Journal of Environmental Management 90 (2009) S124–S131

Contents lists available at ScienceDirect

Journal of Environmental Management

journal homepage: www.elsevier.com/locate/jenvman

Farm diversity, classification schemes and multifunctionality

J.D. van der Ploeg a, *, C. Laurent b, F. Blondeau b, P. Bonnafous b
a
Rural Sociology Group, Wageningen University, P.O. Box 8130, 6700 EW Wageningen, The Netherlands
b
INRA-SAD, UMR SAD APT, 16 rue Claude Bernard, 75231 Paris Cedex 5, France

a r t i c l e i n f o a b s t r a c t

Article history: Diversity is not only intrinsic to agriculture; it can be considered also as one of its main assets as it
Received 17 November 2007 provides a wide range of responses that can help to face uncertain futures. The ongoing encounter
Received in revised form between changing spatial and temporal frameworks and a set of diverse farming strategies is leading to
13 February 2008
the emergence of an ongoing flow of development models that could materialize in a wide range of
Accepted 9 November 2008
Available online 9 January 2009
farming practices, contrasting enterprise models, changing relations between rural households and
agricultural holdings, and differentiated patterns that link farming and farming families to the wider
context in which they are embedded. The many-sided diversity encountered in agriculture is not only the
Keywords:
Styles of farming outcome of the agency and polyvalence of the actors involved; their agency and polyvalence are in turn
Activity systems inspired and strengthened by the material and symbolic diversity, which contributes to a further
Classification schemes unfolding of diversity. A proper understanding of the range, dimensions, significance and causes of
Typologies diversity has been, over the centuries, a main concerndfirst for what is now known as classical
Multifunctionality agronomy, and later on in agrarian sciences. Yet the classification schemes, developed and used for such
an understanding, have increasingly become an Achilles heel as each of them relies on specific
assumptions that will bring out particular features of the overall farm diversity and will result in different
perspectives of what agriculture is and how it fits into societal projects. Consequently, they are at the
core of many debates and struggles, not only within agrarian sciences but increasingly on a wider societal
level. The growing recognition of multifunctionality in agriculture, especially in the context of the
changing EU policy, strengthens the relevance and importance of this debate. In this new context, we
discuss advantages and limits of different classification principles by comparing two methodologies
which have been extensively used in France and the Netherlands.
Ó 2008 Elsevier Ltd. All rights reserved.

1. On schemes of classification: An introduction and three
examples
In order to come to grips, both theoretically and practically, with
the huge diversity that exists among agricultural systems, many
schemes of classification have been developed. Initially, such
schemes particularly reflected inter-regional differences: focusing
on the uniqueness and highlighting the distinctive features of
regional agricultural systems (Hofstee, 1946; Di Medici, 1952;
Dumont, 1970; Bonnieux, 1982). Thus, the differences between
agricultural regions came to the fore. These differences were
related to, and could be explained by, different historical trajecto-
ries and the contrasting town–countryside relations, ecological
conditions, landscapes, man–land relations and the rural institu-
tions embedded in these trajectories. Attention was also put on the
diversity within regions that may be, from an ecological and
* Corresponding author.
E-mail address: jandouwe.vanderploeg@wur.nl (J.D. van der Ploeg).
institutional point of view, more homogeneous. At the same time,
the commonalities existing (or emerging anew) within wider
spaces (that embrace several regions) received more attention. This
complementarity of interregional and intraregional approaches
reflects the growing importance of new parameters in influencing
how agricultural constellationsdwherever locateddare consti-
tuted. At present these parameters include the ongoing supremacy
granted to market and globalization, the worldwide dissemination
of similar technologies, and the widespread processes of deregu-
lation of institutional arrangements, which until recently mediated
between farming and markets, but are increasingly being
annihilated.
Paradoxically, economic liberalization contributed to draw
attention on the multifunctionality of agriculture as politicians and
agrarian studies were quickly confronted with the so-called ‘market
failures’, resulting from the difficulty to merge the requirements of
exacerbated international markets competition with national
environmental and social objectives.
In this new context, description of farm diversity deserves
a renewed attention to better assess the potential of the

