LINGUIST List 15.1916

Thu Jun 24 2004

Review: Psycholinguistics: Nederstigt (2004)

Editor for this issue: Tomoko Okuno <tomokolinguistlist.org>


What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Sheila Dooley Collberg at collberglinguistlist.org.

Directory

  1. Stefan Sudhoff, Auch and noch in child and adult German

Message 1: Auch and noch in child and adult German

Date: Wed, 23 Jun 2004 21:01:26 +0200 (CEST)
From: Stefan Sudhoff <sudhoffrz.uni-leipzig.de>
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
YEAR: 2004
Announced at http://linguistlist.org/issues/15/15-1255.html

Stefan Sudhoff, University of Edinburgh / University of Leipzig


INTRODUCTION

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.


SYNOPSIS

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
B�ring & 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.


CRITICAL EVALUATION

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 B�ring &
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 <Maria, einen Ball> to the set of
pairs of the form <x, y>, 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
 (p. 185)

(6) aber manche M�nner 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
 (p. 185)

(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 M�nner 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.


SUMMARY

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.


REFERENCES

Bayer, Josef (1996): Directionality and Logical Form. On the Scope of
Focusing Particles and WH-in-situ. Dordrecht: Kluwer

B�ring, 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. T�bingen: 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. T�bingen: Niemeyer

Klein, Wolfgang & Christiane von Stutterheim (1992): Textstruktur und
referentielle Bewegung. In: Zeitschrift f�r 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



ABOUT THE REVIEWER

Stefan Sudhoff is a PhD student at the University of Leipzig,
Institute for Linguistics. His research interests include the
syntax-prosody interface and the grammar of focus particles in German.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue