Sie sind auf Seite 1von 18

English Today

http://journals.cambridge.org/ENG
Additional services for English

Today:

Email alerts: Click here


Subscriptions: Click here
Commercial reprints: Click here
Terms of use : Click here

A gay paper: why should sociolinguistics bother with semantics?


Justyna A. Robinson
English Today / Volume 28 / Issue 04 / December 2012, pp 38 - 54
DOI: 10.1017/S0266078412000399, Published online: 03 December 2012

Link to this article: http://journals.cambridge.org/abstract_S0266078412000399


How to cite this article:
Justyna A. Robinson (2012). A gay paper: why should sociolinguistics bother with semantics?.
English Today, 28, pp 38-54 doi:10.1017/S0266078412000399
Request Permissions : Click here

Downloaded from http://journals.cambridge.org/ENG, IP address: 192.41.131.252 on 29 Jun 2016

A gay paper: why should


sociolinguistics bother with
semantics?
J U S T Y N A A . RO B I N S O N
Can sociolinguistic methods shed light on semantic
variation and change in reference to the adjective gay?

1. Introduction1
The study of meaning and changes in meaning has
enjoyed varying levels of popularity within linguistics. There have been periods during which the
exploration of meaning was of prime importance.
For instance, in the late 19th century scholars considered the exploration of the etymology of words
to be crucial in their quest to nd the true meaning of lexemes (Geeraerts, 2010; Malkiel, 1993).
There have also been periods where semantic
analysis was considered redundant to linguistic
investigation (Hockett, 1954: 152). In the past
2030 years semantics has enjoyed a period of
revival. This has been mainly led by the advances
in cognitive linguistics (and to some extent, historical linguistics) as well by the innovations associated with the development of electronic corpora
and computational methods for extracting and tracing changes in the behaviour of the lexicon (cf.
Geeraerts, 2010: 168ff, 261ff). However, there
are still areas of linguistics which hardly involve
lexis in their theoretical and epistemological considerations. One such area is sociolinguistics.
Sociolinguistics as a distinct eld began to
emerge in the US in the late 1950s as a response
to some weaknesses of the then-prevailing structuralist models in accounting for variability within the
linguistic system. With the foundational paper of
Weinreich et al. (1968) and pioneering studies
carried out in the US (Labov, 1963, 1966) and
Europe (Trudgill, 1974), sociolinguistics established
itself as one of the most dynamically developing
areas of linguistics, and has since revolutionised
the way we think about language use. However, it
has done so by investigating mainly phonological

and to some extent morpho-syntactic variation.


Lexis has been mainly studied descriptively and
semantic variation has been rarely addressed within
sociolinguistics. As a consequence, the past 40
years of sociolinguistic research has focused on
exploring the meaning of variation (cf. Eckert,
2012), while leaving the variation of meaning aside.
A possible reason for this situation may be
related to methodological issues in capturing the
complexity of meaning. Sociolinguistic variationist
techniques require rather dened segments of
language that can be objectively derived and

JUSTYNA A. ROBINSON is
a Lecturer in English
Language and Linguistics at
the University of Sussex,
United Kingdom. Her
research focuses on semantic
variation and change,
sociolinguistics, cognitive
linguistics, corpus linguistics
and how these interact.
Recent publications include the co-edited volumes
Current Methods in Historical Semantics (de
Gruyter Mouton 2012), Variation in Language and
Language Use: Sociolinguistic, Socio-cultural and
Cognitive Perspectives (Peter Lang 2012),
Polysemy and Synonymy. Corpus Methods and
Applications in Cognitive Semantics (John
Benjamins, forthcoming) and a special issue of
Review of Cognitive Linguistics entitled Cognitive
Sociolinguistics: Variation in cognition and
language use (John Benjamins, 2012). Email:
justyna.robinson@sussex.ac.uk
doi:10.1017/S0266078412000399

38

English Today 112, Vol. 28, No. 4 (December 2012). Printed in the United Kingdom

http://journals.cambridge.org

Downloaded: 29 Jun 2016

2012 Cambridge University Press


IP address: 192.41.131.252

compared. Phonological and morpho-syntactic


variables comply with this criterion relatively
well. However, the fuzzy nature of meaning can
be seen as a problem when the methodologies
demand clear boundaries. Sociolinguists themselves have admitted that investigating meaning
is problematic since semantic variables do not
t into the notion of a variable proposed by sociolinguistics (Labov, 1972; cf. Lavandera, 1978) and
they have expressed doubts about the benets of
studying lexical change within the variationist
paradigm (Labov, 1978: 8, 2001: 123).
There are some exceptions to the generalisation
that lexical (both onomasiological and semasiological) variation has been overlooked within sociolinguistics. For instance, interest in lexis has
continued in studies of regional variation (Llamas,
1999; Finnegan, 2011). There have also been several attempts to investigate meaning variation within
functional paradigms (Hasan, 1989, 1992, 2009) or
discourse analysis frameworks (Cheshire, 2007;
Macaulay, 2005, 2006; Stenstrm, 2000; Wong,
2002, 2006). It is also worth highlighting that
consideration of meaning variation surfaces in
studies that examine styles and indexical elds
(Eckert, 2008).
The current paper contributes to this modest
body of research by showing the value of sociolinguistic methods in the study of semantic change.
Specically, I implement here the variationist construct of apparent time, which proposes that generational differences in language use reect actual
real-time changes in language. The apparent-time
construct relies on the assumption of the stability
of vernaculars. This means, for instance, that an
older persons language use to a large extent
reects forms this person acquired when they were
much younger, say in their twenties. The validity
of the apparent-time construct has been shown in
investigations of language change in a number of
existing (Bailey, 2002; Sankoff, 2006) and past
communities (Nevalainen & Raumolin-Brunberg,
2003).
In recent work (Robinson, 2010a, 2010b, 2012)
I have demonstrated how the apparent-time construct can also successfully be applied to researching semantic variation and change, and the current
study is based on my ndings in these earlier
papers. The aim of the current paper is to apply
this established method of tracing semantic
change-in-progress to the investigation of an
already well-researched word which is conceptually
complex, and which indexes a range of sociocultural attitudes and stereotypes. There is no better
example to choose for this task than gay. I

demonstrate that a variationist analysis can provide


more insights into the variation and change of
the adjective gay in comparison to what we already
know from other sources and through using
other investigative paradigms. By doing so I am
able to assess the added value of using variationist
sociolinguistics, and ultimately answer the question of why sociolinguistics should bother with
semantics. After introducing the history of gay,
I present and discuss my ndings to conclude
that taking a sociolinguistic perspective is benecial in the investigation of semantic variation
and change.
1.1 Gay as a salient example of semantic change

Gay is one of those words that is frequently mentioned in discussions of semantic change. Gay is
the classic single example of semantic change
that is used to illustrate relevant theories in
textbooks such as Campbell (2004: 281), Carter
et al. (2001: 98), Crowley (1997: 154), Culpeper
et al. (2009: 307), McMahon (1994: 1745),
Meyerhoff (2011: 5962), and OGrady et al.
(2005: Ch.7). A range of other academic publications also use gay to illustrate their arguments.
References to gay are used as an example by scholars working in various areas of linguistics, such as
Cherry (1986), Cook & Hirst (2011), Cymbalista
(2009), Kay & Allan (forthcoming 2013), Lehrer
(1992), Leith (1997: 767); and in encyclopaedias,
compendia and surveys, such as Frawley (2003:
395), Gramley & Ptzold (2004: 35), and
Millward & Hayes (2010: 336). Sometimes these
studies reect shifts in the meaning of gay that
were 'current' for the authors themselves. Thus,
scholars publishing in the early 1990s mention
the change between gay happy and gay homosexual (Radwanska-Williams, 1990: 99; Algeo,
1990: 404; Leith 1997: 76), while others focus
on the change between gay homosexual and
gay lame (Lalor & Rendle-Short, 2007).
Finally, there is a body of linguistic research that
discusses gay more widely or considers gay as
the subject of investigation. Since one of the meanings of gay is homosexual, such studies focus on
gay in the context of gay slang (Coleman, 2012),
gender identity at both synchronic (Rasmussen,
2004) and diachronic levels (Curzan, 2003),
homophobic discourse (McCormack, 2011), and
political correctness (Hughes, 2009), just to name
a few examples.
Considering the range of linguistic contexts in
which gay has been discussed, it makes an interesting case study for the current investigation.
However, before researching the variation and

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

39

IP address: 192.41.131.252

change of this word in apparent time, let us briey


summarise the real-time evidence on its historical
development.
1.2 Etymology of gay

The adjective gay is currently used in several


senses that roughly belong to one of the umbrella
sense groups of: gay happy, cheerful, bright,
gay homosexual, and gay boring, lame, stupid,
weird. The history of gay has never been straightforward, and this may relate to the fact that from
quite early on the denotations of gay have been
on the border of sexual taboo.2
Although etymologically gay can be traced to
Old High German, meaning rapid, fast, surprising, gay entered English as a borrowing from
Middle French, in which it was already used with
a range of meanings including happy, carefree,
bright, or lewd, among many others. According
to the Oxford English Dictionary Online (hereafter,
the OED Online), the rst documented use of the
word (a1310) describes to persons, their actions
and attributes, and is dened as merry, cheerful
and forward. The history of English demonstrates
that the merry, cheerful and forward use of gay
easily yielded itself to various metonymic
extensions. From the 14th century gay started
to signify bright or lively-looking, especially
with reference to colours. A range of more
pejorative sense extensions that usually relate to
sexual activity also developed. Thus, from the
16th century, we see that gay was used to mean
frivolous or addicted to social pleasures and
dissipations, which led to the development of the
meaning promiscuous or even prostitute, as in
gay girls, which was used in early 20th-century
England.
A recent signicant change in the meaning of
gay led to the development of a specialised sense
homosexuals, especially male. According to the
OED Online, the rst record of this use comes
from 1935, although other scholars speculate that
gay could have been used to mean homosexual
as early as the 1880s (Hughes, 1991; Burridge,
2004: 59) or the 1920s (Butters, 1998). There is
also some debate about the emergence of gay
homosexual and its relationship to earlier senses
of gay. The OED Online indicates that gay homosexual derived from gay addicted to social pleasures. On the other hand, Burridge (2004: 59)
seems to suggest that gay homosexual derived
from gay prostitute by arguing that it is common
for contemptuous terms for females to turn into
terms of abuse for male homosexuals. Another
perspective is put forward by Butters (1998), who
40

http://journals.cambridge.org

demonstrates that the contexts of the emergence


of gay homosexual were highly complex and
both the meanings gay addicted to social pleasures and gay prostitute impacted on the development of gay homosexual. Butters (1998) also
argues that gay homosexual was initially a selfreference term that was exclusively used by
homosexuals. However, even in the 1950s, gay
homosexual was not at all an entirely accepted
established term for those groups, as for some
speakers gay had far more derogatory and unpleasant connotations than, for instance, queer or a
more euphemistic my kind/his kind (Butters,
1998: 192, 197). During the 1950s, gay homosexual emerged from more bohemian, cosmopolitan,
homosexual, and subcultural circles, and entered
into more general slang usage (Butters, 1998). By
the 1970s, gay had become the preferred term of
self-reference for homosexuals, and this was
affected by the gay/lesbian/bisexual/transsexual
rights movement. These developments led to a
polysemic clash (Algeo, 1990: 404) and as a consequence, the new semantics of gay has largely
ousted previous happy meanings.
Gay is currently used to refer to homosexual
males and females. More recently, the adjective
gay has undergone yet another sense development.
The OED Online indicates that in the late 1970s
gay started being used in slang American English
as meaning foolish, stupid, socially inappropriate
or disapproved of; lame. The OED Onlines
rst British English quotation of this use is not
until 2002:
At last the rst ofcial conrmation that the term
gay has now ousted pants as the popular youth
slang for naff. (Independent, 2002)

Various sources report that the use of gay lame


is now widespread, especially among young people
(Lalor & Rendle-Short, 2007).3 However, this particular use of gay is not uncontroversial in the context of wider public discourse. In 2006, when Chris
Moyles told his BBC Radio 1 breakfast show audience of 6.3 million that he thought a ringtone was
gay, the BBC was bombarded with criticism and
complaints about homophobic language use. The
BBC Body of Governors defended Moyles, saying
that Children and students use gay as a shorthand
for rubbish (June 6, 2006), and [it] may be used
in a pejorative sense without insulting homosexuals (Telegraph.co.uk 2006).
The claim that Moyles used gay in the sense rubbish could be supported by Lalor & Rendle-Short
(2007), who show that indeed certain speakers
E N G L I S H TO D AY 1 1 2

Downloaded: 29 Jun 2016

D ec e mbe r 2 0 12
IP address: 192.41.131.252

interpret gay as lame, especially when it is used


with inanimate subjects (e.g. ringtone), albeit in
Australian English. There is also some research
into gender identity and homophobic discourse in
American, Australian (e.g. Rasmussen 2004) and
English secondary schools (McCormack 2011)
suggesting that gay lame is not used with homophobic undertones. On the other hand, there are
also studies which argue otherwise and consider
gay to be part of a homophobic discourse, e.g.
Lalor & Rendle-Short (2007: 168). Another perspective is presented by one of the anti-homophobia
campaigns launched in the UK which uses a
slogan Homophobia is gay. Its founders claim
that they are not advocating the use of the word
gay as a derogatory term but use the word gay
ironically.4 In any case, gay lame is an interesting development, which, so far, has been recorded
in various sense instantiations.5 We should certainly keep an eye on this variant for more changes
to come.
1.3 Corpus evidence

In order to complement this chronological picture


of the semantic change of gay, it is worth looking
at available evidence for learning about quanticational aspects of the change. The obvious place to
gain knowledge about measurable aspects of linguistic usage is via corpora. Such an investigation
of the meaning variation and change of individual
words usually involves studying the frequency and
types of collocations the given expression takes
over a period of time. This idea rests on the
assumption that if the meaning of a word changes,
its collocates are also likely to change. For
example, Davies (2010: 461) demonstrates that
between 199094 and 200509 gay/gays started
being used with a new set of collocates (cowboys,
union), reecting socio-cultural changes in
American society. There are also studies that investigate change on the basis of collocates which are
derived via the use of automatic detection algorithms. For example, Wijaya & Yeniterzi (2011)
provide quantitative evidence that the signicant
change of context in which gay has been used
since 1954 occurs around the 1970s, which is
when the word started being used to refer to homosexuals. Another way of investigating meaning in
corpora is to explore synonyms of a given item
(and their change). Comments on gay derived
from such a methodological approach appear in
Aarons (2010: 278) investigation of the adjective
lame. The last, most time-consuming, but perhaps
most intuitive method of investigating the meaning
of words in corpora involves simply going through

hundreds of concordances and manually coding


data for different types of senses and then establishing the frequency of use of different senses across
time. This has been the methodology employed in
the present study.
The distribution of individual senses of the
adjective gay has been traced across two corpora
of British English: the British National Corpus
(hereafter the BNC) and the Oxford English
Corpus (hereafter the OEC). While all the data
from the BNC (19601974) and the BNC
(19751984) are analysed, the analysis of data
from the BNC (19851993) is restricted to several
text types: spoken demographic, spoken contextgoverned, and written-to-be-spoken. For the purpose of the current project I also selected a subcorpus from the OEC. This sub-corpus includes
data from British English which are tagged as spoken and written-unedited.6 While all the concordances from the BNC are used, a random sample
of 500 concordances has been selected from the
OEC. Tables 1 and 2, respectively, demonstrate
the counts of tokens and the relative frequency of
the senses of the adjective gay in the corpora (for
an explanation of the coding of senses, refer to
Section 2).
The corpus evidence indicates that the adjective
gay is undergoing semantic change. The use of the
older sense gay happy is diminishing, balanced
by the increase of use of gay homosexual. The
real-time data also evidences novel uses of the
adjective gay, such as gay unmanly and gay
lame.
1.4 Research questions

The brief account of the meaning of gay presented


so far illustrates that a number of interesting and
intricate issues revolve around this polysemous category. Only within the last three generations has
the adjective gay changed its meaning dramatically
from roughly gay happy to gay homosexual and
now to gay lame. This is not to say that any of
these changes are completed, as available evidence
clearly indicates that all of these senses of gay are
still in active use. A recent survey carried out by
Lalor and Rendle-Short (2007) suggests that the
key difference in understanding gay as happy as
against homosexual and lame is the age of
speakers.7 Since Lalor and Rendle-Short (2007)
focus mainly on exploring the usage of the most
recent sense gay lame, in this paper I examine
the current use of the older senses of gay as well.
The analysis of the usage of gay could also
consider including more socio-demographic information on speakers.

