Making the case: an actor-network
approach to rhetoric in agricultural development

Construyendo el caso: una aproximación actor-red
a la retórica del desarrollo agrícola

Gerad Middendorf^*
Kansas State University

^*Associate professor od Sociology. Deparment os sociology, anthropology and social work. Kansas state university. middendo@ksu.edu
Dirección: 204 waters hall - Kansas state university, Manhattan, Ks 66506 - 4003.

fecha de recepción: febrero 23 de 2013
fecha de aceptación: junio 20 de 2013

Abstract

This paper examines the shifts in language associated with the rise and decline of dominant development models. It draws on Bruno Latour's approach to rhetoric to analyze how actors in development policy networks use language strategically in order to persuade other actors that their development narrative is the most plausible narrative possible. This study focuses on a particular time period - the 1980s-1990s - in which we saw the transition to the neoliberal model. First, the annual reports of the Inter-American Development Bank and other key actors are systematically analyzed to understand the shift in language associated with development models. Then, analysis of interview data shows how individual actors interpret and translate policy change in everyday language. What emerges is a circulation process, in which development language is appropriated by actors in policy networks and refashioned to help each actor "make the case" in order to advance their interests. A strength of the Latourian approach is that it helps us to reveal the strategic discursive practices of authors (scientists, policy makers, researchers). The paper attempts to extend actor-network analysis beyond science studies, where it has been most fully developed, into development studies where it has thus far received relatively less attention.

keywords: Development, language of development, Latour.

Resumen

El presente documento examina los cambios en el lenguaje, asociados con el ascenso y descenso de modelos de desarrollo dominantes. Se apoya en los aportes de Bruno Latour a la retórica, para analizar cómo utilizan el lenguaje, de modo estratégico, los actores en políticas de desarrollo, con el fin de persuadir otros actores de que sus narrativas de desarrollo son las más plausibles posible. Este estudio se enfoca en un período particular de tiempo (1980 - 1990) en el cual vimos la transición del modelo neoliberal. Primero, los reportes anuales del Banco Inter-Americano de Desarrollo, así como otros actores son sistemáticamente analizados para comprender el cambio en el lenguaje asociado con los modelos de desarrollo. De este modo, el análisis de datos de entrevistas muestra cómo los actores individuales interpretan y traducen los cambios de políticas en el lenguaje cotidiano. Lo que emerge es un proceso de circulación en el que el lenguaje del desarrollo es apropiado por actores en redes de políticas y reestructurados para ayudar a cada actor a "llegar al punto", a fin de avanzar en sus intereses. Una fortaleza de la aproximación de Latour es que nos ayuda a revelar las prácticas discursivas estratégicas de los autores (científicos, hacedores de políticas, investigadores). El artículo intenta extender el análisis de actor-red más allá de los estudios científicos, en donde ha sido desarrollado principalmente, hacia estudios de desarrollo en donde ha recibido, hasta ahora, menor atención.

Palabras clave: Desarrollo, lenguaje del desarrollo, Latour.

Introduction

Development policies are based on claims about how development¹takes place. Underlying any prescriptive policy is a theory - implicit or explicit - about how societies "progress" towards greater food security, higher incomes and standards of living, and about how best to organize the productive resources of society in order to achieve that progress. While the policies being promoted by the major actors (in this case, the World Bank and the Inter-American Development Bank) at a given point in time can be seen as the dominant development model, these models do come in and out of favor. This raises the question: If the major actors possess the definitive knowledge about how development takes place, and about how to craft the most effective policies to promote it, then what explains the change in the dominant development models over time?

An important indicator of the rise and decline of such models is the shift in the language associated with them. This language can be found in policy and project documents and in the spoken language of practitioners. While it is important to analyze this language shift, it is equally important to ask: How do some actors convince others that their particular organization of language represents the real version of how development takes place? How do some actors succeed in persuading others that their version of reality is the correct one? In short, how do they make their narrative the narrative? This paper takes a Latourian approach to that question. First I review the substance of the shift in language associated with the rise and decline of development models. To do this I show the change in language use over time in the annual reports of the Inter-American Development Bank (IDB) and the Panamanian Agricultural Research Institute (IDIAP). Then, in order to better understand the relationship between the language shifts of IDB and IDIAP, I examine the rhetorical strategies used by the authors, following Latour's approach to rhetoric (1987). "Rhetoric" is used here to refer to all those means utilized by an author or speaker to persuade an audience in a given situation, i.e., to convince another actor that the presented version is not only the correct version, but the only possible version. It is through rhetorical tactics that "readers" are kept in line, directed and controlled. The analysis of rhetorical strategies is extended beyond IDB and IDIAP to include the chain of actors between these two organizations that link them into a policy network.

On the one hand, a better understanding of rhetorical strategies can help provide a more instructive account of the major actors' source of influence in defining development and the directions it takes. On the other hand, analyzing the rhetoric of the other actors in the policy network also reveals how actors at each link interpret and translate policy language to reflect their own interests. Finally, analysis of interview data examines how individual actors interpret and translate policy change in everyday language. This study focuses on a particular time period - the 1980s-1990s, in which we saw the transition into the neoliberal development model. The contribution of this paper is to extend actor-network analysis beyond science studies - where it has been most fully developed - into development studies, where it has thus far received relatively little attention. Let us turn now to examine the shift in language associated with the rise and decline of development models.

