Parenthetic Insertions

International audience; The present volume consists of contributions presented in a first version at the international conference on Parenthetic Insertions. It was organized by the Research Centre Laboratoire de Recherche sur le Langage of the former Université Blaise Pascal - Clermont-Ferrand in collaboration with the Maison des Sciences de l'Homme de Clermont-Ferrand in March 2014.1TopicParentheses, or in order to guarantee the distinction to the typographical sign, parentheses insertions are already analysed in ancient rhetoric and grammar as phenomena of speaking and writing in different syntactic, linear, semantic, (text-)functional, pragmatic and prosodic-intonatoric perspectives. For more than two thousand years they belong to the inventory of concepts used for the description and analysis of language and speech. Since antiquity, grammarians have stressed the "interruption" or "perturbatio" (interruptio, perturbatio) which cause insertions in the sentence construction (cf. Lähnemann / Rupp 2003; Spitzl-Dupic in this volume and 2016). In rhetoric, insertionss are treated in figure theory: they are analyzed as sentence figures when the syntactic interruption of the construction is in the foreground, or semantical figures when the sense carried by the insert is aimed at. What Breuer (1989: 235) states in general about rhetorical figure theory also applies to definitions of parenthetic insertions: they exhibit a certain "definitional blurriness" which "moves on a middle line between the effort of systematization and practical purpose. [...] What was decisive was the practical value of the terms."A comparable lack of clarity can still be found today in didactic grammar and can probably be explained in parallel: the works are designed for pedagogical effectiveness and therefore hardly attempt an exact definition of the phenomenon. Samples in modern textbooks on the German language show, for example, that parenthetical insertions are either not1 treated at all or very superficially2 or are presented as counter-examples to a 'good style3'. Only 1See e.g. Rug / Tomaszewski 2011.2See e.g. Helbig / Buscha (2001: 570). These lead as the only form the "switching sentence" as "a comment of the speaker on the content of the statement" and define it (sic) "as a form of the unconnected coordinative linkage of main clauses". At the same time, however, they explain that there is no coordination as regards content since 'the principal sentence involved is subordinate in relation to the other principal sentence'.3See, for example, the grammar of Bentin (2006: 45), where the treatment of the style is introduced with a 'deterrent' text to be improved by the pupils, 'peppered' with parenthesic insertions. The text is a translation of a text by M. Twain in which he caricatures the use of insertions through "improbable" sequences. occasionally can a trend 'towards' or 'back to' parenthesis be observed here (cf. Spitzl-Dupic 2016).In addition to the pragmatic-oriented contexts of rhetoric and grammar teaching, parenthetic insertions also remain stepchildren in other contexts of language-theoretical analyses: for example, they do not find detailed treatment in important modern scientific reference grammars of German,4 nor has historiographical linguistic research, with the exception of rhetoric research, dealt with them in detail to date (cf. Spitzl-Dupic 2016).At the same time, however, in the 20th century, against the background of the pragmatic turn in the 1960s, a growing interest in parenthetical insertions can be observed in the scientific literature (cf. Winkler's early investigations in 1969, Jeffersen in 1972, Betten in 1976), whereby spoken language is also included from the outset (cf. Bayer in 1973).5This interest has tended to increase since the 1990s and remains topical in the latest research.6The question of what a comprehensive and generally accepted definition should look like continues to be discussed (cf. Thurmair, Schmale, Dittmar, Haßler, Döring in this volume and Schneider 2007: 3-14), but it is now placed against the background of the increasingly prominent multidimensionality and complexity of the object under investigation, so that, depending on the research perspective - rhetorical-stylistic, linear-topological, syntactic, constructional-grammatical, intonatoric-prosodic, graphematic, semantic, functional-pragmatical and interactional, cognitivistic, text-variety-specific, diachronic – and 4In Zifonun et al. (1997, III: 2363-4) a comprehensive definition can be found, but the authors then do not go further into the problem (ibidem: 2363). Eisenberg (32006, II: 255) deals with parentheses in two sentences and although the online grammar of the IDS Grammis uses the term parenthesis several times, one searches in vain for a definition in all modules (inquiry