Abstract
Alviero Niccacci proposed a Biblical Hebrew syntax based on Harald Weinrich’s text-linguistic framework. I argue Niccacci’s methodology deviates substantially from Weinrich’s proposition. This article critiques Niccacci’s theoretical assumptions about the oppositions between foreground and background and that of comment and narrative. While considering these issues in view of Weinrich’s method, I also suggest that the alternative source of inspiration for Niccacci was the work of Hans J. Polotsky.
Keywords
Biblical Hebrew scholars of the last three decades have been engaging quietly with the work of Harald Weinrich. He has pieced together a proposal which derives its main principles from advancements in philosophy, rhetoric, linguistics, and literary studies. Its results are gathered in a general blueprint of language in his Tempus and two applications to German and French. 1
One of the scholars who has worked with his methodology is Alviero Niccacci. 2 My intention in this article is to produce a critical evaluation of the methodology underlying Niccacci’s syntactical description of Biblical Hebrew (BH). Niccacci confesses that he decided to engage with the work of Weinrich after reading two review articles by Eep Talstra. The latter examines how the textbook of BH grammar written by W. Schneider applies the method of Weinrich. In this context, although he believes Schneider’s BH analysis to be correct, Niccacci also considers ‘Schneider’s application of the Weinrich model to be at an incomplete stage as yet’. 3
Be that as it may, it is worth mentioning that Schneider (in one section of his BH teaching manual) shapes the way in which key concepts of Weinrich are later understood by other scholars, including Niccacci and Talstra. Schneider’s outline contains the seeds of major breaks from Weinrich’s blueprint. Stated explicitly or deduced from the context, one observes that for Schneider: (1) Weinrich’s comment/narrative opposition becomes that of direct/indirect speech 4 and (2) foreground/background is assimilated to the physical spaces within the narrative—while some events are centre stage, others are in the background. 5 These are the main two areas that my contribution will clarify, with regards to how they develop in Niccacci’s work.
We shall now show that because he puts the methodology of Hans J. Polotsky and Weinrich together, Niccacci’s overall stance is no longer retraceable to Weinrich’s principles. Effectively, the linguistic labels apparently common between Weinrich and Niccacci do not reflect the same theoretical reality.
As most of Weinrich’s principles are interpreted differently from how he described them, Niccacci’s proposal, and BH syntax in general, needs serious critical discussion from two perspectives. First, one should clarify the points where Niccacci’s method has taken a different route from Weinrich’s and what the consequences for BH syntax are. This article will look into these issues. Second, my subsequent article explores a proposal regarding how Weinrich’s principles should be employed in the analysis of BH syntax.
This article begins with a description of Weinrich’s method: the general axiom is that tense does not mean time, and that there are three dimensions of language. The result of Weinrich’s proposal is that each tense has a determinate function—there is a fixed relation between tense (as a linguistic sign) and its syntactic function. I briefly touch on the issue at the end of this first part. The second section contains an outline of Niccacci’s main theoretical points about language and how they come to contradict Weinrich’s original stances. In addition, it will be evident that the work of Niccacci is not aware of the abovementioned connection between linguistic sign and linguistic function in the methodology of Weinrich.
Before coming to Weinrich’s method, I briefly mention two other text-linguistic schools which I will not discuss at length in this article. Eep Talstra’s work and that of his followers analyse BH syntax by using a combination of text-linguistic and distributional analysis. This research takes place now in the Eep Talstra Centre for Bible and Computer (Vrije Universiteit Amsterdam). 6 Wolfgang Richter and his students derived their methodology from the advancements of the text-linguistics of L. Tesnière and W. Raible (the latter is one of Weinrich’s pupils). 7 Computer-assisted analysis of the text is the other common trait of these two schools. Each school has its own databases containing the analysis of the Hebrew Bible (see the SHEBANQ project and the Biblia Hebraica transcripta project [BHt], respectively).
1. Main tenets of Harald Weinrich’s syntactical method
Weinrich outlines the main principles of his Tempus in a brief article in English, ‘Tense and Time’. Acquaintance with Weinrich’s core methodology is a prerequisite for any discussion of Niccacci’s work and for similar proposals relating to biblical languages.
(i) it does not touch me immediately and above all, it does not impose on me the need for an immediate action or reaction. [. . .] Thus, one could say that the attitude of the narrator and his listener is relaxed. The speech situation is that of relaxation.
10
By contrast, the comment 11 register represents those communications which ‘concern me directly’, they imply ‘stress’ and ‘tension’ or ‘as listeners, we have to adopt the non-relaxed attitude which corresponds to a situation that touches upon our own existence’. In short, ‘[t]hese two different categories of speech situation are like different levels of alarm in the discourse’. 12 Although there is no clear systematisation of English tenses in his Tempus, I infer from Weinrich’s overall presentation that the narrative tenses are past simple (preterite), past continuous (imperfect), past perfect (pluperfect) and conditional; the comment tenses are present, present perfect, and future tenses.
(ii)
(iii)
As a word of caution, the terms ‘direct’ and ‘indirect speech’ are not the same as Weinrich’s linguistic attitude (register) division of ‘narrative’ and ‘comment’. I discuss this issue further in the critical examination of Niccacci’s work below.
