Public Culture 18(1): 1 – 000 DOI 10.125/10642684-2006-123
© 2006 by Duke University Press
This Is Not a Pipe:
The Treacheries of Indigenous Housing
Tess Lea and Paul Pholeros
There was an important moment in surrealist art when the
Belgian painter René Magritte pointed out the simple truth that his painting of
a smoker’s pipe was not a pipe (see ﬁg. 1). Today there is still a small shock to
be had in recognizing the layers of automaticity embedded
in our conventions of viewing. We react immediately. Of
course, Magritte’s painting is not itself a physical device
that one can ﬁll with fragrant tobacco, ignite, and inhale,
allowing smoke to curl and linger on lip and in lung, but an
image, complete with corrective text that refers to its literal
status: Ceci n’est pas une pipe. Then a belated realization
hits: the tag “This is not a pipe,” while a subtle calligram, is
itself also ﬁgure, seemingly factual script, which we “read”
only to be doubly deceived. Enticing us to elucidate mean-
ing, the referential text deﬂects attention away from its own
status as also, like the pipe, painted image. It is a reﬂection
of text, brushstrokes on a surface fashioned into the shape
of an explanatory legend. The text that seems to reveal also masks, a form of
trickery that is best captured by the painting’s title, La trahison des images (The
Treachery of Images).
Public Culture 22:1 d o i 10.1215/08992363-2009-021
Copyright 2010 by Duke University Press
This article draws both on long-term Healthabitat data and program efforts led by directors Paul
Pholeros, Paul Torzillo, and Stephan Rainow and on wider anthropological ﬁeldwork conducted by
Tess Lea. Lea’s anthropological ﬁeldwork began in 2005 under a scoping grant from the Charles
Darwin University Research Innovation Panel. Fieldwork between 2007 and 2009 was made pos-
sible by an Australian Research Council Industry Linkage Grant. Our thanks go to Caz Comino,
Gillian Cowlishaw, Elizabeth A. Povinelli, Shane Thamm, and Paul Torzillo for useful comments
on earlier drafts.
Figure 1 Words and images, with apologies to
René Magritte. Image by Paul Pholeros
By interrupting culturally engrained habits of viewing and interpretation,
Magritte’s work draws attention to how perception and analysis are subtly directed
through habituated modes of viewing and classiﬁcation. As Michel Foucault put
it, in an essay on exactly this painting, there is a long-standing habit in West-
ern interpretation of suspending further thought when faced with a ﬁgure that
resembles another object or ﬁgure. When an image exerts this hold, a “what you
see is that” effect kicks in: “Resemblance and afﬁrmation cannot be dissociated”
(Foucault 1983: 34). Writes Foucault (1983: 37), “Ceci n’est pas une pipe exem-
pliﬁes the penetration of discourse into the form of things.” Or as Magritte puts
it, “A thing which is present can be invisible, hidden by what it shows” (quoted in
Gablik 1976: 12).
It might seem odd to begin an essay on the contentious politics of indigenous
housing in Australia with a leading surrealist painter. After all, a house is not a
painting. But then neither are indigenous houses, houses. They might look like
houses, most especially when they are newly constructed or refurbished. But the
appearance of new “affordable” houses, buttressed by scripted policy announce-
ments about dollars spent and program achievements, misdirect our interpretation
away from what is literally in front of us: a cheap, partly complete steel shed or
copy of a house of bare utility, which looks like, but is not, a house. It is a non-
house (ﬁg. 2).
The supply and maintenance of affordable housing and infrastructure remains
one of the most vexed issues confronting indigenous public policy. Many per-
Figure 2 Looks like a house. Photograph by Paul Pholeros
This Is Not a Pipe
manent indigenous dwellings are in need of major repair or replacement; are
overcrowded; and lack sufﬁcient water supplies, washing facilities, or sewage
infrastructure. The scale of the poor housing situation for indigenous Australians
is subject to highly fallible calculations; but drawing on information from the
Australian Institute of Health and Welfare (AIHW 2005), the high dependency
on government supply for indigenous accommodation becomes clear. Of the esti-
mated 165,700 indigenous households in Australia, two-thirds are rental situa-
tions, of which nearly 60 percent are public housing (AIHW 2005: 1). Across
Australia, more than one-third, or 58,100, of the permanent houses accommodat-
ing indigenous households have structural problems, a proportion that grows to
44 percent of all houses occupied by indigenous people in the Northern Territory
(AIHW 2005: 94).
Diagnoses of why indigenous housing remains so poor despite seemingly vast
program expenditures may alternately blame the racist state (e.g., Morgan 2006)
or point to inadequate consultation processes (e.g., Lee and Morris 2005). But
by far the most common tendency is to attribute much of the cause to the inca-
pacities of the Aboriginal householders (e.g., Hughes 2007). Whether hard-line
or empathetic, aired in policy backrooms or as eyewitness accounts on talk radio
and Internet postings, there is a dominant mode of interpreting damaged houses
that places indigenous values and behaviors at the center of the housing problem,
an interpretive syndrome that tends to bounce off the seeming look of things
into well-worn grooves of explanation and remediation. The prevailing theory is
that houses become structurally compromised in swift time frames more or less
because of the way householders tend the house.
For the beginnings of a counternarrative, we ask that you hold in mind the fol-
lowing: endemic overcrowding (or high-use load) contributes to rapid wear and
tear, and since much Aboriginal housing is cheaply constructed, repair and main-
tenance issues quickly become a problem. This has a series of material effects.
On the one hand, intermittent maintenance exacerbates degradation of housing
stock. On the other, it swiftly creates a visual image of mess and decrepitude that,
like the portrait of a pipe, invokes an interpretive automaticity for the remediating
viewer. The unkempt house represents a lack of householder pride that becomes
the portmanteau explanation for the substandard nature of indigenous housing
stock. The visual blight is the shaman’s material object: a stone found in the gut
that is responsible for all other afﬂictions.
