Understanding the emergence of a hybrid knowledge production discourse: The case of the Generation Challenge Programme (GCP) drought tolerant rice research in India

Initial ideas of this manuscript were presented at the 14th Global Conference of the International Association for the Study of the Commons, Mt.Fuji, Japan. 3–7 June, 2013 by the first author.

Highlights

•	We study knowledge production process of GCP rice research partnership in India.
•	GCP partnership has been built in a bottom-up and institutionally pluralistic way.
•	GCP research partnership is non-hierarchical and decentralised in its functioning.
•	GCP partnership involves farmers at a peripheral level of research through PVS.
•	Research approach of GCP partnership is interdisciplinary and not transdisciplinary.

Abstract

The Generation Challenge Programme (GCP) is an international platform for agrarian knowledge production for a complex scientific problem, namely, Drought. The GCP ushered in a new form of knowledge production that reconciles both the upstream laboratorial research and its downstream delivery at the farmer's field. This paper aims at understanding the knowledge production process of the GCP. More precisely, it explores the following three research questions: how three processes such as research partnership, transdisciplinary and inclusion of end-user in research manifested within the knowledge production process of the GCP drought-tolerant rice research; what type of knowledge production emerged within the GCP drought-tolerant rice research; and what implications can be discerned for agricultural research and development. Through a qualitative case study method, this paper explores empirically the case of GCP drought-tolerant rice research in the Indian context to elucidate the knowledge production process of GCP to sketch a theoretical as well as a practical position on the agrarian knowledge production discourse. This paper argues that a hybrid knowledge production discourse has emerged within the GCP rice research that has elements from different knowledge production paradigms. Further, this paper also illustrates the implications of such hybrid knowledge production discourse for agricultural research and development.

Keywords:

• Knowledge production
• Hybrid discourse
• Generation Challenge Programme
• Rice research
• Agriculture
• India

1 Introduction

The Generation Challenge Programme (GCP)¹ was an international agrarian knowledge production programme created in 2003 by the CGIAR² as a global crop research consortium (Okono et al., 2013; CIMMYT et al., 2003). Wound down in 2014, as per its 10-year time-framing, the GCP had as its principle objective, the production of knowledge pertaining to drought using plant genetic diversity through advanced genomic science and comparative biology so as to support the efforts of plant-breeders in developing countries to develop drought-tolerant varieties (Wanchana et al., 2008; Basu et al., 2011; Basu, 2016; GCP, 2007; Vroom, 2010). Drought is indeed a major problem for global agriculture (Nelson et al., 2007). In case of rice, some 80 million farmers are working on 60 Mha that are drought prone (Fischer et al., 2012), with drought alone accounting for an 18-Mt annual reduction in production (Bernier et al., 2008). In India, drought is a major concern for the eastern part of the country, where almost 10 Mha of the rice-growing area is drought prone, costing some 36% of the average value of total rice production (Bernier et al., 2008; Kumar et al., 2008; Pandey et al., 2007).

GCP can be seen as a composite knowledge production programme in which divergent approaches of knowledge production have emerged (Basu, 2016; Basu et al., 2017a, 2017b, 2017c; Basu et al., 2011). These divergent approaches include mainstream knowledge production theories such as instrumental and non-instrumental paradigms (see for details Basu et al., 2017b) as well as alternative knowledge production paradigms such as commons based production systems (Basu et al., 2017a) and Commons Based Peer Production (CBPP) mode (Basu et al., 2017c). In this paper, the aim is to understand the knowledge production process of the GCP.

Nomenclature Acronyms
AAU	=	Anand Agricultural University
AICCIP	=	All India Coordinated Crop Improvement Programme
ARBN	=	Asian Rice Biotechnology Network
BAU	=	Birsa Agricultural University
DBT	=	Department of Biotechnology
BF	=	Barwale Foundation
BMGF	=	Bill and Melinda Gates Foundation
DST	=	Department of Science & Technology
CBPP	=	Commons Based Peer Production
CGIAR	=	Consultative Group of International Agricultural Research
CIMMYT	=	The International Maize and Wheat Improvement Center
CRRI	=	Central Rice Research Institute
CRURRS	=	Central Rainfed Upland Rice Research Station
CSIR	=	Council of Scientific and Industrial Research
CURE	=	Consortium for Unfavourable Rice Ecosystem
DBN	=	Drought Breeding Network
EIFPBP	=	Eastern India Farmers’ Participatory Breeding Project
EIRRP	=	Eastern India Rainfed Rice Project
FGD	=	Focussed Group Discussion
GCP	=	Generation Challenge Programme
ICAR	=	Indian Council of Agricultural Research
IDRC	=	International Development Research Centre
IFAD	=	International Fund for Agricultural Development
IRBN	=	International Rice Biotechnology Network
IRRI	=	International Rice Research Institute
JNKVV	=	Jawaharlal Nehru Krishi Vishwavidyalaya
MPUAT	=	Maharana Pratap University of Agriculture and Technology
NAIP	=	National Agricultural Innovation Project
NARS	=	National Agricultural Research System
NATP	=	National Agricultural Technology Project
NDUAT	=	Narendra Dev University of Agriculture and Technology
NGO	=	Non-Government Organisation
NWO-WOTRO	=	Science for Global Development Programmes, Netherlands Organisation for Scientific Research
OUAT	=	Orissa University of Agriculture & Technology
RF	=	Rockefeller Foundation
SAU	=	State Agricultural University
TNAU	=	Tamil Nadu Agricultural University
UAS	=	University of Agricultural Sciences
URRC	=	Upland Rice Research Consortium
URSBN	=	Upland Rice Shuttle Breeding Network
VPKAS	=	Vivekananda Parvatiya Krishi Anusandhan Sansthan

For this, we have relied upon three processes (as organising principles of knowledge production) namely research partnership, transdisciplinary and inclusion of end-user in research to elaborate how these three processes manifested within the knowledge production process of the GCP’s drought-tolerant rice research projects in an Indian context that included the International Rice Research Institute (IRRI), along with several Indian agricultural institutions.³ The case of drought-tolerant rice research in the Indian context was chosen because of the following reasons. First, due to the importance of the rice crop for food security in India (Cornish et al., 2015; Gathorne-Hardy et al., 2016), and the large impact that drought has in curtailing rice productivity in India (Bernier et al., 2008). Second, accessibility of actors and information on GCP rice research projects were available. Third, this research was conceived within the NWO-WOTRO integrated programme between India and Netherlands.⁴

This research is important for the following three reasons. First, hitherto, agrarian knowledge production is studied taking a single perspective in account such as from the research partnership perspective (Hall et al., 2001), from the perspective of transdisciplinarity (Raina et al., 2006; Raina, 2003) and from the perspective of farmers participation (Witcombe et al., 2005); in contrast to these studies, this research takes-up all these three perspectives simultaneously to provide a holistic and comprehensive analysis of agrarian knowledge production; second, there are few studies that reveal the working of science in transdisciplinary settings in India within a North-South research collaboration; third, given the composite nature of GCP knowledge production (Basu, 2016) with a clear non-instrumental alignment of the Indian agricultural research and extension establishment (Raina et al., 2006; Sulaiman and Hall, 2002) it would be interesting to find out the particular knowledge production discourse emerging out of such interface; and finally, the particular type of knowledge production discourse that emerged within the GCP drought-tolerant rice research may have wider implications for agrarian knowledge-producing institutions generally, such as those linked to the CGIAR, NARS and the Indian Council of Agricultural Research (ICAR).

The paper is organised thus: in the following section, we provide the theoretical framework and the context; next, we describe the research methodology employed for this research, and then the empirical findings are presented; this is followed by a discussion, after which the paper is concluded.

2 Theoretical framework

Several theoretical approaches are relevant to an understanding of the process of knowledge production (Basu, 2016). Fundamental in this respect is the binary between the dichotomous understanding of a state-funded, discipline-based, and academia-governed and validated truth-oriented persuasion of science, on the one hand, and transdisciplinary, network-governed, externally validated, problem-solving and impact-driven science, on the other (Hessels and van Lente, 2008; Bouma and Donald, 2010; Knuuttila, 2013; Basu et al., 2017b). Even within the agrarian knowledge production discourse, this dichotomous binary has been expressed through two main paradigms: the linear paradigm and the systems paradigm (Pant and Hambly-Odame, 2009). The linear paradigm, often also known as the technology transfer approach, considers a linear relationship between science and society in which standardised technology packages are developed in the research institutes that are then transferred through extension agencies to the farmer’s field, where, it is expected, it will bring change (Pant and Hambly-Odame, 2009; Rogers, 2003). In contrast to the linear paradigm, the systems paradigm looks at the social processes of knowledge production and technology development, focusing on the transdisciplinary collaboration between formal and informal sectors, engaging with indigenous and local knowledge systems and involving farmers (Basu and Leeuwis, 2012; Röling, 1992; Sulaiman and Hall, 2002).

However, as we have argued elsewhere (Basu et al., 2017b) that this dichotomous binary is apparently not relevant in analysing and/or understanding the knowledge production process of GCP drought tolerant rice research. Especially, taking a closer look at the composite nature of GCP knowledge production and the context (India) (Basu, 2016); one may argue that GCP’s knowledge production embodies attributes that can be found in both the paradigms. For example, GCP’s focus on societal problem solving (drought problem) through building research network (north-south collaboration) and the inclusion of societal representation (resource poor farmers particularly) in actual research clearly relates to a systems paradigm, while the non-profitability, centralised science model in Indian NARS and disciplinary focus are more closely aligned with the linear paradigm. Therefore, in this case, it would be interesting to find out the particular knowledge production discourse emerging out of the interaction between these two dichotomous modes of knowledge production.

Therefore, in order to analyse the knowledge production of GCP drought tolerant rice research in Indian context, this paper relies upon exploring three processes: forming research partnership, directing research towards transdisciplinarity and including end-users into research process. Research partnership can be defined as collaborative effort between several organisations in which each organisation contributes to the planning, resources and activities needed to complete a common objective (Basu et al., 2017b). Generally, research partnerships arise in a particular context and understanding that context is very important while studying the partnership (Hall et al., 2001). Within research partnership, a hybrid overlapping institutional sphere emerges due to the increased interaction and networked structure of collaboration between universities, research institutes, private sector R&D and other organisations (Etzkowitz, 2003; Gibbons et al., 1994). Directing research towards transdisciplinarity refers to the mobilisation of a range of theoretical perspectives and practical methodologies that cannot be accommodated within the framework of existing disciplines (Basu et al., 2017b; Turnhout et al., 2013). Transdisciplinary research also involves serious management challenges in respect to objectives and outcomes (Kragt et al., 2016). Finally, including end-users into research process refers to the representation of societal stakeholders in the research so as to ensure the social problem is understood and most likely to be mitigated (Basu et al., 2017b; Landini, 2016). Gibbons (1999, 2000) famously phrased it as ‘society speaking back to science’. Thus, knowledge is not validated by the scientific community alone, but also by interested parties (farmers in this case), in a process commonly termed ‘external peer community validation’ (Funtowicz and Ravetz, 1993).

We do not claim here that these three processes are coherently discerned from any theory-informed concepts. However, consulting the wider literature on the transformations within knowledge production systems (Basu, 2016), these three processes can be seen as common denominators or organising principles with which to engage with knowledge production. Thus, exploration of these three processes can help us in elucidating the GCP knowledge production process. As it will be elaborated on empirical account that these three processes in configuration generate a hybrid knowledge production discourse that is beyond the dichotomous binaries we have elaborated above.

3 Methodology

This paper aims at understanding knowledge production process of GCP drought tolerant rice research in Indian context by exploring three interrelated processes namely forming research partnership, directing research towards transdisciplinarity and including end-users into research process. The focus here is on studying ‘processes’, hence, a qualitative case-study design is employed. A case-study method involves a detailed examination of an event (or series of related events) that, according to the authors, in some way reveals certain theoretical principles (Mitchell, 1983). Therefore, this method was purposefully chosen for this research because of its ability to provide detailed insights on the complex scientific, social and organisational process of agrarian knowledge production. Moreover, Yin (2009) sees the case study as an ‘empirical enquiry that examines a contemporary phenomenon within its real life context; when the boundaries between phenomenon and context are not clearly evident; and in which multiple sources of evidence are used’. In this study, the phenomenon under investigation is the knowledge production process of the GCP drought-tolerant rice research, and the organising elements (research partnership, transdisciplinary and end-user inclusion and validation of research) of the knowledge production is taken as the explanation of the observed phenomenon; the context cannot be separated from the organising elements, while knowledge production is a contemporary issue that makes good use of case study method as explained by Yin. The priority of this research was to find explanatory patterns through interpretation rather than to measure numerically. Therefore, quantitative methods were not applied.

The research was initiated with an intensive study of GCP’s drought-tolerant rice research-project documents, including research papers, annual reports, project briefs and working papers. From this, key actors (institutional and individual) were identified. Then, a 14-week (April–July 2012) fieldwork in India and the Philippines was organised by the first author. In the Philippines, IRRI was visited and several rounds of interviews with the project leader and other scientists conducted. In India, the author visited and stayed at all the associated institutes (see below), attending meetings (including with farmers), discussing issues at length with scientists (as well as with directors/vice-chancellors), interacting with (masters and doctoral) students, addressing scientists in an interactive session and compiling a large number of participatory observations. A total of 45 in-depth interviews with the scientists, directors and others were conducted. Nevertheless, when it became apparent just how integral the farmers were to this knowledge production process, it was deemed necessary to engage in further fieldwork aimed at talking to them to gain a more holistic understanding of the operation of the system. Thus, a second round of fieldwork focusing exclusively on the farmers was conducted, in the following October and December (2012).

Every institution in this rice research network has its own local farmers’ group. Scientists usually conduct their PVS experiments with these groups, and sometimes the groups also help in seed multiplication and field-level demonstrations. It was not practically possible to study farming communities linked to all the institutes, so we decided to concentrate on the several communities associated with just one institute, the CRURRS, at Hazaribagh, in the state of Jharkhand. Farmers associated with CRURRS were chosen for information availability and accessibility. There, the first author visited three different farming communities situated in three different districts. In each farming community, informal talks were held with a total of around 100 farmers (both men and women). Then, a total of 45 farmers who participated in the CRURRS research activities were interviewed in-depth. Several focus group discussions (FGDs) were held in each farming community (with no scientists present). Together, these informal talks, in-depth interviews and FGDs came to constitute the primary data source.

Then, insights from the knowledge production literature elaborated in Section 2 were applied in order to analyse the data. Three interrelated processes of knowledge production such as forming a research partnership, directing research towards transdisciplinarity, and including end-users in the research process were identified and employed as descriptive models to analyse the data. In case of research partnership, first, the context of the partnership is explored by knowing why and how it was developed, if the partnership is a result of institutional collaboration or individual collaboration, who are the key stakeholders, why scientists participate in this partnership and if they are intrinsically or extrinsically motivated to participate; and second, the organisation of the partnership by exploring the internal dynamics in terms of centralisation, institutional hierarchy and scientific bureaucracy. The issue of transdisciplinarity explored through the research approach of the partnership by exploring the interactions and integrations among the disciplines. And finally, the issue of end-user participation is explored by reflecting on how and why end-users are involved in the research and which level they are involved.

Data analysis was also simultaneously performed in the field during data collection as new perspectives gained from the field also indicated directions for further observations and interviews. Instead of direct quotation, given the high volume of data and large overlapping within it, we have coded the data into coherent narratives that are presented in the next section. Each narrative usually presented within a paragraph that also contain the source of the data (usually the interview) in the footnote. Therefore, the footnote reference is very important for the relevance and legitimacy of the research. In the following section, we organise our findings along with three concepts identified in section 2: forming a research partnership, directing research towards transdisciplinarity, and including end-users.

4 Forming the GCP rice research partnership

The activities of the GCP drought-tolerant rice research partnership were organised among 14 different institutions, 13 in India and one in the Philippines (Fig. 1).⁵

Fig. 1 Location of the 13 GCP partner institutions in India.

In addition to IRRI, the other institutions comprised ICAR research institutes,⁶ state agricultural universities, NGOs and transfer-of-technology centres. These were all situated in or close to drought-prone areas, and in different agro-ecological zones. The GCP awarded four projects between 2005 and 2014 (summarised in Table 1) and produced multiple outputs, both in the form of drought-tolerant varieties – released by the Indian government for commercial cultivation (e.g. Sahbhagi Dhan, Shusk Samrat, Anna R(4) and ARB-6) – and publications (Kumar et al., 2008; Kumar et al., 2012; Mandal et al., 2010; Verulkar et al., 2010; CGIAR Generation Challenge Programme, 2009; CGIAR Generation Challenge Programme, 2011). This section is divided into following three parts: the formation and evolution of the partnership; internal organisational dynamics of the partnership and a reflective section.

Table 1 GCP rice project on drought in India.

Axapta No; Project	Lead Institute; Principal Investigator; Country	Trait	Partners	Start–End (NCE)
G4005.69.01; Developing and disseminating resilient and productive rice varieties for drought-prone environments in India	IRRI; Arvind Kumar; India	Drought tolerance	IGKV, CRRI, CRURRS, NDUAT, UAS, TNAU, BAU, BF	Mar 2005–Feb 2008 (Feb 2009)
G3007.05; Detecting and fine-mapping QTLs with major effects on rice yield under drought stress for deployment (marker-aided breeding)	IRRI; Arvind Kumar; Global	Drought tolerance	BAU, BF, CRRI, CRURRS, IGKV, JNKVV, NDUAT, TNAU, UAS–Bangalore, UoAl, YAAS	Aug 2007–July 2009 (Dec 2011)
G4008.05; Connecting performance under drought with genotypes through phenotype associations	IRRI; Arvind Kumar; Global		IRRI, CIRAD, ARC/WARDA, BIOTECH, CRRI, TNAU, BAU	Jan 2008 –Dec 2010 (June 2012)
G4011.04; Dissemination and community of practice for newly developed drought-tolerant QTLs pyramided breeding lines	IRRI; Arvind Kumar	Drought tolerance	CRRI, CRURRS, DRR, NDUAT	July 2011–June 2014

ARC/WARDA: Africa Rice Center/West Africa Rice Development Association.

BAU: Birsa Agricultural University.

BF: Barwale Foundation.

BIOTECH: National Center for Genetic Engineering and Biotechnology.

CIRAD: Centre de Coopération Internationale en Recherche Agronomique pour le Développement.

CRRI: Central Rice Research Institute.

CRURRS: Central Rainfed Upland Rice Research Station.

CSU: Charles Stuart University.

DRR: Directorate of Rice Research.

IGKV: Indira Gandhi Krishi Vishwavidyalaya.

IRRI: International Rice Research Institute.

JNKVV: Jawaharlal Nehru Krishi Vishwavidyalaya.

NagU: Nagoya University.

NDUAT: Narendra Dev University of Agriculture and Technology.

SUA: Sokoine University of Agriculture.

TNAU: Tamil Nadu Agricultural University.

UAS–Bangalore: University of Agricultural Sciences.

UoAb: University of Aberdeen.

UoAl: University of Alberta.

UoMi: University of Missouri.

YAAS: Yunnan Academy of Agricultural Sciences.s

4.1 Formation and evolution of the research partnership

The GCP drought-tolerant rice research partnership is actually built upon two distinct networks: the Upland Rice Shuttle Breeding Network (URSBN), focusing on rainfed upland, and the Drought Breeding Network (DBN), for rainfed lowland.⁷ The initiation of this partnership goes back to the early 1990s, when the Rockefeller Foundation (RF) took the initiative to improve the rice crop with a major focus on drought tolerance. At that time, within the mainstream of ICAR research, neither was the importance of drought appreciated nor did ICAR have the capacity to deal with it. According to one scientist, ‘Understanding drought research was a visionary step for Indian agriculture by the RF.’ RF’s initiative led to the formation of the International Rice Biotechnology Network (IRBN), and also to the Asian Rice Biotechnology Network (ARBN). Many scientists within the DBN and URSBN are actually come through the RF initiative.⁸

Another institutional setting that played a key role in the formation of the GCP research partnership is IRRI. Almost all the scientists in the above mentioned Indian research institutions have been involved in individual IRRI collaborations. IRRI used to deliver breeding material and training to these scientists.⁹ Several crop improvement projects were completed within this IRRI led individual collaboration with its partner institutions (mostly the GCP ones, as listed in Footnote 5). For example, in 1988, IRRI had led the Eastern India Rainfed Rice Project (EIRRP), which was funded by ICAR along with the International Fund for Agricultural Development (IFAD). The Upland Rice Research Consortium (URRC) was started in 1991 by IRRI comprising with its partners from India, Thailand, Philippines and Indonesia. During the last of its three phases, between 1997 and 1999, the Eastern India Farmers’ Participatory Breeding Project (EIFPBP) was established, partly-funded by the International Development Research Centre (IDRC).¹⁰

In the meantime, Dr Garry Atlin, now with the BMGF, joined IRRI as an upland breeder in 2000. In 2001, the Consortium for Unfavourable Rice Ecosystem (CURE) was established with a special focus on upland and lowland rice conditions. The CURE programme had six working groups on drought, submergence, and salinity, in which Dr Garry Atlin was the leader for drought component.¹¹ Between 2000 and 2002, Dr Atlin several times visited the CURE centres at Hazaribagh, Raipur, Faizabad, Bangalore, Cuttack and Coimbatore. It was his idea to create a network among these partners to continue the drought research. In 2002, a meeting at the CRRI was convened by Dr B.N. Singh, the then director of CRRI, to which all the partners were invited, in order to discuss the possibility of establishing a network. In 2003, the URSBN was formally launched. The funding for URSBN came initially through IRRI’s core funding. In a similar vein, around 2005, Dr Atlin suggested the idea of a similar network for rainfed lowland, the DBN. During this period, two meetings took place, at Raipur and Bangalore. However, at that time IRRI did not have the funding to support the DBN, so Dr Atlin talked to Dr John O’Toole of RF in Thailand about funding. Initially, RF funded DBN and soon GCP also joined. The first GCP project was also funded by the RF. In 2007, RF exited South Asia for Africa, since when GCP has been funding another three projects. Somewhere after 2005, Dr Atlin left IRRI, and Dr Arvind Kumar became the project leader for all projects.¹²

Now in its post-GCP phase, both URSBN and DBN continues to be maintained, funded mainly by the Stress-Tolerant Rice for Africa and South Asia (STRASA) programme, a Bill and Melinda Gates Foundation (BMGF) initiative. It should be added that the funding from external sources referred to applied (and apply) only to research, with other financial requirements, such as the scientists’ salaries and research facilities and other ongoing costs, being met by the respective institutions (i.e. from ICAR). ICAR also regularly evaluates the workings of this partnership.¹³

4.2 Functioning and internal organisational dynamics of the research partnership

The URSBN part of this research partnership is coordinated by the scientists from CRURRS from the Indian side, while the DBN part is coordinated by the scientists from CRRI. The research partnership is led by the scientist from IRRI and it is the main coordinating institution for the entire partnership. Leading this partnership doesn’t imply wielding authoritarian power to control or monitor activities of other scientist’s work, but to perform certain regular activities such as submitting research proposals to funding agencies, maintaining regular communications with partners and visiting the experimental sites of the partner institution. For example, the project leader of this partnership regularly visits all the different locations and partner institutes to observe the experimental work.¹⁴ Moreover, the scientists also connected with each other regularly through personal visits as well as email and Skype.¹⁵

The internal dynamics within this research partnership shaped the forms of research collaboration, which in turn determined the social organisation of the knowledge production. Thus, concerning the work culture within the network, the interviewees generally expressed their sense that relations within the partnership are friendly and based on mutual interest and that everyone is treated equally. They stressed that there was no centralised imposition or planning and that decisions regarding both the research and management of the partnership are taken through consensus, which tended to evolve through intense debate and discussion during the annual review and planning meeting. Once scientists are allocated responsibilities, they chose how to conduct the research themselves. Collaborative projects of ICAR such as AICCIP, National Agricultural Technology Project (NATP) or National Agricultural Innovation Project (NAIP) require a high degree of paperwork and adherence to strict bureaucratic procedures; in contrast to that the major focus in the workings of this partnership is entirely on research. The bureaucratic requirement is very less in the partnership projects as compared to ICAR. The entrenched bureaucratic procedures of ICAR also did not affect the functioning of the network.¹⁶

Regarding funding distribution during the GCP period, the major part was reserved for the CGIAR institute, in this case, IRRI.¹⁷ This is mainly due to the fact that the upstream biotechnology based research is done at IRRI, which is often costlier than conducting the field trials that usually takes place in partner institutions in India. Moreover, as a leading institution, it also has to bear certain logistical costs compared to its partner institutions. However, these decisions – regarding IRRI as a led institute or for that matter the fact that it had kept the major funding – were taken by the partnership through consultation, and it was not imposed by either the GCP or IRRI itself. Interviewees all reported that the leading and coordinator institutions did not behave in a centralised fashion, giving due credit and proper recognition to all the members of the partnership. Inclusion of new scientists into this partnership depends mainly on the following two points: first, the scientists wanting to be the part of this partnership has to be highly competent; and second, the scientist’s institute must have very good research infrastructure and access to drought prone agro-ecology to conduct field research. However, sustaining one’s position in the partnership purely occurs on merit. Usually, inclusion procedure officially occurs once in three years, with inclusion between difficult.

