Language landscapes of children in remote Australia

Language landscapes of children in remote Australia
Jane Simpson
Department of Linguistics, The University of Sydney
Transient Building F12, Sydney, 2006, New South Wales
Tel +61 2 9351 3655
[jane.simpson@usyd.edu.au]

Abstract
Many Indigenous communities in remote Australia are multilingual, and often the languages being spoken in the community are
rapidly changing. Traditional languages are spoken by some people, but at the same time new languages are being developed based
on the interaction of traditional languages, English and an English-based creole. These new languages vary along a continuum. At
one end, the way of talking is close to the way many people in rural Australia talk. At the other end are mixed languages, in which
the structure of the new language contains words and features of several languages. In the middle of the range are varieties of an
English-based creole. Children in these multilingual communities grow up in language landscapes that are often undergoing rapid
change. In this paper I report on some findings from the Aboriginal Child Language Acquisition Project, a longitudinal study
recording interactions between pre-school children and their care-givers in some remote Indigenous communities. I consider the
findings about language shift in the light of national Census data on the demography of remote Indigenous Australia. I also consider
the implications of government policy towards Indigenous Australians for language shift.


Introduction
Children in four communities
Michael Clyne has said that linguists and language
In this section I describe what is happening with respect
professionals have a social responsibility to advise
to children's language landscapes in some remote
people and governments about the advantages of having
Indigenous communities which we studied in the
a multilingual mindset:
Aboriginal Child Language Acquisition project

(Simpson and Wigglesworth, in prep.). The aim of this
“Helping to make Australia a more language-aware
project was to study how people were talking to
society freed of a complacently monolingual
children and what children were saying back to them.
mindset is one of the many exciting tasks
We worked mainly with three communities in the North
confronting Australian linguists today.” (Clyne,
of Australia, with comparisons to a fourth. Samantha
2006)
Disbray, Felicity Meakins, and Karin Moses visited one

community each twice a year for three years and
The few Indigenous languages that are still spoken by
worked with community researchers to collect and
children are at risk from the monolingual mindset of
analyse data from 8 families focussed on a young child
Australia, from the imperatives of demography and
in each family, who were aged between 18 months and
traditional practices of child raising, and from
2 at the beginning of the project. We followed their
government policies. In this paper I discuss some of the
language development over three years, recording data
causes of language endangerment in Australia. I
in school-type settings, home settings and on bush trips.
describe how the language landscapes for Indigenous
The three communities are Kalkaringi, which is in
children in northern Australia have changed over the
Gurindji country (Felicity Meakins and Samantha
last ten years, starting with some examples of language
Smiler Nangala-Nanaku), Tennant Creek which is on
use and language change in four remote Aboriginal
Warumungu country (Samantha Disbray and Betty
communities. I then show how the language landscapes
Nakkamarra Morrison), and Yakanarra, which is on
have changed over the last ten years, using Census data.
Walmajarri country (Karin Moses). We also made
Here I consider traditional Indigenous child raising
comparisons with Lajamanu, a Warlpiri speaking
practices. Finally I look at the likely effects on
community, where Carmel O'Shannessy has been
Indigenous languages of the changes in Government
working.
policy in 2007 which are aimed at encouraging
Indigenous people to leave remote communities for
In all four communities some people speak the
towns and itinerant work. I conclude that the language
traditional languages, but the number of speakers varies.
landscapes are likely to change considerably and quite
(Figure 1).
rapidly, as more people shift away from speaking
traditional Indigenous languages.

---

3,500
3,000
2,500
2,000
1996
2001
1,500
2006
Number of speakers 1,000
500
0
Warlpiri
Warumungu
Walmajarri
Gurindji
Census years
[No figures in 1996 for Warumungu]
Source: Australian Bureau of Statistics1, and National Indigenous Languages Survey

Figure 1: Traditional Indigenous languages in project communities: total speaker numbers 1996 - 2006


