You are on page 1of 12

Geoforum 40 (2009) 269280

Contents lists available at ScienceDirect

Geoforum
journal homepage: www.elsevier.com/locate/geoforum

The spatiality of multifunctional agriculture: A human geography perspective


Geoff A. Wilson
School of Geography, University of Plymouth, Drake Circus, Plymouth PL4 8AA, United Kingdom

a r t i c l e

i n f o

Article history:
Received 27 March 2007
Received in revised form 24 November 2008

Keywords:
Multifunctional agriculture
Geography of multifunctionality
Non-productivism
Multifunctional scales
Nested hierarchies
Gobal-level multifunctionality

a b s t r a c t
Based on reconceptualisations of multifunctional agriculture as a normative spectrum of decision-making
(strong to weak multifunctionality) bounded by productivist and non-productivist action and thought
[Hollander, G.M., 2004. Agricultural trade liberalization, multifunctionality, and sugar in the south Florida
landscape. Geoforum 35, 299312; Holmes, J., 2006. Impulses towards a multifunctional transition in
rural Australia: gaps in the research agenda. Journal of Rural Studies 22, 142160; Wilson, G.A., 2007.
Multifunctional Agriculture: A Transition Theory Perspective. CAB International, Wallingford], this paper
argues that there is currently insufcient research into the geography of multifunctionality. Building on
current human geography debates about issues of scale, the paper suggests that we should conceive of
multifunctionality as a spatially complex nested hierarchy comprising different interlinked layers of
multifunctional decision-making ranging from the farm level to the national and global levels. It suggests
that the notion of multifunctional agriculture only makes sense if it is applied at the farm level as the
most important spatial scale for the implementation of multifunctional action on the ground. Multifunctionality can be interpreted as having direct expression only at the lower geographical scales (i.e. farm,
community and regional levels in particular) while the regional, national and global levels show indirect
expressions of multifunctionality that are mediated by local level actors in order to nd tangible expression on the ground. The notion of global-level multifunctionality is the most challenging, as this level
lacks political and ideological coherence about the required directions necessary for implementation of
strong multifunctionality pathways. The paper concludes by arguing that much work still awaits those
investigating the spatiality of multifunctionality, in particular with regard to the question whether global-level strong multifunctionality is possible, or whether strong multifunctionality in one territory is
predicated on weak multifunctionality in others.
2009 Elsevier Ltd. All rights reserved.

1. Spatial issues and the reconceptualisation of


multifunctionality
The last two decades have seen some of the most interesting
and challenging theoretical debates about the nature, changes
and future trajectories of agricultural and rural systems from a
variety of economic, social, political and environmental stances.
Some commentators have argued that we are witnessing the end
of conventional agriculture that had, as its sole purpose, productivist food and bre production (Marsden, 2003). It is suggested
that a new multifunctional agricultural regime may be emerging
that has much wider purposes, including the production of nature
and new spaces for leisure.
Since the late 1980s, the notion of multifunctional agriculture1
has been used in a variety of contexts. As Hollander (2004) emphasised, the notion of multifunctional agriculture was already implicitly

E-mail address: geoff.wilson@plymouth.ac.uk


Multifunctional agriculture is used here to describe a process that provides
multiple functions, including food and bre production, environmental health, and
social capital (Holmes, 2006; Wilson, 2007).
1

0016-7185/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.geoforum.2008.12.007

apparent in ofcial policy documents of the late 1980s, reecting a


range of issues from political legitimation of the EU farm subsidy
model, changes in the development pathways of agriculture and rural areas, and changes in societal views about what agriculture
should be about. In 1988 the Commission of the European Communities in its publication The future of rural society highlighted the
multiple contributions that agriculture could make to economic
development, environmental management, and the viability of rural
communities (CEC, 1988). One of the rst applications of the notion
of multifunctional agriculture, meanwhile, can be traced back to the
Earth Summit in Rio de Janeiro in 1992, where Article 14 of Agenda
21 stated that agricultural policies should take into account the
multifunctional character of agriculture (Belletti et al., 2003). The
term multifunctional agriculture was ofcially used for the rst
time in 1993 by the European Council for Agricultural Law in an effort to harmonise agricultural legislation across Europe and to provide a legal basis for sustainable agriculture.
Multifunctionality has been approached from various stances
including policy-based approaches that see policy as a key driver
for multifunctionality (e.g. Hollander, 2004; Potter and Tilzey, 2005),
economistic approaches that focus on commodities generated by

270

G.A. Wilson / Geoforum 40 (2009) 269280

modern agriculture and associated externality problems (e.g. Durand and van Huylenbroek, 2003), and more holistic approaches
that also incorporate the strengthening of social capital as a key
component (e.g. Clark, 2005). There have been many critics of the
relatively structuralist policy-based and economistic interpretations of multifunctionality (e.g. Holmes, 2002; Wilson, 2007). This
is particularly reected in recent calls for a more normative evaluation of multifunctionality that suggests that multifunctionality
should not only be seen as a mere concept describing agricultural
change (and as a term appropriated by European policy-makers to
defend the European farm subsidy culture), but as a normative process explaining what is happening on the ground (e.g. Holmes,
2006). This normative view of multifunctionality provides a conceptual framework that argues that transitions are often characterised by non-linearity, heterogeneity, complexity and inconsistency
(Wilson, 2007). As Fig. 1 suggests, several authors have argued that
agricultural transitional pathways should, therefore, be conceptualised as a multifunctional spectrum (e.g. Hollander, 2004; Holmes,
2006) bounded by productivist and non-productivist action and
thought (Wilson, 2008a).
This normative view enables conceptualisation of weak, moderate and strong multifunctionality pathways at various actor scales
(see Fig. 1). As I have argued in detail elsewhere (Wilson, 2007),
strong multifunctionality is characterised by strong social, economic, cultural, moral and environmental capital, where high environmental sustainability plays a key role, as does the focus on
agro-food chains that reduce the need for long-distance food transport. Strongly multifunctional systems will also display low farming intensity and productivity, in most cases characterised by a
reluctance to use Green Revolution or genetically modied crops.
Environmental sustainability is, therefore, a key ingredient of
strongly multifunctional systems, but it needs to be emphasised
that it is only one of many characteristics of strong multifunctionality i.e. environmentally sustainable farming systems are not
necessarily strongly multifunctional per se. In addition, actors in
the strongly multifunctional agricultural regime show strong tendencies for local and regional embeddedness, and strongly multifunctional systems will also be characterised by high(er) food
quality associated with more differentiated food demand and
enlightened visions about food and health. There will also be a
revaluation of existing farm household knowledge, and a greater

likelihood for farms to embark on diversication pathways that


lead to reduced farm activity.
Strong multifunctionality should be seen as the best type of
multifunctionality or, indeed, the type of multifunctionality with
the best quality and weakly multifunctional agricultural systems
should be moved towards strong multifunctionality with the help
of policy and/or changing actor behaviour (Wilson, 2008a). Not
only is strong multifunctionality predicated on ensuring the protection of the environment, healthy farming and rural communities, but it can also be seen as the most moral system.
Intuitively, there is something good about strong multifunctionality, as most of its dimensions resonate positively with what producers, rural stakeholders and wider society would see as the
optimum type of agricultural regime. This means that, in the long
term, economic efciency and survival of farming systems may be
predicated on development or maintenance of strong multifunctionality pathways.
However, there is still much confusion in the literature about
both the constituents of multifunctionality and the geographical
processes associated with making agricultural pathways more
multifunctional (Wilson, 2008a). Thus, in line with Knickel and
Renting (2000), I argue that for multifunctionality to become more
than a mere conceptual issue with potentially limited applicability
on the ground, we need to know in more detail the nature and geographical scale of this ground. From a human geography perspective in particular, multifunctionality should ultimately be about
territorial expression, as different actors and stakeholder groups attempt to impose their specic multifunctional strategies within
specic spatial contexts. In other words, multifunctionality should
have tangible expression rooted in specic localities, in the farmed
landscape, and in what has increasingly been termed multi-level
governance structures that permeate various scales of decisionmaking (Brenner, 2001).
Building on this normative view of multifunctionality, the aim
of this paper is to investigate in detail the spatial implications of
the multifunctionality spectrum. As various human geographers
have emphasised, questions about the geography of multifunctionality assume particular importance if we consider the hitherto
relatively narrow debate based on instrumentalist notions of multifunctionality as a policy concept (geographically largely relegated
to the European policy model of multifunctionality) with virtually

Fig. 1. Farm development pathways and the multifunctionality spectrum (Source: author; after Holmes, 2006; Wilson, 2007).

