|
p.4 management's task is less and less one of guiding performance in a relatively stable environment, and more and more one of organizational development.
p.4 The result of these shifts in social purpose and behavioral complexity is that the traditional ways of organizing social activities in the industrial era are inadequate to serve emerging needs. Society is confronted with an unusual array of unregulated disturbances.
p.4-5 A basic theme of this book is that resulting social problems are compounded by the fact that we tend to misperceive their basic nature. Our view of problems is characteristically superficial. We tend to identify them in "object" terms... our perception of problems and the formulation of remedies depends more upon our capacity to work with symbols than with objects. An unwanted disturbance, whatever its source, is not a problem because it disturbs, but because it is unregulated. Since the heart of social problem solving has to do with the regulation of disturbances, problem solutions are rooted in information processes prior, in sequence and in logic, to material and energy processes - i.e., object manipulation. The regulation of an unwanted disturbance requires a symbolic image of the relationships that form its origin, a representation of that image through the use of symbolic data of observation, evaluating that data through the application of symbolic norms and values, and activating countervailing processes... through the use of appropriate symbolic signals.
p.5 The problems that have become paramount in the present historical situation seem especially disturbing precisely because we are less prepared to regulate them than those characteristic of the industrial era. The organizational (information processing) structures, and the representative forms of data, successful in serving the requirements of that day are not adequate to organize the social processes essential to deal with disturbances that partake a different nature and origin. Behind the emerging set of social problems lies an emerging set of information processing problems.
p.5 this book... holds that far more is required to cope successfully with the current historical situation. It calls for more far-reaching changes in the structure and the symbolic resources of information processing systems themselves. Only now are we beginning to realize that our social problems are, at root, information processing problems of a more fundamental nature.
p.6 Of all the varieties of statistical reform proposed, the most widely based, and currently most active, is the social indicator movement... Some employ the term to designate an indicator of collective welfare... Some insist that these are "indicators of direct normative interest" in setting goals and evaluating programs.
p.7 Every application of the computer to data management requires some form of data banking.
p.9 We cannot safely assume that doing more of the same things will necessarily tell us more of what we most need to know.
p.13 What is not often adequately appreciated is the fact that general purpose data are always used to fill special purpose needs.
p.17 What first step is sufficiently large so as to promise a good prospect for further development?
p.20 The dictionary defines information as an act or process, not a "thing." ...For the purposes of this essay information can be understood as that which informs
p.20 Any element of symbolic data is just that - symbolic - and only represents a meaning or set of meanings... Most of the distorted meanings and misinformation experienced in life stem from the fact that we employ (out of necessity, economy, or laziness) secondhand symbol sources to represent the meaning of our environmental encounters... This exercise in semantics yields three important insights. First, it gives explicit emphasis to the fact that information, the generation of meaning, is the product of an information process. Second, the failure to keep this uppermost in our minds can itself be a source of serious information problems. Third, behind the design problem lies an unperceived theory problem. [JLJ - I would say that information is the product of a trick or technique of some kind that provides insight of a fundamental kind that resists misperception and allows an effective strategic response.]
p.22 When we speak of "existing and potential information processors," we are making an implied reference to our theories and models of social science and social organization. The symbolic representation of those theories and models governs the way we designate symbolic data elements... we need the guidance of a theory of information processing.
p.27-28 What we require is a theory that deals with the way in which social systems process information. Our task presents immediate difficulties: (1) we are not yet well-informed about the nature of social information processing; (2) no adequately developed theory exists explaining the way in which social systems process information; and (3) many of our most intractable information (read also social) problems are a direct reflection of our inadequate understanding of such processes.
p.28 It appears that every behavioral system is regulated by its information processes.
p.30 Although there is substantial literature in the social sciences that deals indirectly or implicitly with aspects of social information processing... I know of nothing that sets forth an explicit, generalized concept of social information processing. It is just such a representation, however impressionistic, that is central to this book.
p.31 it is important to note that human and social information processing is occupied exclusively with symbols (and even symbols of symbols). We cannot describe, understand, or interpret any aspect of information processing without a constant awareness of this fact... Inadequate symbolization and inadequate processing of symbols lie at the root of our most serious information processing problems.