A less obvious element in Weinrich’s Tempus is that there is a fixed relation between each tense (as linguistic sign) and its function. There is one tense of narrative for zero degree of foreground (English past tense or the Italian passato remoto) and another for the background (past continuous or the Italian imperfetto); comment and narrative have different tenses for the zero-degree foreground (present tense or the Italian presente) and background (while English utilises the present continuous, Romance languages do not have one single, specific tense but rather a construction which in Italian, for example, is ‘sta leggendo’). In this context, background does not mean anticipation and/or retrospection: instead, the narrative/comment retrospection and narrative/comment anticipation also have their own tenses. This connection between tense and function is an integral part of Weinrich’s linguistic description (see the conclusion of this article). In discussing particular languages, he never assigns more than one tense to one function without explaining the reasons for this deviation within the confines of his method. 18
Possible combinations of Weinrich’s dimensions and the Italian and English tenses reflecting them. 19
2. Niccacci’s Biblical Hebrew syntax and its methodological guidelines
In this section, I clarify the core of Niccacci’s methodological lines. His syntax is a text-linguistic one insofar as he examines extensive stretches of text and, for the most part, follows Weinrich’s vocabulary. Niccacci’s syntax is a functional syntax because he uses the well-known key terms of theme/rheme (belonging to the Prague School). 20 Speaking in the most general terms possible, one of the aims of functional syntax is to describe how various elements within a sentence contribute differently to transmitting information according to their given or new quality. As already stated earlier, my evaluation of Niccacci’s method explores the following:
his understanding of foreground/background opposition;
his view that the pair of comment/narrative means direct/indirect speech;
Hopefully, it will be evident that Niccacci has not been particularly meticulous in his application of Weinrich’s method. One needs to retain a certain flexibility and adapt the syntactical method in order to make it work with specific languages. However, because of the changes made to the core of the method by Niccacci, one cannot argue that his syntax benefits from the correspondences and mechanisms Weinrich established for modern languages.
2.1. The foreground/background opposition
Niccacci’s theory of foreground–background opposition is intrinsically connected with that of Polotsky’s reading of predicate/subject. Niccacci supposes that BH sentences with the word order predicate–subject are foreground, while those displaying subject–predicate are background. I discuss here the meaning of foreground/background for Niccacci (2.1.1), how it is connected with the work of Polotsky (2.1.2), followed by a short evaluation of Niccacci’s work on this particular matter (2.1.3). Furthermore, I expound on the meaning of Weinrich’s view of foreground–background (2.1.4).
Before delving into the first discussion, one should note that at least two other linguists in addition to Weinrich propose a foreground/background opposition. Paul Hopper thinks that foreground is that narrative form which contains the ‘story line’ or the ‘skeletal structure of the discourse’ as opposed to background, which does not. 21 Tanya Reinhart believes the distribution of tenses has something to do with the way in which one perceives narrative space: some items capture more attention and are in the foreground stage, others are less visible and thus take the background stage. 22 While these two positions are subsequent to and do not draw on Weinrich’s Tempus, they are, however, more familiar to scholars working in discourse analysis. Weinrich is closer to Hopper because both connect foreground to the plot, and less similar to Reinhart, because, for Weinrich, space is a category similar to time.
2.1.1. Foreground/background in Niccacci’s syntax
Niccacci explains with sufficient clarity the wayyiqtol and its narrative status of foreground. By contrast, the less clear item is that of background. In order to understand what this opposition means for him, one needs to note: (a) what background is and what BH sentence types mark it; (b) because Niccacci includes the recovered/anticipated information in the background, 23 one should also indicate which types of BH sentences convey recovered information.
With regards to point (a), Niccacci argues that
the term background is represented by the following types of sentences: waw-x-qatal (p. 31); waw-x-yiqtol (p. 33); wqatal (§15, p 35); the simple nominal clause (§43, p. 65); 24
the meanings attached to background are: ‘a comment on the main event’ (p. 35); ‘peripheral information’; background interrupts the wayyiqtol sequence of foreground whenever one of the four sentence-types expresses: antecedent information (§41), simultaneity (xqatal, §42), contrast (§43), simultaneity/contemporaneity (with simple nominal clause—SNC, §44), a repeated action (§46), or emphasis (§48). Niccacci lists them once again under §49 and asserts: ‘From the aspect of text linguistics every construction which breaks the narrative chain belongs to the background’.
Regarding (b), xqatal and xyiqtol while reflecting recovered and anticipated information respectively, also represent background. 25 In the special case of xyiqtol of discourse (future axis), anticipated information is associated with foreground. 26
2.1.2. Polotsky: non–verbal predicate and non–nominal subject
I begin this subsection with a quote from Niccacci which exemplifies the deep connection his theory and that of Polotsky share. More importantly, it mirrors the confusion created by the binding of the predicate/subject categories to places in the sentence (first/second position), instead of their normal morphological counterparts (verbs/nouns). Niccacci asserts that, First, in Hebrew the first position in the sentence is filled by the predicate, not by the subject. Second, [while it is obvious that the predicate–subject word order is active when a verb occupies the first position in the sentence,] when a Hebrew sentence begins with a noun or an adverb the predicate is not identical with the verb [in second position] but in actual fact with that noun or adverb. Accordingly, what is normally the ‘subject’ becomes the ‘predicate’ and vice versa.
27
Niccacci adds that ‘This transformation is not exclusive to Hebrew as it occurs in other languages, ancient and modern’ and then refers, in note 18, to the examples presented by Polotsky in ‘Les transpositions du verbe en égyptien classique’, §2.5. 28 In order to make the connection between the two authors clearer, I shall discuss the relevant content of this article.
Polotsky argues that ‘in Egyptian, the function of predicative, far from being the prerogative of the verb, is shared equitably among the verb and the substantive, adjective, and the adverb’. Moreover, ‘the verb needs to be able to be expressed in a substantival, adjectival, and adverbial way’, a process called, succinctly, ‘transposition’ or even ‘translation’. 29 While not referring directly to the Aristotelian pair of substance (nouns or ‘particulars’) and property (verb), 30 Polotsky argues for a similar binary division which at the morphological level is noun (which includes the adjective and the adverb) and verb and at that of syntax, subject–predicate.
How does Polotsky explain the possibility that morphological values other than the verb may be a predicate? He suggests that ‘the non-verbal item which is to be emphasised is elevated to the rank of predicate, while the verb is degraded to the rank of subject’. 31 In order to accomplish this reversal, according to Polotsky, Egyptian grammar combines two of its resources: ‘(1) the nominalised verbal forms which turn the verb [into a noun] are already considered a non-predicative form; (2) the different syntactic moulds aim [to create] phrases with non-verbal predicates, with the occurrence of nominals or adverbials (circumstantial)’; that is, I clarify, in the first predicative position of the sentence. He argues that the emphasis is thus either ‘nominal’ or ‘adverbial’ (the latter being of a ‘circumstantial’ type). As a result, there are two types of sentences with non-verbal predicates: 32
those with a nominal predicate (in Middle Egyptian: predicate–the particle pw–subject); the predicate is a nominal construction, while the verb is subject;
and those with an adverbial predicate (in Middle Egyptian: subject–predicate) the adverbial emphasis is the predicate, while the verb in its substantival form 33 is the subject.
Polotsky considers these sentences with non-verbal predicates as emphatic constructions of the ‘cleft sentence’, or in French, ‘phrase coupée’. He believes that in Middle Egyptian the reading of these types of sentences, with non-verbal predicates as cleft sentences, is hindered by ‘peripheral elements’ which hide the two essential elements: ‘(1) the verb is nominalised and (2) a predicative nexus is established between the nominalised verb and the emphasis [in French, “la vedette”]’. 34 This means that these constructions reverse the nexus between the verbal predicate and its subject to that of a nominal predicate followed by a verb as subject.
For his part, Niccacci argues that BH displays three types of sentences: the verbal sentence ‘with a finite verb in 1st position’ (cf. wayyiqtol); the compound (later called also ‘complex’) nominal sentence ‘with a finite verb form in 2nd position’; and the nominal sentence ‘without a finite verb form’. 35
Focusing on the last two types—nominal sentences are sentences either with verb–second, the x–verb sentence (the complex nominal clauses, CNC: xqatal and xyiqtol) or with no verb (the simple nominal sentences, SNC). Considering together Niccacci’s definitions of nominal sentences (CNC and SNC) with that of Polotsky above (nominal predicate sentence and adverbial predicate sentence), it is obvious that Niccacci assumes that the CNC is similar to Polotsky’s sentences with a nominal predicate (which does retain an actual verb in the second position), while the SNC is equivalent to the adverbial predicative sentence (which has only a ‘substantival’ verb).
Taking this into consideration, Niccacci’s affirmations, quoted at the beginning of this section (‘when a Hebrew sentence begins with a noun or an adverb the predicate is not identical with the verb (in second position) but in actual fact with that noun or adverb’), now make more sense. Niccacci means that first position of a noun or adverb in nominal sentences elevates these morphological forms to the state of predicate, one would assume by the same process of transposition or translation.
However, in contrast with Polotsky, who explains how a non-verbal morphological value may be the predicate in Middle Egyptian and Coptic (there is a type of transposition, as explained above), I found no explanation by Niccacci of how BH morphology allows for a noun to be the syntactical predicate and for a verb to be the syntactical subject. Indeed, one finds no such thing, as Niccacci does not understand predicate and subject as ‘syntactical’ values which need that morphological correlation; he reads them as functional values which do not need it. 36 Effectively, Niccacci uses (Polotsky’s) syntactic labels (predicate and subject) to describe functional phenomena (theme and rheme). I discuss more fully the functional interpretation of these two values in the next section on Niccacci’s predicate/subject (see section 2.1.3).
To this point, I have shown how Niccacci’s nominal sentences (SNC and CNC) reflect linguistic structures defined by Polotsky. I now turn to the question: what is the broader framework that connects Polotsky to Niccacci at the level of syntax? Recall that, according to Polotsky, the adjective, the substantive, and the adverb fall under the main category of noun in contrast with verb. However, the adjective, the substantive, and the adverb become the grammatical verb in certain conditions: (1) the adjective acts as a participial sentence, (2) the substantive is part of special ‘emphatic’ constructions (sḏm.f. and sḏm.n.f) or of an infinitival sentence, and (3) the adverb acts as a type of circumstance, or is part of a stative or an infinitival sentence. 37 This is why Polotsky is able to suppose that they are predicates 38 at the syntactic level.
In this context, it is natural for Polotsky to suppose that these (generally called) nouns, while acting in the changed morphological function of verb, are effectively working as proper verbs and, syntactically, as predicates. This is the first element which makes it possible for the general category of noun to act as predicate in Polotsky’s work.
A further element contributing to this is the binding of the status of predicate to the initial position in the sentence. As shown in Niccacci’s examples from Egyptian, the difference between First and Second Tenses, according to Polotsky, is not a morphological difference but one created by word order: First Tense sentences have a verb in the initial position, and Second Tense sentences have a noun in the initial position. 39 In other words, the initial position of the sentence bears a specific ‘emphasis’ which falls, in some cases, on to a predicate represented by a verb or, in other cases, on to a predicate represented by a noun (adjective, substantive, or adverb). Niccacci explains this further: ‘while emphasis is the main (and more frequent) function of the second tenses, it is not the only one. What is common to all the occurrences is that the verb is not the main element of the sentence. It is nominalized morphologically and has the function of a substantive’. 40
Drawing on this theory, Niccacci proposes that the verb’s first position in the sentence influences the way in which a BH sentence is divided: ‘it is possible to posit two fundamental types of sentences with finite verb forms, namely (1) first position verb form + circumstance and (2) circumstance + verb form’. The former type of sentence is called a verbal sentence while the latter is called a nominal sentence. 41 These, in turn, are correlated with the foreground and the background sentences in Niccacci.