At school, children in these communities are taught in
The children in our study have loving caregivers, and
standard English, mostly by non-Indigenous teachers.
large networks of siblings and cousins who play with
At home, the families and children speak in a range of
them, and take great responsibility for younger children.
ways, from the way many country people talk, rural
Many of the children spend times with grandmothers
non-standard English, to varieties that sound like the
and great grandmothers. In Kalkaringi the most
well-known Ngukurr Kriol, to mixed languages
common family grouping is family living with
probably formed by code-switching an English-based
grandmother who takes care of the children when
creole with traditional language, e.g. Gurindji Kriol.
mothers are working or studying, In Tennant Creek

children spend time with grandmothers and great
At Lajamanu, Carmel O'Shannessy has shown that
grandmothers when the mothers are working or
people talk to children in Warlpiri and in a new mixed
studying. In Yakanarra there is no common family
language Light Warlpiri (O'Shannessy, 2006). Children
grouping; some people live in nuclear families
normally talk Light Warlpiri, but can talk Warlpiri, and
(although there is a lot of interaction between families)
this has been supported by school Warlpiri language
and some people live in extended family groups. As
programmes. It seems likely that intervention by
Musharbash points out for Warlpiri children, they learn
Lajamanu people together with a dedicated principal
early the importance of family relations and the need to
and teacher linguist in the late 1980s may have halted
keep contact with them (Musharbash, 2001).
the shift. None of the other three communities have had








1 Cat. No. 2068.0 - 2006 Census Tables, 2006 Census of Population and Housing, Australia (Australia), Language Spoken At Home
(Australian Indigenous Languages Only) By Sex, Count of persons, Based on place of usual residence.

---

the same level of resourcing on language work in
The word order in Gurindji Kriol appears to be
schools.
influenced by information structure requirements which

are likely to match those of the traditional language,
At Kalkaringi a new mixed language, Gurindji Kriol,
Gurindji. Consider the following example, in which
has developed (Meakins and O'Shannessy, submitted),
expressions translating 'a lot of fish' occur in different
whose sources are Gurindji and Kriol. It has auxiliaries
orders.
from Kriol, verbs from Kriol and from Gurindji coverbs,

nouns from both languages, and uses both prepositions
(2)
Pretending to fish.
and case suffixes (Charola, 2002; McConvell and
*CHI: bigija yawu dij mob.
Meakins, 2005). Here is an example of a mother-child
There's a lot of fish
conversation at Kalkaringi:


*GRAN: ma garra big mob wayi yawu.
(1)
MO ca 21, CHI ca 4:
Is there a lot of fish?
MO:
dat guana garra kom gedim yu baitim yu-mob
ma big mob yawu yu mob garram
That goanna's gonna come and get you and bite you all
You've got a lot of fish
i garra kom rarraj dijei nyawa kankula.
big mob yu garram yawuyawu wayi?
It'll come running this way this one above
Have you got a lot of fish?
i garra baitim yu-mob binij
yu garram hiya jarrwa ma.
It'll really bite you all.
You've got many here.
[FM044]


CHI:
i-l be katurl im inti Mam
There is considerable variability as to word order, but
It'll really bite won't it Mum?
initial position is important for emphasising

information. This use of initial position is a property of
MO:
hmm yu-rra katurl im.
the traditional language Gurindji, and of many other
Hmm you'll bite it.
Australian languages, and has been taken over into the

new mixed language.
CHI:
ai-rra katim nyanawu xxx knife-jawung.

I'll cut this thing xxx with my knife. [FM041.C]2
In sum what Meakins found at Kalkaringi is that people

mostly talk to children in Gurindji Kriol, a mixed
The mother starts by using only words shared between
language, and that children mostly talk in Gurindji
Kriol and Gurindji Kriol, but then introduces words
Kriol. Children may hear older adults using Gurindji
shared with Gurindji rarraj (running), nyawa (pronoun)
amongst themselves, and have some understanding of
and kankula (above). Her next utterance has no Gurindji
Gurindji. (Meakins and O'Shannessy, submitted).
words and uses the Kriol verb baitim. The child repeats

the idea, switching to English be, and then a Gurindji
In Tennant Creek, Samantha Disbray found that
word katurl for 'biting'. The mother responds using
children hear varieties of Wumpurrarni English (WE) (a
katurl but switching the subject to address the child with
name some people give to the creole that is used in
the Kriol pronoun yu. The child then utters a sentence
Tennant Creek), and/or English, with a little
with a Gurindji pronoun nyanawu, an English noun
Warumungu and other traditional languages (TIL)
followed by a Gurindji case-ending jawung meaning
(Morrison and Disbray, 2007). Children mostly speak in
'with'. The child's deliberate choice of a partial synonym
WE, but can switch (e.g. when pretending to be doctors)
shared with Gurindji but not Kriol, for words the mother
to speaking close to standard English. (3) shows an
uses which are shared with Kriol shows that the child
adult talking WE.
has mastery of the synonyms, and, at least on this

occasion, is experimenting with the Gurindji source
(3)
ADULT:

form. The mother listened to this sequence and noticed
it no gud fo yu ulkuman,
that the child used the Gurindji word for 'bite' where she
It’s no good for you, old lady
used the Kriol word. She suggested to Felicity Meakins
yu mungku no gud,
that this might have been because the child was
Your stomach’s no good.
spending more time with her grandmothers.
wangu mungku yu gatim.