from 23.07.2017). On the partly very superficial treatment of parentheses in the Duden grammars between 1966 and 2009, see Spitzl-Dupic 2016: 85.5However, Urmson (1952) already investigated the pragmatic function of the 'mitigation' of the claim to truth by means of parenthetic verbal insertions.6Without claim to completeness, see e.g. Schönherr 1993; Pittner 1995; Hoffmann 1998; Pétillon-Boucheron 2002; Caffi 2007; Dehé / Kavalova 2007; Schneider 2007; Corminboeuf / Heyna / Avanzi 2010; Glikman / Avanzi 2012; Kluck / Ott / de Vries 2014; Lazović 2017; etc. thus possibly also depending on the corpus of investigation. Different things are defined or at least focused on.The aim of this short introduction is therefore to briefly outline the current research perspectives for the treatment of insertions in which the contributions in this volume are to be incorporated.According to Schneider (2014: 277-278), there is a minimal consensus in modern linguistics on the following properties of parenthetic insertions:A – A parenthetical (P) is an expression of which it can be argued that, while in some sense ‚hosted‘ by another expression (H), P makes no contribution to the structure of H. (Burton- Roberts 2006: 179, zit. nach Schneider: 2014: 277)B - Parentheticals are expressions of varying length, complexity, function and syntactic category, which are interpolated into the current string of the utterance. (Dehé 2009: 307, see Schneider 2014: 278)Starting from A, the propriety of parenthetic insertions also consists in the fact that they lose their parenthetical character without their direct context H(ost). Contrary to statement A on the independence of the structure of the host from the insert, Pittner (1995: 91) has already explained how inserts can modify the area of different wearers of the host’s scops, so that it must be assumed that the insert can also play a role in the syntactic structure of the host, without this, however, being systematically the case. In any case, the wearer remains functional as a statement without the insert, whereas the insert can lose any statement value without its wearer, depending on its shape. This does not mean, of course, that the insertion is generally without semantic-communicative value for its bearer,7 as early grammatists sometimes assumed who spoke of an inserted 'frömden Sinn' (“extraneaous sens” cf. Spitzl- Dupic in this volume): on the contrary, the meaning carried by the insertion can be central for the meaning of the bearer up to the meaning of the entire utterance situation.Starting from B., it can be specified that the overt syntactic forms of parenthetic insertions are extremely variable, since these can consist of words, phrases, phrase connections, sentences of any form and mode, sentence connections, text blocks (cf. e.g. Kaltenböck 2007: 29-31). The different forms can also be analyzed as different syntactic functions, but also, as is repeatedly suggested, as elliptic propositions (cf. Döring and, from a historiographical perspective, Spitzl-Dupic in this volume). Some, particularly eloquent 7Exceptions are, for example, inserts which are produced due to the utterance situation, e.g. as reactions to disturbances (cf. Schmale in this volume). insertions are also discussed as results (modalizers, adverbs) of grammaticalization processes (cf. Dittmar and Haßler in this volume).A and B show that insertions break through the intrinsic linearity of the wearer / host or better, if more unusual, break it open, insofar as the carrier sequence is resumed after the insertion. The latter puts inserts in a certain proximity to footnotes, digressions, excursions, theme changes, but these differ from those forms in that they normally interrupt a carrier in the structure of its mostly syntactically defined "emergent" structure (Stoltenburg 2007). Conversation-logical approaches also show, however, that the syntactic criterion alone does not always do justice to the interruptive character of insertions: in spoken language prosodic and pragmatic units (gymnastic constructions, communicative activities) must be taken into account (cf. Schmale and Stoltenburg in this volume), in written language graphematic units (cf. Mizony-Schneider, Durand in this volume and below).The decision to realize information as an interrupting sequence is generally understood today as a communicative strategy (cf. e.g. Schneider 2014: 277-279, Lazović 2017: 70). This decision, at least at first glance, does not respect Grice's maxim of clarity and distinctness, since the listener/reader has to process the two-fold transmission of information, a duplication of semantic, syntactic and possibly prosodic patterns as well as possibly a double illocution. It is to be assumed, however, that in the eyes of the speaker / writerobviously another sequential-linear arrangement, where this