G.A. Wilson / Geoforum 40 (2009) 269280

no relevance to the non-European context (Cocklin et al., 2006).


This paper should be seen, therefore, as an extension to human
geography debates on the spatiality of multifunctional agricultural
processes.

2. Situating debates on multifunctionality within spatial


enquiries in geography
Many commentators have emphasised the importance of spatial scale issues in human geography and cognate disciplines (e.g.
Marston et al., 2005). This has been equally important in conceptualisations of multifunctionality as a geographically conditioned
process (Holmes, 2006). Peterson et al. (2002: p. 441), for example,
argued that because of the spatial diversity of the agricultural
landscape, some have advanced the notion that many policies
aimed at addressing the multifunctional nature of agriculture must
be administered at sub-national, regional or local levels. This was
reiterated by Bryden (2005: p. 8) who emphasised that there is a
continuing debate on the relationship between multifunctionality
and territorial development in rural areas. Although there is a long
lineage of thought on the spatial hierarchies in agriculture going
back to the early 1990s (e.g. Munton et al., 1990; Whatmore
et al., 1990; Marsden et al., 1993; Whatmore, 1995), none of the
debates on multifunctionality have yet shed sufcient light on
how the notion of multifunctionality can be approached as a geographical process with potential global repercussions for agricultural and non-agricultural actors.
It is here that human geographers, with their various disciplinary inter-linkages, are situated in an ideal position, as the lens of
human geography possibly more than any other social science
discipline provides an ideal platform from which to re-evaluate
current thinking on the spatiality of multifunctional agriculture.
Indeed, human geographers have highlighted that the social relations which constitute space are not necessarily organised into
scales so much as into constellations of temporary coherences (e.g.
Thrift, 1999; Massey, 2001). These coherences are set within a social space which is the product of relations and interconnections
from the local to the global. Conceptualisations of agricultural multifunctionality should, therefore, be closely intertwined with issues
of scale the geography of multifunctionality and we need to
know to which specic spatial territories the normative notions
of weak, moderate and strong multifunctionality pathways apply.
However, there continues to be confusion about scale-related issues, with some authors suggesting conceptualisations of multifunctionality at the smallest geographical scale (e.g. individual
eld level; Rapey et al., 2004; Winter and Morris, 2004), while others have focused on the global level (e.g. Potter and Burney, 2002).
Recent interventions about human geography without scale
(Marston et al., 2005), reections on the scale debate in human
geography (Jonas, 2006), and the spatialities of contentious human
processes (Leitner et al., 2008), have highlighted both the complex
nature and importance of scale in contemporary debates. Criticising the predominance of hierarchical conceptions of scale, Marston
et al. (2005) even suggest abandoning the concept of scale altogether. On the other hand, many argue that scale continues to have
an ontological hierarchical and nested structure which exists and
is important in the real world (e.g. Brenner, 2001). Critics of the
conception of space as a nested hierarchy have argued that horizontal spatial relations are equally important, especially with regard to highly differentiated actor networks (e.g. Leitner et al.,
2008). Nonetheless, authors have also continued to suggest that
the social ... transformation of the world is inserted in a series
of scalar spatialities (Swyngedouw, 2004: p. 19). Jonas (2006: p.
399), in particular, argued that to reject scale altogether would
be to miss out on an important dimension of thinking about and

271

acting upon contemporary economic, political, social and environmental change. I agree with this latter argument, especially with
Jonas suggestion that important causal processes operate in a scalar dimension of multiple scales, and that viewing human decisionmaking as enmeshed in scalar nested hierarchies is a way of conceptualising complex processes of change that occur around multiple sites and scales, and in ever-changing spatial, temporal and
scalar settings (Jonas, 2006: p. 400).
Thus, while acknowledging Smiths (1992) earlier caution about
scalar hierarchies as socially constructed and problematic entities,
in this study I will side with the hierarchy theory argument of vertical scalar relations based on nested hierarchies, in which multilayered actor spaces are seen to shape possibilities at the local
scale2 (Brenner, 2001; Jonas, 2006). This suggests that vertical scalar
issues continue to be important in any understanding of rural geography processes as a whole, and agricultural multifunctionality pathways more specically. I base this argument on three key
assumptions about the functioning of agricultural spaces. First, agricultural actors continue to be a relatively coherent actor group
operating predominantly at the local level. Second, most agricultural
processes are strongly rooted in a specic locality based on their
dependence on a geographically well dened land base of production (i.e. the farm as the spatial decision-making unit). This pre-denes the local level as an important scale for any investigation of
agricultural processes. Third, agricultural actors, especially the more
than one billion subsistence farmers in the developing world, are the
stakeholder group most affected by top-down decision-making
structures that affect local decision-making processes on the ground
(e.g. through national/international neo-liberal policies or national/
international consumer demand) (Potter and Tilzey, 2005). This
highlights that we need to better understand the complex spatial
interactions between different multifunctionality processes.
3. The geography of multifunctionality
3.1. The nested hierarchy of multifunctional processes
Fig. 2 shows the multiple spatial scales of multifunctionality
based on the concept of a multifunctionality spectrum bounded
by productivist and non-productivist action and thought, ranging
from the farm level, to the rural community, regional, national
and global level. These scales should be seen as constellations of
temporary coherences with relatively open boundaries (Massey,
2001) rather than xed entities in space and time.
The gure highlights several spatial issues. First, and building
on Brenners (2001) scalar structuration and the assumption that
scale exists as an ontological structure, the spatiality of multifunctionality can be seen as a nested hierarchy in which individual spatial components of multifunctionality interact with each other,
resulting in implementation of manifold multifunctionality actions
at the local level. This nested hierarchy is predicated on the
assumption that there is a relationship between levels with hierarchical relations among vertically differentiated spatial units. These
different levels show spatial overlap (e.g. farm/rural community level), suggesting a concept of scale as a nested hierarchy of differentially sized and bounded spaces (Marston et al., 2005: p. 417).
Each of the scales will show spatially differential adoption of multifunctional pathways. Indeed, Brenner (2001: pp. 605606) suggested that scales evolve relationally within tangled hierarchies
and dispersed interscalar networks ... Each geographical scale is
constituted through its historically evolving positionality within
2
As Jones (1998) and Brenner (2001) highlighted, and as discussed further below, a
framework based on vertical nested hierarchies does not negate the existence and
importance of horizontally embedded spatial processes, as amply illustrated in Clarks
(2005, 2006) regional analysis of multifunctional actor spaces.

272

G.A. Wilson / Geoforum 40 (2009) 269280

Fig. 2. Spatial scales and the nested hierarchies of multifunctionality (Source:


author; after Clark, 2003, 2005; Wilson, 2007).

a larger relations grid of vertically stretched and horizontally dispersed sociospatial processes, relations and interdependencies.
Second, as we move away from the local level (farm level)
where multifunctionality nds more direct expression through tangible actions on the ground multifunctionality is expressed more
indirectly (e.g. at national level). Intermediate levels (e.g. regional),
therefore, act as lters for top-down multifunctionality policies/
ideas/decision-making for implementation of multifunctional processes on the ground (cf. Storper, 1997).
Third, while nested hierarchies of multifunctionality up to and
including the nation state level are relatively straightforward to
conceptualise, the link between national-level and global-level
multifunctionality is more complex. The reason for this is largely
related to the geographical scales of national/global policy-making.
As Wilson and Bryant (1997) argued, while the nation state still
provides the main geographical framework within which most policies that affect farmer decision-making on the ground are formulated and implemented, global level policies are (still) mediated
by nation states to suit their own needs (e.g. different implementation of European Union [EU] agri-environmental policies by member states; cf. Buller et al., 2000). The discussion below will
highlight that this global/national policy mediation and scalar
power geometries, makes it more difcult to situate the inuence
of the global on local multifunctional action and thought.
3.2. The spatial levels of multifunctionality
3.2.1. Farm level multifunctionality
Echoing Marston et al. (2005: p. 419) who argued that focus on
the local [can act] as an entry point to understanding broader
processes, I argue that the farm level3 has to be the smallest spatial
unit for the implementation of multifunctional decision-making pro-

The farm level contains any areas within the boundaries of a farm (either in one
block or comprised of several units not necessarily in close vicinity to each other).

