FARMING SYSTEMS RESEARCH
IN THE CONTEXT
*Arising out of Workshop on Farming Systems Research in Mali sponsored by the
Institute d'Economie Rurale (IER) and the Ford Foundation, and held at IER,
Bamako, Mali, November 14th-19th, 1976. D. W. Norman
-,, /Y77 (4 7,er,,,
FARMING SYSTEMS RESEARCH IN
THE CONTEXT OF MALI
1. INTRODUCTION 1
1.1 The Conference
1.2 Objectives of Conference
1.3 Justification for Farming Systems Research
1.4 Organisation of discussion and synthesis
2. SCHEMATIC FRAMEWORK 1
2.1 Need for Schematic Framework
2.2 Stages, elements and processes in the Schematic Framework
2.2.1 Central role of the farmer
2.2.2 Four stages of research
2.2.3 Elements in Schematic Framework
*2.2.4 Technical Element
2.2.5 Human Element
22.214.171.124 Exogenous Factors
126.96.36.199 Endogenous Factors
2.3 Miscellaneous points concerning the Schematic Framework
2.3.1 Dynamic iterative and linkage nature of the
2.3.2 Types of technology
2.3.3 Attune technologies to the realities of the
2.3.4 Potential for shortcuts
2.3.5 The Extension Stage
2.3.6 Relationship between degree of emphasis and
3. IMPLICATIONS FOR DISCIPLINES 5
3.1 Basic disciplines to be involved
3.2 Contribution by other disciplines
3.3 Farmers as a discipline?
3.4 Extension as a discipline?
4. STAGES AT WHICH DISCIPLINES SHOULD BE INVOLVED 6
4.1 Design and Test Stages
4.2 Description Stage
. IMPLICATIONS FOR ORGANISATION 6
5.1 Grouping disciplines in one unit or department
5.2 Interaction with other disciplines and departments not precluded
5.3 Selection of leader of department
6. METHODOLOGY DESCRIPTIVE STAGE
Procedure of data collection
Analysis of data
Proposed changes in infrastructural support system
7. METHODOLOGY DESIGN STAGE
7.2 Screening trials on experiment station
7.3 Time required for Experiment Station Trials
7.4 Modelling cuts down time required for Experiment Station
7.5 Design and backward linkage to government policy
8. METHODOLOGY TESTING STAGE
8.1 Two parts of Testing Stage
8.2 Trials at Farmers Level
8.2.2 Simplicity of trials
8.2.3 Selection of farmers
8.3 Farmer Testing
8.3.3 Selection of farmers
9. METHODOLOGY EXTENSION STAGE
Evaluation units and the Descriptive Stage
Role of evaluation units
figure i. bcnematic kramewo- tor arming systems esea~i .Lu
FARMING SYSTEMS RESEARCH IN
TIE CONTEXT OF MALI
1.1 The synthesis presented in this paper represents what is thought to be
the consensus emerging from a Farming Systems Conference held in Bamako,
November 14th-19th, 1976. The conference was sponsored by IER, Mali and
the Ford Foundation.
1.2 Broadly speaking the objectives of the conference, which kept Mali
specifically in mind, were:
(a). To establish whether or not a Farming Systems Research
Programme is desirable and needed.
(b). If such a Programme is desirable, give guidelines on the
procedures to follow.
1.3 It was agreed that in the Mali setting a Farming Systems Research
Programme could be justified for the following reasons:
(a). The government of Mali has a genuine desire to improve the
welfare of people in agriculture.
(b). It hopes to do this via a somewhat broader basis in terms
of possible crops and livestock that can be grown, compared
with the usual single cash crop orientation of the past.
This implies more of a systems approach.
(c). Although improvement of the welfare of people in agriculture
is an aim, it is intended that this will involve voluntary
rather than enforced change on the part of the farmer. This
implies a greater understanding of the needs of farmers and
the designing of technology that will satisfy these needs.
1.4 In spite of a fair proportion of the conference participants being
engaged in Farming Systems Research, there has, as yet, been no uni-
versally followed approach to such research. It was apparent, however,
that a conventional wisdom is now emerging concerning such work.
Accordingly the following procedure was adopted:
(a). A schematic framework for undertaking research work on Farming
Systems was agreed on.
(b).' This led on to considering the implications for the:
(i). Disciplines to be involved in Farming Systems Research.
(ii). Organisation required for a Farming Systems Research
Programme to be truly operational.
(iii). Methodology to be utilised in undertaking Farming
Systems Research work.
2. THE SCHEMATIC FRAMEWORK
2.1 It was
decided that a schematic framework was desirable because:
It helps channel the thoughts of individuals who will
hopefully engage in Farming Systems research.
(b). It demonstrates the complexities of designing and comple-
menting an effective Farming Systems Research Programme
both in terms of the elements that need to be taken into
account and the steps that have to be undertaken to de-
velop relevant improved technology.
2.2 As a result of a number of papers and extensive discussions, the
Schematic Framework as delineated in Figure 1 was agreed to reflect
the consensus of those present at the conference. A number of points
can be made about the Schematic Framework. They are as follows:
2.2.1 The farmer plays the central role in the Farming Systems
Research Programme. The broad objective of such a programme
is to develop technologies which will overcome the critical
constraint postulated to be faced by farmers, thereby empha-
sising an increase in the productivity of the farmers' re-
sources and his welfare as a result of modifying the existing
system, rather than completely changing it.
2.2.2 To obtain this, four stages of research can be delineated.
(a). The Descriptive Stage in which the actual farming
system as it exists now is examined in order to
ascertain what are the real constraints of farmers
and therefore what types of technology are required
in order to overcome them.
(b). The Design Stage in which a range of technologies
which are thought relevant to overcoming the con-
straints arising from the Descriptive Stage are
tested under Experiment Station conditions.
(c). The Testing Stage in which a smaller number of the
more hopeful technologies arising from the Design
Stage are examined and evaluated on farmers' fields.
(d). The Extension Stage in which the technology (or
technologies), which was (were) found during the
Design and Testing Stages to best overcome the con-
straints delineated in the Descriptive Stage, are
extended to farmers.
2.2.3 In order to describe the existing farming system and indeed to
develop, test and extend relevant technologies, it is important
to understand the:
(a). Technical Element. This reflects what the Farming
System can potentially be, and therefore provides
the necessary condition for the presence of a par-
ticular Farming System.
(b). Human Element. This reflects what the Farming System
will actually be, is a subset of those technically
possible (see 2.2.3(a)), and therefore provides the
sufficient condition for the presence of a particular
2.2.4 The Technical Element has in the past received greatest
attention, particularly by technical scientists. They
have, within certain limits, been able to modify the
Technical Element and thereby improve the potential farm-
ing system by developing technology that will partially
alleviate the effects of one of the Technical Factors which
(a). Physical Factors (e.g., water, soil, solar
radiation, temperature). Technical scientists
can, for example, within certain limits, enhance
the water availability through irrigation, soil
quality through fertilizer application, etc.
(b). Biological Factors (e.g., crop and animal physi-
ology, disease, insect attack, etc.). Examples
of limited intervention of technical scientists
would include breeding earlier maturing varieties
of crops, breeding for disease resistance, etc.
2.2.5 The Human Element, which in the past has received little
attention in "traditional" agricultural research, now plays
a very significant role in Farming Systems Research. This
element can be divided into two factors. They are as follows:
188.8.131.52 Exogenous Factors (i.e., the social environment)
or those that are largely outside the influence
of the individual farmer. Simplistically these can
be considered in two broad groups:
(a). Community structures such as customs of
the community, the way in which the
society is organised at the village level,
(b). Infrastructural factors which in develop-
ing countries are usually financed and
manned by government and therefore reflect
their policies. Very simply these can be
(i).The convincing factor, which involves
an input by extension staff and perhaps
the explicit provision of a market for
the products) produced.
