Verbal Forms Expressing Discourse Dependency in North-Eastern Neo-Aramaic

Geoffrey Khan

1. Introduction

In this paper I shall draw attention to the use of various verbal forms in North-Eastern Neo-Aramaic (NENA) dialects to express discourse dependency. By this I mean that certain verbal forms in certain contexts signal that the predicate of the clause they occur in continues in some way the preceding discourse. This continuation is typically either temporal sequence or some kind of elaboration. I shall propose explanations as to how the function of the expression of discourse dependency developed historically in the various verb forms in question. It will be shown that although the forms are formally different, they exhibit parallels in the historical processes of their semantic change. The data are based mainly on my studies of the C. Barwar and C. Urmi dialects, with occasional references to other dialects.1

2. The bət-qaṭəl Form

Dialects in the northern half of the NENA dialect area and in the Mosul plain have a future construction that is derived historically from the deontic verb ‘to want’ and a subordinate complement:2

The deontic verb has undergone morphological reduction and bonding through grammaticalisation. In the dialect of C. Barwar, for example, the basic form of the construction in slow careful speech is as follows:

The verb has been phonetically contracted and the subordinating complementiser has been affixed to the deontic form and devoiced. In some dialects there is no devoicing, e.g. C. Qaraqosh bəd-qaṭəl.

The process of grammaticalisation has reduced person distinctions in the deontic verb and the particle bət is used before verbs of all persons:3

In normal fast speech, moreover, the particle undergoes further phonetic reduction, resulting in the following allomorphs:

The construction can be shifted into the past to express a future in the past by attaching the past-shifting affix -wa:

2.1. Functions of the bət-qaṭəl Form

In the C. Barwar dialect the following functions of the bət-qaṭəl form can be identified (Khan 2008, 598–608). These functions are the typical functions of the future construction also in other NENA dialects. They can be classified broadly into functions that involve the expression of future tense (§2.1.1.–§2.1.3.) and those that involve the expression of discourse dependency (§2.1.4.). As will be argued below, the discourse dependency function, which is the main focus of this section, has developed from the future function.

2.1.1. Deontic Future

This function retains the deontic meaning of the source construction.4 In such cases, it conveys an element of will and expresses various degrees of intention, obligation, request and permission regarding a future action.

When the verb has an agentive 1st singular subject the bət-qaṭəl form generally has a sense expressing deontic intention, e.g.

When the verb has an agentive 1st plural subject, the form often has a cohortative sense (‘Let’s …’), e.g.

The form may express deontic obligation. In such cases the verb generally has an agentive 2nd person subject, e.g.

Occasionally the form expresses deontic obligation also in other persons, e.g.

2.1.2. Predictive Future

In many cases the construction does not have clear deontic force, but rather expresses a prediction of an eventuality that will happen in the future. This can be regarded as resulting from the grammaticalisation of the deontic construction, parallels to which are found in many languages. The core of this process involves a semantic extension whereby an implicature of the original deontic construction, in particular one with a 3rd person subject, is incorporated into the meaning, e.g. he wants to go to town implies that it is likely that he will go (Bybee 2010, 55).

The predictive future function of the bət-qaṭəl construction is generally found where the subject of the verb is 3rd person or where it is a non-agentive 1st or 2nd person, e.g.

2.1.3. Conditional Constructions

The bət-qaṭəl form is used in the apodosis of conditional sentences. In the majority of cases it expresses an eventuality that is temporally sequential to a hypothetical situation in the future, e.g.

In some cases it expresses a future eventuality that follows logically from a given, real situation in the present denoted by the protasis clause, e.g.

One of the most common uses of the ‘future in the past’ form bət-qaṭəlwa is in the apodosis of conditional sentences. Such sentences may refer to a hypothetical condition in the past that was not fulfilled, e.g.

In some cases the construction may denote a hypothetical condition in the present or future that the speaker assesses to be impossible to fulfil, e.g.

2.1.4. Discourse Dependency

In conditional constructions such as those described in §2.1.3., the apodoses with the bət-qaṭəl and bət-qaṭəlwa forms are dependent syntactically on the preceding protasis. The forms are sometimes used outside of conditional constructions in clauses that are more loosely dependent on the preceding discourse. Various types of discourse dependency are attested. In some cases the forms express events that are temporally sequential to what precedes:

In some cases the bət-qaṭəl and bət-qaṭəl forms do not express temporal sequentiality but only some kind of relevance to a preceding clause, typically elaboration:

In (20) the clause introduced by the bət-qaṭəl form ṱ-ámri constitutes an elaboration of the preceding statement that people would hold a fast, which could be paraphrased ‘with regard to this fasting they say ….’.

In (21) the clause containing the first bət-qaṭəlwa verb, p̂t-̭azíwa, opens a section of discourse that elaborates on the preceding general statement that the festival of Girls’ Sunday took place. The dependency expressed by the verbal forms bind them semantically to what precedes signalling that the descriptions of the specific events in the clauses are intended to be understood as components of the festival.

When the bət-qaṭəl and bət-qaṭəlwa forms have this discourse dependency function, they generally express habitual events, as is the case in the examples above. The construction is sporadically used in narratives where they refer to specific events that are dependent on, and typically sequential to, what precedes, e.g.

