Complex syntax as a window on contrastive rhetoric

Available online at www.sciencedirect.com

Journal of Pragmatics 42 (2010) 744–765 www.elsevier.com/locate/pragma

Complex syntax as a window on contrastive rhetoric Bracha Nir a,*, Ruth A. Berman b,1 a

Communication Disorders, Haifa University, Mt. Carmel, Haifa 31905, Israel b Linguistics Department, Tel Aviv University, Ramat Aviv 69978, Israel

Received 25 December 2007; received in revised form 16 July 2009; accepted 24 July 2009

Abstract The paper concerns complex syntax in the sense of text-embedded clause-combining. We consider different perspectives on why languages employ complex syntax, taking the usage-based view that ‘‘discourse drives grammar’’. Complex syntax is analyzed as shedding light on the nature of ‘‘contrastive rhetoric’’, on the assumption that linguistic typology interacts with rhetorical strategies in the construction of discourse. An innovative methodology is delineated for evaluating syntactic complexity along a hierarchy of clause-combining relations, from isotactic single clauses to paratactic symmetric and asymmetric stringing by coordination and complementation, on to hypotactic layering by adverbials and relative clauses, and endotactic nesting or embedding of one clause inside another. Detailed criteria for each of these levels of clause-combining were applied to 64 narrative texts written by graduatelevel university students, native speakers of four different languages (English, French, Hebrew, and Spanish) on the shared topic of interpersonal conflict. The discourse effects of linguistic typology are analyzed in terms of the linguistic means available to these different languages for combining clauses as well as discursive strategies preferred by speaker–writers in constructing narratives. # 2009 Elsevier B.V. All rights reserved. Keywords: Complex syntax; Clause packages; Contrastive rhetoric; Linguistic typology

1. Introduction This paper is concerned with complex syntax as a facet of cross-linguistic, text-embedded ‘‘contrastive rhetoric’’. Focus is on personal-experience narratives, although the principles delineated below are assumed to apply across genres and sub-genres of discourse. As background, we note different perspectives on the notions of complex syntax (section 1.1) and contrastive analysis (section 1.2), and briefly review previous studies that combine the two domains (section 1.3). 1.1. Complex syntax The topic of ‘‘complex syntax’’ has been of concern to both traditional grammarians (Jespersen, 1924) and contemporary linguists (Bybee and Noonan, 2002; Givo´n, 2009). Traditionally, a complex sentence is defined as being made up of a number of ‘‘simple sentences’’ (Lyons, 1968:178). In a functionalist perspective, Foley and Van Valin (1984) treat the topic in terms of the notion of ‘‘nexus’’, defined by Jespersen (1924:116) as made up of ‘‘two ideas which must necessarily remain separate. . .whereas the junction is more stiff or rigid, the nexus is more pliable; it is, as * Corresponding author. Tel.: +972 4 8288586; fax: +972 4 8249507. E-mail addresses: [email protected] (B. Nir), [email protected] (R.A. Berman). 1 Tel.: +972 3 640 5018; fax: +972 3 640 5109. 0378-2166/$ – see front matter # 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.pragma.2009.07.006

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

745

it were, animate or articulated’’. Contemporary references to ‘‘clause-combining’’, ‘‘clause-linkage’’ (Haiman and Thompson, 1988), or ‘‘clause complexes’’ (Matthiessen, 2002) highlight the role of the clause as the basic unit of analysis in complex syntax, in order to consider constructions that go beyond the boundaries of the simple sentence or clause-internal structures (Shopen, 2007). Different perspectives on complex syntax derive from distinct motivations, ranging from strictly structuredependent to more functional approaches. In the framework of generative grammar from its very outset, complex sentences were of interest as constituting an instantiation of formal principle such as abstract hierarchical structure and recursion, critical to the characterization of Universal Grammar (Chomsky, 1957, 1965). And indeed, crosslinguistically, languages all appear to have some kind of devices for constructing complex sentences. A more functionally oriented typological perspective requires taking into account the different types of structures that may encode the same functional domains even if they are syntactically distinct (Lehmann, 1978; Foley and Van Valin, 1984; Givo´n, 1984; Croft, 1990). From the related perspective of cognitive linguistics, complex sentences are viewed as grammatical constructions that express a relationship between two or more conceptual units. In this framework, for example, subordinate clauses (unlike coordinate clauses) are considered part of the same processing and planning unit as their associated main clauses (Bybee, 2002; Diessel, 2004). In psycholinguistic terms of language processing and comprehension, researchers are concerned with such issues as, for example, the fact that more closely integrated syntactic structures such as center-embedded relative clauses can be viewed as involving lower processing costs, especially in supportive contexts of the kind provided by surrounding discourse (Gibson, 1998; Grodner et al., 2005). Rather different issues are considered in language acquisition research, where studies based on formal generative grammar consider complex syntax as demonstrating children’s early-developing knowledge of formal principles of linguistic theory (Lidz, 2007; Lust et al., 2008). In usage-based child language studies, focus has been largely on use of connectives as reflecting age-related command of complex syntax (for example, Jisa, 1987; Peterson and McCabe, 1991; Berman, 1996; Akinci and Jisa, 2000; Diessel, 2004), while syntactic development is typically characterized as progressing from linear coordination to subordination and embedding (Verhoeven et al., 2002). From yet another perspective, in discourse studies, concern with the flow of information in discourse has led researchers to consider clause-linkage as playing a role in the internal organization of parts of the discourse, for example: to distinguish foreground and background relations, to highlight and focus on some events or ideas in relation to others, and to guide the reader–hearer as to what is going to come (Thompson, 1984; Haiman and Thompson, 1988; Chafe, 1994; Halliday, 1994). In the present context, we adopt the usage-based view of current cognitive and functional linguists, extending it to the idea that ‘‘discourse drives grammar’’ in general (Thompson, 1985; Du Bois, 2003; Tomasello, 2003; Thompson and Couper-Kuhlen, 2005). That is, we view the organization and re-organization of linguistic forms as embedded in discourse, whether conversationally interactive or monologic, spoken or written. And in a developmental perspective, we consider that ‘‘discourse provides children with a developmental mechanism for the acquisition of linguistic devices’’ (Hickmann, 2003:335) – in the present case, complex syntax. A question that as far as we know has not been directly addressed in the linguistics literature is ‘‘why complex syntax?’’. That is, why is it not enough to juxtapose simple clauses in chains, one after another, without syntactic dependency between them? To illustrate, consider the difference between simple stringing of utterances as against using complex syntax to package clauses in the excerpts in (1) and (2) below – taken from two oral texts describing the opening scenes of a picture-book based story (Berman and Slobin, 1994). (A square bracket ] indicates clause boundary, double slashes // indicate new utterances, syntactic connectives are underlined, and xxx stands for unclear strings.) The text in (1) was produced by a 4-year-old child, and consists of 90 words and 15 utterances, all in the form of a single predication (that is, an isolated clause) – except for one relative clause. (1) Opening to Oral, Picturebook-based Story [4;6] they looked at their little pet frog ] and the boy and dog - and he wasn’t there! ] // and then it was nighttime ] // and then they waked up ] // they didn’t see the frog ] // and then they loo. . . and then they looked in the can ] where they to put them ] // then they called after him ] // then they put. . . ] // and then. . . and then he jumped out a window ] the dog jumped ] // and then the boy grabbed him ] // and then they xxx ] // and then they flied xxx ] and he looked for him ] //

746

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

The text in (1) is almost entirely devoid of complex syntax in the sense of syntactic packaging, as a process in which clauses are linked syntactically (Berman and Slobin, 1994:538–554), rather than by utterance-initial discourse markers or segment taggers such as and, and then (Schiffrin, 1987; Berman, 1996; Ravid and Berman, 2006). The only exception is use of the conjunction where introducing a relative. Besides this one case, the passage is made up of simple sentences. In marked contrast, the 9-year-old’s version in excerpt (2) of the situations depicted by the same series of pictures – almost the same in length as the 4-year-old child (110 as against 90 words) – is packaged together into half the number of syntactically linked units. Its 17 clauses – including one that is indicated by angled brackets as center-embedded – are combined into six units of complex syntax, yielding an average of nearly 3 clauses per package.

(2) Opening to Oral, Picturebook-based Story [9;1] once there was a boy ] who had a pet frog and a dog ] and that night he was watching it . . . ] // and . . . when he went to sleep ] the frog got out of his jar ] and got away ] // the next morning he saw ] that the frog was gone ] so he put on his clothes ] // and the dog acci. . . accidentally got his head stuck in the jar ] which the frog was kept in ] // they called out the window ] and the dog fell down ] and the glass jar broke ] and um the boy was mad at him ] // so they went off . . . to find his um ... pet frog ] //

As such, these examples show how, in place of strictly structural analysis, clause-packaging can be shown to have distinct rhetorical effects. Thus, the excerpt in (2), but not that in (1), demonstrates how event components can be highlighted – for example, by contrasting between the preposed (and thus non-narrative, Couper-Kuhlen, 1989) temporal adverbial when he went to sleep and its main clause the frog got out of his jar; and how different phases of an event can be conflated – for example, by specifying what the boy found on waking up as a single ‘‘event complex’’ (Berman and Slobin, 1994). In more maturely proficient narratives, such devices increasingly provide speaker–writers with control over the rhythm and tempo of the narratives that they construct, serving for expression of an individual rhetorical style (Berman, 1988). 1.2. Contrastive rhetoric Consider, next, how complex syntax ties in with the second facet of this study – ‘‘contrastive rhetoric’’. The term ‘‘rhetoric’’ traditionally refers to instruction in the effective use of language. In applied terms, speakers and writers are taught to select those linguistic devices that express the intent of their utterance most appropriately and effectively (Fix et al., 2006). In this sense, use of complex syntax obeys one of Slobin’s (1977) charges for effective communication to be ‘‘rhetorical’’ in the sense of being maximally expressive, since ‘‘languages must provide means for expressing notions and for compacting semantic content. . .’’ (p. 187) even though this may lead to an increase in complexity. Linguists have adopted different approaches in attempting to relate complex syntax to discourse rhetoric. These include Rhetorical Structure Theory [RST] (Mann and Thompson, 1988), Discourse Representation Theory [DRT] (Kamp and Reyle, 1993), and Segmented Discourse Representation Theory [SDRT] (Asher, 1993; Asher and Lascarides, 1995), as well as Coherence relation approaches (Hobbs, 1979; Sanders et al., 1992). Despite important differences, all such approaches view discourse structure as part of the conceptual structure (broadly speaking, the semantics) of a text. Thus both RST and SDRT assume that the process of computing discourse coherence entails relying on these relations to link utterances so that the text will ‘‘make sense’’ (Knott et al., 2001). Such approaches propose a taxonomy of rhetorical relations that hold between parts of a text organized around a rhetorical structure that determines, for example, the selection of connectives (e.g., Degand, 1998; Knott and Sanders, 1998) or referring expressions (Fox, 1987). The structure of discourse is treated as a function of the relations between ‘‘spans of text’’ (Mann and Thompson, 1988) whose thematic unity and hierarchical organization may be (but are not necessarily) signaled by surface use of elements such as connectives. The present study, rather than considering

