BEYOND DUALITY: THE APPLICATION OF CONSTRUCTIVE ATTRIBUTION TO THE CONCEPT OF THE SOCIAL NORMIn: B. A. Banathy, ed. Proceedings of the 43rd Annual Conference of the International Society for the Systems Sciences, Asilomar, CA, June 26 - July 2, 1999, 12 p.
The main purpose of this paper is to illustrate the applicability of the Aristotelian method of constructive attribution to contemporary science. Aristotle used this method to ascend from the abstract duality of opposition to the concrete structural representations of reality. This paper applies constructive attribution to the opposition of norm and deviance, which plays an important role in contemporary social sciences. It develops an abstract theoretical model, based on opposition, into a concrete model of "structural standard". In turn, the effectiveness of the structural standard model is illustrated by two applications: (1) the theoretical interpretation of human activity in conflicting norm situations and (2) deduction of strategies for the evolutionary design of “tolerant-supportive” systems.
Keywords: constructive attribution, structural standard, norm, deviance, normative status of an act.
Following the activity approach, this paper views duality as a result of the application of opposition, the norm of ontological construction traditional for Western science and philosophy. Aristotle was the first to recognize, describe, and systematically use this norm in his work. Aristotle should also be credited with the development of the method of ontological construction which we call constructive attribution (Dubrovsky, 1999). Built upon opposition, constructive attribution permits us to make a step beyond opposition and can be characterized as ascending from an abstract opposition to a concrete structural ontological representation. Opposition and constructive attribution are two consecutive steps of the Aristotelian method of ontological construction.
Developed more then two millennia ago, Aristotle's method of ontological construction has more than historical interest. We previously applied the principles of opposition formulated by Aristotle to the solution of logical problems associated with the concept of norms in contemporary social sciences (Gibbs, 1990). We showed that the problems resulted from violation of the principles of opposition formulated by Aristotle. We repaired the concept by bringing it into conformance with the principles (Dubrovsky, 1996). Although the resulting ontological model based on opposition of norm and deviance was free from the logical flaws, it was too abstract to account for known empirical phenomena or to be applied to practical problems. In this paper, we employ the Aristotelian method of
constructive attribution to develop a concrete ontological representation of norms--the structural standard model. We demonstrate the effectiveness of this model using two applications: (1) theoretical interpretation of strategies people use in conflicting norm situations and (2) deduction of strategies for evolutionary design of “tolerant-supportive” systems.
2. Opposition of Norm and Deviance: Abstract Structural Standard Model.
2.1. Concept of Norm in the Social Sciences
The following are a few typical definitions of norms.
... any standard or rule that states what human beings should or should not think, say, or do under given circumstances (Blake and Davis, 1964, p. 456).
The term norm became accepted in social psychology to refer to accepted standards of social conduct, definitions of correct judgment and appropriate activity or attitude (Jackson, 1988, p. 123).
All societies have rules or norms specifying appropriate and inappropriate behavior, and individuals are rewarded or punished as they conform to or deviate from the rules. The norms are blueprints for behavior, setting limits within which individuals may seek alternate ways to achieve their goals. Norms are based on cultural values, which are justified by moral standards, reasoning, or aesthetic judgment (Broom and Selznick, 1963, p. 68).
Despite the lack of agreement among social scientists in regard to norms, there are two common characteristics of norms accepted by most. First, norm is defined as a standard for behavior which is collectively evaluated as appropriate in opposition to deviance as inappropriate behavior contrary to the norm. Second, conformity to norms is rewarded by society or by groups while deviance is punished. Consideration of these two features is sufficient for our analysis at the level of abstraction corresponding to opposition.
The concept of norm plays a significant role in contemporary social sciences: "The terms norm and normative are favorites in all of social science. They appear as explanatory concepts in discussion of almost every topic" (Jackson, 1988, p. 123). At the same time the concept of norm "has remained peripheral to the main body of theory and research." (Jackson, 1988, p. 110). We believe that this situation is caused by the logical deficiencies of the concept of norm, which Gibbs (1990) characterized as "horrendous". He concludes: "Given the foregoing litany of seemingly insoluble problems, the only solution is to abandon the notion of norms when formulating theories or conducting research..." (Gibbs, 1990, p.
488). Sociology is focused on the norm in its relationship to deviance and social control. Social psychology is focused on the norm in its relationship to conformity and social influence. Despite this difference, they share the same deficient concept of norms.
