Advances in Applied Sociology
2012. Vol.2, No.3, 203-213
Published Online September 2012 in SciRes (
Copyright © 2012 SciRe s . 203
ANT Analysis on the River Management of Urban Taiwan:
A Case Study of Keelung River in Taipei Basin
Tsu-Lung Chou
Graduate Institute of Urban Planning, National Taipei University, New Taipei City, Chinese Taipei
Received April 24th, 2012; revised May 27th, 2012; accepted June 10th, 2012
This article provides a historical investigation on river management in urban Taiwan. It demonstrates the
application of actor network theory (ANT) on the long-term development and evolution of river manage-
ment network through a case study of the Keelung River in Taipei Basin. The study considers the forma-
tion of the actors’ world of the Keelung River and the way the actors have affected flood drainage and the
supply of land in the region. River management and induced disasters in the Basin are explored. Strengths
and weaknesses of the ANT are also discussed.
Keywords: Actor Network Theory; River Management; Keelung River Watershed
Taiwan has 129 rivers, mostly existing side by side with cit-
ies. The natural processes of the river change have significantly
affected Taiwan’s urban development that goes beyond recur-
rent flooding of cities. However, growing literature on the riv-
ers has focused primarily on flood prevention, river pollution,
and hydraulic engineering (e.g. Chen et al., 2001; Hsu et al.,
2003; Huang et al., 2003; Lo et al., 1996; Lu et al., 2001). In
such studies, analyses are limited to technological perspectives,
placing little emphasis on the relationships of natural river with
human management that are attracting growing research inter-
ests recently (Eden et al., 2000; Kelman, 2003; Kortelainen,
1999; Pfadenhauer, 2001; Ryan, 1998; van Diggelen et al.,
2001). In this paper, I point out that a distinct typology of rive
management has emerged in Taipei where the Keelung River
meanders through the city that has had major impact on urban
development and environment. I argue that for individuals in
Taipei whose life has been profoundly affected by the Keelung
River, a distinct river management has emerged that encom-
passes a distinctive way of life with shared values and mean-
ings centered on the river, strategies and collective behaviors on
the part of selected groups of individuals in coping with the
problems of frequent flooding, as well as close relationships
among the key actors involved in the interactions between na-
ture and society, or more specifically, between the Keelung
River and the city of Taipei.
The analytical framework adopted here is the Actor Network
Theory (ANT). Developed by sociologists of science including
Bruno Latour, Michel Callon and John Law (Callon, 1987;
Callon et al., 1986; Latour, 1987; Law, 1991), the ANT has
been extended beyond sociology to the analysis of geographical
and environmental issues (e.g. Kortelainen, 1999; Burgess et al.,
2000; Comber et al., 2003; Eden et al., 2000; Murdoch, 1995,
1997a, b, 1998, 2001). Its analytical framework is rooted in
three central themes: heterogeneity, symmetry and impartiality.
First, heterogeneous associations emphasize actor associations
that can bring together the social and the physical/material to
overcome the traditional dualisms of nature/society and hu-
man/nonhuman (Murdoch, 1997a). It is argued that actors can
join together and able to transcend the binary interactive spec-
trums of local/global, cultural/natural, and social/technical to
form complex associations for mutual benefits. Secondly, the
idea of symmetry dictates that the natural and the social, or
non-human and human, are symmetrically treated as necessary
counterparts in the analyses of any society-nature relationship.
Finally, the ANT highlights analytical impartiality, in which the
analyst is required to be as undecided as the actors they follow
(Latour, 1987: pp. 175-176).
The ANT explores how human and nonhuman actors are
brought together in networks and formulated into associations
that expand across space and time. In this paper, the river man-
agement is identified as a relational product of the key actor
networks who direct the movement of such resources as devices,
technologies, materials and money and to persuade actants
related to or affected by the river to play the roles set by the
dominant actor(s). I focus on the processes of negotiation, rep-
resentation and displacement that affect the establishment of
relations between actors and entities (Murdoch, 1997b). As the
ANT emphasizes, powerful actors usually construct a network
project in an attempt to gain privileges of representation, to
speak for others and to impose particular definitions and roles
on them. In building its network, the actor gives each entity ‘an
identity, interests, a role to play, a course of action to follow,
and projects to carry out’ (Callon, 1986a). The dominant actor
decides the attributes of other actors, links them together, and
draws up the scenarios in which they will take part (Burgess et
al., 2000). It is through this transformation process that a net-
work is formed.
According to Callon (1986b), every transformation involves
a sequence of four moments: problematization, intressement,
enrolment and mobilization. As Woods (1997) summarizes, in
problematization, the potential translator defines other entities,
their goals and problems in a manner presenting the translator’s
objective as the solution to the other entities’ problems. In sub-
sequent intressement, the other entities are attracted into the
project by blocking other possible alignments. The roles of
actors are defined and distributed in enrolment stage. Mobiliza-
tion is the stage at which the representation is enacted. The
transformation may not necessarily engage all four moments,
and in reality, the mome nts may overlap.
Methods and Objects
The river management essentially involves the similar proc-
ess. Drawing on the above framework of ANT, the investiga-
tion attempts to reveal the river management in Keelung River
by exploring how the river was tamed, and its watershed trans-
formed to serve land supplier and flood drainage during inten-
sified urbanization process of Taipei basin. Why is the ANT
selected? With reference to many previous application exam-
ples (e.g. Callon, 1998a & b; Eden et al., 2000; Woods, 1997),
the ANT has two potentials and advantages for the analysis of
river management. First, the acceptance of heterogeneity and
symmetry allows for human institutions (urban community and
government) and non-human entities (river, flood disaster and
relevant prevention devices) to be incorporated into analyses,
overcoming the dualism between nature and society inherited in
mainstream social theories (for detailed discussions, see Mur-
doch, 1997a, b). Secondly, the attention paid to defining actors,
to identifying their long-term goals and to intressement and
mobilization allows for a relational investigation of the river
management in the network construction process.
This research is based on field studies made during a period
from April to June 2004 in Taipei City. Three group interviews
were conducted in Taipei City Government, which involved
selected 9 residents of Taipei City and 11 river managers and
planning officers. Additional 7 personal interviews were also
completed in Central Government, including selected directors
and officers from Ministry of Interior and Water Conservancy
Agency of Economic Affairs, and retired urban planners. The
paper is organized into the following sections. It begins with a
contextual exploration on urbanization developments in the
Keelung River watershed. Subsequently, the investigations
move to an ANT account on Keelung River management in
Taipei basin in which the river management is illustrated by a
discussion of how the government builds the actor world to
tame the river, and for functions as flood drainage and land
supplier in the Keelung-Taipei region. Furthermore, with ref-
erence to the Callon’s four stages framework of ANT, the actor
network of river management is examined by using two cases
of land developments in Taipei City. This is followed by inves-
tigating natural disasters as destructive actants of the original
network occurred, in which a detailed case study of Typhoon
Nari is explored to examine the river disaster and the related
responses of urban community and government. The conclu si on
summarizes the main findings of the study.
Urbanization in the Keelung River
The Taipei Basin is primarily formed by the Tamshui River
and its tributary the Keelung River (KR). The KR drains the
northeastern section in the mountainous northern Taiwan near
the sea. It flows through a narrow valley between the cities of
Taipei and Keelung, southwestward crossing the Taipei Basin
into the Tamshui River at Kwangtu. As one major tributary of
the Tamshui River, the KR is approximately 86 km and its
drainage basin covers 501 km2. The downstream portion of the
river is wide with a gentler slope (1/6700), while upstream is
narrow with a steep slope (1/250) (Huang et al., 2003). The
elevation of KR’s source region is about 500 meter, and in the
downstream it almost stays at the same elevation of 40 meter.
The gentle downstream slope and slow speed of water lead the
river after Nankang to meander in the Taipei Basin, creating a
wide expanse of flood plain there (see Figure 1).
In the earlier 1960s, the Special Assistance Fund in Metro-
politan and Urban Planning from the United Nations arranged a
Taiwan Strait
National Park
Conservation Area
Feitsui Reservoir
Old City
Trade Park
Pacific Ocean
Northern Hsuehshan Range
DownstreamMidstreamUpstreamKeelung River
40m/2100mm 230m/4000mm80m/3000mm
(Taipei Basin)
Pacific Ocean
Shezi Island
Figure 1.
Taipei-Keelung city region.
Copyright © 2012 SciRe s .
team of experts to conduct detailed studies on the metropolitan
development in the Taipei-Keelung region, and recommended
concrete programs for its future urban development (CIECD
1969b; Monson, 1964). As the director Donald Monson docu-
mented in the Regional Plan of Taipei-Keelung Metropolitan in
“Topography and the heavy rainfall accompanying ty-
phoons to which the Taipei Basin is subjected are the ba-
sic restraints which determine where urban development
can take place” (UHDC, 1968: p. 5).
This constraint thus led most of the earlier urban growth in
the Taipei Basin to centralize in Taipei City and its contiguous
suburbs primarily in the Tamshui River watershed (UHDC,
1968: pp. 19-21). Nevertheless, this does not imply that the KR
watershed was free from the urbanization impacts. Owing to its
location advantage adjacent to the Keelung seaport and Taipei,
industrial zones in the KR watershed had emerged as a signifi-
cant part of urbanization in the city region (see Figure 1). The
narrow valley connecting the Keelung harbor with Taipei city,
which provided level road and rail access to the Keelung City,
was also urbanized. The KR watershed had experienced exten-
sive industrial growth, in which industrial zones, cargo yards
and thousand of factories were scattered along the river valley.
Further urban development pressures towards the KR water-
shed had arisen since the late 1980s, because urbanization in
the Taipei Basin had reached its saturation point. The urbaniza-
tion pressures thus embarked on the Basin an ambitious plan of
land reclamation to create new space for industrial and urban
development. As a Japanese map in 1938 shows, areas of the
KR meandering across the Taipei Basin were primarily located
in Neihu area and used by farmers as pasture and arable land
(Taipei City Government, 1970: p. 95). By the 1990s, the me-
andering waterway in the Neihu area was artificially modified
into a straight river channel with embankment and 10 water
pumping stations. The river length in Taipei City was drasti-
cally reduced by 5.3 kilometers. More than 500 hectares of land
were created and transformed into Neihu Science Park for
high-tech industrial developments in Taipei. The land devel-
opment project of the KR was accordingly seen as the most
significant achievement for recent urban developments in
Taipei (see Huang, 2001: p. 118). Moreover, Shezi Island (323
hectares) and Kwangtu (924 hectares) in the downstream flood
plain have since the late 1990s faced increasing development
pressures to serve as the city ’s sub-center.
The urbanization process of Taipei Basin had also led to an
outward spread of land developments into the fringe townships
of the adjacent Taipei County especially in Hsinyi, Neihu, and
Nankang districts, and resulted in population explosion in Hsi-
chih by threefolds from 55,736 to 170,765 during the period
from 1976 to 2003.
The intensified urbanization thus resulted in dramatic land
use changes in the watershed. By the year 2000, industrial
zones totaling 1407 hectares had been developed in Wudoo,
Nankang and Neihu to meet the manufacturing investments.
The urbanized areas in the watershed amounted to 335,691
hectares, and its population reached about 3.24 million by 2003,
growing from 2.56 million in 1976. As Donald Monson (1964:
p. 5) has indicated, during the development of Taipei Basin in
the earlier 1960s, “except where urbanization has taken place,
the flat areas are intensively cultivated while the slopes and
mountains are largely wooded”. However, the intensified ur-
banization subsequently resulted in a dramatic change in its
land cover. The forest areas had since the late 1980s been dra-
matically reduced by about 78 per cents. The areas with
less-than-100 m2 each were reduced by 70 percents, while the
forest areas of between 100 m2 and 1 hectare decreased by 81
per cents (Whong, 2001: p. 78). The huge forest losses consid-
erably increased the surface runoff, and made natural disasters
inevitable in the urbanization process.
The Management of Keelung River
Watershed: An ANT Account
ANT is not only a sociological theory but also a historically
and dynamically sociological theory. Chronological order is
important. As documented, the management plan of KR water-
shed (involved in the Regional Plan of Taipei-Keelung Metro-
politan) began in the 1960s. The policy of the flood Control
Plan for Great Taipei Area was decided in 1971. Levees, chan-
nels, and drainage systems were constructed in the 1970s and
the 1980s. The meandering waterway KR in Neihu area was
artificially modified into a straight river channel with embank-
ment during the 1990s. In the process, all of human and
non-human actors/actants constitute a network by the combina-
tion, recombination, and destruction of relations. As such, fol-
lowing account of ANT on KR management for the whole
process in a historical order is divided into two subsections.
The first part focuses on investigating non-human actants in the
1960s and 1970s. The analysis is done by looking at how
non-human entities, especially hydraulic device, engineering
and knowledge, are enrolled, mobilized and applied in the river
management. The second part discusses on the story of human
actors framed in Callon’s model of four stages happened in the
1980s and 1990s.
The Identification of Actors and Actants
(Or Identities in Action)
As mentioned above, the general limits of urban develop-
ment in the Taipei Basin are set by physiography and the dan-
ger of flood. As the Regional Plan of Taipei-Keelung Metro-
politan documented in 1968, the Taipei Basin
“has a humid sub-tropical climate, with an average an-
nual rainfall of 2118 mm. The rain can come in heavy
down-pours whose violence is reflected in the variation of
flow in the Tamshui River from an average low water
flow of 38 cubic meters per second to a maximum re-
corded flow of 20,050 cubic meters per second, a flow
which flooded most of the basin” (UHDC, 1968: p. 5).
Against this natural constraint, the Water Conservancy
Agency was put in charge in July 1960 of carrying out a de-
tailed survey on the hydraulic character and performance of the
rivers in Taipei Basin, including the precipitation, inundation
estimation, channel survey, flood history, and the effects of
existing flood control facilities. A subcommittee of the Tech-
nological Direction Committee, consisting of hydraulic special-
ists of the Special Assistance Fund from the United Nations and
the Chinese Institute of Civil and Hydraulic Engineering, was
also appointed to provide professional recommendations to the
survey and subsequent flood control plan. Hydraulic engineer-
ing expertise especially for river channelization was brought in
from overseas sources.
Copyright © 2012 SciRe s . 205
With two years’ intensive work, a significant amount of river
information was collected that included geologic maps of
Taipei Basin, precipitation hydrograph of Taipei city from 1900
to 1960, isohyet map of recorded maximum storm rainfall,
hydrographs of the Tamshui River, the KR and its tributaries,
areas subject to flood threat in the Taipei Basin and so on.
Concrete policy outputs included the Yuanshanzi diversion
tunnel plan, a river-mouth widening project at Kwangtu, a jetty
plan at Tamshui River mouth, and so on. Tamshui and KRs in
Taipei Basin were demarcated, classified, quantified, mapped
out, and the records were finally filed in the Survey and Report
of Flood Control Plan for Tamshui River System in 1963 (Tai-
wan Province Government, 1963). All of the paper inscriptions,
maps and the four policy alternatives for flood control in the
Taipei Basin were submitted to the Examination Group of
Taipei Flood Control Plan in the central government. Actually,
the examination was carried out through two reviewing phases
by the Corps of Engineers, arriving Taipei in 1964 at the re-
quest of American International Development (AID). The
Phase I review team visited Taipei in April 1964 and the Phase
II review team was in Taipei from 5 August to 11 September
1964 (see Corps of Engineers USA, 1965: 1). According to the
debates of urban development and flood control, the review
teams reached the following agreements concerning the flood
control plan for the Taipei Basin (see Corps of Engineers USA,
1965; Darling et al., 1964). First, channel improvements and
levees are the most feasible means of obtaining a desirable
degree of protection for the Taipei Basin. Secondly, engineer-
ing designs for flood prevention should take the discharge of
200-year floods into consideration for this densely populated
metropolitan area. Thirdly, long range plan of comprehensive
flood protection for the Taipei Basin should consider changing
the tributary waterway in upstream Tamshui River and widen-
ing its river mouth at Kwangtu which were seen as technically
feasible and economically justifiable.
These recommendations were officially approved as a statu-
tory plan at the end of 1964, or the Flood Control Plan for
Tamshui River, which was subsequently revised as the Flood
Control Plan for Great Taipei Area in 1971 (Ying, 2001: pp
22-32). Nevertheless, the hydraulic engineering plan confronted
a critical opposition by a group of urban planning professionals
led by Dr. Donald Monson who contented that
“the conservation of the flood plains in the Taipei Basin
as overflow lands shall be considered a primary flood
protection measure. Lands subject to flooding shall not be
urbanized except where the regional flood control system
is not adversely affected and where such urbanization is
economically justified” (UHDC, 1968: p. 14).
He further suggests that urban development pressures in the
Basin shall accommodate by establishing new towns and ex-
tending existing satellite towns in the Taipei-Keelung region
(ibid.). However, urban developments spread as usual through-
out the Basin, in that the urban policy proposed by Dr. Monson
was not seriously implemented. As an interviewee retrospect,
“annual river flood disasters caused by typhoons in the 1970s
indeed imposed great political pressures upon the central and
city governments to carry out flood prevention project accord-
ing to the Flood Control Plan” (interviewed in 2004 April).
Nevertheless, implementation of the Flood Control Plan was
not carried out until the 1980s when Dr. Monson had already
left Taiwan.
The Flood Control Plan, drawing from the Review Report,
was accepted as a policy consensus, a “vision plan” for flood
control, and a blueprint for the prevention of 200-year floods in
the Taipei Basin. According to the plan, anticipations of the
river development were mixed up with the goals for land use
gains and the protection of urban communities and properties.
Four important flood control measures were adopted. First,
dikes and levees were commonly used to reduce flood damages.
Second, dredging or channel improvements were employed to
increase the conveying capacity of the river. Third, drainage
systems were also constructed to reduce flood damages in ur-
ban areas. A common practice for drainage improvement was to
install pumping stations in low-lying areas to remove storm
water. Finally, river conveyance was increased by channel re-
location or meander reduction. In short, natural river processes
and their analyses were presented as part of the Flood Control
Plan, but they were subsequently replaced by the engineering
methods of river management, especially channel improvement.
As former Taipei City Mayor Lee Teng-Hui documented in 1981,
“Taipei City Government has effectively carried out the
flood-control measures according to the Taipei Flood
Control Plan …. Under construction are the flood preven-
tion projects in Shezi Island, the dike projects in the
Choumei and Kwangtu areas …. and the pumping stations
at Chungshan …and Yuanshan. With these city-led pro-
jects and efforts, a comprehensive flood control system
was thus constructed. In the next six-year plan, the city
government will move to carry out a drainage system to
ensure safety of city population and their properties” (Lee,
1981: p. 12).
It was expected to reduce the natural forces by using the
channelization engineering solutions to push water through
river channels faster and then reduce flood risk in the urban
areas. The channelization engineering of the river thus became
the dominant mode of river management. One interviewee
comments that “the government-led project did not adopt
eco-based flood control strategies through preserving wetlands
and floodplains, because it was harmful to the riverside land
developments” (interviewed in 2004 April). It was under the
land development logic that the strategies of building levees
and water pumping stations had become the only solution to the
annual flood threats of the KR (Taipei City Government, 1970).
Costing more than US $33 billions, 50 pumping stations as well
as levees, some 12-meter height and 100-kilometer long, had
been completed during fourteen-year period from 1982 to 1996
in Taipei Basin. As Figure 2 shows, the major urbanized areas
in the Taipei Basin had since been tightly enclosed and pro-
tected by the long levees and pumping stations. The river man-
agement hence represented a complex hybrid of nature and
society, with the shapes, flows, and catchments of the rivers
being extensively monitored and modified for urban needs.
However, as an interviewee further comments, “once the levees
were completed, the KR was overwhelmingly isolated from
people’s daily life. The river was not only increasingly sepa-
rated from the urban communities, but also modified to provide
land and flood drainage for the city” (interviewed in 2004 May).
Value and meaning of river to the city are hence functionally
reduced to the land supply and flood drainage.
Against this backdrop, the Commanding Center for Flood
Control and Hydrological Information Center was also estab-
lished by the Water Conservancy Agency (WCA) and filled
Copyright © 2012 SciRe s .
Tamshui River
eelung Rive
Science Park
Shezi Isl an d
Unfini shed Embankment
Inundation Area
Breached Points of Levees
Figure 2.
Riverbanks, water pumping stations and Nari inundation area in Taipei
city (enlargement of a central part of Figure 1).
with professional staff and institutions to provide flood warning,
damage notification and other information management with
GIS for preventing urban communities from violent flood at-
tacks. The WCA ha s thus played as the “center of ca lculation”,
manipulating the Flood Control Plan as the basis for river con-
trol and transformation. However, the story does not end here.