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

747

discourse connectivity in thematic terms, characterizes clause complexes as spans of text that represent different layers of syntactic structure. To this end, our analysis adopts the perspective of ‘‘contrastive rhetoric’’ in cross-linguistic rather than intercultural terms (in contrast, for example, to the studies in Connor et al., 2008). To date, a cross-linguistic approach to contrastive rhetoric has been used mainly in L1–L2 comparisons, in studies that extend the structuralist notion of contrastive analysis to extended discourse (Kaplan, 1966; Bar-Lev, 1986; Ostler, 1988), for example, in essays written by students of English as a foreign or second language (Connor, 2002). But even such studies fail to consider the organization of entire texts as units of discourse or of the syntactic components of a text as ‘‘macrostructures’’ (van Dijk, 1980). Here, we adopt Slobin’s form-function orientation to the cross-linguistic study of language structure and language use (Slobin, 1973, 2001), and interpret contrastive rhetoric as referring to the expressive options that are favored by users of a given language for particular discourse purposes (Berman and Slobin, 1994; Berman, 1995). Contrastive rhetoric is a complex concept that involves several different facets, including: (1) Expressive options – the repertoire of formal devices available in a given language (Lx), including closed and open class lexical items, syntactic constructions and processes, word order, prosody, and so on; (2) Rhetorical preferences – the constructions that are selected by users of Lx out of the total repertoire of expressive options as favored for meeting a particular discourse function at a given time and in given communicative contexts (Tannen, 1980; Jisa et al., 2002; Tolchinsky and Rosado, 2005); (3) Rhetorical strategies – the ways speaker–writers of Lx manipulate the available expressive options through use of such devices as repetition, parallelism, juxtaposing, direct speech, and figurative language (Berman, 1988; Johnstone, 1996; Tannen, 2007). For example, following Jakobson’s ideas on poetic rhetoric, Reinhart (1995) considers ‘‘equivalence relations’’ – of syntactic, lexical, and discursive constructions – as constituting a rhetorical device in the sense of a special means of narrative evaluation; (4) Rhetorical style – how linguistic means are employed to establish a personal manner of expression, for example, prosaic or poetic, concise or verbose, clear or obscure, fluent or disjointed, plain and straightforward or fancy and circumlocutious (Berman and Slobin, 1994:622– 639; Johnstone, 1996); and (5) Rhetorical purpose – the communicative aims of speaker–writers, such as how persuasive or impressive a piece of discourse will be, or which events or ideas they choose to talk or write about – as in intentional approaches that highlight the role of the speaker–writer’s plans, goals, and intentions for achieving discourse coherence (Grosz and Sidner, 1986; Klein and von Stutterheim, 1991). The present study focuses on the first three of these dimensions – options, preferences, and strategies – and relates briefly to the fourth dimension (style) as a basis for possible interpretations of our findings, with little or no reference to individual style or communicative purpose. 1.3. Clause-combining and contrastive rhetoric In the present context, then, contrastive rhetoric refers to how speaker–writers select and deploy the repertoire of expressive options (lexical and syntactic) afforded by their language as their preferred strategies for meeting the discourse function of text-embedded clause-linkage. There are not many studies that combine concern for complex syntax and contrastive rhetoric. But the few that are available are highly suggestive from our point of view. For example, Ostler (1988) identified several unique characteristics of essays written in English by Japanese-speaking students compared with students of Arabic, Spanish, and English-speaking backgrounds. She found that the Japanese students used fewer words per sentence and showed little syntactic elaboration, making less use than the others of relative clauses, nominalized constructions, passive voice, and so on. And recent work on contrastive clause-linkage rhetoric in Norwegian, German, and English suggests that the three languages have distinct strategies of information packaging: Norwegian prefers paratactic structures where English and German tend to prefer hierarchical or hypotactic relations. As a result, in translation from English, Norwegian seems to use coordination (both sentential and verb-phrase) as a preferred means for linking units of discourse units (Fabricius-Hansen, 1999, 2005). These studies are particularly germane to our present interests, but from a different point of departure, since we do not contrast source and target languages in L2 learners or in translations. Rather, we analyzed the syntactic architecture of narrative texts constructed by native speaker–writers of different languages, with the following motivations. Note, first, that concern here is with syntactic architecture as involving both the syntactic construction type of each individual clause (Main Clause, Coordinate Clause, Relative, Adverbial, and so on) and its relation to the clauses that surround it. Second, we adopt a cross-linguistic perspective, in order to

748

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

throw light on the question raised by Dan Slobin following from the ‘‘frog book’’ study of children’s narratives, as to ‘‘whether typological contrasts in rhetorical style found in the frog stories are restricted to this limited genre of pictureelicited narratives intended for children’’ (2004a:117). Our analyses are not based on a picture-book story but on personal experience accounts. In what follows, we analyze strategies of clause-combining favored by speaker– writers of different languages, and we interpret these favored strategies as a feature of the preferred narrative style of their particular language, in the sense of how they choose to deploy the means available in their language for constructing narrative discourse. 2. Inter-clausal syntactic architecture Like earlier studies of complex syntax in terms of clause-combining or clause-linkage, we take the clause as our basic unit of analysis, where this is defined, following Berman and Slobin (1994:660–663), by a combination of syntactic and semantic criteria as ‘‘Any unit that contains a unified predicate. . .that expresses a single situation (activity, event, state).’’ In considering how clauses are combined into complex syntactic units in the text-embedded context of personal-experience narratives, we needed to define the boundaries of such units. Note that, although our analyses relate to written texts, we deliberately avoid reference to the notion of ‘‘paragraph’’, since this is a unit of discourse that, as far as we know, has not proved accessible to empirical specification or to psycholinguistically motivated definitions (Stark, 1988; Hwang, 1989; Heurley, 1997). We also avoid reference to the term ‘‘sentence’’ on two orthogonal grounds. In linguistic analysis, the sentence constitutes an abstract grammatical construct constrained by the stipulations of a particular formal model of syntax. And in terms of usage, the sentence is a unit that is relevant mainly to written, typically edited contexts of language use, whereas the materials we analyze are texts produced by non-expert speaker–writers. Rather, we analyze complex syntax in terms of how clauses are linked or combined together in extended discourse by a process of what we term ‘‘packaging’’. Our concern is thus close to the Berman and Slobin (1994:538) idea of syntactic packaging where the term ‘‘packaging’’ serves as ‘‘a kind of visual metaphor for the various ways in which situations can be analyzed into components and encoded in multiclausal constructions’’. However, we extend the earlier notion of ‘‘syntactic packaging’’ in terms of a unit termed a ‘‘clause package’’ [CP] as our basic unit of analysis. (Details of why and how such units were defined are given in Nir-Sagiv, 2008; Berman and Nir-Sagiv, 2009) A CP is a text-embedded unit of occasionally only one clause but generally of two or more clauses connected by abstract linkage relations that are typically but not necessarily identified by syntactic criteria. Thus, within each CP, relations of clause-linkage are most often explicitly marked by coordinating or subordinating conjunctions, but they may sometimes be inferred from the thematic progression of a text. As such, clause-packaging departs from most linguistic analyses of ‘‘nexus’’ (Foley and Van Valin, 1984), ‘‘clausecombining’’ (Haiman and Thompson, 1988), or ‘‘clause complexes’’ (Matthiessen, 2002). Clause Packages also differ from traditional, pedagogically motivated notions such as a ‘‘T(erminable) Unit’’ (Hunt, 1970; Verhoeven and van Hell, 2008), since they take account of how such units function in the text as a whole.2 For example, we distinguish cases where lexical connectives like and, so, but function as grammatical markers of connectivity or as pragmatically motivated ‘‘utterance-introducers’’ (Schiffrin, 1987; Berman, 1996) or as ‘‘segment-tagging’’ discourse markers (Ravid and Berman, 2006). Moreover, to delimit CP boundaries, we also take account of discourse-topic shift or maintenance – whether the speaker–writer is referring to a different aspect of the same topic or to a distinct topic. Our investigation follows from and elaborates on prior analyses of complex syntax, both traditional and more current. We define five distinct categories of relationships between different types of clauses, summed up as follows: I. Isotaxis = ‘‘equal organization’’ applies to isolated clauses; II. Symmetric Parataxis ‘‘side-by-side organization’’ refers to stringing of clauses; III. Asymmetric Parataxis refers to dependent stringing of clauses3; IV. Hypotaxis ‘‘one under the other’’ stands for layering of clauses; and V. Endotaxis ‘‘one inside the other’’ stands for nesting of clauses

2

Although ostensibly conducted in the same framework as the present study, hence adopting the term ‘‘clause packaging’’, the analysis of Verhoeven and van Hell in fact considers only T-units. 3 Van Valin and LaPolla (1997) specify nexus-internal relationships between clauses as distinguishing between those they define as ‘‘independent’’ (in cases of coordination) versus ‘‘dependent’’ (in cases of what they term ‘‘cosubordination’’ and also subordination).

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

749

Fig. 1. Categories of inter-clausal syntactic architecture.

(Lehmann, 1988; Koch, 1995; Akinci and Jisa, 2000; Viguie´, 2001). These five layers of clause-combining are represented graphically in Fig. 1. I.

Isotaxis [ISO] – refers to lone or isolated clause constructions. These are typically asyndetic, that is, they are syntactically ‘‘unconnected’’ to their surrounding clauses to which they are not linked by a conjunction.4 We identify two main types of isotactic relations: (a) single-clause constructions that are ‘‘autonomous’’ in the sense that they are not linked to the clause that precedes or follows them; and (b) main clauses [labeled as MC] – typically the opening clause of a package that is syntactically related to its adjacent clause(s) – which are analyzed as isotactic with respect to the Main Clauses of their neighboring Clause Packages. These two types of Isotaxis are illustrated by the two CPs in (3), excerpted from the narrative written by an English-speaking graduate student – with CP6 consisting of a lone clause, and CP7 of a main clause followed by two other clauses in the same package. (3) Isotaxis = Single Clause and Main Clause [#05]5 CP6: Over the next week this issue came up time and again with my friend and I. [ISO] CP7: Somehow the issue became personalized [ISO = MC] in that we each thought that the other was being too judgmental and rigid.

II.

Symmetric Parataxis [S-PAR] – refers to clauses that are linked to the clause or clauses that precede them in a relation of equivalence by a process of symmetric stringing. This relation takes the form of either asyndetic juxtaposition or overtly marked coordination. Juxtaposed clauses [labeled as JUX] are syntactically unmarked (asyndetic) in relation to the Main Clause to which they are, however, semantically or pragmatically related (Mann and Thompson, 1988; Koch, 1995; Johnstone, 1996; Akinci and Jisa, 2000; Viguie´, 2001). They are thus not analyzed as isotactic or isolating, bur rather as linked to other clauses inside the same Clause Package [CP]. Coordinate clauses may be asyndetic (where three or more clauses are conjoined in a listing relation) or they may be marked by a conjunction (and, but, or). Coordinated clauses in a relation of equivalence – that is, instances of symmetric stringing – must have an overt subject, either different [labeled as DS] or the same [labeled SS], in the latter case, generally in the form of a pronoun.

4 The term ‘‘conjunction’’ is restricted to syntactically conventional markers of clause linkage. We deliberately exclude from consideration various types of discourse markers of the kind that serve as utterance-initial ‘‘segment taggers’’, in contrast to clause-initiating markers of interclausal relations. The term ‘‘syndetic’’ is used in the traditional (dictionary) sense of ‘‘serving to connect, as a conjunction, copulative or conjunctive’’ or ‘‘connected by conjunction’’. 5 Figures marked by a double bar # indicate participant number from 01 to 20.

750

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

Symmetric stringing by juxtaposition or coordination is typically to a Main Clause, although any clause can in principle embody a relation of equivalence with respect to a preceding clause. That is, symmetric parataxis is not confined to Main Clauses, particularly in the case of coordination. On the contrary, paratactic stringing of complement, adverbial, and/or relative clauses is a hallmark of textembedded complex syntax. Symmetric parataxis is illustrated by Juxtaposition in (4a) – where ‘‘My friend’s perspective’’ and ‘‘his view’’ constitute parallel, periphrastic presentations of the same semantic content; and by Different/Same-Subject Coordination in (4b) – where We is coordinated with the conflict in CP6 and the pronoun I is repeated across the two conjuncts in CP12. (4a) Symmetric Parataxis = Juxtaposition [#10] CP4: My friend’s perspective was [ISO = MC] that the individual was being treated too harshly. It was his view [S-PAR = JUX] that a more thorough investigation plus a more lenient judgment may have been better. (4b) Symmetric Parataxis = Different/Same Subject Coordination [#11] CP6: We both exchanged less than optimal tones with each other [ISO = MC] and the conflict or tension was all too evident [S-PAR = DS] CP12: I next considered [ISO = MC] knocking on the person’s door . . . but I decided against this also [S-PAR = SS] III.

