LINGUIST List 13.2389

Fri Sep 20 2002

Review: Psycholinguistics: Weissenborn & Hoehle (2001)

Editor for this issue: Naomi Ogasawara <naomilinguistlist.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 Simin Karimi at siminlinguistlist.org.

Directory

  1. Annemarie Kerkhoff, Weissenborn & Hoehle (2001) Approaches to Bootstrapping, vol. 1 & 2

Message 1: Weissenborn & Hoehle (2001) Approaches to Bootstrapping, vol. 1 & 2

Date: Fri, 20 Sep 2002 13:54:17 +0000
From: Annemarie Kerkhoff <Annemarie.Kerkhofflet.uu.nl>
Subject: Weissenborn & Hoehle (2001) Approaches to Bootstrapping, vol. 1 & 2

Weissenborn, J|rgen, and Barbara Hvhle, ed. (2001)
Approaches to Bootstrapping: Phonological, Lexical, Syntactic and
Neurophysiological Aspects of Early Language Acquisition, Vol. 1.
John Benjamins Publishing Company, hardback ISBN 1-55619-992-9, vii+294pp,
$86.00, Language Acquisition and Language Disorders 23.

Weissenborn, J|rgen, and Barbara Hvhle, ed. (2001)
Approaches to Bootstrapping: Phonological, Lexical, Syntactic and
Neurophysiological Aspects of Early Language Acquisition, Vol. 2.
John Benjamins Publishing Company, hardback ISBN 1-55619-993-7, vi+331pp,
$91.00, Language Acquisition and Language Disorders 24.

Book Announcement on Linguist:
http://linguistlist.org/get-book.html?BookID=98 
http://linguistlist.org/issues/12/12-2361.html

Annemarie Kerkhoff, Utrecht Institute of Linguistics UiL-OTS, 
Utrecht University, The Netherlands

INTRODUCTION

This book explores the structure and interaction of bootstrapping
mechanisms available to the language- learning child, focusing on the
child's sensitivity to certain properties of the input. The
'bootstrapping' problem is one of the most fundamental issues in the
study of child language: to acquire language a child must parse the
signal, but parsing the signal requires knowledge of the
language. Importantly, the editors have chosen to take an
interdisciplinary and cross-linguistic approach to this topic.

The two volumes of 'Approaches to bootstrapping' together consist of
five separate parts. The first two parts focus on the nature of the
input and its cues to lexical and syntactic knowledge. The third part
concerns early language production, and the role of prosodic and
morphosyntactic knowledge. Part IV addresses neurophysiological
aspects of language acquisition, and part V concludes with some
additional perspectives on the issues raised.

In the following review, a selected number of articles from
'Approaches to Bootstrapping' will be discussed. This selection is
based on the interests of the reviewer, and does not reflect the merit
of these articles in comparison with others.

SYNOPSIS 

Part I deals with early word learning and its prerequisites. In
'Bootstrapping from the signal: Some further directions', Jusczyk
gives an excellent and useful overview of the author's important
contribution to the field. Previously published and unpublished data
are provided concerning infants' word segmentation abilities. The
author reports mainly on studies that use versions of the Headturn
Preference Procedure (Jusczyk and Aslin 1995). These studies have
shown that English-learning 7.5-month-olds have some ability to
segment words in fluent speech. 7.5-Month-old infants (but not
6-month- olds) seem to have acquired crucial information about the
sound organisation of their native language, including word stress,
allophonic cues, phonotactic constraints and distributional
regularities. Infants seem to show a preference for bisyllabic words
with a WS pattern as initial (first-pass) strategy, consistent with
the Metrical Segmentation Strategy (MSS) as proposed by Cutler
(Cutler, 1990; 1994), identifying the occurrence of strong syllables
with the onsets of new words in fluent speech. However, infants do not
detect familiarised WS words (guitar, surprise), which constitutes
evidence for use of distributional cues to determine where the word is
likely to end. However, it is argued that additional information based
on allophonic cues or phonotactic sequences may override a
segmentation strategy based purely on stress cues.

Jusczyk proposes that children use strategies that allow them to
'divide and conquer the input': carving the input up into word-sized
units provides the child with opportunities for observing regularities
in allophonic and phonotactic properties which occur at the onsets and
offsets of these units. The second part of this article addresses the
relationship between word segmentation and syntactic
organisation. From previous work it has become clear that infants are
sensitive to prosodic organisation in clauses and phrases (Gerken et
al., 1994; Jusczyk et al, 1992). The author suggests that infants'
developing word segmentation abilities may play a role in this process
by enabling the learner to track the distribution of grammatical
morphemes within the boundaries of prosodic phrases. However, infants
draw on a variety of sources of information - conceptual and
linguistic - to learn about the syntactic organisation of the
language.

Although this article is one of the most important ones in the book,
and fits in well with the overall aim of the volumes, it does not
contain much new or unpublished data. The articles cited are not very
recent (the latest being (Jusczyk et al., 1999)) and a more extensive
overview of Jusczyk's work can be found in his book 'the discovery of
spoken language' (Jusczyk, 1997). Some of the more problematic issues
that are not dealt with concern the methodology and the precise nature
of the Headturn Preference Procedure. Thus, it is unclear how to
determine whether longer looking times represent a preference for new
stimuli or a preference for familiar stimuli. Also, the difference
between discrimination on the one hand and preference on the other is
not discussed in any detail.

