AUTHOR: Siemund, Peter
TITLE: Pronominal Gender in English
SUBTITLE: A Study of English Varieties from a Cross-Linguistic Perspective
SERIES TITLE: Routledge Studies in Germanic Linguistics
PUBLISHER: Routledge (Taylor and Francis)
YEAR: 2008
Peter de Swart, Center for Language and Cognition Groningen, University of
Groningen (the Netherlands)
SUMMARY
This book presents a revised version of the author's postdoctoral dissertation
(Habilitationsschrift, Freie Universität Berlin). It provides an in-depth study
of pronominal gender in English with particular focus on the use of animate
pronouns (he, she) for inanimate referents. This is done by investigating the
distribution of pronominal forms in different varieties of English which in turn
are put in a wider cross-linguistic perspective. The goal of the book is to show
that this distribution is subject to essentially the same principles in all
varieties of English. The study assumes a variationist approach and combines it
with a functional-typological perspective. At the same time it touches on
various other disciplines in linguistics, such as sociolinguistics and
historical linguistics.
Chapter 1-Introduction
The introductory chapter presents the roadmap to the book introducing the
problem, major claims, and background information. First, it exemplifies and
formulates the problem central to the study: the use of animate pronouns (he,
she) for inanimate referents and vice versa (i.e. the use of 'it' for animate
referents) in different varieties of English. It is indicated that the book is
basically concerned with the former alternation. Its main objective is to
explain the use of 'he/she' instead of 'it' and to a lesser extent the
distribution of masculine and feminine pronouns across different domains of the
inanimate world. The basic claim introduced by the author is that pronominal
gender in English basically depends on the degree of individuation of the
referent. When individuation is conceptualized as a hierarchy or continuum, he
predicts that animate pronouns enter the hierarchy from the left whereas
inanimates do so from the right. Moreover, it is predicted that most cut-off
points on the hierarchy are attested. The chapter continues with an illustration
of the phenomenon under discussion from other Germanic and Romance languages,
some discussion of the status of the category of gender in English (the author
claims that there is no conclusive evidence for the existence of such a category
and therefore speaks of pronominal gender), and a note on the historical
development of the English gender system which shows that the language developed
a semantic gender system from a morphophonological one and that pronominal
reference related to semantic gender existed already in Old English times. The
next two sections introduce the sources for the variation data used in the study
and the generalizations about the distribution of pronominal forms which have
been based on them. In the final section the author discusses how his
variationist approach ties in with the methodology of functional typology. He
claims that his study shows several commonalities with traditional typological
studies. Both systematically compare one specific grammatical domain across
different varieties/languages and they share the assumption that the observed
variation is systematic. Finally, he argues that the study of varieties of a
language can make an important contribution to typology as it makes the
empirical basis of typological studies more complete.
In chapters 2 to 6 the author presents the data and distribution of pronominal
forms in different varieties of English.
Chapter 2-The Southwest of England
This chapter discusses pronominal usage in the southwest of England, in
particular the variety spoken in West Somerset. The choice for this dialect area
is motivated by the amount of information available, in the words of the author
''it is the single most profoundly investigated dialect area of Great Britain''
(p.19). Most importantly, this variety has been argued to draw a clear-cut
distinction between count nouns and mass nouns, using animate for the former and
inanimates for the latter. The author shows that a similar pattern can be found
throughout Great Britain in this way becoming ''a characteristic of all, or at
least most, dialects of Great Britain'' (p.23). However, West Somerset English
has generalized the usage to the greatest extent and has the most data available
therefore becoming the focus of the remainder of this chapter. First the author
provides some background information on West Somerset English grammar through
the discussion of a number of phenomena related to the mass/count distinction
and the pronominal system. Section 2.2 presents an in-depth study of pronominal
usage in 19th-century West Somerset English on the basis of approximately 1000
pronominal references from Elworthy's (1886) _West Somerset Word Book and
Glossary_. These data show that the proposed generalization does not hold
categorically. One finds both masculine pronouns replacing mass nouns and neuter
pronouns replacing count nouns. The former exceptions are only found with
concrete nouns showing that the mass/count distinction does not hold for
abstract nouns which are always taken up by 'it'. Support for count agreement of
animate pronouns is also found in a study of poetry texts although some
differences emerge as well: the neuter pronoun is used both for count and mass
nouns and the author finds use of animate pronouns for abstract nouns. In the
final part of the chapter the author compares his findings from 19th-century
West Somerset English with pronominal usage in modern (20th Century) English of
the Southwest. The conclusion is that the pattern found in West Somerset extends
across the southwest of England although the system is in serious decline. Data
from the British National Corpus even seem to suggest that the traditional
southwestern system has become extinct.
Chapter 3-Newfoundland
In this chapter the author provides a short discussion of pronominal usage in
the variety of English spoken in Newfoundland, Canada. The system of
Newfoundland English shows parallels with that of Southwest England which can
probably be traced back to the fact that many settlers came from that area and
exported their system of pronominal gender. In contrast to the southwest of
England, Newfoundland English still uses this traditional system nowadays. On
the basis of two previous studies, the author shows that this variety follows
the pattern found in West Somerset English: animate pronouns are used for
concrete countable nouns and the neuter pronoun for mass nouns and abstract
nouns. He finds some examples of the masculine pronoun being used for abstract
nouns but they involve only a handful of count nouns (e.g. 'poem', 'song')
making this a phenomenon of restricted scope. The one point in which
Newfoundland English differs substantially from West Somerset English is the
existence of an additional split in the category of inanimate concrete count
nouns. Within this category a distinction is made between nouns that are mobile
('boat', 'car', 'areoplane') which are referred to with the feminine pronoun and
ones that are non-mobile ('paddle', 'wheel', 'wing') which are referred to with
the masculine pronoun. The author discusses two hypotheses as to the origin of
this additional distinction: (i) it is a remnant of the imported variety from
Southwest England, as argued by Paddock (1991); (ii) it is a development of
Newfoundland English due to contact with other English dialects spoken in
Newfoundland, as argued by Wagner (2004). He favors the second proposal as there
does not seem to be any evidence for the first in the traditional dialects from
Southwest England, but there is evidence that in North-American English feminine
pronouns can be used for all sorts of things including moving objects.
Chapter 4-Tasmania and other parts of Australia
The use of animate, in particular feminine, pronouns for inanimate referents in
varieties of Australian English is mentioned in many dictionaries and handbooks
of (Australian) English. There is, however, an enormous scarcity of systematic
studies of this phenomenon. The one exception is Tasmanian English, a variety
which is also different in morphophonological and other syntactic features with
respect to (standard) Australian English. In this chapter, the author embarks on
a systematic study of pronominal usage in this variety of Australian English on
the basis of recordings of natural and spontaneous conversations made by Andrew
Pawley. He shows that the Tasmanian system parallels that found in Southwest
England and Newfoundland in the pronominalization of inanimate count nouns by
animate pronouns. He observes a strong cline from concrete count nouns to
abstract nouns such that nouns with concrete reference are considerably more
likely to be pronominalized by animate pronouns than those with abstract
reference. The difference with the previously discussed varieties is that
Tasmanian Vernacular English makes extensive use of feminine forms instead of
masculine ones. The author shows that the semantic domains in which masculine
pronouns are used are extremely restricted, including animals of male and
unknown sex, plants and trees. Furthermore, some semantic domains (e.g. food and
drinks) show optionality of feminine versus neuter pronoun use. Discussion of
data from other regions of Australian English mainly support the generalizations
made for Tasmanian Vernacular English.
Chapter 5-Informal spoken American English
This chapter presents an excursion into pronominal usage in some informal spoken
varieties of American English. It aims to show that the fundamental claim of the
book (i.e. pronominal gender depends on the degree of individuation of the
referent) has relevance above and beyond specific regional varieties of English.
This is done by re-evaluating the data from a sociolinguistic study by Matthiot
(1979), who collected data in the Los Angeles area and in Buffalo. The
sociolinguistic hypothesis tested by Mathiot was whether the use of 'he' and
'she' reflects the (emotional) attitude of the speaker. She assumes that the use
of animate pronouns for inanimate nouns is tantamount to the upgrading or
assimilation of a non-human entity to a human being. Moreover, she concludes
that there are probably no restrictions on the kinds of non-human entities that
allow upgrading. The author of the present study aims to show that the data
allow for a reinterpretation in terms of individuation of the referent. He shows
that indeed animate pronouns are mainly used for individuated inanimate nouns.
Furthermore, a grouping of the data in terms of semantic domains shows that some
(e.g. animals of unknown sex) are regularly masculine, others are feminine (e.g.
environment), and that for some domains (e.g. tools and instruments) pronoun
choice depends on the sex of the speaker. Thus, the author concludes that the
usage of pronominal gender in informal spoken American English is more
systematic than previously assumed.