(ii).Since most types of improved technology
cost money, ensuring that farmers have
the necessary financial resources at
the time they are required in order to
(iii).Ensuring the improved inputs required
for the adoption of the improved tech-
nology are distributed in sufficient
quantities to the right places at the
184.108.40.206 Endogenous Factors (i.e., the individual decision maker)
or those that are under the influence of the individual
farmer. A farmer initially has access to certain speci-
fied quantities and qualities of the four factors of
production (i.e., land, labour, capital and management).
These may, depending on the circumstances and the
farmer's wishes, be complemented and supplemented
on quantitative and/or qualitiative terms (i.e.,
especially with reference to capital and manage-
ment, see 220.127.116.11(b) above). Subject to his con-
straints and attitudes the farmer then allocates
the factors of production he has at his disposal,
through three possible processes (i.e., cropping
enterprises, livestock enterprises, and other
enterprises) in order to derive a farming system
that attains as far as possible, within the knowledge
he possesses, the goal (or goals) he has in mind (e.g.,
increasing his chances of (social) survival taking
one year with another, maximising net income for a
given mean level and variance of output, etc.).
2.3 There are a number of points arising out of the discussion of the com-
ponents of Farming Systems Research discussed in the preceding section
(i.e., 2.2). They are as follows:
2.3.1 The process of Farming Systems Research is dynamic, iterative and
in the last three stages involves evaluation in terms of whether
the technology designed, tested and extended overcomes the con-
straints delineated in the Descriptive Stage. The evaluation
criteria are applied at each of the last three stages with a
steady narrowing down of the number of possibilities as progress
moves from the beginning of the Design to the end of the Test and
Extension Stages. Failure to obtain any satisfactory result in
terms of technologies to overcome the specified constraints necess-
itates the use of backward linkages or returning to a previous
stage, to repeat preceding step or steps.
2.3.2 The technologies required to overcome specific constraints may
range from very small simple items to large complex packages.
2.3.3 The technologies designed, tested and extended must be attuned
to the realities of the "total" environment (i.e.,technical and
human). Physical, Biological and increasingly Endogenous Factors
are likely to be taken into account in producing improved tech-
nologies but often such work still fails to take into consideration
the Exogenous Factors which include the infrastructural elements
reflecting government policy. For example, the problem of elimi-
nating the presence of a particular disease on a crop may be solved
by spraying or breeding a resistant variety. The former solution
would require considerable managerial expertise in terms of timing
of sprays, would require the purchase of spraying equipment and
materials and, since it is likely to be fairly expensive, may re-
quire an institutional credit programme. This implies the necessity
of having present at the time the technology is extended a fairly
strong infrastructural support system. This is not so in the case
of a resistant variety which simply involves the distribution of
the improved seed. Unfortunately much improved technology has
been developed assuming the presence of a strong infrastructural
support system rather than a low level one which is the reality
'of much of the developing world. Less spectacular technology in
terms of managerial input, quantities of improved inputs and hence
lower cash inputs are required in such situations, although it is
accepted that such technology will of necessity tend to have
a lower economic payoff.
2.3.4 The Schematic Framework presented earlier presents a logical
although a time consuming approach to Farming Systems Research.
Unfortunately the possibilities of short cuts are somewhat limited.
Of key importance in this regard is the Body of Knowledge (see
Figure 1) external to the Farming Systems Research Programme which
reflects knowledge, technologies, etc. that have been accumulated
and developed by other scientists. If relevant technologies
already exist that will overcome the constraint defined at the
Descriptive Stage, it should be possible to reduce or even elimi-
nate the time for Experiment Station Trials and instead emphasis
can immediately be placed on Trials at the Farm Level or even at
the Farmers Testing Level.
2.3.5 It is likely that the Extension Stage will be undertaken by insti-
tutions and individuals external to those in the Farming Systems
Research Programme itself. However it is important that the Farming
Systems Research Programme maintains close links with the Extension
Stage particularly those involved in the evaluation part of that
stage. That evaluation is likely to be undertaken by evaluation
units on development projects (e.g., IBRD, CMDT, etc.) and more
informally by extension workers.
2.3.6 It is recognized that the degree of emphasis on the various stages
and elements will to some extent be a function of the state of
knowledge about the place in which the Farming Systems Research
Programme is located.
3. IMPLICATIONS FOR DISCIPLINES
3.1 It was agreed that in the Mali setting with its emphasis on both crop and
livestock production that the basic Farming Systems Research team should con-
sist of an agronomist, animal scientist and socio-economist.
3.2 Strong cases were also made for other disciplines particularly breeder, soil
scientist and agricultural engineer. However, it was agreed that contributions
by other disciplines would tend to be to a great extent location and/or problem
specific and could be exploited or contributed through the linkage with the
Body of Knowledge (Figure 1). This linkage would be two way with those disci-
plines in the Farming Systems Research Programme requesting help to overcome
specific problems from disciplines outside the programme and those disciplines
deriving relevant solutions to them.
3.3 The inclusion of farmers as one of the disciplines in the Farming Systems
Research Programme was also suggested. However it was generally agreed that it
Swas not satisfactory to consider farmers as a research discipline. However the
spirit of the suggestions was appreciated and it was emphasised once again the
central-role to be played by farmers who are intimately involved in the De-
scriptive, Test and Extension Stages.
3.4 Some also believed that extension should also be included as a discipline
in the Farming Systems Research Programme. However once again it was generally
agreed that extension is not a research discipline as such but that the ex-
tension services should be consulted and their advice sought at all stages of
the research programme. They can also easily be involved at the Farm Testing
4. STAGES AT WHICI DISCIPLINES SHOULD BE INVOLVED
.4.1 It is obvious that all three disciplines should be involved at the
Design and Test Stages.
4.2 There is also a strong argument for the inclusion of all the disci-
plines at the Descriptive Stage. Unfortunately this is rarely done at
the present time. However, it is believed that in order to understand
how, why and what farmers do what they do, which is the underlying basis
of this stage, it is important to appreciate the interaction between
the techncial and human elements of the environment. For example, the
reason why the yields of a particular crop were low may be due-not to
laziness on the part of the farmer but because planting was delayed due
to late rains, a particular disease was present, etc. Therefore unless
proper is achieved between the technical and social science disciplines at
the Descriptive Stage, it is quite likely that the true constraints facing
the farmer will not be correctly identified. This argument which received
overwhelming support means a departure from the traditional approach which
has generally involved only the social scientist at the stage with techni-
cal information being collected largely from secondary sources. Certainly
interaction between the technical and social science disciplines at this
stage has been very rare indeed.
5. IMPLICATIONS FOR ORGANISATION
5.1 As has been emphasised there are three basic disciplines that should be
involved in research on farming systems (See Section 3.1). At the same
time it has been emphasised that this multi-disciplinery team should act
as an inter-disciplinery group. In order to get the control and interaction
required it is essential that the representatives of the three basic dis-
ciplines involved in such Farming Systems Research should not be left in
separate departments, which are usually segregated by disciplines, but
should be placed under one "umbrella" or a department in its own right.
5.2 This organisational arrangement does not of course preclude cooperation
and interaction with other departments and disciplines (See Sections 2.3.4
5.3 The head of the Farming Systems Research Unit or Department could be
either an agronomist, animal scientist or socio-economist. In other words,
selection of the leader should be based on personality, leadership qualities,
and an ability to think in terms of a systems framework rather than simply
in terms of one discipline.
6. METHODOLOGY DESCRIPTIVE STAGE
6.1 The objective of the stage is to identify the major constraints) faced
by farmers, which can be alleviated by the designing, testing and extension
of relevant technologies.during the later stages of the research programme.