One interviewee furt her indicates that “at the nationa l level, the
central government has since the 1990s been pressured by the
worsening natural disasters of typhoon-induced urban floods
and landslides to improve river engineering, hydrologic tech-
nologies, weather knowledge, and disaster relief organizations.
It pumped massive budgets to purchase technologies and con-
duct research on hydrology and weather prediction” (inter-
viewed in 2004 April). More importantly, it has also upgraded
the functional status of the Central Weather Prediction Bureau
(CWPB) to Emergency Relief Center during the typhoon period,
and organized the central and local governments to monitor
typhoon development and carry out disaster rescues. Actually,
the CWPB has functioned as another “center of calculation”,
conducting various weather information collections and predic-
tions using materials/technologies, and provided the analysis
findings to the Relief Center for decision making. These mobi-
lization and transformation processes in typhoon season then
allow ‘the centre of calculation’ to speak on behalf of typhoons
and the river. Against these backdrops, annual budgets of over
US$ 1billion have since 1996 been mobilized by various de-
partments of the government to promote the National Science
& Technology Program for Hazards Mitigation (NSTPHM).
The giant project aims at the production of full knowledge on
natural disaster prevention through researches on meteorology,
flood control, mudslide, natural hazard information collection
and its protection mechanism (see Table 1).
With growing urban flood threats and riverside development
pressures, the government has hence imposed its hegemonic
power upon the actor network of KR with the support of urban
communities. Obviously, the government has built up its power
through the acquisition of weather/hydrological technologies
and knowledge that have enhanced the weather prediction ca-
pability and the knowledge of typhoon formation, its evolution,
path, wind velocity and precipitation. In other words, the gov-
ernment has regarded the hydrological engineering and organi-
zation as important resources to transform and tame the natural
forces of river and typhoon, maintaining order within the KR
network, and finally (re)producing the dominance of the gov-
ernment and urban community over the network.
The Translation of Keelung Rive Network: from
Problematisation to Mobilization
Intensified urbanization of Taipei Basin was accompanied by
the river management that dominated the politics of Taipei and
greatly affected the livelihood of a large number of the city’s
residents. With reference to Callon’s (1986b) four stages of
actor-network theory—problematisation, interessement, enrol-
ment and mobilization—this section investigates two land de-
velopment projects along the river on the Neihu and Shezi Is-
land in Taipei City. As these two areas were subject to flooding,
they were conserved as flood plains downstream. However, the
Neihu project had in the late 1980s translated over 500 hectares
of the riverside into the Neihu Science Park by straightening
out the meandering river way. Meanwhile, the Shezi Island has
since the early 1990s been scheduled by the development plan
of the city government to function as a strategic satellite center
of more than 180 hectares in Taipei city.
By using the case studies of Callon (1986b ) and Wo ods (19 97)
as a model, I hope my case studies of urban development on the
Neihu and Shezi Island (see Figure 1) will contribute to a bet-
ter understanding of the processes of river management that
involve both human (urban community and government) and
non-human (river, flood disasters and prevention devices) ac-
tor-networks and that dominate Taipei’s urban politics.
At the heart of river management during the urbanization
process is the relation between a group of human actors (in the
city government, central government and local community) and
the definitions and interests which the human actors ascribe to
themselves and non-human actants (river, flood disasters and
prevention devices). In Taipei City, the human actors have
identified an urban development problem, which is that the
river and flood disasters are perceived to be a city threat, and
the actors have attempted to resolve the problem through the
coupled devices of an urban development plan and a river flood
prevention project. In addition, the identification of the problem
is dependent on a particular definition of the river and disasters,
and the interests of the actors. Fundamentally, they define the
river as an area whose primary functions is flood drainage for
the city during typhoon season, and whose flood plain awaits
urban development (see Taipei City Government, 1982). The
flood disaster is defined as a natural phenomenon, but a
Copyright © 2012 SciRe s . 207
Copyright © 2012 SciRe s .
Table 1.
National science & technology program for hazards mitigation, 1999-2001.
1999 2000
US$ millions Project numbersUS$ millions Project num bers
US$ millions
Ministry of the Interior (Bureau for Fire Protection) .4 5 .49 5 .35
Ministry of the Interior
(Construction and Pl anning Administration) .18 4 - - .33
Ministry of the Interi o r (Institute of Architecture) 7.3 10 .3 9 19.0
Ministry of Transportation and Communic ation
(Central Weather Prediction Bureau) 2.4 16 .96 17 .63
Ministry of Economic Affairs
(Water Conservancy Agency) 1.9 10 2.17 6 2.67
Ministry of Economic Affairs (Institu te of Geo-
graphical Information Survey) .52 3 2.0 2 .67
Ministry of Finance (Insurance Administration) - - - - .1
Executive Yuan (Public Engineer ing Council) .32 3 .57 6 .6
Executive Yuan
(Agricultural Development Council) .57 2 - - 1.67
Ministry of Education (Consultant Office) .12 15 .12 18 .1
Health Administratio n .15 2 .61 16 .13
Environme nt Administration .3 1 .05 1 .07
National Science Council 2.64 129 2.9 125 3.44
total 7.19 200 10.2 205 11.53
Source: based on NAPHM Newsletter 2000(2): 36.
destructive threat to city development and has to bring under
control. Flood drainage and land supply are identified as the
primary interest of the city to the KR However, the city gov-
ernment views the river itself as incapable of realizing this in-
terest. The government therefore defines itself as representing
the river’s interest by the coupled devices of urban develop-
ment plan (UDP) and flood prevention project (FPP).
However, the city government could not actually have the
full authority to pursue these devices alone. Rather, it recog-
nized that power could only be mobilized by enrolling the cen-
tral government which has the final authority to make decisions
on the master urban development plan of the capital and flood
prevention project proposed by the city government, and which,
through its budgetary and hydraulic engineering supports,
would be able activate a network of flood prevention devices.
The passing of the above UDP and FPP is hence positioned as
an obligatory passage point (OPP) for the objective of urban
development, realizing through the mobilization of all human
actors and non-human actants in river management network
(see Figure 3).
Figure 3.
River development ne t wo rk. OPP = obligatory passage point.
stage involved compatible inter-group definitions of the fol-
lowing five sets of entities by the city government, each of
which would need to be enrolled into an actor-network if the
plan and project for the river development were to be successful.
These entities ar e:
As one interviewee suggests, “the primary interests of the
central and city governments in the river management are
served by gaining legitimacy as the dual tasks of flood draining
and land supply are enforced to facilitate the public interests of
urban development, and to realize the citizen protections from
flood threats” (interviewed in 2004 April). Consequently, an-
other major activity of city government has been to seek public
support from local communities (especially residents and
landowner groups) to carry out the coupled UDP and FPP. At
this point, the city government may be identified as the initiator
of the coupled plan and project. As such, the problematisation
1) The river—defined as above.
2) Flood disasters—defined as undesired threats to city
development and must be brought under control.
3) Flood prevention devices—defined as useful mate-
rial weapons to protect the c it y from flood threats.
4) City government—defined as having the power to
carry out the coupled strategy that bundle UDP with
FPP. The city government sees itself as following the
local wishes of accelerating urban development and
protecting citizen life an d property.