Chapter 6-Fictional texts
This chapter presents data on pronominal usage from written texts making use of
the approximately 300 examples of animate pronouns used for inanimate referents
in a study by Svartengren (1927). The data consist of created examples of
informal speech from novels written in the English language. The examples only
contain feminine pronouns to the exclusion of masculine ones and in certain
cases the animated usage may be analyzed as a stylistic effect. Notwithstanding
these differences with the earlier discussed spoken data, the author wants to
show in this chapter that the data from fictional texts by and large align with
the data from the other varieties. Where Svartengren proposes that the use of
animate pronouns for inanimate referents signals a special attitude of the
speaker towards those referents, he also raises the suspicion that individuation
might play an important role. In order to test this hypothesis the author
analyzes almost all examples discussed by Svartengren which contain mostly count
nouns, some mass nouns, and a considerable amount of abstract nouns making the
ratio of animate pronouns with reference to mass and abstract concepts much
higher in fictional texts than in the naturally occurring data. Nevertheless,
the author concludes that despite these differences the general trend found in
the other varieties is also present in fictional texts: animate pronouns pick
out inanimate referents ranking high in terms of individuation.
Chapter 7-Generalizations across varieties of English
The goal of Chapter 7 is to pull together the commonalities linking pronominal
gender across different varieties of English. It starts with a recapitulation of
pronoun usage in the regional varieties of Southwest England, Newfoundland, and
Tasmania, which shows that all three use animate pronouns for concrete count
nouns with possible further distinctions within the category of inanimates, and
the neuter pronoun for mass or abstract nouns. In the second section the author
presents the following hierarchy or continuum of individuation as ''a useful aid
furnishing our understanding of these systems'' (p.140): proper names > humans >
animals > inanimate tangible objects > abstracts > mass nouns. The systems of
the regional varieties discussed in the book can be directly mapped onto this
hierarchy and each establishes a different cut-off point. The author argues that
the use of this hierarchy has several advantages. It can be used to assess the
degree in which regional varieties differ from standard English through
comparison of cut-off points. It has predictive power concerning general
distributional patterns: due to the Semantic Map Connectivity Hypothesis only
systems with a contiguous or connected area in the hierarchy are expected to
occur and those that divide the semantic space underlying the hierarchy in more
than two parts are ruled out. Finally, the hierarchy is an explanatory
instrument which has proven its usefulness in several other domains of grammar
in this way raising the problem of pronominal gender to a more general and
abstract level. At the same time, the author also voices some concerns in using
the hierarchy as an explanation. First, the hierarchy is an inductive
generalization and therefore an explanation in terms of it 'must perforce remain
relatively weak' (p. 143). Second, there does not seem to be a motivation for
the relationship between grammatical marking and the hierarchy of individuation.
Here the author concurs that ''if the hierarchy of individuation could
convincingly be shown to have basis in human perception or cognition in general,
which I believe it has, this would certainly increase the possibility of it
being used as an explanatory tool for the analysis of grammatical distinctions.
Nevertheless, even if such an analogy can be established, it does not
necessarily follow that grammatical marking obeys the hierarchy of individuation
just because human cognition is based on this hierarchy'' (p. 144). The chapter
closes with a section arguing that pronominal usage in informal spoken American
English as well as in fictional texts is in harmony with the hierarchy of
individuation.
Chapter 8-Modern Standard English
In general the system of pronominal gender in Modern Standard English has a
clear semantic basis involving animacy and sex of the referent. Nevertheless,
there's also a certain amount of variability and this chapter is concerned with
those nouns for which variation in pronoun usage is possible. The chapter starts
off with a discussion of the relationship between specific and generic reference
and high versus low degree in individuation. In section 8.1 the author discusses
and rejects the postulation of a category of gender in English. After a short
overview of nouns for which the use of gendered pronouns is invariable (e.g.
'man', 'woman', 'king', 'queen'), a description of nouns for which variable
pronoun usage is possible follows. Within the class of human referring nouns we
find the so-called personal dual gender nouns (e.g. 'parent', 'friend') which
allow both masculine and feminine reference, common gender nouns (e.g. 'child',
'baby') which allow anaphoric reference with animate and neuter pronouns, and
collective nouns for which the neuter pronoun or plural 'they' can be used. The
other two classes are nouns referring to animals and inanimates referring nouns
which are the central concern of the book. The final section of the chapter
presents a review of proposed factors influencing pronoun usage, including
personification and sympathy of the speaker, and through the discussion of two
empirical studies aims to show that the variation in pronoun usage is determined
by a few well-defined constraints, with special reference to individuation. The
first study by MacKay and Konishi (1980) investigates the use of animate and
inanimate pronouns for non-human antecedents in children's literature. They
examine the correlation of pronoun use with six variables some of which
(specificity, centrality) seem to be related to the notion of individuation as
advanced by the author. He concludes that the variation between animate and
inanimate pronouns ''is more complex than is usually assumed to be the case and
goes beyond a simple contrast in terms of animacy'' (p. 169). The second study
reported is by Newman and contrasts the use of 'he/she' with that of 'they'. It
shows that the choice between the two options is not only related to the wish to
maintain sex-neutrality but also to individuation. It shows that the singular
pronouns pattern with referents high in individuation, whereas 'they' patterns
with those low in individuation. The conclusion the author draws from the
discussion in this chapter is that although the pronominal gender system of
Modern Standard English is based on relatively simple semantic principles, it
also has its intricacies.
