Publishing Partner: Cambridge University Press CUP Extra Publisher Login

New from Cambridge University Press!


Revitalizing Endangered Languages

Edited by Justyna Olko & Julia Sallabank

Revitalizing Endangered Languages "This guidebook provides ideas and strategies, as well as some background, to help with the effective revitalization of endangered languages. It covers a broad scope of themes including effective planning, benefits, wellbeing, economic aspects, attitudes and ideologies."

New from Wiley!


We Have a New Site!

With the help of your donations we have been making good progress on designing and launching our new website! Check it out at!
***We are still in our beta stages for the new site--if you have any feedback, be sure to let us know at***

Review of  Genitive Variation In English

Reviewer: Cristiano Broccias
Book Title: Genitive Variation In English
Book Author: Anette Rosenbach
Publisher: De Gruyter Mouton
Linguistic Field(s): Historical Linguistics
Subject Language(s): English
Issue Number: 14.2272

Discuss this Review
Help on Posting

Date: Thu, 28 Aug 2003 14:59:08 +0200
From: Cristiano Broccias
Subject: Genitive Variation in English: Conceptual Factors in Synchronic
and Diachronic Studies

Rosenbach, Anette (2002) Genitive Variation in English: Conceptual
Factors in Synchronic and Diachronic Studies, Mouton de Gruyter,
Topics in English Linguistics 42.

Cristiano Broccias, Università di Genova and Università di Pavia

Since Chapter 1 offers a short description of the book's topic and main
points, I have not provided a general introduction but have preferred
to rely on a chapter-by-chapter summary. The review ends with some
general remarks.

Chapter 1 (Introduction)

The chapter introduces the topic of the book (i.e. the synchronic and
diachronic study of genitive variation in English) and points out its
main conclusions. Synchronically, Rosenbach studies those genitive cases
where, in principle, both the s-variant and the of-variant are possible
(e.g. "the girl's eyes"/"the eyes of the girl" vs. "a king of
honour"/*"an honour's king"). Crucially, she opts for an experimental
study, which allows her to keep apart the three conceptual factors
influencing the choice of either structure (i.e. animacy, topicality, and
possessive relation) as well as to evaluate their relative importance.
Her main conclusions are (1) that the s-genitive is favoured in what she
calls (user-)optimal cases, i.e. with an animate and topical possessor,
and in a prototypical possessive relation; (2) the relative importance of
the three factors is animacy>topicality>possessive relation; (3) the
s-genitive is on the increase (and especially so in American English);
(4) the extension of the s-genitive to inanimate possessors is more
productive than has been assumed so far. Diachronically, Rosenbach
analyses a corpus of late Middle and Early Modern English (i.e.
1400-1630) and intends to show that the s-genitive became more productive
in such a period. Crucially, she argues that the s-genitive extended
along the hierarchy mentioned above (i.e. animacy>topicality>possessive
relation). In other words, the synchronic functional analysis also sheds
light on language change. She stresses that the factor triggering the
rise of the s-genitive was the development of the relational marker 's
from inflection to clitic (although the change may not be complete yet)
and the s-genitive's acquiring determiner function.

Chapter 2 (The structure of the s-genitive and of the of-genitive: some
theoretical preliminaries)

Rosenbach uses the terms "s-genitive" and "of-genitive" to refer to the
whole constructions and calls the two noun phrases linked via the
relational markers (i.e. 's and of) possessor and possessum (e.g. in "the
girl's eyes", first noun phrase, "the girl", is called possessor and the
second, "the eyes", possessum; in "the eyes of the girls", "the eyes" is
the possessum and "the girl" is the possessor). Building on previous
research (e.g. Plank 1992), she argues that 's cannot be regarded as
either an inflection or a clitic. Rather, 's seems to have developed from
an inflection to a more clitic-like element in Modern English. Still, for
the sake of clarity, she adopts the term "clitic" to refer to 's. Chapter
2 also introduces the distinction (after Biber et al. 1999) between
"specifying genitives" (i.e. genitives having a [+referential] possessor)
and "classifying genitives", i.e. the modifier/compound type (where a
[-referential] possessor occurs). "[A beautiful king's] daughter" (where
the possessor is the determiner) and "the king of England" (where the
possessor is the complement) are examples of the former; "a beautiful
[king's daughter]" and "a king of honour" (where the possessor is in
either case a modifier) illustrate the latter (note that in "a king of
honour", "honour" is called possessor).

Chapter 3 (Grammatical variation)

Rosenbach's empirical study is based on the distinction between
categorical and choice contexts. Her analysis deals with only those cases
where, in principle, both the s-genitive and the of-genitive are possible
(i.e. choice contexts), thus ignoring those instances where the language
user has only one option (i.e. categorical contexts) - although she duly
remarks on the possibly fuzzy nature of the boundary between the two sets
(see p.28). Her choice contexts are restricted to possessive cases (see
Table 10 on p.29 for some examples illustrating this notion) where the
possessor is a full lexical NP, the whole genitive NP is [+definite] and
no reference tracking devices for the whole NP are used (i.e.
demonstratives as in "this head of the king", possessive pronouns as in
"my picture of John", the definite article as in "the head of the king"
are excluded). Rosenbach also briefly discusses some (phonological,
morphological, syntactic, pragmatic and socio-stylistic) factors which
are known to influence the use of either structure (in choice contexts)
and which she intends to disregard in her study. This allows her to
further delimit her choice contexts. In more detail, she only includes
(1) possessors not ending in /s/, /z/, or the voiceless dental fricative,
(2) singular possessor nouns, (3) non-complex, non-branching possessors
and possessums, (4) non-consecutive genitive constructions, and (5) data
which are either balanced or controlled for style.

Chapter 4 (Factors: Animacy, topicality, and possessive relation)

This chapter reviews how animacy, topicality and the notion of possession
- the three factors Rosenbach investigates in her study - have been dealt
with in the literature. It is pointed out that such three factors are
often difficult to define (e.g. what counts as an animate entity can be a
matter of conceptualisation, see p.49). Further, a series of empirical
problems are observed in connection with previous empirical analyses (see
the useful summary on pp.71-72); among them, alongside the definitional
issue mentioned above, is the fact that many studies have been concerned
with absolute rather than relative frequencies for the s-genitive in
choice contexts. Also, since the three factors interact with each other,
they should not be studied independently. Towards the end of the chapter,
Rosenbach acknowledges the importance of Langacker's (1995) and Taylor's
(1996) reference-point analysis of the s-genitive (i.e. the idea that the
possessor functions as an anchor or reference point for the
identification of the possessum). Still, she observes that such an
account does not discuss of-genitives and is essentially theoretical
(i.e. not empirical) and synchronic in nature. Hence, the need to bridge
the gap between empirical studies and theoretical, cognitive accounts.

Chapter 5 ("Variation" versus "choice")

This chapter draws an important distinction between variation and choice.
Variation can be understood, on the level of the language system, as the
availability of variants and, on the usage-level, as having to do with
how such variants are distributed (e.g. with reference to social groups).
Choice involves the usage-level and manifests itself as preferences
within an individual speaker. Rosenbach goes on to detail how variation
and choice are (or could be) dealt with in formal and
cognitive-functional approaches. In particular, only the latter are
regarded to be choice-based (i.e. as dealing with performance). The
chapter concludes with a brief preview of the following material in the
light of the discussion offered in the present chapter. In more detail,
the author makes it clear that her investigation is choice-based (and
hence functionally oriented), that her "speaker perspective [...] is
[...] not sociolinguistic but psycholinguistic/cognitive in nature"
(p.107) and, finally, that the cognitive-psychological factors to be
investigated (i.e. animacy, topicality, and possessive relation) are
regarded as operating largely unconsciously. Further, she intends to make
use of iconicity and naturalness theories by regarding what is more
natural (as well as what is iconic) as what is easier to process (i.e.
more economical) and hence more likely to occur.

Chapter 6 (Modern English data: experimental study)