Now that I have explained how Niccacci’s interpretation of foreground/background grows out of Polotsky’s methodological work, I need to clarify the following issues:
the point of view of the Prague School versus that of Niccacci regarding the pairs predicate–subject and theme–rheme;
the methodological inconsistencies of Niccacci’s Syntax;
the meaning of foreground/background opposition as found in Weinrich.
2.1.3. Evaluation of Niccacci’s definition of foreground/background
This subsection tackles the first and the second issues listed above. By using subject/predicate, Niccacci confuses the syntactic description with the functional description. The confusion increases when one reads further in the same article dedicated to Polotsky that ‘a verb need not automatically be the predicate, and a substantive or a non-verbal element need not automatically be the subject’. 42
However, according to the Prague School, each level of language has its own analysis and vocabulary. First, there is the morphological level (verb, noun, adjective, adverb, etc.). František Daneš outlines the other three levels of analysis which are together part of the syntactical analysis: grammatical (subject–predicate–object), semantic (agent–action–goal), and functional (theme–transition–rheme). 43
Furthermore, from this theoretical discussion of Niccacci’s theme/rheme and Polotsky’s approach, one can assert that Niccacci effectively superimposes the theory of Polotsky’s First and Second Tenses onto the text-linguistic theory of foreground/background: [i]n Egyptian, Polotsky was able to identify the function of the first tenses vs the second tenses in contextual proximity and opposition. As parallel verb forms in Biblical Hebrew we have primarily wayyiqtol for [the] main line of narrative vs x–qatal for the secondary line in narrative.
44
In light of these particular points, Niccacci’s method seems to contain a number of methodological issues which impact profoundly the construal of BH syntax. I explain the importance of some issues already developed (points
(a) The first point is that there is a confusion of the grammatical role of predicate/subject with that of the functional level, theme and rheme. The items in the first pair are communicative units containing the information, while the second pair regulates how that information is distributed in terms of context-dependent/independent information (cf. Firbas). Polotsky indeed argues a legitimate point regarding the functional dimension of the language. Any element of the sentences has a functional role, which is described by the functional phenomenon of rheme/theme. Niccacci borrows a syntactic reading (that of Polotsky and his predicate–subject word order) and puts it together with another functional reading, that of the Prague School, and its theme/rheme theory. The theory of Polotsky and that of the Prague School are not contradictory; on the contrary, they are similar in many ways, with the exception of vocabulary.
The advantages for Niccacci are that he now has the idea of a fixed word order (predicate–subject) and that of theme–rheme very closely linked. That the word order is fixed in language was never implied by the functional stance of the Prague School—in fact, they profess the very opposite—the distribution of the theme/rheme values (the functional analysis) over the syntactical values (subject–predicate–object–adverbial) is flexible and dependent on up to five factors. Each language adheres to one or two principles in the first instance, and only in particular cases, activates the others. 45
(b) Niccacci ignores the fact that if a verb works as subject, that verb should be nominalised (see Polotsky’s essential elements of the ‘phrase coupée’ quoted above). In this context, one could very well argue that Niccacci provides no explanation of how ‘a finite verb form’ (so, a predicative verb form) as in xqatal/xyiqtol sentences, is able to play a nominal role.
(c) The correlation that Niccacci makes between word order with an initial predicate and Weinrich’s foreground has never been independently proven. These concern two different levels of syntax: the functional one and the text-linguistic one respectively. The supposed correlation requires proper methodological discussion of examples before it can be proposed as a core principle of the analysis.
(d) Weinrich’s description of foreground/background and that supposed by Niccacci do not rely on the same theoretical foundations. According to the former, foreground represents that type of information ‘which makes up the reason for which the story is told’. Foreground is that which is usually registered in a [possible] summary, that which the title abstracts or may abstract, that which in substance would induce the people to leave their work for a time to listen to the story of a world which does not belong to daily life
And background is a type of information (cf. the Italian imperfetto) that represents ‘that which by itself would not induce anyone to prick up their ears [to hear it]; that which, nevertheless, helps the hearer in this action [of hearing] by facilitating his orientation in the narrated world’. 46 Weinrich explains elsewhere that background is found in those sections of the story that mark ‘the exposition and final passages, in descriptions and portraits, in marginal scenes, additions, details, and in images’. 47 The way Weinrich looks at this particular pair has nothing to do with the correlation of new/given information or predicate/subject.
2.1.4. Weinrich: foreground and background as tempo movements
Weinrich postulates that tense has nothing to do with space (front stage or back stage). In Knappe Zeit (translated in English as On Borrowed Time), he makes a valuable contribution to our understanding of how time was understood from the days of Aristotle to the present. He concludes that Aristotle’s conception of time is essentially derived from that of space, and they together share a tertium commune which is movement. In contrast with the idea that human time relies on a past–present–future paradigm, Weinrich asserts that [h]uman time, which derives its rhythm from the regular or irregular beating of the pulse, cannot be understood as a movement in space [. . .]. I cannot help but insist [. . .] that time is to be understood in temporal and not in spatial terms.
48
Weinrich’s adjective ‘temporal’ is to be understood as connected to the ‘time’ of the human body which is regulated by pulse or by the tempo of the beating heart. In fact, tenses are types of tempos, a property that is evident in Weinrich’s relievo. The tempo of foreground is presto as it marks the advancing of the plot; background adds the lento details that contribute less to the plot.
There are several important items to note with regard to Weinrich’s theory of foreground/background. The impact of the disconnection between tense and time/space allows Weinrich to incorporate the structuralist position of the zero sign, that is, the sign that marks the absence of something rather than its presence (others call this absence being ‘unmarked’ for a specific quality). Although the connection ‘tense–time’ did not explicitly exclude the opposition between zero degree and retrospection/anticipation, this connection and its derived framework of past–present–future impede the correct perception of the opposition between zero degree and retrospection/anticipation. By removing the source of the confusion from the theoretical description, the connection of tense-time, Weinrich is able to access the true meaning of the zero sign. His zero–degree quality signals the absence or the unmarked position of some tenses with regards to retrospection/anticipation, in contrast with others which do have that mark. Explaining tenses as mirrors of an objectified ‘space’ in language (cf. Schneider, Niccacci, and Reinhart) denies Weinrich’s formula that tense does not mean time. Saying that tense conveys space is the same as saying that tense indicates time.
