Save

t-Forms of the Akkadian Stative

In: Brill's Journal of Afroasiatic Languages and Linguistics
Author:
Iris Kamil University of Edinburgh UK Edinburgh, Scotland

Search for other papers by Iris Kamil in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0000-0002-4441-7049
Open Access

Abstract

This article re-evaluates the status of the Stative paradigm in Akkadian, the language’s only exclusively suffixing paradigm, and argues in favour of a verbal classification on the basis of the possible medialisation of its attested forms. The Stative denotes the state persisting following a perfective-associated event. Thereby, two kinds of Statives are known: a verbal Stative and a nominal Stative. While the root of verbal Statives is mapped onto a Verbal Adjectival base template, the nominal Stative uses either a noun or an adjective’s stem as its base. To the respective bases both forms add the same distinct set of suffixes. Re-evaluating the morphological make-up of the form as well as its semantic connotation, this article proposes a new paradigmatic split between root-derived (verbal) and stem-derived (nominal) Stative forms and adduces previously disregarded evidence for medialisation of both verbal and nominal Statives as the main argument in favour of a verbal interpretation.

1 Introduction

The Akkadian Stative paradigm is the only purely suffixing verbal paradigm in the language, with all other verbal paradigms, namely the preterit, perfective, and durative/imperfective, being both prefixing and (partly) suffixing. Unlike the other conjugations, however, the verbal status of the Stative has been disputed (mostly due to arguments concerning its lack of a specified tense), and its precise effect on the root (or stem) it is applied to remains unclear, too.

Contributing to the debate, this paper has two goals: First, I will try to argue for a new paradigmatic classification of the attested Stative-associated forms. Secondly, and based on this classification, I will argue for the verbal interpretation of all Statives based on the evidence of t-infixation into nominal Statives. Thereby, the focus of this paper is the form’s morphosyntax and its possible effects on phonology and semantics. We will not try to definitively solve the semantic question of the form, but instead propose a new approach to it.

Section 2 will provide an overview over the Stative issue (Section 2.1) and the history of its research (Section 2.2), and will then proceed to discuss both the morphology of the form (Section 2.3) and its semantics (Section 2.4). In Section 3, I will provide a closer look at the t-morpheme (Section 3.1), and then turn to the evidence of t-Statives (Section 3.2) to argue for a verbal interpretation of all Statives. Finally, the implication of the evidence for the morphosyntactic derivation patterns of the Stative will be discussed in Section 4, before I conclude in Section 5.

2 The Akkadian Stative

2.1 Overview of the Stative issue

The debate originates in the morphological composition of the form. In essence, the Stative is made up of two components: a base and a suffix. The base of the Stative looks much like the base of a so-called Verbal Adjective (VA). A VA is a mostly predicatively used adjective that is derived deradically with the fixed template XaY(V)Z- (in the simple/Grundstamm template), the second vowel being the theme vowel of the respective root.1 The main difference between VA and ‘regular’ adjectives is that the roots of VA may be found in other verbal templates such as the preterit, imperfective and perfective. An example would be the root √dmq ‘be good’, which derives the following forms as seen in 1. Onto its base, namely XaYZ-, a VA would then typically go on to take a case suffix (and mimation/nunation2), whereas a Stative takes on the Stative suffixes.

(1) Two examples of the usage of the root √dmq ‘be good’

a. VA: damqu(m), damiqtu(m) ‘(he, she is) good’

b. imperfective: idammiq ‘he/it improves’

Proponents of the “Stative as a nominal”-hypothesis will typically centre their argument around its nominal-looking base, whereas researchers arguing for a verbal interpretation will rely on the Stative’s predicative behaviour and conjugatable paradigm (given above in Table 1).

Table 1

The Akkadian Stative in the ‘simple’ template

Singular

Plural

3. masc

paris

parsū

‘he is cut off’

‘they (m.) are cut off’

3. fem.

parsat

parsā

‘she is cut off’

‘they (f.) are cut off’

2. masc.

parsāta

parsātunu

‘you (m.) are cut off’

‘you (m. pl.) are cut off’

2. fem.

parsāti

parsātina

‘you (f.) are cut off’

‘you (f. pl.) are cut off’

1. com.

parsāku

parsānu

‘I am cut off’

‘we are cut off’

In regard to its suffixes, the paradigm is somewhat patchworky. In the traditional analysis, the suffixes of the second and first persons are derived from the independent personal pronouns (IPP; see Table 2), while the 3rd person plural suffixes likely derived from the verbal endings of the aspectual paradigms (Kouwenberg, 2000). As for the 3rd person singular forms, namely paris and parsat, no consensus exists as to their origin, but as far as the verbal/verbal adjectival 3.SG.M form can be addressed, it would not be unreasonable to presume that the surfacing vowel /i/ is an epenthetic vowel inserted into the nominal base mentioned above (pars- → paris).3

Table 2

IPP in Akkadian

Singular

Plural

3. masc.

šū

šunu

3. fem.

šī

šina

2. masc.

attā

attunu

2. fem.

attī

attina

1. com.

anāku

nīnu

The categorial and semantic nature of the Stative is perhaps even more puzzling than the form’s morpho-historical composition. A Stative can, in theory, be formed for any nominal or adjectival stem,4 and also for the majority of verbal roots. In the case of verbal Statives, the root is mapped onto the nominal base XaYZ-, associated with the VA. In the case of nominal Statives, the suffix is added to the stem. There are multiple nominal templates a Stative suffix can attach to. Verbal Statives, however, always retain the same base skeleton.

As for the forms’ distribution, Kraus (1984, 13 ff.) remarks that in the old Babylonian letters corpus (volumes 1 to 10), of about 1200 Stative forms only some 20 are not derived from verbs or VA and less than half of these 20 are derived from nouns. It would then seem that verbal and VA-Statives, in other words, the Statives associated with a fixed template, are by far the more common type found.

Nailing down the precise semantic effect a Stative has on a root is difficult. While in essence, it denotes the state following a perfective-associated event, this definition becomes problematic once the Stative attaches to a nominal element, not per se associated with an event. The following examples in 2 attempt to illustrate the range of meanings evoked by the Stative.5 In 2a we see both the usage of nominal and adjectival bases with a 1.SG suffix -āku. Each form appears ‘stripped’ of case vowels (u for nominative, a for accusative, and i for genitive), as well as possible mimation,6 e.g. šarr-u(m) ‘king’ → šarr-; ašarid-u(m) ‘first in rank, supreme’ → ašarid-. This behaviour indicates that nominal or adjectival bases do not necessarily conform to a templatic base, the way verbal or VA Statives do (see examples 2c–e). In 2b, we see an adjective combine with the 2.SG.M Stative suffix to ‘you are compassionate’. This example is important as it shows that the Stative suffixes can take on the e-colouring of the stem/root they modify.7 This will become relevant in Section 3.2 and 4 where we will see that the Stative must have access to the roots it modifies. So far so good, it seems that the Stative can turn nominal and adjectival elements to nominal “(x) is y” phrases.

(2) Akkadian Stative examples

a.

šarr-āku

bēl-āku

naʔd-āku

gašr-āku

kabt-āku

šurruḫ-āku

ašarid-āku

king-STAT.1.SG

lord-STAT.1.SG

praised/famous-STAT.1.SG

strong-STAT.1.SG

important-STAT.1.SG

glorious–STAT.1.SG

supreme-STAT.1.SG

‘I am king, I am lord, I am famous, I am strong, I am important, I am glorious, I am supreme’ (AKA 265 i 32)

b.

rēmēn-êta

compassionate-STAT.2.SG.M

‘you are compassionate/merciful’ (BMS 27:18)

c.