cesses such as diversication, intensication or environmental conservation. I, therefore, agree with Knickel and Renting (2000) and
Winter (2005) that multifunctionality should be seen as a land-based
concept in which the notion of multifunctionality applies to a specied unit of land (i.e. the farm). Similarly, Buller (2005: p. 2) highlighted that most policies seeking to promote multifunctionality
. . . focus predominantly upon the farm. Applying multifunctionality
to a smaller scale (e.g. individual elds within a farm; cf. Rapey et al.,
2004) would be relatively meaningless, as the smallest common
denominator for multifunctionality pathways has to be the level of
the decision-making unit in other words, the farm. Whatmore
et al. (1990) already emphasised the importance of the production
unit of the farm as the basic scale, as most payments and subsidies
to farmers remain tied to the farm territory.
Indeed, productivist and non-productivist spaces can lie in close
vicinity within one farm, but as multifunctionality is about the
totality of decision-making processes bounded by productivist
and non-productivist action and thought, picking out individual
spatial subsets of a farm unit (e.g. a eld) may result in compartmentalised multifunctionality that may give a distorted picture of
wider farm level multifunctionality pathways. A farmer may apply
specic farming strategies to one part of the farm that may be close
to non-productivist action and thought (e.g. withdrawing food and
bre production on one upland eld for conservation and biodiversity preservation), while on other parts s/he may apply productivist
strategies (e.g. maximising food and bre production on highly
productive soils with maximum agro-chemical inputs) (Wilson
and Hart, 2001). In other words, farm level multifunctionality is
not only about expression of the physical basis of food and bre
production, but also about unravelling actions and thoughts of the
decision-maker for that space the farmer. Multifunctionality,
therefore, has to be about the link between human decision-making
and spatial expression of these decisions on the ground. Farmers
and farm households will adopt productivist or non-productivist
pathways, depending on factors exogenous to the farm level, but
also based on endogenous factors such as farmer attitudes, ideologies and identities (Burton, 2004; Burton and Wilson, 2006). Using
the farm scale as the smallest common denominator allows the
bringing together of multiple compartmentalised multifunctionality actions, providing a more complete picture of multifunctionality quality on the whole farm.
One farmer may not be the only decision-maker, and ownership
patterns may range from single owners, to family trusts in which
multiple farm household members (e.g. father and daughters)
make joint decisions, to complex ownership arrangements comprising shareholding companies and multiple farm owners. Further, many farms are farmed by tenants or farm managers who
do not have ownership of the land, and where multifunctional
decision-making pathways may be based on a wide spectrum of
decision-making constraints and opportunities, ranging from complete control by the landlord to (almost) complete control by the
tenant. Thus, it is often difcult to disentangle the work of one
farmer from another. In particular, aspects of the farm that contribute to productivity such as soil structure and fertility and eld conditions . . . are not necessarily laid down by a single . . . farmer
(Burton, 2004: p. 206). Yet, complex ownership (even over longer
time scales) still means that the farm should be the smallest unit
for conceptualising multifunctionality strategies. While in many
parts of the world farm ownership patterns have changed considerably over the past few decades towards a more corporate agribusiness model (Marsden, 2003), the notion of a clearly denable
farm property, with a distinctive and mappable boundary, has not.
3.2.2. Multifunctionality at rural community level
The second level in Fig. 2 concerns rural community level multifunctionality. Van der Ploeg and Roep (2003: p. 50) emphasised

G.A. Wilson / Geoforum 40 (2009) 269280

the re-emergence of the community as key in multifunctionality


concepts, while Belletti et al. (2003: p. 68) highlighted that the
collective level is very important in any analysis of multifunctionality . . . [as] multifunctional farm outcomes also have positive effects outside the farm referred to as the multifunctional local
system. Achieving strong multifunctionality at rural community
level is predicated on the need to achieve critical mass to build
up and sustain a multifunctional reputation for the area to external consumers (e.g. tourists), the possibility to gain from scale
economies in providing multifunctional connections between the
local system, markets and public institutions; and scope for the
successful joining of multifunctional elements in the local area.
Adoption of strongly multifunctional pathways is, therefore, not
only predicated on reinforcement of on-farm activities, but also on
synergies between these reproductive strategies and the imperatives of local or community-based territorial development more
generally a process referred to as the new associationalism between farmers and their community (Marsden et al., 2002). This
can be interpreted as the vehicle to transform the public goods
created by farming and agriculture into human welfare at the community level (Bryden, 2005). Thus, agricultural businesses need to
create and maintain new associations with a whole range of external actors and institutions . . . [by] constructing and optimising new
social networks [with] those involved at different points in the various supply chains, and those signicant regional and local actors
who are able to facilitate [multifunctional] economies (Marsden
et al., 2002: pp. 814816). Belletti et al. (2003) even suggested that
the rural community level is the essential dimension for the implementation of multifunctionality, since a large number of functions
connected to agriculture require a territorial concentration of actions and networks (economies of scale) that may not have sufcient weight at farm level.
Not all ingredients for successful rural community level multifunctionality will always be in place. That potential for rural community level multifunctionality is often based on tourism
development is a particularly contentious issue, as not all farm

273

areas are suitable for attracting tourists (Clark, 2003). Nonetheless,


Vanslembrouck and van Huylenbroek (2003) showed that extensication of agriculture as part of strong multifunctionality trajectories can have a positive inuence on attracting tourists.
Establishing strongly multifunctional pathways at the rural community level can, therefore, act as a reinforcing mechanism for
increasing the potential for further multifunctional options
through increased tourism opportunities. It is, therefore, at the rural community level that the distinction between agricultural and
rural multifunctionality is particularly important (Fig. 3). While
agricultural multifunctionality strategies at the rural community
level will be largely concerned with issues surrounding environmental sustainability, farming intensity and productivity, the level
of integration of farms into the global capitalist market, and diversication activities, rural multifunctionality issues will be focused
on local embeddedness between agricultural and rural actors (social capital). The nature, extent and durability of these agriculture-community relationships will depend on weak or strong
multifunctionality pathways adopted by the community as a
whole (Belletti et al., 2003).
Landscape protection and environmental sustainability as key
ingredients of the multifunctional agricultural regime will play a
particularly important role at the rural community level. Strong
multifunctionality pathways are more likely in areas where an
integrated and holistic vision exists that helps farmers look beyond
their own farm boundaries, and that recognises that the choice of
individual multifunctionality pathways also has repercussions for
the rural community as a whole. Here, leadership offered by individuals in the community often proves crucial for acceptance of
innovations such as strong multifunctionality. The Australian
Landcare movement is a particularly apt example. Landcare operates at the rural community level (usually at river catchment scale)
and has been relatively successful in bringing farmers from one or
several districts together to discuss issues of environmental degradation and advantages and disadvantages of adopting different
multifunctionality pathways (Wilson, 2004). The key issue here is

Fig. 3. The nested geography of multifunctional agricultural and rural spaces (Source: author; after Marsden, 2003; Belletti et al., 2003; Wilson, 2007).

274

G.A. Wilson / Geoforum 40 (2009) 269280

for farmers to recognise that implementation of strong multifunctionality is virtually impossible at the farm level alone, and that the
entire community has to provide an enabling environment (van der
Ploeg and Roep, 2003). The rural community level also has a crucial
role for multifunctional food supply chain issues (Clark, 2003). Indeed, (re)localisation of agro-commodity chains often depends on
opportunities for local sale of farm products at the rural community level. Demand for local production will be key in making local
farm shop ventures a success, while community support (practical
and psychological) for farmers choosing strong(er) multifunctionality pathways is crucial for successful farm management
transitions.
3.2.3. Regional level multifunctionality
Recent human geography debates about scale have emphasised the continued importance of the region in scalar hierarchies
(see in particular Storper, 1997). At the regional level, rural multifunctionality issues assume greater importance than agricultural issues (see Fig. 3 above), as agriculture will, in most cases, be only
one of many economic sectors. Nonetheless, the regional level also
has important repercussions for enabling or disenabling multifunctional agriculture pathways on the ground, not least due to recent
trends of the regionalisation of rurality. Clark (2003, 2006) and
Marsden and Sonnino (2005) emphasised how regionalisation in
the EU has been a particularly important driver for different national interpretations of multifunctional policies, while Ward
et al. (2003: p. 21) highlighted recent shifts away from a national,
sectoral and individualized notion of agriculture and agricultural
competitiveness to a regional, territorial and collective notion.
Clark referred to this process as the sub-national promotion of
multifunctional agriculture, and highlighted that for England, for
example, a whole tier of English regional governance now has direct or indirect involvement in promoting multifunctionality
(Clark, 2006: p. 333). Marsden and Sonnino (2005) also emphasised how European Structural Funds have contributed towards a
regionalisation of multifunctionality opportunities since 1988,
while the EU LEADER programme (since 1991) has greatly aided
in reinforcing regionally-based territorial identities. In countries
such as the UK, recent political devolution has provided an additional platform for political empowerment of the regions and, as
a consequence, for more regionally-oriented multifunctional pathway opportunities (Marsden and Sonnino, 2005). Clark (2006) also
highlighted how the development of Regional Development Agencies in the UK has reinforced the importance of regional-level multifunctionality decision-making, while Murdoch et al. (2003)
suggested that the regionalisation of rurality has helped harness
local variety in line with broader policy goals. It is likely, therefore,
that regional level multifunctionality pathways may gain further
importance, at least in an EU context of increasing regionalisation
(Storper, 1997).
Knickel and Renting (2000) emphasised the particular signicance of the regional level in conceptualisations of multifunctionality, especially for job creation and options for pluriactivity, but
positive synergies are only possible at the regional level where
favourable linkages between two or more multifunctional activities can be created. As a result, the development of strong multifunctionality pathways can only be fully understood if we take
into account the synergies and multiplier effects that exist between
the different geographical scales of the multifunctionality spectrum. At the regional level, ve synergistic components will be particularly important, highlighting the existence of manifold tensions
within actor spaces in which multifunctional differentiation (both
in thought and action) are contested through actors attempting to
impose their respective representations of multifunctionality over
others. First, the regional level is often crucial in providing diversication opportunities to farmers (e.g. based on tourism potential of