Furthermore, overview articles like this can make it difficult to
evaluate results since no specific data are given. For example, the
author confines himself to stating ''the overall performance by the
7.5-month-olds suggests that infants do have some capacity to deal
with interference from competing voices''.

Echols' 'Contributions of Prosody to Infants' Segmentation and
Representation of Speech' is very clear and well-structured article,
which further addresses the origin of word-level segmentation, as well
as the availability and use of prosodic cues such as perceptually
salient syllables and rhythm. Experimental findings suggest that the
salience of both stressed and final syllables may derive from acoustic
features used to prosodically highlight these syllables. Furthermore,
both types of syllable are attended to and represented more precisely
by 9-month-old infants. Thus, these perceptually salient syllables may
serve as prosodic cues for word-segmentation, although the relative
roles of these cues may change with development.

Fisher & Church's article links acquisition to perceptual learning and
memory processes. The authors argue that the Perceptual
Representations Systems (PRS) framework makes learning an integral
part of speech perception: long-term auditory priming as phenomenon of
implicit rather than explicit memory, is mediated by a learning
mechanism that could contribute to the creation of a long-term store
of word-sounds in children, as well as to the life-long adaptability
of speech processing to context.

Fernald, McRoberts and Swingley focus on early word recognition,
stressing the graded nature of developmental changes. Increased speed,
accuracy and efficiency of spoken word recognition in the second year
of life are argued to not only reflect changes in cognitive
capabilities (not specific to language processing), but also to
reflect changes in infants' lexical representations. The authors
provide evidence from eye- tracking studies that suggest that infants
have some ability to process speech incrementally, providing evidence
against the assumption that early lexical representations are holistic
or phonetically underspecified. It is not entirely clear however what
is meant by 'holistic' or 'phonetically specified', and whether these
two concepts are in fact opposed to one another.

Part II focuses on input cues to syntactic knowledge. In the first
article, Gerken provides an overview of recent research on the nature
of the input and early acquisition of syntax, arguing for a bridge
between the two. She makes the important point that ''if one type of
grammatical representation of a particular utterance is extractable
from the signal and another representation for the same utterance is
not, the extractable representation should be treated as a priori more
plausible''.

The remaining articles address possible prosodic cues to the
acquisition of syntax. For instance, H�hle, Weissenborn, Schmitz and
Ischebeck found prosodic differences between sentences involving
head-complement constructions as compared to head-modifier
constructions.

Penner, Wymann and Weissenborn address the well-known discrepancy
between the child's perceptive capacity and speech production when
they discuss the prosody-lexicon interface. They propose that the
child makes use of a robust algorithm for prosodic bootstrapping, the
'Rhythmic Activation Principle' or RAP. This will enable the child to
set the head directionality parameter, by mapping the pattern of
relative prominence in the phonological phrase onto the corresponding
complement- head configurations in the syntax. The authors show that
this algorithm is successfully applied by normally developing children
as well as language-impaired children. The production lag follows from
the fact that the child resorts to an underspecified representation of
the phonological phrase, due to a conservative learning principle. The
difference between the two groups is explained by stating that
underspecified representations become persistent in language-impaired
children.

Part III addresses interactions of prosodic and morphosyntactic
knowledge in early language production.

Demuth argues for a constraint-based approach to acquisition. She also
discusses implications of her approach for the identification of
children at risk of language delay.

Freitas, Miguel and Hub Faria show that acquisition of syllabic
structures, like codas in European Portuguese, might depend on the
grammatical features encoded by them. Thus, interaction between
prosody and morphology may have consequences for the order of
acquisition.

In part IV, neurophysiological aspects of language acquisition are
addressed. In the first article, Morfese provides an excellent
overview of studies using ERP (event-related potential) procedures to
study infant and toddler word acquisition. Interestingly, it describes
studies investigating speech perception and phonology, which is a
well-covered area in this type of research, as well as studies on
early word acquisition and word discrimination.

As the two remaining articles show a considerable overlap with this
review article, their inclusion seems unnecessary.

Finally, Part V groups together studies on additional perspectives to
language acquisition, addressing questions of methodology, the nature
of linguistic primitives, and the development of bird song as compared
to human language acquisition.

Plunkett's excellent article on interactionist approaches reviews
contributions of behavioural, neuropsychological and computational
studies to the field of acquisition, focusing on speech perception,
word recognition and the acquisition of inflectional morphology. He
makes a case for a multi-disciplinary approach to language
acquisition, stressing that linguistic development should be seen as
'the interaction of powerful general learning mechanisms with a richly
structured environment that provides the necessary ingredients for the
emergence of mature linguistic representations'.

GENERAL COMMENTS

In general, the selected articles reflect experimental
psycholinguistic work, with little reference to current generative
theories such as Optimality Theory. Still, 'Approaches to
Bootstrapping' presents an excellent (if not state-of-the-art)
interdisciplinary overview of recent developments in first language
acquisition research. It is regrettable that it has taken such a long
time to publish this important book, but it is undoubtedly a valuable
contribution to the field. It will be useful as a reference book for
scholars from a wide range of interests within psychology and
linguistics. It is not introductory, and as such not suitable as a
textbook.

REFERENCES

Jusczyk, P. W. (1997). The discovery of spoken language. Cambridge,
Massachusetts: The MIT Press.

ABOUT THE REVIEWER

I am in the first year of my PhD studies at Utrecht University,
investigating the acquisition of morpho- phonological alternations by
Dutch-speaking children. My interests are the development of
phonological representations for perception and production and the
phonology-morphology interface.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue