Projection in Arabic: a typological overview

3.1 Paratactic elaboration exposition

In the paratactic elaboration exposition type (Example 1), the second clause restates the firsts in different words in order to reinforce it (Halliday and Matthiessen 2004: 397).

1. Paratactic extension alternation

As to the paratactic extension alternation type (Example 2), one clause is presented as an alternative to the other (Halliday and Matthiessen 2004: 407).

1. Paratactic enhancement temporal

The paratactic enhancement type is coordinative with a circumstantial feature. The temporal subtype (Example 3) plays a role in construing an event time line. In all three examples above, the clauses are of equal status (i.e. independent). The examples below are construed by clauses of unequal status. They too can be either elaborating, extending or enhancing as in Examples (4), (5), and (6).

1. Hypotctic elaboration

Hypotactic elaboration (Example 4) is realized by non-defining relative clauses which are descriptive clauses that add general information that can be omitted without affecting the general meaning (Halliday and Matthiessen 2004: 399).

1. Hypotctic extension additive

Example (5) construes a non-finite hypotactic extending additive nexus. These nexuses are typically introduced by a prepositional group functioning as a conjunction similar to ‘besides’, ‘additionally’, ‘apart from’, ‘other than’, etc (Halliday and Matthiessen 2004: 409).

1. Hypotactic enhancement

The nexus type in Example (6) construes what is traditionally known as an adverbial relation of purpose. In general, this type construes relations of time, place, manner, cause and condition (Halliday and Matthiessen 2004: 416). Example (6) brings to an end the examples meant to illustrate some of the expansion relations construed by some clause nexuses in Arabic. The list is short and far from being exhaustive as the focus of this article is not on primary and secondary nexuses of expansion (cf. Bardi 2008 for more information) but rather on primary and secondary relations of projection.

In regards to projection, this relation type yields the distinction between direct speech or thought (parataxis) and indirect speech or thought (hypotaxis), as illustrated in Examples (7), (8), (9), and (10).

In Example (7), the relation between the first and the second clause is one of equality. They are both independent and paratactically combined. The process used in the initial projecting clause ṣāḥa (‘to shout’) is a verbal process that projects and at the same time construes the manner of the ‘saying’. The projected clause is a proposal, an invitation to someone to visit again. The writer uses a colon at the end of the projecting clause and quotation marks to differentiate between the projected and projecting parts of the nexus. Though this way of punctuating citations is quite common, it is important to mention that there is no standard punctuation practices in Arabic. Sometimes the writer would use quotation marks, others just a hyphen or a combination of both. In classical Arabic texts, for instance The Noble Quran, nothing is used to mark the boundary between what is projected and what is projecting. Interestingly, all the recently edited books of hadiths (the saying of the prophet peace be upon him) which appeared around the same time as The Noble Quran are punctuated.

In Example (8), the relation between the clauses in the nexus is not equal, the imbalance and the boundary between the clauses is marked by the particle ʾanna. In this hypotactic nexus, case marking, more precisely ʾism ʾanna which is typically manṣūb ‘in the accusative’ is often a mark of a projected clause (see Section 4.1.2).

Examples (9) and (10) are mental clauses. The relation within the nexus in Example (9) is paratactic whereas in Example (10) it is hypotactic. The projected part in the nexus represents a “semiotic event”. The content of projection, either locutions or ideas is of a higher order of reality than the projecting part. This is reflected in the conventions of comics as illustrated in Matthiessen and Teruya (2014: 6). Projection can thus be realized by a combination of clauses: a clause of saying or thinking projecting a locution (speech) or idea (thought) into existence. Projection may also be realized by a prepositional or adverbial phrase representing the angle of projection in a ‘rankshifted’ way within a single clause as illustrated in Example (11).

The logical construal of projection at paratactic and hypotactic levels was discussed under different headings such as direct and indirect speech. It has also been discussed under the headings of “complementation”, especially at clause simplex level when what is being said is construed as ‘verbiage’. Examples (12) and (13) are an illustration of the latter whereas Examples (14) and (15) are an illustration of the former.

Though in the examples above both mental and verbal projections are construed by typical mental and verbal processes, there are instances of semantic shift when both projection types overlap (cf. Section 4.1.1).

Finally, in addition to being of different order of experience, expansion and projection are different in terms of the order in which the primary (the main/projecting) and the secondary (the subordinate/projected) clause occur in the nexus. In hypotactic nexuses, the primary clause in projection typically occurs in initial position as in Examples (16) and (17) whereas in expansion the hypotactic nexus has more flexibility as to where it occurs as illustrated in Examples (18a) and (18b).

