|EDITORS: López-Couso, María José; Seoane, Elena
TITLE: Rethinking Grammaticalization: New perspectives
SERIES: Typological Studies in Language (TSL) 76
PUBLISHER: John Benjamins
EDITORS: Seoane, Elena; López-Couso, María José
TITLE: Theoretical and Empirical Issues in Grammaticalization
SERIES: Typological Studies in Language (TSL) 77
PUBLISHER: John Benjamins
Heiko Narrog, Tohoku University, Japan
The two volumes reviewed here offer a selection of papers from the Third
International Conference New Reflections on Grammaticalization (NRG 3), held at
the University of Santiago de Compostela in July 2005. Just as the conference
was held for the third time, Benjamins publishes a selection of papers for the
third time (the first was edited by Wischer and Diewald in 2002 with the title
_New Reflections on Grammaticalization_, and the second by Fischer, Norde and
Perridon in 2004 with the title _Up and down the Cline – The Nature of
Grammaticalization_). Also, just as the number of participants and presentations
in the NRG conferences has grown steadily, this selection apparently had to be
split up into two volumes. Together they offer 30 papers on roughly 700 pages of
print. Although the titles of the two volumes may seem to suggest a different
orientation, this is not really the case. Supposedly the articles were divided
as they are for a practical reason.
The first volume (TSL 76), after the introduction by the editors, starts with an
article by Walter Bisang on ''Grammaticalization and the areal factor: The
perspective of East and mainland Southeast Asian languages'' (15-35). Bisang
argues that grammaticalization in these languages shares particular properties,
such as the lack of obligatory categories, predominance of pragmatic inference,
existence of rigid word-order patterns, and no or only limited coevolution of
form and meaning. He suggests that these properties are not sufficiently
accommodated in prevalent theories of grammaticalization and emphasizes the need
for areal studies on grammaticalization.
Philippe Bourdin, in his paper ''On the grammaticalization of 'come' and 'go'
into markers of textual connectivity'' (37-59), dealing mainly with African
languages, discusses clines of grammaticalization of verbs meaning 'come' and
'go' into sequential markers, and the morphosyntactic and semantic variation
seen in these clines.
Zygmunt Frajzyngier pleas for an expansion of the concept of grammaticalization
in his article ''Grammaticalization, typology and semantics: Expanding the
agenda'' (61-102). In his view, the study of grammaticalization should also
include grammatical devices such as word order, phonological means (i.e.,
intonation, vowel reduction etc.) and repetition. Such devices share with
typical grammaticalized items motivations for grammaticalization such as a
'principle of functional transparency', which ''states that the function of every
form in an utterance must be transparent to the hearer'' (72). The author
provides copious examples from African languages.
Livio Gaeta (''Mismatch: Grammar distortion and grammaticalization'' (103-127)),
writes on mismatches in grammar as the result of grammaticalization. These
occur, for examples, when grammaticalization introduces new patterns into a
language which are incompatible with the old ones, or when grammaticalization
fails to extend over a whole category, leaving the category with mixed old and
Anna Giacalone-Ramat, in ''Areal convergence in grammaticalization processes''
(129–167) takes up Bisang's topic of areality under quite different premises.
Four cases of areal grammaticalization in Standard Average European and adjacent
language areas including the development of 'have'-perfects and articles from
demonstratives are studied. The author shows that even if grammaticalization is
contact-induced, basic principles of grammaticalization still hold.
Kaoru Horie's ''The grammaticalization of nominalizers in Japanese and Korean: A
contrastive study'' (169-187) is the first of a number of papers on
nominalization in East Asia, stemming from a workshop on this topic in NRG 3.
Horie compares Japanese and Korean nominalizers and concludes that their
grammaticalization and pervasiveness in grammar is more advanced in Japanese, in
consonance with other categories such as aspect, where grammaticalization is
more advanced in Japanese as well.
