Review of Degrammaticalization
|AUTHOR: Muriel Norde
PUBLISHER: Oxford University Press
Peter M. Arkadiev, Institute of Slavic Studies, Russian Academy of Sciences, Moscow
The book under review treats the phenomenon of degrammaticalization, 'the ugly
duckling of grammaticalization studies', as the author puts it on p. 1.
Degrammaticalization, the process whereby a linguistic unit becomes less
grammatical, runs counter to the common assumption that grammaticalization is
unidirectional, i.e. that 'there is only left-to-right movement' along the
following cline (Hopper & Traugott 2003: 7):
content item > grammatical word > clitic > inflectional affix
Unfortunately, this assumption has sometimes become a dogma rather than a
hypothesis subject to empirical testing. Hence the status of
degrammaticalization has been rather controversial in the discourse of
historical linguistics of the last several decades: some scholars, like
Christian Lehmann (1995 : 16-19), have expressly denied its existence
altogether; others have dismissed it as statistically insignificant (Heine &
Kuteva 2002: 11); still others have used extant examples of degrammaticalization
as a strong argument against unidirectionality in general (Newmeyer 1998: 263).
The aim of Muriel Norde's book is thus twofold. One the one hand, it attempts to
define degrammaticalization in a meaningful way and to classify attested
examples under the proposed definition. On the other hand, this book provides a
critical and informative overview of the major methodological and conceptual
problems faced by contemporary grammaticalization studies, providing a better
understanding of the phenomenon of grammaticalization.
The book consists of seven chapters (including Introduction and Conclusions),
references and indices. The longish Introduction (pp. 1-47), after briefly
presenting the notion of degrammaticalization and problems associated with it,
gives a useful overview of terminology of grammaticalization studies and the
various meanings of the term 'grammaticalization'. In addition, Norde discusses
the notion of 'lexicalization' with its different senses, and its relation with
grammaticalization, especially in the domain of historical development of
derivational morphology. The three notions important for the discussion of
grammatical change, i.e. gradualness, reanalysis and analogy, are also defined
and discussed. A subsection is devoted to less clear types of
grammaticalization, i.e. the so called 'secondary grammaticalization' (when an
already grammatical item becomes more grammatical), 'pragmaticalization'
(development of discourse markers out of content words or grammatical markers)
and clause combining. Norde also briefly discusses the role of context in
grammatical change and the 'constructional' approach to grammaticalization.
A large part of the chapter is devoted to the discussion of various
methodological issues. Norde acknowledges the status of grammaticalization as a
linguistic process in its own right, and not as a mere epiphenomenon of a number
of independent primitive changes such as semantic extension, phonetic reduction
and syntactic reanalysis (for the advocates of this view, see Newmeyer 1998,
Fischer 2000, and Joseph 2001). Though grammaticalization is a process
irreducible to independent changes, the following important caveats are made by
Norde: First, as a process, grammaticalization is not a 'driving force' of
change (p. 31); second, 'there is nothing deterministic about
grammaticalization' (ibid.); third, 'saying that grammaticalization is a process
does not imply that there are universal pathways of grammaticalization' (p. 32),
which means that it is not legitimate to infer the diachronic process whereby a
grammatical item could have developed from a 'superficial comparison of initial
and final states' (p. 33). Norde continues with a discussion of whether there is
a 'grammaticalization theory' and concludes that grammaticalization is rather a
descriptive framework. She also surveys the interrelated issues concerning, on
the one hand, the distinction between diachronic change and correspondence
between the two separate synchronic stages, and the validity of reconstruction
as evidence in grammaticalization studies. In connection to the latter, Norde
discusses several 'false friends in grammar', which have been misanalysed in the
literature, and urges the linguists to be cautious when inferring, e.g., earlier
syntactic constituent order from contemporary morpheme order. Finally, a
detailed case-study of one of the paradigm examples of grammaticalization, i.e.
the development of the Romance adverbial '-mente', is presented on pp. 41-46 to
highlight and exemplify the methodological problems and pitfalls laid out in the
The second chapter, 'Unidirectionality' (pp. 48-105), discusses this notion in
detail along with various related issues. Norde begins with an overview of
different conceptions of unidirectionality, in particular distinguishing between
it being treated as a principle or constraint on grammatical change, on the one
hand, and as a hypothesis about more vs. less characteristic (and
cross-linguistically frequent) kinds of change, on the other. Turning to
counterevidence to unidirectionality, Norde asserts that though a small number
of disparate counterexamples 'would have no bearing on unidirectionality' (p.
