Anthropological Theory

Copyright © 2010 SAGE Publications

(Los Angeles, London, New Delhi,
Singapore and Washington DC)
http://ant.sagepub.com
Vol 10(1–2): 171–178
10.1177/1463499610365383

Number as an inventive
frontier in knowing and
working Australia’s
water resources
Helen Verran
University of Melbourne, Australia

Abstract
Taking number as material and semiotic, this article considers the enumeration of
Australia’s water resources as both a form of audit and a form of marketing. It
proposes that a scientific enumeration utilizes the relation one/many while an
economic enumeration utilizes the relation whole/parts. Working the tension between
these two forms of enumeration can be understood as an inventive frontier in
contemporary Australian life.

Key Words
number as relation • water resources

INTRODUCTION: NUMBERS AS MATERIALIZED RELATIONS

We are used to thinking of numbers used in enumerating the real world as helping us to
work relations – our relations with goods through the market, or with nature through
science for example. This sort of thinking proposes numbers, and mathematical objects
more generally, as cognitive tools. Learning to use mathematical tools is taken as one of the
most important elements of schooling. In my contribution to this collection considering
number as an inventive frontier in social, cultural, political and moral life – as an object of
anthropological interest, I take numbers rather differently: as materialized relations.
To understand this move we can imagine it by analogy to the differing senses of people
having (kin) relations and being relations. Each of us has relations – a mother and a
father for example. Also each of us is a relation in an embodied sense, the relation
between our parents. This sense of a person being a relation is both material in the sense
of being embodied, and semiotic in the sense of expressing a formal relation – that
between husband and wife. Taking numbers as material is to see them as enumerated
materiality – as river water of a specified ecological value for example; taking numbers
as semiotic is to identify numbers as the formal relation unity/plurality.

171

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015

 ANTHROPOLOGICAL THEORY 10(1–2)

Rendering numbers in this way enables a vivid picture of numbers as an inventive

frontier. It sees them as particulars, in place and time, in situ we might say – materi-
alized; realized in specific practical ways. And significantly from an anthropological
point of view, this way of rendering numbers makes them accessible to ethnography.
Indeed, I learned to recognize numbers as embodied materiality and semiotically as
relations by working ethnographically with Yoruba (West African) and Yolngu (Aborig-
inal Australian) teachers and their pupils as they struggled to learn and teach how to
work with numbers. This sort of working with numbers is treated more fully by Jean
Lave. I understand Lave as reflexively engaging the contrasts I make in this paper
(Lave, 2010).
Having numbers as ethnographically accessible in adopting such a material semiotic
framing eschews the more extravagant claims about what numbers are that we find, for
example, in Platonist, intuitionist, or instrumentalist accounts of numbers (Verran,
2001: 178). However, semiotic is an opaque term, being simultaneously both vague and
highly technical, and combining it with material modifies the meanings others attribute
to it. Semiotic can be read as summoning up the baroque complexity of French struc-
turalist and post-structuralist thought, and alternatively the specific categorical propos-
als of Peirce’s philosophy and its offspring, American Pragmatism. Neither of these
discourses is a primary reference in my use of semiotic, however. While not repudiating
these connections, my approach does not use them in justifying my methods. Having
been helped to see some of the insights generally associated with semiotics in the Western
tradition of thought by my Yoruba and Yolngu friends, I found the work of the later
Wittgenstein more helpful in articulating an account of how numbers are ethnograph-
ically accessible (see Verran, 2001: 179).

His attack upon the idea of a private language, which brought thought [and number]
out of the grotto in the head into the public square where one could look at it once
it arrived there – as a set of practices . . . seem[s] almost custom designed to enable
the sort of anthropological studies I, and others of my ilk, do. (Geertz, 2000: xi)

Studying ‘forms of life’ and ‘language games’ as complex clots of signs and collective
actions accepts the paradox of worlds as already/always meaningful, recognizing that
doing worlds as knowable, whether by science or by trading, or for that matter through
Yoruba, or Yolngu knowledge traditions, involves engaging with the world as it is here
and now. And in doing this I suggest that ethnographers of numbers can usefully adopt
Peirce’s typology of the working of signs – as iconic, indexical, or symbolic. Describing
the here-now workings of signs vis-à-vis the collective actions in which the objects asso-
ciated with those signs come to life, these categories are not mutually exclusive (see
Hoopes, 1991: 239). The terms name degrees of reciprocal co-constitution of signs and
collective embodied and embedded actions in which objects come to life. Icons are
deeply co-constitutive with clots of collective actions that generate entities; whereas
symbols and their objects enact a relation of supervenience, objects are accepted as affect-
ing and effecting their signs but not vice-versa. Indexes name the position between these
two. Iconicity, indexicality and symbolism can be understood as ethnographically found
forms of the workings of signs (and depending on how one understands ethnography,
these names work as icons, indexes, or symbols!).

172

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015

VERRAN Number as an inventive frontier in knowing and working Australia’s water resources

In anthropology numbers are for the most part treated ethnographically as symbols
(Crump, 1990; Mimica, 1988, for example). Although as Rotman shows for zero, using
Peirce’s typology they can equally be understood as iconic (Rotman, 1987), and as Mauer
explains in his paper in this collection on a hypothetical ‘finger counting money’ (Mauer,
2010), the numbers of finger counting are most usefully understood as iconic and index-
ical rather than as symbols. One way to understand what I accomplish in Science and an
African Logic is showing that while rendering number as symbols can accomplish certain
important ends like showing the conceptual equivalence of Yoruba and scientific number
and removing the stigma of Yoruba number as primitive, numbers can also be usefully
understood as icons, and that doing so we can learn how to connect Yoruba and
scientific numbers in practice (Verran, 2001).
In this short paper I begin with a situation in which we can understand numbers as
working indexically or symbolically. They are being used to represent the ecological
health of Australia’s creeks and rivers, lakes and billabongs. With an imperceptible shift
such numbers can be used iconically to constitute a water market. This is the context in
which I explore what happens when we understand numbers as materially expressing
formal relations in their generalizing capacities. Refusing to restrict their semiotic
repertoire to that of symbols allows us to recognize that in working markets numbers
enact generalizing iconically in the whole/parts mode (see Guyer, 2010).
In beginning this paper I proposed an analogy between numbers used in enumerat-
ing being taken as materially embodying relations and taking humans as embodying the
relation between their parents. We can immediately recognize that saying that I am
taking a person as a relation does not provide much information about the person, and
that startling as the analogy might seem, we can similarly recognize that saying numbers
are a materialized relation, viz. the unity/plurality relation, does not tell us much about
numbers. Perhaps the first thing we might say in characterizing ourselves is whether we
are female or male. Similarly the first thing we need to know about a number is whether
it takes the one/many form or the whole/parts form. One way to think about the
difference between these relational forms is to characterize one/many as potentially
containing unity within the plurality of a many, and whole/parts as having plurality
contained within a unity. This is a vague (but, as we shall see, a useful) way of asking
whether a number is a cardinal number and works to conserve value by working the
one/many form, or an ordinal number working to conserve order through the
whole/parts form.
The fact that the last couple of sentences are likely to have puzzled many of my readers
alerts us to the fact that thinking about number as materialized relation is novel. Just as
we need more detail in characterizing a person – I have my mother’s chin, my father’s
temperament – so it is with numbers if we are to see how it works as an inventive frontier.
In this short paper I show how in shape-shifting between the form one/many and
whole/parts, and in moving between symbolic/indexical and indexical/iconic modes of
semiosis, number works in inventive ways in mediating a frontier embedded in knowing
and working Australia’s water resources as enumerated.

173

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015

 ANTHROPOLOGICAL THEORY 10(1–2)

SCIENTIFIC ENUMERATION: AUDITING THE STATE OF

AUSTRALIA’S WATER
At 2 pm on the first Sunday of every month Waterwatch Victoria volunteers gather to
test the water quality of Merri Creek on the upstream side of St George’s Road Bridge
in Melbourne’s inner city suburb of North Fitzroy. Their ‘kit’, donated by a sponsoring
water company, has various sized tubes and bottles (some empty, a few already equipped
with reagents, conductivity and pH probes), plastic spoons and bowls, ice cube trays and
plastic syringes. Usually by 3.30 pm a nice set of numbers has been assembled. The
quality of the creek’s water emerges as a complicated arithmetical composite of
aggregated numbers generated in chemical and physical tests set alongside another
complicated arithmetic composite representing aquatic life and derived from estimates
of numbers and types of ‘bugs’ counted in a sample of sludge. Later in the week these
numbers are added to the Waterwatch Victoria database by the group’s coordinator.
Begun in 1993, Waterwatch is an environmental NGO providing services mainly
under contract to governments. It now has many thousands of Australians regularly
attending their streams with bottles, thermometers, and pH meters, peering at tiny
creatures they have scooped up with a net, trying to identify what they are and count
their numbers. In part Waterwatch aims to sensitize citizens to the failing health of our
nation’s rivers. But more substantively and certainly of significance to the people who
do the measuring, the numbers that Waterwatch volunteers produce ‘fill-in spatial and
temporal monitoring gaps’, contributing to Australia’s ongoing environmental audit. In
Victoria there are a mere 270 official water monitoring sites; in contrast, Waterwatch
Victoria collects data from 1454 places.
Animated by the slogan ‘you can’t sustain what you haven’t measured’, the enthusi-
asm that pervades Waterwatch evidences Australians’ new found commitment to eco-
logical sustainability in the face of widespread and ongoing drought. What is much less
evident is the way these numbers so enthusiastically generated by volunteers with the
best of intentions towards Australia’s nature might contribute to constituting water as
commodity, expanding possibilities for ‘doing business with water’.
To the chagrin of Waterwatch Victoria’s thousands of volunteers who go out of their
way to wade about in streams and lakes and understand their labour as generating
important information, the numbers they generate in attesting water quality are not
added to the official register of water quality in Victoria: the Victorian Water Resources
Data Warehouse. That database contains only ‘official data’ gathered by personnel
employed and supervised by state instrumentalities and water companies. Why the
cordon sanitaire? Surely two databases, separate institutions and websites is counter-
productive in a project assembling and disseminating information on the state of the
nation’s water resources.
Unlike the ‘official data’, the confidence limits of the ‘community data’ are not speci-
fiable. While it has managed to develop a network of co-ordinators providing training
in water quality and biological monitoring, the measuring activities of Waterwatch’s
various volunteer groups are still murky. The provenance of numbers arising in their
unreliably disciplined gestures with hands and eyes, probes and tubes, buckets and nets,
words spoken and figures recorded, cannot be guaranteed by the institutional location
nor by attested skill levels of the number generators. These people who in their spare-
time muck about in gum-boots and waders in dams and billabongs, creeks and rivers,

174

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015

VERRAN Number as an inventive frontier in knowing and working Australia’s water resources

do not verifiably possess at least the Level III Community Environment Certificate. The
numbers they record may cohere well enough to give an alarming general picture of the
state of Australia’s waters, but their consistency is not quantifiable.
Nevertheless, the expressed hope is that when the level of discipline embedded in the
hands and eyes of Waterwatch volunteers can be reliably witnessed and quantified, the
data sets will be consolidated. Then the non-equivalence between the two data sets,
currently so carefully maintained, will disappear. If and when this does happen, a
peephole will close up – a peephole that currently allows us to see some of the inner
workings of number as it slips between alternative forms of enumerating.

ECONOMIC ENUMERATION: DERIVING AUSTRALIAN WATER

PRODUCTS
What does this (perhaps) soon to be closed up peephole allow us to see? Among other
things it reveals the wonder of numbers’ working as the relation unity/plurality. The
mis-match between the provenances of ‘community water data’ and ‘official water data’
enables us to see that numbering projects like the Victorian Water Resources Data
Warehouse hold two distinct moments of numbering in tension. The moments articu-
late distinct social ends and express the dual forms of numbers’ remarkable capacities in
enumeration – to develop relations between sameness and difference by working the
relation unity and plurality.
We see that while appearing merely as a means of assembling information, number-
ing Australia’s water resources serves two separate and distinct purposes: first, environ-
mental monitoring which we can understand as a form of cadastral accounting
characterized by a moment of audit and achieved in arithmetic calculation; second,
constitution of an archive to constitute an emergent Australian water market and the
possibility of the complex calculation that far exceeds the simple arithmetic involved in
cadastral numbering.
Constituting an archive to allow for the working of the complex calculative regimes
of the market, the Victorian Water Resources Data Warehouse has to be particular about
confidence limits. It must manage risks and provide data only of quantifiable prove-
nance, for much is at stake in the economic-social-political-moral project of constitut-
ing and developing an Australian water market. Per capita, for each Australian, the
amount of water extracted from the environment and held in storage far exceeds that
for citizens of other nations. Australia has in the past invested vast sums, and in today’s
world this past obsession with storing water is evident as infrastructure which translates
water’s use value into exchange value. Extracted and stored water is a potential commod-
ity. Extracted, stored and audited water is capable of realization as a commodity. Through
its trade, marginal gains and capital gains can be realized.
The Victorian Water Resources Data Warehouse invisibly manages dual moments of
numbering Australia’s water resources: on the one hand registering Victoria’s water
resources with quantifiable accuracy, and on the other contributing to expanding the
markets that exploit those resources. To appreciate the flexibility embedded in the sets
of numbers that populate the Victorian Water Resources Data Warehouse databases, and
to understand why they must be kept separate from the numbers in the Waterwatch
volunteers’ database, we need to remember that number’s remarkable capacities lie in its
being the relation unity/plurality, a relation that might be expressed as the relation