5) Central government—defined as the “representa tive”
of the river, the gatekeeper of the controversial river-
side urban development plan, and as a resource base for
fiscal budget and hydraulic engineering for the flood
prevention project.
6) Local community—defined as residents and land-
owners in Neihu, Shezi and Hsichih districts who are
able to urge governments to facilitate the coupled UDP
and FPP especially after river flood disasters in the ty-
phoon season.
7) Other local population—defined as other citizens in
great Taipei area who wish to live in a safe city without
river flood threat.
The interessement of some of the above entities is straight-
forward and pursues without dispute. The other local popula-
tion does not dispute the definition or interest ascribed to them.
As a newly industrialized country, Taiwan was so wedded to
the capitalist development process that the political culture of
environmental opposition was still at the infant stage in the
1980s. As such, the other population (the public) normally kept
silent towards the riverside land development projects in urban
Taiwan. The river has no means to dissent. For the human ac-
tors—city government, central government and local commu-
nity—it involved linking the issue of flood disasters to wider
discourses of urban development, in order to persuade them that
river flood has to be suppressed on political and economic
Against this backdrop, flood prevention was contextualized
within wider discourses of urban development, suggesting that
the flood disaster violates the legitimacy of the city and central
governments to create a safe haven for local communities, and
it was an obstacle to the right of urban citizens to elevate wel-
fare through urban land developments. As such, the coupled
devices of UDP and FPP discursively became the two sides of a
coin for rive management as the urbanization intensified. As the
former mayor of Taipei City, Xu Shu-Der (1985-1988) docu-
“enforcement of the Neihu development project would
not only reduce flood threats of Eastern Taipei, but also
create a new residential and economic space for l20,000
people and related industrial developments during the city
transformation process. The Neihu project is a win-win-
win project for the city, government and society” (Huang,
2001: p. 118).
The coupled device is consequently identified as being in the
interest of the above three human actors in the KR watershed.
Through this interessement, the definition of the primary enti-
ties in the network was secured and a river management net-
work between the three major human actors formed.
The river was hence enrolled without discussion or negotia-
tion through the devices of the flood prevention. As far as the
Neihu case was concerned, a report on the detailed hydrographs
of the KR was completed by the National Taiwan University in
1983, which concluded that
“the meandering areas of KR in Neihu can be reclaimed
to meet the local requirement of urban development if the
river waterway can be kept at the width of 350 meters,
because this would not increase the river-water level too
much.” (Yet, 1983: p. 40).
The conclusion was further confirmed by a hydraulic simula-
tion experiment of the WCA, which was done at request of and
paid for by the central government. Armed with scientific
technologies, flood prevention devices are fully mobilized and
constructed by the end of 1980s for the city protection. With
these state supports, the city government thus in 1986 submit-
ted the Neihu development project to the central government.
Besides, the development plan of Shezi Island was initiated in
1993 by the city government. Based on flood prevention con-
siderations, low-density development became the primary pol-
icy for the island plan. However, as one interviewee points out,
“the plan was suspended because of ferocious confrontations of
local community in Shezi that wished to have giant capital
gains through a high-density development plan for the island”
(interviewed in 2004 May). In 1996, the city government was
forced to widen the river channel and use landfill to raise is-
land’s elevation by 6 - 9 meters to prevent flooding for the is-
land. Subsequently, a high-density development plan was pro-
posed to translate the island from a flood plain into a satellite
center of Taipei city (Taipei City Government, 2002). The de-
velopment plan was subsequently approved in 2002 by the
central government. There were about 12,000 people living on
the Shezi Island in the late 1990s, while the riverside in Neihu
was occupied by 13,000 residents and numerous factories in the
late 1980s. They were hence fully enrolled into the networks by
a generous offer of city government in high-density benefits of
land development. As such, the local communities were in turn
enrolled into powerful supporters for the coupled UDP and
Mobilization of the network obviously also depends on a full
support of the central government of the coupled UDP and FPP
to develop the riverside areas. In fact, the submitted flood pre-
vention project in the Neihu case encountered difficulty during
the reviewing process of WCA in 1986 due to the resistance of
hydraulic experts who were concerned about flood hazards.
However, a serious flood disaster of the KR caused by Typhoon
Lynn in 1987 (see Table 2) became a legitimacy challenge to
the central government led by Lee Teng-Hui (1987-1999), who
had served as mayor of Taipei City during the period of 1978-
1981. The disaster caused by Lynn typhoon especially imposed
a dramatic political pressure upon the central government to
reconsider the Neihu project proposed by the city government.
Subsequently, President Lee arranged a series of private meet-
ings with his cabinet members and politically pushed through
the coupled UDP and FPP agenda in Neihu (Ying, 2001: p. 38).
It was during this political process that the worried voices of
the hydraulic experts were suppressed and ultimately disap-
peared. The coupled UDP and FPP were approved with a minor
Copyright © 2012 SciRe s . 209
modification in 1988.
Similar mobilization process also happened in the case of the
Shezi Island especially after typhoon attacks. In fact, the cou-
pled device of flood prevention project and high-density de-
velopment plan was announced by Mayor Chen Sui-Ben under
the community pressure of a serious flood disaster caused by
Typhoon Herb in 1996. In the mayoral election of the same
year, the proposal of island development was subsequently
adopted as a strategic propaganda by his political competitor
(Ma Yin-Joe) in order to gain the support of the Shezi Island
community. As Mayor Ma came to power after the election, the
Shezi Island is subsequently scheduled for development by the
city government. Nevertheless, the submitted flood prevention
project has not been approved by the WCA until 2010, although
the urban development plan has been approved. The reason for
its delayed approval, as one interviewee suggests, “is neither
because the Shezi Island was functionally designed as flood
plain, nor because the flood prevention project was regarded
improper by the WCA. Instead, the Shezi development will
expose the central government to fierce political criticisms from
the neighboring urban communities of Taipei County, because
the island development project will significantly put them in
further flood threat” (interviewed in 2004 May). Nevertheless,
the Shezi project finally gains the central supports when Ma
win the 2008 presidential election, and was politically approved
in 2010.
The river management is obviously occurred in a chrono-
logical order, involving to tame the whole KR Watershed by a
long term effort of human actor group to facilitate the urban
development and bring the river under control. The river man-
agement process is discerned in Callon’s sequence of four mo-
ments as above. In problemization, human actors (urban com-
munity and governments) and non-human actants (the river,
flood disaster and prevention devices) are defined and their
interests are described in Nehhu and Shezi cases. The urban
communities constantly enjoy their material benefits in the ur-
banized area. The government reinforces its legitimacy by testing
its flood control plan and the KR continually exists and evolves
with the urbanization development in the watershed. The rela-
tionships in this framework are then tested through intresse-
ment, which refers to the river development tactics by which
the government attempt to impose and stabilize the identities
of the KR. The tactics is organized to mobilize the flood pre-
vention devices, ranging from river channelization engineer-
ing to the improvement of weather technology . Subsequent ly,
the entities are pushed into the river network through a series
of transactions in which the actors and actants must behave in
the ways which they have been assigned. The role of the river
is seen as a land supplier and for flood drainage through the
flood control plan. The government is enrolled through pre-
paring a nd implementi ng the plan. At this stag e of the proce ss
of forming the network of river modification, urban commu-
nities and residents are enrolled as powerful supporters.
