Date: Sun, 17 Apr 2005 17:11:18 -0800
From: Gary Holton
Subject: Adjective Classes: A cross-linguistic typology
EDITORS: Dixon, R. M. W.; Aikhenvald, Alexandra Y.
TITLE: Adjective Classes
SUBTITLE: A cross-linguistic typology
SERIES: Explorations in Linguistic Typology
PUBLISHER: Oxford University Press
Gary Holton, Alaska Native Language Center, University of Alaska Fairbanks
Given the increasing tendency to recognize linguistic categorization as
non-universal (cf. Croft 2001), it may seem somewhat atavistic to publish
a volume with such a bold categorical title as this one. Beginning with
the claim that "a distinct word class 'adjectives' can be recognized for
every human language" (1), this book goes on to present cogent
argumentation for the universality of the adjective class based on data
from a broad range of languages. The papers in the volume were originally
presented at a workshop on adjective classes held at the Research Center
for Linguistic Typology in August 2002. It is comprised of an introductory
chapter, thirteen chapters which survey the adjective class in particular
languages, and a concluding chapter.
Dixon's introductory chapter, "Adjectival classes in typological
perspective", is clearly destined to be a classic in the field, building
on Dixon's (1977) seminal paper on adjectives. Dixon's thinking on the
adjective class has evolved substantially over the past three decades.
Whereas Dixon's earlier work remains agnostic on the universality of
adjectives, here he clearly convinced that appropriate criteria can be
found in all languages to delineate an adjective class. Of course, testing
this hypothesis would necessitate the investigation of adjectives in all
of the world's languages. But the circumstantial evidence presented in
this volume is fairly convincing. This is because many of the languages
selected are difficult cases. That is, the criteria for delineating the
adjective class-differentiating it from either noun or verb-are quite
subtle. One comes away with the feeling that if adjectives can be found in
these languages, then surely they can be found in all languages.
Dixon offers several criteria for distinguishing adjectives. These include
some fairly standard morphological criteria such as restricted
inflectional possibility, as well as more semantically based criteria such
as the ability to participate in comparative constructions. One of the
major difficulties with proposing grammatical criteria by which to
distinguish adjectives lies in showing that such criteria do not
ultimately derive from adjectival semantics. Dixon is very much aware of
this problem. For example, among the criteria Dixon suggests for
distinguishing nouns and adjectives is their semantic behavior under
reduplication: reduplication of nouns tends to indicate plural, while
reduplication of adjectives tends to indicate intensity. One might be
tempted to attribute the effect of reduplication on putative adjectives to
their inherent semantics rather than their word class. Yet Dixon cites as
counter-evidence the example of Dyirbal, in which reduplication of both
nouns and adjectives indicates plural, not intensity (25). In particular,
reduplication does not serve to distinguish nouns and adjectives in
Dyirbal. But more importantly, reduplication of adjectives is not
universally associated with intensity semantics, thus the reduplication
criterion can be considered a grammatical rather than semantic property.
The languages surveyed in the volume are by no means intended as a
representative areal or genetic sample. They consist of 1 language from
Europe (Russian); 1 from Africa (Wolof); 4 from Latin America; and 7 from
Asia (1 Austronesian, 2 Tibeto-Burman, 1 Tai, 1 Austroasiatic, Japanese,
Korean). The editors' intention is not so much to achieve typological
diversity but rather to present a range of criteria and argumentation in
support of adjective categories in individual languages.
The first two papers examine languages with multiple adjective classes.
Anthony Backhouse describes adjectives in Japanese; Carol Genetti &
Kristine Hildebrandt describe adjectives in Manange (Tibeto-Burman). In
both languages one set of adjectives is more nouny while the other is more
verby. Syntactic criteria are used to distinguish verby adjectives from
true verbs. Genetti & Hildebrandt maintain that this is a more
descriptively accurate approach which "elucidates the true nature of the
lexical category" as a hybrid between verb and adjective (95).
Three papers describe languages with small, closed adjective classes.
Alexandra Aikhenvald's "The Adjective Class in Tariana"; Nora
England's "Adjectives in Mam"; and R. M. W. Dixon's "The Small Adjective
Class in Jarawara" each present clear evidence for a distinct class of
adjectives. For example, after carefully discussing the morphological and
syntactic features of nouns and verbs in Tariana, Aikhenvald cites at
least six morphological properties which distinguish adjectives from nouns
and at least three morphological properties which distinguish adjectives
Paulette Levy's "Adjectives in Papantla Totonac" and Ho-Min Sohn's "The
Adjective Class in Korean" each examine a language which has been
previously described as not having a distinct adjective class. Adjectives
in Totonac have previously been described as a sub-class of nouns;
adjectives in Korean have previously been described as a sub-class of
verbs. Both authors provide more subtle criteria by which adjectives can
be recognized as a distinct category. For example, Sohn cites differing
forms of the non-past indicative affix as a feature which distinguishes
verbs from adjectives in Korean.