šumma

amēl-u

qaqqas-su

ašâ

aḫiz

if/when

man-NOM

head-3.SG.M.ACC

ašû-disease.ACC

seize.STAT.3.SG.M

‘When a man’s head is seized by the ašû-disease’ (AMT 6 9:10)

d.

paššūr-u

šamn-a

pašiš

table-NOM

oil-ACC

anoint.STAT.3.SG.M

‘the table has been anointed with oil’ (Arnaud Emar 6 20:19)

e.

ana

dāriš

balṭ-ānu

for

eternal

live.STAT-STAT.1.PL

‘we shall always live’ (ABL 886 rev. 4)

The situation becomes more complicated when we look at the effect of the Stative on verbal elements. In 2c we see an active use of the Stative, as it is sometimes called.8 In 2d the form is clearly passive. Temporally, while in 2c we see a sort of undefined, punctual reference to the time of event, in 2d we see a clear reference to a past one. In 2e a futurate connotation is conveyed.9 It would thus seem that the Stative is not specified for either voice or tense.

This leaves us with the following problem: In Akkadian, there appears to be a conjugable form that is applied to verbal and VA roots in the form of a template (resembling a VA stem), but is also found in a similar form for proper nouns and adjectives, onto the stem of which a suffix of the same conjugable paradigm is added. All forms appear to denote a type of stative situation or act, none are specified for a particular tense or voice. In effect, it is unclear whether the form is verbal or nominal, and whether at all we are dealing with one paradigm, as opposed to possibly two (one verbal/verbal adjectival and one nominal).

New perspectives on the debate may be found in Stative forms containing an infixed t-morpheme. Crosslinguistically, stative verbs are often ascribed perfective features and properties (see for instance Harley (1995)). Yet, in Akkadian, perfective aspect is predominantly expressed through the infixation of the t-morpheme between the first and second root radical in the ‘simple’ and ‘intensive’ templates and between the prefixed -š- or -n- of the ‘causative’ and ‘passive’ templates and the respective first root radical (see Table 3).

Table 3

Akkadian ‘simple’, ‘intensive’, ‘causative’, and ‘passive’ templates with infixed t

Simple

Intensive

Causative

Passive

preterite

iprus

uparris

ušapris

ipparis

perfective/medial

iptaras

uptarris

uštapris

ittapras

Stative

paris

purrus

šuprus

naprus

perfective/medial

pitrus

putarrus

šutaprus

?

The very same morpheme and the exact same strategy is used for medialisation in Akkadian, too (usually evoking medio-passive or reflexive/reciprocal meaning). These two forms are usually (synchronically) indistinguishable.10 Crucial to our debate is that the insertion of the medial/perfective t-morpheme, itself restricted entirely to verbs, is possible for both verbal and nominal Statives. This would imply that nominal Statives, too, are in fact verbal in nature, refuting the long-standing claim made by many Assyriologists that the Stative is a nominal phrase.

This paper thus has two goals in mind. First, I plan to adduce further input to the the debate on the verbality of the Stative, by presenting attestations for the possible but by no means productive infixation of the Semitic t-morpheme in Stative forms across multiple dialects and all periods of attestation. Secondly, after establishing the verbal nature of all Statives, we will try to account for both their differences and similarities in morphology and semantics by investigating the forms’ inherent morphosyntactic build and principles of derivation. The next section will give a broad overview of previous works to illustrate how the form came to be as disputed as it is.

2.2 History of research

The Stative has a long history of research in the Assyriological field,11 but it was first thoroughly discussed by Giorgio Buccellati (1968), who believed the Stative to be a nominal sentence with a predicate in the predicative state.12 Buccellati (1968, 10) argues that “the nominal predicate of the nominal sentence occurs regularly in the predicative state (= [S]tative), except when the predicate is immediately followed by a qualification or complement, or by the conjunction particle -ma, in which case the normal or construct states are used.” Buccellati considered the suffixes to be the grammatical subject of the phrases but because Statives were not (at the time?) attested to occur with ventives,13 he concluded that the clauses were verbless and nominal in origin and function. In light of Huehnergard’s (1987) study, Buccellati (1988) then considered the clause to be equal to the West-Semitic “verbless clause”, or in other words: a copular sentence.

Table 4

Kraus’ (1984) semantic split of the Stative forms, following Carver (2016)

Type

Meaning in Stative

I

Stative verbs

Denote a state, should thus be considered as nominals

II

Fientic verbs

Describe a given situation without reference to “Aktionsart

IIIa

Fientic verbs

Can express the situation as the result of an action

IIIb

Nominal Statives

Denote an attribute held by the subject

Within the scope of his study on the old Babylonian letters corpus (AbB) Kraus (1984) investigated the use of the Stative, too. A statistical evaluation yielded that from 1200 Statives, only 20 were formed of ‘true’ adjectives, participles, numerals, or nouns; in other words not from verbs or VA. Kraus thus understood the Stative to be a verbal clause with a finite verb as its predicate. He raised many important questions regarding the morphology, mophophonology, morphosyntax, and the diachrony of the form, but was unable to solve them at the time (Kraus, 1984, 20 f.).14 He did, however, attempt to categorise the forms semantically, stative verbs constituting the first class, fientic verbs constituting two different classes, depending on their evoked semantics, and finally nominal Statives constituting their own class. Table 4 above illustrates Kraus’ categorizations (here with summaries of Carver (2016, 4 f.)).

Huehnergard (1987), like Buccellati, interprets the Stative as a verbless clause that consists of a VA base and an enclitic subject-pronominal suffix, neither of which is predicative on its own, but only in their combination in the Stative construction. Though still simplified, Huehnergard provides a significant improvement to Kraus’ semantic distinctions.15 Given that the base of the Stative is a VA, he demonstrates the derivational behaviour of three verbal types that may be turned into VA.

Table 5

Semantic types of Akkadian VA

stative roots

active-intransitive roots

active-transitive roots

VA is descriptive

VA is resultative

VA is passive

idmiq ‘improved’ → damiq ‘is good’

ušib ‘sat down’ → wašib ‘is seated’

iṣbat ‘seize’ → ṣabit ‘is seized’

One distinction that Huehnergard attempts to introduce is that of the ‘regular’ Stative and the ‘active’, in other words, transitive Stative. According to him, some Stative forms of transitive verbal roots display two patterns in Stative formation: (I) a passive VA in the predicative construction, and (II) a non-fientic counterpart of transitive finite (i.e. verbal) forms. Consider the following examples for the root √mḫr ‘to face, confront, receive’ and √wšb ‘to sit, dwell.’ 3 shows Huehnergard’s predicted behaviour for the passive type (I) verbs, while 4 shows the predicted “innovative” behaviour of the fientic type (II) verbs (Huehnergard, 1987, 228).16

(3)

a.

ušib-0̷

vs.

3.wšb.PRET-SG.M

‘he sat down’

b.

wašib

wšb.STAT.3.SG.M

‘he is seated’

(4)

a.

i-mḫur-0̷

vs.

3-mḫr.PRET-SG.M

‘he received’

b.

maḫir

mḫr.STAT.3.SG.M

‘he is in receipt of’

Apparently, these active forms of the non-fientic type (II) occur only in a closed semantic group of verbs that are related to the concept of ‘grasping’ or ‘seizing’.17 Contrasted with forms such as wašib, as seen above, Huehnergard suggests a separate treatment of the forms both synchronically and diachronically despite their morphological equivalence, as these ‘active’ Statives clearly behave as verbs (taking on arguments, the enclitic conjunction -ma, the ventive, and the subjunctive), which does not align with his idea of the verbless predicate.

