|Date: Wed, 23 Jun 2004 21:01:26 +0200 (CEST)
From: Stefan Sudhoff
Subject: Auch and noch in child and adult German
AUTHOR: Nederstigt, Ulrike
TITLE: Auch and noch in child and adult German
SERIES: Studies on Language Acquisition
PUBLISHER: Mouton de Gruyter
Stefan Sudhoff, University of Edinburgh / University of Leipzig
The book presents the results of two empirical studies examining the grammatical properties and the acquisition of the German additive focus particles ''auch'' ('also') and ''noch'' ('also/still/another'). In the case of ''noch'', its non-additive meanings are also taken into account. Because
the main focus of the investigation is on spoken language, the analysis of various corpora of spoken child and adult German was chosen as the principal method.
Chapter 1 provides a general overview of the topic of the study, the German focus particles ''auch'' and ''noch'', as well as of the study's aims and methods. The remainder of the book is divided into two parts; the first (chapters 2 to 6) deals with the use of the two particles in adult German, while the second (chapters 7 to 9) examines their acquisition. Some conclusions and implications for future research are given in chapter 10.
Chapter 2 reviews the relevant literature on German focus particles, concentrating on their semantic, syntactic, prosodic, and information-structural properties. It is shown (i) that ''auch'' and ''noch'' differ in many respects, which makes a uniform treatment of the two particles impossible, and (ii) that previous approaches suffer from a number of serious shortcomings. The author criticizes the analyses of Jacobs (1983), Bayer (1996), Reis & Rosengren (1997), and Büring & Hartmann (1999), among others, for not being able to account for the entire range of linguistic
phenomena connected to focus particles and for making incorrect predictions in some cases. A widespread problem identified by Nederstigt is that most research in this field is concerned with only a limited number of particles, which are believed to be prototypical members of certain
subsets. The generalization of the results is often of doubtful value. One crucial criterion of the adequacy of a focus-particle theory seems to be the way it accounts for the complementary distribution of the stressed and the
unstressed variant of ''auch'': the former must precede its domain of application (= DOA; the constituent referring to the entity which is 'added' by the particle), which carries the nuclear accent in this case, whereas the latter obligatorily follows its DOA and carries the nuclear accent itself. Nederstigt argues that this problem has not been solved satisfactorily. One of the main purposes of her book is to provide adequate analyses for constructions with stressed and unstressed ''auch'' and to explain why ''noch'' does not show the same pattern.
Chapter 3 describes the methods and aims of the first empirical study, providing a more detailed examination of ''auch'' and ''noch'' in adult German. The analyzed material consists of 713 utterances originating from three different corpora of spoken German and TV/radio conversations
recorded by the author. First, the DOA of the particles was identified in all utterances by means of the account proposed by Dimroth (to appear), which is based on comparisons with the preceding context. The utterances were then classified with respect to (i) the meaning of the particles, (ii) the syntax of the respective clauses, especially the position of the particle, the position of the DOA, and the relative order of these elements using the
topological field theory (Drach 1939), (iii) their intonation, and (iv) their information structure using the Quaestio model (Klein & von Stutterheim 1992).
Chapter 4 reports the quantitative and qualitative results of the corpus analysis. Only the most important findings can be repeated here. While ''auch'' has only one meaning, namely the additive one, ''noch'' can occur in additive as well as in temporal, repetitive, restitutive, and comparative readings. With regard to the syntax, prosody, and information structure of the utterances, several usage patterns could be established: according to the data, both particles are only possible in the middlefield (the positions between the finite verb and the non-finite verb forms in V2-structures and between the complementizer and the finite verb in verb-final subordinate clauses), whereas the position of the DOA shows a greater variability. Nederstigt argues that an accented DOA following ''auch'' always has focus status. An (accented or unaccented) DOA preceding a stressed ''auch'', on the other hand, is assumed to be a topic, while the particle itself is focused. There is only one exception to this complementary distribution: if the DOA is a finite verb, it is focused (and accented)
and precedes the particle. In the case of ''noch'', the DOA is always focused, irrespective of the syntactic and intonational properties of the utterance. Leaving the non-additive occurrences of ''noch'' aside, the location of the pitch accent differentiates between two similar meanings:
an accent within the DOA indicates that an element of a different type is added, while an accent on the particle itself marks the addition of an element which is of the same type as a contextually given element. The regularities observed for the usage of ''auch'' correspond to the patterns
described in the theoretical literature (Krifka 1999, Dimroth to appear, among others), while ''noch'' had never been analyzed in such detail before.