However, some of them are not actors but audiences, waiting
for final results. When the above entities are involved, mobi-
lization could then be achieved, with all actors organized into
a networ k designed to ach ieve the gover nment’s desir ed goals.
This realization process thus helped turn the river into a man-
ageable entity and made the governmental apparatus the piv-
otal centre of river management focused on a series of flood
control plans.
Disasters as Destructive Actants of the Original
As Murdoch (1997b: p. 741) argues, ‘network building is a
precarious process: the alliances in the network can be con-
tested at any moment’. In the KR case, it is found that the gov-
ernment-led strategy of river development and control using
scientific engineering and knowledge was not wholly effective
in forcing the river and related non-human actants under the
network to play the functional role set by the government. In-
stead, internal tensions within the river taming network had
been deepened as dramatic climate changes accelerated re-
On average, the tropical weather in the West Pacific Region
generates 27 typhoons annually, three to four of which affect
Taiwan (Cheng, 2000: p. 13). With frequent powerful typhoons,
the KR had repeatedly caused destructive disasters, leading to
river crises in Taipei’s urbanized areas. As Table 2 shows, the
river disasters have become more frequent with worsening ef-
fects in recent ye ar s .
Hsichih was inundated with almost every typhoon, because
of its overdevelopment and the lack of investments in flood
mitigation infrastructures. In 1997, Typhoon Winny caused the
flooded areas to go up to 141 hectares. Moreover, 200 house-
holds of Lincoln Mansion in Hsichih were destroyed in massive
mudslides that buried 28 people alive. Subsequently, in 1998
Typhoons Zeb and Babs submerged the city in stormwater
twice in two weeks, inundating about 300 hectares and causing
three deaths. In response to this disaster, the 10th River Basin
Management Bureau of WCA launched a further flood control
project with a budget of US $348.5 millions for Hsichih in
1998, which was completed in 2000 (WCA, 2000: p. 17). Ty-
phoons Toraji and Nari ravaged Taiwan in 2001, claiming 185
deaths with more than 100 people missing islandwide. Hsichih
Table 2.
Floods of KR in selected typhoons.
depth (met ers)Raining hours Influenced
cities Inundation
(1987.10.24) 3.0 48
Taipei 609.6
(1996.7.31) 3.0 24
Major cities i n
Taipei Basin-
(1997.8.18) .2 - 0.3 7 - 24 Hsichih,
Taipei 141.1
(1997.8.29) .2 - 1.0 10 - 15 Hsichih,
Taipei 197
(1998.10.15) .5 - 4.0 10 - 16 Hsichih 291
(1998.10.26) .5 - 3.8 5 - 8 Hsichih 286
(2000.11.1) .5 - 7.5 10 - 16 Hsichih,
Keelung 441
(2001.7.30) - - Hsichih 800
(2001.9.16-18) 1.5 - 8.5 16 Hsichih,
Keelung 800
Copyright © 2012 SciRe s .
was again submerged twice and broke its record in areas sub-
merged. The natural disaster in these two typhoons almost du-
plicated the river disasters occurred before. This shows that the
government-led costly projects of flood control cannot com-
pletely shield the urbanized areas from serious flooding. The
river and flood disasters had not played their roles as expected
and controlled by the government strategies. In particular, Ty-
phoon Nari of 2001 has played as a destructive actant of origi-
nal network constructed by the UDP and FPP, and leads a new
network and a new translation to rise after the disaster.
Being one of the deadliest typhoons, Nari wreaked havoc
mainly on northern Taiwan. It broke many records in Taiwan’s
weather history, including precipitation in a single-day as well
as the longest period for a typhoon to linger over the island
which was more than 48 hours. Nari dropped in Taipei, Hsichih
and Keelung 425 millimeters, 881 millimeters, and 900 milli-
meters of rainfall, respectively. Taipei city area received 800
millimeters of rainfall in a single day on 17 November, about
one third of the annual precipitation in Taipei. On the same day,
mountainous areas in northern Taiwan received over 1000 mil-
limeters of rainfall. Rainwater is more dangerous than violent
gusts in Taiwan because many urban communities are located
on slopelands subject to debris and landslide threats. The rain-
fall and path of Nari Typhoon was beyond the technological
capability of the government to predict (China Times 2001/9/
18). Its record-breaking precipitation in major urbanized areas
resulted in serious disasters. 800 hectares were submerged and
60 people died in the Nari disaster.
In fact, the KR is firmly walled by long embankments as
well as 50 pumping stations, among which the Yucheng pump-
ing station was claimed to be the largest in Southeast Asia.
However, since Typhoon Nari had brought too much stormwa-
ter, the highest water level in Hsichih reaches 15.4 m that ex-
ceeded the embankment height of 12 m (Shiu, 2001). In Taipei
City, the river spilled from the embankment at lower positions
(about 9 meter height), and leaked from its breaches. It subse-
quently submerged almost half of the pumping stations, and
resulted in a severe river disaster in Taipei city (see Figure 2
for the breaching points). Owing to the failure of pumping sta-
tions, Nari’s floodwater not only submerged Keelung, Taipei
and Hsichih cities, but also rushed into underground tunnels of
the Mass Rapid Transit (MRT) and destroyed its central control
system. It cost more than US $150 million in losses and seri-
ously affected Taipei’s traffics for a month. The flood control
projects of US $3 billion suffered a crushing defeat in the Nari
disaster. The heartland of Taipei Basin was, although highly
protected, seriously submerged in water. In Taipei County, not
just Hsichih was submerged as usual, Juifang and Pingsi in the
upstream were also inundated up to the second floor. Nari again
smashed the government-led flood control projects in 2001.
As one interviewee comments, “the natural forces of typhoon
and river had inflicted so much damage to peoples’ lives and
properties that severe criticisms in Taiwan’s newspapers and
electronic media went on for more than three months. The Nari
flood was not only described as a natural disaster but also em-
phasized as a man-made catastrophe” (interviewed in 2004
May). Various in-depth reports and discussions in the media
argued that the natural disaster was made worse by improper
land use regulations in the watershed, uncompleted investments
in the flood prevention infrastructure, and governmental mis-
management in the disaster rescue process (United Daily News
2001/10/1-3). The media pressure thus translated the river dis-
aster into a fierce political storm targeted at the central and
local government. It is condemned by one interviewee that
“Taipei city government was in mismanagements of the river
and thus must take full responsibility to compensate for the
losses of lives and properties in the river disaster” (interviewed
in 2004 May). The political repercussions finally forced Mayor
Mar Yin-Jeo to make an apology in the City Parliament. Sub-
sequently, the Control Yuan filed an investigation, also charg-
ing him with mismanagement. In September 2002, an investi-
gation bill was passed by the Legislative Yuan requiring the
city government to upgrade the river engineering infrastructure
and reinforce its administrative capability to carry out natural
disaster relief.