2.2. Analysis

2.2.1. From Apodosis to Discourse Dependent

When used in the first three functions described above, viz. deontic future, predictive future and apodosis of conditionals (§2.1.1–2.1.3.), the bət-qaṭəl(wa) form expresses future tense. There is a crucial difference, however, between the deontic future and predictive future, on the one hand, and conditional constructions, on the other, with regard to the reference point of the future tense. Following the temporal analysis proposed by Reichenbach (1947), we should be careful to distinguish event time (E), speech time (S) and the temporal reference time (R). The original system of Reichenbach has undergone various modifications in more recent research, but the ‘neo-Reichenbachian’ approaches still distinguish these three components of analysis. The reference time (R), sometimes referred to as the ‘evaluation time’ (Hatav 2012), is the contextual temporal anchor to which the future verb form relates. One may say that the future form is temporally ‘bound’ to this anchor (Hatav 2012). In the case of the deontic future and predictive future functions, the reference time overlaps with speech time, i.e. the contextual temporal anchor is the speech situation. For the bət-qaṭəl(wa) form in the apodosis of conditional constructions, however, the reference time is that of the eventuality expressed in the protasis clause. In such cases the bət-qaṭəl(wa) form expresses an eventuality that is posterior to this reference time but this reference time does not necessarily overlap with speech time.

According to the original Reichenbachian system of representation, the analysis of these functions of the bət-qaṭəl(wa) form would be as follows (where a comma indicates temporal overlap and a dash — indicates temporal separation):

According to some Neo-Reichenbach approaches (e.g. Johnson 1981; Dinsmore 1982; Verkuyl 2012), rather than consisting of a single triple system, the analysis should consist of two pairs of components, namely S and R, on the one hand, and E and R, on the one hand. The relationship between S and R would correspond to tense, whereas the relationship between E and R would be one of posteriority or anteriority independent of tense. The analyses, therefore, would be:

Table 1: Analysis of bət-qaṭəl and bət-qaṭəl-wa

As can be seen, according to this temporal analysis all of these three future constructions share the common feature of R—E, i.e. the event time is posterior to the reference time.6 What this Reichenbachian temporal analysis does not show, however, is that the reference time in the three constructions has different locations. In the deontic and predictive future constructions the reference time is internal, i.e. it coincides with the utterance of the clause. The reference time of the verb of the apodosis, however, is external to the clause and is located in the preceding protasis clause. This distinction is referred to by Hatav (2012) as local versus long distance semantic binding of tenses.

As for the aspect of the bət-qaṭəl(wa) form, in the examples cited above for its functions of deontic future, predictive future and apodosis of a condition the verb denotes a specific temporally bounded event and so is perfective. The form in these constructions may also denote iterative events (Khan 2008, 599, 606), e.g.

It is important to distinguish iterativity from habituality (Dahl 1985, 97; Bertinetto and Lenci 2012). Verbs expressing iterativity assert the occurrence of the event on multiple occasions, typically specified by an adverbial (‘He visited us three times’, ‘He visited us every day’). Such predicates are perfective and express repeated temporally bounded events, i.e. events that are viewed as a whole typically from a reference time that is external to it (G. Carlson 2012, 835). Verbs expressing habituality present an event as a characterizing property of an individual, which occurs on the majority of occasions during a particular time interval (He usually visits us every week). Unlike iterative predications, habitual predications are not completely ‘lawlike’ (Dahl 1985, 97) and are contingent on circumstances (He usually visits us every week, but he did not come last week because he was ill). A habitual predicate is imperfective in aspect since it includes the reference time within it and is viewed from within (G. Carlson 2012, 835).

The bət-qaṭəl and bət-qaṭəlwa forms in deontic future, predictive future and apodosis constructions may express iterative predicates but not habitual predicates.

When expressing the deontic future, predictive future or the apodosis of a conditional the bət-qaṭəl and bət-qaṭəlwa forms are modal. Their modality is either root modality or epistemic modality. Root modality qualifies the subject of the clause indicating that some factor is inherent in the subject (e.g. ability, volition) or is operative upon the subject (e.g. obligation, circumstances) that influences the occurrence of the event. Epistemic modality involves the speaker’s assessment of the truth value of the propositional content of the sentence as possible, probable or certain. The various types of modality inherent in the three aforementioned functions can be identified follows:

Table 2: Root and Epistemic Modality

Turning now to the discourse dependency function of bət-qaṭəl(wa), this has a close family relationship to the apodosis function in conditional constructions. Indeed, I shall argue that it developed historically by a process of extension of conditional constructions. The bət-qaṭəl(wa) form in discourse dependency constructions exhibits long distance semantic binding, as is the case with the verbs in apodoses. The reference point precedes the event in the discourse, but, unlike in apodoses, this is not a temporal relationship. Rather the bət-qaṭəl(wa) verb is bound to a topical reference point that has been invoked by the preceding discourse. It depends on this and continues it in some way.

This phenomenon can be represented in the dependency framework of mental spaces proposed by Fauconnier (1994) and Dinsmore (1991). According to this model, knowledge can be represented in a network of mental spaces. These spaces are constructed by the listener, interpreting grammatical or lexical cues. Spaces contain information belonging to distinct times, locations or realities. ‘Space builders’ are cues that construct new mental spaces. Dependent verbs such as the bət-qaṭəl(wa) form express events that belong to a current, already constructed mental space.

The differences from the conditional construction, therefore, involve (i) the change of the temporal reference point to a topical reference point and (ii) the dependency on preceding discourse rather than on a preceding syntactically subordinate clause. This can be explained using a model of linguistic change through schematisation of constructions (e.g. Fillmore, Kay and O’Connor 1988; Goldberg 1995; 2006; Langacker 1987; Bybee 2010). This involves extensions of components of constructions by a process of substitutions of items with a family resemblance, thereby making the slots of the components more schematic, i.e. abstract. Another feature of the extension of constructions is their incorporation of pragmatic associations and implicatures into their meaning (Bybee 2010, 48). An example of this process of extension of constructions that is often cited (e.g. Bybee 2015, 124; 2010, 55) is the development of future constructions consisting of movement verbs, e.g. English he is going to eat. This originated as a construction that expressed real physical movement of an animate agentive subject, but it became schematised as subject + be + going to + verb, whereby any subject or verb could fill the subject or infinitive slots. Moreover, when used in the third person, although it originally expressed an intention, it implied that the predicate would be carried out. This implicature became conventionalised in the construction and so its meaning was extended to include prediction, e.g. The branch is going to fall.