4.3 Reflection on forming the GCP rice research partnership

Summarising the above findings, the following issues become apparent. First, it can be stated that this partnership was not created by GCP rather GCP inherited a research partnership that has been evolving over last two decades at the interface of several projects (EIRRP, EIFPBP, URRC and CURE) in which resources have been mobilised from a pluralistic institutional initiatives such as RF, ICAR, STRASA etc. (Basu and Leeuwis, 2012). Second, it is evident that the emergence of this partnership is basically initiated by individual scientists, and it is not a result of institutional initiatives. Therefore, it can be stated that this partnership is developed in a bottom-up fashion (Ingram, 2015). Third, this partnership has developed its own rules and regulations by which it functions. These rules and regulations have been developed through rigorous consultations among its members. These regulations are not centrally imposed by the apex organisations such as ICAR, CGIAR or GCP, and these are quite diametric to the organisational culture of these apex bodies. For example, ICAR itself is a very rigid, bureaucratic, hierarchical institution with centralised planning (Raina et al., 2006; Hall et al., 2001; Mruthyanjaya and Ranjitha, 1998), whereas the functioning of this partnership is decentralised, non-hierarchical and absence of scientific bureaucracy. Therefore, this partnership is fairly autonomous in its functioning (Basu et al., 2017c).

5 Directing research towards transdisciplinarity

In order to understand transdisciplinarity within the partnership, we investigated its approach to the research. This was shaped by two factors: first, the presence of drought as an uncertain phenomenon and the scientific strategy applied to deal with this, and second, the presence and development of many drought-tolerant varieties that went uncultivated due to farmers’ dislike of varieties with low yield potential.¹⁸

Because of the inherent complexity of drought as a very uncertain phenomena (in terms of occurrence, stage and intensity), it was necessary to develop a particular scientific strategy and vision with which to do research and then operationalize that strategy within the ICAR system. In developing this strategy, scientists reached a consensus on the following three issues. First, it was agreed to take grain yield as a primary consideration, with comparison of yields in both stressed and irrigated conditions. This became the basic principle for subsequent breeding activities. Second, in order to investigate the effect of the various locations on the reduction of grain yield, the decision to check the developed breeding lines at different places and at different levels of stress conditions was taken (in order to reduce environmental effects, breeding lines have to be developed over several years at one place or else through multi-locational trials). The location-dependent variation of genotype-environment (G*E) interaction made this a scientifically difficult job. Third, the partnership decided to carry out multi-locational trials in its various agro-ecological zones.¹⁹ At that time, in fact, the performance of multi-locational trials was not an option available even within ICAR, and even within the AICRIP, only five trials in two to three locations were possible, compared with the much larger trial numbers that were required in many, ecologically different locations.²⁰

Consensus was reached on these three issues during the annual review and planning meeting, and a complex, scaled-up system of trialling was organised to improve varieties to become drought and also disease resistant. Accordingly, crosses were made mainly at IRRI and breeding material collected from the partners at CRRI. Off-season multiplication was performed at CRRI, and the different breeding material was then sent back to the partners, including IRRI. Generally, 300–400 initial breeding lines were tested through observation yield trials (OYTs) in all the locations, in both irrigated and stressed conditions. Those lines that performed well in both irrigated and drought conditions were then tested in Advance Yield Trials (AYTs), which generally comprised 50–70 lines. After the AYTs, 10–12 promising breeding lines were nominated for the ACRIP trials and these lines evaluated in the farmers’ fields, through PVS.²¹

Alongside the (non-academic, extra-disciplinary) participation of farmers’ groups (described above), the partnership also involved professionals engaged in a range of disciplinary activities. In addition to conducting the PVS and associated breeding work, these involved field screening (phenotyping), genotyping (often with advanced level molecular and biotechnological tools) and screening (for disease and pest resistance). In order to perform all these activities, the partnership comprised a high number of breeders, made up of professionals across several disciplines (in plant breeding, genetics, physiology, pathology, entomology, biostatistics and social sciences). At each step of such variety development, the disciplinary input involved is essential for the next and somewhat intertwined. Thus, while single disciplinary inputs were applied at each stage – or, stages could be defined by the input (therefore discipline) involved – the process of producing knowledge related to rice-variety drought tolerance actually comprised multiple inter-disciplinary interactions.

5.1 Reflection on directing research towards transdisciplinarity

The findings showed the approach of the GCP drought-tolerant rice research to be embedded within the disciplinary structure of modern agricultural sciences. Instead of a transdisciplinary approach (Gibbons et al., 1994), this case illustrates rather the application of a disciplinary plant-breeding approach, albeit executed through cooperation among a range of various specific (sub-) disciplines leading to a kind of integration among several (sub-) disciplines, such as plant-breeding genetics, agronomy, plant physiology, entomology and plant pathology, biotechnology and social sciences. These (more specifically their representatives, the scientists involved) were all committed to developing a plant-breeding solution to and thus understanding of the drought problem, but perceived from different perspectives, insights, data, concepts, theories and methods as related to the particular (sub)disciplines of the plant-breeding framing of the drought problem in which they had specialised. Therefore, although the partnership’s research approach and practice may be said to represent a disciplinary integration, that was not one, in our opinion, that transcended disciplinary boundaries so as to qualify as transdisciplinary (Maasen et al., 2006; Repko, 2008).

6 Including the end-users in research: the involvement of farmers

As end-users, the place of farmers in relation to the partnership’s knowledge production is particularly important in Instrumental paradigm and systems paradigm. Aiming thus to consider the GCP rice research partnership from a societal perspective, we posed two types of questions to farmers who use drought-tolerant varieties for cultivation: the first regarding the research process conducted inside the institutes, and the second about the research conducted with their help through PVS. In respect of the former, we received unanimously negative responses. Every one of the 100 farmers questioned said that they neither participated and nor had the expertise to participate, and nor were they interested in setting the research agenda. In respect of PVS, on the other hand, we received almost unanimously positive responses. All the farmers questioned considered that they participated, felt involved in this participation and had the expertise for it, with 87 of the 100 stating that they were being sponsored (paid) for their participation.

These results indicate that, firstly, farmers did not have any interest in contributing to the research activities that occurred inside the institute; they tended not to possess the academic skills, expertise or language skills, especially in English, to contribute to the research activities or agenda formulation. They were satisfied with their inclusion in the research process only at a much later phase, in which respect, the investigation revealed that farmers generally did participate in the PVS trials and field level demonstrations (FLD), and sometimes also in village seed programmes (e.g. the seed multiplication programme). Their opinions were taken very seriously by the scientists during the PVS trials, and that these were usually fully sponsored by the institutes, making them willing to participate. Indeed, the farmers were compensated by the institutes when there was a crop failure, so the programme acted as a kind of price-support mechanism, guaranteeing their income, which would have otherwise been at risk.²² PVS operated through farmers at the level of variety performance (and rice quality), as determined not only verbally through interaction with the researchers but also in action, by farmers’ dissemination, related extension activities and seed multiplication and sale. Moreover, researchers stress how they felt that their research was validated precisely when the varieties they had tested were accepted and became used by farming communities.²³

The farmers did not consider themselves as stakeholders in the research partnership. In this respect, farmers might be considered as outside the partnership (and hence were not included in the consideration of internal dynamics, above). However, they did see themselves as important actors in it. This was unanimously expressed during the FDGs in respect of the PVS trials, since these gave them access to new and improved breeding lines. Farmers mostly expected good varieties from the scientists, defined as varieties that would both withstand abiotic-biotic stresses and be productive. Most of the farmers held the scientists in high regard and testified to sharing very cordial working relations with them. Interestingly, they often disagreed with the scientists, and they also often modified the agronomic practices to their convenience. Farmers generally discussed these changes with the scientists, and they claimed that both sides learnt greatly from each other, which was confirmed by several scientists.²⁴

Sahbhagi Dhan is one of the drought-tolerant varieties that was developed through this partnership, with the cooperation with farmers particularly in the Hazaribagh region. The formal, public release for cultivation of the Sahbhagi Dhan through government organisations by the Central Varietal Release Committee (CVRC), in October 2010, took place only after several years of trials (Mandal et al., 2009). The farmers identified with their involvement in developing this variety, and, through PVS, they enjoyed access to the seed throughout the testing period. In fact, they independently and informally chose to share the successful variety with their friends and neighbours, so many local farmers who were not involved in the PVS also managed to gain access to it. This further indicates the role of the farmers within the network, both as agents (insofar as they acted autonomously) and for dissemination of the material end product (insofar as they effectively extended the trial community, since any issues with the new variety would most likely have been communicated by word of mouth, back to the identified farmers and thence the scientists).²⁵

6.1 Reflection on including the end-users in research

This case study revealed the end-users (farmers) as involved at a peripheral level of the actual research (through PVS), which may perhaps be categorised as research-driven farmer participation to maximise the research impact (Sumberg et al., 2003). In line with other research on farmers participation (Courtois et al., 2001; Sumberg et al., 2003), this study also confirms that the non-involvement of farmers in the actual research or in the agenda setting need not negatively affect its societal relevance if the problem addressed is widespread and well known (like drought) and if the research process involves the application of advanced sophisticated technologies (such as, here, molecular breeding techniques). Our findings suggest that actually farmers do not want to participate in the upstream activities, and – again in line with other research, particularly in the developing world (Jones et al., 2014) – that the main reasons for farmers’ participating in the knowledge production process through PVS are the perception that it is likely to help them increase yield and gain access to improved varieties. The regular benefits of PVS – identifying preferred varieties and gaining access to the new variety (such as the Sahbhagi Dhan) before national release (Singh et al., 2014) – were also reaped in this case. However, PVS here was slightly different than other PVS trials (Joshi et al., 2012), as the full sponsorship by the partnership, even in the case of crop failure, was also a major benefit for farmers and encouraged their participation.

7 Discussion

In this section, we discuss the above findings. This section is divided into three parts. First, we indicate the learning outcomes from the GCP drought-tolerant rice research case. Second, we typify the nature of the knowledge production discourse emerging from the case. Third, we outline the implications that can be discerned from the case on research organisation.

7.1 Learning outcomes from the GCP drought-tolerant rice research case

The following learning outcomes in relation to knowledge production discourse can be discerned from the GCP drought-tolerant rice research case as elaborated in the previous section. First, the basic rationale behind building-up the research partnership was not to optimise resources or to minimise transaction costs of doing research – a consideration often introduced in support of Instrumental paradigm and systems paradigm (Gibbons et al., 1994; Spielman and von Grebmer, 2004) – rather, it only aimed at tackling the complexity of drought problem through addressing the issue of G*E interactions (by conducting multiple trials in various agro-ecological zones). Second, complex scientific problems (such as drought) often require the application of advanced, cutting-edge technology (e.g. using molecular markers) in an advanced science (genetic engineering) that is embedded strongly in a particular disciplinary structure (agricultural biotechnology, or agri-tech). This is qualitatively different from the transdisciplinary approach, which mobilises a range of theoretical perspectives and practical methodologies to solve problems that are not necessarily derived from pre-existing disciplines or formative of new disciplines (Repko, 2008). Third, hitherto, agricultural research partnerships – those are able to deliver and sustained for a longer period in India – have either had private sector involvement (see Hall et al., 2001) or centrally led by the apex public institute (for example, the AICCIPs, NATP and NIAP) (Hall et al., 2001; Raina et al., 2006). In contrast to this, this research partnership was essentially led by itself (scientists) and continues to exist over last two decades and manages to deliver in terms of knowledge production (such as developing drought tolerant varieties and producing research papers). Fourth, mobilisation of resource from diverse source is one of the necessary preconditions for a successful research partnership building (Basu and Leeuwis, 2012). Fifth, end-user participation in the research process does not necessarily correspond to higher impact in societal problem-solving (Sumberg et al., 2003). Sixth, validation of scientific knowledge by an extended peer community (farmers here) does not replace scientific validation process or methods, but rather supplements and extends it. Scientific community validation still is very important for the scientific community, but increasingly in a way that acknowledges the importance of extended peer community evaluation.