Asymmetric Parataxis [As-PAR] refers to clauses that are linked together in a relation of partial equivalence or dependent stringing such that the second clause can only be interpreted by reference to the preceding clause, but is not subordinated to it.6 We identify three main types of asymmetric paratactic relations: (a) coordination with same-subject ellipsis [labeled as SSE] (grammatically obligatory in languages like Spanish or Italian)7 or verb-gapping [labeled as GAP]; (b) appositions [labeled AP]; and (c) complement clauses [labeled CMP]8 – including embedded or indirect questions, direct speech constructions, and complement clauses in copular constructions. In the latter case, both the matrix and the dependent clause cannot occur in isolation, since the argument is obligatorily required by the complement-taking predicate (Cristofaro, 2003). These categories are illustrated by the examples in (5), with an elided subject and that marked complement in (5a), verbgapping in (5b), appositions in (5c), direct speech in (5d), and that marked complement of copula in (5e). (5a) Asymmetric Parataxis = Same Subject Ellipsis/Complement [#20] CP3: I confronted her about it [ISO = MC] and Ø demanded [As-PAR = SSE] that she tell me the truth [As-PAR = CMP] (5b) Asymmetric Parataxis = Verb Gapping CP5: she would tell on us [S-PAR = JUX] and we on her [As-PAR = GAP] (5c) Asymmetric Parataxis = Appositions CP3: usually I’m very nonconfrontational [ISO = MC] I kinda just like let things slide [As-PAR = AP] (5d) Asymmetric Parataxis = Direct Speech Complement [#06] CP3: She shot back with [ISO = MC] ‘‘Don’t tell me . . . !’’ [As-PAR = CMP] (5e) Asymmetric Parataxis = Complement of Copula [#05] CP4: My friend’s perspective was [ISO = MC] that the individual was being treated too harshly [As-PAR = CMP].

6 Analogous references to the distinction between symmetric and asymmetric inter-clausal relations are found in Foley and Van Valin (1984) and Diessel (2004). The term ‘Asymmetric Parataxis’ in our analysis is not necessarily restricted to constructions analyzed as ‘Asymmetric Coordination’ in the sense of Ho¨hle (1990) and Frank (2002). Their focus, unlike ours, is on coordinated sentential structures that differ with respect to the order of the finite predicates they contain. 7 This means that, in fact, Same-Subject coordination with elision has a different status in Spanish than in subject-requiring languages like English or French, with Hebrew lying between these two extremes (Berman, 1990). 8 Van Valin (2005) relates to complement ‘that’ clauses in English as symmetrical linkage, since it is in what he terms an ‘‘extra-core’’ position.

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

IV.

Hypotaxis [HYPO] refers to clauses that are related to the preceding clause by a non-equivalence relation of asymmetric layering, where one clause is dependent on the preceding clause, in which it is ‘‘integrated’’ (Cristofaro, 2003; Diessel, 2004) or with respect to which it is ‘‘downgraded’’ (Ohori, 1994; Raible, 2001). The two main types of such constructions are (a) relative clauses and noun complements and (b) adverbial clauses [labeled as RC and ADV respectively].9 These clauses may be either finite (tensed) or non-finite (infinitive, participial, gerundive, labeled as NF), where nonfiniteness is viewed as a scalar notion that involves different levels of integration with its adjacent clauses (Givo´n, 1980). These categories are illustrated in (6a) by adverbial clauses that are preposed to the main clause in CP1 and that follow it in CP2; in (6b) and (6c) by a finite relative clause and a noun complement; and in (6d) by non-finite adverbial and relative clauses respectively. (6a) Hypotaxis = Preposed/Postposed Adverbial [#11] CP1: As I stated just moments prior in an interview [HYPO = ADV], a story about a conflict . . . occurred recently [ISO = MC]. CP2: Within my research laboratory there are certain times . . . for example [ISO = MC], when one is under the time pressure of a grant submission deadline [HYPO = ADV]. (6b) Hypotaxis = Relative Clause [#02] CP1: Several years ago I had a room-mate [ISO = MC] who was also a close friend [HYPO = RC] (6c) Hypotaxis = Noun Complements CP8: The conflict arose from this situation [ISO = MC] because my parents made the mistake [HYPO = ADV] of telling their 14-year-old son [HYPO = NCMP] that it was because of Rick’s earring [As-PAR = CMP] (6d-1) Hypotaxis = Non-finite Adverbial [#08] CP8: Eventually Michelle Parullen came running down the hall [ISO = MC] to inform me [HYPO = ADV-NF]. . . (6d-2) Hypotaxis = Non-finite Relative Clause [#16] CP3: This was an inevitable source of conflict [ISO = MC] resulting in a few harsh words, much hurt, and a great deal of tension [HYPO = RC-NF]

V.

Endotaxis [ENDO] refers to a process of nesting, by which constructions, mainly Adverbial – as in example (7a) – and Relative Clauses (7b), are inserted inside another clause. Nesting occurs typically in center-embedding, but also by constructions introduced parenthetically, as in (7c). (7a) Endotaxis = Center Embedded Adverbial Clause [#02] CP7: I did not pressure him about his debt for a few days [ISO = MC] but [ENDO = ADV] he assured me. . .[S-PAR = DS] (7b) Endotaxis = Center Embedded Relative Clause [#18] CP1: One of the most recent conflicts [ENDO = RC] occurred [ISO = MC] . . . (7c) Endotaxis = Parentheticals [#16] CP5: and I would not judge them as lesser of people.

751

This analysis treats inter-clausal relations as a categorial hierarchy. As such, it is consistent with Van Valin’s (2006) insight that syntactic linkage relations may be ranked hierarchically in terms of the strength of the syntactic bond between the units. He, too, takes into account how integrated the units are into a single unit or whether they need to be coded as distinct units, with clausal and sentential coordination representing the weakest bonds. Although both approaches are discourse-oriented, our analyses apply to extended monologic texts, relating textual connectivity (in the sense of Berman, 1998) to rhetorical processes of isolating, stringing, layering, and nesting of clauses within the unit of a Clause Package. Further, our notion of nesting means that center-embedded and parenthetical clauses are set apart from regular processes of subordination. And we take into account non-syntactically specified relations such as thematic juxtaposition, while excluding from analysis non-syntactic ‘‘discourse marker’’ types of lexical connectives. 9

Adverbials may also relate to a following clause, when pre-posed to the main clause, indicated in our coding by ADV-PRE.

752

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

3. Description of study The texts analyzed for this study were collected in the framework of a large-scale cross-linguistic project. The project was funded by a Spencer Foundation major research grant to Ruth Berman as PI (for details of aims and elicitation procedures, see Berman and Verhoeven, 2002; Berman and Katzenberger, 2004; Berman, 2005), and involved 20 subjects at four different age groups – from grade school, junior-high school, high school, and graduate school. The population consisted of native speakers of different languages from monolingual, middle- to upper-class backgrounds in seven different countries: France, Iceland, Israel, the Netherlands, Spain, Sweden, and the United States. All subjects were first shown a 3-min wordless video clip depicting different (unresolved) conflict situations. The video clip was neutralized for cultural content – by eliminating all verbal cues, spoken or written, and having typical teenager participants clothed in ‘‘universal’’ jeans and tee-shirts – and proved usable across the countries in our study. It showed scenes of interpersonal conflict in a school setting – for example, a moral conflict of whether to cheat in an exam or return a purse someone dropped, a social conflict of how to treat a new kid who interfered in a conversation, and a physical conflict of fighting during recess. After seeing the video clip, participants were each required to produce four texts in balanced order. They were asked to write and tell a story about an incident where they had been involved in a situation of ‘‘problems between people’’ (a personal-experience narrative) and to write a composition and give a talk in which they discuss the topic ‘‘problems between people’’ (an expository discussion). That is, both narrative and expository texts were elicited on the shared, socially relevant theme of interpersonal conflict. A total of 320 written and spoken texts was collected at each of the seven sites. The present study deals with a subset of the larger cross-linguistic sample, analyzing 74 narrative texts written by 16–20 (university-educated) adults in each of four target languages: Californian English, Israeli Hebrew, Castilian Spanish, and French [France].10 Confining the sample for the present analysis to written narrative texts and to adult participants, enabled us to focus on cross-linguistic facets of clause-packaging by neutralizing the other variables of the larger cross-linguistic project – age, modality, and genre (Berman and Nir-Sagiv, 2007; Berman and Ravid, 2009). The adult participants were all graduate-level university students in their 20s and early 30s, specializing in the humanities and social sciences or in the mathematical and life sciences.11 Participants were thus all well-educated and highly literate, but they were non-expert language users in the sense that they were not professional writers, editors, or translators. 4. Text-based analyses Each of the 74 texts in the sample was transcribed by trained linguists, native speakers of each of the target languages, following the conventions specified for CHAT (MacWhinney, 2000), and further elaborated by special procedures standardized for cross-linguistic and cross-modal comparisons across the international project (Berman and Verhoeven, 2002). Quality control of all transcripts was conducted at the data-collection site by native speakers of the four languages in the present study. Explicitly motivated criteria were specified for division of texts into clauses and Clause Packages, as detailed in Nir-Sagiv (2008). All texts were verified for reliability of clause division, interclausal relations, and of Clause Package segmentation by three raters, at least one of whom was a native speaker of the language in which the texts were written.12 Inter-rater reliability was based on consensus estimates (Stemler, 2004) of the ‘‘amount of agreement or correspondence among two or more coders’’ (Neuendorf, 2002:141), and is reported here as a simple percentage value.13 Agreement reached over 90% between raters, and cases of disagreement were resolved in discussion with independent judges trained in linguistics. Extensive experience with division of oral and written texts, both narrative and expository, in a range of cross-linguistic projects, has shown that linguistics students can be 10

Data collection for the present analysis was supervised for English in San Diego by Judy S. Reilly, San Diego State University, for Spanish in Co´rdoba by Liliana Tolchinsky, University of Barcelona, and for Hebrew in Tel Aviv by the authors of this paper. 11 That is, all participants had to have had experience in academic writing at either or both the under-graduate and post-graduate level. 12 The two authors, native speakers of Hebrew and English respectively, conducted reliability for these two languages together with graduate students of linguistics at Tel Aviv University; for French, we worked together with Harriet Jisa of Universite´ Lumie`re, Lyon 2, and for Spanish, with Judy Kupersmitt of Haifa University and Melina Aparici, Universidad de Barcelona. 13 Percent agreement is calculated by adding up the number of cases that received the same rating by both judges and dividing that number by the total number of cases rated by the two judges.

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

753

trained to reach between 90% to 95% agreement on division into clauses within a matter of a week or two. As for division into Clause Packages, the vast bulk of these were uncontroversial, since they were overtly marked by coordinating or subordinating conjunctions in all four languages. The remaining 10% or so that were either unmarked or introduced by discursively motivated rather than by syntactically and/or semantically conventional elements were the only source of disagreement in this respect, and here too, difficulties were resolved by consultation. Following Nir-Sagiv (2008), each Clause Package in the sample was treated as a ‘‘construction’’ in the sense of Construction Grammar (Fillmore et al., 1988; Goldberg, 1995; Kay and Fillmore, 1999) extended to the view that ¨ stman, 2001; O ¨ stman, discourse, too, can be considered part of the repertory of linguistic constructions (Halmari and O 2005). Here, a CP is a construction that takes the form of a complex of clauses revolving around the same discourse topic (as reviewed by Goutsos, 1997). Each CP consists minimally of a Main Clause, constituting the ‘‘head’’ to which the other clauses relate in one of the four linkage types specified earlier. This method of segmentation and analysis is illustrated in (8)–(11) below by typical sample texts in French, English, Hebrew, and by an excerpt from a text in Spanish. With the exception of the latter, all the texts share similar thematic content (interpersonal conflicts at a work place), with approximately the same number of Clause Packages. Clauses in each clause package are specified for syntactic architecture in terms of (a) its relation to adjacent clauses and (b) clause-type. Center-embedded clauses are enclosed in angled brackets.

