Keywords

1 Introduction

This paper claims that the metalinguistic richness of negation in Standard Arabic (abbreviated SA) has not triggered any significant research that distances itself from the traditional account of negation. Rather, traditional approaches to Arabic syntax still dominate the grammatical landscape and continue to exercise absolute authority in pedagogical grammar. Whether approached from a prescriptive, descriptive, explicative or typological perspective, pre-verbal and pre-nominal negators have been treated essentially as conveying a temporal value that accounts for their working in discourse: negation in the past, in the present, and in the future.

Based on a corpus of utterances collected from different sources, such as the International Arabic Corpus, the Quran, and literary texts, this study questions the chronological treatment of negation in the dominant theoretical and pedagogical grammar. It also shows that negators in SA do not function as time locators of the predicative relation (R) or work in free variation. Rather, they constitute a micro-system of interrelated units governed by an enunciative logic and contextual factors.

2 The Traditional Approach to Negators

The Arabic grammatical tradition should not be understood as a homogeneous school but as an episteme indicative of an autonomous stage in human linguistic thought. It shares with western traditional grammars their prescriptive, semantic, atomistic, taxonomic, context-insensitive, and writing-oriented approach that envisages not language at work but language as an end-product. These epistemic features are detectable in the treatment of negation and other grammatical operations in SA. In spite of its heterogeneity, the traditional approach to negation reflects a consensus on several premises and theoretical presuppositions:

  1. i.

    The main linguistic corpus used by all traditional grammarians is collected either from authentic Quranic and poetic texts or made of intuition-based sentences generally constructed with Zeid and ʻamr as hypothetical subjects.

  2. ii.

    Negation was not researched as an autonomous linguistic category but as a “linguistic style” associated with affirmation, its opposite. Compared with other grammatical operations, negation received scant mentions, often taxonomic and semantic, in the context of non-affirmation and reference to time. The most influential grammarians, such as Sibaweihi [54], Al Mubarrad [12], Al Zamakhshari [23], Ibn Hisham [42], Ibn Al Sarraj [41], Ibn Yaʻiish [45], and Ibn Jinni [43], to name a few, touched on “particles of negation” but never elaborated on negation.

  3. iii.

    All grammarians, except Al Jurjani [10, p. 417-418], a prominent rhetorician, considered affirmation to be the origin of speech, and negation extrinsic to the sentence’s basic structure, always affirmative.

  4. iv.

    Although some grammarians, Sibaweihi [54, vol. 2, p. 116] and Al Khaliil, [11, vol. 8, p. 350] assigned a corroborative ‘meaning’ to some negators, such as lam, and lan, Al Zamakhshari, [24, p. 407] and Al Suyuti, [14, vol.2, p.287], negators were always associated with extralinguistic temporal (present, past, future) values.

  5. v.

    The traditional approach was focused on the all-pervading theory of governance. Negators were described and classified according to their declensional potential or operative force (Versteegh [57, p. 6]. A typical traditional definition of a negator, such as lan, generally includes three functional properties: it negates, puts the verb in the accusative, and locates the event in the future.

  6. vi.

    Some grammarians, namely Ibn Jinni [43], adopted a morpho-semantic approach exploring forms of verbal and nominal negation other than negative particles. The case of morphological patterns, like /’afʻala/ and /faʻʻala/, which, by interlocking with a root, assign negative properties to the new lexical unit. This phenomenon is studied in Al-Sajustaanii [24], Ibn Saiyidih [44], and Al Zajjaaj [22].

Often based on Quranic, poetic, and contextless sentences, negators are associated with temporal “meanings”. Thus, the negator laa (لا) is said to “negate the event in the future” (Ibn Hisham, [42, vol. 1, p. 6). Other grammarians argue that laa may negate present states, as well (Al Muraadii, [13, p. 296]; leisaيس / negates “future and sometimes present events or states” (Al Istiraabaadi, [9, p. 197]; lammaa (لمّا) is used “to negate past events related to the present time” (Sibaweihi, [54, vol. 4, p. 223); lam / لم affects verbs in the imperfective and puts them in the past and the jussive mood Ibn Al Sarraaj, [41, vol. 2, p. 157); and lan /لن is defined as “a particle of negation, futurity and the accusative [mood]” (Ibn Hisham, [42, vol. 1, p. 464]. It is noteworthy that Ibn Hisham [42] and other grammarians, like Ibn ʻusfuur (in Al Muraadii, [13, p.274], disagrees with Al- Zamakhshari’s [24, p.407] claim that lan conveys corroboration and perpetuity of negation; maa / ما is used to negate present states [53, vol. 5, p. 24]. When it collocates with the so-called expletive min / من, it is said to corroborate negation [41, p. 374]. Finally, the archaic negator ‘in (إن) works in nominal and verbal past and imperfective sentences to denote a present temporal value. It is defined as synonymous to and interchangeable with the negator maa / ما ! [12, vol. 1, p. 188].