As the river crisis intensified into political controversies in
Taipei, the Executive Yuan ordered the WCA to improve the
200-year flood defense policy and to raise the height of the
river embankment, with the designed capacity of embankment
for peak river discharge increased from 3600 tons to 5200 tons
per second. Moreover, a self-help civilian organization was also
initiated by the residents of Hsichih who resorted to a street
demonstration demanding the central government to accelerate
the KR flood control project. Above all, Liao Shei-Kwan, a
member of the Legislative Yuan and former mayor of Hsichih,
proposed a special law for the KR flood control to reinforce
flood mitigation infrastructure and disaster prevention engi-
neering in Taipei County. He tactically boycotted the important
bills of Taiwan’s WTO entrance and connected national eco-
nomic restructuring projects as a political weapon to push
through the special law on the KR in the Legislative Yuan. The
central government was accordingly forced to further spend
huge amounts of money on channel engineering for flood con-
trol. Combined with a growing political pressure of the coming
general election, the special law was enacted in 2002, and a
large budget was also swiftly passed for embankment construc-
tion and flood diversion. Four flood control programs were
quickly launched to bring the KR under control after the Nari.
The entire project costing about US $3 billion aims to 1) con-
struct a flood control dam in upstream Pingsi; 2) facilitate the
embankment construction program in the Taipei-Keelung re-
gion; 3) carry out flood-diversion projects at Yuanshanzi and
Badoo areas; and 4) complete the flood diversion tunnel at
Yuanshanzi to intercept floodwater upstream and guide it to the
Pacific Ocean directly, which was originally proposed by the
Survey and Report of Flood Control Plan for Tamshui River
System in 1963 (Taiwan Province, 1963: 47). It is estimated
that such projects would reduce the water level of KR by 1.5 m.
The above four strategies that constitute the Comprehensive
Plan of Flood Control for the KR is completed in 2008. As one
interviewee concludes, “the river crisis has further driven the
government to intensify its efforts to bring the KR under con-
trol by reinforcing the hydraulic engineering. The city govern-
ment, urban residents and other actors repeatedly attempted to
institutionalize the river management under the notion of ‘man
over nature, while resorting to more engineering techniques to
suppress and control the KR’” (interviewed in 2004 June). This
development has hence led the KR transformation into a new
stage of translation of original network, in which existing rela-
tions are destructed and recombined to further mobilize the
non-human actants for flood prevention devices.
Concluding Remarks
With reference to the actor network theory, this paper has at-
Copyright © 2012 SciRe s . 211
tempted to explore river management in urban Taiwan by ex-
amining the historical transformation processes of KR in Taipei
Basin. It is suggested that urbanization developments in Taipei
Basin has increasingly formulated the governments and urban
communities into dominant actors in the actor network of the
KR. During the development process, they have intensified
network construction in an attempt to gain privileges of repre-
sentation, to speak for the river and flood disaster, and to im-
pose land supply and flood drainage as the functions for the
river. In building this network, the government paired UDP
with FPP as a strategy to develop the river’s floodplains. The
river course was changed to make land available for urban de-
velopment. In response to the intensive riverside land develop-
ment and in order to avoid river flood threats, the government
has played a central role, constantly pumping monetary re-
sources to mobilize various scientific knowledge and technol-
ogy in the weather prediction and flood control engineering to
transform the river and disasters. In the spiraling interaction
between city and river disasters within the network, the ‘man
over nature’ idea was constantly facilitated as an engineering
conviction to further tame the river. In the process, all of human
and non-human actors/actants constitute a network by combina-
tion, recombination, and destruction of relations. The river
management is obviously not produced solely by natural nor
social processes. Instead, they are a hybrid of the processes of
social, technical and natural change, of decades of urbanization,
land development and flood defense. However, during the river
management process, the contradictions and conflicts between
the river disaster, government and urban community are inten-
sified as the river is extensively transformed to meet the de-
mand of urbanization. The Nari disaster has played as destruc-
tive actantto the orginal network constructed by the UDP and
FPP, and lead to a further transformation of the river wate rshed.
As Latour (1994: p. 32) argues, transformation is “displacement,
difference, invention, mediation, the creation of a link that did
not exist before and that to some degree modifies two elements
or agents”. It is a mutual and interactive change among hetero-
geneous elements, guided neither by determinism (technologi-
cal, social, or environmental) nor (evoluti onary) progress shaped
by any powerful individuals or structures (Callon, 1991; Labour,
1987; Eden et al., 2000). The river can still challenge the city.
In fact, the city of Hsichih was seriously inundated again by
Typhoon Haima in September 2004.
The above discussions suggest that the actor network theory
is a useful framework providing an alternative view on nature
and man as actors in man-land relationship. This is primarily
because of its distinctive ability to incorporate non-human ac-
tors on the same basis as humans, which cannot be realized in
conventional social theories (Murdoch, 1997a, b). Nevertheless,
the ANT also faces considerable limits. A significant limitation
of ANT relates to its impartiality rule that requires the analysts
to be as undecided as the actors they follow. According to La-
tour (1991: p. 130), “we refuse to accept judgments that tran-
scend the situation”. According to Eden, Tunstall and Tapsell
(2000), this implies that the analyst can only look backwards to
what has happened and not what might happen, because the
contingent character of transformation makes it unpredictable.
The analyst cannot judge the outcome of the process. This thus
causes Winner (1993: p. 368) to complain about its “disdain for
anything resembling an evaluative stance or any particular
moral or political principles”. In a nutshell, the ANT “becomes
empty when asked to provide policy, pass judgment or explain
stable feature” (Latour, 1996: p. 304). Eden, Tunstall and
Tapsell (2000) further take all these together to conclude that
the ANT is less helpful in evaluating the environmental policy.
Therefore, the ANT is theoretically unable to come up with any
future policy recommendations for flood control in the KR
watershed. Nevertheless, what is clear is that engineering pro-
jects to river management cannot be implemented effectively
without considering the policies governing land use change in a
watershed such as the KR where unpredictable hydrologic ex-
tremes can occur under the unfavorable physiographic and
geologic conditions of the watershed.
Burgess, J., Clark, J., & Harrison, C. M. (2000). Knowledge in action:
An actor network analysis of a wetland argi-environment scheme.
Ecological Economics , 35, 119-132.
Callon, M. (1986a). The sociology of an actor-network: The case of the
electric vehicle. In M. Callon, J. Law, & A. Rip (Eds.), Mapping the
dynamics of science and technology (pp. 19-34). London: Macmillan
Callon, M. (1986b). Some elements of a sociology of translation: Do-
mestication of the scallops and fishmen of St. Brieuc Bay. In J. Law
(Ed.), Power, action and belief: A new sociology of knowledge (pp.
196-233). London: Routledge and Kegan Paul.
Callon, M. (1987). Society in the making: The study of technology as a
tool for sociological analysis. In T. Huges, & T. Pinch (Eds.), The
social construction of technological systems: New directions in the
sociology and history of technology (pp. 83-103). London: MIT
Callon, M. (1991). Techno-economic networks and irreversibility. In J.