Bad stomach you have
[SD104C]



There is some use of Warumungu nouns, wangu and

mungku. It is much rarer to find people using

Warumungu verbs or Warumungu words expressing

actions or feelings. Notice that the object of the sentence

wangu mungku 'bad stomach' is put at the start of the

sentence. Items which are emphasised may be placed
2 Codes such as FM041.C refer to the recording and transcript
initially in sentences as in Gurindji Kriol and in
made in the ALCA project, ultimately to be archived at the
Warumungu, but there is considerable variation.
Australian Institute of Aboriginal and Torrest Strait Islander

Studies..

---

In Yakanarra, Karin Moses found that, while people are
when they have children, they will be unlikely to pass
multilingual and shift with varying degrees of facility
on the traditional language to those children.
between Walmajarri, Kimberly Kriol and English,

people mostly talk to children in Kriol and English and
The shift from traditional languages follows roughly the
a limited number of Walmajarri words. Language
same pattern. First, the words for actions and feelings
directed to children is determined by situational factors
go, along with the auxiliary system if there is one (or
including location, purpose, participants and language
else, as in Lajamanu Light Warlpiri, the Kriol pronoun
skills, and age of the interlocutor. Children normally
system is pressed into service to do something similar).
talk in Kriol, but they can switch to English.
Then, the case endings go. Then, the nouns are reduced

to those expressing objects, such as animals, plants,
Of the Walmajarri words used, twelve were used by
bodyparts. However, the use of initial position for
only one speaker. Only eleven were used by ten or more
expressing salient and prominent information appears to
speakers. These words are all words for things or
stay for a while.
people. They don't include words for actions or feelings.
In fact most of the 48 words denote objects.
Census data on languages spoken at home

What is happening in these communities is happening


Number of words in the category
across Australia. The number of speakers of traditional
Animals

16
Indigenous languages is declining. An index of
Bodyparts

11
language endangerment based on whether children are
Environment and plants 8
speaking Indigenous languages was developed in the
Humans and spirit world 4
most recent major work on the state of endangerment of
Food and drink
3
Indigenous Australian languages, the National
Material culture
2
Indigenous Languages Survey Report 2005 (NILS
Actions

2
report) (Australian Institute of Aboriginal and Torres

Strait Islander Studies and Federation of Aboriginal and
Table 2: Semantic categories of the 48 most frequently
Torres Strait Islander Languages, 2005), prepared by
used words
Patrick McConvell, Douglas Marmion and Sally

McNicol. This report contains a study of the 1996 and
The influence of traditional ways of presenting
2001 census data for Indigenous languages, comparing
information can still be seen in the way people talk at
them with other figures on numbers of speakers of
Yakanarra. In this example, the mother first puts the
Indigenous languages.
object, the big yellow cake, at the end of the sentence,

then in the next sentence she puts it at the beginning.
Since the NILS report, some of the results of the 2006

Census have appeared online on the Australian Bureau
(4)
MOTHER

of Statistics (ABS) website (http://www.abs.gov.au/).
ai meik _im _bat big yelo yelo keik .
The relevant question is "Does the person speak a
I'm making a big yellow cake.
language other than English at home?" Unfortunately
big yelo keik ai _l meik _im.
the online site doesn't yet give the crucial information
A big yellow cake I'll make.
for the language endangerment index of the ages of the
yelo _wan keik fo ola kid dei angri.
speakers of Indigenous language. However, enough
A yellow cake for all the kids who are hungry. [KXM
data is given to provide a preliminary picture of the
092B]
changes in numbers of Indigenous languages over the

last ten years. Table 3 compares the language figures
In summary, in our data, leaving aside Lajamanu, across
from 1996 to 2006.
the other three communities, no one talks to children

solely in a traditional Indigenous language, or solely in
standard English. Instead, people use a broad range,
from mixed languages like Gurindji Kriol, to traditional
languages, to a variety which is close to non-standard
rural English. They often switch between these
languages, sometimes in a single sentence. This has
consequences for the children's future language
development. While children in Lajamanu can still
produce traditional Warlpiri, in the other communities
children are not likely to use the traditional languages in
full sentences. Particularly in Kalkaringi, children
understand everyday talk in their traditional language.
But, if present trends continue, the children in all three
communities are unlikely to develop a good active
command of these languages. This in turn means that