This temporal approach is approximated by Al Mabkhout [18, p. 119] in the following visualization (Fig. 1):

Fig. 1.
figure 1

Temporal values assigned to negators

Full size image

3 Contemporary Research on Negation in SA

The unchallenged prevalence of traditional grammatical views in current pedagogical grammar is indicative of the severe limitations of contemporary linguistic research on SA. The direct assignment of a chronological ‘meaning’ to formal categories without any referential value in the extralinguistic, such as lam, lan, maa, etc., is largely detectable in contemporary views, from different theoretical frameworks, on negation. Al-Makhzumi [19], though he defined negation as “a linguistic style governed by the contexts of speech” [19, p. 244], did not seem to take the context factors into consideration by adhering to the traditional temporality of negators. Amaira’s definition pertinently associates negation with the speaker’s intentions [25, p. 154] but his analysis of negators reiterates the same chronological treatment. Hamasa [38] approaches negation as a category “extrinsic to the structure of the sentence. It denotes the non-validity of the predicative relation in verbal and nominal sentences” [38, p.280]. His approach does not break with the predominant views on negation; it reduces the working of negators to distinctions of tense [38, p. 285-301]. By adopting a pragmatic approach, Al-Mabkhout [18] distances himself from the predominant grammatical orientation. First, he starts from negation not from negators and considers that the non-referentiality – literally “ the referential emptiness” [18, p. 485] - which specifies negation relates it to the categories of expressive language acts (‘al ‘ifsaaḥiyaat الافصاحيات.) [18, p. 485]; thus, a negative sentence does not necessarily presuppose an affirmative one [18, p. 451]. Second, he assumes that the non-referentiality of negation presupposes a complex structure made of two components; the first expressing negation and the second conveying its specification [18, p. 421], as exemplified in:

figure a

According to Al Mabkhout, this binary structure, reminiscent of the structures of the conditional, oath, and the vocative in SA, is based on a semantic link [18, p. 421]. The second clause “is understandable only in the context of the negative one” [18, p. 423]. It assumes different context-dependent functions, such as resumption, corroboration, justification, specification, or restriction. However, this line of demarcation from predominant grammatical orientations, does not seem sharply drawn when it comes to the working of negators. “The differences between negators are basically temporal” [18, p. 484].

Negation in Standard and Dialectal Arabic has also received considerable theoretical attention in the different stages of generative linguistics, notably the Minimalist approach to Universal Grammar, Shlonsky [53]), Benmamoun [26], Ouhalla [51]. This theoretical framework has been associated with a formalist and typological approach which has long stressed the primacy of thought over its external realization in languages. Negation is therefore investigated within a parametric approach to the linguistic differences permitted by the human language faculty. Attention is mainly devoted to the underlying representations of negation, not to how negators work in real contexts of communication. The traditional Past/non-Past temporal distinction has resurfaced in Minimalist literature to account for the differences between negators in Arabic. Fassi-Fehri [37, p. 163] proposes that the negators laa, lam, lamma and lan should be treated as modal negatives. Section 6 demonstrates that the validity of this claim is restricted to two negators only.

4 The Metaoperational Framework: From Enunciation to Metaoperation

Negation in SA has not to date received any systematic analysis from a Metaoperational perspective that takes into consideration the utterance’s context of production and reception. The conceptual framework used in this study is based on the findings of the applications of the Metaoperational theory on different natural languages, such as in Adamczewski [2–6], and [8], Delmas [31], Delmas & Girard [32], Delechelle [30], Delmas, Adams, Deléchelle, Girard, Lancri & Naudé [33]. Santin-Guettier & Toupin [52], and Toupin [55]. Adamczewski’s theory has developed in the wake of the major anti-mentalist shift in linguistics from “langue” (collective intelligence) to “parole” (individual act of language) which dethroned the study of “langue” as a self-contained system. The shift is initiated essentially by Benveniste [27] and [28], a student of Saussure, who formulated the Enunciation theory. Benveniste claims that the Saussurean Structuralist paradigm, by reducing language study to an over-emphasis on language as a code, eliminated the speaking subject (utterer) and the relationship he/she maintains with his/her interlocutor (co-utterer); two parameters without which utterances cannot be properly decoded. The theory holds that “before enunciation, language is but the possibility of language. After enunciation language is realized in an instance of discourse which has its origins in the speaker” [28, p.80]. Benviniste maintains that every utterance bears on its surface permanent and variable formal traces of its utterer and his/her co-utterer. Such traces are the confirmation that subjectivity and inter-subjectivity are interdependent properties of language and language use [27]. The status of the speaking subject in discourse will later constitute the basis of Culioli’s Enunciative Operations theory [29] and Adamczewski’s Metaoperational grammar [4].