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

41

IP address: 192.41.131.252

42
Table 1: Number of tokens of senses of the adjective gay in corpora
Count of meaning variants
Gay

Lame

Unmanly

BNC 19601974

BNC 19751984

Happy

Reported
Lame

Reported
Homosexual

Reported
Happy

Other/
NA

Sum

27

27

21

19

12

52

Homosexual

BNC 19851993

124

22

155

OEC 20002006

462

13

500

Sum of senses

517

58

47

E N G L I S H TO D AY 1 1 2

Table 2: Relative frequency of tokens of senses of the adjective gay in corpora


Frequency of meaning variants
Gay

D ec e mbe r 2 0 12

Lame

Unmanly

Homosexual

Happy

Reported Lame

Reported Homosexual

Reported Happy

Other/ NA

BNC 19601974

0.00

0.00

0.00

1.00

0.00

0.00

0.00

0.00

BNC 19751984

0.00

0.00

0.40

0.37

0.00

0.00

0.00

0.23

BNC 19851993

0.01

0.01

0.80

0.04

0.00

0.00

0.00

0.14

OEC 20002006

0.01

0.01

0.93

0.01

0.01

0.00

0.00

0.03

http://journals.cambridge.org

Downloaded: 29 Jun 2016

IP address: 192.41.131.252

The question of the innovation and diffusion of


change could be addressed when investigating
such an interesting case of polysemy. How does
change progress at the level of the community?
Who are the innovative speakers and who are the
most conservative ones? Also, one may consider
to what extent the processes of change are
accompanied by an awareness of different meanings of gay, their positive/negative evaluation,
and other overt comments. Perhaps answers to
these questions could help in nding out how it
is possible for a word to change its meaning to
the opposite within a relatively short space of time.
A multiplicity of issues that have been affecting
gay at the levels of age, gender, and overt and covert attitudes makes this word a suitable candidate
for testing the value of a sociolinguistic paradigm
and assessing whether this paradigm can add to
previous insights in understanding the variation
and change of gay.

comprises 223 examples of use of gay. All of


these uses are the most salient senses for each individual and not necessarily all the senses of gay that
the given participant knew. In order to ensure that
speakers were using gay in the way that was most
natural to them, I tried to make them feel comfortable, encouraged them to reply to my questions
quickly and spontaneously, and assured them that
there were no bad answers.
The individual uses of gay were afterwards
classied into sense clusters on the basis of grouping similar usage (referents and explanations of use
for these referents) and matching usage with dictionary denitions. However, the use of gay
unmanly did not correspond to a separate dictionary denition.
The usage of gay in apparent-time data and in
corpora is grouped in the following way:

2. Method and data


The data for the current investigation of the adjective gay comes from a larger project in which
I explored the usage of polysemous adjectives
that have been undergoing change in present-day
British English (Robinson, 2010b). In order to
collect data suitable for variationist research I carried out a survey among speakers who were all
born and bred in South Yorkshire, UK. Although
I interviewed more than 100 speakers, in my quantitative analysis I could only include a sample of
72 participants who were representative of different
age groups (age range: 1194), genders and socioeconomic classes. The interviews took place in
20052006.
In order to elicit the salient usage of all investigated adjectives (including gay) for each participant, I asked a series of carefully designed
questions. The questions followed the format of
asking for a referent that could be described by
that adjective and then asking for a justication
for the use of a given referent. In this way, I managed to obtain suitable usage data without explicitly asking for denitions of the adjectives. The
following example illustrates how the usage of
the adjective gay was elicited:
Question: Who or what is gay?
Answer: My school.
Question: Why is your school gay?
Answer: Because it is boring.

There were on average three referents elicited per


speaker and the overall corpus of gay responses

gay lame: often used to describe something as


stupid or boring. For example, in the interviews a number of younger participants describe
their school as gay.
gay unmanly: refers to activities or objects
which may traditionally be associated with
females; effeminate. For example, eating salads, or carrying an umbrella. Although dictionaries do not provide explicit evidence for this
use of the adjective gay, there are several
instances evidencing this usage in interview
and corpus data.
gay homosexual.
gay happy: uses of gay happy, merry and gay
bright.
gay other or not applicable: overlapping uses
of the adjective and uses which cannot be
coded as any of the above, e.g., Gay, the female
personal name; gay in Gay Gordons, a type of
traditional dance; or in nosegay, a ower.

Additionally, so-called reported uses of the


adjective were recorded. These uses indicate that
participants admit to knowing gay lame or gay
homosexual or gay happy, but deny using the
adjective gay in this way at all or admit using it
with some stylistic-pragmatic constraints.
The structured part of the interview was followed by a more relaxed conversation about the
use of the word and its meanings during which a
range of perceptions and attitudes to variation
and change was expressed. These comments turned
out to be invaluable data in explaining the observed
variation in usage.
In the following sections I present results from
the variationist sociolinguistic analysis of the
data and interpret them in the context of the

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

43

IP address: 192.41.131.252

discussion of the semantic change of the adjective


gay.

3. Gay: the variationist analysis


3.1 Gay and age

The age of speakers is arguably one of the most


important factors used in systematising and analysing language change. A number of studies have
shown that by looking at generational differences
in linguistic usage (variation in apparent time),
we can draw conclusions about linguistic change
in progress (for a summary, see Sankoff, 2006).
Thus, the analysis of the adjective gay starts from
considering age-related variation. The mean frequency of use of individual senses of the adjective
gay has been recorded and plotted across a horizontal axis representing the age of participants, and
this is represented in Figure 1.
An initial inspection of Figure 1 indicates that
the individual senses of the adjective gay are nonrandomly distributed across different age groups.
In order to verify whether the differences between
groups are actually statistically signicant I carried
out Kruskal-Wallis non-parametric tests.8 The
results indicate that signicant age-related differences are observed in the uses gay lame (p
< .001), gay lame reported (p = .004), gay
unmanly (p = .015), gay happy (p < .001), and
gay happy reported (p = .02). Although gay
homosexual is the central sense for the community as a whole, each generation uses a different
range of additional senses of gay. Thus, happy
uses of gay are most prominent for older speakers
and the recent sense gay lame is most frequently
used by younger generations. This observation not

Figure 1. Distribution of senses of the adjective


gay across age groups
44

http://journals.cambridge.org

only reects the trends that are recorded in realtime data, but also allows for more ne-grained
conclusions on how semantic change takes place
in a community.
However, before we plunge into a discussion of
the observed age-related variation, it is worth investigating whether other social categories, such as
gender and social status, play any role in the semantic variation and change of the adjective gay.
3.2 Gay: social class and gender

Since a number of variationist studies point to the


signicance of gender and social class alongside
age in explaining change (see the review in
Coates, 2004, Eckert, 1998, or Kerswill, 2006),
the importance of these factors must be considered
in the context of the usage of gay.
The socio-economic status of participants was
established on the basis of their education, occupation and current place of residence. Speakers
occupation is measured against the National
Statistics Socio-Economic Classication (hereafter
NSEC) system, which is used for ofcial surveys in
the United Kingdom.9 The professional status of
speakers was assigned to one of three NSEC
groups, with 1 referring to higher-level professions
and 3 to lower-level occupations.10 The value of
the current place of residence was measured on
the basis of postcode information provided by
each participant. These postcodes were used to
obtain the average house prices within the respective areas from the Land Registry House Prices
Database (2008). In this way, the relative status
of each neighbourhood was determined and speakers were divided into those living in more and less
afuent areas.
Figure 2 presents the difference in the use of gay
between males and females in each generation. In
order to assess whether sense extensions of the
polysemous adjective gay are used signicantly
differently by a particular gender or a socioeconomic group, Kruskal-Wallis statistical tests
have been carried out.11 While gay lame is used
by young males (p = .001) who are still in education, gay lame reported is used by young
females who are still in education (p = .012)*.
Gay unmanly is used by males who are in the
<1930> age group (p = .015)*. Gay happy is
used by speakers who completed their education
by the age of 16 (p = .000), especially if they are
males (p = .004), as well by speakers occupying
lower occupations (NSEC3) in the <3160> age
group (p = .027)*. Gay happy reported is only
marginally signicantly used by educated speakers
(p = .054), especially educated females (p = .059).
E N G L I S H TO D AY 1 1 2

Downloaded: 29 Jun 2016

D ec e mbe r 2 0 12
IP address: 192.41.131.252

Figure 2. Distribution of senses of the adjective gay across gender and age

Females also report on all senses of gay taken


together (p = .011). Gay other is used by professionals (p = .019).
The quantitative analysis of the sociolinguistic
variation of the adjective gay indicates that the
use of individual sense variants relates to the age,
gender and the socio-economic status of speakers.
In the following sections, these results will be interpreted in the context of patterns of semantic change
of the adjective gay.