As to paratactic projection nexuses, there is a flexible variation in the order of the projecting and projected clauses as illustrated by Examples (19), (20), and (21). The quoting clause could even be placed within the quote. However, this is less common in Arabic than it is in English and more frequent in translations into Arabic than in Arabic novels. This last comment is based on the analysis of the data for this paper, further investigation is needed to confirm these findings.

Arab grammarians mostly discussed projection under the heading of “complementation” in spite of the fact that they developed a delicate verbal group taxonomy (cf. Section 2 above). So far this article has discussed not only Halliday’s notion of projection (mental and verbal) but also that of expansion (elaboration, extension, and enhancement). This was done without major analysis though. In the following section, there will be more focus on only one logico-semantic relation which is projection (for more information about expansion cf. Bardi [2008]).

4.1 The level and mode of projection

There are two types of projection: projection of wording (locution) and projection of meaning (idea). The level of projection and the mode intersect resulting in four subtypes. This basically means that projection could be realized as a quote or a report as illustrated in Figure 1 below. The distinction between quotes and reports has often been referred to in traditional grammar in terms of the contrast between direct and indirect speeches. In SFG, this contrast is extended to cover direct thoughts (quote) and indirect thoughts (report) as well.

Figure 1:

The level and mode of projection.

4.1.2 Projection markers

Projection is associated with a definite set of markers/particles. These markers are ʾanna, ʾan, ʾinna, and ʾin. The first three signal that the status within the projection nexus is unequal. Most importantly, some of them are more likely to be used in construing propositions, while others are more likely to be used in construing proposoals. The sub-sections below provide a comprehensive description of what each is capable of doing.

4.2.1 Reporting statements

There is a range of verbal processes one can choose from to introduce a report. The most generic process is qāla, it can be used to report a proposition or a proposal without depicting much else in regards to manner, attitude, etc.

The other options available to construe reports verbally vary extensively in Arabic. Some processes can be perceived as either objective or subjective as illustrated by Examples (56) and (57).

1. Subjective report

1. Subjective report

Both examples above are subjective. Example (56) is quite negative as it suggests that the person who is relaying the utterance does not have much trust in its authenticity whereas Example (57) is quite positive as it suggests that what is being reported is of some benefits to the receiver (cf. Table 1 below for more options).

Some reporting verbs depict the way the quote or the report is delivered. The options vary from least to most visual. Interestingly this type of projection is often realized paratactically (i.e. as quotes) by behavioural processes ‘pressed into verbal services’ as in Example (58) below. The ‘visual effects’ may be enhanced by adding a prepositional group to the behavioural process as illustrated by Example (59) or an adverbial as in Example (60). The same modulation of the ‘visual options’ apply to verbal processes as in Example (61), the only difference is that verbal processes could project with more ease both paratactically and hypotactically.

1. Manner of delivery: Behavioural process

1. Manner of delivery: Behavioural process + prepositional group

1. Manner of delivery: Behavioural process + prepositional group

1. Manner of delivery: process + adverbial

While the personal and spatial deictic variations between the quote and the report are quite simple and predictable; the deictic differences in temporal values are modal and much less straightforward. They are also more complex in reporting propositions than proposals as will be illustrated by Examples (62)–(64).

In Example (62), the verbal group in the projected clause is in the imperfective indicative. It is very likely that the ‘tense’ in the original quote is the same. The reason that there is no change in the report is that at the moment of speaking the newspaper reporting on the event believes that the financial package still includes grants with the same value. In Example (63), the projected clause is relational, it is also very likely that the original clause construes the same temporal value. The reason why there is no change is the fact that the clause construes a characteristic which is perceived as permanent at the moment of reporting the action i.e. those groups do not act randomly and have their own strategic plans. This example should be contrasted with the following one that is Example (64), where the clause is relational too but with a verbal group kāna (‘to be’) in the perfective. Unlike Example (63) in which there is no change in terms of temporal values between the report and the quote, in Example (64) the report has a different temporal value from the quote because the state of ‘being thirsty’ is finished at the moment of the report [he asked if I was thirsty (then)]. In brief, although these examples barely scratch the surface of the temporal shifts in quotes and reports, it seems there are instances when the person reporting the citation feels that these temporal shifts are warranted and others when they are not. This is often based on the way the action is perceived at the moment of the report, and is not systematic as in languages such as English.