For some commentators, Aborigines neglect their houses out of ineptitude,
indolence, and ﬁnancial incompetence, borne of too many years of passive wel-
fare. “I dunno, Tess,” a maintenance ofﬁcer working in an indigenous housing
organization put it. “A lot of what outsiders say happens, really does. My sister
took photos of a house on Groote Eylandt. It was like watching white ants eat the
tree — one day a door missing, the next a window; in six months there was noth-
ing but a brick shell. I don’t blame whitefellas for getting cross. We have created
expectations. Now it’s too far gone.” For anthropologists and other well-meaning
commentators, a curiously analogous theory dominates. When the topic of
Aboriginal housing moves into conversation, anthropologists have a pious lesson
in cultural evolution ready to hand (see, e.g., Sutton 2007). They will recall that
Aboriginal people were and are hunter-gatherers, who roamed in small family
bands and cooked and fed outside. In customary times, thick social relations and
a demand sharing economy were automanaged via sensitivity to nuance and close
monitoring of the emotional tenor of everyday events. Domiciles were not perma-
nently composed but could be split up to avoid conﬂict, with a strategic parting
of company being the principal dispute resolution mechanism. A makeshift camp
could be destroyed in a ﬁt of rage or in an outpouring of monumental grief, with
little lasting damage. With no ﬁxed property to be owned, neither the disciplines
of domestic hygiene nor the coveting of ﬁxed assets needed to be recognized.
The transition to a sedentary, settled lifestyle has unraveled these underpin-
nings. Three-bedroom houses groan under the strain of untenable numbers as
extended family transition in and out. Aggravations between household members
across gender, age, and family lines are exacerbated either by the close proxim-
ity of houses that are arranged to mimic urban subdivisions or by the lack of
traditional spatial separation, leading to stress, vandalism, disillusionment, and
abandonment. Only now, when houses get damaged, they are on permanent dis-
play, and kin live cheek by jowl in the mess. The mattresses, camp beds, and
outside cooking ﬁres to be found in sparsely grassed yards are pointed to as fur-
ther proof that suburban-style houses conﬂict with the different cultural desires
and social histories of indigenous people. Continuing the cultural discontinuity
reasoning, anthropologists join sympathetic interpreters in explaining that toilets
are ruined and stoves broken because they remain foreign impositions. Statistics
concerning their higher-than-average breakdown simply conﬁrm for the omni-
scient anthropologist the poor lay understanding of indigenous living preferences,
the preferences of a nonaccumulative people who to get by really do not need
much more than secure storage space for the odd artifact and a shedlike shelter for
Contrasting with these accounts, which, using a cultural teleology, still pin
housing dysfunction to householder attributes, this essay argues for a new material
literalism to demonstrate the surreal nature of indigenous nonhousing in Austra-
This Is Not a Pipe
lia. Drawing on ethnographic material, housing data, and a visual archive span-
ning the previous two decades, we hope to suggest how pulling the aesthetic logic
of discourses out of indigenous habits and habitats pulls the housing from the
house. Indigenous houses are composite deceptions. They are material non sequi-
turs, the penetrative effect of discourse in things, ﬁgures that produce an aesthetic
response with material consequences. But whereas Magritte’s painting intends to
unsettle the lock between representational orders, to disrupt the governance of
expectations by ideas, the aesthetic order this essay examines ﬁrmly tethers image
to meaning. The surreal nature of indigenous housing construction becomes the
real moral deﬁcit of indigenous people. While hardware breakdowns are more the
result of poor initial construction and material choice, inadequate maintenance,
and high levels of wear and tear, the residents themselves are routinely consid-
ered the main reason houses (and the items that make houses habitable) are in
In this discursive environment a new level of liter-
alism, a dirty literalism about materiality, is needed
to reintroduce for the interpreter the surreal nature of
indigenous dwellings, to show how moral orders of
responsibility and culpability are socially directed.
For this reason we assert that a level of functionality
is required before the objects of Aboriginal housing
policy should be considered materially real, before a
house is deemed a house. If a pipe is not a pipe if it can-
not be smoked, a length of polyvinyl chloride tubing is
not a pipe when, as is the case with much indigenous
housing in Australia, it might be materially and three-
dimensionally present, but it does not connect the toilet
to an efﬂuent disposal system. That is, it is literally not
a pipe — deﬁned as a hollow cylinder for transporting
ﬂuids — when it stops a meter from the house, buried
under dirt, without deliberate connection to a subter-
ranean plumbing network, unable to ﬂoat detritus away
(see ﬁg. 3). The sign of the material house is an illusion
when no system of institutionalized expectation is in
place to connect the physical structure (house) and the
range of functions it is assumed to be able to provide
to the resident: safety, security, and health beneﬁt (which might then deserve the
title “public housing”).
Figure 3 When is a pipe not a pipe?
Image by Paul Pholeros
Seeing Is Believing
Houses-that-are-not-housing are endemic across indigenous Australia. Special-
izing in driving improvements in Aboriginal housing function, Healthabitat
Pty Ltd has generated independent data on housing function in more than ﬁve
thousand dwellings across Australia from 1999 to the present.
Run by directors
Paul Pholeros, Paul Torzillo, and Stephan Rainow since 1985, Healthabitat devel-
oped an approach for surveying and ﬁxing what they call “health hardware” in
Aboriginal allotments, via a methodology that is now known as the “Housing for
The program is complex but is based on the straightforward
yet strangely contentious premise that householders’ ability to practice speciﬁc
healthy living practices is dependent on their means — the health hardware — to
do so. Using standard, repeatable measurement of housing performance in terms
of safety and the basic requirements for healthy living (such as a working shower,
a working toilet and wastewater system, a safe electrical system, and a kitchen
where food can be stored, prepared, and cooked), Healthabitat surveys show
repeatedly that the condition of Aboriginal houses is extraordinarily poor. We
report here on data collected from projects performed in 132 communities sur-
veyed between January 1999 and November 2006 (see Torzillo et al. 2008; also
SGS Economics 2006). Of some 71,869 items assessed as requiring repair or
replacement, 65 percent of the problems were due to maintenance (or the lack
thereof), 25 percent to faulty installation or equipment, and only 10 percent to
householder damage or misuse.
1. The Housing for Health method conducts a survey on each house to determine whether the
health hardware and other features of that house are functioning (Survey Fix 1). In each case, a visual
check alone is insufﬁcient. Each outlet has to be tested, temperatures taken, ﬂow rates physically
measured, buttons pressed, and toilets ﬂushed with wads of paper. The nonfunctioning elements
of the house are then ﬁxed, either on the spot or shortly after the initial survey and, for larger jobs,
over a six- to twelve-month period. Houses are then surveyed again (Survey Fix 2) to assess the
functioning of all items and whether the urgent items commissioned to licensed trades have indeed
been ﬁxed. Over the years Healthabitat has collected a large amount of detailed information on the
functionality of houses before and after repairs are completed. For more details, see Pholeros et al.