---

2006
noted there was under-counting and over-counting

1996 Census 2001 Census Census
because Indigenous respondents were quite mobile.
Speaks English 82%
80%
78.5%
There was also under-counting of children. They also
only
14,564,924
14,875,072 15,581,334
noted that some of the questions were not understood by
Speak other
the census administrators and the people responding to
2,657,767
2,841,210
3,146,194
language:
the census. As well, the ABS note that there has been
Australian
some miscoding of the census data to do with
Indigenous
48,193
50,978
55,698
languages. For example, the 2006 census lists 118
Languages
Dhay'yi speakers, compared with 3 people in 2001,
Language not
apparently because 84 persons who reported that they
530,138
872,026
1,127,760
stated
spoke 'Dari' or 'Thai' at home were miscoded as 'Daii',
and included in the Yolngu language Dhay'yi4.
Total
17,752,829
18,588,308 19,855,288

Number
On the interpretation of the results, I point to three
identifying as
problems. The first has to do with differences of names
Indigenous
352,970
401,916
455,028
of languages - the numbers of speakers of Yolngu
Source: ABS 20063
Matha, Djambarrpuyngu, Dhuwaya have varied greatly

across the three census counts, according to which name
Table 3: Census data 1996-2006 on languages spoken at
was most popular at the time. Some languages,
home in Australia, and on number of people identifying
especially the new ones, don't have well-established
as Indigenous.
names. The new mixed language Gurindji Kriol has 4

speakers in the 2006 census; it has far more speakers,
The table shows that most people claim they speak
but people have not had a name to describe the language
English only at home, although this has decreased
they use. As a result they may have recorded themselves
slightly over the last ten years. There are far more
as speaking Gurindji or perhaps Kriol or English. This
speakers of immigrant languages like Chinese in
would result in overcounting of Gurindji and Kriol
Australia than there are of Indigenous languages. Note
speakers.
also that the 2006 figure of 55,698 Indigenous

languages speakers covers more than a hundred
The second problem concerns what speaking a language
languages. There is apparently a small increase in
at home means: language of regular communication,
speakers of Indigenous languages over the 10 years.
versus language of occasional or ceremonial
But this is an illusion.
communication, versus language which is being learned

(e.g. the 2006 census lists 34 people as speaking Kaurna
Before explaining why the number is an illusion, I shall
at home, but, since Kaurna is a language which has been
briefly discuss some necessary cautions on relying on
revived largely from nineteenth century sources, the
census data on Indigenous languages. The caveats fall
range of these people's use of Kaurna is different from
into three types - problems with the data collection,
that of the 25% of Yolngu Matha speakers who say they
problems with the coding, and problems with
don't speak English well, or the 48 people who speak
interpreting the data.
Warlmanpa at home, and who heard Warlmanpa spoken

around them as children). As well, there may be an
On the data collection and coding, a research team who
overestimate of the number of children speaking
shadowed the administration of the 2006 census in some
Indigenous languages, since parents who speak an
northern Indigenous communities (Morphy, in prep.)
Indigenous language at home may be taking their

children's ability to understand an Indigenous language

as equivalent to speaking that language.



The third problem has to do with willingness to admit

speaking a particular language, which also relates to

having a name to describe the variety a person speaks.

There has been a large increase in the number of people
3 Cat. No. 2068.0 - 2006 Census Tables 2006 Census of
Population and Housing Australia, Language Spoken At

Home(A) By Sex For Time Series. Count of persons. Based

on place of usual residence.

Cat. No. 2068.0 - 2006 Census Tables 2006 Census of
Population and Housing Australia. INDIGENOUS STATUS

BY AGE BY SEX FOR TIME SERIES. Count of persons

(excludes overseas visitors) Based on place of usual residence.

ABS 2001 Census of Population and Housing Australia.

