- Email: [email protected]
Identifying New Subsystems Caused by Technological Impact
IDENTIFYING NEW SUBSYSTEMS CAUSED BY TECHNOLOGICAL IMPACT
Sydney F. Love Designectics 327 Batavia Place Waterloo, Ontario, Canada
THE ON-GOING WAY OF LIFE People who share a way of life have a set of enduring beliefs and values which give them cultural unity. The relative importance given to values results in a specific value orientation, or characteristic pattern, which guides the people in their way of life. For example, it is generally accepted that the orientations toward man, nature and time of North Americans are different from the orientations of Far Easterners. The study of the enduring beliefs and values is known as Cultural Anthropology. Noted cultural anthropologists like Radcliffe-Brown and F. Kluckhohn have shown how specific cultures are integrated by beliefs and values which are reinforced by ceremony and group behaviour. Each identifiable group behaviour functions in some way to meet personal or group needs (Malino'Nski et al.). According to some (Jarvie et al.) the function of an act or group behaviour may be visibly "manifest" or it may be subtly "latent." For example, a manifest function of public dancing is to provide entertainment. A latent function of public dancing could be the provision of courting opportunities. Nations mayor may not be made up of one people sharing a con~on culture. Usually several identifiable cultures are present in a nation, often having different languages and different racial backgrounds. Thus the beliefs and values within a nation are usually heterogeneous rather than homogeneous. Moreover, within an identifiable culture, there may be subcultures. For example in New York there are the subcultures of Jewish, Puerto Rican, American Negro, and a subculture of poverty.
norm of behaviour. (1) Everett ROgers(2) has shown how innovations can diffuse within a culture by an adoption process. It is generally accepted that change in a culture is gradual, say over years rather than months, and that the enduring beliefs and values continue for years, even for centuries. It follows that any innovative change which is not in accord with an enduring value will have difficulty in gaining acceptance. For example, birth control methods are being adopted very slowly in some cultures where it is in opposition to some enduring value. Adoption of innovations will vary from culture to culture, from fast adoption to outright rejection. The main concern of this paper is with the impact on society of a large technological change. It is generally accepted now that technology and society interact to change each other. Thus, if a more or less stable culture is confronted with a massive technological change, the society itself is under pressure to change. The reaction will vary from yielding to outright opposition. Changes in one part of a way of life have secondary changes in other parts of the culture (Fig. 1). Given time, the culture adapts, but before doing this there may be turbulence which leads to disorientation and apathy. Some tribes of North American Indians are in a state of apathy and physical disintegration, due to conflict of their values and those required to live in a technological society. The unmitigated good intended by a technological aid program is not always seen that way by the recipients. For example, a poor caste of peasants in India did not care to m~ke use.of tbe)water from a huge irrigatl.on proJect. l3
TECHNOLOGICAL IMPACT The normal way for a culture to change is for some individuals to stretch the limit of a specific behaviour just short of the point of where severe sanctions would be applied by others. If the new behaviour functions for the benefit of the majority, it ultimately is adopted and becomes the
THE INCOMPREHENSIBLE WHOLE To understand how beliefs and values affect the way of life of a people is to understand their culture. This is not a simple matter. A study reported on by Laura Thompson(4) concerned only one tribe of
59
Some systems have feedback which tends to stabilize the output. A system which controls itself has cybernetic control which is similar to the organic controlling of blood temperature in mammals. An analysis of any societal system can be improved by finding cybernetic controls which tend to stabilize some output. The systems approach focuses on the relationships between the elements of re~lt'ty, rather than to merely classify them.~5 Thus a subsystem may be formed from interrelated variables which function together to transform inputs to outputs. A system is made up of subsystems which are interrelated in achieving some overall outputs. A system is not a fixed part of reality, but is a construct which is appropriate to the problem being analyzed or solved
North American Indians, the Hopi. This report was the result of six years of study by a multidisciplinary team of 75 persons. A national technological project, such as electrification, may affect several cultures living together, making the problem even more difficult. A complete understanding of the way of life in all sectors of a society is nor~ally beyon? the scope of most project des1gners and 1mplementers. Societel problems everywhere seem to be difficult to correct. The rich natio~s spend millions on research and are.st111 beset with problems of transportat10n, pollution, poverty and drugs. An i~prove ment in one sector is often a loss 1n another sector. The same is true of developing countries. Take public health for example. Better medical facilitie~ and eradication of disease may result 1n a . later population increase of such a magn1tude as to threaten the food supply, increase crime, and put the country in another kind of crisis. To properly assess the effect.of a.directed technological change on a soc1ety 1S to consider the whole society as one system. Yet this is still beyond the scope of any tea~ of scientists that can feasibly be assembled. There are thousands of significant variables in societal processes, and many of the relationships between them are intangible and unquantifiable. A computer cannot be used to find the effect of changes on so large a system, because a mathematical expression for a whole societal system is still beyond the reach of science. Moreover, it is beyond the scope of human minds to assess the simultaneous change of hundreds of variables.
(Fig. 2).
HIERARCHICAL PROBLEM SOLVING Suppose a system be subdivided as much as possible into more or less independent subsystems, each containing two or more variables, and that this be called the first hierarchical level. Usually a subsystem will be less independent of some than of others. The less independent ones may then be combined into new subsystems at the second hier.a rchical level, resulting in larger but fewer sUbsystems. This can be repeated until all subsystems are combined into a whole. This process permits an hierarchical problem-solving procedure (Fig. 3) and is part of the methodology described in this paper. The methodology has utility to the problem solver only to the extent that the subsystems are independent and can be treated as subproblems. At each succeeding hierarchical level, the subproblem solutions need to be revised to account for the residual relationships between the variables of different subsystems which are to be combined. Strictly speaking, only a simultaneous solution of all subproblems is correct and subproblem solutions cannot be added to make a solution to the whole. Socioeconomic-technical systems do not, however, lead to unique solutions. One is fortunate to be able to get a solution to any analysis or synthesis problem at all, and a wide range of solutions is acceptable. Hierarchical problem solving with minor revisions at each level is often the only feasible method of finding a reasonable solution to the whole. The utility of any methodology for subdividing a system for problem solving depends on the nature of the subsystems which result. In the systems approach, the functional relationship between variables is the primary determinant of their combination into subsystems. A useful methodology, therefore, would seek to combine into a subsystem the variables which are strongly related to each other, but which are weakly related to those in other subsystems.
THE SYSTEMS APPROACH TO PROBLEM SOLVING An important aspect of the systems approach is the delimiting of the system by a permeable boundary which is appropriate to the problem at hand. In the case of the impact of a new technology, only those variables which are affected significantly need be considered. Inside the boundary are the variables which can vary in order to achieve the outputs of the system. Outside the arbitrary boundary is the significant environment which has inputs to the system, and receives outputs from it. The problem system itself is thus much ~malle7 th~n the society as a whole, as cons1derat10n 1S only given to those variables having significant relationship to the problem. The system may be modelled in descriptive, tabular, graphical or quantitative form. The system can be further simplified if more or less independent functional subsystems can be found. The relationships between the functional subsystems are important to the overall objectives of the system.
60
POLITICAL
H
~
r.::l H
CULTURAL
~
U H
t5
SOCIPJ.
~
ffi H
::r:
~oc.\E.T"(
I
TECHNOLOGY AND SOCIETY INTERACT Figure 1
0\
I-'
+ + + STRONGEST RELATIONSHIPS
- - - RESIDUAL RELATIONSHIPS PERMEABLE
BOUNDARY
SIGNIFICANT ENVIRONMENT
).
@~;';:UTS) INCIDENTAL
THE SYSTFJ;IS CONSTRUCT Figure 2
~COMBINING
SUBSYSTEMS
HIERARCHICAL PROBLEM SOLVING Figure 3
every other variable in a project. Dominant variables may often be removed by inspection of the data, since the sums of their relationship weights tend to be high. Dominant variables may also be identified by normalizing the relationship weights.
A radically new approach to hi~rarchic~l problem solving was developed ~n the f~eld of architecture by Christopher Alexander. This is embodied in computer programs HIDECS, HIDECS-2 and HIDECS-3. Another approach has been to use FACTOR ANALYSIS computer programs to subdivide multivariate data into smaller meaningful groups. It was the author's dissatisfaction with these techniques that led to the dev~lQpment of a new methodology called SUBSYS.~6}
A MEXICAN VILLAGE A~D ITS VARIABLES To illustrate the SUBSYS methodology, the way of life in the rural Mexican village of San Pedro (not its real name) was studied with respect to the changes that would be caused by electrification of the village. It will be seen that the methodology illustrated herein is generalizable to most socio-economic-technical systems. The first step was to construct a relevant model of the on-going way of life. Initial preparation by the author included a year of study on Cultural Anthropology and related subjects. An opportunity came to be a co-ordinator for a group of university students on a rural Mexican project. The Mexicans in the area were accustomed to students and co-ordinators gathering data and participating in the life style of the co~munity so that the author was able to take on a participant-observer role. This is one of the methods used by Cultural Anthropologists to gather data. San Pedro was a village which might have got electric power within five years, yet was small enough for an in-depth study. It met the requirement of being a place for the perception of a large number of variables which would relate to the addition of electric power, with the variables having tangible and intangible relationships. Data was gathered on site over a three-month period. Some assistance was obtained from Cultural Anthropologists in Mexico. The first task was to develop a classificatory descriptive model of the way of life. It contained 2,800 words and a summary follows:
AN ELECTRICAL SIMULATION OF A SYSTEM STRUCTURE For the purpose of this paper, a socioeconomic-technical system is an aggregate of social, cultural, economic, political, demographic and technical variables which interact together to achieve some technical and non-technical outputs. (In other terminology, a system of interrelated elements has objectives of achieving specific goals.) For example, the addition of a major technical change to a cownunity, such as electric power, water, sewers, roads, irrigation, etc., is a class of problem that can be studied as a socio-economic-technical system. In such a system, there will be a large number of variables and the relationships between them will be both tangible and intangible. A variable may be more strongly related to some than to others. In the SVBSYS method a numerical value is assigned to the estimated strength of relationship between any two variables, while the others remain constant. These relationship values become conductances in an electrical network. In Fig. 4, the vertices represent the variables of the system, and the links represent the relationship values which can be transformed to conductances. Groups A and B represent subsystems. Because of the branches in subsystems, the equivalent conductance between vertices in a subsystem is increased. The data required to assign variables to different subsystems can be acquired from analysis of the electrt·c~l network.which simulates the problem. 6} SUBSYS ~s a computer program that does this and finds the subsystems. It is also possible to construct a physical network of resistors, and acquire the data with a Wheatstone Bridge. In the latter case, the data is used to manually assign variables to subsystems. One of the advantages of the SUBSYS method is that it can be used in developing countries where a large computer is not readily available. SUBSYS has been s~bjected to empirical and pragmatic testing~6} and found to be quite effective in identifying the structure of a problem. In some real-life problems, there are dominant variables which mask the structure of the problem. For example, the variable of cost tends to be related to
San Pedro is a village of about 1,000 persons living in 128 households. It is connected to a town on a highway by a very rough truck road, and to neighbouring smaller villages by rocky mountain trails. The economic subsistence of the village is based on the cultivation of about 1,000 acres of nearby valleys and mountain slopes. The main crops in this semi-arrid zone are maize (corn), beans, peanuts and maguey. The maguey is a large cactus plant that provides some animal fodder and rope fibres, but its main use is to provide a sap which is fermented into a nutritious and mildly intoxicating beverage called "pulque." Taking into account the market value of all the things produced, being mainly agricultural, the average household income in San Pedro is about V.S. $200 per year, with no single household earning more than about $1,000 per year. The village has piped
62
water, a school and a church, but little else of physical community value. Social activity is important to their way of life. The men meet every day just before dusk in family-related drinking groups. The women visit family-related households during the day while the men work in the fields. Religious processions are held frequently. There is one major fiesta during the year and several minor ones on national holidays. Although the population is mainly of mixed Spanish and Indian blood, the language is Spanish. However, the agriculture and housing style have been mainly acquired from the culture of the Otomi Indians. There is an elected head of the village who deals with the outside government. Real power is with the heads of the major family groupings. These groups include all households related through the father. There are no law enforcement officers. Social control is exercised within and between family groupings by social sanctions. Violence between members of different family groups usually leads to vengeance slayings. Many men carry pistols, but slayin9s probably averaged only about one per year at the time of the study. However, there was considerable animosity between some family groupings, and co-operative activity for the good of the community was temporarily paralyzed. San Pedro had many of the characteristics of the genefatized peasant culture outlined by Rogers 2 while at the same time it had uniqueness. The second task was to develop a framework from which significant variables could be developed. This took the form of a classificatory-ordinal model of community functions. A community function is a result of a set of community variables which function together to meet individual and group needs of the people in the community. In Fig. 5 there are two classifications of community functions, and within each the functions are rank ordered. The rank ordering was done according to the degree to which they appeared to give satisfaction to the people who chose to live there (e.g. this puts commerce ahead of waste disposal). The raison d'etre of a community varies. San Pedro's PQPulation could have lived in a dispersed pattern close to their fields. Even water access was not a compelling reason for San Pedro's existence. For some people the reason might be mainly for what is called an ancillary community function. However, observations of the time spent. in social interaction, the priority given to it, and the complaints about how shootouts interfered with it led the author to conclude that for the people of San Pedro, social interaction ranked highest for the majority of the community. Since San Pedro did not actually have electric power, the technical variables and
63
their relationships could not be developed from observations. It was found convenient and advantageous to develop a set of 48 technical variables for a conceptually generalized electric power installation. Each community function was further analyzed for variables which would significantly interact with any of the technical variables. The final set of 98 variables chosen for analysis consisted of 48 technical, 11 economic, 10 demographic, 12 social, 5 political and 12 cultural variables. ESTIMATING RELATIONSHIPS In order to reduce the effect of subjectivity, rules were established for estimating the relationship weights (which in turn become conductances). In the San Pedro problem, the following rules were used to estimate the weights on a four-point scale:
1.
2. 3. 4.
Relationship Between Variables Weight, u ij The state of one is related to the state of the other, such that if the states were measurable and changes were possible, a 10% variation in one would be functionally related to a 10% variation in the other in the problem context, all other variables being held constant 10 As above, but 10% functionally related to 5% 5 As above, but 10% functionally related to 1% 1 No significant functional relationship
°
In estimating weights for the relative functional relationships between variables, the four-point scale shown was found to be about as discriminating as could be used with intangible relationships. If the weights are divided by 10, they will be somewhat analogous to the coefficients of partial correlation. Estimation requires a good mental model of the problem context. In the San Pedro case it was necessary to have in mind the approximate expected states of the technical variables. From a pragmatic viewpoint, the relationship weights do not change much except near threshold and saturation levels. An example of such an estimate follows: Variable #50 is "No. of Households" and Variable #53 is "No. of Places for Team Sports." What is the estimate of u(50, 53) for the context of the present-day San Pedro, having just added typical electric power to the community? In general, the more the households, the more the place for
GROUP TYPE COMMUNITY FUNCTIONS 1. Social Interaction (such as occurred in male drinking groups, by senoras visiting other households, in fiestas, parties and market days . 2. Protection of Person and Property {the family-related residence pattern, surveillance of strangers, kindred vengeance of violence, pooling of fire-arms, etc.). 3. Ideological Continuity (the transfer of values and customs by the teachings in home, school, church and by group sanctions for deviant behaviour) • 4. Reproduction (the begetting and rearing of cnildren; the regulation of sex by social sanction; courtship arrangements).
SUBSYSTEMS OF VARIABLES Figure 4
ANCILLARY TYPE CG~UNITY FUNCTIONS (mainly for household and personal needs) 1. Shelter for people. 2. Access to food suP?ly. 3. Access to water supply. 4. Cooking facilities. 5. Space for livestock and poultry . 6. Space for gardens. 7. Commerce within and beyond. 8 . Fuel su?ply . 9. Corn grinding facility. 10. Storage space. 11 . Transport to other co~~unities. 12. Waste dis posal. 13. Dispute settling by elected "Juez." 14. Medical aid. A CLASSIFICATORY-ORDINAL MODEL OF THE CO~~UNITY OF SA~ P£DRO FIGURE 5
.........
,
n
......
~ tl
!§
~~~ ~
-:l,,~t:~r;; ;<~~ig~~86
tl!:l
",~IIll o... ~
'"
0;
0;
~ !
!§O
"...
;;;
-
h
..
. I It ~I I I~ i n~ ii6~ ~; I ~ ~ Ii d I! ~ fi ti U I , i ! i ~I 8 " ".. .. . U. .. .. co
~
'" "1Ii:i ~ ~o .e ~I;!i~
5
1:1
~
~
:i
I-
t:
a~~i
...:
J-
"
,
"'-;;:;
~-
;;;
~
,,~
"
a!o
~~ ~
J.
"
'0
I:l
~
~
0;
~
~
g ...
°
N
tl
. .~, ..'" .. . 6
o ,J,
•
J
I . R. • ELECTRIC RISOURCI R• • RISOURCI (.) • lI\JMBZR
or
VARIABLIS
WIDTH OF BWCK I S PROPORTIONAL TO NO. OF VARIABLES IN SUBSYSn>\
THE SUBSYSTEMS
a=
SAN PEDRO WITH ELECTRIC PCM'ER FIGURE 6
64
sports. There will be a partial correlation between the two variables, although the underlying functional relationship is vague, complex and involves many other variables. Observation of nearby communities indicates some relationship between the two variables in question. The community of Barranca Arriba has about 20 households and zero sports fields. El Carrizel has about 40 households and one sports field. San Pedro has 128 households and two sports fields. There is a strong correlation and it is evidently more than coincidental covariance. If we now add electric power to San Pedro, does the relationship change? The states of these demographic variables may change, but the underlying relationship is not seriously changed. The author gave this relationship a medium weight, whence
u(oO, 03) • 5.
The steps in this methodology which involve the study of the way of life, the development of variables, and estimation of relationship weights are not easily done. The author was assisted by linguists and cultural anthropologists. A multidisciplinary study is required, and where resources permit it, a team study should be done.