additional effort for the recipient is avoided, would cause - even - greater decoding difficulties or would not do justice to his communicative intention. Parenthetic insertions are generally regarded as secondary with regard to the weighting of information and enable the speaker/writer to formulate content that is not directly presented to the recipient (cf. e.g. Potts 2005: 6-7). In this way they also offer speakers / writers the opportunity to say something in them that they do not want the reader / listener to question (cf. e.g. Gautherot in this volume). Beyond this essential quality,insertions can play a role for the "optimization of the illocutive, argumentative, epistemic, interactive, evaluative quality of the unit of expression and for the progression of discourse" (Lazović 2017: 42) (cf. also the contributions in this volume).Hoffmann (1998: 325) speaks of the "paradox of dependent autonomy", de Vries (2009) of behindence, by which he means a special form of coordination, and Kügelgen of structures operating on each other (2003: 224-228) in order to do justice to the complexity of the relation between insertion and host. Thus, at various levels of analysis, attempts are made to take into account the fact that the recipients stores the previously incomplete syntactic, semantic, prosodic, pragmatic information of the carrier at the beginning of the insertion process and, at its resumption, re-analyzes this information retrospectively and projectively, taking the insertion information into account.The production and processing of parenthetic insertions seem to be favoured by the existence of syntactically defined "parenthesis niches" (cf. Altmann 1981: 64-65,8 Zifonun et al. 1997, III: 2363). However, corpus-based research in recent years in particular has also shown that speakers/writers also use insertions outside these privileged positions, e.g. within functional links (cf. Pittner 1995: 89 and Stoltenburg, Dittmar in this volume).As already indicated, lexical, prosodic and graphematic markings can also control reception.Lexical means can introduce or terminate insertions, indicate the resumption of the carrier or characterize it as such within the parenthetic sequence (cf. Schmale in this volume).Prosodic and intonatory markings can be realized by pauses, speech tempo changes, accentuation, tone strength and pitch changes, etc. (cf. the contributions by Thurmair, Schmale, Stoltenburg, Dittmar in this volume and Döring 2011).In written language, double commas, dashes, brackets or typographical variations can signal the presence of insertions (cf. Mizony-Schneider, Durand in this volume and Pétillon- Boucheron 2002).Recent research results make it increasingly clear that these markings are by no means unambiguous. For example, literary texts in particular play with typographical norms (cf.Spitzl-Dupic, 2018), graphematic markings can be omitted completely, and they are not completely standardized in everyday writing (cf. Mizony-Schneider in this volume). And for spoken language, Döring (2011) shows that the boundaries of the inserts are often 8According to Altmann, this is the position before or after the finite verb, the borders of the sentence elements in the midfield and the end of the sentence. Hoffmann (1998: 316) specifies and modifies this as follows: "In German, the positions between outer field and apron and between apron and left sentence bracket are to be determined as parenthesis niches, in midfield any position between borders of phrases or particles can form a parenthesis niche. Finally, a parenthesis may also appear between the right record bracket and the next field and between the next field and the right outer field. [...] This is a premature or postmature realization of a speaker plan at specific positions that break down the linear structure." prosodically unclearly marked. And finally there are also inserts that are prosodically integrated (cf. Dittmar in this volume and e.g. for the English Dehé 2007: 270-274).At the same time, the presence of these possible markings is by no means a guarantee for the actual presence of an parenthetic insertion. In spoken language, for example, a pause may correspond to hesitation, and for the written, a flagrant example is Proust's research, for which Serça (2010) has shown that numerous segments marked typographically as insertions correspond to sequences of utterances that inscribe themselves excursus-like into the context and thus no longer correspond to the relatively consensual definition of parenthetic insertions (see above).And finally, all these markings can also be used to realize syntactically integrable sequences as a parenthetic insert, which again shows that a syntactic definition of inserts alone does not always prove to be operational (cf. also Pittner 1995: 89).These few observations make it clear