The-objective of this is to increase the farmers' welfare which may, depend-
ing on the circumstances, be articulated in terms of production, security
6.2 The following steps were agreed as being necessary for implementing
the Descriptive Stage:
(a). First of all an inventory should be made of the existing in-
formation concerning the geographical area in which the
Farming Systems Research Programme is to be located. When
necessary and possible such existing data should be further
analysed in an attempt to fulfill the objective in Section 6.1
(b). This should be followed by a field tour by the three team
members who as a group interact with each other and with
government officials, extension officials, other research
workers, but above all with farmers themselves. Methods such
as observing, interviewing, etc., will help familiarise them
with the area and hopefully will bring to their notice major
problems facing farmers.
(c). If baseline data arising from (a) and (b) above are not suf-
ficient, then a baseline study should be undertaken, consisting
of only one interview per farmer, which should complete a basic
quantitative description of the technical and human elements
of the farming systems in the area under investigation.
(d). Information arising from (a), (b), and, if undertaken, (c) above,
can form the basis of more formal detailed sample surveys to fill
in gaps and/or confirm, or result in the determination of, the
major constraints) faced by farmers. The type of sample surveys
required will be to a great extent location and/or problem specific.
They can, depending on circumstances, take many forms, for example,
agro-economic (i.e., technical-human), farm management, market
opinion surveys, etc. The type of survey and the questions it was
seeking to answer will largely determine the sampling procedure and
stratification system to be used. Interviewing at frequent inter-
vals may be desirable in these surveys, although this can be very
expensive in terms of time, money and manpower.
(e) An alternative approach to (d), which was agreed to be more real-
istic in the Mali setting was to undertake a few detailed case
studies, thereby emphasising quality rather than quantity in terms
of data collection. This of course ignores the desirability of
reducing sampling errors implicit in (d) above but assigns a major
role to the importance of reducing measurement errors, which is
important if an in-depth understanding of the farmers constraints)
is to be achieved. The skill of course lies in selecting farmers
for such case studies that are representative of the largest group.
For this purpose some subjective modification of the sampling
procedure may be necessary.
6.3 It is important that because of the importance of the time factor in
delineating the constraints) facing the farmer as soon as possible, that
processing and tabulation of the data is commenced immediately the survey
commences (see (d) and (e) in Section 6.2) and that analysis should be
completed as soon as possible after the data collection phase has finished.
Consideration of this may well be a factor that should be considered in
deciding the sample sizes in (d) and (e) above.
6.4 Although the Descriptive Stage concentrates on what the situation
is now, there is merit when moving into the Design and Testing Stages
to take into account possible changes in government policies with re-
spect to infrastructural systems. For example, the definite initiation
of a development project within the next two years, which will greatly
increase the infrastructural support system, can be taken into account
in designing and testing the type of technology that will most easily
overcome the constraints) determined at the Descriptive Stage.
7. METHODOLOGY DESIGN STAGE
7.1 The objective of this stage is to develop technologies to solve farm
level constraints(s) identified in the Descriptive Stage.
7.2 To achieve this objective initial screening of possible technologies'
is on the basis of inter-disciplinary trials undertaken on the Experiment
Station and Sub-Stations. Evaluation in terms of the technologies that
should be taken to the Test Stage is based on those that best overcome
the constraints, delineated in the Descriptive Stage, at least cost.
7.3 Due to the complexity of the trials (e.g., Latin squares, randomised
split blocks, etc.), it is usually necessary to undertake these trials
on Experiment Stations. Nevertheless the time for such "artificial"
trials should be minimised as much as possible and progress to the Test-
ing Stage be encouraged. However, as has been emphasised earlier (see
Section 2.3.4), the time required for Experiment Station Trials will be
largely determined by the availability of relevant technologies developed
outside the Farming Systems Research Programme.