This chapter describes the experimental study carried out by the author
in order to assess animacy, topicality, and possessive relation as
factors influencing the speaker's choice of the s-genitive vs. the
of-genitive. Rosenbach first of all offers operational definitions of the
three factors under scrutiny. Animacy is taken as referring to the
animacy status of the possessor only. As examples of [+animate]
possessors, she limits her data to [+human], [-animal], [-collective]
personal nouns (e.g. "girl", "mother", "boy", "man"); as [-animate]
possessors, she opts for [-human], [-collective] concrete nouns only
(e.g. "chair", "bed", "door", "table"; note that temporal and
geographical nouns are excluded). Moving on to topicality, she points out
that the topicality of the possessum is always kept new in her study. As
for the topicality of the possessor, she distinguishes between a
[+topical] possessor, which is [+referential], implies second mention and
is definite (e.g. "the girl"), on the one hand, and a [-topical]
possessor, which is [+referential], implies first mention and is
indefinite (e.g. "a girl"), on the other. Since animate and topical
elements usually (at least in English) occur first in linear order (i.e.
the serialization or linear sequencing principle, which is iconic in
nature, in that the linear order reflects the order of
conceptualisation), the s-genitive is predicted to be preferred with
[+animate] and/or [+topical] elements. Finally, possessive relations are
divided into [+prototypical] and [-prototypical]. With [+animate] and
[+human] possessors, [+prototypical] possession includes body parts, kin
terms and permanent/legal ownership, whereas [-prototypical] possession
includes states and abstract possession. With [-animate] possessors,
[+prototypical] possession includes part/whole relations, whereas
[-prototypical] possession includes non-part/whole relations. Rosenbach
argues that the iconic principle of conceptual distance (i.e. formal
distance between X and Y reflects conceptual distance between the two)
allows her to predict that [+prototypical] possession should favour the
s-genitive and [-prototypical] possession the of-genitive. This is so
because "[+prototypical] possession relations are close relations between
possessor and possessum" (p.124) and, structurally, the s-genitive
employs a marker ('s) which, for example, is "more bonded with the
possessor that the preposition of" (p.124).

The second part of chapter 6 details how the experiment was carried out
and its results. 56 native speakers of British English and 48 native
speakers of American English, aged between 18 and 81 (most of them having
a university education), had to choose between an of-genitive and an
s-genitive (both possible in theory) in 93 short extracts from
contemporary novels. The prediction that the s-genitive should occur more
often with a [+animate] possessor, a [+topical] possessor, and a
[+prototypical] possessive relation (see above) is further specified into
two predictions. Prediction I requires that the s-genitive is more
frequent that the of-genitive in each of the three cases. Prediction II
requires that s-genitives for positive values of each of the three
factors are more frequent than s-genitives for negative values (i.e. we
are dealing with the relative distribution of the s-genitive). Both
predictions are confirmed (in both varieties) although the difference
between s-genitives and of-genitives in [+prototypical] possessive
relations turns out to be not statistically significant.

Rosenbach then analyses how the three factors interact with each other,
that is she studies how s-genitives and of-genitives are distributed
according to the eight conditions [+/-animate], [+/-topical], [+/-
prototypical] (abbreviated as [[+/-a], [[+/-t], [[+/-p]). Her data
indicate that (in both varieties) animacy is more important than
topicality, which in turn is more important than possessive relation
(i.e. animacy > topicality > possessive relation). Her findings are
summed up in a preference hierarchy (see Figure 17 on p.153) where the
eight possible conditions (starting with those where the s-genitive is
more frequent) are arranged as follows: [+a, +t, +p] > [+a, +t, -p] >
[+a, -t, +p] > [+a, -t, -p] > [-a, +t, +p] > [-a, +t, -p], [-a, -t, +p],
[-a, -t, -p]. Importantly, Rosenbach found that of-genitives become more
frequent starting with [+a, -t, -p] and that the relative frequency of
the s-genitive is not significantly different in the last two cases in
either variety. However, American English data, unlike British English
data, indicate that the difference in frequency of s-genitives between
the "worst" animate condition (i.e. [+a, -t, -p]) and the "best"
inanimate condition (i.e. [-a, +t, +p]) is not significant either.

Rosenbach also divides the British and American subjects into two age
groups and observes a higher frequency of s-genitives in the younger
groups, particularly in the British English one - although, as far as
British English is concerned, the [+animate] environment does not show a
significant difference and, in American English, only the [-animate] and
[+prototypical] conditions are significant. Moving on to the interaction
of the three factors, the author concludes that the behaviour of the two
age groups in both varieties conforms to the preference hierarchy above.
As for British English, she observes that the increasing use of the
s-genitive has mainly to do with [-animate] conditions (and that the
effect of topicality and possessive relations is secondary). As for
American English, a similar picture emerges although only in the [-a, +t,
+p] condition is the difference between the two age groups significant
(and note that the increasing frequency for [+prototypical] conditions
mentioned above is largely restricted to [-animate] cases). In other
words, the increase in s-genitives (in [-animate] conditions) is
particularly strong in British English.

