LINGUIST List 13.2116
Fri Aug 16 2002
Review: Linguistic Theories: Bybee & Hopper, ed. (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 simin
linguistlist.org.
Directory
Ahmad R. Lotfi, Bybee & Hopper (2001) Frequency and the Emergence of Linguistic Structure
Message 1: Bybee & Hopper (2001) Frequency and the Emergence of Linguistic Structure
Date: Mon, 12 Aug 2002 23:04:21 +0000
From: Ahmad R. Lotfi <arlotfiyahoo.com>
Subject: Bybee & Hopper (2001) Frequency and the Emergence of Linguistic Structure
Bybee, Joan, and Paul Hopper, ed. (2001)
Frequency and the Emergence of Linguistic Structure.
John Benjamins Publishing Company, vii+480pp, paperback ISBN 1-58811-028-1, USD 42.95,
hardback ISBN 1-58811-027-3, USD 125.00, Typological Studies in Language 45.
Book Announcement on Linguist:
http://linguistlist.org/get-book.html?BookID=2388
Ahmad Reza Lotfi, Azad University at Khorasgan
SYNOPSIS
"Frequency and the Emergence of Linguistic Structure" is a collection
of revised papers presented in a symposium at Carnegie Mellon
University (1999). In addition to the introductory paper by Bybee and
Hopper, the volume contains 19 articles organised in 4 parts: (1)
patterns of use, (2) word-level frequency effects, (3) phrases and
constructions, and (4) general.
Part One: Patterns of use (3 articles)
Sandra A. Thompson and Paul J. Hopper in their "Transitivity, clause
structure, and argument structure: Evidence from conversation" examine
conversational data from an English database and argue that (English)
conversation is very low in transitivity to the effect that the number
of participants is low (only in 27% of cases there are 2 or more
participants), and the clauses mainly depict non-action, atelic,
non-punctual, non-volitional events. They further argue that more
useful constructions have a better chance to become structuralised.
Joanne Scheibman's "Local patterns of subjectivity in person and verb
type in American English conversation" is an analysis of
conversational data from audiotaped informal conversations among
friends and family members. Subjectivity--the speakers' ability to
view themselves as subjects--is usually characterised with the use of
1st person singular and also such verbs as feel, believe, and
suppose. Scheibman shows that the most frequent subjects are 3rd
person singular, 1st person singular, and 2nd person singular
respectively. Relational, material, cognition, and verbal verb types
are the most frequent ones. Finally, most of the predicates are
expressed with the present tense.
"Paths to prepositions? A corpus-based study of the acquisition of a
lexico-grammatical category" by Naomi Hallan focuses on what Bowerman
(1996) calls path morphemes--multi-functional word forms such as
'over' and 'on' in English that seem to be primarily locative in
interpretation. Child and adult data are examined to show which
function (phrasal or prepositional) is acquired first.
Part Two: Word-level frequency effects (5 articles)
Betty S. Phillips in her "Lexical diffusion, lexical frequency, and
lexical analysis" offers a refinement of the Frequency-Action
Hypothesis to the effect that "sound changes which require
analysis--whether syntactic, morphological, or phonological--during
their implementation affect the least frequent words first; others
affect the most frequent words first" (p.123). Diachronic data from
English are given in support of this hypothesis.
Janet B. Pierrehumbert's "Exemplar dynamics: word frequency, lenition
and contrast" provides a formal architecture that is claimed to
describe how word-specific phonetic detail interacts with general
phonological principles in terms of exemplar theory. She concludes
that we learn phonological categories by remembering their labelled
tokens.
In their "Emergent phonotactic generalizations in English and Arabic",
Stefan A. Frisch, Nathan R. Large, Bushra Zawaydeh, and David
B. Pisoni maintain that emergent phonotactic grammar is grounded in
the lexicon with its effects observed at multiple levels of
abstraction. While well-formedness judgements in English reveal that
the probability of a new word implies its phonotactic well-formedness,
Arabic data are ambiguous in that "it is difficult to draw a clear
line between influences due to the phonotactic grammar and influences
due to the use of that grammar in a metalinguistic task" (p. 176).