Chapter 9-A Cross-Linguistic View on English Varieties
This chapter puts the use of pronominal gender in English in a wider
cross-linguistic perspective. It contains two sections on Germanic languages,
one on Romance languages, and a third one on other languages. The scope of the
study is all agreement triggered by a nominal head that involves contrasts which
can be related to the hierarchy of individuation. In the first section on
Germanic the author presents data from varieties of Germanic which exhibit a
system of pronominal gender based on the mass/count distinction comparable to
the ones reported for English. The dialects and languages discussed include the
Danish dialect spoken in West Jutland, the variety of Frisian spoken on
Helgoland, Dutch, and Afrikaans. The next section shows that in many Germanic
languages and dialects the assignment of a noun to a particular gender class can
be partially motivated by the mass/count distinction and the hierarchy of
individuation. The section on Romance languages shows that central Italian
dialects make a distinction within the non-feminine gender between count
(masculine) and mass (neuter) nouns which is signaled on articles,
demonstratives, pronouns, and adjectives. Likewise, in Asturian Spanish there is
mass/count agreement on pronouns and adjectives. The final section of the
chapter presents data from a wide variety of languages illustrating gender
agreement influenced by the degree of individuation of the noun (e.g. in Arabic
and Bantu), the correlation between gender and the mass/count distinction (e.g.
in Bantu and Daghestanian), and gender systems organized on the basis of the
animate/inanimate distinction (e.g. Algonquian and Niger-Congo).
Chapter 10-The Categorial Status of Pronominal Mass/Count Agreement
The aim of this chapter is to contribute to the correct categorization of
agreement systems based on the mass/count distinction. Previously these have
been analyzed in terms of gender or number but the author wants to argue that
they possess certain aspectualizing functions as well. He first points out that
a conception of the mass/count distinction in terms of a binary opposition
associated with different nominal entries in the lexicon cannot hold true.
Instead, the mass/count distinction seems to be a property of noun occurrences.
Mass/count agreement systems are often interpreted in terms of gender, the
author argues that such an analysis is problematic for several reasons including
the fact that mass/count gender systems are rare cross-linguistically and only
involve pronouns whose deictic use is not compatible with an agreement category
like gender, but most importantly the fact that the mass/count distinction is
not a lexical property of nouns. A similar story holds for the interpretation of
the systems as number systems. From this discussion, the author concludes that
''the mass/count agreement systems at issue here are more likely to be analyzed
as number systems than as gender systems'' (p. 236). Finally, the author puts
forward an analysis of the mass/count agreement systems in terms of nominal
aspect. This is inspired by the fact that speakers can use pronominal agreement
to portray an entity as individuated or non-individuated. This is reminiscent of
verbal aspect which can be used to express different perspectives on a
situation. Moreover, there is a close semantic relationship between verbal
aspect and the mass/count distinction. Nevertheless interpretation of these
mass/count agreement systems in terms of nominal aspect remains questionable on
both formal and distributional grounds. The conclusion of this chapter must then
be that 'we cannot but admit that the mass/count systems discussed in this study
apparently have some properties of gender systems, number systems and perhaps
even aspectual systems' (p. 241).
Chapter 11-Conclusion and Outlook
The final chapter of this study summarizes the previous chapters and presents
some open issues and loose ends including historical, sociolinguistic and
theoretical problems.
EVALUATION
Siemund's book is a well-written and clearly structured study. With its
accessible style and explanatory clarity the book is extremely reader friendly.
Throughout the book the author displays an impressive patience in his
descriptions of previous research and the discussion of the data, and almost
nowhere does he rush to premature conclusions. Altogether this makes for an
extremely readable book. Also content-wise there is much of interest. For one,
the book establishes that the distribution of pronominal gender forms in the
varieties of English is much more complex, and hence much more interesting, than
one might think at first sight. As with any other publication, some issues
remain that deserve further discussion, two of which will be addressed below.