175

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015

 ANTHROPOLOGICAL THEORY 10(1–2)

between one and many, or alternatively as the relation between a whole and its parts.
Working number as the relation one/many expresses one form of numbers’ generalizing;
working it as the relation whole/parts mobilizes another.
On the one hand, the numbers that constitute the Victorian Water Resources Data
Warehouse are about knowing the state of Australia’s natural resources. This purpose
depends on numbers being the relation one to many: single instances of defined
measurement are cumulated, and numerals are taken to represent enumerated abstract
entities. This is the form of generalizing that the Waterwatch volunteers’ numbers effect.
Simultaneously, the numbers that constitute the Victorian Water Resources Data
Warehouse make a whole – a marketplace for Australia’s emergent water trade. This
whole, ‘articulable water resources’, is the water market. Because the confidence limits
of the Waterwatch numbers are not quantified, those numbers cannot mix-up with the
official numbers in contributing to generating the water market. The water market has
many parts: surface water, ground water, return water, and increasingly reuse water; and
each of these parts has manifold sub-parts. These parts and sub-parts are not given but
continually proliferating as new, derived configurations which are ingeniously designed,
while other configurations wither, or are killed off. The whole, the Australian water
market, is necessarily vague and emergent, but the numbering on which it depends must
have quantified and specified confidence limits.

COMPARING AND CONTRASTING ONE/MANY AND WHOLE/PARTS

GENERALIZING
Let me consider in more detail the contrast between the forms of enumerating that the
quarantining of the Waterwatch numbers alerts us to. The numbers representing water
quality for the purpose of exhaustive environmental audit work the relation one/many;
semiotically they are symbols. As part of a precise, specifying and definitive process, both
the official numbers of the Victorian Water Resources Data Warehouse and the commu-
nity numbers of the Waterwatch programme contribute to registering the quality of
located water in all Australia’s catchments. Ones – specified units of flowing or stored
water, with specified space-time co-ordinates and specified physical, chemical and
biological properties – are collected together as a many. They add up to general picture
of the states and places of Australia’s waters. In some ways a fantasy of exhaustive cumu-
lation, the purpose of the audit project is nevertheless identified as a common good
endeavour: developing better representations of the evidently deteriorating state of the
waters that are crucial in sustaining Australia’s nature.
Yet sets of consistent ‘official’ numbers stored in the likes of the Victorian Water
Resources Data Warehouse, soon (perhaps) to be supplemented by the vast stores of
numbers that Waterwatchers produce, also serve quite different ends. Another fantasy
of quite a different type is also sustained by the existence of these numbers: the vague,
emergent and unspecifiable notion of ’ the ‘Australian water market’. Here the sets of
numbers are no longer a representation but are now constitutive of the entity itself. They
work as icons.
The wholeness of the Australian water market emerges as its continually re-invented
and enumerated parts. Water commodities are conjured into existence by various
ingenious means of bundling, tracking, partitioning, and so on. These parts of the whole
are literally made feasible by the cumulated numbers. Surface water, ground water, return

176

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015

VERRAN Number as an inventive frontier in knowing and working Australia’s water resources