suasys
COMPUTED RESULTS As shown in an earlier paper,(6) it is possible for dominant variables to mask the basic form of the problem. The first computer runs with SUBSYS on the 98-variable system indicated only one large group and a few trivial groups. Subsequently, the 8 variables with the largest sum of data rows (largest sum of relationship weights) were set aside and the 91 remaining variables were processed by SUBSYS. The result indicated 3 large groups and 11 smaller ones, all being more or less independent enough to be subsystems. The 8 variables were then assigned to these groups by an iterative process of calculating the relationships to each group. The author originally had difficulty in making sense of the subsystems indicated by SUBSYS. They did not fit any of the patterns with which he was familiar, so an experienced cultural anthropologist was consulted. This led to the development of a resource model which fitted the subsystems quite well. Only a small percentage of the variables in each subsystem did not fit its general category. The subsystems became recognizable as entities, thus simplifying the problem. From the residual relationships between the 14 subsystems, a 3-level hierarchy of subsystems was constructed. At level 11, combinations reduced it to 9 subsystems, 5 of which remained as separate but trivial subsystems which are combined into the whole at level Ill. These are shown in Fig. 6. A list of the variables and their subsystems are included in the Appendix. More
details on the San Pedro variables and on the quantitative aspect~ Qf the system are given in another paper. l7 ) USING THE SUBSYS METHODOLOGY The SUBSYS methodology is a tool for the analysis or synthesis of socio-economictechnical systems. It can help to identify the problem by indicating the underlying structure as subsystems. The problem is not solved, but it is simplified by at least an order of magnitude. Hence, the impact of new technology on a developing country is one class of problem to which SUBSYS may be applied. All communities of people, like San Pedro, have some uniqueness about their way of life. It is the recognition and understanding of the uniqueness of each situation that can lead to a better systems approach to developing countries. APPENDIX A Complete Listing of the San Pedro Subsystems of 98 Variables for Hierarchical Level I l(a) Electrical Resource Variables of the Input (6 variables) Voltage of power source. KW peak taken from power source. % regulation of power source. No. of phases of power source. Conductor size of primary lines. Primary to secondary voltage ratio. l(b) Electric Resource Variables Within Community and Value of Community (22 variables) KVA of distribution transformers. Size of conductors for distribution. Secondary to use voltage ratio. % regulation of distribution. Use voltage to household and co~nercial. No. of phases to household and commercial users. % regulation at use point. Minimum size of household service. Peak power available to a household or commercial user. Minimum conductor size for users. KW used for household refrigeration. KW used for household heating. KW used for street lighting. D KW used by commercial users. KW used for religious activities. Track of street lights. D Household density to the community. Average co~~unity expense per year for grinding corn. Community debt payment for next five years. Index of real estate values. Value placed on modern replacing traditional. No. of masses held per year.
No. of team sports games per year. Value placed on co-operative action to improve the community. 0 o Community News Diffusion (2 variables) No. of social contacts per day of household senoras with other households. Diffusion rate of community news. 6 Electric Resource Value to Education (2 variables) % population learning trade or profession other than farming. Hours per week of electrical aids at school. 7 Electrical Resource Cost to Educatio~ (2 variables) % of San Pedro resources going into schools. % of population who are literate. 8 Electric Resource Cost to Agricultural Resource (2 varfa1>Ies) Length of primary lines. Area of land used for primary lines. 9 Electric Resource Variables Affectin Agricultura Resource 0 var~a Minimum height of primary lines. Spatial tract of primary lines. Capital cost of power source. Spatial location of substation switches. % of community projects not getting state aid because of rival town X. 10 Electric Resource Secondar Distribution Varia 4 var~a es Minimum height of conductors for distribution. Height of supports for secondary conductors. Average distance between secondary conductor supports. Spatial track of secondary lines. 11 Alternate Electric Power as a Resource ( 3 variables) KW peak of alternative power. Standby minimum illumination of streets. Value placed on continuous power with few outages. l2 Gardens as a Resource (3 variables) Annual income per household from gardens. Value placed on community aesthetics. Value placed on flowers in gardens. 13 Household Resource of Cooking (2 var.) KW used for household cooking. Price of firewood. 14 Electric Resource Variable of Frequency (2 variables) Frequency of power source. Use frequency.
l(c) Electric Resource Value to Household (9 variables) Phases of secondary lines. KW used for household water pumping. KW used for household small appliances. Length of streets. No. of households. 0 Agricultural land per household. Economic value per household of home-made artifacts. Diffusion rate of non-local news. % of households with radios. o
2
Electric Resource Cost to Household (10 variables) Cost per KWH to community. KW used for household lighting. KW used for household yard lighting. Minimum illumination for eating. Minimum illumination for cooking. Minimum illumination in bedrooms. Minimum illumination for reading. Average area of kitchens. Cost per household of lighting by candles, kerosene and resin wood. Projected cost per household per year for electric power. 3 Electric Resource as Modifier to Homeostasis of Social and Anti-social Activities (20 variables) Minimum illumination of streets. No. of places suitable for a public dance. Centroid of residential grouping of largest extended family. Locations of male drinking groups. Average no. of hours after sunset for male drinking groups. Size of after-work male drinking groups. No. of public celebrations per year. No. of private celebrations per year. % of people visiting others after dark. 0 No. of extended families with more than la~ of households. % of population who co-operate on co~nunity projects. No. of killings, wounding and rODberies per year of San Pedro people. % of killing, wounding and robberies punished by legal means. % of killings, wounding and robberies punished by traditional means. No. of dogs per household. Value placed on settling interfamily disputes by violence. Value placed on firing pistols into the air. % of criminals in area who are punished by initiative of outside government. No. of public meetings using night lighting. No. of night meetings of community government. 4 Electric Resource Value to Communi~ Co-operative Action (4 variables) KW used for outdoor sports. No. of places for team sports.
Note: A "0" after a variable number indicates one of the dominant variables which was assigned to the group to which it had the most relationship. However, the dominant variables are strongly related to other groups as well. 66
REFERENCES (1) Barnett, H. G., INNOVATION: THE BASIS OF CULTURAL CHANGE, McGraw-Hill Book Co. Inc., New York, 1953. (2) Rogers, Everett, MODERNIZATION AMONG PEASANTS, Holt, Rinehart & Winston, New York, 1969. (3) Nair, Kusum, BLOSSOMS IN THE DUST, Gerald Duckworth, London, 1961. (4) Thompson, Laura, CULTURE IN CRISIS: A STUDY OF THE HOPI INDIA~S, Harper Bros., New York, 1950. (5) Thompson, Laura, TOWARD A SCIENCE OF MA~KIND, McGraw-Hill, New York, 1961. (6) Love, S. F., "A New Methodolo9Y for the Hierarchical Grouping of Related Elements of a Problem,u IEEE TRANS. ON SYSTEMS, MA~ A~D CYBERNETICS, SMC-2, No. 1, January, 1972, pp. 23-29. (7) Love, S. F., "Socio-Economic-Technical Subsystems for Addition of Electric Power to a Mexican Village," September, 1971, unpublished. A summary is published in the PROCEEDINGS OF THE 1972 INTE&~ATIONAL CONFERENCE ON CYBERNETICS AND SOCIETY, October 9-12, Washington, D.C.