2.2. The Repaired Concept: Abstract Structural Standard Model
We analyzed the deficiencies of the concept of norms catalogued by Gibbs (1990) in the spirit of the activity approach and by means of the Aristotelian principles of opposition. We repaired the concept by bringing it to conformance with the principles of opposition (Dubrovsky, 1996). The resulting Abstract Structural Standard model (Figure 1) represents the concept of social norms at the level of opposition of norm and deviance as contraries.
Figure 1. Abstract Structural Standard Model.
As the model shows, in order to solve the logical problems of the concept of norms, we had to modify the existing concept in several significant ways. Three of them were most important: (1) we defined norms and deviance in opposition to each other; (2) we introduced the concept of social standard as a common basis for opposition of norm and deviance; and (3) we used norms of sanctions instead of sanctions in definition of norm- deviance.
2.2.1. Opposition of norm and deviance upon the basis of standard
With exception of the Labeling Theory (e.g., Becker, 1963), social science defines deviance as behavior contrary to a norm. But even Labeling Theory, according to its critics (e.g., Polner, 1974 and Gibbs 1990) implicitly treats deviance as a behavior contrary to a norm. This way of opposing norm and deviance does not conform to Aristotelian principles of opposition, and is deficient in several respects.
Aristotle identified three components of opposition: two opposites, of polar species of a same genus and the genus as a basis for the opposition (Metaphysics, Book 10, Chapters 4 and 7). In the social sciences, norm and deviance do not belong to the same genus. "Norm" is usually understood as an imperative, a standard or rule for appropriate behavior, while deviance is characterized as behavior itself. We believe that it is obvious that standard of behavior and behavior itself do not belong to the same genus. It is not surprising that a common genus or basis for the opposition was never mentioned in the discussion of norms or deviance.
We solved this problem in the spirit of the activity approach (Shchedrovitsky, 1975). First
"deviance" does not denote a particular behavior but a type of behavior (Gibbs, 1990).
For example, such labels as "nut", "pervert", "criminal", or "hooker" are deviant behavior types, which can be conferred on particular instances of behavior. In other words, types of deviance are used as standards for identification of instances of deviance, or inappropriate behavior, in exactly the same way as norms are used as standards for identification of instances of appropriate behavior. Criminal law has precise definitions of different crimes, according to which the accused are charged, tried, judged, and sentenced, i.e., which are used as standards in the judicial process. Medical definitions of diseases are the standards against which a particular illness is diagnosed and then treated. In human factors engineering, error taxonomies provide reference types of errors that are used as standards according to which particular errors are identified
Thus we suggest that "standard" is a more basic concept than "norm", and that "norm" and "type of deviance" are polar species of the genus standard. Thus instead of opposing norm (which is a standard) to behavior (which is not a standard), we oppose norm (as standard) to types of deviance (which are standards). More precisely, we oppose norm as standard for appropriate behavior, to types of deviance as standards against which inappropriate behavior must be identified and evaluated. So we oppose norm and types of deviance upon the basis of standard. Now every act corresponds to a certain standard, or every act has a normative status of either norm or identified deviance; the same is true for the objects involved in the acts.
2.2.2. Standards of social sanctions instead of sanctions
In the social sciences, norms and deviance are associated with positive and negative sanctions, or rewards and punishments: "Sanctions are positive or negative reactions to behavior that attempt to alter that behavior, or to increase or decrease its frequency... People receive physical or psychological rewards and punishments for their behavior, which encourages or discourages them from acting in a similar way in the future (Labovitz, 1977, p. 9). Without sanctions, norms and types of deviance would lose their imperative modality and could not function as standards. This prompted some authors to suggest that sanctions should be included in the definition of norms: "Neither the behavior nor the sanction is the norm; rather the norm is the entire rule that connects the two" (Johnson, 1995, p.191).
Although we agree in principle that sanctions should be part of the definition of norms, we see a major flaw in the treatment of sanctions as particular behaviors or acts in such definitions: "Social norms are defined either as evaluations or as sanctions. Evaluations are should or should not statements, while sanctions are actual rewards and punishments attached to particular behaviors" (Labovitz, 1977, p.39). The deficiency of such definitions is that they make it possible for an act to be normal and deviant at the same time: It is normal because it conforms to a norm and deviant because it happens to be punished. Or it is deviant because it can be identified as a type of deviance and it is normal because it is rewarded. But it only appears that the contradiction reflects the empirical world of everyday life where normal acts are not necessarily rewarded and deviant acts are not necessarily punished.
