OVERDUE FINES:
25¢ per day per item
RETURNIMS LIBRARY MATERIALS:

Place in book return to remove
charge from circuhmon records

   
   

 
 

\

 
 

‘T‘lkI-lm 4“ s
.4
X

1"” ’

I ‘HQP ,’ 31150
‘H‘r’égfi

MAY 1 6 2004

v

 

 

 

 

 

 

FORMATION OF HUMAN RESOURCES FOR AGRICULTURAL RESEARCH
IN LATIN AMERICA, AND PARTICIPATION IN RESEARCH NETWORKS:

The Case of Former CIAT Trainees.

By

Jairo Cano

A DISSERTATION

Submitted to
Michigan State University
in partial fulfiITment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Educationa] Systems Deveiopment

1981

 

G) Copyright by
JAIRO A. CANO GALLEGO

1981

ABSTRACT

FORMATION OF HUMAN RESOURCES FOR AGRICULTURAL RESEARCH
IN LATIN AMERICA, AND PARTICIPATION IN RESEARCH NETWORKS:
Training researchers at CIAT
By

Jairo Cano

This is a study of two aspects of the performance of persons who
received training in research at the International Center of Tropical
Agriculture (CIAT), Cali, Colombia, during the period l969-79. The
study focuses on two questions: 1) To what extent is the Latin American
region building inventories of intermediate level agricultural research-
ers as a result of CIAT's training?, and 2) To what extent are former
CIAT trainees participating in third cultural networks of science and
technology?

The study population is predominantly composed of young applied
researchers who occupy full—time professional positions at the middle

levels of large, complex agricultural research and development institutes

in Latin America.

a The main findings on inventories of researchers are: a) this
has been an effective process which exhibits some inefficiencies with
respect to the ability of the region to retain trained persons in the

sponsoring organizations and in the activity of research;

Jairo Cano
b) inefficiencies relate to a growing migration process; and c) efficiency
in retention of trained persons is low during the first years of CIAT's
training, but since l975 tends to increase.

Additional findings indicate that research networks are highly
promising for institutionalizing science and technology for agriculture
in Latin America: a) about two thirds of the persons in the sample
participate in networks of researchers; b) most linkages are at the
loci of organizations and country, but almost one third of the signifi-
cant connections are beyond these national boundaries; and c) the net-
works generated by the interaction among these researchers are highly
structured: about one half of the participants are organized in twenty-
two research groups (“invisible colleges"); most of these groups are
specialized by fields of research; over three fourths of the groups
are interorganizational, and one half of them are transnational. About

one third of former CIAT trainees appear to be in a state of scientific

isolation.

To Amparo,
Jairo Alonso,
Juan Carlos, and

David Alejandro.

ii

ACKNOWLEDGEMENTS

In the course of advancing my education to the doctoral level
I have received the support of many persons and organizations. I wish
to thank aTlof'them, even if their names are not printed here.

First, I wish to acknowledge the members of my guidance
committee. During the last two years, I have had the invaluable
opportunity of continuously interacting with two persons of the highest
intellectual caliber that I have encountered during my educational
career: Drs. George H. Axinn and John H. USeem. As co-directors of
this thesis they have both worked hard to open my mind to the complex
of intervening forces which characterize human and social development
in our present interdependent world. They have both spent countless
hours trying to instill in me a sensitivity and scholarly creativeness,
in an interdisciplinary field which still does not reach a distinctive
identity in most higher education organizations: agricultural develop-
ment. I want to express my deepest appreciation also to Dr. James
L. Page, who, as chairman of my doctoral committee, created a flexible
environment which allowed me to work my dissertation topic with
freedom, and with the participation of relevant intellectuals, irrespec-
tive of their departmental affiliation in the university. I wish to
express my gratitude to Dr. Castelle G. Gentry, my department chairman;

during my course work at Michigan State University, he helped me enter

the field of Educational Systems Development. At the dissertation stage
of my education, Dr. Gentry also helped me by making valuable suggestions.
My interest in the topic of this dissertation is not new. In
keeping alive this interest I have received the encouragement of many
persons. I especially want to acknowledge two of them. Dr. R. Vincent
Farace introduced me to the field of organizational communication, and
personally taught me the methods of computerized communication network
analysis; he was one of the first persons who supported-~as a worthwhile
pursuit--my interest in the topic of communication among agricultural
researchers in developing countries, viewed from a network perspective.
Dr. Lawrence Sarbaugh, who acted as my major advisor when I worked for
a Master's degree in Communication, guided me in preparing an earlier
paper in which I was still very ambiguously introducing many of the
ideas on research networks presented in this thesis. In addition,
Larry has been a colleague and a friend on whom I have counted each
time I have needed support during five years of my stay at Michigan
State University.
With the highest appreciation, I want to say thanks to Dr.
Francis C. Byrnes. He has been more influential and supportive of my
professional career than anyone else. Dr. Byrnes selected me, from
the 1968 generation of "Ingenieros Agrdnomos“ at the Universidad Nacional
de Colombia, Medellin, and awarded me a CIAT scholarship to advance my
education at the Master's level in Chapingo, Mexico. Later on, during
five years of work at CIAT, he was not a boss to me, but a mentor.
He will identify many of his own thoughts when reading this dissertation.
I am very grateful to the Rockefeller Foundation, which awarded
me a five-year scholarship to advance my education at the doctoral level.

iv

The Rockefeller Foundation also funded, jointly with CIAT, the field work
carried out for this thesis. Mr. Joseph R. Bookmyer, my advisor at the
Foundation, was totally supportive in helping me make my studies at
Michigan State University an experience personally meaningful for me,
and socially useful for the region to which I belong: Latin America.

I also want to acknowledge the Instituto Colombiano de Estudios
Técnicos en el Exterior (ICETEX) for supplementing, with funds, the last
part of my stay at Michigan State University, enabling me to have the
opportunity to complete this manuscript.

I am deeply indebted to many persons at the Centro Internacional
de Agricultura Tropical (CIAT), and to this organization as a whole.
There is no person or section of the Center who did not help me when I
sought assistance. In particular I wish to mention the names of several
persons. Dr. John L. Nickel, CIAT's Director General has supported this
study in many ways, from its funding to its time-consuming implementa-
tion details; such as the signing, one-by-one, of about three hundred
personalized letters addressed to Directors and Department Heads of
Latin American organizations of agricultural research. Dr. Jose
Valle-Riestra, Director of the CIAT International Cooperation, recom-
mended financially backing the last stage of my work on this thesis,
when the Rockefeller Foundation scholarship expired. Dr. Fernando
Fernandez, Coordinator of CIAT Postgraduate Training and Conferences,
has been a partner in this study from the beginning of the project,
since the middle of T979; in December T979, he came to Michigan State
University to participate in a meeting with the co-directors of this
thesis. Later on, during the field work, Dr. Fernandez gave me full
access to the CIAT training files, and supported me with the facilities

V

and clerical help I needed. Dr. Jesds A. Cuellar, CIAT's Executive
Officer, provided me the administrative assistance required for complet-
ing my work. All the Training Associates made contributions to this
study, especially Mr. Carlos Dominguez, Mr. Elias Garcia, and Mr.
Eugenio Tascon. Dr. Maria Cristina Amézquita and Ingeniera Maria del

R. Henao were also very helpful in the processing and analysis of the
data on inventories of researchers.

I want to convey many thanks to the Directors (If agricultural
. research organizations in Latin America who were responsive to the
request of data for this study,In particular, I want to mention the
names of Dr. Eduardo Alvarez-Luna, Director of the Instituto Nacional
de Investigaciones Agricolas (INIA), Mexico, and Ingeniero Mario Lalama,
Director of the Instituto Nacional de Investigaciones Agricolas y
Pecuarias (INIAP), Ecuador, who collaborated with we when I collected data
for in-depth case studies which, unfortunately, are not reported in this
thesis because time constraints impeded it. I hope that I will soon
produce the documents which incorporate these case studies.

I would like to express my gratitude to all Latin American

agricultural researchers who filled out the network questionnaires, as
well as to those whom I had the opportunity to personnalTyinterview.
I hope that this piece of research may in some way be useful to the
development of agricultural research in the region and to their
personal professional development.

Last but not least, I wish to say thanks to Miss Sharon Wilson

for her invaluable editorial help.

vi

Chapter
I.

II.

TABLE OF CONTENTS

Page

INTRODUCTION ........................ 1

A. Focus of the study .................. 1

8. Independent variables ................. 3

1. Training content ............... 3

2. Length of training. . ............. 3

3. Support facilities to do research . . . 4

C. Research questions. ............ 4

l. Inventories of researchers. . . 4

2. Participation in research networks. 6

D. Limitations of the study ................. 7

l. Limitations on design . . . . . .......... 3

2. Limitations on measurement ..... . 8

3. Practical limitations ........ 9

LITERATURE REVIEW AND THE STUDY CONTEXT . . . 10

A. Rise of Western science and technology (S&T) ...... 11

l. Western science: from magic to scientific knowledge . 11

2. Technology: from craft to science~based technology. . 14

a. Product-embodied technology . . . ........ 15

b. Process-embodied technology ....... . . . 15

c. Person-embodied technology .......... 16

3. Threats and hopes related to Western S&T ..... 16

8. Spread of Western S&T for Latin America ....... . 17
C. Institutionalizing Western S&T for Latin American

agriculture . . . .................... 21

l. Institution building and related concepts . . . . . 21

2. Actions for agricultural research in Latin America. . 27

a. Special Studies Offices . . . . . I . . . . . 23

b. National Agricultural Research Institutes . . 28

c. International Centers ......... . 30

vii

Chapter
D. Developing human S&T capabilities for Latin American
agriculture ....................
1. Training opportunities ............. . . .
2. Training follow-ups .................
3 PROTAAL, an in-depth, comprehensive research project
on S&T in Latin America . . . . . . . . . ......
E. Social structures and third cultural networks of S&T.
l. Notion of network ..............
2. Research networks: organization among S&T
professionals ..................
a. Social structure among researchers .....
i) Research community. . . . . . . . . . . .
ii) Research networks ..... . . . ......
iii) Research groups ..... . . . . . . .
iv) Network structure ...........
b. Cultural elements within research networks.
i) Values. . ..... . . . . . . . . .
ii) Norms . . . . ............
iii) Sanctions . . . . . . . .......
3. Third cultures of S&T . . . .'. . . . . . .
a. The concept of third cultures . . . . .
b. The third cultures of science ......
c. Types of linkages . ..........
Extensive transnational networks .....
Intensive transnational networks. . . . . . . .
Slim transnational networks ...........
No networks . . . . . . . ............
F. The study setting ........ . ...........
l. CIAT as a research and training organization. . .
a. Origin. . ...............
b. Objectives ............. . . .
c. Research approach ............ . . . .
d. Training activities . . . . . . . : .......
2. The food problem in Latin America ......
a. The nutrition problem ..............
b. The production gap .............

c. Heterogeneity of Latin American agriculture : . .

viii

Page

33

33
34

39
44
44
44
45
45
47
48
49
50
5o
52
53
55
S6
57
58
58

59
6O

60
6O

6O
61
62
63

64
64

65
65

Chapter Page
3. Crops as contextual setting ...... 68
a. Rice ................... . 68
b. Bean .................... . . . 69
c. Cassava ..................... 71
d. Tropical pastures .............. 71
4. The trainees' home organizations. . ...... 72
III. STUDY DESIGN ...... 76
A. Strategy ..... . . ........ . . ..... 75
B. Delineating the population ................ 77
1. Definition. . . . . . ............ . . . . 77
2. Data gathering ...... . . . . . ......... 77
3. Population size ..... . . . ..... . . 77
4. Srata for independent variables .......... 78
a. Strata for training content . . . . . ...... 78
b. Strata for length of training . . 78
c. Strata for support facilities to do research. 78
5. PopUlation distribution in conjunction with each of
the independent variables . . . ....... . . . 80
6. Profile of the study population in terms of age,
gender, marital status, education, and country of
origin .................... 84
C. Inventories of researchers: Data collection and analysis. 87
l. Objectives ......... . . . . ..... 89
2. Instruments ........ . . . . . . . . . . . . 89
3. Tactics for data collection ............ . 91
4. Analysis of inventories ........ . . ..... 92
a. Effectiveness in inventory formation. . . . 93
b. Efficiency ............. . . . . . . . 94
c. Migration ........... _ ....... . . 94
d. Retention and stay ................ 94
e. Research experience ............... 95
5. Relationships between stay and two independent
variables . . . ................. . 97
a. Stay and length of training ........... 98
b. Stay and research support facilities ....... 99

ix

Chapter

IV.

D. The researchers' networks: Data collection and analysis .

l. The instrument. . . . . ...........
a. Interchange of research .............
b. Personal data . . . . . . ............
2. Tactics for data collection . . . . . . . . . .
3. Network analysis ...................
a. Participants and non-participants . . . . . . . .
b. Third cultural character of networks. . .....
i) Locus of incidence. . . . . . . . . . . .
ii) Types of linkages ..............
c. Network structure ..... . . . . .......

4. Relationships between participation and three
independent variables ......... . . .....

Findings of the Study, Part I:

INVENTORIES OF HUMAN RESOURCES FOR AGRICULTURAL RESEARCH. .
A. Inventory formation . . . .....

l. CIAT's contribution . . ........

2. Effectiveness . . . . ....... . . . . .

3. Efficiency. . . . . . . . . . . . .....

4. Migration . ...... . . . . . . . . . . .

5. Retention and stay. . . . . . ....... . . . . .

6. Research experience . . . . . . . ..........
B. Relationships between stay and two independent

variables ............. . ......

l. Stay and length of training ..........

2. Stay and facilities to do research ......

Findings of the Study, Part II:
THE SOCIAL STRUCTURES AND THIRD CULTURES OF THE RESEARCHERS'

NETWORKS .............

A. Population and sample . . . . . .............

8. Participants and non-participants ..... - .......

C. Third cultural character of former CIAT trainees'
networks .........................
l. Locus of incidence ............... . . .
2. Type of linkage ...................

X

Page
99
99

100
100

100
101

102
103

103
103

104

106

107
108
108
108
111
114
117
124
128

128
132

135
136
138

138

139
140

Chapter Page
D. Network structure . . . . . . . . . . . . . . . . . . . . 141

E. Relationships with three selected factors . . . . . . . . 177

l. Participation and training content. . . . . . . . . . 177

2. Participation and length of training. . . . . . . . 181
3. Participation and facilities to do research . . . . . 187

VI. SUMMARY, IMPLICATIONS, AND INTERPRETATIONS. . . . . . . . . . 188
A. Distinctive character of the study‘s population. . . . . . 188
B. Summary of findings . . . . . . . . . . . . . . . . . . . 190
1. Findings on inventories of researchers. . . . . . . . . 190
2. Findings on research networks . . . . . . . . . . . . . 192
C. Implications . . . . . . . . . . . . . . . . . . . . . . . 195
1. Implications for CIAT . . . . . 195

2. Implications for international sponsors of agricultural
research . . . 196
3. Implications for country-based science policy makers. . 197
4. Implications for scholars in agricultural development . 198

5. Implications for the field of Educational Systems

Development . . . . . . . . . . . . . 199
6. Two issues on education in general . . . . . . . . . . 199
D. Interpretation . . . . . . . . . . . . . . . . . . . . . . 199
1. Assessing the inventory formation process . . . . . . . 200
2. Conjectures on reasons for migration . . . . . . . . . 202
3. Interpreting findings on research networks. . . . . . . 212
4. Non-participation . . . . . . . . . . . . . . . . . . . 222
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236

A. Letter of the Director General of CIAT for
legitimizing the study and the investigator, and for

sending the census formats . . . . . . . . . . . . . . 236
8. Census instructions . . . . . . . . . . . . . . . . . . 237
C. Letter to send the network questionnaire. . . . . . . . 239
D. General data about CIAT training program. . . . . . . . 243
E. Additional tables on inventories of researchers . . . . 245

xi

Table

10.

ll.
l2.

13.

l4.
l5.

LIST OF TABLES

Training programs offered by CIMMYT, CIAT, and CIP, in
l976, l977, or l978 ...................

INTA, ICA, and Universidad Agraria de La Molina: Personnel
who initiated postgraduate studies in general and in the
national programs ....................

Agricultural research expenditures and expenditures on
research as a percentage of the value of agricultural

product ...........................

Estimated percentages of the population consuming
inadequate amounts of calories in various Latin American
countries ........................

Former CIAT research trainees, distributed by number of
trainees per sponsor organization .......

Classification of Latin American countries by availability
of facilities to do research in the areas in which training
was carried out at CIAT ..............

Distribution of the study population by training content,
length of training, and facilities to do research at home
organizations ...... . . . . . ...........

Number of research trainees from Latin America, distributed
by content and length of training .........

Latin American research trainees distributed by training

content and facilities to do research at home organizations .

Distributions of the study population in terms of age,

gender, marital status, and education . . ..........

Research trainees from Latin America distributed by country .

Former CIAT trainees from Latin America distributed by
training content. . . . ..... . . . . . . . ......

Population and sample of former CIAT research trainees,
distributed by training content and year in which training
was completed ................. «.

Inventory formation for the period l969-79 ..........

Efficiency in the formation of human resource inventories
for agricultural research . . . .......

xii

Page

35

41

43

66

75

81

82

83

83

85
88

109

110
112

113

Table Page

16. Cumulative effects of migration, drop-out, and non-return
up to the year l979 ................ . ..... 115

l7. Cumulative migration, drop-out, and non-return expressed
as percentages of cumulative trained personnel ........ 116

18. Cumulative retention, distributed by calendar years . . . . . 118

T9. Cumulative stay distributed by number of years after
training .............. . . . . . . . . . . . . . 122

20. Number of person-years and efficiency in the formation and
retention of research experience, up to T980 ....... . . 125

2l. Cumulative non-return, drop-out and retention with respect
to research activities for the ll-year training period. . . . 127

22. Summary of the analyses of variance for the dependent
variable stay as a function of the independent variable
length of training ...... . ............ . . . 130

23. Summary of the analyses of variance for the dependent
variable stay as a function of the independent variable
facilities to do research at sponsor institutions . . . . . . 133

24. Former CIAT trainees who compose the sample for network
analysis ...... . . . ................ . . 137

25. Locus of incidence of links reported by network participants. 139

26. Number of links reported by network participants. . . . . . . 140
27. Linkages of network participants ............ . . . 141
28. The distribution of the performance of network roles. . . . . 144
29. Members of the 22 research groups by type of organizational
affiliation ..................... . . . . 147
30. Research groups identified through the network analysis . . . 150

3l. Former CIAT trainees who are members of research groups,
distributed by age, gender, marital status, and education . . 176

I 32. Participation in networks of researchers, distibuted by

training content ................. . . . . . . 178
33. Linkages of particpants at four loci of incidence, distributed

by training content .................. . . . 180
34. Network roles, distributed by training content. . . . . . . . 182

xiii

Table Page

35. Participation in networks of researchers, distributed by

length of training. ................ . . . . . 183
36. Linkages of participants at four loci of incidence,

distributed by length of training . . . ...... . . . . . 184
37. Group members, distributed by length of training. . . . . . . 185

38. Group members, distributed by content and length of training. 185

39. Participation in networks of research, distributed by
facilities to do research at home organizations . . . . . . . 187

xiv

Figure

0301-99.)

11.

12.

13.

14.

15.
16.
17.
18.
19.

LIST OF FIGURES

Research community, research network, and research groups. .

Growth rates of demand and production of food in the Latin
American countries . . . . . ..........

Network roles. ........ . . . . . .....
Cumulative migration, drop-out, and non-return . .
Rates of cumulative retention for the period 1970—79 . . .

Rates of cumulative stay, distributed by number of years
after training ..................

Cumulative retention, drop-out, and non-return for the
eleven-year training period ..............

Average number of years of stay within sponsor organizations,

fields of training, CIAT commodities, and research in
general, for the four training contents .....

Group number 3, integrated by bean researchers . .

Group number 20, integrated by tropical pastures
researchers. . . . . . . . . ...........

Group number l, integrated by tropical pastures
researchers. . . . . . . ..... . . . . .....

Group number 5, integrated by tropical pastures
researchers ..... . . . ..............

Group number 14, integrated by tropical pastures
researchers. . . . ..............

Group number l9, integrated by tropical pastures
researchers .....................

Group number 10, integrated by rice researchers.

Group number 17, integrated by rice researchers ......

Group number l3, integrated by Dean and rice researchers . .

Group number l5, integrated by bean researchers. . .

Group number 7, integrated by cassava and bean

researchers. . . . . . . . . . . ..............

XV

Page

46

67
105
119
120

123

126

131

153

154

155

156

157

158
159
160
161
162

Figure Page

20. Group number 2, integrated by cassava researchers. . . . . . 164
2l. Group number 9, integrated by cassava and bean

researchers ........... . . . . .......... 165
22. Group number 6, integrated by rice and bean researchers. . . 166
23. Group number 8, integrated by bean researchers ..... . . 167
24. Group number 24, integrated by bean researchers. . . . . . . 168
25. Group number 16, integrated by rice and bean researchers . . 169

26. Group number 4, integrated by rice and bean researchers. . . 170

27. Group number 12, integrated by bean researchers. ...... 171
28. Group number 23, integrated by bean researchers. . . . . . . 172
29. Group number 25, integrated by bean researchers. . . . . . . 173

30. Group number ll, integrated by tropical pastures
researchers ................... . . . . . . 174

3l. Research network mapped with base in a network analysis of
data provided by 207 former CIAT trainees ....... . . . 175

32. Percentages of participants in research networks,
represented in relation to training content ..... . . . . 179

33. Percentages of group members represented in relation to
length of training ... . ...... . . . . . . . . . . . . 186

xvi

CHAPTER I

INTRODUCTION

A. Focus of the study.

 

This is a study of training for agricultural research. The
availability of effective researchers is a pressing difficulty faced
today by people and organizations involved in the evolution of the
indigenous science and Western science partnership in developing
countries. During the last two decades scholars in the development
literature have been emphasizing the need for less developed countries
to increase the quantity and quality of their national scientific
human resources (for example see Price, 1963; Moravcsik, 1975; Street,
1979; United Nations, 1979). It is argued that scientific capability
is one of the factors which will enable less developed countries to be
more self-reliant in solving their internal problems, to decrease
their technological dependence on countries scientifically more
advanced, and to more effectively participate in an increasingly
interdependent world community.

The interest in focusing this study on one type of educational
system development, research training in specific fields of agri-
culture, grew out of the hope of many people and organizations for the

potential contribution of science-based technology in increasing

1

r

2
agricultural productivity. Such potential is regarded by both people
from more developed countries (MDC's) and less developed countries
(LDC's) as instrumental for the broader aim of supplying food for a
growing global population.

One of the lessons learned by people engaged in international
agricultural work during the past two decades is that, particularly
for the Tropics, the generation of the biological components of
production technology must be tailored to the local conditions in
which farming is going to take place (Jennings, 1976). Such a
requisite implicitly carries the need of having available task forces
of skilled researchers at national and regional levels, who
effectively work in the generation, validation, and adaptation of
agricultural technology. This requires the deveIOpment of appropriate
educational systems. The number of researchers required for such a
task is not always available in LDC's, and when available they do not
always have the adequate skills (Byrnes, 1974; Wortman and Cummings,
1978; IADS, 1979a; IADS, 1979b).

One action designed and implemented to prepare scientific
personnel for national organizations is non-degree training held in
LDC's at international agricultural research centers (IARC's). The
purpose of this study is to follow-up on post-training experiences of
professionals from the Latin American and Caribbean countries who
participated in training programs at the Centro Internacional de
Agricultura Tropical (CIAT), Cali, Colombia, during the period 1969-
I979.

The study selectively examines two segments of former CIAT

trainees' performance as researchers: 1) reincorporation of former

trainees into their organizations and stay in the activity of
research, and 2) their participation in person-to-person networks for
research interchange.

B. Independent variables.

 

Relationships between the above-mentioned aspects of former
trainees' research performance and three selected factors are
explored: 1) training content, 2) length of training, and 3)
facilities available at home organizations for doing research in the
fields in which people carry out their training. Other features of
the training experiences are assumed to be similar for all
participants: the philosophy and goals of the training program; its
physical and organizational setting; the training facilities available
at CIAT; and the quality of instruction, e.g., quality of instructors,

delivering methods, and training materials.

1. Training content is assumed to reflect a substantive

 

variable of the structure of science: the knowledge and technology
basis available for doing research in different fields of science.
Variation due to training content is explored by grouping the
population in research-related commodities. Content is assumed to be
similar within each research commodity, but different between them.
The research and training work at CIAT is organized in four
commodities: a) rice, b) beans, c) tropical pastures, and d) cassava.
Therefore, these four categories of training content'are separated in
this study.

2. Length of training is defined as the time from physical

arrival at CIAT to departure from CIAT, measured in number of months.

'1

It is grouped in three levels: a) short training, when training took
up to two months; bl intermediate training, when it took more than two
months up to six months; and c) long training, for programs longer
than six months.

3. Support facilities to do research. The degree of facility

 

support needed for carrying out research once trainees go back to
their home organizations is a structural variable which reflects the
national priorities assigned to research within the countries as well
as internal constraints within organizations. To account for this
variable, the Latin American* countries which have sent professionals
for training to CIAT were classified in eight groupings, two per
commodity: those with more research facilities in each commodity, and

those with less facilities. (See pp. 78-81).

C. Research questions and rationale.

 

The study is centered on the exploration of the two following
questions: 1) To what extent is the region building inventories of
human resources for agricultural research as a result of CIAT's
training?, and 2) to what extent are former CIAT trainees
participating in third cultural networks of science and technology?

1. Inventories of researchers. In the context of this thesis

 

this expression means the number of persons available each year, who

have appropriate training to carry out specific research tasks. The

 

*The term Latin America is applied loosely in this
thesis to mean the Western Hemisphere with the exclusion of
the U.S.A. and Canada. The term Latin America is used
interchangeably with thw words "the region".

_,1

5
building of effective research organizations has been postulated as
one of the foundations of the institutionalization of science for
agricultural development (Moseman, 1970). At the core of the building
of agricultural research organizations is the problem of staffing
(Drillon, 1977; Wortman and Cummings, 1978). The aim of building
"critical masses" of trained researchers for staffing agricultural
organizations in Latin America is difficult to achieve because of a
high mobility of researchers out of the organizations which sponsor
their training, and out of the activity of research.

If the goal of training programs at IARC's is to strengthen
the research capabilities of national research organizations by
helping them to train their research staff, immediate concerns with
respect to training outcomes are: What has been CIAT's contribution
in this respect? How effective and efficient has the region been in
building human resource inventories for agricultural research with
basis in CIAT's training? How is migration of researchers toward non-
research activities affecting those inventories? For how long have
professionals stayed in their sponsoring organizations after training?
A crucial feature of scientific work is that the experience required
for effective performance is developed by researchers through long
periods of involvement with particular problems. Consequently, a
question relevant to the issue of human research inventories is: How
efficient has the region been in forming and retaining research
experience in relationship to the personnel trained,at CIAT?

There is no simple way to answer the above-stated questions.
The formation of human research inventories may be appraised from the

perspectives of different persons and interests. Four views are

6

envisioned and examined in this thesis: a) the perspective of the
national organization, public or private, which sends members of its
staff to advance their training with the expectation that they will
return to the organization and utilize that training for better
performing their duties; 6) the perspective of the coordinator of a
commodity program at the training organization, who expects to
increase the number and quality of researchers working in a given
subject, thus establishing a basis for future research collaboration;
c) the perspective of the directors of the training organization, who
expect that former trainees will engage in research work related to
the center's mission and priorities; and d) the perspective of
country-based and international research sponsors and science policy
makers, who expect that trained personnel will engage in research-
related work in the region. From this final perspective, it does not
matter that professionals work for a particular organization or in a
specific field; what matters is that they apply their training and
contribute to the advancement of science and technology.

2. Participation in research networks. Interfaced with the
building of formal research organizations, it is assumed that there is
another substantive factor involved in the institutionalization of
science in devel0ping countries: the building of formal and informal
channels for the interchange of research findings and resources, e.g.,
scientific literature and germ plasm. This study does not focus on
the larger formalized channels, which are a part of organizations and
programs, and which set the stage and form the base for all
interpersonal relationships, but is limited to the person-to-person

informal networks among working scholars.

7

In recent years, an increasing awareness and interest in the
informal communication system of science has developed. The relevance
of this system has been supported by findings in the sociology of
science, which consistently assigns a central role to the personalized
interchanges in the advance of research areas.

Such communication occurs among the most effective members of
any field of research. These networks include interpersonal ties
between researchers within a country and across national boundaries.
The relationships between people from different societies create
opportunities for “third cultural networks" to emerge; the cognitive
nature of interchanges among researchers allow reference to those sets
of people and relationships as third cultural networks of science
(Useem, Donoghue, and Useem, 1963; Useem, 1967, 1971; Restivo and
Vanderpool, 1974).

This study explores the extent to which former CIAT trainees
are participating in third cultural networks of science and
technology. Three aspects of networks are studied: a) the extent of
their participation in research networks; b) the third cultural
character of the networks in which former CIAT trainees participate;

and c) the social structures of those networks.

0. Limitations of the study.

This descriptive study aims at developing a data-based
analysis of the impact of training at an IARC upon two selected
aspects of the research performance of agricultural researchers who
took part in that form of training. The study does not pretend to

test hypotheses or determine causal relationships. This follows

methodological limitations in design and measurement of post-training
research performance in the particular population being studied in
this thesis, and also practical limitations of time and financial
resources.

1. Limitations on design. A comparative study of research
performance of trainees and non-trainees would have been desirable.

It would have required an experimental or at least quasi-experimental
design. Experimental and quasi-experimental designs were found
unfeasible for this study. Limitations were: a) independent
variables of interest could not be manipulated in order to have random
assignment of treatments to experimental groups; b) control groups
were found unfeasible for at least two reasons: first, inventories of
human resources in agricultural research were not found for most Latin
American countries; and second, the identification of cohorts (an
alternative to sampling from the population of Latin American
agricultural researchers) is costly and there were concerns regarding
equivalence of cohorts; and c) pre-test measures of research
performance are not available for the population of former CIAT
trainees, nor for candidates which were not admitted to the training
program.

2. Limitations on measurement. Measurement of research
performance in this particular population was another source of
limitation for the study. Difficulties were found in applying three
types of measures to the population under study: a) citations in the

scientific literature; b) number of research papers; and c) number of

 

professional and work associated activities. Citations are not

available for most members of this population. Former CIAT trainees

9

usually publish their work in reports of their organizations where the
identity of authors very often is not established. The same
limitation applies to the second measure, research papers; in
addition, this measure presents problems of equivalence. The number
of professional and work-related activities carried out by
agricultural researchers after training is a feasible measure for this
population, but the problem of equivalence remains, and is even more
critical because these professionals practice a wide diversity of
research activities, depending on their speciality.

3. Practical limitations. A source of practical limitation
was the circumstance that the study population is scattered across 26
countries and 182 organizations covering a wide geographical area from
Mexico to Argentina. Time and financial limitations were related to
the fact that the author was conducting this study under the
relatively fixed boundaries of an appointment as a graduate student.
Such constraints made the exclusion from this thesis of the reports of
case studies carried out in Mexico and Ecuador necessary. Data
collected for these case studies refer to a deeper examination of
aspects of the research performance of former CIAT trainees in the
sites in which they are presently carrying out their professional
duties. Both personal interviews and ethnographic data about the work
setting were gathered for this purpose. The author hopes to find
opportunities to analyze and report those case studies after

completing his graduate studies. .

CHAPTER II
LITERATURE REVIEW AND THE STUDY CONTEXT

This chapter presents a review of pertinent literature in
relation to the development of human resources and networks of
researchers in the context of the institutionalization of science and
technology for agriculture in Latin America.

The purpose is to provide a framework of concepts, facts, and
interpretations--drawing upon a diversity of intellectual traditions--
which are relevant to this thesis.

In as much as possible, literature selected for this chapter
refers directly to developments that occurred in the region. However,
ideas and lessons from studies focused on other regions of the world--
more and less developed regions--are also included. This is the case
of the sociological construct about third cultures of science,
generated by John and Ruth Useem and several of their students (1963,
1967, 1971); the main empirical referents for these studies being four
countries of South East Asia. General writings about S&T and about
research networks are another example of material utilized for
structuring this chapter, even though their authors in
most cases did not focus specifically on Latin America.

These materials are arranged in six sections. The first two deal
with general concepts about S&T, and about its spread in Latin
America. The third section refers to concepts of institution building

10

11

and efforts at institutionalizing agricultural research in the region.
The fourth part focuses on training opportunities and assessments of
them, for develOping human S&T capabilities for Latin American
agriculture. The next section is a framework for examining the social
structures and third cultures which develop among professionals who
participate in research networks. Finally, the last section presents
overviews on the empirical setting for this study: CIAT as a research
and training organization; the food problem in Latin America; four

crops as contextual setting; and the trainees' home organizations.

A. Rise of Western science and technology (S&T).

1. Western science: from magic to scientific knowledge.
Western science is a social phenomenon which evolved through a long
period out of ancient cultures in the East, and crystallized in
several countries of Western Europe about five hundred years ago. It
was derived both from the ordered speculation of the magician priest,
or phiTOSOpher, and from the practical Operation and traditional lore
of the craftsman. This phenomenon was backed in the Hellenic period,
continued through the Roman and Middle Ages, incorporated Arabic
contributions, and absorbed the intellectual advances ot the
Renaissance (Bernal, 1939, pp. 13-34; Sagasti, 1979). Basalla (1967)
mentions seven Western EurOpean nations as providing the original home
for science during the 16th and 17th centuries: Italy, France,
England, the Netherlands, Germany, Austria, and the Scandinavian
countries. '

The words "Western science" appear in the literature designating

at least four different manifestations of this phenomenon: a) a

product, e.g., scientific knowledge; b) a set of systematic

12

procedures, e.g., scientific method; c) a type of human work, e.g.,
scientific research; and d) a social institution, e.g., the
collectively supported, specialized enterprise of scientific knowledge
production.

A common definition of research is that of careful reseach or
close searching. In Western science such search has been classified

according to its purposes: when the purpose is understanding of

 

natural phenomena without reference to any practical end, it has been
called basic or pure research; on the contrary, when its aim is 32
solve a practical, specific problem, it has been named applied
research. Such classification is not as clear cut as it seems at
first glance. Two reasons to support such a view are: First, what is
defined as a "natural“ phenomenon tends to vary over time and from one
social grouping to another; and second, research which is aimed at
solving practical problems may result in basic understanding, and vice
versa (Mulkay, l977). However, the distinction basic-applied is
useful for looking at science conceptualized as a social system.
Building on the work of Merton, Norman Storer developed a
sociological construct which sees science as another primary social
system, such as the family, or religion (Storer, 1966). Storer's
central argument is that science has a commodity of its own, which
regulates transactions among scientists. Such a commodity is
recognition to scientist's work given by competent colleagues. In
their search for recognition, scientists must adhere to a series of
institutional norms: universalism, communality, disinterestedness,

and organized skepticism (Merton, l942).

13

Under this conception, legitimate members of the system of
science are those people whose only concern is the description and
understanding of empirical reality. Those are the "basic," or "pure“
scientists. In contrast, applied scientists are tied to the solving
of practical problems. This tie exposes them to developing interests
in rewards other than recognition, such as money and power.
Consequently, following Storer's conception, applied scientists tend
to be regarded with suspicion by basic scientists, who sometimes
reject applied scientists or at least do not accept them as legitimate
members of the social system of science. The view of the normative
structure of science was, in general, accepted uncritically throughout
the l950‘s and most of the 1960's, but in recent years it came under
considerable criticism (Mitroff, T973; Duncan and Zaltman, 1973).
Central implications of the Storer's conceptualization is the unity
and internationality of science.

It seems that Western science has been concerned since its
beginning with basic and applied purposes. The statement that Nature
can be mastered through understanding has been attributed to Francis
Bacon. Underlying this statement are several assumptions related to
Western science, e.g., that order exists in Nature, that humans can
"discover" such order and translate it into scientific laws, that
Nature proVides the objective criteria to test the validity of
scientific laws, that such laws allow humans to make predictions, and
that the capacity of humans to understand and predict natural
phenomena gives to people and nations unprecedented power over Nature.
Such power over Nature materializes in terms of what is widely known

as technology.

14

2. Technology: from craft to science-based technology. Not all
technology, however, is science-based. Sociologists and
anthropologists have used the word technology to designate in general
the "know-how" that human beings have developed, wherever and whenever
they have lived, which has been at the same time a by-product of and a
tool for survival. In fact, in this sense, perhaps most technology
used today in human affairs is not science-based.

According to Sunkel (l977) science-based technology is the result
of a complex process of interaction between science and production,
which took place in Western Europe concurrently with the emergence of
capitalism as the dominant mode of production. He points out that in
the same countries where sciencecngstallized, there was a parallel
development in production technologies, out of the crafts practiced by
artisans. Crafts were gradually transformed into manufactured ac-
tivities, and later into industrial production. The merger of both
currents, science and production technology, constitutes what is known
as the scientific and technological revolution (Sagasti, l979).

Goulet (1975) defines technology as the systematic application of
collective human rationality to the solution of problems by asserting
control over Nature and over human processes of all kinds. Goulet
states that implicit in his definition there are several character-
istics of technology: a) technology is normally the fruit of system-
atic research, which is disciplined and cumulative, not merely ac-
cidental or serendipitous; b) technology is not merely intellectual
speculation but rather knowledge applicable to practical problems; c)

this systematically applied human reason must operate in a collective

15

social context, so that a practical invention which originates in a
solitary mind does not qualify as technology unless it is expressed in
a tool, process, or object which can be used by others; and d)
technological activity aims at expanding and improving the ability of
human beings to control the natural and social forces which surround
them.

Technology interfaces with knowledge, but it is not totally
knowledge. It has abstract and concrete components. Abstract
components of technology include symbols, concepts, models,
principles, theories, and laws, that together may be designated as
data, information and knowledge.

In relationship to problem solving, the concepts of knowledge,
information, and data may be seen as different states of refinement in
a continuum of abstractions. A target problem leads the orientation
of the continuum. The continuum starts with empirical facts of Nature
and goes to abstractions which attempt to generalize over time and
space. In this context, data are defined as the facts of Nature as
perceived by humans through their senses; information is defined as
data processed to solve a specific problem; and knowledge is defined
as information of future use in general, not "here and now," but in
any place and at any time (McDonoughs, 1963).

Technology includes not only abstract “know-how“ knowledge, but
physical states in which knowledge is "embodied". Hall and Johnson
(1971) conceptualized technology as know-how derived from scientific
knowledge and incorporated in some object, process, or activity. They
say that, a) certain technologies are incorporated in concrete Eggl§_

or products, e.g., fertilizers, tractors, seeds; b) process-embodied,

 

 

16

in turn, are technologies incorporated in plans, formulae, blueprints,
and directions for processing materials into finished products, e.g.,

farming practices; and c) person-embodied or decisional technologies,

 

embraces the practical knowledge used by planners, designers,
managers, scientists, engineers, and technicians in analyzing bodies
of data to determine what practical consequences may be drawn from
them.

3. Threats and hopes related to Western S&T. Western S&T has
been regarded both as a destructive and a constructive tool for
humankind. In 1931, Bertrand Russell wrote a book anticipating
several of the threats associated with uncontrolled S&T. He says that
one of the serious drawbacks of S&T is that the power which it
generates is not always accompanied by wisdom; Russell refers to
"wisdom" as a sound conception of the supreme ends of human life.
Russell advises not to abandon S&T, but to be sure that its
utilization is guided by wisdom.

In his well known book, The Social Function of Science (1939),

 

Bernal also treats the topic, attributing the opposition to the use of
science to a blind attitude of "romantic reactionaries and
conservative economists“ who fail to distinguish between the necessary
effects of science and its abuse.

Many others have examined the issue. Recently, Goulet (1975)
points out three dangers involved in the use of S&T: a) S&T is a
powerful instrument of domination, allowing its owners to exercise
social control in various forms; b) it decisively affects modes of
decision making; and c) it relates directly to patterns of alienation

characteristic of affuent societies. From the constructive side,

17

Goulet sees S&T as a major resource for creating development through
greater production and productivity.

In relation to agriculture, most peOple working in international
assistance from more developed countries (MDC's) to less develOped
countries (LDC's) regard S&T as the route for economic growth, as the
engine of change in agriculture, as an essential for agricultural
deveTOpment, or at least as a potential contributor to improve the
conditions of the urban and rural poor in developing countries (for
example, see Schultz, 1964; Mosher, 1966; Ruttan, 1971; Moravcsik,
1975; Wortman and Cummings, 1978).