52), the growing body of firmly established instances of degrammaticalization
calls for the 'downgrading' of unidirectionality to the status of a statistical
tendency. The most important consequence is that once the existence and
significance of counterdirectional change is recognized, linguists can no longer
use the hypothesis of unidirectionality as a 'handy reconstructional tool' (p. 51).
Among the important issues discussed in this chapter I would single out first
the distinction between directional and non-directional linguistic change. The
latter contains so-called 'lateral shifts' not affecting the position of a
linguistic element on the 'cline' of grammatical status, exemplified, inter
alia, by English noun-to-verb conversion. Second is the recognition of the fact
that all primitive changes comprising grammaticalization are in fact not
irreversible. On pp. 66-89, Norde discusses in turn changes on the semantic
level (bleaching or desemanticization as well as resemanticization, whereby
linguistic units gain in semantic content), on the category level
(decategorialization and recategorialization, i.e. transfer from a minor to a
major lexical class), univerbation of two former separate words, phonological
attrition vs. strengthening, subjectification vs. objectification. In particular
with respect to the latter, Norde concludes that objectification, i.e. loss of
speaker-oriented meaning, can co-occur both with grammaticalization (in
particular, with secondary grammaticalization) and degrammaticalization, hence
neither subjectification nor objectification can be used as reliable diagnostics
of either type of change. In general, it is shown that though the primitive
processes of change are 'overwhelmingly unidirectional' (p. 89), they are all
not unexceptionally so, and, importantly, different types of change are
irreversible to different degrees. The overall preferred unidirectionality of
grammaticalization results from the cumulative effect of the directional
preferences of the primitive changes comprising it. The last section of this
chapter is devoted to the explanations of unidirectionality proposed in the
literature, among which Norde singles out Rosenbach and Jäger's (2008)
experimental psycholinguistic approach.
In chapter 3, ''Defining degrammaticalization'' (pp. 106-134), Norde addresses the
issue of delineating degrammaticalization in a meaningful way, first giving a
useful historical sketch of the research in this field and terminology used. The
following general definition of degrammaticalization is proposed (p. 120):
'Degrammaticalization is a composite change whereby a gram in a specific context
gains in autonomy or substance on more than one linguistic level (semantics,
morphology, syntax, or phonology).'
Notably, the following (sometimes only theoretically) possible changes are
excluded from the notion of degrammaticalization:
(i) Loss of grammatical meaning by a morpheme (e.g. the transition from the
Latin inchoative suffix '-sc-' to the French stem formative '-ss-' marking
certain cells in the paradigm of some conjugation types).
(ii) 'Mirror image reversal' with 'grams stepwise moving up the cline' (p. 111),
a phenomenon concrete examples of which seem most implausible to be discovered,
which means that 'degrammaticalization is not the mirror image of
grammaticalization in the sense that it cannot be the complete reverse of a
grammaticalization chain' (p. 112).
(iii) Lexicalization of function words and affixes, such as English 'isms',
'ups' and 'downs'. This point is especially important because perhaps the
majority of the grammaticalization skeptics have adduced such instances as the
most obvious counterexamples to unidirectionality. However, since
grammaticalization involves not isolated words or morphemes, but whole
constructions, whose identity is generally preserved in the course of the
change, lexicalizations of the 'ism'-kind, where items are taken out of their
context, are not degrammaticalizations proper.
(iv) Replacement of a morpheme by another synchronically co-existing morpheme
under (partial) homonymy, as well as exaptation, i.e. 'the use of relatively
marginal grammatical material as more productive morphology with a different
function' (Traugott 2004). Both these changes may interfere with
grammaticalization and degrammaticalization, but are essentially different from
Norde evaluates instances of degrammaticalization against Lehmann's (1995/1982)
six 'parameters of grammaticalization' and Andersen's (2008) four 'levels of
observation'. After discussing both taxonomies in some detail, in particular
focusing on the rather controversial parameter of structural scope, Norde
distinguishes the three major types of degrammaticalization (p. 133):
--Degrammation: shift from grammatical content to lexical content
--Deinflectionalization: 'movement out of a paradigm accompanied by a change in
--Debonding: shift from a bound morpheme to a free morpheme.
Deinflectionalization and debonding are considered secondary
degrammaticalization, while degrammation is taken to be primary
degrammaticalization, by analogy with primary vs. secondary grammaticalization.
