| AUTHOR: Malchukov, Andrej L.
TITLE: Dependency Reversal in Noun-Attributive Constructions
SUBTITLE: Towards a typology
SERIES TITLE: LINCOM Studies in Language Typology 03
PUBLISHER: Lincom GmbH
Antonis Polentas, Department of Language and Linguistics, University of Essex (UK)
This is a short monograph, more like the print-out of a long paper, about the
typology of a rarely investigated phenomenon, dependency reversal in
noun-attribute (DRNA) constructions (this review refers to the study as
published in 2000). DRNA constructions present a reversal of the dependency
relation between the attribute and the noun in a noun phrase (NP). Specifically,
the attribute is marked as the (formal) head of the NP and the noun, which is
the semantic head, is demoted to the role of the dependent. In the cases
discussed in this study and elsewhere (Ross 1998), the DRNA construction looks
like a possessive NP, with the original attribute being the possessed and the
noun as the possessor.
The first half of the work is devoted to the terminology used by Malchukov
(section 1) and to the description of his data. Due to the rarity of DRNA
constructions, he decided to present all six cases where he had spotted this
type of construction (sections 2-7) and in a further short section he discusses
the possibility of similar data in other languages (section 8). The second part
of the work contains a discussion of how to build a typology of DRNA
constructions (section 9) and some suggestions of possible explanations of the
patterns observed (section 10).
In the introductory section, Malchukov situates his study in the context of
related typological work (Croft 1995 on the typology of genitive constructions,
Plank's survey of ''Suffixaufnahme'' constructions, work on possessive
constructions, and Nichols' (1986) study of head vs. dependent marking
languages). He defines the domain of his inquiry as the reversal of dependency
marking between the noun and its attribute in the NP without any effect on the
meaning of the NP. In English and other European languages this phenomenon does
not manifest itself in such a clear way as in the languages examined by
Malchukov, but examples such as (1) and (2) come close to it:
(1) Foolish Hobbit!
(2) Fool of a Hobbit!
(see Foolen 2004 for an examination of such constructions -which he calls
''expressive binomial NPs'' - in Germanic and Romance languages with reference to
Malchukov's work).The definition of the domain of inquiry limits to a certain
extent the data that fall under its scope, as DRNA without effect on meaning and
marked in a formally clear way, is quite rare crosslinguistically. In sections
2-7 Malchukov presents the data from six languages that can be classified as
exhibiting this type of dependency reversal in the NP. Because of the small
amount of data, there is no discussion of sampling or statistical analysis, as
is usually the case in typological studies.
Section 2 presents DRNA data from Tungunsic languages (on which Malchukov has
worked extensively). It is shown that there is variation among the members of
this relatively small (12) language family with respect to DRNA constructions:
Oroch exhibits a clear case of adjective-noun dependency reversal, but in the
other languages this is possible only with participles.
Section 3 contains a 6-page long discussion of DRNA construction data from
Aleut, a highly agglutinative, paleosiberian language where the DRNA pattern is
the only option in attributive NPs (i.e. they are like the possessive NPs with a
Section 4 presents data from Hausa (western Chadic), where the prenominal
adjectives in attributive NPs are marked with a linker, very much like the head
in the possessive NPs (although Malchukov points out a number of differences
between the DRNA and the possessive constructions).
Section 5 contains a brief discussion of DRNA data from Gude, another Chadic
language. Although data like the following appear to exhibit a clear DRNA instance:
(2) minä nga Musa
wife of Musa
(3) buuta nga darä
cheap of hat
''Buuta'' is classified as an adjectival noun. Adjectival nouns form a closed
class in this language. Nevertheless, as Malchukov points out, the data from
Gude are not complete, so it is not clear how they fit in the DRNA typology.
Data from Chinook (Penutian) are discussed next. Based on remarks by Boas,
Malchukov points out that the DRNA pattern seems to be the most common one to
render property words in this language (following the possessive pattern).
The Latin data presented in the 7th section are likely to be the more familiar
ones - as Latin is the best described among the languages discussed in this
study - but they may also be the most puzzling seen from the perspective of a
DRNA typology The type of data presented are of the type sapientia Laeli (lit.
'wisdom of Laelus') meaning 'wise Laelius' (sapiens Laelus).
Section 8 concludes the description of the data with a brief discussion of a
number of candidates for the DRNA typology from a variety of other languages.
Unfortunately, the section is very brief (at about 1.5 pages long) without much
discussion of the relevant data. It is more of a section with suggestions for
further research on the topic. In a postscript Malchukov refers to Ross (1998)
on ''possessive-like attributive constructions'' in Western Melanesian languages
presenting a wealth of data related to Malchukov's study. It's a pity that time
limitations did not make it possible for the author to take them into account in
The second part of the study contains two sections: section 9 which is about the
typology of the data discussed in the previous sections and section 10 which
presents a sketch of potential explanations of the typology of DRNA.