Sydney F. Love Designectics 327 Batavia Place Waterloo, Ontario, Canada
THE ON-GOING WAY OF LIFE People who share a way of life have a set of enduring beliefs and values which give them cultural unity. The relative importance given to values results in a specific value orientation, or characteristic pattern, which guides the people in their way of life. For example, it is generally accepted that the orientations toward man, nature and time of North Americans are different from the orientations of Far Easterners. The study of the enduring beliefs and values is known as Cultural Anthropology. Noted cultural anthropologists like Radcliffe-Brown and F. Kluckhohn have shown how specific cultures are integrated by beliefs and values which are reinforced by ceremony and group behaviour. Each identifiable group behaviour functions in some way to meet personal or group needs (Malino'Nski et al.). According to some (Jarvie et al.) the function of an act or group behaviour may be visibly "manifest" or it may be subtly "latent." For example, a manifest function of public dancing is to provide entertainment. A latent function of public dancing could be the provision of courting opportunities. Nations mayor may not be made up of one people sharing a con~on culture. Usually several identifiable cultures are present in a nation, often having different languages and different racial backgrounds. Thus the beliefs and values within a nation are usually heterogeneous rather than homogeneous. Moreover, within an identifiable culture, there may be subcultures. For example in New York there are the subcultures of Jewish, Puerto Rican, American Negro, and a subculture of poverty.
norm of behaviour. (1) Everett ROgers(2) has shown how innovations can diffuse within a culture by an adoption process. It is generally accepted that change in a culture is gradual, say over years rather than months, and that the enduring beliefs and values continue for years, even for centuries. It follows that any innovative change which is not in accord with an enduring value will have difficulty in gaining acceptance. For example, birth control methods are being adopted very slowly in some cultures where it is in opposition to some enduring value. Adoption of innovations will vary from culture to culture, from fast adoption to outright rejection. The main concern of this paper is with the impact on society of a large technological change. It is generally accepted now that technology and society interact to change each other. Thus, if a more or less stable culture is confronted with a massive technological change, the society itself is under pressure to change. The reaction will vary from yielding to outright opposition. Changes in one part of a way of life have secondary changes in other parts of the culture (Fig. 1). Given time, the culture adapts, but before doing this there may be turbulence which leads to disorientation and apathy. Some tribes of North American Indians are in a state of apathy and physical disintegration, due to conflict of their values and those required to live in a technological society. The unmitigated good intended by a technological aid program is not always seen that way by the recipients. For example, a poor caste of peasants in India did not care to m~ke use.of tbe)water from a huge irrigatl.on proJect. l3
TECHNOLOGICAL IMPACT The normal way for a culture to change is for some individuals to stretch the limit of a specific behaviour just short of the point of where severe sanctions would be applied by others. If the new behaviour functions for the benefit of the majority, it ultimately is adopted and becomes the
THE INCOMPREHENSIBLE WHOLE To understand how beliefs and values affect the way of life of a people is to understand their culture. This is not a simple matter. A study reported on by Laura Thompson(4) concerned only one tribe of
59
Some systems have feedback which tends to stabilize the output. A system which controls itself has cybernetic control which is similar to the organic controlling of blood temperature in mammals. An analysis of any societal system can be improved by finding cybernetic controls which tend to stabilize some output. The systems approach focuses on the relationships between the elements of re~lt'ty, rather than to merely classify them.~5 Thus a subsystem may be formed from interrelated variables which function together to transform inputs to outputs. A system is made up of subsystems which are interrelated in achieving some overall outputs. A system is not a fixed part of reality, but is a construct which is appropriate to the problem being analyzed or solved
North American Indians, the Hopi. This report was the result of six years of study by a multidisciplinary team of 75 persons. A national technological project, such as electrification, may affect several cultures living together, making the problem even more difficult. A complete understanding of the way of life in all sectors of a society is nor~ally beyon? the scope of most project des1gners and 1mplementers. Societel problems everywhere seem to be difficult to correct. The rich natio~s spend millions on research and are.st111 beset with problems of transportat10n, pollution, poverty and drugs. An i~prove ment in one sector is often a loss 1n another sector. The same is true of developing countries. Take public health for example. Better medical facilitie~ and eradication of disease may result 1n a . later population increase of such a magn1tude as to threaten the food supply, increase crime, and put the country in another kind of crisis. To properly assess the effect.of a.directed technological change on a soc1ety 1S to consider the whole society as one system. Yet this is still beyond the scope of any tea~ of scientists that can feasibly be assembled. There are thousands of significant variables in societal processes, and many of the relationships between them are intangible and unquantifiable. A computer cannot be used to find the effect of changes on so large a system, because a mathematical expression for a whole societal system is still beyond the reach of science. Moreover, it is beyond the scope of human minds to assess the simultaneous change of hundreds of variables.
(Fig. 2).
HIERARCHICAL PROBLEM SOLVING Suppose a system be subdivided as much as possible into more or less independent subsystems, each containing two or more variables, and that this be called the first hierarchical level. Usually a subsystem will be less independent of some than of others. The less independent ones may then be combined into new subsystems at the second hier.a rchical level, resulting in larger but fewer sUbsystems. This can be repeated until all subsystems are combined into a whole. This process permits an hierarchical problem-solving procedure (Fig. 3) and is part of the methodology described in this paper. The methodology has utility to the problem solver only to the extent that the subsystems are independent and can be treated as subproblems. At each succeeding hierarchical level, the subproblem solutions need to be revised to account for the residual relationships between the variables of different subsystems which are to be combined. Strictly speaking, only a simultaneous solution of all subproblems is correct and subproblem solutions cannot be added to make a solution to the whole. Socioeconomic-technical systems do not, however, lead to unique solutions. One is fortunate to be able to get a solution to any analysis or synthesis problem at all, and a wide range of solutions is acceptable. Hierarchical problem solving with minor revisions at each level is often the only feasible method of finding a reasonable solution to the whole. The utility of any methodology for subdividing a system for problem solving depends on the nature of the subsystems which result. In the systems approach, the functional relationship between variables is the primary determinant of their combination into subsystems. A useful methodology, therefore, would seek to combine into a subsystem the variables which are strongly related to each other, but which are weakly related to those in other subsystems.