Language Spoken at Home by Sex. Count of persons. Based
4
on place of usual residence
http://www.abs.gov.au/AUSSTATS/abs@.nsf/Latestproducts/
ABS 1996 Census of Population and Housing Australia. Basic
0260B4B1CD4563A6CA25729E0008A883?opendocument.
Community Profile.
Checked 16/10/07

---

saying they speak Torres Strait Creole (Yumpla Tok)
Arrernte
2,835
and Kriol, as I shall discuss below.
Kriol
4,213

Bearing these reliability concerns in mind, we return to
Aboriginal English
4885
the question of understanding the apparent increase in
Torres Strait Creole
6,042
language speakers. Considering the number of
Total people Indigenous languages
55,695
Indigenous people overall, and comparing the 2006
Source ABS6
census figures with the census and NILS report figures

for 1996 and 2001, it appears that the number of people
Table 4: 2006 Census: Number of people speaking
identifying as Indigenous is increasing rapidly (13%
major Indigenous languages (1000+ speakers)
increase from 1996 in 2001, and also in 2006), but the

number of people saying they speak an Indigenous
A point to notice from Table 4 is that over 10,00 of the
language at home (5% increase from 1996 in 2001, and
speakers are in fact speakers of new Indigenous
10% increase in 2006) is not keeping pace with the
languages, Kriol or Aboriginal English or Torres Strait
increase in people saying they are Indigenous. There's
Creole (Yumpla Tok). Comparing the figures for 2006
no evidence of a drop in birth-rate in remote Aboriginal
with 2001 (Table 5), we can see that there has in fact
communities where Indigenous languages are spoken,
been a drop in the number of people saying that they
and so we might posit that fewer children are learning
speak traditional languages.
Indigenous languages. However, the difference in rate

of increase of speakers and rate of increase of people

2001
2006
identifying as Indigenous is merely suggestive, since the
speakers of "Kriol"
2990
4,213
latter could be affected by greater willingness among
speakers of Aboriginal English

488
adults to identify as Indigenous. A proper study would
speakers of Torres Strait Creole
1240
6,042
require a community by community study comparing
speakers of other Indigenous 46,748
44,952
numbers of Indigenous language speakers in different
languages
years.
Total:
50,978
55,695

Source: ABS and my calculations
We turn now to those languages which have more than

1000 speakers, given in Table 4. Even for these
Table 5: Comparison between speakers of new and
relatively large languages, the absolute number of
traditional Indigenous languages, 2001 and 2006
speakers is small. For these languages, however, other

evidence from community members and linguists
The number and proportion of speakers saying they
suggests that there is a core of people still speaking
speak a new language (Kriol, Torres Strait Broken,
these languages as a first language, and almost all have
Aboriginal English or Gurindji Kriol) has increased
a solid number of children speaking them (NILS Report
substantially since 2001. This could be because more
2005).
people are feeling confident enough and knowledgeable

enough about what they speak to claim it. Or it could be
Language name
2006 census
because there are more people speaking it.
Ngaanyatjarra
1,000

Figure 2 shows the changes in numbers of speakers of
Anmatyerr
1,002
the larger Indigenous languages since 1996 (not all
Wik Mungkan
1,050
languages are named the same way in each census year
Burarra
1,074
which results in some strange distributions, particularly
Kalaw Kawaw Ya/Kalaw Lagaw Ya
1,216
with respect to the Yolngu languages).
Anindilyakwa
1,283
Luritja
1,480
Alyawarr
1,664
Tiwi
1,716
Murrinh Patha
1,832
Warlpiri
2,507

Pitjantjatjara
2,657

Djambarrpuyngu
2,766





5 Aboriginal English is included because it is likely to include
both creoles and non-standard English.
6 Cat. No. 2068.0 - 2006 Census Tables 2006 Census of
Population and Housing Australia Language Spoken At Home
(Australian Indigenous Languages Only) By Sex.

---

Major Indigenous languages
Kalaw Kawaw Ya/Kalaw Lagaw
Ya
Anindilyakwa
6000
Tiwi
5000
Murrinh Patha
Warlpiri
4000
Pintupi
3000
Kukatja
Pitjantjatjara
2000
Number of speakers
Dhuwaya
1000
Alyawarr
0
Anmatyerr
1996
2001
2006
Arrernte
Census years
YOLNGU MATHA FAMILY
Source: ABS and NILS Report

Figure 2: Changes in numbers of speakers of the larger Indigenous languages since 1996