Kouwenberg (2000) returns to the idea of the Stative as a verbal form in his very long and comprehensive treatment of the form. Though he acknowledges that the forms were diachronically nominal, he argues that synchronically they were not. The pronominal suffixes marking the subject of the clauses would have made them functionally equivalent to other verbal affixes (referring to the prefixing conjugations), making them verbal forms, and their marking for gender, number, and person, would make the verbal forms finite.

He further argues for the verbal nature of the Stative by noting its clause-final position in the sentence, which would align with the SOV observed in Akkadian.18 Especially then with the existence of the ‘active’ stative, which takes arguments, he concludes that the Stative must be a finite verb.

In order to account for the nominal Statives, Kouwenberg (2000, p. 50 ff.) proposes a semantic differentiation between the ‘verbal’ Statives and the nominal predicates. While the verbal Stative denotes ascriptive states, nominal predicates denote equative states. Nouns in the Stative are then no longer nouns, but have so to say acquired the grammatical status of a verb.

Finally, addressing the tense-question, Kouwenberg (2000, p. 31) claims that the Stative is “tenseless and denote[s] a state rather than an event.”

The last treatment of the Akkadian Stative discussed is that of Carver (2016). Carver, too, believes the Stative to be a verb, albeit a non-finite one, following Dik’s (Dik, 1997, pp. 144–168) criteria for finiteness: a finite verb is (a) morphologically inflected for person and number, (b) the syntactic predicate of the clause, and (c) marked for tense, (viewpoint) aspect, or mood.19 Carver’s important semantic analysis of the Stative will be discussed in Section 2.4. He concludes that the Stative is not marked for T/A/M, given that although the Stative is marked for aspect, it is not marked for viewpoint aspect, and is thus a non-finite verb.

To summarise, previous accounts have mostly attempted to define the nature of the Stative as a verbal or nominal element, using rather vague semantic categorisations as their leading arguments. Undisputed is the predicative nature of the Stative, as is its basic morphological structure of VA + enclitic IPP for verbal and verbal adjectival Statives, and nominal/adjectival stem-base + enclitic IPP for nominal and adjectival Statives. The precise categorisation of the forms within the paradigm is still disputed, as is the nature of the forms themselves.

2.3 Morphology of the Stative

In this section, we will review the knowledge we have of the Stative’s morphology and begin motivating a different way of categorising the forms. As was summarised before, the morphology of the Stative construction, specifically of the verbal or VA Statives is generally agreed upon to be at least overtly the same:

(5) Formation of the Stative

XaYZ- + suffixed IPP (2nd and 1st person), or verbal ending (3th person)

The nominal and adjectival Statives on the other hand, retain their base form without any case marking or mimation. To the base form, the Stative suffixes are then added, as can be seen in 6b below. Previous studies such as Buccellati (1968), Lancellotti (1962), Diakonoff (1965), and Reiner (1966) have interpreted the verbal Statives to be nominal forms where the ‘base form’ is their predicative state.

(6)

a.

zikār-u(m)

man-NOM

b.

zikār-āku

man-STAT.1.SG

‘I am a man’

An in-depth discussion as to what the predicative state is and why it should be rejected goes beyond the scope of this paper.20 We shall renounce this interpretation here and suppose a skeletal base similar to the VA for VA and verbal Statives. Nominal and adjectival Statives are, as already claimed by Kouwenberg and Huehnergard, a secondary development, wherein Stative suffixes seemed to have become somewhat productive when added to a stem. Kraus’ statistical evaluation would appear to back this claim, too.

Given the different bases or stems to which the Stative suffixes may be added, I suggest the following paradigmatic split for the Akkadian Stative, as seen in Table 6. The examples given here mean to illustrate that while ‘primary’ Statives consist of an unchangeable base skeleton XaYZ-, a template, onto which the root consonants are mapped, ‘secondary’ Statives have their nominal stem as their base without conforming to a given template.

Table 6

Paradigmatic division of the Akkadian Stative

Primary Statives

Secondary Statives

Verb

Verbal adjective

Adjective

Noun

pars-

damq-

datn-

šarr-

‘cut off’

‘be(come) good’

‘strong’

‘king’

rēmēn-

zikar-

‘merciful’

‘man’

paḫallān-

bēl-

‘with large thighs’

‘lord’

etc.

Morphologically, we can then separate two general types of Statives based on their form: ‘primary’ Statives of the verbal type, and ‘secondary’ Statives of the nominal type. Within these two groups, we can again differentiate between two sub-groups based on the roots in question. Verbal roots derive Statives of the pars-type, while certain adjectival roots may derive ‘primary’ Statives of the damq-type. As for the ‘secondary’ Statives, we differentiate between nominal and adjectival roots deriving different kinds of nominal and adjectival stems. Regarding two different outputs for adjectival roots, I hope to show their difference in Section 4 once we address the implications for a morphosyntactic structure. For now, we hold onto the morphologically-based classification above.

In this current take on the Stative, it remains undecided whether or not forms such as šar ‘(is) king’, which have often been argued to be 3.SG.M Statives (here of the noun šarru(m) ‘king’), should be regarded as nominal Statives or not, though I can, at this time, not offer a better explanation for them other than the forms possibly being status absolutus-forms used in copular clauses.

On the notion of copulas, it is important that a clear distinction between them and the Statives is made. If we consider for instance the two sentences below in 7, both supposedly meaning ‘I am king’, it would stand to reason to expect at least a pragmatic difference between the two. I would argue that 7a refers to the state or perhaps status of the person being king without a specific reference to a time period in which this statement is true, while 7b is a regular copular sentence, referring to the person being king in a more specifically referenced point in time.21

(7) Two constructions using the 1.SG as a subject and šarr- ‘king’ as an ascriptive property

a.

šarr-āku

king-STAT.1.SG

b.

anāku

šarr-u

I.NOM

king-NOM

With the base of the Stative covered, we now turn to the suffixes. The Stative suffixes, especially in their paradigmatic appearance bear striking similarities to other verbal suffixing paradigms in Semitic, despite not being ‘perfect’ correspondences (Carver, 2016, 10). Compare, for instance, the Akkadian Stative paradigm with the Gəʕəz perfect and the Arabic perfective paradigm:

Table 7

Suffixal paradigms in Akkadian, Gəʕəz and Arabic

Akkadian Stative

Gəʕəz perfect

Arabic perfective

singular

3. masc.

paris

gabra

labisa

3. fem.

parsat

gabrat

labisat

2. masc.

parsāta

gabarka

labista

2. fem.

parsāti

gabarki

labisti

1. com.

parsāku

gabarku

labistu

plural

3. masc.

parsū

gabru

labisū

3. fem.

parsā

gabrā

labisna

2. masc.

parsātunu

gabarkəmu

labistum

2. fem.

parsātina

gabarkən

labistunna

1. com.

parsānu

gabarna

labis

Cohen (1984, 109–110)

The Akkadian pronominal system features a combination of both the pronominal -k- (as found in Ethiosemitic) and -t- (as found in Arabic) forms. In that regard, it really does diverge from other paradigms. But as far as the somewhat ‘questionable’ suffixes of the 3th persons are concerned, we can see a near perfect correspondence to the Gəʕəz and Arabic paradigms.22 While the Arabic 3.PL.F suffix would have to be be accounted for separately, the initially problematic 3.SG.M forms show the same behaviour, namely the insertion of an epenthetic vowel between or after the last two root consonants in their base form. A reduced or ‘less marked’ 3.SG.M form is by no means unexpected and can, in fact, be observed in the aspectual paradigms, as well (e.g. i-prus-∅). The ‘strategy’, in any case, is the same, and this would lead to the presumption that the Akkadian Stative could perhaps be an inherited paradigm. Most crucially, however, this crosslinguistic comparison shows us that the paradigm is ineherently verbal and that the suffixes serve as an inflectional device.