p.32 we know the world only through our symbolic representation of our images of it, and these must be constructed through active, purposive interaction with the world.
p.36 A hierarchy of specialized information processing structures is one of nature's way of coping with complexity. [JLJ - perhaps 'ways' was intended here]
p.36-37 Schroder, Driver, and Streufert (1967) [JLJ - Human Information Processing] articulate the concept of levels of information processing in a way quite consistent with the concept of the information processor. They relate these levels to the complexity of the rule structures of the central processor. They identify four levels. In the first, the information processing structure is characterized by a single rule... It tends to yield polarized readings of the environment... Stimuli are always given the same interpretation... Black and white thinking and little adaptive capacity is typical. Adaptation is purely reactive... The second level permits two or more rules... At a single time one or the other is operative but not both. In this case, perceptions not only trigger the operation of the rule but influence which rule is operative. Once the rule determination is made, level two behaves like level one... The third level allows for a second-order set of rules added to the evaluative criteria so that the contingent operation of level-one rules can be eliminated... the environment can be tracked simultaneously from a variety of value perspectives. Perceptual data becomes richer and more complex... The whole process is less deterministic, more responsive to internal guidance, and more actively adaptive. The fourth level allows for a third-order set of rules that permit the combination and recombination of complex evaluative structures organized by the second-order rules... This level permits creativity and experimentation with concepts as a limited substitute for experimentation with behavior. [JLJ - the third level corresponds in my mind to the concept of critical success factors, and is an appropriate level of abstraction to consider for the game theory of complex games of strategy. The fourth level is ideally not really necessary, because it will be achieved anyway with evolutionary experimentation and the third level concepts, combined. Schroder, Driver and Streufert's book Human Information Processing is on order, stay tuned for a future review. Compare with Ashby's Law of Requisite Variety, which simply states that the complexity of the decision rule must match that of the environment. No matter how simple the cup-and-ball game looks, the team of participants on London Bridge is determined to complexify what looks simple and operates a clever scheme to depart with your wagered money.]
p.37 One can identify a further characteristic of higher-order levels of information processing. They tend to become more heuristic than algorithmic. The distinction is made nicely by Beer.
...An heuristic specifies a method of behaving which will tend towards a goal which cannot be precisely specified because we know what it is but not where it is.... To think in terms of heuristics rather than algorithms is at once a way of coping with proliferating variety. Instead of trying to organize it in full detail, you organize it only somewhat; you then ride the dynamics of the system in the direction you want to go. (Beer, 1972, p. 68-69.)
p.39 social systems initially come into being as extensions of human psychological structures.
p.40 Social information processors work with symbols.
p.42 The only way social information processors can cope with burgeoning complexity is to subdivide the total information processing task into specialized, quasi-autonomous information processors organized in a instrumental hierarchy... Beer (1972) and others point out that more complex problems of social regulation increasingly force social systems to employ higher-level heuristic information processing methods in the place of simpler algorithms.
p.43 Heuristic processes constitute another line of defense against system overload in the face of complexity.
p.43 A social system... consists of the coordinated activities of individual selves... social systems must enlist their components... to the support of group purposes, that this is often possible is a consequence of the fact that the product of individual behavior can often be amplified through social action.
p.44 a managed ecosystem... involves those situations where information processors (human and/or social), occupying a common environment, try to directly affect each others behavior.
p.45 an engineering metaphor... is useless when novel forms of social adaptation are required. The best that planners can do under these circumstances is to set up some initial objectives and developmental hypotheses. It is then up to individuals operating (singly and socially) as high-level information processors to experimentally develop novel structures. Terminal state designs give way to a form of evolutionary experimentation.
p.46 Environments made up of the relationships of social information processors are especially troublesome to represent symbolically.
p.50-52 There are a number of reasons why social environments confront social systems with increasing need to function as high-level information processors. [JLJ - formatting below added for readability]
- First, social systems... need to process information in the developmental mode... larger and more rapid environmental changes... often generate disturbances that stretch the adaptive capacities of the performance mode.