8. Spread of Western S&T for Latin America.

 

According to Basalla (1967, p. 612) the spread of Western science
to a wider world took place through military conquest, colonization,
imperial influence, commercial and political relations, and missionary
activity. From the 16th century through the 17th and 18th centuries,
there was a constant stream of Spanish, French, German, Dutch,
Swedish, and English naturalists traveling on scientific expeditions
to South America, many of them as members of scientific movements wich
culminated in the early decades of the 19th century in the works of
Alexander von Humboldt and Charles Darwin (Basalla, 1967, p. 612).

In most Latin American countries, scientific activites started in
colonial times, performed by science pioneers. They were gifted
individuals, usually financed by their personal fortunes. They
followed the trends imposed by the leading EurOpean'scientific
community (Flit, 1979). However, an autonomous community of Latin

American scholars never flourished, and S&T did not root in these

18

societies. This anomalie is explained in terms of the dependency
theory by Sagasti (1979) as follows:

At the same time (as the rise of the scientific and
technological revolution) Latin America became
incorporated into an international division of labour
as colonies, thus helping to sustain the industrial
revolution through the supply of cheap raw materials
and the provision of markets for manufactures.

As a consequence of these historical processes,
(Latin America) did not establish a basis of productive
technologies linked to scientific findings of their
own. There was no organic linkage between the
development of activites devoted to generation of
knowledge and the evolution of production techniques,
with these two areas remaining isolated from each
other.

The diffusion of Western science (to these
countries) was an irregular process, entailing a
partial acceptance of results, but without full
awareness of the cumulative process that originated
them. Science in these countries was an activity
limited to a few isolated pioneers whose efforts were
inherently out of phase in time, since the frontiers of
knowledge were being explored in other parts of the
world. As a consequence, the pursuit of science did
not grow roots in the majority of these countries until
the first decades of the twentieth century, and even
then it acquired a fragmentary and imitative character,
divorced from productive sphere.

The nature of modern productive activities was
conditioned first by the interest of the colonial
powers and then, after some regions became independent,
by the way in which their economies were incorporated
into the international division of labour that
accompanied the expansion of the capitalist system.
This meant that they were oriented primarily towards
the extraction of natural resources, and to the
generation of surpluses to be transferred abroad.

The implanted or modern productive activities
employed imported technologies that brought with them
skill requirements, use of materials, organizational
habits, and technical traditions that were alien to the
local envirnment. Furthermore, the technological
capabilties associated with modern production were
expanded primarily through new technology imports, wich
meant that the technological traditions--developed
slowly and cumulatively over a long time-~were left
aside and even eliminated. This led to a reduction in
the variety of indigenous technological responses.

19

Even admitting that dependency theory explanations are difficult
to test empirically, the preceding paragraphs seem to illuminate the
context in which an intensive transplant of agricultural technology
was carried out from the United States to Latin America after the
Second World War, via agricultural extension, or the "servicio" as it
was known in several countries of the region. Such activities
included transplants of product and process emodied technologies. An
example of products are, tractors, fertilizers, and seeds. Some
processes involved are, farming practices, and the whole conception
about how to extend the new technologies to Latin American farmers.
This conception was taken and directly transplanted to Latin America
from the experiences which evolved in the United States Agricultural
Experimental Stations.

Basic assumptions underlying the transplant of agricultural
extension to Latin America were: a) technology to increase
agricultural productivity was already available, the basic problem was
one of communicating it to Latin American farmers; b) "rational"
farmers in Latin America should adopt the new technology, as U.S.
rational farmers do; c) Latin American farmers should "modernize"
following the paths that U.S farmers followed; d) new technology was
something inherently "good" and desirable, which would benefit any
people who would adopt it; and e) powerful tools for extension were
the massive communication media, e.g., newspapers and radio. All
these assumptions were inadequate. Inadequacies have been widely

discussed by scholars in the agricultural development literature (for

20

example, see Myren, 1964; Grunig, 1968; Felstehausen, 1971; Rice, 1971;
Diaz Bordenave, 1976; Beltran, 1976; Rogers, 1976; Axinn, 1978).

Results of technology transplants to Latin America, via
agricultural extension, started to be assessed in the 1960's through
numerous studies done by members of the diffusionist school and by
economists. The general appraisal of those studies was summarized by
Rogers with a sense of failure in an article titled "Communication and
DevelOpment: The passing of the Dominant Paradigm,“ which appeared in
an issue of the “Communication Research" (April, 1976) entirely
devoted to examining the t0pic.

Although all these papers are focused on the role played by
communication strategies and actions, they allow one to draw a
conclusion which is not confined to the field of communication, but it
rather represents a general lesson: social change and modernization
is a very complex process affected by many factors internal and
external to the pepple who are supposed to be the main beneficiaries
of change and modernization.

Not the most important factor, but one which, with others, is at
the center of social change and modernization, is the nature of new
technologies, including where and by whom such technologies were
developed. It seems that some Latin American leaders and
international experts involved in technical assistance programs
perceived the problem very early in the stages of the extension work,
and consequently started carrying out activities to finstitutionalize

S&T for Latin American agriculture as an internal capability of those

societies.

1

21

C. Insitutionalizing Western S&T for Latin American Agriculture.

1. Institution building and related concepts. The word
institution has been used with different meanings in the social
sciences. For the purpose of this dissertation it means a relatively
organized and enduring social system that has established roots in a
society in response to some basic and persistent problems.

As a social system, an institution is comprised of social

 

relationships in intimate interplay with cultural elements. The
social relationships include a set of distinct positions and the
corresponding interrelations between them. Attached to those
positions are sets of rights and obligations, whose meanings and
definitions are tied to the cultural elements of the system and its
wider environment. The set of relationships constitute the internal
structure of the system, structure which is characterized by recurrent
and regularized patterns of interaction among the occupants of the
system positions.

Three cultural elements are central to the functioning of a
social system: values, norms and sanctions. .Vglggs are generalized
beliefs about what is desirable and what is undesirable in a society;
such beliefs are of maximum importance because they legitimize the
very existence of the system. Nggflg are standards of conduct which
regulate the interaction among individuals in a society. Sanctions,
including both rewards and deprivations, involve the use of various

resources to control the behavior of persons in a society.

22

The basic unit of a social system is not an individual, but
selected aspects of the individual's actions in relationship to the
system and its wider environment. A sociological category suitable to
study those selected aspects is the concept of social role. A £919
consists of the expected behaviors of occupants of a specific
position. Social systems do not exist in a vacuum, but are embedded
in larger cultures and societies. Consequently, a role is determined
by internal characteristics of the system to which it belongs, and by
external factors of the environment in which the system is embedded.

A person who occupies a specific position learns the expected
behaviors attached to that position (its social role), and at the same
time interprets those expectations in personal terms. Therefore, that
person's role performance includes both the role as socially defined

and his/her interpretation of that role. The term role performance

 

applies to what individuals as occupants of positions do in face-to-
face or mediated, formalized or informal encounters with other people.
Institutionalization is a concept which allows one to relate a
given social system (an institution) to its larger social and cultural
environment. As used in this dissertation, it does not imply a fixed
state but a goal, the goal of inserting a complex innovation into a
society. In other words, the goal of organizing an enduring set of
expectations (roles) in terms of essential components (positions,
internal structure, values, norms, and sanctions) of a social system
(institution). Specifically, in the context of the present study
institutionalization refers to the goal of introducing Western S&T as

an internal social system of Latin American nations.

23

During the past two decades many people and organizations have
been working for the goal of the institutionalization of new or
"modern“ social systems in developing countries. Parallel with those
activities, a conceptual framework emerged under the leadership of
Professor Milton J. Esman and with contributions of numerous scholars
inside and outside the United States (Blaise, 1973; Axinn, 1978).
This paradigm was named “institution building" by its developers. It
is strongly rooted in the concepts of social engineering, based on the
proposition that the most significant, contemporary changes--
especially in developing countries--are deliberately planned and
guided. The word institution is synonymous to the word organization.

Consequently, institution building consists of the process of

 

designing, implementing, installing, and fostering new formal

organizations or changes in old ones (Esman, 1967).

 

In terms of the concepts presented at the beginning of this
section, institution building may be seen as a strategy to
institutionalize a given social system in a society. The word
institutionalizing is used here to indicate the actual process of
carrying out such a strategy. Seven crucial events may appear in the
process of institutionalizing a given social system: innovation,
rejection, legitimation, acceptance, normality, entrenchment, and
rigidity. In Axinn's words:

The idea is that a new organization, or a change

in an old organization might, at first, be seen

as an innovation. It is natural for the environ-

ment to reject it. If institution building is

successful, it will be legitimized, and eventually
be acce ted. Later, usually, it is viewed as
normal ln the society, and then can become
entrenched. If it also becomes so rigid that it no
longer serves its changing environment, new

 

24

innovations will come along and either replace it
or modify it (Axinn, 1978, p. 161).

When has that innovation become institutionalized? The
conditions mentioned in the literature may be summarized in two
general criteria: a) survival, and 6) social value. Survival of the
organizations concerned with the innovation is a necessary but not a
sufficient condition for institutionalization. If organizations
dissolve, the innovation will not become an institution. But they can
continue operating and yet fail to be institutionalized (Esman, 1967).
The innovation becomes insitutionalized when the organizations
involved have demonstrated the 13123 of its functions over time, and
others have accepted those functions as important and significant. In
the process, the involved organizations become more stable and secure,
more capable of performing those functions and they ensure that their
rising productivity is incorporated into the society's regular
activities and beliefs (CEDA, 1971).

Social value is not a unidimensional and easily determinable
category. It may manifest in society in several ways. For example,
value may take the form of social recognition, e.g., increased or
decreased prestige and monetary compensations given to the persons who
perform the roles implicit in the corresponding institution. Value
may also be socially expressed as increased or decreased allocation of
resources, e.g., higher or lower budgets for the corresponding
organizations. Another manifestation of social value is the
utilization of the institution by other systems of a society, e.g.,
the production sector or the consumption sector. If producers use the

outcomes of the institution, e.g., new technology, that is an

E

25

indication that some social value is assigned to the functions of the
institution in that society. Consumers also may indicate the value
they assign to the institution, by accepting or rejecting the goods
and services, e.g., food products, which incorporate outcomes of the
institution.

In summary, an innovation has been institutionalized in a society
when it has developed deep roots in the functioning of that society.
In return, the society shows appreciation (value) for the new
functions by supporting the institution in a variety of ways. But
this process does not take place easily or quickly. The process of
social change that steadily evolved and resulted in the Western
institution of S&T took place in a series of small steps, the
cumulative effects of which were not apparent to their initiators.
Increasingly complex and formal organizational structures emerged as
the cumulative consequence of their activities at the grass roots
level (Morse et a1., 1969).

However, as pointed out by members of the institution-build ng
school, institutionalization need not be a "natural" or evolutionary
process, which occurs independently of deliberate human will and
direction. Institutionalization, in present times, is achieved "by
design” and typically following an "inverse model." In this case the
hope is that eventually the set of new or reformed organizations or
"institutes" will evolve in well-established and structured
constellations of roles which fulfill specific functions for societies
or groups within societies, and which will be valued as "natural"

institutions, such as family or religion.

26

In the long route to institutionalization, an essential requisite
will be to secure continuity to the operation of those organizations.
And continuity will probably not be sufficient. Changes and
"institution building" in other aspects and sectors of the concerned
societies will be necessary. In recent days, a general preoccupation
of LDC's leaders has been not only to have available internal S&T
capabilities, but to be sure that those capabilities are indigenous.
That means to have control of the whole process of S&T importing,
testing, adapting, generating, and utilizing (for instance, see
documents of the "Group of 77" to the 1979 United Nations Conference
on Science and Technology for Development, Vienna 1979; and the
Background Study on Suggested U.S. Initiatives for the same
conference).

In Latin America the regional Organization of American States has
been concerned with the topic for years. Its work has concentrated on
the development of national 5&1 policies. In relationship to
agriculture an important action has been the establishing of and
permanent support of the Instituto Interamericano de Ciencias
Agricolas (IICA), which has conducted social and biological research
and largely contributed to the development of indigenous human S&T
capabilities through its Turrialba Graduate School.

A very influential work with respect to institutionalization of
science in Latin America has been the concept called "Sabato's
triangle," which is based on the proposition that the development of
indigenous S&T is the result of a deliberate process of

interrelationships among the vertices of a triangle: the government

15“"!

27

vertex, the vertex of S&T infrastructure, and the productive structure
vertex (Sabato and Botana, 1975). Following a similar line of
thought, Sagasti (1979) has identified three groups of actions
necessary for the development of indigenous S&T capabilities in Latin
America: a) the expansion and reorientation of the S&T system, b) the
selective and systematic recovery of the traditional technological
base, and c) the transformation of the productive system.

In general, it may be said that the design of strategies to
institutionalize Western S&T in developing countries may result in
different modalities of institution building. For instance, in
centrally planned economies of socialist countries, the strategy may
consistcfiibuilding a centralized governmental organization. In mixed
and market economies, it may be to build a set of public and private
organizations. In a predominantly market economy, it may put emphasis
on private organizations whose survival will depend on the
profitability of their outcomes.

With reference to Latin America, a set of strategies and actions
for institutionalizing S&T were initiated in the post-war period, and
since then they have been dynamically evolving. Such actions were
initiated with participation of national leaders and supported by the
U.S. Department of Agriculture and Rockefeller Foundation. Later on,
other governments and private organizations joined the enterprise.

2. Actions for agricultural research in Latin America. Attempts
to institutionalize Western S&T for Latin American agriculture appear
to have started during the 19th century and the first part of the
present century, but appropriate documentation was not found in the

literature search carried out for this dissertation. In Latin America

14...

28

there are S&T agricultural organizations whose foundations are traced
to that period. For instance, the Escuela Nacional de Agricultura de
Chapingo, Mexico, installed during the first years of the Mexican
Revolution, has it roots in the Escuela Nacional de Agricultura de San
Jacinto, founded in 1854 (Fernandez y Fernandez, 1976; Gomez, 1976).
Pichilingue, in Ecuador, which was originally tied to plantation crops
like banana and cacao, took in 1943 the name of Estacidn de
Investigacidn Agricola del Ecuador, and in 1963 was incorporated to
what today is the Instituto Nacional de Investigaciones Agricolas y
Pecuarias, INIAP (EMA and AID, 1979).

a. Special Studies Offices. The efforts to install 5&1
capabilities for agriculture in Latin America with U.S. assistance
started with the organization of Special Studies Offices, which
appeared in several countries like Mexico and Colombia. In terms of
Hall and Johnson's typology, those programs began with a transfer of
person-embodied technologies represented in the staffing of those
offices with well-known agricultural scientists like Richard
Bradfield, P.C. Mangelsdorf, E.C. Stakman, J.G. Harrar, and Norman
Borlaug. Subsequently, the development of indigenous human S&T
capabilities started with in—service training programs and with the
sending of promising young nationals to U.S. universities, in order to
continue their education at Master's and Ph.D. levels (Wortman and
Cummings, 1978).

b. National Agricultural Research Institutes. ‘put of these
seminal experiences, a process of institution building was initiated
in the late 1950's. Most of the present Latin American national

agricultural research institutes come from that time. With small

0"

29

variations, but within the same general organizational model INTA was
created in Argentina in 1957, INIAP in Ecuador in 1958, INIA in Mexico
in 1960, ICA in Colombia in 1962, INIA in Chile in 1964, and EMBRAPA
in Brazil. More recently, in the 1970's, four additional national
institutes appeared in the scenario: IBTA in Bolivia, ICTA in
Guatemala, INTA in Nicaragua, and INIA in Peru (Trigo, Pineiro, and
Ardila, 1979).

The organizational model for Latin American national agricultural
research institutes was inspired in the experiences generated during
more than one hundred years in the U.S. Land Grant Colleges. At the
core of the lessons learned from those experiences is the principle
that education, research and extension are parts of the same
continuum, and that it is not convenient to have them separate. This
feature of the model was difficult to implement. Most Latin American
institutes are devoted only to research, with exclusion of the other
two functions.

The closest example to replicate the model seems to be the
Colombian case. There the Escuela de Graduados de Tibaitata, which
has been conceptually linked to the Universidad Nacional de Colombia,
was operationally attached to the Instituto Colombiano Agropecuario
(ICA). Within its multiple responsibilities, ICA has the other two
functions: research and extension. Another example, but far more
different, took place in Mexico, where three separate organizations
corresponding to the three functions had a common home for their
central offices in the same physical place. They are the Colegio de
Postgraduados de Chapingo (education), the Instituto Nacional de

Investigaciones Agricolas, INIA, (research), and Extensidn Agricola

30

(extension). This arrangement, however, did not last, and at present
each organization has its own home at a different place. The Peruvian
is a third case, where at present the establishing of a national
research institute (INIA) is taking place in close connection with the
Universidad Agraria La Molina.

c. International Centers. A further phenomenon which appeared in
the Latin American scenario of Western S&T institutionalization was
the “invention" of the international agricultural research center
(IARC). According to Wortman and Cummings this model
grew out of the accumulated experience of many organizations and
individuals over half a century. Examples of organizations which
provided inspiration for the model are the Experiment Station of the
Hawaiian Sugar Planter's Association, the Pineapple Research Institute
of Hawaii, and the Central Rice Research Institute of India.
Typically, international centers are production oriented, and use a
multidisciplinary problem solving approach to remove the more critical
restrictions of biological nature which limit productivity increases
of a single crop.

Wortman and Cummings (1978, pp. 131-133) point out eight
attributes as the unique characteristics of international agricultural
research centers: 1) each center is governed by an autonomous, self-
perpetuating board of trustees; 2) international centers are staffed
by the best talent available, regardless of nationality, and they are
being increasingly accepted by nations of diverse ideologies as
apolitical institutions; 3) they are mission-oriented (the mission
being to help nations develop their capabilities to increase

agricultural production); 4) a center can initiate a new program

31

guickl ; 5) international centers have unusual flexibility in
responding to national requests, flexibility which enables them to
arrange cooperative programs on a scientist-to-scientist basis as well
as an organization—to-organization basis; 6) training of national S&T
personnel is central to the conception and operation of international
centers, and they can tailor training to fulfill specific needs; 7)
international centers support institutionalization of S&T by keeping
their staff acquainted with advanced work done anywhere in
relationship to their research programs, by systematically collecting
and preserving the associated knowledge and materials, and by making
the knowledge and materials available to all organizations and
persons; and 8) international centers are financed by a growing nunoer
of national and international assistance agencies financial support
which allows them to have continuity of research work.

The system of international centers began with the International
Rice Research Institute, established in the Philippines in 1960. In
1978 this system was integrated by twelve centers, three of them
operating in Latin America: the Centro Internacional de Mejoramiento
de MaTz y Trigo (CIMMYT) founded in Mexico in 1966, the Centro
Internacional de Agricultura Tropical (CIAT) in Colombia in 1969, and
the Centro Internacional de la Papa (CIP) in Peru in 1972. Financial
support for the entire system is coordinated through a consortium of
donors, the Consultative Group on International Agriculture Research
(CGIAR). This organization evolved out of a meeting at Belagio,
Italy, in 1969, where leaders of some of the world's major assistance
agencies decided to concert their resources for the expansion of the

activities of the four existing international centers, and for

32

establishing new ones as might seem appropriate (Wortman and Cummings,
1978; IADS, 1976, 1977, 1978, 1979, 1980, 1981).

Wortman and Cummings see international centers not as isolated
corps, but as components of a worldwide research network, which has
three interlocking components: national research and production
systems, the international agricultural research centers, and research
organizations of more developed countries (Wortman and Cummings, 1978,
pp. 127-135).

During the last five years, the system of international centers
has been enlarged with the creation of two new organizations,
specialized in providing assistance to nations interested in
developing indigenous S&T capabilities for agriculture: the
International Agricultural Development Service (IADS) initiated by the
Rockefeller Foundation in 1975, and the International Service for
National Agricultural Research (ISNAR), established by the CGIAR in
1980. IADS, with headquarters in New York, offers assistance in
agricultural planning, strengthening research systems, training
personnel, and implementing production programs (IADS, 1981, preface).
ISNAR, with headquarters in the Hague, Netherlands, has as its
ultimate goal "to enable developing countries to plan, organize,
manage, and execute development-oriented research more effectively
from their own human, natural, and financial resources, supplemented
where necessary by effectively incorporating external resources"
(Gamble, 1981, p. 6). As emphasized in the objectives of both
centers, the development of human capabilities is one of the main

limitants in the building of indigenous S&T organizations.

1'. .—_. _
I

33

0. Developing human S&T capabilities for Latin American agriculture.

1. Training opportunities. Training for providing Latin
American agriculture with its own S&T capabilities started in a
systematic way in the 1940's with several different approaches to
educational system development: in-house in-service training,
followed by postgraduate training at Master's and PhD levels in MDC's,
mainly carried out in U.S. universities and to a lesser extent in
Europe. At the same time the building of agricultural higher
education organizations was initiated in the region. Many such
programs began parallel to the efforts of building national
agricultural research institutes, and heavily relied on external
sources for financial support. It was expected that internal monies
would gradually replace external aid, to eventually gain a relative
self-reliance (Trigo, Pineiro, and Ardila, 1980).

Perhaps the oldest Latin American center for degree postgraduate
training is the graduate school of the Instituto Interamericano de
Ciencias Agricolas (IICA) which has operated in Turrialba, Costa Rica
since 1946. The building of IICA started in 1940 as an initiative of
the Organization of American States (also known as the Pan-American
Union), which appointed an Inter-American Committee on Tropical
Agriculture. This committee obtained the consulting assistance of a
U.S. Department of Agriculture Committee, which, after visiting Latin
Ameican countries, suggested Turrialba as the appropriate site for the
field headquarters. Naranjo (1966) says that it is hard to
distinguish the graduate school from the total structure of IICA
during its initial years. In his opinion, IICA basically appeared as
an organization devoted to teaching and research in agriculture and

related fields, in order to encourage and promote the development of

34

agricultural sciences in Latin American countries. In 1946 the IICA
Graduate School started with eight students from Bolivia, Colombia,
Costa Rica, Guatemala, and the Dominican Republic.

Associated with the national research institutes, the following
graduate schools have been in operation: the Colegio de Postgraduados
de Chapingo in Mexico, the ICA-Universidad Nacional Programa de
Graduados (PEG) in Colombia, and the graduate school of INTA in
Argentina. A tradition of agricultural postgraduate studies existed
and has continued in the Universidad Agraria La Molina in Peru. In
addition, postgraduate programs have been established during the last
two decades in universities of Brazil and Chile.

These efforts for developing human 5&1 capabilities have been
complemented with non-degree training programs carried out at
different organizations of MDC's and within the region, particularly
at international agricultural research centers. At IARC's there is a
wide range of training opportunities from in-service internships to
post-doctoral fellowships (see Table 1, based on Fernandez, 1977).
Although some of these opportunities involve fieldwork for Master's
and PhD theses, it may be classified as non-degree because IARC's do
not grant academic degrees.

2. Training follow-ups. Concerns wih effectiveness and
relevance of such training efforts have resulted in several follow~up
studies. Emphasis on topics has ranged from attitudinal studies to
cognitive aspects of work performance, but a dominant theme has been
the so-called "brain-drain." No studies were found'in relationship to

the participation of former students in third cultural networks of S&T

in Latin America.

35

Table 1. Training programs offered by CIMMYT, CIAT, and CIP, in 1976,
1977 or 1978.

 

International center

 

 

Program CIMMYT CIAT CIP
In-service internships in research X X X
In-service internships in production X X
In-service internships in management

of station operations X X
Short internships on special research

or production skills X
Summer students trainees (undergraduates) X
M.S. thesis research X X X
PhD dissertation research X X X
Advanced research fellowships (visiting

research associates), post M.S. X X X
Postdoctoral fellowships X X X

 

Adapted from Fernandez, 1977.

36
In 1966 Naranjo followed up the opinions of former IICA graduate

students for a period of nineteen years, regarding the value and
achievement of the school's educational objectives, academic program
and related aspects. Naranjo's most important findings were, a) a
high value was awarded by graduates of IICA's school; b) graduates
recognized that IICA has met its educational objectives in an
appropriate way; c) Turrialba school strengths consist of an ideal
combination of teaching and research, endowed with good library
facilities and faculty with PhD degrees and research experience; d)
most of the graduates following doctoral studies in U.S. universities
have received recognition for their "Magister" degree granted by IICA;
and e) job dissatisfaction among the IICA graduates as expressed by
them was very low.

Another follow-up study about former students' opinions on their
training program was carried out by Morris (1974) about CIAT's
training programs. The main conclusions were, a) there were problems
in the process of trainees' recruitment, mainly associated to lack of
and vagueness of pre-interview information for potential candidates
and institutions; b) the selection process presented weaknesses in
terms of the role that trainees' home institutions played in selecting
candidates; c) the area of planning training to fit specific needs of
individuals and institutions was considered one of the major
weaknesses in CIAT programs by both staff and trainees; d) many
trainees expressed their need to have more involvement with small
farmers in trainees' learning activities; e) trainees were in general
very satisfied with the length and level of their training; f)
practical experience was by far the most valuable aspect of the

training period as pointed out by trainees; g) in estimating the

"1

37

degree to which they believe they are applying their training in their
work, 59% indicated extensively, 27% said moderately, 10% thought
little was applicable, and 4% none; and j) the trainees assigned a
high value to the opportunity they had while at CIAT to meet other
research workers from throughout the region; 93% indicated they have
since had some contact with such new acquaintances and 74% of these
were to some extent professional in nature.

Swanson's study (1974) compares the training approaches of CIMMYT
and IRRI, as well as a set of aspects of the training aftermath of
professionals who attended research and production programs at those
two centers. Swanson's main concern is the internalization of the
normative structure of science by foreign students who are trained at
MDC's universities. His central research question is: Can the
international agricultural research centers offer the type of training
that will result in trainees working to solve local production
problems? Swanson found that a training which emphasized knowledge-
generating types of reseach resulted in a job performance
characterized by a relatively high output of technical research
papers; in contrast, a training which stressed technology development
led to trainees' performance more oriented toward the production
processes and problems. He concludes that serious organizational
bottlenecks may occur in national research and extension programs due
to the lack of appropriately trained personnel.

With reference to degree training in the United States, Fienup
and Riley (1980) conducted a worldwide survey among agricultural
economists from developing countries. Their main conclusions were,

a) students from LDC's represented 30 percent of the total of new

entrants in U.S. university graduate programs in agricultural

38
economics over the period 1969-1977, the largest absolute numbers
being from Asia and Latin America; b) 82 percent of the survey
respondents were still living and working in their native regions, but
the percentage ranged from 73 percent in Asia to 85 percent in Africa
and 90 percent in Latin America; c) U.S. graduate training in
agricultural economics is highly regarded by LDC alumni and their
employers; d) the traditional professions such as law, medicine and
engineering are typically better developed and more prestigious in
Latin America than those in agriculture; e) although there is a
sizeable number of trained agricultural economists in Latin American
countries, significant capabilities for graduate-level training
currently exist only in Brazil, Chile and Mexico; graduate training
programs that were developed in several other Latin American countries
during the 1960's have either ceased or have become very weak. Based
on their own and former studies, Fienup and Riley (1980, p.87) suggest
adOpting three broad goals which provide a general orientation to the
profession of agricultural economics in relationship to international
develOpment:
1) To substantially increase LDC professional
capacities to train agricultural economists at
the M.S. level and to conduct research and related
extension programs on agricultural and rural
development programs.
2) To strengthen the U.S. university faculty
capabilities to train LDC and U.S. professionals
for effective international development work.
3) To establish and maintain professional networks
of LDC, U.S., and other deveoped-country .
agricultural economists that would facilitate

collaborative programs of research, education, and
public service.

39

3. PROTAAL, an in-depth, comprehensive research project on S&T
in Latin America. A major research project focused on the nature of
the technological processes involved in Latin American agriculture has
been pursued by IICA with the cooperation of the Ford Foundation, the
United Nations Development Program (UNDP), and the International
Development Research Center (IDRC) of Canada. Its name is Proyecto
Cooperativo de Investigacidn sobre Technologia Agropecuaria (PROTAAL).
It started operations in January, 1977, and has developed conceptual
and methodological frameworks; conducted in-depth case studies in
Colombia, Argentina, and Peru; and published preliminary reports.
PROTAAL looks at technological processes as endogenous phenomena,
inherent to the society in which they develop. Its aim is to
understand the problem and subsequently to suggest policies,
organizational models, and actions to contribute to agricultural
development in the region.

Most of PROTAAL initial work has been concerned with the
development of indigenous S&T human resources for Latin American
agriculture. In view of the fact that the researcher plays an
extremely important role in S&T activities, PROTAAL, at this initial
stage, has emphasized the study of the development of human resources
for S&T, as a necessary condition for putting in operation all the
other infrastructural components.

The main findings of the project so far increasingly suggest that
a serious deterioration of the national reserch an éxtension
institutes is generalizing in the region, with the exception of

Brazil, Venezuela and Mexico. Associated to such deterioration there

“_u_ 1

40

are many factors in one way or the other tied to the historical
background of those societies and the resultant duality of their
economies. Such duality has impact on the research institutes via
changes in national policies, sometimes emphasizing commercial
agriculture and at other times, traditional, subsistence production.

Three components of the problem have been indentified; a) from
the time in which the national institutes were established, starting
in the late 1950's, a change in emphasis of national development
policies has occurred, passing from importance of technology for
commercial agriculture to importance of structural changes such as
land redistribution (agrarian reforms), and from these reforms to
rural development focused not on production but on welfare; b) the
biases of science-based technology, which is neutral only in some
components, e.g., new varieties, but which at the level of whole
production systems favors commercial production; and c) the rise of
alternative source of technology at the national level, tied to
' agricultural input industries, organizations of commercial
agricultural producers, and agribusinesses.

Deterioration of the national agricultural research institutes is
manifest, according to PROTAAL findings, in the decreasing social
value of research, e.g., lack of political support, decreased budgets,
and so on. This complex of forces has affected the development of
human S&T capabilities in three ways: a) the postgraduate schools are
ceasing to exist (see Table 2); b) the national institutes are losing
their human resources by changes to non-research activities or

migration to other organizations and countries; and c) there is a high

41

Table 2. INTA, ICA.and Universidad Agraria de La Molina: Personnel
who initiated postgraduate studies in general and in the
national programs (l965-78).*

 

 

 

 

 

INIA I§A_ LA MOLINA

Started National Started National Started National

YEAR studies _program studies program studies program
1965 15 2 ll - 15 2
1966 22 3 22 - 13 2
1967 34 - 24 8 27 l
1968 28 10 35 7 24 6
1969 23 2 4O 8 l6 2
1970 21 5 51 15 20 4
1971 39 21 37 8 10 3
1972 24 8 110 59 10 1
1973 24 16 96 52 ll 5
1974 4 - 57 40 13 5
1975 l - 53 51 7 l
1976 2 - 28 23 6 2
1977 l - 7 - 1 -
1978 - - 4 3 - -

 

*Source: Trigo, Pineiro, and Ardila (1980, p. 10).

42

rotation in research posts, with the consequent lack of continuity in
research work and loss of opportunity for researchers to develop more
experience and professional maturity (Trigo, Pineiro, and Ardila,
1980; Ardila, Trigo y Pineiro, 1980; Trigo, PiNeiro, and Ardila, 1979;
Ardila, Trigo, Torres, Pifieiro, and Rincon, 1980; Ardila, Torres, and
Trigo, 1980; Ardila, Reichart, and Rincon, 1980; Pineiro and Trigo,
1977).

Trigo, Pineiro, and Ardila (1980) see IARC's as a contextual
element which may have had influencein the deterioration of national
agricultural research institutes. Three possibilities are, a) the
rise of IARC's implied a loss of international financial support to
national institutes; b) the perception of IARC's by users as an
alternative source of S&T may have contributed to loss of political
support to national institutes; and c) IARC's may be affecting the
research priorities as defined by national institutes, which do not
necessarily coincide with the research priorities of IARC's.

The above-mentioned studies throw doubts on the viability of
institutionalizing indigenous and endogenous science capabilities for
the agriculture of the region. Figures about the evolution of
agricultural research expenditures show increases from (in U.S.
dollars) 73 million in 1965, to 146 million in 1971, and to 170
million in 1974. These figures allow one to think that agricultural
research is receiving increasing priority as a basic function of these
societies. Nevertheless, Latin America continues to rank last among
the regions of the world in terms of research expenditures as a
percentage of the value of agricultural output (see Table 3, from a

study by Boyce and Evenson, 1975).

43

Table 3. Agricultural research expenditures and expenditures on
research as a percentage of the value of agricultural
product, by region (l965-74).*

 

 

 

 

Total annual Percentage of total
expenditures research expenditures
(millions of 1971 to value of

constant USS) agricultural product
Region 1965 1971 1974 1965 1971 1974
Western Europe 407 671 733 1.4 2.0 2.2
Eastern Europe & USSR 627 818 861 1.5 1.7 1.8
North America & Oceania 806 1203 1289 1.9 2.6 2.7
Africa 114 139 141 1.3 1.4 1.4
Asia (excluding China) 356 610 646 1.2 1.8 1.9
Latin America 73 146 170 0.6 1.1 1.2
Total 2383 3588 3841

 

*Source: J. Boyce and R. Evenson (1975).

44

Another perspective for looking at these phenomena has been
emerging, based on studies of the sociology of science and on Useem's
sociological construct about third cultural networks of science and

technology.

E. Social structures and third cultural networks of S&T.

l. Notion of network. Network is a word commonly used in the
daily language. It can suggest a structure of cords, threads, or
wires that cross each other at regular intervals and are knotted or
secured at the crossings. It can indicate a set of lines or channels
that interlace, like a network of highways, a network of rivers, or a
network of veins. Sometimes this word is used to describe a system of
electrical conductors in which electricity flows between certain
points by more than one path, e.g., a network of telephones. Today it
is common to hear about networks of radio, networks of television,
networks of computers, and so on.

With regard to scientific research, the words "research network"
may refer to a variety of entities such as networks of papers,
networks of authors, networks of libraries, networks of organizations,
or networks of experiments. In this dissertation the concern is with
networks of persons in which the nodes are individual researchers or
groups of researchers, and the links are relationships among
reseachers. With this meaning, the words "research networks“ are used
here interchangeably with the words "networks of reseachers."

2. Research networks: organization among S&T professionals.

The study of social organization in science has been a research area

in which sociologists have focused particular interest during the last

45
twenty years. The first empricial studies, to the knowledge of the
author of this dissertation, came from the early 1960's when Derek de
Solla Price (1963) started looking for evidence about the existence of
organization in science. He studies mutual citations which appeared
in published research papers. Later studies concentrated not on the
formal communication system of science (the scientific journal and
related printed media), but on the multitude of personal contacts
through which researchers keep themselves up to date on the work of
their colleagues.

Diana Crane (1972) found evidence about the existence of some
groupings with clear functions in the production of research. She
regarded those groupings as ”natural" organizations and called them
"invisible colleges," following Price's use of the term in former
writings. Price had taken the term from writings of the early Royal
Society of London. An author attributes the first use of the term
“invisible college" to Sir Robert Boyle. According to Greene (1969)
Boyle used that term to distinguish his own group from the visible
Gresham College.

a. Social structure among researchers. The conventional
conception of science saw research as the activity of an individual
pursuing one-sided dialogue with Nature. However, studies from
sociology of science suggest that research is rather a corporate
activity. Starting from a macro perspective, three concepts help look
at the social structure among researchers: research community,
research networks, and research groups (see Figure 1).

i) Research community. This is the all-encompassing concept.

The research community comprises all those persons who occupy research

positions in their societies as a central part of their professional

46

 

Research
community

 

Research
network -——— .p

 

 

 

groups ‘
("Invisible
colleges")

I
Research \ .
\
‘ \

 

 

Fl'gure 1. Research communitys research network, and research grOUPS-

47

activities. There is no distinction in this definition about "basic"
or "applied" researchers. It simply allows one to identify in a given
society who are and who are not members of the research community.
However, the research community is not a uniform entity, but a
multitude of loosely defined subsets. A central feature of the
research community which will be treated in the final section of this
chapter is the tendency of its members to transcend national
boundaries and to create third cultures in terms of their cognitive
interests and cultural backgrounds.

ii) Research networks. This is a concept which looks for a more
workable specification of subsets of the research community.
Subdivisions based on distinctions between science and technology,
between basic and applied research, or between disciplines and
specialities are too ambiguous to provide a coherent focus for
studying social organization among researchers (Mulkay, 1977). Such a
focus is provided by much smaller sets of reserchers, who are
concerned with a fairly narrow range of closely related problems (Law,
1973). These sets are conceptualized as problem networks, within
which basic processes of innovation and social control in science take
place (Mulkay, 1972; Crane, 1972). Under this conceptualization, the
basis of establishing the boundaries of a reseach network is a
cognitive goal: the research problem. Given that a reseach problem
may be perceived in different ways and from different perspectives, it
seems more appropriate to talk about sets of closely related problems
rather than about a well defined specific research problem.

Therefore, research networks are defined here as subsets of the
research community integrated by researchers whose attention is

focused on closely related problems.

48

Diana Crane uses interchangeably the terms "communication
networks" and "invisible colleges." In this dissertation the word
network has been associated with a wider entity which includes members
of invisible colleges and people playing other roles, such as
"liaisons" and ”isolates." Thus, the term research network is used
here to communicate the idea of sets of researchers interested in
closely related problems, but without the requirement that they all
constitute a group, be aware of each other, be present in the same
place, belong to the same formal organization, have direct
relationship with each other, nor that they develop clear leadership.
All these circumstances could be present within a network but they are
not necessary conditions. The term “invisible colleges" is used here
as synonymous to "research group."

iii) Research groups. Several studies have identified a
tendency among the most productive and influential researchers of a
field to develop intensive communication interchanges among themselves
(for instance see Price and Beaver, 1966; Garvey and Griffith, 1967;
Crane, 1972). These are members of invisible colleges, called here
research groups. Price (1963) argues that such groups are an
inevitable product of the remarkably rapid exponential growth in he
number of scientists and in the size of the literature. Research
groups play prominent roles in the development of research networks
and in the institutionalization of S&T. Indeed, they provide
leadership and orientation within their own research network (Crane,
1972), they represent their field in the wider research community, and
they contribute to gain social recognition for S&T activities (Cole

and Cole, 1968).

49
Research groups are operationally defined here as "communication

groups" in the terminology of communication network analysis.
Communication groups are composed of members whose communication is
devoted more to each other than to persons outside the group, and to
group linkers who serve the function of uniting or binding two or more
groups, thus moving messages from group to group (Farace et a1., 1977,
p. 185).

iv) Network structure provides a basis for looking at the
internal dynamics of research networks. The patterned set of
relationships among researchers is what allows one to talk about a
paticular set of researchers as a network. If there are no
relationships, there is no network.

Relationships among the members of a network can be studied in
terms of different dimensions, such as power and communication.
Communication relationships among researchers have received
preferential attention from sociologists, perhaps because it is
considered that authority relationships are contrary to the
professional norms which underlie scientific activity (Hagstrom, 1965)
and scientists are supposed to advise and criticize but not to command
each other (Crane, 1972), or because in communication processes, with
the flow of messages is also conveyed a flow of influence.

The fact is that researchers behave differently as communicators,
and that different communication behaviors result in the establishment
of different sets of relationships within a network. When interaction
is repetitive over time, it is likely that some regularities or
"patterns" appear in the configuration of those relationships. Such

regularized patterns of relationships constitute the sturucture of a

network.

50

b. Cultural elements within research networks. In the preceding
section a conceptual framework was presented for looking at social
relationships among researchers. This framework regards networks of
researchers as social systems. The conceptualization is complemented
in this section with three cultural elements of research networks:
values, norms, and sanctions.

i) 231333. With reference to the above-mentioned three
cultural elements of networks of researchers, the following framework
is mainly built on the works of Kuhn (1962) and Ziman (1968). The
central point here is that values, norms, and sanctions in research
networks are highly influenced by general agreement or consensus which
is created within invisible colleges.

In his book, The Structure of Scientific Revolutions, Kuhn
provides a perspective for examining the issue of consensus in
science. He conceives the emergence of general levels of consensus in
science as a result of interaction processes among researchers who
share specific definitions about their work. Thus, members of
research groups establish criteria and standards to govern their own
research field. They constitute ”paradigms" which influence the
setting of research priorities and the selection of appropriate
techniques and procedures to conduct research in their own field.
Therefore, what is to be regarded as a significant research problem
and the decision about what is the appropriate methodology to study
it, more than an "objective" issue, is a value socially shared by a

specific group of researchers.

51

From this perspective, the dominant values in a particular
research network may be seen as negotiated and imposed under the
pressure of the most influential researchers of the network. If they
are well connected with other networks and other social systems, those
influential researchers will represent not only the views within
networks, but views predominant in other research networks, and views
about the role of science in society held outside the research
community by influential members of researcher sponsor groups,
governmental agencies, political movements, and so on.

The values shared by sets of researchers result in some degree of
consensus among them about their research activities. Consensus among
researchers is not based on the universal right to vote, the key
principle of political democracy, but in the agreements of
scientifically competent persons.

How can such a system avoid falling into a self-penetrating
pattern of doctornal orthodoxy and error? Ziman (1968) says that the
protection against a self-penetrating circle in science is the size of
the scientific community and the interlocking of the various fields so
that new critical attitudes, expertise, and ideas can always percolate
from one grouping to another, allowing for correction of local errors.

Because invisible colleges are not organized on any a priori
plan, they overlap one another in a multidimensional array, so that
many scholars find themselves at the junctions of two, three, or more
fields. The problem of creating a consensus of ideas becomes
transformed into the creation of a consensus on the personal standing

and credibility of the individuals who speak for these ideas.

52

In the following paragraph Ziman points out his view about the ‘
role of consensus in science:

What I have tried to show. . . is that the criteria of

proof in Science are public, and not private; that the allegiance
of the scientists is toward the creation of a consensus. The
rationale of the "scientific attitude" is not that there is a set
of angelic qualities that guarantee the validity of their
thought--as if they were, so to speak, well-tuned computing
machines whose logical circuits precluded them from error--but
that scientists learn to communicate with one another in such
tones as to further the consensible end to which they are all
striving, and eventually train themselves to construct their own
internal dialogues in the same language. A private psychological
censor takes over from the policeman or parent, and conforms our
behavior to social norms. But he does not keep whispering into
our ear, "Be honest, be objective," in a chorus of pious
aspiration; he says, "Have you checked for instrumental errors?
Is that series convergent? Would anyone understand that
sentence? What is the present status of that old bit of theory?"
and so on (Ziman, 1968, p. 78).

Consensus is also seen in this thesis as influencing the norms and
sanctions predominant in a research network.

ii) Norms. In his formulation about the norms of science,
Merton (1942) argued that the extension of scientific knowledge is
possible only in a community where there is widespread conformity to
four institutional imperatives: universalism, communality,

disinterestedness, and organized skepticism. Universalism requires

 

that information presented to the scientific community be assessed
independently of the personal characteristics of the source of

information; communality requires that scientific knowledge be held in

 

common, for scientific information belongs to the scientific
profession and not to those who were resonsible for its discovery;
disinterestedness requires that researchers pursue scientific
knowledge without considering their career or reputation; and

organized skepticism requires never taking results on trust.

53

Scientists should be consistently critical of their own and others'
research.
Two additional norms appear in the writings of Storer (1968) who

built upon the works of Merton and Barber (1963): rationality, which

 

requires that reason guide all scientific work; and emotional
neutrality, which requires that data are not distorted, and research
approaches are not taken in an inflexible way.

Mertonian norms of science are expressed at a very abstract and
idealized level. These norms were in general accepted uncritically
throughout the 1950's and most of the 1960's, but in recent years they
came under considerable criticism (Mitroff, 1973; Duncan and Zaltman,
1973).

Restivo (1971, p. 119) suggests that "the norms, whatever their
origin, are today part of an ideology among scientists that has
emerged concomitantly with the professionalization and
bureaucratization of scientific activities.“

Ziman (1968, p. 96) regards norms as consequences of consensus
among scientists. In this sense, the persons participating in a
research network develop and interpret their own sets of norms and
standards, which may or may not be tied to a worldwide, generalized
ideology of science.

iii) Sanctions. Mulkay (1977) says that it has become widely,
although not universally, accepted by sociologists that recognition by
competent colleagues is the basic reward within the research
community.

To admit that recognition is the main reward for doing research

does not necessarily imply that other rewards such as professional

54

promotion and increases in salaries are not within the motivation of
scientists. These may be results of higher reputation, achieved via
recognition by comptent colleagues.

Quality of research is suggested by Mulkay (1977) as the most
important single determinant of recognition. In other words,
professional rewards seem to be distributed partly in response to the
amount of information that participants supply to the research
community, but more significantly, on the basis of the perceived value
of this information.

However, the value of the information will be determined by the
consensus about what is considered “good research," "appropriate
technique," etc., in particular research networks.

With reference to agricultural research in developing countries,
Swanson (1977, pp. 338—340) regards as problematic the reward system
for producing science-based technology. Swanson suggests that a
substantial number of agricultural researchers in these countries
might give first priority to individual projects with potential for
publication in journals of more developed countries, but not
necessarily relevant to the needs of their societies. He supports his
view with the following factors: First, many research workers in the
LDC's obtained their advanced degrees at universities of MDC's, where
they were socialized in a "publish or perish" tradition which usually
rewards only theoretical inquiries. Second, economic compensations,
but not professional recognition, constitute the main reward system of
agricultural research organizations in LDC's; increased in salaries
are assigned according to bureaucratic procedures and criteria, but

not according to the type of research carried out. The third factor

55

relates to the social status of agricultural activities. Swanson
says:

in most cultures agricultural work is considered a low-

status occupation; therefore, there are no positive

social rewards to encourage a highly trained research

worker to work on practical problems--the result of

which would be of direct importance only to a peasant or

cultivator and which would not result in any significant

professional recognition. Even if individual scientists were
motivated to do this type of research, the research organization
has no effective way formally to reward such successes.

Finally, Swanson points out why he thinks that the reward system
of research in LDC's would make agricultural researchers prefer
individual projects over teamwork. To develop science-based
agricultural technology requires the cooperation of specialists in
various disciplines, but the credit accrues to the team, not to the
individual. "If this team credit is usurped by the research director
or the team leader, instead of being shared by the team members,
individual research workers will not be encouraged to work together on

future endeavors" (Swanson, 1977, p. 340).

3_ Third cultures of SBT. During the last two decades John and

 

Ruth Useem have developed a paradigm for the study of cross-cultural
interactions. The focus of their attention has been on the new social
structures and cultures which emerge out of the interchanges among
members of different social systems in our present, increasingly
complex, conflicting, and interdependent world. Those communities
which have no national or geographical boundaries have been called
"third cultural networks" by the Useems. Following are presented the
aspects of Useems' conceptualization which are relevant to this

thesis. The first part deals with the concept of third cultures, and

56

the second one makes reference to a specific kind of those cultures,
the third cultures of science.

a. The concept of third cultures refers to the patterns of
“relationships (including norms and values) created, maintained, and
shared, generic to communities of men who related their societies, or
segments thereof, to each other" (Useem, 1963). Third cultures
reflect the cultural backgrounds of their creators and impinge on
themselves and on segments of their societies:

Third cultures are created, carried and changed by

persons who are relating segments of one society with

segments of one or more societies. Although affected

by macro-level changes in the relationships between

societies and by redefinitions of the position of the

segment within a society, still each third culture has

a history of its own which can be traced. These third

cultures, in turn, impact on other segments of societies

through the roles which carriers play (Useem, Useem,

0thman, and McCarthy, 1980, p. 4.)

Implicit in the above-stated concepts are several features
inherent to the third cultural paradigm. First, the unit for cultural
interchanges between societies is the individual human being. The
carriers of "third cultures" are "individuals who have received a
modern and high education, have an occupation or a profession that is
part of the modernizing-developing nation-building institution, and
typically have been recruited as adults" (0thman, 1977, p. 6).
Sec..i. third culture creators and carriers do not act in a vacuum,
but they are "emc-*ded" in particular settings. With respect to these
two features of their cont atuaiizat..' the Useems and some of their
disciplines recently emphasized:

We discern one largely neglected yet crucial ‘omain

in most discussions, namely, recognition that world

systems, policies, and programs are carried out by

individual human beings, whose lives, scientific careers and
professional roles are embedded in particular historical,

57

economic, technological, and political settings (Useem, Useem,
0thman, and McCarthy, 198, p. 3).

Third, the interchanges among members of third cultural networks
are personalized but do not require face-to-face interaction. Fourth,
what holds these networks tegether is sharing of interests, but not
kinship or location; particular networks may emerge around a variety
of transnational interests, e.g., religion, ideology, political
movements, business and industry enterprises, science and technology
endeavors, and so on. These two aspects of third cultures arise from
"the ability of people, widely scattered around the world, to maintain
and sustain shared interests via communication systems and other ways
not necessitating continuous face-to-face contact“ (McCarthy, 1972, p.
40).

Fifth, repetitive interaction over time is an essential requisite
for network formation; this requisite implies the need for continuity
in roles which are central to the main interest of the network. It
does not imply occupying a particular position in a given
organization. However, having an organizational affiliation is
instrumental for network participation.

And, sixth, the scope of third cultures may encompass people from
two or more societies. Participants from two societies form a bi-
national third culture; if members come from a cluster of more than
two societies, they generate multi-national third cultures; and when
people from all over the world are involved, they create worldwide
third cultures.

b. The third cultures of science. Restivo and Vanderpool (1974,
p. 461), building upon a formulation previously presented by Useem

(1971, p. 14), define the third cultures of science as "the cultural

58

(including intrascientific) patterns created, shared, and learned by
scientists of different societies who are in the process of relating
their societies (or sections thereof) to each other." These authors
point out that at least three factors determine the emergence of third
cultures out of the transactions among scientists of different
societies: i) the existence of adequate media of communication; ii)
the sharing of views about "reality" and scientific goals; and iii)
compatibility in the definition of scientific activities in different
societies.

Ideally, third cultural networks of science will open avenues for
increased communication and cooperation among segments of different
societies, which hopefully may facilitate the creation of new
appropriate views and opportunities for modernization and social
progress. These new scientific communities may include mutually
influencing linkage systems between newcomers to S&T endeavors and the
centers of world science. Through these personalized connections flow
ideas, skills, research techniques, and present and future images
about humankind. ‘

c. Types of linkages. Based on their studies in South East

 

Asian countries, the Useems, 0thman, and McCarthy (1980, pp. 38-41)
have differentiated four types of links in third cultural networks of
science (extensive, intensive, nascent, and latent) and have
elaborated profiles for scientists who predominantly have each type of
linkage. They are described, in the authors“ words, in the following
paragraphs:

Extensive transnational networks. Scientists with

extensive third cultural networks are persons who have
commanding knowledge of their field and are considered to

 

59

be leading representatives within...the worldwide scientific
community. Their daily conversations touch upon the latest
problems and issues of interest in their specialty and, to some
degree, in science and technology in general. They are
nationally and often internationally recognized scientists, have
established reputations in their specialization but also
understanding of the problems in adjacent fields. They attend
foreign-held international conferences, teach as exchange
scholars or serve as guest speakers at a series of institutions
abroad. Typically they are members and fellows of foreign
scientific societies and have given papers both locally and
abroad. Their continuous communicaton with foreign colleagues is
the primary source of information for keeping up to date. They
often exchange reprints and preprints with their "significant
others" abroad in the same specialty. Scientists with extensive
international ties know what is being worked on in their field
and who is working on those projects.

Within their countries, they can be described as influentials
or brokers--those persons who are likely to influence the
direction of scientific research by controlling appointments,
promotions, and the distribution of special subsidies and awards;
they can secure admission for their best students in foreign
universities, and are instrumental in establishing arrangements
between their home university and a foreign foundation, agency,
or institution. They are "insiders" among highly mobile people
who know about each other before meeting, and hence are aware of
the norms of moving swiftly into "shop talk.“

Those with extensive networks interact with many others, but
only in narrowly defined segments of the total person. The
personal dimenson is not so much ignored as respected and used
for making realistic judgements concerning the "business" at
hand.

Intensive transnational networks. Intensive networks are both
professional and highly personalized. They occur with one or a
few scientists in foreign countries, and encompass more of the
totality of the individuals involved. They are multi-bounded
relationships, often involve parts of the "private selves," and
often include family members. Many have a long history of being
together in various places and times-~as fellow graduate
students, as first teacher-students and subsequently colleagues
in a study, as collaborators in a major research project, and as
intimate friends who meet together while attending
international conferences.

Slim transnational networks. Slim networks between our sample
scientists and fOreign counterparts refer to more tenuous and
often more protean relationshps than is true of the extensive or
intensive networks. Some are nascent ties of young scientists
with former teachers and fellow graduate students which have not
yet had time or opportunity to develop. Some are those being
developed at mid-career as individuals enter new fields of
research or administrative roles, or have more opportunities to
attend international conferences. Others are fading extensive or
intensive networks as the participants shift away from research

60

interests which were previously shared, or as changes in the
allocation of funding have contracted opportunities to carry out
research or directly interact with foreign colleagues. Other
slim relationships might be classified as latent, and could be
renewed should priorities of support change once more or
political differences betwen the nations of their citizenship
take a different turn.

No networks. The scientists with no interpersonal, third—
cultural networks include: some who have the potential of
establishing personalized ties abroad and hope to do so in the
future, but have not done so up to now because of their newness
to their academic and professional roles; others who have entered
into full-time administrative roles, or extra-scholarly pursuits
in private and public life and have curtailed their research and
active communicaton with scientists in foreign countries; a few
who for ideological-nationalistic reasons prefer to minimize
further contacts with foreign scientists, institutions, and
foundations; and some whose foreign counterparts have retired,
died, or moved into other research emphases.

 

F. The study setting.

 

This context includes overviews about four settings: 1) CIAT as
a research and training organization; 2) the food problem in Latin
America, to the solutions of which agricultural organizations and
researchers are expected to contribute; 3) four specific crops as foci
of their research; and 4) the trainees' home organizations.

1. CIAT as a research and training organization. The Centro

 

Internacional de Agricultura Tropical (CIAT) is one of the
organizations supported by the Consultative Group on International
Agricultural Research (CGIAR) with the mission of helping nations
develop their capabilities to increase agricultural production (see
Section c in the Review of Literature, pp. 30-32).

a. 951312, CIAT grew out of an initiative by the Ford and
Rockefeller Foundations, which commissioned Dr. Lowell S. Hardin (Ford
Foundation) and Dr. Lewis M. Roberts (Rockefeller Foundation) to study

the problems of agricultural productivity in the tropics of Latin

61

America, and to recommend a course of action. As a result of the
study, Hardin and Roberts submitted in October, 1966, the document "A
Proposal for Creating an International Institute for Agriculture
Research and Training to Serve the Lowland Tropical Regions of the
Americas" (CIAT, 1981).

On the 12th of May, 1967 an agreement to establish the center in
Colombia was signed between the Colombian government and the
Rockefeller Foundation. The first meeting of the Board of Trustees
was held on October 17, 1967. In 1968 CIAT's headquarters moved from
Bogota, Colombia's capital city, to a 522 hectare dairy farm provided
by the Colombian government without cost to CIAT. This farm, located
between the cities of Cali and Palmira in the Cauca Valley, has been
the main station for CIAT's operations. Research and training
activities started there in 1969 (CIAT, 1981).

b. Objectives. The center began its activities with a broad

 

focus on agricultural development, operationalized in terms of the
priorities identified by Hardin and Roberts. Two divisions for
research were established at the beginning: animal sciences and plant
sciences. Under the Animal Sciences division two programs were
formed, one with emphasis on rumiants and the other on swine
production. Plant Sciences had three programs: the Cereals Program,
which included rice, maize and sorghum; the Grain Legumes Program,
with soybeans and field beans; and the Tropical Roots Program, with
cassava, sweet potatoes, and yams. Parallel to the two research
divisions, a strong Training and Communication Program was created.
Through the years a continuous refinement of philoSOphies,

scope, strategies, and operations has characterized CIAT's evolution.

62
At present its statement of objectives is:

To generate and deliver, in collaboration with national
institutions, improved technology which will contribute
to increased production, productivity, and quality of
specific basic food commodities in the tr0pics--princi-
pally countries of Latin America and the Caribbean--
thereby enabling producers and consumers, especially
those with limited resources, to increase their
purchasing power and improve their nutrition (CIAT, 1981,
p.39).

This statement emphasizes that: a) the product of CIAT's
work is improved technology to increase production, productivity,
and quality of selected basic foods; b) CIAT's activities are

collaborative in nature, the main collaborator being the national

 

institutes; c) the tropics of Latin America and the Caribbean
constitute its geographic scope; and d) limited resource
producers and consumers are expected to be the principal
beneficiaries of CIAT's work.

c. Research approach. Current CIAT research work is

 

focused on four commodities: rice, common beans, cassava, and
tropical pastures. Although research work on these commodities
was carried out from the beginning of CIAT's operations, a sharp
focus on them arose at different points in time. Emphasis on
rice and cassava started in 1969. Formation of a program focused
only on common beans dates from 1973. The present concentration
on tropical pastures has resulted from successive approximations
to a broad set of problems which initially encompassed the areas
of animal health, animal management, and cattle production
systems. This initial approach operated under the name Beef
Production Systems Program for six years. Between 1975 and 1977

these activities were concentrated on soils problems which are

63

characteristic of the acid infertile savannas of Latin America.
The name at that time was changed to Beef Production Program.
From 1978 to the present, the focus has been geared to obtaining
low-cost, grass-legume associations fo the acid soil savannas.
To reflect this focus a new name was adopted in 1979: Tropical
Pastures Program (CIAT, 1981).

Underlying CIAT's focus on a few cr0ps is the assumption
that such strategy will enable the center to make not only better
additive contributions to generation of technology, but to serve
catalytic functions which may help overcome institutional
constraints imposed on agricultural research in the region. The
rationale is that: a) concentration of efforts in problem
solving, production oriented, interdisciplinary research will
result in more rapid progress in the development of appropriate
agricultural technologies for the region; b) more effective
research will stimulate governments to better recognize the value
of agricultural research; and c) increased value for agricultural
research will result in higher levels of funding for national
research organizations as well as higher social recognition of
the work of agricultural researchers (CIAT, 1981, pp. 24-26).

d. Training activities. Higher investments in
agricultural research would produce increased need for more and
better trained researchers. Universities are the major
contributors to increasing the number of highly educated
personnel. International centers, within their specific foci,
can also make a contribution in this respect. Training

activities have been an integral component of CIAT's philosophy

64

and strategies since its initial conception by Hardin and Roberts.
Training orientations and emphases have followed the development trends
of the center. At present training is the largest of CIAT's outreach
services. Its major goal is "strengthening the capacity of national
institutions to carry out, cooperatively as well as independently,
their key technology adaptation and transfer functions" (CIAT, 1979,
p. 7). In the implementation of this goal training activities have
focused on building "critical masses" of professionals trained in
research on CIAT's commodities, who are able to effectively conduct
multidisciplinary, team work in relevant organizations of selected
countries; the expectation has been that these researchers will
establish collaborative links within and among national organizations,
and between such organizations and CIAT (CIAT, 1981).

2. The food_problem in Latin America. An overview of some
aspects of this highly complex problem is presented as an additional
contextual reference. Three components of that problem which are
related to agricultural research are: a) the nutrition problem;

b) the production gap; and c) the heterogeneity of Latin American
agriculture as a socio-economic activity.

a. The nutrition problem. Caloric and proteinic deficits are
present in large segments of the population of Latin America. Figures
for fifteen countries show that where deficits are less severe at
least one third of the population consumes diets below minimum

"reguirements. In other countries of the region the problem

65

problem affects over two thirds of the population (Table 4).
Nutritional deficiencies are directly tied to low income or lack of
it. Improved technology by itself is not a solution of the problem of
nutrition. However, technology can have indirect effects on the diet
of low income population through its effects on prices and qualities
of food products (CIAT, 1981).

b. The production gap. Looking at aggregate figures for the
region as a whole, food production has grown consistently with food
demand in Latin America, at a rate of 3.6% per annum since 1960.
However, individual figures reveal that this has not been the case for
sixteen of twenty-one countries (Figure 2). All Latin American
countries located within the tropics are net importers of food
staples. If current rates of growth in production are projected to
1990, food deficits in all tropical countries are expected to increase
by 50%, except for Brazil and Paraguay (IFPRI, 1977; CIAT, 1981).

c. Heterogeneity of Latin American agriculture. The hetero-
geneity of agriculture in Latin America may be realized by considering
two interrelated contradictions: i) while there is a high pressure on
agricultural land currently under the plow, such land represents only
somewhere between 18 and 35 percent of Latin America's potential land
resources for agriculture (CIAT, 1981, p. 2); and ii) while small
farms represent a high proportion of agricultural enterprises, and
while small farmers are the most efficient producers in combined
utilization of land, labor, and capital (Berry, 1979), their
contribution to total agricultural output is very limited. These

contradictions are tied to the facts that: i) the cost of production

66

Table 4. Estimated percentages of the population consuming inadequate
amounts of calories in various Latin American countries

 

 

(1973).*

Below 90% of

Below Minimum Recommended

Caloric Levels Caloric Levels
Country % %
Honduras 60 50
Ecuador 70 61
El Salvador 72 61
Colombia 61 46
Dominican Republic 58 44
Guatemala 69 48
Brazil 45 31
Mexico 34 22
Jamaica 30 21
Peru 53 41
Costa Rica 34 20
Panama 51 38
Chile 33 10
Uruguay 33 20
Venezuela 56 47

 

* Data are available from the Food Balance Sheets of FAO on mean caloric
consumption, and FAO/WHO have made estimates of caloric requirements.
The distribution of the population by caloric consumption utilizes
income distribution data and the functional relationship between
calorie consumption and income.

Source: 55. Reutlinger and H. Alderman, The Prevalence of Calorie
Deficient Diets in Developing Countries, World Bank Staff Working
Paper No. 374, Washington, D.C., March 1980, pp. 27 and 28.

Taken from: CIAT, 1981, Appendix 3.

67

 

 

Demand
5 _ Growth
Rate
5" Mexico BraziJ
.Ecuador .
Peru. Nicara ua .Costa Rica
El Salva orC t 1 A
° - ra merica
4 -— BOhWfiC Co omgmao . .Guafiemala
. araguaw‘lenezuea
AhdeanC untries atl" America
on ura;
3 —
rribe n
oEountrie
.Chile
2 __ .Argentina
11—
Uruguay
0 Food
Production
Growth Rate
1 I L 1 l l
l 2 3 4 5 6

Figure 2. Growth rates of demand and production of food in the
Latin American countries (1966-77).

Source: CIAT, 1981, Appendix 10.

~

68

for infrastructure and soil correction appear to be presently too high
to support a more rapid expansion of land use at the agricultural
frontier; and ii) small farmers control only a minor portion of land
and capital resources. The picture is complicated by the high
variability in ecosystems, the increasingly young age profile of the
population, and a heavy migration from rural areas to cities where
unemployment is already a grave problem. These complexities of Latin
American agriculture impinge on reseachers' work performance. They
are expected to generate appropriate technology for both the
commercial and the subsistence sectors. Technology for the commercial
sector usually is capital intensive. On the contrary, small farmers
require labor-intensive technologies.

3. Crops as contextual setting. CIAT's four commodity programs

 

constitute another contextual setting for this dissertation. In terms
of research each commodity has characteristics of its own, linked to
production ecosystems and cultural and socio-economic traits of those
who produce them, as well as to previous research work conducted
within and outside the region in those commodities. Following are
some points of reference.

a. Rigs. The importance of rice as a basic food crop in Latin
America has been increased over the past 15 years, as a rapidly
urbanizing population has shifted from other staples to rice as a
principal calorie source. In some countries of the region, this
tendency is related to an apparent decline in per capita direct
consumption of maize (CIAT, 1981).

The increase in consumption has been largely satisfied within

the region, with imports remaining around 150,000 tons per year, and

69

intra-regional trade increasing to 360,000 tons per year, 36% larger
than the level in 1963-65. About two thirds of the production growth
has come from an increase in area (mainly in the upland sector) and
the remainder from increases in yields (mainly in the irrigated
sector) (CIAT, 1981).

There is a large diversity in the ways in which rice is produced
in Latin America, varying between intensive irrigated systems to
extensive or subsistence upland systems at the frontier. The main
ecological factor determining a production system is the rainfall
patterns and possibilities of improved water management. In terms of
output, rice is predominantly a larger farm crop, but it is practiced
by many operators of small holdings; farms of five hectares or less in
Latin America represent about one third of the total number of
production units under non-irrigated conditions, and about 13% of
those operated under irrigation (Scobie and Posada, 1977).

Rice is perhaps the commodity with a longer research tradition,
building upon a knowledge base fostered by IRRI and other research
centers in Asia as well as in other parts of the world. Compared to
the other three commodities emphasized in CIAT training activities,
rice enjoys acceptable, developed research programs in national
organizations. Ten countries are classified for this thesis within a
group with more facilities to carry out rice research: Argentina,
Brazil, Colombia, Costa Rica, Ecuador, Guatemala, Mexico, Peru, and
the Dominican Republic. .

b. (Begg. The common bean is the most important grain legume

for direct human consumption in the world. It is a traditional food

70

in Latin America, providing the cheapest source of protein and a
relatively inexpensive calorie source (CIAT, 1981).

Bean production in Latin America is characterized by low
profitability and high risks, which result from the interaction
between biophysic, political, and socio-economic factors. For
example, some common traits of bean production are: high variation in
yields in relation to climatic variations, high susceptibility to the
attack of insects and pathogens, high price fluctuations, low economic
incentives to producers, and production carried out mainly by poor
farmers on poor lands and with poor water supply.

There is a wide range of cropping systems in Latin America for
bean production, going from bush mono-culture to climbing beans in
association with maize. Bean production has been stagnant, with
yields generally declining. A slight expansion in area has occurred,
but it has only compensated for yield declines. A large gap exists
between experimental and farm productivity, experimental yields
ranging between three and five tons/hectare, when farm yields average
only 0.6 tons/hectare (CIAT, 1981).

The complex of factors associated with the crop make it
difficult to generate appropriate technologies for bean production.
Staff members of CIAT consider that all of the major environmental and
biological constraints to increased production are researchable (CIAT,
1981, p. 71). National research programs have existed in Latin
America for years, but progress in technology generation has been
limited. The programs of Mexico and Colombia have made considerable
advances. Other countries of the region which are classified for the
purpose of this dissertation as having facilities to do bean research

are Argentina, Brazil, Chile, Guatemala, and Peru.

71

c. Cassava. At present this is a basic source of calories for
large segments of the rural population in 26 tropical countries of the
world. For Latin America it is very popular in Brazil, Colombia, and
Paraguay. Consumption in urban areas is more limited because of its
high perishability and high marketing margins as a fresh product.
However, an important potential is attributed to cassava for human
consumption as well as for industrial utilization, such as production
of a wheat flour substitute, feed concentrates, ethanol, and starch.
Fulfillment of such potential is dependent on more agricultural
research, and on development of infrastructure capacities for
processing and marketing. Cassava strengths reside in its remarkable
efficiency in carbohydrate production, adaptation to soil and water
stress, indeterminate harvest period, high yields per unit of land and
labor, and compatibility in association with a variety of crops (CIAT,
1981).

In Latin America, cassava crops can be found under a wide range
of ecological conditions, from the lowland tropics with high
temperatures to the cool highlands of the Andean cordillera.
Production modes range from the slash-and-burn system of the Amazon
jungle, to planting cassava as an introductory crop in pasture
establishment, and to small farm multi-crop systems, frequently in
association with a legume or maize. It has comparative advantages in
areas where there are major constraints on the growth of other craps.

Systematic research on cassava does not have a long historical
tradition. National research programs in Latin America are very new,
most of them in operation for less than ten years and still in the

formative stage. Nevertheless, experimental yields have considerably

71a

increased. CIAT has had experimental yields up to 80 tons/hectare
which in comparison with average farm yields of 10 tons/hectare
suggests high potential for raising farm productivity (CIAT, 1981).

d. Tropical pastures. In the preceding paragraphs, three

 

single crops--rice, beans, and cassava—~were presented as contextual
frameworks for training. The case of tr0pical pastures is different
in two respects.

First, tropical pastures have not been a continuous focus during
the period of training analyzed here. They constitute a more recent
stage in a research area which started by working on beef production,
and since then has been making successive redefinitions of its
phi1050phy, aims, approaches, and priorities.

And second, what is called here "tropical pastures" is not a set
of already defined species. It is, rather, a research focus aimed at
developing low-cost/low-input technology for the production of
pastures in extensive tropical regions of Latin America, which are
characterized by acid, infertile soil conditions. The rationale
behind this selective focus in tropical pastures is that such
technology would provide economically and ecologically sound
alternatives for the utilization of those regions. Food producers and
consumers would be the main beneficiaries. Improved production of
pastures would result in increased animal production and productivity.
These developments would allow the agricultural frontier to expand in
the tropics of America, and to release more fertile Jands, today
utilized in animal production, for the production of other crops.

CIAT estimates that the tropical and subtropical areas of

America have nearly a billion hectares of underutilized savannas and

72

forests, 75 of which are occupied by acid and infertile soils. These
areas have great agricultural potential since they have abundant solar
radiation with adequate rainfall and favorable temperatures for
extended growing seasons. Topography and soil physical conditions are
also generally favorable. The current average stocking rate in the
acid savannas is 0.12 animal/hectare. This can potentially be
increased more than tenfold. In addition, beef production per
animal/year could be more than doubled. However, the available
technology is not profitable in these areas. Furthermore, another
source of constraints in such regions is infrastructure; but beef
produced extensively does not require a well developed infrastructure
for inputs or marketing outputs (CIAT, 1981, pp. 93-102).

Similar to the case of cassava, there is no long research
tradition in Latin America in tropical pastures. CIAT considers the
main constraint to expanding research collaboration in this subject as
the absence, in most countries, of strong pasture research programs
working in areas with acid infertile soils. Existing programs tend to
concentrate on more fertile soils and areas where animal production
has expanded in the past, and where more of the cattle population is
still located (CIAT, 1981, p. 111). Only four countries of the region
were classified for the purpose of this thesis as enjoying research
facilities for tropical pastures: Argentina, Brazil, Colombia, and
Venezuela.

4. The trainees' home organizations. A growing number of
organizations in Latin America send members of their professional
staff to CIAT to advance their training in agricultural research. The

main user of these training opportunities is the national agricultural

73

research institute. In addition, a diversity of governmental and
private organizations are showing interest in fostering the research
competencies of their personnel. Included among them are agricultural
extension agencies, banks and financial organizations, agricultural
universities, associations and federations of crop producers, and a
number of private agro-businesses related to production and marketing
of inputs such as seeds, pesticides, and fertilizers.

A large number of organizations (182 private and public
agencies) scattered acrosss 26 Latin American and Caribbean countries
have utilized CIAT's opportunities for preparing research personnel.
This figure suggests a growing interest in the region for establishing
agricultural research capabilities in relationship to the four
commodities of CIAT's mandate. However, the number of persons trained
is not evenly distributed in those agencies. Twenty-one out of the
182 sponsoring organizations account for half of the personnel trained
in research during the period 1969-79. This concentration reflects
CIAT's policy of building "critical masses" of researchers in selected
organizations, and it may be also interpreted in the sense that a
small cluster of Latin American organizations have implemented more
assertive policies of human resource development for S&T in the fields

of agriculture (Table 5).

-1 ...

PLEASE NOTE:

Page 74 is lacking in number
only per school. No text is
missing. Filmed as received.

University Microfilms International.

75

Table 5. Former CIAT research trainees, distributed by number of
trainees per sponsor organization (N = 783).*

 

 

 

Number of
trainees per Organizations Trainees
organization Number % Number 1
l - 3 130 71.4 204 26.0
4 - 8 31 17.1 172 22.0
9 - 23 15 8.2 206 26.3
24 - 48 6 3.3 291_ 25.7

 

182 100.0 783 100.0

 

* This distribution indicates the organizational origins of
trainees, and not their present locations.

‘i‘ “Win—Ti,
i

CHAPTER III

STUDY DESIGN

This chapter describes how data reported in this thesis were
collected and analyzed. The study population is defined and
delineated by size, training content, length of training, facilities
to do research at sponsor organizations, and a set of characteristics
of its members (age, gender, marital status, education, and country of
origin). The rest of the chapter presents details about instruments
and procedures applied in gathering data for the analyses of human

resource inventories and networks of researchers.

A. Strategy.

The general strategy for data collection consisted of three
activities: 1) to gather all relevant data available at CIAT on the
study population; 2) to collect data for the question of inventories
of human resources through a mail census to all Latin American
organizations which sponsored trainees at CIAT during the period 1969-
1979; and 3) to obtain data for network analysis by means of a mail
questionnaire sent to all members of the study population who had done

research after training and whose addresses were known at the time of

the survey.

76

F‘

77
B. Delineating the population.

1. Definition. The study population is defined as composed of
all former research trainees from Latin America who completed their
training between December 31, 1969, and December 31, 1979, in any of
the four CIAT commodity programs, and for whom there were records in
the CIAT training files.

2. Data gathering. Data for delineating the population were
obtained from both the files and staff of CIAT. The total data
collected from these two sources made it possible to: a) determine
the size of the population; b) describe its distribution by the three
independent variables of interest; and c) characterize it in
relationship to age, gender, marital status, formal education, country
of origin, and year of training completion. In addition,
organizational reports obtained during this phase of the study
provided the basic materials for writing a description of the training
setting.

3. Population size. The determination of the population size
consists of identifying and counting those professionals who fit the
population definition. Two types of persons are excluded: a) all
students whose names are in the files because they were financially
supported by CIAT, but whose actual training was carried out at
educational organizations other than CIAT; and b) all postdoctoral
fellows because they differ from the study population in status and
activity at CIAT. The size of the population so determined consists
of 783 Latin American agriculturists. It represents 62.2% of the total
number of 1259 persons who had completed their training at CIAT

between December 31, 1969, and December 31, 1979.

78

4. Strata for independent variables. The following strata were
constructed in order to classify the population according to training
content, length of training, and facilites to do research in the
fields of training at sponsor organizations.

a. Strata for training content were readily available in the
records. They constitute the four CIAT commodity programs: rice,
beans, tropical pastures, and cassava.

b. Strata for length of training were constructed on the basis
of the following patterns which have emerged over time in the CIAT
training programs: i) In general, periods of training up to two
months correspond to structured short intensive courses, mainly
carried out within classrooms, and usually focused on a commodity,
e.g., rice, or a field of research within a commodity, e.g.,
multiplication of cassava germ plasm by means of meristems. ii)
Training periods from two months to six months usually are less
structured than short courses, take place almost totally at labs,
greenhouses, and reserch fields, and are designed for gaining
specialization in a discipline or field of research within a
commodity, e.g., cassava phytopathology, rice plant breeding. iii)
Periods longer than six months commonly involve the pursuit of a
research project under the guidance and supervision of a researcher of
the staff of CIAT; and sometimes the projects are connected with
fieldwork for M.Sc. and PhD theses.

c. Strata for support facilities to do research at home
organizations were more difficult to construct and apply to this
population. Several possibilities were examined for macro-data, for

instance, investment in agricultural research in each Latin American

79

country, or resources available for agricultural research at Latin
American organizations. The data found were insufficient to construct
a comprehensive set of strata inclusive enough to classify all members
of this population. In addition, those data were not directly related
to the commodities around which research training is organized at
CIAT. Hence, the decision was made to try using the judgement of
competent researchers who know the research facilities available at
agricultural organizations in Latin America. Two researchers from the
staff of each CIAT commodity program were asked to classify all
countries of the region according to two categories: those having
more facilities to do research in the fields in which CIAT has had
training programs, and those with less facilities. These two
categories are referred to below as the "more" and the "less"
facilities. The category "less" does not mean inadequate facilities.
Experts who were asked to make appraisals were selected with the
criterion that they had been personally involved in CIAT international
cooperation activities, so they would have first-hand knowledge about
the research facilities available at national organizations in their
specialized fields.

There was little disagreement amongst the two separate sets of
appraisals about the classification. When it appeared, disagreement
was resolved by contacting other experts and discussing with them the
issues until an acceptable level of consensus was reached. Experts
did not judge other fields of research but only their own. For
example, cassava researchers classified the countries according to
their judgement on the facilities available to carry out cassava

research, but did not judge research facilities for other commodities.

 

80

The resultant classification is in Table 6. Twelve countries of
the region were classified as having "more" facilities for at least
one of the four CIAT reseach commodities. These 12 countries are:
Brazil and Colombia, which have "more" research facilities for all
four commodities; Argentina and Mexico, for three commodities;
Guatemala, Peru, and the Dominican Republic, for two commodities; and
for only one commodity, Cuba, Costa Rica, Chile, Ecuador, and
Venezuela. All other Latin American countries were classified as
having "less" facilities to carry out research in the fields in which
CIAT offers research training.

5. Population distribution in conjunction with each of the
independent variables (Table 7). Out of the 783 research trainees,
173 (22.1%) carried out their training in rice, 245 (31.29%) in beans,
172 (21.97%) in tropical pastures, and 193 (24.64%) in cassava.
Although differences in the proportions of trainees who were exposed
to each training content are not large, a slight concentration is
noticeable in the number of persons in beans research. Rice and
tropical pastures training takes place approximately in the same
proportions, and the proportion of training in cassava was in between
the other three contents.

With respect to length of training, almost half of
the population (49.43%) had a short period of training of two months
of less, about one third (32.43%) had an intermediate period of more
than two up to six months, and slightly more than one sixth (18.14%)

had a training period longer than six months (Table 8).

81

Table 6. Classification of Latin American countries by
availability of facilities to do research in
areas in which training was carried out at CIAT.

 

Research facilities

 

Rice Beans T. pastures Cassava
Country More Less More Less More Less More Less

 

Antigua.................X..........X...........X.........X
Argentina...........X..........X...........X.............X
Belize..................X..........X...........X.........X
Bolivia.................X..........X...........X.........X
Brazil..............X..........X...........X.........X....
Colombia............X..........X...........X.........X....
Costa Rica..........X..............X...........X.........X

Cuba....................X..........X...........X.....X....
Chile...................X......X...............X.........X
Ecuador.............X..............X...........X.........
El Salvador.............X..........X...........X.........
Guatemala...........X..........X...............X.........
Guiana.................
X

0000......x00000000000 0.0000000

0X
Haiti00000000000000.0000x0000000000x0000000.... .00000000
X

X
X
Honduras................
Jamaica.................X..........X...........X.........
X
X
X

00.0.0.0..X00.0000.000 000.0000.

XXXXXXX

MeXiCO.0000000000000X00.00000.0X000000000.000.0 00000X0000
Nicaragua00000.00.00.000x0000000000x.0.000.... X000000000

Panama....0000000000000.X00000000.0x00000000000X000000000

Paraguay00.00000.0000000x00.000....x0.00000.00

Peru....000.0.0.0..0x0000000000x..000000000000 0000000..

XXXXX

X

Puerto Rico.00000000000.x..00000000X.00000000.. 000......
Dominican RepUb]ic00x0.00000.00.0.0x00000000000x0000x0000.
Trinidad0000000000.000.0x0000000.00x00000.00.00X000000000

Urugua‘y0000000000.00000.X00000.0... 0......00..X0....0000

X
venezue1a00.0.000.000.00x00000.0...X0000000X000,-.00000000

xxx

82

Table 7. Distribution of the study population by training
content, length of training, and facilities to do
research at home organizations.

 

 

Independent variable N %
Training content 783 100.00
Rice 172 22.10
Beans 245 31.29
Tropical Pastures 172 21.97
Cassava 193 24.64
Length of training 783 100.00
Up to two months 387 49.42
More than two months,
up to six months 254 32.44
More than six months 142 18.14
Facilities to do research 783 100.00
"More“ facilities 514 65.65

"Less" facilities 269 34.35

 

83

 

 

 

 

 

 

 

 

 

 

 

 

Table 8. Number of research trainees from Latin America,
distributed by content and length of training.
Length of Training
2 months 2<m36 >69months TOTAL
Content N % N % N % N %
1 2 3
Rice 73 9.32 81 10.34 19 2.43 173 22.09
Beans 135 17.24 84 10.73 26 3.32 245 31.29
Pastures 36 4.60 67 8.56 69 8.81 172 21.97
Cassava 143 18.26 22_ 2.81 .28 3.58 193 24.65
TOTAL 387 49.42 254 32.44 142 18.14 783 100.00
Table 9. Latin American research trainees distributed by
training content and facilities to do research at
home organizations.
Facilities
More Less TOTAL Ratio
Content N i N i N % More/Less
Rice 133 16.99 40 5.11 173 22.10 3.3
Beans 150 19.16 95 12.13 245 31.29 1.6
Pastures 87 11.11 85 10.86 172 21.97 1.0
Cassava 144 18.39 '49 6.25 193 24.64 2.9
TOTAL 514 65.65 269 34.35 783 100.00 1.9

84

As for facilities to do research at home organizations, the
proportions were two to one: about two thirds (65.65%) of the
population came from countries classified as having more research
facilities in the research area in which training was carried out, and
approximately one third (34.35%) came from countries classified as
having less facilities (Table 9).

