Peter J. Brumby
Director General, ILCA, Ethiopia
May I warmly welcome you here this morning. This workshop on our research in Pastoral Production Systems has been in the making for several years, plus a bit longer, the bit longer being the additional time we needed for the hard work in preparing the papers for this meeting, as well as the time needed to develop the courage to expose our experience and ideas to you.
Very little has been written on the analysis of the livestock aspects of farming systems research (FSR). The 1981 compendium of Farming Systems Research in Africa contains virtually no references on the subject. But this situation is changing. In 1982, IDRC published an excellent monograph on issues and policies influencing livestock development in Asia and this publication contains several valuable papers on the analysis of livestock systems. Winrock International are also increasingly productive in this subject, particularly in respect of Latin America. We at ILCA have now published nine major reports on various livestock farming systems in different parts of Africa. Excellent journals on the subject of FSR generally are also available; unfortunately, however, they contain very little livestock material.
Systems analysis is a term, like so many others that will be used in this workshop, that means different things to different people. To a computer engineer the meaning of systems analysis is very clear, but we use the term in a broader and less precise manner. The phrase 'systems analysis' or 'farming systems research', or the fancier semantic derivatives that describe a fancier research focus that extends beyond the farm level, such as 'farming systems perspectives' and 'farming systems research and development', provide little more than an innocuous umbrella under which people of different disciplines and background can work together, interactively, on the task of improving livestock and crop output and income.
Our discussions this week will focus upon rather specific and technical parts of our approach to livestock systems The techniques we use in this work are obviously very important, but it is the approach we take to livestock and pastoral farming systems that I would like to see characterizing your views of ILCA's work, rather than the use of any of the specific research techniques we shall discuss this week.
The ILCA approach consists of two main items: a characteristic philosophy and a collection of techniques, I shall spend a few minutes talking about the philosophy that underlies our work; my colleagues will discuss mainly the techniques used and the results obtained. In brief, the philosophy is that there are such things as systems, and that things that operate as systems have to be regarded rather differently from things that operate in some other manner.
What is a system, are systems important and in what way do systems differ from non-systems? Let me try to explain by referring to ILCA as an institution. At ILCA we regard our organisation, ILCA, as a system. We, the staff, interact and work together for a common purpose and thereby have the key characteristics of a system. Thinking unkindly, other organisations may be regarded as a collection of individuals, and a collection does not constitute a system. With a collection you can take out a component and it makes little difference, similarly you can add one without making much difference. But if you interfere with one component of a system such as ILCA, remove it, alter it, or add another, you are likely to have important effects on the whole system, because it is a characteristic of a system that the components interact, It follows that grey hair and system management go well together !
There are two other important and inter-linked characters of a system I want to emphasise. The first is that a system has a boundary and that boundary must be clearly established in our description of the system. The second is that a system is purely a concept, By this I mean that an adequate description of a specific system will depend on the purpose of the description and the boundaries that are perceived for the system; a naturalist defines the same aggregation of interacting activities in quite a different way to an economist, a politician or an agriculturalist Obviously you need a clear notion of the specific purpose of your description of a system in order to describe either the system or its boundaries.
You will also appreciate that a description of a system is highly biased by the training of the person you have asked to describe it. If you doubt me set loose an agriculturalist, an economist and a sociologist to describe a specific system such as ILCA Each will give an appropriate description that suits their particular purpose, but none is likely to be appropriate for the purpose of understanding the ILCA system as a whole. The first task is to think carefully about what is to be described, and even more importantly about what does not need to be described if you try to describe everything you will never finish and will never produce a description that is of use to a specific audience
To decide if a component observed in a system is important, one has to relate it to the purpose for which the system exists. In the ILCA system, for example, one finds components such as sociologists, chemists and librarians. It is only when we know the purpose of ILCA that we can start to decide if describing these specific components is important. If it is argued that the purpose of ILCA is to increase milk production, then we can decide on the basis of the likely effect of our sociologists etc. on milk production whether they are an important part of our system, or whether they can be disregarded in our description. If the purpose of ILCA is perceived as something different than simply increasing milk production, our description and judgement of the vital component parts must be different.