0301-4797/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jenvman.2008.11.022
 J.D. van der Ploeg et al. / Journal of Environmental Management 90 (2009) S124–S131
different types of farms to adapt to theses changes and to meet
the expectations that society articulates towards the farming
sector.
2. Classification as a reflection and a carrier of modernization
An ordinary and useful way of describing farm diversity consists
in specifying the structural characteristics of different farm types. A
good example of such an approach is given by the European farm
typology proposed by Eurostat which allow to classify each Euro-
pean agricultural household according to its productive orientation
(‘type of farming’) and its economic size. Such descriptions give
a general overview of a farm population and can be a useful starting
point for most of the agrarian studies. Fig. 1 gives an example of this
kind of structural description. It reflects the distribution of
a representative sample of Dutch dairy farms in terms of their scale
and the intensity of farming.1 The reference year is 1969. The graph
shows some diversity, which, at that time, had some significance.
Levels of intensity (here measured in terms of the gross output in
Dutch guilders per standard farm unit2) ranged from 400 to 900,
whilst the scale dimension (the number of standard farm units per
unit of labour force) ranged from 30 to more than 120.
Simplistic visions of agriculture evolution can be drawn on these
sole structural indicators as shown by the proposals of the Man-
sholt Plan (1969) for the modernization of European agriculture.
The subsequent discussions generally envisaged modernization as
a pure quantitative process, modernization occurring through (and
resulting in) an enlargement of farm sizes and an increase of the
productivity of the production factors, leading to a growth of the
Gross Value of Production (GVP) per farm and to increased levels of
income. For instance, the arrow in Fig. 1 symbolizes this movement
towards the new optimum promulgated in the Netherlands by
agrarian science and policy, which saw convergence as the key to
agricultural development (Van der Ploeg, 2003).
As a result little importance was attached to the empirical
diversity that existed at that time, which was seen as merely rep-
resenting different points of departure for the jump or ‘structural
development’ towards the new optimum. This is reflected in the
classification scheme used in the 1970s, 1980s and 1990s, which
took the relative ‘distance’ of farms from the newly defined
optimum, and on this basis divided the farming population into
three categories. Hence specific attention was given to the farms
that already were realizing the highest levels of intensity levels and
scale, and which consequently had the highest levels of GVP/
enterprise. These farmers were considered to be able to ‘stay’ in
agriculture, by adapting the newest technologies and realizing
further expansion (including reliance on externally provided credit
to finance the required growth). These farmers were often regarded
as ‘vanguard farmers’.
In the same period, at the beginning of the 1960s (1960 and
1962 blueprint laws), French agricultural policy was radically
reformed in a similar mood, after intense discussions between
Christian farmers’ organizations and the government. The core of
the modernization policy was the model of a family farm employ-
ing two full-time workers (meant to be the farmer and his spouse),
and having an economic size sufficient to generate an income
similar to what could be obtained by the couple if they were sala-
ried in other occupational branches. The underlying philosophy of
the compromise of the modernization policy was that adequate
1
The working definitions of scale and intensity used here follow the general
approach suggested by Hayami and Ruttan (1985).
2 3
This was a former unit for the measurement of economic size. It is comparable,
though not identical, to current units for economic size used in European statistics.
S125
2000
1800
1600
intensity (gross output/
standard farm unit)
1400
1200
1000
800
600
400
0 30 60 90 120 150 180 210 240 270 300
scale (standard farm unit/labour unit)
Fig. 1. Diversity and the contours of the modernization project, 1969.
support would be given to each individual willing to reach this
target: land regulations were set up to secure land access, credit
policy was established to reduce capital inequalities, extensions
services were organized to compensate lack of technical knowl-
edge. This anchored the alliance between the state and modernist
family farmers who at the time were gradually taking hold of
farmer unions. This model, which is simultaneously ideological and
economic, met with great success and contributed substantially to
unprecedented production increases.
However, both in the Netherlands and in France, and in spite of
supporting measures a large group of farmers were incapabledor
unwillingdto make the required ‘jump’ into the modernization
project, or refused to make it according to the prescribed path.
There were several manners to deviate from the dominant models:
the farms could remain too small in scale, their level of intensity
could remain too low and their total output insufficient, the head of
the farm might combine several gainful activities, etc. According to
the prevailing modernization theories of the time most of these
farmers were considered as ‘traditional’, ‘non professional’,
‘residual’ or ‘leavers’ but in any case they were assumed to have no
future in agriculture.3 The diffusion-of-innovations approach easily
classified them as ‘laggards’ while many economic analyses
considered that they could ignore them and focus on serious
farmers.
Between these two groups lay an intermediate category of those
whose future was uncertain. The farmers in this category could
either stay in agriculture (by following the ‘road’ embodied in the
practices of the ‘vanguard farmers’), or their fate would also be to
disappear.
So, in both countries the farm models targeted by moderniza-
tion policies could be considered as ‘dominant’ in the sense that
they were a reference for all the farmers, they functioned as self-
evident and incontestable notions and became central to the
language used in extension, research and the formulation of
interests and prospects by the farmers’ unions. They were also
dominant in the sense that all economic mechanisms favoured
the selection of the first type of farms to the detriment of the rest of
the peasantry as they led to a classification scheme that became the
cornerstone of agrarian policies. Spatial planning and interventions,
the distribution of subsidies, extension activities, the definition of
price levels, and the delineation of programmes for applied
research were all structured to encourage the required ‘jump’. Thus,
This introduced a normative dimension into the modernization project.
‘Leaving’ was increasingly considered as a personal failure.
S126 J.D. van der Ploeg et al. / Journal of Environmental Management 90 (2009) S124–S131
agrarian policies became a self-fulfilling prophecydbut only to
a certain degree.