THE SYSTEMS APPROACH TO PROBLEM SOLVING An important aspect of the systems approach is the delimiting of the system by a permeable boundary which is appropriate to the problem at hand. In the case of the impact of a new technology, only those variables which are affected significantly need be considered. Inside the boundary are the variables which can vary in order to achieve the outputs of the system. Outside the arbitrary boundary is the significant environment which has inputs to the system, and receives outputs from it. The problem system itself is thus much ~malle7 th~n the society as a whole, as cons1derat10n 1S only given to those variables having significant relationship to the problem. The system may be modelled in descriptive, tabular, graphical or quantitative form. The system can be further simplified if more or less independent functional subsystems can be found. The relationships between the functional subsystems are important to the overall objectives of the system.
60
POLITICAL
H
~
r.::l H
CULTURAL
~
U H
t5
SOCIPJ.
~
ffi H
::r:
~oc.\E.T"(
I
TECHNOLOGY AND SOCIETY INTERACT Figure 1
0\
I-'
+ + + STRONGEST RELATIONSHIPS
- - - RESIDUAL RELATIONSHIPS PERMEABLE
BOUNDARY
SIGNIFICANT ENVIRONMENT
).
@~;';:UTS) INCIDENTAL
THE SYSTFJ;IS CONSTRUCT Figure 2
~COMBINING
SUBSYSTEMS
HIERARCHICAL PROBLEM SOLVING Figure 3
every other variable in a project. Dominant variables may often be removed by inspection of the data, since the sums of their relationship weights tend to be high. Dominant variables may also be identified by normalizing the relationship weights.
A radically new approach to hi~rarchic~l problem solving was developed ~n the f~eld of architecture by Christopher Alexander. This is embodied in computer programs HIDECS, HIDECS-2 and HIDECS-3. Another approach has been to use FACTOR ANALYSIS computer programs to subdivide multivariate data into smaller meaningful groups. It was the author's dissatisfaction with these techniques that led to the dev~lQpment of a new methodology called SUBSYS.~6}
A MEXICAN VILLAGE A~D ITS VARIABLES To illustrate the SUBSYS methodology, the way of life in the rural Mexican village of San Pedro (not its real name) was studied with respect to the changes that would be caused by electrification of the village. It will be seen that the methodology illustrated herein is generalizable to most socio-economic-technical systems. The first step was to construct a relevant model of the on-going way of life. Initial preparation by the author included a year of study on Cultural Anthropology and related subjects. An opportunity came to be a co-ordinator for a group of university students on a rural Mexican project. The Mexicans in the area were accustomed to students and co-ordinators gathering data and participating in the life style of the co~munity so that the author was able to take on a participant-observer role. This is one of the methods used by Cultural Anthropologists to gather data. San Pedro was a village which might have got electric power within five years, yet was small enough for an in-depth study. It met the requirement of being a place for the perception of a large number of variables which would relate to the addition of electric power, with the variables having tangible and intangible relationships. Data was gathered on site over a three-month period. Some assistance was obtained from Cultural Anthropologists in Mexico. The first task was to develop a classificatory descriptive model of the way of life. It contained 2,800 words and a summary follows:
AN ELECTRICAL SIMULATION OF A SYSTEM STRUCTURE For the purpose of this paper, a socioeconomic-technical system is an aggregate of social, cultural, economic, political, demographic and technical variables which interact together to achieve some technical and non-technical outputs. (In other terminology, a system of interrelated elements has objectives of achieving specific goals.) For example, the addition of a major technical change to a cownunity, such as electric power, water, sewers, roads, irrigation, etc., is a class of problem that can be studied as a socio-economic-technical system. In such a system, there will be a large number of variables and the relationships between them will be both tangible and intangible. A variable may be more strongly related to some than to others. In the SVBSYS method a numerical value is assigned to the estimated strength of relationship between any two variables, while the others remain constant. These relationship values become conductances in an electrical network. In Fig. 4, the vertices represent the variables of the system, and the links represent the relationship values which can be transformed to conductances. Groups A and B represent subsystems. Because of the branches in subsystems, the equivalent conductance between vertices in a subsystem is increased. The data required to assign variables to different subsystems can be acquired from analysis of the electrt·c~l network.which simulates the problem. 6} SUBSYS ~s a computer program that does this and finds the subsystems. It is also possible to construct a physical network of resistors, and acquire the data with a Wheatstone Bridge. In the latter case, the data is used to manually assign variables to subsystems. One of the advantages of the SUBSYS method is that it can be used in developing countries where a large computer is not readily available. SUBSYS has been s~bjected to empirical and pragmatic testing~6} and found to be quite effective in identifying the structure of a problem. In some real-life problems, there are dominant variables which mask the structure of the problem. For example, the variable of cost tends to be related to
San Pedro is a village of about 1,000 persons living in 128 households. It is connected to a town on a highway by a very rough truck road, and to neighbouring smaller villages by rocky mountain trails. The economic subsistence of the village is based on the cultivation of about 1,000 acres of nearby valleys and mountain slopes. The main crops in this semi-arrid zone are maize (corn), beans, peanuts and maguey. The maguey is a large cactus plant that provides some animal fodder and rope fibres, but its main use is to provide a sap which is fermented into a nutritious and mildly intoxicating beverage called "pulque." Taking into account the market value of all the things produced, being mainly agricultural, the average household income in San Pedro is about V.S. $200 per year, with no single household earning more than about $1,000 per year. The village has piped
62
water, a school and a church, but little else of physical community value. Social activity is important to their way of life. The men meet every day just before dusk in family-related drinking groups. The women visit family-related households during the day while the men work in the fields. Religious processions are held frequently. There is one major fiesta during the year and several minor ones on national holidays. Although the population is mainly of mixed Spanish and Indian blood, the language is Spanish. However, the agriculture and housing style have been mainly acquired from the culture of the Otomi Indians. There is an elected head of the village who deals with the outside government. Real power is with the heads of the major family groupings. These groups include all households related through the father. There are no law enforcement officers. Social control is exercised within and between family groupings by social sanctions. Violence between members of different family groups usually leads to vengeance slayings. Many men carry pistols, but slayin9s probably averaged only about one per year at the time of the study. However, there was considerable animosity between some family groupings, and co-operative activity for the good of the community was temporarily paralyzed. San Pedro had many of the characteristics of the genefatized peasant culture outlined by Rogers 2 while at the same time it had uniqueness. The second task was to develop a framework from which significant variables could be developed. This took the form of a classificatory-ordinal model of community functions. A community function is a result of a set of community variables which function together to meet individual and group needs of the people in the community. In Fig. 5 there are two classifications of community functions, and within each the functions are rank ordered. The rank ordering was done according to the degree to which they appeared to give satisfaction to the people who chose to live there (e.g. this puts commerce ahead of waste disposal). The raison d'etre of a community varies. San Pedro's PQPulation could have lived in a dispersed pattern close to their fields. Even water access was not a compelling reason for San Pedro's existence. For some people the reason might be mainly for what is called an ancillary community function. However, observations of the time spent. in social interaction, the priority given to it, and the complaints about how shootouts interfered with it led the author to conclude that for the people of San Pedro, social interaction ranked highest for the majority of the community. Since San Pedro did not actually have electric power, the technical variables and
63
their relationships could not be developed from observations. It was found convenient and advantageous to develop a set of 48 technical variables for a conceptually generalized electric power installation. Each community function was further analyzed for variables which would significantly interact with any of the technical variables. The final set of 98 variables chosen for analysis consisted of 48 technical, 11 economic, 10 demographic, 12 social, 5 political and 12 cultural variables. ESTIMATING RELATIONSHIPS In order to reduce the effect of subjectivity, rules were established for estimating the relationship weights (which in turn become conductances). In the San Pedro problem, the following rules were used to estimate the weights on a four-point scale:
1.
