260 – Farmer Research Networks

[00:00:07] Lucie: Hi, and welcome to the IDEMS podcast. My name’s Lucie Hazelgrove Planel, I’m a social impact scientist and anthropologist, and I’m here today with David Stern, one of the founding directors of IDEMS.

Hi, David.

[00:00:18] David: Hi, Lucie. Keen to have another discussion about presumably our CRFS work today.

[00:00:24] Lucie: Yeah, exactly. So there’s been discussions about what we call FRNs, which are Farmer Research Networks, which is one of the key things that, well, it’s an amazing organisation strategy, I would say, with the community of practise that we work with in West Africa. It’s not only used there, it’s used throughout the Global Collaboration for Resilient Food Systems’ work.

And I think they’ve been rethinking about principles. I mean, should we think about the new principles or should we talk about just what farmer research networks are in general?

[00:00:55] David: It’s a good question and I think we should talk about the history a little bit because it will put into perspective what’s happened and what this could be.

[00:01:07] Lucie: Sounds good.

[00:01:08] David: I feel the idea for me really solidified with a presentation that a lady called Rebecca Nelson and a Cornell professor and a statistician called Rick Coe presented over a decade ago now. They explained in a beautiful way how FRNs could lead to a new type of research. They explained how, in agricultural research, traditionally you had experiments in a very controlled environment, on a station, and you had a relatively small number of experiments. And from that you were able to draw out the traditional agricultural knowledge of how plants grow and how things work at the biological level, if you want.

[00:02:02] Lucie: This is sort of an ideal environmental situation perhaps too. 

[00:02:06] David: I think the better word, it’s not about being ideal, it’s about being controlled. You control the environment so that the thing you are observing, the experiments you put in place, the only things that change is the thing you are experimenting on. And so you are actually gaining deep causal relationships, and you are understanding them in great detail.

This is the scientific method at play, and it was very successful. Traditional agricultural research has transformed the nature of crops that we grow, it’s had huge impacts on our society, many of which are positive. But they then explained how there was a limit to what actually, to the value that that brought. And for 40, 50 years, there have been a whole range of participatory approaches which have been evolving and emerging, which look at actually saying, well, what happens when you put this into context? You remain with a small number, but you get very high levels of participation and co-design.

And so it is moving in your direction towards an anthropological study. It’s not anthropology, but it is very much a social study where you are actually understanding the context and the complexity and you’re understanding how things work in context. There’s many different approaches, but the most well known is called PAR, Participatory Action Research, and there’s really established methods on this.

And this has transformed the way that we understand agricultural context, and that they put that on one axis, which was, in their case, the Y axis, where it was having a real depth of understanding. So you still have small numbers, but you have a depth of understanding of the context and of the agriculture within the context.

And then the other innovation, which is slightly more recent, but still 20 to 30 years old, one name for it is “precision agriculture”. It’s where you use lots of data, large amounts of information, to understand micro context even, you know, sensors which determine variable levels of fertilisation because you have micro information about the soil and how the soil changes in different places.

[00:04:36] Lucie: Yeah, precision research. I always associated it with mechanised agriculture, but I guess that’s why, because they’re able to get so much more, it’s about the data then.

[00:04:45] David: It’s all about the data, and it’s about this immense amount of data, and yes, broadly it is when things are mechanised, that this has been extremely successful and this has been used in high resource environments in particular for high value crops. And it remains a really exciting area where data is being used to make decisions in ways that are really interesting.

And the graph, that was then along the x-axis, the amount of data, the amount of information that you get, was then what they were saying, well, okay, you started with small amounts of information in your traditional context, you only had small amounts of information in some sense for small amounts of people for your PAR context.

And then, when you actually get to the scale, you actually have lots of information. And FRN, a Farmer Research Network, was in theory a way of getting high levels of participation, that deep contextual understanding at a large scale. And that was really the original presentation of what FRN was and why it was important. And I understood that very deeply and was really taken in, and have been really enamoured with this approach since.

That was how it was originally presented. However, when it came down to actually the discussions on this and the implementation and the development of the principles around it, there was a nervousness from certain people of including this large end, this scaled approach as part of the FRN definition. 

[00:06:25] Lucie: When was the definition? Did the CRFS start talking about FRNs before trying to develop principles?

[00:06:31] David: Yes.

[00:06:32] Lucie: Okay. So this is perhaps why there was nervousness. And how was it used before having principles?

[00:06:38] David: Well it was an idea and the idea had really emerged from these discussions of large end and how we can get large end participatory trials, that’s what it emerged from. And it was really well thought out. And there were good reasons why people were nervous because of actually the implications of working at large scale and what this means, particularly on the limited budgets within a CRFS project. This isn’t like Gates who have huge amounts of money to put it into this, these are sort of small projects doing this.