1993, 2000, 2004; Lea 2008b; Torzillo et al. 2008; and also www.healthabitat.com.
2. “Health hardware” was a phrase introduced by Fred Hollows to describe the physical equip-
ment necessary for disease-free living in ﬁxed domiciles: “In a water supply system, for example,
health hardware includes both the bore and the basin plug, as well as the shower rose [showerhead],
taps and drain” (DFCSIA 2007: 9). Hollows is also responsible for the mantra “No survey without
service” incorporated in the Housing for Health process.
3. Under the Housing for Health methodology, 250 items are checked in total in each house and
living area. Through such detailed empiricism in more than ﬁve thousand houses — including many
that look to all intents and purposes as if they should work — a functioning shower was available in
This Is Not a Pipe
only 35 percent of them. The criteria for functioning nutritional hardware included storage space for
food, preparation space, a working stove, and a sink. Only 6 percent of houses met these criteria at
the initial assessment. One-quarter of identiﬁed plumbing and electrical shortcomings were due to
faulty installation, the ﬁtting of wrong parts or components, or the absence of an essential item from
the house (Torzillo et al. 2008: 8). For more on the functionality of essential hardware in indigenous
houses assessed as part of Healthabitat’s work, see Australian Institute of Health and Welfare data-
cubes at www.aihw.gov.au/indigenous/datacubes/index.cfm.
4. It is interesting that this push toward the creation of private home ownership has abated in the
current context, where the former cultural habit in American banking of providing unlimited credit
expansion for private housing purchases regardless of one’s ability to pay has helped bring the global
ﬁnancial system to its knees.
There is, then, a different picture available to be seen. Poor construction by
contracted building ﬁrms, coupled with lack of supervision from the contracting
body and associated funders, leads to houses that break down, are poorly main-
tained, and do not function. Yet, conventionally, it is rare to see poorly superin-
tended public infrastructure programs and the micropractices of policy makers,
regulators, tradesmen, and manufacturers placed in the explanatory frame for
failing housing stock. Beginning with the former conservative Liberal govern-
ment and showing no signs of abatement under the current Labor government, the
prevailing policy argument is that Aboriginal people display shallow regard for
the dependency-causing gifts of welfare (e.g., public housing) and, further, that
this nonvaluing will be realigned only when the tenant learns to value the poorly
functioning asset via the motivations of home ownership and forms of indebted-
ness through mortgage arrangements.
Like most interventions in the indigenous
domain, it is an ill-considered or ersatz “free market” on offer: mortgages will be
brokered by the government, and houses will be paid for through subsidized bank
loans rather than subsidized rents. Why and for whom it is better to buy a small
and poorly functioning house than to rent public housing for the same amount are
rarely debated questions. But the dream itself is simple: the disproportionately
high levels of rental arrears, the poor maintenance, and the antisocial behavior
that observers associate with publicly funded indigenous housing will vanish with
the psychological possession that comes with home ownership, which will mani-
fest itself as houses that are treated with more respect.
In his inﬂuential blueprint for remote community reform, the indigenous
reformer Noel Pearson develops this argument explicitly. His central recommen-
dation “is to shift from the current system of exclusively public provision of hous-
ing to a system based on private property markets” to counter “the passive welfare
problems of the public housing model” (Cape York Institute 2007: 108). However,
the approach has to be comprehensively managed:
Families’ legal ownership of their own homes is a necessary condition for
responsibility, but in and of itself may not be sufﬁcient. Simply trans-
ferring the title and converting rent into mortgage repayments may not
work. [It] will not necessarily cause individuals or families to become
more engaged in, and committed to, the maintenance and protection of
their homes over the longer term. The experience of the “Katter leases”
in Cape York in the 1980s supports this view. These houses were gifted
in poor condition and nearing the end of their lifespan; the families were
not required to make any ﬁnancial or other commitments; and no educa-
tion programs were provided to inform people about their responsibilities
as home owners and how to fulﬁl them. As a result, the families’ sense of
responsibility and ownership remained unchanged and the houses contin-
ued to deteriorate. (Cape York Institute 2007: 109; emphasis added. See
also Hughes 2005, 2007.)
As these excerpts illustrate, despite the opportunity to consider the active role
played by houses that are “gifted in poor condition” to later deterioration, even to
possible devaluing, the dominant interpretation concerns a suspected irreverence
for the idea of the house itself, a failing that stands in need of comprehensive
forms of correction.
Of concern to us here is the way portraits of substandard housing are automati-
cally correlated with presumptions about deliberate householder damage and the
need to remediate, with the combined effect that evidence of poor practice in
house construction and the impact of irregular maintenance glides past the col-
lective eyeball. Aiding this elision are the serial images of damaged houses that
outsiders either see for themselves or access via media spectacle. The apparent
squalor of indigenous housing allotments plays a key aesthetic role in provoking
the compulsion to have a ready-made answer. In the more remote parts of north-
ern and central Australia, broken toys, hardware, and appliances litter unkempt
yards, together with soiled disposable diapers, car bodies, and general junk. Dis-
carded dishes, cemented with food remains, might lie tossed with casual abandon
amid piled laundry and stained mattresses. Holes are punched into walls in drug-
addled ﬁts of rage. People might sleep in dilapidated lounge rooms and on kitchen
ﬂoors, using bedrooms for visiting family members, refrigerators, or food storage.
Importantly, while this description does not hold for the majority of indigenous
households across Australia, these images of dereliction operate as proxy for the
dire nature of indigenous living conditions across the country. That is, without
necessary understanding of the diverse realities of indigenous households, a
gestalt image invokes a ritualistic round of causal explanations and declarations of
This Is Not a Pipe
more resolved intervention, a view that is reinforced by the sage advice of anthro-
pologists offering reiﬁed cultural history to explain the situation, paradoxically
reasserting an indigenous terminus point for the onus of responsibility.