- Second, social systems are confronted with the growing complexity of the performance mode...
- Third, the regulation of conflict is a growing task of the social order... increasing activity in the developmental mode creates, in itself, new possibilities for conflict. The regulation of this conflict requires higher levels of information processing, relying less upon proscribed behavior to maintain social coherence.
- Finally, higher levels of information processing are essential to create symbolic representations of the environment... the structure and behavior of social entities become more complex and interrelated, more transitory, less precisely bounded, and hence, more difficult to perceive and symbolize...
In sum, our need for higher-level information processors to deal with the changing requirements of development, conflict regulation, the management of complexity, and environmental representation is outrunning achieved levels of information processing.
p.53 structures, like biological species adapted to a narrow environmental niche, may die out in the face of changes beyond their adaptive capacity.
p.60 Public systems increasingly confront problems that require higher levels of information processing. They involve regulating interrelationships that are manifest at system levels not matched by currently developed information processing systems.
p.61 mismatches in the interrelationships... The solutions to the problems these mismatches represent are not accessible to the organized components of the social ecology. They involve the coherence and stability of the ecosystem itself and must be addressed at an ecological level of regulation.
p.61-62 it was thought adequate for public systems to adapt reactively and incrementally to each disturbance originating in the social ecology by building "fences" that constrained in some fashion a limited subset of the interactions that comprise the ecological domain. But increasingly the disturbances we seek to avoid are the indirect consequence of our disjointed and incremental responses to earlier disturbances.
p.63-64 Forrester (1969)...
...complex systems are counterintuitive.... A complex system... behaves in many ways quite the opposite of the simple systems from which we have gained our experience... When we look for a cause near in time and space to a symptom, we usually find what appears to be a plausible cause. But it is usually not the cause. The complex system presents apparent causes that are in fact coincident symptoms.... As a result we treat symptoms, not causes. The outcome lies between ineffective and detrimental (pp. 9-10.)
p.65 At the outset, social science attempted to conform to the physical science model, feeling that this was necessary to gain sanction as a science. However, this had the effect of restricting the practice of social science to certain areas: [JLJ - formatting added for readability]
- where the issues of human purpose can be treated as a priori forms arising outside science;
- to those areas where social entities are largely "linear" in form...
- to those areas where social entities and their transactions are somewhat stable in time;
- to those systems whose relationships are largely homeostatic [JLJ - the ability or tendency to maintain internal stability in an organism to compensate for environmental changes] in character and whose regulatory mechanisms can be reduced to a simple rule structure; or, finally,
- to those areas where interpersonal relationships are either complementary or where conflict can be resolved by the influence of
- (a) free-market parameters,
- (b) submission or subjection to higher-order rule structures, or
- (c) as a later development, through classical bargaining games.
These conditions define fairly mechanical, deterministic social systems operating in relatively simple performance modes. [JLJ - social scientists rarely encounter complexity isolated from forms where humans are grappling with systems for simplification] It is small wonder that social science has concentrated heavily upon those issues of social management that focus upon social maintenance.
p.66 As social science is moving to consider the social processes of human relationship, along with the nature of social development and the management of complexity, it is being forced to adopt an evolutionary epistemology. Concepts of linear causation yield to concepts of contingent information processing relationships. Hypothesis testing shifts from the testing of formal relationships to the testing of evolutionary hypotheses through experimental simulations and social actions.
Under this view the emerging role of social science appears to be that of providing the epistemological roots and the representational techniques for modeling contingent social structures.
p.69 Our most insistent social information processing problems are not a simple function of inadequate data. They stem from the inadequacy of the information processing structures that form our social systems.
p.70 The different levels of information processing required by adaptive situations are associated with different kinds of representational problems. These, in turn, yield problems of perception and data generation. The fact that we are currently plagued with low-level information processors in high-level adaptive situations is, therefore, associated with the fact that existing data systems are incapable of supporting higher levels of information processing. Behind every problem having to do with the appropriate structure of information processors is a problem having to do with the representation of its environment.
p.70 The first characteristic of human information processors is that they process symbols. They work exclusively with data that are symbolic representations of those entities of the real world with which the information processor chooses to interact.