that each candidate for classification as a parenthetic insertion must be examined individually not only for its interrupting characteristics, but also for its specific relationship to the bearer, context, co-text and to the participants in the utterance situation. Since parenthetic insertions are neither graphematically nor prosodically, nor semantically nor functionally-pragmatically homogeneous phenomena, their multidimensionality should be taken into account in the analysis, which presupposes a precise definition of the analytical perspective. These precautions are taken into account in the contributions to this volume, thus avoiding what Calude (2010: 205) calls "the slipery nature of the term of 'parenthetical'".2 To this volumeOn different theoretical backgrounds, the contributions published in this volume deal with parenthetic insertions from different perspectives:-Typography and writing in connection with syntax, semantics and text functions,-Writing and orality: parenthetical insertions in spoken language, also in comparison to written language,-Discursive Properties and Functions of Parenthetic Inserts-also individual diachronic studies on the syntax, pragmatics and functionality of parenthetic insertions.These contributions are supplemented by a historiographical study on the treatment of parentheses in the history of German language theory.Ad I. Odile Mizony-Schneider examines the different use of the marking of parenthetic insertions by double dashes and brackets, as it presents itself on the one hand in the formal and semantic-expressive criteria of historical and current codification works and on the other hand in contemporary writing (supra-regional press, non-fiction prose & fiction). She shows the different functions in the orientation of the reading and understanding focus for the use of one or the other sign combination and the change in their use.Marie-Laure Durand analyses so-called loose nominal appositions to the right of a nominal phrase, which are enclosed or excluded by commas or brackets. It shows that bracket constructions have such a much higher degree of autonomy that the entity develops its own illocution and cannot be anaphorically resumed in the following text. In addition, the hypothesis is put forward that the more pronounced autonomy of the parenthesis- NP with nominal reference is due to the circular grapheme of the parentheses, which clearly excludes the separated nominal phrase from the sentence and the main strand of the text.Ad II. Starting from the observation that most parenthesis definitions are based either on oral or on written language, Maria Thurmair examines the insertions into lectures a university and thus into a type of text from the field of scientific communication, which can be characterised by the fact that it is generally regarded as written, but is realised orally. M. Thurmair shows that quantity, quality and functionality of parentheses are indications for a constitutive oralness or constitutive writtenness of this type of text and thus make it possible to categorize lectures into those which are closer to the pole of orality and those which are closer to the pole of writtenness.Within the framework of an incremental syntax that takes into account the specific properties of spoken language - temporality and processuality of expression production, prosody - Benjamin Stoltenburg points out the proximity and difference between parenthetic interruptions of gymnastic construction units before possible termination points and different types of extensions of gymnastic construction units after possible termination points of gymnastic constructions.G. Schmale starts from the conversation-analytical approach of "side sequences" (Jefferson 1972) and stipulates that in the spoken language primarily not the syntactic interruption is the characteristic that characterizes parenthetic insertions, but the suspension of a previous activity by another. On the basis of three German and French conversation corpora, he develops ten different types of side-sequences (securing understanding, corrections / repairs, metadiscursive activities, theme-changing activities, etc.) and examines the often interacting lexical, prosodic and paraverbal indicators that indicate the existence of a side-sequence for the interacting subjects.Based on an analysis of the “Berliner Wendekorpus”, Norbert Dittmar concentrates his investigation on inserts in oral speech that interrupt the current coding of syntactic patterns in a reduced form and in a speech-accompanying function. These are sequences such as weeß ick ↑, sagn= wa= mal / sag =ick =mal, which are described here as speech commenting routines that maintain the conversation process and control it projectively or retrospectively. He shows how these formulas enable the speakers to, among other things, reduce