copulas developing from personal pronouns are attested in many languages. In most cases, the copula construction developed from a reanalysis of precisely this kind of topic-comment construction, namely a nominal sentence where a noun was topicalized through left dislocation. The original grammatical subject, a pronoun, was reanalyzed as a copula morpheme, a link between the topicalized subject and the predicate (Li & Thompson 1977). We can crudely represent the process in the chart below, where the final stage is a result of frequent use, which eventually erodes the semantic load of the topicalization. In the final stage, the formerly topicalized NP is integrated into the sentence as an argument and the pronoun is no longer anaphoric: Pred Subj NP Pron Stage 1 \ Top Pred Subj NP Proni / NPi Stage 2 \ Subj Pred Cop / NP NP Pron Given that the surface structure of stage 1 and 2 are identical, several criteria are needed to determine whether a language is at stage 1 or stage 2. The question we should ask is whether Akkadian has fully grammaticalized its copula or whether it still reflects a topic-comment construction. The linguistic literature on this issue offers a number of possible criteria,8 not all of which are useful in the case of Akkadian: 7 8 For example, Hämeen-Anttila comments that “the pronoun… could be analyzed as subject…, but it seems to me preferable to analyze it as copula which includes the subject as in verbal clauses” (2000: 107). E.g., Devitt 1994, Lohndal 2009. Copyright Institut für Orientalistik Wien 2022 The Missing Link 27 ● Intonation: the initial noun in a copula construction is not separated intonationally from the rest of the sentence. This is not the case in topic-comment construction, where the topic is more stressed than other constituents. ● lack of anaphoric reference: a copula will not necessarily agree with the subject. ● distribution: the copula should be obligatory, or near obligatory, under certain conditions. ● specificity: the subject in a copula construction, but not in a topic-comment construction, can be indefinite. Akkadian does not lend itself easily to such an examination for two reasons: (1) it is a dead language, and its writing system gives no indication of sentence stress; (2) it does not formally mark the category “definite”, which if it existed, would help distinguish topics and subjects. Even with these hurdles we believe that it is possible to evaluate whether the 3rd person pronoun in Akkadian is a subject or a copula. In the following discussion we examine several criteria in order to evaluate whether the pronoun indeed functions as a copula. While the arguments presented in previous discussions of this issue are either misleading or inconclusive, our discussion below provides a robust body of evidence that supports a topic-comment analysis for Neo-Assyrian. 1) Anaphoric reference: In Semitic languages which developed a pronominal copula (Hebrew, Arabic, and Aramaic) the copula has no anaphoric value and can mark predication in sentences with non-3rd person subjects; e.g. Egyptian Arabic inta huwwe il-mudarris ‘You are the teacher.’ (Choueiri 2016: 106) Biblical Hebrew ˀattā hû ha-ˀělōhîm ‘You are God.’ (2 Samuel 7:28) In Hebrew and Arabic, if the copula had still retained 3rd person anaphoricity, these sentences would be ungrammatical. Since the copula is a mere syntactic marker, it is possible to use it to mark the relationship between 1st and 2nd person pronouns and a predicate.9 The loss of the person feature would be a strong argument in favor of a copula analysis (Devitt 1994: 140). In Akkadian, however, when the subject in a nominal clause is 1st or 2nd person, 3rd person pronouns are disallowed, even when there is no agreement between the subject and the predicate; see, for example, the first example above where the 9 Furthermore, Eid (1991: 45–46) argues that all agreement features are suspended with the pronominal copula, and even sentences such as inti huwwe inti ‘you (fs) are you (fs)’ where the copula is a reflex of the 3ms pronoun are preferred. Copyright Institut für Orientalistik Wien 2022 28 Ø. Bjøru & N. Pat-El subject is singular, but the predicate is plural. In other words, there are no contexts in which the 3rd person pronouns have no 3rd person referent; therefore, when the subject is a 1st or 2nd person pronoun there will never be a “redundant” 3rd person pronoun in the clause. at-ta NUMUNmeš GIN ša md30-PABmeš-SU atta zarʾū kēnūte ša Sin-aḫḫê-eriba ‘You are the true seed of Sennacherib.’ (SAA 16, 96: r 1–2) lú ARAD ša AD-ia at-ta urdu ša abīya atta ‘You were the servant of my father.’ (SAA 10, 90: r 19’) lú ARADmeš ša LUGAL at-tu-nu urdāni ša šarre attunu ‘You are the servants of the king.’ (SAA 5, 210: r 6-7) lú un-za-ar-⸢ḫu⸣ lúARAD ša LUGAL EN-ia a-na-ku unzarḫu urdu ša šarre bēlīya anāku ‘I am a house-born slave, a servant of the king my lord.’ (SAA 5, 243: o 5–6) Since the 3rd person pronoun in Akkadian always occurs in appropriate contexts and carries the expected agreement, there is no reason to assume that its original function has changed. 2) Position of the pronoun: As was noted above, several scholars relied on the final position of the pronoun, the same position that predicates generally occupy in this language, to argue that the pronoun is a copula (e.g., Vinnichenko 2016). This is misleading because nominal sentences with pronominal subjects in all stages of Akkadian show a different order than other types of predications and are consistently Predicate-Subject. Additionally, although the pronoun typically appears in the far-right periphery, where we otherwise find predicates, it can also occupy other positions in the sentence, where we otherwise find subjects. For example: UD-mu ša ši-i-[a-re] ⸢šu⸣-ú ú-de-e-šú ūmu ša šiʾāre šū udēšu ‘The day of tomorrow is the only one.’ (SAA 10, 46: b.e. 14–r 1) ù šu-ú is-se-e-a ina lúrak-sumeš u šū issēya ina raksūte ‘He was with me in recruiting (i.e., as a recruit).’ (SAA 1, 205: o 13–14) šu-nu an-na-ak UNmeš am-ma-ka šunu annāk nišū ammāka ‘They are here, the people are there.’ (SAA 15, 121: r 14–15) Copyright Institut für Orientalistik Wien 2022 The Missing Link 29 In this position we can find other pronouns as well, and this is also a possible position for the pronoun in bipartite nominal sentences. For example: at-ta LUGAL be-lí atta šarru bēlī ‘You are the king, my lord.’ (SAA 16, 36: r 9) ù a-na-ku ⸢an⸣-[na-(kam)-ma] u anāku annākamma ‘I am there.’ (SAA 15, 1: r20’) A relatively flexible word order is a feature of subjects in NA, but not of predicates. A comparison with a sentence with left dislocation is instructive. In the example below, the subject, a demonstrative pronoun, stands between a preposed prepositional phrase and the predicate. ina UGU ta-mar-ti 30 u 20 [an-ni]-u pi-še-er-šú ina muḫḫi tāmarti Sîn u Šamaš annīu pišeršu ‘Concerning the observation of the moon and sun – this is its interpretation.’ (SAA 10, 94: o10–11) The position of the pronoun in nominal sentences is, therefore, consistent with the pronoun serving as a subject, and not being part of the predicate. 3) Allomorphy: In many languages, and certainly in Semitic languages, copulas exhibit allomorphy with zero. While the exact distribution is not always predictable, semantic and structural conditions play a role; for example, in Israeli Hebrew, when both nouns are definite the copula is near-obligatory.10 In Akkadian, however, there is no context in which either construction, zero copula or topiccomment with a pronoun, are required or disallowed; however, zero is by far more common. To illustrate the lack of distributional criteria, consider the following examples, where the predicate is a prepositional phrase. In the first example the order is Subject-Predicate, while in the second, it is Predicate-Subject in a topic-comment structure. šúm-mu ḫi-ṭa-a-a pa-an LUGAL EN-ia LUGAL EN li-du-kan-ni šummu ḫīṭāya pān šarre bēlīya šarru bēlī lidūkanni ‘If my error is before the king, the king should kill me.’ (SAA 19, 91: o 6–7) ki-la-li-šú-nu ina IGI-šú šu-nu kilallīšunu ina pānīšu šunu ‘Both of them are in his presence.’ (SAA 19, 195: e 3) 10 But see Shor (2020) who argues that the pronoun in Israeli Hebrew is not a copula but subject doubling. Copyright Institut für Orientalistik Wien 2022 30 Ø. Bjøru & N. Pat-El In Semitic languages which developed a copula (Hebrew, Arabic, and Aramaic11), the pronominal copula correlates with a suppletive verbal copula in non-present TAM categories: kāna in Arabic, and hyy/hwy in NWS; for example: Biblical Hebrew ˀēlle hēm mišpәḥōt hal-Lēwī ‘These are the families of the Levites.’ (Lev. 3:20) wә-ˀēlle hāyû bәnê ˀohŏlîbāmā ‘and these were the offsprings of Oholibama.’ (Gen 36:14) Akkadian, however, never developed a verbal copula. This does not negate the possibility that Akkadian has a restricted present-tense copula, but it certainly reduces our ability to securely identify one. 4) Distribution: in languages with pronominal copulas, the copula is obligatory or near-obligatory in certain cases. For example, in Hebrew when both the subject and the predicate in equative clauses are noun phrases, the copula is near-obligatory: Biblical Hebrew wә-Yôsēp hû haš-šālîṭ ˁal hā-ˀāreṣ ‘Joseph is the ruler over the land.’ (Gen. 42:6) Vinnichenko (2016) argues that in Akkadian the pronoun is obligatory with nominal predicates (but not with nominalization), but that does not hold with predicative prepositional predicates (p. 95). If indeed the pronoun were obligatory, this would be a strong indicator that the copula analysis is right. But the pronoun is, in fact, never obligatory in Akkadian, including in NA, where in most nominal sentences, the pronoun is missing. This is especially telling in cases where a misunderstanding would be possible, for instance when both subject and predicate are nominal or when the predicate is a prepositional phrase. The following is not an exhaustive list but is meant to show the variety of non-verbal predicates, all of which do not condition the use of an explicit pronoun. ● When the predicate is a substantive: LUGAL DINGIR-a-a ù LUGAL ⸢du-ma-qi⸣ šarru ilāya u šarru dumāqī ‘The king is my god and the king is my jewelry.’ (SAA 16, 126: r 17) mì-i-nu ḫi-iṭ-ṭu mīnu ḫiṭṭu ‘What’s wrong?’ (SAA 10, 48: r 5) [x]-lim-2-me LAL-e x-lim-2-me muṭê 11 Ethiopian languages use a verbal copula also in the present tense, see Crass et al. 2005. Copyright Institut für Orientalistik Wien 2022 The Missing Link 31 ‘[1]200 is the deficit.’ (SAA 1, 143: r 6) an-ni-i-u ṭè-en-šú-nu annīu ṭēnšunu ‘These are their news.’ (SAA 5, 2: r 8–9) ● When the predicate is an adjective: ⸢LUGAL⸣ be-lí liš-al-šú-nu ma-a ra-ma-ni-šú-nu IRImeš šarri bēlī lišʾalšunu mā ramanīšunu ālāni ‘My lord should ask: are the cities independent?’ (SAA 19, 195: b.e. 20–r 1) ma-ʾa-⸢da⸣ lúDUMU KÁ.DINGIRki lu-u lúDUMU BÁR.SIPAki ša ina ŠÀ-bi e-ti-qu[ni] maʾda mār Babile lū mār Barsip ša ina libbi ettiqūni ‘The Babylonians or Basippans who proceed there are many.’ (SAA 15, 223; o 11– 14) mu-ma-a-a i-ba-ši la ḫa-an-šú-ti Mumaya ibašši lā ḫanšūti ‘The Mumaya aren’t submissive.’ (SAA 5, 78: r 11–12) iri ● When the predicate is an adverb: ⸢ú-ma⸣-a már-[ba-a-a] ⸢an⸣-na-[ka] ūmâ Arbaya annāka ‘Now Arabaya is here.’ (SAA 10, 175; r 4–5) šu-nu an-na-ak UNmeš am-ma-ka šunu annāk nišū ammāka ‘They are here, the people are there.’ (SAA 15, 121: r 14–15) ma-a ḫa-⸢na-ka⸣ šu-u-tú ḫannāka šūtu ‘He is here.’ (SAA 1, 204: o 7) ● When the predicate is a PP: ḫa-nu-u-te lu ina pa-ni-ka ḫannūte lū ina pānīka ‘These should be under you.’ (SAA 5, 121: o 8) ṣa-ra-ḫu ša SAG.DU-su Áii-meš-šú GÌRii-meš-šú i-ṣar-ḫu-u-ni TA pa-an ZÚmeš-šú ṣarāhu ša kaqqassu ahīšu šēpēšu iṣarhūni issu pān šinnīšu ‘The feverishness by which his head, arms, (and) feet are burnt is connected to his teeth.’ (SAA 10, 302: o 11–r 2) Copyright Institut für Orientalistik Wien 2022 32 Ø. Bjøru & N. Pat-El di-bi ina KA-ia dibbī ina pîya ‘I have claims.’ (SAA 5, 241: o 7–8) The fact that there is no environment in which the pronoun is obligatory weakens the claim that it has grammaticalized to a copula. 6) Negation: In unmarked negated nominal sentences, the negation particle is typically positioned before the predicate, unless the subject is specifically negated. This is, of course, similar to the position of negation with verbal predicates. ṭè-mu an-ni-u la a-ke-e ša pi-i lúA.BA É.GAL šu-tu-u-ni ṭēmu annīu lā akê ša pî ṭupšar ēkalle šūtū-ni ‘This report is not how it is according to the palace scribe.’ (SAA 15, 32: r 5’–7’) When the predicate is in final position, the negation particle will precede it. As was noted above, in sentences with a pronoun, the predicate is typically pre-posed, and that is where the negation will be positioned. For example: la-a DUMU lúEN IRI ⸢ša iri[x x] a-na-ku lā marʾu bēl āle ša irix x anāku ‘I am not the son of the lord of the city [...].’ (SAA 5, 243: o 4–5) la-a ŠEŠ-a-a ⸢at⸣-ta-a lā aḫāya attā ‘Are you not my brother?’ (SAA 5, 81: r 3) la lúARAD-ka ⸢šu⸣-u lā urdaka šū ‘He is not your servant.’ (SAA 19, 91: r 11–12) The same is true for sentences with 3rd person pronouns in tripartite clauses. In the examples below, the core sentence includes a negated predicate and a subject pronoun. The position of the negation indicates that the pronoun at the far right is not a copula. The noun to the left of the core clause should be understood as a preposed subject functioning as a topic and not as the subject of the core clause. m da-da-a la-a EN de-ni-⸢ia šu-ú⸣ Dada lā bēl dēnīya šū ‘Dada is not my legal opponent.’ (SAA 5, 243: r 5–6) [KUR an]-ni-ú la-a ia-ú šu-ú mātu anniu lā yāʾu šū ‘This land is not mine.’ (SAA 1, 250: o’4) The position of the negation therefore supports the analysis of the pronoun as a subject, and not a copula. Copyright Institut für Orientalistik Wien 2022 The Missing Link 33 6) lū: The “precative” modal particle is in a complementary distribution with the functionally equivalent and etymologically related verbal prefix l-, e.g., liprus. There are a number of examples of 3rd person pronouns preceded by lū, which could be an indication that the pronoun is part of the predicative nucleus: ⸢ka-a⸣-du ina UGU-ḫi-ni lu šú-[nu] kādu ina muḫḫīni lū šunu ‘Let them be a stronghold for us.’ (SAA 15, 199: e. 2) But a thorough examination reveals that this is not the case. In the example below, the pronoun must be the subject, but it is still preceded by lū. Clearly, in nominal sentences lū may precede the predicate or the subject. UD 29-KÁM ina iri⸢ŠÀ IRI lu-u at-ta ūm 29 ina Libbi-āle lū atta ‘Be in the Inner City on the 29th!’ (SAA 1, 91: o 8–9) mi-il-[ki] lu-u šu-ú milkī lū šū ‘Let this be my advice.’ (SAA 10, 240: r 16–17) It is, therefore, more accurate to say that lū tends to be positioned before the final constituent in a sentence. Indeed, other examples support this conclusion. ḫa-nu-u-te lu ina pa-ni-ka ḫannūte lū ina pānīka ‘These should be at your disposal.’ (SAA 5, 121: o 8) The position of the precative particle before the pronoun does not provide support for the copula analysis. 7) Subordination: While nominal clauses can be subordinated, there are almost no examples of subordinate tripartite nominal clauses with a pronoun in NA.12 Under the assumption that the pronoun is a copula, this distribution would be difficult to explain. There is no objective reason to exclude one type of predication from this environment. However, under the assumption that tripartite sentences with a pronoun are types of topicalization, this exclusion is clear; cross-linguistically, subordinate clauses are far less likely to contain topicalization (Bybee 2002: 2). So far we have supplied arguments to support a topic-comment analysis of the personal pronoun in Akkadian. A possible argument for the existence of a copula would be a tripartite sentence where a left dislocation is an unlikely reading. Such is the case with conditional clauses, where a topic-comment within the scope of a 12 This was also mentioned by Hämeen-Anttila, who notes that in subordinate and interrogative clauses, the copula / pronoun is almost never used (2000: 107). Copyright Institut für Orientalistik Wien 2022 34 Ø. Bjøru & N. Pat-El šumma-clause would be highly unlikely. We have found only one such example in the Neo-Assyrian letter corpus: šum-mu mdPA-še-zib ina ŠÀ-bé šú-u ana šá-a-šú EN UNmeš-šú ŠUii MAN ŠÚ lu-šak-šid-su šummu Nabû-šēzib ina libbe šū ana šāšu adi nišēšu qāt šar kiššate lušakšissu ‘If Nabu-šezib is in there, I will have the hand of the king of all reach him and his people.’ (SAA 21, 155: r 3–5) However, the following conditional clause has the same construction, only negated. The position of the negation is indicative that the prepositional phrase is the predicate, not the pronoun. The negation therefore makes a copula reading less likely (though not impossible in this one case). u šúm-mu la ina ŠÀ šú-u UNme šá i-na É-a-mu-kan ina ŠUii MAN GAR-un u šummu lā ina libbe šū nišē ša ina Bit-Amukan ina qātē šarre laškun (SAA 21, 155: r 5–7) ‘But if he is not in there, I shall position the men in Bīt-Amukani under the authority of the king.’ We, therefore, suggest that Neo-Assyrian, like previous Akkadian dialects, did not have a functional pronominal copula. Given the evidence presented here, there are no compelling reasons to assume that the pronoun in Akkadian functions is a copula and therefore there is no objective reason to assume that the construction in NA has moved beyond a topicalization with an anaphoric pronoun, an inherited construction common in other dialects and stages of Akkadian. Bibliography Bybee, Joan L. 2002. “Main Clauses Are Innovative, Subordinate Clauses Are Conservative: Consequences for the Nature of Constructions.” In Joan L. Bybee & Michael Noonan (eds.), Complex Sentences in Grammar and Discourse: Essays in Honor of Sandra A. Thompson. John Banjamins, 1–18. Carver, Daniel E. 2016. “The Akkadian Stative: A Non-Finite Verb.” Ancient Near Eastern Studies 53: 1–24. Choueiri, Lina. 2016. “The Pronominal Copula in Arabic.” Brill's Journal of Afroasiatic Languages and Linguistics 8/1: 101–135. Cohen, Eran. 2000. “Akkadian -ma in Diachronic Perspective.” Zeitschrift für Assyriologie 90: 207–226. Crass, Joachim, Girma A. Demeke, Ronny Meyer, & Andreas Wetter. 2005. “Copula and Focus Constructions in Selected Ethiopian Languages.” Leipzig: Institut für Afrikanistik. de Ridder, Jacob Jan. 2018. Descriptive Grammar of Middle Assyrian. Harrassowitz. Devitt, Daniel C. 1994. Copula Construction in Cross-linguistic and Diachronic Perspective. PhD diss., State University of New York at Buffalo. Copyright Institut für Orientalistik Wien 2022 The Missing Link 35 Edzard, Dietz Otto. 2013. Sumerian Grammar. Brill. Eid, Mushira. 1983. “The Copula Function of Pronouns.” Lingua 59: 197–207. Hämeen-Anttilla, Jaakko. 2000. A Sketch of Neo-Assyrian Grammar. Neo-Assyrian Text Corpus Project. Hecker, Karl, Guido Kryszat & Lubor Matouš. 1998. Kappadokische Keilschrifttafeln aus den Sammlungen der Karlsuniversität Prag. Filozofická Fakulta Univerzity Karlovy. Holmstedt, Robert D. & Andrew R. Jones. 2014. “The Pronoun in Tripartite Verbless Clauses in Biblical Hebrew: Resumption for Left-Dislocation or Pronominal Copula?” Journal of Semitic Studies 59/1: 53–89. Huehnergard, John. 1986. “On Verbless Clauses in Akkadian.” Zeitschrift für Assyriologie 76: 218–249. Kienast, Burkhart & Konrad Volk. 1995. Die sumerischen und akkadischen Briefe des III. Jahrtausends aus der Zeit vor der III. Dynastie von Ur. FAOS 19. F. Steiner. Kouwenberg. 2000. “Nouns as Verbs: the Verbal Nature of the Akkadian Stative.” Orientalia 69: 21–71. Li, Charles N. & Sandra A. Thompson. 1977. “A Mechanism for the Development of Copula Morphemes.” In Charles N. Li (ed.), Mechanisms of Syntactic Change. University of Texas Press, 419–444. Lohndal, Terje. 2009. “The Copula Cycle.” In Ely van Gelderen (ed.), Cyclical Change. Philadelphia, 209–242. Luukko, Mikko. 2004. Grammatical Variation in Neo-Assyrian. The Neo-Assyrian Text Corpus Project. McWhorter, John. 1995. “Looking into the Void: Zero Copula in the Creole Mesolect.” American Speech 70/4: 339–360. Pustet, Regina. 2003. Copulas: Universals in the Categorization of the Lexicon. Oxford University Press. Reiner, Erica. 1966. A Linguistic Analysis of Akkadian. De Gruyter. Schroeder, Otto. 1920. Keilschrifttexte aus Assur verschiedenen Inhalts. Hinrichs. Shor, Leon. 2020. Reassessing the Third Person Pronominal “Copula” in Spoken Israeli Hebrew. Linguistics 58: 1807–1837. Stassen, Leon. 1994. “Typology Versus Mythology: The Case of the Zero-Copula.” Nordic Journal of Linguistic 17: 105–112. Vinnichenko, Olga. 2016. The Reassessment of the Influence of Aramaic on Assyrian Syntax. PhD diss., University of Cambridge. von Soden, Wolfram. 1995. Grundriss der akkadischen Grammatik. Pontificium Institutum Biblicum. Zielenbach, Maria. 2021. Grammaticization of Pronominal Copulas in Aramaic. MA thesis, Institut für Linguistik, Universität zu Köln. Copyright Institut für Orientalistik Wien 2022