Mary L. Hare, Michael Ford, and William D. Marslen-Wilson in
"Ambiguity and frequency effects in regular verb inflection" report
two experiments (writing to dictation, and primed lexical decision)
designed to examine the question of whether relative past
tense/homophone frequency has any effect on the speed of access for
irregular verbs. "Both experiments show effects of past tense
frequency that are, if anything, stronger in the regularly inflected
items than in irregulars" (p. 196). This means that no dual mechanism
is needed for the lexical representation of regular and irregular
verbs.
"Frequency, regularity and the paradigm: A perspective from Russian on
a complex relation" (Greville Corbett, Andrew Hippisley, Dunstan
Brown, and Paul Marriott) is also concerned with the question of
whether frequency affects only the irregular forms of a lexeme or all
its manifestations. The data come from Russian, a language with
sufficient cells on its noun paradigms for this type of
investigation. The data support the hypothesis that "there is a
relation between absolute plural anomaly and irregularity" (p. 219)
while no evidence is found to relate irregularity to the high relative
frequency of any cell in the paradigm.
Part Three: Phrases and constructions (8 articles)
Daniel Jurafsky, Alan Bell, Michelle Gregory, and William D. Raymond
in "Probabilistic relations between words: Evidence from reduction in
lexical production" examine the Probabilistic Reduction Hypothesis
according to which word forms with a higher probability are more often
reduced. The results of their corpus study suggest that all in all
"more probable words are reduced, whether they are content or function
words" (p. 246).
"Frequency effects and word-boundary palatalization in English" by
Nathan Bush is an analysis of the corpus from CHILDES project. The
palatalization of /d/ and /j/ in *would you* is more regular and
typical than that in *good you*. The study shows that word- boundary
palatalization is more probable when the words occur together with
high frequency.
Catie Berkenfield's "The role of frequency in the realization of
English *that*" addresses the phonological structure and
representation of four types of *that* in English, namely,
Demonstrative pronoun, Demonstrative adjective, Relative clause
marker, and Complementizer. The study shows that as the functional
category of the token becomes more frequent, its vowel duration
decreases. Berkenfield concludes that each function is represented
independently in the lexicon.
"Frequency, iconicity, categorization: Evidence from emerging modals"
(Manfred G. Krug) is concerned with the interaction of discourse
frequency, iconicity, and categorization during early stages of
grammaticalization. He proposes that such structures as BE GOING TO,
HAVE TO, and WANT TO are going through a process of changing their
categorial status under the influence of discourse
frequency. "Frequency of use facilitates (phonetic) ... variation
... . In an iconicity-driven cognitive process, structurally similar
variants are selected, which leads to a convergence of items belonging
to a category" (p. 328).
Joan Bybee in her "Frequency effects on French liaison" makes a
distinction between two types of sandhi variation: (a)
phonetically-conditioned sandhi applying whenever the appropriate
phonetic environment is available, and (b) (external)sandhi NOT
phonetically conditioned so that it applies across word boundaries
only in specific constructions. In case of French liaison, Bybee shows
that the second type only occurs in high frequency constructions. She
concludes that phrases and constructions share many properties with
morphologically complex words: "[i]n a model in which memory storage
includes not just individual words, but also phrases and
constructions, lexicon and grammar are not strictly separated, but are
integrated and subject to the same organizational principles"
(p. 357).
"The role of frequency in the specialization of the English anterior"
by K. Aaron Smith is a diachronic study of the BE/HAVE + PP
constructions in Old and Modern English. Smith argues that the
takeover of the HAVE construction in Modern English is a case of
specialization in which HAVE replaced BE with low frequency words
first and then with high frequency ones. He concludes that morphology
can in many cases be "a diachronic reflex of a more grammaticized
syntax" as frequency makes it possible for language users to store an
entire syntactic construction in their memory.
Joyce Tang Boyland in "Hypercorrect pronoun case in English? Cognitive
processes that account for pronoun usage" examines non-standard
syntactic constructions such as 'the possible misunderstanding between
you and I' or 'thanks to all whom helped me' that show prestige forms
'X AND I' and 'WHOM'. The corpus study of the phenomena suggests that
hypercorrect usage is not only sociolinguistically but also
cognitively motivated: "More frequently encountered and thus highly
activated constructions are more likely to be used subsequently, by
other speakers, in other utterances, and in other clauses" (p. 402).