In his groundbreaking work on “Be + ing” in English, Adamczewski [3] asserts that the direct assignment of meaning to meaningless categories, such as the so-called progressive form, is the main reason for the failure of the descriptive approach to account for the working of language. For him, the linear utterance is the final product of a complex and multi-faceted phonological, lexical and syntactic process [7]. Utterances exhibit on their surface observable traces of an invisible activity and codify the mental operations whose main object is not to enable the speaker to refer to the world, but to indicate how the utterance was processed in a given context, as well as the speaker’s position relative to both the propositional content and the co-utterer (co-U).

The significance of surface binary operators, such as (Ø/ Be + ing), (V-s/do), (nearly/almost), (shall/will), (may/can), (this/ that), (too/ also), etc., to cite just a few English oppositions, is that they constitute a natural metalanguage indicative of the working of language itself, hence their metalinguistic status. Adamczewski [8] refers to them as real “portholes” to the underlying language activity. For him, these operation tracers constitute the real subject of languages study. Most, perhaps all, grammatical phenomena are organized in pairs based on the Rhematic (phase 1 / open paradigm) Thematic (phase 2 / closed paradigm) vector. According to Adamczewski [7], “[this] basic principle is repeated cyclically to create different grammatical tools that are necessary to the working of languages”. Contrastivity as a systematic intra- and inter-lingual analysis of authentic data collected from languages is a methodological prerequisite.

The following section is restricted to six formal Arabic negators working in verbal and/or nominal utterances: lam, maa, lammaa, leisa, laa, and lan.

5 A Metaoperational Analysis of Negation in SA

Lam, maa, leisa, lammaa, laa, and lan constitute the nucleus of the Arabic negation system and behave as a micro-system governed by inter-related binary oppositions. These oppositions are not only intra-operational, i.e. within negation, but also inter-operational, i.e. in symmetry with their functional correspondents in affirmation. Consequently, intra- and inter-contrastivity is the approach adopted to investigate the working of the following pairs: (lam vs. maa), (lam vs. lammaa), (leisa vs. maa,) and (laa vs. lan).

5.1 Lam vs. Maa : Intra- and Inter-Operation Analysis

5.1.1 Corpus

  1. (1)

    Hiina wasaltu ‘ilaa beitina, ‘aḥsastu shei’an gheira `aadiyyin, fa-khashiitu qaliilan `indamaa lam ‘ara ‘ummii tusri`u lifatḥi-l-baabi.

    When I arrived home, a feeling of apprehension came over me. Thus, I was slightly worried when I did not see my mother rushing to open me the door.

  2. (2)

    Laa! lam yugheiyyir mawqifahu, wa laakinnahu bada’a -l-’aana yaftaḥu ‘udhuneihi.

    No! he did not change his attitude, but he now started to lend attentive ears.

  3. (3)

    lam yakdhib wa maa kaana kaadhiban `aleihaa.

    He did not lie and he would not.

  4. (4)

    Laakin man huwwa-l-jaani-l-ḥaqiiqii -l-ladhii ḥawwala ḥayaata raaniya ‘laa jaḥiimin wa hiya ṭiflatun lam tata`addaa-s-sanawaati-l-’araba`a?

    But, who is the real criminal who made the life of Raniya, a child who did not exceeded four years, a living hell?!

  5. (5)

    Hiyya tadda`ii bi’annahaa tagheiyarat wa maa tagheiyarat bal izdaadat ta`aṣṣuban

    She claims she has changed; she did not! She has just become more intolerant.

  6. (6)

    ‘a lam tara-s-maka `alaa qaa’imati-l-intiḑhaari ?

    Didn’t you see your name on the waiting list?

  7. (7)

    ‘amaa ra’eita-ḑaw’a -l-’aḥmara ?

    Haven’t you seen the red light on?

  8. (8)

    He should not have believed such nonsense.

5.1.2 Intra-Operation Analysis

  1. (1)

    Hiina wasaltu ‘ilaa beitina, ‘aḥsastu shei’an gheira `aadiyyin, fa-khashiitu qaliilan `indamaa lam ‘ara

    ‘ummii tusri`u lifatḥi-l-baabi.

    When I arrived home, a feeling of apprehension came over me. Thus, I was slightly worried when I did not see my mother rushing to open me the door.