We believe that definitions should be congruent with everyday life experience and offer the following solution to the problem. Definition of norm and type of deviance as standards
requires sanctions to be a part of the definition to support the imperative modality of the standards. As part of the standards definition, sanctions themselves should be interpreted as standards, or imperative modal statements--norms of sanctions. For the concept of norm and deviance, it is not important that in the real world normal acts sometimes are punished or deviant ones are rewarded. What is important is that according to the norms of sanctions, normal acts ought to be rewarded and deviant ones ought to be punished. Also, for specific norms and types of deviance, the norms of response prescribe corresponding specific rewards and punishments. Sanctions violating such prescriptions should be considered as deviance and they themselves should be negatively sanctioned.
The argument that this is an abstract theoretical statement that does not correspond to everyday reality, where injustice is common, misses the point. In the activity perspective, the theoretical statement of justice is not a description of reality, but a prescription for practical action. A particular unjust sanction should be identified as deviant and ought to be resisted and punished appropriately. One who does not resist injustice should not cry injustice: abstention from resistance is a deviant act and is itself unjust.
Thus a correct opposition is the opposition of norm associated with the norm of appropriate positive sanction to types of deviance associated with the norms of corresponding negative sanctions, upon the basis of standard. Indeed, now every act can have one, and only one, normative status. It is either an act which is evaluated as normal and, therefore, ought to be rewarded, or an act which is identified as a type of deviance and, therefore, ought to be punished. By associating standards of acts with the standards of responses we made all constituents of the model species of social standard. According to this model, every act must be evaluated in the context of social interaction. The Abstract Structural Standard Model (Figure 1) represents the correct definition of norm and deviance.
3. Constructive Attribution of Norm and Deviance: Concrete Structural Standard Model.
Although the concept of the abstract structural standard solves some logical problems of the traditional concepts of norm and deviance, it has a major problem of its own. As a theoretical model constructed by opposition, it is too abstract to represent empirical facts or to be applied to practical problems. For example, according to the model, people interact as "cultural and judgmental dopes" (Garfinkel, 1967): Since any act is either normal or deviant, it would elicit reward or punishment, respectively. The model does not account for the fact that, typically, people cooperate, exchanging products of their work, or exchanging productive acts leading to the achievement of a common goal, rather than spending all their time rewarding or punishing each other.
Being part of culture, standards play the role of blueprints for social interaction, but these blueprints often are “drawn in blurred lines” (Jackson, 1988) only “setting limits within which individuals may seek alternate ways to achieve their goals” (Broom and Selznick,
1963). Usually the alternative ways are determined via normative processes of negotiation
and coordination. (Jackson, 1988). Concluding his study of conflicting social norms deriving from conflicting role obligations, Stouffer (1949) concluded:
In essay writing in this field it is common and convenient to think of a social norm as a point, or at least as a very narrow band on either side of a point. This probability is quite unrealistic as to most of our social behavior. And it may be precisely the ranges of permissible behaviors which most need examination, if we are to make progress in this realm which is so central in social science. For it may be the very existence of some flexibility of social slippage--but not too much--which makes behavior in groups possible. (Stouffer, 1949, p.717).
To be able to account for the above and many other empirical facts, the model of the abstract structural standard should be made more concrete. It should be unfolded beyond the level of opposition to accommodate what Aristotle would call intermediates between norm and deviance. For this task, we will use the Aristotelian method of constructive attribution (Dubrovsky, 1999).
3.1. Aristotelian Method of Constructive Attribution
The method of constructive attribution uses an abstract opposition model as its initial material and includes two phases: cross-attribution and same-side attribution. At the cross- attribution phase, two intermediates that correspond to certain empirical facts should be formally constructed out of the contraries by interpreting opposites as attributes of one another. For example, Aristotle opposed Form and Matter as contraries upon the basis of Substance. Then he constructed two intermediates, Thing and Element. Thing, e.g. bronze sphere, an empirical "particular substance" consisting of sphere (form) and bronze (proximate matter), was defined as "bronze sphere", where bronze is an attribute of sphere, or in general terms, "material Form". Symmetrically, Element (fire, air, water, or earth) was defined as "formal Matter", where Form is an attribute of Matter.