There are two ensuing clarifications for zero degree and retrospection/anticipation. First, Weinrich’s reading of zero sign hides the true meaning of the foreground/background core value: both foreground and background are zero-degree values, which can be of narrative or of comment. The message of the foreground/background zero-degree values looks to the next thing in a subsequent line of events. This is precisely one of the points where Niccacci’s work becomes most confusing: by losing the perception of the zero degree as foreground and background, one is unable to see that background does not mean retrospection and anticipation.
Second, Weinrich’s foreground/background impacts on the understanding of the recovered/anticipated information. The relievo concerns only the zero-degree narrative or comment. This means that recovered and anticipated information are not equal to the background because they do not contribute to the plot of the foreground (by advancing it).
Moreover, the retrospective/anticipated information does not qualify as being part of the plot. Instead, this retrospective/anticipated information interrupts the foreground/background of presto/lento tempo with information that, while inserted in the current plot, is part of another storyline. I turn now to the other often misunderstood pair in Weinrich’s model, the opposition between the comment and the narrative registers.
2.2. On comment and narrative
Niccacci takes for granted that the pairs of comment/narrative and direct/indirect speech are equivalent. 49 The former pair is actually the most important theoretical point in Weinrich’s Tempus and probably one of the most difficult concepts to apply to a new language. 50
One can hold Weinrich responsible for not stressing enough the difference between one pair and the other. There is also the issue of BH, for which scholars have not yet clarified the difference between direct/indirect speeches properly. Consequently, proposing a new concept like comment/narrative—which seems so closely related to direct/indirect speech, sets up an almost inescapable methodological trap. For these reasons, one needs to examine what each pair means and how they feature in Weinrich’s work. 51
Gideon Goldenberg is one of the few scholars who raised the question of indirect and direct speech in BH. 52 Following the bibliographical notes of Florian Coulmas, 53 he asserts that there are three styles of speech or discourse: (1) ‘Direct speech, oratio recta’; (2) ‘Indirect speech, oratio obliqua, reported speech, [. . .] expressing [. . .] the content of what is/was said etc’; and (3) The ‘independent form of indirect discourse, style indirect libre, [. . .] oratio tecta, [. . .] inner monologue’. 54
Leaving aside the style indirect libre, I suspect that the difference between direct and indirect speech is formal, and easier to determine than that of comment/narrative, which is a qualitative difference. 55 The use of speech verbs like אמר and change of personhood (from third to first/second person) are enough to tell that what is before these formal signs is indirect speech and what is after them is direct speech.
Weinrich seems to support a similar formal relation between direct and indirect speech. According to him, ‘[i]n direct discourse, we quote an opinion in a more or less exact manner. The sign of citation almost always precedes the direct discourse’:
56
The indirect discourse is a recounted opinion which is not cited to the letter. The person recounting rephrases in his own way the opinion he recounts. In the process, he may be more or less “faithful”, i.e. he may move away to different degrees from the exact terms of the recounted opinion and more or less colour it with his own ideas. So, more than one semantic nuance occurs and the passage from direct to indirect speech may provide a glimpse of the ideas that the person recounting holds.
57
Reading more closely these few lines from Weinrich is enough to dispel the notion of equality between direct/indirect speech and comment/narrative. In the case of direct/indirect speech, there is a formal difference between a quotation ‘in a more or less exact manner’ versus a communication in which the ‘person recounting rephrases in his own way the recounted opinion’ or, more generally, the narrator outlines the events in the story. By contrast, with regards to the comment/narrative registers, the point in case is the quality of the message—the existence of involvement or distance between the people communicating (speaker/listener) and the message (‘this involves or does not involve us’). 58
However, the mixing of these pairs originates with Weinrich’s Tempus itself; so, in order to clarify the careful stratification of Weinrich’s vocabulary, a section of this earlier book should be investigated to shed some light on the original confusion. Here, I also look more closely at a pair of terms that Schneider uses as an equivalent to Weinrich’s narrative/comment—narrative/dialogue. 59 The untrained eye may be easily confused by the explanations from the section in question, ‘The Tenses of Dialogue’ of Chapter 8 in Tempus. One needs to stress the three opposing pairs of terms that Weinrich describes: dialogue versus monologue; direct speech versus indirect speech; comment versus narrative.
I begin with the assertion that might be the original source of Schneider and Niccacci’s confusion. To be clear about what it is actually saying, it argues that the change from narrative to comment is a ‘syntactic’ one: ‘Individual literary genres, as the various literary ages, make different use of this literary device [he refers here to the term “a dialogue” which occurs in the preceding sentence], which is a syntactic, not a stylistic device’. 60
By ‘dialogue’, Weinrich does not mean direct speech but the particular category of comment. One must interpret the assertion in this way, since reading his words otherwise (i.e. dialogue is a syntactic device) would render the whole passage incomprehensible. Indeed, after recounting some of the plot with past simple in indirect speech, a narrator could let his characters tell the story in a dialogue with past simple in direct speech which is not a syntactic change by any means (the direct and the indirect speech display the same tense). Past simple retains its narrative syntactic value regardless of the type of discourse (indirect or direct) where it occurs. A closer look at Weinrich’s wording on the two pages under discussion (226–227) shows that he builds on the most general meaning of the word ‘dialogue’. I have rearranged his arguments into the following points from 1 to 4 for the sake of clarity. 61
Weinrich begins this section by stating that ‘Dialogue in itself does not create particular tense problems’. 62 This means that the use of dialogue does not affect his theory of the comment/narrative with regards to tenses. Weinrich’s argument starts with the general situation of two people interacting through dialogue. He proceeds by saying that the characters use the same verbal tenses to facilitate their exchange of information—the range of tenses in an exchange between two people tends to be less than that used normally in ‘the continuous speech of a single person’, or, for us, a monologue. Weinrich asserts here that the opposite of dialogue is not indirect speech, but monologue.