To further our verbal argument, we could also draw on the fact that the forms take on typically verbal markers such as the -ma-particle, functioning as a conjunction 8a, the ventive 8b, or the subjunctive suffix -u 8c, as well as direct and indirect objects. Likewise, once a form has taken on Stative suffixes, it cannot take any nominal suffixes, such as case markings, proving as further evidence. In other words: the suffixes of the Stative are both the primary and secondary Statives’ verbalising elements, verbalising even the nominal Statives, as shall be argued in Section 4. This should suffice to evidence the verbal nature of the Stative.

(8)

a.

erm-um

ša

ṭupp-i

ḫepi⸗ma

ṭuppa-ša

i-šrum-ū⸗ma

envelope-NOM

of

tablet-GEN

break.STAT.3.SG.MCONJ

tablet.CSTR-ACC.3.SG.F

3-peel(off).PRET-PL.MCONJ

“The envelope of the (plaintiff’s) tablet happened to be broken, so they (= the judiciary) broke open her (the sued party’s) tablet” (RA 9, 22:19–25)

b.

ištūma

kasp-um

ana

qātī-ka

maqt-an-ni

if indeed

silver-NOM

DAT

hand.CSTR-ACC.2.SG.M

NEG

fall.STAT.3.SG.M-VEN-ACC.1.SG

“Since the silver has not come into your hand” (CCT 4 30b:8–22)

c.

ina

balṭ-u

u

šalm-u

in/DAT

live.STAT.3.SG.M-SUBJ

and

be healthy/whole.STAT.3.SG.M-SUBJ

“When he is alive and sound” (RA 13, 131–132:7–8, Translation from Kouwenberg (2010a, 221))

2.4 Semantics of the Stative

Perhaps the greatest difficulty posed by the Stative is the understanding of its intended semantics. Especially when dealing with extinct languages, we are assigned the difficult task of translating the language material at hand into modern understandings. The translation difficulties that arise are worsened in Akkadian by the verbal system’s entirely aspectual nature, which is not easily compatible with the mostly tensed languages used for its interpretation. In this paper, I do not attempt to solve the semantics question, but instead briefly present previous accounts of it and conclude with a tentative generalisation of my own before we move on to the central claim in Section 3.

Two more recent accounts have tried to narrow down the semantics of the Stative. Loesov (2012) gives a very detailed description of the verbal Stative across different points of attestation. The work was conducted within his research on “Akkadian Sentences about the Present Time” and thus also investigates the possibilities of temporal references of such Statives. The second account is by Carver (2016) who also manages to give a very detailed, close look into the form’s implied semantics.

Starting off with Carver’s work, he follows Lyons (1977) and Comrie (1976, 49) in their definition of stative verbs: “The Stative, like the nominal predicate, is a predicate that denotes a static situation, which ‘is one that is conceived of as existing, rather than happening, as being homogenous, continuous and unchanging throughout its duration.’ ” Taking this definition and applying it to the Akkadian Stative, we get a rather unique output in meaning, especially given the Stative’s possibility to stativize fientive verbs with AGENT-subjects. To illustrate this, Carver (2016) presents a number of examples taken from Illingworth’s (1990) study of the syntax of old Babylonian letters.

(9) Stative uses in old Bablyonian letters

a.

awilt-um

ana

mutī-ša

2

mār-ī

ald-at

woman-NOM

DAT

husband.CSTR-3.SG.F.ACC

2

son-PL.GEN

bear.STAT-STAT.3.SG.F

‘The woman has borne two sons for her husband.’ (AbB 7, 106: 19–21)

b.

qāti

PN1

u

PN2

nasḫ-ā⸗ma

hand.PL.CSTR

PN1

and

PN2

tear/pull (out).STAT-STAT.3.PL.F⸗conjunction

‘The claims of PN1 and PN2 are withdrawn.’ (CT 4 31b: 6–8)

The crucial meaning relayed in 9a is that the woman, at the time of the letter’s composition, was (legally) in the state of having borne two children for her husband (thus fulfilling her wifely duties). The form does not mean to imply that the woman was currently in the state of (actively) giving birth to two sons. It fares similarly with 9b: the Stative does not denote an imperfective connotation of the claims “currently being withdrawn”. Rather, at the time of writing, both claims were in the state of already having been withdrawn, not to be put forth again.

In other words: the Stative would appear to denote a continuous state of resultative being or ‘having achieved’. I join von Soden (GAG § 77e) and Kouwenberg (2010a, 163 f.) here in assuming that in its core, the Stative denotes the state of a (resultative) condition, rather than the state of an ongoing action. In that respect, it reminds us of the perfect(ive) paradigms in other Semitic languages in which the completed action is more clearly set in the past from the speaker’s point of view. Diachronically, one may suppose that the Stative’s inner-Akkadian development saw a development along the lines of perfectivestate following perfective action.

Crucially, however, we now turn to Loesov’s (2012, 93 f.) study, in which he has found a significant amount of non-resultative, but dynamic Statives. This leads him to establish four general observations (adapted from Loesov (2012, 95)):

(10) a. The application of semantic constraints on the RESULTATIVE-feature, allows for the detection of a sizable number of verbal Stative forms with dynamic readings.

b. The morphological shapes of Statives of basic telic verbs display very uneven token frequencies, from very common to zero.

c. For telic and punctual verbs, both the availability of the Stative and whether it is resultative or not further depends on certain fine-tuned parameters of meaning of the given lexeme.

d. Lexicon-based observation starts from inflection-morphology evidence: the rarer a PARIS23 of a given frequent telic lexeme is, the likelier it will be used in nonresultative senses.

The examples Loesov lists as ‘non-resultative’ but instead ‘dynamic’ may be disputed, however. The main point of dissonance in the interpretations of the resultativeness of the forms may perhaps lie in the different perceptions of how the Stative references time.24 The following sentences are a selection of examples Loesov (2006, 147; 2012, 90, 116) provides for non-resultative Statives.

(11)

a.

šeʔ-am

a-maḫḫar⸗ma

kašd-ak-ki

barley-ACC

1.SG-mḫr.‘face/receive’.IMPFCONJ

kšd.‘reach’.STAT.3.SG.M-VEN-ACC.2.SG.F

‘I will receive the barley and (then) come to you’ (AbB 13 87:8’f)

b.

ana

U4

2.KAM

kašd-ak??-ka

in

day

2.COUNT.DET

kšd.‘reach’.STAT-STAT.1.SG??-ACC.2.SG.M

‘Within two days I will reach you’ (AbB 12 42:14)

c.

šumma

A׊A

šaddagdam

eriš⸗ma

nad-i

if

field.NOM

last year

NEG

ʔrš.‘sow/cultivate’.STAT.3.SG.MCONJ

fallow-GEN

‘If the field has not been tilled last year and is fallow’ (AbB 2 92:15 f.)