the region). If diversication activities are not available to a farm


household at the regional level, it is likely that economically marginal farms may give up farming altogether instead of moving their
farm business elsewhere, or seeking pluriactive income sources far
away from the farm. As Clark (2003) emphasised, a buoyant region
that places great emphasis on strong multifunctionality should
provide farmers with opportunities to embark on strongly multifunctional pathways.
Second, the positionality of a region in localglobal processes of
interaction (Swyngedouw, 2001) inuences the embeddedness of
farm holdings in the global capitalist system. Regions that are
weakly embedded in the global capitalist system will often have
strong agricultural multifunctionality. Highly globalised regions,
meanwhile, will often contain weakly multifunctional intensive
productivist-driven agro-industrial farms geared towards income
maximisation and export of food and bre (Goodman, 2004).
Third, the region also plays a key role in inuencing food supply
chains. Geographically remote regions are more likely to have locally integrated food supply and consumption structures (i.e.
strongly multifunctional), while centrally located or highly urbanised regions provide farmers with opportunities for globalised agrocommodity chains (i.e. often weakly multifunctional) (Marsden,
2003). However, location near urban agglomerations may also
open multiple and different multifunctionality opportunities for
farmers due to both higher purchasing power of urban populations
and more varied consumer demand for agricultural products.
Fourth, at the regional level open-mindedness of farming and
rural populations will be crucial. Implementation of strong multifunctionality on the ground may be difcult for farmers if the regional culture is reluctant to support such strategies. Although
there are examples where individual farmers go against the grain,
empirical evidence suggests that it is difcult for farmers to implement what may be perceived as radical actions that are not compatible with the way things are done4 (Burton, 2004).
Fifth, the region will also be important as the level in which
state-led policy is mediated through various lter and gatekeeping processes (Wilson and Juntti, 2005). In particular, the region
often acts as a vehicle for policy implementation in which interpretation of policy as weakly or strongly multifunctional by regional
policy stakeholders may have repercussions for multifunctional actions on the ground. Although policy intentions at the state level
may be strongly multifunctional (e.g. some agri-environmental
policies), regional policy-makers may interpret such policies in a
different way. Key policy decisions may be made at regional level
that distort national policy messages or that may lead to non-implementation of policy on the ground. The latter is particularly true if
certain policies do not t regional priorities. Although examples of
this regional policy lter are manifold, Mediterranean countries
exemplify particularly well the importance of this spatial layer
through frequent failure of implementation of strong multifunctionality policies which are often perceived as leading to a reduction in regional revenue. In many Mediterranean regions,
implementation of agri-environmental policies, for example, is often seen as the imposition of strong multifunctionality that does
not t a weak multifunctionality model aimed at increasing agricultural intensity, exports and prots (Buller et al., 2000).
3.2.4. Multifunctionality at national level
As with the regional level, national level action and thought can
have substantial repercussions for the implementation of strong
agricultural multifunctionality on the ground, in particular through
4
See in particular Pretty (1995), Potter (1998), Marsden (2003) and Wilson (2007)
for detailed analyses of the tensions between how farmers would like to manage their
land/farms and the global pressures for change that often go against the grain and
intuitions of farmers.

G.A. Wilson / Geoforum 40 (2009) 269280

national political development pathways and the role that national


governments, non-governmental organisations and institutions
play in policy formulation and implementation (Wilson and Bryant, 1997). Three key issues are of particular relevance here. First,
the political orientation of a country will dene the chosen level
of a countrys embeddedness into the global capitalist market.
While many countries have now adopted neo-liberal free-marketeering ideologies that permeate every aspect of society, including
many agricultural development pathways (e.g. Australia and New
Zealand; Primdahl and Swafeld, 2004), the example of North Korea highlights how political isolation of a country will also substantially inuence opportunities for the selection of different
multifunctionality pathways (Hale, 2005). While countries that
are globally well embedded would often spawn weakly multifunctional agricultural regimes, the globalisation of agricultural trade
and production may also open multiple and different multifunctionality opportunities for farmers. Countries such as North Korea,
meanwhile, may only offer very limited multifunctional opportunities to their farmers (based on a weak multifunctionality model
predicated on satisfying national food self-sufciency). In these
countries, opportunities for strong multifunctionality pathways
at the farm level are, therefore, intrinsically linked to political decision-making pathways at the national level.
Second, the national level most often provides the framework in
which policy affecting multifunctional agriculture is formulated
(e.g. the Department for Food and Rural Affairs in the UK). This is
also largely the case in the supra-national EU, as subsidiarity ensures that most policy implementation powers are delegated to the
level of the nation state (Buller et al., 2000). In the strong state
political model (i.e. where the state has much coercive power), in
particular, multifunctionality is often orchestrated by the state by
command-and-control regulatory legislative instruments. We
can invoke the example of North Korea again as the most extreme
example of state control over human decision-making, and where
multifunctionality is equivalent to national agricultural multifunctionality (Woo, 2006). Here, the boundaries between the nested
hierarchies in Fig. 2 are particularly blurred, and it may be difcult
to differentiate between farm-level and state-level multifunctionality pathways. Yet, many other examples exist where the nested
hierarchies of multifunctionality also remain fuzzy. In China, for
example, while some decision-making powers are granted to the
hundred millions of small-scale farmers, the state still retains control over food policy, farming structures and, indeed, agricultural
ideologies. At the other extreme are countries such as Australia
and New Zealand, where neo-liberal agendas have led to a virtual
removal of state inuence over individual farm level multifunctionality decisions (Primdahl and Swafeld, 2004). There, farmers
are essentially left to fend for themselves, exposed to the vicissitudes of the global market, resulting in often weakly multifunctional farm trajectories driven by the need to survive
economically in what is often a harsh and unforgiving globalised
economic climate (Wilson and Memon, 2005). Countries in the
EU, meanwhile, occupy the middle ground. Here, states provide
an enabling policy framework (which can span entire multifunctionality spectrum), mediated by regional policy stakeholders,
where farmers are often left with substantial exibility about
which multifunctionality pathways they may wish to adopt.
Third, at the state level societal ideological multifunctionality
dimensions play the most important role often inuenced by
the media, art and literature. The strong multifunctionality model
is predicated partly on open-minded societies who accept that the
nature of farming and agriculture is in the process of change.
While such ideologies may vary considerably from region to region
(i.e. a highly urbanised region is likely to have a different view of
the importance of agriculture than a remote agriculturally-based
region), it is often ideologies on the position of agriculture in soci-

275

ety that dene constraints and opportunities for the adoption of


strongly multifunctional pathways. Countries where society continues to see agriculture essentially as a guarantor of self-sufciency and high export earnings are unlikely to provide an
enabling environment for strong multifunctionality pathways. Just
as farmers will nd it difcult to swim against the stream within
their own communities, they will also nd it hard to implement
agricultural pathways that may seem to contradict societys wishes
and aspirations. In democratic pluralistic societies this will be less
of an issue, as manifold expressions of multifunctionality will be
encouraged. But in countries with more restrictive socio-political
cultures, productivist societal ideologies may be the key factor in
dissuading farmers from leaving productivist farm development
trajectories (Wilson, 2008b).

