Publication patterns of the Agricultural Sciences in Brazil: influenced by research evaluation and financing systems?
Authors:
MSc. Alejandro Caballero Rivero; Federal University of Pernambuco (UFPE), Brazil. Email: caballero.alecaba@gmail.com
PhD. Raimundo Nonato Macedo dos Santos; Federal University of Pernambuco (UFPE), Brazil. Email: rnmacedo@uol.com.br
1 INTRODUTION
Studies on the sociology of scientific knowledge show the existence of epistemic cultures within academic communities. For Knorr-Cetina (2005, p.68), they are "machineries of knowledge construction", a combination of cognitive, rational or technical elements (e.g. nature of studied phenomena, methodologies, models) and social ones (e.g., evaluation systems, selection of communication channels).
Several authors (TRZESNIAK, 2014; ALEXANDER, 1999) show consensus on the existence of two major epistemic cultures: "hard" sciences (Engineering, Exact and Natural sciences) and "soft" sciences (Human, Social and Arts). “Hard” sciences deal with physical, more universal and predominantly deterministic phenomena, which can be observed and verified by experiments; thus, it is easier to reach consensus on theoretical paradigms. In these sciences, research is more quantitative, and scholarly communication is carried out in a highly coded language, enabling faster publication. Articles in scientific journals are more appropriate: short, standardized, synthetic publications, enabling faster writing, publishing and reading; journals periodic diffusion provide researchers a fast updating on the state of the art and their broad dissemination reaches a more international audience.
“Soft” sciences deal with mental states or conditions for them, mainly stochastic phenomena, which are quite dependent on the socio-cultural contexts; therefore, it is more difficult to reach consensus and several theoretical paradigms coexist. Qualitative research methods are broadly employed, and researchers use a less codified and uniformed language, which demands more elaboration and argumentation, slowing down scholarly communication. In this case, monographs are more appropriate; their writing, editing and reading takes more time and effort and permit to express better the sort of complex knowledge produced by these sciences.
The selection of the communication channel also depends on the social components of the epistemic cultures, particularly, on the recognition of peers and the external pressures associated with evaluations. Research evaluation systems typically link researcher’s reputation and funding to their quantity of publications. Additionally, emphasize citation indicators for evaluation purposes, particularly the Impact Factor (IF) of the Journal Citation Report (JCR). This way, changes in researcher’s publication patterns are promoted, especially, a progressive concentration in journal articles, even in the "soft" sciences (WARE; MABE, 2015; ADAMS; GURNEY, 2014).
In a previous study (CABALLERO RIVERO; SANTOS; TRZESNIAK, 2017), we identified the publication patterns of the CNPq’s eight broad disciplinary knowledge areas (Exact and Natural sciences; Agrarian sciences; Biological sciences; Health Sciences; Engineering and Computing science; Linguistics, Letters and Arts; Human Sciences; Applied Social Sciences) during the 2000-2014 period, and grouped them into two hyper-areas ("hard" and "soft") according to their similarities. “Soft” sciences showed a balance between monographs, full papers published in proceedings and articles, predominantly, those published in national journals. In the "hard" sciences, articles published in international journals significantly prevailed over other channels.
Nonetheless, a peculiarity was observed in the Agrarian Sciences (AS). During the 2000-2006 period, articles published in national journals and full papers in proceedings accounted for 75%-80% of the total volume of publications. In its turn, during the 2008-2014 period, articles published in international journals matched those published in national ones, while the contribution of full papers in proceedings underwent a significant decrease.
This paper deepens analyzes carried out in that previous study in order to better understand this shift in the publication patterns of the AS in Brazil. Discussions are based on the graphical representation of the historical series of the Brazilian scientific production during the 2000-2016 period for identifying and comparing publication patterns. A documentary analysis of the evaluation criteria used by the main funding bodies in Brazil (Capes and CNPq) is also carried out in order to identify their possible influence on publication patterns of the AS.
2 METHODS
This is a quantitative-qualitative, non-experimental and longitudinal study. Data of the Brazilian scientific production was collected manually (on May 2019) from the eight censuses developed by CNPq during the 2000-2016 period and available at the Directory of Research Groups; specifically:
(a) quantity of national (An) and international articles (Ai); full papers in proceedings (T); book chapters (Mc); books (Ml);
(b) quantity of Articles (A) (national + international); monographs (M) (books + chapters); total production (P);
c) percentage contribution of national (an) and international articles (ai); full papers papers in proceedings (t); book chapters (mc); and books (ml);
d) Ratios: articles / monographies (RA/M); national / international articles (RAn/Ai).
Data was transferred into eight Excel spreadsheets (one for each census). Variables mentioned in the (b), (c) and (d) items were calculated from that data. The historical series of all variables were graphically represented to identify and compare publication patterns. This paper highlights, specifically, some series for the AS, “soft” sciences and “hard” sciences:
• articles (national + international); monographs (books + chapters) and full papers in proceedings;
• percentage contributions of articles (national + international), monographs (books + chapters) and full papers in proceedings.