After several years it appeared that the farms that were not in
accordance with these reference models did not disappear as
quickly as expected.
In the Netherlands subsequent research has shown, that in
practice a considerable number of Dutch farmers ‘escaped’, as it
were, from the state-imposed classification scheme and the
associated developmental trajectories (for a synthesis see Van der
Ploeg, 2003). Fig. 2 (which uses the same criteria as Fig. 1) shows
that there was a far greater diversity of development in empirical
terms than assumed by the dominant model. The same occurred
in France. In the 1980s, farms deviating from the dominant model
were still the majority (Lacombe, 1983) and still are (Laurent,
2005). In both countries, over time there was no convergence
towards the reference models but, instead, a clear differentiation.
Farmers created, in practice, different modernization paths that
translated in ever so many styles of farming, many of which can be
interpreted as actively constructed responses to the moderniza-
tion trajectory induced in each country by the state and the
common agricultural policy (CAP). Other households invented
new ways of integrating farm activities in rural livelihood to
get complementary income, to secure patrimony or to develop
new rural enterprises based on agro-tourism or environmental
services.
Therefore the descriptions of agriculture based on the refer-
ence models are strongly challenged by such differentiation along
two major lines. First there is not ‘one best way’ for producing
agricultural commodities; an array of differentiated practices can
be observed even in farms showing similar structural character-
istics. This array reveals a potential of innovation that cannot be
ignored when designing development policies. Secondly, a large
number of farms, mostly small-scale farms or part-time farms, are
still there, even if ignored by most of the analyses based on the
reference models; these households contribute to rural develop-
ment, their farm activity may generate income and impact on
ecosystems. While such expressions of differentiation were
ignored when the farm reference model was the unquestioned
Fig. 2. Empirical development trajectories in the 1969–1981 period.
target of modernization policies they should be reintegrated in
analyses dealing with the multifunctionality of agriculture. This
was done using two distinct methodologies which have been
extensively used in the Netherlands and in France, the farming
styles approach and the typologies of households activity
systems.
3. Farming styles: A critique on, and step beyond
modernization
Dutch research into farming styles brings three interconnected
and relevant perspectives to this debate. Firstly, it considers
a style of farming to be a coherent set of strategic notions about
the way in which farming should be practised. It is a mode of
ordering: a coherent repertoire that guides practical actions and
informs farmers’ judgements. In this respect, a farming style
provides a specific model for decision-making, one where the
strategic notions are repeatedly shared by groups of farmers.
Hence, it is possible to refer to the networks in which these
notions circulate and are discussed (the French research tradition
building on the work of J.-P. Darre refers to this as groupement
professionel local; see for instance Beaudeau, 1994; Deuffic and
Candau, 2006).
Secondly, a farming style also appears as a particular practice: as
an internally consistent mode of farming. The structure and the
internal coherence of this practice are informed (‘structured’) by
the cultural repertoire mentioned above. At the same time, practice
reconfirms and/or modifies the cultural repertoire. Crucially this
implies that the decisive relations (e.g. cattle density, ratio of
heifers/milking cows or milking cows/labour force) and the objects
(cows, fields, manure, etc) are moulded in a particular way (Ver-
hoeven et al., 2003). A farming style crafts a specific productive
constellation that is reflected in, and through, the specific charac-
teristics of the objects of labourdin e.g. the productivity and
longevity of cattle, the production levels of grassland and the
nitrogen content of manure (see respectively Groen et al., 1993;
Sonneveld, 2004; Reijs, 2007).
The particular ordering of farming practice can also be regarded
as a specific model for income generation. Each of the development
trajectories shown in Fig. 2 contains a particular strategy (and
hence a set of empirical cost-benefit relationships) for generating
income. In dynamic terms, a farming style emerges, at this level, as
a particular development pattern.
Thirdly, a farming style can be defined as a set of particular
relations between markets and technology supply on the one hand,
and farming on the other. When the ordering of the market and
technology form explicit parts of a government’s agricultural
policies, we can also conceptualize farming styles as implying
strategic positions vis-à-vis government policy.
Taking these three levels together, it can be argued that
a farming style is a systematic and consistent attempt to create
congruence both at the single level, but moreover between the
various levels (see Van der Ploeg, 2003 for a fuller discussion and
illustrations).
In the countryside, the different farming styles that have
emerged over time are well known and are often described through
a broad range of folk concepts (Leeuwis, 1993). For example, for
a given productive orientation (dairy cows) and for similar struc-
tural characteristics (size, labour units) different farming styles are
distinguished. Farmers that put their cattle central in their process
of production (individual care for the milking cows, trying to ach-
ieve high milk yields and attractive economic margins per cow, etc)
are typically known as cow men. Those who centre on mechani-
zation (and on ongoing increases in scale) are defined as machine
men. And opposed to large farmers there are the so-called
 J.D. van der Ploeg et al. / Journal of Environmental Management 90 (2009) S124–S131
economical farmers.4 These are the ones that try to restrict the use of
external inputs as much as possible (while improving the use of
internal resources) and who carefully manage the balance between
loans and savings. They are, in short, the ones that aim at, and
consequently realize, low financial costs. Typically they develop
their farms through a step-by-step approach.
An interesting feature of this ‘emic’ classification scheme is that
it is not based on an a priori hierarchization (as was the case in the
scheme that underpinned agrarian policy). Rather, it recognizes the
different social, economic and productive logics that are encoun-
tered in the countryside. In identifying different (and viable) ways
of farming it thus also acts as an implicit critique of the assumptions
of a single modernization path. At the same time, though, there
clearly is a ‘classification struggle’. People who might refer to
themselves as large or vanguard farmers might be referred to by
others as e.g. subsidy hunters, just as economical farmers might be
described by others as those who are missing the boat. The point is,
though, that the implied ‘struggle’ is an ongoing processdone that
is unfinished and where the outcomes are contested and insecure.
This contrasts sharply with the earlier classification which pre-
determined the ‘fate’ of stayers and leavers. In the former classifi-
cation scheme the fate of these groups was very largely influenced
by the intertwining systems of farm enterprise, state policies and
agribusiness. In the latter scheme farmers have far more scope to
exercise their own agency.
4. Activity systems: Going beyond virtual boundaries
While the styles of farming approach represented the major
renewal in farm classification within the Netherlands during the
last two decades of the 20th century another, and at first sight
highly contrasting development materialized, in the same epoch, in
France. It departed from the premise that the reference model of
the farm of the 1960s and 1970s had lost part of its credibility and of
its reality and could no more structure the analyses of agriculture
transformations. At the end of the 1980s a research on land aban-
donment dynamics (Laurent, 1991) showed that predictions strictly
based on this model resulted in dramatic mistakes. While in
a region like Normandy, these predictions forecasted a huge land
abandonment (about 30% of the agricultural area in 15 years; Le
Boterff, 1988) studies taking into account the actual needs and
strategies of all the existing farms, including the small ones,
showed a high pressure on land occupation and a need to keep land
regulation active. This demonstrated that approaches ignoring the
actual diversity of agricultural households were strongly
misleading for policy makers, especially when they had to deal
with environmental and rural development issues.
This approach was part of a long tradition of French studies in
rural studies that used to link farm surveys and statistical data
analyses (Brun et al., 1970; Blanc et al., 1990; Laurent and Remy,
1998), contributing to the recurrent debate on the actual impor-
tance and role of the ‘dominant farm model’ (Aubert et al., 1985).
Descriptions of the structural evolution of French farms through
census data of 1979, 1989 and 2000 confirmed that the majority of
French farm households did not match with the canonical model of
the specialized farm household put forward in the 1960s (a
household associated to a full time farm that is its sole source of
4 [.] irrespective of income procured by the agricultural holding and work time
An intriguing element is that farming styles are about repertoires, practices and
sets of inter-relations. Hence, reference should be made to the style of farming
economically (etc.). However, in everyday life language people nearly always refer
to the identity of those who are involved. Thus the economical farmer becomes the
logo for the style of farming economically, just as the cow man becomes the logo for
the style of fine tuning and the large farmer (sic), the pars pro toto that symbolizes
the large-scale and continuously expanding farm.
S127
income). The analyses showed that full-time farms associated to
households with no ‘other gainful activities’ (OGA) and no pension,
decreased from 31.4% of the total number of farms in 1979 to 20.8%
in 2000. The share of farms associated to households benefiting
from other gainful activities grew from 39.1% in 1979, to 41.1% in
1989 and 49.0% in 2000. Most of this increase was on full time
farms without any pensions: in 1979 this type of farm accounted for
15.4% of French farms, but by 2000 this had grown to 21.4% and they
now outnumber the full time farms without OGA. Full time farms
with OGA account for 34.8% of the total agricultural area (AA), while
full time farms without OGA account for 31.6% (Laurent, 2005).
Such results strongly contributed to orientate a new generation
of research works in the early 1990s, when policy makers at
national and EU levels were stating that agricultural activity could
no longer be considered as exclusively devoted to commodity
production. Besides productive functions (and the associated
objective of competitiveness) other main functions or objectives
had to be taken into account: environmental functions (natural
resource management) and social functions (contribution to intra-
and inter-regional cohesion). Hence the need to better assess the
environmental and social dimensions of agriculture and its roles for
rural development especially in countries like France where large
areas fell under the objective 5b of the EU policy (MacSharry,
1990)dthis objective was specifically designed to boost rural
development. Within French agrarian sciences, this shift in
emphasis was accompanied (if not partly preceded) by the devel-
opment of new conceptual and methodological tools that (a)
recognized the multifunctionality of agricultural activities, (b)
allowed for multidisciplinary approaches and (c) were based on
broader geographical frameworks (Laurent and Bowler, 1997).
Multiple research projects were then organized (e.g. Laurent
et al., 1994, 1998; Bonnafous and Revel, 2004) that aimed at rein-
tegrating in the analyses of the transformation of agriculture all the
forms of agricultural activity, whatever their degree of socio-
political recognition. The objective was twofold: to better identify
the different functions that agricultural holdings have for the
households that live and work on them5 and to have an assessment
of the different functions fulfilled by agriculture at a societal level
by diverse types of farms.
This concern resulted in the elaboration of typologies of
households having an agricultural activity,6 typologies that could
be built at different territorial levels (small region, region, country).
In this kind of classification, the types are mutually exclusive (each
farm belong only to one type) and together they are exhaustive
(taken together, the types include the totality of the farms of the
territorial unit that is concerned). In that aspect it differs from the
farming style approach that never intended to describe the totality
of the farm population and could link a farmer to two farming styles
if he had two independent productions (e.g. dairy and cereals) each
deserving specific practice analysis.
As summarized in Table 1, this new approach resulted in
a typology of ‘activity systems’ that not only reflect different ways
of combining income sources within farm households and differ-
ences in economic size, but also differences in the aims of, and
motivations for, the activities and functions.7
5
Defined in sensu lato as ‘‘all households associated with an agricultural holding
allocated to it’’ (Laurent and Remy, 1998, p. 420, note 8).
6
There is a certain tradition of describing farm diversities in French approaches
(Perrot and Landais, 1994), although very few works consider the totality of the
existing agricultural holdings.
7
This is an interesting commonality between the Dutch and French approaches
as they both include references to the strategies of the actors involved.
S128 J.D. van der Ploeg et al. / Journal of Environmental Management 90 (2009) S124–S131

Table 1
Diversity of the ways of practising agricultural activity.

Institutional dimensions Types of agricultural activities Main objective of the agricultural

Macro-economic function
1. Commodity production
2. Combined economic
activities in rural areas
3. Income distribution
system/ social welfare
4. Consumption
Institutions considered as
legitimate to regulate conflicts
of interests (for ex. for land access)
Market regulation
Sector based institution
Local/rural authorities
State/Institutions in charge
economic and social inclusion
Local/rural authorities
Market regulation
Local/rural authorities
Skill: The head of the
agricultural holding
qualifies him (her)self as:
1. Employee-run companies (1%)a
Business manager 2. Capitalistic agriculture (3%)
Farmer 3. Agriculture as a structured
profession (20%)
Farmer 4. Agriculture based on a
traditional farmer logic (21%)
Rural entrepreneur 5. Rural enterprises (8%)
Various 6. Non integrated multi-activity (7%)
Farmer 7. Subsistence farming for
retired farmers (13%)
Various 8. Qualifying to social welfare
coverage/ old age pensions (9%)
Various 9. Agricultural activity for home
consumption and barter (2%)
Various 10. Luxury agriculture (4%)
Various 11. Small-scale recreational
agriculture (12%)
activity for the households
Income, profit
Income, profit
Income, taste for farming
Income, self-employing
profession
Associated income, patrimony
Associated income, to keep an
inherited family farm
Compensation of a low pension,
subsistence and barter
Access to social scheme (access
to pension scheme, etc.),
subsistence and barter
Subsistence and barter
Leisure, prestige, patrimony
Leisure, subsistence and barter

Sources: Laurent et al., 1998; Laurent and Rémy, 1998.

a
Estimated percentage in the total farm population at national level in France in 1998.