Weinrich describes accurately the situation of BH literature: the narration is interrupted by a ‘dialogue in the form of direct speech’ or ‘the narrative flow is interrupted to insert a passage of dialogue in direct speech’. In this context, dialogue means any exchange between two people (as opposed to monologue, as discussed above), not direct speech or comment. So, when Weinrich says (on page 227) that ‘this literary device [“a dialogue”], [. . .] is a syntactic, not a stylistic device’, he does not refer to dialogue (as the opposite of monologue) or to direct speech. He actually refers to the syntactic exchange from narrative tenses to comment tenses which dialogue or direct speech mirror. The change is syntactical as it means a change from the narrated world to the commented world, both understood as registers (or linguistic attitudes); their syntactical individuality derives from the tenses that compose their logical set of tenses (past, past perfect, conditional versus present, present perfect, and future, respectively).
Weinrich’s argumentation on the matter of tense also states: ‘When the characters of an account communicate in direct speech by dialogue, 63 in the majority of cases they have something to comment together and consequently they use tenses of the commented world’. So, the syntactical difference here is of tenses and it does not associate tense with direct speech or dialogue, but with the ‘commented world’. Tenses are linguistic signs for the commented world and narrated world, not for indirect/direct speech or monologue/dialogue. Weinrich continues by explaining how the change from direct to indirect speech (a change in form of the communication) is translated in most cases at the syntactical level: ‘This is the moment of the heteronomous 64 tense transitions which pass from the group of the narrative tenses to that of the comment tenses’. 65 Again this is about tense, a category analysed by Weinrich with comment/narrative opposition, not with the other two pairs. It is clear now that the syntactic device is not the transfer from indirect to direct speech (or the other general exchange from the monologue of the narrator to the dialogue of the characters) but that of exchange of narrative tenses with comment tenses. 66
Consequently, putting into perspective Weinrich and Goldenberg’s description, the latter places in the same category an array of concepts which may be further delimited—I refer here to the list under no 3: the ‘independent indirect discourse, style indirect libre, [. . .] inner monologue’ 67 ). Following Weinrich, there is an actual difference between these three items cobbled together. Monologue and dialogue suppose one speaker’s communication versus a communication between two or more speakers. Direct speech and indirect speech refer to telling information by ‘quoting’ or not, respectively. These two (I called them formal) differences are both underlined by the syntactic difference between two sets of tenses of comment and narrative, described by Weinrich. They define the way the speaker/listener transmits and receives information: tension and involvement, on the one hand; relaxation and distance, on the other.
3. Conclusion
There is sufficient evidence to question whether Niccacci’s BH Syntax is as methodologically sound as previously thought. By importing syntactical mechanisms from Polotsky without making the necessary adjustments—for example, explaining how the BH morphology allows a predicative verb to be subject, Niccacci has effectively created his own theory of language. Ultimately, his analysis of BH syntax cannot be traced back to Weinrich’s Tempus. The terminology and arguments which Niccacci uses may appear similar to those used by Weinrich, but in truth, they convey meanings which Weinrich never envisaged. In Weinrich’s model, the oppositions foreground/background, zero degree/recovered or anticipated information, and comment/narrative are intimately linked to tense form, each individual tense having its own place in the overall account of the system of tenses. He uses these three oppositions to create a fixed connection between the linguistic function and linguistic sign (tense, or tense and word order). Essentially, each morphological tense receives a specific analysis which does not apply to another tense.
To explain, Weinrich determines his syntactic values by analysing pieces of literature in various languages. This, for example, leads to the conclusion that the French passé simple is the equivalent of the Italian passato remoto. The two tenses share a common value in their respective languages which Weinrich described under three coordinates: under relievo, they are foreground; under register, they are both narrative, and under linguistic perspective, they are zero degree. It is not that two tenses are equivalent to each other; it is the text-linguistic analysis of literary works which designates them as equivalent to each other.
The meaning of this state of facts for BH syntax is that the application of the text-linguistic analysis imposes a fixed relation between one BH sentence type and its function. Subsequently, if a certain function is proposed for one BH sentence type, the function is to be attached to one BH sentence. Moreover, the BH sentence should be translated with its equivalent in modern languages as determined by Weinrich’s Tempus—this is because the translation would not otherwise reflect the reality described in the analysis. While Weinrich had the examples from literature to argue the similarity between passé simple and passato remoto, the syntactic examination of BH completes the cycle of the analysis 68 by stating the unique appropriate translation. If any of these two methodological points are not met in BH, the examination becomes irrelevant as it does not produce the intended outcome of Weinrich’s model—the true and actual connection between function and sign.
In reference to these two points, the end result of Niccacci’s Syntax is exactly this: there is no precise correspondence between function and BH sentence type; and the translation does not reflect appropriately the text-linguistic analysis. He rightly proposes wayyiqtol as the main line/foreground in BH, an analysis which reveals wayyiqtol functions as equivalent of passé simple and passato remoto. However, he assigns the single function of narrative background to four BH sentences: wqatal, xqatal, xyiqtol and the simple nominal sentence. 69 In modern languages, according to Weinrich, background narrative is represented by imparfait/imperfetto and past continuous. Conversely, if one accepts that background narrative in BH takes all these forms, the other possible forms of narrative retrospect and anticipation, and all the meanings that comment contains (zero degree, retrospect/anticipation; foreground/background) are left with no visible possibility of being expressed. 70
In summary, this article raises awareness of the precise vocabulary that Weinrich uses in his writings. It argues that the results of Niccacci’s Syntax are not supported by the work of Weinrich, but also leave aside the connection that the latter makes between sign and function. In addition, it lays the groundwork for a reconsideration of BH syntax within the text-linguistics method. My next contribution proposes a new reading of BH word order based on Weinrich’s work and advancements in post-structuralism.