"Variability, frequency, and productivity in the irrealis domain of
French" by Shana Poplack is a study of natural conversation with the
analytical tools of Variation Theory. The data are taken from Corpus
du francais a Ottawa Hull (1989). Three areas of French
irrealis--the subjunctive mood, the inflected future, and conditional
modality--are examined. As far as the subjunctive is concerned, a
frequency-based analysis is a reasonable account of the facts. For the
other two, however, the analysis fails.
Part Four: General (3 articles)
Gertraud Fenk-Oczlon in "Familiarity, information flow, and linguistic
form" focuses on how cognitive costs, frequency, and linguistic forms
are related. Frequency is the only quantifiable independent variable
in this respect. It lowers cognitive costs, which in turn influence
linguistic forms.
"Emergentist approaches to language" by Brian MacWhinney is a review
of some major issues in the study of language as an emergent behaviour
with six different levels of emergence, namely, evolutionary
emergence, epigenetic emergence, emergence from local maps, emergence
from functional circuits, grounded emergence, and diachronic
emergence.
Oesten Dahl's "Inflationary effects in language and elsewhere" is a
comparison of economy and language in monetary terms: people multiply
conventionally-valued objects (e.g. by over-titling someone: using
*man* and *gentleman* or *woman* and *lady* synonymously) to "buy"
positive reactions. "[I]nflation is the unintended result of (such)
intentional actions" (p. 472).
CRITICAL EVALUATION:
The collection contains very insightful articles on the issues of the
highest interest to phoneticians, morphologists, syntacticians,
cognitive linguists and psycholinguists. They represent the very
healthy attitude of the recent years to focus on the question of
possible relationships between abstract linguistic structures and
issues in performance captured in empirical terms. Although
functionalists have been concerned with language use and its
realisations in grammar, an emergentist, frequency-based analysis of
the issue had never received the attention the topic deserves. This
volume of papers on such issues sheds light on the importance of the
issue, and how a target as easy (in terms of empirical inquiry) as
frequency of linguistic elements can explain structural
complexities. The articles in this volume remind us of the fact that
there are times when the true explanation of the phenomena under study
is so readily available to the researcher that the they tend to simply
ignore it and dig the unexplored corners of world to find an answer.
Although the authors never claim that frequency can explain everything
there in linguistic structures, the very empirical availability of the
issue now seems to pave the way for the unjustified temptation to
stick to frequency as the sole solution to our problems in the study
of structure. There seem to be two major issues left out in this
respect. Firstly, nowhere in the collection they explain why these and
not other linguistic elements have happened to be so highly frequent
in the discourse. Is the frequency of use to be explained in terms of
communicative demands of speakers (an E-language approach), or on the
other hand, some innately available elements of Universal Grammar are
behind it (an I-language approach), or both? In other words, the
authors ignore the question of whether frequency is a cause or an
effect when it comes to linguistic structures. The alternative
explanation is to claim that some cognitive/innate predisposition
towards certain linguistic structures brings about the high frequency
of certain elements but not others. Secondly, no mention is made of
those cases where frequency has no tangible effects on grammar: some
structures stay with us although they are not necessarily high in
frequency. For instance, such structures as "the mouse the cat chased
ran away" (let alone those structures with three NPs followed by three
VPs) do not seem to be very high in frequency nor very readily
available in terms of mental processing (hence not very useful, one
dares say), but still quite grammatical when it comes to
native-speakers' judgements.
Personally I did not like the arrangement of the articles. Perhaps the
collection would be more friendly to a non-specialist if general
articles had appeared first (or perhaps if a more comprehensive
introduction to the subject had opened the volume). Finally, some
articles are still in need of editing and proof-reading. For instance,
on page 438 it reads "[s]ome examples from Fenk-Oczlon (1989a) are
figured in Table 1" while the table in question (p.437) figures some
data from Josselson (1953) and Thorndike and Lorge (1944) instead. Or
Dahl (pages 471 and 473) drops the definite article in the expression
*on the one hand* two times.
REFERENCE
Bowerman, M. 1996. The origins of children's spatial semantic
categories: cognitive versus linguistic determinants. In: Rethinking
linguistic relativity, J. J. Gumperz and S. Levinson, eds. Cambridge:
Cambridge University Press.
ABOUT THE REVIEWER
Dr. Ahmad R. Lotfi, Assistant Professor of linguistics at the English
Department of Azad University (Khorasgan, IRAN) where he teaches
linguistics to graduate students of TESOL. His research interests
include (minimalist) syntax, second language acquisition, and Persian
linguistics.