emotionally difficult content to a partial aspect and at the same time signal the provisional nature of the expression.Ad III. From a discourse-analytical perspective, Laure Gautherot develops the concept of ethical parenthetic insertion by relying on the analysis of crisis communications in the political public sphere. Taking into account the prediscursive context, inserts are analyzed as possible ethical markers of the speaker ethos and their performance is examined on a communicative-pragmatic and discourse-ethical level. L. Gautherot shows that the use of parenthetic inserts for the speaker offers the possibility of supporting his credibility, but their use also bears the danger of working in the opposite direction if the previous ethos and self- image of the speaker conflict with the semantic-pragmatic interpretation of the insert by the addressees.Elisabeth Malick Dancausa's contribution, which is also discourse-analytically oriented, treats parentheses as an excellent means of expressing irony. Based on texts by Karl Kraus and Kurt Tucholsky, she reveals the semantic and pragmatic parallels between irony and insertions, relying on the concepts of break and continuity. Finally, in the analysis of her examples, she explains in more detail how Ducrot's concept of polyphony must be used to precisely analyze the interplay between ironic parenthetic insertions and context.Ad IV. Gerda Haßler examines the use of epistemic verbs in insertions by examining the validity of three competing explanatory approaches: a. the assumption of matrix sets with the epistemic verb as the basis on which a complement set depends; b. the assumption of a parenthetic insertion; c. the assumption of a grammaticalization of the insertions as modalizers. On the basis of also diachronic corpus investigations on French, Spanish, Portuguese and German, she rejects the hypothesis a. and points out the syntactic, semantic and pragmatic advantages and disadvantages of the other two analyses. She concludes that the significance of a type of insertion depends, among other things, on its position and emphasis, so that the assumption of a polysemy of epistemic verbs in insertions should be made.Based on the Mannheimer Corpus, Hiroyuki Miyashita concentrates on the analysis of parenthetic um-zu constructions, whose topological, syntactic, semantic, functional, typographical, and illocutive properties he examines and contrasts with the final um-zu -constructions. For the parenthetic um zu constructions, he points out the existence of four different functional types that inscribe themselves in the framework of metacommunication and text organization. His research concludes with the question of the emergence of the parenthetic um-zu construct from the final um-zu construct, whereby the hypothesis is investigated that the parenthetic um-zu is a change of reference from the main clause to the speech action originating from the origo.It is Christian Fortmann's intention to examine whether the thesis formulated by Reinhart (1983) for English, according to which parenthetic constructions with verba dicendi et sentiendi (can) have two readings from the point of view of the attribution of the utterance of the carrier theorem, applies to German. In the subject-oriented reading the utterance is attributed to the subject referent of the parenthetic predicate, in the speaker-oriented reading to the speaker. By means of different tests it is determined whether the possible disambiguation can be related to syntactic conditions and differences between parenthetic constructions and regular complementation are shown.On a generativist background Sandra Döring examines the syntax of parentheses and loose appositions. By means of different tests it is demonstrated, how parentheses can be analyzed as sentences, which are possibly realized in elliptical form, and which syntactic operations are to be assumed with a form not corresponding directly to a sentence form. In a second step, parentheses and loose appositions are compared in their syntactic status and behavior.My own historiographical contribution examines the concept of parenthesis in a corpus of German-language texts of various genres published between 1641 and 1839 (stylistic and orthographic doctrines, rhetoric, poetics, encyclopaedias, grammars). It is shown how the analyses of parentheses deepen during this period, how they gain objectivity against the background of the "pragmatization" of grammar and the refinement of grammatical analytical instruments, and how - temporarily - in the first third of the 18th century some linguists put the pragmatic-communicative functions of insertions at the centre of their analysis.With its wide range of mostly corpus-based analyses, this volume should serve as a further helpful building block in the current discussion on parenthetic insertions.