The last part of the experimental study is devoted to a comparison
between British and American English. By looking at the single factors,
Rosenbach concludes that, in general, s-genitives are more frequent in
American English in all conditions (except for the [+animate] one) but
only in the [-animate] context is the difference significant. The
analysis of the interaction of the three factors shows that the
s-genitive is more frequent (but not significantly so) in British English
in the first three [+animate] conditions in the preference hierarchy
given above. On the other hand, s-genitives are more frequent in American
English in all [-animate] conditions although only in the [-a, -t, +p]
condition is the difference between the two varieties significant.
Finally, Rosenbach studies how the two varieties interact with the two
age groups in the use of the s-genitive in [-animate] cases. She
concludes that, although American English still uses more s-genitives in
[-animate] cases, the difference between the two varieties is fading,
which might indicate an influence of American English over British

A problem in her analysis, as she herself points out, is however that
in the [-a, -t, +p] condition the two varieties differ (significantly)
in the present usage but not in the past (whereas the opposite might be
expected). She points out that this might be a recent American English
innovation which has not spread into British English yet (see p.167 but
also note 124 where Rosenbach argues that this line of explanation may
not be correct). In the last part of the chapter (where she draws the
conclusions), Rosenbach also stresses the fact that the somewhat
problematic distribution of s-genitives in the two "worst" cases
(i.e. the [-a, -t] conditions) might be due, among other things, to the
occurrence of more "car" items in the [-a, -t, -p] condition than in the
[-a, -t, +p] case and to the shortness of such a word. The latter
observation leads her to a short discussion of the role played in general
by word length in her study (i.e. the s-genitive should be more frequent
in the "short>long" condition and the of-genitive in the opposite
condition). She concludes that word length may have biased her results
but only a little. Still, she admits that further research into this
factor is needed.

Chapter 7 (Historical development of the genitive variation)

Drawing on Rosenbach and Vezzosi (2000) and Rosenbach, Stein, and Vezzosi
(2000), the author shows that, contrary to what is usually accepted, the
s-genitive, after having been on the decrease in early Middle English,
started to increase from around 1400. Further, from the middle of the
16th century s-genitives are more frequent than of-genitives in
[+animate] contexts (see below). She argues that this trend does not
necessarily run counter to general laws of language change because it may
be linked to the structural change of 's from an inflection to a more
clitic-like determiner (i.e. we are not dealing with the same type of
s-genitive over the span of time under consideration).

In her historical analysis, Rosenbach only considers [+animate] contexts,
[-animate] environments being the almost exclusive province of
of-genitives. She then analyses how topicality and possessive relations
interact in such [+animate] cases. As for topicality, it is also worth
pointing out that she includes [++topical] cases (e.g. the possessor is a
proper name, as in "John's book"). Her results for the 1400-1630 period
are summed up in the preference structure [+a, ++t, +p] > [+a, ++t, -p] >
[+a, +t, +p] > [+a, +t, -p] > [+a, -t, +p] > [+a, -t, -p], i.e.
s-genitive contexts extend from the left to the right and the relative
frequency of the s-genitive increases in each context. Crucially, such a
hierarchy resembles the one proposed for the synchronic analysis in the
previous chapter.

The second part of the chapter attempts to offer an explanation of why
the s-genitive became productive again in the 16th century. Rosenbach
makes two important points: (1) the precondition for the s-genitive to
become productive again is the change of 's from inflection to clitic and
(2) the increase in the relative frequency of the s-genitive along the
preference structure mentioned above is to be related to the function of
the clitic s-genitive as a determiner. She also explores what caused the
change in (1) and suggests that a variety of factors might have
contributed to it, although she admits that this question remains
unsettled (see for example page 231). First of all, the inflectional
s-genitive was not replaced by the s-less genitive (which is still
attested in northern British English dialects nowadays) possibly because
of elliptic genitives (e.g. "at Mary's"), which strongly favoured the
s-genitive. Alongside existing accounts for the change in (1), namely the
collapse of the genitive paradigm, deflexion and the influence of the
his-genitive (i.e. "John his book" for "John's book"), Rosenbach mentions
systemic influences such as the shift towards phrasal compounding (cf.
Modern English phrases like "an off-the-rack dress") and "the close
interconnections between specifying genitives on the one hand and
classifying genitives/compounds and nominal premodification on the other"

As for the determiner function of the s-genitive, this may be related to
the evolution of the definite article in the late Middle/early Modern
English period, in that the s-genitive came to occupy the new structural
determiner position available in English. Since a definite determiner
functions as a referential anchor, we expect the s-genitive to be
preferentially used with animate and topical possessors in prototypical
possessive relations and not to be used in those contexts incompatible
with the referential function, i.e. descriptive and partitive genitives.

