Ajric* Media Review Vol. 1. No. 1. 1986 
African Council on (Communication Education. 
Social Marketing as a Strategy for 
Communicating Agricultural 
Innovations in Rural Development 
by John C.G. Isoba 
Abstract 
This article is a systematic assessment of how nine key concepts in social 
marketing can help us in drawing up effective rural development 
campaigns in Africa. The focus is on how to use these concepts in 
formulating a strategy for communicating agricultural innovations in 
African rural settings. The article recognises the fact that there is no 
single overall reliable method for communicating with farmers 
everywhere but believes that the social marketing principles can be more 
successful with the farmer in most rural situations. 
Re'sume' 
L'article est une etude de la maniere dont 9 concepts cles du marketing social 
peuvent aider a mettre au point une campagne de developpement efficace en 
Afrique. 
L'attention est ici palarisee sur la maniere d'utiliser ces concepts pour 
formuler une strategic de communication des innovations agricoles au sein des 
communautes rurales en Afrique. 
L'article reconnait qu'il n'y a pas une methode unique pour communiquer 
avec les paysans partout, mais neanmoins, affirme que les principes du 
marketing social peuvent reussir avec les agriculteurs dans un grand nombre de 
cas. 
*John C.C. Isoba is a Ugandan Communication Researcher, formerly Senior Research Fellow, 
International Institute of Tropical Agriculture, Ibadan, Nigeria. 
66
While the theme of this paper is communication strategies, let us start 
by looking at the general concept of development since we are concerned 
with using communication for development purposes. However, here 
we are not going to discuss the various theories of development. Rather, we 
shall discuss a few assumptions underlying some development theories to 
see how they relate to strategies of communication. We shall be looking at 
the question, why theories of social change fail. In trying to answer this 
question, Boudon (1983)1 presents four postulates. These are the postulate 
of the coherence of social structure, the nomological postulate, the 
structural postulate and the ontological postulate. 
According to the postulate of the coherence of sociaF structure, the 
various features in a society cohere, i.e. they are bound firmly together so 
that if we identify one feature, we expect to find all the others that are 
bound with it. The implication of this assumption in development terms is 
that introducing change in one of the features will lead to change in the 
others cohering with it. In other words, change in one feature will set off a 
chain reaction in the whole system. This assumption is not necessarily 
correct. 
Let us take a simple example. In Uganda in the late 1960s, it was 
recommended that cotton growers should spray their cotton four times at 
weekly intervals. Using the postulate of the coherence of social structures, 
we would say that when a farmer is convinced about the value of spraying 
his cotton, he will buy the insecticide and spray his cotton four times at 
weekly intervals. However, things did not turn out that way. Not all 
farmers sprayed their cotton four times even though the spray was sold in a 
package to spray four times so that all farmers bought enough chemical to 
spray four times. 
Therefore, the chain reaction did not occur. In this case we should 
speak of interdependence rather than coherence. Spraying four times 
depends on whether the farmer buys the insecticide, which in turn depends 
on whether the farmer is convinced that he should spray his cotton which 
also depends on whether the farmer has enough money to buy the spray. As 
we have seen , buying the insecticide does not necessarily lead to spraying 
four times. In more complex situations, change in one feature may even 
lead to change in the opposite direction. 
The nomological postulate assumes that there are universal regularities 
in all societies and that one of the main tasks of the social scientist is to 
indetify them. This assumption may lead to making generalizations about 
situations based on information obtained from a single observation. Under 
such circumstances, we may make misleading statements. To realize the 
importance of this in terms of development we only have to remember the 
67
number of agricultural development schemes that failed in Africa because 
they depend on technology imported whole sale from temperate countries. 
Under the structural postulate, we assume that there are features in a 
system which are essential in detecting and explaining the main trends in 
development. In other words, we assume that if we identify these essential 
characteristic features, we are in a position to say that this system will 
change through a given characteristic process. In this case, all the features 
within the system not considered essential in bringing about change are 
ignored. Here again, we may run into problems when we are planning 
trends in agricultural development. 
Let us go back to our example of growing cotton. We know that the key 
features in producing more cotton in Uganda are increased acreage, growing 
varieties, proper weeding and, spraying against pests and diseases. 