2. 3. 4.
Relationship Between Variables Weight, u ij The state of one is related to the state of the other, such that if the states were measurable and changes were possible, a 10% variation in one would be functionally related to a 10% variation in the other in the problem context, all other variables being held constant 10 As above, but 10% functionally related to 5% 5 As above, but 10% functionally related to 1% 1 No significant functional relationship
°
In estimating weights for the relative functional relationships between variables, the four-point scale shown was found to be about as discriminating as could be used with intangible relationships. If the weights are divided by 10, they will be somewhat analogous to the coefficients of partial correlation. Estimation requires a good mental model of the problem context. In the San Pedro case it was necessary to have in mind the approximate expected states of the technical variables. From a pragmatic viewpoint, the relationship weights do not change much except near threshold and saturation levels. An example of such an estimate follows: Variable #50 is "No. of Households" and Variable #53 is "No. of Places for Team Sports." What is the estimate of u(50, 53) for the context of the present-day San Pedro, having just added typical electric power to the community? In general, the more the households, the more the place for
GROUP TYPE COMMUNITY FUNCTIONS 1. Social Interaction (such as occurred in male drinking groups, by senoras visiting other households, in fiestas, parties and market days . 2. Protection of Person and Property {the family-related residence pattern, surveillance of strangers, kindred vengeance of violence, pooling of fire-arms, etc.). 3. Ideological Continuity (the transfer of values and customs by the teachings in home, school, church and by group sanctions for deviant behaviour) • 4. Reproduction (the begetting and rearing of cnildren; the regulation of sex by social sanction; courtship arrangements).
SUBSYSTEMS OF VARIABLES Figure 4
ANCILLARY TYPE CG~UNITY FUNCTIONS (mainly for household and personal needs) 1. Shelter for people. 2. Access to food suP?ly. 3. Access to water supply. 4. Cooking facilities. 5. Space for livestock and poultry . 6. Space for gardens. 7. Commerce within and beyond. 8 . Fuel su?ply . 9. Corn grinding facility. 10. Storage space. 11 . Transport to other co~~unities. 12. Waste dis posal. 13. Dispute settling by elected "Juez." 14. Medical aid. A CLASSIFICATORY-ORDINAL MODEL OF THE CO~~UNITY OF SA~ P£DRO FIGURE 5
.........
,
n
......
~ tl
!§
~~~ ~
-:l,,~t:~r;; ;<~~ig~~86
tl!:l
",~IIll o... ~
'"
0;
0;
~ !
!§O
"...
;;;
-
h
..
. I It ~I I I~ i n~ ii6~ ~; I ~ ~ Ii d I! ~ fi ti U I , i ! i ~I 8 " ".. .. . U. .. .. co
~
'" "1Ii:i ~ ~o .e ~I;!i~
5
1:1
~
~
:i
I-
t:
a~~i
...:
J-
"
,
"'-;;:;
~-
;;;
~
,,~
"
a!o
~~ ~
J.
"
'0
I:l
~
~
0;
~
~
g ...
°
N
tl
. .~, ..'" .. . 6
o ,J,
•
J
I . R. • ELECTRIC RISOURCI R• • RISOURCI (.) • lI\JMBZR
or
VARIABLIS
WIDTH OF BWCK I S PROPORTIONAL TO NO. OF VARIABLES IN SUBSYSn>\
THE SUBSYSTEMS
a=
SAN PEDRO WITH ELECTRIC PCM'ER FIGURE 6
64
sports. There will be a partial correlation between the two variables, although the underlying functional relationship is vague, complex and involves many other variables. Observation of nearby communities indicates some relationship between the two variables in question. The community of Barranca Arriba has about 20 households and zero sports fields. El Carrizel has about 40 households and one sports field. San Pedro has 128 households and two sports fields. There is a strong correlation and it is evidently more than coincidental covariance. If we now add electric power to San Pedro, does the relationship change? The states of these demographic variables may change, but the underlying relationship is not seriously changed. The author gave this relationship a medium weight, whence
u(oO, 03) • 5.
The steps in this methodology which involve the study of the way of life, the development of variables, and estimation of relationship weights are not easily done. The author was assisted by linguists and cultural anthropologists. A multidisciplinary study is required, and where resources permit it, a team study should be done.