In 11a, I understand the use of the Stative to refer to a state at a future time point, again following an action: “First I receive the barley, then I will be coming to you, and then I will be in the state of having reached you (so we can, for instance, discuss further transactions).” Similarly, in 11b: “Following the process of reaching you for two days, after two days, I will be in the state of having reached you.” But also Statives in modal contexts can be used that way. I understand the use of the Stative in 11c the following way: “If the field is currently in a state resulting from not having been tilled last year and if it is (thus) fallow (then, for instance, it will be not cultivatable this season).” The Stative arguably being able to refer to these resultative states in future, past, and present, and seemingly also various modal contexts leads us to now raise the question of how the Stative references time. Particularly: how is it TAM-categorised?

As has been acknowledged by many Assyriologists, but only rarely written down, Akkadian is not a temporal language, but employs a rather complex tense-aspect system. As I will claim here, Akkadian is a fully aspectual language, meaning none of its paradigms (that includes all prefixing paradigms) are tensed. Given the lack on consensus on the language’s TAM-system, it is only natural then that the debate on the Stative’s categorisation has come up, too. While Illingworth has attributed perfective and imperfective aspect and past tense to the Stative, Kouwenberg (2010a, 163 f.) claims the Stative is marked for imperfective aspect, as “it does not envisage the beginning or the end of the state.” Kouwenberg (2010a, 164) also claims that the Stative can in principal refer to the past, the present, and rarely also the future. Loesov relates the ‘tense’-specifications to the individual roots that derive Statives (see for instance Loesov (2012, 116)). Finally, Carver (2016, 17) finds that the Stative “fits the non-progressive, continuous category of imperfective aspect which views the situation (i.e., state/condition) as ongoing without change.” Further, while the situations described by the Stative are states, sometimes defined in temporal scope, they cannot refer to any other kind of imperfective aspect such as habitual, progressive, or iterative situations (unlike the imperfective for instance).

As of yet, I believe that we need to establish a more precise theory of the aspectual system of the prefixing paradigms, before moving on to assessing the aspectual range of the Stative. Efforts of this nature are being undertaken by a number of Assyriologists and linguists at the moment, myself included, so that a more accurate TAM-analysis should hopefully be viable in the nearer future. For now, while my analysis (to follow in Section 4) will not yet attempt to make such precise predictions for the semantic/aspectual outcome of a (verbal) Stative, I propose the conflation of Carver’s and Loesov’s approaches: while I maintain that a Stative cannot be specified for tense, the state denoted by the form must be attributable to some phase on the timeline. This point or phase is determinable through the context in which the Stative appears and thus relies on the sentence proposition. That being said, whether or not the Stative can be associated with a previous, perfective event is, perhaps as Loesov suggests, dependent on the root the Stative is applied to.

To sum up, it would appear that the two points of discourse, namely the precise effect of the (primary) Stative on a root and the form’s TAM-categorization appear to intersect. This is also discussed in Loesov’s article, where he introduces the concept of ‘observability’ into the equation. He writes: “The availability and the productiveness of the resultative SC [Stative] P [passive]25 do not depend on the prototypical transitivity, but rather on the observability of the trivial resultant state of the former patient” (Loesov, 2012, 100).

A last comment is due on the semantics of ‘secondary’ Statives, which we have not yet addressed. The main difficulty that arises in defining the Stative’s effect as a ‘resultative’ state is that nominal elements are usually not associated with an event. As was mentioned in the discussion of example 7 above, however, there might be a level of applicability of the principle of observability as we apply our generalisations of the Stative onto ‘secondary’ Statives. Again, more precise predictions will hopefully be feasible once the TAM-system of Akkadian is better understood. For now, we shall leave the question of the nominal Statives unanswered.

Nonetheless, in Section 4 we will attempt to account for the more broad semantic differences that arise in the attested forms through the morphosyntactic differences of the different Stative constructions. For a more reasoned discussion and elaboration on the various readings of the Stative, the reader is encouraged to reference Loesov (2012), but also related work such as Loesov (2006). We now turn to the discussion of the t-morpheme in our goal of proving that all Statives, ‘primary’ and ‘secondary’, are verbal in nature.

3 t-Statives

3.1 The t-morpheme

A morpheme that is probably shared by all Semitic languages is the affixing t-morpheme used as a medialising/reflexive-reciprocal marker in verbal forms. It is typically inserted after the first overt consonant, meaning it is inserted between the first and second root radical in the ‘simple’ and ‘intensive’ templates and before the first root radical in templates characterised by a prefix, just as the ‘causative’ and ‘passive’.

(12) Examples of the t-infix in Semitic

a. Modern Hebrew

hi-t-pareq-0̷

CAUS-MID-prq.‘unload/dismantle’.PAST-3.SG.M

‘fell apart, broke’

b. Gəʕəz

ʔas-ta-rʔya

CAUS-MID-rʔy.‘see’.PAST.3.SG.M

‘showed oneself’

c. Akkadian

i-p⟨t⟩ašar-0̷

3-p⟨MID⟩šr.‘release/free’.PRET-SG.M

‘mutually disclose’

In Akkadian, unlike in the other Semitic languages, the morpheme can be inserted into any verbal form in any template (simple, intensive, causative, passive; see for instance 12c where the morpheme is infixed into a ‘simple’ form of pašāru ‘release, free’). In contrast, in Modern Hebrew, Arabic, and Gəʕəz, it is restricted to the ‘causative’ template (see 12a–b. Further, the t-morpheme developed a secondary function as a perfective marker (see Kouwenberg (2010b)). In the intensive and causative templates, it is also indistinguishable from the pluractional/diminutivising infix -ta(n)- (marked here as PLUR) when not in the imperfective.

Table 8

ta(n)-forms in Akkadian

Simple

Intesive

Causative

Passive

preterite t

iptaras

uptarris

uštapris

ittapras

preterite tan

iptarras

uptarris

uštapris

ittapras

imperfective t

iptarras

uptarras

uštapras

ittaparras

imperfective tan

iptanarras

uptarras

ušta(na)pras

itta(na)pras

(13) Akkadian -ta(n)-Forms in the simple, intensive, causative, and passive templates

a.

i-š⟨tan⟩appar-0̷

3-š⟨PLUR⟩pr.‘write/send’.IMPF-SG.M

‘he writes again and again (he ceases not to write (me))’ (ARM 2 137:42)

b.

u-l⟨ta⟩ḫiš-0̷⸗ma

3-l⟨PLUR⟩ḫš.‘whisper’.INTENSIVE.PRET/PERF-SG.MCONJ

‘he murmured’ (ARM 2 23 rev. 6’)

c.

šu-ta-ktumu

CAUS-PLUR-ktm.‘cover’.INF

‘to cover tightly’ (AfO 19 64 iii 5)

d.

i-t-tan-aqlap-ū

3-PASS-PLUR.qlp.‘peel’.IMPF-PL.M

‘they (the water) filtered (through the dam)’ (ARM 3 7:9)

While the ambiguity of these forms and their at times near impossible differentiation is a hotly debated phenomenon, treated in Streck (1995, 2003) and Kamil (to appear), it is unanimously agreed upon, perhaps not even contested, that this morpheme is verbal: it attaches only to verbal forms. As such, it serves as an ideal diagnostic to test the verbality of lexical items.

3.2 t-Stative forms

Given the t-morpheme’s distributional restriction solely to verbal forms, we now return to the Statives to test out the applicability of the morpheme in the forms. Indeed, we find a multitude of attestations for t-Statives:

(14) t-Stative Forms

a.