Chapter 8 (A diachronic scenario: the extension of the s-genitive from
Middle to Modern English - economically-driven language change?)

Rosenbach assumes that principles of (cognitive) economy play a role in
language processing and uses the term economy to refer to "any states
and/or processes which are (i) easier to conceptualise or process for the
human mind (=synchronically user-optimal construction) and/or (ii) run in
the automatic processing mode" (p.237), the latter mode being akin to
Cognitive Grammar's notion of entrenchment. She also distinguishes
between speaker and hearer economy in that the speaker wants to use short
utterances whereas the hearer wants them to be as explicit as possible.
She stresses that her study focuses on speaker optimality (see for
example p.242); still, the choice of the s-genitive vs. the of-genitive
may also satisfy the economic needs of the hearer (p.242). In order to
account for the spread of the s-genitive, the author makes use of three
economical principles. (1) Synchronic user-optimality: the s-genitive in
[+animate], [+topical] and [+prototypical] contexts is synchronically the
more economical option since it satisfies the iconic/natural principles
of serialisation and conceptual distance (introduced in chapter 6). (2)
Automatization: since the s-genitive in [+animate], [+topical] and
[+prototypical] contexts is synchronically user-optimal, it is
unsurprising that it becomes more frequent (given the structural
precondition mentioned in the previous chapter) and, hence, entrenched in
this context. In this way, a diachronic change comes about since
entrenchment is another type of user-optimality. (3)

Analogical/metaphorical extension: analogy and metaphor motivates why the
s-genitive extends along the preference hierarchy and, in particular,
becomes more and more frequent in [-animate] contexts. Rosenbach also
deals with the question of whether grammaticalisation is involved in the
spread of the s-genitive. If we understand grammaticalisation in a broad
sense as a process by which a construction becomes more and more fixed,
then this is indeed is the case. If we interpret grammaticalisation in
the strict sense (i.e. a lexical, less bound element progressively
develops into a grammatical, more bound element), then it may be useful
to distinguish between two processes: (1) the development of 's from
inflection to clitic, which may be a case of degrammaticalisation, and
(2) the spread of the (clitic) s-genitive, which may be analysed as a
case of grammaticalisation (although evidence is at present not
conclusive in either process).

The final part of the chapter examines other factors which may have
influenced the extension of the s-genitive, in particular stylistic and
structural factors. Since the s-genitive is more frequent in informal
texts, its rise may be linked to a trend towards more informality.
Further, the increasing use of nominal premodification (vs.
postmodification) also ties in with the rise of the s-genitive.

Chapter 9 (Summary and conclusion)

In the final chapter Rosenbach summarises her main conclusions and
stresses some theoretically important points, such as the notion that an
economically-driven language change approach does not necessarily result
in a more optimal system (this may be due to analogical/metaphorical
extensions as well as competing needs on the part of the speaker vs. the


Rosenbach's book is stimulating and fascinating for a variety of reasons.
First, her investigation is highly commendable because she deals with
genitive variation both synchronically and diachronically (and in the
latter case both within a limited time span, i.e. within different age
groups, and over the course of the history of the English language from
the 15th to the early 17th century). Second, when examining contemporary
variation she pays attention to different varieties (i.e. American and
British English). This, I think, should pave the way to analyses of more
varieties, which might prove useful in order to seek confirmation of
Rosenbach's findings. Third, she devises a very interesting experimental
method for the study of genitive variation, which had never been
attempted before. Fourth, she is always aware of the possible problems
and drawbacks in her methodology, analyses and conclusions, thus offering
us a very balanced account and evaluation of her findings. Fifth, she
demonstrates a very good knowledge of general linguistics and is capable
of tying in her very specific study with such more general work on the
nature of language. Sixth, she stresses the importance of a systemic
approach to language change in that a variety of factors are dully
recognised as conspiring to the rise of a certain pattern; in other
words, it is acknowledged that any phenomenon must be evaluated within
the language system as a whole.