Tables 8 and 9 show how proportions for length of training and
research facilities distribute in relationship to training content.
Training in rice or cassava is more concentrated in those countries
classified as having more facilities to do research: the proportion
more/less, as indicated in Table 9, is three to one in rice and
cassava, while it is two to one for the whole population. For the
other two commodities, in contrast, the proportions more/less are
below the general proportion, two to one, for the whole population,
being almost even (one to one) for tropical pastures. As for length
of training, cassava and beans programs have concentrated in short
training periods, rice programs have focused on short and intermediate
training with a slight tendency toward the intermediate ones, and
tropical pastures have emphasized intermediate and longer periods with
a slight tendency toward longer training (Table 8).

6. Profile of the study population in terms of age, gender,

 

marital status, education, and country of origin. Most of the
research trainees are relatively young males, married, and have an
education at the 8.5. level (Table 10). People under 34 years of age
represent sixty percent of the population. That nine out of ten are
males reflects the historical fact that females have been entering

agricultural research careers in Latin America only recently. With

F

85

Table 10. Distributions of the study population in terms of
age, gender, marital status, and education.

 

Cumulative
Characteristic N % N %

 

 

Age in years

 

Less than 30 182 23.24 182 23.24
30 - 34 290 37.04 472 60.28
35 - 39 196 25.03 668 85.31
40 - 44 69 8.81 737 94.12

More than 44 46 5.88 783 100.00

Gender

Female 52 6.64 52 6.64

Male 731 93.36 783 100.00

Marital Status

 

Single 283 36.14 283 36.14
Married 500 63.86 783 100.00
Education

Less than B.S. 25 3.19 25 3.19
8.5. or equivalent 678 86.59 703 89.78
M.S. or equivalent 74 9.45 777 99.23

Ph.D. or equivalent 6 0.77 783 100.00

 

86

respect to marital status, the proportion of married to single people
is almost two to one. A few former CIAT trainees (3.19 percent) do
not have an education at the B.S. level or the equivalent; most of
them come from countries where agricultural education at the
university level has been established only recently. But the bulk of
the study population (86.59 percent) has formal education at the B.S.
level. One out of ten has studied beyond the B.S. level; most of
these held a Master's degree at the time of training or were at CIAT
carrying out fieldwork for their theses; some of the persons
classified at the PhD level had not graduated by the time of their
training at CIAT, but were in the last stages of their doctoral work.

The ages and educational levels predominant in this population
suggest that most of its members were entering research careers when
they went to CIAT for their training; perhaps many were in their first
job; and a high proportion of former CIAT research trainees probably
occupied low, or at most, intermediate positions in their
organizations before training.

With reference to country of origin, a high proportion of the
study population came from two countries, Brazil and Colombia. In
fact, Brazil sponsored nearly one fourth and Colombia one fifth of
former CIAT research trainees. Seven countries sponsored percentages
between four and six percent: Ecuador, Mexico, Peru, Guatemala,
Venezuela, Honduras and the Dominican Republic; these seven countries
altogether account for 36.14 percent of the population.

There are seven other countries, each one sponsoring between one
and four percent of former CIAT research trainees: Costa Rica,

Bolivia, Cuba, El Salvador, Chile, Panama, and Argentina; they

87

supported almost 17 percent of the population. The rest of the Latin
American countries (ten out of 26) account, in total for less than
five percent of former CIAT reseach trainees (Table 11).

The concentration of CIAT research training in Brazil and
Colombia during the past decade may be associated to country science
policies and easy access to CIAT. Brazil gives high priority to the
development of indigenous scientific capabilities for its productive
sectors. Since the 1960's Colombia has been actively engaged in
preparing human resources for agricultural research. The circumstance
that Colombia is the host country for CIAT activities has facilitated
the access of Colombian agriculturists to CIAT's training programs.

Other factors which might explain the uneven proportions
presented in Table 11 are: lack of awareness in some countries about
the research training opportunities available at CIAT; less relevance
of CIAT's work for non-tropical countries, e.g., Argentina and
Uruguay; preference of science policy makers and professionals of some
countries for degree training or for the educational facilities .
offered by more developed countries; and low or no priority assigned

to agricultural research in other countries.

C. Inventories of researchers: Data collection and analysis.

A population survey was preferred over a probabilistic sample.
Two considerations made advisable this decision. First, the nature of
the research question (inventory formation) seemed to be more amenable
for a census, particularly in relationship to data ahalysis. And
second, criteria for designing an adequate sampling schedule for this

population were not available. For instance, the author was unaware

r? ‘ifi'iLoi-w V

88

Table 11. Research trainees from Latin America distributed
by country.

 

Cumulative

 

 

 

Country N % N %
Brazil 185 23.63 185 23.63 '
Colombia 147 18.77 332 42.40
Ecuador 52 6.64 384 49.04
México 47 6.00 431 55.04
Peru 44 5.62 475 60.66
Guatemala 36 4.60 511 65.26
Venezuela 36 4.60 547 69.86
Honduras 35 4.47 582 74.33
Dominican Republic 33 4.21 615 78.54
Costa Rica 27 3.45 642 81.99
Bolivia 27 3.45 669 85.54
Cuba 21 2.68 690 88.12
El Salvador 19 2.43 709 90.55
Chile 16 2.04 725 92.59
Panama 13 1.66 738 94.25
Argentina 10 1.28 748 95.53
Nicaragua 7 0.89 755 96.42
Guiana 7 0.89 762 97.31
Belize 6 0.77 768 98.08
Haiti 5 0.64 773 98.72
Paraguay 4 0.51 777 99.23
Uruguay 2 0.25 779 99.48
Puerto Rico 1 0.13 780 99.61
Antigua 1 0.13 781 99.74
Jamaica 1 0.13 782 99.87
Trinidad 1 0.13 783 100.00

 

89

of previous assessments of several factors likely relevant to
formation of inventories of human resources for agricultural research
in Latin America, e.g., country of origin, organizational affiliation,
date of training, among others.

1. Objectives. The main objective of the census is to gather
quantitative data about formation of human resource inventories for
agricultural research with basis on the personnel trained at CIAT. In
addition, the census provides an opportunity to fulfill two
instrumental needs. First, to legitimize with the national sponsor
organizations, the study and the investigator who carried it out.
Second, to generate a mailing list required for the subsequent stage
of the study.

2. Instruments. The instrument for the main objective of the

 

census consists of a check list and its instructions. The check list
is designed in a matrix format (See Appendix B). It has at the top
the name and year of training completion of a particular professional.
Years from 1969 to 1979 are the headings of the matrix columns.
Information requested for each particular professional during each
year after training is presented in short sentences in the matrix
rows. Meanings assigned to such sentences are explained in the
instructions.

By means of these census instruments the investigator asks
directors of organizations to provide the requested data by checking
inside the cells of the matrix when the answer to each particular
question is positive, and not checking when it is negative. 50,
answers are nominal in the form "yes," "no," with each answer
representing one year. Census instructions are presented in a letter

format.

I

90

Instructions ask for responses to three types of questions with
respect to after-training performance of the person whose name and
year of training completion is at the top of each matrix check list:
a) whether or not that professional worked for the sponsor
organization after training, in any kind of activity (row 1 of the
matrix check list); b) whether or not that professional did research
after training in rice, beans, tropical pastures, or cassava (rows 2-
5); and c) whether or not that professional worked after training
doing research on species other than the ones mentioned in literal b,
or performing research related roles, such as instructor of other
researchers, research administrator, or research consultant (row 6).

Below the matrix check list space is provided for comments or for
reporting lack of information about that professional. For building
the mailing list directors are asked to write the present address of
the professional on the back of the check list.

An example showing how to fill out the census instrument is
presented in the instructions. A request is also made for completing
the appropriate forms in the case that a professional did not return
to the sponsor organization or resigned after training, but for whom
pertinent information is available.

The instrument for legitimizing the study and the investigator of
country-based agricultural organizations in Latin America consists of
a personalized letter sent by the Director General of CIAT to
directors of all the organizations which sponsored agriculturists for
research training at CIAT. Spanish versions of the census instruments

are in appendices A and B.

-n' V

 

_._ _

.F'

91

3. Tactics for data collection. Census instruments are pre-

 

tested, revised, and tested again until gaining evidence that they
work satisfactorily. Such testing is carried out with the cooperation
of Colombian organizations which operate in Cali and Palmira in the
vicinities of CIAT. Census materials are arranged in sets and put
inside mail envelopes, one per organization or sectional when
appropriate.

Each set consists of the letter by the Director General of CIAT,
the census instructions, a set of check list forms with the names of
the professionals whose data were requested, and an addressed return
envelope. The possibility of sending pre-paid return envelopes was
examined but not implemented in view of the large number of countries
included in the census.

During the testing of instruments, it was found advisable to send
census materials to a more local level of each organization, because
it increased the possibility of getting the specific data requested
for each former CIAT trainee. Given that most agricultural research
organizations in Latin America have their headquarters office in the
country's capital city, and sections in different places of the
country, it was decided to send a set of materials to the central
office only, mainly in terms of legitimization of the study.
Overlapping of names was avoided by first trying to allocate data
requests at the local level and then sending the remainder to the
central office. .

In this way, approximately three hundred sets of materials were
sent, corresponding to 182 organizations and their divisions. The

sending of the first round of census materials took one and a half

92

months, from the beginning of July to the middle of August, 1980. A
recall was carried out by sending the materials again to those
organizations whose answers had not been received. Materials were
accompanied by a small card making reference to the original request,
the likelihood that it got lost in the mail, and the importance of
including such data in the study.

More than ninety percent of the organizations responded. A
complementary tactic was to ask people at that time with CIAT about
those professionals whose data were missing. Staff members of CIAT as
well as participants in training and conferences provided data about
after-training performance of former trainees whom they personally
knew. Eventually, after a screening procedure, there were complete
data for 580 former CIAT trainees, accounting for 74 percent of the
study population.

4. Analysis of inventories. The type of analysis of inventories
presented in this thesis is analogous to those of accounting
techniques. In this sense, this part of the study follows a
methodology similar to the approach of the PROTAAL studies (see pp. 39-
43). However, the scope of the data is different, and consequently
specific definitions for the analytical tools were required.

In each case of PROTAAL studies the focus is a single
organization. The formation of inventories of researchers is examined
throughout the history of the organization under study. All
researchers who entered the organization are included in the analysis,
irrespectively of the location of their training. In contrast, in
this thesis the focus is on professionals who have been employed by

many organizations, and who have carried out some portion of their

93

research training at a specific place, the Centro Internacional de
Agricultura Tropical.

Analytical tools applied for this analysis are: final
inventories, initial inventories, annual net balances, trained
personnel, migration, efficiency in the formation of inventories, non-
returnees, drop-outs, retention and stay of trained personnel, and
formation and retention of inventories of research experience.
Following are the definitions for each of these analytical tools, as
well as the pertinent relationships between them.

a. Effectiveness in inventory formation. Effectiveness of the
region in forming inventories of human resources for agricultural
research is indicated by the tendencies--increasing, stagnant,
decreasing--and absolute values of final inventories (FI). These
inventories are calculated by adding (or subtracting, when negative)
annual net balances (ANB) to initial inventories (II).

FI = II + ANB

The number of trained persons available at the beginning of each
year forms the initial inventory of the immediately preceding year.

Annual net balances reflect the movements of personnel, e.g., the
persons who enter and leave the organizations (or research on their
fields of training, on CIAT's commodities, or research in general,
according to the perspective under analysis) during a year. They are
calculated by subtracting from the personnel trained by CIAT each year
(TP) the number of researchers who migrate (M) during that year.

ANB = TP - M
The trained personnel in a given year is defined as the number of

persons who completed their training in research at CIAT up to

94

December 31 of that year. When trained personnel is higher than
migration, annual net balances are positive, and formation of
inventories will increase. When trained personnel is equal to
migration, net balances are zero, and formation of inventories will be
stagnant. And, when trained personnel is lower than migration,
balances are negative, and formation of inventories will decline.

b. Efficiency is defined as the percentage which final

 

inventories represent with respect to the cumulative number of trained
personnel. It is calculated by the formula:

Final inventory

EffiCiency = Cumulative trained personnel

 

c. Migration. The component of inventories which diminishes
efficiency is migration (M). It is defined as consisting of those
persons who, after training, do not return to work for their sponsor
organizations (or to do research in their fields of training, on CIAT
commodities, or on research in general, according to the perspective
analyzed), plus those researchers who leave their organizations
(fields of training, etc.). The first group is called here "non-
returnees" (NR), and the second category is named "drop-outs" (DO).
Migration is computed by the formula:

M = NR + 00.

d. Retention and stay. Retention (R) designates the number of

 

persons who, after training, have continued working for their sponsor

 

organizations (fields of training, etc.). Retention rates are the
percentages which retention absolute values represent with respect to
the number of people trained during a given year. (Retention absolute
values are the figures occupying the main diagonal in Tables E-l to

E-4, Appendix E.)

95

This study is focused on the analysis of inventory formation
over the period 1969-79, and not on comparing figures for individual
years. Therefore, retention is not presented here on a separate
annual basis, but cumulatively; figures corresponding to a given year
incorporate retention in the preceding years.

Stay indicates the total number of persons who have not migrated
up to a given year. Retention figures are not enough for exploring
the question, "For how long have professionals stayed in their
organizations and done research after training?" Retention is a
subset of stay. Its complement is the number of persons who, after
training, reentered their organizations (fields of training, etc.),
worked there for one or several years, and later dropped out.
Consequently, "stay“ up to a given year is computed by adding to the
figures of retention corresponding to that year, the number of persons
of the preceding training years who still have not dropped out. (For
the arrangement of data presented in Tables E-l to E-4, in Appendix E,
the procedure of computing "stay" consists of adding the totals of the
columns, starting from left to right.)

In addition, the length of the interval after training needs to
be considered. It is not equal for all trainees, provided that their
training took place at different dates. The size of this interval
might influence tendencies of stay in different directions. For
instance, a positive correlation might exist between the size of the
interval and the magnitude of drop-out, the underlying force being a
propensity of people to move with the passing of time. In other

words, the higher the size of the interval, the higher the probability

96
of drOpping out. From the other side, a propensity to stay might also

exist. It seems reasonable to think that the longer a person has
stayed practicing a professional activity or working for a given
organization, the more well established that person would become in
that activity or organization, and correspondingly the higher would be
his/her propensity to remain working on that activity or for that
organization. The preceding considerations made advisable not to give
a single rate of stay for the whole period analyzed, but to present
rates annually, according to the size of the interval between the year
of training and 1980.

e. Research experience. Retention and stay are closely related

 

to One variable of much importance in scientific research:
experience. In its essence, experience is generated cumulatively by
doing. An unrefined indicator of experience is the number of years a
person stays doing a given activity. This is an oversimplification
because it assumes that every person is devoting 100% of their work
time to that activity, and because it also assumes that one year of
experience of person A is equivalent to one year of experience of
person 8, which might not be the case. However, for the purpose of
getting a notion about what occurs in this sample with regard to
experience formation and retention, this indicator is considered
adequate.

The following analysis of formation and retention of research
experience is based on five concepts: potential experience (PE),
actual experience (ER), and efficiency in experience retention (EER).
None of these refer to the number of persons, but to the number of
person-years. One person-year is the stay of one person during one

year in a given organization or activity.

97
This analysis of experience is not conducted on an annual basis

because the year-by-year figures do not produce more information about
trends than what is provided by the series of retention and stay.

This is so because the data about number of persons are the same used
for preparing those tables; the only difference is that in the
analysis of experience such figures are weighted by the size of
intervals between training years and 1980.

To summarize the figures about former trainees' performance with
respect to the activity of research, the following relationships are
applied to findings on non—return to research, drop-out of research,
and retention for the activity of research. The difference between
efficiency in retention and 100 percent represents the lost efficiency
corresponding to the cumulative effects of migration. The difference
between efficiency in stay and retention (and between experience
formation and experience retention) gives the losses corresponding to
drop-out. And the difference between migration and drop-out is the
loss of efficiency corresponding to non-return.

5. Relationships between stay and two independent variables.
Two of the three independent variables of interest for this study are
analyzed in their relationships to stay. These variables are length
of training and research support facilities at former trainees' home
organizations. The third variable of interest, training content, is
not analyzed because of a methodological problem related to the
circumstance that emphases on training have not been equal at
different points in time for the four CIAT commodity programs.
Consequently, average lengths of stay are not directly comparable

between the corresponding four training contents.

98

The same four perspectives examined for the analysis of
inventories are applied to the relationships between stay and
independent variables. Therefore, “stay" will refer to the average
number of years during which former trainees have worked for their
sponsor organizations, fields of training, CIAT commodities, and
research in general. Data have been processsed through analyses of
variance which were conducted within each perspective and training
content. They do not apply to comparisons between perspectives or
between training contents.

a. Stay and length of training. An assumption in the design of

 

training programs at CIAT with different durations has been that
length of training is positively correlated with post-training
performance. The rationale is that not only with more time is it
possible to acquire more information and develop new skills, but that
an "enduring" effect of training occurs as length increases. The
intent of this study is not to examine in detail how length of
training affects research perfonmance, but to "check" in an
exploratory way the presence or absence of the relationship with
regard to stay as a single indicator of a dimension of research
performance (stay).

Length of training is defined as the time from physical arrival
at CIAT to departure from CIAT, measured in number of months. It is
grouped in three categories: a) short training, when the duration was
up to six months; b) intermediate training, when it took more than two
months up to six months; and c) long training, for programs longer

than six months.

99
b. Stay_and research support facilities. The degree of

 

facilities to do research in the fields in which former trainees
carried out their training is a structural factor supposedly very
influential in post-training research performance. It is assumed that
stimuli as well as constraints may arise from the degree of facilities
available to a person for continuing to actively engage in research.
In order to examine the presence or absence of a relationship
between stay and research facilities, the Latin American and Caribbean
countries which have sent professionals for training to CIAT were
classified in eight groupings, two per commodity: those with more
facilities to do research in each commodity, and those with less

facilities (see pp. for details about this classification).

0. The researchers' networks: Data collection and analysis.

Data for the question on network participation was also
designed to be collected from the whole population. This decision was
made in regard to the nature of this type of analysis. In general,
current network analysis procedures are better suited to population
studies than to samples (Farace and Mabee, 1980, p. 384); attempts to
develop sampling techniques for network analysis have been advancing,
but they still are at an exploratory stage (Granovetter, 1976). The
population for this part of the study is defined as composed of those
former CIAT research trainees who, after training, engaged in researCh
even for a brief period.

1. The instrument for collecting network data is a
questionnaire (a Spanish version is in Appendix E) on the interchange

of research with colleagues and personal data.

100

a. Interchange of research. Former trainees are asked to write

 

names, organizational affiliation, and country of those colleagues
with whom they seriously interchanged views and resources about
research work and professional interests during the last year.
Interchanges are asked at four levels: within respondent's
organization, or country, within Latin America, and in the rest of the
world. They were also asked to indicate the intensity of each
interchange in frequency and depth. Frequency of interchanges is
reported by checking one of two categories: often, or not often.
"Often“ meant several times during the year. "Not often“ meant once a
year or less. Depth of interchanges is indicated by checking one of
two categories: deep, or superficial. In order to record such data,
an empty table is provided. A hypothetical example is presented about
how to answer the network question.

b. Personal data. Five items are asked about the respondent:

 

i) name; ii) institution where currently working; iii) country where
currently working; iv) number of years of experience doing research;
and v) current research area, e.g., rice, beans, tr0pical pastures,
cassava, other.

2. Tactics for data collection. Instruments for network data

 

were pre—tested at Michigan State University. In Colombia they were
tested again with persons carrying out their training at CIAT by that
time (May to July, 1980). Although the item about research
interchanges seems complex when presented as a research technique,
respondents usually do not find difficulties in answering it. That
was the original experience at Michigan State University, confirmed

later during its testing in Colombia.

101
There was some concern about the use of the Spanish version of

the network questionnaire for Brazilian pe0ple, whose native language
is Portuguese. Instruments were translated, but during the testing it
was realized that more difficulties appeared when using both languages
than when using only Spanish. Therefore, Spanish was the language
used for data collection.

Attached to the network questionnaire was a personalized letter
(see its Spanish version in Appendix D). It followed the same format
of the letter signed by the Director General of CIAT and was mailed
with the census. In the case of the network questionnaire the letter
was signed by the investigator who carried out the study, jointly with
the coordinator of the CIAT training programs.

Mailing of network instruments is dependent on receiving census
data, not only because lack of trainees' current personal addresses,
but also because network analysis results are relevant only for those
people who performed research related roles after training. Mailing
started very slowly by August, 1980. Due to time and financial
constraints, it was decided to analyze only the network data available
at the end of January, 1981. After a screening procedure, 207
complete questionnaires (38 percent) resulted out of the 542 which
were sent to those former trainees who performed research related
roles after training, and for whom addresses were known. At the
moment of writing this dissertation approximately 400 answered
questionnaires have been received.

3. Network analysis. A persistent interest in.analyzing social

 

phenomena in terms of networks has resulted in a set of concepts and
research techniques called network analysis. It has received

contributions from a diversity of disciplines such as anthropology,

102
sociology, social psychology, demography, geography, mathematics,

communication, and others (Asch, 1975, pp. 26-45).

In general, concepts and analytic techniques are built on Barnes'
idea of network (1954) as a set of points, some of which are joined by
lines. The points are people, or sometimes groups, and the lines
indicate which pe0p1e interact with each other. When the aim is to
decipher network structure, the apprOpriate techniques must be
relational in nature (Farace and Mabee, 1980; Rogers, 1981). A
technique of this type with its corresponding set of concepts was
elaborated by Richards (1975) and packaged in a computer program
called NEGOPY.

Network data for this dissertation were processed by the NEGOPY
program. Following are pertinent concepts and definitions based on
Richards' description of that computer program (1975).

The points representing persons in the network are called ppdes,
and the relationships between them links. Whenever a person reports a
relationship with another person, there would be a link between the
corresponding pair of nodes.

Links may be directional or undirectional. Directional links
leave Open the possibility of unreciprocated relationships.
Undirectional links imply reciprocity. Directional links which are
reciprocated and have equal weight imply symmetry in the relationship.
Lack of reciprocation or unequal weight in directional links implies
assymetrical relationships. .

a. Participants and non-participants. The analysis of the

 

network by NEGOPY starts by classifying all nodes in terms of two
general roles: participants and non-participants. Participants are V

the nodes that take part in the bulk of interchanges with other

103
participants. The non—participants include all the nodes having

either no connection or only minimal connection to participants.

b. Third cultural character of networks. Once participants in

 

the network are identified, the next step in this analysis is to
examine the third cultural character of linkages among participants.
This is not made by NEGOPY. This is accomplished by studying the loci
of incidence and types of linkages of participants.

i) Locus of incidence. If researchers have links beyond their

 

organizations and countries this is an indication that their
interchanges are not totally domestic, but transnationality exists in
some pr0portion of those connections.

ii) Types of linkages. Three of the four of the Useems'

 

conceptual categories for classifying and enumerating types of
networks (pp. 59-60) are selected for guiding the analysis of types of
linkages: latent, nascent, and intensive ties. They are
operationalized for this study in terms of two dimensions of intensity
of interactions: frequency and depth.

At the operational level, latent links are characterized by
person-to-person interactions occurring frequently and SUperficially.
Nascent links correspond to more frequent but still limited in sc0pe
interactions. And intensive links are those which are deep in nature;
two sub-categories of intensive personalized ties are further
differentiated here: those which do not occur very often, labeled
intensive links of sub-type one; and those interactions which, in
addition to being deep in nature, take place frequently; the latter
are called in this study intensive links of sub-type two. An
assumption implicit in this entire operationalization is that the same

person can have all types of linkages; with some colleagues the

104
relationship can be of a latent type; with others, it can be nascent;
and so on for the intensive types.

Finally, the intensive nature of transnational linkages needs to
be examined for gaining evidence about the third cultural character of
networks. If transnational linkages are predominantly nascent or
latent, then third cultures are not very likely to emerge within those
networks. This is checked by cross-tabulatng locus of incidence and
type of linkage.

c. Network structure. The social structures emerging out of the

 

networking activities are identified by classifying participants and
non—participants in more Specific network roles (see Figure 3 taken
from Richards 1975, p. 7). Non-participants include four types of

roles: a) isolates type 1 are nodes which have no links whatsoever;

 

b) isolates type 2 or attached isolates are nodes having only a single

 

link, and thus their participation in the network is very restricted;
they may, however, function as sources of information and other
resources if they have links outside the network; c) members of

isolate dyads are pairs of nodes linked to each other, but without

 

connections to the rest of the network; and d) if there is a subset of
nodes with minimal connections, the subset will be a tree structure
composed of isolates and tree nodes; the isolates will be the nodes at
the ends of the structure, and will have only single links; the t:ee_
22922 are the other nodes in the structure.

Participants include three types of roles: a) group members,

 

which are nodes with more than fifty percent of their linkage with
other members of the same group; group members who are linked to
members of other group(s) constitute a special subcategory called

bridges; b) liaisons are those nodes which have very intensive

24

‘ Liaisons: 19, 20.

 

 

/
\ /
\
\ /
\ l 8
) l
GROUP A 3 " \ 11
\
' \
’ \ GROUP B
’ \
/ 1‘ 9
/
/ 0
4 / \\ 1 /
\ \ ‘ I / / 19 \ \ \ //

\‘N ’/
6 ..

 

_‘-— _--

I/ 18 13 ‘ \ 22 23

GROUP C \

\

\ 17 1
1

‘ 15 1

Figure 3. Network roles. (Taken from Richards, 1975, p. 7).

 

 

 

Participants
Group members. Group A:

1, 2
Group C: 12,

, 3, 4, 5. Group B: 6, 7, 8, 9, 10, ll.
13, 14, l , l , l , 1

Others: 21.

Non—participants
Isolate type 1: 27.
Isolate type 2: 23, 24.
Isolated dyad: 25, 26.
Tree node: 22.

106
interchanges with members of several groups without being members of
any of those groups; and c) type lggheggl are participants in the
network which are not group members or liaisons.

To be called a group, a set of nodes must satisfy the following
criteria: a) more than half of the communication of the members of a
group is with each other; b) each member is linked to all members on a
direct or indirect basis; c) no single link nor member can be
eliminated and have the group break apart; and d) each group should
consist of at least three members.

4. Relationships between participation and three independent
variables. These relationships are tested by means of chi-square
tests. This test is appropriate because of the nominal nature of the
structural categories (participation, non-participation; specific
network roles). When relationships are supported by statistical
significance in the tests, additional more detailed examination is
carried out by cross-tabulating each independent variable with the
pr0portion of group members, which represent the more salient role in
a research network. (Liaisons are also of great importance in a
research network, but they usually are very few, so comparisons are
only rarely feasible). Another analysis of relationships is made by
computing the number of links per person at ;ach locus of incidence,
and then cross-tabulating loci of incidence with independent variable

strata.

CHAPTER IV
Findings of the Study, Part I:
INVENTORIES OF HUMAN RESOURCES FOR AGRICULTURAL RESEARCH

This chapter focuses on the return of former CIAT trainees to
their organizations and their length of stay in the activity of
research. It examines the question of, to what extent are the
Latin American countries building inventories of human resources for
agricultural research, based on CIAT's training?

The first section deals with the central question of inventory
formation: 1) CIAT's contribution to preparing agricultural researchers
for the region; 2) effectiveness, and 3) efficiency of the region in
forming researchers' inventories based on CIAT's training; 4) migration
of researchers as a source of inefficiencies; 5) stay in and retention
for the activity of research; and 6) inventories of research experience.

The last portion of the chapter explores the presence of
relationships between stay in research and two independent variables:
length of training, and facilities to do research in the fields of

training at sponsor organizations.

107

108

A. Inventory formation.

 

1. CIAT's contribution. The contribution of CIAT to the
formation of inventories of agricultural professionals in Latin
America during the period 1969-79 is indicated by the number of
persons trained. The review of training files carried out for the
purpose of this dissertation shows that during the period a total of
1413 professionals from various parts of the world utilized the training
opportunities provided by CIAT. Most of them (89.1%) are from Latin
America. The rest are distributed almost evenly between Africa, Asia,
and Austrailia (4.6%), and North America and Europe (6.3%).

Figures corresponding to Latin American professionals are
arranged in Table 12 in terms of three types of training: a) production
and extension, b) research support, and c) research. A segment of
this last type of training (enclosed within discontinuous lines in
Table 12) is the center of attention for this dissertation. It
comprises training in research in CIAT's present four commodity
programs. A set of 783 persons, representing 62% of the total for the
period 1969-79, constitute the study population.

2. Effectiveness. The source of data for the analysis of human

 

resource inventories is a census conducted for this thesis in the 182

organizations which sent personnel to CIAT for research training. From

the population of 783 professionals who completed their training in

research between 1969 and 1979, data were obtained for 580 persons.

The average rate of response for the total period was 74% (Table 13).
The extent to which former CIAT trainees have reentered their

sponsor organizations and stayed active in research is analyzed here

from the envisioned points of view of four different sets of people:

109

Table 12. Former CIAT trainees from Latin America distributed by
training content.

 

 

Training content N %
Production and extension 201 16.0
Crop production* 57 4.5
Beef production* 46 3.7
Seed production 98 7.8
Research support 116 9.2
Biometrics and data processing 6 0.5
Communications 9 0.7
Documentation 41 3.3
Station operations 38 3.0
Others 22 1.7
Research 942 74.8
0.61.117"; Kori-modifies 783 62.2 :
: Rice 173 13.7 .
. Beans 245 19.5 ,
, Tropical pastures 172 13.7 .
, Cassava 193 15.3 ,
Other research 159 12.6
Economics** 12 1.0
Swine production*** 85 6.7
Corn 20 1 6
Weed control 42 3.3
TOTAL 1259 100.0

 

* Actual number of participants in crop and beef production is
larger because some trainees participated in both production
and research training. They are assigned to research in order
to not double count the number of participants.

** These are only the participants in training in economics who
cannot be specifically classified within any of the other research
categories. The actual number of trainees in economics is
larger.

' *** Swine training includes both research and production. It was
decided to classify this as research training. '

.g3Study population.

110

 

 

 

 

 

 

 

 

 

 

 

 

 

as mmm omm m5 map Nmp mo NNP vFF NN ch PON mm m“, NFF 4<pop
om mPN oN— mm we Fe on mm oN Pm on we om me am asap
mm FN— mNP 0N we mm «m Fm mN mm mm mm om oN mp wump
Nu _m on cop a w Nu _m m_ mm Ne um om «F NF NNmF
o“ —m we mo ov mN mo mp NF Fm NN ON cm or N asap
no Fm am mm PF o no NF m cu mN op ow m a mum_
on Pm me mm mN mN um NP m oe m N mm m? m cum?
mm MN my cor N N mm o m mu w m so 0 a mnmp
me me Pm ooF m N am a m oop m m om em up Nump
Po _P OF 00’ P F mm a m COP F P cop m m pump
No FN m— cop N N o o o o o o mm N m osm—
oop F P o o o o o o o o o cop F _ mmmp
e z c s oz w: s a2 4: s mz m: & a2 a: cam>
4<H0h <><mm<u mumaem<¢ .» mz<mm NuHm
.umumpaeoo mm: mcwcwmcu cows: cw cam» new ucmpcou mcwcpmgp
an umpanwcumwu .Amnnmcva mwmcwmcp :ugmmmmc e<~o smacom mo mpnsmm use copumpzaoa .mp upon»

111

a) directors of sponsor organizations, b) coordinators of programs at
the training organization (CIAT), c) directors of the training organiza-
tions,and d) country-based and international sponsors and science policy
makers (see p. 6). For convenience of expression, these four perspec-
tives will be respectively labeled, throughout the chapter: "sponsor
organizations," "fields of training," "CIAT commodities," and "research
in general."

The formation of inventories has increased during the period
analyzed (Table 14). Annual net balances are positive for each of the
four perspectives examined, and in all of them, increase final inven-
tories. This process, however, has been more effective in the building
of human resources for'research in general," and less effective for
building inventories in specific "fields of training." The other two
perspectives, "sponsor organizations" and "CIAT commodities," show
intermediate effectiveness.

3. Efficiency. While final inventories show effectiveness in

 

the process of forming resources for agricultural research, we can ask
the further question: How efficient have these processes been? To
examine this question, efficiency is defined as the percentage which
final inventories represent with respect to the cumulative number of
trained personnel. Efficiency is not calculated for the whole ll-year
period because the personnel trained during the first years was
relatively low in number. Efficiency rates appear in Table 15 starting
~ with the year 1972.

Two tendencies are observed. First, efficiency in the formation
of inventories for "research in general" and for "sponsor organizations"

has been over 70 percent, while efficiency in building a "critical mass"

112

Table 14. Inventory formation for the period 1969-79, showing initial
inventories, annual net balances, and final inventories for
the four perspectives: organizations, fields of training,
CIAT‘s commodities, and research in general (n = 580).

 

Years
Perspectives 1969 70 71 72 73 74 75 76 77 78 79

 

Sponsor organizations

Initial inventory 0 1 12 20 49 61 101 123 172 218 308
Annual net balance 1 ll 8 29 12 4O 22 49 46 90 121
Final inventory 1 12 20 49 61 101 123 172 218 308 429

Field of training

 

 

Initial inventory 0 l 7 14 29 4O 64 81 107 143 222
Annual net balance 1 6 7 15 11 24 17 26 36 79 89
Final inventory 1 7 14 29 4O 64 81 107 143 222 311
CIAT commodities

Initial inventory 0 l 7 14 32 44 69 88 115 158 235
Annual net balance 1 6 7 18 12 25 19 27 43 77 94
Final inventory 1 7 14 32 44 69 88 115 158 235 329

Research in general

Initial inventory 0 1 14 22 51 66 108 133 185 241 341
Annual net balance 1 13 8 29 15 42 25 52 56 100 129
Final inventory 1 14 22 51 66 108 133 185 241 341 470

 

113

 

 

 

 

 

 

 

Table 15. Efficiency in the formation of human resource inventories
for agricultural research (n = 580).
Years

Perspectives 1972 1973 1974 1975 1976 1977 1978 1979
Sponsor organizations
Cumulative trained

personnel 55 74 117 151 215 281 410 580
Final inventories 49 61 101 123 172 218 308 429
Efficiency (%) 89.1 82.4 86.3 81.5 80.0 77.6 75.1 74.0
Fields of training
Cumulative trained

personnel 55 74 117 151 215 281 410 580
Final inventories 29 4O 64 81 107 143 222 311
Efficiency (%) 52.7 54.0 54.7 53.6 49.8 50.9 54.1 53.6
CIAT commodities
Cumulative trained

personnel 55 74 117 151 215 281 410 580
Final inventories 32 44 69 88 115 158 235 329
Efficiency (%) 58.2 59.5 59.0 58.3 53.5 56.2 57.3 56.7
Research in general
Cumulative trained

personnel 55 74 117 151 215 281 410 580
Final inventories 51 66 108 133 185 241 341 470
Efficiency (%) 92.7 89.2 92.3 88.1 86.0 85.8 83.2 81.0

 

114

of professionals for establishing collaborative linkages with respect

to CIAT commodities and particular "fields of training" has been much
lower, oscillating between 50 and 60 percent. Second, from the four
perspectives--"organizations," "fields of training," "CIAT commodities,"
and "research in general"--there seems to be a slight tendency toward
decreasing efficiency in formation of inventories.

4. Migration. Tables 16 and 17 present migration broken down
in its non-return and drop-out components (raw data for preparing these
tables appear in Appendix E, Tables E-l to E-6). Examining the cumula-
tive effects of migration up to 1979, two conditions are evidenced:

a) for the four perspectives, migration in the sample studied has been an
extended phenomenon, ranging in magnitude between 19 percent, in
"research in general," and 46 percent, in "fields of training," of the
total trained personnel; and b) this process has eroded more heavily the
human resources prepared for establishing collaborative research linkages
among CIAT and counterpart organizations in the region (in "fields of
training" and "CIAT commodities"), than for preparing personnel for
country—based "organizations" and "research in general" (Table 16).

Looking at how the phenomenon has grown over the Years, some
tendencies are observed: migration is increasing from the perspectives
of "sponsor organizations" and "research in general," while it seems to
be stagnant--in spitetrfits high magnitude--for "fields of training"
and "CIAT commodities" (Table 17).

The disaggregation of migration into its two components helps
one to gain some insights about the preceding findings. Two immediate
facts surface: a) drop-out is an extended phenomenon for the four

perspectives studied, its cumulative effect up to 1979 being higher

115

Table 16. Cumulative effects of migration, drop-out, and non-return,
up to the year 1979, expressed in absolute numbers and
percentages of the total number of trained personnel
(n = 580), and as percentages of migration within each

 

 

 

 

perspective.
Drop-out Non-return Migration
Perspectives Number % Number % Number %

 

Numbers and percentages of total trained personnel

Sponsor organizations 102 17.6 49 8.4 151 26.0
Fields of training 86 14.8 183 31.6 269 46.4
CIAT commodities 90 15.5 161 27.8 251 43.3
Research in general 83 14.3 27 4.7 110 19.0

Percentages of migration

 

Sponsor organizations 67.5 32.5 100.0
Fields of training 32.0 68.0 100.0
CIAT commodities 35.9 64.1 100.0
Research in general 75.5 24.5 100.0

 

116

 

 

 

 

 

 

 

Table 17. Cumulative migration, drop-out, and non-return expressed
as percentages of cumulative trained personnel (n = 580).
Years
Perspectives 1972 1973 1974 1975 1976 1977 1978 1979
Sponsor organizations
Migration 10.9 17.6 13.7 18.5 20.0 22.4 24.9 26.0
Drop-out 1.8 5.4 5.1 8.6 9.8 12.8 15.9 17.6
Non-return 9.1 12.2 8.6 3 9 10.2 9.6 9.0 8.4
Fields of training
Migration 47.3 45.9 45.3 46.4 53.0 59.1 45.8 46.4
Drop-out 3.6 4.1 2.6 4.6 7.0 8.9 11.2 14.8
Non-return 43.7 41.8 42.7 41.7 46.0 40.2 34.6 31.6
CIAT commodities
Migration 41.8 40.5 41.0 41.7 46.5 43.8 42.7 43.3
Drop-out 3.6 4.1 2.6 4.0 6.5 7.8 11.7 15.5
Non-return 38.2 36.4 38.4 37.7 40.0 36.0 31.0 27.8
Research in general
Migration 7.3 10.8 7.7 11.9 14.0 14.2 16.8 19.0
Drop-out 5.5 8.1 5.1 7.3 8.4 8.9 11.2 14.3
Non-return 1.8 2.7 2.6 4.6 5.6 5.3 5.6 4.7

 

117

than 14 percent of the total number of persons trained; and b)drop-
out has its heaviest effects for "sponsor organizations" and "research
in general," while non-return affects more the formation of inventories
for "fields of training" and "CIAT commodities" (see absolute numbers
and percentages of migration in Table 16). Is there any tendency over
the years for the cumulative effects of drop-out and non-return?
Percentages in Table 17, graphed in Figure 4, clearly suggest that, in
general, drop-out is increasing, and non-return is decreasing or
remains stagnant,with the exception of "research in general," where
non-return has slightly increased.

5. Retention and stay, Retention (R) designates the number

of persons who, after training, have continued working for their

 

sponsor organizations (fields of trainingfi'etc.). Retention rates
are the percentages which retention absolute values represent with
respect to the number of peOple trained during a given year.
(Retention absolute values are the figues occupying the main diagonal
in Tables E-l to E-4, Appendix E).

This study is focused on the analysis of inventory formation
over Unaperiod 1969-79, and not on comparing figures for individual
years. Therefore, retention is not presented here on a separate
annual basis, but cumulatively; figures corresponding to a given year
incorporate retention in the preceding years (Table 18 and Figure 5).