ūl

a⟨ta⟩mur-at

ṭupp-āt-im

NEG

ʔ⟨PLUR⟩mr.‘see’.STAT-STAT.3.SG.F

tablet-FEM.PL-GEN

‘she has not acquired much experience’ (ARM 1 85:7)

b.

mi⟨t⟩gur-at

amas-si

m⟨PLUR⟩gr.‘consent/agree’.STAT-STAT.3.SG.F

word-ACC.3.SG.F

‘her word is friendly’ (JAOS 88 127 ii b:13 & 20)

c.

ši⟨tak⟩kun

š⟨PLUR⟩kn.‘put’.STAT.3.SG.M

‘(it) was laid/ placed again and again’ (YOS 10 41:33)

Example 14a could be interpreted as a medial or a pluractional. The vocalisation of the form changes with the insertion of the ta-infix to Xit(a)YuZ-. The form is a little odd given the absence of the contraction of the aleph with the /i/ (phonologically, one may expect êtamurat, or even just ʔitamurat), but at least semantically, it seems to be reasonable: The person to which it refers was in the state of not having revised or seen too many tablets (thus not having, up until this point, acquired much experience). To the base form the expected 3.SG.F Stative suffix is then attached. In 14b, we see a typical t-Stative in its expected vocalisation. The form can be taken medio-passively, likely indicating that the person’s word is in the state of being agreeable, thus friendly. Finally, in 14c, we find a clearer pluractional in which the -tan-’s -n- assimilated to the following root consonant. Crucially, we see that even an event that appears to denote a state may be pluractionalised or broken up into multiple such stative events. In this case, to differentiate from a preterite, the form should likely be read along the lines of “having been laid / placed again and again.”

All these examples show t-Forms of ‘primary’ (meaning templatic or deradical) Statives. Hence, these examples in particular were less problematic in their verbal definition and we are not too surprised to see their successful medialisation. The insertion of the t-morpheme becomes more interesting when we consider ‘secondary’ (stem-derived) Statives. And indeed, we find the following Old Assyrian attestation in AKT 6 309:34:

(15)

anāku

u

attā

a⟨ta⟩ḫ-w-āni

I

and

you

a⟨MID⟩ḫ.‘brother’-PL-STAT.1.PL

‘You and I, we are brothers to one another.’

This example is remarkable in several regards. For one, we have a medialised nominal, or ‘secondary’ Stative: The state of an existing brotherly relation between two participants is turned into a reciprocal state. The t-insertion thus adduces another (equal?) argument into the event structure. Such a t-insertion, however, is only possible for verbal items, given that voice may only be applied to events, not to entities or adjectives.

We find further evidence for the morphosyntactic make-up of the form in its phonological form. In Akkadian, a masculine noun may be pluralised by suffixation of a long ū, e.g. aḫu ‘brother’, aḫū ‘brothers’. Now, crucially for us, Akkadian phonology dictates the contraction of almost any two neighbouring vowels; in the case of u/ū+a/ā, the expected outcome would be a long û.26 In the present form ataḫwāni, we would thus expect to find ataḫûni. It seemes, however, that the plural marker and the Stative suffix do not directly ‘interact’ that way, both morphemes remaining preserved: the plural marker is turned into a glide, and the Stative suffix is found in its expected ‘original’ form.

Stative suffixes may very well be altered, for instance by e-colouring of their underlying root, but phonological or “morphological” reasons are less likely to be responsible for the preservation of the full Stative suffix. A likelier reason for the retention might be the disambiguation from the ventive/dative -ni (which interestingly for the verbal definition of the form is an exclusively verbal suffix, too). A third explanation might be found morphophonologically, by supposing that the plural suffix and the Stative suffix are processed in different morphophonological cycles reflecting different syntactic domains. One possible approach to this could be found with the concept of syntactic phases Chomsky (2001), which we will turn to in the next section.

This ultimately suggests that nominal Statives, too, must be verbal and not just ‘predicative’ nouns or even copulas. In keeping with a uniform theory for the Akkadian Stative, wherein not every slightly diverging form must be explained separately, this should apply to non-medialised nominal Statives, too. While they are certainly nominal in origin, forms like 15 leave no doubt about their synchronically verbal usage.

4 Discussion

We have now reviewed both previous literature and introduced some new data into the equation. Having established that nominal Statives must be verbal in nature, too, we can now begin to to think about their theoretical implications. The most important findings are summarized in 16.

(16) a. The Akkadian Stative differentiates a ‘primary’ form and a ‘secondary’ form.

b. The characterising element of both the ‘primary’ and ‘secondary’ Stative is the same set of suffixes, attached either to the Stative template (for ‘primary’) or to the stem (for ‘secondary’).

c. Both Statives have access to information encoded in the root, such as potential e-colouring.

d. The reading of ‘primary’ Statives is dependent on the Aktionsart of the verb/root and on the observability of the event preceding the Stative. Generally, aspect and semantics appear to be correlated.

e. ‘Secondary’ Statives can be shown to be verbal, too, which is shown through their ability to take verbal affixes. New evidence for this claim comes from the possibility of their medialisation.

Given these claims, we would now like to see a potential explanation that could account for most of, if not all the multiple readings of the various Akkadian Stative constructions. The morphosyntactic theory employed here is that of Distributed Morphology (DM; Halle and Marantz (1993)).27 In essence, the characterising trait of the theory is that morphology, too, follows the principles of syntax. This allows us to closely follow the steps involved in derivational word-formation processes.

The first consideration that we have to make is how to represent the similarities and the differences of the forms. Let us begin with the similarities. I suggest that the common Stative traits, namely the suffixes, the verbal properties, and the influenceability of the forms’ readings by aspectual/higher syntactic structures, are structurally related in the following way: -features (person, number, grammatical gender) are contained within the Stative suffix, which is adduced into the derivation via v. That is where the verbalisation of the form takes place. This verbalisation-unit then Merges with what we will call BaseP (see below). With the verbal layer we Merge onto BaseP, we can then account for the possible t-infixation under VoiceP, the influence of aspect on the forms, but also for the possibility of later Merge-operations with morphemes such as the ventive, subjunctive, or the -ma-conjunction.

These structures, identical for both ‘primary’ and ‘secondary’ Statives, can be represented in the following way, as seen in 17.

(17)

The similarities covered, we then have to address how to represent the differences. Following our paradigmatic split in Table 6, the crucial differentiation can be traced back to the forms of ‘primary’, verbal Statives, and ‘secondary’, nominal Statives. In morphosyntactic terms, our main difference lies in BaseP. I suggest that ‘primary’ Statives are derived with BaseP1, which contains the Stative template in V, while ‘secondary’ Statives are base-derived with BaseP2, wherein the root does not Merge with a template, but forms an abstract stem with N/ADJ.

(18) a. BaseP1:

b. BaseP2:

By assuming the same mechanism for both verbal and VA Statives, namely the Merge-operation of a root with one specific template, we can explain the morphological similarity of both ‘primary’ Statives. Whether BaseP1 then derives a verbal or a VA Stative depends on the root. An adjectival root such as √dmq ‘to be(come) good’ naturally derives a VA-Stative. An action root such as √škn naturally derives a verbal Stative. Just as Loesov (2012) suggests with his analysis, different semantic root types will derive different meanings. A deradical approach can further explain possible idiosyncrasies that may arise with ‘primary’ Statives.