Shifting Language: The Inter-American Development Bank

To begin, let us compare two sources of development language: annual reports of IDB and IDIAP, respectively. The shift in language can be readily observed in the table of contents of the annual reports of IDB. A systematic review of the table of contents provides an indication of IDB's emphasis at any given point in time, and how the Bank has framed categories of interest. Table 1 contains a listing of the substantive categories in the table of contents of available IDB annual reports from 1968 to 2000. By "substantive categories" I mean those headings that indicate a substantive category of interest and activity on the part of the Bank, such as "Environment" and "Women in Development." These categories signify that in that year the IDB organized a subset of its lending activities around a specific topic. The point here is not that a change in language necessarily indicates an associated change in the substance of practices. It might be, for example, that a new linguistic framework does not relate strongly to a change in practice. Categories left out of Table 1 include those which deal solely with the operations of the bank, such as "Evaluation and Internal Audit" and "Borrowings."

Table 1 gives a fairly complete listing of the substantive table of contents categories. This provides a sense of the range of language used from 1968 to 2000 in order to describe categories of interest. While some categories show remarkable endurance over time, e.g., "Economic Integration" (1973-2000), others disappear almost as suddenly as they appear, e.g., "Indigenous Groups" (1996 only). Other categories, such as "Women in Development," have an intermediate life span (1987-1996). Table 2 below summarizes this data to show the shift in language over time. Again, the year and category are presented. A dot in the cell (•) indicates that the category appeared in the table of contents of the IDB annual report of the year shown. A number of changes are instructive. From the earliest annual reports, the emphasis on technology as a linguistic framework for organizing development efforts is obvious. "Technical Assistance" (1968) changed to, "Technical Cooperation" by 1973, then remained unaltered through the 2000 report. Given the ebb and flow of development language, the persistence of this particular category for more than twenty years is indicative of the strength of the belief in technology as an important approach to resolving development problems.

Two categories that endured throughout the 1980s were "Financing for Small Projects" and "Support for Low Income Groups." Interestingly, soon after "Financing for Small Projects" disappeared from the table of contents (after 1989), the term "Microenterprise" appeared (1991), and remained through 1996. "Microenterprise" is arguably more in tune with a focus on the individual entrepreneur and the rise of the neoliberal development paradigm. This contrasts with the "small projects" language, which is more suggestive of collective projects. Moreover, "Financing for . . ." places the emphasis on what the IDB is doing for some group, as opposed to "microenterprise," which shifts the focus to the activities of the enterprising individual. In 1991, what was "Support for Low Income Groups" becomes simply "Low Income Groups," which may seem inconsequential, but it certainly is less of an advocacy stance. The latter is dropped altogether in 1994.

The environment entered the IDB table of contents for the first time in 1985, where it remained through 2000. "Women in Development," entered the table of contents in 1987 and remained a category for about a decade, after which it was dropped. The inclusion of women in the table of contents was not the first time that a specific group had appeared as a category. "Low Income Groups," (1980-1981) "Central America" (1983-1984) and "Indigenous Groups" (1986) are other groups that emerged and disappeared as specific categories.

Two categories added in the 1990s, perhaps the clearest indicators of the shift in language toward the neoliberal model, are "Modernization of the State" and "Private Sector." In contrast to the minimalist language of other categories in the late 1990s, such as "Private Sector" and "Environment," "Modernization of the State" is action-oriented and less neutral. It is more clearly suggestive of a development agenda. If the decade of the 1980s was representative of an approach to development that emphasized support for marginalized groups (low income groups, small projects, women), then the language from the 1990s on in particular is representative of the linguistic shift toward the neoliberal model (e.g., "Microenterprise," "Modernization of the State" and "Private Sector").

Shifting Language: The Panamanian Agricultural Research Institute (IDIAP)

A similar shift is discernible in IDIAP's documentation from 1979 to 2001. Table 3 contains excerpts from IDIAP's annual reports from 1979 to 2001. The reports were reviewed systematically in order to treat reports from all years uniformly. The excerpts were selected only if there was explicit reference to IDIAP's mission or objectives. Reference to mission and/or objectives was chosen because of the high likelihood that issues, themes and goals important to IDIAP would be expressed in these statements.

There are at least three key linguistic transitions evident in these excerpts that deserve mention: 1) the shift in IDIAP's language defining its clientele, 2) the shift from a narrow to a broad definition of objectives for IDIAP, and 3) the shift from the language of supplying technologies to responding to client (market) demand for technologies. Let us briefly consider each of these in turn.

Redefining clientele. As shown in Table 3, in 1979 the primary clientele was identified as agricultural producers, ". . . with emphasis on small producers." This early definition of clientele is consonant with the populist state agricultural policy at the time, which focused attention on the reform sector and the rural poor. In the 1989 annual report, IDIAP's mission statement identifies ". . . farm workers, principally those that are marginalized, and small and medium farmers" as the primary beneficiaries of research efforts. Even though medium size farmers appear in this definition, it is somewhat surprising that marginalized farm workers would have appeared in the mission statement, given that Panama was already under structural reform programs by that time. Yet, even as late as 1990, producers, and especially ". . . small producers and marginalized campesinos" were the stated target of IDIAP's efforts. In 1994 the shift became more obvious, with the clientele constructed more broadly as "producers." Moreover, one of the specific objectives in that year was to "Ensure the participation of the sector (producers, industrial suppliers, public and private entities) . . ." in the process of generating technologies. Finally, in 1996 and 2001, the mission statement identifies "agribusiness" as a main clientele, with a reference to broader benefit for "Panamanian society."

Thus, in IDIAP's own language, the range of definition of its clientele is from "small producers" in 1979 to "agribusiness" in 2001. This certainly reflects the shift from a state-led development model to the neoliberal model, which emphasizes the role of the private sector in spurring the growth of the sector.

From narrow to broader objectives. In IDIAP's early years, the clearly defined objectives were to carry out research programs leading to knowledge and technologies "... to raise the production and productivity, as well as the income level of agricultural producers . . ." The basic language of production and productivity remained fairly stable, while the end goals of research progressively broadened. For example, in 1990 research was not only to increase production, productivity and incomes, but to "improve domestic supply as well as export possibilities." In the same year conservation and rational use of agricultural resources (i.e., environment) enters into the language of objectives. The 1990 objective of improving "domestic supply" is supplanted from 1992 on by the somewhat weightier objective of achieving "food security." Now agricultural research will be responsible not for merely improving domestic supply, but for achieving food security. The objectives become still broader. In 1994, increased economic and productive efficiencies from new technologies are to guarantee "desired levels of sustainability." In that same year, no longer focused on just agricultural producers, IDIAP stated one of its objectives as: "Promote the industrialization of the sector." Thus, agricultural research was now claiming broader responsibility for the growth of the whole agricultural sector. Finally, in the latter half of the 1990s, included in IDIAP's mission statement was its broadest framing of goals: to contribute to ". . . food security, competitiveness and the sustainability of agribusiness in benefit of the Panamanian society." The scope of IDIAP's mission and objectives has shifted from raising production and productivity to benefit small producers to research that will benefit the entire society! It is not necessarily that the actual research or its impact changed, but that the claim of societal benefit was broadened. This broadening of objectives is partly a response to increased and broadened demands on agricultural research organizations. I argue below that their linguistic response is a sign that in an era of "modernization of the state" they were increasingly being forced to "make the case" for their own existence. An important part of making the case is broadening their claims through stronger rhetorical strategies, in order to show that they are beneficial not just to a group of producers, but indeed to the larger society.

From supply to demand orientation. Evident in the 1984 and 1989 mission statements is the notion that IDIAP was supplying the agricultural sector with technologies. No mention was made of the demands of the clientele. The clientele seems to be "out there," rather distant from the organization. IDIAP presumed to know the needs of producers, and saw its role as generating technologies for producers, in what appears to have been a largely internal process. On its own, IDIAP will ". . . design, promote, stimulate, coordinate and execute research activities to generate knowledge and technologies . . ." (IDIAP 1987). The shift towards seeing the clientele as "demanders" of technologies, to which IDIAP should respond became apparent in 1994. In that year, part of IDIAP's objectives was to ". . . respond to the needs of the producers and the demands of the market." Moreover, since at least 1990, part of the mandate of the sector has been to increase agricultural exports. In order to compete in international markets, farmers need to be competitive. Thus, the logic is that (a) Panamanian producers are trying to compete in international markets, (b) to do so they need cost-reducing technologies and technologies that will help them produce commodities that will meet the standards of international markets, c) producers know what technologies they need in order to compete in these markets, and thus (d) by responding to the demands of the producers and others down the commodity chain (e.g., processors, traders), IDIAP is in effect responding to the market. This keeps it closely in line with the government's development model for agriculture - growth through export of non-traditional commodities. Indeed, in 2001, IDIAP's objectives include "responding] to the demands of the clients, users, and beneficiaries . . ." and "contribute[ing] to increasing efficiency" [and] competitiveness . . ."

In sum, the shift at the IDB has been from language that emphasized support for marginalized groups to the language of the neoliberal model. The shift at IDIAP has similarly been from language that emphasized a clientele of small producers, marginalized campesinos and farm workers to language that emphasizes the private sector. Figure 1 depicts how these are related over time, with IDB's language on the upper time line and IDIAP's language on the lower time line.

During the same years that the IDB language reflects an interest in support for low income groups and small projects, IDIAP defined its clientele as small producers, marginalized farm workers and marginalized campesinos. Likewise, during the same years that IDB language showed a shift toward emphasis on microenterprise, modernization of the state, and the private sector, IDIAP defined its clientele more in terms of private sector interests. The next section examines the rhetorical strategies - of these two organizations as well as others that are linked to them - as a way to better understand the language shift associated with changing development models.

Rhetorical Strategies

What accounts for the association between the shift in language at IDB and IDIAP? Is it that a small agricultural research organization is discovering "truths" about agricultural development at about the same time as a major actor, and therefore making similar linguistic adjustments? Is it that the IDB is forcing its models and their concomitant language on developing countries like Panama, using their financial leverage? Are some actors cleverly reorganizing their development language in order to capture external funding? Are changes in development policy, in the latter case, mostly about reorganizing language? Let us turn now to examine the rhetorical strategies of various actors, including IDB, the Republic of Panama, the IDB President, IDIAP researchers, and farmers and other end users.

The IDB and Agricultural Services Modernization Program. To address these questions, let us consider the case of the Agricultural Services Modernization Program in Panama. The Modernization Program was a comprehensive project aimed at improving the profitability and productivity of Panama's agricultural sector. As an entry point to examine the rhetoric surrounding of this project, let us begin with the primary document of one of its subprograms, the "Agricultural Services Modernization Program: Subprogram of Technology Generation and Transfer" (BID 1995). First, a comment on how this project proposal was constructed is necessary. Government officials in charge of Panama's agricultural sector expressed an interest to IDB in a project to modernize the organizations that make up the institutional matrix of the sector. To formally express this interest the Panamanian team developed an initial proposal, which was submitted to IDB for consideration. The Panamanian team was led by the Minister of Agricultural Development, and otherwise comprised the Directress of Economic and Social Planning in the Ministry of Planning and Economic Policy (MIPPE), the Director General of IDIAP and a consultant from MIDA. In actuality, the collection of data and drafting of the document was most likely carried out by a team of planners and economists from MIDA, IDIAP and MIPPE. After submission of this initial proposal, the IDB expressed an interest and sent a team of technicians on a mission to Panama. It was this team, not the Panamanian officials, who drafted the full proposal that went to the IDB.

Individual to Corporate Authorship

From the standpoint of rhetorical analysis, one of the first questions we confront is authorship. Who is the author of "Agricultural Services Modernization Program: Subprogram of Technology Generation and Transfer?" Given the process of proposal development at IDB, it is reasonable to conclude that the document was produced by a team of professionals from MIDA, IDIAP and MIPPE. Yet, authorship on the final document is attributed to a corporate author, IDB, not the individuals who drafted it. This has an impact as a rhetorical device, since the reader is now faced not with fallible, subjective, individual authors, but rather the IDB. If the reader wants to challenge the document, then she needs to take into account not only the substance of the claims, but the status, values, goals and interests of the IDB, a corporate body.

The status of the author is not a trivial issue (Latour 1987). Consider the standing of a document of the same title authored by a small Panamanian NGO. The reader might ask whether a small NGO would have the necessary expertise to make an "authoritative" statement on such a topic. The political interests of the NGO would also become an issue. Likewise, consider the standing of a document with the same title authored by an individual, say Rafael Ortiz. The reader, perhaps even before engaging the text, would immediately want to know: Who is Ortiz? What is his institutional affiliation? Is he with the Bank? An academic? What are his qualifications? With what authority does he write about this topic? The reader takes into account not only the substance of the claims, but the reputation of the author and the author's institution. If the reader is able to answer the above questions, it conditions the manner in which she engages the document - with more skepticism or trust, for example. Without this knowledge, the document is more of a "black box" (Latour 1987), in the sense that the reader has less knowledge about the processes and conditions of production that led to the final document. The social relations of production are obscured. Thus, authorship matters, as the reader must take into account not only the substance of the claims being made, but also her views of the reputation, skill, objectivity, experience and goals of the author and the author's institution. Some of the information the reader learns may be contradictory. For example, the reader may know that the IDB has considerable economic expertise, but also that the IDB is a profitable bank with its own interests (Inter-American Development Bank 2000). Making the author a corporate one (i.e., removing information about individual authors, their institutional affiliations, their qualifications) can be understood as the first rhetorical tactic here, because it can make dissent on the part of the reader more difficult.

Enrolling Benefits to Avert Criticism

At the outset of the text we immediately encounter the next rhetorical tactic. The author says in the opening statement:

As a consequence of the macroeconomic policy and of the participation of Panama in GATT, its agriculture will confront greater competition due to the gradual lowering of tariff protection, the access competing countries have to improve new production technologies, transformation and marketing, and a growing environmental regulatory framework (BID, 1995: 1).

This is a strategy that the IDB uses successfully throughout the document. Panama's agriculture sector will face increasing competition, but it will not be because of the particular development model that the IDB is promoting; rather, it will be due to the macroeconomic policy path that Panamanian officials themselves have chosen. This strategy is more evident in the next paragraph:

The national authorities have proposed to carry out "The Program of Modernization of Agricultural Services," which includes the Subprogram A: Generation and Transfer of Technology. Said modernization program proposes to ". . . contribute to the creation of the conditions that will permit the Panamanian agricultural sector to strengthen its capacity to contribute to general economic growth, increasing production and productivity of the commodities and production systems of small and medium-size producers, and contributing to the preservation of renewable natural resources of the rural sector." The above will be achieved as the legal, institutional and technological restrictions that limit agricultural development are removed (BID 1995: 1).

This contributes to the legitimation of IDB's case because it says that Panamanian authorities have conducted their own analysis, identified their country's needs, drawn their own conclusions, and developed a program to address their needs. To great effect, IDB inserts in quotations Panama's own wording in the above passage to state the proposition of the program, then ends with an assertion in its own wording. Enrolling the Panamanian national authorities deflects critical attention away from IDB's own interests in promoting the project. Through this rhetorical strategy IDB enrolls Panama as an ally in the project - an ally that the Bank is simply helping to accomplish its own goals. It creates an alliance that strengthens the document. The reader who wishes to dissent will have to confront this IDB-Panama association. The argument that IDB is imposing its development model on a dependent southern nation is weakened when Panamanian officials are brought in to show that it is indeed they who are pursuing the project and financing. Politically, it also suggests a sensitivity on the part of the Bank to criticisms stemming from the debt crisis of the 1980s, in which the development banks were criticized for encouraging Latin American governments to finance growth by assuming increasing amounts of external debt.

The maneuver of enrolling the Panamanian authorities from the start averts that criticism.

Having brought the Panamanian government in to legitimate its involvement, IDB now needs a credible narrative of the agroecological and socioeconomic context. To build this credibility IDB increases the number of associations in the narrative, i.e., it calls on more actors to help support its case. In the process, IDB solidifies its associations with the government of Panama by enrolling MIDA, MIPPE and IDIAP, and it enlists a new actor in the cause - IICA (the Instituto Interamericano de Cooperación para la Agricultura, a leading policy research agency in the region). What actors other than the local ministries and research institutions could speak with greater authority on the local/regional agroecological and socioeconomic context? Enlisting these actors in an affirmative manner helps to strengthen and legitimate IDB's case, while it also turns the statements in the documents cited more into facts. For example, when IDB wants to show that poverty is primarily a rural problem, and then suggests the importance of agriculture in generating employment and producing export value (1995: 5), it cites an IICA study (1992) to strengthen this claim. Likewise, in its diagnostic of the national socioeconomic situation in Panama, IDB calls on MIPPE's (1994) documentation of poverty (50% of the population), extreme concentration of income, high levels of unemployment, low competitiveness, excessive protectionism and inadequate infrastructure (BID, 1995: 4). This diagnostic is immediately followed by prescriptive measures which are strategically attributed to the central government:

Based on this type of diagnostic of the present situation, the central government has proposed various strategies to induce change. First, the internationalization of the markets for goods and services as a measure to create new investment and employment opportunities. Second, generation and transfer of technology . . . to improve the competitiveness of the domestic producers (BID 1995: 4).

Once the context of the problem has been convincingly established, the narrative moves on to the next step in the logic of the argument - showing the current limitations and deficiencies in the sector. To identify the elements that limit the development and contribution of the agricultural sector to the economy, it is much more effective to draw on MIDA's own analysis of these factors. Thus, the IDB states:

According to the MIDA document, the potential for development and for the contribution of the sector to the national economy. . . is limited by: [among others] a) a lack of adaptation of the sectoral policy to the macroeconomic policy; b) institutional inefficiency; c) the lack of adequate technology; d) low competitiveness in production; e) high costs of manual labor and agricultural inputs; f) imperfections and high levels of market protection; g) low level of public and private investment in port infrastructure, commercialization, etc. (BID, 1995: 6, citing Ministerio de Desarrollo Agropecuario, 1994).

In this way, the reader traces the language of limitations and deficiencies to MIDA rather than IDB, which helps to avoid the impression that IDB is overstating the deficiencies of the sector in order to show a greater urgency for the financing. IDB's message in the subtext here is, "We are not imposing our view on Panamanian agricultural officials. They are telling us what their own limitations are."

Finally, how does IDB make the case that this particular project will effectively address the deficiencies identified in the diagnostic process? Its main strategy in the Subprogram document is to bring the propositions of the project closely in line with the objectives outlined by IDIAP in 1994. IDB states the specific objectives of the subprogram to strengthen IDIAP in the following terms:

Increase the supply of agrotechnological innovations, which have been biologically, economically, environmentally and socially validated . . . . Promote the participation of the private sector in the process of generation of technologies . . . . Transfer the resulting technologies to Extension so that they might be diffused en masse to the producers (BID, 1995: 75).

In its 1994 annual report, IDIAP (1995: 3) had established the following objectives (among others) for a modernized system of generation and transfer of agricultural technologies:

(a) Increase the supply of agrotechnological innovations . . . .; (b) Increase economic and productive efficiency . . .; c) Promote the adoption of innovation . . . .; (d) Ensure the participation of the sector (producers, providers, agribusiness, public and private entities), in the process of agrotechnology generation . . .; (e) Promote the agroin-dustrialization of the sector . . . .

The IDB statement of objectives parallels closely what IDIAP cites above as the orientation that the government has defined for its agricultural research policy (BID 1995: 9). Thus, one of the ways in which IDB attempts to make the case for the Modernization Program is to use the language already in circulation in official government documents.

Republic of Panama. This raises an additional point. What becomes apparent in the above analysis is that the IDB cites documents from the Panamanian government - where language in accord with IDB's perspective was already available - already circulating in agricultural sector networks. Similar language was used by both IDB and the Panamanian government organizations to state objectives, limitations, problems and prescriptions. As noted earlier, the major actors had begun implementing a neoliberal development model in the early to mid 1980s. Yet, it is the government's 1991 statement in the "Program of Development and Modernization of the Economy of Panama" that displays the full embrace of the neoliberal language. A brief excerpt is sufficient to witness the flavor of this document:

The government of Panama has made the decision to transform the productive structure of the country and modernize its economic system. Experience indicates that a self-sustaining process of economic and social development is based in various principles, salient among which is the creation of a market economy. . . Competition between firms and a free system of prices are the best mechanism for an optimal allocation of resources. . . . To achieve a full market economy, the following must be done: (i) Eliminate all interventions in the pricing system; (ii) Expand competition through the introduction of imports, with a reasonable tariff; (iii) Liberate the restrictions that limit access to markets; (iv) Create a flexible labor market; (iv) Reduce State intervention (MIPPE, 1991: 2-3).

An important issue here is the audience. The MIPPE document above is essentially an expanded letter to the President of the World Bank. The objective of the rhetoric is to normalize relations with the international finance community. As such, the document presents the strong case, i.e., the idealized version of neoliberal policy reform. Sweeping, unsubstantiated statements are made based on the presumed expectations of the audience:

The Private Sector is more efficient than the Public Sector in productive activities. This is explained by the lack of bureaucracy, the flexibility to act and a better mechanism for decision making and allocating resources. Therefore, the expansion of private activity increases the efficiency of the whole economic system (MIPPE, 1991: 3).