Whether compassionate or hostile, the “rubrics of public safety and social
pathogenesis” (Feldman 2001: 60) that are correspondingly debated in policy
and research forums on the problem of indigenous housing share a key uniting
feature: they free policy makers, departments, designers, and builders of much
of the responsibility for the poor performance of what is there now. And what is
there now, whatever else can be said of indigenous living conditions, is also the
end product of an entire genealogy of previous expenditure, design, and imple-
mentation decisions. Of great signiﬁcance, but seldom regarded in images of the
indigenous housing problem, is what Primo Levi (2000: 129) calls the “frighten-
ing anesthetic power of company papers.” It needs remembering that the pipe that
is not a pipe, since it is unconnected to an efﬂuent disposal system, exists despite
techniques of regulation, audit, and administration that themselves support the
ambiguous, contradictory, and nonrepresentational relationship between words
and things. Well-crafted, thoughtful guidelines do exist that specify what and
how things should be installed to ensure safety and function under the harsh mix
of pressures that indigenous housing can be subject to (see DFCSIA 2007). For
example, drains with inadequate slopes are more likely to become blocked by
heavy use; thus, ensuring that national standards exist and that they are met for
indigenous housing are important technical interventions.
But even the most exacting speciﬁcations are undermined by a raft of local
conditions, from challenges experienced by materials themselves in climatically
harsh environments (think rust, rats, and calciﬁcation), to the inability to rely on
competent tradesmen to complete assigned jobs, to substandard work that is rou-
5. In other words, we are not claiming here that good design and detailed speciﬁcations are irrel-
evant to the performance of houses. Indeed, the robustness of the house’s ﬁttings, hot water system,
and waste system in response to overcrowding and water that is often aggressive (corrosive through
salt, alkalinity, acidity, etc.) in indigenous communities should be considered. Areas needing better
design and careful speciﬁcation and detailing include wastewater systems able to cope with large
numbers of people; hot water systems that take into account water quality, running costs, and house
population; bathroom layouts and ﬂoor drainage to cope with large numbers of people; showerheads
manufactured with water quality in mind; light ﬁxtures and energy-saving bulbs or tubes resistant to
insects, vapor, and vermin; doors and hardware, particularly locks; efﬁcient windows and new ways
of insect screening; stove tops and ovens that survive high-use load; kitchen-counter backsplashes;
kitchen storage units; solutions to keeping food cool and pest-free; usable yard areas with cooking,
sleeping, and storage potential; and thermal performance equivalent to sitting outside the house
under a tree.
tinely approved as complete.
In the laissez-faire administrative ﬁefdoms that often
“govern” indigenous communities, being insensitive to the letter of standards,
guidelines, and certiﬁcation requirements is a tactic of organizational survival,
while being punctilious about who is hired and the quality of what they deliver
represents the path of most conﬂict. So it is that contracts for indigenous housing
repair and maintenance, for example, to ﬁx repeatedly identiﬁed problems with the
plumbing, will be signed as done when a tradesman has only half completed the
job, having attended to the quickest-effort, highest-charge items over unrewarding,
longer-standing, and more complex repairs. Deliveries of ordered parts, whether
used or not, complete the appearance of fulﬁlled action to any casual scrutinizer.
This ﬁrst approval of a contract delivered on triggers others, as reliance on the
look of function shifts from the original certiﬁcation site to desks and in-boxes
in linked administrative tributaries, with various types of paperwork snowballing
the assertion “This is no longer a problem.” Eventually, at the top of the policy and
program food chain, ministers can report against aggregate program expenditure
on indigenous housing, maintenance, and governance efforts as the penultimate
sign of discharged pastoral responsibility.
6. For all that, regulatory frameworks and guidelines are necessary but not sufﬁcient ingredients
for functionality; the biography of indigenous housing guidelines is itself a story of struggle and
contest. After the Building Code of Australia and the Australian Standards, the National Indigenous
Housing Guide (DFCSIA 2007), ostensibly endorsed by all states, territories, and the Common-
wealth government in Australia, is the nationally accepted guide. It stresses safety and health and
is compiled from the collective practice wisdom of the trades, ﬁeld ofﬁcers, community residents,
engineers, and architects. While innocent enough in appearance, and containing suggested rather
than legislated codes, the most recent edition, volume 3, was released only after threats of litigation
from the ensemble editors against the Commonwealth of Australia.
7. Much is glossed over in this condensed description of the conspiracy of effects that create sub-
standard housing — for instance, reliable tradesmen. Indigenous infrastructure is hot, thirsty work
that sometimes attracts men who are as interested in the trade opportunities offered by the informal
sex and cash economies characteristic of high-poverty environments the world over as they are in
the tasks at hand. One answer to such conundrums is to build housing off-site and truck it in for
installation, to reduce not only costs but also reliance on potentially rough external labor. Such
efforts attract criticism for limiting the full-time employment of local indigenous workers in the total
project. At various times in the history of government responses to these systemic issues, there have
been attempts to make indigenous infrastructure work more attractive to larger, established ﬁrms via
composite contracts, with extra inducements for the time and ﬁnancial inconvenience of involving
indigenous employees. In all this, the dirty little jobs that require relentless levels of inspection at
low proﬁt-to-time-required ratios are at risk of being avoided. An answer to this, as pioneered by
Healthabitat, is to insist on highly regulated programs of assessment and repair. But then the addi-
tional supervision costs required for quality assurance are routinely held in policy dispute in efforts
to reduce the price of infrastructure programs, which gives the cycle of argument and (compromised)
solution another spin of the wheel.
This Is Not a Pipe
Chances are that at this stage of the cycle a new sur-
vey will be called for, because while effort has been
made and program dollars announced and allocated,
a backlog will still remain. A new survey will enable
government to better understand the nature of the prob-
lem and researchers to better establish a connection
between housing supply and indigenous health (ﬁg. 4).
This issue of endless surveying as a reﬂex policy
response requires separate treatment (but see Lea
2008a: chap. 9). For now, let us quickly note that as
part of the ﬁeldwork conducted for the larger study
from which this essay is drawn, seven distinct survey
teams examining the condition of the housing and
compositional densities appeared and disappeared
over a twelve-month period in one community alone.
These surveys not only create the look of action in
policy defenses, but also rely on the cursory glance for their completion: simply
poke your face into the house and look for a wall socket to answer the question
“Access to power?” (ﬁg. 5).