4. The changing meanings of gay


from the variationist perspective
4.1 The overall pattern of change of the adjective
gay

In order to provide greater insight into how semantic variation and change happens in a community, I
have closely inspected the variation in the use of
different senses of gay in the context of the age,
gender, and social class of participants, and supplement this discussion with a range of qualitative
comments provided by participants of the study.
After commenting on the general ndings of the
survey, I focus on a detailed analysis of each
sense of the adjective gay.
The variation in gay presented in Figure 1 indicates that different senses of gay are active for
different generations. Although gay homosexual
is the key sense for every generation, there is variation in how other senses of gay are used. For
instance, while for the youngest speakers gay
happy is a peripheral sense, the oldest speakers
would consider gay happy as the central reading
of gay. Additionally, within each generation differences in the use of gay relate to the gender or social
class of speakers. For instance, gay unmanly is

only used by males, whereas educated females


more frequently report on using gay happy.
These ndings already demonstrate that the variationist paradigm provides a more ne-grained
look into the variation and data for the analysis
of this variation. But what about the change?
In this analysis we are drawing on the assumption that age-related differences in use are indicative of real linguistic changes (the apparent-time
construct). Thus, in order to make conclusions
about the diachronic development of gay we can
examine how the frequency of use of individual
senses changes from generation to generation
(Figure 1). It is clear that the use of different senses
of gay across different generations is not random,
but seems to be structured. The use of individual
senses gradually alters with decreasing age, forming clear trends of increased or decreased use of
gay, cf. gay lame and gay happy, respectively.
These trends correspond to the general trends for
the semantic development of gay presented in the
real-time data, with gay happy being historically
an older sense and gay lame a more recent
innovation.
It is worth adding that apart from giving detailed
information about types of senses used and indicating the general directions of change, the variationist analysis presents us with information on the
relative frequency of use of gay in each generation,
and thus allows us to measure the rate and pace of
change. Thus, while gay happy is the central
sense for speakers over 60, its frequency of use
drops by an overall amount of 20% for the next
younger age group and maintains even more peripheral status for the remaining generations. This
quantitative information allows us to build a
usage prole for each speaker and each generation,
consisting of information on the central and

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

45

IP address: 192.41.131.252

peripheral senses of a given polysemous category


expressed quantitatively. Comparison of this information for speakers of different ages allows us to
demonstrate more precisely how semantic change
happens. So I turn now to discussing the details
of how the change happens by surveying the
change of individual senses.
4.2 Gay happy

Figure 1 clearly indicates that the use of gay


happy diminishes with the age of speakers.
While the oldest generation used gay as happy
in nearly 50% of cases, this happy usage of gay
diminishes gradually to 5% among the youngest
speakers. On the one hand, this trend is unsurprising since it conforms to the general observations
from real-time data which indicate a diminishing
use of gay happy. On the other hand, it may be
surprising to nd that so many speakers (including
the youngest ones) use a sense that pragmatically
clashes with gay homosexual (cf. Algeo, 1990:
404). Closer inspection of the responses provides
more context for this situation. Alongside the spontaneous uses of gay happy in each generation, there
are a proportion of responses that are stylistically
constrained and presented here as reported uses.
So there are speakers from each age group who mention gay happy, but they explicitly distance themselves from using this sense. Reporting on gay
happy is evidenced in each generation. However,
the proportion of reported vs. non-reported uses
of gay happy decreases with increasing age of
speakers.
Language change is not only reected in quantitative differences in uses of gay happy but also in
additional comments provided by participants of
the survey. A number of speakers from older age
groups admitted that they hardly use gay happy
or that they do not use it now that it has developed
the sense gay homosexual, which happened in a
time that many of them can precisely locate, e.g.
gay homosexual has been used from 1970s, in
the past 30 years. There is a mixed reaction to this
change. Some speakers say that they are now conscious of other senses of gay and are careful when
they use gay happy. There are older speakers
who seem to be upset about the fact that gay
happy has ceased to be commonly used, at the
expense of gay homosexual. These speakers
would often comment on the change, saying it is a
pity that gay changed its meaning, [the change] is
sad. It is a pretty word, [gay] has only one meaning.
And its sad. Weve lost a good word. Sometimes
they continue to express negative opinions on the
change of gay from happy to homosexual by
46

http://journals.cambridge.org

saying gay homosexual has taken over; I really


ght against the use of this word [as homosexual].
I object strongly to hijacking of the word, [gay] has
been hijacked.12 On the other hand, there are older
speakers who continue to use gay happy but they
use it only in appropriate contexts and situations,
depending on audience. For example, certain speakers say they use gay happy only with older people.
An elderly middle-class woman says With my
group of friends we say lets have a gay day, but
we know what gay means so we make a play on
the word. To summarise these comments, it is
apparent that for older speakers a common, nostalgic
narrative emerges. The English language used to
have a good, pretty13 word which then we lost
because this word was hijacked and this is a sad
thing. Certain speakers object strongly and even
ght against the use of this word [as homosexual],
whereas others still use it, but in a very conscious
manner.
For the vast majority of younger speakers, gay
happy has not been part of the usual usage. In
fact, the large majority of them only report on
gay happy. So while they seem to know that
gay can mean happy; they often discuss it in a
temporal frame, e.g. by saying the term phased
out, now it means, it used to mean, older
meaning. They often point to hearing this use
from older generations, e.g., my grandparents
say Gay old time, people over 50 use gay
happy, people 6080 use gay happy, my
grandma does. It is interesting to see that while
in the <1930> age group all of the uses of gay
happy are reported, in the youngest age group
there are speakers who use gay happy in an
unconstrained manner. One of the users of this
sense is the youngest participant, an 11-year-old
girl, who most likely uses the meaning that she
had heard from older family members. Apart
from orienting themselves towards adult norms,
younger speakers may also be inuenced by the
way gay is used in childrens literature. A quick
review of more recent childrens stories (available
via the OEC) indicates that gay happy is used
in this genre unexpectedly frequently.
Statistically distinct socio-demographic groups
who use gay happy or report on this sense emerge
only among the two older groups. While the use of
gay happy is mainly associated with older uneducated males, reporting on gay happy is most frequent among educated females. One could argue
that educated females report more frequently
because they are more sensitive to overt linguistic
norms than uneducated males (cf. Trudgill, 1994).
This nding can also provide insights into how
E N G L I S H TO D AY 1 1 2

Downloaded: 29 Jun 2016

D ec e mbe r 2 0 12
IP address: 192.41.131.252

semantic change happens. The apparent-time


interpretation of the observed variation suggests
that these females reporting on gay happy are
also the rst speakers in the community who
break from the uniform use of this sense. They
are the rst ones among their peers to start putting
constraints on their usage of gay happy. This process may be motivated by the emergence of the
newer gay homosexual. Their suppressing of the
older sense could be interpreted as the rst signs of
change happening. It arguably makes educated
females the most innovative speakers among their
peers, while males are the most conservative ones.
What now remains to be seen is how the new
sense gay homosexual is incorporated into the
speech community while gay happy disappears.
4.3 Gay homosexual

Although gay homosexual is the sense that is


used most frequently in the community as a
whole, it is the central meaning only for the two
middle generations (1930 and 3160). This is
unsurprising when one considers the historical contexts of the emergence of gay homosexual discussed in Section 1.2. It was not until the 1970s
that gay homosexual entered the mainstream discourse and adolescent speakers living at this time
would be the rst ones to be growing up with the
homosexual sense of gay being openly used. The
use of gay homosexual is still very frequent in
the fringe generations, but it competes with gay
lame for the youngest speakers and gay happy
for the oldest ones.
However widespread the use of gay homosexual is currently, this sense is marked for certain
speakers, usually the older speakers. To some
extent, reservations about its use are related to the
taboo sphere of sexuality that gay belongs to. For
example, during the interviews, a number of participants reacted with (what I interpreted as) slight
embarrassment when asked about gay. Sometimes
their responses were preceded by laughter, hesitation and laughter, or a long pause. Occasionally,
they talked for a while about the current and past
use of gay homosexual without explicitly referring to sexuality but saying these things, or
People are now gay in the way that they werent
before. One participant did not even mention gay
homosexual as one of the uses, which would not
be odd apart from the fact that he mentioned that
he hears children using gay lame. In addition, several participants mentioned gay happy in the rst
response to my questions, despite then admitting
that they normally would not use that sense. I interpreted these reactions as a way for these speakers to

manage their face when confronted with a taboo


sense in this semi-formal situation they found themselves in. They also may think of the earlier meaning as somehow more correct because it is older.
Other constraints emerge from the fact that the
social and legal acceptance of homosexual orientation in society does not always translate to individual acceptance of the homosexual usage of gay.
This is visible in the comments quoted in the previous section in which participants express dissatisfaction with the development of gay homosexual.
Several participants distance themselves from gay
homosexual to the extent that they deny using
this sense altogether. For example, a 79-year-old
lower-working-class female says:
I mean, this is a new word to me in the other sense of
being gay. We always said They look nice and gay,
clothes; and people who are happy. That is what I
think of this gay. I think they spoiled the word.
Wrong choice of a word, but I am a bit old fashioned.
Homosexuals used to be called queers, pansies. To
be quite honest a lot of these things I have learnt as I
got older. I knew nothing about it when I was young.
The other words become xed in my mind for certain
things.