According to the structural postulate, we would now say that when we 
fulfill these four conditions we shall increase cotton production in the 
country. Put this is not necessarily so. There is one element which we have 
ignored. Yet it can contribute substantially to increased cotton production 
in the country. This element is the limited labour available to the farmer. 
Since the cotton grower in Uganda depends on family labour, there is a limit 
to the amount of cotton he can grow because he may not be able to increase 
his cotton acreage. 
The ontological postulate assumes that there are categories of variables 
in which the main determinants of change are found. In its extreme form, 
this postulate assumes that the same category of variables is always 
responsible for change. In its mild form, it assumes that the nature of the 
variables for change can be different under varying situations, but there are 
always these main determinants of change. However, change is a result of 
individual reactions within the system so that it becomes difficult to isolate 
the variables responsible for change. 
Suppose a research station recommends that the government should 
import and sell to the farmer a particular insecticides for spraying cotton to 
increase production in the country. The government imports the spray and 
gives it to the cooperatives to sell to the farmer. But the farmer does not buy 
the insecticide because it is too expensive lor him. The research station 
comes up with a cheaper, but equally effective insecticide. The farmer 
sprays his cotton. Cotton production in the country goes up. In this case, 
who has been responsible for the increase in production? Is it the 
researcher who made the recommendations? Is it the government that 
imported the spray? Is it the cooperative that sold the insecticide to the 
farmer? Or, is it the farmer who sprayed the cotton? This example points 
out the pitfall in regarding some features of society as being unimportant in 
the development process. 
68
The foregoing show that some projects fail or can fail because we start 
planning using the wrong premises. In the process we neglect or ignore 
certain features of the social system that are actually important for the 
success of our programs. One element we tend to forget to include in our 
plans is the person who has to change. 
For example, even though several approaches were used, rural 
development programs in Ghana had little impact on the rural population in 
the 1960s. One of the main reasons is that local participation in conception, 
planning and implementation was minimal (Brown, 1978).2 At Glacier 
Valley in South Australia, the Department of Agriculture mounted a 
campaign to improve pastures. After five years of continuously running the 
this program, it was found that farmers showed mild interest in improving 
their pastures and had adopted little of the advice they wer given. A major 
reason for this was that farmers were not involved in planning the pasture 
regeneration programme. (Blencows, el al., 1981).3 
These and other campaigns fail because they take the notion of 
distributed benefits too literally so that they never address the potential 
individual benefits which would be gained from participating in a national 
programme (Solomon, 1983)4. In other words, putting excessive emphasis 
on community or national interest at the expense of individual welfare is 
counterproductive (Liu, 1983).5 To try and overcome this problem, 
campaign designers are turning to the principles of marketing in which 
social marketing is defined as: 
The design, implementation and control of programs seeking to increase 
the acceptability of a social idea or practice in a target group(s). It 
utilizes concepts of market segmentation, consumer research, idea 
configuration, communication, facilitation, incentives, and exchange 
theory to maximize target group response (Kotler, 1975 as quoted by 
Solomon, 1983). 
In this paper we discuss nine concepts of social marketing to see how 
they can help us in drawing up effective rural development campaings. The 
nine concepts are: 
1. The marketing philosophy 
2. The "Four Ps" of marketing 
3. Hierarchies of communication effects 
4. Audience segmentation 
5. Understanding all the relevant markets 
6. Information and rapid feedback systems 
7. Interpersonal and mass communication interactions 
8. Understanding the competition 
9. Expectations of success. 
69
The Marketing Philosophy 
The philosophy in social marketing is based on two main principles. The 
first is that the goal ofan organization should be to meet the needs and 
wants of the consumer. The second is that for an exchange to occur, the 
organization and the consumer must interact roughly as equals with each 
having something to offer the other and each one being in a position to 
benefit from the exchange. The two principles taken together say that the 
exchange is mutual. In campaign terms, the mutual exchange is between 
those mounting the campaign and those supposed to adopt the idea or the 
practice. 
However, this mutal exchange does not always exist. Most often the 
campaign is mounted to fulfill the wishes of the organizers while the needs 
of the person who is to change are ignored leading to failed programs. For 
example, in the 1960s, the Ghana government decided to modernize 
agriculture through mechanization by importing a large number of tractors 
and various implements (Nyanteng, 1978).b But farmers could not 
mechanize economically because their farms were small. In the meantime, 
the cutlasses farmers needed were in short supply. 
A similar situation occurred in Samoa where inflexibility on the part of 
the organizations led to misunderstanding village life and consequently to 
failure of rural development programs (Pitt, 1981).7 Here, the developers 
explained the slow growth in export earnings and in village income in terms 
of traditionalism, communalism, authoritarianism and absence of 
individualism. However, research in the villages showed that village life 
was flexible allowing for rewarding individual achievement. Researchers 
also found that traditional institutions could adopt quickly, extended 
families could be turned into cooperatives while chiefs could be used as 
managers. But such was the inflexibility on the part of the developer that 
the good side of village life was not recognized. They continued to 
introduce change according to what they felt was good for the rural people. 
They even proposed a land tax which, in previous years, had led to a general 
revolt. They also proposed growing crops virtually unknown to farmers. 
The "Four Ps" of Marketing 
The "Four Ps" of marketing are the product, the price, the place an& 
promotion. The product is at the centre of the transaction between the 
marketer and the target market. It may consist of ideas, practices or 
services. The price includes more than cash. It covers other costs such as 
time, effort as well as opportunity cost. In planning a campaign, these costs 
70
should be taken into account because they are factors that could lead to a 
project being rejected or accepted. 
This is one of the main reasons why tick control programs in Eastern 
Africa have failed (Ferguson and Poleman, 1973) .8 Here little attention 
was given to the true costs to cattle owners in relation to improvements in 
their standards of living which could be directly attributable to tick control. 
Since cash cost per head appeared low, it was assumed that cattle owners 
would willingly pay fees to achieve the presumed benefits of major increases 
in income. Therefore, insufficient attention was given to minimizing the 
noncash costs to owners. For example, tick control centres were widely 
spaced - up to 8 km radius. This made the costs for tick control higher to 
cattle owners psychologically apart from the labour costs of walking these 
long distances. 
The place refers to the channels through which the idea, product or 
practice reaches the buyer. In campaign terms, the place means the 
channels of communication information to the buyer. However, in 
campaigns for rural areas the campaign designer must be very careful so 
that he selects the media that are available to the audience as Table 1 below 
shows: 
Table 1 
Per cent Farmers Receiving Information by 
Source of Information and Improved Practice 
Seed 
Dressing 
Improved 
Seed 
38.4 (1) 
10.3 (4) 
17.9 (3) 
GhRRM 
Ministry of Agriculture 
Friends 
Relations 
Local Farmers 
Cooperative Union 
Radio or School 
Other 
18.7 (1) 
18.7 (1) 
— 
12.2 (2) 
4.0 (4) 
2.7 (5) 
1.3 (6) 
5.3 (3) 
26.9 (2) 
1.3 (6) 
- 
2.6 
1.3 (3) 
Source: 
Notes: 
(5) 
Compiled from Adoo, N.O., et al, 1978.9 
1. A dash means no data given. 
2. GhRRM means the Ghana Rural Reconstruction 
Momvement. 
3. Figures in brackets give the column rank. 
71 
Fertilizer 
Use 
47.5 (1) 
10.8 (3) 
17.5 (2) 
3.3 
3.3 
6.7 
(6) 
(6) 
(5) 
10.0 (4) 
-
Table 1 gives farmers' sources of information about improved seed, seed 
dressing and fertilizer use in the Manpong Valley, Ghana. As campaign 
designers, there are two main lessons to learn from the data in Table 1. The 
first is that the mass media, in this case, play a very small role in 
transmitting information to farmers. It could be that newspapers, 
television and publications are not available to these farmers. The second is 
that channels of communication are not always equally important in 
transmitting information about different practices. 
For instance, according to rank scores in Table 1, the Ghana Rural 
Reconstruction Movement and the Ministry of Agriculture come first in 
passing information to farmers regarding improved seed. However, as far as 
seed dressing is concerned, the Ghana Rural Reconstruction Movement still 
comes first, but the Ministry of Agriculture drops to fourth place. The last 
of the "Four Ps" is promotion which covers publicity about the campaign as 
well as designing the message and monitoring the progress of the campaign 
to be able to make modifications as required. 