4.2.2 Reporting polar and elemental interrogatives

In Arabic interrogatives can be reported at clause simplex level. This is typically done by using saʾala ‘to ask’as a projecting process and a prepositional phrase construing a circumstance Matter. The preposition of choice which is used as head of the prepositional group is typically ʿan ‘about’ as in Example (65) below. In these cases, the report is typically a general enquiry about a topic and not a proper report of the utterance even when the prepositional group is complex and may have a qualifying post posed embedded clause in it.

Reporting yes-no questions is typically done at clause complex level using saʾala + ʾin (‘to ask if/whether’). This version requires a change in personal, spatial and temporal deixis as illustrated by Example (66) where clearly there is a temporal shift between the quote and the report. The temporal value is realized in the original by the imperfective and in the report by kāna + the imperfective.

In reports, elemental questions are altered not only in terms of intonation patterns but also in terms of personal, spatial and temporal deixis depending on who is speaking and the position of the reporter vis-à-vis the Sayer and what was said. There is no change construed by particular markers or change in the placement of mood elements (Subject + Finite) like that which is done in English as in Example (67).

The projecting clause is in the perfective while the projected one is a nominal clause which means that there is no temporal shifts. These examples are quite challenging to translate as one has to look at the context for clues about the right temporal value construed by the verbless clause.

4.2.3 Reporting commands and suggestions

Reporting commands, advice and suggestions is realized at clause complex level by (1) a verbal process in the projecting clause, the most typical in commands is ʾamara ‘to order’, in advice is naṣaḥa ‘to advise’ and in suggestions ʾiqtaraḥa ‘to suggest’ + (2) ʾan + (3) a verbal group in ʾalmuḍāraʿ ʾalmanṣūb. This ‘tense’ is often compared to the subjunctive in some western descriptions of Arabic as it reflects the non-actualized nature of the action or in this case the proposal it construes as in Examples (68), (69), and (70).

Arabic can also report commands at clause simplex level. This is realized by (1) a verbal process + (2) a preposition + (3) a noun or nominalized verb form as in Example (71).

Arabic may also construe proposals in a gray area of the system of logico-semantic relations. The report in those instances is realized by a nominalized verbal form construing a non-finite dependent clause as illustrated by Example (72).

In Example (72), the unequal status is clearly marked by the use of the nominalized verbal form in the accusative. The non-finite status of the clause leaves it open to interpretation. It can be treated as a dependent clause projected hypotactically or simply a complex nominal group ‘head’ with a post-posed qualification construing verbiage at a clause simplex level.

5.1 Logical manifestation

As mentioned, the distinction between the two modes of projection, paratactic and hypotactic stems from the general contrast in taxis between parataxis and hypotaxis. Telling the difference between paratactic and hypotactic projection may represent a problem in certain texts in Arabic especially when quoting and reporting overlap. This has been particularly noticed in classical Arabic, in different verses of the Quran as in Examples (73) and (74) below and particularly when ʾinna is used (cf. Section 4.1.2.1.).

In the translations of Examples (73) and (74), the translator used quotation marks which are not in the original text. Though Example (73) can be a direct quote and therefore the translator made the right decision when he added the quotation marks, Example (74) is less likely to be a direct quote. In fact, as the words were said by a group of people (i.e. they said), this means that unless everyone was in sync and said the exact same words at the same time, what was reported is very unlikely to be a direct quote. The challenge of differentiating “direct quotes” and “indirect quotes” has been largely discussed (Buchstaller 2013: 55–64 in Matthiessen and Teruya 2014).

Matthiessen and Teruya (2014: 13) say that the different logico-semantic realizations of projection are similar to any of the other realizations in the other systems in that they are organized around a cline. In other words, although it is common that quoting and reporting may overlap with “free indirect speech” as an intermediate category, they too can be aligned along ‘a cline of integration’. At one end of this cline, there are quoted propositions and proposals that are completely “free”, without an accompanying projecting clause. Their projected status is normally established by the speaker, who for instance may take on the persona of a different individual, as in stand-up comedy. These utterances have all the interactive properties of a proposition or proposal serving as a dialogic move, including tone, mood tags, and modal assessment as in Example (75).

At the other end of the cline, one should find propositions and proposals that are totally removed from the line of dialogue and cleared of their interactive value. These are typically realized by clauses that are embedded in other clauses as in Example (76) below.

In-between these two poles, we find the realm of the nexus of projection, with three degrees of agnation which are:

1. “freedom” – paratactic projection as in Example (77).

1. Paratactic projection: reporting. This is common in projecting questions and commands, which typically lose their intonation patterns and as such are misleadingly paratactic as in Example (78) below.