A key objective we all share is to improve livestock production; we also probably share a belief that using a systems approach is a sensible way to tackle this objective. This accepted, the key questions we then need to ask are what systems are to be improved and what constitutes an improvement.
In many livestock systems of concern to us there is very little information available and it is tempting to start extensive, time consuming baseline surveys to try to build the data necessary for initial modelling and intervention, If this exercise is followed we frequently find, several years later when modelling and analysis of the data collected is started, that some key data required have not been collected. My very important point here is that it is essential that data collection, data analysis, and modelling go hand in hand. We must start with a very simple model at the outset to help determine what information is actually required and not wait until we have accumulated large lumps of information that are likely to be both too much and too little.
Having described in adequate terms the systems to be improved one then has to decide what constitutes an improvement. Improving milk production per cow may be technically possible in several different ways, but each way suggested may be financially or organizationally impractical. Clearly we need to consider not only increased outputs but also increased inputs; we need to consider if the inputs proposed are likely to be available, can they actually be safely used by our target farmers, and what do they cost? One also quickly realizes that livestock systems do not produce just one thing, and that they are not using just one resource. "Improving" a system that produces several products and uses many resources is a matter beset with many complications that are frequently overlooked in an otherwise simple technical approach to livestock improvement.
We must also ask who benefits from any improvement that may be made, If output is increased does the farmer benefit, or do prices decrease? In the latter case does the nation as a whole reap the rewards, or just one small part of the nation?
If the objective is to increase overall food production, obviously a most important task in Africa, we still need to ask if fulfilling this objective will feed hungry people. Hungry people are poor people; if they weren't poor they wouldn't be hungry; if they had the money to buy food, that food would certainly be produced and available. Clearly, a food production objective alone is inadequate and we need to consider equity distribution, cash flows, labour requirements, multiplier effects and all the rest of the bits and pieces that are critical in deciding on the value of our improvement.
Farming systems are rarely static, they are usually being continually changed in many ways by a multiplicity of external forces. We as systems researchers must make judgements as to where our target systems may be headed. Research appropriate to improving the productivity of large farms is frequently inappropriate to improving farming systems on small farms, and we need to be reasonably sure that the target system for our research effort has an appreciable life span.
ILCA's task is twofold : to apply existing knowledge to improve livestock production, and to undertake research on major gaps in that knowledge. The technical base for improving livestock production in Africa has proved to be less extensive than was perceived when ILCA was established just a few years ago. In consequence, we have recently been placing greater emphasis in our work at ILCA on what we loosely call component research. Yet I urge great caution on your part in regard to the use of the term "component research". Our primary interest and concern is not in understanding how a legume grows, or how a cow copes with a protein deficiency, but how key changes in vital parts of livestock production systems, such as improving pasture quality and dry-season nutrition, might improve the productivity and efficiency of the farming system as a whole. Our main task in many of the systems we work with lies in assessing the relationships among their key components. Strictly speaking this is not component research, rather it is process research. Others, however, may regard it as a development activity. May I suggest that the distinction between research and development is largely in the eye of the beholder. In line with our purposive philosophy, we regard the difference between the two as the reason one gives as to why the particular work is being done.
May I make a final comment about organising the staff of any institution to carry out a systems research approach. Many ways of doing this have been tried and the virtues of multidisciplinarity are usually loudly emphasised. I want to stress, however? there is little merit in having many disciplines represented if they are not required, and that the pursuit of multidisciplinarity as such can be overdone. In our experience what is required is staff with knowledge that is highly relevant to the systems you are studying, and this staff needs to encompass a very wide range of training and experience. First class generalists backed up by specific selected specialists is the formula we seek. I am sure you will note this amongst our staff at ILCA this week.
During the course of the next few days we expect to learn a great deal from you. We appreciate greatly the time you have given to share your experience and your wisdom, We also value highly your willingness to work with us in helping develop our overall programme. IDRC have made a sizeable cash contribution to the organisation, of this workshop. Their assistance has made it possible to invite more participants than would otherwise have been the case and it has been a spur to us to hold this important meeting. We are most grateful to them. We hope that you will benefit from the discussions of the next few days, and that you greatly enjoy your stay with us.