In chapter 5, which discusses some consequences of the findings for a theory of additive focus particles, the author argues that the observed differences in the usage patterns indicate the need not only for a different
treatment of ''auch'' and ''noch'', but also for independent analyses of the stressed and the unstressed variant of ''auch''. A new proposal concerning the latter distinction is presented in chapter 6; according to Nederstigt, stressed ''auch'' should be analyzed as the head of a functional projection located between VP and IP, whereas unstressed ''auch'' is assumed to be an adjunct to the maximal phrase serving as its DOA. ''Noch'', which unlike ''auch'' does not occur in two complementary patterns, is assigned the same syntactic representation as the unstressed ''auch''.
Chapter 7 introduces the data and method of the analysis of ''auch'' and ''noch'' in child language. The study is based on the Caroline corpus, a collection of more than 23000 utterances of a German girl recorded when she was between 10 and 39 months old. The methods used for analysis were basically the same as for adult language, although the high number of verbless utterances and the non-applicability of the Quaestio model in many cases posed some problems for the analysis. The findings, presented in chapter 8, suggest that child German shows for the most part the same regularities as adult German, particularly with respect to the complementary distribution of the two variants of ''auch''.
Chapter 9 describes the acquisition of the particles. They appear in the Caroline corpus in the following order: stressed ''auch'', unstressed ''noch'', stressed ''noch'', unstressed ''auch''. Nederstigt attributes the differences in the acquisition times to individual properties of the particles, such as their intonation, their semantics, and their contextual requirements. The syntactic complexity of the respective clauses seems to be of minor importance. Chapter 10 summarizes the main issues of the book, arguing
once more for a separate analysis of the stressed and the unstressed variant of ''auch''. The results of the child-language study are interpreted as supporting this distinction.
It is unquestionably a great merit of Nederstigt's book that it employs an empirical approach to the grammar of focus particles. With regard to ''auch'', it can be shown that the patterns discussed in the theoretical literature correspond to the linguistic reality. Concerning ''noch'', on the other hand, this method makes it possible to develop a comprehensive classification of the particle's different usages, which had been systematized only incompletely before. Moreover, in showing that ''auch'' and ''noch'' behave quite differently in many respects, the author uncovers a serious deficiency of previous theories: analyzing just one member of a subgroup of focus particles and generalizing the findings for the whole group often leads to questionable results. The differences in the grammar of individual particles such as ''auch'' and ''noch'' must be accounted for by an adequate linguistic theory of these elements. With respect to the different variants of ''auch'', some inconsistencies of previous approaches could also be detected. Stressed ''auch'', which is often considered an exceptional usage of the particle, is shown to be the more common of the two patterns. Not only does it occur more frequently in child and adult German, but is clearly acquired earlier than the unstressed variant. Another positive characteristic of Nederstigt's account is the important role that it ascribes to information structure, although the respective ideas, which seem to provide straightforward explanations for many phenomena connected with focus particles, are not always developed as far as they could be.
Apart from the advantages of using corpora of spoken German, as mentioned above, the employed empirical methods are problematic in many ways. Based on the data, Nederstigt excludes occurrences of ''auch'' in the German prefield (the position before the finite verb in main clauses) from her analysis. As can be seen in (1) and (2), examples taken from Reis & Rosengren (1997), ''auch'' is perfectly possible in this position, be it
alone or together with some topicalized constituent:
(1) Auch hat Peter einen AUFsatz geschrieben.
Also has Peter an essay written
(Reis & Rosengren 1997: 257)
(2) Auch PEter hat das Buch gelesen.
Also Peter has the book read
(Reis & Rosengren 1997: 241)
In fact, Nederstigt uses the very same pattern in chapter 6 to argue for her structural analysis, but denies its existence in chapter 4 ('''auch' and 'noch' are restricted to the middlefield.''; p. 122). As shown by this example, there is no guarantee that a given corpus contains tokens of all possible usage patterns of a certain linguistic expression. To dispense with constructed examples completely and to rely on natural discourse instead, as suggested by the author, seems therefore not to be a good solution. On the other hand, Nederstigt rightly emphasizes the importance of the context for focus-particle utterances. One potential solution to this dilemma is to give explicit contexts for constructed examples in order to control their information structure.
Also susceptible to criticism are the different structural representations of the stressed and the unstressed variant of ''auch'' proposed in chapter 6. First of all, the employed terminology is quite confusing: in the case of stressed ''auch'', Nederstigt argues for an analysis in terms of a functional projection AUCHP serving as the complement of I°, but calls the involved structural relation adjunction (''this analysis assumes that 'AUCH' is an adjunct to the VP''; p. 196). As for unstressed ''auch'', on the other hand,
the particle, which is obviously an adjunct in the syntactic representation given, is argued to be a co-constituent of its DOA (''we will assume that in 'auch'-utterances, the particle is a co-constituent of the phrase that functions as domain of application for the particle rather than being an adjunct to this phrase''; p. 199). Leaving these inconsistencies aside, the analysis still seems inadequate, because it cannot account for the whole
range of linguistic data connected with ''auch'', and the phenomena regarded as evidence for it can be explained in more straightforward ways.
Let me first consider some constructions which cannot be accounted for by the adjunction-to-XP analysis of unstressed ''auch''. Nederstigt's proposal suggests that only maximal projections can serve as DOA of the particle.
However, this prediction is not born out, as shown by (3) and (4).
(3) (Peter hat dem Otto ein BUCH geschenkt und) er hat
auch [der Maria einen BALL] geschenkt.
(Peter has Otto a book given and) he has also [Maria a ball] given.
(4) Peter [SCHLUG] den Hund auch.
Peter [beat] the dog also.
In (3), the indirect and the direct object together form the DOA of ''auch'', whereas in (4) it is only the finite verb that is associated with the particle. An alternative proposal along the lines of Büring & Hartmann (1999), which analyses focus particles as adjuncts to VP but identifies the DOA with the focus of the utterance, is able to account for both constructions. In (3), the participle ''geschenkt'' obviously cannot belong to the DOA, because it has been mentioned before. ''Auch'' only adds the pair to the set of pairs of the form , where Peter gave y to x. However, the defocused participle has to remain in its base position in VP for syntactic reasons. Therefore, ''auch'' does not associate with the whole VP to which it is adjoined, but only with the focus constituents remaining within VP. In (4), the finite verb is the only focused element and precedes the unstressed focus particle. It has to move out of the focus domain to C° for syntactic reasons, but because of its focus status it is interpreted in its base position within VP. Again, the focus of the utterance serves as DOA of ''auch'', although the particle is not adjoined to it.
If unstressed ''auch'' is, like its stressed counterpart, analyzed as an adjunct to VP, both variants of the particle can be derived from the same deep structure, the main difference between the sentences being their information-structural properties. Nederstigt argues against such a uniform analysis, claiming that constructions with unstressed ''auch'' cannot always be transformed into constructions with the stressed variant of the particle and vice versa. Some of her examples are repeated here in (5) - (7).
(5) und zwar so dass man die Beschreibung auch [LEsen kann]
and in fact so that the description can also be read
(6) aber manche Männer sind dann so neugierig und wollen
einfach dabei sein [...] oder auch [ihre MEInung sagen]
but some men are so curious and want to be there or
also give their opinion
(7) Hab ich AUCH schon erledigt.
Have I also already finished
(p. 205; from Reis & Rosengren 1997: 249)
On closer examination, however, these examples do not seem to provide evidence against a uniform analysis, because in all cases the impossibility of using the complementary pattern can be attributed to independent factors. Deriving structures with stressed ''auch'' involves the movement of
its DOA into a position preceding the particle. This movement is blocked in (5) and (6) by the syntax of the clauses. In (5), the relevant part of the construction is an embedded clause, in which the verb cluster cannot leave
its base position in German. In (6), the elliptic coordination by means of ''oder'' ('or') requires the two coordinated elements to be of parallel structure. Without this restricting syntactic context, the use of stressed
''auch'' is possible, cf. (8).
(8) [Ihre Meinung sagen] wollen manche Männer AUCH.
[their opinion give] want some men also
(7) is a case of preverbal ellipsis, the omitted element being the DOA of stressed ''auch''. This construction has no counterpart with the unstressed variant of the particle because (i) only the DOA of stressed ''auch'' can occur in sentence-initial position (which is the only possible position for the element to be omitted in this type of construction) and (ii) only the DOA of stressed ''auch'' can be established in the previous context, which makes its elision possible. This suggests that constructions with the two variants of the particle differ considerably in their information structure, but it does not provide a reason to reject a uniform analysis.
Apart from these problems, Nederstigt's account leaves a number of questions unanswered. Consider (9), an example taken from Reis & Rosengren (1997).