Section 9 is the longest one of the monograph and discusses ways in which the
data presented in the previous sections can be classified. A first subsection,
builds on Plank's study of ''Suffixaufnahme'' (1995) to attempt a taxonomy of DRNA
constructions with respect to four parameters: linear order (in all cases DRNA
constructions show the same patterns as possessive NPs), locus of marking, type
of marking, and degree of fusion. There are no clear findings in this section
although some correlations seem to emerge (the one between the locus and the
type of marking). Next Malchukov discusses the classes of attributes that take
part in DRNA constructions. As far as the six cases he examines are concerned
these can be adjectives and/ or participles. Quite interesting is the discussion
of the degree of attribute's upgrading and recategorization where the author
places five of the six languages (i.e. all except Gude due to lack of data) in a
scale according to these two parameters. Once more variation is observed, even
among this small number of languages. Finally, it is shown that languages differ
also with respect to whether the DRNA pattern is a basic or a marginal/
restricted one in the language. The next subsection (9.2) explores how one could
draw an ''upward taxonomy'' of the DRNA patterns found in the sample. Malchukov
shows that two (binary) parameters, upgrading of the attribute and downgrading
of the noun, can be combined giving a four way classification of dependency
patterns in attributive NPs. The DRNA pattern is one where both parameters have
positive value (so it involves both upgrading of the attribute and downgrading
of the head noun). The ''normal'' (iconic) pattern is the reverse, where both
values are negative. The interesting aspect of this way of looking at
attributive NPs is that it provides a way to look at the relation of DRNA
pattern with other constructions which show diversions from the ''iconic''
pattern. In the rest of the subsection Malchukov elaborates further on the
upward taxonomy of DRNA constructions. The final subsection of section 9 (9.3)
contains remarks about constructions that share some characteristics with the
Section 10 presents the sketch of three attempts to provide a motivation for the
DRNA construction. First, a trivial observation is that the
under-differentiation between adjectives and nouns is a condition for the DRNA
pattern, but it cannot provide a clear explanation of its appearance. Functional
motivations seem more promising as they analyze DRNA constructions as the result
of the pragmatic salience of the attribute. Finally Malchukov discusses briefly
how the DRNA patterns may have emerged in the history of the languages as the
result of recategorization and reanalysis resulting in the attributive
constructions becoming similar to the possessive ones.
In the endnote section the book contains a list of the abbreviations used (the
only omissions have to do with glosses explained in the main text) and a list of
references. The size of the text does not make necessary an index of linguistic
terms, authors and languages, although this might have been useful.
In spite of its small size, Malchukov's monograph is of interest as it discusses
a puzzling phenomenon that hasn't been analyzed at any depth. The work is worth
reading as it can be seen as offering a window on a ''project in progress,'' a
glimpse at the typologist's workbench. The number of languages discussed may be
small, but they have different typological profiles and they belong to different
families. Malchukov's discussion touches on issues related to parts of speech
(specifically adjectives and their relation to nouns), possessive constructions,
internal relative clauses, etc.
This short study looks somehow incomplete and probably it needed a longer
discussion of the cases that Malchukov identifies as potential candidates for
DRNA or of cases that come close to meeting his definition. The discussion of
the six cases of DRNA constructions is rather short and maybe a more extensive
analysis of the data from each language would have made it easier for the reader
to follow the author's points.
Typos are very few and do not disturb the reading of the text. The quotations
from German texts are not translated, which might be a problem for some readers.
The exposition is clear and it presents a good case of how it is to plan a
typological project. The cross-linguistic rarity of the DRNA as discussed by
Malchukov raises some questions for theoretical linguistics as well (and this is
also the case with ''Suffixaufnahme'' cases, as discussed in Plank 1995). One
would wish that the author (or someone else) would undertake the more ambitious
task to incorporate his findings on DRNA in the typological discussions of the
structure of possessive NPs, attribute constructions and the general discussion
of the nature of syntactic categories and of the notion of ''head'' (e.g. Corbett
et al. 1993). More recently Foolen (2004) used analyzed expressive binomial NPs
(i.e. constructions of the type ''a dream of a car'', ''a hell of a job'' etc.)
using insights from Ross (1998) and Malchukov's typology of DRNA patterns.
Generally, reading Malchukov's explorative study makes one wish for a longer
monograph incorporating findings that have been reported in the last ten years
and maybe a longer discussion of the motivation of the DRNA pattern and its
relation to possessive NPs. It should be of interest to linguists working on the
typology of attributive NPs, on the nature of syntactic and semantic heads, on
the structure of the NPs, on the nature of syntactic categories, etc. The text
as it is offers an interesting discussion of a relatively rare phenomenon and
explores ways for its typological classification. It opens the way for further
research in an under-explored area of linguistic typology.
Corbett, G.G., Fraser, N.M., and S. McGlashan (eds.). (1993). _Heads in
grammatical theory_. Cambridge: Cambridge University Press.
Croft, W. (1995). Modern syntactic typology. In Shibatani, M. and Th. Bynon
(eds.), _Approaches to Language Typology_. Oxford: Clarendon Press. (pp. 85-145).
Foolen, A. (2004). Expressive binomial NPs in Germanic and Romance languages. In
Radden, G. and K.-U. Panther (eds.), _Studies in linguistic motivation_. Berlin:
Mouton de Gruyter. (pp. 75-100).
Nichols, J. (1986). Head-marking and dependent-marking grammars. _Language_ 62:
Plank, F. (1995). (Re-)Introducing Suffixaufnahme. In Plank, F. (ed.), _Double
case: agreement by Suffixaufnahme_. New York: Oxford University Press (pp. 3-113).
Ross, M. (1998). Possessive-like attribute constructions in the Oceanic
languages of Northwest Melanesia. _Oceanic Linguistics_ 37, 2, 234-275.
ABOUT THE REVIEWER
Antonis Polentas is a PhD student at the University of Essex (UK). His
dissertation is about the application of the theory of Simpler Syntax in the
case of Modern Greek clitic constructions. His interests include
Constraint-Based linguistics (HPSG, Simpler Syntax), Morphology, and linguistic