The speakers of most languages have remained the same
Speakers of Warlpiri
488
or gone down since 1996. Murrinh Patha and Kalaw
Speakers of unidentified Indigenous 5
Kawaw Ya/Kalaw Lagaw Ya are the only languages
languages
with substantial increases. Languages whose speaker
People who speak English only
61 (10%)
numbers have gone down, or remained about the same,
Language not stated
57
since 1996 include Tiwi, Warlpiri, Anindilyakwa. Of
Total
614
the three big language groups, among the Arandic
Source: ABS7
languages, Arrernte and Anmatyerre have gone down,
.
while Alyawarr has a slight increase. Among the
Table 6: Language spoken at home by Indigenous
Western Desert languages, Pintupi and Kukatja have
people, Lajamanu, 2006 Census
decreased, while Pitjantjatjara has a slight increase. It

isn't clear what's happened to the Yolngu Matha group,
Leaving aside the questions of whether the people who
since Dhuwaya speakers appear to have started calling
claim they speak Warlpiri are speaking traditional
their languages by other names in the 2006 census, and
Warlpiri or the new mixed language Light Warlpiri
so I have given the overall Yolngu Matha figure for
(O'Shannessy, 2006), and of whether the people who
2006. The fact that even these strong languages have
claim they speak English are speaking standard English
mostly failed to increase their numbers is a matter of
or non-standard English or Kriol, 10% of the Indigenous
grave concern, since the number of Indigenous people
has increased.



A final important factor to consider is the age

distribution of the Indigenous population. All the speech

communities are so small that the survival of the

languages is precarious. Take a community such as

Lajamanu in the Northern Territory. Table 6 shows

language use by Indigenous people from the 2006
7 Cat. No. 2068.0 - 2006 Census Tables, 2006 Census of
census.
Population and Housing, Lajamanu (CGC) (Indigenous Area)

- NT, Language Spoken At Home By Proficiency In Spoken
English/Language. Count of Indigenous persons. Based on
place of usual residence. Figure excludes Indigenous speakers
of languages other than English and Australian Indigenous
languages

---

population claim they speak English only at home. This
Source: ABS 20069 and own calculations
is not such a concern for language maintenance when

the community is large, but it is when the community is
Table 8: Percentage of population under 15 [excluding
small, as their influence can be quite strong. Table 7
those who made no comment on Indigenous status]
shows the age distributions at Lajamanu.


Thus for the last ten years and perhaps earlier, just
Total Indigenous population 613
under half of the Indigenous population have been
People under 15
214 (35%)
language learners. A related point is that lots of girls
People under 15:
Min. in year: 5
are having babies in their teens. The percentage of
Range of number in year Max. in year: 20
young mothers (under 24) is much higher than in the
cohort
general population. Many of the under-15-year-old
Source: ABS8 and my calculations
children recorded in the 1996 census are now parents.

Their children will be listening to how they talk, as well
Table 7: Age of Indigenous population, Lajamanu, 2006
as how their playmates talk. A possible brake on
Census
language shift are if the primary caregiver is actually a

grandmother or great grandmother who is a strong
35% of the population is under 15. But the number of
speaker of the traditional language. However, this effect
children at each year level is small - for each year there
is often cancelled out or diminished by the fact that
may be from between 5 and 20 children of that age in
many older people endure poor health, including
class at school. One or two popular children who insist
dementia at younger ages (Broe, Jackson Pulver, Flicker
on speaking English may be all it takes for a whole class
and Curnow, 2007), and untimely deaths. They also
to shift to speaking more English, or, equivalently, for
suffer from the social dysfunction in some of the
maintaining Warlpiri.
communities and the large number of demands on the

time of capable people.
Warlpiri is a relatively strong language, and Lajamanu

is a relatively large and homogenous language
A final factor relates to child-rearing practices. It
community. But even so, it is very easy to see how little
appears that in many Indigenous communities,
it would take to cause a language shift there.
children’s independence and right to decision-making
O'Shannessy's thesis documents a shift over the last 25
are highly valued (Hamilton, 1981; Hamilton, 1982;
years or so from children speaking traditonal Warlpiri to
Shaw, 2002), and force is rarely used by parents against
children speaking a mixed language, 'Light Warlpiri',
children. Parents accede to children’s requests
which has a Kriol verb spine and Warlpiri case suffixes.
(Kaberry, 2004), and are unlikely to enforce speaking a
Nonetheless, they can still speak traditional Warlpiri to
traditional language instead of a creole or English. As
some extent, and this may be attributable to the school
well, children take responsibility for younger children,
Warlpiri language development programme.
and spend a lot of time with other children. The peer

group pressure is very strong. (Kaberry, 2004).
We have seen how young the population at Lajamanu is.