7.4 Another possible way of reducing the time for such trials, and there-
fore increasing the efficiency of deriving relevant technologies, is
through modelling such as simulation, linear programming, etc. The
question can be raised in the Mali setting as to whether its limited
availability of resources (i.e., skilled manpower and equipment) would
permit the realistic use of this method. Unfortunately for livestock
with their relatively long life spans, this approach may be very im-
portant. If such an approach was deemed necessary it would be highly
desirable to have close collaborative links with an external institution
which has skills in modelling.
7.5 Although the design of relevant technologies usually should take into
account the existing, or definitely expected, infrastructural support
system (i.e., Exogenous Factors), there is a role occasionally at this
stage for communicating to government the potential returns from a tech-
nology if a given support system is implemented (e.g., spraying regime).
If an attempt at such moral suasion fails to convince government to change
its support system in order to accommodate the particular technology, then
that technology should not be carried on into the Testing Stage, since it
will be impossible for farmers to adopt the technology in the absence of
the relevant infrastructural support system.
8. METHODOLOGY TESTING STAGE
8.1 The Testing Stage, it was agreed, should be divided into two parts,
namely Trials at the Farmer Level and Farmer Testing.
8.2 The following points apply to Trials at the Farmer Level:
8.2.1 The objective of this phase is to evaluate the technologies
screened from the Design Stage in a real world situation using
farmers' land and labour but with the managerial input still
being provided by the research workers. The evaluation will
be based on whether or not the technologies overcome the con-
straints delineated in the Descriptive Stage.
8.2.2 These trials, which must represent what are thought to be the
most relevant options arising from the Design Stage, must be
limited in number because it is necessary to obtain:
(a). An idea of the variation of the treatments which
requires emphasis on replications rather than more
(b). The farmer's impression of the technologies which
would be difficult for him to do adequately if there
were a large number of treatments.
8.2.3 A point of disagreement, which was not resolved was concerning
the type of farmers to be selected for these trials.
(a). View 1. One group preferred selecting the keener or
more enlightened farmers to participate in the trials.
While this has the advantage of maximising the interaction
between the research workers and the farmers, it has the
potential problem that if the technology receives a posi-
tive evaluation it may still not be truly relevant for
the "average" farmer.Therefore the adoption process may
be biased towards the enlightened farmers therefore causing
inequalities to arise in terms of benefits in the long run.
(b). View 2. The other group advocated selecting a cross-section
of farmers representative of the target group. This approach
has the disadvantage of not maximising the interaction between
farmers and research workers since not all the farmers are
likely to be particularly keen or enlightened. However it
does have a big advantage in that it gives a more satisfactory
idea of whether the technology is likely to be adopted by
the "average" farmer.
8.3 The following points apply to Farmer Testing:
8.3.1 The objective is to evaluate the performance of the proposed technology
in selected farm environments using farmers land, labour and manage-
8.3.2. Tlere appeared to be a slight divergence of opinion between
those who wish to evaluate the proposed technology on a dif-
fusion basis (i.e., numbers of farmers who adopt it over time.)
and those who wanted a quicker final test prior to extension.
The former approach is certainly important at the Extension
Stage but the advantage of the latter approach is that a final
test of relevancy is imposed to determine whether or not its
extension should be officially supported or whether the backward
linkages to trials at the farmers' level,etc., need to be ex-
ploited. In terms of the latter approach, which appeared to re-
ceive more support, it was suggested that evaluation could be
expressed in internal and external terms. In internal terms,
which comes into the purview of the Farming Systems Research
Programme itself, the technology is basically evaluated to ascertain
whether or not it overcomes the constraints) delineated in the
Descriptive Stage. The technology should also be looked at in
(a). Technical feasibility as far as the farmer is concerned.
(b). Assurance that it is profitable and dependable when ex-
pressed in terms of the most limiting resource (e.g.,
land, labour, money, labour input during June and July,
(c). Social acceptability as far as the farmer is concerned.