For the sample studied, rates of cumulative retention follow
a similar pattern of increase-decline-increase for the four perspectives,
although decline lasts longer for “fields of training". and "CIAT
commodities." The similarity in trends from the four perspectives might

be associated with the impact that retention has on the high rates

118

 

 

 

 

 

 

 

Table 18. Cumulative retention, distributed by calendar years (n = 580).
Calendar years

Perspectives 1970 71 72 73 74 75 76 77 78 79
Sponsor organizations
Trained personnel 14 24 55 74 117 151 215 281 410 580
Cumulative retention 5 12 29 34 56 75 112 159 271 429
Rate of cumulative

retention 36% 50% 53% 46% 48% 50% 52% 57% 66% 74%
Fields of training
Trained personnel 14 24 55 74 117 151 215 281 410 580
Cumulative retention 3 8 18 22 31 37 59 92 182 311
Rate of cumulative

retention 21% 33% 33% 30% 27% 25% 27% 33% 44% 54%
CIAT commodities
Trained personnel 14 24 55 74 117 151 215 281 410 580
Cumulative retention 3 8 19 23 33 40 64 101 193 329
Rate of cumulative

retention 21% 33% 35% 31% 28% 26% 30% 36% 47% 57%
Research in general
Trained personnel 14 24 55 74 117 151 215 281 410 580
Cumulative retention 6 14 36 41 71 95 140 190 304 470
Rate of cumulative

retention 43% 58% 66% 55% 61% 63% 65% 68% 74% 81%

 

119

a. Sponsor organizations c. Research in general

%

/ 25
20 / 20 .

%
15 /\/ 15 / i?
10

25

 

 

 

 

o
10
H ./""——.
5 5 It”/’ Ikczw.
F.
197273 I4 75 If: n 18 79 197T73 74 15 lb 77 78 7‘9
%
50 l///.r”.\\\\ %
45 ' '. 9” ° 45

40 \ 4o '\6/'“\'/

35 \ 35
3o 1 30 '
25 25

I
20 . 20
15 15 /‘
10 10 /‘/

r,‘

 

 

5 5.t———A~2\‘r,,ar//

1972 73 74 75 76 77 78 79 1972 73 74 75 76 77 78 79

 

 

b. Fields of training d. CIAT commodities

0 = Total migration I = Non-return A= Drop-out

Figure 4. Cumulative migration, non-return and drop-out expressed as
percentages of cumulative trained personnel, and graphed
in terms of the corresponding calendar years (n = 580).

120

Cumulative retention (%)

100
90 Research
in general
80
ponsor
organizations
70
CIAT
60 commodities
Fields of
'50 training
40

30 .‘

20

 

10

 

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 Calendarfiyear

Figure 5. Rates of cumulative retention for the period 1970-79.

121

of non—return,_particular1y for the first half of the period. Non-
return has affected more heavily "fields of training" and "CIAT commodi-
ties;" however, its tendency to decline for these two perspectives
between 1976 and 1979 has allowed a relatively rapid recovery in the
rates of retention.

Stay is a different factor from retention. Stay indicates the
total number of persons who have not migrated up to a given year. Rates
of cumulative stay decrease until an interval of six years for"fields of
training" and "CIAT commodities," and until the seventh year for
“research in general" and "sponsor organizations." Beyond these inter—
vals, rates of cumulative stay appear to increase (Table 19 and Figure 6).
These tendencies seem to support both directions of propensities to
migrate and stay, suggested as possiblities in the study design (pp. 95-6).
During the first six to seven years after training there is a propensity
to migrate. In the other direction, data suggest that beyond six to
seven years after training there is a propensity-among those professionals
who still have not migrated--to stay in their organizations and fields of
research.

The question of length of stay does not have a single answer for
each perspective; it depends on the size of the interval after training.
Rates of cumulative stay range between 95% and 34% for the different
perspectives and intervals studied. Nine out of every ten former CIAT
trainees have stayed for at least one year doing research after training;
. but out of every three professionals who have had the opportunity to
stay at least six years, two have migrated from their'fields of training.
However, for those who have had the chance to stay nine years, rates

range between 38% and 75% according to the different perspectives (Table

19 and Figure 6). These results raise the question of the extent to

122

 

 

 

 

 

 

 

Table 19. Cumulative stay distributed by number of years after training
(n = 580).
Number of years after training,

Perspective 1 2 3 4 5 6 7 8 9
Sponsor organizations
Trained personnel* 580 410 281 215 151 117 74 55 24
Cumulative stay 531 345 210 146 98 70 41 32 15
Rates of cumulative

stay 92% 84% 75% 68% 65% 60% 55% 58% 58%
Fields of training
Trained personne1* 580 410 281 215 151 117 74 55 24
Cumulative stay 397 243 133 90 56 4O 28 19 9
Rates of cumulative

stay 68% 59% 7% 42% 37% 34% 38% 35% 38%
CIAT commodities
Trained personne1* 580 410 281 215 151 117 74 55 24
Cumulative stay 419 258 146 98 61 45 31 21 9
Rates of cumulative

stay 72% 63% 52% 46% 4 % 38% 42% 38% 38%
Research in general
Trained personnel* 580 410 281 215 151 117 74 55 24
Cumulative stay 553 363 230 170 116 86 49 4O 18
Rates of cumulative

stay 95% 89% 82% 79% 77% 74% 66% 72% 75%

 

* These figures represent the number of persons who meet the criterion
of having at least 1, 2, . .

. 9 years after training.

123

 

 

 

Stay (%)
100
90
80
Research
70 in general
60
I Sponsor
organizations
50
CIAT
40 commodities
Fields of
30 training
20
IO
1 2 3 4 5 6 7 8 9 Years after

training

Figure 6. Rates of cumulative stay, distributed by number of
years after training (n = 580).

124

which the region is building and retaining research experience with basis
on the training carried out at CIAT. The following sections address
this question.

6. Research experience. Retention and stay are closely related

 

to one variable of much importance in scientific research: experience.
In its essence, experience is generated cumulatively by doing. An
unrefined indicator of experience is the number of years a person stays
doing a given activity. This is an oversimplification because it
assumes that every person is devoting 100% of their work time to that
activity, and because it also assumes that one year of experience of
person A is equivalent to one year of experience of person 8, which
might not be the case. However, for the purpose of getting a notion
about what occurs in this sample with regard to experience formation
and retention, this indicator is considered adequate.

Cumulative figures for experience formation and retention are
presented in Table 20. They show--more vividly than figures expressed
in number of persons--the erosion that migration causes on the invento-
ries of human resources for agricultural research which Latin American
countries have been building with basis on CIAT's training.

Efficiency in retention measured in person-years (experience)
diminishes ten to fifteen percent in comparison to retention measured
in number of persons. The impact on retention of the two components
of migration is visualized in Figure 7 by means of circles, each one
representing 100 percent. The percentages which appear inside the
circles result from applying the following relationships to data from
Tables 18, 19, and 20. The difference between efficiency in retention

and 100 percent represents the lost efficiency corresponding to the

125

Table 20. Number of person-years and efficiency in the formation and
retention of research experience, up to 1980 (n = 580).

 

 

 

 

Experience formation Experience retention
Perspectives Person-years Efficiency Person—years Efficiency
Sponsor
organizations 1494 78% 1183 62%
Fields of
training 1019 53% 764 40%
CIAT
commodities 1092 57% 814 42%
Research
in general 1632 85% 1368 71%

 

cumulative effects of migration. 'The difference between efficiency
in stay and retention (and between experience formation and experience
retention) gives the losses corresponding to drop-out. And the
difference between migration and drop-out is the loss of efficiency
corresponding to non-return.

Figure 7 indicates that approximately 45 percent of the profes-
sionals who left their fields of training and research on CIAT commodi-
ties account for a loss of about 60 percent of the research experience
which would have been built during the ll-year training period. And
those who migrated from their sponsor organizations and research in
general represent losses in research experience of about 40 percent and
30 percent, respectively.

Drop-out accounts for 13 to 16 percent of such losses, and
non-return losses have ranged between 15 and 47 percent. The higher
numbers for non-return do not mean that drop—out is a phenomenon of

less importance. It should be recalled that drop-out shows a tendency

to increase, when non-return seems to decline (Figure 4, p. 119).

126

Number of persons Number of personjyears

 

 

Retention Drop
out=16%

 

Sponsor organizations

 

 

    
      
 

Retention

Retention = 40%

Fields of training

 

 

= 54%
Non-return Non-return = 47%

= 32%

 

    
    
     

    

Retention Drop-

out=1 %

Retention Drop-
= 42% out=15%

 

 

CIAT commodities
Non-return '

= 28% Non-return = 43%

Retention

Retention Dr0p-W
Research in general out=145

 

= 81%

 

 

Figure 7. Cumulative retention, drop—out, and non-return for the
eleven-year training period, measured as percentages of
number of persons and number of person-years (experience).

127

A condensation of these findings (Table 21) reveals that, during
the ll-year period studied, five percent of former CIAT trainees never
worked in research, representing a total loss of 16 percent of the
potential experience which could have been accumulated. Fourteen
percent dropped out of research toward other activities, making up a
loss of 13 percent of potential experience.

Table 21. Cumulative non-return, drop-out, and retention with respect
to research activities for the 11-year training period

expressed as percentages of trained personnel and percentages
of person-years (n = 580).

 

 

Percentages of trained Percentages of
personnel person-years
Non-return 5 16
Drop-out 14 14
Research other than CIAT 24 29
Migration within CIAT commodities 3 2
Retention in fields of training _55_ _40.
TOTALS 100 100

 

Table 21 also presents condensed figures to answer two broad
and central questions in relation to the formation of human
inventories for agricultural research in Latin America and the Caribbean
countries: What proportion of former CIAT trainees were still doing
research in 1980? What proportion still continued working in the four
areas in which they carried out their training?

The number of persons who have stayed doing research reaches

a total of 81 percent, representing 71 percent of the cumulative

128
experience for the 11-year period. This 81 percent is distributed in the
following way: 24 percent have kept doing research in areas different
than CIAT endeavors; three percent moved from their original training
area toward others within CIAT research commodities; and finally, 54
percent of the trained personnel have remained working in their fields of

training, accounting for only 40 percent of the total potential experience.

8. Relationships between staygand two independent variables.

Two of the three independent variables of interest for this study
are analyzed in their relationships to stay. These variables are length
of training and facilities to do research at former trainees' home
organizations.

The same four perspectives examined in the preceding analysis are
also studied in this section. Therefore, "stay" will refer to the average
number of years during which former trainees have worked for their
"sponsor organizations," "fields of training," "CIAT commodities," and
"research in general." Data have been processed through analyses of
variance which were conducted within each perspective and training
content. They do not apply to comparisons between perspectives or
between training contents.

1. Stay and length of training. An assumption in the design of
training programs at CIAT with different durations has been that length
of training is positively correlated with post-training performance.

The rationale is that not only with more time is it possible to acquire
-more information and develop new wkills, but that an 9enduring" effect
of training occurs as length increases. The intent of this study is not
to examine in detail how length of training affects research performance.

but to "check" in an exploratory way the presence or absencetrfthe

129

relationship with regard to stay as a single indicator of a dimension of
research performance (stay).

Length of training is defined as the time from physical arrival
to CIAT to departure from CIAT, measured in number of months. It is
grouped in three categories: a) short training, when the duration was
up to two months; b) intermediate training, when it took more than two
months up to six months; and c) long training, for programs longer than
six months.

Results of analyses of variance (Table 22 and Figure 8) support
the existence of a positive relationship between stay and length of
training. Most comparisons of means are statistically different at an
alpha level of 5%. However, the high heterogeneity of the data should
be noticed, revealed by large standard deviations.

Such high variability may be tied to individual, organizational,
and socio-economic-cultural factors which may affect average stay and are
not controlled in this study. Large standard deviations may also be
related to the way in which analyses were carried out, aggregating
training in only three categories: short, intermediate, and longer
training. A further analysis of these issues is beyond the scope of
this thesis.

A closer look at the patterns of the relationship stay-length
suggests some tendencies for the perspectives and commodities within
which the analysis is carried out (Figure 8). In general, the relation—
ship is clearer for "research in general" and "sponsor organizations."
This is more evident in the cases of rice and beans. .For the perspec-
tive "fields of training" there are no statistical differences in the

cases of beans and cassava, which indicates that the relationship does not

130

.mo.o n ~w>m~ asap<

.ucmcmwewu xppcouwwwcmwm yo: men copump macm mzp saw; mcmmz ”muoz

 

 

 

 

 

 

Po.N Fm.N < oa.m < om.m m am.~ Pacacam c. eucaamam
mm; om: < am < a: < 8; 35.8.5.8 28
mN.F mo._ < no.” < FN.N < oo._ meccwacp co ma_a.2
om._ No.N a Na.m ma No.m m.mmwm meowpaawcamco comcoam
.NM .N. :V $W :N

ANmP 1 .cv Raw 1 .CV Aa_ u ==v he._ 1 =cv a>ammau
_m.~ mp.m < e_.a m mp.N m< mo.m _acacam a. sucaamam
mN.N om._ < om.~ < mm.~ < mN.F ma.3_aoesou e428
VP.N mo._ < am.~ m mm.F m $0.? meecwacp co muPa.a
oa.~ oo.~ a eo.m m No.m m mo.N meowpawwcamco camcoam
.IMW .1“: :M :WAI :M.

A¢__ 1 .cv Ame u =cv A03 1 .cv ANN u =cv maczpmaa,_auwaace
Na.F P~.N a mm.m m om.~ u wo.~ Pacacmm a. socaamaa
aa._ am.~ a mm.~ < om.~ m mm.~ ma.u_aoEEOU 4(20
N¢.F _m.F < mm.P < me.r < mm.~ m=_:.acp co mc_a.a
mm.~ No.~ < 3N.N < mm.~ m mm._ meo.pawwcamco camcoam
.MW .m. .w ;m .m .1111.

AFON n .Cv AVN H :CV AON H :Cv ANOP N =CV mcwwm
ow.~ em.m < mo.o m No.m m NG.N Pacmcam c. guacamaa
m~.~ om.~ < mm.~ < _P.m < aa._ ma.3_aoesou F<Hu
on.m ma.~ m< N©.~ < _P.m m m¢._ mcwcwacp to ma_a_a
mm.~ mm.m < mo.m m oa.m m mm.~ meowwa~wcamco camcoam
.mm .m .m. .m. =w

AMPF fl .Cv AMP " :Cv fiNm N ::v Am? N :Cv Mme

Pouch mcucos xwm xwm op a: .mcpcoe mmmp Lo mzucoe m>wpumqmcmm\pcmu=ou
song use: oz“ cusp mcoz ozh

 

A.mm>wpumamcma co mucmpcoo :mmzumn
.mmm.wum mcomwcmaeouv .mcwcwmcu so gamcmp mpnmwco> acmucmamocw mg» so cowaocaw a ma
Amcmmx we cmnszcv xmwm mpnmwcm> “cavemamu mg» cow mucmwcm> No mmmxpmco mzu No xcm553m .NN mpnmh

6Stay (years)

Research in general”Z—'
Sponsor organizations

CIAT

 

3 /’)Q2

Fields of training

1 2 3
a. RICE (Length)

Stay (years)

Research in general-«ELfiL

     
 

Sponsor organizations

 

 

 
    
 

131

commodities

 

3 CIAT com-
\ f/ modities
2 "
M

. 0r*“"'”"" Fields of
‘l— training

1 3

c. TROPICAL PASTURES (Length)

(30

Stay (years)

 
 
 
 

Research in general‘?h>
Sponsor
organizations

 

 

 

M:
/ cml‘
Fields of commodities
training
1 2 Y17
b. BEANS (Length)

Stay (years)

Research in general

Sponsor
organizations

 
    
 

 

 

 

CIAT
commodities
.Ffiggsszzsfidt:\cc2~‘~\fi:
Fields of/A/fl
training
___ 1 2 3
d. CASSAVA (Length)

Figure 8. Average number of years of stay within sponsor organizations,
fields of training, CIAT commodities, and research in general,
for the training contents: rice, beans, tropical pastures, and
cassava. Stay is graphed as a function of length of training.

intermediate training, and 3 = longer

(1 = short training, 2
training).

132

exist or it is very weak. This is also the case for the comparisons
between short and intermediate training in tropical pastures, and
between intermediate and longer training in rice. However, for “fields
of training" the positive relationship is very clear in the comparison
between short and intermediate training, in the case of rice.

2. Stay and facilities to do research. The degree of facilities
to do research in the fields in which former trainees carried out their
training is a structural factor supposedly very influential in post-
training research performance. It is assumed that stimuli as well as
constraints may arise from the degree of facilities available to a
person for continuing to actively engage in research.

In order to examine the presence or absence of a relationship
between stay and research facilities, the Latin American and Caribbean
countries which have sent professionals for training at CIAT were
classified in eight groupings, two per commodity: those with "more"
facilities to do research in each commodity, and those with "less”
facilities (see pp. 78-81 for details about this classification).

None of the means compared exhibited statistical differences at
an alpha level of 5% (Table 23). Therefore, data do not support the
existence of a relationship between "stay" and "facilities to do
research.” Or, at least, such a relationship is very weak in these
data, and consequently significantdifferences could not be detected.
Similar to the case of "length of training," the size of standard
deviations for "research facilities" is large.

In spite of the fact that differences are not'statistically
different, a slight relationship seems to be suggested by the data.

Two thirds of the actual values of average stay are higher for trainees

133

 

 

 

 

 

 

 

 

 

 

 

Table 23. Summary of the analyses of variance for the dependent variable
stay (number of years) as a function of the independent
variable facilities to do research at sponsor organizations.

Facilities
Content/Perspective More Less Total
Rice (n" = 91) (n" = 22) (n‘ = 113)
R" R" 7| S'x'l
Sponsor organizations 3 58 2 54 3.38 2.78
Fields of training 2.51 2.09 2.43 2.70
CIAT commodities 2.60 2.09 2.50 2.73
Research in general 3.81 2.54 3.57 2.86
Beans (n" = 126) (n“ = 75) (n' = 201)
TI ‘x'll Y. S-x'l
Sponsor organizations 1 99 2.08 2.02 1.38
Fields of training 1.53 1.48 1.51 1.47
CIAT commodities 1.64 1.64 1.64 1.44
Research in general 2.21 2.23 2.21 1.41
Tropical pastures (n" = 64) (n" = 50) (n' = 114)
in ‘ill 2| Si!
Sponsor organizations 2.73 2.64 2.69 2.46
Fields of training 1.50 1.84 1.65 2.14
CIAT commodities 1.59 2.06 1.80 2.25
Research in general 3.27 2.96 3.13 2.51
Cassava (n“ = 110) (n" = 42) (n' = 152)
2" ‘ill i0 32"!
Sponsor organizations 2.61 2.64 2.62 1.96
Fields of training 1.72 1.50 1.66 1.78
CIAT commodities 1.83 1.74 1.80 1.88
Research in general 2.84 2.74 2.81 2.01
Note: All means are not statistically different at alpha level = 0.05.

134

whose sponsor organizations are classified as having "more" research
facilities than those with "less" facilities.

A possible explanation for the lack of significant differences
is that the categories applied in this study are very gross categories.
Another possibility is that all members of this sample have the minimum
level of research facilities, under which people are restricted from
doing research. This last explanation might be supported by the
circumstance that agricultural research tends to be non-capital
intensive. Speaking in general tenms, most agricultural research
activities do not appear to require costly investments in labs and
50phisticated equipment. Even the ownership of land and buildings for
experimental stations does not seem to be essential, as illustrated by
the case of the Instituto de Ciencia y Tecnologia Agricola (ICTA) of
Guatemala, where researchers have been effectively conducting most of

their research work off stations on farmer's land.

CHAPTER V
Findings of the Study, Part II:
THE SOCIAL STRUCTURES AND
THIRD CULTURES OF THE RESEARCHERS' NETWORKS

This chapter examines former CIAT trainees as practitioners
who participate in varied types of networks among agricultural
researchers. It starts with a brief description of the population
and sample for network analysis.

The analysis begins by partitioning individual researchers in
two categories, participants and non-participants.

Then the "third cultural" nature of these networks is delineated
through the analysis of participants' linkage on two dimensions: "locus
of incidence" and ”types of links." Transnationality, the special
characteristic of these networks, emerges from sorting the sample
linkage with research colleagues into four mutually exclusive loci
of incidence: 1) within respondents' organizations, 2)outside their
organizations but inside their country, 3) in the rest of Latin America
but outside respondents' country, and 4) in the rest of the world but
outside the region. The issue of types of linkages characterizing this
sample is analyzed by using the Useems' conceptual categories of
intensive, nascent, and latent ties. (see pp. 59-60).

Following that, network structure is examined in terms of the
roles that individual researchers perform in these third cultural
networks as members of research groups, liasons between groups, other

135

136
participants, as well as components of various types of non-participants.
Finally, former CIAT trainees' participation in research networks

is related to the three selected independent variables of this study:

training content, length of training, and research facilities.

A. Population and sample.

 

The population for this study of the networks of researchers
consists of former CIAT research trainees who, after training, engaged
in research, even if for a brief period, and for whom current addresses
are known. From the 580 persons for whom data were obtained in the
census of organizations (p. 109 and Table 13), 553 professionals did
research after training. Current addresses were identified for 542
of these persons, to whom network questionnaires were mailed. ’

In view of time and financial limitations a relatively short
time interval was set for initiating data processing. Hence, the
sample analyzed in this chapter includes only 207 questionnaires
received by the end of January, 1981. (At the moment of writing this
thesis, approximately 400 answered questionnaires have been received.)

According to the three variables, proportionate distributions
by "training content," length of training," and "research facilities"
are not substantially discrepant to the proportions in which the total
number of persons trained are distributed (Table 24). "Length of
training" is represented in exactly the same proportions. "Training
content" shows a 13 percent over-repreSentation for beans, which
correspondes to under-representation for rice, tr0pical pastures, and

_cassava. As for "research facilities," the individuals with "more"

facilities are 7 percent over-represented, and individuals with "less"

facilities are under-represented in the same proportion.

137

Table 24. Former CIAT trainees who compose the sample for network
analysis, distributed by training content, length of
training, and facilities to do research at home organizations.

 

 

 

 

 

Variable n % Total %*
Training content 207 100 100
Rice 32 16 22
Beans 92 44 31
Pastures 41 20 22
Cassava 42 20 25
Length of training 207 100 100
Two months or less 102 49 49
More than two, up to six months 69 33 33
More than six months 36 18 18
Facilities to do research 207 100 100
More facilities 151 73 66
Rice 27 13 17
Beans 67 32 19
Pastures 24 12 11
Cassava 33 16 19
Less facilities 56 27 34
Rice 5 2 5
Beans 25 12 12
Pastures 17 8 11
Cassava 9 5 6

 

* Percentage distribution of trainees within the total number of persons
trained in research during the period analyzed.

138
8. Participants and non-participants.

These 207 former CIAT trainees reported that altogether, at the
time of answering their questionnaires, they has ties with a worldwide
conmunity of 1413 agricultural researchers. The multitude of inter-
relationships among the members of this community are classified by the
intensity and direction of their interchanges by using the computerized
NEGOPY program (p. 102). Roughly, about two thirds of them (141 trainees)
were classified as participants in networks of researchers, and the rest

were identified as non-participants.

C. Third cultural character of former CIAT trainees' networks.
According to the Useems' sociological construct (pp. 55-60)
members of third cultures are people who relate their societies to each
other through continued interactions which, in essence, are personalized
but do not require continuing face-to-face encounters. These relation-
ships prevail among scientific researchers who do not have the same
nationality or organizational affiliation, but are widely scattered
throughout the world, working for different countries and organizations.
What keeps those interactions alive is not kinship or location ties, but
the sharing of cognitive goals and experiences: their research problems.
By means of these interchanges, researchers create new social groupings;
they are members of small transnational communities. Participants carry
into these communities cultural elements of their respective indigenous
societies; from the interaction between cultural elements of the inter-
‘T,dependent societies, new cultural forms arise--third pultures--which are
shared by network members and passed on to newcomers to these new social
groupings. In such ways bi-national, multi-national, and worldwide third
cultures of science and technology are created, maintained, and shared by

participants in networks of researchers.

139

To examine the nature of the networks in which former CIAT
trainees participate, two dimensions of their linkages are analyzed:
locus of incidence and type of linkage. Locus of incidence refers to
the location of the interacting persons mentioned by respondents. Type
of linkage refers to the frequency and depth of the interactions.

1. Locus of incidence. An aggregate of links are presented in

 

Table 25. The 141 former CIAT trainees who participate in networks
report a relatively high incidence of links with their significant
others; on the average, 9.3 links per person. Nearly three out of every
four ties are with other researchers within their immediate organizations
and home countries. Within Latin America one in every five has personal
ties with researchers working for organizations and countries other than
their own. Close to seven percent have linkages outside the Latin

American region in some part of the rest of the world.

Table 25. Locus of incidence of links reported by network participants.

 

 

Number of Percentage Links per
Locus of incidence links of links individual
Own organization 529 40.4 3.75
Home country 429 32.8 3.04
Latin America
(Outside their country) 261 20.0 1.85
Worldwide
(Beyond Latin America) 89 6.8 0.63

—— _— _—

' TOTAL 1308 100.0 , 9.28

 

140

These results indicate that the total of former CIAT
trainees' linkages are primarily oriented toward their own work
organization and home country. However, these networks are not
totally domestic. To put it in the larger transnational setting, over
a fourth of trainees have connections with fellow researchers beyond
their home land. These encounters open up the possibility fOr third
cultures of science and technology to emerge within segments of these
agricultural researchers.

2. Type of linkage. Six out of every nine links are of an

 

intensive nature. Relationships of sub-type two--more intensive--are
about twice as common as those of sub-type one. The other three out
of nine links constitute latent and nascent interactions, with the

latent being approximately double the nascent ties (Table 26).

Table 26. Number of links reported by network participants,
distributed by type of linkage.

 

Number of Percentage Links per

 

Type of linkage links of links respondent
Latent 286 21.8 2.02
Nascent 150 11.6 1.07
Intensive, sub-type one 284 21.7 2.01
Intensive, sub-type two 588 4§;Q_ flglz
TOTAL 1308 100.0 9.28

 

However, hypothetically it is possible that intensive ties
concentrate within organizations and countries, and transnational
relationships might be only of a nascent or latent character. To

check these possibilities, and explore in more detail the third

141

cultural nature of the networks in which former CIAT trainees participate,
the relationships reported by participants are cross-tabulated according
to "locus of incidence" and "type of linkage" (Table 27). This cross-
tabulation shows that almost one third of the intensive relationships
occur at the regional or world levels, that is to say with people from
outside of respondents' own organization and country.

Table 27. Linkages of network participants, distributed by locus of
incidence and type of linkage. 1

 

 

 

 

15‘
Type of linkage
Locus of Latent Nascent Intensive 1 Intensive 2 Total
incidence Links % Links % Links % Links % LinEs % ‘_

 

Organization 131 10.0 61 4.7 100 7.6 237 18.1 529 40.4
Country 88 6.7 57 4.4 84 6.4 200 15.3 429 32.8
Latin America 59 4.5 22 1.7 65 5.0 115 8.8 261 20.0
World _8 0.6 10 _9;§_ 35 2.6 36 2.8 89 6.8

-——-.———_ _._—_—

 

TOTAL 286 21.8 150 11.6 284 21.6 588 45.0 1308 100.00

 

In summary, the analyses of locus of incidence and type of
linkage of CIAT-trained researchers with relevant colleagues provides
evidence that third cultural networks of science and technology are
being created, maintained, and shared in particular fields of agricultural

research in the region.

0. Network structure.

 

This analysis follows the conceptual framework presented in
pages 45 to 60 and the methodology described in pages 102 to 106. The
research community is the wider concept; it is composed of all those

persons who occupy research positions in their societies as a central

142
part of their work activities. The research community is not a uniform,
homogeneous entity, but a multitude of loosely defined clusters of
interrelated individuals.* Some of these persons participate in more
than one cluster, acting as the junctures of two or more intersecting
clusters. This overlapping, as well as other factors such as the size
of the research community, its lack of geographical or political
boundaries, and the non-formalized nature of the relationships among
researchers, creates a high complexity which makes it difficult to
study the research community in its entirety.

The concept of research networks provides a more workable
specification of the above-mentioned clusters. A research network is a
subset of the research community integrated by researchers whose
attention is focused on closely related research problems. This
definition allows one to include in the study of a given network the more
active members who are well interconnected with each other (participants),
and also those potential members (non-participants) who have very few or
no connections within that particular network.

When interchanges among researchers are continued and intense
enough, their networks become more structured; differentiation of roles
emerges as evidence of increasing social organization. A high degree of
structure in a research network is characterized by the development of
some special groupings called "invisible colleges" in the sociology of
science.

In this thesis the terms "invisible college" and "research
groups" are used interchangeably. Research groups aré operationally

defined in this study as "communication groups" in the terminology of

 

*Because of such heterogeneity, it would be more appropriate to talk
in plural about "research communities."

143
communication network analysis. They may be interconnected with each
other through some of their respective members, who, while being group
members, perform the role of bridges between groups. However, the
particular role which is often salient in linking disparate groups is
that of "liaison.“ These individuals in research networks serve as
natural research leaders; they are informal coordinators, who intensively
engage in interchanges with members of several "invisible colleges,"
mediate new findings, opportunities, and opinions between groups, and
also bind separate groups to a common paradigm. Liaisons in association
with influential members of research groups exercise leadership and
guidance in their research network, providing orientations in the selec-
tion of research problems and techniques, and sanctioning research work
with competent assessments, recognition, and rewards.

The rest of the participants constitute a role designated in
communication network analysis by the term "others." They are those
people who actively engage in research interchanges but remain
”independent," and those who are in the process of forming new groups
or entering already established groups.

The remaining part of the network is constituted by non-
participants; these range from total isolation to a limited contact with
participants or with other non-participants. Four different roles among
non-participants are: isolates type one, isolates type two, dyad members,
and tree nodes (the corresponding definitions are on p. 105).

To decipher network structure, data was processed by the Richards'
computerized NEGOPY program (1975). Half of the participants are group
members, and four persons perform the role of liaisons; on the other

side, out of the 66 non-participants, 49 are tree nodes (Table 28).

144

Table 28. The distribution of the performance of network roles.

 

 

 

 

Role N Percentage
Participants 141_ 68.1
Group members 69 33.3
Liaisons 4 1.9
Others 68 32.9
Nonaparticipants §§_ 31.9
Tree nodes 49 23.7
, Dyads 4 1.9
Isolates type 2 3 l 4
Isolates type 1 _1g_ 4.9
TOTAL 207 100.0

 

The specificities of network structure are more clearly visual-
ized when results are discriminated and mapped in reference to particular
research groups (Table 29 and Figures 9-30). Twenty-two* invisible
colleges were identified. Most of them are specialized: seven in bean
research.six in tropical pastures, two in rice, and one in cassava. The
other six groups combine researchers of different fields: four groups
are integrated by rice and bean researchers. and two are composed of
cassava and bean researchers. The size of research groups varies between
three and fourteen persons; groups with three, four, or five people are

most frequent (thirteen groups altogether).

 

*Groups are named with numbers from 1 to 25, as they were labeled by
the NEGOPY program. This means that 3 clusters initially identified as
' groups did not meet one or more of the group criteria. The numbers used
as group labels by the program were kept to facilitate the drawing of
the maps, which is a relatively cumbersome task, given that NEGOPY does
not plot these diagrams, but only gives numerical information.

145

Most research groups include professionals with and without CIAT
training; two exceptions occur: one group in bean research (Figure

9) and the other group in tropical pastures (Figure 10), which consists
exclusively of former CIAT trainees. Among the 135 persons in the
twenty-two research groups, about one half are CIAT alumni. The

group members without CIAT training include fifty-six researchers who
work in Latin America, and ten professionals who work in other parts

of the world, outside Latin America.

As for the transnational identities of group members, they split
in two halves; eleven are entirely in-country groups, and eleven groups
have members of various nationalities. However, in all twenty-two
groups there is within-country dominance in the proportionate number of
group members. The eleven groups with membership from various countries
include five of the six tropical pastures groups (Figures 10-14), plus
six groups integrated by rice researchers (Figures 15 and 16), rice and
bean researchers (Figure 17), bean researchers (Figure 18), bean and
cassava researchers (Figure 19), and cassava researchers (Figure 20).
There is only one specialized group in cassava (Figure 20). There are
two additional cassava trainees who are group members. The other two
cassava trainees who are group members belong to groups numerically
dominated by bean researchers (Figures 19 and 21). However, in one
of these groups (Figure 19) the cassava researcher is the more central
person in the group (identified by the number 53); the other four group
- members are bean researchers; and three of them are not former CIAT
trainees. In another group the cassava researcher is the peripheral
member (Figure 21, node number 2), but at the same time it is noticeable

that this cassava researcher performs as a bridge helping to bind his

146

group (Group 9 in Figure 21) to a tropical pastures group (Group
14 in Figure 13).
Organizational affiliation and the country where members
of research groups are currently working are not factors selected as
independent variables in this study. Nevertheless, a look at these
maps immediately suggests that these factors, in addition to field of
research, are among the most relevant traits in tracing the identities
of members of invisible colleges in the population of this study.
Twelve Latin American countries have representation in these
twenty-two research groups: BoliVia, Brazil, Colombia, Costa Rica,
Chile, Ecuador, Honduras, Mexico, Nicaragua, Paraguay, Peru, and
Venezuela; there are no group members from the Caribbean countries.
Eight countries in the rest of the world, outside the region, also
have participants in these twenty-two invisible colleges: Japan,
the Philippines, India, Syria, the Netherlands, Denmark, England,
and the U.S.A.
The eleven within-country groups are: two in Brazil (Figures
22 and 23), three in Colombia (Figures 21, 24, and 25), one in Costa
Rica (Figure 26), two in Ecuador (Figures 27 and 28), one in Honduras
(Figure 9), one in Mexico (Figure 29), and one in Venezuela (Figure 30).
With the exception of three groups, invisible colleges normally
include persons from differing research organizations. These three
groups are: one in ICA, Colombia (Figure 24), one in INIAP, Ecuador
. (Figure 27), and one in the Ministry of Natural Resources of Honduras
(Figure 9). All members of these three groups are bean researchers.
The distribution of organizational affiliation of the 135

persons who compose the 22 research groups identified by this study is

147

presented in Table 29. About half of them work for national agricultural
research institutes (this type of institute is briefly described on p. 29).
Almost one third belong to universities and governmental agencies; two of
those are U.S. universities, and the rest are universities of seven Latin
American countries: Brazil, Colombia, Chile, Mexico, Peru, Puerto Rico,
abd Venezuela. Researchers from four international centers--CIAT, IRRI,
ICRISAT, and ICARDA--participate in these groups; however, it is called

to the attention of readers that the other two international centers
working the region--CIMMYT and CIP--are not represented in these invisible
colleges. Finally, about one in every nine group members is affiliated
with a country-based or an international private organization.

Table 29. Members of the 22 research groups by type of organizational
affiliation.

 

 

Organization Number Percentage
National research institute 69 51.1
University 22 16.3
Other governmental agency 20 14.8
International center 12 8.9
In-country private organization 9 6.7
International private organization __;3 __2;2
TOTAL « 135 100.0

 

By relationship to the identities of group members, the relevant
interplay of fields of research, countries where currently working, and
organizational affiliation may be realized in describing one of these

invisible colleges, e.g., group 7 (Figure 19). This group is integrated

148

by five researchers. Four are bean researchers. The other one is a
cassava trainee who has been studying the crop association bean-cassava.
The group includes four nationalities and four organizational affilia-
tions: two people are from the Universidad Tecnica del Piura, in Peru;
one is from ICA, the Colombian institute of agricultural research;
another from FONAIAP, the Venezuelan national organization for agricul-
tural research; and the last one is a researcher in a Japanese organiza-
tion. This invisible college is connected with groups 14 (Figure 13)

and 15 (Figure 18) through person number 54, who plays the role of
liaison. This professional is a woman from Peru. She was trained at
CIAT in cassava research. Group 7 is also bound to group 10 (Figure 15)
by person number 697, a bean researcher. In addition, group 7 is
connected by groups 19 (Figure 14) and 23 (Figure 28) by means of members
of the corresponding groups who perform as bridges between these groups.
Group 7 is described here as a matter of example, but not as typical of
the research groups existing in this network; in fact, there are no
typical groups; each one has its own characteristics. Some commonalities
among them are mentioned elsewhere.

Participants in this research network are totally interconnected
with each other. An all-encompassing picture of these inter-dependencies
is presented in Figure 31 (p.175»). With the purpose of simplifying the
visual representation, only groups and liaisons are shown in this map.
Some clusters of groups by field of research can be recognized. Such is
.the case of groups 1, 11, l4, l9, and 20, composed of tropical pastures
researchers. Specialized groups in rice, beans, and cassava are well
interconnected wtih groups including researchers of two commodities--

in some cases cassava and beans, in others, rice and beans.

149

The roles of liaisons are performed by four former CIAT trainees,
four members of the staff of CIAT, and one member of the staff of INIA,
in Mexico.

A profile of some personal characteristics of the segment of this
analysis sample who are members of research groups follows. (Table 31,
[L 176). They range between 24 and 55 years of age, similar to what
occurs in the rest of the population; however, frequencies in the ages
between 24 and 34 years are higher for group members than for the rest
of the population. These frequencies characterize sample group members
as relatively young researchers. Group members are, with one exception,
males. The proportions between single and married people are the same in
this sub-sample as in the population. As for education, there is a
slight tendency for practitioners holding higher formal degrees to be
group members. There is a group member whose education is under the

B.S. level.

 

*The text continues after the presentation of Table 31 and Figures
9 through 30, on page 177).

150

 

.o .Fwchm ”zcucsou .m .eowm memou ”xcpcaou

 

.m .mcomm ”mmmcwmcu LmELom .P .muwc mm .mcmmm "mwmcwmxg cmscom

.m:0mcma o ”mNWm .mcomcma m "mNPm

m .505. e .595

.F .cmamw “F .m_neo_ou MP .mpm:~m:m> MN .3cma "mmwcpczou .v .mmcsuco: "acucaou
.P .mcmmn up .m>mmmmu ”mmmcwmcp LmELON .5 .mcmmm ”mmmcwmcu gagged
.m:0mcma m ”mNPm .m:0mcma e ”mNWm

N macaw m macaw

.mp ._wchm ”xcpczou .P .AF<HU ucmpmcm no .PPchmummr»uczou

._ .mowc mm .mcmmm "mmmcwmcp cmELou .m .m>ommau ”mmmcwmcp emacou
.mcomcma m? ”mem .m=0mcwa m ”wNwm

m macaw N 120mm

._ .xmzmmcma ”N .ouwxmz ”mmwcpcsou .P ..<.m.: we .Loumaom ”mmwcucsou

.P .mmcapmoa Pmquoch "mmmcwmcu LmEcoN .m .mmcsumma _muwaoce "mmmc_mcu cmEgON
.m:0mcmq m nm~_m .mcomcma m um~wm

m gnome _ anomo

 

.mcmnems azocm No swmwco mo xcucaoo new .pcmpcou mc_:wmcp
ma umwmwumam mazocm :wzpwz mmmcwmcp cmEcow mo Longs: .mNPm aaocm No magma cw mpnmu m_;a cw
um~_cmuumcm;u mew masocm cucmmmmm .mwmxpmcm xcozpm: mcu gmzoczu umwmwpcmuw maaocm zucmmmmm .om mpnme

 

151

.— .xcmscmo my .AH<HQV mwnsorou mm .mwnsopou

P .mwwcH

.m .mwnEOFQU ”xcuczou

.P .mu_c ”F .mcmmm "mmmcwmcu LmELom
.m:0mcma m ”mN_m

©_ enema

.P .macaycmcpmz mN .scma ”mmwcpcsou
.N .mcmmm ”mmmcwmcp coacom

.mcomcmg m ”mNWm

mp macaw

“mmwcuczou
"mmmc_mcu coacoa
.m:0mcma m "wNwm

.m .mmczumma Foquocp

«P macaw

.2 2%
m_ .AH<HUV a_aso_ou ”_F .62.;u ”mmacpcaou
.F .mu?c no .mcmmm ”mmmcwmcu emacom

.v .coumzum

.m .mcmmm ”mmmcwmcp LmELom

.mcomcma e "mNTm

NP macaw

._ .ouwm opcmza we .mpm:~m:m> "mmwcucaou

.m .mmcaumma quwaoc»

.m:0mcwa m um~wm

FF gnome

.m .acma ”xcucaou

.N .muwm ”mmmcwmcu cases;
.mcomcma m ”mem

o— gnome

.c .mwneoFou ”xcpczoo

.N .mcmmn mp .m>mmmmu

”xcpcaou

“mmmcwmcp Lasso;

"mmmcwacu emacom

 

 

.mcomcma 5P um~_m .mcomcma m “wNwm
m_ gnome m azomo
.Aum::_pcouv .om mpnce

152

 

.NN .ouvxmz uchczoo
.m .mcmmm "mmmcwmcu cmELOg
.mCOmcmg NP ”mNNm

.mN gnome

.v .mwneopou "xguczou

 

.N .mcmmm "mmmcwmgu cmELog
.m20mcmg v ”mNNm

eN gnome

.N .Lovmsum ”xgpcaou

.m .mcmmm ”mmmcwmcu LmEcog

.mcomcmg m ”mem

mN gnome

.angN

NN gnomo

.xngN

NN gnome

.N .mpmstcm

MN .Acaguv

_- nc<omo
erase—o

> ”N .mzmmgmuwz mp .
m .mmcspmmg NmuwgocN

.P .ge<guv .
2 MN .ge<guv stag .
U “N .agaeogog “F .