At the phase of same-side attribution, original opposites should be reinterpreted in the light of intermediates thus ascending the entire construct (two opposites and two intermediates) to a more concrete level. This is done by attribution of an intermediate, with "subtracted" opposite side attribute, to the same side opposite. For example, Aristotle attributed being a Thing (material Form), to the Form, "subtracting" the attribute "material" from "material Form", and defining Form as pure thing, or disassociated from any matter thought-of- thought, or "form of forms" (On the Soul, Chapters 4 and 8). On the other side, Aristotle attributed being an Element to Matter, but "subtracting" the attribute "formal" from "formal Matter", thus characterizing matter as a pure, or prime matter--an ultimate substratum of transmutations of elements (Metaphysics, Book 9, Chapter 7).
Figure 2 depicts the formal structure of the constructive attribution. In this figure A and B are opposites; bA and aB designate cross-attribution of B to A and A to B, respectively. I1 and I2 are intermediates while A(I1b) and B(I2a) stand for reinterpreted opposites A and B by means of same-side attribution of I1 to A and I2 to B, respectively; "b" after, not before,
I1 in A(I1b) means that the attribute "b" is "subtracted" from I1 in the reinterpretation of A, "a" after I2a in B(I2a) has the same meaning. The "link" stands for the ontological relationship between the opposites determined by the phenomena being explained, e.g. change, motion, creation, and destruction.
3.2. Cross-Attribution: Intermediates of Regular Norm and Remedied Violation
Social sciences are not the only areas where the concepts of norm, deviance, and standard, are used. Although under different names, laws, organizational procedures and policies, work standards, task descriptions, safety restrictions, product specifications, and alike play the same role as norms, i.e., prescribe what ought to be done and/or proscribe what ought not to be done. Acts and their results that correspond to norms and those that violate norms, law obedience and criminal activities, adhering to and violation of organizational procedures, clean air and water and air pollution, correct performance and errors, good and poor product quality, timely and late deliveries should be considered as well. Moreover, according to the activity approach, human interactions with natural objects and engineering devices should be considered as social ones, i.e. as acts that correspond to or deviate from norms of appropriate human activities. In these cases, conformance to norms or deviations from norms are not necessarily socially sanctioned. Instead a correct action is "rewarded" by good technical results and violations or errors are "punished" by bad technical results or a system’s failure. This expansion of the scope of application of the concepts of norm, deviance, and standard may give us more empirical material to reflect upon and from which to draw additional ideas for intermediates.
3.2.1. Regular Norm
Sherif (1936) characterized norm as "a range of tolerable behavior rather than a single value." Jackson (1966) proposed the "Return Potential Model" that represents the structural characteristics of norms graphically as a curve. He represented a range of tolerable actions as the part of a behavioral curve that has positive group evaluations. In work design, allowances, or allowed unproductive worker’s time, are usually established via negotiation between management and labor to account for personal needs (e.g. restroom), fatigue, and delays (e.g. machine breakdown) (e.g., Konz, 1983). Another example is tolerance in engineering, or permitted maximum amount of variance from stated nominal specifications (e.g. Brumbauch, 1982). Still another example is sales quotas, which assign minimal acceptable sales.
These examples point to a special type of norm that specifies a range of acceptable performance or acts. We will call this type of norm Regular Norm; we will call a range of acceptable performance tolerance. At our level of abstraction, several characteristics of Regular Norm should be emphasized. First, on one hand, Regular Norm is a norm established by a special institutionalized agency, or authority, e.g. industrial engineering department, team of designers, quality control department, or owner of a store. On the other hand, Regular Norm allows deviation from what can be characterized a desirable performance, e.g., "standard time", in the case of worker’s allowances, nominal specifications in the case of engineering devices, or maximum possible sales in the case of
a store. Thus, in terms of the opposites, Norm and Deviation we can categorize Regular
Norm as deviative Norm and characterize it as a "desirable" norm loosened by tolerance.
C o n c r e t e L e v e l
Genus - Basis
A b s t r a c t L e v e l
Figure 2. Formal Structure of the Aristotelian Method of Ontological Construction
Second, acts within the limits of tolerance usually produce acceptable results: products of acceptable quality, acceptable sales amount, or system’s performance on an acceptable level. In return, acts conforming to Regular Norm elicit regular positive social response, e.g., workers are paid their wages, salespeople meeting the quota receive their commissions, average students receive their diploma. One should distinguish an acceptable result and regular positive social response from reward. Sales people significantly exceeding quota get a bonus as a reward and excellent students receive a diploma with distinction as a reward.