Heirachies of Communication Effects 
Adoption is a process a person goes through from the time he is aware of 
a practice, an idea or service. This process consists of a number of steps 
which vary from three to 12 according to different authors. (McGuire, 
1983).10 For example, Wilson (1955)ll gives seven steps. He says that a 
person starts by having something he wants. Then his attention is drawn 
to it after which he develops interest and his desire becomes aroused. He 
then becomes convinced and takes action or adopts it and finally, he is 
satisfied. 
In the adoption process, the interaction and reaction between the 
communicator and the recipient of the communication is very complex 
indeed. This is illustrated in Figure 1. Here, the columns present the 
communication variables of source, message, channel, receiver and 
destination. For each of these, there is a number of factors which affects the 
interaction between the communicator and the recipient. The rows 
represent the 12 steps in the adoption process. This matrix shows the 
factors to consider at each step in the adoption process. 
Audience Segmentation 
A campaign designed to satisfy everybody in the community is a bad 
one. This is because such a campaign will not have a specific aim in which 
case it will be difficult to isolate the elements to be used in the campaign 
72
such as the channels of communication. Further, a campaign aimed 
generally at a community assumes that the community is homogeneous 
which is not likely to be the case. For example, we may want a particular 
community to grow more maize by opening up more land. The campaign, 
therefore, should be directed at those who have additional land on which to 
grow more maize. 
In this connection, it is worthwhile to remember that in a rural 
community there are several classes of farmers. There are small-scale 
farmers, medium-scale farmers and large scale farmers. There are also 
young farmers, progressive farmers, tenants, absentee farmers, and 
sharecroppers. In planning a campaign, it is essential to decide which of 
these is your target farmer because each one requires a different approach 
when you are introducing an innovation. 
Understanding all the Relevant Markets 
When introducing an innovation, it is not farmers only that may benefit 
from it. There may be people other than farmers who might be interested. 
For example, in a rural community there are such people as farmers, church 
leaders, traders, school teachers and medical assistants. All these are 
potential markets for the campaign designer. 
Suppose that we want to introduce a higher yielding variety of soybean 
to increase production. How can the campaign designer involve all the 
potential markets? In this situation, the farmer will be the centre of the 
campaign since he is the one to grow more soybean. However, churches 
might want to grow this soybean as part of their agricultural projects. The 
traders, too, might be interested in higher production because they have a 
lucrative market for soybean in another part of the country. On the other 
hand, the hospitals might be interested in higher soybean production in the 
area to provide a cheaper source of protein to combat malnutrition. The 
schools might be looking for a suitable crop to use in their agricultural 
syllabus. For them, the soybean might be what they want. 
A campaign properly designed should involve all these people as far as 
possible in a manner that satisfies everyone so that the reaction of each one 
goes to helping the farmer grow more soybean. 
Information and Rapid Feedback Systems 
One of the most important components of a campaign is evaluation. 
This helps in monitoring the progress of the campaign so that we are able to 
modify our strategy as required. In this case, we have to establish 
73
communication channels to pass information around as quickly as possible. 
The information we collect will also _ be very useful at the end of the 
campaign when we want to answer the question, why did the campaign 
succeed? Or, why did the campaign fail? The answer to this question shows 
us the strengths and weakness of our campaign which in turn help us to 
organize better campaigns in the future. 
Interpersonal and Mass Communication Interactions 
One of the functions of the mass media is setting the agenda for people 
in a community. Mass communication researchers have found that 
increasing emphasis on a topic in the mass media leads to increase relevance 
of that topic to the public (Becker, et al., 1975).12 In other words, people 
rely on the mass media to inform them about issues, but not necessarily to 
sway them (Meadow, 1980)1.3 In this way, the media are giving people what 
to think about. This research has implications for the campaign designer 
trying to communicate agricultural innovations to a rural community. For 
instance, there are people regularly in towns whereas they live in rural areas 
around the towns. These people have access to the mass media from which 
they pick ideas. In this situation, the notion of media agenda-setting comes 
in. The urban worker picks up the ideas from the media and when he goes 
home in the rural community he discusses these ideas with his friends. 
Understanding the Competition 
In planning a campaign, it pays to know the competition you will face. 
Some ofthe competition will be for the resources you want to use. For example, 
you will want to know who is competing with you for finances, personnel, 
media use and audience attention. From the point of view of your audience 
you will want to know such things as other activities which will coincide 
with your campaign and what crops or ideas or practices are competing 
with yours. 