7.2 Typifying the nature of GCP knowledge production

In this section, based on the findings presented in the previous section on the knowledge production process of GCP drought-tolerant rice research, we typify the nature of knowledge production emerging out of the three configured process. First, the knowledge production evolved (still continuing to evolve) within an agro-ecology mediated and institutionally pluralistic network (Hall et al., 2001). Thus, the knowledge production is situated within in a supra-institutional organisational composition and/or institutional hybridisation that were nevertheless essentially transient in nature (Gibbons et al., 1994; Etzkowitz, 2003). Second, the knowledge production process successfully forged a continuing informal system of linkages through repeated interactions between a range of institutions (Thune and Gulbrandsen, 2011) that extended (two) previously developed systems over a period of more than decades. However, the linkages did not expand beyond the scope of ICAR institutions or state agricultural universities. It did not grow to encompass, for example, (public or private) non-agricultural universities, Department of Biotechnology (DBT) or Department of Science & Technology (DST) bodies or (the autonomous but DST-linked) Council of Scientific and Industrial Research (CSIR) institutes, private R&D institutions or extension agencies, or other knowledge-based policy organisations, civil society groups or NGOs. Similarly, media forums (traditional or Internet-based) that could have played a major role in dissemination and adoption of the drought-tolerant varieties developed were not incorporated in this partnership. Third, the knowledge production process can be considered as autonomous as it is marked by absence of centralised planning, institutional hierarchy and scientific bureaucracy that are hall mark of ICAR (Mruthyanjaya and Ranjitha, 1998). Fourth, the knowledge production process has extended its boundaries to include a societal representation at the periphery. This offered possibilities for the tailoring of technology to context specific needs, for the development of an arena where technology is negotiated and disseminated. Thus, this knowledge production manifests a middle-ground route that allows a negotiated approach beyond the black-and-white acceptance-versus-rejection of technology debate (Ruivenkamp, 2005; Ruivenkamp, 2008; Landini, 2016). Fifth, the knowledge production process of GCP drought-tolerant rice research is embedded within the disciplinary structure of modern plant breeding simultaneously working in cooperation with various allied disciplines such as agronomy, plant physiology, entomology and plant pathology, biotechnology and social sciences. The dominance of plant breeding disciplinary trajectory is understandable as the sheer number of plant-breeders within the knowledge production process was very high. This tended to have the effect of making breeder perspectives dominant. Thus, other trajectories of drought mitigation such as root intensification got little attention (Basu et al., 2017c). Sixth, validation of scientific knowledge emerging out of the knowledge production process of GCP drought-tolerant rice research is performed not only by the scientific community (academically trained experts) but also by the end-users, the farmers as practitioner experts (Funtowicz and Ravetz, 1993; Gibbons, 1999; Gibbons et al., 1994). Of course, the fact that the validation of scientific research is increasingly performed by extended peer communities or end-users at large does not in itself undermine the role of scientific validation as performed by the scientific community.

7.3 Implications

Bearing this overall discussion in mind together with the specific empirical results of the study conducted, the following policy implications may be drawn as applying to ICAR and other, similar NARS and networks, as well as to this, continuing network itself. First, insights from this case, such as sustaining a research partnership through decentralised planning, absence of scientific bureaucracy and an interdisciplinary approach can become key issues for institutional reform, to move away from rigid, hierarchical institutionalism and overly mono-disciplinary orientations for a systems approach that, in the case of the functioning of ICAR, is long overdue. Second, when researching complex scientific problems such as drought, conducting the research within an agro-ecologically diverse set-up can assist understanding (the different manifestations of drought in different places) and providing structural restriction (to analyse and overcome the genetic and environmental interactions within a limited timeframe). Third, ICAR should ensure the meaningful participation of end-users at an appropriate phase of research so as to achieve socially relevant research outputs and increased dissemination and adoption for the improved impact of research results in farmers’ fields. Fourth, PVS trials should be further encouraged, as PVS provides an arena in which two different knowledge systems – roughly, theoretical and practical – meet and shape each other, which is beneficial for overall agrarian knowledge systems. Fifth, the partnership should include more actors such as policy makers, innovations brokers, social scientists and other agricultural professionals along with better, Internet-based media communications to make a greater impact at the farmer’s level.

8 Conclusions

In view of the above discussion on the knowledge production process by taking into account three processes: forming a research partnership, directing research towards transdisciplinarity, and including end-users; it can be now concluded that the way these three processes are expressed in this case do not adhere to their definitions provided in the non-instrumental & Instrumental paradigm and linear & systems paradigms. The way these three processes emerged goes beyond the regular ambit and often contradicts to its existing understandings. For example: the partnership is neither state led (as per non-instrumental and Linear paradigm) nor it is market led (as per Instrumental paradigm and systems paradigm), but led by the partnership itself irrespective of the fact that the partnership is situated within the public research institutions; the research approach is neither disciplinary (as per non-instrumental and linear paradigm) nor it is transdisciplinary (as per Instrumental paradigm and systems paradigm), but interdisciplinary in nature leading to the formation of new disciplines; and the research is neither entirely guided by the science establishment (as per non-instrumental and linear paradigm) not it is guided entirely by the end-user or by extended peer community, rather it includes end-users at a peripheral level that provides opportunity to attune knowledge in a context specific manner. Therefore, the knowledge production discourse that has emerged in this case is ensemble, extended and hybridised in nature taking elements from all the different paradigms in a diverse way. Moreover, this hybridised discourse also contradicts the observation that Instrumental paradigm has replaced the Non-instrumental paradigm as some of the features of non-instrumental paradigm also resurfaced in this hybridised discourse in a different manner. As such, this hybridised discourse can be seen as a third way that goes beyond the traditional dichotomy between either Non-instrumental or Instrumental paradigm or linear and systems paradigm respectively.

Finally, we would like to end the paper by highlighting about the alter-institutional space that has been created within the actual institutional space of several national and international institutions where the knowledge production is situated. The actual institutional space could be characterised as institutions with entrenched scientific bureaucracy and top-down managerialism, whereas this alter-institutional space is characterised by decentralised organisational base and volunteerism in which scientists co-operatively pursue innovative research with colleagues from several other organisations. For instance, this case showed that drought related research started within this alter-institutional space long before ICAR understood or had the capacity to engage in it. Therefore, this particular alter-institutional space required further exploration to understand its role in setting the research discourse within the ICAR.