The temporal reference point of the bət-qaṭəl(wa) apodosis that was in the preceding clause was schematised to being a more abstract cognitive reference point, referring to the general situation rather than specifically to a point in time. The bət-qaṭəl(wa) predication, therefore, is cognitively but not temporally bound to this preceding point. It is linked to it through discourse coherence analogous to a topic—comment relationship, or, to use Fauconnier’s model, it is a continuation of the mental space of the preceding discourse. This does not mean that the topical situation itself may not have a reference time, but rather the bət-qaṭəl(wa) form no longer expresses temporal posteriority to this reference time but rather communicative posteriority to the more abstract topical situation, i.e. topical reference point—comment. One of the consequences of this is that the bət-qaṭəl(wa) may express discourse dependency on a non-propositional topic constituent without a temporal reference time, as is the case in (28) below.

The process of extending the location of the reference point of the bət-qaṭəl(wa) form from specifically the subordinate protasis clause to a broader component of preceding discourse that establishes a topic can also be identified as schematisation. This may have been facilitated by the fact that protasis clauses can in some contexts be used pragmatically as strategies for introducing a topical frame for what follows (Khan 2008, 1005). This pragmatic usage would then have been conventionalised (see Bybee 2010, 48 and the discussion above), e.g.

The preceding topic-establishing discourse may be propositional, as in (17, 19–23), or it may be a non-propositional constituent such as an adverbial, as in (18) (repeated here as (27)), or a noun phrase (28—the example is from the C. Mawana dialect):

Conditional clauses and topics are coded identically in a number of unrelated languages. This reflects the fact that their semantic analysis is very similar (Haiman 1978; Ebert, Ebert, and Hinterwimmer 2014). This would have facilitated the proposed development of the bət-qaṭəl.

2.2.2. Sequentiality and Habituality

It was noted above that when the bət-qaṭəl(wa) form expresses discourse dependency, the eventuality it presents is sometimes temporally sequential to what precedes but other times is an elaboration without temporal sequentiality. This can be understood as arising from the fact that its reference point in the preceding discourse is not temporal but rather topical. The fact that it is often used to express temporally sequential events is, therefore, an epiphenomenon arising from the fact that events expressed in successive clauses are typically temporally sequential. The construction, however, does not express temporal sequentiality directly.

When the bət-qaṭəl(wa) form is used to express discourse dependency, it most frequently expresses the imperfective aspect of habituality. It is very commonly used in expository discourse describing customs. As we have seen, the bət-qaṭəl(wa) form does not express habituality in its other functions of future and apodosis constructions. Why does the discourse dependent bət-qaṭəl(wa) form most commonly have habitual meaning?

Since the discourse dependent bət-qaṭəl(wa) does not have a future meaning, we must assume that it has acquired a reference time that coincides with the eventuality that it denotes. The development can be represented as follows, where T = Topic:

Table 3: Sequentiality and Discourse Dependency of bət-qaṭəl(wa)

A chain of bət-qaṭəl(wa) forms that comment on a topical situation would share the same topical reference point. This can be represented thus:

T1—R1,E1, + T1—R2,E2 + T1—R3,E3 + T1—R4,E4

This can be regarded as the resumption of the topic by a form of anaphora, analogously to the way topical referents are resumed by anaphoric pronouns.

These anaphoric topics are variables that are bound by and dependent on the antecedent topic, just as anaphoric pronouns are variables bound by an antecedent topic. This anaphoric binding of the topic can be regarded as a type of modality, so the construction is modal, just as a bət-qaṭəl(wa) form in an apodosis is modal. Indeed according to some approaches, the binding of anaphoric pronouns to antecedents is also a type of modality (Roberts 1987; 1989).

It is noteworthy that the bət-qaṭəl form in C. Barwar is not used in generic predicates such as (29).

This is because such generic predicates are typically independent of discourse context and do not comment on a preceding topical situation.7

The normal habitual meaning of the discourse dependent bət-qaṭəl form most likely arises from a retention of the contingent semantics of a conditional apodosis. As remarked, habitual predications are not completely ‘lawlike’ (Dahl 1985, 97) and are contingent on circumstances.

As we have seen, the bət-qaṭəl(wa) form is attested occasionally in narratives (22–23). Following the analysis that has just been proposed, we may say that they have the communicative function of expressing a comment on a previously mentioned situation, which has been set up as a topic. This analysis is appropriate for (22) (repeated below as (30)), since it consists of two situations that are set up in contrastive opposition. Contrastive oppositions are typically expressed by contrasting topics. The topical situations can be glossed by ‘when’–clauses:

Example (23) (repeated below as (31)) can be given a similar analysis of topical situation—comment, with the topical situation glossed by a ‘when’–clause:

These constructions in narrative contain what can be termed an ‘episodic topic’ with the status of an adverbial expression that sets the spatio-temporal frame for what follows. They appear to be used to mark boundaries in the discourse. In (30), as remarked, the two episodic topics set up two episodes in contrastive opposition. In (31) the episodic topic marks the onset of a new section of narrative.

In some NENA dialects the discourse dependent form with the original future particle bət/bəd has developed further and can be used as an actual present without first presenting a situation as its topic. This was the case in the now extinct Jewish dialect of Dobe (on the north bank of the Zab river). In this dialect, for example, the form bd-ez (fut-go.sbjv.3ms) can be used as

This extension of the construction to the actual present can be explained as having arisen by a process similar to Greenberg’s (1978) ‘cycle of definiteness’, whereby anaphoric pronouns develop into non-anaphoric definite articles. The anaphoric topical component of the dependent bət-qaṭəl construction has come to be used where the speaker assumes the hearer can identify the situation that is being referred to without explicitly presenting an antecedent topical situation in the preceding discourse. It appears that the speaker assumes that the hearer can identify the speech situation as the situation that is being talked about, i.e. it expresses situational immediacy.