As seen in sections two and three above, grammarians agree that lam is a verbal negator; though it affects an imperfective verb, it conveys negation in the past. This means that the verb is morphologically in the present (muḍhaariʻ) and grammatically in the past. However, negation in the context of the past is also conveyed by maa. The affirmative sentence (ra’eitu ‘ummii / I + see + past + my mother) has two possible negative realizations in Arabic:

  1. i.

    lam ‘ara ‘ummii / I did not see my mother.

  2. ii.

    maa ra’eitu ‘ummii / I did/have not see(n) my mother.

Lam and maa share two grammatical features: they negate and both are used in the context of the past. Yet, they are not interchangeable as they are produced in two different contexts. In (1) for instance, the use of maa would be ungrammatical. The temporal adverb ʻindamaa / when announces an open paradigm “when I did not see [my father/ sister/ mother (√), etc.]”, i.e. a new piece of information selected by the linguistic subject from a set of choices in a specific context. The fact that the predicative relation (R) is posed and not presupposed invalidates the possibility of maa due to its presupposing properties. In Adamczewski’s terms [4], compared with maa, lam is a phase 1 negator. It has a rhematic status, encodes the non-validity of the predicative relation in the context of the past, and is governed by a speaker’s informative strategy. A possible context for maa in (ii) is when the utterer does not negate to address an informational deficit but to deny a claim or refute a wrong view held by the co-utterer.

(2), (3), (4) and (6) also include negation with lam and provide further clarifications about its working in discourse. The context of (2) لا! لم يغير موقفه، ولكنه بدأ الآن يفتح أذنيه) / No! he did not change his attitude, but he now started to lend attentive ears.) includes a questioner seeking information. The answer therefore is reducible into a negative short form.

figure l

And did he change his attitude towards his colleagues after that?

  • لا !/ laa / No! (He did not).

The non-validation of the predicative relation in the past reflects a detached strategy of the linguistic subject. Whereas the use of maa, generally triggered by an intervenient strategy of the speaker, would convey a judgment, not a piece of information. (3) is an interesting utterance as it includes both lam and maa in the context of the past. The first clause introduces a neutral negative statement where the linguistic encodes the extralinguistic, while in the maa-clause the linguistic takes the lead over the extralinguistic; the linguistic subject (the speaker), in the light of his previous knowledge of (He), the grammatical subject, categorically excludes the possibility of the predicative relation (X lying to Z in the past). The filter of the speaker’s judgement obstructs the utterance’s direct reference to the extralinguistic event; hence the metalinguistic function of maa which works not to inform about the non-validity of R, but to codify a processing strategy of the speaker. The shift from did to the modal would in the English translation provides a revealing insight into the working of both negators in Arabic. In (4), the phase-one negator lam is associated with an additional information about the grammatical subject (هي / hiyya / she). Maa is possible only in a context of denial including a speaker rejecting and rectifying the claim that the girl is over the age of four. (6) includes a neutral interro-negation with lam (Did you not see your name on the waiting list?) awaiting a yes/no response from the co-utterer. The same sentence with maa would presuppose a different presumptive context, such as when the utterer does not understand how that can be, and therefore awaits explanation, not information.

(5), (7) and (8) shed more light on the working of the phase-two negator maa. (5) is an interesting compound sentence. The negative clause (ما تغيّرت / She did not!) is preceded and governed by the affirmative clause (هي تدّعي بأنها تغيّرت / She claims she has changed.) which conveys the utterer’s judgment that ‘her change is just pretense’. The verb (ادّعى / ‘iddaʻaa / claim) is the trigger of the Arabic modal operator أن / ‘anna / which has a corroborative value comparable to the emphatic do in English. The speaker’s conviction that change is a mere claim is the co-textual factor which justifies the use of maa instead of lam. (7) is an interro-negative utterance that introduces a doubtful speaker who finds it hard to believe the co-utterer’s claim. An interro-negation with lam would be possible only if the questioner was seeking information, such as in (6) above. Therefore, what is negated in (7) and (5) is not the extralinguistic event itself, but the co-utterer’s claim that R did not take place. In both cases, maa functions as a meta-negator codifying a metalinguistic activity. The translation of (8) into a sentence with a negated modal past sheds more light on the status of the grammatical subjectهو) /He) as an object of discourse, not an autonomous agent. The utterer considers that it was wrong for X to believe such nonsense. The direct object (such nonsense) serves as a justification of the judgement passed by the major player in the utterance, i.e. the linguistic subject. It is noteworthy that the triggers of any phase-two meta-negator differ from one utterance to another according to the enunciative strategy at work in the context of production and reception alone.