(8) Narrative Written by French-speaking Woman [#17] CP1: CP2:

CP3: CP4:

CP5:

CP6:

CP7:

CP8:

CP9:

J’ai eu un conflit avec une personne dans le milieu professionnel [ISO]. ‘I had a conflict with a person at my place of work’ Je travaille dans un service export [ISO = MC] ou` la flexibilite´ entre les personnes est indispensable [HYPO = RC]. ‘I work in an export firm where flexibility between people is essential’ Nous devons nous remplacer lors des absences des uns et des autres [ISO]. ‘We have to fill in for one another in case one of us is away’ Une de mes colle`gues est partie une semaine en vacances [ISO = MC]. J’ai duˆ la remplacer [S-PAR = JUX] alors que son poste et le mien e´taient tre`s charge´s [HYPO = ADV]. ‘One of my colleagues went on vacation one week. I had to replace her while both her desk and mine were very heavily loaded’ Sur le poste de ma colle`gue j’ ai eu un proble`me informatique [ISO = MC] qu’ il e´tait urgent de re´soudre [RC-HYPO] pour pouvoir travailler [HYPO = ADV]. Cet incident est arrive´ a` un mauvais moment [S-PAR = JUX] puisque j’ e´tais de´borde´e [HYPO = ADV]. ‘At my colleague’s desk I had a computer problem that had to be urgently resolved so that I could work. This incident came at a bad time because I was drowning in work’ J’ai donc contacte´ la personne au service informatique [ISO = MC] qui e´tait suceptible de m’ aider [HYPO = RC]. ‘So I contacted the person at the computer center that was supposed to be able to help me. Malheureusement cette dernie`re e´tait elle aussi surcharge´e de travail [ISO = MC] et e´galement tre`s e´nerve´e [As-PAR = SSE]. Elle a donc refuse´ de m’ aider [S-PAR = JUX] et m’a par contre re´pe´te´e pendant dix minutes [As-PAR = SSE] qu’ elle e´tait surcharge´e de travail [As-PAR = CMP]. ‘Unfortunately, the latter was herself swamped with work and also very upset. She therefore refused to help me and instead complained repeatedly for ten minutes that she was swamped with work.’ Le fait [ENDO = NCMP] m’a davantage e´nerve´e [ISO = MC]. ‘The fact that she wasted ten minutes telling me her troubles made me even more upset’. Nous en sommes arrive´es a` une situation conflictuelles [sic.] [ISO = MC] qui a dure´ dix minutes [HYPO = RC] avant que nous parvenions a` un accord [HYPO = ADV]. ‘We reached a point of conflict that lasted for ten minutes before we could reach an agreement.’

754

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

The French text in (8) contains 24 clauses, packaged together into 9 CPs, with an average of 2.6 clauses per package. One-quarter of the clauses (6/24) are packaged by means of isotactic or symmetric paratactic relations with their surrounding clauses – with CP#1 and CP#3 each containing only one isolated clause. CP#5 and CP#7 are marked in bold to illustrate different strategies: both contain five clauses packaged together; CP#5 demonstrates a strategy of stacking dependent (hypotactic) clauses that are layered on top of one another; while CP#7 uses a strategy of paratactic ‘‘stringing’’ of clauses one after another. The text in (9), written in English, contains 41 clauses that are packaged into 10 CPs, with an average of 4.1 clauses per package. (9) Narrative Written by English-speaking Woman [#11] CP1: CP2:

CP3: CP4:

CP5: CP6: CP7: CP8:

CP9:

CP10:

As I stated just moments prior in an interview [HYPO = ADV], a story about a conflict [ENDO = RC] occurred recently [ISO = MC]. Within my research laboratory there are certain times [ENDO = RC], for example [ISO = MC] when one is under the time pressure of a grant submission deadline [ENDO = ADV]. It was during one of those crunch times [ISO = MC] when [=that] it became readily apparent [HYPO = NCMP] that our research staff was under a great deal of stress [As-PAR = CMP]. Concomitantly, groups of people began to snap at each other [ISO = MC], show labile mood [As-PAR = SSE], and/or have certain tones in conversation [As-PAR = SSE] that were found [HYPO = RC] to be condescending by others [As-PAR = CMP-NF]. In one instance, this type of scenario occurred between me and another lab colleague [ISO]. When alone in my office [HYPO = ADV], we both exchanged less than optimal tones with each other [ISO = MC] and the conflict or tension was all too evident [S-PAR = DS]. We decided [ISO = MC] that we should close the door [As-PAR = CMP] and discuss what [As-PAR = CMP] was going on [HYPO-RC]. Although it was difficult [HYPO-ADV] and I felt vulnerable in many ways [HYPO-ADV], it was helpful [ISO = MC] to talk things through [As-PAR = CMP-NF] and resolve the situation [As-PAR = CMP-NF]. Our resolution was [ISO = MC] to warn each other [As-PAR = CMP-NF] when feeling stressed out tired or irritable [HYPO-ADV-NF] and to assert the fact [As-PAR = CMP-NF] that this had nothing [ENDO-RC-NF], but all [HYPO-NCMP]. Since then our interactions have been much much better [ISO = MC] and during stressful periods we are much clearer about what [S-PAR = DS] is going on [HYPO-RC] without just snapping at each other [HYPO-ADV-NF] or allowing the pressure [HYPO-ADV-NF] to manifest itself in other hurtful ways [As-PAR = CMP-NF].

Relatively few of the clauses in the English text in (9) are isotactic or strung together by symmetric parataxis, only around 10% (5/41), compared with the French example in (8). Instead, around one-fifth of the clauses (9/41) are related by asymmetric stringing through complementation. In fact, nearly every Clause Package contains one or more complement clauses. Favored strategies are illustrated by the two CPs that end the story: CP#9 and CP#10 illustrate ‘‘stacking’’ types of strategies – by asymmetrically strung (complement) clauses and by dependent (hypotactic) clauses layered one on top of the other. In addition, no fewer than half of the clauses linked together in these two CPs are non-finite (e.g., to warn each other, when feeling . . ., and to assert, snapping, allowing, to manifest). The text in (10), written in Hebrew on the same topic of problems at the work place, contains 27 clauses packaged into 9 CPs, with an average of 3 clauses per package. As many as one-third of the CPs (3/9) take the form of lone isolating clauses. In general, many clauses are packaged in isotactic and paratactic relations, accounting for over onequarter of the total (9/27).

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

755

(10) Narrative Written by Hebrew-speaking Man [#13] CP1: CP2:

CP3:

CP4: CP5:

CP6:

CP7: CP8: CP9:

mikre sˇe-kara li im amit la-avoda [ISO] ‘(An) incident that-happened to-me with (a) colleague at-work’ le-orex kol ha-tkufa [ENDO = RC], ani margisˇ kin’a [ENDO = PAR] ve-ayin cara [ISO = MC], omnam loh rak klapay [As-PAR = GAP] gam klapey axerim [As-PAR = GAP] ‘Throughout the entire period, that-I work with-him, I feel envy and-ill will, although not only towards-me, towards others as well’. yom exad ha-menahel loh higia [ISO = MC] ve-hayiti amur lehaxlif oto [As-PAR = DS] ve-la’asot et avodato sˇel ha-menahel [As-PAR = SSE-NF] heyot ve-gam ani yodea ota [HYPO = ADV] ‘One day the-boss (did) not arrive, and- (I)-was-supposed to-replace him and-to-do the boss’s job, since I also know it’. hu ka’as meod alkax [ISO = MC] sˇe-ha-menahel loh higia [HYPO = NCMP] ‘He was-furious about (it) that-the-boss not arrived’. ax hu sovev et ha-inyan [ISO = MC] ve-ta’an [As-PAR = SSE] sˇe-ha-siba le-kax [ENDO = NCMP] neuca bi [As-PAR = CMP] misˇum sˇe-hirgalti et hamenahel [HYPO = ADV] sˇe-yesh lo maxlif dehaynu ani [As-PAR = CMP] ‘However, he twisted the matter around, and claimed that-the-reason for it (is)-occasioned by-me as I-(had) accustomed the-boss (to the fact that) he has a substitute, namely me’ hu higdil ve-amar [ISO = MC] sˇe-ani loh ose klum [As-PAR = CMP] ve-ani nehene lasˇevet ve-la’asot et avodato sˇel ha-menahel [As-PAR = CMP] ‘He went-far and said [=went so far as to say] that I do nothing and-I enjoy sitting and-doing the boss’s work’. ze hixis oti meod [ISO] ‘It annoyed me very-much’. hayu xilufey dvarim kasˇim, tonim gvohim [ISO] ‘(There) were harsh exchanges, loud tones (of voice)’. ve-levasof hixlateti [ISO = MC] sˇe-eyn yoter cxokim ito [As-PAR = CMP] ve-yaxasay imo yiyu inyaniyim bilvad [As-PAR = CMP] ve-kax yitav li [As-PAR = CMP] ‘And-eventually (I) decided that-(there’s) no more fun with-him, and-my-relations with-him will-be strictly business-like, and-thus I’ll benefit’.

The Hebrew text in (10) shows similar patterns and preferences to those illustrated by the French text in (8), although there are more complement clauses in Hebrew (7/27). On the other hand, CPs #2 and #5 illustrate ‘‘Hebrewtype’’ clause-packaging, since they demonstrate a typically Semitic strategy of paratactic stringing. Even where the text displays a more mixed syntactic architecture, this still reflects predominantly stringing strategies, using coordination where English prefers non-finite subordination. For example, the string ‘‘he twisted . . . and claimed’’ sounds rather odd in English, but constitutes not only acceptable, but preferred style in Hebrew! The last set of examples, in (11), illustrate two typical CPs taken from a personal-experience narrative written in Spanish.

(11) Excerpt from Narrative Written by a Spanish-speaking Woman [#15] CP1:

En lo [ENDO = RC] [ENDO = RC] y [ENDO = RC] tengo muy variadas [ISO = MC] pero no so´lo del ambiente educativo [As-PAR = SSE], sino tambie´n experiencias del ambiente social [S-PAR = DS] que no he vivido directamente [HYPO = RC] pero que han sido evidentes [HYPO = RC].

756

CP2:

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

‘As for and , I have many different ones, but not only in the educational domain, but also experiences in the social domain that I have not lived directly (myself) but that have been clear (to see).’ La primera experiencia [ENDO = RC] es un tipo de discriminacio´n intelectual [ISO = MC]. me refiero [S-PAR = JUX] a que el profesor nos etiquetaba a los alumnos por inteligentes o ma´s torpes [HYPO = NCMP] y lo demostraba [HYPO = NCMP] resaltando [HYPO = ADV] y valorando siempre a aquellos alumnos [HYPO = ADV] que acababan pronto sus actividades [HYPO = RC] y realizaban buenos exa´menes [HYPO = RC] y por el contrario desprestigiaba a los [HYPO = NCMP] que tenı´an dificultad [HYPO = RC] para resolver ciertos problemas de aprendizaje [HYPO = NCMP]. ‘The first experience is a kind of intellectual discrimination. I refer to it [=the fact] that the teacher would label us as smart or as clumsy, and he proved this (by) always highlighting and praising those students that completed their assignments quickly and (that) did well on tests, and on the other hand, he would deprecate those that had difficulty with solving certain learning problems’.