Chapters 4-6 (pp. 135-227) present a compendium of particular
degrammaticalization changes, subdivided into types just noted. Each change,
along with the relevant research history, is explicitly evaluated against the
six parameters of degrammaticalization, the mirror-images of Lehmann's
parameters of grammaticalization: resemanticization, phonological strengthening,
recategorialization, deparadigmaticization, deobligatorification, scope
expansion, severance (debonding) and flexibilization (increase in syntagmatic
variability). In Chapter 5 on deinflectionalization, Norde also discusses the
not unproblematic distinctions between inflection and derivation and between
affixes and clitics, using Zwicky and Pullum's (1983) criteria to determine the
status of the relevant morphemes at different stages of their history.
The examples presented in these three chapters come from a wide variety of
languages (Germanic, Slavic, Celtic, Finno-Ugric, Kwaza, Hup, Japanese, Tura),
with an obvious strong bias towards European languages with well-documented
histories. To a reader preferring wide-scale cross-linguistic generalizations
some of these case-studies may seem to be over-detailed, but I consider Norde's
emphasis on particularities of each individual development to be not only
well-justified but also conditio sine qua non of an empirically adequate study
in diachronic typology. As Norde herself states several times,
grammaticalization theorists have too often overlooked the minor historical
facts in favour of broad and sometimes aprioristic conceptions of grammatical
change, and this has led to an unfortunate propagation of misconceptions, e.g.
concerning the development of the 'group genitive' in English and Mainland
Among the numerous case studies presented in these three chapters, I would like
to single out the peculiar development of the Dutch '-tig' / German '-zig',
originally forming the numerals denoting tens ('twenty', 'thirty', etc.). In
contemporary, especially colloquial, Dutch and German, these suffixes have first
undergone debonding and resemanticization, turning into quantificational
adverbials ('many'), and then have further grammaticalized into intensifiers
('very'). This seems to be the only known example of a gram which has first
degrammaticalized and then (re)grammaticalized, albeit in a function having
nothing to do with its original meaning.
In the concluding chapter (pp. 228-238) Norde summarizes the data presented in
the preceding chapters and reviews the significance of Lehmann's parameters as
diagnostics of various types of degrammaticalization. As expected, for different
kinds of degrammaticalization different parameters are relevant, and in
different ways. While resemanticization is a defining characteristic of
degrammation whereby a former gram acquires a lexical meaning, it turns out that
in most cases of deinflectionalization the new function is grammatical.
Phonological change is not a necessary part of degrammaticalization, but where
it occurs, 'it always comprises strengthening' (p. 228). Recategorialization is
again relevant mostly for degrammation, but irrelevant for
deinflectionalization, since here the grams remain bound. In debonding,
recategorialization is attested only in a few cases, such as Irish 'muid' (1PL
suffix > 'we'). Deobligatorification and deparadigmaticalization are of little
relevance for debonding of derivational affixes and clitics. The parameter of
structural scope, which has been shown to be controversial already for
grammaticalization proper, is almost as problematic for degrammaticalization,
clearly relevant only in debonding where scope expansion is generally observed.
The parameter of bondedness applies only to secondary degrammaticalization,
where it yields slightly different results for deinflectionalization (shift to a
'looser' morpheme boundary) and debonding (shift from a morpheme boundary to
word boundary). Finally, the parameter of syntagmatic variability is irrelevant
for deinflectionalization, again for the reason that in this change morphemes
remain bound. In degrammation and debonding, syntagmatic variability usually
increases, though it is not necessary, cf. Japanese connectives, 'which always
occupy a specific syntactic slot, albeit different ones for enclitic and free
connectives, respectively' (p. 231).
Further, Norde discusses the defining characteristics of each type of
degrammaticalization, again stressing that 'as a general requirement for all
types of degrammaticalization, the change needed to be construction-internal'
(p. 233). Summarizing the values of different parameters, she observes that
'there are virtually no negative scores' (e.g. putative instances where an item
would acquire a lexical meaning and in the same time become more bound), which
means that 'degrammaticalization is as consistent in the directionality of its
primitive changes as is grammaticalization' (p. 233). A section of this chapter
is devoted to mechanisms of degrammaticalization changes. In degrammation and
deinflectionalization, reanalysis is the primary mechanism, while in some
instances of debonding, analogy also plays an important role. Norde stresses
that it is necessary not to confuse mechanisms of degrammaticalization with its
motivating forces. The latter are rather obscure and always depend on the
particular historical circumstances. In a number of the cases discussed in the
book, deinflectionalization resulted from the more general process of the loss
of former rich inflectional systems (especially relevant for the English and
Scandinavian genitive), and in the case of the Old Estonian particles 'es' and
'ep', phonological factors must have played a role. The chapter concludes with a
brief look at further research perspectives on degrammaticalization. Of
particular interest, in my view, is the hypothesis (yet to be proven or
falsified) that in polysynthetic languages, where affixes cover large arrays of
diverse functions, it is probably 'easier for (some) suffixes ... to
degrammaticalize, simply because they are more contentful' (p. 238).
Over the past decades, grammaticalization has become a popular subject and a
topic of intensive discussions between adherents of different conceptions of
language change. In these discussions, the controversial notion of
unidirectionality has assumed a central role, and putative counterexamples to
unidirectionality have been subject of much debate. Though the number of such
examples assembled so far is already quite impressive, there has been no
agreement even among the adherents of 'grammaticalization theory' as to how to
treat these facts, which terms to apply to them, and how to classify and
motivate them. Norde brings much awaited order into this domain, providing a
clear and uncontroversial definition of degrammaticalization and a useful
classification of types thereof. However, I believe that the impact of this book
is not limited to the study of degrammaticalization. Once degrammaticalization
is rigorously defined and classified, grammaticalization becomes better
understood, too. Besides this, and perhaps most importantly, this book, in my
opinion, settles the burning issue of unidirectionality, which, in the light of
the unequivocal examples of changes 'up the cline', evaluated against the same
parameters which are relevant for grammaticalization, can no longer be regarded
as a universal constraint on grammatical change. Nevertheless, it must be
stressed that this book makes a strong case for directionality of change: both
grammaticalization and degrammaticalization are directional processes in that
both involve cumulative effect of intercorrelated changes on different
linguistic levels. The existence of degrammaticalization does not in anyway
invalidate the notion of grammaticalization, neither does it cast doubt on the
systematic nature of the latter in comparison to the exceptionality of the former.
Except for a few typos and editing shortcomings, as well as several references
which could have been included (e.g. among the criteria of inflection vs.
derivation distinction discussed on p. 153, Dressler's (1989) approach should
have found its place) I have no critical remarks about this book.
Muriel Norde ''Degrammaticalization'' is, in my opinion, among the most important
contributions to the field of grammaticalization research published in the
course of the last few years, and an indispensable guide not only to the more
'exotic' phenomenon of degrammaticalization, but also a critical and
well-informed review of all the major conceptual issues in the study of
grammatical change, as well.
Andersen, Henning (2008). Grammaticalization in a speaker-oriented theory of
change. In: Thórhallur Eythórsson (ed.), Grammatical Change and Linguistic
Theory: The Rosendal Papers, 11-44. Amsterdam, Philadelphia: John Benjamins.
Dressler, Wolfgang (1989). Prototypical differences between inflection and
derivation. Zeitschrift für Phonetik, Sprachwissenschaft und
Kommunikationsforschung 42:1, 3-10.
Fischer, Olga (2000). Grammaticalization: Unidirectional, non-reversable? The
case of to before the infinitive in English. In Olga Fischer, Anette Rosenbach,
and Dieter Stein (eds.), Pathways of Change: Grammaticalization in English,
149-69. Amsterdam, Philadelphia: John Benjamins.
Heine, Bernd & Tania Kuteva (2002). World Lexicon of Grammaticalization.
Cambridge: Cambridge University Press.
Hopper, Paul J. & Elizabeth Closs Traugott (2003). Grammaticalization. 2nd ed.
Cambridge: Cambridge University Press.
Joseph, Brian D. (2001). Is there such a thing as grammaticalization? Language
Sciences 23:2-3, 163-86.
Lehmann, Christian (1995 ). Thoughts on Grammaticalization. München,
Newcastle: LINCOM Europa.
Newmeyer, Frederic J. (1998). Language Form and Language Function. Cambridge,
MA: MIT Press.
Rosenbach, Anette & Gerhard Jäger (2008). Priming and unidirectional language
change. Theoretical Linguistics 34:2, 85-113.
Traugott, Elizabeth Closs (2004). Exaptation and grammaticalization.
Zwicky, Arnold M., Geoffrey K. Pullum (1983). Cliticization vs. inflection:
English n't. Language 59:3, 502-51.
ABOUT THE REVIEWER
| ABOUT THE REVIEWER:
Peter M. Arkadiev, PhD in linguistics (2006), is a senior research fellow
in the Department of Typology and Comparative Linguistics of the Institute
of Slavic studies of the Russian Academy of Sciences, Moscow. His main
interests are linguistic typology with a focus on event and argument
structure and its formal realization, tense-aspect-modality and case
marking. He works mainly on Lithuanian and Adyghe.