Berg, Thomas (2001) Linguistic Structure and Change: An Explanation
from Language Processing. Oxford University Press, paperback ISBN
0-19-829985-0, 352pp, $27.95 (hardback ISBN 0-19-823672-7, $99.00,
published in 1998).
Publisher's announcement: http://linguistlist.org/issues/12/12-1998.html
Alexander T. Bergs, Heinrich-Heine-University Duesseldorf
The (im-)possibility of explaining language change seems
to have fascinated generations of linguists. This book
represents one of the latest and also one of the
more ambitious attempts to come to grips with that "ever-
whirling wheel" (Aitchison 1998).
Although the book is formally divided into ten chapters,
it has three main sections with respect to content.
Section one contains what may be called the theoretical
and philosophical background and comprises the following
Chapter One "On the 'Art' of Explanation"
This chapter contains a very brief discussion of the
status of explanations in the theory of science in general
and linguistics in particular. Having refuted any
generative "explanations" in linguistics,
Berg suggests, contrary to Lass (1980), that linguists
must (and should) often content themselves with more
pragmatic modes of explanation, including 'statistical'
explanations. Explanation in his study (with reference
to Popper)is understood as the connection of two hitherto
unconnected things from two more or less unrelated domains.
Chapter Two "Explanations from a Macrolinguistic
In this chapter Berg reviews no fewer than nine different
approaches to language structure and change that somehow
seem to connect intra- and extralinguistic factors in
their explanations of change. These include
"the neurological, the phonetic, the psychological, the
semiotic, the functional-communicative, the pragmatic,
the sociocultural, the historical and the system-internal
approach". Berg concludes that all these approaches and
their relationships are characterized by "complementariness
and competition": while backing each other to a certain
extent, they also compete for the rank of 'highest
explanatory power'. He proposes to start with a
psycholinguistic account of language change which is, on
the one hand, necessarily eclectic in nature, as it is
informed by many of the other approaches mentioned above,
but on the other hand also restrictive enough, and
therefore, evaluable('gaugable'), if it proceeds from one
individual processing principle to a more comprehensive
theory that is only as eclectic as necessary.
Chapter Three "Method"
Berg proposes the use of predictions about structure and
change based on externally and independently developed
principles instead of post hoc explanations, which seem to
have been favoured so far. His explanans is the
psycholinguistic point of view, his explanandum certain
linguistic patterns. This study is based on the
interactive-activation model. This concludes the
introductory section of the book.
The second section contains the main body of the study. It
Chapter Four "Language Structure"
Berg describes and discusses numerous examples of language
structures from various languages (mostly English, however)
and from all levels of language, ranging from (segmental)
phonology through morphology to syntax. The structures and
phenomena discussed include, inter alia, onset/coda
(a)symmetries, assimilation phenomena, the order of
inflectional and derivational suffixes and that of nouns
and adjectives in NPs. He shows that the majority of
attested structures and phenomena can be explained with
reference to constraints of the language processor.
Chapter Five "Language Change"
Starting from his psycholinguistic analysis of frequent
inadvertent linguistic slips Berg suggests that these
errors actually may be regarded as entry points of
linguistic change. A number of notorious language change
processes and problems are elucidated from this point of
view. These include changes based on assimilation,
differences in consonant and vowel changes, the
susceptibility of certain consonant classes and clusters
to change, paradigmatic pressures and pattern symmetries,
and the role of frequency and word classes in linguistic
Chapter Six "Poetic Language"
This chapter discusses processing constraints and the
development of particular rhyming patterns in English and
Arabic. Contrary to what is commonly assumed, namely that
rhyming patterns are nothing but arbitrary cultural
conventions, Berg shows that these can also be explained
on the basis of psycholinguistic constraints on language
processing. This chapter concludes the main body of the
Section Three presents a broad theoretical discussion and
the development of a new research program on the basis of
the preceding chapters. It comprises
Chapter Seven "Discussion"
in which the overall results of chapters four and five are
Chapter Eight "A Psycholinguistic Model of Language
Structure and Change"
In this chapter it is suggested that psycholinguistic
processing constraints facilitate certain changes. They do
not cause changes as such, but they delineate the
direction of the change and, therefore, also exert certain
pressures on the linguistic structure. Thus, attested
linguistic structures are often the result of (sometimes
counteracting) psycholinguistic processing forces at work
on various language levels.