These reported uses of gay homosexual came


up solely among the oldest generation of speakers,
which is actually an unsurprising nding if we consider this synchronic usage to represent a stage in
the diachronic development of gay (cf. the
apparent-time construct). It is during their adult
life that gay changed its meaning from a relatively
neutral sense happy to a taboo sense homosexual, and the evaluative reaction to this change
emerges in the form of reported usage.14 Tracing
the sociolinguistic contexts of the more conservative usage events allows us to pinpoint those sections of the community in which semantic change
is perhaps most resisted.
4.4 Gay lame

Tracing semantic change in apparent time also


allows us to track innovations in the language as
they are happening. One such innovation is the
use of gay as a general term of displeasure or disapproval, whose meaning, according to the participants of this survey, ranges from boring (in
reference to school, lessons, writing, reading,
friends, dad), and not interesting, to weird (in
reference to various politicians), annoying (in
reference to my sister), bad, rubbish, and stupid.15 It can also be heard in the phrase Stop
being gay, which means Stop annoying me.

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

47

IP address: 192.41.131.252

Although there is some concern in the literature


(Lalor & Rendle-Short 2007) and also in educational circles (e.g. in guides for teachers)16 that
this new sense of gay projects homophobic undertones, the data I obtained do not seem to support
this assumption in a straightforward way. The
types of referents that can be described as gay
refer to the key areas in a young persons life
such as school, family and friends, and it is not
obvious that all of those would evoke the notion of
homosexuality. Also, during and between individual
interviews at one secondary school, I occasionally
asked students whether the new uses of gay carry
any homophobic meaning, and they denied this.
This nding corroborates McCormacks (2011:
3478) observations that teenage boys in England
maintain that using gay as in That is so gay is not
homophobic and that the word gay in general is
never used in an aggressive manner as an insult.
Overall, these observations suggest that gay now
functions as a non-homophobic term of disapproval or an insult (see below) and can be read as
ironic, self-referential, habitual, or even deployed
without a knowing relation to gayness as a sexual
signier (Rasmussen, 2004: 304). Gay lame can
have various socio-psychological functions, including one of social bonding (McCormack, 2011: 349).
This awareness of the multiple instantiations of gay
in the speech of younger generations (including
references to homosexuality) can actually provide
the basis for more socio-culturally sensitive pedagogies, especially since recent evidence suggests that
cultural homophobia (Anderson, 2009) and actual
homophobic discourses are diminishing in English
secondary schools (McCormack, 2011).
Apart from age, signicant gender differences
also emerge in the use of gay lame. While this
sense most frequently occurs in the speech of
young men, who can arguably be considered innovators of this sense, young females who are still in
education tend to report on gay lame. These
women and girls distance themselves from using
gay lame by saying that it is an insult that is
used by 1316-year-old teenagers and that it
can often be heard in the speech of boys. These
perceptions reinforce the idea that gay lame
could have been innovated among teenage boys,
but perhaps not very young boys: there was one
12-year-old middle-class boy who said that he
heard gay being used at school, but he did not
really know what it meant.
When it comes to females, again we see a situation where they report on a pragmatically risky
sense that is undergoing a change (cf. discussion
of gay happy). In a way, this observation is
48

http://journals.cambridge.org

unsurprising in the context of a range of sociolinguistic studies which indicate that women tend to
be more sensitive to overt norms (cf. Trudgill,
1974). This sensitivity to pragmatically and also
stylistically risky uses can surface as (conscious)
distancing from these uses, and this is what happens via reporting in this case study.
4.4.1 Gay lame on the Internet
Although usage evidence from the OEC indicates
that gay lame is used rather infrequently in
British English, a quick look at the interview data
presents a different picture. The apparent-time
data indicates a much higher frequency of the use
of gay lame. Gay lame occurs mostly in the
language of the youngest age group (up to 18
years old). Eleven out of 18 of the speakers use
gay lame in 60% of their total responses (18 out
of 28). Five out of the remaining seven speakers
in this group report on knowing this meaning,
although they deny using it. So why does the corpus usage differ so much from the interview usage?
One possible explanation may be related to the
composition of the corpora. Any corpus is just a
sample of language used in a given period and
there is a chance that this sample might not be
very representative in some cases. However, after
probing this issue further, both in the OEC and
after using Google as a corpus, I discovered that
gay lame is usually spelled in on-line sources in
more cryptic ways. For example, gay can be spelled
as gh3y, g4y, jey, g@y, ghey. The idea behind this
cryptic spelling is that such words are not picked
up by taboo and sensitive word lters on internet
forums.17 By using advanced Google search
options, gh3y was searched on websites from the
United Kingdom, and a vast majority of the results
returned used gh3y in the sense lame. Many of
the examples come from gaming and computer
equipment forums, such as digitalspy.co.uk,
gameon.co.uk, euro-chimaera.co.uk, forums.multiplay.co.uk, and techangel.co.uk.
The real-time evidence obtained from the
Internet indicates that gay lame may be more frequently used than the OEC suggests. This nding
would corroborate the apparent-time usage data
obtained from the interviews.
4.5 Gay unmanly

While teenage boys use gay lame frequently, the


analysis of responses from young adult males indicates that they employ gay in a slightly different
way. These males use gay to describe activities,
such as eating salads, carrying an umbrella, or
E N G L I S H TO D AY 1 1 2

Downloaded: 29 Jun 2016

D ec e mbe r 2 0 12
IP address: 192.41.131.252

to describe looks, such as wearing two earrings.


These activities, according to participants, are gay
because they are effeminate, not masculine
(hence we use a reference tag of gay unmanly).
It is clear that gay unmanly projects connotations
of the speakers masculinity through distancing from
anything that can be seen as feminine or not masculine enough (cf. Pollack, 1998). Gay unmanly
demarcates the borders of masculinity and disapproves of the transgressing of those borders. It is
used and was probably innovated by males to challenge the masculinity of other men, and it is noticeable that in this survey gay unmanly does not
surface in the speech of women. However, one
could wonder to what extent gay unmanly builds
on the stereotypes of heterosexual masculinity and
even homosexuality. Previous studies suggested
that because homosexuality is stigmatised, men can
use homophobic discourse as a central mechanism
in the maintenance of their heterosexual identities
(Mac an Ghaill, 1994). So should the use of gay
unmanly perhaps be considered as an instantiation
of a homophobic discourse?
When analysing the change of the adjective gay
in apparent time (Figure 1) one can observe that the
unmanly sense of gay followed on from gay
homosexual. Considering that semantic change
typically involves some overlap between an older
and a newer meaning, we cannot exclude
the possibility that gay unmanly reects and
encourages homophobic attitudes in the community. However, the data obtained in the current
survey do not provide support for that usage.
Participants in the survey who used gay unmanly
clearly indicated that they do not think this sense is
used as a homophobic pejorative. Indeed, the referents that they describe as gay, such as carrying an
umbrella and eating salads, are not traditionally
discussed in the context of sexuality. What these
examples of the use of gay may project, though,
is behaviour that goes beyond traditionallyunderstood masculinity. These examples may
promote the idea of masculinity as based on the
culturally salient archetypes of local (here northern) toughness. This idea of masculinity can be
described as (regionally) localised masculinity.
For example, one can argue that while in the
south of England eating salads would project positive cultural connotations of healthy eating, in the
north it is often referred to with a more derogatory
phrase, rabbit food. References to carrying an
umbrella can also be interpreted in the context of
not being manly/tough enough to get on with
what are not unusual weather conditions in that
area.

4.6 Gay other

Several middle-class speakers (older participants in


the <3160> and <over 60> age groups) used gay
in reference to owers (nosegay), the girls name
Gay, or to a local coach company called Gay
Tours. Although these examples were lumped
together in the gay other category, their use is no
less important. The presence of these uses in the
older generations could actually be predicted from
the apparent-time hypothesis. Employing Gay as a
name could only be a pragmatically safe choice in
historical periods and for speakers whose major
senses of gay are also happy, bright, cheerful,
etc. And this is precisely what emerges in the variationist analysis of the adjective gay. Although the
use of gay other is infrequent, the applied method
located it and showed that it sits in the predictable
place on the demographic map of gay.