3. Past Forms with the Prefix qam-

3.1. Attested Constructions

In many NENA dialects, a perfective past tense is expressed by combining the prefixed particle qam-, or variant forms of this, with the subjunctive verb form qaṭəl. The variant forms of the prefix attested across the dialects arise from a range of phonetic reductions, e.g. qəm (C. Barwar, etc.: vowel centralisation), kəm (C. Qaraqosh, etc.: vowel centralisation and fronting of the uvular), qa (C. Koy Sanjak: elision), tam, ta (C. Sulemaniyya and C. Sanandaj: fronting of the uvular and elision). In the documented dialects this construction is restricted to verbs with pronominal suffixes expressing the direct or indirect object. It is used predominantly to express past perfective events in narrative (33–34) or the occurrence of a punctual event in the recent past in conversational discourse (35). In the examples the particle is given the gloss qam:

Various hypotheses have been proposed for the historical origin of this verbal construction.9 Pennacchietti (1997) argued that it developed from a construction consisting of the verb qayəm ‘he gets up’ followed by the subjunctive, e.g. qayəm qaṭəl-le. He found support for this in the grammar of Rhétoré (1912, 225–26), who states that such a construction could be used in the sense of ‘aussitôt il le tua’ (‘he immediately killed him’), i.e. the event happened immediately after the event mentioned before it. Rhétoré, who unfortunately does not specify in which dialect(s) he found this construction, states that its original meaning was ‘se levant, il le tue’ (‘getting up, he kills him’), i.e. a sequence of events in the present.10 Pennacchietti, however, proposes that it originated as a construction expressing the immediate future ‘he will immediately kill him’, comparing constructions such as qemən ʾazən ‘I shall immediately go’, which are found in various NENA dialects. He draws attention to the fact that several languages use a construction that originated as the expression of the immediate future to narrate a sequence of events in the past, e.g. Catalan11 (36) and late medieval French (14th-16th century) (37):

The second verb in the original construction qayəm qaṭəl-le would, therefore, have been a subjunctive in a purpose clause ‘he gets up in order to kill him’ rather than an indicative present, as suggested by Rhétoré’s translation ‘se levant, il le tue’.

Here I would like to present some additional data that strengthen Pennacchietti’s hypothesis.

In some NENA dialects a subordinating particle regularly occurs before initial /ʾ/ verbs after the qam in the qam-qaṭəlle construction, e.g.

This reflects the fact that the verb after the qam was originally a subordinate subjunctive. Some isolated cases of the subordinating particle are found before initial /ʾ/ verbs in C. Barwar, (Khan 2008, 609), e.g.

The crucial missing link in the evidence, however, is provided by the dialects from the north-western sector of NENA. In some dialects in this region, such as the dialects in the area of the Cudi mountain and Billin, the initial inflected verb in immediate future constructions such as qemən ʾazən has become grammaticalised to qam without person inflection. Most of my data come from the Harbole dialect (Cudi), e.g.

Harbole (field notes)12

This immediate future construction can be used with both intransitive and transitives verbs. An example with a transitive verb is (41):

The future construction may be given a past reference time (‘was about to’) by adding the past shifting suffix -wa, e.g.

This can be used in initial position in a discourse such as the following:

The construction is also used in narrative discourse in the Harbole dialect to express an event that occurred immediately after the event mentioned in the preceding clause. According to speakers, an event expressed by the construction in this context is typically unexpected and surprising, i.e. it has a mirative function. The immediate future form is used with or without the past shifting -wa suffix, i.e. qam-ʾaxlən or qam-ʾaxlənwa, e.g.

As can seen in (45), the lexical verb of the construction does not have to have a pronominal object suffix, as is the case in most other NENA dialects. Indeed the verb can be intransitive, e.g.

A further variant of the construction is the substitution of the qaṭəl form by the qṭəlle form after the qam in past narratives, e.g.

Speakers judge constructions such as qam-mətle in (48) to express an unexpected and surprising event, but to be in the less immediate past than qam-mayət (46).

3.2. Analysis

In this section I should like to propose an explanation as to how the immediate future constructions with qam came to be used to express past events in narrative.

An example such as (44) above is likely to have originated in a juxtaposition of the immediate future construction with the previously mentioned situation, which meant ‘I went out and I was about to see my wife’. The reference time of the immediate future, therefore, coincided with the reference time of the first event. This can be represented as follows, with the index on R indicating the coincidence of reference time:

This was a strategy for expressing the immediate succession of the events. The preparatory onset phase of the second event overlaps with the first event. The events are connected together cognitively in the same mental space. It was also a strategy for giving prominence to the second event by anticipating it before it had occurred in the narration of events.

An immediate future construction such as qam-ʾazən ‘I am about to go’ would have been a deontic future, but as with other future constructions, when used with a third person, i.e. qam-ʾazəl ‘he is about to go’, there is an implicature that the event will happen, and so the construction comes to function also as a predictive future, presenting the event as a certainty. When combined with a past event, as in (44), the certainty of the future occurrence shifts to the assertion of the occurrence of the event in close succession to the preceding event. This comes about by the qam-construction acquiring a reference time that coincides with its event:

The original reference time of the qam-construction is retained (R1) and this expresses a sequential or continuative meaning, i.e. the event took place at reference time R2 in relation to reference time R1. The past tense of qam-xazənna is not expressed morphologically. The form can be assumed to have had a ‘relative tense’ that was determined by being bound to the R1 of the past verb pləṭli. This past verb was marked morphologically as past, i.e. its reference time was before speech time (R1—S).13

When the construction developed the meaning of immediate sequence, this marked type of sequence was associated with the implicature of mirativity (surprise), and also with salience and, in turn, with the recent past. Speakers of the Harbole dialect report that the construction has these associations.