Other papers from this workshop include ''Nominalizations in Bodic languages''
(219–237) by Michael Noonan, ''On the rise and fall of Korean nominalizers''
(239–264) by Seongha Rhee, ''The grammaticalization of clausal nominalizers in
Burmese'' (265–288) by Andrew Simpson, and ''The development of nominalizers in
East Asian and Tibeto-Burman languages'' (309–341) by Foong Ha Yap and Stephen
Matthews. As stated before, the reasons for the particular division of the
articles into the two volumes are not really apparent, but at least it can be
said that all the papers from the nominalization workshop are found in the first
volume. Noonan shows that nominalization in Bodic languages is no less prolific
than in Japanese or Korean. Several specific phenomena such as the
nominalization-relativization syncretism are discussed, and their spread is
attributed to areal rather than genetic factors, as in previous research. Rhee
more or less provides a history of nominalizers in Korean, their origin and
development, their functional competition, and various ways in which they
specialize. Simpson follows the grammaticalization of nominalizers in Burmese.
At the end of their development, they lose nominal properties and mark verbal
categories, specifically mood. Finally, Yap and Matthews take the broadest
perspective in trying to identify common paths of the grammaticalization of
nominalizers across East Asian languages. Beyond this task, they shortly discuss
contentious questions concerning nominalizers, such as their relationship with
tense/aspect and evidentiality/mood and with copulas.
TSL 76 has still two other papers. Tania A. Kuteva, one of the invited speakers
at the conference, writes ''[o]n the _frills_ of grammaticalization'' (189–217).
She shows that contact-induced grammaticalization is often accompanied by
redundant accretion of grammatical material. However, as also claimed by
Giacalone Ramat, the process of grammaticalization follows the same principles
as in purely language-internal grammaticalization. The overlap of various means
of marking the same category can be observed both in diachronic stages of a
language, and synchronically in buffer zones in geographical space.
Lastly, Ferdinand von Mengden concerns himself with ''The grammaticalization
cline of cardinal numerals and numeral systems'' (289–308). Cardinal numerals
often start out as body parts, then become a numeral, and later even an affix
that encodes internal arithmetic relations in complex numerals such as –ty in
forty. The author suggests that there could be bidirectionality in the
development between low numerals and pronominal elements.
The second volume (TSL 77), after an introduction which summarizes the papers,
starts with Peter Andersson on ''Swedish _må_ and the (de)grammaticalization
debate'' (15-32). The modal verb _må_ has been claimed by some scholars to be an
example of degrammaticalization as it appears that a lexical meaning developed
out of a grammatical meaning. Andersson shows that this claim is due to an
insufficient consideration of the actual historical data by those who have put
Laurel J. Brinton's ''Where grammar and lexis meet: Composite predicates in
English'' (33-53) deals with the tricky question whether composite predicates are
to be viewed as lexicalizations or grammaticalizations. Brinton argues for a
gradient view of lexicality and grammaticality, and suggests that some type of
constructions exhibit more features of lexicalization, while others, especially
those involving light verbs, are rather grammaticalizations.
In the next paper, by Bert Cornillie ''[o]n the grammaticalization and
(inter)subjectivity of evidential (semi-)auxiliaries in Spanish'' (55-76), the
author challenges Traugott & Dasher's (2002) notion that in semantic change
subjectification always precedes intersubjectification. The Spanish
(semi-)auxiliaries _parecer_ 'seem' and _resultar_ 'turn out to' apparently
offer counterevidence. It should be noted, though, that Cornillie's concept of
intersubjectivity and intersubjectification do not necessarily coincide with
that of Traugott but is closer to that of Nuyts (2001).
Gabriele Diewald and Gisella Ferraresi are the authors of ''[s]emantic, syntactic
and constructional restrictions in the diachronic rise of modal particles in
German: A corpus-based study on the formation of a grammaticalization channel''
(77-110). In this article they analyze the historical development of two German
modal particles, _schon_ and _eben_, particularly focusing on the specific
contexts of their grammaticalization. As a result they submit a common path for
the grammaticalization of German modal particles, which are functionally and
semantically quite diverse.