In marked contrast to the other three languages – where the mean length of a clause package ranged from around 2.5 to 4 – each CP in Spanish averages as high as ten clauses packaged together. And these typically show a highly intricate, layer-on-layer-on-layer type of syntactic architecture, in the kind of ‘‘interwoven patterning’’ that we found to be typical across the Spanish-language sample, even in the texts produced by school-age children (Berman and NirSagiv, 2009). Quantitative analyses show the different rhetorical means used for clause-packaging illustrated in texts (8) through (11) to be consistent across the populations in the different languages. Fig. 2 displays the mean number of clauses linked together in Clause Packages in each of the four languages. A Kruskal–Wallis Analysis of Variance test showed statistically significant differences in the average number of clauses packaged together as a function of language (x2 = 28.395, p < 0.001). Additional tests revealed a ranking of the four languages on this measure, with texts written in Spanish packaging together significantly more clauses compared to English (Mann–Whitney U = 35.5, p < 0.001 two-tailed), English packaging more clauses per Clause Package than French (Mann–Whitney U = 99.5, p < 0.05 two-tailed) or Hebrew (Mann–Whitney U = 86, p < 0.01 two-tailed), with no differences found between Hebrew and French. Thus, Spanish writers combine the largest number of clauses together in a single unit of narrative discourse. This finding is in line with differences that emerged between the languages in the type, as well as the amount, of clause-combining options that they tended to favor – as shown in Fig. 3. Fig. 3, depicting the internal constituency of these packages in each language, shows that in all four languages, layering of clauses by hypotaxis is a highly favored means of clause-linkage. On the other hand, the slices in these piefigures show that the cake divides up differently in the four languages in terms of ‘‘preferred clause-combining strategies’’. Hebrew relies significantly more on isolating clauses when compared to English (Mann–Whitney

Fig. 2. Mean number of clauses per Clause Package, by language [N = 74].

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

757

Fig. 3. Preferred strategies for Clause-Packaging in four languages.

U = 86.5, p < 0.01 two-tailed) and Spanish (Mann–Whitney U = 41.5, p < 0.001 two-tailed), but no significant difference was found between Hebrew and French. French is also more isotactic than English (Mann–Whitney U = 95.5, p < 0.05 two-tailed) and Spanish (Mann–Whitney U = 27, p < 0.001 two-tailed). Moreover, French shows the least reliance on layered packaging through adverbial and relative clauses, with significantly less use of hypotactic clauses compared to Spanish (Mann–Whitney U = 71, p < 0.01 two-tailed) and Hebrew (Mann–Whitney U = 102.5, p < 0.01 two-tailed). English favors asymmetric parataxis by complementation compared to Hebrew (Mann–Whitney U = 37, p < 0.001 two-tailed) and Spanish (Mann–Whitney U = 70, p < 0.05 two-tailed) with no difference between English and French; and Spanish speaker–writers tend to use heavily embedded hypotactic packaging of units of discourse into an endotactic nesting of adverbial and relative clauses one inside the other compared to English (Mann–Whitney U = 56.5, p < 0.01 two-tailed) and French (Mann–Whitney U = 22, p < 0.001 two-tailed), with no difference between Spanish and Hebrew. Spanish also appears to rely more heavily on symmetric paratactic stringing by juxtapositioning and coordination than the other languages. We interpret these findings as shedding light on the contrastive rhetoric of complex syntax, in the sense of the expressive preferences demonstrated by writers of different languages for how they package together the pieces of information in constructing narrative texts. Narrative rhetoric thus turns out to be more than a feature of linguistic structures per se. In terms of available linguistic devices, the languages in our sample share the same major types of coordinating constructions (with different subjects, pronominalizations, and same-subject ellipsis); complement clauses (in direct speech, with a complementizer like that, or in indirect questions); and subordinate adverbial and relative clauses. As a result, all four languages can and do combine clauses on each of the levels we defined, from isolating lone main clauses via paratactic stringing of clauses and on to nesting and even nested stacking of subordinate clauses. However, these shared resources are differently distributed and deployed across narrative texts constructed in each language. What we find, rather, are expressive preferences for which of their available structural options speaker– writers tend to favor for combining parts of the texts that they construct, from relatively flat stringing to greater reliance on interwoven patterning. 5. Discussion The analyses presented in this paper, and the ideas proposed in the following discussion, are most closely associated to the discipline of discourse studies. As such, they depart somewhat from the bulk of research on complex syntax in both linguistics and language acquisition, where concern is mainly with the expression of local semantic relations – temporal or logical – between adjacent clauses or groups of clauses.14 Our concern is, rather, with the strategies that speaker–writers employ when packaging together clauses as a means of instantiating text-embedded complex syntax. 14 These are defined by Tomasello in the following terms: ‘‘The linking of clauses – whether more loosely or more tightly – serves a variety of discourse functions, from expressing speaker attitudes about things (as in infinitival and sentential complements) to specifying referents in more detail (as in relative clauses), to indicating the spatial-temporal-causal interrelations among events (as in adverbial clauses)’’ (2003:242).

758

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

To this end, the study examines the nature of syntactic architecture in monologic narratives, relating to ‘‘contrastive rhetoric’’ by comparing how speaker–writers of different languages deploy syntactic packaging in the context of extended discourse. Our approach to the topic of contrastive rhetoric is thus typological, not cross-cultural (as, for example, in the recent collection of studies in Connor et al., 2008). Rather, our work derives from and aims to extend prior cross-linguistic research noted in section 1. The present study is thus congruent with research that applies analytical discourse models such as Rhetorical Structure Theory (RST) to uncover different rhetorical means used for expressing procedural relations in a text by writers of, say, English, French, and Portuguese (Scott et al., 1999). Specifically, our analysis is anchored in a cross-linguistic approach that has proved valuable for developmentally oriented research, as shown by Berman and Slobin’s work on oral narratives of preschool compared with school-age children and adults (1994, especially pp. 622–641) as well as in comparing narrative and expository discourse in later language development across adolescence (Berman, 2008). The value of such cross-linguistic research is perhaps most markedly demonstrated by discourse-based extensions of Talmy’s (1991, 2000) typological insights into the distinction between verb-framed and verb-satellite languages in the expression of motion events (e.g., Guo et al., 2009:121–292; Slobin, 2003a, 2004a). The potential contribution to contrastive discourse research of detailed, text-embedded cross-linguistic analysis of features of linguistic structure (in expression of motion events or, in our case, of clause-packaging) is highlighted by Slobin’s (1996, 2000, 2003b) investigations of what he terms ‘‘typological differences in narrative style’’ and the ‘‘discourse effects of linguistic typology’’ on the rhetorical style of novels, newspaper reports, conversations, and multilingual translations (2004b). Against this background, we attempt to explain some of the cross-linguistic differences revealed by our analysis of clause-packaging in personal-experience narratives produced on the same topic of discourse in four different languages. Particularly challenging in this respect is the marked contrast we found between the French favoring of more isolated and less layered constructions compared with the layering and nesting types of clause packages in Spanish, also a Romance language. That this is not a mere artifact of our data-base or methodology is shown by the fact that wide reliance on paratactic constructions was also found in other studies of French discourse: in the oral narratives of young children (Jisa, 1987), in written narratives of school-age children (Schneuwly et al., 1989), and in an extensive sample of narrative and expository, both oral and written, texts elicited in the framework of the larger crosslinguistic project referred to in section 3 (Viguie´, 2001). Earlier cross-linguistic analyses of extended discourse suggest that variation between languages may depend on language typology not only in the sense of the structural means available to speaker–writers of a given language, but also in terms of the expressive options that they favor when constructing discourse. That is, contrastive rhetoric in the sense used here derives from a complex interplay between the repertoire of linguistic constructions in a given target language, on the one hand, and rhetorical preferences governing how speaker–writers of the language select to deploy and alternate these structures to meet particular discourse functions, on the other. Explanations for the distribution of clause-linking devices that have been proposed from the perspective of contrastive rhetoric typically appeal to notions of rhetorical style and purpose (Grosz and Sidner, 1986; Klein and von Stutterheim, 1991). For example, Johnstone (1996) suggests that clarity in speech involves explicitly indicating relationships among propositions, by such means as subordination and coordination to express relationships between ideas and parallelism for lining up ideas with equal weight. Relatedly, Levy’s (2003) work on children’s narrative retellings suggests that increased use of clause-linking devices such as subordinate and embedded clauses is a mark of compacting or compressing processes corresponding to increased textual cohesion. However – in line with Dan Slobin’s (1996, 2003b) cross-linguistic perspective on ‘‘thinking for speaking’’ and ‘‘thinking for writing’’ – a contrastive analysis like ours suggests that to consider subordination or embedding as clearer or more cohesive – and hence, somehow, ‘‘better than’’ juxtaposing or coordinating – may derive from a monolingual or developmentally motivated bias. For example, in earlier studies, non-finite subordination in English was more common in adult than in children’s usage both in oral picture-book narratives (Berman, 1998) as well as in data-bases similar to the one used here (Kupersmitt, 2006; Berman and Nir-Sagiv, 2009). Rather, we propose that what governs the choices made by speaker–writers for expressing a given discursive function, in this case of clause-packaging, is the availability of other, highly accessible alternative means of expression for achieving the same discourse purpose. Such an interplay between structural options and rhetorical preferences was earlier demonstrated for children’s oral narratives in five languages for expression of event construal (Berman and Slobin, 1994) and discourse connectivity (Berman, 1998). And it has also been shown for the larger cross-linguistic