Law (Ed.), A sociology of monsters (pp. 132-161). London: Routledge.
Callon, M., Law, J., & Rip, A. (1986). Mapping the dynamics of sci-
ence and technology. Basingstoke: Macmillan.
Chen, K. S., Wang, J. T., & Mitnik, L. M. (2001). Satellite and ground
observations of the evolution of Typhoon Herb near Taiwan. Remote
Sensing of Environme nt, 75, 397-411.
Cheng, J. D., Lin, L. L., & Lu, H. S. (2002). Influences of forests on
water flows from headwater watersheds in Taiwan. Forest Ecology
and Management, 165, 11-28. doi:10.1016/S0378-1127(01)00626-0
Cheng, L.-F. (2000). Typhoons in this year. NAPHM Newsletter, 2,
13-16 (in Chinese).
CIECD (Council for International Economic Cooperation and Devel-
opment) (1969). Working reports of urban construction and housing
planning. Taipei: International Economic Corporation and Develop-
ment Council.
Comber, A., Fisher, P., & Wadsworth, R. (2003). Actor-network theory:
A suitable framework to understand how land cover mapping pro-
jects develop? Land Use Policy, 20, 299-309.
Corps of Engineers USA (1965). Report on review of Taipei area flood
control planning phase II. Taichung: Water Conservancy Agency,
Taiwan Province.
Darling, W. D., & Buswell, J. M. (1964). Preliminary review of flood
protection plan for Tam-shui river, Taipei Plain, Taiwan. Taichung:
Water Conservanc y Agency, Taiwan Pro v i n c e .
Eden, S., Tunstall, S. M., & Tsapsell, S. M. (2000). Translating nature:
River restoration as nature-culture. Environment and Planning D, 18,
257-273. doi:10.1068/d180257
Hsu, M.-H., Fu, J.-C., & Liu, W.-C. (2003). Flood routing with real-
time stage correction method for flash flood forecasting in the
Tanshui River, Taiwan. Journal of Hydrology, 2 83 , 267-280.
Huang F.-S. (2001). Oral history of Taipei Municipality by mayors and
councilors. Taipei: Taipei City Council (in Chinese) .
Huang, K.-M., & Lin, S. (2003). Consequences and implication of
Copyright © 2012 SciRe s .
Copyright © 2012 SciRe s . 213
heavy metal spatial variations in sediments of the KR drainage basin,
Taiwan. Chemosphere, 53, 1113-1121.
Kelman, A. (2003) A river and its city. Berkeley, CA: University of
California Press.
Kortelainen, J. (1999). The river as an actor-network: The Finnish
forest industry utilization of lake and river systems. Geoforum, 30,
235-247. doi:10.1016/S0016-7185(99)00019-6
Latour, B. (1996). Social theory and the study of computerized work
sites. In W. J. Orkilowske, G. Walsham, M. R. Jones, & J. I.
DeGross (Eds.), Information technology and changes in organiza-
tional work. London: Chapman and Hall.
Latour, B. (1991). Technology I society make durable. In J. Law (Ed.),
A sociology of monsters. London: R o ut l e d g e a n d K e g a n P a ul .
Latour, B. (1987). Science in Action: How to follow scientists and en-
gineers through society. Milton Keynes: Open University Press.
Latour, B. (1986). The power of association. In J. Law (Ed.), Power,
action belief: A new sociolog y of knowledge? London: Ro utledge and
Kegan Paul.
Law, J. (1991). Sociology of monsters. London: Routledge.
Lee T.-H. (1981). Development plan for Taipei City. Taipei: Taipei
City Government.
Lo, S. L., Kuo, J. T., & Wang, S. M. (1996). Water quality monitoring
network design of KR, northern Taiwan. Water Science and Tech-
nology, 34, 49-57. doi:10.1016/S0273-1223(96)00853-0
Lu, S.-Y., Cheng, J. D., & Brooks, K. N. (2001). Managing forests for
watershed protections in Taiwan. Forest Ecology and Management,
143, 77-85. doi:10.1016/S0378-1127(00)00507-7
Monson, D. (1964). Special fund assistance to the Republic of China
for regional and urban planning and housing in Taiwan. Taipei:
Murdoch, J. (1995). Actor-networks and the evolution of economic
forms: combining description and explanation in theories of regula-
tion, flexible specialization and networks. Environment and Planning
A, 27, 731-757. doi:10.1068/a270731
Murdoch, J. (1997a). Towards a geography of heterogeneous associa-
tions. Progress in H uman Geography, 21, 321-337.
Murdoch, J. (1997b). Inhuman/nonhuman/human: Actor-network the-
ory and the prospects for a nondualistic and symmetrical perspective
on nature and society. Environment and Plann i n g D , 15, 731-756.
Murdoch, J. (1998). The spaces of actor-network theory. Geoforum, 29,
357-374. doi:10.1016/S0016-7185(98)00011-6
Murdoch, J. (2001). Ecologising sociology: Actor-network theory, co-
construction and the problems of human exemptionalism. Sociology,
35, 111-133. doi:10.1017/S0038038501000074
Pfadenhauer, J. (2001). Some remarks on the socio-cultural background
of restoration ecology. Restoration Ecolo g y, 9, 220-229.
Ryan, R. T. (1998). Local perceptions and values for a Midwestern
river corridor. Landscape and Urban Planning, 42, 225-237.
Shiu, K.-L. (2001). A special report on Nari Typhoon. Taipei: United
News, 18 September 2001.
Taipei City Government (2002). Master plan for Shezi island develop-
ment. Taipei: City Development Bureau.
Taipei City Government (1982). Taipei city master plan. Taipei: City
Construction Bureau.
Taipei City Government (1970). A brief history of Taipei urban devel-
opment. Taipei: Archives Councils Press (in Chinese) .
Taiwan Province Government (1963). Survey and report of flood con-
trol plan for Tamshui River system. Taichung: Water Resources
Management Bureau, Taiwan Province (in Chinese).
UHDC (Urban and Hosuing Development Committee) (1968). Taipei-
Keelung metropolitan regional plan. Taipei: UHDC.
Van Diggelen, R., Grootjans, Ab. P., & Harris, J. A. (2001). Ecological
restoration: Government of the art or government of science? Resto-
ration Ecology, 9, 115-118.
Water Conservancy Agency (2000). Profile of water conservancy
agency, ministry of economic affairs: Man and water—Mutually de-
pendent. Taipei: Water Conservancy Agency.
Whong, Y.-L. (2001). Watershed planning based on the management of
ecological system: A case study on KR. Taipei: Master dissertation of
Graduate Institute of Resource Management, National Taipei Uni-
versity (in Chinese).
Woods, M. (1997). Researching rural conflicts: Hunting, local politics
and actor-networks. Journal of Rural Studies, 14, 321-340.
Yet, C.-L. (1983). A hydrographic studies on KR. Taipei: Taiwan Uni-
versity (in Chinese).
Ying, W.-Z. (2001). Flowing hope/disaster: Political economy of KR
flood control project. Master Dissertation of Graduate Institute of
Building and Planning Institute, National Taiwan University (in