Andreas Dufter and Elisabeth Stark, in ''Double indirect object marking in
Spanish and Italian'' (111-129) compare indirect object doubling (e.g. Spanish
_Me gusta a mí_ 'I like it') in Spanish and Italian, showing that it is more
common in Spanish. A historical investigation of the phenomenon reveals that in
Spanish its frequency has been more or less stable, these constructions have
become more and more restricted over the past centuries in Italian.
''The emergence of particle clusters in Dutch: Grammaticalization under adverse
conditions'' by Jack Hoeksema (131-149) deals with combinations of two or more
particles, which form units in their own right, and are more specialized
functionally and stylistically than the single particles from which they are
derived. Hoeksema sees their development as part of a general drift towards
specialization in the lexicon.
Dmitry Idiatov, in ''Antigrammaticalization, antimorphologization and the case of
Tura'' (151-169) argues that grammaticalization should be separated as a concept
from morphologization. The morphological development of affixes, for example,
would fall under the latter and not under the former label, since
grammaticalization is only concerned with the development of grammatical meaning
obligatory in a grammar. Accordingly, most cases traditionally labeled as
'antigrammaticalization' or 'degrammaticalization' would be in fact mere
'antimorphologization'. That process is exemplified with the development of a
suffix in the Eastern Mande Language Tura.
Jurgen Klausenburger asks ''Can grammaticalization be parameterized?'' (171-182).
In this article, he confronts a generative analysis of the development of the
Romance future by Roberts and Roussou (2003) with a functional analysis based on
an updated version of the synthesis/analysis cycle and the invisible hand
concept of Keller (1994). He argues that the latter approach is more economical
Ekkehard König and Letizia Vezzosi, in ''Possessive adjectives as a source of
intensifiers'' (183-206) trace development of Modern English intensifier _own_
from _agen_, the past participle of the verb _agan_ 'own, possess'. The fact
that similar developments can be found in numerous other languages as well
points to the existence of a cross-linguistic path from possession to
''Information structure and grammaticalization'' is the topic of a paper by
Christian Lehmann (207-229). In this paper, he shows that information structure
determines the course of grammaticalization but at the same time information
structure, as coded through grammatical constructions, is subject to
grammaticalization itself, leading to a relationship of mutual dependence.
Torsten Leuschner, in ''From speech-situation evocation to hypotaxis: The case of
Latin _quamvis_ 'although''' (231-252) follows the path of _quamvis_ from
free-choice quantification to concessive subordinator. Free-choice quantifiers
are a typical source for concessive subordination and Leuschner highlights a
number of specific aspects of their development, such as layering in the
inventory of subordinators and the role of speech-situation evocation.
In ''Grammaticalization waves: The Russian subjunctive mood and person/number
marking'' Jens Nørgård-Sørensen (253-268) argues for a concept of waves of
grammaticalization. A specific change in grammar seldom comes alone but triggers
changes in other parts of the grammar, as the value of one item in a grammatical
paradigm affects the value of other items. This point is illustrated with the
development of Russian subjunctive mood.
Regina Pustet writes on ''[d]iscourse frequency and the collapse of the
adposition vs. affix distinction in Lakota'' (269-292). She calls into question a
strict distinction between adposition and case markers. In Lakota, the same
element may operate as both postposition and case-marking suffix, with discourse
frequency as a determining factor. Interestingly, Lakota postpositions may
become affixes without phonological reduction.
''On the grammaticalization of the Spanish expression _puede que_'' (292-314) is
the title of a paper by María José Rodríguez Espiñeira and Belén López Meirama.
In this paper, they trace back the development of _puede que_ 'it may be that'
to the construction _puede ser que_ with the same meaning. Its development
exhibits typical features of grammaticalization such as formal reduction and
increased subjectivity. An intriguing question is why this construction has
stopped short of becoming an adverb and the authors cite competition from other
common epistemic adverbs as a probable reason.