Footnotes
1.
H. Weinrich, Tempus: Besprochene und erzählte Welt (Stuttgart: Kohlhammer, 1964/1985). There is no English translation, but there are translations into French (1973) and Italian (1978). He develops this theory further through two substantial applications to German and French (H. Weinrich, Textgrammatik der deutschen Sprache (Mannheim: Dudenverlag, 
) and H. Weinrich, Grammaire textuelle du français (Paris: Didier/Hatier, 1982/1989)).
2.
3.
5.
Cf. the introduction of section 2.1 below.
6.
7.
9.
10.
Weinrich, ‘Tense and Time’, 34–35.
11.
Weinrich translates the German term Besprechen as ‘discursive’ in ‘Tense and Time’. I use the term ‘comment’ instead as it is in line with the Italian and French translations of Tempus and avoids confusion with the term direct discourse. To be clear, the labels ‘comment tenses’ and ‘narrative tenses’ come from the German ‘Besprechende Tempora’ and ‘Erzählende Tempora’, cf. Weinrich, Tempus: Besprochene und erzählte Welt, 57.
12.
Weinrich, ‘Tense and Time’, 35.
13.
Weinrich, ‘Tense and Time’, 36.
14.
Weinrich, ‘Tense and Time’, 36.
15.
These two terms refer to the third dimension of Weinrich’s description of tenses. They both translate the German term Reliefgebung; relievo belongs to Weinrich while the other belongs to Watson, cf. Niccacci, The Syntax of the Verb, 14. As ‘prominence’ bears an array of linguistic associations, I opt for the use of relievo in this article.
16.
Weinrich, ‘Tense and Time’, 37.
17.
In order to avoid burdening the reader with unnecessary items, I leave aside the possibility of comment foreground/background which is discussed in Weinrich, ‘Tense and Time’, 38–39. The notion of comment foreground/background is indeed used by Niccacci; however, he does not share Weinrich’s description of comment foreground/background in ‘Tense and Time’ (cf. point 2.1 below).
18.
This is relevant because Niccacci suggests that multiple BH sentence types function as narrative background.
19.
Cf. the Italian translation of Tempus, 79. Not all possible tense paradigms are evaluated fully in the limited space of his Tempus (where these two languages are discussed). This, however, does not mean that the syntactic slots missing in the table are not signified. For example, the English background narrative retrospection slot is most probably filled by the present perfect continuous.
20.
21.
22.
T. Reinhart, ‘Principles of Gestalt Perception in the Temporal Organization of Narrative Texts’, Linguistics 22 (), 787–789. She follows in her view the Gestalt psychology of Kurt Koffa combined with ideas from William Labov. A reading of BH syntax on the lines of Reinhart is Robar, The Verb and the Paragraph in Biblical Hebrew.
23.
Weinrich keeps them separate as they are part of two dimensions: relievo (foreground/background) and perspective (zero perspective versus recovered/anticipated information).
24.
The four types of sentence are also summarised together under ‘Interruption of the WAYYIQTOL chain (with constructions denoting background)’, see Niccacci, The Syntax of the Verb, §39–49.
25.
Cf. Niccacci and Watson [tr], , §14–15 (xqatal is considered both retrospective and background), § 27 (xqatal retrospective); §51 and §88 (xyiqtol represents anticipated information); §135 Table 2 records various combinations of xqatal and xyiqtol as being recovered/anticipated information and background at the same time.
26.
This is another instance where Niccacci mixes relievo with perspective.
27.
Niccacci, The Syntax of the Verb, 29. The square brackets represent my clarifications.
28.
29.
Polotsky, ‘Les transpositions du verbe en égyptien classique’, 1976, 2. Footnote 1 of this paragraph reveals that the theorisation of the concept of transposition goes back as far as L. Tesnière, Elements de syntaxe structurale (Klincksieck, 1959/1982), §151–163 (theory) and Ch. Bally, Linguistique générale et linguistique française (Berne: A. Francke, 1944). I presume that Polotsky’s general approach to morphology is the same in Middle Egyptian and Coptic.
30.
31.
Polotsky, ‘Les transpositions du verbe en égyptien classique’, 15.
32.
Polotsky, ‘Les transpositions du verbe en égyptien classique’, 15–16. The square brackets are mine.
33.
This morphological form is called the ‘Second Tense’.
34.
Polotsky, ‘Les transpositions du verbe en égyptien classique’, 18. Polotsky, ‘Nominalsatz und Cleft Sentence in Koptischen’, 420–421, seems to agree that the labels ‘logical subject’ and ‘predicate’ have no place in the description of verbal sentences. However, he asserts that these concepts in their logical meaning suitably describe the relationships of the two members of the nominal sentence; to be more precise, Polotsky interprets ‘predicate’ in the sense of Demourette–Pichon’s ‘vedette’ or ‘emphasis’, see also J. Damourette and É. Pichon, Des mots à la pensée: Essai de grammaire de la langue française, Vol. 5 (Paris: Collection des linguistes contemporains, ).
36.
In all fairness, neither Niccacci nor Polotsky seem to be aware of the difference. To take the example of Polotsky, while the discussion of the Greek κεκαλυμμένον is a functional one, his proposal for Middle Egyptian and Coptic is a syntactic one, although for both languages he uses the same predicate–subject pair. The confusion of levels may well be originally attributed to Polotsky, cf. H. J. Polotsky, ‘Études de syntaxe copte I–II’, in Collected Papers, ed. E. Y. Kutscher (Jerusalem: Magnes Press, ), §5, p. 25. See also below under point 2.1.3 for the linguistic background of these sentences the demarcation of language levels proposed by František Daneš.