Even the survey forms achieve the initial appearance of instruc-
tive matrices but, on closer inspection, often turn out to be simula-
tions deserving the tag “This is not a survey.” They might have all
the serious ﬁdelity of a data grid yet invite annotations that cannot
be used to correct any particular faults in the hardware (see ﬁg. 6).
After all, what is anyone meant to make of a cell with “yes” scripted
in response to three possible questions and no standardized rating
system for deﬁning the answer?
Meanwhile, when — after all that assessment and accumulated
institutional response — the “health hardware” defaults to a predict-
able dysfunctional state, a handy culprit to explain the visibly spoiled
interior is the unhygienic and undisciplined indigenous tenant who
needs further tutelage in the arts of living in a house (being so new
to the concept of sedentary living and all).
It is this wider chain of interconnected acts of translation and
interpretation (and not simply some kind of vaguely assigned societal
prejudice) that has to be apprehended if we are to explain the grip of
the single most common explanation offered for the poor performance
Figure 4 We need more data. Image by Paul Pholeros
Figure 5 Hint . . . it might not be the
old or ugly one. Image by Paul Pholeros
of houses: namely, that it is due to tenant damage, misuse, or misunderstanding
of how to use the house.
To illustrate a similar interplay among surface appearances, “company papers,”
and conventions of interpretation, we offer a short case study of the life of a recent
“new idea.” In this story, a big company wanting to help the situation offers up
affordable houses, to be built by local people, with the aim of yielding a new
national housing model deserving of the title “This is a house.”
8. In addition to the relatively low incidence of vandalism, there is little evidence that poor
functionality is due to householder ignorance. This speaks to another long-standing “ethnologism”
in Australia, where the idea that indigenous people are mystiﬁed by modern Western houses — in
particular, appliances such as refrigerators and stoves — exerts a stranglehold on bureaucratic and
public imagination. A related and equally disputable commonplace is that education campaigns are
required to reduce household wear and tear. Interestingly, a recent study of the effects of mass sanita-
tion in Salvador, Brazil, represents an unusual contemporary instance in which the health impact of
mass investment in connected neighborhood sewers could be isolated. This work showed no connec-
tion between “hygiene behaviour” and reductions in diarrhea prevalence (Barreto et al. 2007: 1623).
Rather than education programs, it was the provision of more than two thousand kilometers of sewer
pipes for more than three hundred thousand homes in linked neighborhoods that caused diarrhea to
drop by 22 percent citywide — and by 43 percent in districts with the highest diarrheal prevalence
before the intervention (see also Cohen 2008).
Figure 6 This is not a “Housing Condition” form. Image by Northern Territory Department of
Health and Families. Reproduced by Paul Pholeros
This Is Not a Pipe
Let’s Build More Houses!
One hot day in December 2006, a group substantially composed of middle-aged
white men, organizational intermediaries representing assorted “stakeholder”
groups, traveled by bus to a remote community located in the tropical savannah
country southeast of Darwin in the Northern Territory of Australia. Their excur-
sion was organized around the unveiling of a cost-effective solution to the endemic
problem of indigenous housing: namely, an affordable kit house system. The proto-
types up for inspection that day had been designed by a private industry manufac-
turer, were economically trucked in to the remote area, and had used the labor of
unskilled indigenous locals to frame and complete external wall cladding on-site,
providing much-needed community employment. The intent of the prototype is that
structurally superﬁcial design changes will be easily accommodated within a lim-
ited number of overall housing designs, thus ensuring customization and cost effec-
tiveness at the same time. While it constituted a speculative project, at this point, we
join the men at the peak of a collectively shared enthusiasm for the ideal solution.
That is, attendees are predisposed to reading surface appearances to see a greater
materiality, a “total housing project,” their interpretive gaze predirected by a com-
pany brieﬁng paper proclaiming the hard-to-dispute worthy aims of the project.
What do they see, after their long and thankfully air-conditioned bus ride? The
prototype gleamed in the hot tropical sunlight (recall ﬁg. 2), displaying enough
of the look of a house with framing and external wall sheet materials in place to
allow a gallery of photos to be taken of pleased company men, locals, and visi-
tors standing side by side, with a sponsorship banner spanning the veranda posts
providing a photogenic backdrop (ﬁg. 7).
But as we alight from the bus for a closer inspection, while the majority of the
visitors complete business card exchanges, hand out gifts of footballs to the eager
local children, and linger for chitchat over cups of tea and iced bun, the image
of the house disassembles to become — like the pipe — a simulation. Inside the
structure, there is as yet no electrical wiring, apart from wall-socket mounting
brackets. That is okay: this is ordinarily the time when the space between the
studs, ﬂoor joists, ceiling joists, and every other nook and cranny is most easily
accessed and redone, should an inspection against electrical code requirements
necessitate rewiring. But no, there is no rough installation of the electrical wiring,
boxes, breaker panel, and subpanels to speak of. Nor have any pipelines or outside
yard works (trenches and footings, drainage or site leveling) to connect the house
to mains services commenced. Most of the internal linings have yet to be com-
pleted, the reﬂective foil sarking being no substitute for cladding (ﬁg. 8).
The type, location, and power draw of the hot water system are not decided.
Ceiling fans, essential in the melting tropical heat, are apparently intended, but
where and when has not been documented. There is no meter box, circuit breaker,
or grounding rod, placing a question mark around future electrical safety (see
DFCSIA 2007: 25); in the case of ﬁre, the security-screened bedroom windows
have no internal emergency release openings. The external veranda and breeze-
way ﬂoor have no slope to drain water away from the building, and well before
any resident is in place, the sheet ﬂooring material for the veranda (marked “for
internal use only”) is already showing signs of wear at the edges.
Inside the most important area for health gain (and the part most prone to
failure in much indigenous public housing), the steel wall frames of the bathroom
sit directly on the ﬂooring, foretelling a story of future corrosion, if the as-yet-
uninstalled waterprooﬁng membranes fail and the (inevitably cheap) faucets, also
yet to arrive on-site, leak in the wall cavity.
Figure 7 Pleased with
the product. Photograph
by Paul Pholeros
So what is there to
see? Certainly no one was
on the lookout for such
prophesies of serious fail-
ure points. The company
with the good intentions
was focused on selling
a material product that
could be beneﬁcial to the
ﬁnal goal of “indigenous
housing.” The adult excur-
sion was therefore timed
to inspect and sign off on
a “completed” material
product, a house, creating
a photo opportunity and a
good news story that soon
would become a fast-cir-
culating factoid in stories
of solutions to the holy
grail of cheap yet sturdy
housing for remote areas. But without the signiﬁcant additional work of super-
vised plumbers and electricians (not part of the “offered package”), the nonhouse
before us will quickly dissolve into an unsightly shelter, and nearby household-
ers will be held to account for their failure to tend it, this being the treachery of
images, texts, and meanings in indigenous public housing.
The Nonhouse: A Repeating Cycle
The issue of indigenous community housing in remote Australia and on the fringes
of outback towns ebbs in and out of view on the tides of national attention. To take
a recent example, in February 2008 the Australian Broadcasting Commission’s
(ABC) Radio National Program PM featured an interview with Paul Torzillo,
lead author of a scientiﬁc journal article describing Healthabitat’s historical data
on the poor functionality of indigenous houses (Torzillo et al. 2008). To recap: the
data show that vandalism is a cause of breakdown in 10 percent of cases, while
poor choice of initial materials, poor construction, inadequate maintenance, and
normal wear and tear account for the rest. Four hours and fourteen minutes after
Figure 8 Still not
housing. Photograph by
the interview transcript had appeared on the ABC Web site, seventy-eight com-
ments were posted, forcing the editor to block further reactions.
“I just love this quote,” sneered “John Connor,” the ﬁrst person out of the com-
ment box at 9:40 a.m., February 18, 2008: “ ‘Professor Torzillo says contrary to
popular myth, the vast majority of problems were not due to damage, but rather
faulty construction and a lack of maintenance.’ So the houses just smashed their
own windows, ripped off their own doors, broke their own toilets? This is con-
cerning if buildings are becoming sentient and choosing to kill themselves???”
“Tim” countered at midday: “But why would they want to throw bottles through
windows, why do some indigenous people carry such a lack of respect . . . ?
It’s easy just to blame them, but much more difﬁcult to look into reality. . . . In
short these people have been neglected and oppressed by the white powers that be
for far to [sic] long. Now we have an obligation to right the wrong.”
Finishing the frenzy, and echoing Pearson’s analysis of why early attempts at
gifting also failed, “Rob” asserted the most popular verdict of all:
The fact is that new houses in communities are not only continually dam-
aged but basic cleaning is not done at all. Even where all the facilities are
working, kids are not washed or showered regularly and the house often
becomes a disgusting, ﬁlthy mess in a few months — toilets, kitchens, etc.
are never cleaned as the occupants are too busy sitting around drinking
and gambling. Then in a few years the government does a “refurbish-
ment” and the cycle continues. Before the Federal Government puts more
millions into housing, there has to be some programs to ensure that they
[indigenous householders] look after them properly[;] otherwise it’s good
money after bad.
During these recurrent discoveries of “the problem of indigenous housing,” con-
cern about the vast sums of money that will have to be procured if state gov-
ernments are to meet the quantum required tends to dominate agitated public
discourse and policy ruminations alike.
The huge price tag associated with the
“unmet housing need” has the effect of transforming the ideal of adequate hous-
9. Australian Broadcasting Commission, www/abc.net.au/pm/content/2008/s2163237.htm
(accessed February 19, 2008).
10. A report of the Australian Housing and Urban Research Institute estimates a backlog require-
ment of $705 million, or $141 million annually for ﬁve years, in indigenous housing (Hall and Berry
2004: iv). But citing ﬁgures in this way implies an accuracy uncharacteristic of indigenous housing.
Uncanny costs are part of the aesthetics of failure feeding the loop of diagnosis and suggested rem-
edy that enables the recurrence of shoddy housing. It would be just as accurate and just as arbitrary
to add two zeroes to the ﬁgures every few years.
This Is Not a Pipe
ing into a different kind of pipe dream. At ﬁrst, the issue is simply that of ration-
ing. According to policy orthodoxy, scarce government resources have to be spent
efﬁciently. But as Bruno Latour (2002: 32) reminds us, “No technological project
is technological ﬁrst and foremost,” and certainly, the moral discourse embedded
in indigenous program considerations adds a distorting ingredient to otherwise
normative budgetary talk in the public sector.
The inability of abjectly poor householders to pay rent to tenancy managers
(whether these are banks, government agencies, or that beleaguered invention of
indigenous social policy, the Indigenous Community Housing Organization) at the
levels that would be required to restore inadequately designed and badly installed
houses is at the center of the recurrence of nonhousing in indigenous Australia.
But when the costs for a robustly designed house nudge above and beyond what
is (equally inadequately) spent on mainstream public housing, bureaucratic con-
cern about the public “look” of expensive public housing for indigenous people
kicks in. This concern soon leads to broad suggestions for reducing the costs of
administering and delivering housing programs, a discussion that just as quickly
narrows to ideas for reducing the size or speciﬁcation of the house. That is, before
a program of indigenous housing works is decided, the cost of a well-ﬁtted and
durable house is disputed in the backrooms of policy, with the result that details
of houses that would ensure their function, performance, or ongoing maintenance
(the very attributes that would transform the image and words about a house into
actual housing) are among the ﬁrst corners to be cut. With the specter of disor-
derly householders as backdrop and, lately, accusations of molestation and pedo-
philia behind closed doors in indigenous communities,
clear and compelling
policy notions such as “affordable,” “cost-effective,” “sustainable,” and “tenant
responsibility” percolate in endlessly variable combination in policy debates.
These helpful logics redirect attention from what are straightforward cost-cutting
agendas in welfare spending based on expediency (ﬁg. 9).
Importantly, the pressure to have more and therefore cheaper houses has mul-
tiple sources: from politicians who need to announce numbers by way of answer
to accusations about the continuing dire state of indigenous living conditions, to
indigenous community leaders who desperately need houses to meet the very real
problems of overcrowding. But reducing the cost of new houses to build more
11. Citing child sexual abuse and rampant violence against women as cause, in June 2007 the
former Australian government announced that the condition of indigenous people in the Northern
Territory constituted a national emergency and warranted intervention. For an account of these ini-
tiatives at the time of their unfurling, see Altman and Hinkson 2007; also Toohey 2008.
houses within strict budgets often reduces the money spent on items necessary
for long-term house function. Counterintuitively, at least at ﬁrst blush, increasing
house numbers but not housing function makes little overall difference to over-
crowding, for people have the inexplicable habit of “crowding” into the house that
has functioning electricity, water, and efﬂuent disposal.