Chapter Nine "Implications for Psycholinguistic Theory"
This brief chapter (2 pages!) basically advocates a
parallel, not a serial processing model.
Chapter Ten "The Overall Perspective: Reductionist or Non-
Here Berg returns to the question put at the outset: can
psycholinguistic theories explain language change? Berg
argues they do - in contrast to generative theories. With
an eye on the alleged accusation (by some, mostly
theoretical linguists) that many 'external' approaches to
language are 'reductionist', Berg counters that his study
shows - in the most extreme generalization - that there is
in fact no competence grammar over and above performance
grammar. The non-existence of the former, however, seems to
be impossible to prove.
This is a tremendously ambitious work that covers a lot of
ground. The idea that certain structures in language are
more susceptible to change as they are psycholinguistically
less preferable seems well argued for and the corresponding
analyses are carried out with great care. But whether these
"weak spots in language" indeed also show up in or as slips
of the tongue is a different question. To my mind, Berg
succeeds in demonstrating the importance of psycholinguistic
and/or cognitive structures, principles, and mechanisms for
both language structure and language change - even though I
am sure that many linguists, including myself, would not
subscribe to each and every analysis put forward here.
A stronger incorporation of similar strands of research (such
as OT, Sapir's drift, Natural Morphology, or D. Stein's
concept of diachronic vectors in synchrony) would have been
interesting, its lack, however, does not ruin the argument as
a whole. On the contrary, it seems this would have increased
eclecticism and thus would have been counterproductive in the
light of chapter two.
Once one has accepted that there are other theoretical
positions than Lass's (1980) with regard to the status and
structure of explanations, this book is a treasure chest of
insights, new ideas and starting points for new research
A few critical remarks must also be made, though. The
first turns one of the main positive features of the book
into a negative one. The breadth of coverage and the
sheer amount of analyses (45 all in all, 21 synchronic, 21
diachronic, 3 poetic, with 83 subanalyses on the whole)
makes some of the single analyses seem shallow in parts.
In other words, at some points one would have wished for
greater depth. Also, it creates the impression of a
patchwork study (a two page chapter seems odd, somehow).
But this, again, is also one of strengths
of this complex book. A second minor remark: the second
chapter (an overview of alternative approaches), helpful as
it is for placing this study in the right context, clearly
suffers from the little space that has been devoted to it
(nine approaches in 35 pages). Some of the lines of enquiry
are much better characterized than others: the subsection
on the semiotic approach or the psychological one, for
instance, are much better than the one on the
sociocultural or the historical approach (the latter
seemed particularly infelicitous and its labelling is still
a mystery to me). This may give less experienced readers a
slightly tendentious idea of what other scholars are doing
in their fields and should thus not be left uncommented on
if the book is used in class. The latter seems to be
possible in advanced graduate courses on language
change or psycholinguistics. For these, however, the book
seems to be ideal as it leaves students enough space to
develop their own ideas and hypotheses and in doing so
maybe even find an interesting topic for their theses.
Summing up, it must be said that this book is indeed a
very valuable and innovative contribution to the study of
both linguistic structure and change. It is to be hoped
that it will stimulate a whole range of follow-up studies
in a similar vein in an area which is as old as linguistics
and as topical as can be: the question of the whys and
hows of language structures and their change.
Aitchison, Jean. 1998. Language Change: progress or decay?
Lass, Roger. 1980. On explaining language change. Cambridge:
Alexander T. Bergs is lecturer in English Language and
Linguistics at Heinrich-Heine-University Duesseldorf and
in General Linguistics at Bonn University. His main areas
of research include historical linguistics and language
change, sociolinguistics and syntax.