(9) weil ja [Peter] es ihm AUCH zeigen wollte
because yes [Peter] it him also show wanted
(p. 204; from Reis & Rosengren 1997: 249)
Following Reis & Rosengren, Nederstigt assumes ''Peter'' to be in its base position, because it follows the modal particle ''ja''. But if stressed ''auch'' is the head of a functional projection above VP and if Peter and, consequently, the two object pronouns ''es'' and ''ihm'' are in their base positions and precede the ''auch'' at the same time, where should those base positions be located? The analysis implies that the direct and the indirect object as well as the subject of the main verb are generated outside VP. Furthermore, this stands in direct contradiction to
Nederstigt's assumption given in chapter 9 that the constituent associated with stressed ''auch'' does move: ''In AUCH-utterances, the particle does not move, but given that the particle is always following the contrasted element in the utterance, AUCH-utterances require the movement of this
contrasted element out of the VP to a position preceding the particle.'' (p. 334)
Another problem of Nederstigt's argumentation concerns the relation between ''auch'' and ''noch''. The fact that ''noch'' does not show the same complementary distribution as ''auch'' is taken as evidence for the different analyses of the two variants of the latter. According to Nederstigt, ''noch'' is, like unstressed ''auch'', always a VP-adjunct and cannot serve as a functional head. However, this analysis cannot explain why ''noch'' not only lacks a variant corresponding to stressed ''auch'', but also behaves very differently from unstressed ''auch''. What is needed here are detailed lexicon entries for both particles.
To sum up, Nederstigt is not able to provide convincing evidence for a different syntactic analysis of stressed and unstressed ''auch''. A uniform account deriving constructions with the stressed variant of the particle by movement of its DOA still remains a valid option to be seriously examined, as the observed differences (e.g. with respect to the character of the contrast, the possibility of omitting the DOA, and the distribution of accents) can be attributed to the different information structure of the sentences.
What would be necessary under these conditions is, of course, a separate analysis for occurrences of ''auch'' in the prefield, which seem to behave similarly to contrastive negation. This topic cannot be explored in more detail here.
Because the reviewer has no expertise in the field of language acquisition, the evaluation of the book's second part is restricted to some comments on the general argumentation. From the theoretical analysis provided in chapter 6, contradicting predictions can be derived: on the one hand, stressed ''auch'' should be acquired later than its unstressed counterpart, because it involves a more complex phrase structure as well as additional movement operations. On the other hand, the unstressed variant of the particle requires a longer MLU (mean length of utterance) and has a
more complicated accentuation pattern, which predicts the reverse acquisition order. The analysis of the child-language corpus shows that stressed ''auch'' is clearly acquired earlier than unstressed ''auch''. Nederstigt interprets this result as confirming her theory (''The
acquisition of 'auch', 'AUCH', 'noch' and 'NOCH' in German provided clear evidence for an analysis of 'auch' and 'AUCH' in which the two particles are analyzed as two different linguistic elements.''; p. 367) and concludes that
the structural factors leading to the first of the predictions mentioned above are not important in the process of acquisition. This way of arguing is obviously circular: if the corpus had shown that the variants of ''auch'' are acquired in the reverse order, the same conclusions with respect to the theory could have been drawn.
The book is without doubt an important publication in the field of focus-particle research, because it employs an extensive empirical analysis of ''auch'' and ''noch'' for the first time. However, the theoretical conclusions are not entirely convincing, and in many cases the observed data allows different interpretations. A fully satisfying theory of the grammar of (additive) focus particles in German is therefore still to be developed.
Bayer, Josef (1996): Directionality and Logical Form. On the Scope of Focusing Particles and WH-in-situ. Dordrecht: Kluwer
Büring, Daniel & Katharina Hartmann (1999): The Syntax and Semantics of Focus-Sensitive Particles in German. In: Natural Language & Linguistic Theory, 19, 229 - 281
Dimroth, Christine (to appear): Fokuspartikeln und Informationsgliederung im Deutschen. Tübingen: Stauffenberg
Drach, Erich (1939): Grundgedanken der Deutschen Satzlehre. Frankfurt am Main: Diesterweg
Jacobs, Joachim (1983): Fokus und Skalen. Zur Syntax und Semantik der Gradpartikeln im Deutschen. Tübingen: Niemeyer
Klein, Wolfgang & Christiane von Stutterheim (1992): Textstruktur und referentielle Bewegung. In: Zeitschrift für Literaturwissenschaft und Linguistik, 86, 67 - 92
Krifka, Manfred (1999): Additive Particles under Stress. In: Proceedings of SALT 8. Cornell: CLC Publications, 111 - 128
Reis, Marga & Inger Rosengren (1997): A Modular Approach to the Grammar of Additive Particles: the Case of German Auch. In: Journal of Semantics, 14, 237 - 309