As we have seen, common reductions (for houses that authorities know from
their repeated surveys will have to accommodate more people than “the Aus-
tralian average”) include little or no insulation; smaller and lower-capital-cost
hot water systems; lower-quality faucets and reduced door and window quality;
fewer and lower-quality light ﬁxtures; fewer electrical outlets; no landscaping,
12. This pattern of moving to the functioning house should put to rest the insistence of cultural
literalists that remote indigenous people might not want houses. Yes, there are culturally distinctive
uses of domiciliary space, which typically involve “diurnal/nocturnal behavior patterns for different
seasonal periods . . . external orientation and sensory communication between domicile, sleeping
behavior, cooking behavior and use of hearths” (Memmott 2004: 46), but these different sociospatial
patterns do not exclude a desire for functional hardware. That is, people can cook and sleep outside
and still desire a secure shelter, a stove that works, a toilet that ﬂushes, walls that stay up, and ﬂoors
that do not fray.
Figure 9 Houses to nonhouses. Image by Paul Pholeros
This Is Not a Pipe
site drainage, or fencing; smaller verandas; and, per-
haps most critically, fewer inspections of the works dur-
ing construction. Each of these abridgments has serial
entailments, including housing hardware failure. The
expedient solution of smaller hot water systems, for
instance, not only promotes less frequent showers and
a consequent reduction of health beneﬁt, facilitating the
spread of infectious and parasitic diseases (Bailie 2006).
It also entails higher running costs for a hot water sup-
ply in extended households, the greater likelihood that a
poor family cannot pay the ensuing electricity charges,
and the related likelihood that the power will be dis-
connected. In other words, when data on failure rates
of inadequate hot water systems are linked to forms of
householder stress, the burden of cost shifting sharpens
End-of-ﬁnancial-year dollars that need to be quickly
spent add other distortions (ﬁg. 10).
Even the ostensibly reasonable proposition that new
houses are the answer to the overcrowding and immis-
eration replayed in scandalizing images of indigenous
housing invokes the premise that new houses perform
better than old houses, which in turn implies that new-
ness equates with quality design, a (nonvalidated) belief
as visual as it is commonsensical. Yet while much indig-
enous housing is old and cruddy, there is no clear evi-
dence that new housing performs better than old. Nor
is it necessarily the case that more or better-informed
consultation at the design stage will improve the perfor-
mance of the houses, another popular catch cry. While
this noble idea might be argued for on other grounds,
it is not an explanation for housing breakdown (see also
Dovey 2000). The houses do not necessarily fail simply
13. For a related case analysis of the impact of “utility stress” in indigenous Australia — that is,
the hardship that accrues when the radically poor are unable to pay electricity, telephone, water, gas,
or other utility charges — see Willis et al. 2006.
14. While “more or better-informed consultation at the design stage” does not necessarily lead
to better performance, the residents are blamed for poor performance and have no forum in which
Figure 10 Being strategic, efcient, and coordinated.
Image by Paul Pholeros
because they fail forms of cultural distinction. They most often fail because of
poor initial construction and lack of routine maintenance, a dirty materialism that
seems the hardest of things to see.
As Magritte’s work reminds us, seeing is an act, “in the course of which it can
happen that a subject escapes our attention” (Gablik 1976: 13). So it is that pipes
may be laid but not connected, while policy words reassert that indigenous hous-
ing is the nation’s foremost national priority, and other words remind us who is
really to blame. There is a different kind of deception involved in a pipe that does
not function, of relevance to this essay. Magritte’s image does not reveal whether
his model for the pipe had a cavity through the stem to the bowl, or, more exactly,
no mouth. Yet for the original shock to have effect, function needs to be imagi-
natively imputed to an original material form. This implicit contrast subtends the
reference to the trickery of the image. The image of indigenous housing also con-
jures an imaginary referent, and similarly, the functionality of the original mate-
rial form is presumed. The aesthetic orientation carries with it a heavily loaded
set of responses concerning the meaningfulness of social life itself. It is a morality
that apprehends a certain unruliness in indigenous domiciles and seeks either to
describe the values of the imagined cultural alterity, with its seeming rejection of
the middle-class house fetish (as liberally inclined anthropologists might), or to
amend that imputed social world (as policy makers might). Either way, these hab-
its of association at work in the verdicts surrounding indigenous housing neglect
as much as they act on the material problems at hand.
The newly demoted opposition spokesman for indigenous affairs, Tony Abbot,
responding to the Rudd government’s calls for bipartisan action on the scandal of
indigenous housing, reasserts the loop simply: “There is no doubt that indigenous
housing is very poor quality, but there’s also no doubt that there’s many indig-
enous people who don’t respect the housing they have” (Karvelas and Kearney
2008: 7). Abbot is referring, of course, to those bad householders who do not
deserve or even need well-constructed shelters because of their myriad shortcom-
ings. These include the relativizing anthropologized verdict that indigenous mobs
were once hunter-gatherers who not so long ago roamed and slept in small kin
groups and thus have not evolved the right cultural norms to habitually manage
to defend themselves. House images (which conjure householder images) are powerful, easily trans-
muted, and autoproduced; they also function out of context and are voiceless. While images are
always a factor of concealment, householder voices might tell a truth that even data cannot.
This Is Not a Pipe
new material forms. We have argued that a different regime of cultural expec-
tations and habits of mind is at stake and that while it lies adjacent to, it is not
reducible to, indigeneity. Put simply, the habitual and entrenched insistence on
the centrality of (imagined or actual) Aboriginal habits to the endemic housing
problems in indigenous Australia, whether manifest as nostalgia for notions of
culturally appropriate design, justiﬁcation for a national intervention, or anger at
what appears a rejected public beneﬁt, obeys a “what you see is that” effect.