Acknowledgements

The authors would like to acknowledge financial support provided by the Global Development Programmes, Netherlands Organisation for Scientific Research (NWO-WOTRO) through its integrated project titled ‘Open Source and Commons in India and the Netherlands: New Perspective for Development’ (Grant Number: W01.65.326.00). The authors would like to thank all the scientists in DBN and URSBN for sharing their story with us. Farmer’s contributions from the Hazaribagh region are also gratefully acknowledged. We are also highly appreciative of the input from two anonymous reviewers who made constructive comments on earlier versions of this manuscript. Finally, we would like to express our sincere thanks to the Editor (Sietze Vellema) of NJAS for his extremely valuable suggestions that were very helpful in improving the quality of the manuscript.

Notes

Initial ideas of this manuscript were presented at the 14th Global Conference of the International Association for the Study of the Commons, Mt.Fuji, Japan. 3–7 June, 2013 by the first author.

1 http://www.generationcp.org/.

2 http://www.cgiar.org.

3 These are mostly Indian Council of Agricultural Research (ICAR) research institutes, State Agricultural Universities (SAUs) and NGOs.

4 This research was one of the four research projects under the integrated programme titled ‘Open Source and Commons in India and the Netherlands: New Perspectives for Development’ funded by the Science for Global Development Programmes, Netherlands Organisation for Scientific Research (NWO-WOTRO), which generally funds interdisciplinary, North-South research projects focusing on sustainable development and poverty reduction to strengthen the bridge between research, policy and practice; see http://www.nwo.nl/wotro.

5 In India, Indira Gandhi Krishi Vishwavidyalaya (IGKV), Raipur; Central Rainfed Upland Rice Research Station (CRURRS), Hazaribagh; Narendra Dev University of Agriculture and Technology (NDUAT), Faizabad; Birsa Agricultural University (BAU), Ranchi; Central Rice Research Institute (CRRI), Cuttack; Tamil Nadu Agricultural University (TNAU), Coimbatore; Barwale Foundation (BF), Hyderabad; University of Agricultural Sciences (UAS), Bangalore; Anand Agricultural University (AAU), Anand; Orissa University of Agriculture & Technology (OUAT), Semiliguda; Maharana Pratap University of Agriculture and Technology (MPUAT) Banswara; Vivekananda Parvatiya Krishi Anusandhan Sansthan (VPKAS), Almora; and Jawaharlal Nehru Krishi Vishwavidyalaya (JNKVV), Rewa; and in the Philippines, International Rice Research Institute (IRRI) (Los Banos) (CGIAR Generation Challenge Programme, 2011).

6 ICAR is the Indian NARS; it is a huge network focussed on several crops throughout India that is quite unique among other NARS. See http://www.icar.org.in.

7 Interviewee 1, Hazaribagh, IN, 2–4 June 2012.

8 Interview, Interviewee 3, Bengaluru, IN, 19–23 June 2012.

9 Interviews: Interviewee 4, Coimbatore, IN, 25–26 June 2012; Interview, Interviewee 5, Raipur, IN, 12–13 July 2012.

10 Interview, Interviewee 1, Hazaribagh, IN, 2–4 June 2012.

11 Interviewee 1, Hazaribagh, IN, 2–4 June 2012.

12 Interviews: Interviewee 1, Hazaribagh, IN, 2–4 June 2012; Interviewee 2, Hazaribagh, IN, 2–4 June 2012; Interviewee 6, Ranchi, IN, 6–8 June 2012; Interviewee 7, Cuttack, IN, 15–18 June 2012; Interviewee 8, Cuttack, IN, 16–18 June 2012; Interviewee 5, Raipur, IN, 12–13 July 2012.

13 Interviews: Interviewee 1, Hazaribagh, IN, 2–4 June 2012; Interviewee 7, Cuttack, IN, 15–18 June 2012.

14 During an interview with the project leader (at IRRI, Los Banos), he received a Skype call from a network scientist, in which they discussed the watering of breeding lines at the scientist’s station.

15 Interviewee 1, Hazaribagh, IN, 2–4 June 2012; Interviewee 2, Hazaribagh, IN, 2–4 June 2012; Interviewee 6, Ranchi, IN, 6–8 June 2012; Interviewee 7, Cuttack, IN, 15–18 June 2012; Interviewee 8, Cuttack, IN, 16–18 June 2012; Interviewee 5, Raipur, IN, 12–13 July 2012.

16 Interviews: Interviewee 1, Hazaribagh, IN, 2–4 June 2012; Interviewee 2, Hazaribagh, IN, 2–4 June 2012; Interviewee 6, Ranchi, IN, 6–8 June 2012; Interviewee 7, Cuttack, IN, 15–18 June 2012; Interviewee 8, Cuttack, IN, 16–18 June 2012; Interviewee 5, Raipur, IN, 12–13 July 2012.

17 Interviewee 1, Hazaribagh, IN, 2–4 June 2012.

18 Interviews: Interviewee 5, Raipur, IN, 12–13 July 2012; Interviewee 3, Bengaluru, IN, 19–23, June 2012.

19 Interviews: Interviewee 5, Raipur, IN, 12–13 July 2012; Interviewee 9, Los Banos, Philippines 2–4 May 2012.

20 Interview, Interviewee 7, Cuttack, IN, 15–18 June 2012.

21 Interviews: Interviewee 9, Los Banos, Philippines 2–4 May 2012; Interviewee 10, Ranchi, IN, 6 June 2012; Interviewee 11, Coimbatore, IN, 25 June 2012.

22 FGD with Farmers, Koderma District, Jharkhand, 14 December 2012; FGD with Farmers, Hazaribagh District, Jharkhand, 17 December 2012; FGD with Farmers, Chatra District, Jharkhand, 18 December 2012.

23 Interviews: Interviewee 12, Hazaribagh, IN, 4 June 2012; Interviewee 2, Hazaribagh, IN, 2–4 June 2012.

24 FGD with Farmers, Koderma District, Jharkhand, 14 December 2012; FGD with Farmers, Hazaribagh District, Jharkhand, 17 December 2012; FGD with Farmers, Chatra District, Jharkhand, 18 December 2012.

25 Sahbhagi Dhan’s impressive performance in successive PVS trials between 2005 and 2007 also compelled the state National Seeds Corporation Ltd. to undertake an intensive seed production programme in collaboration with some of the institutional partners of the network, thus extending the farmer involvement (Mandal et al., 2009).