Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
National scientific communities and styles of science
Before illustrating the different styles of science by referring to studies on specific countries, it may be useful to look at the way national scientific communities have emerged or are still struggling to emerge in smaller, lesser-known countries that are representative of many developing countries.
(a) Costa Rica, Senegal, and Thailand
These three countries are representative of a number of small (Costa Rica and Senegal) or medium-size (Thailand) developing countries. All three are agricultural, although the economies of Costa Rica and Thailand have been changing structurally as industry accounts for an increasing part of GNP. They are characteristic examples of young scientific communities. The first traces of research institutes and schools of higher learning started appearing at the end of the nineteenth century and became really visible during the first part of the twentieth century. And it was not until the 1960s and even more the 1970s that national scientific communities started taking root and becoming bona fide institutions.
As universities grew, research activities went through a process of institutionalization, and science policy-making bodies were created. This process started in Thailand at the end of the 1950s, in Senegal in the 1960s, and in Costa Rica at the beginning of the 1970s. Their scientific communities are small and difficult to evaluate. The Senegalese and the Costa Rican scientific communities probably numbered 800-1,000 scientists and the Thai community just over 5,000 in the mid-1980s [33]. The scientific potential is concentrated in universities and focused on agriculture, social sciences, and health. University scientists are the best trained and have the highest qualifications. The number of scientists in engineering is low in all three countries. Unlike the other two, the large number of scientists in Senegal working on what could be called "general advancement of knowledge" is rather unusual and can only be explained by historical reasons going back to the colonial period: the Senegalese scientific community is still heavily dependent on expatriate scientists 30 years after independence [33, 35].
Of the three, Costa Rica offers its scientists, especially those working at the University of Costa Rica (UCR), the best salaries and the most incentives to be involved in scientific activities and to publish their findings. Proposals have recently been made through the UCR internal promotion system to encourage academic staff members to do more research. Obviously there is still room for improvement, and different institutes offer different opportunities. The major devaluation of the colon early in the 1980s seriously affected salaries and lowered their purchasing power. But this situation was not unique to people working in science.
In Thailand despite a political decision to support scientific work and an atmosphere propitious to its research development, there are inherent difficulties in the profession that are far from being solved. In the public sector, Thai research scientists and university teachers are badly underpaid. Promotions and salary increases depend almost exclusively on seniority, with very little credit for educational level, work performance, and services rendered, be it in research, education, or administration. This situation is of course deleterious to normal research schedules, since far too many research scientists and teachers doing research have to supplement their salaries by accepting unrelated jobs on the side.
The status of the research scientist in Senegal is precarious. Since there is no career stability, a future in this profession is fraught with uncertainty. Institute regulations make no allowances for the unique characteristics of the profession. Career paths and positions are very different from each other. The scientists who are classed as civil servants are probably the worst off, since the professional scale in the government civil service does not provide for a Ph.D. level. This means that scientists with a doctorate, a master's, or a bachelor's degree are all in the same category and move up the seniority-driven scale at the same pace. Attempts to provide the research scientists with common professional statutes were nearing completion when the Ministry of Research was dissolved in 1986. In view of the present institutional structures, the economic crisis, and the budgetary restrictions in Senegal, there is little chance that research will be given statutes of its own in the near future.
(b) Different styles of science
How can the profession of scientist distinguish itself in society and how do scientific communities gain legitimacy? The history of science in a number of developing countries shows that the professionalization of third world scientists, the emergence of national scientific communities, and the legitimation of scientific activities are often associated with the creation of an active professional association. This is the case in India, where the launching of the Indian Science Congress Association (ISCA) "afforded a widely scattered scientific community a much needed common meeting ground" [51]. In Brazil, the Brazilian Society for the Progress of Science (SBPC) played a crucial role in the professionalization of Brazilian scientists between 1950 and 1960. In the case of the SBPC, "scientific interests and political legitimacy were closely imbricated" [11]. Modelled on the American Association for the Advancement of Science, the Venezuelan Association for the Advancement of Science (ASOVAC), created in 1950, has brought a stronger cohesion to the young, emerging Venezuelan scientific community while legitimating its role [82]. It is also responsible for a distinct, rather academic, style of science.
The charismatic role of leading political and/or scientific figures must also be taken into consideration. Thus, in India, the important role of Nehru - "a passionate believer in modern science" and "the main architect of India's science policy" - is widely recognized (e.g. [52]). Other, lesser known examples are Clodomiro Picado in Costa Rica, Cheikh Anta Diop in Senegal, and King Mongkut, considered as the father of science in Thailand. All of them are venerated by their national scientific communities; universities, research centres, and scientific awards have been named after them [33]. Conversely, dictatorships and anti-science political systems always have devastating effects on the emergence and growth of scientific communities everywhere. One of the most extreme examples is the Chinese Cultural Revolution (1966-1976), during which civilian scientific research came to a halt. The situation hopefully changed radically in 1978, when the Party declared science and technology to be one of the four modernizations, but much time had been lost while Chinese researchers were sent to the countryside and isolated from the rest of the world.
The (social) origin of the scientists may also be linked to the emergence of different styles of science. In Africa, for instance, the social origin of the scientist is less distinctive than in the other continents, although many come from rural areas. In Latin America, the scientific communities mainly recruit among the middle classes and in particular among immigrants.
By contrast, in many Asian countries, scientists are often drawn from groups traditionally associated with learning and/or power. In Thailand, as already mentioned, the royal family played, and is still playing, a very important role in the birth, growth, and dissemination of science. Princess Maha Chakri Sirindhorn, for instance, who recently received her Ph.D. from Srinakharinwirot University, is considered a symbol and often participates in public scientific events [35]. In India, ever since the British brought Western science to Bengal in the nineteenth century, the scientific community seems to have been dominated by the upper Hindu castes, especially the Brahmans. Kapil Raj describes how the Brahmans in their own way "appropriated occidental ideas and science to give credence to their new dominant - status in the Indian society" [67]. This, according to Raj, explains why Indian science tends to pay more attention to basic disciplines and to be what he calls a "clean" science. Krishna [51, 52] also describes how an Indian national science developed in the 1920s in opposition to the British colonial science around a few scientific leaders, mainly in "pure" sciences, such as physics, chemistry, and mathematics. This may also partly explain why the Indian scientific community is clearly influenced and attracted by the international scientific community and why Indian scientists tend to publish in mainstream journals.
Conversely, Brazil, the second largest scientific community of the third world after India, is characterized more by an "inward-looking" or "inbred" research approach, and the Brazilian scientists tend to publish, particularly within agricultural sciences, in Portuguese and in local journals [78, 19]. The development of the Brazilian scientific community in the field of informatics illustrates how young, jobless Brazilian university graduates succeeded in convincing national banks and the government to launch a national plan to develop a Brazilian computer industry [10]. The future of the Brazilian scientific community is, however, far from secure. "There is certainly more science and technology in Brazil today than only twenty years ago; and yet, it is clear that a space for science, in terms of socially defined, accepted and institutionalized scientific roles, is barely there. What we have at most are islands of competence, niches where science was able to develop for some time, but always precariously, and threatened by an unfriendly environment" [74].
The example of Singapore, where R&D activities are now developing under the influence of a strongly technocratic political approach, is attractive to many developing countries because of its impressive results, but it leaves us with a number of unanswered questions. In a recent article, Goudineau [41] shows that Singapore reached a rather advanced stage of socioeconomic development before thinking of developing a national science and technology policy. This occurred in a context where a technological potential existed but where local scientific elites had not been trained nor had a true national scientific community emerged. This is compensated for by a massive importation of knowledge, experts, and know-how in a limited and carefully selected number of areas (informatics, biotechnology, and new materials). A more and more vigorous attempt to encourage Singaporean scientists to return from abroad is also taking place.
The origins, behaviours, and conditions of scientists
Origins
Higher education and research training
Brain
drain and brain gain
Research scientists in search of statutes and
status
Choosing research topics and practicing research
The results reported below are mainly derived from a survey carried out during 1985 and 1986. A questionnaire was sent to the 766 scientists in 78 countries who had received research grants from the International Foundation for Science (IFS) between 1974 and 1984. (The IFS, which was founded in 1972 in Stockholm, is a non-governmental organization, multilaterally funded by a number of countries and development agencies. It provides support and guidance to young scientists in and from developing countries. To date, over 2,000 scientists in more than 90 countries have benefited from IFS support.) The results obtained from the questionnaire survey were supplemented by interviews.