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

759

project underlying the present study in the expression of temporality in narrative as compared with expository texts written in English, Hebrew, and Spanish (Kupersmitt, 2006). This interplay between available forms and those preferred for a particular discursive function is perhaps most dramatically demonstrated by comparison of the means selected by writers of different languages for expressing a more detached or impersonal ‘‘discourse stance’’ (Berman et al., 2002). In terms of structurally determined typological factors, studies examining expository texts written in different languages show a clear divide in the favored means for agency downgrading in languages tolerant of subjectless constructions like Hebrew and Spanish compared with languages that require a surface subject like English and French (Berman, 2005). Hebrew relies on subjectless 3rd person plurals, subjectless modal operators, and generic 2nd person subject, or on use of intransitive, middle-voice verb morphology (Berman, 1979, 1993), far less frequently, on passive voice constructions (Berman, 2004). In Spanish, means for expressing a depersonalized discourse stance include, in addition to a generic 2nd person pronoun subject, reliance on se marked subjectless impersonals and, less commonly, se marked middle and passive voice constructions and periphrastic passives with a non-agentive subject (Tolchinsky and Rosado, 2005). In contrast, English speaker–writers use passive voice constructions or expletive pronouns it, there and generic subjects such as people alternating with generic pronouns like you, they, one used in what appear to be rather haphazard ways (Reilly et al., 2005). English differs from other subjectrequiring languages in the sample in not having a dedicated generic pronoun such as French on (Jisa, 2004b; Jisa and Viguie´, 2005) and the analogous Swedish item man (Ragnarsdo´ttir and Stro¨mqvist, 2005). Importantly, from the point of view of the interplay between what is structurally available and what is rhetorically favored, reliance on passive voice for expressing a less involved, non-agentive discourse stance revealed clear crosslinguistic distinctions across parallel samples: Icelandic speaker–writers made greater use of passive voice compared with their counterparts in the closely related Swedish (Ragnarsdo´ttir and Stro¨mqvist, 2005), while Dutch and English made far greater use of passive constructions as against Hebrew and Spanish, with French lying somewhere in the middle (Jisa et al., 2002). These differences in distribution of use of passive voice for agency downgrading in different languages are noteworthy since, just as all four languages in the sample analyzed in the present study share much the same repertoire of structural options for clause-packaging so, analogously, do all the seven languages in the larger cross-linguistic sample have structurally productive passive voice constructions. We explain the differences in deployment of these constructions for agency downgrading, rather, by the availability of alternative rhetorical options for expressing this same discourse function: a dedicated generic pronoun in subject-requiring languages (French on, Swedish man), and highly accessible, early acquired subjectless impersonals and middle voice in Spanish (with se) and in Hebrew (with 3rd personal plural verbs and intransitive verb morphology respectively). We propose that an analogous explanation may throw light on our findings for differences in (narrative) clausepackaging between the two Romance languages in our sample. Baron and Herslund (2005), for example, suggest that French is more ‘‘exocentric’’ than a language like English or Danish. Relevant differences are also revealed by Korzen’s (2005) comparisons of manner and circumstantial attributes in these three languages. While Bickel (2006) suggests that many apparent differences in clause-linkage are not necessarily due to differences in syntax per se, but derive, rather, from their interaction with other typological factors such as choice of affixal versus conjunctional marking for linkage, choice of level of verb finiteness and the deployment of referring devices. These observations point to possible directions for explaining the relative ‘‘flatness’’ of clause-packaging we found in French, indicating the need for further investigation of competing means preferred by speaker–writers of French for achieving textual connectivity in their language. In the case of Hebrew, we suggest that the penchant for paratactic stringing is explainable along two lines. One is the existence of a rich range of nominalized constructions in the language (Berman, 1976, 2004; Seroussi, 2004) and the generally very ‘‘nominal’’ character of Hebrew discourse (Ravid and Berman, 2009). This involves, for example, use of verbless copular clauses and extensive verb-gapping of the kind illustrated in the Hebrew narrative text in (10) above, along with, most relevantly to the topic of analysis here, a tendency to use nominalized forms in favor of finite verbs and predicative adjectives for achieving textual cohesiveness and density (Ravid and Cahana-Amitay, 2005; Kupersmitt and Laks, 2007). A second factor promoting reliance on nominalized alternatives in preference to nonfinite subordination of the kind common in English and of mainly finite subordinate clauses in Spanish reflects the tendency of Hebrew speaker–writers to adhere to classically Semitic options in their written usage, by use of parallel equivalence structures typical of Biblical narrative rhetorical style (Rubinstein, 1980; Waltke and O’Connor, 1993; Goldfajn, 1998; Polak, 1998) and Arabic (Johnstone, 1987; Ostler, 1987). Independent analyses of both the current sample and of the oral picture-book narratives revealed extensive use of non-finite subordination by participials and

760

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

gerunds in English compared with Hebrew, although Hebrew, too, has structurally productive analogous constructions.15 Berman (1998) relates to this as a favored strategy for particularly close knitting together of two or more events in English, in marked preference to coordination by same-subject ellipsis favored in Hebrew, while Kupersmitt (2006) notes the same finding for an extended sample of the data-base used here as a favored stylistic means for English speaker–writers for expressing the temporal flow of events in narrative discourse. Observations in the disparate domains of discursive temporality and agency downgrading for expression of an impersonal discourse stance indicate that typologically motivated cross-linguistic analyses provide an important basis for discourse-based contrastive rhetoric. We have suggested, however, that the domain needs to go beyond obligatory aspects of the grammar of the target languages to explain rhetorical preferences shown by speaker–writers in the course of text construction. This proposal is strongly supported by the differences that emerged in clause-packaging syntactic architecture in the narratives we examined, since all four languages, as noted, share much the same structural devices for this purpose: same- and different-subject coordination, and complement, adverbial, and relative clauses. Just as the ready availability of other, highly accessible means of expression for achieving an impersonalized discourse stance (by subjectless impersonals and middle voice in Hebrew and Spanish and of dedicated generic pronouns in French and Swedish) explain the relative avoidance of passive voice for agency downgrading in these languages compared with Dutch and English, so, too, for purposes of clause-packaging, English can rely on a rich range of nonfinite participials and gerunds and Hebrew on nominalizations. Finally, we suggest that rhetorical preferences interact not only with what Berman (1986, 2002) terms ‘‘typological imperatives’’ in the sense of grammatical features of the language that are both obligatory and pervasive, but also with various other discursive factors in determining what forms of expressions will be favored or avoided by speaker– writers of different languages. Two such factors that are particularly relevant to the data-base under consideration here are modality (spoken versus written discourse) and genre (here, personal-experience narratives compared with expository discussions). Findings from our data-base, mainly for English and Hebrew, in domains other than clausepackaging, indicate that native speaker–writers are not only peculiarly attuned to the ‘‘spirit of the language’’ when producing a piece of discourse, they are also attentive to the way rhetorical style is constrained by communicative context and discourse genre. Modality-driven distinctions relevant to clause-packaging are suggested for French by Labbe´’s (2003) study showing that texts written in French contain fewer subordinating than coordinating conjunctions compared with data from speech.16 The relative monoticity or flatness of French might be the result of the impact of conventions of written style inculcated from an early age in the French school system, to date considered largely in terms of particular linguistic devices or constructions rather than of extended textual connectivity (Viguie´ and Jisa, 1999; Jisa, 2004a,b). This clearly needs to be considered in relation to other well-documented comparisons of written and spoken French (e.g., Blanche-Benveniste, 1994, 1995). In relation to the factor of genre, work in progress on expository essays produced in the cross-linguistic project that formed the framework for the present analysis shows that rather different trends emerge for clause-combining syntactic architecture in non-narrative discourse. For example, in expository essays, Hebrew and English speaker– writers favor hypotaxis to much the same extent (24% and 21% respectively), but the Hebrew group relies more on asymmetric parataxis (24%) than their English-speaking peers (18%). And an independent analysis of the French database (Viguie´, 2001) revealed that, compared to narratives produced by the same people, expository texts contain less symmetric parataxis (by different- and same-subject coordination) and more hypotactic linking (e.g., by circumstantial adverbials and relative clauses). These observations from parallel samples of expository essays compared with the personal-experience narratives analyzed in the present study point to the impact of text type and the intersection of modality and genre in expression of inter-clausal connectivity. This accords with findings from work on the rhetorical structure of text and on speaker– writers’ knowledge about a particular genre from the point of view of theories of discourse structure such as RST 15

In fact, university level Hebrew-speaking students with an excellent command of English typically over-use and constructions to combine clauses where a non-finite form with –ing would be stylistically more appropriate to express an English-like cohesiveness. 16 Labbe´ examined over 300 conversations over 2000 written texts in Modern French. The French example given by Slobin (2004a:121) from a translation of Tolkien’s The Hobbit further illustrates a marked reliance on clause ‘‘stringing’’: Il continua d’avancer au hasard, sortit du haut vallon, en franchit le bord et descendit la pente au-dela` ‘He continued to advance aimlessly, exited from the high small valley, crossed the edge of it and descended the slope beyond’.

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

761

(Taboada and Mann, 2006). For example, Benwell’s (1999) analysis of spoken tutorial discourse in physics compared with English literature revealed that the two disciplines structure knowledge differently – as embedded relations in physics and as coordinated in literature. In sum, the cross-linguistic findings revealed for narrative connectivity in the present study appear to be best explained by a complex interplay between factors of linguistic typology and rhetorical preferences, as further constrained by different discourse genres and communicative settings. This claim echoes Berman’s earlier (1978) insight to the effect that any exercise in contrastive analysis needs to consider different types of systematic differences between languages – at the time couched in terms of obligatory versus genetic or idiosyncratic differences – to explain contrasts in rhetorical style. What current research indicates is that contrastive analysis can valuably be extended to ‘‘rhetorical typology’’, as a usage-based, text-embedded domain of linguistic inquiry. Acknowledgements This paper is based on a talk presented to the conference on Language, Culture, and Mind II: Integrating Perspectives and Methodologies in the Study of Language, held in Paris, July 17–20, 2006. The authors are indebted to participants in the conference and to the journal editor and two anonymous reviewers for their helpful comments and queries. References Akinci, Mehmet Ali, Jisa, Harriet, 2000. Development of Turkish clause linkage in the narrative texts of Turkish–French bilingual children in France. In: Go¨ksel, A., Kerslake, C. (Eds.), Studies in Turkish and Turkic Languages: Proceedings of the 9th International Conference on Turkish Linguistics, Oxford University. Harrassowitz Verlag, pp. 317–324. Asher, Nicholas, 1993. Reference to Abstract Objects in Discourse. Kluwer, Dordrecht. Asher, Nicholas, Lascarides, Alex, 1995. Lexical disambiguation in a discourse context. Journal of Semantics 12, 69–108. Bar-Lev, Zeev, 1986. Discourse theory and contrastive rhetoric. Discourse Processes 9, 235–246. Baron, Ire`ne, Herslund, Michael, 2005. Langues endocentriques et langues exocentriques. Approche typologique du danois, du franc¸ais et de l’anglais. Langue Franc¸aise 145, 35–54. Benwell, Bethan, 1999. The organisation of knowledge in British university tutorial discourse: issues, pedagogic discourse strategies and disciplinary identity. Pragmatics 9 (4), 535–565. Berman, Ruth A., 1976. On derived and deriving nominals in Modern Hebrew. In: Cole, P. (Ed.), Studies in Modern Hebrew Syntax and Semantics. North-Holland, Amsterdam, pp. 57–98. Berman, Ruth A., 1978. Contrastive analysis revisited: obligatory, systematic, and incidental differences between languages. Interlanguage Studies Bulletin 3, 212–233. Berman, Ruth A., 1979. Passives, middles, and impersonals in Modern Hebrew. Berkeley Linguistic Society 5, 1–27. Berman, Ruth A., 1986. The acquisition of morphology/syntax: a crosslinguistic perspective. In: Fletcher, P., Garman, M. (Eds.), Language Acquisition. Cambridge University Press, Cambridge, pp. 429–447. Berman, Ruth A., 1988. On the ability to relate events in narratives. Discourse Processes 11, 469–497. Berman, Ruth A., 1990. Acquiring an (S)VO language: subjectless sentences in children’s Hebrew. Linguistics 28, 1135–1166. Berman, Ruth A., 1993. Developmental perspectives on transitivity: a confluence of cues. In: Levy, Y. (Ed.), Other Children, Other Languages. Lawrence Erlbaum, Hillsdale, NJ, pp. 189–241. Berman, Ruth A., 1995. Narrative competence and storytelling performance: how children tell stories in different contexts. Journal of Narrative and Life History 5 (4), 285–313. Berman, Ruth A., 1996. Form and function in developing narrative abilities. In: Slobin, D.I., Gerhardt, J., Kyratzis, A., Gus, J. (Eds.), Social Interaction, Social Context and Language. Lawrence Erlbaum, Hillsdale, NJ, pp. 343–367. Berman, Ruth A., 1998. Typological perspectives on connectivity. In: Dittmar, N., Penner, Z. (Eds.), Issues in the Theory of Language Acquisition. Peter Lang, Bern, pp. 203–224. Berman, Ruth A., 2002. First language acquisition: crosslinguistic. In: Smelser, N.J., Baltes, P.B. (Eds.), International Encyclopedia of Social and Behavioral Sciences. Pergamon Press, Oxford, pp. 5681–5686. Berman, Ruth A., 2004. Between emergence and mastery: the long developmental route of language acquisition. In: Berman, R.A. (Ed.), Language Development Across Childhood and Adolescence. John Benjamins, Amsterdam, pp. 9–34. Berman, Ruth A., 2005. Introduction: developing discourse stance in different text types and languages. Journal of Pragmatics 37 (2), 105–124. Berman, Ruth A., 2008. The psycholinguistics of developing text construction. Journal of Child Language 35, 735–771. Berman, Ruth A., Katzenberger, Irit, 2004. Form and function in introducing narrative and expository texts: a developmental perspective. Discourse Processes 38, 57–94. Berman, Ruth A., Nir-Sagiv, Bracha, 2007. Comparing narrative and expository text construction across adolescence: a developmental paradox. Discourse Processes 43 (2), 79–120.