5.1.3 Inter-Operation Analysis

In the previous section, the approach to lam and maa is intra-operational, i.e. the two negators are contrasted as different realizations of the same operation. The present section explores negation (NEG) in relation to affirmation (AFF), the polar opposite of negation. Such extension is expected to provide further elucidation on the status and the working of the two negators in Arabic. However, this inter-operational contrastivity should on no account lead to considering lam and maa, or any other negators, as the negative duplicates of the affirmative markers they are contrasted with. In fact, each linguistic utterance is governed and justified by its unique context of production and reception. In the present case, the NEG-AFF opposition is rooted in the assumption that the symmetry detected is not between affirmative and negative markers, but between the two phases in both poles.

What is said about the strategies and the role of the linguistic subject in negation is valid for affirmative utterances too. Consider (9) and (10) below:

  1. (9)

    هي تشتري ملابسها من محلات هارودز.

    hiyya tashtarii malaabisahaa min maḥallaati Harrods

    She buys her clothes at Harrods.

  2. (10)

    إنّها تشتري ملابسها من محلاّت هارودز.

    ‘innahaa tashtarii malaabisahaa min maḥallaati Harrods

    She’s buying her clothes at Harrods.

If the difference between the two utterances is particularly noticeable in the absence of ‘inna (Ø) in (9) and its presence in (10), what really distinguishes this grammatical pair is their context of production and their structures. A possible context of (9) is the co-utterer’s question (Where does she buy her clothes from?). A short answer (From Harrods) is expected given the most important piece of information it includes. The marker (Ø) encodes the non-intervenient speaker strategy where the linguistic subject acts just as information provider, whereas in (10) the utterer endorses the predicative relation {S-P}:{hiyya – tashtarii malaabisahaa min maḥallaati Harrods}

{she – buys her clothes at Harrods}

In fact, the context of (10) includes not an information seeker, but a co-utterer expressing ‘amazement at the great elegance of the lady’. Therefore, if (9) informs about the place where (she) buys her clothes, the purpose of (10) is completely different: the speaker talks about the grammatical subject (she) and assigns her the property (buying-clothes-at-Harrod’s), a phrase nominalized by the operator (-ing) and in which (at Harrods) is not a new piece of information. A phase-two metaoperator encoding a justification strategy of the utterer, ‘inna has a metalinguistic status, works outside the predicative relation, obstructs reference to the extralinguistic, puts the grammatical subject in the accusative and announces that the strategy engaged by the utterer is the key to understand and therefore to translate the utterance. As shown in a previous work [46, p. 181] and [48, p. 224], the working of ‘inna in Arabic offers a comfortable validation of this analysis; in fact, its emergence in discourse automatically puts the grammatical subject in the accusative.

The grammatical subject is stripped of its original mark of agentivity (the nominative) to bear the accusative case which marks grammatical objects in Arabic. In English, the metaoperator (Be + ing) is the effective solution to render (10). Thus, the major difference between (9) and (10) is related neither to tense nor to aspect, but to the two phases at work in discourse. This leads to the provisional conclusion that the two phases define the line of symmetry between affirmation and negation markers in SA.

5.2 Lam Vs. Lammaa : Intra- and Inter-Operation Analysis

5.2.1 Corpus

  1. (11)

    استوفيت الكتاب ولمّا يتنفّس الفجر. / Istawfeitu-l-kitaaba wa lammaa yatanaffasi-l-fajru

    Dawn had not come up yet when I finished the book.

  2. (12)

    . لا يزال يذكر حينما انتزعوه من أمّه ولمّا يبلغ الخامسة من عمره

    Laa yazaalu uḏkuru ḥiinamaa intazaʻuuhu min ‘ummihi

    He can still recall when they snatched him away from his mother, he not yet having reached five years old.

5.2.2 Analysis

Lam and lammaa are approached in opposition because they share the following properties: they are verbal negators, they affect a verb morphologically in the mudhaariʻ and grammatically denoting the past, they work to fill an information deficit, and both have a phase-one status. However, if lam, as seen above, encodes the non-validity of the predicative relation in the context of the past and introduces a closed relation {R}, i.e. a finished act; lamma, often an inter-verbal negator, as in (13):

  1. (13)

    / حلّ الضيف ولمّا يصل المضيّف! halla-ḍ-ḍaifu wa lammaa yaṣil almuḍhayyifu

    The guest was here but the host had not arrived yet!

emerges in contexts always implying a probable prospective validation of the predicative relation. An aspectual negator, lammaa denotes a verbal event which has not occurred up to the time of speech but is likely to happen in the future. The predicative relation is therefore awaiting realization {R} and this is what explains why lamma, unlike lam, is incompatible with conditional markers, such as إن / ‘in and إذا / ‘iḏaa. The aspectual behavior of lammaa is detectable in (11) and (12) which include an implied aspectual adverb (baʻdu/yet). Whether explicit or not, this adverb is inherent in any utterance with lammaa. In the English translation, a negated present perfect is often the most viable solution. In fact, the adverb (baʻdu/yet) belongs to a class of negative polarity items conveyors of an aspectual value and associated with the past, present and future times:

  • lam … qaṭtu (lam +past time-oriented never)

  • lammaa… baʻdu (lammaa + not yet)

  • lan…’abadan (lan +future time-oriented never)

Qaṭtu and ‘abadan convey maximization values in the past and the future, respectively.