In many contexts, such assertions would need at least some substantiation. Yet, once it is understood that the primary intended audience for this document is the international finance community, and the World Bank in particular, it becomes obvious that no supporting evidence is necessary. The authors understand that they are in friendly territory, and that these will not be points of contestation. To identify the points of contestation - those issues that are most sensitive - one need only search in the document for the places where the layers of supporting evidence become thicker. For example, the MIPPE document is designed to convince the international finance community that Panama is serious about structural reform, that important measures have been taken, and that they are having the desired impact. Part of the document is a letter from the Second Vice President of Panama and the Minister of Finance and Treasury to Lewis Preston, then President of the World Bank. The letter summarizes the main points of the document. In the letter, the evidence becomes more layered around the most sensitive points - those of importance to the relationship between Panama and the international finance community:

We have achieved considerable progress in all three interrelated areas [reestablishing democratic institutions, resuscitating the economy and restructuring the economy]. These positive developments have restored Panama's position within the international community. We have complied with the [International Monetary] Fund-Monitored Program since September 1990. The public sector deficit was reduced from 11.5 percent of the GDP in 1989 to 2.9 percent in 1990. We have initiated the implementation of a sound public investment program addressing critical infrastructure needs. We lifted the deposit restrictions which had been introduced to avoid deposit withdrawals . . . . Deposits in the banking system have increased by US$3.5 billion in 1990. Strengthened public confidence contributed to real GDP growth of 3.4 percent in 1990. . . Since April 1990, we have made all payments on maturing debts owed to the IFIs [International Finance Institutions] . . . In November 1990, an agreement was reached with the Paris Club for the rescheduling of official debt service payments in arrears . . . (MIPPE, 1991: annex II)

The most sensitive issues are identifiable by the amount of effort the authors expend in supporting the claim. Compare, for example, the differential effort the authors invested in defending, "We have achieved considerable progress in all three interrelated areas [reestablishing democratic institutions, resuscitating the economy and restructuring the economy]," versus "The Private Sector is more efficient than the Public Sector in productive activities. . . ." Clearly, supporting evidence becomes thicker around the claims the author feels will be the most scrutinized by the reader, or perhaps most vulnerable to dissent. An additional tactic used above to make the claims more incontrovertible is to bring in allies for support - independent authorities that could verify their claims. Thus, the authors call on the IMF, international finance institutions and the Paris Club, in case Mr. Preston is not fully convinced.

Recirculating Language

The IDB President Speaks: Authorship and Authority. Let us now consider an example as a step towards understanding the circulation of language in policy networks. Returning to the Modernization Program example, imagine the scenario leading to the drafting of the proposal and its eventual fruition. It is the early 1990s. Officials of the Panamanian agricultural sector are contemplating a proposal to infuse significant investment into the sector. The two leading candidates for the source of such financing are the World Bank and the IDB. The challenge is to draft an initial proposal that will attract the attention of one of the major lenders. Strategically, the next step is to network with key actors/decision makers at the lending institutions in order to learn what kinds of projects are being sought, what is currently in favor and what is possible. Yet, to review precise proposal language, one would also need to go to the documents of the lenders to investigate the language being used, the model being promoted, and the vision the major lenders have of the kinds of projects they would like to see advanced. In short, investigate what the actors you need to convince are saying.

An example of a document that the Panamanian officials and technicians would want to consult is Reflections on Economic Development: Toward a New Latin American Consensus, published by the IDB and authored by then IDB President, Enrique Iglesias (Iglesias 1992). In this document Iglesias outlines a number of objectives to serve as an underpinning for the ". . . design of the region's economic and social policies for the 1990s" (Iglesias, 1992: 144). Appropriate to our analysis here is Iglesias's objective concerning technology, which he states as: "Promote the incorporation of the most advanced technology into the productive processes in order to enhance the regions' international competitiveness" (Iglesias 1992: 144). Here we see a familiar theme, in which technology's primary role is defined as enhancing competitiveness in an international economy. Later, the author expands this to say:

Our countries' potential for producing competitively and penetrating world markets depends on their capacity to keep up with international technological trends and incorporate the new knowledge into the production of tradable goods and services. This, in turn, depends not only on the existence of programs designed specifically to promote scientific and technological development, but also on the organization of enterprise, on its relationship with the various productive sectors, on financial and marketing services . . . (Iglesias, 1992: 151)

As we have seen repeatedly in documents from IDIAP, MIDA and MIPPE, Panamanian officials have tailored the role of science and technology in agriculture precisely along these lines. Note also that this language was already circulating in development networks at about the time that Panamanian officials were beginning to develop the proposal for the Modernization Program to IDB. One plausible explanation of how this circulation of language takes place is that local authorities draw on statements such as this one by Iglesias, they appropriate the language, "translate" it to accommodate their circumstances, then recirculate the language in project proposals. A more cynical interpretation is that local officials appropriate the language of the major actors in order to feed back to them what they want to hear in order to increase the likelihood of a successful proposal. Yet, this is an overly simplistic reading. The language of the neoliberal model is simultaneously circulating in networks that extend globally, and so there are many different possible sources. Government officials, planners and researchers are connected into multiple overlapping networks where various versions of this language are circulating. Finally, it is reasonable to view Iglesias as a spokesperson for the Bank, and to assume that his views are likely similar to those of the top decision makers in the IDB. His words will be taken by many as authoritative, and they are made possible in part by his position of authority. They will be influential in setting the development agenda in Latin America.

IDIAP Researchers. Let us now move to a different point in the network to consider how the language of the neoliberal model is interpreted, translated and recirculated by IDIAP researchers. One of the key documents written by IDIAP researchers is the project document. Particularly in terms of rhetoric, the project document is key because it is here that the researcher frames the project in terms of the larger objectives of the institution, provides justification for expending resources on the project, states specific project objectives, identifies specific beneficiaries, and states the expected results. The project document is a quintessential exercise in rhetorical tactics.

In 1995, IDIAP went through a process of rethinking and developing all new research project proposals. The aim of these efforts was to recast research projects from a more holistic, integrated management perspective, and to bring all research projects in line with sector policy. For example, in a research project on watermelon for export, the project opens with the following justification:

The national plan for modernizing the Panamanian economy considers . . . entering into external markets, through the processes of agricultural diversification, with emphasis on production for export. In this sense, the agrotechnological innovations that are promoted must respond to the necessities of external markets, without ignoring the sustainability of production systems (IDIAP, 1995d: 2).

The proposal goes on to argue that watermelon is one of the commodities with high potential for export, due to the opening of markets in the US. This opening of markets, the author argues, has caused the acreage devoted to watermelon to nearly double, and the production to more than double. In short, the specific research project in watermelon is framed within IDIAP's objectives, the objectives of the sector, and in turn, those of the national economic modernization plan. The logic of the case being made here is from broad (Panamanian economy) to narrow (watermelon's contribution). The translation of the language in this case seems fairly straightforward. Cleverly, the researcher establishes the importance of IDIAP in removing the barriers to increasing exports in this subsector, while simultaneously ensuring the role of research by identifying "lack of information . . . about the technologies for integrated management of watermelon . . ." as a primary problem in the subsector (IDIAP 1995d: 11).

A project in pineapple research adopts a similar strategy, in which the author strategically locates research as a critical link in getting the commodity chain to function effectively:

[...] the prices obtained in recent years by countries that sell fresh pineapple to the US have been good and have been improving in recent months. This offers an alternative to our producer[s] and stimulates interest in export activities. Yet, the necessary adjustments must be made in order to enter into said market, which includes technological changes that must emerge from research and be carried immediately to the farmer [...]. (IDIAP, 1995c)

Again, the author is making the case for the project by inserting research as a critical link in this subsector - critical if the national goals (e.g., competing in international markets) are to be met. A project on highland onions takes no chances in locating its project within the relevant framework adopted by the sector. In stating the project objectives, the author combines (nearly verbatim) the 2001 mission statement (see Table 3 above) with IDIAP's 1995 statement of what characteristics their technologies should have (also cited in (BID, 1995: 9)), thereby linking together a number of documents to derive:

The project will permit the strengthening of the national technological base to contribute to food security, competitiveness and the sustai-nability of agribusiness, in benefit of the Panamanian society through technologies that are technically solid, economically feasible, socially desirable and environmentally safe and stable (IDIAP, 2001c).

Farmers and End Users. Finally, farmers and other end users also take up, interpret, translate and recirculate the language of the new model. For example, a member of the Board of Directors of APACH (Association of Rice Producers of Chiriquí) argued that the private sector should have a closer relationship with IDIAP. APACH would be willing, according to this board member, to provide support in terms of inputs and plots, as long as IDIAP "carried out research in what is needed." (#26: 1) He portrayed APACH and IDIAP as close allies. He related an instance in which there was discussion that the central government planned to privatize IDIAP. Representatives from APACH met with administrators of MIDA to express their objection:

For us IDIAP is exceedingly important. If you want to reduce the bureaucracy, eliminate the [Agricultural Development Bank]. We can find financing elsewhere, but we cannot move forward without the technology.

Similarly, a highland potato farmer, who produced for the markets in David and Panama City, portrayed his interactions with IDIAP as closely collaborative:

I work in a very tight relationship with IDIAP. We compare plots; I am always trying new products that come out of IDIAP and the commercial houses. We compare and discuss what is going well, what is not . . . . IDIAP has had a tremendous impact here with technology. The one thing I will say is that IDIAP needs to be more integrated with the producer.

This is the view of a farmer who was well educated, comfortable discussing the latest technologies, and so comfortable in the IDIAP research station where the interview took place that he sat down at the director's desk to make a phone call to his workers in order to plan the work day.

What has been shown above is a circulation process, in which development language is appropriated by actors in policy networks and refashioned to help each actor "make the case" in order to advance their interests. For example, the following circulation of language is one plausible scenario: (1) the IDB president publishes his views on the development agenda for the next decade, (2) Panamanian officials pick it up as leading edge development language, knowing they will need to cast their plans in this language in order to convince the major actors, (3) IDIAP leaders cast the goals, objectives and activities of the institution in similar language, (4) IDIAP researchers pick up the signals and represent their research projects in the image of the emerging model in order to show their relevance. Yet, this circulation of language does not necessarily occur in a linear fashion; rather, it might be more aptly described as a recirculation or translation process, as suggested in Figure 2 below.

Latour (1993) uses the notion of translation as synonymous with network. Thus, translations link in a continuous chain the power relations, strategies and language. Likewise, networks are not "things"; they are relationships or sets of practices. The translation relationships depicted above suggest associations in which actors appropriate language, adapt it to their circumstances, then recirculate it in such a way as to situate their own actions strategically, oftentimes making their own project or institution a critical link, or what Latour refers to as an "obligatory passage point." This is not to suggest that (re)presentation of one's activities is disingenuous, but rather to say that it is strategic. Casting one's actions in various lights depending on the circumstances is a common, often necessary, strategy.

We have also seen rhetorical tactics used to create the appearance of objectivity in the formation of policy in order to legitimate the process. For example, IDB was careful to create associations with local actors in order to legitimate its involvement, appear as an objective outsider, and avert criticisms that they were acting in self interest. Rhetorical strategies may also be used to rationalize a particular policy. For example, actors who were speaking for the state argued for their policy based on the need for the agriculture sector to be competitive. By claiming national interest as the objective of policy, the state attempts to situate itself as beyond the specific interests of any group or class that may benefit from the policy.