762

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

Berman, Ruth A., Nir-Sagiv, Bracha, 2009. Clause-packaging in narratives: a crosslinguistic developmental study. In: Guo, J., Lieven, E., Budwig, ¨ zc¸alis¸kan, S. (Eds.), Crosslinguistic Approaches to the Psychology of Language: Research in the Tradition N., Ervin-Tripp, S., Nakamura, K., O of Dan I. Slobin. Lawrence Erlbaum, Mahwah, NJ, pp. 149–162. Berman, Ruth A., Ragnarsdo´ttir, Hrafnhildur, Stro¨mqvist, Sven, 2002. Discourse stance. Written Language and Literacy 5 (2), 255–290. Berman, Ruth A., Ravid, Dorit, 2009. Becoming a literate language user: oral and written text construction across adolescence. In: Olson, D.R., Torranced, N. (Eds.), Cambridge Handbook of Literacy. pp. 92–111. Berman, Ruth A., Slobin, Dan I., 1994. Relating Events in Narrative: A Crosslinguistic Developmental Study. Lawrence Erlbaum, Hillsdale, NJ. Berman, Ruth A., Verhoeven, Ludo, 2002. Cross-linguistic perspectives on the development of text production abilities: speech and writing. Written Language and Literacy 5 (1), 1–43. Bickel, Balthasar, 2006. Towards a multivariate typology of clause linkage, with particular reference to non-embedded structures. In: Symposium Lenca-3, Tomsk, Russia. Blanche-Benveniste, Claire, 1994. The construct of oral and written language. In: Verhoeven, L. (Ed.), Functional Literacy: Theoretical Issues and Educational Implications. John Benjamins, Amsterdam, pp. 61–74. Blanche-Benveniste, Claire, 1995. De la rarete´ de certains phe´nome`nes syntaxiques en franc¸ais parle´. French Language Studies 5, 17–29. Bybee, Joan, 2002. Main clauses are innovative, subordinate clauses are conservative: consequences for the nature of constructions. In: Bybee, J., Noonan, M. (Eds.), Complex Sentences in Grammar and Discourse: Essays in Honor of Sandra A. Thompson. John Benjamins, Amsterdam, pp. 1–17. Bybee, Joan, Noonan, Michael (Eds.), 2002. Complex Sentences in Grammar and Discourse: Essays in Honor of Sandra A. Thompson. John Benjamins, Amsterdam. Chafe, Wallace L., 1994. Discourse, Consciousness and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago University Press, Chicago. Chomsky, Noam, 1957. Syntactic Structures. The Hague, Mouton. Chomsky, Noam, 1965. Aspects of the Theory of Syntax. MIT Press. Connor, Ulla, 2002. New directions in contrastive rhetoric. TESOL Quarterly 36 (4), 493–510. Connor, Ulla, Nagelhout, Ed, Rozycki, William (Eds.), 2008. Contrastive Rhetoric: Reaching to Intercultural Rhetoric. John Benjamins, Amsterdam. Couper-Kuhlen, Elizabeth, 1989. Foregrounding and temporal relations in narrative discourse. In: Schopf, A. (Ed.), Essays on Tensing in English, Vol. II: Time, Text and Modality. Max Niemeyer, Tu¨bingen, pp. 7–29. Cristofaro, Sonia, 2003. Subordination. Oxford University Press, Oxford. Croft, William, 1990. Typology and Universals. Cambridge University Press, Cambridge. Degand, Liesbeth, 1998. On classifying connectives and coherence relations. In: Proceedings of Coling/ACL Workshop on Discourse Relations and Discourse Markers, Montreal, Canada, pp. 29–35. Diessel, Holger, 2004. The Acquisition of Complex Sentences. Cambridge University Press, Cambridge. Du Bois, John W., 2003. Dialogic Syntax: The Syntax of Engagement. Lecture Series: Stance and Dialogic Syntax in Interaction. Oulu University. Fabricius-Hansen, Cathrine, 1999. Information packaging and translation. Aspects of translational sentence splitting (German–English/Norwegian). In: Doherty, M.M. (Ed.), Sprachspezifische Aspekte der Informationsverteilung. Akademie-Verlag, Berlin, pp. 175–213. Fabricius-Hansen, Cathrine, 2005. Coordination, discourse relations, and information packaging—cross-linguistic differences. Reports of the project Languages in Contrast. Fillmore, Charles J., Kay, Paul, O’Connor, Mary Catherine, 1988. Regularity and idiomaticity in grammatical constructions: the case of let alone. Language 64, 501–538. Fix, Ulla, Gardt, Andreas, Knape, Joachim (Eds.), 2006. Rhetorik und Stilistik. Ein internationales Handbuch historischer und systematischer Forschung. Mouton de Gruyter, Berlin/New York. Foley, William A., Van Valin, Robert D., 1984. Functional Syntax and Universal Grammar. Cambridge University Press, Cambridge. Fox, Barbara A., 1987. Discourse Structure and Anaphora: Written and Conversational English. Cambridge University Press, Cambridge. Frank, Annette, 2002. A (discourse) functional analysis of asymmetric coordination. In: Butt, M., King, T. (Eds.), Proceedings of the LFG02 Conference. pp. 174–196. Gibson, Edward, 1998. Linguistic complexity: locality of syntactic dependencies. Cognition 68 (1), 1–76. Givo´n, Talmy, 1980. The binding hierarchy and the typology of complements. Studies in Language 4 (3), 333–337. Givo´n, Talmy, 1984. Syntax. A Functional-Typological Introduction. Benjamins, Amsterdam (2 volumes). Givo´n, Talmy, 2009. The Genesis of Syntactic Complexity. John Benjamins, Amsterdam. Goldberg, Adele E., 1995. Constructions: A Construction Grammar Approach to Argument Structure. Chicago University Press, Chicago. Goldfajn, Tal, 1998. Word Order and Time in Biblical Hebrew Narrative. Oxford University Press, Oxford. ¨ zc¸alis¸kan, Seyda (Eds.), 2009. Crosslinguistic Approaches Guo, Jiansheng, Lieven, Elena, Budwig, Nancy, Ervin-Tripp, Susan, Nakamura, Keiko, O to the Psychology of Language: Research in the Tradition of Dan I. Slobin. Lawrence Erlbaum, Mahwah, NJ. Goutsos, Dionysis, 1997. Modeling Discourse Topic: Sequential Relations and Strategies in Expository Text. Ablex, Norwood, NJ. Grodner, Daniel, Gibson, Edward, Watson, Duane, 2005. The influence of contextual contrast on syntactic processing: evidence for stronginteraction in sentence comprehension. Cognition 95, 275–296. Grosz, Barbara, Sidner, Candace, 1986. Attention, intentions, and the structure of discourse. Computational Linguistics 12 (3), 175–204. Haiman, John, Thompson, Sandra A. (Eds.), 1988. Clause Combining in Grammar and Discourse. Amsterdam, John Benjamins. Halliday, Michael A.K., 1994. An Introduction to Functional Grammar. Edward Arnold, London. ¨ stman, Jan-Ola, 2001. The soft-spoken, angelic pickax killer: the notion of discourse pattern in controversial news reporting. Halmari, Helena, O Journal of Pragmatics 33, 805–823.

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

763

Heurley, Laurent, 1997. Processing units in written texts: paragraphs or information blocks? In: Costerman, J., Fayol, M. (Eds.), Processing Interclausal Relationships, Studies in Production and Comprehension of Text. Lawrence Erlbaum, Amsterdam, pp. 179–200. Hickmann, Maya, 2003. Children’s Discourse: Person, Space and Time Across Languages. Cambridge University Press, Cambridge. Hobbs, Jerry R., 1979. Coherence and Coreference. Cognitive Science 3 (1), 67–90. Ho¨hle, Tilman N., 1990. Assumptions about asymmetric coordination in German. In: Mascaro´, J., Nespor, M. (Eds.), Grammar in Progress: Glow Essays for Henk van Riemsdijk. Foris, Dordrecht, pp. 221–235. Hunt, Kellog W., 1970. Syntactic Maturity in School Children and Adults. Monographs for the Society for Research in Child Development, vol. 35, no. 1. Chicago University Press, Chicago. Hwang, Shin Ja J., 1989. Recursion in the paragraph as a unit of discourse development. Discourse Processes 12, 461–477. Jespersen, Otto, 1924. The Philosophy of Grammar. George Allen and Unwin, London. Jisa, Harriet, 1987. Sentence connectors in French children’s monologue performance. Journal of Pragmatics 11, 607–621. Jisa, Harriet, 2004a. Growing into academic French. In: Berman, R.A. (Ed.), Later Language Development: Typological and Psycholinguistic Perspectives, Trends in Language Acquisition Research (TiLAR). John Benjamins, Amsterdam, pp. 135–161. Jisa, Harriet, 2004b. Developing alternatives for indicating discourse stance. In: Ravid, D., Shyldkort, H. (Eds.), Perspectives on Language and Language Development: Essays in Honor of Ruth A. Berman. Kluwer Academic Publishers, Dordrecht, pp. 357–374. Jisa, Harriet, Reilly, Judy S., Verhoeven, Ludo, Baruch, Elisheva, Rosado, Elisa, 2002. Passive voice constructions in written texts: a cross-linguistic developmental study. Written Language and Literacy 5, 163–181. Jisa, Harriet, Viguie´, Anne, 2005. A developmental perspective on the role of ‘on’ in written and spoken expository texts in French. Journal of Pragmatics 37, 125–142. Johnstone, Barbara, 1987. Parataxis in Arabic: modification as a model for persuasion. Studies in Language 11, 85–98. Johnstone, Barbara, 1996. The Linguistic Individual. Oxford University Press, New York. Kamp, Hans, Reyle, Uwe, 1993. From Discourse to Logic. Kluwer, Dordrecht. Kaplan, Robert B., 1966. Cultural thought patterns in intercultural education. Language Learning 16, 1–20. Kay, Paul, Fillmore, Charles, 1999. Grammatical constructions and linguistic generalizations: the what’s X doing Y? Construction. Language 75, 1– 33. Klein, Wolfgang, von Stutterheim, Christiane, 1991. Text structure and referential movement. Sprache und Pragmatik 22, 1–32. Knott, Alistair, Sanders, Ted, 1998. The classification of coherence relations and their linguistic markers: an exploration of two languages. Journal of Pragmatics 30 (2), 135–175. Knott, Alistair, Oberlander, Jin, O’Donnell, Michael, Mellish, Chris, 2001. Beyond elaboration: the interaction of relations and focus in coherent text. In: Sanders, T., Schilperoord, J., Spooren, W. (Eds.), Text Representation: Linguistic and Psycholinguistic Aspects. John Benjamins, Amsterdam, pp. 181–196. Koch, Peter, 1995. Subordination, inte´gration syntaxique et ‘‘oralite´’’. In: Andersen, H.L., Skytte, G. (Eds.), La subordination dans les langues romanes, E´tudes Romanes n8 34. Museum Tusculanum, Copenhague, pp. 13–42. Korzen, Hanne, 2005. Attributs directs et indirects en franc¸ais, en danois et en anglais: diffe´rences typologiques et proble`mes de traduction. Langue Franc¸aise 145, 55–72. Kupersmitt, Judy, 2006. Temporality in texts: a crosslinguistic study of form-function relations in narrative and expository discourse. PhD Dissertation. Bar-Ilan University. Kupersmitt, Judy, Laks, Lior, 2007. Missing categories in L1 and L2: between verbs and nouns in narrative and expository texts. Paper presented at the 10th International Pragmatics Conference, Gotenborg. Labbe´, Dominique, 2003. Coordination et subordination en franc¸ais oral. ERLA. Universite´ de Bretagne Occidentale, Brest. Lehmann, Christian, 1988. Towards a typology of clause linkage. In: Haiman, J., Thompson, S.A. (Eds.), Clause Combining in Grammar and Discourse. John Benjamins, Amsterdam, pp. 181–225. Lehmann, Winfred P. (Ed.), 1978. Syntactic Typology: Studies in the Phenomenology of Language. University of Texas Press, Austin. Levy, Elena T., 2003. The roots of coherence in discourse. Human Development 46, 169–188. Lidz, Jeffrey, 2007. The abstract nature of syntactic representations: consequences for a theory of learning. In: Hoff, E., Shatz, M. (Eds.), Handbook of Language Development. Blackwell, London, pp. 277–303. Lust, Barbara C., Foley, Claire, Dye, Cristina, 2008. The first language acquisition of complex sentences. In: Bavin, E. (Ed.), Handbook of Child Language. Cambridge University Press, Cambridge, pp. 463–505. Lyons, John, 1968. Introduction to Theoretical Linguistics. Cambridge University Press, Cambridge. MacWhinney, Brian, 2000. The CHILDES Project: Tools for Analyzing Talk, 3rd edition. Lawrence Erlbaum, Mahwah, NJ. Mann William, C., Thompson, Sandra A., 1988. Rhetorical structure theory: toward a functional theory of text organization. Text 8, 243–281. Matthiessen, Christian M.I.M., 2002. Combining clauses into clause complexes: a multi-faceted view. In: Bybee, J., Noonan, M. (Eds.), Complex Sentences in Grammar and Discourse: Essays in Honor of Sandra A. Thompson. John Benjamins, Amsterdam, pp. 235–319. Neuendorf, Kimberly A., 2002. The Content Analysis Guidebook. Sage Publications, Thousand Oaks, CA. Nir-Sagiv, Bracha, 2008. Clause packages as constructions in the development of narrative discourse. PhD Dissertation. Tel Aviv University. Ohori, Toshio, 1994. Diachrony of clause linkage: TE and BA in Old through Middle Japanese. In: Paliugca, W. (Ed.), Perspectives on Grammaticalization. John Benjamins, Amsterdam, pp. 135–149. Ostler, Shirley E., 1987. English in parallels: a comparison of English and Arabic prose. In: Connor, U., Kaplan, R. (Eds.), Writing Across Languages: Analysis of L2 Texts. Addison Wesley, Reading, MA, pp. 169–185. Ostler, Shirley E., 1988. A study in the contrastive rhetoric of Arabic, English, Japanese, and Spanish. Doctoral Dissertation. University of Southern California.