4. Nested hierarchies: direct and indirect expressions of


multifunctionality
Application of the multifunctionality concept is ultimately associated with implementing multifunctional pathways on the
ground at farm level. Yet, the discussion so far has also highlighted
that expressions of multifunctionality differ depending on scale issues, as multifunctional action and thought at any level will have
repercussions for how multifunctionality is implemented at the local level. These different spatial expressions represent the nested
hierarchies of multifunctionality, implying that the different multifunctionality levels in Fig. 1 that constitute geographically conditioned processes are interlinked and form mutually constituted
decision-making spaces that reect individual national and regional governance structures.
This addresses Winter and Gaskells (1998) assertion that the
contemporary emphasis on the local (i.e. farm level) may enhance
certain kinds of sensitivities, but may also erase others and thereby
truncate rather than emancipate rural research. I also agree with
Bullers (2005: p. 20) suggestion that the assessment of multifunctionality requires a holistic analytical framework that extends
beyond the farm gate and beyond the individual farm and that
the true multifunctionality of agriculture [needs to be seen] as a
set of embedded and socially/spatially interlinked activities. Further, as Marsden (2003: p. 29) emphasised, the relative signicance of food and food networks in and through space is
variable, according to the main activities in those networks, and
the relative power to capture the social value of food products.
Echoing recent human geography debates (e.g. Marston et al.,
2005; Jonas, 2006), Marsden (2003) argued that any agricultural
region has to be seen as vertically embedded with national governments, horizontally through its engagement with the wider rural
community, and diagonally with sub-regional frameworks. Such
hierarchies and spatial networks can be seen as nested inasmuch
as they each form the basis for the existence and functioning of the
other both in ascending and descending order from the local to
state level. This means, for example, that it is futile to conceptualise multifunctionality at the farm level without also considering
multifunctionality drivers at other levels. Similarly, understanding
national level debates about multifunctionality (e.g. expressed
through policy formulation) will be difcult without understanding the implementation of geographically conditioned multifunctional ideas at the regional, rural community and farm levels.
However, expressions of multifunctionality are not similar
across all scales highlighted in Fig. 2. Multifunctionality is, rst
and foremost, about implementation of productivist and non-productivist processes on the ground. This means that agricultural multifunctionality nds its most direct expression at the farm level.
Rural multifunctionality, meanwhile, has its most direct expression
at the rural community level (but also, at times, with strong inu-

276

G.A. Wilson / Geoforum 40 (2009) 269280

ence at the farm and regional levels) (see Fig. 3). National-level
multifunctionality, meanwhile, should be seen as the scale of decision-making for policy formulation and general societal interpretations of what multifunctionality means in a national context.
Thus, multifunctional agriculture will have more direct expression
at the farm level (i.e. where multifunctionality ideas and policies
are implemented), while pathways at the national level are characterised by indirect expression of multifunctionality. This emphasises
an important difference between the local and state levels, highlighting that national level policies and ideas on multifunctionality
need to be translated into multifunctional actions on the ground. As
Hoggart and Paniagua (2001: p. 51) argued, change experienced at
the local level might be driven by broader forces, but this does not
mean it is accepted uncritically or unaltered at the local level, nor
does it deny the prospect of inter-local disparities in response to
the same inputs. Thus, to speak of multifunctional policies (e.g.
Potter and Tilzey, 2005; McCarthy, 2005) only makes sense
through the consideration of the effects of such policies on the
ground.
The regional and rural community levels are important mediators/conduits for both the trickling down of national multifunctionality ideologies to the farm level and, conversely, for relaying
action and thought about multifunctionality at farm level to national decision-makers. Such mediation between the national and
farm levels can take various pathways. On the one hand, we witness frequent re-interpretation of national multifunctionality policies at the regional level in most democratic societies through
intermediate actors. On the other hand, we may see outright imposition of national multifunctionality ideologies on the ground in
non-democratic countries with strong coercive state powers and
intermediate actor spaces closely aligned with state ideologies
(e.g. North Korea; China; Russia to some extent). In complex governance structures where many intermediate actors may mediate
multifunctionality policies formulated at the national level, the
further removed from the farm level decision-makers are, the more
difcult it may be for them to inuence multifunctional trajectories on the ground. Recent empirical evidence from Southern Europe regarding non-implementation of strongly multifunctional
policies would tend to support this distance-decay effect of strong
multifunctionality at the local level (Wilson and Juntti, 2005).
However, Clark (2003) also highlighted that even regional expression of multifunctionality pathways may vary substantially
between regions based on differences in mediation and re-interpretation of national multifunctionality ideals by powerful regional
decision-makers. As a result, some regions may adopt more
strongly multifunctional development pathways than national-level multifunctionality ideologies (e.g. some of the German regions).
It can, therefore, be surmised that the more complex mediating
governance structures are, the more likely it will be that multifunctional action on the ground will be highly varied and span the entire spectrum of productivist and non-productivist action and
thought shown in Fig. 1 (e.g. in most EU member states). This
may also have important repercussions for the exibility in adoption of changing multifunctionality pathways at different levels
of nested spatial hierarchies. In complex governance environments, farm level actors may not only have more possibilities in
the adoption of highly varied multifunctional farm trajectories,
but they may also be able to change the direction of these trajectories more quickly than actors at other levels. It will be more difcult for regional and national level decision-makers to change
multifunctional ideas and ideologies an assertion supported by
the quick adaptation of many farmers to changing market forces
and the rather lethargic pace of agricultural policy change at national and, indeed, supra-national levels (Wilson and Juntti,
2005). The more complex and multi-scalar a system in other
words, the further removed it is from the local or farm level

the more difcult it will be for that system to change its multifunctionality trajectories. This is probably best highlighted through
innovations such as genetically modied crops, which give added
inuence to the often weakly multifunctional downstream sectors
(e.g. multinational agri-businesses) in shaping the nature of local
food production and controlling production schedules at arms
length (Pretty, 1995). This also explains why the embeddedness
of farmers in globalised agro-commodity production chains largely
denes whether farmers will be behind or ahead of regional or
national-level multifunctionality trajectories.
The notion of direct and indirect expressions of multifunctionality at various scales highlights one of the key strengths of those
advocating that a multifunctionality spectrum best describes current agricultural/rural development trajectories, in particular as
the spectrum has an in-built localising tendency (i.e. focus on the
farm territory for direct implementation of multifunctionality action) that allows temporal and spatial portability. The notion of a
normative multifunctional decision-making spectrum is, therefore,
more sensitive to local geographies than largely policy-based or
economistic interpretations of multifunctionality.

5. Global-level multifunctionality?
So far, we have only explored scales of multifunctionality to the
national scale, largely because it is still at the nation state level that
most policy decisions with tangible expressions on the ground are
formulated. Yet, socio-political developments over the past few
decades have highlighted that, increasingly, global-level multifunctionality drivers need to be taken into account when explaining multifunctionality actions at local level (Losch, 2004). The
nation state may be getting weaker regarding policy inuence
(hollowing-out of the nation state), while supra- or international
decision-making structures may be gaining ground. However, as
Buller (2005: p. 18) emphasised, there is a general absence of
any systematic attempt to globally assess the multifunctional impacts of agriculture and agricultural support mechanisms. In this
nal section I, therefore, wish to explore whether we can conceptualise global-level multifunctional agriculture based on normative conceptualisations of multifunctionality.
Four key issues need to be considered. The rst relates to the
question of global-local spatial interactions (Swyngedouw, 2001).
Indeed, direct expression of multifunctionality on the ground can
not be understood without acknowledging global level decisionmaking pathways that affect national, regional and, at times, rural
community level development pathways. Notwithstanding the
possibility that global phenomena emerge unbidden from interactions among lower-level entities (OSullivan, 2004: p. 285), and
Urrys (1984) cautionary note that the local cannot be taken as
the national or global writ small, increasing globalisation trends
suggest that the global level is assuming ever increasing importance for multifunctional decision-making pathways on farms
(Shiva and Bedi, 2002). The boundaries of the global/local are
unstable and blurred, with the global and local increasingly understood as embedded with one another rather than as dichotomous
categories. As a result, when global and local are inevitably intertwined, approaches to the study of multifunctionality which look
at only one spatial scale will miss much of interest. Global level
studies may, thus, fail to take into account local outcomes and responses to global processes by individual actors in the multifunctionality spectrum, while local studies may omit global economic
and socio-cultural inuences. Gupta and Ferguson (1997) argued
that the reterritorialisation of space requires researchers to reconceptualise fundamentally the politics of community, solidarity,
identity and cultural difference. Such glocalisation (Swyngedouw,
2001), characterised by mutually constituting sets of practices, also