Günter Rohdenburg, in ''On the history and present behaviour of subordinating
_that_ with adverbial conjunctions in English'' (315-331) analyzes the
development of adverbial conjunctions with _that_, such as _given that_,
_assuming that_, or _except that_. Contrary to the assumption that these
conjunctions start out with accompanying _that_ and then eventually shed it in
the course of grammaticalization, Rohdenburg finds examples where _that_ has
become obligatory again. It turns out that this apparent degrammaticalization
can be linked to a decline of the use of these forms as conjunctions.
In the final paper of the collection, on ''[t]he regrammaticalization of linking
elements in German'' (333-355), Heide Wegener deals with linking elements between
compound nominals in German, the origin of which has been contested. Wegner
argues that they constitute a case of regrammaticalization and exaptation of old
Indo-European stem suffixes that indicated noun class and later lost this
original function. The same suffixes were also reanalyzed as plural markers in
German. Each of the volumes ends with a language index, a name index, and a
Given the scope of this volume, an evaluation of each paper is clearly not
feasible. Instead, I wish to comment on the enterprise as a whole. The
conference series _New Reflections on Grammaticalization_ has seen a constant
increase in participants and presentation since its start in 1999. These two
volumes reflect this increase both in quantity but also in variety and depth of
topics studied under the label of grammaticalization. For example, half of the
papers selected here primarily deal with non-Indo-European languages of Africa,
America and Asia. Most of the articles on Indo-European languages make use of
diachronic corpora, contributing to issues of grammaticalization with historical
evidence. Particularly ''hot'' topics according to the selection presented in this
volume were the question of the areal nature of grammaticalization and
concomitantly the relationship between language-internal grammaticalization and
contact-induced grammaticalization, the boundaries of grammaticalization,
nominalization, and the perennial question of counterexamples to the
unidirectionality of grammaticalization, i.e. so-called degrammaticalization or
antigrammaticalization. The relatively large number of papers on nominalization,
which make up about half of the first volume, might be due to the fact that the
workshop for some reason decided against publishing its own volume. With respect
to de- or antigrammaticalization, it seems that the number of valid
counterexamples presented does not significantly increase. On the contrary, some
authors argue against antigrammaticalization analyses in cases where they were
previously suggested. That is, the tendency of unidirectionality in
grammaticalization is still generally regarded as valid by scholars in this
field. Rohdenburg, however, presents an apparent reversal of grammaticalization,
and Wegener a plausible case of exaptation. One tendency which is not, or only
indirectly, reflected in these volumes is the increasing interest of generative
scholars in grammaticalization (in fact, Paul Kiparsky was one of the invited
speakers at the conference but apparently chose not to contribute to this
Readers may also note that only relatively few of the founding figures of
grammaticalization theory are represented among the contributors. Bernd Heine
and Elizabeth Traugott, for example, have recently ventured into adjacent and
related areas such as contact-induced language change, and language evolution
(Heine) or semantic change and constructions in change (Traugott). Nevertheless,
the contributions in this publication show that grammaticalization has clearly
caught on among younger generations of scholars and has become a vibrant field.
The quality of the articles is, as far as this reviewer can judge, generally
excellent, and they have been carefully edited by Seoane and López-Couso.
Keller, Rudi. (1994) _On Language Change. The Invisible Hand in Language_.
Nuyts, Jan. (2001) _Epistemic Modality, Language, and Conceptualization: A
Cognitive-Pragmatic Perspective_. Amsterdam: John Benjamins.
Roberts, Ian and Anna Roussou. (2003) _Syntactic Change. A Minimalist Approach
to Grammaticalization_. Cambridge: Cambridge University Press.
Traugott, Elizabeth & Richard D. Dasher. (2002) _Regularity in Semantic Change_.
Cambridge: Cambridge University Press.
ABOUT THE REVIEWER
Heiko Narrog is an associate professor at Tohoku University, Japan. His research
interests include historical linguistics, syntax and semantics, modality,
linguistic typology, and the Japanese language.