The qam-constructions in examples such as (45), with a preceding imperfective circumstantial expression (‘When I was coming here’) and (47) with a posterior temporal adverbial clause (‘After I and my friend went out’) cannot be so easily analysed as having the temporal structure R1—R2,E with two reference times, one anterior to the other. This is because from a temporal point of view the eventuality in the qam-clause is overlapping with, rather than sequential to, what is expressed by the circumstantial construction and the ‘after’ adverbial clause. It appears that the first reference point is no longer temporal but rather has become a cognitive topical anchor, similar to the process described in §2.2. in the development of the discourse dependency bət-qaṭəl(wa) form. This can be represented as T—R,E, where T = the episodic topical situation that acts as the spatio-temporal frame for what follows:

When I was coming here (topical frame)—I saw a lion.

After I and my friend went out (topical frame)—he fell into the canal.

Note that in (45) the past shifting suffix -wa is added to the qam-construction (qam-xazənwa), indicating that it has its own tense marking, i.e. its tense is absolute and it is not bound as a relative tense to the tense of a preceding verb. This is likely to be because it is preceded by a tenseless expression (ʾana bitaya ‘I coming’).

The qam-construction is not obligatory in constructions of the type illustrated in the preceding examples. Contrast the following:

The (a) sentences of (49) and (50) with the past perfective qṭəlle form are matter-of-fact descriptions of events. The (b) sentences present the events as surprising and unexpected.

As has been remarked, a hybrid construction exists, in which the qaṭəl form in the qam-construction is substituted by the qṭəlle form ((48) repeated below as (51)). This still has a different pragmatic association from a corresponding construction with qṭəlle without the preceding qam- (52):

According to my consultant for the dialect, (51) implies a causal connection between the going out and the death whereas there is no such implicature in (52). This could arise from the fact that the qam-qṭəlle construction, like the qam-construction with the qaṭəl form, expresses immediate succession and close cohesion in the same mental space, an implicature of which could be causal connection.

In Harbole, as we have seen, the qam-construction with the qaṭəl form can be used in narrative in both transitive and intransitive clauses. There is no obvious feature of behaviour of the construction in this dialect that could explain why it became restricted to transitive verbs with pronominal objects in most of NENA. This specific distribution in other dialects appears to have been a strategy to avoid using internal pronominal objects in the qṭəlle form, especially 1st and 2nd person objects.14 In such dialects the qam-construction became a general past perfective form substituting for qṭəlle where the verb would have had pronominal objects. One can compare this to, for example, to the vaig cantar construction in Catalan (see (36)), which became a general past perfective form. This development of the qam-construction would doubtless have been facilitated by the fact that already in Harbole the temporal reference point of the construction in the preceding clause came to be analysed as a topical reference point in some contexts (i.e. R1—R2,E became T—R,E). As with the bət-qaṭəl(wa) form, this would have facilitated the use of a series of qam-constructions with the same spatio-temporal topical frame in narratives.

The generalised qam-construction did, however, retain some of the features of its embryonic form seen in the Harbole dialect. Firstly, when it takes 3rd person pronominal objects in narrative, these are anaphoric to the preceding discourse so the verb is sequential to or at least continuative of what precedes, as it is in Harbole. Secondly, speakers of some dialects report that in conversational discourse where the expression of a pronominal object is possible in both a qṭəlle or qam-construction, the qam-construction refers specifically to the recent past whereas the qṭəlle form does not have this restriction, e.g. qam-xazən-a (qam-see.sbjv.1ms-3fs) ‘I have just seen her’ vs xəzy-a-li (see.pst-3fs-1s) ‘I saw her’ (not necessarily recently) (Shaqlawa dialect, field notes). This is reflected by the fact that the particle na, which is used in the Shaqlawa dialect to express immediacy, is more frequently used with the form qam-xazəna than with xəzyali.15

4. Narrative Subjunctive

4.1. Attested Constructions

In NENA dialects the bare present stem qaṭəl functions as a subjunctive. This is used in a variety of irrealis contexts, including jussive main clauses, irrealis subordinate clauses and conditional clauses. An example of a subjunctive clause in a subordinate purpose clause from the C. Barwar dialect is given in (53):

The indicative present, such as the progressive and habitual, are typically expressed by innovative forms, such as the following in the C. Barwar dialect:

Indicative habitual

Progressive

hole qṭala (deictic copula + progressive stem) or qṭalɛ=le (progressive stem + enclitic copula). The progressive stem is derived historically from the infinitive:

In the C. Barwar dialect the qaṭəl subjunctive form is often used in narratives as a perfective sequential form. It typically continues an event or events that are expressed by a narrative past form (qṭille or qṭilɛle), e.g.

Sporadically the qaṭəl form is used as a sequential habitual:

The use of the bare qaṭəl form in narratives looks prima facie like the active participles that are commonly found in narratives in earlier types of Aramaic, such as Biblical Aramaic, e.g.

This is, however, a false analogy, since the Barwar qaṭəl form, although derived historically from the active participle of earlier Aramaic, now has a subjunctive function. A closer analogy to the use of active participles in earlier Aramaic narratives is the occasional use of the progressive form inflected by a copula (qṭalɛ=le) or sometimes without a copula (qṭala) in C. Barwar narratives, e.g.