With BaseP2 and the absence of the Stative template in V, we can explain the greater variation in base-templates. Instead, N/ADJ adduce their respective templates, with which the root then forms a nominal stem. Allowing for the possibility of pluractionalisation as seen in 15, we also include a possible DP-layer. The full nominal stem then Merges with the Stative suffixes in the same manner as BaseP1.

These morphosyntactic differences can also account for some of the semantic differences that arise between ‘primary’ and ‘secondary’ Statives. It is here that we need to introduce the concept of ‘phases’ Chomsky (2001). A phase in syntactic theory refers to a type of barrier that occurs after certain ‘checkpoints’ in the derivation. The idea is that at these ‘checkpoints’, the derived material is sent to spell-out, both on the Logical Form (LF) level, as well as on the Phonetic Form (PF) level. As a result, any item that is added to the derivation after a phase, has no complete access to the information contained within it. Phases are crucial to us in so far as they occur at C, v, and D/n.

A crucial difference between BaseP1 and BaseP2 is that BaseP2 is likely contained within such a phase. If we look at the form in 15, namely ataḫwāni, we can see why the Stative suffixes, added after D/n, could not react phonologically with the plural marker -ū (contained within D). Semantically, we can thus motivate a higher regularity in ‘secondary’ Statives that are stem-derived vis-à-vis ‘primary’ Statives that are root-derived and thus more susceptible to idiosyncrasy.

Finally, if we assume phases for this derivation, we would also have to assume one following v, meaning after the introduction of the Stative suffixes. This should pose no problem for t-infixation, however, given that the morpheme appears to select its point of insertion depending on the phonological information it can access at that time. It appears that it is inserted in the first consonantal position following the first overt consonant it can detect. That would be following the first root consonant in the ‘simple’ and ‘intensive’ templates and following the respective prefixes in the ‘causative’ and ‘passive’ templates (see Section 3.1 above).

5 Conclusion

This paper set out to briefly review the status of the Akkadian Stative and provide further evidence for its verbal nature. To the previous arguments made by previous studies on the form, namely of some Statives taking arguments, and many taking verbal suffixes like the ventive or conjunctive, we now added evidence of t-insertion both in verbal and nominal forms. This shows that even nominal Statives must in fact be verbal in order for t-insertion to take place. They can thus no longer be used to argue for an interpretation of the Stative as a nominal clause. The issue of whether or not the form is finite or not, remains, at least for me, unsolved.

We have also tried to account for the form’s morphosyntactic build, differentiating between ‘primary’ root-derived and ‘secondary’ stem-derived Statives. The morphosyntactic differences can also help us better grasp semantic differences that arise.

Moreover, while its unique characteristics, most notably the Stativization of nominal elements, may initially lead one to believe that the Stative is an Akkadian invention, its morphological and semantic similarity to other perfect(ive) paradigms in Semitic, should inspire us to assume a Proto-Semitic origin of the paradigm, but a diverging inner-Akkadian development resulting in the puzzling form we find attested. In any case, the notion of a completed action relayed by the perfective aspect holds still for Stative, too, the major difference being the Stative denoting the state that holds after the completion of said action or condition.

The difference between nominal Statives and copular phrases remains somewhat obscure, but with the new definition given of the Stative, we can probably assume these copular constructions to be more similar to copular sentences in West-Semitic, while the Stative maintains the more specific meaning of a (resultative) state.

On a final note, a future study setting out to determine the precise semantic scope of the forms, verbal, VA, and nominal, must work both quantitatively and qualitatively with a well-defined body of texts. Indeed such studies are being undertaken at the moment by various researchers, strengthening the prospect of bettering our future understanding of the form. I hope to have shown that the Stative is perhaps not as obscure as it may appear at first glance, but also that a study of the Stative must be data-based and context-grounded.

Acknowledgments

I would like to thank Noam Faust, Itamar Kastner, Pavel Iosad, Michael Jursa, Laura Grestenberger, Maya Rinderer, Peter Hallman, Ur Shlonsky, Ruth Kramer, and the two reviewers for their helpful comments and support in developing the idea for this contribution. In particular, I would like to thank Matthias Adelhofer for showing me the Old Assyrian form, which supports my main argument and analysis of the Stative.

1

Akkadian roots can be classified into the following five vowel-Ablaut-patterns: (a/u), (a/a), (u/u), (i/i), and (a/i). The majority of adjectival roots are found to have the pattern (i/i), hence making the majority of VA also appear as XaY(i)Z-.

2

Mimation refers to a word-final m suffixed after the case vowels. It was strong in Old Babylonian and Akkadian but was gradually dropped in the second half of the millennium BCE. Whether or not it had meaning or function remains unclear.

3

Note for instance the use of /i/ as an epenthetic vowel in construct state formations in Akkadian, but also the cross-linguistic evidence of /i/ in epenthetic contexts in other Semitic languages, such as Arabic (see for instance Haddad (1983)).

4

This includes non-predicative adjectives. See for instance wēd-āku mamman ša ina rēš-ēa i-zzazz-u=ma … laššu alone-STAT.1.SG somebody REL in/by head-GEN.POSS.1.SG 3.SG.M-stand.IMPF-SUBJ=CONJ NEG.EXISTENTIAL ‘I am alone, there is no one to stand by me’ (BIN 6 104:15) <(w)ēdu ‘solitary, individual, single, only.’

5

Lemmas in capitals represent Sumerograms as found in the original texts. As for the case-glosses, Akkadian marks in principal three cases, the nominative with -u, the “genitive” with -i, and the “accusative” with -a. These older labels for the latter two cases mostly do correspond to the cases they are named after, but both overlap with the syntactic dative. The phenomenon was described, among other takes, as the “ablative accusative” (Jacobsen, 1963).

6

Note that mimation was dropped around the middle of the second millennium BCE both in Babylonian and Assyrian.

7

In Akkadian, a number of Proto-Semitic gutturals have conflated into one simple aleph. Nonetheless, reflexes of the original gutturals remain in the language, especially in their phonological effect on words. The following alephs are differentiated: *ʔ > ʔ1, *h > ʔ2, * > ʔ3, *ʕ > ʔ4, and *ġ > ʔ5 (Kouwenberg, 2010a, 516). In respect to the so-called e-colouring, ʔ3–5 differ from the other two in that they tend to colour all templatic /a/ and some /i/ vowels to /e/. That being said, not all roots that trigger e-colouring seem to contain a guttural (at least overtly), just as could be seen with rēmēnêta above.

8

The “active” Stative refers to a Stative that can take on objects. Huehnergard refers to it also as the parsāku-type. As can be seen from the glosses and their translation, these Statives are by no means “active” in voice, which is why we shall not adopt this term here.

9

Small capitals in the transcription refer to Sumerian logograms, and have not been transliterated into Akkadian here.

10

It is at this point perhaps interesting to note that the perfective use of the cross-Semitically medial -t- is a secondary development in Akkadian.

11

A more comprehensive summary of the history of research conducted on the Akkadian Stative may be found in Carver (2016, 3–8).

12

Akkadian nominals know three so-called states: the status rectus, the status constructus used for genitive constructions, and the very rarely used status absolutus. Some grammarians of Akkadian such as Ungnad, Buccellati, and Huehnergard assume another fourth state called the “predicative state”. While the rectus and absolutus are considered ‘free forms’, the construct state is considered a bound form, meaning it relies on the presence of another item, and the predicative state is considered to be a mixture of both.

13

A ventive is a directional morpheme found in various Semitic languages. In Akkadian it is most often attached following the suffixes of a verb involving motion, but later spreads to other verbs not denoting any sort of movement, too.