If you are planning to promote growing more maize, how would you feel 
when you read a report which says, "The staple food crops required in large 
amounts (in Nigeria) consist of maize, millet, guinea corn, rice, cassava, 
yams, cocoyams, plantain, cowpeas and groundnuts. By 1975, deficiencies 
are expected in maize, the millets, guinea corn, wheat, Irish potatoes, yams, 
groundnuts, vegetables, fruits, various oils and butter, which adequate 
planning, funds, research and effective extension support there is no reason 
why Nigeria should not be able to produce enough quantities of all these 
crops" (Okigbo, 1972)." 
74
Expectation of Success 
Having designed and implemented a campaign in the best manner 
possible, we may sit back and wait for an agricultural revolution. However, 
such optimism may be short lived. As past experience shows, immediate 
and uniform adoption of agricultural innovations is rare (Feder, 1981).ls 
This is because adoption of new technology is affected by the social system, 
attitudes, knowledge, skills and the physical environment. In each of these 
aspects, change tends to be slow (Svennilson, 1968).16 There are, 
however, a few points which can help us avoid making some common 
communication errors (McGuire, 1963).17 
1. The attenuated effects fallacy - in the adoption process, the farmer 
moves from one step to the next until he adopts the innovation, i.e he moves 
progressively from the first step (awareness) through the intermediate steps 
until he arrives at the last step (adoption). However, the chances that the 
farmer will move from one step to the next depend on his having accepted 
the preceding step. But the chances of adopting the preceding step are not 
100%. This happens all along the line of steps in the adoption process so 
that the chances that the farmer will adopt an innovation are small. The 
implication for the campaign designer is that he should guard against 
exaggerated expectations that his campaign will produce large changes in 
behaviour. 
2. The distal measure fallacy - sometimes the failure or success of a 
campaign is judged at the wrong step. For instance, if the aim of a campaign 
is to change behaviour, it would be wrong to say that the campaign was a 
success because the overwhelming majority of the audience is aware of the 
required new behaviour. The question is, they are aware of the new 
behaviour, but are they behaving according to the recommendation? 
3. The neglected mediator fallacy - it is possible to mistakenly put 
emphasis on a particular step in the adoption process at the expense of other 
steps. For instance, in a campaign to increase soybean production, emphasis 
may go on the value of soybean in the diet at the expense of teaching the 
skills of preparing soybean in the way that will enrich the diet. 
4. The compensatory principle - Let us suppose that 
the farmer can get a higher income from cotton than he gets from soybean. 
We may decide to produce more soybean at the expense of cotton for 
nutritional reasons. However, we might have overlooked the fact that with 
a much higher income from cotton the farmer will have enough cash to buy 
the required amount of soybean for a balanced diet and at the same time 
have money left for other purposes. In this case, cotton has compensated 
for its nutritional disadvantages. 
75
5. The golden mean principle - in designing our campaign 
message, we may find that people tend to put more attention to our message 
if we tell them the dangers of the present practice. Where this is so, we 
should guard against throwing a scare into the people saying that since a 
little awareness of the danger produces some attention, then scare will more 
or less automatically get the people to adopt the recommended practice. For 
instance, discussing a case study of a cotton project in Tanzania. Keregere, 
et al..18 say that where farmers "resisted" change, a common response on 
the part of the officials was to put as much pressure as possible on the 
farmer because the authorities assumed that the researchers .were right. 
In this paper we have looked at a strategy for communicating 
agricultural innovations to the rural community. However, we should bear 
in mind that there is no single strategy that is correct. All depends on the 
situation you are working in. All we have in this paper are principles that 
can help us communicate more successfully with the farmer. 
INPUT: Independent 
(Communication) 
Variables 
OUTPUT: 
Dependent 
Variables 
(Response 
Steps 
Mediating 
Persuasion) 
1. E: Exposure to the 
communication 
2. A: Attending to it 
3. L: Liking, becoming 
interested in it 
4. C: Comprehending it 
(learning what) 
5. S: Skill acquisition 
(learning how) 
6. Y: Yielding to it; 
attitude change 
7. M: Memory storage of 
content andy»r 
agreement 
8. I: Information searcfc 
and retrieval 
9. D: Deciding on basis 
of retrieval 
10. B: Behaving in accorc 
with decision 
11. R: Reinforcement ol 
desired acts 
12. P: Post-behavioral 
consolidating 
Fig igure 1. The Communication/Persuasion Model as an Input/Output 
Matrix. 
mber 
nimity 
mographics 
attractivenes 
credi dibility 
type a 
type 
inclu 
organ 
repet 
modality 
irectness 
ontext 
demo 
abili 
perso 
life s 
immediacy/delay 
prevention/cessation 
direct direct immune 
SOURCE MESSAGE 
DESTINATIOIS 
VER 
CHAN- RECEI^ 
NEL 
76
2. 