The second characteristic identifies the purposive nature of human information processing. The activities and events that constitute the real world in any comparative sense... vastly exceeds the capacity of any information processor to record them... An information processor can extract meaning from this confusion only if it guides sensors to record those activities that generate meaning in relation to its purposes.
p.71 our own information processors are heavily conditioned by the fact that they work most naturally with "objects" accessible to direct perception... We tend to overlook the true nature of the entities that social information processors must work with... It is hard to get it straight in mind that we don't "see" the social entities we work with; we invent them. They are creatures of our "image-ination" (shades of Boulding, 1961). They are often given symbolic representation through the use of instrumental artifacts that may yield not a direct image of a set of relationships, but one only inferred from a fragmentary set of observations
p.71-72 Vickers... has emphasized the importance of the entity concept as follows.
The entity is in fact a pattern of relationships, subject to change but recognizably extended in time... the object of attention is a dynamic system, a configuration of forces
p.72-73 The neurophysiologist Maturana (1970) describes cognitive processes in general as a process of constructing and reconstructing entities. Anything beheld by an observer takes the form of an entity. But an entity is not defined or described in object language. It is depicted as a set of meaningful interactions perceived by an observer.
p.73 In sum, an entity is formed by any set of relationships that yields meaning in terms of the purposes of observer system qua information processor. Thus, the operating environment is a set of quasi-ecological relationships which forms an entity.
p.73 The symbolized observations employed in the construction of operative representations of entities are data.
p.74 We begin by designating a social information processor as an observer system.
p.75 organized social information processors of any complexity have accounting and statistical subsystems. Such subsystems... count, measure and make qualitative observations that are converted into symbolic data amenable to further processing.
p.76 Patterns of relationships are formed by the transaction linkages among the organized social information processors that form the operating environment... These relationships do not form an image of another organized information processor. Rather they form an image of a set of quasi-ecological relationships. Representing these more generalized concepts of the operating environment is often desirable from the point of view of an observer because the indirect consequences of the coadaptations of a wider set of entities may be essential to guide its own purpose motivated behavior. Intelligent social regulation requires that we move from reactive adaptation to anticipatory adaptation, and the social problems themselves require that we bring patterns of ecological relationships under some form of regulation.
p.76 the pattern of relationships to be represented... are often so multifaceted and complex that they defy representation in detail. Observer systems are often required to deal with entities formed by the behavior of a population of similar and dissimilar component entities. The relational patterns represented must often be crude abstractions of underlying relationships. The representational task is complicated by the problem of identifying the levels of aggregation and abstraction that best serve the adaptive needs of the observer system.
p.77 The representation of the periphery serves several purposes. First, the same set of operating environmental observations may be interpreted differently against a different or changing background... Second, this is the area within which unexpected disturbances arise... The wise information processor maintains a surveillance of the peripheral environment because these less-focused observations add to its adaptive sensitivity.
p.77 Data assume different symbolic forms appropriate to differences in the representational tasks dictated by the adaptive situation encountered by an information processor.
p.78 The simplest form of symbolic data are those that designate the states (conditions), that characterize an entity at some point in time. They can be recorded by the simplest sensors, named or designated in the simplest form, stored in the simplest form, and used with the fewest complications.
p.79 It is useful to emphasize some aspects of the organizational logic of hierarchically organized information processors in order to appreciate the enormously effective power such simple indicator signals can have under appropriate conditions. Consider the design of a simple automatic heating system... The thermostat is the central processor of such a system... When combined with a normative goal, and a few information processing rules, the thermostat can effectively induce an appropriate performance by the heat source without any knowledge as a central processor of how the heat source subsystem processes either energy or information. (It does both.) ...As the internal and external operating environments become more complex, the simplest state indicators become inadequate. More complex representations of entity states must be developed.
p.80 Information processors that simply adapt reactively to immediate observations often encounter time-related anomalies... These time-related problems can appear in all kinds of operating variants... Such situations induce information processors to practice anticipatory rather than reactive adaptation. They attempt to set time-lagged adaptations in motion in anticipation of future states of the operating environment.