Conclusions

In a seminal study of elite scientists Latour and Woolgar (1979) showed that rather than discovering facts about nature, rather than revealing nature as it "really" is, scientists are engaged in a game of literary tactics which are designed to convince readers that their version of reality is the correct one. They followed the production of facts from the hypothesis, to the production of images through various inscription devices, eventually to the final product-the scientific paper. Similar literary tactics are at work here. In science and technology studies, discourse analysis has been underway for perhaps three decades, with some important results (see Ashmore et al. 1995 for a review). A strength of the Latourian approach to rhetoric is that it attempts to show the practical answers: What are the discursive practices that authors (scientists, policy makers, researchers) actually use? What strategies do they use to convince others? It keeps the analysis grounded in action, and can tell us about discursive practices. Yet, a weakness of the approach is that while we learn about rhetorical tactics, the connection from these practices to how material and social relationships are transformed is weaker. In other words, discourse analysis by itself is not sufficient.

Moreover, the documents analyzed in this paper are, to some extent, black boxes. They are for the most part finished products.

To accomplish a more thorough ethnography of policy making, one would need access to the discussions, arguments, debates and battles that took place during the production of the documents. What should be included in them, what should be excluded, and what language should be used? With this kind of data, one could show the social processes of policy formation, and more clearly reveal the interests embedded in them.

¹ While there is an enormous literature on "development," this paper is concerned with the multinational efforts that continueglobally under the rubric of international development.

² Entries for "Year" separated by a comma indicate that the two years have identical categories in the table of contents. Entries for "Year" separated by a hyphen indicate that all the years in the range have identical categories.

References

Ashmore, M., Myers, G. & Potter, J. (1995). Discourse, rhetoric, reflexivi-ty: seven days in the library. In: Jasanoff, S., Markle, G., Petersen, J., & Pinch, T. (Eds.). Handbook of science and technology studies, p. 321 - 342. Thousand Oaks, CA: Sage.         [ Links ]

BID (Banco Interamericano de Desarrollo). (1995). Programa de Modernización del Sector Agropecuario: Subprograma de Generación y Transferencia de Tecnología. Panamá: Banco Interamericano de Desarrollo.         [ Links ]

IDIAP. (1979). Informe Anual 1979. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1987). Memoria Anual 1983-1984. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP.(1989). Memoria Anual 1989. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1991). Memorial Anual 1990. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1993). Memorial Anual 1992. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1995). Memoria '94. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1995c). Proyecto de Investigación en Manejo Integral del Cultivo de Piña en el Sistema de Producción para la Exportación. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1995d). Proyecto de Investigación en Manejo Integral del Cultivo de Sandía en el Sistema de Producción para la Exportación. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1996). Memoria 1995."Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (1997). Memoria 1996. Panamá: Instituto de Investigación Agropecuaria de Panamá         [ Links ].

IDIAP. (2001c). Proyecto de Investigación y Transferencia en Manejo Integral del Sistema de Producción de Cebolla de Tierras Altas de la Provincia de Chiriquí. Panamá: Instituto de Investigación Agropecuaria de Panamá. <http://www.mida.gob.pa/idiap/cebo-llas.html> (Abril 15, 2002).         [ Links ]

IDIAP. (2002). IDIAP: Quienes Somos: Panamá. Instituto de Investigación Agropecuaria de Panamá. <http://www.mida.gob.pa/idiap/quie-nes.html> (Abril 6, 2002).         [ Links ]

Iglesias, E. V. (1992). Reflections on Economic Development: Toward a New Latin American Consensus. Washington, D.C.: Inter-American Development Bank.         [ Links ]

IICA. (1992). Panamá: Sector Agropecuario, diagnóstico y propuestas. Panamá: Instituto Interamericano de Cooperación para la Agricultura.         [ Links ]

Inter-American Development Bank. (1996). Panama: AgriculturalSer-vicesModernizationProgram. Report, 924/OC-PN. Washington, D.C.: Inter-American Development Bank.         [ Links ]

Inter-American Development Bank. (2000). Annual Report 2000. Washington, D.C.: Inter-American Development Bank.         [ Links ]

Latour, B. (1987). Science in Action: How to follow scientists and engineers through society. Cambridge, MA: Harvard University Press.         [ Links ]

Latour, B. (1993). Wehave never been modern. Cambridge, MA: Harvard University Press.         [ Links ]

Latour, B. & Woolgar, S. (1979). Laboratory Life: The construction of scientific facts. Princeton: Princeton UniversityPress.         [ Links ]

MIDA. (1992). Propuesta para desarrollar el Plan Nacional de Promoción de las Exportaciones de Panamá. Panamá: Ministerio de Desarrollo Agropecuario.         [ Links ]

Ministerio de Desarrollo Agropecuario. (1994). Marco Orientador de la Política Agropecuaria, 1994-1999. In IDIAP (Ed.), Marco Orientador de la Generación y Transferencia de Agrotecnologia. Panamá: IDIAP.         [ Links ]

MIPPE. (1991). Programa de desarrollo y modernización de la Economía. Panamá: Ministerio de Planificación y Política Económica, República de Panamá         [ Links ].

MIPPE.(1994). Políticas Públicas para el Desarrollo Integral: Desarrollo Social con Eficiencia Económica. Panamá: Ministerio de Planificación y Política Económica, República de Panamá         [ Links ].