And so actually enforcing that you are doing things at a large scale, it wasn’t clear that for the programme that was what would’ve been appropriate. So from a programme perspective, there was pushback about the scale and the fact that you’re working at large scale being a component or a principle within FRNs.

[00:07:33] Lucie: Or even just socially, it’s hard to organise people into large groups in order to do similar work. 

[00:07:41] David: And in the West African context, the farmer federations had already done all that hard work. But in the other contexts many projects were starting from scratch, and they were building groups up and so on. And the large end was not really realistic. So, very practically, working at large scale was not realistic.

And so as a programme, the concept of FRN never included this scale component as part of the definition of what it meant to be a farmer research network. And that was very positive because it led to a very inclusive definition. But it also led, I believe, to quite a lot of confusion as people find it difficult to distinguish between the existing methods, which are used very successfully, like PAR, Participatory Action Research, and Farmer Research Networks.

I even heard very recently someone who was doing FRNs, oh, we added PAR to FRN and we extended. And that to me is exactly the opposite of what’s actually, how it was conceived, PAR was part of FRN. FRN is extending PAR by having these networks, and the scale element is really what FRN was conceived to be.

And so the fact now that there’s so much confusion on the terminologies, and it’s never gone beyond the programme, I believe is because it is not well defined enough as a concept. And I believe the only thing missing is the idea of scaling in the principles of FRN. And the implications of that we could maybe go into in another episode.

Actually, a number of us who believe scaling should have been since the beginning, we talk about embedded scaling. And that’s what we think is the missing part of the principles, FRN should have a component of embedded scaling. Now, it is not something where others will agree with me on this, others within CRFS have different opinions.

And one of the things I love about CRFS is that it’s okay, it is a community where different opinions are welcomed and it’s okay to have these differences of opinions, and it’s okay to have that debate around it. But I do think that my hypothesis is, if we could get that as a principle, FRNs could be more easily identified and you could say, well, that’s not an FRN more easily. So actually being slightly less inclusive with the definition. That is PAR, that’s good, it’s PAR, but it’s not an FRN.

[00:10:32] Lucie: Exactly. There could be a sort of clear scale or journey between starting off with PAR and then what the CRS does is to support projects in scaling, in going to this FRN concept.

[00:10:44] David: For example. But the thing that I think would be powerful about that, it’s that it is not for CRFS to support projects, it’s that others could pick up the concept. I believe one of the reasons others have not taken up the concept of FRN is not because others aren’t interested. I think many people are interested in it, but it’s not well defined enough that it is not clear what it would mean to take it up.

And if somebody were to take it up, then they are likely to interpret it in a way which is very different, and therefore it wouldn’t actually have cause and effect. If you do an FRN, if you do this in an FRN approach, then these things will happen.

[00:11:24] Lucie: Yeah.

[00:11:25] David: Now the work on principles is really powerful and I’m really happy with the extra work which is happening on the principles. I love the work that’s gone into the counter principles, which really does help to sort of articulate this is how it does, these are some of the differentiating factors.

[00:11:46] Lucie: Exactly. So if we just give an example of that, is it diverse farmers are involved in research?

[00:11:52] David: Yes. And the opposite of that, or not the opposite, a counter principle to that could be that you work with the farmers who are engaged and interested in becoming part. 

[00:12:03] Lucie: Which is a very sensible thing to do. 

[00:12:06] David: This is the PAR approach. So the PAR approach is often to find and identify the research farmers in the community and to work with them and to encourage them to then interact in their community. This is a great approach and this is a distinguishing feature, done well, to understand how the FRN approach can or could be different from a PAR approach. And therefore people could be making decisions to go in the PAR direction.

[00:12:33] Lucie: Yeah.

[00:12:33] David: And it’s not that one is better than the other, it’s that there is a clear distinction, which can be used to distinguish between them.

I do think that all the principles, as they’re articulated, are useful and powerful. So I’m not at all in disagreement with the principles that exist. However, as we’ll discuss in another episode, I would love it if “embedded scaling” was another principle, because that would have huge implications for the research, in which I think, at the moment, the researchers are pushed out of their comfort zone along one axis, but not the other.

If we think about it in the original framing of the participatory element and the scale element, with the current principles, we’re pushing researchers out of their comfort zone towards more participatory inclusive elements of research, but we are not pushing them on the scale and actually thinking what it means to be doing things at scale. 

[00:13:32] Lucie: But this also has repercussions on the data that they actually get too, ’cause when they’re doing agricultural experiments with just a small handful of farmers who are in different ecological contexts and who work their farms differently, then it’s hard to actually identify the learnings. 

[00:13:50] David: This is exactly the key, not the problem, but the key point, where if you don’t include embedded scaling, it’s not clear to me that you can get this transformative research. And so I’m not saying that there isn’t something valuable, I’m just saying that the really valuable piece is being missed, and that actually our efforts are not being put enough into what would make FRN truly distinct and really valuable.