All the respondents (489 in 67 countries) are third world scientists, and all are working in their home country. The majority of them (71.4 per cent) work for universities or other academic institutions; 83.4 per cent are men, and 80 per cent are between 30 and 45 years of age. More than 60 per cent of the respondents have a Ph.D. or the equivalent, for which most of them (76 per cent) have studied in an industrialized country. They are mainly working in the agricultural and biological sciences, which are high priority and dominant areas of research in developing countries. The most distinctive feature of the survey population may be that it is composed of internationally selected scientists, chosen according to criteria that have become ever stricter over the years.
The main conclusion of the study [33, 36] is that scientists from developing countries find themselves faced with a dilemma: whether to participate in solving local problems or to follow the models and reference systems more or less imposed by the international scientific community. They are highly dependent on countries in the centre, as well as on the international scientific community. They often rely upon outside sources for education and training, institution building, research financing, etc. To a large extent, third world scientists use international scientific literature as their reference, choose research topics on the basis of essentially the same criteria as their colleagues in the centre, and tend to select the same equipment that they grew accustomed to during their Ph.D. studies in the laboratories of the industrialized countries.
But importing equipment manufactured in the North into the developing countries of the South, even with clear instruction manuals, is not enough to ensure equal quality service [37]. Similarly, scientists who studied in the North often discover that the subject of their thesis, their course curricula, knowledge, and experience are not directly applicable upon return to their home country. It is becoming increasingly obvious that applying major international criteria on scientific communities of the periphery, especially in the developing countries, will not guarantee the latter's integration into the international scientific community. Furthermore, it may detract from the relevance of research to local needs and problems.
Close to one-third of the researchers who responded to our questionnaire come from farming families (many of them from small subsistence-level farms), and one-fifth spent their childhood in a small village. This rural background is even more widespread among the African scientists. Eisemon provided further evidence for these results through interviews he conducted in Kenya and in Nigeria in 1978: "African scientists, like most other Africans with higher education, are usually the first in their families to receive secondary and higher education. Many, particularly in Kenya, come from rural backgrounds" [26, p. 512]. Thus a relatively large number of scientists from developing countries have experienced a rapid social rise, going from a small village to a big (capital) city. At the end of this socio-intellectual adventure, they go on to become members of the intelligentsia, leaving their home village behind them.
For the other categories of social origin, results unquestionably prove that the grade system and then the university system have selection criteria that are hardest on the least favoured classes, albeit without totally excluding them. The intermediate categories (especially crafts and commerce), with approximately one-fifth, are rather well placed. Close to one-tenth have a father in the "office staff" category, whereas the percentage of sons and daughters of "labourers" was lower (3.7 per cent). This last low percentage reflects the inequality of opportunities for the lower social classes; it can also be partly explained by the lower rate of industrialization in most of these countries. The high percentage of researchers (close to one-fourth) whose parents are in liberal professions or senior management positions - a social category that accounts for a small percentage of the population in most developing countries - confirms the inequality of opportunity.
With 16.6 per cent of the overall population, women appear to be underrepresented. However, a quick comparison with the situation in the industrialized countries of the world softens this initial reaction. For example, in 1982 only 13 per cent of the scientists and engineers in the United States were women, and this was a 200 per cent increase over the 1972 figure. In a country like Sweden, which is well known for its efforts in favour of equality of the sexes, women accounted for only 12 per cent of the research scientists in 1982. The use of averages obscures regional disparities and important differences between countries. Women researchers in our population figure as follows: 9 per cent for Africa, 15 per cent for Latin America, and 23 per cent for Asia. The Philippines (36 per cent) and Thailand (33 per cent) had the highest percentages. Some African countries such as Tunisia (27 per cent) and Tanzania (23 per cent) have a laudably high percentage compared to the continent as a whole, while countries like Burkina Faso, Morocco, and Senegal rank far below the average. Our results also brought out a strong degree of disciplinary specialization: women tend to choose disciplines that involve laboratory work and that offer jobs in the capital. Women are often reluctant to live outside urban areas, not only because of their discipline; other factors such as marital status, the number of children to support, and the spouse's profession can also affect the researcher's region of residence.