Table 8 shows that this is true more generally of
When the fact that in many traditional Aboriginal
Indigenous communities. And note that in remote
communities children are encouraged to be independent
communities the number of children is likely to have
and to spend time with other children is coupled with
been under-estimated rather than over-estimated
the fact that television and street lights make it easier for
(Morphy, submitted).
children not to spend time listening to older family

members at night, this all reduces the time children

Australia
Northern Territory
spend listening to traditional languages. It is thus likely
Non-
Non-
that the effect of the peer group on Indigenous childen

Indigenous Indigenous
Indigenous Indigenous
in remote commmunities like Lajamanu will be even
stronger than on migrant children living in nuclear
1996
40%
21%
38%
23%
families in cities.
2001
40%
20%
36%
22%

2006
38%
19%
35%
20%
Thus the small size of the communities, the young age














8 Cat. No. 2068.0 - 2006 Census Tables 2006 Census of
9 Cat. No. 2068.0 - 2006 Census Tables 2006 Census of
Population and Housing, Lajamanu (CGC) (Indigenous Area)
Population and Housing Australia. Indigenous Status By Age
- NT Age By Indigenous Status By Sex. Count of persons.
By Sex For Time Series. Count of persons (excludes overseas
Based on place of usual residence. Excludes those who made
visitors) Based on place of usual residence. Converted to
no comment on Indigenous status.
percentages.

---

of the population, young mothers, the independence of
There is already a drift to towns, triggered by greater
children and the importance of the peer group mean that
access to services such as dialysis machines, as well as
if language shift takes hold among children, the spread
to entertainment and alcohol. Moving to town will have
to the next generation can be very rapid.
a great impact on children, as they will suffer from the
greater access to alcohol, as well as from reduced access
Social landscapes
to activities such as gathering and hunting in the bush.
Having parents who work mean that children will be
The state of Indigenous languages in Australia is
even more influenced by their peer group. What
precarious at the moment. But their state is likely to
happens at school may also influence them, depending
become more precarious in the next few years, as a
on the effectiveness of the home liaison officer and the
result of the interplay of new government policies with
punishing of parents if children miss school.
the factors listed above (the monolingual mindset of
Australia, the demography of communities, and child
rearing practices). In 2007 the Australian Government
Moving to town will also result in fragmentation of
started introducing social welfare policies which are
speech communities, as it may not be possible for
aimed at making it hard for Indigenous people to stay in
people to find houses near family speaking the same
remote communities, where most children who speak
language. It is also unlikely that, given the number of
Indigenous languages as their first language live.
language groups involved, town schools will have the

resources to run proper Indigenous language
Remote communities are attractive to many Indigenous
programmes. Unless the strong effects of peer group
families as places to raise children, because, while some
pressure are countered by effective English as a Second
are dysfunctional, many offer the security of home,
Dialect or English as a Second Language teaching, and
cheap accommodation, free child-care from family
by a feeling of partnership between Indigenous and non-
members, a safety net of relations to provide support,
Indigenous people, it is likely that children will switch
ready access to the bush for gathering and hunting, and
to speaking a creole, rather than standard English. This
a relatively safe place for children to roam around, since
will not help the children's access to further education
there are few strangers. Remote communities also have
and employment.
more Indigenous people controlling organisations and
services. There are, however, undeniable social
problems. People are poor, they have poor housing, and
Conclusion
poor health. They are in constant mourning for relations
Fewer and fewer children are growing up speaking a
who have died young. There are few jobs on remote
traditional Indigenous language of Australia. Their
communities. Many children miss many days of
number is set to decrease sharply as a result of a lack of
schooling. There are many causes for this: ill health,
support for Indigenous languages at schools,
poor schools, high mobility of parents, mourning, as
demographic factors, child-rearing practices, and the
well as a failure to enforce school attendance by parents
likely effects of government policies aimed at moving
and by home liaison officers. There is considerable
Indigenous people out of remote communities and into
violence.
towns or itinerant work. By the time the Australian