N
a
a .mmcspmag Nmu_goge

NNchm "mmwcuczou
”mmm:_mcu cmELOg
.chWng m umem

ON gscgo
.<.m.:
.ccmg
w>w_om "mmwcucsou

”mmmc_mcp cmELOg
.m:0mcmg oN ”mNNm

mN gnome
.xNgEm

NF gnome

.N .mmcmggwpwzg do .ouwxwz "mmwgucsou

.a .auwm

”mmmcwmcp gmeLOg
.mcomgwg N um~_m

 

NP gnome

Aaaacwucouv .om m_aae

153

 

 

 

GROUP 3

/’—-—§

 

 

 

 

 

 

-
i
416
\
‘\
\ 1
l
g l
l
\
’I 494
I
I
I
456 ,
* /
./
l
’/
Liaison for
Groups 3
.nd 4

 

 

Figure

*402
*416

9. Group number 3, integrated by bean researchers.

Organiza-

tional affiliations of groups members are:

Min. R. N., Honduras
Min. R. N., Honduras

*456
*462

* = Former CIAT

Min. Rec. Nat., Honduras
Min. Rec. Nat., Honduras

trainee

154

 

 

GROUP 20 ®

 

/ \
/’. ‘\\.
I
1’ * * \
I x
, 306 312 ®
,1 \
I 1
I 1
I I
| I
‘ I
I
1‘ I
\
\, /

 

J

 

Figure 10. Group number 20, integrated by tropical pastures researchers.

Organizational affiliations of group members are:

*306 = FONAIAP, Venezuela

*312 = EMGOPA, Brazil

*321 = INTA, Nicaragua
*z

Former CIAT trainee

Tl

155

 

 

GROUP 1

 

 

 

 

 

Figure 11. Group number 1, integrated by tropical pastures researchers.
Organizational affiliations of group members are:

*207 - INIAP, Ecuador 1267 - U. of Florida; U.S.A.
*223 - INIAP, Ecuador 1424 - U. Catolica, Chile
*239 - INIAP, Ecuador

* = Former CIAT trainee

156

 

 

GROUP 5

 

 

Figure 12. Group number 5, integrated by tropical pastures researchers.

*303
1386
1522

Organizational affiliations of group members are:
Colegio Superior de Ag. Tropical, Tabasco, Mexico
Colegio Superior de Ag. Tropical, Tabasco, Mexico
Ministerio de Agricultura y Ganaderia, Paraguay

* = Former CIAT trainee

 

 

 

 

Liaison for n
856 Groups 14, 19 and 20 ,ugoiali‘
I ’r‘ N N N
d’ \\
I’ \
.I' ‘
.’ \\
’l \
\
I, \
,’ 1.587 \
I * \
I 341 \
' \
I
I l
1 l
i l
| 1
1 ,'
“ * 342 II
343
\ * l
\ I
\ 1265
\ 1229‘ /
GROUP 9 /’
A " “ \ \ /
\ \ ’/
\ - ,,/
* \ “-~_ ,."
2 1 * ~"""""‘—-' * 344
i 54 Liaison for
, Groups 7, 14, and 15
Figure 'L3. Group number 14, integrated by tropical pastures researchers.
Organizational affiliations of group members are:
*341 = FG,Valle,Colombia 1229 = D. I. S. P., Denmark
*342 = SDF, Valle, Colombia 1587 = CIAT, Colombia
= FG, Putumayo, Colombia

*343

* = Former CIAT trainee

158

 

 

 

 

r
1’
/
* I
w GROUP 19 ‘ GROUP 11
a- “" “a“ ~ ~ u.
z” s“ \ k
x 1448 \ \
/
I/ \\ ~-
[I
I’ *0
I
I / \
’I 1260 \
I \
I 1
I | ‘
,' 1
1 l
I I
{ , 307 ‘1
1‘ I
\ I
\ I
‘ 0
. ®
I
\\ 809 / I , --
\ I’
\ I /
\ I
® ~\ ,’ I 9
\\ ” I
Liaison for ‘5 ”a, (GROUP 7
Groups 14, 19 and 20 ‘~-______....--
1
\

 

 

Figure 14. Group number 19, integrated by tropical pastures researchers.
Organizational affiliations of group members are:
*243 = CIAT, Colombia 810 = CIAT, Colombia
*307 = CIAT**, Bolivia 818 = North Carolina U., U.S.A.
-*360 = CIAT, Colombia 1266 = CIAT, Colombia
*365 = INIA, Peru 1401 = CIAT, Colombia
809 = CIAT, Colombia 1448 = IVITA, Peru

Former CIAT trainee

This is a national organization, different from CIAT, Colombia,
which is international.

159

 

 

 

 

GROUP 10
I”‘-_\\ J
l’ ‘\
z" \‘\ i"::”
/ \
/ ‘\
,’ 780 * Liaison for
I \ Groups 7
, 1068 \ and 10
g 1
' 1
. I
I
\ I
\ I
\ I
\ /
\ I
\ I
\ /
928 x 749 4 /
\ ,’
\\ ,

 

 

Figure 15. Group number 10, integrated by rice researchers. Organiza-
tional affiliations of group members are:

*749 = ACARPA, Brazil
*780 = INIA, Peru
1068 = CIAG-Norte, Peru

* = Former CIAT trainee

 

16O

 

 

 

 

828
G139113~17
866 a “‘ J ~‘ ~ \
x‘ 713 * \
.r ‘.‘
a’ ‘x
/ \\
a ’ \
435 I, \
’ \
7’ \ 868
I \
l
f 797 \
I l
I 711 * ‘
1 I
I I
\ I
1 I
\\ I
I
‘\ 724 I
\ I
\ 619
.. _ \ 800 735 * I
[I ‘ \ \ / *
GROUP 25 \ \ \ /
, 459 \ \ \ 997 / ,,
\ \ I I .
1 ~5~~ __»v' M6
t. 1L

 

Figure 16. Group number 17, integrated by rice researchers. Organiza-
tional affiliations of group members are:

*711 = INIA, Mexico 797 = INIA, Mexico
*713 = INIA, Mexico 800 = Col. Post. Chapingo, Mexico
*724 = INIA, Mexico 997 = IRRI, Philippines

= INIA, Mexico

*735

* = Former CIAT trainee

161

 

\ I I Liaison
\ 949 Q for
x / GROUP 13 Groups 13
\ l ‘ fl.“ fl -' C. ~
“‘ \ ‘- ~ "'~ and 23

 
 
 
 

\
(EROUP 101

I.
1068//
"/;’
w/

866

  

 

 

 

 

 

 

 

 

/ " \ ~\ / Liaison for
/ . .
/ GROUP \ ~2~ ~~~ ”xx @Greups 13

I 23 ,, 455 ‘1 and 25

 

 

Figure 17. Group number 13, integrated by bean and rice researchers.
Organizational affiliations of group members are:

”*405 = INIA, Chile 893 = INIA, Chile
*418 = INIA, Chile 994 = INIA, Chile
*467 = INIA, Chile 995 = INIA, Chile
*511 = INIA, Chile ' 899 = ICRISAT, India
*529 = Soc. Nal. Ag., Chile 901 = ICAROA, Syria
*595 = INIA, Chile 960 = INIA, Chi1e
*728 = INIA, Chile 1069 = CIAT, Co1ombia

* = Former CIAT trainee

162

 

 

GROUP 15

 

 

I \ Liaison for
* \ Groups 7, 14
,’ 1118 410 s and 15
, \
I \
l 1
| l
| I
\
\ I
\ I
‘ /
\\ /
\\ l/
\ * / *
\\ 539 ,’ 482

 

 

Figure

*4l0
*539
lll8

18. Group number l5. integrated by bean researchers.

Organiza-
tional affiliations of group members are:

Universidad Agraria La Molina, Peru

Ministerio de Agricultura y Alimentacion, Peru
I. V. T., Netherlands

Former CIAT trainee

163

 

 

I
{GROUP 19
1 o
\
\
GROUP 7 \e_ ,.
/"’" ‘M~“‘~ ..
/ \ /GROUP
/ s 1’ 23

z'
/’
I

’0 ".---..‘
-----’

  

é

,, ' * @Liaison for
@ Liaison for Groups 7 and 10
Groups 7, 14
and 15

 

 

Figure 19. Group number 7, integrated by cassava and bean researchers.
Organizational affiliations of group members are:

*53 = U. Tec. Piura, Peru l369 = ICA, Colombia
*4l9 = FONAIAP, Venezuela 1370 = Universidad Nacional Technica
907 = JICA, Japan de Piura, Peru

* = Former CIAT trainee

164

 

 

 

o x
.1, ‘\‘
\
\
\
\
69 \
\
o 1
\
*0 ‘
' I
I
9* ,
l
*0 ,
I
l
.l
l
e ,
s I
x e
"\.‘ a"’
s ,”

 

 

Figure

*33
*69
*7l
*lll

20. Group number 2, integrated by cassava researchers.
zational affiliations of group members are:

Organi-

EMCAPA, Brazil Qll = EMBRATER, Brazil
EMBRAPA, Brazil 1035 = CIAT, Colombia
EMBRAPA, Brazil 1335 = Inst. Ag. Campinhas, Brazil

Inst. Ag. Campinhas, Brazil

* = Former CIAT trainee

165

 

 

Figure

*2
*523

GROUP 9

fl“ .-

 

 

l

Group number 9, integrated by cassava and bean researchers.
Organizational affiliations of group members are:

21.

U. de Narifio, Colombia
ICA, Colombia

*557
l084

* = Former CIAT

ICA, Colombia
Universidad de Narifio, Colombia

trainee

 

 

 

 

 

 

 

 

 

 

 

1'0" 441 ., w
Groups 489

4 and 6.

 

 

Figure 22. Group number 6, integrated by rice and bean researchers.
Organizational affiliations of group members are:

*485 = EMCAPA, Brazil *756 = EMATER, Brazil

*486 = EPA, Minas Gerais, Br. 948 = Emp. Peso. Ag. Minas Gerais, Br.
*504 = ESA, Lavras, Brazil 949 = U.Federal do Vicosa, Brazil
*547 = EMBRAPA, Brazil ll27 = PESAGRO, Rio do Janeiro, Brazil
*548 = ESA, Lavras, Brazil l132 = Esc. Sup. Ag., Lavras, Brazil
*556 = EPA, Minas Gerais, Br. ll36 = Esc. Sup. Ag., Lavras, Brazil
*59l = EMBRAPA, Brazil

* = Former CIAT trainee

 

167

 

 

 

897
o *
GROUP 8
z’fi” ‘\‘\‘\\
./" '\
/ \
f \
/ \ ' it Q
./ ‘\
.I ‘\
I 1
, o o x
i, ‘
I \
1
’ 1
’ 1
y 542
l
i 1
l I
l * l
1 ® :
\ I
\ II
\
l
‘\ 968 ,
\| I
\ l
\\ II
\N\ If/
\ /
\ l
-“ ',v’
-.______ _‘_,.-"
Figure 23; Group number 8, integrated by bean researchers. Organiza-
tional affiliations of group members are:
*542 = IAPAR, Brazil 968 = IAPAR, Brazil
*549 = ACARPA, Brazil 969 = IAPAR, Brazil
“*552 = IAPAR, Brazil l4S4 = IAPAR, Brazil

* = Former CIAT trainee

168

 

 

 

 

 

 

e
e
3”, “\.
a’ '\
1’ * \
\
I \
’ 1237 \
I \
I, \
, \
, l
I l‘
1
1 l
I
I I
1
l
‘ I
I
l I
l I
\ I
\ I
\ 366 /
i * ’
® ‘ 503 g
\ l
\ l’d’
- s / Liaison fOr
‘s \ (a’ Groups 24
® ~z~~ ...»"' . and 25
* '"m' 860

 

J

 

Figure 24. Group number 24, integrated by bean researchers. Organiza-
tional affiliations of group members are:

*366

ICA, Colombia *503 ICA, Colombia
*363

ICA, Colombia l237 ICA, Colombia
* = Former CIAT trainee

169

 

 

 

6129913. .16
I I, ‘ ‘ \
/ \
t / \s
@ ’ “ ®
/ \
I \
’ 938 *
I \
I 422 \
I i
l' 1
l I
1 1
\ I
\ I
\ /
\ /
\ I
I
\ I
\ /
\ * /
\.\ .x’

 

 

Figure :25. Group number l6, integrated by rice and bean researchers.
Organizational affiliations of group members are:

*422 = ICA, Colombia
*746 = FEDEARROZ, Colombia
938 = FEDEARROZ, Colombia
*=

Former CIAT trainee

 

llfll-ll.‘

 

 

 

 

170

 

 

 

 

 

 

@

(Egiibfor Groups

3 and 4

Liaison for
Groups 4 and 6
4

iaison
;

 

 

Group number 4, integrated by rice and bean researchers.

Organizational affiliations of group members are:

Figure 26.

-*452 = Min. A&G, Costa Rica *560
*494 = Min. ABG, Costa Rica *769
*524 = Cons. N. P., Costa Rica 952

*525

Cons. N. P., Costa Rica 1619
* = Former CIAT

Cons. Nal. Prod., Costa Rica
Cons. Nal. Prod., Costa Rica

Of. Nal. Semillas, Costa Rica
Con. Nal. Produccion, Costa Rica

trainee

171

 

 

®

Liaison for
Groups 12,
23 and 25

1k
*
’gsgoup 12 a
I ~~~
I’ w \
I
I
I

 

 

Figure 27. Group number l2, integrated by bean researchers.
tional affiliations of group members are:

*370 = INIAP, Ecuador *SlO = INIAP, Ecuador

*461 = INIAP, Ecuador 784 = INIAP, Ecuador

* = Former CIAT trainee

Organiza-

11" m . . ‘

172

 

Liaison for
Groups 13 and 23

 

I s '
\\ 1 GROUP 7 "
GROUP 13 ‘ 905 1
, I
o , -.~ /
‘--v' ,’ Liaison for
’z’ \\ Groups 12, 23
- ‘1
I, QM] 25 ®
I
I
I

   

   

 

 

~. .,r’

~~ "
@ ~~~~IIII ’9’.

 

 

Figure 28. Group number 23, integrated by bean researchers. Organiza-
tional affiliations of group members are:

*465 = INIAP. Ecuador *601 = INIAP, Ecuador

*478 = INIAP, Ecuador 903 = INIAP, Ecuador

*53l = INIAP. Ecuador l0l6 = CESA, Ecuador

*56l = INIAP, Ecuador ll47 = CESA, Ecuador

* = Former CIAT

trainee

173

 

Liaison for * Liaison for Groups 24 and 25'
Groups 12, 23 399

”.0-~--.

anS 905 ,.. GROUPZS ‘~~.\ 931

,2 o i¢®
Mt; \\'®‘ /® \‘11
“i ‘3' “‘®
“1 ®' * ’1;
1‘4k @“ [I]!

\ 1485
\ ’®
607 t \ I,
\
\. w ” Liaison

 

 

 

 

 

 

\ x, for
/ 7”- ‘ 2 Groups
’ \\\ ‘ ~ ’ 13 and 2
’GROUP 1 \ §‘~.~ ,. o” 427 Liaison for
’ ‘ "“ *** " " * Groups 5 and 2

 

‘fi

Figure 29. Group number 25, integrated by bean researchers. Organiza-
tional affiliations of group members are:

*458 = INIA, Mexico 1226 = INIA, Mexico

*459 a INIA, Mexico 1461 = INIA, Mexico '

*568 = INIA, Mexico 1464 = INIA, Mexico

*569 = INIA, Mexico 1485 = Col. Post. Chapingo, Mexico
*608 = INIA, Mexico 1567 = INIA, Mexico

l023 = INIA, Mexico

* = Former CIAT trainee

 

 

Figure

*352
*353
*356

174

30.

”V.

I H

1‘ "
1‘

844

 

Group number 11, integrated by tropical pastures researchers.

Organizational affiliations of group members are:

U. de Oriente, Venezuela 1013
FONAIAP, Venezuela 1296
FONAIAP, Venezuela

* = Former CIAT

FONAIAP, Venezuela
FONAIAP, Venezuela

trainee

175

7. astures
aexico, 2. ice

Paraguay. 1. Mexico. 6.

a Beans

x

 

CBTBEbia. 4.
Beans 117
Ecuador, 4.0 Q

 
 

Beans

BFEETl, 6.

  
   

Mexico, 11

   
  
   
 
 

  

 

    
     

 

 

 

Beans
ECuador, 8
Cassava a d b
Beans eru, 2. n e_an_s 13 gflP‘TMinE
{EFGT'2. Colombia, 1, -'i“ ' 13°
Netherlands, 1. Venezuela, 1. 39505 and rice
Japan, 1. Chile, 11.
Colombia, 1.
India, 1.
Syria. 1.
@ Rice
0 Peru. 3..
o @ Beans and rice
. aastures ‘Ta 0 C Costa Rica. 3.
To :mnia, _Eéfava Rice and beans
Denmark. 1. Cassava and beans. BraleT 5- C—lcmbia, 3.

pastures
0T Venezuela. 4. 0

Puerto Rico. 1.

amL as_t_____ures
astures ivia. 1.
3'11”. 1. Colombia. 4. o
Beans

Nicaragua. 1. Peru. 3. —-———...
Venezuela. 1, U. S. A.. 2. Honduras, 4-
l'. Pastures 0
Ecuador, 4.
u. s. ‘0’ 1.

Figure 31. Research network mapped with base in a network analysis of

data puovided by 207 former CIAT trainees. To simplify the map, "others

and non-participants are not diagrammed. Numbers within circles are

labels to identify groups. Small circles represent liaisons. Figures

in front of the names of countries indicate number of group members
working in those countries.

Table 31.

176

Former CIAT trainees who are members of research groups,
distributed by age, gender, marital status, and education.

 

Characteristic Number Percentage

 

Age (years)

 

Less than 30 21 30.4
30 - 34 33 47.8 1
35 - 39 11 15.9
40 - 44 3 4.4

More than 44 ‘_1 1.5

9 100 0

Gender

Female 1 l 5

Male 68_ 98.5

69 100 0

Marital status

Single 25 36.2

Married 44‘ 63.8

69 100 0

Education

Less than 8.3. l 1.5

B.S. or equivalent 59 85.5

M.S. or equivalent 7 10 1

PhD _i; 2 9

 

177

E. Relationships with three selected factors.

 

1. Participation and training content. The selection of content

 

as a factor of interest for this study assumes that a researcher's
performance is affected by the cognitive nature of his/her research
tasks. Another related assumption is that the nature of these tasks is
tied to the differential historical development of scientific fields and
modes of production. For instance, when cassava researchers perform
differently from rice researchers, this is associated not only with the
contrasting character structures of the persons who carry out research

tasks in these specialized fields, and with particular conditions of the

 

organizational environments in which they work; differential performance
is associated also with the disimilar characteristics and stage of
development of research traditions and modes of production of rice and
cassava. This is not to say that doing research on cassava is more or
less difficult than doing research on rice, but simply that the two
differ. (An overview of the research traditions and modes of production
of rice, beans, cassava, and tropical pastures is presented on pp.66-73).
A chi-square test suggests that proportions of network partici-
pants and non-participants distributed by training content are different
among the sample studied (Table 32 and Figure 32). There is a higher
rate of participation for beans and rice former trainees who have
engaged in research (approximate ratios between participants and non-
participants are 4:1 in beans and 3:1 in rice), and lower participation
"for cassava and tropical pastures trainees (ratios are 2:1 in cassava

and 1:1 in tropical pastures).

178

Table 32. Participation in networks of researchers, distributed by
training content.

 

 

 

 

 

Participants Non-participants Total
Variable n' % n' % n %
Training content 141 68.1 66 31.9 207 100.0
Rice 24 75.0 8 25.0 32 100.0
Beans 71 78.9 19 21.1 90 100.0
Pastures 19 45.2 23 54.8 42 100.0 1‘
Cassava 27 62.8 16 37.2 43 100.0 E
l
Chi—square = 32.17. Statistically significant at alpha level 0.05. h‘

To further pursue the exploration of this relationship, linkages
are dissaggregated by locus of incidence. Given that the number of
respondents differs within each category of independent variables, a
standard index is constructed by dividing the number of links for each
category by the number of respondents in each category (Table 33).
Before generalizing about the class of data, a word of caution is
presented here. Methodological considerations preclude applying statis-
tical tests to the data arranged in this way. Instead, the writer
merely looks for patterns in the actual values of the indexes; differ-
ences in these indexes may well have occurred by chance.

Looking at the total number of links per respondent for each
field, the pattern observed in the preceding analysis emerges again,

(”with those trainees in rice and beans better connected than their
colleagues in tr0pical pastures and cassava. However, within specific

loci of incidence new patterns emerge. Interpersonal contacts at the

worldwide level show higher indexes in the fields of rice and tropical

179

% Participants
100

90
BEAN

 

'30 RICE

 

70

CASSAVA
r
60

TROPICAL
50 PASTURES

 

40

 

30

20

 

 

 

 

 

 

 

 

 

 

Training content

Figure 32. Percentages of participants in research networks,
represented in relation to training content.

1'

180

Table 33. Linkages of participants at four loci of incidence,
distributed by training content.

 

Training content

 

 

 

 

Locus of incidence Rice Bean T. Pastures Cassava Total
Organization

Number of links 98 283 58 90 529

Index 4.08 3.99 3.05 3.33 3.75
Country

Number of links 86 213 47 83 429

Index 3.58 3.00 2.47 3.07 3.04
Latin America

Number of links 47 132 40 42 261

Index 1.96 1.86 2.11 1.56 1.85
World

Number of links 23 34 16 16 89

Index 0.96 0.48 0.84 0.59 0.63
Total

Number of links 254 662 161 231 1308

Index 10.58 9.32 8.47 8.56 9.28

Number 0. respondents 24 71 19 27 141

.m‘ 1" V

 

.IT;

 

181

pastures than in beans and cassava. It is also noticeable that
researchers in the field of tropical pastures show the lowest index of
connections compared with researchers in the other three fields at the
level of country and organization, and slightly higher for the Latin
American region. Researchers in the fields of beans and cassava exhibit
the lowest indexes of linkage at the worldwide level.

A last examination of the relationship between participation in
research networks and training content arises from disaggregating
participation in terms of network roles, and then focusing on the
proportions of group members within each content. This focus on group
members is based on the high significance that “invisible colleges”
have within research networks. (Liaisons are also very important, but
their reduced number does not allow one to make comparisons.) The
proportion of group members is higher in the fields of bean (44%) and
tropical pastures (36%), intermediate for rice (28%), and lowest for
cassava (12%). The high proportion of group members in tropical pastures

sharply contrasts with the circumstance that this same field exhibitis

the highest proportion of non-participants (55%) in this sample (Table 34).

2. Participation and length of training. Unexpectedly, length
of training is inversely related to participation in networks of
researchers (Table 35). This result is contrary to the general assump-
tion made in the design of training programs at CIAT (see p. 98), and
also contrary to the relationship observed formerly between length of
. training and number of years that practitioners stay doing research

0

after training (see pp. 128-131).

F‘

182

Table 34. Network roles, distributed by training content.

 

 

 

Rice Beans Pastures Cassava TOTAL
Role n' % n' % n' % n' % n %
Participants 24 75.0 71 78.9 19 45.2 27 62.8 141 68.1
Group members 9 28.1 40 44.4 15 35.7 5 11.6 69 33.3
Liaisons 1 3.1 2 2.2 0 0.0 l 2.3 4 1.9
Others 14 43.8 29 32.3 4 9.5 21 38.9 68 32.9

Non-participants 8 25.0 19 21.1 23 54.8 16 37.2 66 31.9

 

Tree nodes 7 21.9 14 15.6 19 45.2 9 20.9 49 23.7
Dyads 0 0.0 3 3.3 1 2.4 0 0 0 4 1.9
Isolates type 2 O 0.0 2 2.2 0 0.0 1 2 3 3 1.4
Isolates type 1 l 3.1 0 0.0 3 7.2 6 14.0 10 4.9
TOTAL 32 100.0 90 100.0 42 100.0 43 100.0 207 100.0

 

T

183

Table 35. Participation in networks of researchers, distributed by
length of training.

 

 

 

 

Participants Noneparticipants Total
Variable n' % n' % n %
Length of training 141 68.1 66 31.9 207 100.0
Two months or less 76 73.1 28 26.9 104 100.0
More than two months
up to six months 47 71.2 19 28.8 66 100.0 '
More than six months 18 48.7 19 51.3 37 100.0 -

 

Chi-square = 14.82. Statistically significant at alpha level 0.05.

This puzzling relationship is examined again in terms of linkage
locus of incidence. Practitioners who had an intermediate period of
training exhibit higher linkage indexes than people in the other two
categories, at three loci of incidence with the exception of the level
of country, in which the lowest index of linkage is associated with
intermediate training (Table 36).

Altogether, trainees in all three categories of length show
high indexes of connections within their countries as a whole. In the
region and on a global scale, persons who had intermediate and longer
training have higher proportions of linkages than those who had only
two months or less of training (Table 36).

In Table 37 the role of the group member is disaggregated by

..length of training. A clearer pattern emerges suggesting a curvi-linear
relationship, in which intermediate training is correiated with the

highest proportion of group members.

184

Table 36. Linkages of participants at four loci of incidence,
distributed by length of training.

 

Length of training

 

 

 

 

Locus of incidence méZ 2(mg—6 m>6 Total
Organization

Number of links 283 179 67 529

Index 3.72 3.81 3.72 3.75
Country

Number of links 272 113 44 429

Index 3.58 2.40 2.44 3.04
Latin America

Number of links 117 109 35 261

Index 1.54 2.32 1.94 1.85
World

Number of links 25 50 14 89

Index 0.33 1.06 0.78 0.63
Total

Number of links 697 451 160 1308

Index 9.17 9.60 8.89 9.28
Number of respondents 76 47 18 141

 

 

Altogether, trainees in all three categories of length show
high indexes of connections within their organizations. Short term
trainees have a higher index within their countries as a whole.

In the region and on a global scale, persons who had intermediate
and longer training have higher proportions of linkages than

- those who had only two months or less of training (Table 36).

185

Table 37. Group members, distributed by length of training.

 

 

Group Percentage Sample
Length of training members of sample size
Two months or less 32 31% 104
More than two months,
up to six months 30 45% 66
More than six months 7 19% 37

 

Finally, this relationship is examined within training contents,
selecting again as a focus the number of group members, given the
' relevance of this role within a research network. The curvilinear
pattern suggested in the preceding paragraph consistently emerges

within all four training contents (Table 38 and Figure 33).

Table 38. Group members, distributed by content and length of training.

 

Length of training

 

 

 

 

Training Short Intermediate Longer Total

content Number % Number % Number % Number %
Rice 4 29 5 33 0 0 9 28
Bean 23 43 13 50 4 36 40 44
T. pastures 2 40 ll 52 2 13 15 36
Cassava __3 _;g l 25 1_ l4 5 12

TOTAL 32 31 3o 45 7 19 ' 69 33

186

 

 

 

 

 

     
 

 

 

 

 

% Group members 3 Group members
60 60
RICE BEAN
50 50
40 40
30 30
20 20
10 10
i
1 2 3 1 2 3
(Length) (Length)
3 Group participants % Group participants
50 50
40 40
3o TROPICAL 30 CASSAVA
PASTURES
20 20
10 10
1 2 3‘7— 1 2 3
(Length) (Length)

~Figure 33 Percentages of grou °
. p members re resent d '
to length of training. p e 1? relation

187

3. Participation and facilities to do research. Data grouped in
terms of more and less facilities to do research in the fields of
training at home organizations do not show statistically significant
differences(Table 39). This result conforms with what was observed
before in the relationship between research facilities and number of
years that practitioners stay doing research after training (p. 132-4)_
This is an unexpected finding which may be associated with the gross

categories used for the variable research facilities. In spite of the

‘

lack of statistically significant differences, a slight difference

,7!

seems to exist in the data suggesting a possible positive correlation

f

between participation in networks of researchers and facilities to

conduct research at home organizations.

Table 39. Participation in networks of researchers, distributed by
facilities to do research at home organizations.

 

 

 

 

 

Participant Nonjparticjpants Total
Variable n' 6 n' % n %
Facilities to
do research 141 68.1 66 31.9 207 100.0
More facilities 104 68.9 47 31.1 151 100.0
Less facilities 37 66.1 19 33.9 56 100.0

 

Chi-square = 0.23. Not statistically significant at alpha level 0.05.

CHAPTER VI
SUMMARY, IMPLICATIONS, AND INTERPRETATIONS

This study of research training at an international agricultural
research center has focused on two aspects of the former trainees'
performance: 1) to what extent is the Latin American region building
inventories of human resources for agricultural research based on CIAT's
training?, and 2) to what extent are former CIAT trainees participating
in third cultural networks of science and technology?

The relationships of three independent variables to both in-
ventories and network participation were examined. The three factors are:
training content, length of training, and facilities to do research in

the fields of training at home organizations.

A. Distinctive character of the study's population.

The researchers and their work related experiences and struc-
tures-—the center of attention of this thesis--are not members of those
communities which have received primary attention in studies of the so-
ciology of science, nor in the accounts of institution building of agri-
cultural science in the developed and developing countries. Such studies
- have focused on Nobel laureates, on professional doing "pure" or "basic"
research, on the careers of famous people in "glamorous" fields like
high energy physics, on the "fathers" of the green revolution, and on

members of academia in MDC's and LDC's who are influential in the central

188

 

189

bureaucracies of their nations and international agencies, as well as
integrated in the world—encompassing networks of research and develop-
ment enterprises.

By contrast, this study's population is composed of young
scholars who usually were in their first job assignments when they
entered CIAT's training; for many of them their stay at CIAT in Colombia
was their first trip out of their home land, their first opportunity to
interact with members of other cultures, and their first chances to
become personally acquainted with people whom they regard as highly com-
petent professionals. The vast majority of them still do not hold formal
degreesbeyond the B. S. level, and occupy middle level professional po-
sitions within large, complex agricultural research organizations. Their
job assignments are usually defined as relatively specialized full time
research. In most cases they are not part of an agricultural technology
delivery system.

In their work they often are not the conceptualizers of research,
but rather they are mediators between those who produce new S&T for agri-
culture and the users of such knowledge (farmers, extension staff, people
in agrobusinesses). As mediators they usually do not live in the capital
cities but in the countryside close to the production units (farms) for
which they test and make adaptations to already available technology.
Most of them have not published in the international journals cn’the
world centers of science, are not listed in the science citation index,
and few are influential in the policy-making of the bureaucracies of their
countries. They are seldom thought about in writings of macrolevel plans
for generating an endogenous science for appropriate technology. Not-
withstanding, these are the researchers who are trying to make techno-

logical options relevant and applicable for increasing agricultural

190
production and productivity in their immediate localities and countries.
Given the location-specific nature of farm problems in conjunction with
part of the agricultural technology, without the continuing work contri-
butions of these researchers a large segment of new S&T may remain
useless for development. As professionals and human beings they have an
identity of their own, they have dignity, and they deserve respect. In
sum, they are more than simple aggregates of low level scientific man-
power who are marginal appendices of the upper level hierarchies of their

research organizations or of the international research communities.

8. Summary of findings.

 

1. Findings on inventories of researchers. The data for this
study of inventories come from a census of 182 agricultural organizations
in 26 Latin American and Caribbean countries which sponsored persons for
research training at CIAT during the period 1969—79.

a. The forming of inventories of human resources for staffing the
intermediate level of agricultural research organizations in Latin America
through CIAT's training has been an effective process. It exhibits some
inefficiencies in regard to the ability to retain trained persons in the
sponsoring organizations and in the activity of research. However, beneath
the surface of obvious loss are more complex and unexamined consequences
of the internal brain drain. Individuals who have left their sponsoring
organizations may nevertheless perform significant roles in other posi-
tions tied to agricultural research and development. Effectiveness of the
process is supported by annual net balances and final-inventories which
during the entire period are positive and increasing. Inefficiencies are
due to a growing process of migration of trained persons towards sites

and tasks other than those of their sponsoring organizations, fields of

 

191
training, and the activity of research. This migration has directly
constrained in particular the forming of a "critical mass" of qualified
professional to establish collaborative research linkages between CIAT
and counterpart organizations in the region.

Four out of every five are still agricultural researchers, and
over half (57%) of them work in those training commodities (rice, bean,
tropical pastures, and cassava). Another fourth (24%) shifted to other
types of research, either within their sponsoring organizations or else-
where. Slightly one fifth (19%) have left research activities.

b. Migration, disaggregated in two components (non-return and
drop-out) shows different magnitudes and tendencies. The non—return to
sponsoring organizations and fields of training is higher than drop-out.
But the trend in the proportion of non-return during the last four years
is showing a decreasing tendency. 0n the contrary, drop-out represents a
smaller proportion of migration than what is accounted for by non-return,
but during the whole period 1969-79 drop-out presents an increasing
tendency.

The impact of the components of migration on inventories varies
with the perspectives from which these processes are examined. Thus, from
the view of "sponsoring organizations“ and "research in general” the
major proportion of migrants consists of professionals who return to their
organizations and their research role, but later on drop-out. This phe-
nomenon has been one of relatively low magnitude during the first half of
~the decade analyzed, but rapidly increased during the second half of this
period. From the perspective of the “fields of training" and "CIAT com-
modities" most of the migration is due to non—return of professionals to

this activites immediately after training; non-return shows a high

192

magnitude during the first half of the training period, but has rapidly
decreased during the latter years.

c. Put in a time frame, efficiency in retaining trained persons
and research experience was low during the first years of CIAT's training,
but since 1975 has shown a consistent tendency to increase, from the four
perspectives analyzed.

d. The length of stay in sponsoring organizations, specialized
fields of training, and in research activities manifest a decreasing trend
throughout the first five to six years after training, but beyond this
interval there is a slight increasing tendency for staying in the same
organizations and fields among those practitioners who still have not left
research.. '

e. The total span of years that members of the sample continue
in their sponsoring organizations, fields of training, and activity of
research, is positively related to the length of their CIAT training ex-
periences. It is not related to the facilities available in their work
organizations in their specialized fields of training.

2. Findings on research networks. A substantive proportion of

 

former CIAT trainees participate in research interchanges with relevant
colleagues and form networks which have a third cultural character. There
is solid evidence that identifiable social structures have emerged out of
the research interactions in which these agricultural researchers par-
ticipate. However, not all former CIAT trainees are well connected to
these networks: among the proportion who are non—participants there are
different degrees of isolation.

a. Participants. About two thirds of the persons in the sample

network analyzed participate in communities of researchers.

193

b. Third cultural networks. Slightly over one out of every four
(27%) of the participants linkages extend beyond their home country, with
most of these ties being with their counterpart researchers working in
Latin America (20%). Among the participants in transnational networks,
seven percent have personal professional ties beyond the region with
colleagues working in some part of the rest of the world. Two thirds
of all their linkages are of an intensive nature, and about one third of
the intensive relationships occur at the region or worldwide level.

c. Social structures. The networks generated thorugh the

recurrent interaction of participants with professional colleagues are

 

highly structured. Twenty-two research groups ("invisible colleges")
were identified; most of them are specialized: seven in bean research,
six in tropical pastures, two in rice, and one in cassava. The size of
the groups varies between three and fourteen persons. Most groupings are
interorganizational and half of them are transnational. Some of the
transnational linkages include researchers of international centers,
but it is noticeable that these research groups do not include profes-
sionals working for two of the centers which operate in the region,
CIMMYT and CIP. These twenty two different research groups are inter-
connected with each other through the links of some of their individual
members who play the role of bridges. Nine persons perform as liaisons;
four of these liaisons are former CIAT trainees. Similar to the total
population of former CIAT trainees, participants are young males with
an education at the B. S. level.

d. Relationships with independent variables. Participation in
research networks varies with training content and length of training,

but not with facilities to do research at home organizations. The

 

194

highest proportions of participants are in the fields of rice and bean,
and the lowest in tropical pastures and cassava. However, when partici-
pants are disaggregated by network roles, bean and tropical pastures are
the fields with highest proportion of group members, rice is intermediate
and cassava is the lowest. Linkages are more transnational in the fields
of rice and tropical pastures, and more localized within organizations
and countries for bean and cassava researchers. These relationships
reflect the structural character of the content factor. In fact, there
is a more developed infrastructure in Latin America for enterprises
related to rice and bean production and marketing, and a less developed
infrastructure for cassava. Tropical pastures are an intermediate input
for beef production, which also has a more developed infrastructure than
cassava. Both cassava and tropical pastures are newer fields of research
than rice and beans. Rice and tropical pastures are more worldwide
concerns which is.reflected in the apparently more "cosmopolite"
linkages of researchers in these fields. In contrast, bean and cassava
research are more domestic concerns at present, and this appeansto be
reflected in a more "localite" orientation of the professional linkages
in these fields.

There is a curvilinear relationship between length of training
and participation of research networks with those persons who have an
intermediate training presenting the highest proportion of participation.
This is a puzzling relationship whose understanding requires further
i investigation. I
e. Non-participants. One third of this study's sample do not

participate in networks of researchers. The highest number and pro-

portion of non-participants are in the field of tropical pastures.

 

195

C. Implications.

 

The following section attempts to suggest practical implications
of the empirical results for CIAT, for international sponsors of agri-
cultural research, for country-based science policy makers, and for
scholars in agricultural development and education.

1. Implications for CIAT. A selective aspect of the two research

 

questions examined in this study can be pursued further in relationship
to these findings. On one side, a heavy migration outside the fields of
training was clearly evident in the study's data; on the other an in-
tensive process of networking emerged. The facts about migration suggest
a strategy of continuously training twice the number of persons estimated
as the "critical mass" required. Findings on networks indicate that
numerous research groups have emerged across the region in relationship
to CIAT's training and communication strategies. These "invisible col-
leges" are social structures which have been associated in the sociology of
science with the most productive individuals in specialized fields of
research. The existence of these social structures among Latin American
agricultural researchers opens a new option to CIAT for policy-making

and resource allocation with regard to training. An appropriate route

for strengthening agricultural research in Latin America seems to be to
inVest more in networking. CIAT already has in its own structure excel-
lent facilities for stimulating the emergence and functions of these
social structures: documentation centers, publications, training and
conferences facilities, as well as an active component of international
cooperation for research interchanges of materials and expertise. A
series of newsletters have been published and distributed in recent

years by each commodity program. Although some directories of researchers

196

have been prepared and distributed, this might be an area which invites
more recurrent attention in the future. All these components might be
explicitly and formally integrated in a systematic strategy for en-
couraging and sustaining research network development.

With regard to the issue of the length of training, the duration
of training experiences relates to the content of learning objectives
formulated for particular training programs. Perhaps, allocation of
resources to longer durations should be considered only after conducting
further analysis of instructional design.

Finally, major potential sources of improvement in the length
that practitioners stay in the activity of research must be discerned
and managed from a better understanding of the complex of reasons that
professionals have for leaving their organizations and the activity of
research.

2. Implications for internationalsponsors of agricultural re-
sgapgh. What emerged out of the study of interpersonal professional
linkages among former CIAT trainees were findings about a new and
distinctive type of network which is role-related and professionally
meaningful for these researchers at work. These networks are not simply
sets of friends who occasionally meet to politely reaffirm their inter-
personal ties with one another. These networks are structured com-
munities of working scholars. Research networks regularly provide to
their members important resources and social-psychological support in
their work roles.

This finding on the existence of active social structures among
agricultural researchers in Latin America allows one to make the

following suggestion to research and development sponsors:

 

 

 

197

the encouraging_and supporting of networks of intermediate level
pgricultural researchers across the region may be as important for

a genuine institutionalization of S&T for agricultural and social de-
velopment aims as is the more Specialized and planned building of
formal organizations for the same purposes.