Our recourse to surrealism hints at the confounding principles that need to be
reckoned with. The service world inhabited by indigenous householders is one
that operates in deﬁance of rational analysis. A Kafkaesque universe of mad hap-
penings and unbearable logics is being brought to bear on fundamentally old-
fashioned issues of inequality, poverty, and race, with a dollop of policy ineptitude
for good measure. As sewer pipes struggle to carry away wastewater as a result
of poor design, indifferent (or no) inspection, and allowed shoddy construction,
the pipeline or conduit for myths, rumors, and shock from indigenous Australia to
the mainstream population ﬂows without even the coarsest ﬁlter of critical assess-
ment or hard evidence. There may well be no, some, or many unruly indigenous
householders, but this is beside the point. When the technocratic solutions are in
fact the original form of delinquency, we might turn to surrealism’s concern with
reaching beyond registrations of the real to resee what is directly in front of us
(see ﬁg. 11).
Altman, Jon, and Melinda Hinkson, eds. 2007. Coercive reconciliation: Stabilise,
normalise, exit Aboriginal Australia. Sydney: Arena Publications.
Australian Institute of Health and Welfare (AIHW). 2005. Indigenous housing
needs 2005: A multimeasure needs model. AIHW catalog no. HOU 129. Can-
berra: Australian Institute of Health and Welfare.
Bailie, Ross. 2006. Housing and health in indigenous communities: Key issues
for housing and health improvement in remote Aboriginal and Torres Strait
Islander communities. Australian Journal of Rural Health 14: 178.
Barreto, Mauricio, Bernd Genser, Agostino Strina, Maria Gloria Teixeira, Ana
Marlucia Assis, Rita F. Rego, Carlos A. Teles, Matildes S. Prado, Sheila M. A.
Matos, Darci N. Santos, Lenaldo A. dos Santos, and Sandy Cairncross. 2007.
Effect of city-wide sanitation programme on reduction in rate of childhood
diarrhoea in northeast Brazil. Lancet 370: 1622–28.
Cape York Institute for Policy and Leadership, and Cape York Partnerships Proj-
Figure 11 With further
apologies to Magritte.
Image by Paul Pholeros
ects. 2007. From hand out to hand up: Cape York Welfare Reform Project;
Aurukun, Coen, Hope Vale, Mossman Gorge; Design recommendations, www
Report.pdf (accessed October 11, 2009).
Cohen, Jon. 2008. Pipe dreams come true. Science, no. 319: 745–46.
Department of Families, Community Services and Indigenous Affairs (DFCSIA).
2007. National indigenous housing guide: Improving the living environment
for safety, health, and sustainability. Vol. 3. Canberra: Australian Government
Dovey, Kim. 2000. Myth and media: Constructing Aboriginal architecture. Jour-
nal of Architectural Education 54: 2–6.
Feldman, Allen. 2001. Philoctetes revisited: White public space and the political
geography of public safety. Social Text, no. 68: 57–89.
Foucault, Michel. 1983. This is not a pipe, edited and translated by James Hark-
ness. Berkeley: University of California Press.
Gablik, Suzi. 1976. Magritte. Boston: New York Graphic Society.
Hall, Jon, and Mike Berry. 2004. Indigenous housing: Assessing the long-term
costs and the optimal balance between recurrent and capital expenditure.
Australian Housing and Urban Research Institute, Royal Melbourne Institute
of Technology–National Centre for Social and Economic Modelling Research
Centre. Final Report no. 93.
Hughes, Helen. 2005. Policies entrench poverty. Australian, September 23.
———. 2007. Lands of shame: Aboriginal and Torres Strait Islander “home-
lands” in transition. Sydney: Federation Press.
Karvelas, Patricia, and Simon Kearney. 2008. Division hits “war cabinet.” Aus-
tralian, February 15.
Latour, Bruno. 2002. ARAMIS; or, The love of technology, translated by Cath-
erine Porter. Cambridge, Mass.: Harvard University Press.
Lea, Tess. 2008a. Bureaucrats and bleeding hearts: Indigenous health in North-
ern Australia. Sydney: University of New South Wales Press.
———. 2008b. Housing for Health in indigenous Australia: Driving change when
research and policy are part of the problem. Human Organization 67: 77–85.
Lee, Gina, and David Morris. 2005. Best practice models for effective consulta-
tion: Towards improving built environment outcomes for remote indigenous
communities. Australian Housing and Urban Research Institute, Southern
Research Centre, April. Final Report no. 76.
This Is Not a Pipe
Levi, Primo. 2000. The periodic table, translated by Raymond Rosenthal. Lon-
Memmott, Paul. 2004. Aboriginal housing: Has the state of the art improved?
Architecture Australia 93: 46–48.
Morgan, George. 2006. Unsettled places: Aboriginal people and urbanisation in
New South Wales. Kent Town, South Australia: Wakeﬁeld.
Pholeros, Paul, Stephan Rainow, and Paul Torzillo. 1993. Housing for health:
Towards a healthy living environment for Aboriginal Australia. Newport
Beach, Sydney: Healthabitat.
Pholeros, Paul, Paul Torzillo, and Stephan Rainow. 2000. Housing for Health:
Principles and projects, South Australia, Northern Territory, and Queensland,
1985–97. In Settlement: A history of Australian indigenous housing, edited by
Peter Read. Canberra: Aboriginal Studies Press.
———. 2004. Healthabitat: A brief history. Newport Beach, Sydney: Healthabi-
SGS Economics and Planning and Tallegalla Consultants Pty Ltd. 2006. Evalua-
tion of ﬁxing houses for better health projects 2, 3 and 4. Australian Govern-
ment Department of Families, Community Services and Indigenous Affairs
Occasional Paper no. 14.
Sutton, Peter. 2007. Mainstreaming: Equal opportunity or discrimination? Paper
presented at the Royal Australian Institute of Architects conference “Which
Way? Directions in Indigenous Housing,” Alice Springs, Northern Territory,
Australia, October 26–27.
Toohey, Paul. 2008. Last drinks: The impact of the Northern Territory interven-
tion. Quarterly Essay 30: 1–97.
Torzillo, Paul, Paul Pholeros, Stephan Rainow, Geoffrey Barker, Andrew Irvine,
Tim Short, and Tim Sowerbutts. 2008. The state of health hardware in aborigi-
nal communities in rural and remote Australia. Australian and New Zealand
Journal of Public Health 32: 7–11.
Willis, Eileen, Meryl Pearce, Carmel McCarthy, Tom Jenkin, and Fiona Ryan.
2006. Utility stress as a social determinant of health: Exploring the links in
a remote Aboriginal community. Health Promotion Journal of Australia 17: