Past, Present and Future 173neurs’ (which in fact has already been raised by Constandse in 1964). They are
peasants, and precisely because they are peasants they have proved to be so success-
ful through the ages. The fact that this is also extremely topical and relevant nowa-
days is discussed in Chapter 5.
The transformation or conversion of resources into end products is discussed
at length in Chapter 4. This transformation (for example from grass and cows into
milk and meat) is conceptualized in the agricultural expert system as regulated by
production functions: a fixed relationship between inputs (grass and cows) and
outputs (milk and meat). It is assumed that new, more productive and/or more
sustainable production functions can only be created after technological break-
throughs.
In contrast to this thesis, I want to develop another image. Indeed, there are
regularities and patterns of coherence. Unmistakably, the relation between inputs
and outputs (the I/O relation) displays a certain range and contains a certain, as
yet uncrossable, frontier function (the most efficient I/O relation) in certain, tem-
porally and spatially restricted situations. But what is important here is that these
patterns or regularities are a product, a result of the labour process in agriculture.
Furthermore, I regard the labour process as locus of co-production, of the con-
tinuous encounter between, and mutual transformation of, the social and the
natural. Precisely because the results of this co-production are rooted in, and arise
from, the labour process, it will always involve extremely variable I/O relations.
New production functions are constantly created in and through the process of
farm labour: new patterns of coherence that correspond closely with farmers’ own
interests, perspectives, insights and knowledge.
In previous studies of farming styles, researchers concentrated initially on the
above-mentioned dimensions: the conversion and mobilization of resources. This
initially limited approach was later broadened, thanks to the work of Kerkhove
(1994), Ventura and Van der Meulen (1994), De Rooij et al (1995), Wiskerke
(1997), Van der Meulen (1999a, 1999b) and Roep (2000). The studies identified
the sale of end products and, hence, the specific circuits connecting production
and consumption as ordering principles, as essential types of interlocking, and
therefore as crucial parts of socio-technical networks. I will return to these new
insights in Chapter 9, where new development opportunities for Dutch agricul-
ture will be discussed.
What is important in this first storyline is that the interconnected domains of
mobilization, conversion and sale contain considerable variability – each sepa-
rately, but above all collectively. They are malleable, flexible, changeable. The
resulting variability cannot be understood as a ‘fixed’ range: it is constantly enlarged
in the practice of farming – that is, by farmers’ innovative capacity.
Similarly, the potential variability is constantly restricted, if not reduced. If
two keywords describe agricultural history and its development of production
capacity, they are variation and selection. On the one hand, there is a constant
search for new patterns, and for new combinations, to further increase the already
available variation. On the other hand, some patterns, and some combinations,