5. Discussion: semantic change


through a sociolinguistic lens
The sociolinguistic analysis of the semantic variation in the adjective gay provides a number of
insights into why different speakers in the same
community use different senses of the same polysemous word. The key nding of this analysis is
that the usage of different senses of the adjective
gay is not random, but relates to the sociodemographic characteristics of speakers (cf.
Robinson, 2010b, 2012). This non-randomness of
usage can be explained in terms of the experiential
nature of meaning which surfaces at the sociodemographic level in the form of varying frequencies of the use of individual senses of gay (cf. Ptz
et al., 2012, Reif et al., 2012).
The investigation of the semantic variation in
gay also yields a range of observations that
enhance our understanding of how semantic
change happens. This is carried out by employing
the apparent-time construct, which assumes that
generational differences in linguistic use reect
the actual diachronic developments in language.
The key nding that the older generations retain
older meaning variants of gay, while younger generations innovate, is not surprising. However, what
this analysis has additionally shown is that an
observation of more subtle usage differences
helps to point to areas of linguistic innovation
and conservatism in that speech community.
These areas of conservatism/innovation surface in
relation to such factors as gender, social class, the
education of speakers, the conscious constraining
of usage contexts, attitudes to change, attitudes to

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

49

IP address: 192.41.131.252

the usage of other speakers, perceptions of change


in terms of timing, perceptions of innovative and
conservative groups, and nally the actual referential meaning of a particular variant and speakers
(subconscious) assessment of the pragmatic need
for using that variant. In order to contextualise
these observations, in this section I summarise
the key ndings of how the semantic change of
gay happens. I rst focus on gender, age, and
socio-economic status, the classic sociolinguistic
categories for learning about language change.
Among the older generations, the happy uses
of gay are retained among speakers belonging to
lower socio-economic classes, especially males.
These speakers are less mobile in sociogeographic terms and probably belong to more
established, stronger networks which are not
easily penetrable by more innovative uses of
language (cf. Milroy and Milroy, 1992). On the
other hand, more educated speakers (especially
females) are most clearly distancing themselves
from the older senses and doing what I have
referred to as reporting on gay happy. However,
there is no clear evidence that these speakers are
also the ones leading in the use of a newer sense
(gay homosexual).
Among the younger generations, males, rather
than females, are innovating gay unmanly and
gay lame. Why might this be the case? Let us
look again at the demographic distribution of
senses in Figure 1. Mens conservative and innovative uses of gay are present in generations that are at
both ends of the <3160> age group, which is the
generation representing the highest frequency of
gay homosexual. Perhaps, the increased rates of
retention or innovation of non-homosexual senses
of gay (happy, unmanly, lame) could be read
as a way for these speakers to react to, draw
upon, or create distance from the notion of homosexuality (cf. Anderson, 2009).
The current study also provides empirical evidence for the gradual nature of semantic change,

and argues for a more ne-grained analysis of


polysemy in investigating semantic change. Data
show that many changes that seem to be abrupt
are in fact preceded by a phase of overlapping
uses (cf. Nerlich & Clarke, 1992), with newer
meanings often retaining elements of older senses
(cf. Bybee and Pagliucas (1987) retention or
Hoppers (1991) persistence). For instance, the
development of gay homosexual to gay lame
has been accompanied by the emergence of gay
unmanly. Figure 3 illustrates how this change,
which emerged at the demographic level, could
have taken place on a conceptual level. The change
from gay homosexual to gay unmanly was
marked by the shift of focus from homosexuality
to masculinity. This stage was probably preceded
by the development of negative attitudes or even
homophobic uses of gay (marked with a question
mark) which I have not accounted for in my survey.
The indication of disapproval is the key conceptual element for further development of gay
unmanly and lame. By the time gay lame
develops, connotations of sexuality are lost and
gay remains to be used as a negative evaluative
adjective. Since gay lame has a considerable
pragmatic impact, it is probable that in time this
negative intensier will continue to widen its range
of contexts. However, judging by the fate of other
frequently used intensiers, gay lame may start losing the pragmatic impact it has now (cf. Macaulay,
2006; Jing-Schmidt, 2007). Alternatively, gay
lame may start being used to mean the opposite
(as awesome, wicked, sick have), which would certainly be an interesting development to witness!
The apparent-time analysis of recent changes of
gay allows for less frequent uses of this adjective
to emerge in between gay lame and gay homosexual. An awareness of the existence of these bridging senses helps to explain how semantic change
from homosexual to lame happened at the socioconceptual level. These ndings also show that the
emergence of gay lame did not necessarily have

Figure 3. The chain of semantic change from gay homosexual to gay lame

50

http://journals.cambridge.org

E N G L I S H TO D AY 1 1 2
Downloaded: 29 Jun 2016

D ec e mbe r 2 0 12
IP address: 192.41.131.252

its roots in gay or even homophobic discourse, as at


the diachronic conceptual level gay lame and gay
homosexual do not seem to be directly related,
as they are mediated by other sense(s).
A wealth of information on how semantic
change happens is available in the metalinguistic
comments provided by participants of the survey.
The analysis of reported uses of sense variants
is particularly useful for learning about the types
of meaning change that are happening as well as
learning about the direction of change. Another
interesting observation comes from examining the
demographic distribution of reported senses presented in Figure 1. Reported senses appear in
those places in the socio-demographic structure
which are sensible or even predictable from the
point of view of the historical development of the
word. Reporting on a change seems to be demographically adjacent to the actual change going
on. Thus, one can observe that as gay starts being
used to mean homosexual, the oldest generation
reacts by reporting on it. They often refer to gay
homosexual as the new meaning of gay. On
the other hand, the majority of these speakers are
not aware of the newer meaning lame and certainly none of them report on gay lame. The
analysis of the reporting on usage further supports
the idea of the social grounding of semantic
change. There needs to be some contact, e.g. at
the demographic level, between speakers for the
change that is taking place to be noticed and commented upon. A nal observation here is that overall females comment upon pragmatic or stylistic
restrictions of the usage of gay signicantly more
frequently than males do. This observation can
lead to the conclusion that females in this speech
community exhibit an increased awareness/responsiveness to overt sociolinguistic norms. This
nding is unsurprising if one considers a range of
sociolinguistic studies carried out at other levels
of language and in different communities which
corroborate these conclusions (cf. Coates, 2004).18
Finnegans (2011) recent study carried out among
speakers from Shefeld also demonstrates that
females are leading phonological change in that
city.
Apart from reporting on the contexts of use of
the adjective gay, which then help in locating the
type and the direction of change, participants provided a range of metalinguistic comments that
allow us to situate the change in time and social
space. Thus, speakers point to particular sociodemographic groups that use gay in a characteristic
way, and also provide comments on the timing of
the change. For instance, speakers of all ages

comment on the timing of the key changes gay


has gone through during their life. For example, a
young middle-class teenager notices that he heard
gay lame for the rst time ve years ago, while
another teenager says that gay lame is used by
year 7 and 8 teenagers. Older participants mainly
comment on the change from gay happy to homosexual, e.g. gay homosexual is used from 1970s,
gay changed its meaning in the past 30 years.
These perceptions correspond to real-time information on the change of gay. This very accurate
reporting may be related to the fact that this adjective
is exceptionally salient in that speech community as
it evokes sexual taboo, prejudices, stereotypes, politically shaped usage, etc.
Metalinguistic comments also indicate that the
evaluative load that accompanies the usage of
gay also leads speakers to consciously alter the
way they use gay. In the analysis of gay happy
and gay homosexual we have seen that some
people change the way they speak during their lifetime by adopting new senses (gay homosexual)
or reducing the use of older senses (e.g. gay
happy), sometimes concurrently. At rst sight,
this goes against the apparent-time hypothesis,
which largely assumes the stability of vernaculars.
Here, I would like to suggest that changes in individual semantic usage provide evidence to further
strengthen the validity of the apparent-time
hypothesis (cf. Boberg, 2004). Older speakers
who adopt innovative variants actually contribute
to the acceleration of semantic change. In the
same way, shedding older variants can accelerate
the disappearance of some uses. This age-graded
behaviour can explain why semantic change can
sometimes be rather abrupt and why cases of a
change to an opposite meaning (such as gay
happy to lame) can occur within just a couple
of generations.

6. Conclusions
Sociolinguistics has rarely investigated semantics,
and semantics has rarely considered sociolinguistic
insights into the variation and change in meaning.
In this paper, I have challenged this research custom by placing one of the most iconic examples of
semantic change onto the socio-demographic map.
By employing the apparent-time construct I was
able to trace a semantic change as it happens, and
take account of all sorts of evaluative comments
and perceptions accompanying the change. The
conclusions of this study contribute to analyses
of an already well-investigated word, by providing
insights into how semantic change emerges from

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

51

IP address: 192.41.131.252

the interaction of individual and community usage.