Greville G. Corbett's "The Russian Adjective: A pervasive yet elusive
category" is a unique contribution in that it examines a well-studied
language which unarguably has a distinct adjective category. Corbett's
exhaustive survey reveals a rather surprising lack of homogeneity in the
Russian adjective class.
The final five papers actually argue against the universality of the
adjective category by providing evidence that adjectives in the languages
in question should be considered a subcategory of verbs rather than a
separate word class. These papers include: Fiona McLaughlin's "Is there an
Adjective Class in Wolof?"; Catriona Hyslop's "Adjectives in North-east
Ambae"; Nicole Kruspe's "Adjectives in Semelai"; Randy J. LaPolla &
Chenglong Huang's "Adjectives in Qiang"; and N. J. Enfield's "Adjectives
in Lao". The authors of these papers generally find at least some criteria
by which to distinguish adjectives from verbs but ultimately choose to
reject these criteria, citing the overwhelmingly verby nature of the
class. McLaughlin's conclusion is typical. She notes of the adjective-like
words that there are "some subtle differences in their behavior,
especially in relative clauses, which set them apart from verbs" but "the
extent of class membership is not completely clear since many lexical
items may be used either as adjectives or as non-adjectival verbs" (261).
Similarly, Kruspe classifies Semelai adjectives as a sub-class of verbs
based on their "strong verb-like behavior" (305). The editors clearly
disagree with the conclusions reached by the authors of these last five
papers. Nevertheless, the decision to present these papers is commendable
and offers the reader direct insight into the difficulties of applying
The volume concludes with a summary essay by John Hajek, who acknowledges
that the central thesis of this volume--that all languages have an
adjective class--is "challenging and provocative" (348). To some extent
the claim can always be religiously defended: presented with a lack of
evidence for an adjective class in a particular language, proponents of
the universality hypothesis can respond simply that "more subtle testing"
(348) is required. Ultimately, Hajek labels the reluctance on the part of
some authors to recognize a separate adjective class as Eurocentric. While
this label may be needlessly inflammatory, Hajek does well to point out
that the tendency to ascribe adjectives as subclass of a major word class,
rather than a distinct class, occurs more often in languages in which
adjectives share many properties with verbs. Where adjectives share many
properties with nouns, as in European languages, authors seem much more
willing to ascribe them to a distinct class.
Hajek goes on to review the criteria employed by the individual authors to
distinguish adjectives. Since the language sample is not representative,
this review should be treated with some caution, but, Hajek nevertheless
draws some interesting conclusions regarding the relative importance of
the various criteria employed for delimiting the adjective category. For
example, all of the authors of the individual language studies use the
distinction between intransitive predicate and copula as a basis for
delimiting word classes.
Hajek closes his summary with a discussion of attributive-only adjective
classes, a subject which receives little attention in this volume. In
fact, the phenomenon may be more common than the evidence of the thirteen
languages surveyed in this volume suggests. Athabaskan languages provide
one example. In Tanacross adjectives are generally related to verbs but
can be distinguished from them by their lack of verbal prefix morphology.
Verbs are used predicatively; adjectives are used attributively.
This volume will clearly be useful to anyone interested in adjectives and
the nature of linguistic categorization, but Dixon & Aikhenvald's primary
contribution is to encourage a reexamination of adjectives in the world's
languages. This reviewer has found that to be a useful exercise. After
overcoming my initial skepticism and reading on through the book, I
eagerly returned to some of my own field data and was surprised to find
myself looking anew at adjective classes. In Holton (1999) I argue quite
strongly against the existence of an adjective class in Tobelo (West
Papuan), noting that words referring to property concepts may be expressed
either as nouns or as verbs, depending on discourse context. In Tobelo
most roots can occur with either noun-like or verb-like morphology, and
adjectives are no exception. However, property concept words can be
distinguished by their ability to participate in scalar constructions,
even though they may occur in both nouny and verby scalar constructions.
By adopting the scalar construction as criterial, it is possible to
delimit a distinct class of adjectives in Tobelo.
The choice to recognize a distinct adjective class in Tobelo--as in any
language--remains to some extent an artifact of the descriptive framework
chosen. Ultimately, our understanding of adjectives must be informed by a
clear understanding of just how to delimit lexical categories. This volume
will surely encourage further work toward that end.
Croft, William. 2001. Radical Construction Grammar: Syntactic Theory in
Typological Perspective. Oxford: Oxford University Press.
Dixon, R. M. W. 1977. Where have all the adjectives gone? Studies in
Holton, Gary. 1999. Categoriality of property words in a switch-adjective
language. Linguistic Typology 3.341-360.