In my opinion, the kind of variation data used by the author are at the same
time the strongest and weakest part of the book. It is astonishing to see the
kind of robust patterns recurring in such a diverse collection of data sets
which have hardly anything in common in terms of genre, style, register or the
way in which the data have been collected. The emergence of such clear
generalizations from such a heterogeneous data set seems to strongly support the
grammatical relevance and reality of the phenomenon under discussion. One can,
however, bargain about the presentation, interpretation and comparison of the
different data sets examined. My main objection concerns the fact that the
author only presents tables with counts and percentages but does not corroborate
his findings by the use of statistics. Instead the reader is often presented
with impressionistic descriptions of the data and qualifications such as ''there
is no significant (in the non-technical sense) difference'' (p. 111) which are
rendered rather meaningless without a proper grounding. Although I am aware that
statistical modeling is not everyone's cup of tea, I believe the data analysis
and conclusions presented in the book could have benefited from some descriptive
statistics and simple tests (chi-square, Fisher's exact). More importantly,
given that the different data sets are hardly comparable and numbers within data
sets often highly skewed, application of such methods would have allowed the
author to draw stronger conclusions than he was able to do now in many instances.
Statistical methods might also have been of help in finding a solution to the
second point of discussion which is the definition of the notion of
individuation. Although the concept of individuation is very central to the
book, its definition seems to change from chapter to chapter. As stated in
chapter 1 and detailed in chapter 7 individuation is conceptualized in terms of
a hierarchy ranging from proper names to mass nouns, cf. above. The distinction
in the hierarchy most relevant to pronominal usage in the varieties of English
is that between count and mass nouns. However, in many instances this
distinction does not suffice and further parameters, with no place on the
hierarchy, are needed, e.g. the notion of mobile in Newfoundland English. In
chapter 8 the author himself notes that maintaining his line of argumentation in
terms of individuation ''necessitates a concept of individuation that
significantly goes beyond undisputed notions (or qualities) like discreteness or
boundedness ... Many of the factors discussed there can plausibly be subsumed
under the notion of 'individuation''' (p. 163). Instead of subsuming all kinds of
different factors under the notion of individuation one could also keep this
notion simple and investigate how its interaction with these other factors
results in the observed patterns. Although under the latter scenario it may seem
that there is no unified explanation for the phenomenon of pronoun usage, one
can wonder how unified the explanation is with a broad notion of individuation.
Whichever way one chooses, and I would opt for the latter, I think statistical
methods may help to elucidate and to better understand pronominal usage in the
varieties of English.
The kind of micro-typological approach advocated by the author is still in its
infancy. In his book he has demonstrated that the combination of this
methodology with a variationist approach can result in new and interesting facts
about the grammar of a well-studied language like English. These empirical
results make a strong case for the methods applied in the book to be used in
future research. I am convinced that taking into account the remarks raised
above this will yield many more interesting results.
REFERENCES
Elworthy, F.T. (1886)[1965]. _The West Somerset Word-Book and Glossary_. Vaduz:
Kraus Reprint LTD.
MacKay, D. and T. Konishi. (1980). Personification and the pronoun problem.
_Women’s Studies International Quarterly_ 3, 149-163.
Matthiot, M. (1979). Sex roles as revealed through referential gender in
American English. In: M. Matthiot (ed.), _Ethnolinguistics: Boas, Sapir and
Whorf revisited_. The Hague: Mouton, 1-47.
Paddock, H. (1991). The actuation problem for gender change in Wessex and
Newfoundland. In: P. Trudgill and J. K. Chambers (eds), _Dialects of English:
Studies in Grammatical Variation_. London: Longman, 29-46.
Svartengren, T. (1927). The femine gender for inanimate things in
Anglo-American. _American Speech_ 3, 83-113.
Wagner, S. (2004). Gender in English Pronouns: Myth and Reality.
http://deposit.ddb.de/cgi-bin/dokserv?idn=97211100x.
ABOUT THE REVIEWER
Peter de Swart received his PhD in Linguistics from the Radboud University
Nijmegen, the Netherlands in November 2007. His dissertation entitled
''Cross-linguistic variation in object marking'' examines the influence of
semantic features, in particular animacy, on the marking of direct objects.
Until the end of 2008 he continued his research as a postdoctoral researcher in
an NWO financed project 'animacy' at the same university. Since January 2009 he
is a postdoctoral researcher at the Center for Language and Cognition Groningen
of the University of Groningen, the Netherlands. In his NWO VENI-project ''The
status of hierarchies in language production and comprehension'' he investigates
the role of hierarchies of individuation in grammar and use.
|