1. Hypotactic projection: reporting. It is the norm for reported projection to be construed hypotactically and as mentioned above it typically takes a projection marker as in Example (79) below.

ʾalfatḥaẗu ‘the ʾa sound that construes the accusative’ is a sign of hypotactic projection as the particles that tie the main clause to the subordinate one, require either a nominal group or a verbal group in the accusative.

Finally, there are instances where both quotes and reports are mixed, as in Example (80) below which is from a newspaper article.

In the example above the writer uses both direct and indirect speech in the same citation. As mentioned earlier, in Halliday’s account, the three degrees are all related by tactic patterns within the domain of the clause nexus whereas in traditional accounts both paratactically and hypotactically projected clauses that is quoted and reported speech and thought are analysed as ‘complement clauses’. The next section will deal with the experiential manifestation of projection.

5.2 Experiential manifestation

The experiential manifestation of projection appears in the role the verbal group plays in construing a projection nexus. It also appears in the role some circumstances play at clause simplex level.

5.3 Interpersonal manifestation

The manifestation of projection in the interpersonal mode in Arabic is either congruent or incongruent i.e. metaphorical. The congruent aspect may overlap with the experiential especially when construing the angle/source of projection. The incongruent is in tandem with the logical resources as it involves projecting clauses. Both will be briefly described in the two sections below.

5.3.1 Congruent realization

The congruent interpersonal manifestation of projection is realized through the system of comment adjuncts in particular comments of presumptions (Matthiessen and Teruya 2014). These manifestations occur when the speaker interferes into a proposition or proposal with a comment. Comments of presumption show that the source of the information in a proposal is not the speaker but somebody else. The speaker in these situations does not take responsibility for the validity of the information but rather distances him/her-self from it (Matthiessen and Teruya 2014), as in Example (90) below.

(90)

ḥasaba maṣādira -mawṯūqaẗin- liـ“ʾaššurūqi”	“ʾannahḍa”	tadʿamu	ʾassibsī.
according indf.source-fpl.acc indf.trusted-fsg.gen to-“alshuruq-gen”	‘“alnahdha-nom’”	3fsg:ipfv-support	alsibsi-gen”
Angle: viewpoint	Actor	Process: material	Recipient
‘According to some ‘Al-Shuruq’ reliable sources: ‘Annahdha’ supports Al-Sibsi.’
(Al-Ajrudi 2014)

Like in English (cf. Matthiessen and Teruya 2014), comments of presumption in Arabic are one category within the system of comments (cf. Figure 2); related types include: Prediction (e.g. mina ʾalmufāǧiʾi, ‘surprisingly’/mina ʾalmutawaqqiʿi, ‘predictably’), Desirability (e.g. lisūʾi, ʾalḥaẓẓi ‘unluckily’/min ḥusni ʾalḥaẓẓi, ‘luckily’/lilʾasafi, ‘regrettably’). Comments of presumption are also related to comments concerned with the act of exchanging speech functions. These include: Validity (e.g. ʿumūman, ‘generally’/biṭarīqaẗin ʿāmmaẗin, ‘in a general way’), Personal engagement (e.g. šaẖṣiyyan, ‘personally’/biṣarāḥaẗin, ‘frankly’), Persuasion (bikulli mawḍūʿiyyaẗin, ‘objectively’/bikulli nazāhaẗin, ‘honestly’), factuality (e.g. ḥaqqan, ‘really’).

Figure 2:

The system of comment adjuncts (Halliday and Matthiessen 2004: 129).

The interpersonal manifestation of projection vary according to register where certain text types have their own preferred comment Adjuncts. For instance, in newspaper articles comment Adjuncts about the sources credibility (e.g. ḥasba maṣdarin mawṯūqin, ‘according to a trusted source’, ḥasba maṣdarin rasmīyyin, ‘according to an official source’, fī ʾanbāʾin mutaḍāribaẗin, ‘based on some conflicting news’, ḥasba baʿḍi ʾattasrībāti, ‘according to some leaked news’). are quite common. Some of these same Adjuncts could be slightly altered to accommodate the speaker’s negative or positive appraisal of the source and its credibility. For instance, ḥasba maṣdarin mawṯūqin (‘according to a trusted source’) could become ḥasba maṣdarin ġayri mawṯūqin (‘according to an untrustworthy source’) and ḥasba maṣdarin rasmīyyin (‘according to an official source’) can become ḥasba maṣdarin ġayri rasmīyyin (‘according to an unofficial source’) etc.