In some NENA dialects the progressive construction is, indeed, the verb form that is most commonly used in narratives. This is the case, for example, in the C. Urmi dialect:

4.2. Analysis

I should like to argue here that the narrative qaṭəl form is indeed a modal subjunctive, which has been extended from its use in subordinate clauses, in particular purpose clauses. In C. Barwar, purpose clauses are introduced by a particle, typically a directional preposition such as ta ‘to’ as in (62), or are asyndetic without an introductory particle (63) (Khan 2008, 582–83, 667, 995–95):

Purpose clause constructions are also used to express the final outcome or result of a preceding action, whereby the speaker presents the chain of events from the viewpoint of this outcome (Khan 2008, 995), e.g.

Such result clauses appear to have developed from the common implicature of purpose clauses that the event took place, especially after verbs of movement, e.g. I went to buy some bread has the implicature that I did in fact buy bread (Schmidtke-Bode 2009, 178). This conventionalisation of an implicature as the expression of a real event that is reflected in result clauses is likely to have been the pathway of development also of dependent narrative qaṭəl forms. This may be represented as follows:

This analysis is similar to that of the development of the qam-construction forms discussed above. In the purpose clause construction the subjunctive verb is irrealis and takes as its reference time that of the main clause (represented by the repeated R1 in (65a)). The event time of the purpose clause, therefore, is posterior to its reference time and so the predicate is analogous to a future construction such as the immediate future qam-construction. In (65b) and (65c) the subjunctive clause has been reanalysed as a real asserted event. This involves the acquisition of a reference time coinciding with the event (R2). It can be assumed that the clause retains the R1 reference time, to which it is posterior. This reflects its reanalysis as a real asserted event that is sequential to what precedes.

As remarked, in the C. Urmi dialect the normal narrative verb form is the progressive. The narrative subjunctive is, however, sporadically used in this dialect. It is significant that in the attested examples in my text corpus it occurs predominantly after verbs of movement:

Cross-linguistically purpose clauses are very commonly preceded by verbs of movement (Schmidtke-Bode 2009, 98) and this is also the case in NENA dialects, see (60)–(61) from C. Barwar. An example from C. Urmi is (67), where the purpose clause is asyndetic:

This can be taken as evidence, therefore, that the sequential narrative subjunctive had its origin in subordinate purpose clauses.

The subjunctive qaṭəl form in purpose clauses is neutral as to aspect and can be used in habitual contexts, e.g.

This could explain the sporadic use of qaṭəl as a habitual sequential (see (58)).

According to (65) the narrative subjunctive, which developed from subordinate clauses, retained the temporal structure of result clauses, i.e.

The ‘main clause’ in this model of its development would be a clause in the preceding discourse. As can be seen in the examples (56) and (57) the narrative subjunctive can be used in chains of clauses. This could be represented thus:

R1—R2,E2 + R2—R3,E3 + R3—R4,E4 + R4—R5,E5

Each subjunctive form would take as its anterior reference time the reference time of the previous verb.

Alternatively, it could be proposed that by a further development the anterior reference time in the structure R1—R2,E has become schematised to a cognitive topical reference point analogously to the analysis we have proposed for the development of the bət-qaṭəl and qam-qaṭəl forms, i.e. T(opic)—R,E. This, therefore, would not be a temporal point in the preceding discourse but rather a topical frame, which in narrative would be typically a spatio-temporal frame. The chain of narrative subjunctives would all cohere together in the same ‘mental space’ and be linked to this spatio-temporal frame. This is clear in (57), in which the chain of narrative subjunctives begins after the adverbial ʾaṣərta ‘in the evening’, which sets the spatio-temporal frame. The clauses would anaphorically resume this topical frame, thus:

T1—R1,E1 + T1—R2,E2 + T1—R3,E3 + T1—R4,E4 + T1—R5,E5

It has been remarked that the progressive form is occasionally used in C. Barwar as a narrative form and that this is the normal narrative form in C. Urmi. It is relevant to note that the progressive form can also express purpose, similarly to a subjunctive clause, e.g.

This usage possibly originated in the use of the progressive in circumstantial constructions such as the following

The sequential narrative use of the progressive may, therefore, have also developed through the pathway of a purpose clause.

5. Conclusions

In this paper I have discussed various verbal forms in NENA dialects that express dependency on the preceding discourse beyond the syntactic confines of a sentence. These include the bət-qaṭəl(wa) form, the qam-qaṭəl form and the narrative subjunctive form. These can be used to express continuity of the preceding discourse, which can be interpreted as temporal sequence or elaboration according to the context. The proposed historical development of the three verbal forms with these functions share several common features. In all cases the verbs originally expressed some kind of future or, to be more precise, an event time that was posterior to its reference time. From the perspective of this reference time they expressed contingent events that were modally dependent on a preceding eventuality.

Parallels to such discourse dependent verbal forms have been documented in a variety of other languages. Numerous languages of Africa have special verbal forms for the expression of continuity in discourse. These are used, for example, for the chaining of clauses in narratives and descriptions of habitual procedures.16 This continuity may be temporal sequence or elaboration. Such forms are often identical to forms that express modal subordination in subordinate clauses and so have been referred to as narrative subjunctives (R. Carlson 1992; Seidel 2015, 180). In some African languages the consecutive forms can be used independently of preceding discourse as a future or modal form denoting an unrealised action (e.g. Seidel 2015, 186). Historical reconstructions of Oceanic languages have revealed connections between narrative continuity devices and future verbal forms (Lichtenberk 2014).

Within Semitic one can find some parallels to what has been described in this paper. Owens (2018) argues persuasively that the preverbal particle b- that is found in a variety of Arabic dialects originates in the deontic verb baġa ‘to want’ (cognate with Aramaic baʿe). What is of interest is that although it has retained its deontic or modal sense in some dialects of the Gulf, in some dialects it has developed into an indicative (e.g. Levant). The missing link, Owens claims, is its use in Nigerian Arabic to express what he calls ‘propositional adjacency’, which corresponds to what I have been calling here discourse dependency. The situation in the J. Dobe dialect, where the bət-qaṭəl form can be used as an indicative, would represent the same advanced stage of development as indicative b- in Levant Arabic.

