It is well known that Brazilian research has been getting more attention in the area of world science – the country’s contribution to international scientific production rose from 1.6% of all articles published in 2002 to 1.9% in 2006, according to data from the Thomson Reuters database. However, the weight of the various fields of knowledge in the Brazilian performance is uneven. A study released in the July issue of the journal Scientometrics produced an unprecedented survey about the performance of each set of disciplines in the sum total of the Brazilian articles published in international periodicals. The work shows that the fields of exact sciences and of Earth sciences account for almost 70% of the researchers that published more than half of their articles in international periodicals, meaning articles published in languages other than Portuguese. However, in fields such as linguistics and the arts, more than 80% of the researchers only publish in national journals (see exhibit). The study was signed by Paula Leite, who has a PhD from the Federal University of Rio de Janeiro (UFRJ), and by her two advisors, Jacqueline Leta, a professor at UFRJ, and Rogério Mugnaini, from the University of São Paulo (USP).
One of the highlights of the article concerns the unprecedented nature of the source: it was based on an analysis of a database with more than 51 thousand researchers from the Lattes Platform of curricula, run by the CNPq (National Council for Scientific and Technological Development). “The Lattes Curriculum [platform] is a unique example worldwide. It covers the entire population of Brazilian researchers and is updated often,” says Jacqueline Leta. The group had access to a wide range of data covering thousands of Brazilian researchers and their scientific production from 1997 to 2004. The information also included the name of each researcher, plus gender, field of expertise, year of graduation, institution, whether or not this person has headed any research, has a CNPq productivity grant, and his or her proficiency in foreign languages. The scientific productivity, measured by the number of articles, was divided into two categories: national publications, in Portuguese, and international publications, in other languages, in particular in English. Across the entire sample, 7,076 researchers have published no articles in eight years.
To estimate the weight of the international publications in each sector of the Brazilian scientific community, Rogério Mugnaini, a statistician and professor of quantitative methods in public policy at the USP Leste campus, proposed using the International Publication Ratio (IPR), determined by the ratio between the number of international publications and the total number of publications. As performance was shown to be quite heterogeneous, the researchers chose to discard those researchers who had produced fewer than three articles during the researched period, which caused the number of researchers to drop to 34,390 names. This universe was then broken down into five groups. The researchers in IPR 1 were the “highly international” ones, with 80.1% to 100% of their articles published in foreign language journals. Those in IPR 2 were rated as “Mainly international,” with 60.1% to 80% of international publications. Then came the intermediate group (IPR 3), with 40.1% to 60% of international publications, the “Mainly national” group (IPR 4), with 20.1% to 40% of international publications, and the “highly national” group (IPR 5), with no more than 20% of international publications.
The analysis of the performance based on areas of expertise shows that the exact sciences and the Earth sciences, which bring together disciplines such as physics and chemistry, lead the most internationalized group. Half of the researchers in such fields fall under the “highly international” category and, together with the “Mainly international” group, they form a cluster of almost 70% of the total number of researchers in such areas. Fewer than 20% fit the “highly national” profile. Then come biology and engineering, with about 30% of IPR 1 researchers, and the health sciences, with about 10%. The other areas have fewer than 10% IPR 1 researchers, in the following order: agricultural sciences, humanities, sociology, linguistics and art. In the case of agricultural sciences, there is a contingent of at least 10% of the researchers that fits the category “Mainly national” and “Intermediary.” “It’s important to note that agriculture is an important field in Brazilian science. It has a specific pattern of publications that reflects a strong regional research focus,” wrote the authors. As for the last three categories, which comprise the human sciences, 80% of their researchers are “highly national.”
The research data provide strong evidence of the different publishing and citation cultures in various fields of knowledge. Whereas the publication of articles in international journals, those with higher visibility and renown, is an unavoidable requirement for researchers in fields such as physics, astronomy and medicine, their colleagues in the agricultural sciences usually divulge most of their production in national periodicals. This is the case in the human sciences as well, which also favor books. These differences are not limited to publication type or origin. Rogério Mugnaini, who previously worked for the SciELO electronic library, is the author of another article, published in December of 2010, showing the differences of behavior in citations made in the articles. He analyzed the pattern of citations in Brazilian journals on the SciELO base and found that books are considerably more cited in a journal of applied social sciences, whereas the collective health area makes use of books in a proportion comparable to its use of scientific articles. In physics and medical journals, the citations in international journals are much more prevalent. And in veterinary medicine and science journals, theses and periodicals are more common. “One question that guided this analysis was the strong dependence on indicators of impact, that are calculated based strictly on the citations that articles get,” states Mugnaini. “On analysing the databases of journals, one finds that other types of documents are relevant for the literature published in different types of journals,” he says.
One of the discussions that takes place among authors concerns the appropriateness of the strategy of forcing researchers to publish their findings preferentially in international, high-impact journals. “Each field has its own dynamics. Even renowned Brazilian sociology or human sciences researchers publish most of their articles in Brazil and in Portuguese,” says Jacqueline Leta. “Is it appropriate to force researchers in certain fields to do something that they don’t want to do or are not qualified to do? This is the question that must be discussed by the scientific community and by those responsible for determining policies,” states the professor. Rogério Mugnaini notes that each area has been negotiating its own criteria – and eventually drawing inspiration from the example of others – to underpin its internationalization strategy. He mentions the different rules that the area coordinating offices of CAPES adopted to avoid trauma and distortions in the Qualis system. This is a tool used to classify the periodicals in which the postgraduate programs publish their scientific production and it values the journals with a greater impact factor, which applies to many of the international publications. “In the field of probability and statistics, a decision was made to take into account not only the impact factor, but also the journal’s half-life, in order to avoid equating consolidated journals with those that, in their quest for impact, favor only research that is of immediate interest,” states Mugnaini.
Other fields, such as computer science, associate the impact factor with Google Scholar citations, an indication of the application of knowledge. “There are journals in the medical and biological fields that are highly valued for their importance, regardless of their impact factor, whereas in the social sciences the only thing that counts is whether the journal is indexed, because it doesn’t make sense to think in terms of the impact factor,” states the researcher. These negotiations, however, are not an obstacle for the visibility of scientific production abroad, notes Mugnaini. “Every three years, Capes reevaluates the postgraduate courses and raises the bar a little,” he observes. The Scientometrics article shows that no matter how hard it may be to develop policies that involve all the fields of knowledge, the internationalization path is having an impact. Despite each field’s publishing culture, all of them are showing growth in the ratio of international publications relative to the 2001- 2004 and 1997-2000 periods.Republish