Compared with the national average in their countries, third world scientists marry late. In our study population, 70 per cent in the 25-29 age group were unmarried, as are close to one-third in the 30-34 age group, and one-fifth in the 35-39 age group. One reason may be that many of them had long years of schooling and extended journeys abroad. Another reason may be the contact with Western models during their studies outside their home countries. The Western standard also seems to have been adopted for the number of children, since two-thirds of the scientists in our population had at most two children. Close to half the scientists in the 30-34 age group and over one-fourth in the 35-39 age group had no children at all. Who do the scientists marry? There is a strong endogamous trend, since half of the spouses are scientists and teachers. The marriage strategy (late marriage, strong endogamy, Malthusian behaviour) seems to characterize a very rational approach to reproduction. Under the influence of the Western model, which holds that small families are more mobile and do better socially than large families, the scientists produce as many children as they think they can establish at a level they would be satisfied to occupy themselves. The investment required for research quite clearly implies postponing marriage and the first child. Since in research the social status that accompanies the profession seems to take more time to acquire than in other professions, scientists have to - and seem prepared to make the relevant sacrifices.
Higher education and research training
Student populations stayed small and relatively few diplomas were awarded by higher learning institutions in most developing countries until the end of the 1960s. During the 1970s student enrolment grew substantially in all countries. By the early 1980s a number of countries, mainly in Asia and in Latin America, boasted a student population comparable with the OECD countries in relative terms (2 per cent or more of the total population). Part of the explanation lies in the creation of many new public universities (most of them outside capital cities) during the 1970s and in the overpopulation of most universities. The proliferation of private universities, mainly specializing in business and administration, also contributed to this spectacular development. The student boom and the large number of graduates produced, combined with the economic crisis and the budgetary cuts, gave rise in the late 1970s to a new phenomenon: unemployment among the intellectuals. This does not mean that all the employment needs have been fulfilled. The situation is quite the opposite. But the key employer, i.e. the state, is no longer able to keep up with the need to create new posts. This is particularly true in Africa, where nearly all branches suffer and where associations of unemployed university graduates have been created.
Until relatively recently - except for certain countries such as India many third world students had to leave their home countries to attend a university and obtain the education needed to become scientists. Studying abroad is nothing new and is not limited to young people from developing countries, but it is noteworthy that the percentages of such students in the total foreign student population has increased considerably in most Western countries since the 1960s [58, 59, 63, 32]. During the colonial period, most of the (very few) students who were sent abroad for their education studied in the colonizers' country. During the pre-independence years, increasing numbers of students applied to study abroad, and the number of scholarships made available by industrialized countries rose considerably. This showed increased awareness of the importance of higher education in development-oriented science; it also reflected the donor countries' desire to maintain - or acquire - political and economic influence in newly independent states.
At the time of independence, there were some universities in the developing countries, but they did not go as far as the doctoral level and did not offer a full range of science and technology courses. In some countries, after the first university was created, change took hold very quickly, especially in the 1960s. By way of illustration, a country like Brazil now offers hundreds of graduate programmes in some 30 independent institutions and universities. Two-thirds of the programmes lead to a master's degree, one-third to a doctorate. At the end of the 1980s, the University of São Paulo alone offered 100 master's programmes and 66 doctoral programmes in a great variety of disciplines.
Although the proportion of doctorates conferred in the developing countries has been constantly increasing since the beginning of the 1970s, research scientists, especially the most active ones, still rely heavily on foreign education. Among the countries that train third world students to the doctoral level, three stand out on the international scene: the United States, Great Britain, and France [32]. A student who has the choice between studying at home or abroad will generally choose the latter. Besides the economic benefits that accompany a stay in an industrialized country, a diploma obtained there is usually rated higher than a diploma from a developing country. The quality of the doctoral programmes in third world universities is also often questioned by the officials of the same countries. It is also claimed that their graduate programmes provide a very slow rate of training; chemistry training in Brazil, for example, requires on average 4.5 years to complete a master's programme and a further 6.5 years to take a Ph.D. [17]. In Thailand, too, it takes an abnormally long time to finish a Ph.D., because many Ph.D. candidates work at the same time, and it is often difficult for them to meet their supervisors, who have many other commitments outside the university [35].