5.2.3 Inter-Operation Analysis

Traditional grammarians, such as [52, vol. 3 p. 117] and [41, vol. 1, p. 458], have already detected a bilateral symmetry between the negator lammaa and the affirmative verbal marker qad when it affects a verb in the past. Yet, the opposition should include lammaa and the aspectual laqad, instead of the epistemic modal qad. The latter always affects either verbs in the past to encode a high degree of speaker certainty, or with verbs in the mudhaariʻ (imperfective) to encode uncertainty or doubt. Therefore, like the affirmative ‘inna, qad is a speaker visibility marker in discourse and assumes a modal not an aspectual function. Laqad, however, indicates completeness of the verbal event in the past and that is why an English perfective, often with already, emerges in translation, like in (13), (14), and (15):

  1. (13)

    Sayidii, laqad waṣala ḍhaabiṭu-sh-shurṭati wahwa-l-’aana fi-l-mukhtabari

    Sir, the police officer has (already) arrived. He is now in the lab.

  2. (14)

    عفوا ! لقد نفد رصيدكم./ ʻafwaa laqa nafada raṣiidukum

    Sorry, you have exhausted your balance.

  3. (15)

    لقد أذاعوا البيان العسكري للتّو. / Laqad ‘adhaaʻu-l-bayaana li-t-tawwi

    They have just broadcast the military communiqué.

Intralingually, the aspectual behavior of lammaa is in symmetry with that of the affirmative laqad. It is noteworthy that aspect and modality in Arabic have been under-researched in traditional and contemporary research, and are entirely absent from pedagogical grammar. Viable Arabic equivalents for the terms aspect and modality and their derivatives are still to be coined [47]. Didactically, this has not been without adverse implications on translation students who encounter difficulties in negotiating effective translation to the English aspectual and modal markers [1, 16, 20, 35, 36, 50, and 58].

5.3 Leisa vs. Ma : Intra- and Inter-Operational Analysis

5.3.1 Corpus

  1. (13)

    Not everything you read in history books is true.

  2. (14)

    ا تنتظره منها ليس بالأمر الهيّن. / maa tantaḍhiruhu minhaa leisa bi-l’amri-lhaiyyini

    What you expect from her is not a trivial matter.

  3. (15)

    ما كل ما يتمنى المرء يدركه. / maa kullu ma yatamanna-l-mar’u yudrikuhu

    Nothing ever happens exactly as one would wish.

  4. (16)

    ما هذا بشرا. / maa haḏaa basharan/

    This is not a human being! (This is not other than a noble angel!)

  5. (17)

    ما أنا بمهمل لواجباتي . / maa ‘anaa bi-muhmilin li-waajibaatii

    I am not careless about my duties.

5.3.2 Analysis

Traditional grammarians disagreed about leisa as a grammatical category. Some, like Ibn Al-Sarraj (in [15, vol. 2, p. 73], argue that it is a particle, while for Sibaweihi and his followers it is a verb. In Al-Horais [17], leisa is categorized as a verb, a particle and an adverb. Using a corpus of artificial sentences, Al-Horais incorrectly considers that leisa “can be replaced by maa kaana” [17, p. 10]. Leisa and maa, as explained below, have different statuses and are opposable, but not interchangeable.

Leisa is a partly conjugable negator that works in three different structural contexts: as an auxiliary verb in verbal sentences, like in (18) below, a lexical verb in nominal sentences, such as in (19), and as an operator in stripping constructions (20).

  1. (18)

    ليس يعلم ما تخبّئه له الأيّام. / leisa yaʻlamu maa tukhabbi’uhu lahu-l-’ayyaamu.

    He doesn’t know what the future holds for him.

  2. (19)

    هذه معضلة أخرى ليس لها تفسير. / haaḏihi muʻḍilatun ‘ukhraa laisa lahaa tafseerun.

    This is another inexplicable dilemma.

  3. (20)

    هي تنتظر منك اعتذارا وليس تبريرا. / Hiyya tantaḍhiru iʻtiḏaaran wa leisa tabriiran

    She is expecting an apology, not a justification.