A very simple example of this, which is within our own remit: you know, we work with the agroecology hub in Kenya, Beth Medvecky who we’ve discussed before is a key person on that. And she’s got this amazing approach to FRNs. She’s also someone who believes in embedded scaling, but what we found is that we haven’t been successful at embedding scaling.

So this is actually really hard. So we believe in it, we want to do it, but the point is, it’s been so hard. We’ve not embedded, despite many discussions, we’ve failed to embed scaling within this. And this is the key point, I understand the reasons why scaling has not been included, because it’s hard. And actually, it’s not to say that doing things without the scaling component isn’t valuable. The FRNs that Beth has been involved in coming out of Manor House as part of the agroecology hub, I would argue all of them are valuable, but they would not satisfy the definition of FRN yet if we included embedded scaling.

And so we would be pushed to actually work on the hardest part of it, which is to understand how we actually embed this within the scaling of the interventions that they’re finding. They’ve had a workshop just the last few weeks, which was on mole rats and the trapping of mole rats and so on. But it’s never gone beyond the community of farmers who did that work. It’s been very effective for them in their communities, their communities have benefited, but the impact of it has not gone beyond. The same with the fermented chicken feed, the same with the tephrosia.

[00:16:08] Lucie: But this is partly, again, because of the structures, the social structures that exist at the moment in Kenya, compared to our work in West Africa, where there aren’t the farmer organisations, so that there isn’t that existing network. 

[00:16:22] David: There are NGOs, there are CBOs, so there are mechanisms which could be used for scaling. 

[00:16:29] Lucie: There’s churches too.

[00:16:30] David: There’s churches, there’s radio, there’s all sorts of mechanisms. But there isn’t the mechanism that if things start scaling to continue the research, to have the research embedded. So that’s what I want to, again, talk about in a separate episode, to dig into what that would look like, what the implications would be.

But I do think that there’s a real opportunity to be a bit braver as CRFS with FRNs and to push and to challenge ourselves and our projects to think harder about the really hard problem, not just the fact that FRNs are valuable and they’re useful, and that’s enough. Because if they’re not well defined as actually being more than just that, then we haven’t developed something new.

And what I do believe is that FRNs in McKnight have evolved slowly because the finances are small and that’s a good thing. But if we’re taking a concept which could then get taken up by others who have larger amounts of funding, we want something which is more ambitious and actually is building from the learnings we’ve had over the last decade.

So it is absolutely okay, in my opinion, to be brave in how we frame this,, in how we evolve in our language, and to actually move into this more ambitious discussion around what a FRN is and could become. 

[00:18:04] Lucie: And of course, listeners to the podcast won’t be at all surprised that you want to be more ambitious. You want to encourage people to go for the things which are their dreams.

[00:18:13] David: It’s my role in life, to push people out of their comfort zone. But more seriously, I actually think within CRFS, there’s a readiness for this. And the reason that I feel that more strongly now actually comes out of the group who presented on behalf of the Andes, because they presented in a very nice way about how they started off essentially with this PAR model, and how they’re now using the networks to extend into them. And so they are thinking about scaling.

So I believe everyone is thinking about scaling, especially when you think of it as being not just scaling out or scaling broad, depending on how people like to call it, but also scaling deep. And that’s something that is useful, where you are actually enabling people to be more, the scaling to go into the society, into the community in terms of how they’re thinking about these things.

So, I think, in almost all successful cases there have been, and there are, these elements of scaling, but they are different types of scaling. And so the key thing is that it might be that a decade ago we couldn’t have included “embedded scaling” because the embedded scaling would’ve only been scaling out, we didn’t have the language to do it properly.

Whereas I believe now we have a new language around scaling, which is coming into CRFS, but more broadly into the way scaling is discussed. And I think with those broader definitions of scaling, I think now “embedded scaling” is easier for people to accept that it should be a component of FRN. I think it would help us to have something concrete that others could pick up and use and run with and explore and play with. 

[00:19:59] Lucie: And it’s more inclusive because it means that everyone can scale in the way that’s most appropriate to them, whether that’s going really deep with just a small group or whether it’s expanding out, as you say.

[00:20:10] David: But the key point is that it doesn’t matter whether you are going deep, out, or up. In each of these cases, you are increasing the amount of data. This is why it comes back to that original graph that Rick and Rebecca presented, which had this element of large N where large N is not necessarily large numbers of farmers, but large amounts of data.

This is maybe the evolution of our thinking, that this is really what’s important, that we need to be thinking about this with respect to large amounts of data. And that’s where we should stop because I’m starting to get stuck into what I would like to discuss in an episode dedicated to embedded scaling and what the implications of it could be.

[00:20:51] Lucie: Sounds good. Well, thank you very much, David. It’s been interesting.

[00:20:54] David: No, thank you.