In some of these activity systems farming is basically oriented
toward home consumption and/or leisure (7, 9, 10 and 11). Other
types are primarily aimed at facilitating access to social welfare
schemes or supplementing inadequate pensions (8 and again 7).
Together with the part-time or pluriactive farms (type 6) and the
newly emerging ‘rural enterprise’ (very similar to the multi-
functional farms of the Dutch approaches) (type 5), it was esti-
mated that these different activity systems make up 55% of all
farming households. Entrepreneurial farming (type 3) and
peasant-like farming (type 4) cover 20% and 21% of all farm
households respectively. Strictly profit oriented agriculture (1 and
2) accounts for only 4% of all holdings included in the research
project. These data illustrate complex inter-relations between
agricultural holdings and farming households. Such shifts (and
the classification schemes that make them visible) need to be
taken into account when considering issues relating to land-use,
environment and rural development and/or the contribution of
agriculture to rural society as a whole (Laurent and Remy, op
cit.:423–4).
5. The French and Dutch classifications compared
At first sight these two classification schemes appear radically
different, but each of them highlights a specific part of the reality
that is masked by representations based on the sole dominants
models. In short:
– the first one shows the variety of farming styles within
a given population of farms having similar structural
features, hence demonstrating the limits of analyses postu-
lating the convergence of farmers’ practices towards ‘one
best way’;
– the second one provides evidence that the structural diversity
of farms persists and is simultaneously reconstructed over
time, hence putting down the idea that all farms that did not
meet the structural requirements of the modernization model
(small-scale farms, part-time farms, pluriactive farmers, etc.)
would vanish after few years.
Therefore, these classification schemes cannot be compared
unless their specific objectives as well as their context are taken
into account.8
The main objective of the Dutch farming styles approach was to
show that specialized and full time farming was not converging
into one assumed (or scientifically assessed) optimum, but instead
was unfolding into a rich array of contrasting constellations. It
showed that in practice farming families deconstructed the
assumed ‘virtual farmer’ constructed by policy and simulation
modelling and actively created identities, practices and trajectories
that went beyond the narrow limits of the model associated with
the modernization paradigm.
The same applies to the French conceptualization of activity
systems. Its specific objective was to show that ‘‘the notion of the
agricultural holding is [.] not inclusive enough to comprehend
ongoing change’’ (Laurent and Remy, 1998: 416). It showed that in
practice many rural families relate their agricultural activities with
other activities, thereby creating new systems of activity that
transform the boundaries of the agricultural holding or enterprise
that are taken as fixed by the scientific and policy communities. It
more specifically shows that many of the (smaller) farmers
assumed to disappear (to leave agriculture) have reconstructed
a rural existence through combining and intertwining different
activities and sources of income.
The differences in objectives partly reflect differences in context.
For a long time the attention of Dutch agricultural policy and
agrarian sciences was focused almost exclusively on commodity
production and very little attention was paid to additional activi-
ties. Pluriactivity was thought to be a temporary phenomenon that
hardly had any relevance. These assumptions are reflected in the
relatively narrow range of farms studied within the initial farming
8
It is precisely the deliberate ‘de-contextualization’ that characterizes e.g. the
formal separation of Dutch farmers into leavers, stayers and an in-between group. It
was supposed to reflect ‘objective truth’ and these groups were presented as
non-negotiable and non-discursive categories.
 J.D. van der Ploeg et al. / Journal of Environmental Management 90 (2009) S124–S131
styles approach.9 In France, on the other hand, the context was
quite different. Not only are there are large areas in which particular
combinations of activity systems 5 to 11 (see Table 1) represent
a significant fraction of the farm population. There is also far more
recognition of the intrinsic multifunctionality of farming, the roles
played by all types of farm activity to help sustain social and
economic cohesion in the rural areas, and the innovative diversi-
fication of farms that is characterized by unconventional produc-
tion (transformation of products and direct sale, the supply of
non-agrarian services and environmental practices) that takes
advantage of the multifunctional character of agriculture (income,
employment, social relations, rural development, environmental
protection) (Bonnafous and Revel, 2004).
By taking into account the specificities of objectives and context
of these two classification schemes, it turns out that they are not
conflicting representations as they might initially seem. On the
reverse, the two are potentially complementary and could be well
integrated into more comprehensive approaches if necessary.
Before discussing this possibility, another issue needs to be
addressed: that of whether classification schemes can be made to
‘travel’ from one place to another.
6. Making schemes of classification ‘travel’
A direct transfer of a classification scheme from one context to
another cannot but produce frictions, misunderstandings and
distortions. That has not only been the case when classification
schemes that turned out to be relatively successful were transferred
to other countries of the Mediterranean basin (such as the classi-
fication focusing on stayers, leavers and an in-between-group; see
e.g. Van der Ploeg and Saccomandi, 1995; Remmers, 1998) or the
typologies of households’ systems of activities (Auclair et al., 1996).
It also became clear in the work of Vanclay et al. (1998) who tried to
repeat the Dutch farming styles analysis in Australia. A straight-
forward application cannot but generate the conclusion that ‘‘styles
of farming cannot be identified [just as in the Netherlands]’’
(Vanclay et al., 1998: 104).
For classification schemes to travel successfully, a process of re-
contextualization is needed. That is to say, the specific objectives
underlying the approach need to be taken into account (and if
needed, reformulated), while the specificities of context (i.e. the
differences between Dutch dairy farmers and Australian vine
growers) need to be appropriately introduced into the analysis. This
preliminary step is an absolute necessity. It implies, first, to choose
the classification methodology that is best adapted to concern of
the research. For instance one must keep in mind the ability of the
French approaches to show ‘invisible parts of the farm population’
while the farm style approach aims at providing in depth knowl-
edge of the actual functioning of the farms and motivations of the
farmers. Then, there is a need to adapt the chosen method to the
specific situation of the study. If these conditions are met, classifi-
cation schemes might very well be used in settings other than those
in which they were originally intended for, and might well serve to
highlight specificities of time and place. This was demonstrated by
the work of Laurent et al. (1999), Anseeuw et al. (2001) and Mod-
iselle (2001), who applied the classification of activity systems in
the South African context of the post apartheid period. In that case
this type of method was chosen because the priority was to
describe the various patterns of the existing farm activities in the
so-called ‘black’ and ‘coloured’ fractions of the population, what-
ever the aim of the activity (e.g. subsistence, market-oriented) or
9
This initial bias was later on corrected by De Vries (1995) and Oostindie and
Van Broekhuizen (2004).
S129
their economic size; hence the interest of an approach aiming at
describing all types of farm activities. But some key points of the