The indicative preverbal particle ka- in Moroccan Arabic appears to have its origin in the modal use of the auxiliary verb kān in conditional clauses (Corriente 1977, 140–41; Stewart 1998, 111–12; Hanitsch 2019, 256–58). This also, therefore, may have followed a similar pathway of development as Neo-Aramaic bət-qaṭəl.

Tsukanova (2008) has identified the use of dependent subjunctive forms containing the modal auxiliary čān in Gulf Arabic as a continuative form in narratives.

The Neo-Aramaic discourse dependent bət-qaṭəl and the narrative subjunctive exhibit close parallels also with continuative verbal forms in Biblical Hebrew.17

Finally, in addition to cross-linguistic typological parallels one should also take into account the factor of language contact. It is noteworthy, for example, that the NENA dialects that exhibit the bət-qaṭəl future forms are/were mainly in the region of Northern Kurdish (Kurmanji). The NENA dialects in the region of Central Kurdish (Sorani) do not generally exhibit a specific future marker. This distribution corresponds to the presence of a dedicated future verbal prefix in Northern Kurdish and the absence of such a prefix in Central Kurdish. What is of particular interest is that in Northern Kurdish verbs with the future particle are sometimes used as discourse dependency habitual forms just like the bət-qaṭəl(wa) form in Neo-Aramaic (Haig 2018, 292).

References

Bertinetto, Pier Marco, and Alessandro Lenci. 2012. ‘Habituality, Pluractionality, and Imperfectivity’. In The Oxford Handbook of Tense and Aspect, edited by Robert I. Binnick, 852–80. Oxford: Oxford University Press.

Boneh, Nora, and Edit Doron. 2013. ‘Hab and Gen in the Expression of Habituality’. In Genericity, edited by Alda Mari, Claire Beyssade, and Fabio Del Prete, 176–91. Oxford: Oxford University Press.

Bybee, Joan L. 2010. Language Usage and Cognition. Cambridge: Cambridge University Press.

———. 2015. Language Change. Cambridge: Cambridge University Press.

Carlson, Gregory. 2012. ‘Habitual and Generic Aspect’. In The Oxford Handbook of Tense and Aspect, edited by Robert I. Binnick, 828–51. Oxford: Oxford University Press.

Carlson, Gregory N., and Beverly Spejewski. 1997. ‘Generic Passages’. Natural Language Semantics 5: 101–65.

Carlson, Robert. 1992. ‘Narrative, Subjunctive, and Finiteness’. Journal of African Languages and Linguistics 13: 59–85.

Coghill, Eleanor. 2016. The Rise and Fall of Ergativity in Aramaic: Cycles of Alignment Change. Oxford Studies in Diachronic and Historical Linguistics 21. Oxford: Oxford University Press.

Comrie, Bernard. 1985. Tense. Cambridge University Press.

Corriente, Federico. 1977. A Grammatical Sketch of the Spanish Arabic Dialect Bundle. Madrid: Instituto Hispano-Árabe De Cultura.

Dahl, Östen. 1985. Tense and Aspect Systems. Oxford-New York: Blackwell.

Dinsmore, John. 1982. ‘The Semantic Nature of Reichenbach’s Tense System’. Glossa 16: 216– 239.

———. 1991. Partitioned Representations. A Study in Mental Representation, Language Understand and Linguistic Structure. Dordrecht: Kluwer.

Ebert, Christian, Cornelia Ebert, and Stefan Hinterwimmer. 2014. ‘A Unified Analysis of Conditionals as Topics’. Linguistics and Philosophy 37: 353–408.

Fassberg, Steven. 2015. ‘The Origin of the Periphrastic Preterite Kəm/Qam-Qāṭəlle in North-Eastern Neo-Aramaic’. In Neo-Aramaic and Its Linguistic Context, edited by Geoffrey Khan and Lidia Napiorkowska, 172–86. Piscataway, NJ: Gorgias Press.

Fauconnier, Gilles. 1994. Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge: Cambridge University Press.

Fillmore, Charles J., Paul Kay, and Mary C. O’Connor. 1988. ‘Regularity and Idiomaticity in Grammatical Constructions’. Language 64: 501–38.

Fox, Samuel E. 2015. ‘The History of the Future’. In Neo-Aramaic in Its Linguistic Context, edited by Geoffrey Khan and Lidia Napiorkowska, 162–71. Piscataway, NJ: Gorgias Press.

Goldberg, Adele E. 1995. Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press.

———. 2006. Constructions at Work: The Nature of Generalization in Language. Oxford: Oxford University Press.

Greenberg, Joseph. 1978. ‘How Does a Language Acquire Gender Markers?’ In Universals of Human Language, edited by Joseph Greenberg, 3:47–82. Stanford: Stanford University Press.

Haig, Geoffrey. 2018. ‘The Iranian Languages of Northern Iraq’. In The Languages and Linguistics of Western Asia: An Areal Perspective, edited by Geoffrey Haig and Geoffrey Khan, 267–304. The World of Linguistics 6. Berlin: De Gruyter Mouton.

Haiman, John. 1978. ‘Conditionals Are Topics’. Language 54: 564–89.

Hanitsch, Melanie. 2019. Verbalmodifikatoren in den Arabischen Dialekten: Untersuchungen zur Evolution von Aspektsystemen. Porta Linguarum Orientalium. Neue Serie 27. Wiesbaden: Harrassowitz.

Hatav, Galia. 2012. ‘Bound Tenses’. In The Oxford Handbook of Tense and Aspect, edited by Robert I. Binnick, 611–37. Oxford: Oxford University Press.