water, and reuse water are phenomena generated in disciplined interrogation of

knowable water in place. Each of these parts of the water market has its sub-parts, each
is made separable as a product, a commodity, by the social, literary, and material tech-
nologies through which it is born. Ingenious social and technical tinkering – separating
water entitlements from land rights, rebundling water as low, medium, and high security
licences, inventing tamper-proof flow meters – is continually coming up with new sorts
of parts in the vague emergent whole of Australia’s water market: multiple water products
designed for multiple markets. Here the waters and the numbers are one and the same,
equally crucial in constituting the entity, the commodity. The work, the human labour
that constitutes the water market’s parts as socio-material entities, becomes invisible; or
it would if only the confidence limits of the ‘community water data’ were specifiable and
the community numbers did not need to be quarantined.
Recognizing the origins of numbers’ capacities in carrying human endeavours is
important, for the ease with which they enable one project to become quite another is
uncanny. We see number working as inventive frontier in the shape-shifting evinced by
the official numbers of the Victorian Water Resources Data Warehouse. Numbers,
coming to life originally as rule-bound sequences of words, are radically incomplete. In
use they are both agile, a property deriving from their rule-boundedness, and needy, a
mode bequeathed by their origins in the patterns of events, not in the events themselves.
In this combination of needy agility is found numbers’ unnerving capacity to continu-
ally evert themselves, flipping imperceptibly from their one/many manifestation to their
whole/parts form of working, shifting between signing as symbols and signing as icons.
In this lies numbers’ capacities to simultaneously and seamlessly work relations between
sameness and difference and unity and plurality, and to dissemble – eliding projects with
very different moral and political resonances.
A one, invoked as a defined unit, can be collected together with many other similar
units. This plurality can with an almost imperceptible alteration in the criterion of
completion invoke a whole. The cumulated defined units of water in many linked loca-
tions can with a self-evidently useful alteration in the defining criterion become ‘regu-
lated water’. The usefulness of the slide obscures the flip from cumulus to whole, and
more significantly the change in numbers’ role from representation to constitution. The
elision accomplished, however, parts can be evoked utilizing differential completing
criteria: ‘dam water bodies’ evoked in one criterion can be the subject of trade with land-
holders; ‘regulated river flows’ evoked through another criterion can be subject to trade
as environmental flows, and ‘weir pools’, though yet another criterion might allow trade
in high security water licences.
If not for the glitch that the numbers so earnestly recorded by Waterwatch volunteers
do not have specifiable confidence limits because of the unreliability of the volunteers’
gestures with hands and eyes, tubes and meters, words and figures which generate those
numbers, the elision of a many – numerous observations/measurements, to a whole –
Australia’s tradeable water resources, would be less visible. The conflation of a project
whose explicit purpose is for the public good – to get a better picture of the rapidly dete-
riorating state of the waters of Australia’s rivers and creeks, swamps and lakes – with a
project whose stated purpose is to trade those waters and enable a few to reap private
capital gain from that trade would be more seamless.

177

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015

 ANTHROPOLOGICAL THEORY 10(1–2)

Acknowledgements
I am indebted to Christian Clark. The ethnographic work from which this paper grew
was carried out by Christian as part of his undergraduate honours year project which I
supervised. The responses of participants in the number as an inventive frontier
workshop were helpful, I am grateful to the organizers and funders of that event. I
acknowledge in particular a fruitful conversation over many years with Jane Guyer
around the many puzzling aspects of West African numbers and markets.

References
Crump, T. (1990) The Anthropology of Numbers. Cambridge: Cambridge University
Press.
Geertz, Clifford (2000) Available Light: Anthropological Reflections on Philosophical
Topics. Princeton: Princeton University Press.
Guyer, Jane (2010) ‘The Eruption of Tradition? On Ordinality and Calculations’,
Anthropological Theory 10(1).
Hoopes, James (ed.) (1991) Peirce on Signs: Writings on Semiotic by Charles Sanders
Peirce. Chapel Hill: University of North Carolina Press.
Lave, Jean (2010) ‘Math Lessons from Liberia’, Anthropological Theory 10(1).
Mauer, Bill (2010) ‘Finger Counting Money’, Anthropological Theory 10(1).
Mimica, Jadran (1988) Intimations of Infinity: Mythopoeia of Iqwaye Counting System
and Number. Berg: Oxford.
Rotman, Brian (1987) Signifying Nothing: The Semiotics of Zero. London: Macmillan
Press.
Verran, Helen (2001) Science and an African Logic. Chicago: University of Chicago
Press.

HELEN VERRAN is a Reader in History and Philosophy of Science at the University of Melbourne. Her
interest in numbers began in Nigeria in the 1980s when working at Obafemi Awolowo University. She used
differences in numbers as a way to think about difference differently in writing her book Science and an African
Logic. She is currently working on a book called Numbers and Nature where she uses numbers, and differences
in numbers, in order to rethink the concept of nature. Address: University of Melbourne, Rm 204, Old Quad,
VIC 3010, Australia. [email: [email protected]]

178

Downloaded from ant.sagepub.com at UNIV OF CONNECTICUT on May 14, 2015