The findings of the study of these social structures imply
that opportunities now exist for selectively allocating resources for
a better understanding by administrative leadership of the process of
research network development (RND), and a comprehension of how these
networking processes function to complement other organizational
procedures for creating realistic supportive systems for the generating
and sustaining of these communities of researchers. These efforts would
constitute complementary strategies for institution building of science
development, and might result in mutually supportive coexistence of
both the formal organizations of research and the communities of
researchers.

3. Implications for country-based scienceppplicy makers. The
outcome of this study on inventories of researchers show an increasing
migration of trained persons out of the sponsoring organizations and
out of the activity of research. This migrationsupports the assertion
made in other studies about an apparent declining trend of agricultural
research as a powerful social value in Latin American societies. (This
loss of value, of course, refers to the region as a whole. It is well

known that individually several countries are presently engaged in

 

ambitious programs of agricultural research, e. 9. Brazil, Mexico, and

Venezuela).

 

198

On the other hand, the findings of this study on research
networks allow- one to have an alternative view of the process of
science development. A perspective on research networks opens fresh
ways of looking more closely at some of the pervasive macro-level
concerns of many science policy makers throughout the region on the
phenomenon of "brain drain". From a network perspective, the horizontal
mobility of researchers from organization to organization, or from
country to country within the region, does not necessarily need to
be assumed to constitute a total loss of skilled persons. Researchers
in general are mobile persons; their continued interchanges with
colieagues in other societies and cultures enables them to elicit
reactions to their own thinking and cope reciprocally to keep abreast
of the new achievements in other parts of the world. As long as
researchers continue actively interchanging their findings, paradigms,
and methodological breakthroughs with other members of their research
networks, the circumstance of the specific location where they are
working may be of secondary importance, and calls for a more sophis-
ticated search for understanding the cross-cultural transfer of
technology.

4. Implications for scholars in agricultural development. A
central implication of this study for scholars in agricultural develop-

ment is the need to reconceptualize the agricultural research system.

 

Current conceptualizations tend to assume that the only components of
this system are the formal organizations of research.and education.
These conceptualizations implicitly ignore the existence and functions
of the "invisible" communities of researchers, to say nothing of the

farmer-researcher networks.

199

5. Implications for the field of Educational Systems Develppment.
A main implication of this exploratory study for E30 is the potential for
use of social-psychological variables in the design of educational systems.
These variables may be particularly relevant in the area of faculty de-
velopment, but instructional development and organizational development
strategies may also be improved with the inclusion of those social-
psychological factors. Learning has been largely conceived and applied
in system design as an individual process, restricted to the psychological
sphere. The inclusion of social factors involved in learning may constitute
a source of improvement in the formulation of systematic strategies for
educational goals and achievement. Two additional implications of this
study for E80 refer to the use of network analysis as a tool for prac-
titioners in this field and the potential of ethnography as an alternative
research method.

6. Two issues on education inpgeneral. From this study of networks
among former participants in a non-degree training program at least two
issues arise for higher education: a) how to integrate non-degree training
programs--which appear to be highly effective for selected educational
goals--with formalized degree programs in higher education organizations
of LDC's and MDC's; and b) how to integrate these processes of development
of communities of intellectual workers in the formalized conceptions
about the role of education in society. These two issues are suggested for

future study by scholars interested in higher education.

0. Interpretation.
This interpretive section attempts to advance one step beyond the

'empirical data collected for this study, and hypothetically look for

200

additional meanings about the main findings of the study. These inter-
pretations reflect three types of experiences of the author of this
thesis: 1) in-depth interviews with selected members of the study pop-

ulation,* 2) the author's personal interchanges with Latin American
agriculturists during several years of involvement with CIAT's training
programs, and 3) the study of the literature on the sociology of science
and on the emerging third cultures of scientific communities in de-
veloping countries. Readers of this section may keep in mind that all
the following comments are speculative in nature.

1. Assessing the inventory formation process. How does one
assess this inventory formation process? Are figures too high or too
low? Are there some findings to be concerned with? It depends on who
makes the assessment. Standard criteria based on previous research are
not available, to the knowledge of the author.

First of all, the general picture of the inventory analysis,
as well as field observations made during the data collection process,
suggests that most former CIAT research trainees had had no trouble
finding a job after training. This is remarkable in a region where
unemployment exists among agricultural professionals. Perhaps the
contrary has occurred: immediately after training, they have not only
had the option of returning to their research duties in their sponsoring
organizations, but have found opportunities to engage in activities

other than research and with organizations other than those which

 

* These interviews were conducted as part of the data collection for
this thesis; however, they were not systematically reported in this
thesis due to time and funds limitations.

~51: ' “v

1:.

 

201

sponsored their training. If this has been the case , it may indicate
that CIAT's training is highly regarded, not only for preparing personnel
for research, but for other professional work in agriculture, e. 9.,
extension, teaching, commercialization of agro-chemicals, credit,
technical assistance to and administration of commercial farms, among
others. It is likely that from positions like those, the 19% of

former CIAT research trainnes who left the practice of research

may be contributing to food production and economic growth.

Second, from the perspective of preparing research workers
for the region, the accomplishments seem very good. Four fifths of
the people prepared during the 11-year period still participated in
research activities in 1980, including those who were working on
research areas different than their training fields. Specialized
training is likely to be useful for working with a variety of
species (commodities) and research approaches. It is probable that
research techniques, the attitudes toward the conduct of research,
the organizational frameworks of research, and other learnings
acquired by professionals during their training at CIAT, as well
as the contacts they made there with significant colleagues from
other organizations and countries, have enabled CIAT trainees to
do research in fields other than those of their specialized training.
Therefore, these persons-~24% of whom have engaged in research
with other than CIAT commodities--may also be contributing to
increased food production and economic growth in their countries.
Moreover, they may be performing instrumental roles in the institution-

alization of S&T as endogenous systems of Latin American societies.

202

Third, from the perspective of the sponsor organizations, the re-
entering, stay, and retention of former CIAT trainees seems to offer a
more encouraging picture than the general tendencies observed in the
PROTAAL studies (pp.39-44 ). Three-fourths of the persons trained between
1969 and 1979 were still in their sponsor organizations in 1980. This does
not suggest that a loss of twenty-six percent has no significance for
sponsor organizations. Studies of the reasons why people leave their or-
ganizations should be pursued, and corrective actions, when feasible,
should be taken.

Fourth, from the perspective of CIAT efforts to prepare skilled re-
searchers in the focus of its mission and with the purpose of solidly es-
tablishing collaborative linkages in CIAT's four specific commodities, the
findings of this study on inventories constitute a concern. About half of
the persons trained during the period have migrated toward activities dif-
ferent than their fields of training and CIAT commodities.

2. Conjectures on reasons for migration. Why do people leave their
organizations, the activity of research, or the specific fields in which they
were trained? Several speculative comments are advanced in the following
paragraphs about this complex issue. Before advancing any hypothetical ex-
planation, it should be emphasized that these conjectures do not refer to
the majority of former CIAT trainees. Most of them, as pointed out in the
preceding section, remain in their sponsoring organizations and in the
activity of agricultural research. In addition, a proportion of those who
are not counted in the inventories may have had to abandon their spon-
soring organizations and the activity of agricultural research because of

improper professional performance.

203

From the perspective of social systems which underlies the whole
conception of this thesis, human behaviour is determined by the inter-
action between the desires, expectations, and characteristics of persons,
and the structures and cultures of the social systems in which these
persons are embedded. Although systems overlap and interact with each
other, they are not necessarily convergent in the demands made upon their
members. In addition, social systems are not equally responsive to and
do not equally fulfill the particular desires and expectations of each
of their members. These circumstances generate conflicts within indi—
viduals, among individuals, and between individuals and social systems.

How are these conflicts resolved? This is an empirical question
ikn‘Particular'settings and persons. With respect to agricultural research,
the complexities of the situation may be simplified by reducing the more
relevant social systems to three categories: a) the research organization;
b) the profession of agricultural research; and c) the primary systems in
which the individual's socialization originally took place, e. g., his/her
country, local communities, and family.

a. Divergent systems’ demands. When the demands of the organi-
zation, the profession, and the primary systems are not convergent, agri-
cultural researchers need to make compromises between their loyalties to
these systems. They may even have to break their ties with a particular
system. For instance, if a professional perceives that the research goals

of the organization are irrelevant to the needs of local communities to
which he or she expects to make contributions with improved technologies,
this researcher may resolve the conflict by leaving the organization and/or

the profession, or by lessening his/her ties with local communities.

204

Conflicts of this nature may be frequent in Latin America, where
agricultural research must respond to contradictory demands of dual eco-
nomies. Conflicting developmental goals often alternate in relative short
intervals within the same organization, e. g., the goals of generating
technologies to implement national policies of import substitution, ex-
portation of agricultural products,feeding increasing urban populations,
or making available technological options for subsistence farms. Some
agricultural researchers may have kinship or economic ties to either
commercial or subsistence agriculture which bring these conflicts to
critical points. Crises may be resolved by leaving the organization or
the profession. An alternative route in the resolution of this type of
conflict is to strengthen the loyalties to the organization or the
profession, and to decrease the importance of developmental goals. This
option may imply, for instance, dwindling researchers' ties with local
communities. A consequence may be diminished relevance of researchers'
technological outcomes to the problems of the compromised social system,
e. g., communities of farmers who practice subsistence agriculture.

The demands of different social systems upon agricultural re-
searchers may act as counterveiling forces. What may be dysfunctional
for researchers' role performance is the total orientation toward a given
social system with exclusion of the others. For instance, a total orien-
tation toward the profession may result in research outcomes lacking
relevance to organizational goals and agricultural development needs;
. a total orientation toward the organization may lead researchers to con-
centrate on the goal of getting the highest positions of the bureaucratic

hierarchies with the result of losing continuity in their research work;

205

and, a total orientation toward the farmer without taking into con-
sideration the needed loyalties to the organization and the profession

may result in researchers who understand very well their clientele needs
but do not have the research support provided by the organization, and the
knowledge basis generated by other colleagues.

b. Difficulties in making a living. From the personal side, re-
searchers have their own needs, aspirations, and expectations, in the
whole range from physiological primary requirements to social psychological
and cultural needs, such as belonging, identity, self-esteem, and
creativity.

In reciprocity for their work, researchers expect to make a decent
living from the practice of their profession and their loyalty to a given
organization. When agricultural researchers realize that by staying in
their organizations and professions they cannot afford to have adequate
housing, food, clothing, medical services, education, and recreation for
themselves and their families, they will probably be compelled to move
toward other activities and organizations. Sometimes what makes a re-
searcher leave his/her organization or profession is a prospect of a
better education for his/her children, or a minimum additional facility,
e. g., the opportunity to have a vehicle for personal and family use.

An alternative that seems to be emerging in several countries is
the engagement of agricultural researchers in small businesses parallel
to their work at a given organization. In some cases this option takes
place through ownership of a small farm or growing of.a cash crop on
rented land. These income earning activities are not always well accepted
by administrators of agricultural research organizations, because by

doing that professionals may demphasize their research commitment to

206

their work. This course of action, however, may have indirect but un-
planned advantages; carrying out crop production gives researchers
opportunities to test in real conditions the technologies they recom-
mend to farmers, and to be aware first hand of the need to modify those
technologies according to local bio-physical, economic, and cultural
environments.

Salaries and other economic compensations for agricultural re-
searchers in Latin America are very restricted. One reason for this is
the general stage of development of national economies. Another is the
relatively low status in any country of agricultural activities. Other
factors relate to organizational constraints, such as very limited
budgets and rigid administrative procedures. Many administrators of
Latin American agricultural research organizations are conscious of the
need of improving the economic conditions of researchers. Administrators
may be willing to take these actions, yet find obstacles in doing it.
Most agricultural research in Latin America takes place in governmental
agencies: ministries, research institutes--where bureaucratic practices
usually do not permit compensation to a creative person for their work in
correspondence with their contributions and accomplishments, but on the
basis of narrowly fixed criteria which apply uniformly to all public
servants in a specific status position within the hierarchy. When a
given researcher finds better job conditions than those of the organi-

zation or the activity of research, administrators do not usually have a

realistic option but to accept with regret the departure of qualified and

talented members of the organization.

207

c. Personal professional development. It is entirely legitimate
according to the norms of the system of science for a person to seek personal
development as a member of a given profession. Furthermore, in the case
of agricultural research in Latin America, this seems highly desirable
for the long run institutionalization of S&T as an internal capability
of these societies.

d. Training and formal education. The population of this study is
composed of relatively young persons, many of them in their first job and
just starting careers as agricultural researchers. They need opportunities
to pursue their training and formal education at higher levels, e. g.,
participation in workshops, seminars, conferences, and other professional
meetings, M. S. and Ph. 0. studies, postdoctoral appointments, sabbatical
leaves, and so on. When researchers perceive that in other organizations
they will have better training and educational opportunities, it is very
likely that more venturesome researchers will try to move there.

e. A research career. The vigorous pursuit of a productive
professional career requires not only training, formal education, and
degrees. It requires the recurrent engaging in research on problems
which are significant for oneself and for others to whom one regards as
significant colleagues. Implicit in this fulfilling of a research career
are i) the opportunities to develop a long term commitment to specific
fields of research, ii) the availability of a minimum level of resources
and support for carrying out research in those fields, and iii) open
possibilities for acquiring an identity as a member of a self-generating

community of researchers.

208

i. Long-term commitments. Opportunities for developing long
term commitments to specific research fields may be restrained in the
environments of Latin American agricultural research, because of com-
pounded effects of the conceptions and values held at the societal level
with respect to scientific activities, and interferences of bureaucratic
mandates.

On an abstract level, it is very likely that SGT endeavors are
highly valued by most sectors of Latin American societies. Nevertheless,
these abstract valuations may result in mainly idealized conceptions of
scientific activities without sensitivity to the conditions that enable
them to occur. Perhaps the more concrete images of S&T in Latin America
are associated with labs, sophisticated equipment, computers, journals,
books, and libraries, surrounding eminent scholars. What is observable
by outsiders about the actual work of an agricultural researcher is not
the creative individual and the collective intellectual processes, but
the pnysical activities inherent in agricultural research. The daily
routines of a middle level agricultural researcher who spends most of his
time planting, growing and harvesting crops might be regarded as too
prosaic a set of tasks to fit in theidealized conception of science.
Consequently, people-~particularly decision makers at the national level--
could wonder why they should give much thought and attention to the long-
term commitments of a professional career of a number of public servants

whose more visible activities are traveling in a jeep and making visits
to crop fields. Ironically, the idealized conception of science may lead
to very simplistic understandings of research work. Continuity of research
may be considered as identical to continuity in a job, ignoring that what

matters in scientific research is not the inherent physical activities,

209

but a set of cognitive and social processes. These are the processes by
which a person internalizes specific research concerns; keeps aware of
past and current work done in the field; creates and tests images about
empirical relationships, elaborates conceptual frameworks; engages in
collaborative work with others interested in similar or related cognitive
goals; shares approaches, findings, interpretations, assessments, and
accomplishments with significant colleagues; and, in so doing, contributes
to the building of SGT. These cognitive and social processes tsle time.
Without open opportunities for researchers to engage in long-term com-
mitments to specific fields of research, these cognitive and social
processes are not likely to develop.
Long-term commitments may be hampered by bureaucratic mandates

which obligate researchers to switch from one research problem or field
to another; there are occasions when those problems and/or fields may be
totally unconnected. The emphasis here is not with autonomy of profes-
sionals in the selection of research problems, although intellectual
freedom is also an important factor in the pursuing of a scientific career.
By its very nature, agricultural research is expected to make important
contributions to socio—economic development in LDC's; this reality implies
that, in self-consciowsselecting of cognitive goals, researchers need to
make compromises with their personal preferences and with the dominant
concerns within their professions. However, what cannot be compromised,

if research effectiveness is desired, is continuity in the pursuit of a
research goal. Managerial mandates which obstruct continuity frequently
emerge not as purposive will to interfere with research, but as the
result of a diversity of political and economic pressures from inter-

related social systems. There are cases in which researchers seem to be

210

equated with computers which are fed instructions and data, and are
expected to immediately give out the desired information. Jhis "com-
puterized" image of a researcher does not take into account the social-
psychological and cultural forces involved in becoming a creative re-
searcher. Unfortunately, continuity of research-~which is one of the
main functions a formal organization of research is expected to
maintain--is sometimes disrupted by managerial practices. Perceived
lack of opportunities for having continuity in research may induce
professionals to abandon their organizations and/or the activity of
research.

ii. Research resources and support. Here is where research or-
ganizations are expected to play a substantive role in the institution-
alization of S&T. Organizations constitute a fundamental infrastructure
for the activity of research. They provide a home, a collective identity,
and resources, so researchers can work with support, security, and
continuity.

There is much heterogeneity in the level of development of re-
search organizations in Latin America. The range goes from no existence
of organizations expressly designed for research in some countries to
complex systems of local, state, and national research centers (such as
EMBRAPA in Brazil).

This heterogeneity makes one expect that there is a relationship
between researchers' behaviour and the level of research facilities. In
contrast to this expectation, this study disclosed no statistically
significant difference to support this belief. As pointed out in
chapter four, this lack of statistically significant differences may

be due to the gross categories used in the comparisons. The consideration

V

‘F ‘ W A.‘ is

 

211

that most agricultural research is a non-capital intensive activity,
and that most members of the sample studied have at their disposal

at least the minimum requirements of resources for conducting research
may explain the lack of statistical differences in the variable re—
search support facilities.

There are more subtle aspects of research support that may
stimulate migration: a) unrealistic expectations from research ad-
ministrators and segments of a society about project datelines, or
about the quality and quantity of research outcomes; these expectations
may reflect lack of understanding of the uncertainties inherent in
research and can take the form of very short-term planning and/or
inadequate reserch implementation; b) hidden organizational policies
against the building of professional careers in agriculture; c) per-
ception by researchers that they and their work are not held in esteem
in their organizations, profession, countries, or local communities;

?) political influences over professional appointments and promotions;
d) lack of identity of researchers with the product of their work
and/or loss of intellectual property; f) lack of a communication en-
vironment which facilitates interaction among colleagues; and g) per-
ception by agricultural researchers that they lack a feeling of self-
confidence and security as members of their organizations and
professions.

iii. Self-supporting communities of researchers. A factor
which appears not to be well recognized for facilitating the building

of research careers is the existence and possibility of participation

212

in self-supporting communities of researchers. This topic refers to the
researchers' networks. Interpretations of the findings on networks are
the substance of the following section.

3. Interpreting findings on research networks. Three questions
of the social structures and third cultures created, maintained, and
shared by this segments of Latin American agriculturists are especially
noteworthy: a) how do these linkages arise, grow, and develop?,

b) what are the uses and purposes of these networks?, and c) why do
these research communities arise, grow, and develop?

a. How do linkages emerge and evolve? This section attempts
to speculate about the process of research network development (RND).
The arousal of nascent linkages may presumably begin in a diversity
of occasions. For instance, some links are established very early
in the life of a person' as member of primary systems; these include
kinship, religious, local community, first schooling, and other
kind of ties.

A substantive number of interpersonal links among middle
level agricultural researchers in Latin America are created during
the B. 5. studies with teachers and peers. However, many of these
connections do not evolve; they remain latent or eventually fall
apart. This, in particular; may be the case of latent or no
’significant ties with former B. S. teachers, who in most Latin

American organizations of agricultural education do not have
adequate opportunities and rewards for pursuing research
endeavors.

Probably a high proportion of ties which become intensive

are created when practitioners are in the process of entering

 

 

 

213

specialized fields of research, and when they are trying to keep them-
selves up-to-date on the developments of their current fields of
specialized work. This may occur during postgraduate studies, but for
this particular sample, many of what eventually evolved as intensive
ties are relationships that were initiated with instructors and fellow
students during training programs at CIAT for entering specialized
fields of research. Ties established with former professors at the
Master and PhD level are highly influential, often resulting in deep
interchanges, but are low in the frequency of interactions (intensive
sub-type one) and in the number of persons engaged. The low frequency
might be attributed to lack of a facilitative communication environment
for making interchanges feasible, but the evidence gained in this study
in this respect is mixed. The low number of people engaged in intensive
interchanges with former professors at the Master of PhD level is
explained by the fact that most members of this study population do not
undertake postgraduate studies.

Some links emerge from literature reviews when a professional
becomes aware of the existence of persons conducting significant work
in his/her field of interest, makes the decision of trying to contact
those colleagues, has the opportunity to do it, and finds reciprocity
in some of them. Other connections are established at the moment of
entering a new organization of research, in the planning and executing
of inter-organizational formal contracts for conducting coordinated
- research, during the interaction with consultants, and at professional
meetings.

The number of research groups identified in this study, and the

fact that most of them are specialized, supports the conception that

214
what brings these people together is the sharing of their cognitive
goals and experiences, that is, their research problems and findings.
The point of departure for starting the building of functional linkages
with significant colleagues is, then, the awareness by practitioners of
the existence of other researchers who have similar cognitive interests
to theirs.

A second step is the acquaintance with the researchers personally
and with the work they have completed or that is in progress. This is
done through face-to-face encounters, by correspondence, or by studying
published work. At the level of acquaintance, in addition to knowing
work characteristics, practitioners learn about each other's national,
organizational, and professional identities. Availability of directories
of researchers with names, organizational affiliations, current
addresses, and indications about work interests facilitate these pro-
cesses of awareness and acquaintance. This is the level that Miller
(1975) calls socio-cultural in the development of an interpersonal
relationship. Nascent ties are typically at this level.

Additional and continued interactions allow researchers not only
to learn more about each other's work activities and interests, but to
get insights about each other's personalities. Then relationships
become established at a more psychological level. Nascent ties start
transforming into intensive linkages. Functional interchanges for work
activities and for personal support begin to appear.

As interchanges gain frequency and depth, inconsistencies among
practitioners' conceptions of research work may emerge. These incon-
sistencies are frequently identified with communication barriers, such

as problems in terminology, imcompatible personalities, lack of

215

communication skills, negative attitudes toward team work, or excessive
individualism on the part of some researchers. However, communication
difficulties are only symptoms; these inconsistencies reflect deeper
philosophical disagreements about the whole conception and practice of
scientific research; disagreements which often materialize as conflicting
views on research priorities, approaches, and techniques. As a result,
at this stage ties may decay and extinguish or stay at a latent level.

These disagreements are not mere confrontations of personalized
views and individualized patterns of behavior. They arise from differing
views predominant in segments of the larger social structures and
cultures in which researchers are embedded, and with which they identify
themselves, e.g., the profession, the organization, sectors of the
economy, political parties, nationalistic views, and so on.

An intense and open debate or submission on the part of main—
tainers of the conflicting views is required so a minimum of consensus
can begin to emerge among members of the research network. An example
of these types of conflicts has existed for several years among sectors
of this study sample in the fields of rice and beans. In rice the
central issue has been the generation of appropriate technologies for
farmers who produce rice under non—irrigated conditions; in beans a
recurrent debate is the generation of appropriate technologies for
small farming, particularly in reference to crop associations between
beans and other compatible Species.

Interpersonal relationships which survive these confrontations
and continue to be intensive gain endurance and functionality. "Invisible
colleges" begin to be structured by those practitioners who have not

only a common cognitive focus in their research interests, but who

216
share views about the relevant priorities and the appropriate ways of
conducting research in their specialized fields.

At this stage, collaboration in research ceases to be an abstract
aim, and starts to be translated into concrete sets of concerted
activities. Well structured research groups come to exert leadership
and influence within their fields. Up to this point, a community of
researchers is no longer a metaphor, but a viable, concrete, self-
supporting entity whose existence and continuity of operation represents
a solid step in the institutionalization of S&T in a given society or
in a set of inter—dependent societies.

At all stages in the devolopment of a research network links
continue to emerge and transform. Some of them become intensive, others
decay and stay latent or even disappear. In turn some latent links
may be activated by new research opportunities and events among members
of the research network. A high proportion of extinction of intensive
ties may indicate the decline of the entire research paradigm as a
consequence of drastic internal and/or external changes.

b. The uses and purposes of these research networks. The
content of research interchanges mentioned by the sample members inter-
viewed give origin to a long list from which the uses and purposes of
these networks can be inferred. These uses and purposes are summarized
in two sets: practitioners engage in network interchanges to obtain
i) research resources, and ii) socio-psychological support.

i) Resources. There is a variety of research resources that
practitioners with nascent and intensive ties interchange: materials,
particularly plant germ plasm; services, especially lab analyses and

data processing; information about research sponsors; names and

217

addresses of other colleagues in the field; and published research
literature (it seems that many people get access to the relevant litera-
ture more by personal contacts than through libraries).

Other resources are interchanged mainly by members of invisible
colleges: pre-prints, most of the time asking for comments and sugges-
tions; raw data about experiments and field observations which are just
finished, but for which analysis and interpretations has not been
completed by the investigator; interpretation and assessment of findings;
and views about research priorities and approaches.

An item of particular importance in the more intensive inter-
changes among members of research groups is what may be designated as
interchange of work. More active members of these research networks
sometimes engage in collaborative personal research agreements in which
an experiment is replicated by different practitioners with differing
organizational affiliations and, on occasions, working in different
countries.

ii) Socio-psychological support. The category called here
socio-psychological support is perhaps what keeps these sets of people
together and makes possible the development of social structures and
cultures which eventually emerge as identifiable communities of research-
ers. Socio—psychological support at the beginning of a relationship--
when links are still at a nascent stage or starting to gain intensity--
consists of help provided by more mature researchers to new entrants.

In many cases this type of support is given by helping new practitioners
identify training, educational, and career opportunities; providing
letters of recommendation for entering graduate schools; helping secure

funds and invitations for participation in professional meetings, local,

218
national, and international; instructing about where and how to apply

for scholarships and assistantships; and helping newcomers develop their
own personal identities as practitioners of particular fields of research.

At subsequent stages of development of these interpersonal
relationships, socio-psychological support implies some degree of mutual
reliance; more mature researchers take the risks involved in recommend-
ing fellow researchers for job appointments in places where those more
mature persons have achieved prestige. A recommendation for a research
position means much more to a practitioner than an employment opportunity;
it means that the one who is recommending him/her is showing a signal of
reSpect and appreciation; recipients of this type of social-psychological
support feel that they are held in esteem by people in what they start
to consider "their" fields of research.

Finally, more developed linkages--typical of invisible colleges--
are both highly professional and highly personalized. At this stage the
interpersonal support implies high reciprocity, and it is characterized
by the presence of a strong social-psychological factor, which in
Spanish is well described by the word "confianza." "Confianza" implies
more than reliance and confidence; perhaps "trust“ is its equivalent
in the English language. Trust among fellow researchers opens opportuni-
ties for them for interchanging specifics about research approaches and
techniques that they use, although these techniques and approaches are
still not widely accepted or standardized; opportunities arise for
- interchanging details of projects which failed; details about what would
be considered "negative" results if published; about problems and
confrontations with research administrators and politicians; about

forthcoming opportunities perceived as promising in the field or in

219

neighboring fields; about projects in the fields perceived as very risky
undertakings; and, particularly, about philosophical views on what

should and should not be research priorities in their specialized fields.
However, trust in the interpersonal relationships among members of
research networks does not imply a blind loyalty which avoids objectivity
in pointing out the professional concerns with each other's research
activities and outcomes. Researchers who trust each other openly analyze
their work and make competent criticisms about it; but those interchanges
are personalized, and remain among the interactants who trust in each
other's loyalties, so this type of information will not be used against
any one of them in the profession, in their organization, or in their
other social milieus.

c. Why do these research communities arise and develop? Four
sources of explanations are presented in the following paragraphs in
relation to the emergence and development of research networks:

1) dysfunctionalities in the formal structures; ii) the nature of some
research goals which are inherently transorganizational and transnational;
iii) costs and economies of scale and size; and iv) risks and uncertain-
ties inherent to scientific research.

i) Dysfunctionalities. At the intra-organizational level, the
emergence of non-formalized structures may be associated with dysfunction-
alities in the distribution of power, influence, and resources through
the formal organization chart; or the organigram may seem adequate but
: for some reason it simply does not work. This seems to be the case in
some Latin American organizations of agricultural research having a
highly structured distribution of positions from the national to the local

level. These structures usually establish as "link pins" between the

220
central hierarchy and the research teams at the local level a set of
“national coordinators" (the equivalent of "liaisons" in the research
networks) who are not always adequately equipped with the necessary
budgetary flexibility to operate, or lack formal authority to
"coordinate" research activities and projects.

Dysfunctionalities of the formal structures in research
organizations are a complex difficulty, which involves more than simply
systematic arrangements of positions and job descriptions. Recently,
Theodore W. Schultz (1979) has pointed out how the entrepreneurial
function, which is responsible for the dynamics of innovation and
progress in society, operates differently in businesses and research
organizations. In business enterprises, the centralized body of high
level executive officers performs the entrepreneurial function, while
the skilled factory workers do the unchanging, routine tasks. But
research is otherwise. Administrators may be entrepreneurs in fostering
the organizational infrastructure for research activities but the
dynamics of innovation and progress in the specialized fields of research
is beyond administrators' possibilities. Research entrepreneurship is
the function of the research worker. A critical question concerning the
design of formal organizations of research, then, is: To what extent
is it viable and advisable to "coordinate" by mandatory acts? In the
words of Professor Schultz:

The convenient assumption is that a highly organized research
institution firmly controlled by an administrator will perform
this important function. But in fact a large organization that
is tightly controlled is the death of creative research,
regardless of whether it be the National Science Foundation,

a government agency, a large private foundation, or a large,
research-oriented university (Schultz, 1979, p. 7).

 

 

 

221

Dysfunctionalities in formal structures of research, in addition,
may relate to lack of or weak mechanisms for providing adequate recom-
pensations to researchers for their work and accomplishments. Another
important component of this problematic situation is related to the
extent to which the formal organization has adequate mechanisms for the
allocation of professional recognition and rewards to its research workers.
As pointed out by Swanson (1975, quoted on p. 75), on occasions research-
ers in Latin American organizations of agricultural research lose their
identities in theirwork. This loss of identity sometimes occurs as
the result of very limited availability of channels for publishing
research findings. There are organizations in which the annual report
constitutes almost the only channel available. As a general practice,
most annual reports do not associate the work reported with any person
in particular. Practitioners involved in these situations feel that they
are not held in esteem by their organizations. Interpersonal relation-
ships with significant colleagues may be regarded as a substitute for the
social-psychological support that their organizations are failing to
provide them.

Beyond the boundaries of organizations, at the country, region,
and worldwidelevels, dysfunctionalities in formal structures which
stimulate the emergence of research networks may be related particularly
to lack of active professional associations, lack of or weak formal
communication systems of science, and lack of accessibility of these
. sample researchers to the international journals of the world epicenters
of science. These journals may have very low priority for publishing
the type of outcomes generated by these practitioners, which are for the

most part adaptations of previously generated technologies.

222

ii) The nature of research goals. This is another possible
explanation causing high networking activity. There are research concerns
which transcend the boundaries of a particular organization, local
community, country, and even region. This seems to be the case of
practitioners engaged in research in rice and tropicalpastures.

iii) Costs. Related to the above-mentioned factor is the issue
of research projects which may be extremely costly if performed by a
single organization or country, but may be jointly pursued by members of
different organizations who are already scattered through wide geographi-
cal regions and in a diversity of ecological conditions.

iv) Risk and uncertainty. In its own essence research is a
human enterprise conducted under conditions of high risk and uncertainty.
As pointed out elsewhere, researchers may not be willing to share their
uncertainties with formally appointed research leaders in their organiza-
tions, because this information could be used against researchers'
professional and bureaucratic careers. So they do not trust in the
formal structures to reduce the psychological stresses and tensions
induced by the inherent uncertainties and risks of their research
projects. Consequently, they develop interpersonal relationships deep
enough so they can trust and receive personalized understanding and
support from others who have experienced similar risks and uncertainties.

4. Non-participation. This last part of the chapter turns the
attention to the phenomenon of non-participation. Two aspects are
~interpreted: a) why a proportion of research practitioners do not
participate in these communities, and b) What may be some consequences

of non-participation?

223

a. Why people do not participate. This is a much more specula-
tive section than the preceding, because non-participants have not been
interviewed for this study. Before presenting a series of possible
reasons for non-participation a methodological note related to the
quantification of non-participation is introduced here as a kind of
brief digression.

The number of persons who do not participate in networks of
researchers (sixty-six out of the 207 in the sample of this particular
study) suggests an extended pattern of isolation for subsectors of the
population. _However, the magnitude of non-participation may only be
apparent, given that a large portion of the population is not included
in this analysis ( p. 146). It is possible that the extent of non-
participation is lower in the population than in the sample. The reason
is that non-participants in the presence of more respondents may
receive nominations, increasing their linkage in that way. Depending on
the intensity of such connections, non-participants might emerge, in
reclassification, as participants. The highest change in this respect
may occur in tropical pastures, which shows 45% of respondents identified
as non-participants. The opposite may also occur. The 207 questionnaires
included in the analysis correspond to the earlier respondents. A bias
could exist associated to response of questionnaires. For instance,
delay in responding could be positively correlated with non-participa-
tion. Therefore, if this be the case, the addition of more respondents
~ to the analysis probably would increase the proportion of non-participants.

Returning to factors likely to be associated With non-
participation, the following is a list of possibilities. Under the

assumption that non-participation is relatively high (about one third

224

in this study sample) the first reason may be lack of a facilitative
communication environment. This includes lack of actualized directories
of persons in the research fields surveyed, low activity of professional
associations within countries and in the region, organizational policies
consciously or unconsciously militating against research interchanges
among colleagues, lack of funds for traveling or using telephone and
mail systems, mail system difficulties, and visa problems, among others.

Additional reasons for non-participation in networks of
researchers may include: personality characteristics of some researchers
who prefer work and stay in relative isolation, excessive nationalism
of some researchers who regard with suspicion the transnational
activities involved in networking, ideologies in conflict with inter-
national collaboration and with a view of inter-dependence among nations,
patterns of authoritarianism and dogmatism perceived by some non-
participant researchers as predominating among groups of very active
colleagues, excessive competence in some fields (which may result in
patterns of secrecy), excessive bureaucratization which frustrates any
attempt to participate in professional meetings, and fields composed of
"one-of—a-kind" researchers (that is, fields where a "critical mass" of
colleagues still has not developed).

b. Consequences of non—participation. Several possible con-
sequences may result from non-participation. First, lack of contribution
to 5&T. As pointed out by Goulet (1975), 5&1 activities normally occur
- in a collective social context. Non-participants, as defined in this
study, are not transmitting by personalized interchanges the research
findings to the mainstream of colleagues. Accordingly, unless non-

participants publish their findings in the formal system of communication--

225

journals, books, etc.--their isolation deprives their scientific
communities of their contributions. Given that most members of this
study population are young professionals in the first stages of their
research careers, and given the limited number of journals and scienti-
fic publications available in Latin America, it is very likely that the
research findings of non-participants are not accessible to their
colleagues.

Second, SGT obsolescence. Non-participation in networks of
relevant researchers could also be interpreted as suggesting that a
segment of former CIAT trainees may be unaware of the technical progress
that their research associates are achieving in research endeavors
pertinent to theirs. Such isolation may result in rapid obsolescence of
previous training.

Third, SGT lag. Non-participants may be getting access to the
SGT progress through the formal communication system, via printed and
other media. This system is not consistently characterized by speed.
While its better performed functions are giving professional recognition,
assessment, and permanence to scientific advances, the last issue of a
scientific journal often reports research carried out and findings
obtained several years ago. In addition, the formal system is not
commonly the best way to get full access to the specifics of research
technology. Fine descriptions of research instruments, techniques, and
procedures are usually not published in scientific journals, but
' typically diffuse via personalized interchanges among colleagues. This
is particularly the case in applied agricultural sciences. In most
applied fields much of the flow of knowledge comes not through formally

published books and journals, but through correspondence, pre-prints,

226
mimeographs, findings, and data, which circulate by personalized inter-
changes. Researchers in rapidly growing fields, who rely entirely on
journals and books for their information may confine themselves by
continually lagging behind in scientific knowledge.

Fourth, cultural dependence. One consequence of over-relying
on established international journals and books published in the world
epicenters of science is cultural dependence. Latin America still has
not developed an adequate formal system of scientific communication.
Consequently, most literature comes from other cultures. A mere
consumerism of foreign scientific literature does not provide a full
opportunity for entering in the creative process of interaction between
cultures, which results in the development of third cultures of science
and technology. An over-dependency on the foreign scientific literature
can impede the generation of indigenous capabilities for SGT in the
region.

Fifth, lack of identity as practitioners of research. Non-
participants who cannot get recognition and rewards through the formal
communication system, and who--by definition of non-participation--do not
receive personalized socio-psychological support, respect and esteem,
are open to disillusionment in their research activities which, added to
other consequences of isolation, might precipitate migration of non-
participants toward activities other than research. Non-participation
and propensity to migrate from research might be two inter-dependent
-factors which reinforce each other through time.

In summary, non-participation in networks of researchers is
regarded in this study as a symptom of scientific isolation. And

isolation might be impinging on the broader aims of encouraging task

 

227

forces of researchers to contribute to agricultural development in

particular, and to the institutionalization of SGT in general.

 

 

 

BIBLIOGRAPHY

 

 

BIBLIOGRAPHY

Ardila, J., N. Reichart x A. Rincon (1980). "El Caso del Instituto
Nacional de Tecnologia Agropecuaria en Argentina (INTA)," in
Sistemas Nacionales de Investigacioh Agropecuaria en America Latina:
Analisis Comparativo de los Recursos Humanos en Pafses Seleccionados.
Bogota, IICA, PROTAAL, Documento No. 48.

Ardila, F., E. Trigo y M. Pineiro (l980). “Resumen Metodologico y Planteo
Operativo de la Investigacion," in Sistemas Nacionales de Investiga-
cidn Agropecuaria en America Latina: Analisis Comparativo de los
Recursos Humanos en Paiges Seleccionados. Bogota, IICA Publicacidn
Miscelanea No. 222, PROTAAL, Documento No. 46.

Ardila, J., R. Torres y E. Trigo (1980). "El Caso de la Universidad
Agraria de La Molina del Peru," in Sistemas Nacionales de Investiga-
cion Agropecuaria en América Latina: Analisis Comparativo de los
Recursos Humanos en PaiSes Seleccionados. Bogota; IICA, PROTAAL.
Documento No. 49.

Ardila, 0., E. Trigo, R. Torres, M. Pifieiro y A. Rincdn (l980). "El Caso
del Instituto Colombiano Agropecuario (ICA)," in Sistemas Nacionales
de Investigacion Agropecuaria en América Latina: Analisis Comparativo
de los Recursos Humanos en PaiSes Seleccionados. Bogota, IICA
Publicacion Miscelanea 223, PROTAAL Documento No. 47.

 

Asch, S. M. (1975). "Transnational Networks: A Multimethod Study of the
American Associates of Foreign Students," PhD Thesis, East Lansing,
Michigan State University.

Axinn, G. H. (1978). New Strategies for Rural Development. DeWitt,
Michigan, Rural Life Associates.

 

, (1976). Non-Formal Education and Rural Development. Supple-
mentary Paper No. 7, Program Studies in Non-Formal Education,
East Lansing, Michigan State University.

Barnes, J. A. (1954). "Class and Communities in a Norwegian Island
Parish," Human Relations, 7 (February): 39-58.

Barber, 8. (1963). Science and the Social Order. New York, Collier
Books.

Basalla, G. (1967). "The Spread of Western Science," Science, 156
(May 5): 61l-62l.

228

 

 

229

Beltran, L. R. (1976). "Alien Premises, Objects, and Methods in Latin
American Communication Research," in Comnunication Research, 3 (2):
107-134.

Bernal, J. D. (1939). The Social Function of Science. Cambridge,
Massachusetts, The MIT Press, edition 1967.

 

Berry, R. A. and W. Cline (l979). Agrarian Structure and Productivity
in Developing Countries. Baltimore, John Hopkins University Press.

Blaise, M. G. (1973). Institution Building: A Source Book. East Lansing,
MUCIA, Michigan State University.

Boyce, J. K. and R. E. Evenson (1975). Agricultural Research and Exten-
sion Programs. New York, Agricultural Development Council.

 

 

Byrnes, F. C. (1963). "Americans in Technical Assistance: A Study of
Men's Perceptions of Their Cross-Cultural Experience," PhD Thesis,
East Lansing, Michigan State University.

, (1974). I'Agricultural Production Training in Developing
Countries: Critical but Controversial," in Strategies for Agricultural

 

 

Education in Develpping Countries: Agricultural Education Conference I,
New York, The Rockefeller Foundation, working paper.

, (1968). "Some Missing Varables in Diffusion Research and
Innovation Strategy." New York, Agricultural Development Council,
Report.

CEDA, Center for Economic Development and Administration (1971).
"Proceedings: Seminar in Institution Building and Development,
June 26 to June 30, 1971,” Kintiput, Kathmanda, Nepal, The Centre,
Study Series Seminar Paper No. l.

CIAT, Centro Internacional de Agricultura Tropical (1981). CIAT en la
Decada de los Ochenta. Cali, CIAT.

 

 

, (l979). CIAT's Strategy for Outreach Services. Cali, CIAT.

Cole, S., and J. Cole (1968). "Visibility and the Structural Bases of
Awareness of Scientific Research," American Sociological Review,
33: 397-412.

 

Crane, D. (1972). Invisible Colleges. Chicago: Chicago University Press.

 

Diaz Bordenave, J. (1976). "Communication of Agricultural Innovations
in Latin America: The Need for New Models," Communication Research,
3 (2): 135—154.

 

Drilon, J. 0. Jr. (1977). Agricultural Research Systems in Asia. College,
Laguna, Philippines, Southeast Asian Regional Center for Graduate
Study and Research in Agriculture.

 

 

230

Duncan, R. and G. Zaltman (1973). "Values and Ethical Dilemmas in
Implementation," presented at the Implementation of OR/MS Models
Theory, Research and Application Conference, University of Pittsburgh.

 

EMA, Ecuadorian Ministry of Agriculture, and A10, the Agency for
International Development (1979). "Baseline Study of Agricultural
Research, Education and Extension in Ecuador," Quito, Ecuador,
prepared by a team involving the Ecuadorian Ministry of Agriculture,
the United States Agency for International DevelOpment, the United
States Department of Agriculture. Oklahoma State University, and
the University of Missouri.

Esman, M. J. (1967). "The Institution Building Concepts: An Interim
Appraisal," Pittsburgh, University of Pittsburgh, Graduate School
of Public and International Affairs, University Research Program
in Institution Building.

Farace, R. V. and T. Mabee (1980). "Communication Network Analysis
Methods," in Monge, P. R. and N. Capella (eds.), Multivariate
Technigpes in Human Communication Research. New York, Academic
Press, pp. 364-391.

 

 

Farace, R. V. et a1. (1977). Communicating and Organizing. Reading,
Massachusetts, Addison-Wesley.

 

Felstehausen, H. (1971). "Conceptual Limits of Development Communica-
tion Theory,‘I a paper presented at the Association for Education in
Journalism. Columbia, South Carolina.

Fernandez, F. (1977). "Objectives and Content of Training at the
International Centers of Agricultural Research," paper prepared for
the CGIAR at the Forum on Training, International Centers Week,
September 1977. Washington, D. C., mimeographed.

 

Fernandez y Fernandez, R. (1976). Chapingo Hace 50 Anos. Chapingo,
Mexico, Escuela Nacional de Agricultura, Colegio de Postgraduados,
Centro de Economia Agricola.

 

Fienup, D. F. and H. M. Riley (1980). Training Agricultural Economists
for Work in International Develppment. A report based on a study
sponsored by the American Agricultural Economics Association,

New York Agricultural Development Council.

 

Flit, I. (1979). La Investigacion Tecnologica y el Desarrollo Industrial.

 

Lima, Peru, ITINTEC.

Gamble, W. K. (1981). "National Agricultural Research Systems and the
International Service for National Agricultural Research (ISNAR)."
Novi Sad, Yugoslavia, paper prepared for the Seventh Ministerial
Session of the World Food Council.

231

Garvey, W. D. and B. C. Griffith (1971). "Scientific Communication:
Its Role in the Conduct of Research and Creation of Knowledge,"
American Psychologist, 26 (April): 349-362.

Garvey, W. D. and B. C. Griffith (1967). "Scientific Communication as a
Social System,“ Science, 157: 1011-1016.

Gomez, M. R. (1976). Episodios de la Escuela Nacional de Agricultura.
Chapingo, México, Escuela Nacional de’Agricultura, Colegio de
Postgraduados, Centro de Economia Agricola.

 

Goulet, D. (1975). "The Paradox of Technology Transfer," Bulletin of
the Atomic Scientists, 31 (6): 39-46.

 

 

, (1977). The Uncertain Promise: Value Conflicts in Technology
Transfer. New York, IDOC in Cooperation with the Overseas
Development Council.

 

Granovetter, M. (1976). "Network Sampling: Some Fist Steps," American
Journal of Sociology, 81: 1287-1303.

 

 

Grunig, J. (1968). "Communication and the Economic Decision Process of
Colombian Farmers." Madison, University of Wisconsin, Land Tenure
Center.

Hagstrom, W. O. (1965). The Scientific Community. New York, Basic Books.

 

Hall, G. R. and R. E. Johnson (1971). "Transfer of U.S. Aerospace
Technology to Japan," in Vernon, Raymond (ed.) The Technology Factor
in International Trade. New York, Columbia University Press.

 

 

IADS. International Agricultural Development Service (1976). Report 1975.
New York, IADS.

 

, (1977). Report 1976. New York, IADS.

 

, (1978). Report 1977. New York, IADS.

 

, (l979a). Report 1978. New York, IADS.

 

 

, (l979b). Preparing_Professional Staff for National Agricultural
Research Programs. Report of a workshop, Bellagio, Italy. New York,
IADS.

, (1980). Report 1979. New York, IADS.

 

, (1981). Report 1980. New York, IADS.

 

IFPRI, International Food Policy Research Institute (1977). "Food Needs
for Developing Countries: Projections of Production and Consumption
to 1990.“ Research Report No. 3.

Jennings, Peter 1976). “The Amplification of Agricultural Production."
Scientific American, 235 (3): 181-194.

 

232

Kuhn, T. S. (1964). The Structure of Scientific Revolutions. Chicago,
University of Chicago Press.

Law, J. (1973). "The Development of Specialities in Science: The Case
of X-ray Protein Crystallography," Science Studies, 3 (3): 275—303.

 

McCarthy, F. (1972). "Third Cultural Networks of Philippine Physical,
Life, and Social Scientists.“ PhD Thesis, East Lansing, Michigan
State University.

McDonoughs, A. M. (1963). Information Economics and Management Systems.
Tokyo, Kogahusha.

 

Merton, R. K. (1942). "Science and Democratic Social Structure," in
Merton, R. K., Social Theory and Social Structure (1949). New York,
The Free Press.

 

Miller, G. R. and M. Steinberg (1975). Between People: A New Analysis
of Interpersonal Communication. Chicago, Science Research Associates.

 

 

Mitroff, I. I. (1973). "On Mutual Understanding and the Implementation
Problem: A Philosophical Case Study of the Apollo Moon Scientists,"
presented at the Implementation of OR/MS Models Theory, Research and
Application Conference, University of Pittsburgh.

 

 

Moravcsik, M. J. (1975). Science Development: The Building of Science
in Less Developed Countries. Pasitam, Bloomington, Indiana University
Press.

 

 

Morris, R. C. (1974). "Report to the Staff Seminar on CIAT Training
Programs and Policy." Cali, CIAT.

Morse, C. et al. (1969). Modernization by_Design: Social Science in the
Twentieth Century. Ithaca, New York, Cornell University Press.

 

 

Moseman, A. H. (1970). Building_Agrjcultural Research Systems in the
Develppjng Nations. New York, Agricultural Development Council.

 

 

Mosher, A. T. (1966). Getting_Agriculture Moving. New York, Praeger.

 

Mulkay, M. J. (1977). "The Scientific Research Community," in Spiegel-
Rosing, I. and D. de S. Price (eds.) Science,_Technology and Society.
London, Sage Publications, pp. 99-148.

 

, (1972). The Social Process of Innovation: A Study in the
Sociology of Science. London, MacMillan.

 

 

Myren, D. T. (1964). "El Papel de la Informacidh en las Decisiones de
Agricultores Bajo Condiciones de Riesgo e Incertidumbre," in Report
of First Interamerican Symposium of Research on the Functions of
Communication in Agricultural Development. Mexico, The Rockefeller
Foundation, Report 21.

 

 

 

 

 

233

Naranjo, G. E. (1966). "The Graduate Program of the Inter-American
Institute of Agricultural Sciences: An Evaluation of Certain Aspects
Through a Follow—up of Graduates." PhD Thesis, East Lansing,
Michigan State University.

National Research Council (1978). "U.S. Science and Technology for
Development: A Contribution to the 1979 United Nations Conference."
prepared by the U.S. Deparment of State. Washington, D. C.,

U.S. Government Printing Office.

0thman, Abu Hassan (1977). “The Roles, Work Patterns and Third Cultural
Networks of Academic Scientists in Malaysia." PhD Thesis, East
Lansing, Michigan State University.

Pineiro, M. y Trigo (l977). " n Marco General para el Analisis del
Progreso Tecnologico Agropecuario: Las Situaciones de Cambio
Tecnoldgico." Bogota, IICA, Publicacidn Miscelanea 149, PROTAAL,
Documento No. 3.

Price, 0. J. de S. (1963a). "Networks of Scientific Papers," Science,
149 (July 30): 510-515.

, (1963b). Little Science, BiggScience. New York, Columbia
University Press.

 

Price, 0. J. de S. and D. de 8. Beaver (1966). "Collaboration in an
Invisible College," American Psychologist, 21: lOll-1018.

 

Restivo, S. P. and C. K. Vanderpool (eds.) (1974). Comparative Studies

 

in Science and Society. Columbus, Ohio, Charles E. Merrill Publishers.

Rogers, E. M. (1976). "Communication and Development: The Passing of
the Dominant Paradigm," Communication Research, 3 (2): 213-240.

Rogers, E. M. and D. L. Kincaid (1981). Communication Networks: Toward
a New Paradigm for Research. New York, The Free Press.

 

Russel, B- (1931)- The Scientific Outlook., Spanish translation by
G. Sans Huelin and M. Sacristan, published under the title "La
Perspectiva Cientifica“ (1980). Barcelona, Ariel.

Ruttan, V. W. (1971). "Induced Innovation Hypothesis," in The Annual
Report 1971, Economic Development Center.

 

 

Sabato, J. E. and N. Botana (1975). "La Ciencia y la Tecnologia en el
Desarrolo Futuro de America. Latina," in Sabato, J. A. (compilador)
El Pensamiento Latinoamericano en la Problematica'Ciencia-
Tecnologia-Desarrollo-Dependencia. Buenos Aires, Paidos.

a

Sabato, J. E. (1980). "Las Politicas delicas Frente a1 Estancamiento
y a la Transformacién del Agro Pampeano." San Jose, Costa Rica,
IICA, PROTAAL, Documento No. 38.

 

 

 

234

Sagasti, F. R. (1979). "Towards Endogenous Science and Technology for
Another Development," in Development Dialogue 1979 (l).

 

Schultz, T. W. (1979). The Economics of Research and Agricultural
Productivit . New York, IADS, Occasional Paper.

 

, (1978). "What Are We Doing to Research Entrepreneurship?,"
in Hueg, W. F., Jr. and C. A. Gannon (eds.), Transforming Knowledgg
into Food in a Worldwide Context. Minneapolis, Miller Publishing
Company, pp. 96-105.

 

 

, (1964). Transforming_Traditional Agriculture. New Haven,
Connecticut, Yale University Press.

 

Scobie, G. y R. Posada (l977)3 E1 Impacto de las Variedades de Arroz
de Alto Rendimiento en America Latina. Cali, CIAT.

 

 

Shatuck, G. M. (1965). Between Two Cultures: A Study of the Social
Adaptation of Foreign Students to an American Academic Community.
Department of Rural Sociology, Cornell International Agricultural
DevelOpment, Mimeograph 12. Ithaca, Cornell University.

Useem, J. and R. H. Useem (l980). "Generating Fresh Research Perspec-
tives and Study Designs for Transnational Exchanges Among the
Highly Educated,” presented to the German—American Conference:
International Exchanges. Bonn, November 1980.

 

 

, (1967). "The Interfaces of a Binational Third Culture: A
Study of the American Community in India," The Journal of Social
Issues, 23 (January): 130-143.

 

, (1955). The Western-Educated Man in India. New York, Holt,
Rinehart and Winston.

 

Useem, J., J. Donoghue and R. H. Useem (1963). "Men in the Middle of
the Third Culture: The Roles of American and Non-Western PeOple
in Cross-Cultural Administration," Human Organization, 22: 169-179.

Useem, J., R. H. Useem, A. H. Othman and F. E. McCarthy (Forthcoming).
"Transnational Networks and Related Third Cultures: A Comparison
of Two Southeast Asian Scientific Communities," forthcoming in
L. Kriesberg (ed.) Research in Social Movements, Conflicts and
Change, Volume 4.

Vanderpool, C. K. (1971). “Visiting Foreign Scientists in the United
States: The Impact of Systematic and Role Circumscription and
Dissociative Experiences on the Homogeneity of the International
Scientific Comnunity." PhD Thesis, East Lansing,'Michigan State
University.

Wortman, S. and R. W. Cummings, Jr. (1978). To Feed This World: The
Challenge and The Strategy. Baltimore, The Johns Hopkins
University Press.

 

 

 

 

Ll

235

Ziman, J. (1976). The Force of Knowledge. London, Cambridge University
Press.

, (1968). Public Knowlegge: The Social Dimension of Science.
Cambridge, Cambridge University Press.

 

 

1
‘1

APPENDICES

 

 

APPENDIX A

Letter of the Director General of CIAT for legitimizing the study
and the investigator, and for sending the census formats.

236 «11*-

:enrro Internacional de Agriculluro Tropical

Desde sus inicios e1 CIAT ha venido desarrollando
una activa labor de apoyo a la preparacion y actualizacion
de investigadores agricolas en América Latina por medio del
entrenamiento postgrado.

Ahora deseamos hacer una revisidn sistematica de
nuestros esfuerzos de entrenamiento en investigacion, con el
prepfsito de identificar si es necesario introducir cambios
que hagan este entrenamiento mas significativo para los
profesionales, y eSpecialmente mas fitil para los programas
de investigacion que usted dirige.

Con tal fin vamos a realizar un estudio que incluye
tres fases de recoleccion de datos: 1) e1 cuestionario que
adjuntamos a esta carta, cuyo objetivo es determinar la
poblacion del estudio; 2) un cuestionario que sera enviado
subsecuentemente a todos los profesionales que han estado
vinculados a la investigacion deSpués de su entrenamiento en
CIAT; y 3) entrevistas con algunos de esos mismos profesio-
nales. Las entrevistas seran llevadas a cabo por el investi-

ador principal del estudio Ing. Jairo Cano. Se espera que
os resultados estén listos a mediados de 1981, época en la
cual haremos llegar a usted un resumen del mismo.

El éxito del estudio depende en gran medida del
apoyo que usted nos brinde. Al efecto quiero solicitarle que
nos suministre la informacion relativa al cuestionario que
le estamos haciendo llegar con esta carta. Le anticipo mis
agradecimientos por su valiosa ayuda y le reitero nuestro
agrado en continuar colaborando con su institucion.

Atentamente,

QOHN L. NICKEL
i

rector General

 

Apariado Aéteo 6713
Cali . Colombia

 

;;nnuxnmr

Telex: 05769. CIAT CO

Teléfono: Palmira 27231.1

f;21

 

Census

APPENDIX B

instructions.

 

 

‘wm— _._ ‘__~_.

 

237 527 * /./
ML @2129

q
A
‘. . l
\ "
- \ ‘1 11"
1

1...:

V_..
4
h
l

Cenlro Internacional de Agriculluro Tropical .

Call - Colombia

 

“1w"

Estudio del entrenamiento en investigacion, CIA:

Apreciado senor:

Por favor informcnos acerca del profesional cuyo nombre
aparece encabezando cada tabla. Deseamos obtener los siguientes
datos para el periodo posterior a su entrenamiento en CIAT:

a. Si este profesional ha trabajado en la institucion que usted
dirige. aunque su actividad no haya estado relacionada con la
investigacion (punto 1 en la tabla).

b. Si este profesional ha trabajado en investigacion en una de
las siguientes especies: arroz, frijol, pastas tropicales y
yuca (puntos 2-5 en la tabla).

c. Si este profesional ha trabajado en actividades relacionadas
con la investigacién aunque no haya sido en las especies men-
cionadas antes (punto 6 en la tabla).

Usted puede darnos los datos del caso simplemente haciendo
una marca (I) en la celda correspondiente a cada pregunta y afio
para los cuales 1a respuesta sea positiva. Para reducir posibles
confusiones hemos hecho las siguientes definiciones, las cuales
1e rogamos leer antes de llenar las tablas.

Definiciones:

l. Trabaio en el afio X: Significa que el profesional trabajé al
menos parte de ese afio en lo indicado por la pregunta (su
instituciSn, arroz, frijol, pastos tropicales, yuca 0 en
investigacian on general).

- 2. Trabaip en su institucian: Significa que el profesional trabajé

en eseTEfiérpara su institucion, bien sea.en investigacion 0 en
cualquiera otra actividad.

3. Trabaj6 en investigacién on general: Significa que el profesional
trabajoTen una o varies de las siguientes actividades en relacion
con cualquier tipo de investigacion:

(Sigue al reverso)

 

6;;mcmnmw

Telex: 05769. CIAT CO lelelono; Palmira 27C44

 

 

238

a. Como investigador, aunque no haya tenido que ver con las
cuatro especies mencionadas en la tabla.

b. Como instructor de otros investigadores.
c. Como administrador para dar apoyo a la investigacién.

d. Como consultor acerca de asuntos relacionados con la
investigacion.

 

Bjemplo: Supongamos que el profesional Luis Rodriguez estuvo
en CIAT en 1974 participando en entrenamiento. Regrosé a su
institucién y trabajé el resto do 1974, 1975 y 1976 como
investigador en frijol. En 1977 ocupo una posicion adminis-
trativa dando apoyo a la invostigacién. En 1978 y 1979 per-
manecié en la institucién pero so dodicé a actividadcs no
relacionadas con la investigacién. Los datos correspondientcs
a este profesional se anotan en la tabla asi:

 

 

 

 

 

 

 

 

 

 

 

 

F 1
films RODRIGUEZ (1974) W
ffiabajé on e1 afio ‘ 69 7o 71 72 73 74 75 76 77 73 7§
1) en esta institution v’ v7'v/ v’ v’ v’
2) como investigador en ARROZ

) como investigador en FRTJOL t/rv’ v’
4)_como investigador en PASTOS
S) como investigador en YUCA
6 en investigacién en general v’ v’ r/ 9’ 4)

 

 

 

 

 

 

 

 

 

 

 

 

a. / / carccemos por completo dc informacién acerca de este
profesional.

K b. Otras observacioncs: 2)

 

 

 

Es posible que algunos profesionales no hayan regresado
a su institucién después del entrenamiento en CIAT, o que des-
pués de algunos afios pasaron a otras instituciones. Si usted
esta informado acerca de cllos por favor complete las tablas
respectivas. En caso de que carezca por completo de informacién
acerca de ese profesonal, indiquelo en el espacio correspon-
diente al pié de cada tabla. ‘

A continuacién 1e adjuntamos la lista dc profesionales
de su institucién que participaron en cursos de entrenamiento
en CIAT:

 

 

APPENDIX C

1. Letter to send the network questionnaire.

2. Network questionnaire

 

239

   
   

'\ ‘ . . . €E{:::§$? ‘* '
,enrro Iniemocuonol de Agricultura Tropical «nanwoiemem

Cali - Colombia

Desde sus inicios el CIAT ha venido desarrollando una

activa labor de apoyo a la preparation y actualizacion de investigadores
agricolas en América Latina por medio del entrenamiento postgrado.

Ahora deseamos hacer una revisidn sistemética de nuestros
esfuerzos de entrenamiento en investigacion, con el prOpésito de
identificar si es necesario introducir cambios que hagan este
entrenamiento mas significativo para los profesionales y las

instituciones que se dedican a la investigation agricola en América
Latina.

 

Con tal fin estamos realizando un estudio que incluye tres
fases de recoleccién de datos: 1) un cuestionario que ya fue enviado
a las instituciones, cuyo obJetivo es determinar la poblacion del
estudio; 2) el cuestionario adjunto que estamos enviando a usted y a
todos los profesionales que han estado vinculados a la investigation
después de su entrenamiento en CIAT; y 3) entrevistas con algunos de
ustedes. Las entrevistas seran llevadas a cabo por el investigador
principal del estudio Ing. Jairo Cano. Se espera que los resultados
estén listos a mediados de 1981.

En busca de su valiosa ayuda queremos solicitarle que
responda el cuestionario y nos lo envie de regreso. Las respuestas de
cada persona se mantendran estrictamente confidenciales por el
investigador principal quien sera la finica persona con acceso a los
datos. Estos seran procesados en forma agregada, de tal manera que
las reSpuestas finalmente seran presentadas en términos de promedios,
porcentajes y otros estadisticos. No se haran comparaciones que
permitan identificar a personas o institutiones.

Deseamos insistir en la importancia que para el planeamiento
_de las futuras actividades de entrenamiento en CIAT tiene la information

que usted nos proporcione. Contamos con que usted nos brindara tan
valiosa cooperacion. <——~

.M @6120 Atentamente, {VGA—Jfloflai‘ifl .7...

Jairo Cano, Fernando Fernandez

llnve ador Principal Coordinador, Entrenamiento
Michigan State University Postgraduado, CIAT.

 

  

F

2 alum tau: 05769. cut on mm; mm 27w

 

240

Cuestionario No. 2

PREGUNTA 1. SU PARTICIPACION EN INTERCAMBIO DE INVEST/GA-
CION CON SUS COLEGAS

Deseamos conocer su participacidn en intercambio de investigacion con otros
colegas que estén trabajando en su campo profesional. Tal informacion nos
permitira’ explorar la formacién de redes de personas que comparten sus intereses
y recursos en investigacién agricola en América Latina.

Permitanos conocer los nombres de las personas con quienes usted SERIAMENTE
intercambia investigacion. Deseamos mantener tales nombres separados a cuatro
niveles: Centro de su institucién. su pais, América Latina, y el mundo. Por favor
indiquenos con qué frecuencia y profundidad tiene usted tales intercambios.

Por intercambio de investigacion estamos designando no solo el intercambio de
ideas acerca de problemas y metodologias de investigacion, sino también el hecho
de compartir otros recursos tales como especimenes (por ejemplo, germoplasma),

acceso a servicios de equipo y laboratorio. u otros recursos significativos para llevar ‘1
a cabo investigacion.

 

Usted puede darnos una indicacion de la frecuencia de sus intercambios de
investigacion al sefialar una de las siguientes dos categories: FRECUENTE. POCO
FRECUENTE. Por “frecuente” queremos decir varias veces durante un ano. Por
”poco irecuente" queremos decir una vez por afio o menos. USted puede indicarnos
Ia profundidad de sus discusiones e intercambios chequeando una de dos
categorias: PROFUNDO, SUPERFICIAL. Para anotar su informacién utilice la tabial

que esta en la siguiente pagina. En la parte superior de la tabla usted encontraré un
ejemplo.

PRIMER NIVEL: SU lNSTiTUClON. Empiece con su institucion u organizacién. Por
favor anote los nombres de las personas con quienes usted SERIAMENTE
intercambia puntos de vista y recursos acerca de su trabajo de investigacion. Si no
hay ninguna persona, escriba en la coiumna de nombres: "Nadie de mi institu-
cion". Antes de seguir al nivel de pais. trace una iinea bajo el ultimo nombre que
usted escribio o bajo Ia frase "Nadie en mi mstitucién" si ese es el caso.

SEGUNDO: SU PAIS. Por favor continue a nivel de su pals, perofuera de su institu-
cién. Si no existe ninguna persona. escriba ”Nadie de mi pals" y trace una linea
como lo hizo antes para mantener Ios niveles separados.

TERCERO: AMERICA LATINA. Continue al nivel de América Latina, pero fuera de su
pa is. Si no existe ninguna persona, escriba ”Nadie de América Latina". Por favor,
una vez mas trace una linea para mantener los niveles separados.

CUARTO: EL RESTO DEL MUNDO. Finalmente, indique sus intercambios con
investigadores del resto del mundo. pero fuera de América Latina. Si no existe
alguno. escriba "Nadia del resto del mundo”. -

_.n

 

..,..‘.....a, :.-.. .. 1......i
. . i u. . it. \~.u..-\I~ I. \I“\ in\ v....‘|. h I .I\h\il\

 

 

CIIzOU-Illluitiiill -
P 1.1....- 2.. vu,.~i.-u.uqn.ul\ ..1.. 1

 

7.5.1.3.: z... .5366: 235mm.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

— N p
P N P
— N F
p N _
_ N P
— N F .
_ N a
p N P
P N —
1
M. p N p
. N
P N
— N
. N
[llll‘ Ill
IlllllilllilIIIII-IJIIIIIIIII [III]
a G 6 36.523 <5. $8.82 93
cacao 0% as? m? mum?“ mmzoanthmz. mummfioz
a
new 00/ 090 %v 090
9A.! eoo so: 0 ea: 55:2: 2:236 m. an .2295 mm 5.09:5: m.
at 525203 > $5038: 88 com 3563.95 no.2 mac on c2263.;
umEcchi Eucmauoi 2 2mm .NmDoEoOm m5... no @560: 9:6 8:232 .259.ng
c: coo (h: _on 953 0563.25 98.39: > 2.535 8.8 on

 

 

 

“com mcomcom 8mm coo
cofimazmoé. 2.
mo.nEmo.oE_ 9.2

 

 

 

 

 

 

.5683? .25 5660530 .<.5_ on Exam? ”29:06 6239595
on 8568.35 mam an 358 5.86.25 2 .293 23 283m

.. 2%:

PREGUNTA 2. LA DIS TRIBUCION DE SU TIEMPO DE TRABAJO

En términos de la posicion que usted ocupo inmediatamente después de su
participacién en entrenamiento en CIAT. por favor diganos como distribuyé su
tiempo de trabajo durante aquel afio entre las actividades profesionales que se
seflalan mas abajo. Dénos sus respuestas en porcentajes:

Actividades
Por ciento

a. Haciendo investigacion ................................................................. _____%.._.
b. En actividades de apoyo a la investigacion (administracion,

consultoria, entrenamiento, operaciones de campo.

documentacion. u otra Similar) _._—96 _
c. En actividades N0 RELACIONADAS directamente

con investigacion ___%__

Por favor, revise las cifras que aceba de escribir y asegurese de que suman en
total 1 00%.

PREGUNTA 3. EVALUACION DE SU DESEil/IPEIVO PROFESIONAI.

Por favor evaILie su prepio desempeno profesional en investigacién durante el afio
inmediatamente después de su participacién en entrenamiento en CIAT.
ENCIERRE CON UN CIRCULO el numero que en su Opinion describe mejor su
desempefio profesional, de acuerdo a la siguiente escala:

 

O - No es mi caso pues no desempefié este tipo de actividad.
l =- Entre los mas bajos en comparacio’n con el personal profesional
de mi institucion en ese afio.

   

2 = Bastante bajo.

3 = Promedio.

4 = Bastante alto.

5 = Entre los mas altos.

3. Mi desemper’io profesional como investigador ................

b. Mi desempeflo profesional en actividades de
apoyo a la investigacion (administracién,
consultoria. entrenamiento. operaciones de

campo. documentacién, u otra similar) ............................ 0 l 2 3 4 5
c. En actividades NO RELAClONADAS
directamente con investigacién ........................................ O l 2 3 4 5

 

 

 

 

 

 

 

 

PREGUNTA 4. SUS DA 705 PERSONALES

Nombre:

 

a.
b. lnstitucién donde trabaja actualmente:

 

c. Pais:

 

.0-

Numero de aflos de experiencia en investigacion:

 

e. Areals) de investigacién:

E] Arroz D Frijol [:1 Pastos [:IYuca

D Otra(s). Por favor, indique cualles):

 

 

GRACIAS POR SU COOPERACION

242

 

 

 

APPENDIX D

General data about CIAT training program.

 

 

 

 

 

 

243
Table 0-1. Total number of former CIAT trainees distributed a;
geographical regions of the world. '
Region N %
Africa, Asia and Australia 65 4-6
North America and Europe 89 b.3
Latin America 1259 "§2;l
TOTAL 1413 100.0
Table 0-2. Former CIAT trainees from Latin America distributed

by type of training.

 

 

Type of training N 1
Production and extension 201 16.0
Research support 116 9.2
Research 942 74.8

TOTAL 1259 luO.U

 

* This does not include postdoctoral fellows, nor other students whose

training has financially supported by CIAT but carried out at other
_educational organizations.

‘

 

 

 

244

 

 

[l‘l "

 

 

 

 

NNN NHN HNH HN HN HN HN NN NN HH HN H NNNNN
NN N N N N N _Ni Ni .N: .Ni .Ni .mi NHNNNNNNN
N H N N N H N N N N N N NNNNNLN
H N H N N N N N N N N N ._NNN_NNLN
NN NH N N N H N N H N N N NNNUNNNENN NNNHNNNNN
H N N N H N N N N N N N NNNN NNLNNN
NN NH NH N N H N H N N H N . NLNN
N N N N H N N H N N N N NNNNNNNN
NH N N N N N N N H N N N NNNNNN
N N N H N N N N N N N N NNNNNNUNZ
NN NH N N NH N N N N N N N NNNXN:
H N N N N N N N H N N N NNNNNNN
NN N N N N H N N H H N N NNNNNNNN
N N N N N N N N N N N N NN_NN
N H N H H N N N N N N N NNNNNNN
NN N N NH N N N H N N N N NHNNNNNNN
NH N N N N N N N N N N N LNNN>HNN HN
NN NH NH N H H N N N N N N LNNNNNN
NH H N H N N N N N N N N NHNNN
HN NH N N N N N N N N N N NNNN
NN NH N N N H N N N N N H NUNN NNNNN
NNH NN NN NH NH N NH N N N NH N NNNaoHoN
NNH NN NN NH NH N HN N . NN N H N HNNNLN
NN N N N N N H N H N N N NNNNHNN
N N N N N N N N N N N N NNHHNN
NH N N H N N N N N N H N NNHNNNNNN
H H N N N N N N N N N N NNNNNNN
NNNNN NNNH NNNH NNNH NNNH NNNH NNNH NNNH NNNH HNNH NNNH NNNH NLNNNNN
. 2 C C Q.- .
:oNNoHNHou N N N .u .mua NNNNN

No Lama ucm Ncucaou \N NNNNN_NNN_N .NoNLNE< N_NNN sec» NNNNNNLN zucNmNNc No cmNeNz

 

APPENDIX D

General data about CIAT training program.

243

 

 

 

 

Table 0-1. Total number of former CIAT trainees distributed a}
geographical regions of the world. *
Region N %
Africa. Asia and Australia 65 4-5
North America and Europe 89 6.3
Latin America 1259 .99-1
TOTAL 1413 100.0
Table 0-2. Former CIAT trainees from Latin America distributed

by type of training.

 

 

 

Type of training N Z
Production and extension 201 16.0
Research support 116 9.2
Research 942 74.8

TOTAL 1269 luO.U

 

* This does not include postdoctoral fellows, nor other students whose

training was financially supported by CIAT but carried out at other
.educational organizations.

244

 

 

 

 

 

 

 

NNN NHN HNH HN HN HN HN NN NN HH HN H NNNNN
NN N N N N N N a: N: Ni _mi .mi NHNNNNNNN
N H N N N H N N N N N N NNNNNNN
H N H N N N N N N N N N .NNNNNHLN
NN NH N N N H N N H N N N NNNUNNNENN NNNHNNNNN
H N N N H N N N N N N N NNNN NNLNNN
NN NH NH N N H N H N N H N . NLNN
N N N N H N N H N N N N NNNNNNNN
NH N N N N N N N H N N N NNNNNN
N N N H N N N N N N N N NNNNLNUNZ
NN NH N N NH N N N N N N N NNNXNN
H N N N N N N N H N N N NNNNNNN
NN N N N N H N N H H N N NNLNNNNN
N N N N N N N N N N N N _NHNN
N H N H H N N N N N N N NNNNNNN
NN N N NH N N N H N N N N NHNNNNNNN
NH N N N N N N N N N N N LNNN>HNN HN
NN NH NH N H H N N N N N N LNNNNoN
NH H N H N N N N N N N N NHHNN
HN NH N N N N N N N N N N NNNN
NN NH N N N H N N N N N H NUNN NNNNN
NNH NN NN NH NH N NH N N N NH N NHNNNHNN
NNH NN NN NH NH N HN N . NN N H N HNNNLN
NN N N N N N H N H N N N NNNNHNN
N N N N N N N N N N N N NNHHNN
NH N N H N N N N N N H N NNHNNNNNN
H H N N N N N N N N N N NNN_N=<
NNNNN NNNH NNNH NNNH NNNH NNNH NNNH NNNH NNNH HNNH NNNH NNNH NLHNNNN
. E 55.8
NNNNNHN No N . . ,H .NNN NHNNN

No .53. 2:. 32:8 3 25323.6 .3295 :33 EN: 32.8.5 :88ch No $95.2

APPENDIX E

Additional tables on the formation of

human resource inventories for agricultural
research in Latin America and the Caribbean
countries.

246

 

NNN NNN NNH NNH NHH NN NN NH NH N NH N N H NNN NNNNN
NNI NNN HINI NNIH. q: q: q: Ni NI NI N... q: NI N. .3 NNNH
NN NNH NN NH NN N N N N N N N N N HNH NNNH
NN NN NN N N NN N N N N N N N N HN .NNNH
NN NN NN N N N NN N N N N N N N HN NNNH
NH NN NH N N N N N N N N N N N HN NNNH
NH NN NH N N N N N N N N N N N HN NNNH
N NH N N N N N N N N N N N N NN NNNH
NH HN NH H N H H N N N NH N N N NN NNNH
H NH N H N N N N N N N N N N HH HNNH
N NH N H N N N H H N N H N N HN NNNH
N H N N N N N N N N N N N H H NNNH

 

NNONN N0 N H .M. .m: N N N .Ni .Ni .mi .NN .NN NNN LNNN

 

 

 

 

zocx HNNNN - i -NNNLN NNNNNNNN
p.20o N No; NNNNNNNN No NNNNNN NN.mNNNLN3 NLmelwo LmNeNz NNNNHN
NNNN -ENU

 

.coNumHNEou chcNNNu No cum» No UNNNNNLNNNN .NLNNN No LNNeN: cm>Nm N ou a: NUHNNN NNNNNNLN
.505 5 3.3.33 333. on; 3.29% F53 so: NNNENS :83ng 23 3.8.8 No .5952 .Nim «EN;

249

Table E-5.Inventory formation for the period 1969-79, showing initial
inventories, trained personnel, non-returnees, drop-outs,
and final inventories (n = 580).

 

 

 

For sponsor
organization

 

Initial inventory
Trained personnel
Non-returnees
Drop-outs

Final inventory

1 12 20 49 61 101 123 172 218 308
13 10 31 19 43 34 64 66 129 170
1 5 7 5 10 12

D 1 D 3 2 7 8 15 29 37
121 20 49 61 101 123 172 218 308 429

HOOO—‘O
N
H
N
4:.

Fnr fie1d of -

 

 

 

training
Initial inventory 0 1 7 14 29 4O 64 81 107 143 222
Trained personnel 1 13 10 31 19 43 34 64 66 129 170
Non-returnees O 7 2 15 7 19 13 30 20 29 41
Drop-outs O 0 1 1 1 O 4 8 10 21 40
Final inventory 1 7 14 29 4O 64 81 107 143 222 311

For CIAT commodities

(together)
Initial inventory 0 1 7 14 32 44 69 88 115 158 235
Trained personnel 1 13 10 31 19 43 34 64 66 129 170
Non-returnees O 7 2 12 6 18 12 29 15 26 34
Drop-outs O O 1 1 1 0 3 8 8 26 42
Final inventory 1 7 14 32 44 69 88 115 158 235 329

For agricultural re-

search in general
Initial inventory 0 1 14 22 51 66 108 133 185 241 341
Trained personnel 1 13 10 31 19 43 34 64 66 129 170
Non-returnees O O O 1 1 1 4 5 3 8 4
Drop-outs O O 2 1 3 O 5 7 7 21 37
Final inventory 1 14 22 51 66 108 133 185 241 341 470

 

 

250

Table»E-6.Cumulative trained personnel, final inventories. cumulative
migration, cumulative number of drop-outs, and cumulative
number of non-returnees (n = 580).

 

 

Perspective 196 70

I“
\l
N
\l
w
\l
4:
IN
\1
\J
\1
\1
so

 

 

Sponsor organizations

Cum. trained personnel
Final inventories

Cumulative migration
Cumulative drop-out
Cumulative non-return

14 24 55 74 117 151 215 281 410 580
12 20 49 61 101 123 172 218 308 429
16 28 43 63 102 151
0 l 1 4 6 13 21 36 65 102
2 3 5 9 10 15 22 27 37 49

OOOHH
N
J}
0‘
g...
00

Fields of training

 

 

Cum. trained personnel 1 14 24 55 74 117 151 215 281 410 580
Final inventories 1 7 14 29 40 64 81 107 143 222 311
Cumulative migration 0 7 10 26 34 S3 70 114 138 188 269
Cumulative drop-out 0 0 1 2 3 3 7 15 25 46 86
Cumulative non-return O 7 9 24 31 SO 63 99 113 142 183
CIAT commodities
Cum. trained personnel 1 14 24 55 74 117 151 215 281 410 580
Final inventories 1 7 14 32 44 69 88 115 158 235 329
Cumulative migration 0 7 10 23 30 48 63 100 123 175 251
r‘urnulative drop-out 0 0 1 2 3 3 6 14 22 48 90
Cumulative non—return 0 7 9 21 27 45 57 86 101 127 161
Research in general
Cum. trained personnel 1 14 24 55 74 117 151 215 281 410 580
Final inventories 1 14 22 51 66 108 133 185 241 341 470
Cumulative migration 0 0 2 4 3 9 18 30 40 69 110
Cumulative drop-out 0 0 2 3 5 5 11 18 25 46 83
Cumulative non-return 0 0 0 1 2 3 7 12 15 23 27

 

232

Kuhn, T. S. (1964). The Structure of Scientific Revolutions. Chicago,
University of Chicago Press.

 

Law, J. (1973). "The Development of Specialities in Science: The Case
of X-ray Protein Crystallography," Science Studies, 3 (3): 275-303.

 

McCarthy, F. (l972). "Third Cultural Networks of Philippine Physical,
Life, and Social Scientists." PhD Thesis, East Lansing, Michigan
State University.

McDonoughs, A. M. (l963). Information Economics and Management Systems.
Tokyo, Kogahusha.

 

Merton, R. K. (1942). "Science and Democratic Social Structure," in
Merton, R. K., Social Theory and Social Structure (l949). New York,
The Free Press.

 

Miller, G. R. and M. Steinberg (l975). Between People: A New Analysis
of Interpersonal Communication. Chicago, Science Research Associates.

 

 

 

Mitroff, I. I. (l973). "On Mutual Understanding and the Implementation
Problem: A Philosophical Case Study of the Apollo Moon Scientists,"
presented at the Implementation of OR/MS Models Theory, Research and
Application Conference, University of Pittsburgh.

 

 

Moravcsik, M. J. (l975). Science Development: The Building of Science
in Less Developed Countries. Pasitam, Bloomington, Indiana University
Press.

 

 

Morris, R. C. (l974). "Report to the Staff Seminar on CIAT Training
Programs and Policy.” Cali, CIAT.

Morse, C. et al. (1969). Modernization by Design: Social Science in the
Twentieth Century. Ithaca, New York, Cornell University Press.

 

 

Moseman, A. H. (l970). Building Agricultural Research Systems in the
Developing Nations. New York, Agricultural Development Council.

 

 

 

Mulkay, M. J. (1977). "The Scientific Research Community," in Spiegel-
Rosing, I. and D. de S. Price (eds.) Science,,Technology and Society.
London, Sage Publications, pp. 99-l48.

 

|
l
Mosher, A. T. (I966). Getting_Aggiculture Moving. New York, Praeger. I

. (1972). The Social Process of Innovation: A Study in the ;
Sociology of Science. London, MacMillan.

 

 

Myren, D. T. (1964). "El Papel de la Informacioh en las Decisiones de
Agricultores Bajo Condiciones de Riesgo e Incertidumbre," in Report 1
of First Interamerican Symposium of Research on the Functions of ;
Communication in Agricultural DevelOpment. MERico, The Rockefeller ‘
Foundation, Report 21.