G.A. Wilson / Geoforum 40 (2009) 269280

inuences multifunctionality decisions on individual farms. Marsden et al. (1999: p. 299) suggested that these glocalisation processes are associated with the conict between globalised
aspatial systems of production and locally situated ecological
systems.
There is ample literature that discusses the effects of globalisation on individual farm decision-making pathways (e.g. Shiva and
Bedi, 2002; Goodman, 2004). What is important for multifunctional agriculture conceptualisations is that inuences of globalisation affect different agricultural territories in different ways. While
decision-making pathways on super-productivist agro-businesses
often occur entirely within the realm of the global, there still are
many farming communities that have not, or only barely, been
touched by globalisation (Wilson and Bryant, 1997). Although such
remote communities are increasingly rare, their existence emphasises that global inuences on local level multifunctionality pathways cover a wide spectrum. At the regional level, meanwhile, it
is almost impossible to nd examples where entire regions are
shielded from globalisation processes, while at the national level
even the most protectionist and isolationist approaches can not
shield states from the vicissitudes, fashions and trends of the global market (Shiva and Bedi, 2002). Even multifunctional national
policy trajectories of North Korea will be inuenced by market
and political developments beyond its borders (the recent need
for food imports to stave off starvation in North Korea is a case
in point) (Hale, 2005).
Second, global level inuences play a key role in conceptualisation of weak and strong multifunctionality, as globally embedded
agricultural stakeholders may be more likely to adopt weak(er)
multifunctionality pathways (Wilson, 2008b). This may occur out
of choice (e.g. an agro-business in New Zealand opting for intensication through opportunities of direct sales of live animals to the
Middle East) or out of necessity (e.g. an Indian farmer caught in the
globalisation treadmill by having to opt for Green Revolution
seeds and fertilisers on his/her formerly more strongly multifunctional subsistence-oriented farm) (Shiva and Bedi, 2002). Rural
sociologists (e.g. McMichael, 1995) have highlighted how such
scaling-up processes to supra-national decision-making often entail the loss of democracy and local/regional decision-making powers. Globally weakly embedded farms, on the other hand, often
have strongly multifunctional characteristics, as they are often
extensively farmed and more likely to produce (healthier) foods
for the local market or for subsistence (Pretty, 1995). Global level
drivers, thus, have repercussions for decision-making on the
ground, and often these inuences lead towards a weakening of
multifunctional processes at farm level. Globalisation, yet again,
emerges as the great leveller leading to a more uniform, more
monofunctional and less socially rich agriculture.
Third, global level inuences can not be fully understood without considering the rising importance of political and policy-related drivers of agricultural change coming from supra-national
decision-making frameworks. As Potter and Tilzey (2005) argued,
the role of the state as a protector of national interests is
increasingly undermined by neo-liberal globalisation. The rising
importance of World Trade Organisation (WTO) discussions needs
to be particularly acknowledged, as these are increasingly shaping
ideological and practical approaches of national agricultural policies. WTO negotiations have become an important inuence on
policy decisions at the nation state level. For example, Common
Agricultural Policy (CAP) reforms can partly be interpreted as a
response to WTO pressures to reduce and, eventually, abolish
farm subsidies. As McCarthy (2005) argued, the recent rowing
back by the EU from multifunctionality in trade discussions
exemplies the increasing global pressure exerted by powerful
WTO actors such as the United States or the Cairns Group
attempting to force a liberalisation of EU agricultural markets.

277

This, in turn, is likely to reduce multifunctional quality of European farms in the long term, and will inuence multifunctional
decision-making pathways further down the spatial levels (Potter
and Burney, 2002).
However, there is a reason why in Fig. 2 the state level boundary has been drawn as a relatively rm line separating it from the
global level. As with the complex issue of trickle down of multifunctionality drivers from the state to the farm level, so too the
complexity of global level policy inuence on national decisionmaking needs to be acknowledged. Many commentators have
highlighted the continuing lack of a global Leviathan that dictates
what nation states should do (Wilson and Bryant, 1997). This
makes it difcult to conceptualise global-level multifunctionality,
as it continues to be unclear whose global multifunctionality notion
is advocated (e.g. that of the EU, the Cairns Group, the WTO, the
United Nations, or just the G7/G20 countries?). Although recent
ratication of the Kyoto Protocol on climate change indicates a
gradual change in how national environmental policy is increasingly inuenced by global level decision-making, enforcement
and monitoring of implementation of such international policy remains weak. Yet, even the more structured CAP does not ensure
implementation of EU-based policies in individual member states
(see above). This highlights that the inuence of the supra-national
and global level on multifunctional decision-making at national level is highly complex and that this inuence will vary greatly
depending on specic policy contexts (e.g. whether suggested policy change ts broadly with national policy goals), the coercive
powers underlying supra-national policy suggestions, and the general willingness of a nation state to work towards the greater good
beyond narrow needs and expectations of the state and its citizens.
In particular, there continues to be much exibility concerning
acceptance of WTO guidelines by individual nation states.
Although this is likely to change over the next decades, at the
beginning of the 21st century it is too early to argue that the
WTO inuences multifunctionality actions on the ground of all
agricultural actors. WTO negotiations can, therefore, at best be
seen as an enabling factor for implementation of multifunctionality
on the ground that will often lead to weaker multifunctionality
pathways. For the time being, national-level multifunctionality
decisions partly mediated by global pressures may, thus, continue to have more inuence on direct expressions of multifunctionality at the local level.
Finally, any discussion of global-level multifunctionality needs
to take into account multifunctional interdependencies between
countries and territories. Although the nation state still provides
the platform for implementation of multifunctionality pathways,
national-level multifunctionality may be highly dependent on the
movement of food in and out of a country. This global spatiality
of multifunctional processes highlights the importance of conceiving multifunctionality as composed of all individual parts of an
agro-food system. Thus, just as calculations of the ecological footprint of an individual or nation state have to take into account food
miles, international air travel, or the importing of energy that go
well beyond the boundaries of the nation state, so too do we need
to factor in processes of imported and exported multifunctionality possibly best expressed through the notion of the multifunctional footprint (Wilson, 2008b: p. 11; original emphasis). This
may well mean that a country is able to maintain relatively strong
levels of multifunctionality, precisely because this country imports
large amounts of food and bre from weakly multifunctional countries. Indeed, the latter countries may follow a weak multifunctionality model precisely because their agricultural systems are
predicated on intensive export-oriented food and bre production
(Wilson, 2008b). Japan (importer) and New Zealand (exporter) are
examples to invoke in this context. The former has been able to
maintain a relatively traditional agricultural sector that may be

278

G.A. Wilson / Geoforum 40 (2009) 269280

moderately or even strongly multifunctional (low multifunctional


footprint). However, this national level strong multifunctionality
has only been possible due to massive food imports from less multifunctional countries with higher multifunctional footprints, such
as New Zealand or Australia that have partly adjusted their agricultural trajectories to suit consumer needs of countries such as Japan
or EU member states (Wilson and Memon, 2005). As I suggested
elsewhere, global multifunctionality may, therefore, only be a
zero-sum-game, characterised by multifunctional competition in
which strongly and weakly multifunctional territories balance
each other out, rather than a winwin scenario in which all territories can simultaneously move towards strong multifunctionality
(Wilson, 2008b). This suggests that it may only be possible to
implement strong multifunctionality pathways in one region or
country at the expense of other territories.