The key nding of this study is that meaning variation and change is a complex socio-conceptual
phenomenon, and therefore semantics and sociolinguistics can and should inform one another in
future research.
Notes

1 I would like to thank Kathryn Allan, Joan C. Beal,


and M. Lynne Murphy for providing generous comments on previous versions of the paper.
2 For a more comprehensive overview of the etymology and history of gay see Butters (1998), Coleman
(2012), Hughes (2009), and the etymology note for
gay in the OED On-line.
3 Although sometimes it can be spelled as as gh3y,
g4y, jey, g@y, ghey (see more discussion of this issue
in Section 4.4.1).
4 See http://www.homophobiaisgay.com/
5 Gay is recorded as 'hard, difcult' in New Zealand
Playground Language Project (19992002) (http://
www.victoria.ac.nz/lals/research/Playground/index.
aspx). Lalor & Rendle-Short (2007) even split the usage
of what they call negative uses of gay into groups:
bad/lame/pathetic/stupid/boring and weird/different/
strange/unconventional. Also, see a discussion on
LinguistList on the varied usage of gay lame
(Milligan, 2002).
6 The decision to use selected registers was motivated
by the desire to focus on texts that reect more spontaneous speech, since these are a likely source for
slang, regionalisms, and neologisms (cf. Beeching,
2005: 172).
7 Within the same age group, the immediate context
will also point to one meaning over another. There is
some suggestion that gender may play a role.
8 The Kruskal-Wallis statistics compare two or more
groups of cases on one variable. From the
Kruskal-Wallis one-way analysis of variance, one
might learn whether the four age groups of speakers differ in the average use of different senses of the adjective
gay. The tests were run via SPSS18.
9 The reduced method was used, the summary of
which is available via http://www.ons.gov.uk/aboutstatistics/classications/current/ns-sec/deriving/
reduced-method.pdf.
10 For further details of the classication, see
Robinson (2010b).
11 For the Kruskal-Wallis tests I partitioned data
according to the socio-demographic groups of speakers
and carried out the statistical test on these groups. A
more efcient way of carrying out a multifactorial analysis is to let the statistical test choose a signicant
predictor for the usage of gay. Thus, I have also carried
out decision tree analysis and a logistic regression analysis. These inferential statistics tests yielded fewer signicant results than the descriptive statistics of the
Kruskal-Wallis test, which is unsurprising. These results
are marked with an asterisk placed next to the p-value.

52

http://journals.cambridge.org

12 Similar attitudes to gay homosexual are recorded


in Burcheld (1996: 335); for example I have deplored
for years the homosexual adoption of the word; Can
we have our word back, please?
13 Cf. Butters (1998: 196) who quotes Prescott saying in
Newsweek (27 March 1969) A series of quotations cant
show the damage that a current usage can do: how gay a
delightful and necessary word is now virtually lost.
14 Refer to similar comments and attitudes to gay
homosexual presented in Burcheld (1996).
15 This multiplicity of meanings associated with the
recent uses of gay is also reported in other studies,
such as Lalor & Rendle-Short (2007), McCormack
(2011).
16 Cf. http://www.stonewall.org.uk/at_school/education_for_all/quick_links/education_resources/4003.asp
17 Spelling that involves a combination of letters and
certain numbers often indicates the use of internet
language and is called leet. Leet (from the word elite)
was originally used by hackers to encrypt words so
that they would not be picked out by keyword searches.
Currently, leet is frequently used by virtual communities, especially by gamers (see Blashki and Nichol,
2005 or Squire, 2002).
18 There are studies indicating that males also lead
change, e.g. see Cheshire et al. (2008) for account of
recent change in Multicultural London English.

References
Aaron, J. E. 2010. An awkward companion: disability and
the semantic landscape of English lame. Journal of
English Linguistics 38 (1), 2555.
Algeo, J. 1990. Semantic change. In E. C. Polom (ed.),
Research Guide on Language Change. Berlin: Mouton de
Gruyter, pp. 399408.
Allan, K. & Robinson, J. A. (eds.), 2012. Current Methods in
Historical Semantics. Berlin/ Boston: Mouton de Gruyter.
Anderson, E. 2009. Inclusive Masculinity: The Changing
Nature of Masculinities. London: Routledge.
Bailey, G. 2002. Real and apparent time. In J.
K. Chambers, P. Trudgill & N. Schilling-Estes (eds.), The
Handbook of Variation and Change. Oxford: Blackwell,
pp. 31232.
Beeching, K. 2005. Politeness-induced semantic change.
The case of quand mme. Language Variation and
Change 17, 15580.
Bierwisch, M. 1989. The semantics of gradation. In
M. Bierwisch & E. Lang (eds.), Dimensional Adjectives.
Berlin: Springer-Verlag, pp. 71261.
Blashki, K. & Nichol, S. 2005. Game geeks goss:
Linguistic creativity in young males within an online
university forum. Australian Journal of Emerging
Technologies and Society 3, 7786.
Boberg, C. 2004. Real and apparent time in language
change: late adoption of changes in Montreal English.
American Speech 79, 25069.
British National Corpus, Version 3 (BNC XML Edition)
(2007). Online at <http://www.sketchengine.co.uk>
(Accessed January 1, 2009).
Burcheld, R. W. (ed.) 1996. The New Fowlers Modern
English Usage. Oxford: Oxford University Press.

E N G L I S H TO D AY 1 1 2
Downloaded: 29 Jun 2016

D ec e mbe r 2 0 12
IP address: 192.41.131.252

Burridge, K. 2004. Weeds in the Garden of Words. Further


Observations on the Tangled History of the English
Language. Cambridge: Cambridge University Press.
Butters, R. R. 1998. Cary Grant and the emergence of gay
homosexual. Journal of the Dictionary Society of North
America 19, 188204.
Bybee, J. & Pagliuca, W. 1987. The evolution of future
meaning. In A. Giacalone-Ramat, O. Carruba & G.
Bernini (eds.), Papers from the 7th International
Conference on Historical Linguistics (ICHL). Amsterdam:
John Benjamins, pp. 10922.
Campbell, L. 2004. Historical Linguistics: An Introduction.
Second edition. Edinburgh: Edinburgh University Press.
Carter, R., Goddard, A., Reah, D., Sanger, K. & Bowring, M.
2001. Working with Texts: A Core Introduction to
Language Analysis, Second ed. London: Routledge.
Cherry, R.L 1986. English Words from Latin and Greek
Elements. Arizona University Press.
Cheshire, J. 2007. Discourse variation, grammaticalisation
and stuff like that. Journal of Sociolinguistics 11,
15593.
. Fox, S., Kerswill, P. & Torgersen, E. 2008. Ethnicity,
friendship network and social practices as the motor of
dialect change. Linguistic innovation in London.
Sociolinguistics 22, 123.
Coates, J. 2004. Women, Men and Language. 3rd edn.
London: Routledge.
Coleman, J. 2012. The Life of Slang. Oxford: Oxford
University Press.
Cook, P. & Hirst, G. 2011. Automatic identication of
words with novel but infrequent senses. In H. H. Gao &
M. Dong (eds.), Proceedings of the 25th Pacic Asia
Conference on Language Information and Computation
(PACLIC 25), Singapore, pp. 26574.
Crowley, T. 1997. An Introduction to Historical Linguistics.
Third edition. Oxford: Oxford University Press.
Culpeper, J., Katamba, F., Kerswill, P., Wodak, R. &
McEnery, T. (eds) 2009. English Language: Description,
Variation and Context. New York: Palgrave Macmillan.
Curzan, A. 2003. Gender Shifts in the History of English.
Cambridge: Cambridge University Press.
Cymbalista, P. 2009. Cleopatra wore a bracelet gay How
words change their meanings? In G. Kleparski,
R. Kietyka & E. Wicawska-Szymaska (eds), Mielec
Anglistentag. Mielec, Rzeszw: NKJO in Mielec and
Rzeszw University, pp. 2037.
Davies, M. 2010. The Corpus of Contemporary American
English as the rst reliable monitor of English. Literary
and Linguistic Computing 25 (4), 44764.
Eckert, P. 1998. Gender and sociolinguistic variation. In
J. Coates (ed.), Language and Gender: A Reader. Oxford:
Blackwell, pp. 6475.
. 2008. Variation and the indexical eld. Journal of
Sociolinguistics 12, 45376.
. 2012. Three waves of variation study: the emergence of
meaning in the study of sociolinguistic variation. Annual
Review of Anthropology 41, 81100.
Finnegan, K. 2011. Linguistic variation, stability and
change in middle-class Shefeld English. Unpublished
PhD dissertation. The University of Shefeld.
Frawley, W. 2003. Oxford International Encyclopedia of
Linguistics (Vols 14). New York: Oxford University
Press.