3. 
5. 
6. 
7. Pitt, David, 1981, 'Planned change and communication in Samoa' in Bruce R. Crouch 
and Shakaria Chamala (eds.) Extension Education and Rural Development, 
8. 
9. 
10. 
11. 
12. 
REFERENCES 
Boudon, Raymond, 1983. 'Why theories of social change fail: some methodological 
thoughts.' Public Opinion Quarterly, Vol. 47: 143 - 160. 
Brown, C.K., 1978, 'Rural development strategies with special reference to Ghana' in 
John R. Schott (ed.) An Experiment in integrated Rural Development: The 
Manpong Valley Social Laboratory in Ghana Institute of Rural Reconstruction 
and Ghana Rural Reconstruction Movement. New York, N.Y. 
Blencowe, J.P., et at., 1981. 'A technique for involving farmers in planning extension 
programmes' in Bruce R. Crouch and Shankaria Chamala (eds.) Extension 
Education and Rural Development, Vol. 2, John1 Viley and Sons, Chichester, 
New York, Brisbane, Toronto. 
Solomon, Douglas, S. 1983. 'A social marketing perspective on campaigns' in Ronald 
R.Priceand William J. Paisley (eds.) Public Communication Campaigns, Sage 
Publications, Beverly Hills, London. 
Liu, Alan P., 1983, 'Mass Campaigns in the People's Republic of China' in Bruce R. 
Crouch and Shankaria Chamala (eds.) Extension Education and Rural Development 
Vol. 2, John Wiley and Sonds, Chichester, New York, Prisbane, 
Toronto. 
Nyateng, V.K., 1978. 'Problems of implementation: the macro-level' in John R. Schott 
(ed.) An Experiment in Integrated Rural Development: The Mampong 
Valley in Ghana, Institute of Rural Reconstruction and Ghana Rural Reconstruction 
Movement, New York, N.Y. 
Vol. 2, John Wiley and Sons, Chichester, New York, Brisbane, Toronto. 
Ferguson, Donald S. and Thomas T. Poleman, 1973, Modernizing African Animal 
Production: The Uganda Tick Control Project Department of Agricultural 
Economics, Cornell University, Ithaca, New York. 
Ado, N.O., eta/., 1978, An evaluation survey in John R. Schott (ed.) An Experiment 
in Integrated Rural Development: The Mampong Valley Social Laboratory 
in Ghana, Institute of Rural Reconstruction and the Ghana Rural Reconstruction 
Movement, New York, N.Y. 
McGuire, William J., 1983. 'Theoretical foundations of campaigns' in Ronald E. 
Price and William J. Paisley (eds.) Public Communication Campaigns, Sage 
Publications, Beverly Hills, London. 
Wilson, Meredith, C. 1985. Extension Teaching Methods, Federal Extension 
Service, US Department of Agriculture, Washington, D.C. 
Becker, Lee R., et al., 'The development of political cognitions' in Stephen H. Chaffee 
(ed.), Political Communication: Issues and Strategies for Research, Sage 
Publications, Beverly Hills, London. 
77
13. Meadow, Robert C , 1980. Politics as Communication, Able* Publishing Corporation. 
Norwood, New Jersey. 
14. Okigbo, Bede N., 1972. 'Problems and prospects of increasing food production in 
Nigeria,' in Symposium on Science and Technology in National Development, 
Lagos. 
15. Fedder, Cershon, 1981, Adoption of Agricultural Innovations in Developing 
Countries: A Survey, World Bank Staff Working Paper No. 444, The World 
Bank, Washington, D.C. 
16. Svennilson, Ingvar, 1968. 'The Strategy of transfer' in Daniel L. Spencer and 
Alexander Woroniak (eds.) The Transfer of Technology in Developing 
Countries, Frederich A. Praeger, Publishers, New York, Washington, London. 
78