Jacobs, Bart. 2011. ‘Present and Historical Perspectives Onthe Catalan Go-Past’. Zeitschrift für Katalanistik 24: 227–25.

Johnson, Marion R. 1981. ‘A Unified Temporal Theory of Tense and Aspect’. In Syntax and Semantics, 14: Tense and Aspect, edited by Philip J. Tedeschi and Annie Zaenen, 145–75. New York: Academic Press.

Khan, Geoffrey. 2002. The Neo-Aramaic Dialect of Qaraqosh. Studies in Semitic Languages and Linguistics 36. Boston, MA: Brill.

———. 2008. The Neo-Aramaic Dialect of Barwar. 3 vols. Leiden: Brill.

———. 2016. The Neo-Aramaic Dialect of the Assyrian Christians of Urmi. 4 vols. Studies in Semitic Languages and Linguistics 86. Leiden-Boston: Brill.

———. 2017. ‘Ergativity in Neo-Aramaic’. In Oxford Handbook of Ergativity, edited by Jessica Coon, Diane Massam, and Lisa Travis, 873–99. Oxford: Oxford University Press.

———. 2021 (to appear). ‘The Coding of Discourse Dependency in Biblical Hebrew Consecutive Weqaṭal and Wayyiqṭol’. In New Perspectives in Biblical and Rabbinic Hebrew, edited by Aaron D. Hornkohl and Geoffrey Khan. Cambridge Semitic Languages and Cultures. Cambridge: University of Cambridge & Open Book Publishers.

Langacker, Ronald. 1987. Foundations of Cognitive Grammar: Theoretical Prerequisites. Vol. 1. Stanford: Stanford University Press.

Lichtenberk, Frantisek. 2014. ‘Sequentiality-Futurity Links’. Oceanic Linguistics, 53: 61–91.

Longacre, Robert E. 1990. Storyline Concerns and Word Order Typology in East and West Africa. Studies in African Linguistics Supplement 10. Los Angeles: University of California.

Mutzafi, Hezy. 2002. ‘Barzani Jewish Neo-Aramaic and Its Dialects’. Mediterranean Language Review 14: 41–70.

Noorlander, Paul. 2017. ‘The Proximative and Its Correlatives in North Eastern Neo-Aramaic’. In Prospective and Proximative in Turkic, Iranian and Beyond, edited by Agnes Korn and Irina Nevskaya, 188–210. Wiesbaden: Reichert Verlag.

———. 2018. ‘Alignment in Eastern Neo-Aramaic Languages from a Typological Perspective’. Ph.D. Thesis, Leiden: University of Leiden.

Owens, Jonathan. 2018. ‘Dialects (Speech Communities), the Apparent Past, and Grammaticalization: Towards an Understanding of the History of Arabic’. In Arabic Historical Dialectology, edited by Clive Holes, 206–56. Oxford Studies in Diachronic and Historical Linguistics 30. Oxford: Oxford University Press.

Palmer, Frank R. 1986. Mood and Modality. Cambridge: Cambridge University Press.

Payne, Doris L., and Shahar Shirtz, eds. 2015. Beyond Aspect: The Expression of Discourse Functions in African Languages. Typological Studies in Language 109. Amsterdam-Philadelphia: John Benjamins.

Pennacchietti, Fabrizio. 1994. ‘Il Preterito Neoaramaico Con Pronome Oggetto’. Zeitschrift der Deutschen Morgenländischen Gesellschaft 144: 259–83.

———. 1997. ‘On the Etymology of the Neo-Aramaic Particle Qam/Kim-’. Massorot 9–11: 475–82.

Persohn, Bastian. 2017. The Verb in Nyakyusa: A Focus on Tense, Aspect and Modality. Berlin: Language Science Press.

Reichenbach, Hans. 1947. Elements of Symbolic Logic. New York: Macmillan.

Rhétoré, P. Jacques. 1912. Grammaire de la Langue Soureth ou Chaldéen Vulgaire. Mossoul: Imprimerie des Peres Dominicains.

Roberts, Craige. 1987. ‘Modal Subordination, Anaphora, and Distributivity’. Ph.D. Thesis, Amherst, N.Y: University of Massachusetts, Amherst.

———. 1989. ‘Modal Subordination and Pronominal Anaphora in Discourse’. Linguistics and Philosophy 12: 683–721.

Schmidtke-Bode, Karsten. 2009. A Typology of Purpose Clauses. Amsterdam-Philadelphia: Benjamins.

Seidel, Frank. 2015. ‘Rethinking Narrative Tenses Based on Data from Nalu (Atlantic) and Yeyi (Bantu)’. In Beyond Aspect: The Expression of Discourse Functions in African Languages, edited by Doris L. Payne and Shahar Shirtz, 177–217. Typological Studies in Language 109. Amsterdam-Philadelphia: John Benjamins.

Stewart, Devin J. 1998. ‘Clitic Reduction in the Formation of Modal Prefixes in the Post-Classical Arabic Dialects and Classical Arabic Sa-/Sawfa’. Arabica 45: 104–28.

Tsukanova, Vera. 2008. ‘Discourse Functions of the Verb in the Dialect of Sedentary Kuwaitis’. In Between the Atlantic and Indian Oceans. Studies on Contemporary Arabic Dialects. Proceedings of the 7th AIDA Conference, Held in Vienna from 5–9 September 2006, edited by Stephan Procházka and Veronika Ritt-Benmimoun, 447–55. Münster-Vienna: LIT-Verlag.

Verkuyl, Henk J. 2012. ‘Compositionality’. In The Oxford Handbook of Tense and Aspect, edited by Robert I. Binnick, 563–85. Oxford: Oxford University Press.