Studying abroad is expensive, though the cost obviously varies depending on the country. In the late 1980s it ranged from US$3,000 per annum in the USSR to US$7,400 in the United States and US$10,800 in Japan. These figures do not include registration and tuition fees nor travel expenses. With US$4,500 as the average tuition fee for a semester in the United States and annual living expenses of $7,440 (adjusted for inflation between 1985 and 1988), two and a half years of study for a master's degree in the United States would cost about $44,000; a doctorate requiring four years of study would "cost" $70,000. By way of comparison, a "maestria" at UCR in Costa Rica would cost only 227,600 colones (including living expenses), which, at the 1987 exchange rate, comes to slightly less than US$4,000, in other words, one tenth of the cost in the United States [68]. For African countries with small budgets, and small developing countries in general, the cost of scholarships for training abroad represents a high proportion of the budget for higher education. Thus, in Mali up to 85 per cent of the higher education budget is spent on scholarships for training abroad [24].
Nevertheless, research training is too heavily reliant on foreign facilities and countries, and training abroad usually does not satisfy the needs of the third world scientists. It would be more realistic, efficient, and, in time, productive to allocate the considerable sums of money now being used to train these scientists abroad to reinforce and establish doctoral programmes leading to a Ph.D. in priority fields within the national universities. Doctoral programmes could also be organized on a regional basis. Strengthening national academia would contribute to improving the structure of scientific communities in developing countries, thanks to added input from both the national scientific potential and the student body. This implies that the countries of the North would have to remodel their educational aid policy, but obviously does not mean cancelling all opportunities for doctoral or postdoctoral education abroad in certain very highly specialized fields. Another important aspect that has to be taken into account is that studying abroad for a long period increases the risk of not returning to the home country.
The survey showed that, logically enough, offers for positions abroad were made more frequently to scientists who had spent longer periods of time for training abroad. There is also a clear correlation between acceptance of the job offer and the number of years spent studying abroad. In other words, the longer one has studied abroad, the greater the chance of receiving an offer to work abroad and the greater the tendency to accept the offer. The motives for returning home or remaining abroad after studying there for several years are diverse. The full potential for emigration is not fulfilled, however, because most of the scientists are attached to their country and home environment. As Bernardo Houssay, the Argentinian Nobel Prize recipient, said, "Science does not have a country, but the scientist does . . . the country where he was born, or raised and educated, the country that gave him a place in his professional career, the country of his friends and family" (quoted in [20, p. 450]). This confirms some of the most important findings of a UNITAR study on emigration and return, namely that "the most common pulls back home are family, friends, and patriotic feelings" [40].
Paradoxically, economic and material factors - even if they may influence the outflow of scientists- are not the strongest determinant of a decision to emigrate. The possibility of obtaining a much higher income may, however, cause a scientist to emigrate for a short period of time. Family ties and children's future are believed to play an even stronger influence than salary or working conditions upon an individual's decision to return to the home country or to emigrate. Another important finding of the UNITAR study, which might have direct policy implications, is that students with scholarships or special grants from their home countries are more likely to return home than those who study abroad with a foreign grant or privately [40]. Similarly, having made the decision to go back home, few scientists plan to emigrate again. Out of close to 500 scientists supported by the International Foundation for Science to carry out a specific research programme in their home country, more than 95 per cent were still active within their national scientific communities in 1985, i.e. 15 years after the first grant was given [33]. Racial, ethnic, and political discrimination may also strongly influence the decision to emigrate or to return home.
In addition, the UNITAR study suggests that "in forecasting whether nationals of a particular country might become part of the brain drain, a more important factor than the stage of development is the extent to which a country trains an excess of professionals in a particular field" [33, p. xxvi]. If the country of the scientist is "the country that gave him a place in his professional career," it is at the same time clear that if the key employer, i.e. the state, does not offer him a position, he will not have much choice but to leave his country if he wants to remain a scientist. A heavy load of teaching and administration, not enough time for research, poor equipment and facilities, and isolation from the international scientific community are among the most important factors in a decision to emigrate, particularly among experimental (biological) scientists and engineers. When planning to emigrate, scientists always prefer to go to the industrialized countries they know best, i.e. the one where they studied. Thus, the United States is clearly the favourite, followed by Great Britain, France, Canada, and Australia. But there are also developing countries to which third world scientists migrate: Nigeria in Africa, some oil-producing countries, Singapore, etc.