p.81 It turns out, however, that even these more elaborated state descriptions give adequate representations of the entity only under appropriate conditions. What constitutes "appropriate conditions" has to do with the stability and predictability of the activities or performances that generate the observed entity states and their changes over time... state descriptions of the kind discussed provide no explicit representation of the performance or activities of the entities being described.
p.82 The coordination of multiple options requires that some representation of the activities themselves be made to the central processor. State descriptions may no longer be sufficient servants of purpose... In short, while state descriptions in their ramified forms are adequate to perform many representational functions, they serve only the situations where the representation of the activities that formed them can be overlooked by the information processor without loss of function. This condition is fulfilled only in those situations where the underlying activity relationships are sufficiently stable that they can be represented by inference. In all other situations we need to generate activity descriptions that are representations of the performances underlying the resulting conditions.
p.83 Single-act descriptions can be combined to construct representations of more complex entity performances... one can elaborate the representations of acts into representations of performance patterns. This is a much more complex form of representation tan can be shaped with state descriptors. A wider range of meaning may be represented and these represented acts can be employed to construct an explicit network of relationships.
p.84 we can have no certain knowledge of the behavioral pattern that might emerge in a situation not yet observed. And we can project the same behavioral pattern under similar circumstances only if the underlying behavioral design is stable. [JLJ - I have previously looked at this concept in the light of using typical consequences to guide our actions, out of necessity, when we look into distant future and attempt to peer beyond the "horizon" effect of what lies beyond.]
p.84 The third form of entity representation involves the designation of the process (or the structure of the processor) that transforms patterned observations of environmental states and acts into adaptive actions designed to maintain or modify existing states... This form of entity representation is precisely a programmatic description.
p.84 Just as activity descriptions are more difficult to generate and employ than state descriptions, programmatic descriptions involve more difficult representational tasks than activity descriptions.
p.85 Our ability to evolve social systems capable of managing complexity... may well rest upon our success in conceiving, perceiving, representing, designing, and operating developmental information processors.
p.86 By now it should be clear that solutions to our information processing problems must rest upon our ability to generate adequate representations of operating entities and of the entities that make up the observational fields which constitute their internal and external operating environments.
p.86-87 The social information processing task of representing entities is a great deal more difficult when the entities are formed by human-social relationships... Since no real-world entity is completely isolated from relationships with internal and external entities... its designation requires specifying an arbitrary boundary between those relationships that describe one entity and those describing all other entities forming its environments. The “drawing” of this relative boundary is guided by the purposes and intentions of the observer... social entities tend to be less precisely bounded than material entities... human-social entities are artifact constructs of human purpose... Social entities tend to be less permanent and stable in form.
p.88 Precisely because social entities are artifacts of joint human behavior, they are repeatedly redesigned in response to adaptive encounters with changing environments, and to the transformations of technology and human purpose... the social entities one seeks to represent tend to display stable forms over shorter periods of time. Not only must the representations of social entities be reworked more frequently, but, because of the nature of our adaptive problems, we must increasingly employ higher-order forms of representation. Inferences previously supported by state descriptions must now be supported by activity descriptions... the observational fields within which social information processing takes place are being continuously transformed as the meaning and functional structure of each human-social entity is altered through a process of human-social learning.
p.89 How many business enterprises have failed for a want of a capacity for developmental information processing? There are many such examples of fixation. They are a product of our adherence to low-level performance modes of information processing in situations where they are inadequate.
p.89 Even where an information processor wishes to represent the entities of its operating environments differently, the ready accessibility of certain forms of symbolic data often reinforces the tendency to work with established representational images.
p.89 When management is confronted by the fact that its traditional processes are not well enough informed to deal with current disturbances, it often demands more “information.” This is often interpreted as a need for more of the same kinds of data accumulated in the same way... All too frequently, information processors seek to resolve their problems by accumulating masses of traditional, easily available state indicators, when these problems might be much more easily resolved by acquiring a few well-selected activity descriptors.