method had to be adapted. The range of the functions that farm
activity could play for the people was enlarged (e.g. livestock can be
kept for saving, or for religious purposes). The unit of the analysis
was still the household but embedded in a wider family network
interfering with the major decisions of this household (Anseeuw,
2000). The building of new typologies according to these principles
helped expose households systems of activity with a farm
component that had previously remained more or less invisible and
helped in the formulation of new support mechanisms: ‘‘The
diversity, which has been identified [.], implies that different
types of farming households may need different kinds of support.
In reverse, it implies that a given policy measure will have very
different impacts according to types’’ (Anseeuw et al., 2001: 10).
Such approaches were also used by Brazilian researchers to make
visible parts of the farm population that was set aside by policy
makers but also by researchers (Cazella, 2001, 2006).
These studies show that the application of a re-contextualized
classification scheme does not necessarily block the identification
of meaningful differences. Instead, a typology of farming situations
may allow being more specific about these differences for example
when it shows how a similar objective (to get a combined income
from farm and non-farm activities) might concern very different
parts of the farm populations according to the countries. These
approaches that take into considerations all the forms of farm
activity, whatever their degree of insertion in the market, make
visible, each time, a significant part of agriculture that cannot be
ignored if environmental and social objectives are to be considered.
Agricultures observed in that way appear as diverse combinations
of heterogeneous systems of activities whose structures and soci-
etal functions can be described and compared. The same has been
shown in the successful (re-contextualized) application of farming
styles analysis within South Africa (Van Averbeke and Mohamed,
2006; Hebinck and Monde, 2007) and Brazil (Cabello Norder,
2004).
7. Combining the two approaches in order to understand
agricultural multifunctionality
As argued before, these two approaches result in some conver-
gent conclusions. They both show how the practices of farming
flow beyond the boundaries implied by the dominant models of the
modernization period. The farming styles approach shows how
agricultural resources and the interrelations actively constructed
between them are moulded and remoulded in a far wider range of
patterns than the ‘optimum’ specified by the dominant models.
Parallel to this, the activity systems approach shows how different
activities and functions are combined and intertwined in ways that
cross the boundaries of the classical farm ‘enterprise’. Cross-
application of both approaches could thus considerably enrich each
other’s representation and understanding of agriculture (Renting
et al., 2008).
One might consider that an encouragement to describe a farm
population through two different approaches is an admission of
helplessness. It is all the reverse. Classification schemes are only
tools for inquiry. They should not to be reified (as occurs in state
managed classification schemes) into objectified and non-disput-
able representations of reality as it is supposed to be. How many
people have left agriculture because they believed in the classifi-
cation that considered them as ‘leavers’? It is impossible to count
them but each definition of a class may modify the behaviour of the
people who are classified once they know about it. For this reason it
is important to make evident that each classification highlights
S130 J.D. van der Ploeg et al. / Journal of Environmental Management 90 (2009) S124–S131
a specific part of the reality, and to assess and let know the limits of
each approach.
The will to recognize the multifunctionality of agriculture makes
this cross assessment even more necessary. Multifunctionality is
increasingly becoming a major framework for the debates on
European agriculture and its future development (OECD, 2000; EC,
2003; Van Huylenbroeck and Durand, 2003). Crucially multi-
functionality not only implies redefining the many functions of
agriculture, it also implies a material transformation of agriculture
itself. It implies (re-)linking agriculture to society at large through
a far wider range of interrelations than just large commodity
markets. Hence, the diversity of the activity systems of farm
households need to be assessed in a different way, taking into
account the roles of this diversity for meeting new social and
environmental purposes.
It implies also that agricultural processes of production need to
be (re-)moulded so as to create a new synergy between the
different functions. As source of a new economic strategy at
household level, multifunctionality is not about a simple juxtapo-
sition of different activities10; rather it is about their intertwining
and their integration in new frameworks that allow for economies
of scope (Saccomandi, 1998).
Technically speaking the intertwinement of the two discussed
approaches might proceed along two interconnected pathways.
Firstly, the material dimensions of different activity systems can be
explored: how is co-production, i.e. the interaction and mutual
transformation of man and living nature, of the social and the
material, being shaped within different activity systems? How are
e.g. soils, animals, biodiversity, product qualities, local knowledge,
organizational forms, patterns of labour division within farming
families and the circuits that connect production to consumption
shaped and reshaped within different activity systems? In short,
how do different activity systems translate into the distinctive and
distinguishable socio-material constellations that together char-
acterize the impressive heterogeneity of Europe’s rural areas?
Secondly, the relational patterns of different farming styles should
be re-studieddnot only in terms of differential relations with
markets and the supply of technology but, above all, in terms of the
different (and distinguishable) activity systems in which they are
rooted. This simultaneously implies that the farming style analysis
should be broadened so as to cover not only type 3 and type 4
activity systems (see Table 1) but also all the other activity systems.
8. Conclusions
Through the systematic combination of the French ‘activity
systems approach’ and the Dutch ‘farming styles approach’ that aim
to understand and classify complementary aspects of farm diver-
sity, significant progress can be made in understanding newly
emerging rural constellations throughout Europe and making the
most of the potentials entailed in these new constellations. The
specific objective of such a new and enlarged scheme of classifica-
tion is to contribute to the exploration and specification of new
levels of both competitiveness and societal support that are being
constructed through new forms of multifunctionality. The specific
context is a highly diverse European Union in which a wide range of
new opportunities are being tested in practice while, at the same
time, new contradictions are emerging that need to be addressed
politically. Within this framework the implied travelling of
10
It is not about multifunctional land-use at e.g. the regional level. Multi-
functionality implies that specific areas are simultaneously used for farming, for the
reproduction of landscapes, for the conservation of biodiversity, for environmental
contributions and for the other gainful activities entailed in rural enterprises, etc.
experiences and insights (obtained through the different classifi-
cation schemes) throughout Europe could provide a fruitful
instrument for both scientific inquiry and the communication of
novel experiences and insights.
References
Anseeuw, W., 2000. Household, Family or Economic Unit: A discussion about the
identification of the pertinent economic units for development (A case study of
the ‘coloured’ population of South Africa). Master thesis, 74 pp. þ annexes.
Anseeuw, W., Laurent, C., Modiselle, S., Carstens, J., Van der Poll, S., 2001. Diversity
of the rural farming households and policy issues: an analysis based on a case