The situation in the United States has reached such an extreme, particularly in engineering, that a few people are starting to wonder if the presence of foreign graduate students is a boon or bane [7]. Beginning in 1981, and for every year since then, more than one-half of the engineering doctorates awarded in the United States have been to foreigners, nearly 70 per cent of them Asians. Furthermore, foreigners, and particularly Asians, comprised about two-fifths of total post-doctoral employment in 1985, up from one-third in 1979 [63]. Thus, in given scientific fields (chemistry, physics, mathematics, and computer science) there is a clear shortage in the supply of high quality US applicants and a surplus of high quality foreign (mainly Asian) applicants [21]. A potential reinforcement of the repatriation schemes in some Asian countries and the possible subsequent return of scientists to their home country may pose a threat to the long-term competitiveness of US universities and firms. Thus, a number of countries in Asia and to a lesser extent in Latin America have started to rethink the problem of brain drain and tend to consider that working abroad for a while can represent a gain to the home country, rather than a loss, if the scientist returns with increased skills directly related to the needs of national research groups. Measures should be taken to identify these needs and the scientists concerned. Mechanisms should also be implemented to bring them back home, with attractive research careers and proper professional status.
Research scientists in search of statutes and status
Thus far, research scientists in developing countries long to have a proper professional status; draft texts have often been prepared and then stored away in anticipation of better times to come. Research is often carried out as part of some profession or system designed to uphold professional standards or value systems that are not specific to research. Furthermore, in most developing countries, research scientists do not have high social standing or prestige. Doctors and lawyers and other professionals of that level, with at most the same amount of education as the research scientists, are not only better paid but also enjoy a much higher social status.
Speaking about Venezuela, Roche said, "I know many examples of young people whose rich parents forbade them to major in sciences or to devote themselves to research often because of the low salaries or uncertain career opportunities. The bourgeois attitude to careers in science is much the same as the attitude to professions in the arts; success is reserved to very outstanding people alone, all the others being condemned to a Bohemian life of uncertainty. The profession has probably changed since the Sputnik was invented, but research is still not seen as a fully acceptable profession" [70]. The low wages explain why many of them supplement their incomes by working overtime on side jobs that include anything from working as a consultant, a teacher, or a taxi driver. Anyone who has spent time with third world scientists quickly realizes that a second (or even third) job and income are vital. These additional jobs are of four main types: consultancy, teaching, agriculture, and commerce. Consultancy is nearly always related to the expertise developed in the research activity. Agriculture can mean anything from working on a coffee plantation to raising layer hens. Scientists working in commerce usually help in a family business.
How attractive research can be as a profession depends very much on the country. In Kenya, according to Eisemon, scientists have enjoyed a place of special pride in society since the European colonization period, when close relationships were established between the scientific and the politico-economic circles. Thus, a scientific career, which brings an individual into proximity with the elite, is pursued for social advancement [27]. In India, although the scientific community seems to be dominated by the upper Hindu castes [67], there is paradoxically not much prestige attached to the profession of research scientist; and, like most intellectual positions in the public sector, it is poorly paid [27]. In an effort to better understand the professional choice made by the research scientists in our population, we found out that social status ranked very low, whereas intellectual stimulation and social utility were rated first and second respectively. Actually, an a priori, carefully considered choice seemed to explain a career in scientific research after higher education less than the fact that students were selected or had access to a scholarship at the right stage of their education, even when it meant studying subjects that initially did not interest them [36].
The strategies adopted by the scientists are the result of negotiations carried out in a socio-economic, cultural, and political environment that is not always conducive to scientific perspectives and societal recognition of research science as a profession. Up to the present, science in the developing countries, especially in Africa, has been essentially controlled by government. The first step for the newly independent countries was to build up the state and its institutions. Education was given top priority in order to train civil servants for the state. Careers have, however, often been constructed without considering diploma qualifications. Success in the power struggle has been given more importance than professional specialization. Because of this situation, it has often been difficult to develop research science as a profession, or even as a vocation. As a career, it is not very appealing, and urgently needs statutes.
Choosing research topics and practicing research
The conditions described above affect the way research subjects are chosen and research activities in general are practiced. In an attempt to determine the different factors that may play a part in the choice of research subject, I found that third world scientists have more or less adopted the same reference systems as American researchers working in comparable fields. (For the comparison, I adapted a list of criteria tested in the United States: [15, p. 45].) The leading criterion, "importance to society," takes us back to the criterion that was in second position in the list on choosing research as a profession, namely "social utility." When I asked the scientists what this concept meant to them, they indicated that social utility was more or less the capacity of research to solve the economic and social problems facing their country. The fact that the criterion "demand raised by clientele" is at the bottom of the list no doubt reflects the marginal position of science in developing countries and supports the theory that research scientists and scientific institutions are kept out of the economic and production system.