(d). Its compatibility with the farming system the farmer
(e). Its compatibility (i.e., as a final cross-check) with
the existing infrastructural support system or sometimes
with what is definitely expected in the very near future
(see Section 6.4). There is also the complex question
of the external evaluation of the technology being pro-
posed. This essentially brings into consideration a
longer run perspective than that of the internal evalu-
ation. As a result criteria for such evaluation are
not so easy to articulate but they would attempt to
evaluate, for example, the potential effects of adopting
the technology on the balance of the physical eco-system,
welfare of urban residents, etc.
8.3.3 The question concerning the selection of farmers for field testing
received analogous arguments to those for selecting farmers to
participate in trials at the Farmers' Level (see Section 8.2.3).
Once again it was not resolved satisfactorily. What was agreed,
however, was that research work designed to overcome farmers con-
straint(s) should result in minimal technological packages that
are applicable to as wide a range of farmers as possible.
9. METHODOLOGY EXTENSION STAGE
9.1 The objective of the Extension Stage is to ensure that as large a section
possible of the target group farmers should benefit from the technology being
9.2 The actual implementation and evaluation of the Extension Stage is ob-
viously not the responsibility of those involved in the Farming Systems
Research Programme (see Section 2.3.5). However close links with units
responsible for evaluation at this stage could be of considerable value in
helping to assess the success of technology already recommended for ex-
tension, and for helping ascertain future priorities for research based
on the new constraints which result. Evaluation units should, for example,
be-able to answer the following questions:
(a). Who is benefiting from the improved technology?
(b). To what degree are they benefiting?
(c). If some have not adopted the improved technology, why is this
the case? Is it because the technology was not relevant for'
some farmers (i.e. did not overcome the most pressing constraintss?
This would imply that the target group of farmers were not as
homogenous in the constraints) they faced as was originally thought.
This would also imply the need for developing different technologies
for different strata of the farming population. Or is the failure
to adopt on the part of some of the farmers due to the fact that
there has been differential access to, and hence benefits from, the
infrastructural support system (e.g., it being biased towards larger
farmers). In such cases evaluation units can be of use in giving
government some indications of the trends occurring and can hope-
fully propose a more equitable infrastructural support system.
9.3 Although evaluation units on development projects should usually be
involved in the type of work discussed in the preceding section (Section 9.1)
it is also true, that owing to a lack of knowledge about the areas, when
development projects commence, that they also can play a role in the Descrip-
tive Stage (see Section 6.2). This can be extremely useful in spite of the
fact that this information is collected at a later stage than would be
9.4 The evaluation units can therefore provide an invaluable role in ensuring
the dynamic iterative role of Farming Systems Research.
At the conference it was apparent that the majority of participants were
convinced of the importance of a Farming Systems approach to agricultural re-
search. This was based on the recognition that in the past little improved
technology developed by technical scientists, who largely concentrated on the
Technical Element and often only undertook trials on the Experiment Station,
was in fact adopted by the mass of small farmers. The Farming Systems approach
attempts to ensure that the improved technology that is developed is more
relevant to the actual situation, thereby increasing the prospects of it being
adopted by farmers. This is done by taking into consideration, in addition,
the Human Element and by doing as much as possible of the development work
at the farmers level.
It is obvious that the synthesis presented here has many gaps, for example,
particularly with respect to sampling procedures, survey and trial designs,
analysis and evaluation. It also needs considerable refinement but at least
it provides a starting point for further discussion.
A note of caution is appropriate at this point. It is likely that such a
team as envisaged in the synthesis would have to confine its activities to a
part of Mali in order to have any impact. However such a team could be repli-
cated to work in other areas of the country. Also it should be appreciated
that this type of research is time consuming and it would be unreasonable to
expect any major breakthroughs in the short run. However, it is believed that
this approach provides the most rational way of developing relevant im-
proved technologies in the realities of the present day world.