6. Conclusions
Based on reconceptualisations of multifunctional agriculture as
a normative spectrum of decision-making bounded by productivist
and non-productivist action and thought, this paper has argued
that there is currently insufcient research into the geography of
multifunctionality. The paper suggested that we should conceive
of multifunctionality as a spatially complex nested hierarchy comprising different interlinked layers of multifunctional decisionmaking ranging from the farm level to the national and global
levels.
I argued that the notion of multifunctionality has so far been
analysed mainly from a policy-based or economistic perspective
and that only few studies have acknowledged the importance of
the spatial territory to which multifunctionality should apply.
Referring to recent human geography debates on issues of vertically and horizontally embedded scales, this paper has challenged
the notion that agricultural multifunctionality should be interpreted simply as a policy-driven top-down process used by European policy-makers to justify the continuation of a subsidy-based
farm support model. Simultaneously, the paper has challenged
economistic approaches that see agricultural multifunctionality
as a relatively aspatial externality issue. Instead, I have argued
that multifunctional agriculture has to be understood as a process
characterised by complex geographies ranging from the local to the
national and global in which issues of scale and the understanding of actor space inter-linkages at different scales play signicant roles. Indeed, geography matters in multifunctionality
debates, and the notion of nested hierarchies provides a robust scalar model with which to explain the complex spatial interrelationships between different layers of multifunctional decision-making.
In particular, we witness complex interactions between the farm
level where multifunctional processes are operationalised, and
other levels ranging from the regional to the global levels in the
nested hierarchies of multifunctionality. This emphasises that multifunctionality should be about territorial expression of actions, and
that it should have tangible expression at the local level of the
farmed landscape. At the same time, I acknowledged the fallacy
of scale rigidities (Marston et al., 2005) and that the notion of
nested hierarchies does not negate the fact that within each horizontal plain there can be multiple spatial expressions of multifunctional decision-making processes. As Clark (2005, 2006)
emphasised, all depends on the agenda of those empowered by a
given scalar strategy (e.g. the quest for regional-level strong
multifunctionality).
With these caveats in mind, I began with the suggestion that the
notion of multifunctional agriculture only makes sense if it is applied at the farm level as the most important spatial scale for the

implementation of multifunctional action on the ground. As a result, multifunctionality can be interpreted as having direct
expression only at the lower geographical scales (i.e. farm, community and regional levels in particular) while the regional, national and global levels should be seen as having indirect
expressions of multifunctionality that need to be mediated by local
level actors to nd tangible expression on the ground. We should,
therefore, consider different levels of multifunctionality, rather
than referring to a multifunctionality. The notion of global level
multifunctionality is, arguably, the most challenging, as this level
lacks political and ideological coherence about the required directions for strong multifunctionality pathways. It is still at the national level that most policies with implications for local level
multifunctionality actions are formulated, and, despite of the rising
importance of supra-national decision-making institutions such as
the WTO, the global level still only acts as a broad indirect framework for multifunctional action and thought.
Many questions related to the geography of multifunctionality
still remain. It is here that human geographers are ideally placed
to both re-spatialise work on multifunctionality by often aspatial
rural sociologists, agricultural economists, environmental scientists and policy analysts, as well as to provide empirically rich
studies based on both quantitative and qualitative methodologies
needed to unravel the full spatial complexities of the multifunctional transition (Knickel and Renting, 2000). It is here that I disagree with recent critical human geography studies that argue
for abandoning the notion of scale altogether (e.g. Marston
et al., 2005). Instead, specic focus of future research should be
placed on the interrelationships within nested hierarchies highlighted in Fig. 2 and, as Brenner (2001) emphasised, the possible
horizontal scalar issues within each of the hierarchical plains
(see also Clark, 2005, 2006). Indeed, as different site-scalar congurations and territorial structures create opportunities for a variety
of different site-scalar-strategic actions (Jonas, 2006: p. 404),
more work is awaiting those interested in scalar issues of multifunctional agricultural processes. In particular, as both the boundaries of what agriculture entails and the limits of policy decisionmaking at the nation state level are becoming increasingly blurred
(Marsden, 2003), a constant reassessment of the strong multifunctionality dimensions that are important for the farm, rural community and regional level will be necessary. This will also necessitate
further work on the spatial constituents of the (rather articial)
boundary between agricultural and rural multifunctionality alluded to in Fig. 3. For example, could the same normative view
be adopted to understand different rural multifunctional quality
and at what scales would such a concept be most appropriate?
Finally, while much work exists on assessing the impact and
nature of state-level multifunctional policies, challenging questions still relate to global multifunctionality issues, in particular
with regard to the applicability and exportability of Northernbased concepts (such as multifunctionality) to the South (Wilson
and Rigg, 2003). These questions assume ever greater importance
if we consider the possibility that one countrys choice of strongly
multifunctional agricultural pathways may be predicated on another countrys weak multifunctionality based on productivist export-oriented ideologies (Wilson, 2008b). Although researchers
will most likely be frustrated by methodological problems associated with the assessment of national-level (let alone global-level)
multifunctionality issues (Knickel and Renting, 2000), it is at the
nation state/global level interface that the most interesting geographical questions lie, in particular with regard to identifying
common indicators to assess and use empirical data on multifunctionality across different cultural and institutional spheres. Ultimately, it is hoped that such work will help better target state
policy and local action towards stronger multifunctionality.

G.A. Wilson / Geoforum 40 (2009) 269280

References
Belletti, G., Brunori, G., Mascotti, A., Rossi, A., 2003. Multifunctionality and rural
development: a multilevel approach. In: van Huylenbroek, G., Durand, G. (Eds.),
Multifunctional Agriculture: A New Paradigm for European Agriculture and
Rural Development. Ashgate, Aldershot, pp. 5580.
Brenner, N., 2001. The limits to scale? Methodological reections on scalar
structuration. Progress in Human Geography 15, 525548.
Bryden, J.M., 2005. Multifunctionality, rural development and policy adjustments in
the European Union. Paper Presented at the Policy Research Institute, Ministry
of Agriculture, Fisheries and Forestry, Tokyo, Japan, February 2005.
Buller, H., 2005. Evaluation of Policies with Respect to Multifunctionality of
Agriculture: Observation Tools and Support for Policy Formulation and
Evaluation. (UK National Report [WP 6], EU Multagri Project). University of
Exeter, Exeter.
Buller, H., Wilson, G.A., Hll, A. (Eds.), 2000. Agri-Environmental Policy in the
European Union. Ashgate, Aldershot.
Burton, R.J., 2004. Seeing through the good farmers eyes: towards developing an
understanding of the social symbolic value of productivist behaviour.
Sociologia Ruralis 44 (2), 195215.
Burton, R.J., Wilson, G.A., 2006. Injecting social psychology theory into
conceptualisations of agricultural agency: towards a post-productivist farmer
self-identity? Journal of Rural Studies 22, 95115.
CEC [Commission of the European Communities], 1988. The Future of Rural Society.
CEC, Brussels.
Clark, J., 2003. Regional innovation systems and economic development: the
promotion of multifunctional agriculture in the English East Midlands.
Unpublished PhD thesis, Department of Geography, University College
London, UK.
Clark, J., 2005. The New Associationalism in agriculture: agro-food diversication
and multifunctional production logics. Journal of Economic Geography 5, 475
498.
Clark, J., 2006. The institutional limits to multifunctional agriculture: subnational
governance and regional systems of innovation. Environment and Planning C:
Government and Policy 24, 331349.
Cocklin, C., Dibden, J., Mautner, N., 2006. From market to multifunctionality? Land
stewardship in Australia. The Geographical Journal 172 (3), 197205.
Durand, G., van Huylenbroek, G., 2003. Multifunctionality and rural development: a
general framework. In: van Huylenbroek, G., Durand, G. (Eds.), Multifunctional
Agriculture: A New Paradigm for European Agriculture and Rural Development.
Ashgatec, Aldershot, pp. 118.
Goodman, D., 2004. Rural Europe redux? Reections on alternative agro-food
networks and paradigm change. Sociologia Ruralis 44, 316.
Gupta, A., Ferguson, J., 1997. Culture, power, place: ethnography at the end of an
era. In: Gupta, A., Ferguson, J. (Eds.), Culture, Power, Place: Explorations in
Critical Anthropology. Duke University Press, Durham (NC), pp. 129.
Hale, C.D., 2005. Real reform in North Korea? The aftermath of the July 2002
economic measures. Asian Survey 45 (6), 823842.
Hoggart, K., Paniagua, A., 2001. What rural restructuring? Journal of Rural Studies
17, 4162.
Hollander, G.M., 2004. Agricultural trade liberalization, multifunctionality, and
sugar in the south Florida landscape. Geoforum 35, 299312.
Holmes, J., 2002. Diversity and change in Australias rangelands: a post-productivist
transition with a difference? Transactions of the Institute of British Geographers
27, 362384.
Holmes, J., 2006. Impulses towards a multifunctional transition in rural Australia:
gaps in the research agenda. Journal of Rural Studies 22, 142160.
Jonas, A.E., 2006. Pro scale: further reections on the scale debate in human
geography. Transactions of the Institute of British Geographers 31 (3), 399406.
Jones, K., 1998. Scale as epistemology. Political Geography 17, 2528.
Knickel, K., Renting, H., 2000. Methodological and conceptual issues in the study
of multifunctionality and rural development. Sociologia Ruralis 40 (4), 512
528.
Leitner, H., Sheppard, E., Sziarto, K.M., 2008. The spatialities of contentious politics.
Transactions of the Institute of British Geographers 33 (2), 157172.
Losch, B., 2004. Debating the multifunctionality of agriculture: from trade
negotiation to development policies by the South. Journal of Agrarian Change
4 (3), 336360.
Marsden, T., 2003. The Condition of Rural Sustainability. Van Gorcum, Assen (NL).
Marsden, T., Banks, J., Bristow, G., 2002. The social management of rural nature:
understanding agrarian-based rural development. Environment and Planning A
34, 809825.
Marsden, T., Murdoch, J., Lowe, P., Munton, R., Flynn, A., 1993. Constructing the
Countryside. UCL Press, London.
Marsden, T., Murdoch, J., Morgan, K., 1999. Sustainable agriculture, food supply
chains and regional development: editorial introduction. International Planning
Studies 4 (3), 295301.
Marsden, T., Sonnino, R., 2005. Setting up and Management of Public Policies with
Multifunctional Purpose: Connecting Agriculture with New Markets and
Services and Rural SMEs. (UK National Report [WP 5], EU Multagri Project).
Cardiff University, Cardiff.
Marston, S.A., Jones III, J.P., Woodward, K., 2005. Human geography without scale.
Transactions of the Institute of British Geographers 30 (4), 416432.
Massey, D., 2001. Talking of space-time. Transactions of the Institute of British
Geographers 26, 257261.