The findings also confirm the fact that a choice of subject depends more on a series of factors - some of which are external to the science involved than on any single factor. While saying that third world scientists have more or less adopted the same reference systems as their American colleagues, I do not mean that they orient their scientific activity toward research problems defined in the industrialized countries. On the contrary, they do tend to choose research topics that they perceive to be relevant to local problems. This is demonstrated by the fact that a large number of third world scientists who studied in an industrialized country had to change research subjects when returning to their home countries so as to match research work with perceived national needs.
Thus, a researcher who had to work on the problem of nutrition linked to obesity in the United States quite obviously had to change subjects upon her return to Thailand; she decided to work on controlling the thiamine (vitamin B1) deficiency caused by consuming too much tea and tannin. The degree of relevance of the selected research topic to local problems may, however, vary among disciplines. While Lea Velho presents evidence in a recent paper that agricultural scientists in Brazil select research topics directly relevant to local agricultural problems [80], this may not be the case for other scientific disciplines such as physics. More studies would be needed on criteria for choice of research topic in developing countries to come to more definite conclusions.
Time devoted to research depends on various factors. One of them is the nature of the researcher's home institution. Obviously the researchers with the heaviest teaching load work in universities. This is the case of the majority of third world scientists. In an attempt to compare the sample scientist population with their American colleagues, I found that American university researchers on average spent less time teaching than their colleagues in the third world (27 per cent as against 37 per cent), and, above all, more time doing research (57 per cent as against 34 per cent). The differences are much less significant for researchers working in research institutes, although American researchers again spend more time (77 per cent) doing research in these institutions than do their third world colleagues. As for the size of their research budget, the differences are of another magnitude. While American researchers in government research institutes have an average annual budget of US$209,000 and their university colleagues have US$68,000, I found that researchers in developing countries on the average have only between US$5,000 and US$15,000 depending on the level of foreign funding. Even if we are dealing with estimates given by the researchers themselves, who very often do not know the precise total of their budgets, the differences observed are such that they require no further comment.
Other disparities also bring out the fact that third world scientists are at a significant disadvantage compared with their colleagues in scientifically more advanced countries. Lack of equipment, vehicles, technicians, and scientific documentation are among the most frequently observed and described. Another disadvantage that is perhaps even more critical and at the very centre of the scientific enterprise is communication. Many scientists suffer from a feeling of isolation, especially when they have just returned from studying abroad and are trying to fit into the scientific community at home. Moravcsik [61] describes how difficult, and in some cases impossible, it is for scientists in developing countries to communicate with their peers and colleagues by drawing a comparison with birds whose wings have been clipped. The feeling of isolation is probably heightened by the fact that these scientists have been trained in a large variety of universities located throughout the industrialized countries. Furthermore, during this early period, when the young national scientific communities are just "taking off," the scientists often have to cope with being the only specialists in their field within their institution, or even within their country. All the authors agree, however, that science cannot exist without communication, and that a colleague's criticism is vital to progress in any scientific endeavour: "an isolated person builds only dreams, claims and feelings, not facts" [54, p. 41]. Here again these scientists are enduring a handicap little known to their colleagues in the industrialized countries. Other handicaps relate to the visibility and the recognition of their scientific production.
Scientific production: Not very visible
The place of third world science in mainstream
science
Mainstream science and local science: A needed
revision
Developing countries are credited with approximately 5 per cent of the world's scientific production. But science produced there is inadequately reflected in the international databases. International databases, and particularly that of the Institute for Scientific Information (ISI), are very selective and screen only the world's most prestigious scientific journals, the ones that publish the most frequently cited articles. Thus, the Science Citation Index (SCI), developed by ISI, focuses on what has become known as "mainstream science," i.e. the most internationally visible science carried in about 4,000 scientific journals. Since we know that there are about 70,000 scientific journals in the world, we can measure the ISI's selectivity in building up a database; less than 2 per cent of the scientific journals selected come from the developing countries. In general, journals that are not in English are at a disadvantage.