study in the Northern Cape Province. African Institute for Southern Africa,
‘South Africa since 1994’.
Aubert, D., Lifran, R., Mathal, P., Perraud, D., Viallon, J.B., 1985. Systèmes de
Production et Transformation de l’Agriculture, 2 vol. INRA-ESR, Paris.
Auclair, L., Ben Cheikh, K., Laajili-Ghezal, L., Pontanier, R., 1996. The use of sylvo-
pastoral resources and production systems of the ‘haut-tell’ (Tunisia). In:
Budelman, A. (Ed.), Agricultural R&D at the Cross-Road. Merging Systems Research
and Social Actors Approaches. Royal Tropical Institute, the Netherlands, pp. 55–67.
Beaudeau, F., 1994. Cow’s Health and Farmer’s Attitude Towards Culling Decisions in
Dairy Herds. Ph.D. thesis, Wageningen University, the Netherlands, 215 pp.
Blanc, M., Brun, A., Delord, B., Lacombe, P., 1990. L’agriculture française est-elle
encore familiale? In: Coulomb, P., Delorme, H., Hervieu, B., Jllivet, M.,
Lacombe, P. (Eds.), Les Agriculteurs et la Politique. Presses de la Fondation
Nationale des Sciences Politiques, Paris, pp. 310–327.
Bonnafous, P., Revel, A., 2004. La diversification innovante des exploitations agricoles,
une alternative au modèle spécialisé? Les Cahiers de la multifonctionnalité 7, 71–84.
Bonnieux, F., 1982. Les disparités départementales de l’agriculture française de 1962
à 1980. Economie Rurale 152, 34–39.
Brun, A., Lacombe, P., Laurent, C., 1970. Les Agricultures à Temps Partiel dans
l’Agriculture Française. INRA-SCEES, Paris, 144 pp.
Cabello Norder, L.A., 2004. Politicas de Assentamento e Localidade; os desafios da
reconstitucao do trabalho rural no Brasil. PhD thesis, Wageningen University,
the Netherlands, 313 pp.
Cazella, A.A., 2001. Les Instalations Agricoles Nouvelles: les Cas des Agriculteurs
Néo-ruraux dans l’Aude (France). Espace, Populations, Sociétés, Liège, Belgique.
Cazella, A.A., 2006. Contribuições metodológicas da sócio-antropologia para o
desenvolvimento territorial sustentável. Eisforia (UFSC) 4, 225–247.
Deuffic, P., Candau, J., 2006. Farming and landscape management: how French
farmers are coping with the ecologization of their activities? J. Agric. Environ.
Ethics 19, 563–585.
De Vries, W., 1995. Pluri-activiteit in de Nederlandse landbouw. In: Studies van
Landbouw en Platteland, 17. Wageningen Agricultural University, Wageningen.
Di Medici, G., 1952. Italy, Agricultural Aspects. Il Mulino, Bologna.
Dumont, R., 1970. Types of Rural Economy: Studies in World Agriculture, [s.n.], London.
EC (European Commission), 2003. Planting seeds for rural futures. Rural policy
perspectives for a wider Europe. Second European Conference on Rural
Development, 12–14 November 2003, Salzburg, 204 pp.
Groen, A.F., De Groot, K., Van der Ploeg, J.D., Roep, D., 1993. Stijlvol fokken, een
orienterende studie naar de relatie tussen sociaal-economische verscheiden-
heid en bedrijfsspecifieke fokdoeldefinitie, Bedrijfsstijlenstudie no.9. Vakgroep
Veefokkerij en Vakgroep Rurale Sociologie, Landbouwuniversiteit, Wageningen.
Hayami, Y., Ruttan, V.W., 1985. Agricultural Development: An International
Perspective (revised and expanded edition). John Hopkins, Baltimore and
London.
Hebinck, P., Monde, N., 2007. Production of crops in arable fields and home gardens.
In: Hebinck, P., Lent, P.C. (Eds.), Livelihoods and Landscapes: The People of
Guquka and Koloni and their Resources. Brill, Leiden.
Hofstee, E.W., 1946. Over de oorzaken van de verscheidenheid in de Nederlandse
landbouwgebieden. Rijksunversiteit Groningen, Groningen, the Netherlands.
Lacombe, P., 1983. Ou va l’Agriculture Française? Questions et Débats. Economie
rurale 155, 64–69.
Laurent, C., 1991. L’Agriculture et son Territoire dans la Crise. PhD Thesis. Economics.
Univ. Paris VII.
Laurent, C., 2005. Les agricultures de l’Union Européenne. In: Laurent, C., Thinon, P.
(Eds.), Agricultures et Territoires. Hermes Sciences, Paris, pp. 31–55.
Laurent, C., Bowler, I., 1997. CAP and the Regions, Building a Multidisciplinary
Framework for the Analysis of the EU Agricultural Space. INRA, Sciences up-
date, Paris.
Laurent, C., Remy, J., 1998. Agricultural holdings: hindsight and foresight. Etudes et
Recherches des Systemes Agraires et Developpement 31, 415–430.
Laurent, C., Langlet, A., Chevallier, C., Jullian, M., Maigrot, J.-L., Ponchelet, D., 1994.
Ménages, activité agricoles et utilisation du territoire: du local au global à travers
les recensements Généraux de l’Agriculture. Cahiers Agriculture III (2), 93–107.
Laurent, C., Cartier, S., Fabre, C., Mundler, P., Ponchelet, D., Remy, J., 1998. L’activité
agricole des ménages ruraux et la cohesion économique et sociale. Économie
Rurale 224, 12–21.
Laurent, C., Van Rooyen, J., Madikizela, P., Bonnal, P., Carstens, J., 1999. Household
typology for relating social diversity and technical change. Agrekon 38, 190–208.
Le Boterff, M., 1988. Le secteur agricole et agro-alimentaire et la crise économique:
des réponses économiques et sociales diversifiées. Rapport présenté au nom du
Conseil économique et social. Journal Officiel, Paris, 96 pp.
 J.D. van der Ploeg et al. / Journal of Environmental Management 90 (2009) S124–S131
Leeuwis, C., 1993. Of Computers, Myths and Modelling: the social construction of
diversity, knowledge, information and communication technologies in Dutch
horticulture and agricultural extension. PhD Thesis, Wageningen University,
Wageningen, the Netherlands.
MacSharry, R., 1990. Préface. Cadre Communautaires d’Appui pour le Développe-
ment des Zones Rurales (Objectif 5b). Commission de la Communuté Europé-
enne, Luxembourg.
Modiselle, S., 2001. Rural Households Diversity in the Liliefontein District of the
Northern Cape Province of South Africa. master thesis, Dept of Agricultural
Economics, Extension and Rural development. University of Pretoria, 112 pp.
OECD, 2000. Multifunctionalitydtowards an analytical framework. http://www.
oecd.org/. AGR/CA/APM(2000)3/FINAL, Paris.
Oostindie, H., Van Broekhuizen, R., 2004. Landbouw en platteland in de Wolden:
een studie naar agrarische ontwikkeling, verbrede landbouw en nieuwe bed-
rijvigheid in voormalige boerderijen in de gemeente Wolden. Wageningen
Universiteit, Wageningen.
Perrot, C., Landais, E., 1994. Research into Typological methods for Farm Analysis.
The Why and the Therefore. Systems Studies in Agriculture and Rural Devel-
opment. Brossier, de Bonneval, Landais Ed. pp. 373–392.
Reijs, J., 2007. Improving Manure through Diet Adjustments: A Novelty to Reduce
Nitrogen Losses from Grassland Based Dairy Farms. Ph.D. thesis, Wageningen
University, Wageningen.
Remmers, G., 1998. Con cojones y maestria: un studio sociologico-agronomico
acerca del desarollo rural endogenmo y proceses de localizacion en la Sierra de
la Contraviesa. Wageningen University, Wageningen.
S131
Renting, H., Oostindie, H., Laurent, C., Brunori, G., Barjolle, D., Jervell, A.,
Granberg, L., Heinonen, M., 2008. Multifunctionality of agricultural activities,
changing rural identities and new institutional arrangements. Int. J. Agric. Res.
Govern. Ecol. 7 (4/5), 361–385.
Saccomandi, V.,1998. Agricultural Market Economics: a Neo-Institutional Analysis of the
Exchange, Circulation and Distribution of Agricultural Products. Van Gorcum, Assen.
Sonneveld, M.P.W., 2004. Impressions of Interactions: Land as a Dynamic Result of
Co-Production between Man and Nature. PhD Thesis, Wageningen University,
Wageningen.
Van Averbeke, W., Mohamed, S.S., 2006. Smallholder farming styles and develop-
ment policy in South Africa: the case of Dzindi Irrigation Scheme. Agrekon 45
(2), 136–157.
Vanclay, F., Mesiti, L., Howden, P., 1998. Styles of farming and farming subcul-
tures: appropriate concepts for Australian rural sociology? Rural Society 8,
57–160.
Van der Ploeg, J.D., 2003. The Virtual Farmer: Past, Present and Future of the Dutch
Peasantry. Royal van Gorcum, Assen.
Van der Ploeg, J.D., Saccomandi, V., 1995. On the impact of endogenous develop-
ment in agriculture. In: Van der Ploeg, J.D., Van Dijk, G. (Eds.), Beyond
Modernization. Van Gorcum, Assen, pp. 10–27.
Van Huylenbroeck, G., Durand, G., 2003. Multifunctional Agriculture: a new para-
digm for European agriculture and rural development. Ashgate, Aldershot.
Verhoeven, F.P.M., Reijs, J.W., van der Ploeg, J.D., 2003. Re-balancing soil-plant-
animal interactions: towards reduction of nitrogen losses. NJAS Wageningen.
J. Life Sci. 51, 147–164.