37.
38.
A similar association may be derived from Polotsky’s definition of the verb in Coptic: ‘a word which is capable of filling the last position in the Tripartite conjugation system’ cf. H. J. Polotsky, ‘The Coptic Conjugation System’, in Collected Papers, ed. E. Y. Kutscher (Jerusalem: Magnes Press, ), 239. The term ‘conjugation’ is not limited to the verb but is used in relation to entire sentences and clauses (he states earlier ‘We thus have Sentence Conjugations and Clause Conjugations’, p. 238).
39.
Niccacci, ‘Polotsky’s Contribution to the Egyptian Verb-System’, 402–405.
40.
Niccacci, ‘Polotsky’s Contribution to the Egyptian Verb-System’, 415. This statement explains the state of facts in Middle Egyptian.
41.
42.
Niccacci, ‘Polotsky’s Contribution to the Egyptian Verb-System’, 438.
43.
44.
Niccacci, ‘Polotsky’s Contribution to the Egyptian Verb-System’, 455.
45.
J. Firbas, ‘Some Thoughts on the Function of Word Order in Old English and Modern English’, SMFPUB A5 (), 73 asserts that there may be up to five principles that can influence word order: that of functional sentence perspective (FSP), grammatical, the principle of coherence of certain sentence elements, the rhythmical principle, and the emphatic principle. Firbas argues that in English the grammatical principle and that of coherence are most influential, leaving a secondary role for FSP. The way in which these principles apply to BH has never been discussed in the literature as far as I know.
46.
Weinrich, Tempus. Le funzioni dei tempi, 129.
47.
Weinrich, ‘Tense and Time’ (1970), 38.
49.
He uses only direct-speech passages to support his discussion of ‘Discourse’ in Niccacci, The Syntax of the Verb, §51–65. Schneider before him had the same opinion as is evident from the examples of comment, all of direct speech, presented in the section ‘In dialogue’, cf. Schneider and McKinion [tr], , §44.2.
50.
This is because whereas Weinrich’s narrative register was developed independently by theorists of discourse analysis (see also the work of Paul Hopper, Hellen Dry, and Reinhart), the other equally significant register of comment is unique to him.
51.
The theoretical discussion of comment/narrative contributes to the new interpretation of the BH syntax proposed in the subsequent article.
52.
53.
54.
55.
Both comment/narrative and direct/indirect speech are communications in their own way. To explain the difference I propose between formal and qualitative communication, the speaker/writer pours this communication of narrative and comment quality into direct/indirect speech formal moulds. With the former pair s/he identifies how the message should be understood (involved/non-involved; tensioned/not tensioned), with the latter s/he assigns speaking roles: monologue for the voice of the narrator in indirect speech; dialogue when one or more voices are brought to testify to events.
56.
Weinrich, Grammaire textuelle du français, 565.
57.
Weinrich, Grammaire textuelle du français, 566. Reporting is a key term in Weinrich’s framework; comment retrospective is a report of past events, cf. H. Weinrich, Tempus. Le funzioni dei tempi nel testo (Bologna: Società Editrice il Mulino, ), 104. This is why I did not translate the French words opinion raportée, rapporteur and rapporte with the English equivalents based on the root word report.
58.
Cf. Weinrich, Grammaire textuelle du français, 124–129 (general definitions of commentaire and récit) and 637 (a type of glossary of the terms used in this book).
59.
Schneider, Grammar of Biblical Hebrew, 140–141.
60.
Weinrich, Tempus. Le funzioni dei tempi, 227.
61.
The actual sequence of points is 1, 3, 2, and 4 in Weinrich’s Tempus.
62.
Weinrich, Tempus: Le funzioni dei tempi, 226. The right translation here is not ‘temporal’ but tense, see also the title of this section, ‘The Tenses of Dialogue’, which precedes the quoted sentence.
63.
On three occasions in these two pages, Weinrich implies that dialogue is not the same thing as direct speech.
64.
A homogenous transition is that from a tense to another which implies the change of one dimension, for example, passé simple to imparfait (from foreground to background within the relievo dimension where the register and perspective stay the same, i.e. narrative and no recovered/anticipated perspective). The ‘heteronomous tense transitions’ means any transition that implies changing two dimensions, for example from passé simple to passé compose – narrative to comment; and zero degree to recovered information), see Weinrich, Tempus. Le funzioni dei tempi, 250.
65.
Weinrich, Tempus. Le funzioni dei tempi, 227.
66.
Weinrich lists several other signs which mark the change from narrative to comment (‘the [narrative or comment] structure of the account should be evident also when read aloud’): (1) the passage from third to second person (‘Marcel soupirait. ‘Tu peux être sûre qu’il n’a jamais vu …’); (2) the transition from French imparfait (a narrative tense) to a present (a comment tense); (3) the use of a verb of communication (the French ‘soupirait’ introduces Marcel’s speech). Weinrich, Tempus. Le funzioni dei tempi, 227–228, the square brackets are mine.
67.
Goldenberg, ‘On Direct Speech and the Hebrew Bible’, in Studies in Hebrew and Aramaic Syntax, 80.
68.
For Weinrich, the cycle starts with observing the function of tense in French and Italian literatures and ends with the proposal of unique syntactic analysis for each tense.
69.
70.
Supposing that BH anticipation and retrospection are achieved by subordination is not part of Weinrich’s original proposal. Subordination does not qualify the background—the former may be used as ‘cumulative’ evidence for the latter, cf. Weinrich, 1964/1978, 196.