Third, an important component of "technical" or social response associated with regular norm is tolerance absorbing/offsetting. Consumers absorb products of less than perfect quality by tolerating some inconveniences in using them. A car can withstand, or "absorbs," handling by a rough driver. Workers absorb and/or offset lack of precision in parts by additional efforts during assembling. Management offsets worker’s allowances by assigning more workers for the same job.
3.2.2. Remedied Violation
In many cases, an act that deviates from the norm beyond the tolerance limits can still be corrected by "remedial work" (Goffman, 1971), or performance of an additional remedial act that depends on the extent of the violation of tolerances. For example, one "English for Foreigners" course explains: "If you are ten minutes late for a meeting, you should say ’I am sorry I am late’. If you are fifteen minutes late, you should say ’I am very sorry I am late’." While in the case of Regular Norm, tolerance ought to be absorbed and/or offset by objects and people other than the actor, a remedial act must be performed by the actor him/herself. A remedial act can save the situation only if the deviation does not exceed some limits of remediability. For example, being too late for a date may result in the date’s leaving before one arrives. We will call deviation that can be remedied remediable violation. As the above example of being late to a meeting illustrates, remedial act depends on remediable violation, e.g., "if you are five minutes late, you ought to do this; if you are ten minutes late, you ought to do that".
Norms for remedial acts are determined according to the type and extent of remediable violation. So these two standards work in combination: for a certain type of remediable violation one ought to perform a certain remedial act. We will call this combination Remedied Violation. Remedied Violation is a standard, and, therefore is an intermediate between Norm and Types of Deviance. On one hand, remedied violation is a violation, because it exceeds the limits of tolerance. On the other hand, it contains a norm for a remedial act in the situation of remediable violation. On this basis, we categorize Remedied Violation as normalized Deviance, or following Aristotelian "attributive" language, normal Deviance.
Unless a remedial action is performed, remediable violation is deviance. It should result in a product that does not meet the standards, the system’s failure, and negative social sanction. So the purpose of the remedial act is to avoid this kind of "response" and make up for a product which does not meet the standards, prevent the system’s failure, and countermand negative social sanctions.
3.3. Same-Side Attribution: Ideal Norm and Violation
The phase of constructive attribution following cross-attribution is same-side attribution. The essence of this phase is concretizing reinterpretation of opposites by attribution of an intermediate to the same-side opposite, or attribution of Regular Norm to Norm and Remedied Violation to Deviance but with "subtraction" of tolerance and remedial acts, respectively.
Ideal Norm can be viewed as a pure Norm, or a Regular Norm with the tolerance limits shrunk to zero. It can be the achievement of some ideal fixed value, such as a 4.0 GPA by a student, or a zero-defect level by an organization. It can be the best performance, relative to other actors, such as winning the Olympic Gold Medal, or the largest possible sales (March, 1964). Also it can be performance at the "golden middle" level, such as the right level of participation, earning the highest approval of discussion group members (Jackson,
1966). In all these cases, the ideal, or close to ideal, performance ought to produce optimal technical results and/or elicit social reward. At Clarkson University, a student graduated with the GPA above 3.75 receives a diploma with Great Distinction, an Olympic athlete gets a Gold Medal, and a salesperson receives the Best Salesman award with a larger bonus. Since the Ideal Norm does not assume any tolerance, no absorbing or offsetting is involved.
We can view Deviance as "pure" Violation, or a remediless violation, deviation beyond tolerance limits to such an extent that no additional action can be a remedy or a remediable deviation but without the appropriate remedial act following. In either case, the "response" ought to be technical failure (unacceptable product, unacceptable performance, or system’s failure) and/or negative social sanction.
Figure 3. Concrete Structural Standard Model.
The Concrete Structural Standard model (Figure 3) overcomes some deficiencies of the abstract model.. For example, it allows representation of actors who are not "cultural and judgmental dopes" any more and who participate in "real -life" cooperation, performing their acts at the regular norm level, in exchange for a regular positive response, leaving rewards and punishments for ideal norm acts and violations. The concrete model permits explanation of the behavioral strategies actually used by people in practical situations, such as a conflicting norms situation (Stouffer, 1949). In our study of computer-mediated group consensus development (Dubrovsky, 1987), the members of decision groups, according to their answers to a post-experiment questionnaire, expected to conform to two conflicting
norms. One norm was "compromise and be willing to hear the other side". A conflicting norm was: "if one believes that she is right, she must argue for her point, or the group may go with somebody who is wrong." Correspondingly, in computer-mediated group discussions, students, who were unwilling to compromise were pressed to do so ("Please agree!!! You are the only hold out!!!"), called names (e.g., "pig headed"), sworn at, and threatened with physical punishment (e.g. "I’ll get to your face"). Students who agreed too easily were called "too agreeable", or being "without a back bone". So in common language, a correct strategy would be "do not be either too agreeable nor too stubborn". In terms of the Structural Standard Model, this can be translated into the following strategy: "Conform to both structural standards by maintaining actions at the levels of regular norm and remedied violation. Seldom compromise and seldom disagree. With more compromising or disagreeing, use explanations, new arguments, and apologies as remedial acts." "I will go along with it to get the hell out of here. I have a test tomorrow " is an example of the latter.