The identified publication patterns were analyzed considering the evaluation criteria of the Capes and CNPq for the AS. Documents specifying the evaluation criteria for post-graduation programs (Capes), as well as those used to evaluate researchers that have received research grants (CNPq), were collected (on January 2019). Evaluation criteria were identified and classified according to: funding body (Capes or CNPq); evaluation period; evaluation area; communication channels considered in the evaluations and their specific requirements.
3 RESULTS AND DISCUSSION
Similar publication patterns of the AS and the “hard” sciences over the analyzed period were identified (Fig. 1, 2, 3, 4, 5). In both cases articles are the main communication channel; its percentage contribution in the AS increased from 54% in 2000 to 76% in the 2016 and in the "hard” sciences from 52% to 70%. Likewise, the contribution of full papers in proceedings decreased significantly; in the AS from 37% (2000) to 14% (2016) and in the “hard” sciences from 39% to 20%.
Monographs contribution was similar, ranging from 9%-11% in both cases. When analyzing the ratio RA/M, an increase was also observed in both cases: in the AS from 5,67 (2000) to 8,03 (2016) (mean = 6,23) and in the “hard” sciences from 5,96 to 7,21 (mean = 5,88).





However, the analysis of the RAn/Ai (Fig. 6) show that the publication patterns of the AS did not always match those of the "hard" sciences. During the 2000-2006 period, RAn/Ai ranged 2,40-3,63 (mean = 2,88) in the AS. These values neither correspond those of the “hard” sciences (mean = 0,67), nor of the "soft" sciences (mean = 6.09), but indicate a national orientation, characteristic of the "soft" sciences. This situation could be explained by the peculiarity and diversification of Brazilians soils, which promotes important research efforts, but mainly of national interest (TRAJANO et. al., 2013).
Yet, this argument is equally valid for the 2008-2016 period, when RAn/Ai in the AS changed from 2,11 to 0,83 (mean = 0,84), indicating a predominance of international publications and clear approximation towards “hard" sciences’ patterns (mean = 0,55).

Obviously, causes of this change are multifactorial, and this paper does not intend to present a definitive answer. However, the analysis of the evaluation criteria used by Capes and CNPq indicates they might be influencing this shift. When evaluating the intellectual production of the post-graduation programs, the four Capes Committees of the AS punctuate articles, while only one (Food Science) considers monographies and only two (Food Science and Zootechnics / Fisheries Resources) include full papers in proceedings. Additionally, all Committees consider the IF (JCR) as an essential criterion for stratifying scientific production in the Qualis system. On the other hand, all seven CNPq Committees of the AS demand publications in journals with IF for researchers with research grants; monographies are considered only for defining tie-break among researchers in one of the Committee (Aquiculture and Fisheries Resources). No Committee demands full paper in proceedings.
4 FINAL CONSIDERATIONS
Publication patterns of the AS show a continuous increasing of the percentage contribution of articles published in international journals and a decrease of those published in national journals, as well as of that of monographs and full papers in proceedings. This situation seems to be influenced by the evaluation criteria used by CAPES and CNPq, which promote, almost exclusively, publication of articles in journals with IF. Evidently, this strategy is directed towards raising the quality and visibility of the scientific output of the AS. However, scientific research does not only have an intellectual impact, but also a social one. Favoring scientific production in high-FI journals, could certainly increase the intellectual impact of the AS, but it could also negatively impact research focused on national problems.
References
ADAMS, J.; GURNEY, K. Evidence for excellence: has the signal overtaken the substance? London: Digital Science, 2014.
ALEXANDER, J. C. A importância dos clássicos. In Anthony Giddens e Jonathan Turner (Org.), Teoria social hoje. São Paulo: Editora UNESP, 1999. p. 23-89.
CABALLERO RIVERO, A.; SANTOS, R. N. M.; TRZESNIAK, P. Caracterización de las prácticas de publicación de las grandes áreas de conocimiento em Brasil. Revista Cubana de Información en Ciencias de la Salud, v. 28, n. 4, 2017.
KNORR-CETINA, K. Culture in Global Knowledge Societies: Knowledge Cultures and Epistemic Culture. In: JACOBS, M. D.; HANRAHAN, N. W. (Eds). The Blackwell Companion to the Sociology of Culture. Blackwell Publishing Ltd: Malden, MA, 2005. p. 65-79
TRAJANO, M. A. B.; RAZUCK, F. B.; CERETTA, C. A.; SCHETINGER, M.R. C. Evolução da produção
científica em Ciência do Solo no Brasil: um olhar sobre o Qualis. Geografia (Londrina),
2013
TRZESNIAK, P. Indicadores quantitativos: como obter, avaliar, criticar e aperfeiçoar. Navus – Revista de Gestão e Tecnologia, v. 4, n. 2, 2014. p. 5-18.
WARE, M.; MABE, M. The STM Report: An overview of scientific and scholarly journal publishing. 4th Edition. The Hague: International Association of Scientific, Technical and Medical Publishers, 2015.