suasys
COMPUTED RESULTS As shown in an earlier paper,(6) it is possible for dominant variables to mask the basic form of the problem. The first computer runs with SUBSYS on the 98-variable system indicated only one large group and a few trivial groups. Subsequently, the 8 variables with the largest sum of data rows (largest sum of relationship weights) were set aside and the 91 remaining variables were processed by SUBSYS. The result indicated 3 large groups and 11 smaller ones, all being more or less independent enough to be subsystems. The 8 variables were then assigned to these groups by an iterative process of calculating the relationships to each group. The author originally had difficulty in making sense of the subsystems indicated by SUBSYS. They did not fit any of the patterns with which he was familiar, so an experienced cultural anthropologist was consulted. This led to the development of a resource model which fitted the subsystems quite well. Only a small percentage of the variables in each subsystem did not fit its general category. The subsystems became recognizable as entities, thus simplifying the problem. From the residual relationships between the 14 subsystems, a 3-level hierarchy of subsystems was constructed. At level 11, combinations reduced it to 9 subsystems, 5 of which remained as separate but trivial subsystems which are combined into the whole at level Ill. These are shown in Fig. 6. A list of the variables and their subsystems are included in the Appendix. More
details on the San Pedro variables and on the quantitative aspect~ Qf the system are given in another paper. l7 ) USING THE SUBSYS METHODOLOGY The SUBSYS methodology is a tool for the analysis or synthesis of socio-economictechnical systems. It can help to identify the problem by indicating the underlying structure as subsystems. The problem is not solved, but it is simplified by at least an order of magnitude. Hence, the impact of new technology on a developing country is one class of problem to which SUBSYS may be applied. All communities of people, like San Pedro, have some uniqueness about their way of life. It is the recognition and understanding of the uniqueness of each situation that can lead to a better systems approach to developing countries. APPENDIX A Complete Listing of the San Pedro Subsystems of 98 Variables for Hierarchical Level I l(a) Electrical Resource Variables of the Input (6 variables) Voltage of power source. KW peak taken from power source. % regulation of power source. No. of phases of power source. Conductor size of primary lines. Primary to secondary voltage ratio. l(b) Electric Resource Variables Within Community and Value of Community (22 variables) KVA of distribution transformers. Size of conductors for distribution. Secondary to use voltage ratio. % regulation of distribution. Use voltage to household and co~nercial. No. of phases to household and commercial users. % regulation at use point. Minimum size of household service. Peak power available to a household or commercial user. Minimum conductor size for users. KW used for household refrigeration. KW used for household heating. KW used for street lighting. D KW used by commercial users. KW used for religious activities. Track of street lights. D Household density to the community. Average co~~unity expense per year for grinding corn. Community debt payment for next five years. Index of real estate values. Value placed on modern replacing traditional. No. of masses held per year.
No. of team sports games per year. Value placed on co-operative action to improve the community. 0 o Community News Diffusion (2 variables) No. of social contacts per day of household senoras with other households. Diffusion rate of community news. 6 Electric Resource Value to Education (2 variables) % population learning trade or profession other than farming. Hours per week of electrical aids at school. 7 Electrical Resource Cost to Educatio~ (2 variables) % of San Pedro resources going into schools. % of population who are literate. 8 Electric Resource Cost to Agricultural Resource (2 varfa1>Ies) Length of primary lines. Area of land used for primary lines. 9 Electric Resource Variables Affectin Agricultura Resource 0 var~a Minimum height of primary lines. Spatial tract of primary lines. Capital cost of power source. Spatial location of substation switches. % of community projects not getting state aid because of rival town X. 10 Electric Resource Secondar Distribution Varia 4 var~a es Minimum height of conductors for distribution. Height of supports for secondary conductors. Average distance between secondary conductor supports. Spatial track of secondary lines. 11 Alternate Electric Power as a Resource ( 3 variables) KW peak of alternative power. Standby minimum illumination of streets. Value placed on continuous power with few outages. l2 Gardens as a Resource (3 variables) Annual income per household from gardens. Value placed on community aesthetics. Value placed on flowers in gardens. 13 Household Resource of Cooking (2 var.) KW used for household cooking. Price of firewood. 14 Electric Resource Variable of Frequency (2 variables) Frequency of power source. Use frequency.
l(c) Electric Resource Value to Household (9 variables) Phases of secondary lines. KW used for household water pumping. KW used for household small appliances. Length of streets. No. of households. 0 Agricultural land per household. Economic value per household of home-made artifacts. Diffusion rate of non-local news. % of households with radios. o
2
Electric Resource Cost to Household (10 variables) Cost per KWH to community. KW used for household lighting. KW used for household yard lighting. Minimum illumination for eating. Minimum illumination for cooking. Minimum illumination in bedrooms. Minimum illumination for reading. Average area of kitchens. Cost per household of lighting by candles, kerosene and resin wood. Projected cost per household per year for electric power. 3 Electric Resource as Modifier to Homeostasis of Social and Anti-social Activities (20 variables) Minimum illumination of streets. No. of places suitable for a public dance. Centroid of residential grouping of largest extended family. Locations of male drinking groups. Average no. of hours after sunset for male drinking groups. Size of after-work male drinking groups. No. of public celebrations per year. No. of private celebrations per year. % of people visiting others after dark. 0 No. of extended families with more than la~ of households. % of population who co-operate on co~nunity projects. No. of killings, wounding and rODberies per year of San Pedro people. % of killing, wounding and robberies punished by legal means. % of killings, wounding and robberies punished by traditional means. No. of dogs per household. Value placed on settling interfamily disputes by violence. Value placed on firing pistols into the air. % of criminals in area who are punished by initiative of outside government. No. of public meetings using night lighting. No. of night meetings of community government. 4 Electric Resource Value to Communi~ Co-operative Action (4 variables) KW used for outdoor sports. No. of places for team sports.
Note: A "0" after a variable number indicates one of the dominant variables which was assigned to the group to which it had the most relationship. However, the dominant variables are strongly related to other groups as well. 66
REFERENCES (1) Barnett, H. G., INNOVATION: THE BASIS OF CULTURAL CHANGE, McGraw-Hill Book Co. Inc., New York, 1953. (2) Rogers, Everett, MODERNIZATION AMONG PEASANTS, Holt, Rinehart & Winston, New York, 1969. (3) Nair, Kusum, BLOSSOMS IN THE DUST, Gerald Duckworth, London, 1961. (4) Thompson, Laura, CULTURE IN CRISIS: A STUDY OF THE HOPI INDIA~S, Harper Bros., New York, 1950. (5) Thompson, Laura, TOWARD A SCIENCE OF MA~KIND, McGraw-Hill, New York, 1961. (6) Love, S. F., "A New Methodolo9Y for the Hierarchical Grouping of Related Elements of a Problem,u IEEE TRANS. ON SYSTEMS, MA~ A~D CYBERNETICS, SMC-2, No. 1, January, 1972, pp. 23-29. (7) Love, S. F., "Socio-Economic-Technical Subsystems for Addition of Electric Power to a Mexican Village," September, 1971, unpublished. A summary is published in the PROCEEDINGS OF THE 1972 INTE&~ATIONAL CONFERENCE ON CYBERNETICS AND SOCIETY, October 9-12, Washington, D.C.