Constitution10 is rewritten to include a mention of
The Australian Government's 2007 solution to this rests
Indigenous languages as part of heritage, it may be that
on the assumption that the cause for the social problems
most of those languages will no longer be used by
is the lack of employment and education and services
Indigenous people in everyday talk.
available on remote communities. The cost of providing
these on remote communities is high. And so, instead
Acknowledgements
they have decided to make the costs of living on
This paper owes much to my fellow researchers in the
communities higher, so as to outweigh the benefits
ARC-funded Aboriginal Child Language Acquisition
Indigenous people find from living there. This is
project (Samantha Disbray, Patrick McConvell, Felicity
happening in several ways. First, they are cutting off the
Meakins, Betty Nakkamarra Morrison, Karin Moses,
supply of money and services to people living on
Carmel O'Shannessy and Jill Wigglesworth), as well as
remote communities, by abolishing 'work-for-the-dole'
to the organisers and participants in the Indigenous
schemes in communities of fewer than 100 people, and
Languages Conference (Adelaide, September 2007) and
by
abolishing
the
Community
Development
Employment Programme which provides extra wages

and services in many communities. Incidentally, this

will probably also result in the loss of Indigenous

language workers' jobs in language centres and schools.

Second, they are exerting more control over people's

lives in remote communities, by installing government

business managers and quarantining welfare payments.

This is to encourage people to find paid work, which,
10 John Howard "A new indigenous settlement". 12/10/2007
almost inevitably, will lead to a move to town.
The Australian.

http://www.theaustralian.news.com.au/story/0,25197,2256993
9-7583,00.html. Accessed 23/10/07.

---

the AIATSIS seminar (October 22 2007). I thank
code-switching. Australian Journal of Linguistics:
Maggie Brady, Patrick McConvell and David Nash for
Special number edited by Margaret Florey and
valuable discussion.
Patrick McConvell, 25(1), 9-30.
Meakins,
Felicity
and
O'Shannessy,
Carmel.
References
(submitted). Ordering arguments about: The
interaction of word order and ergative marking in
.Australian Institute of Aboriginal and Torres Strait
two Australian mixed languages. In W. McGregor
Islander Studies and Federation of Aboriginal and
and J.-C. Verstraete (Eds.), Special Issue of Lingua
Torres Strait Islander Languages. (2005). National
Morphy, Frances. (in prep.). Agency, contingency and
Indigenous Languages Survey Report 2005: Report
census process: Observations of the 2006 Indigenous
submitted to the Department of Communications,
Enumeration Strategy in remote Aboriginal
Information Technology and the Arts. Canberra:
Australia. Canberra: Centre for Aboriginal Economic
Department
of
Communications,
Information
Policy Research, The Australian National University.
Technology&theArts.
Morrison, Betty Nakkamarra and Disbray, Samantha.
http://www.dcita.gov.au/__data/assets/
(2007). Children talking in Tennant Creek.
pdf_file/35637/NILS_Repport_2005.pdf.
Indigenous Languages Conference, Adelaide:
Broe, Tony, Jackson Pulver, Lisa , Flicker, Leon and
Musharbash, Yasmine. (2001). Indigenous families and
Curnow, Venessa. (2007). Is dementia a problem for
the welfare system : the Yuendumu community case
Aboriginal
communities?
National
Dementia
study, stage two. Centre for Aboriginal Economic
Research Forum, Sydney.
Policy Research, Australian National University.
Charola, Erika. (2002). Verb phrase structure in
O'Shannessy, Carmel. (2006). Language contact and
Gurindji Kriol. Honours. University of Melbourne,
children's bilingual acquisition: learning a mixed
Melbourne.
language and Warlpiri in northern Australia. PhD.
Hamilton, Annette. (1981). Nature and nurture:
University of Sydney, Sydney.
aboriginal child-rearing in North-Central Arnhem
Shaw, Gillian. (2002). An ethnographic exploration of
land. Canberra: Australian Institute of Aboriginal
the development in child rearing style among the
Studies.
Ngaanyatjarra people from the pre-contact era to the
Hamilton, Annette. (1982). Child health and child care
present. Masters. University of New South Wales,
in a desert community 1970-71. In J. Reid (Ed.),
Sydney.
Body, land and spirit : health and healing in
Simpson, Jane and Wigglesworth, Gillian. (Eds.) (in
Aboriginal society (pp. 48--71). St. Lucia, Qld.:
prep.). Children’s language and multilingualism:
University of Queensland Press.
Indigenous language use at home and school.
Kaberry, Phyllis. (2004). Aboriginal woman: sacred
London: Continuum International.
and profane. London and New York: Routledge.

McConvell, Patrick and Meakins, Felicity. (2005).
Gurindji Kriol: a mixed language emerges from

---