The very few observations I have mainly have to do with some decisions
which might reflect editorial choices. I would have liked to see the
questionnaire used by Rosenbach in her experimental study included in the
book. This would allow the reader to have a better understanding of the
author's treatment of the data, see also below, as well as render her
experiment easily replicable. Unfortunately, her appendix only includes
the genitive phrases in isolation (and also note that the frequencies
breakdowns do not offer percentages for each of the examples employed).
Similarly, the historical data taken into consideration should also have
been listed (even if this might have resulted in many more pages). Going
back to the synchronic data (as they are presented in the appendix in
10.1, pp.278-281), I must say that I find some of the examples
problematic. For instance, Rosenbach repeatedly stresses (see note 41 for
example) that she has not included subjective and objective genitives in
her questionnaire (since subjective genitives, for example, tend to occur
in the s-variant). Still, examples like "a door's shoosh", "a body's soft
thud", "a car's fumes", "a motorbike's sound", "the driver of a car",
"the renovation of a flat" and "the transport of the body" seem to me
good candidates as subjective and objective genitives (i.e. the first
four examples could be analysed as subjective genitives in that the
"possessor" "emitted" a sound or some other substance, whereas the
remaining examples, especially the last two, could easily be regarded as
objective genitives). Note also that some of the examples are so similar
to one another that analogy effects cannot be excluded.

Further, since Rosenbach acknowledges the problematic categorisation of "car" as an inanimate noun (see pp.172-173), one may wonder why she has decided to include it (and why she has done so in more than one instance). Similarly, the noun "tree" (cf. "a tree's shadow") designates an entity which is intuitively much more animate than a room; some readers may therefore object to its categorisation as [-animate] (although, admittedly, Rosenbach explicitly states that by [+animate] she means [+human]). It is also not clear to me why "a motorbike's sound" is included twice, both in the [+animate] set and in the [-animate] set. Finally, another problematic example may be "a cart's creaking wheels" (assuming that this is the correct ordering of the elements vs. "the wheels of a creaking cart", which is given as the of-genitive variant). Apart from word-length considerations, one may observe that, conceptually, the example in question implies two "components": one pertaining to a part-whole relation between "possessor" and "possessed" (which justifies the [+prototypical] possession classification) and the other involving the emission of a sound. The latter may influence (strengthen?) the choice of the s-genitive in analogy with what is the case with subjective genitives, as was pointed out above. Surely, the inclusion of the questionnaire might have dispelled some of these doubts. One final observation concerns the notion of "choice contexts". Rosenbach does not consider prepositional variants other than the of-genitive (for the very obvious reason that they would not count as genitives). Still, in some cases (not necessarily those included in her questionnaire), the availability of other prepositional variants could also bear on the use of the s-genitive. A case in point is Rosenbach's example "her notebook's pages". Why should we exclude a structure like "the pages IN her notebook" and only consider "the pages OF her notebook"? It remains to be seen whether such alternatives also influence genitive variation.

With the proviso concerning the questionnaire in mind, I think
Rosenbach's book is an invaluable contribution to the study of genitive
variation in English and opens up interesting avenues to further
research. (e.g. What was the role played by animacy, topicality, and
possessive relations in the rise of the of-genitive in the period between
late Old English and early Middle English? Or, to give one more example,
the idea that s-genitives are more bonded than of-genitives should
perhaps be investigated further).


Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward
Finegan. 1999. Longman Grammar of Spoken and Written English. London:

Langacker, Ronald. 1995. Possession and possessive constructions. In:
John Taylor and Robert MacLaury (eds.). Language and the Cognitive
Construal of the World, 51-79. Berlin: Mouton de Guyter.

Plank, Frans. 1992. From cases to adpositions. In: Nicola Pantaleo /ed.).
Aspects of English Diachronic Linguistics: Papers read at the Second
National Conference of History of English, Naples, 28-29 April 1989,
19-61. Fasano: Schena.

Taylor, John. 1996. Possessives in English. Oxford: Clarendon.

ABOUT THE REVIEWER Cristiano Broccias teaches at the Universities of Genoa and Pavia (Italy). His research interests include English grammar and cognitive linguistics. He has recently published a book on English change constructions (Broccias, C. 2003. The English Change Network. Berlin: Mouton de Gruyter).