Compared with the negator maa approached in 6.1 above, leisa is not associated with an intervenient speaker strategy. Whenever used, it denotes the non-occurrence of the predicative relation, hence its referential or world-oriented, not metalinguistic, value. In (13) and (14), leisa is a phase-one inherently negative copula (be + not) associated with the time of discourse in (14), and with a timeless event in (13). The examples (15–17), which include negation with maa, are also possible with leisa, but with different contextual triggers and meanings. As seen above, maa is an anaphoric negator that presupposes an antecedent contextual element, whereas leisa poses and negates a new relationship anchored in the time of discourse. (16) is an interesting example as it is possible with the two negators, maa and leisa. Embarrassed by the structural similarities between the two sentences, Sibaweihi claims that maa in (16) and leisa are interchangeable [21]. (16) is in fact a case of what Ducrot [34] and Horn [39] and [40] call metalinguistic negation, i.e. when negation is not a truth-functional statement. In this case we have “a formally negative utterance which is used to object to a previous utterance on any grounds whatever, including the way it was pronounced” [40, p. 374]. This is confirmed by the context of the Quranic narrative about prophet Joseph; when he was introduced to the Egyptian women - who are the linguistic subject in (16) - they were too mesmerized by his unearthly beauty to believe he was an ordinary man. Leisa, a world-oriented negator, would require a totally different context:

  1. (20)

    أنظري إلى عينيه الغائرتين وجبهته النّاتئة. لا لا! ليس هذا بشرا. لابدّ أنّه مخلوق فضائي.

    ‘unḏurii ‘ilaa ʻainaihi-l-ghaa’irataini wa jabhatihi-n- naati’ati. Laa laa! leisa

    haḏaa basharan; Laa budda ‘annahu makhluuqun faḍaa’iyyun.

    Look at his receding eyes and swollen forehead! No, no! this is not a human being. This must be an extraterrestrial!

The importance of (17) is that the negator maa works with the preposition bi-:

maa ‘anaa bi-muhmilin li-waajibaatii / I am not careless about my duties.

This discontinuous phase-two negator [maa …bi…] has its correlative discontinuous marker in affirmation, [‘inna …la…], such as in:

  1. (21)

    إنّني لمهمل لواجباتي! / ‘inna-nii la-muhmilun li-waajibaatii

    Definitely, I am careless about my duties.

In traditional grammar, bi- in negation with maa or leisa is treated as a redundant preposition, and la- a corroborative particle. In fact, these two operators, which have no clear equivalents in English, constitute the predicative node and the scope of the phase-two markers maa and ‘inna respectively.

5.4 Laa vs. Lan : Intra- and Inter-Operation Analysis

5.4.1 Corpus

  1. (22)

    Laa taduuru-sh-shamsu ḥawla-l-’arḍhi لا تدور الشمس حول الأرض /

    The sun does not revolve around the earth.

  2. (23)

    أراك لا تبالي حتى بما يجري من حولك

    I see that you don’t even care what’s going on around you.

  3. (24)

    لم يستهينوا بك يا سيّدي، انّهم فقط لا يفهمون شيئا ممّا تقول.

    They are not disrespectful sir, they just haven’t understood anything of what you are saying.

  4. (25)

    لا يدخل الجنة قاطع رحم. / laa yadkhulu-l-jannata qaaṭiʻu raḥimin

    He who severs family bonds will not go into paradise.

  5. (26)

    .فلن أكلّم اليوم إنسيا / Fa-lan ‘ukallima-l-yawma ‘insiyyan

    Therefore, I shall not speak this day to any human being.

  6. (27)

    .ولا تمش في الأرض مرحا إنّك لن تخرق الأرض ولن تبلغ الجبال طولا

    And walk not in the land exultant, for you cannot cut through the earth nor stretch to the height of the mountains.

  7. (28)

    لن أدع ما حصل يفسد طموحي وآمالي…لن أفعل. لن أستسلم، سأقاوم!

    I won’t let what happened ruin my hopes… I shall not. I shall not surrender. I shall resist!

5.4.2 Analysis

If in verbal negation lan works always with a muḏaariʻ commonly associated with reference to futurity - ‘al ‘istiqbaalu / الاستقبال - laa is possible in four temporal contexts: in timeless events (22), in the present - ‘al-ḥaalu / الحال - (23 and 24), in the future (25), and not often in the past (30) when it joins alternatives:

  1. (30)

    فلا صدّق ولا صلّى. / fa-laa ṣaddaqa wa laa ṣallaa

    For he neither believed, nor prayed.

    In nominal negation, the use of laa is associated with the time of speech:

  2. (31)

    لا حاجة لي بكل هذه الأدوات لأفكّ برغيا!

    laa ḥaajata lii bi-kulli heḏihi-l-’adawaati li ‘afukka burghiyyan

    I don’t need all these tools to unscrew a screw.