p.90 The symbolic data that social information processors require for both self-representation and the representation of their internal environments are predominantly data that each processor must generate through its own resources, using its own perceptual apparatus.
p.97 The end result of operating in a developmental mode is the establishment of new or modified performance modes.
p.97-98 higher levels of information processing necessary to meet the challenge of changing social environment dictate that... The developmental experiment needs to be monitored through an active, continuing representation of the external operating environments associated with its intentions. The systems' objectives must look beyond the requirements of adjustments in performance; they should focus on whether program purposes are being served. Beyond this, an effort should be made to detect any residual disturbance not regulated by the program, or secondary disturbances that result from the effects of the program upon its operating environment.
p.98 In any case, the practice of the developmental mode needs to be viewed as an iterative, never-ending process of entity representation and evolutionary experimentation.
p.103 all data are employed by information processors to "indicate" something.
p.105 The concept of social information processing makes it clear that, if one is speaking about developmental policy formulation, symbolic data restricted largely to state descriptor time series cannot make much of a contribution to the policy formulation process. Higher-order forms of representation are essential... Higher levels commonly require both the representation of structural patterns of state descriptors and the various representational forms of activity descriptions. Social indicator enthusiasts have not concerned themselves with identifying forms of symbolic data most needed to meet the present day requirements of policy formulation.
p.106 Sheldon and Freeman (1970, p. 104) emphasize that the paramount problems of the statistical indicator movement are not technical problems, but conceptual problems of knowing what to measure... Sheldon and Moore (1969) propose that we inventory available data in various substantive areas, evaluate their inadequacies and incompatibilities, and then, against this backdrop, refine existing series and add supplementary ones as needed. Such an approach can proceed only a limited way without employing a concept of systemic relationships to support the analysis.
p.107 [Olson]
A social indicator... may be defined to be a statistic of direct normative interest which facilitates concise, comprehensive and balanced judgments about the condition of major aspects of a society. It is in all cases a direct measure of welfare and is subject to the interpretation that, if it changes in the right direction, while other things remain equal, things have gotten better, or people are "better off." (Towards a Social Report, U.S. Dept. of H.E.W., 1969. p. 97.)
p.108 the meaning of perceptual data is never intrinsic to the data itself, but is supplied by a purposive information processing structure.
p.108 Olson (1970) has said:
If we have a meaningful interest in knowing what the changes in the Net National Product are, it can only be because the statistics measure, or approximate some quantity or quantities in which we have a direct normative interest. (p.203.)
p.109 Social accounting structure is not limited to aggregating indicators into composite measures. It can also be utilized as a framework for disaggregating social system descriptors. In this role, social accounting techniques facilitate the representation of a variety of patterned environmental complexes, a role to which the concept of social information processing gives much emphasis [JLJ - yes, they can importantly be used to estimate critical success factors, and once estimated, can be used as part of a strategy to indicate a direction or course of action which might yield an overall system improvement. Hmmm...]
p.110 [Jaszi]
GNP should be regarded as the center of a tableau [JLJ - tableau: a representation of a picture, statue, scene, etc., by one or more persons suitably costumed and posed] that represents a comprehensive and detailed statistical picture of the economic process in terms of the production of total output and its distribution among, and use in, the major markets whose dynamics [reflect] the functioning of the economy.
p.113 Social information processors are decomposed into subordinate instrumental systems and superordinate guidance systems that permit information structures to be partially and arbitrarily uncoupled at each system level. Consequently, a higher-order system's relationship with an instrumental subsystem is restricted largely to across-boundary transactions having to do with shaping instrumental purposes and monitoring subsystem behavior. All the relationships the subsystem employs to achieve the result, and the data that represent them, need not be known to a supersystem operating in performance mode. Similarly, the subsystem often needs only to receive parametric signals from the supersystem to perform its instrumental role.
p.113 Even if an organized social system needs to represent some aspect of its external operating environment, it does not wish to represent all of its ecological relationships. Rather, it seeks to represent only those relationships, or environmental complexes, that are meaningful in terms of its intentions... So we see that social accounting structures are no more than useful techniques for organizing data into patterned representations of entity concepts that are formed to serve social purposes. The structure of any such set of accounts derives from the purpose of its use. The governing principle is the logic of social information processing and not the logic of accounting structures.