Geeraerts, D. 2010. Theories of Lexical Semantics. Oxford:


Oxford University Press.
Glynn, D. & Robinson, J. A. (eds) Forthcoming. Polysemy
and Synonymy. Corpus Methods and Applications in
Cognitive Semantics. Amsterdam: John Benjamins.
Gramley, S. & Ptzold, K.-M. 2004. A Survey of Modern
English. London: Routledge.
Hasan, R. 1989. Semantic variation and sociolinguistics.
Australian Journal of Linguistics 9, 22175.
. 1992. Meaning in sociolinguistic theory. In
K. Bolton & H. Kwok (eds), Sociolinguistics Today:
International Perspectives. London: Routledge,
pp. 80119.
. (ed.) 2009. Semantic Variation: Meaning in Society and
in Sociolinguistics. The Collected Works of Ruqaiya
Hasan. London: Equinox.
Hockett, C. F. 1954. Two Models of Grammatical
Description. Indianapolis: Bobbs-Merrill.
Hopper, P. J. 1991. On some principles of
grammaticization. In E. C. Traugott & B. Heine (eds),
Approaches to Grammaticalization. Amsterdam: John
Benjamins, pp. 1735.
Hughes, G. 1991. Swearing. Oxford: Blackwell.
. 2009. Political Correctness: A History of Semantics and
Culture. Maldon, Mass.: Wiley-Blackwell.
Jing-Schmidt, Z. 2007. Negativity bias in language: a
cognitive-affective model of emotive intensiers.
Cognitive Linguistics 18 (3), 41743.
Kay, C. J. & Allan, K. Forthcoming. English Historical
Semantics. Edinburgh: Edinburgh University Press.
Kerswill, P. 2006. Socio-economic class. In C. Llamas &
P. Stockwell (eds), The Routledge Companion to
Sociolinguistics. London: Routledge, pp. 5161.
Labov, W. 1963. The social motivations of a sound change.
Word 19, 273309.
. 1966. The Social Stratication of English in New York
City. Washington: Center for Applied Linguistics.
. 1972. The social motivation of a sound change. In
W. Labov (ed.), Sociolinguistic Patterns. Philadelphia:
University of Pennsylvania Press, pp. 142.
. 1978. Where Does the Linguistic Variable Stop? A
Response to Beatriz Lavandera. Austin, TX: Southwest
Educational Development Laboratory.
Lalor, T. & Rendle-Short, J. 2007. Thats so gay: a
contemporary use of gay in Australian English.
Australian Journal of Linguistics, 27 (2), 14773.
Lavandera, B. 1978. Where does the sociolinguistic variable
stop? Language in Society 7, 17183.
Lehrer, A. 1992. A theory of vocabulary structure:
retrospectives and prospectives. In M. Ptz (ed.), Thirty
Years of Linguistic Evolution: Studies in Honour of Ren
Dirven on the occasion of his sixtieth birthday.
Amsterdam/Philadelphia: John Benjamins, pp. 24356.
Leith, D. 1997. A Social History of English. London:
Routledge.
Llamas, C. 1999. A new methodology: data elicitation for
social and regional variation studies. Leeds Working
Papers in Phonetics and Linguistics 7, 95118.
Mac an Ghaill, M. 1994. The Making of Men. Buckingham:
Open University Press.
Macaulay, R. K. S. (2005). Talk that Counts: Age, Gender,
and Social Class Differences in Discourse. New York and
Oxford: Oxford University Press.

A GAY PA PER: WHY S HOUL D SOCIOL INGU ISTICS BOTHER WITH S EMANTICS ?

http://journals.cambridge.org

Downloaded: 29 Jun 2016

53

IP address: 192.41.131.252

. 2006. Pure grammaticalization: the development of a


teenage intensier. Language Variation and Change 18,
26783.
Malkiel, Y. 1993. Etymology. Cambridge: Cambridge
University Press.
McCormack, M. 2011. The declining signicance of
homohysteria for male students in three sixth forms in the
south of England. British Educational Research Journal,
37 (2), 33753.
McMahon, A. M. S. 1994. Understanding Language
Change. Cambridge: Cambridge University Press.
Meyerhoff, M. 2011. Introducing Sociolinguistics. Second
edition. London: Routledge.
Milligan, K. (ed.) 2002. New uses of gay. LINGUIST
List 13, 23 February. Online at http://linguistlist.org/
issues/13/13-498.html (Accessed July 1, 2012).
Millward, C. M. & Hayes, M. 2010. A Biography of the
English Language, Third edition. Cengage Learning.
Milroy, L. & Milroy, J. 1992. Social network and social
class. Toward an integrated model. Language in Society 21,
126.
Nerlich, B. & Clarke, D. D. 1992. Semantic change: case
studies based on traditional and cognitive semantics.
Journal of Literary Semantics 21, 20425.
Nevalainen, T. & Raumolin-Brunberg, H. 2003. Historical
Sociolinguistics: Language Change in Tudor and Stuart
England. London: Longman.
OGrady, W., Archibald, J., Aronoff, M. & Rees-Miller, J.
(eds) 2005. Contemporary Linguistics: An Introduction.
Fifth edition. Boston: Bedford / St Martins.
Ofce for National Statistics 2008. Online at http://www.
statistics.gov.uk/pdfdir/fng0908.pdf (Accessed January 1,
2010).
Ofce for National Statistics: Standard Occupational
Classication 2000, Volume 2 Coding Index 2000.
London: The Stationery Ofce.
Oxford English Corpus 2008. Online at http://www.
sketchengine.co.uk (Accessed January 1, 2009).
Oxford English Dictionary (OED) Online 2000. Online at
http://dictionary.oed.com (Accessed June 1, 2012).
Pollack, W. 1998. Real Boys. New York: Henry Holt and
Company.
Ptz, M., Robinson, J. & Reif, M. (eds) 2012. Cognitive
sociolinguistics: social and cultural variation in cognition
and language use. Special issue of Review of Cognitive
Linguistics 10 (2).
Radwanska-Williams, J. 1990. Expression and
communication as basic linguistic functions. Intercultural
Communication Studies 3(1), 91100.

54

http://journals.cambridge.org

Rasmussen, M. L. 2004. Thats so gay!: a study of the


deployment of signiers of sexual and gender identity in
secondary school settings in Australia and the United
States. Social Semiotics, 14 (3), 289308.
Reif, M., Robinson, J. A. & Ptz, M. (eds) 2012. Variation in
Language and Language Use: Sociolinguistic,
Socio-cultural and Cognitive Perspectives. Frankfurt am
Main: Peter Lang.
Robinson, J. A. 2010a. Awesome insights into semantic
variation. In D. Geeraerts, G. Kristiansen & Y. Piersman
(eds), Advances in Cognitive Sociolinguistics. Berlin:
Mouton de Gruyter, pp. 85110.
. 2010b. Semantic variation and change in present-day
English. Unpublished PhD Dissertation. The University
of Shefeld. Available via http://etheses.whiterose.ac.uk/
2232/
. 2012. A sociolinguistic perspective on semantic
change. In K. Allan & J. A. Robinson (eds), Current
Methods in Historical Semantics. Berlin/ Boston: Mouton
de Gruyter, pp. 191231.
Sankoff, G. 2006. Apparent time and real time. In
K. Brown (ed.), Elsevier Encyclopedia of Language and
Linguistics. Amsterdam: Elsevier, pp. 11016.
Squire, K. 2000. Cultural framing of computer/video
games. Games Studies 2, 111.
Stenstrm, A.-B. 2000. Its enough funny, man: intensiers
in teenage talk. In J. M. Kirk (ed.), Corpora Galore:
Analyses and Techniques in Describing English.
Amsterdam: Rodopi, pp. 17790.
Trudgill, P. 1974. The Social Differentiation of English in
Norwich. Cambridge: Cambridge University Press.
Weinreich, U., Labov, W. & Herzog, M. I. 1968. Empirical
foundations for a theory of language change. In W.
P. Lehmann & Y. Malkiel (eds), Directions for Historical
Linguistics: A Symposium. Austin: University of Texas
Press, pp. 97195.
Wijaya, D. T. & Yeniterzi, R. 2011. Understanding semantic
change of words over centuries. Proceedings of the 2011
International Workshop on Detecting and Exploiting
Cultural Diversity on the Social Web. New York: ACM,
pp. 3540.
Wong, A. D. 2002. The semantic derogation of tongzhi: a
synchronic perspective. In K. Campbell-Kibler,
R. Podesva, S. J. Roberts & A. D. Wong (eds), Language
and Sexuality: Contesting Meaning in Theory and
Practice. Stanford, California: CSLI, pp. 16174.
. 2006. Tongzhi, ideologies, and semantic change.
Proceedings from the Annual Meeting of the Chicago
Linguistic Society 42, 30317.

E N G L I S H TO D AY 1 1 2
Downloaded: 29 Jun 2016

D ec e mbe r 2 0 12
IP address: 192.41.131.252

Das könnte Ihnen auch gefallen