Whatever its context of use, laa conveys a core grammatical value: it signals that the non-validity of R is a new piece of information not endorsed by the linguistic subject. A comparison with lan is expected to elucidate the working of both negators.

A yes/no question is a possible context for laa in (22); the questioner seeks information that is provided by the questioned. An answer with lan would be ungrammatical, even though well-formed. Lan is a modalizing negator typical of contexts conveying a guarantee of the linguistic subject that R will not take place. In (28), where it is used three times, lan is a phase-two negator which encodes the speaker’s sheer determination ‘to achieve her goals’, thus conveying a deontic modality. laa is not compatible with such a context. In English, the negator not and the modal marker are always discrete even when they are in a contracted form (mustn’t), but in Arabic the distinction is between modalizing and non-modalizing negators. This may induce translation students into confusion, especially in comparison with English. (32), (33) and (34) are pertinent examples of the interplay between negation and modality (M):

  1. (32)

    يجب أن لا يتخلّف عن الموعد. / yajibu ‘an laa ytakhallafa ʻani-l-mawʻidi

    / Must - not – he – miss the appointment /

  2. (33)

    لا يجب أن يتخلّف عن الموعد. / laa yajibu ‘an ytakhallafa ʻani-l-mawʻidi

    / Not – must – he – miss the appointment /

  3. (34)

    أعرفه جيّدا، إنّه لن يتخلّف عن الموعد مهما كانت العواقب.

    ‘aʻrifuhu jaiyyidan, ‘innahu lan ytakhallafa ʻani-l-mawʻidi

    I know him very well, he won’t be missing the appointment, no matter what happens.

The following configurations visualize the difference between (32) and (33) (Fig. 2):

Fig. 2.
figure 2

Interaction of negation and modality

Full size image

Even though the figure clearly shows the scope of the negator, R in (32) and M in (33), it does not totally eliminate semantic ambiguity resulting from the mobility of the negator and the modal; in fact 24 out of 25 translation students surveyed for this article provided the same English translation for both sentences (He must not miss this meeting.). Only one trainee suggested mustn’t for (32) and must not for (33).

Lan in (34) is possible with the same predicative relation (He / miss the meeting), but it is neither triggered by the same context of (32) and (33) nor does it convey the same grammatical value or translation. The presence of the epistemic modal marker ‘inna, which encodes certainty of U, in the context of lan, itself a modal negator, over-modalizes the utterance and complicates the task of the translator. 19 students out of 25 (76%) provided the same English translation to (34), an utterance with inna + lan, and (35) an utterance without ‘inna (Ø).

  1. (35)

    I know him very well, he will not be missing the appointment, no matter what happens.

The utterances were given separately for translation in two different tasks. The respondents who suggested different translations used a modal adverb, such as definitely, to render ‘inna. In (23), (24) and (25), Laa has the same phase-1 status and conveys the same referential value. The modal ‘inna in (24) takes scope over a predicative relation negated by laa. The context of (23), (24) and (25) are not permeable to lan. (26) is a very interesting example as it refutes the claim of temporality associated with negators. The time locator of R is the adverb ‘al-yawma / today, lan is there to convey a deontic modality, like in (28). (27) is slightly different as lan is an epistemic modal negator conveying certainty, not determination. A phase-1 negator, laa would be ungrammatical in the contexts of (26), (27) and (28).

6 Conclusion

The paper has proposed a metaoperational analysis of the major verbal and nominal negators in SA and has demonstrated that the metalinguistic richness of negation in SA, compared with the single formal negator (not) in English, is governed by an underlying binary microsystem (phase 1 phase 2 vector) that accounts for this diversity. Lam, maa, leisa, lammaa, laa, lan, and kallaa are in fact more effectively understood, taught, learned, and translated when they are approached from the perspective of the language user and the processing strategy at work in discourse. The dominant approach in theoretical and pedagogical grammar is too reductionist to account for the working of negators in SA. By direct assigning of meaning to formal markers, it simply confounds the linguistic with the extralinguistic and induces learners and translators into error and mistranslation. Yet, the diversity of negation markers in Arabic provides a highly significant case of operators which convey referential, metalinguistic, modal, and aspectual values. The interaction of operations, such as reference to time, aspect, and modality with negation is still to be investigated from a corpus-based explicative perspective. This is expected to provide valuable assistance especially to language learners, translation trainees and computational linguistics.

Tables 1 and 2 recapitulate the key findings related to the application of the binary microsystem underlying discourse:

Table 1. Intra-operation Contrastivity
Full size table
Table 2. Inter-operation Contrastivity
Full size table