p.135 social information processing can itself be described as a generic inferential process. Its generic task is that of inferring the appropriateness of its images of real-world entities on the basis of the adaptive value of the symbolic representations it constructs of them.
p.141 Entity concepts guide the construction of representational windows for observing the social world. The process begins with designating the entities, then designating the states, activities, and processes that describe them. This is achieved by naming the identifiers and descriptors that guide perceptors in counting or measuring the dimensions of the states and activities... These identifiers are then employed to guide observers or respondents in providing... counts or measures... of the attributes so specified... Observing the social world through a set of related representational windows yields a recurrent picture of the dimensional changes in a configuration of interest.
The representational windows that can be constructed is limited only by our ingenuity in naming those attributes related to our entity concepts and intentions.
p.142 traditional descriptors and identifiers usually yield data capable of serving only low-level information processing requirements... They are a mixed bag of fragmented images... and are not defined in ways that permit the representation of patterns of relationship required by higher levels of information processing associated with managerial complexity and developmental processes.
p.143 The entity problem is a reflection of that fact that adaptive social behavior requires a flexible choice of observation windows, so that we can perceive our changing realities.
p.152 We are at a time in the history of developing statistical systems when we face a substantial problem with these forms of obsolescence... A methodology can be employed with this class of problems as well... We need, for this purpose, a special kind of instrumental data, sometimes spoken of as longitudinal data. Since this term is ambiguous, it may be better to speak in terms of entity studies in time.
p.152 Entities designated in the data base should be kept under some surveillance. We need to keep track of their behavioral transformations.
p.153-154 There are no data forms more critical to the conduct of higher levels of information processing - or more notable by their absence from the data base - than those that facilitate representation of activity and process characteristics of social entities. Longitudinal studies focus on the changing behavioral patterns of the entities through time. But it is our present curse that the forms of data we require to represent patterns of activity are the most difficult to come by... this is a matter of considerable importance. Before we can achieve more comprehensive and valid inferences about the nature of the social world, we must achieve higher levels of information processing, based upon better entity representations, which require developing higher-order forms of symbolic data.
p.158 As we have dealt in this chapter with the problems of entity representation and the various remedial techniques that might serve to alleviate them, we have focused upon the creation of data elements which are designed and managed in ways that facilitate the flexible use.
p.159-160 No blueprint is available. We are not confronted with a conventional engineering design problem that specifies a complete, integrated system to be constructed by a preselected date. It is not a project for a year or a decade. We are confronted with a problem in experimental evolutionary development that has no boundaries. We must work towards capabilities which are more flexible and more integrated - that is, towards higher levels of information processing... First, it needs to be built around an established base of recurrent multiple-purpose data... Second, an integrated statistical service system needs a capability for generating more specialized nonrecurrent data, and for linking these data and more specialized data from other sources with the data elements in the recurrent data base... Revitalizing the data base would become a continuing consequence of a use-response system.
p.162 Improving the levels of social information processing of social entities is basically a matter of reforming their information processing structures.
p.164 If social science can look beyond its self-imposed limits, it will find it has positive and essential roles in solving our social information processing problems.
p.165 Social science should be working vigorously to solve the problems associated with observing, symbolizing, and storing activity and process descriptions.
p.165 We recall our earlier point that social information processing... can itself be described as a generic inferential process. We are reduced to inferring the meaning of our environments from our representations of them; our inferences and our adaptive behavior can be no better than our representations of the world.
p.165-166 Social science... at the level of representing processes, there has been a marked tendency to ignore all processes that do not lend themselves to neat specification as algorithms. We have avoided the exploration of heuristic processes that might support inferences in areas where our clues to the nature of adaptive behavior are weakest.
p.166 Our discussion of information processing problems and their remedies traces an arduous path, in keeping with the nature of the problem thrust upon us.
p.166 Our old planning paradigm presupposes that we can engineer or predesign social systems which will adequately serve some set of terminal requirements... Attempts to operate in this way have accumulated an impressive record of failure. What we need most is enhancement of the adaptive capacities of social systems so that they can become better social learners. This, in turn, calls for information resources of a different character, systems of symbolic data that can themselves be developed and adapted in response to change.
p.167 The ultimate test of expense is the cost of doing without something. If social systems are to achieve the levels of social information processing required to cope with their increasing complexity, we may indeed have no recourse to developing more adequate information resources. The cost may be trivial compared to the cost of social systems that lack adaptive capacity. In the concept of social information processing, information is the basic resource. No life system, including social systems, can stretch beyond the strict behavioral limits imposed by its information. [JLJ - yes, but information can be ambiguous, or it can merely hint or suggest, in tip-of-the-iceberg-fashion - what lies below. Strategically we constantly must ask ourselves, "how much should I care about that?", to the richly-detailed, prominently displayed, but often vague and ambiguous information cues of our social world. The answer to that question drives how we manage our attention. We must develop a posture or a stance, against both the known and the unknown, and both the vaguely present and briefly glimpsed. We must strive to develop a general-purpose adaptive capacity, to respond to those situations we cannot directly see, but that might/must nevertheless transpire, when other agents in our social world respond to our responses by likewise acting strategically. I am fascinated by the bid-ask posturing of the US stock markets, as brokers continuously seek to make money by establishing precise buy/sell positions, and then riding the dynamics of the market demands.]
p.167 If our overriding social problem is to raise the level of social information processing, we cannot continue to draw on statistical systems still structured to serve earlier simpler needs. And paradoxically, although our approach to the development of statistical systems is more complicated than in the past, its goal is to simplify emergent forms of social information processing.
p.190 We have never focused sufficiently on understanding our most difficult representational problems, those that limit adaptive behavior in a social world where important relationships are rarely directly accessible to perception and representation at the human level of observation. We have not devoted our concerns and talents to developing the concepts, methodologies, and data systems needed to represent our social operating environments in a way adequate to our adaptive requirements. We quite commonly misuse data by employing them to represent some aspect of our operating environment which, on reflection, we find they are clearly not adequate to represent. We often do this without hesitation or doubt, and even without much concern about the "reality" of our symbolic images.
p.194 [Friedrich]
All order, I would submit, is related to the ends for which any congeries of entities is assembled... political order... is a term suited to designate the political situation of the community in which component parts... are arranged in such a way that the actions required for the attainment of the purposes of the community will be taken. The crucial issue is whether things are arranged in such a way that there is enough order so that the actions occur which are needed in terms of the values, interests, and beliefs of the community. Even that amount of order is never fully achieved
p.195 It is an inevitable characteristic of systems structured as low-level information processors that they emphasize order at the expense of freedom. The lowest-level systems, as we have seen, tend to be organized to perform single-purpose missions by employing a single rule (or set of rules in fixed relationship) which determines the activities of human and social system components. They play the role of automata, and the system itself is essentially a closed one, adapting reactively to the signals of environmental parameters... A higher level of information processing structures permit the system to track multiple purposes simultaneously.
p.196 As Friedrich points out, a great deal of order is essential to provide for the exercise of higher-order freedoms. Conversely, we can say that in a complex, changing environment, social order has a way of slipping into disorder unless social entities are free to recreate order - to bring emerging sources of disturbance under control. [JLJ - another argument that supports my diagnostic test of the adaptive capacity to mobilize coercion concept - we must coerce our opponent (and resist his coercion) if we are able to make any progress towards our end goals in a complex game of strategy.]
p.196 In summary, all who have a stake in the maintenance of social order and freedom in our rapidly changing environment have a stake in the development of the higher levels of information processing necessary to create and manage open, adaptive systems. Whether, in fact, social information processors can evolve advanced information processing structures in time to regulate the disturbances of our time is very much an open question.
p.198 It seems to me that, at this time in history, our primary concerns have become misplaced... The fundamental question is whether we have time enough to gain a sufficient understanding of the nature of social information processing, so as to acquire the information resources and achieve levels of processing sufficient to deal with these vital problems. We fail to comprehend that we must first resolve the information problems of society.
|