279

McCarthy, J., 2005. Rural geography: multifunctional rural geographies


reactionary or radical? Progress in Human Geography 29 (6), 773782.
McMichael, P. (Ed.), 1995. The Global Restructuring of Agro-Food Systems. Cornell
University Press, Ithaca (NY).
Munton, R., Marsden, T., Whatmore, S., 1990. Technological change in a period of
agricultural adjustment. In: Lowe, P., Marsden, T., Whatmore, S. (Eds.),
Technological Change and the Rural Environment. David Fulton, London, pp.
114132.
Murdoch, J., Lowe, P., Ward, N., Marsden, T., 2003. The Differentiated Countryside.
Routledge, London.
OSullivan, D., 2004. Complexity science and human geography. Transactions of the
Institute of British Geographers 29 (3), 282295.
Peterson, J.M., Boisvert, R.N., de Gorter, H., 2002. Environmental policies for a
multifunctional agricultural sector in open economies. European Review of
Agricultural Economics 29 (4), 423443.
Potter, C., 1998. Against the Grain: Agri-Environmental Reform in the United States
and the European Union. CAB International, Wallingford.
Potter, C., Burney, J., 2002. Agricultural multifunctionality in the WTO: legitimate
non-trade concern or disguised protectionism? Journal of Rural Studies 18, 35
47.
Potter, C., Tilzey, M., 2005. Agricultural policy discourses in the European postFordist transition: neoliberalism, neomercantilism and multifunctionality.
Progress in Human Geography 29 (5), 120.
Pretty, J.N., 1995. Regenerating Agriculture: Policies and Practice for Sustainability
and Self-reliance. Earthscan, London.
Primdahl, J., Swafeld, S., 2004. Segregation and multifunctionality in New Zealand
landscapes. In: Brouwer, F. (Ed.), Sustaining Agriculture and the Rural
Environment: Governance, Policy and Multifunctionality. Edward Elgar,
Cheltenham, pp. 266285.
Rapey, H., Lardon, S., Josien, E., Servire, G., Fiorelli, C., Klingenschmidt, F., Matter, E.,
2004. Multifonctionnalit de lespace agricole dun territoire: premires
conclusions issues dune demarche de recherche. Les Cahiers de la
Multifonctionnalit 5, 7186.
Shiva, V., Bedi, G., 2002. Sustainable Agriculture and Food Security. Sage, London.
Smith, N., 1992. Contours of a spatialized politics: homeless vehicles and the
production of geographical space. Social Text 33, 5481.
Storper, M., 1997. The Regional World: Territorial Development in a Global
Economy. Guildford, London.
Swyngedouw, E., 2001. Elite power, global forces, and the political economy of
global development. In: Clark, G.L., Feldman, M.P., Gertler, M.S. (Eds.), A Reader
in Economic Geography. OUP, Oxford, pp. 541558.
Swyngedouw, E., 2004. Scaled geographies: nature, place and the politics of scale.
In: Sheppard, E., McMaster, R.B. (Eds.), Scale and Geographic Enquiry. Blackwell,
Oxford, pp. 129153.
Thrift, N., 1999. The place of complexity. Theory Culture and Society 16, 3169.
Urry, J., 1984. Capitalist restructuring, recomposition and the regions. In: Bradley, T.,
Lowe, P. (Eds.), Locality and Rurality. Geo Books, Norwich, pp. 4564.
van der Ploeg, J.D., Roep, D., 2003. Multifunctionality and rural development: the
actual situation in Europe. In: van Huylenbroek, G., Durand, G. (Eds.),
Multifunctional Agriculture: A New Paradigm for European Agriculture and
Rural Development. Ashgate, Aldershot, pp. 3753.
Vanslembrouck, I., van Huylenbroek, G., 2003. The demand for landscape amenities
by rural tourists. In: van Huylenbroek, G., Durand, G. (Eds.), Multifunctional
Agriculture: A New Paradigm for European Agriculture and Rural Development.
Ashgate, Aldershot, pp. 8399.
Ward, N., Lowe, P., Bridges, T., 2003. Rural and regional development: the role of the
regional development agencies in England. Regional Studies 37 (2), 201214.
Whatmore, S., 1995. From farming to agribusiness: the global agro-food system. In:
Johnston, R.J., Taylor, P.J., Watts, M.J. (Eds.), Geographies of Global Change:
Remapping the World in the Late Twentieth Century. Blackwell, London, pp.
3649.
Whatmore, S., Munton, R., Marsden, T., 1990. The rural restructuring process:
emerging divisions of agricultural property rights. Regional Studies 24, 235
245.
Wilson, G.A., 2004. The Australian landcare movement: towards post-productivist
rural governance? Journal of Rural Studies 20, 461484.
Wilson, G.A., 2007. Multifunctional Agriculture: A Transition Theory Perspective.
CAB International, Wallingford.
Wilson, G.A., 2008a. From weak to strong multifunctionality: conceptualising
farm-level multifunctional transitional pathways. Journal of Rural Studies 24,
367383.
Wilson, G.A., 2008b. Global multifunctional agriculture: transitional convergence
between North and South or zero-sum game? International Journal of
Agricultural Sustainability 6 (1), 321.
Wilson, G.A., Bryant, R.L., 1997. Environmental Management: New Directions for
the 21st Century. UCL Press, London.
Wilson, G.A., Hart, K., 2001. Farmer participation in agri-environmental schemes:
towards conservation-oriented thinking? Sociologia Ruralis 41 (2), 254274.
Wilson, G.A., Juntti, M. (Eds.), 2005. Unravelling Desertication: Policies and Actor
Networks in Southern Europe. Wageningen Academic Publishers, Wageningen
(NL).
Wilson, G.A., Memon, P.A., 2005. Indigenous forest management in 21st century
New Zealand: towards a postproductivist indigenous forest-farmland
interface? Environment and Planning A 37, 14931517.
Wilson, G.A., Rigg, J., 2003. Post-productivist agricultural regimes and the South:
discordant concepts? Progress in Human Geography 27 (5), 605631.

280

G.A. Wilson / Geoforum 40 (2009) 269280

Winter, M., 2005. Geographies of food: agro-food geographies food, nature,


farmers and agency. Progress in Human Geography 29 (5), 609617.
Winter, M., Gaskell, P., 1998. The Effects of the 1992 Reform of the Common
Agricultural Policy on the Countryside of Great Britain. Countryside Community
Press, Cheltenham.
Winter, M., Morris, C., 2004. Multifunctionality and farming systems: the noncommodity outputs of organic and integrated systems of agriculture. Paper

presented at the 2nd meeting of the Anglo-German Rural Geographers Rural


multifunctionality: perspectives from policy-making, implementation and
practice, University of Exeter, Exeter, UK, July 2004.
Woo, M.J., 2006. North Korea in 2005: maximising prot to save socialism. Asian
Survey 46 (1), 4955.