764

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

¨ stman, Jan-Ola, 2005. Construction discourse: a prolegomenon. In: O ¨ stman, J.-O., Fried, M. (Eds.), Construction Grammars: Cognitive Grounding O and Theoretical Extensions. John Benjamins, Amsterdam, pp. 121–144. Peterson, Carole, McCabe, Alyssa, 1991. Linking children’s connective use and narrative macrostructure. In: McCabe, A., Peterson, C. (Eds.), Developing Narrative Structure. Lawrence Erlbaum, Hillsdale, NJ, pp. 29–53. Polak, Frank H., 1998. The oral and the written: syntax, stylistics, and the development of Biblical prose narrative. JANES 26, 59–105. Ragnarsdo´ttir, Hrafnhildur, Stro¨mqvist, Sven, 2005. The development of generic maður/man for the construction of discourse stance in Icelandic and Swedish. Journal of Pragmatics 37, 143–156. Raible, Wolfgang, 2001. Linking clauses. In: Haspelmath, M., Ko¨nig, E., Oesterreicher, W., Raible, W. (Eds.), Language Typology and Language Universals. Mouton de Gruyter, Berlin/New York, pp. 590–617. Ravid, Dorit, Berman, Ruth A., 2006. Information density in the development of spoken and written narratives in English and Hebrew. Discourse Processes 41, 117–149. Ravid, Dorit, Berman, Ruth A., 2009. Developing noun phrase complexity across adolescence: A text embedded analysis. First Language. Ravid, Dorit, Cahana-Amitay, Dalia, 2005. Verbal and nominal expression in narrating conflict situations in Hebrew. Journal of Pragmatics 37, 157– 183. Reilly, Judy S., Zamora, Anita, McGivern, Robert F., 2005. Acquiring perspective in English: the development of stance. Journal of Pragmatics 37, 185–208. Reinhart, Tanya, 1995. From text to meaning: strategies of evaluation. In: Shen, Y. (Ed.), Cognitive Aspects of Narrative Structure. Porter Institute, Tel Aviv University, (in Hebrew), pp. 4–37. Rubinstein, Eliezer, 1980. ha–ivrit shelanu ve–ha–ivrit ha–keduma. (Contemporary Hebrew and Ancient Hebrew).University on the Air, Tel Aviv. Sanders, Ted J.M., Spooren, Wilbert P.M., Noordman, Leo G.M., 1992. Toward a taxonomy of coherence relations. Discourse Processes 15, 1–35. Schneuwly, Bernard, Rosat, Marie-Claude, Dolz, Joaquim, 1989. Les organisateurs textuels dans quatre types de textes e´crits (e´le`ves de 10, 12 et 14 ans). Langue Franc¸aise 81, 40–58. Schiffrin, Deborah, 1987. Discourse Markers. Cambridge University Press, Cambridge. Scott, Donia, Delin, Judy, Hartley, Anthony, 1999. Identifying congruent pragmatic relations in procedural texts. Languages in Contrast 1 (1), 45–82. Seroussi, Batia, 2004. Hebrew derived nouns in context: a developmental perspective. Folia Phoniatrica et Logopaedica 56, 273–290. Shopen, Timothy (Ed.), 2007. Language Typology and Syntactic Description. 2nd edition. Cambridge University Press, Cambridge (3 volumes). Slobin, Dan I., 1973. Cognitive prerequisites for the development of grammar. In: Ferguson, C.A., Slobin, D.I. (Eds.), Studies of Child Language Development. Holt, Rinehart, Winston, New York, pp. 175–208. Slobin, Dan I., 1977. Language change in childhood and history. In: Macnamara, J. (Ed.), Language Learning and Thought. Academic, NY, pp. 185– 214. Slobin, Dan I., 1996. From ‘thought and language’ to ‘thinking for speaking’. In: Gumperz, J.J., Levinson, C. (Eds.), Rethinking Linguistic Relativity. Cambridge University Press, Cambridge, pp. 70–96. Slobin, Dan I., 2000. Verbalized events: a dynamic approach to linguistic relativity and determinism. In: Niemeir, S., Dirven, R. (Eds.), Evidence for Linguistic Relativity. John Benjamins, Amsterdam, pp. 107–138. Slobin, Dan I., 2001. Form-function relations: how do children find out what they are? In: Bowerman, M., Levinson, S.C. (Eds.), Language Acquisition and Conceptual Development. Cambridge University Press, Cambridge, pp. 406–449. Slobin, Dan I., 2003a. How people move. In: Moder, C.L., Martinovic-Zic, A. (Eds.), Discourse Across Languages and Cultures. John Benjamins, Amsterdam, pp. 195–210. Slobin, Dan I., 2003b. Language and thought online: cognitive consequences of linguistic relativity. In: Gentner, D., Goldin-Meadow, S. (Eds.), Language in Mind: Advances in the Investigation of Language and Thought. MIT Press, Cambridge, MA, pp. 157–191. Slobin, Dan I., 2004a. The many ways to search for a frog: linguistic typology and the expression of motion events. In: Stro¨mqvist, S., Verhoeven, L. (Eds.), Relating Events in Narrative, Vol. 2: Typological and Contextual Perspectives. Lawrence Erlbaum, Mahwah, NJ, pp. 219–257. Slobin, Dan I., 2004b. Relating events in translation. In: Ravid, D., Shyldkrot, H.B. (Eds.), Perspectives on Language and Language Development: Essays in Honor of Ruth A. Berman. Kluwer, Dordrecht, pp. 115–130. Stark, Heather A., 1988. What do paragraph markings do? Discourse Processes 11, 275–303. Stemler, Steven E., 2004. A comparison of consensus, consistency, and measurement approaches to estimating inter-rater reliability. Practical Assessment, Research & Evaluation 9 (4) Retrieved from http://PAREonline.net/getvn.asp?v=9&n=4. Taboada, Maite, Mann, William C., 2006. Applications of rhetorical structure theory. Discourse Studies 8 (4), 567–588. Talmy, Leonard, 1991. Path to realization: a typology of event conflation. Berkeley Working Papers in Linguistics 480–519. Talmy, Leonard, 2000. Toward a Cognitive Semantics, vol. 1. MIT Press, Cambridge, MA. Tannen, Deborah, 1980. A comparative analysis of oral narrative strategies: Athenian Greek and American English. In: Chafe, W. (Ed.), The Pear Stories: Cognitive, Cultural and Linguistic Aspects of Narrative Production. Ablex, Norwood, NJ, pp. 51–87. Tannen, Deborah, 2007. Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse, 2nd edition. Cambridge University Press, Cambridge. Thompson, Sandra A., 1984. Subordination in formal and informal discourse. In: Schiffrin, D. (Ed.), Meaning, Form, and Use in Context: Linguistic Applications. Georgetown University Press, Georgetown, pp. 85–94. Thompson, Sandra A., 1985. ‘Subordination’ in formal and informal discourse. In: Schiffrin, D. (Ed.), Meaning, Form, and Use in Context: Linguistic Applications. Georgetown University Press, Washington, pp. 85–94. Thompson, Sandra A., Couper-Kuhlen, Elizabeth, 2005. The clause as a locus of grammar and interaction. Discourse Studies 7 (4–5), 481–505. Tomasello, Michael, 2003. Constructing a Language: A Usage-based Theory of Language Acquisition. Harvard University Press, Harvard. Tolchinsky, Liliana, Rosado, Elisa, 2005. The effect of literacy, text type, and modality on the use of grammatical means for agency alternation in Spanish. Journal of Pragmatics 37 (2), 209–237.

B. Nir, R.A. Berman / Journal of Pragmatics 42 (2010) 744–765

765

van Dijk, Teun, 1980. Macrostructures: An Interdisciplinary Study of Global Structures in Discourse Interaction and Cognition. Erlbaum, Hillsdale, NJ. Van Valin, Robert D., 2005. Exploring the Syntax–Semantics Interface. Cambridge University Press, Cambridge. Van Valin, Robert D., 2006. On the relationship between syntactic theory and models of language processing. In: Bornkessel, A., Schlesewsky, M. (Eds.), Semantic Role Universals and Argument Linking: Theoretical, Typological and Psycho-/neurolinguistic Perspectives. Mouton de Gruyter, Berlin, pp. 263–302. Van Valin, Robert D., LaPolla, Randy J., 1997. Syntax: Structure Meaning and Function. Cambridge University Press, Cambridge. Verhoeven, Ludo, Aparici, Melina, Cahana-Amitai, Dalia, van Hell, Janet, Kriz, Sarah, Viguie´-Simon, Anne, 2002. Clause packaging in writing and speech: a cross-linguistic developmental analysis. Written Language and Literacy 5, 135–162. Verhoeven, Ludo, van Hell, Janet G., 2008. Clause packaging in Dutch narrative and expository text writing in children and adults. Discourse Processes 45, 387–405. Viguie´, Anne, 2001. Etude de la connexion syntaxique dans les textes narratifs et expositifs a` l’e´crit et a` l’oral chez les enfants et les adultes issus de classes sociales diffe´rentes. PhD Dissertation. Universite´ Lyon2. Viguie´, Anne, Jisa, Harriet, 1999. Schooling in France. In: Aisenman, R.A. (Ed.), Working Papers in Developing Literacy Across Genres, Modalities, and Languages. Tel Aviv University. Waltke, Bruce K., O’Connor, Michael P., 1993. An Introduction to Biblical Hebrew Syntax. Eisenbrauns.