14

Kraus does, however, pose a secondary development for the nominal Statives given their scarce attestation in AbB. This view was later adopted and further developed by Huehnergard and Kouwenberg.

15

Unfortunately, Huehnergard does not comment on possible secondary derivations, such as deadjectival verbs, just as idmiq would appear to be.

16

Note that the difference in annotation between (3-a) and (4-a) regarding the 3th person prefix derives from the morpheme in (3-a) combining with the primae w to u. It is thus not entirely separable from the root.

17

Huehnergard also provides an example of the root ṣbt. The Stative form ṣabit would have, according to him, both the meaning of ‘he seized’ and ‘he is in possession of’. We shall turn the issue of semantics in Section 2.4. For now, according to the present interpretation, the form is to be read in only one way, wherein ‘he seized’ and ‘he is in possession of’ refer to the same result of the same event. Since the form refers to the state following a perfective-associated event, it would implicitly reference both the action of having seized as well as the possession of the item seized.

18

This is also true for pronouns when they are the subject of a verbless clause: ul bēlt-i atti NEG lady-1.sg.poss you. 2.f.sg ‘you are not my mistress’ (CT 8 22b: 8) (Carver, 2016, p. 7).

19

For instance, in divination, we often find sentences such as “If a man is asleep (STATIVE): his wish will not be achieved (IMPERFECTIVE)” (Rachel Lerculeur, pc).

20

To date, the assumption of a predicative state has been either accepted or ignored, but never actually debated. Likewise, this state has not truly ever been properly argued for.

21

As we will see in the next Subsection 2.4, Loesov (2012) argues for a factor of ‘observability’, which I am at this point inclined to consider as a relevant mechanism in determining the more precise difference between the two examples in 7.

22

See also Kuryłowicz (1973) who first suggested a verbal origin for the Suffixes.

23

It is unclear whether Loesov refers to the general form of the Stative, or the VA-base, or perhaps specifically to the 3.SG.M-form of the Stative. The terms PARIS and SC (suffix conjugation) are used interchangeably and usually all stand for the (verbal) Stative construction.

24

Loesov’s claims as to why certain forms are resultative and others are not is at present not fully clear to me.

25

This distinction between the active and the passive Stative as it was introduced by Huehnergard is adapted here, an active Stative again indicating a Stative that takes objects.

26

Note that here a circumflex indicates a contracted long vowel. The orthography is used to differentiate contractions from ‘inherently’ long vowels.

27

To accomodate the readers not familiar with the theory, I will provide a very general walkthrough of the thought-processes underlying the theory for the Stative’s structure that is proposed here.

References

  • Buccellati, G. (1968). An Interpretation of the Akkadian Sentence as a Nominal Sentence. Journal of Near Eastern Studies, 27:112. University of Chicago Press.

    • Search Google Scholar
    • Export Citation
  • Buccellati, G. (1988). The State of the ‘Stative’. In FUCUS: A Semitic/Afrasian Gathering in Rememberance of Albert Ehrman. Current Issues in Linguistic Theory, 58:153189. John Benjamins Publishing Company.

    • Search Google Scholar
    • Export Citation
  • Carver, D.E. (2016). The Akkadian Stative: A Non-Finite Verb. Ancient Near Eastern Studies, 53:124. Peeters Online Journals.

  • Chomsky, N. (2001). Derivation by Phase. In Kenstowicz, M., editor, Ken Hale: A Life in Language, 150. MIT Press.

  • Cohen, D. (1984). La phrase nominale et l’évolution du système verbal en sémitique: études de syntaxe historique, Collection Linguistique de la Société de Linguistique de Paris (73). Peeters Publishers.

    • Search Google Scholar
    • Export Citation
  • Comrie, B. (1976). Aspect: An Introduction to the Study of Verbal Aspect and Related Problems, volume 2. Cambridge University Press.

  • Diakonoff, I. (1965). Semito-Hamitic Languages: An Essay in Classification (Languages of Asia and Africa). In Moscow 1984, “Letter to the Conference …”, in J. Bynon (éd.), op. cit, 110. Nauka.

    • Search Google Scholar
    • Export Citation
  • Dik, S.C. (1997). The Theory of Functional Grammar: The Structure of the Clause. Walter de Gruyter.

  • Haddad, G.F. (1983). Epenthesis and Sonority in Lebanese Arabic. Studies in the Linguistic Sciences, 14(1):5788.

  • Halle, M. and Marantz, A. (1993). Some Key Features of Distributed Morphology. MIT Working Papers in Linguistics, 21(275):88.

  • Harley, H. (1995). Subjects, Events, and Licensing. PhD thesis. Massachusetts Institute of Technology.

  • Huehnergard, J. (1987). ‘Stative, Predicative’, Pseudo-Verb. Journal of Near Eastern Studies, 46:215232. University of Chicago Press.

    • Search Google Scholar
    • Export Citation
  • Illingworth, N.J.J. (1990). Studies in the Syntax of Old Babylonian Letters. PhD thesis, University of Birmingham, UK.

  • Jacobsen, T. (1963). The Akkadian Ablative Accusative. Journal of Near Eastern Studies, 22(1):1829. University of Chicago Press.

  • Kamil, I. (To appear). New Roads to Proto-Semitic Reconstruction. Exemplified by the Old Babylonian t-Forms in the Archives Royales de Mari. Master’s thesis, University of Vienna.

    • Search Google Scholar
    • Export Citation
  • Kouwenberg, N.J.C. (2000). Nouns as Verbs: the Verbal Nature of the Akkadian Stative. Orientalia, 69:2171. GBP Press.

  • Kouwenberg, N.J.C. (2010a). The Akkadian Verb and its Semitic Background. Languages of the Ancient Near East 2. Eisenbrauns.

  • Kouwenberg, N.J.C. (2010b). The Recycling of the t-Infix in Pre-historic Akkadian: A Case of Exaptation. In Proceedings of the 53e Rencontre Assyriologique Internationale (1): Language in the Ancient Near East (2 parts), 617646. Penn State Press.

    • Search Google Scholar
    • Export Citation
  • Kraus, F.R. (1984). Nominalsätze in altbabylonischen Briefen und der Stativ.

  • Kuryłowicz, J. (1973). Studies in Semitic Grammar and Metrics, Prace Językoznawcze (67). Curzon Press.

  • Lancellotti, A. (1962). Grammatica della lingua accadica. Analecta hierosolymitana. Franciscan Press.

  • Loesov, S. (2006). Akkadian Sentences About the Present Time (ii/1). Babel und Bibel 3: Annual of Ancient Near Eastern, Old Testament and Semitic Studies, 3:133. Eisenbrauns.

  • Loesov, S. (2012). The Suffixing Conjugation of Akkadian: In Search of Its Meaning. Babel und Bibel Annual of Ancient Near Eastern, Old Testament, and Semitic Studies, 6:75147. Eisenbrauns.

  • Lyons, J. (1977). Semantics: Volume 2. Cambridge University Press.

  • Reiner, E. (1966). A Linguistic Analysis of Akkadian. De Gruyter Mouton.

  • Streck, M.P. (1995). Zahl und Zeit. Grammatik der Numeralia und des Verbalsystems im Spätbabylonischen. Cuneiform Monographs. Brill.

  • Streck, M.P. (2003). Die akkadischen Verbalstämme mit ta-Infix, Alter Orient und Altes Testament (303). Ugarit-Verlag.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 299 254 9
PDF Views & Downloads 480 405 18