4. Tolerant-Supportive Versus Restrictive-Punitive System Design
4.1. Normative Status of an Act.
According to the Structural Standard Model, each performance of an act corresponds to a certain standard type, or has one, and only one, normative status with the ideal norm as the highest and violation as the lowest. Each performed act ought to get a type of response that corresponds to its normative status. Quantitatively each standard type should be viewed as a range. The closer an action is to the ideal norm, the larger is the reward, e.g., a diploma with distinction for 3.5 GPA and with great distinction for 3.9 GPA. The larger the tolerance, the larger is the offset required on the part of a system, e.g., the larger the worker’s allowances, the more workers who must be assigned for the same job. The more a deviation exceeds the tolerance limits, the stronger the remedial action required on the part of the actor, e.g. "I am sorry I am 10 minutes late" and "I am sorry indeed for being 15 minutes late". The more severe a violation, the stronger the negative sanction. In other words, as a measure, normative status of an act has its qualitative and quantitative components.
Because technical result or social response is represented in the model by its norm, or what results or response ought to be and not what they actually are, empirical investigation cannot stop at evaluation of an act and observation of the results or responses. It should proceed with evaluation of the results and responses as appropriate or inappropriate, and, if possible, observation of secondary responses to the results and responses. In other words, the same Structural Standard Model should be applied to an act and to the response to the act. Or still in other words, the actual result and social response should be, in turn, considered as an act and its normative status should be determined. Abstractly speaking, this "chain" analysis should be continued until the normative status of the response to a given act is clear. In our experimental observations of uninhibited verbal behavior in computer-mediated group decisions, we observed either no secondary response to a "fair" social sanction or an apology. However, if a group member perceived a sanction as an unfair or excessive one, they responded to it negatively, e.g. threat for threat. Another
example of such secondary reaction was an apology by a person who applied an unfair sanction: the apology was issued because this member had called another one "a stubborn pig head" right before he found that the "culprit" had already agreed but his electronic message came with a delay. Another aspect of empirical application of the model is tracing evaluations and sanctions according to their agents’ social status, power, strata, or subculture.
4.2. Tolerant-Supportive Versus Restrictive-Punitive Systems
Normative status of an act depends not only on the act itself, but also on its social and technical context. "...Whether a given act is deviant or not depends in part on the nature of the act (that is, whether or not it violates some rule) and in part on what other people do about it" (Becker, 1963, p. 14). Jackson (1966) distinguished between tolerant-supportive and restrictive-punitive environments. For example, an act can be a regular-norm (correct) act with a regular result, if a user works with a “tolerant-supportive” computer system permitting multiple reversals; and the same act can be violation, or an error, in a “restrictive-punitive” system which does not permit reversals. Another example is a tolerant-supportive organization with liberal allowances for childcare and, due to it, higher loyalty and dedication on the part of the workers versus its restrictive-punitive counterpart.
4.3. Strategies of upgrading human performance by system support.
As a theoretical model the Structural Standard permits deduction of the strategies for evolutionary design of tolerant-supportive systems. We will illustrate such strategies using a computer system as action context and user errors as deviations from norms of correct actions. There are two complementary approaches for upgrading normative status of user action. One approach is the improvement of user actions themselves by better training, retraining, and continuous education. A complementary approach is to develop technical means and organizational procedures, making the system more error-tolerant and forgiving and, thus, contextually upgrading the normative status of same user actions. Typically both approaches should be used in concert.
The model suggests that a general strategy for upgrading normative status of user performance could be stretching the system’s tolerance and remediable limits, on one hand, and supporting learning of better performance on the other. The model also suggests types of appropriate technical means. We use examples of software for upgrading user performance with computer applications assuming that most readers are familiar with them.
4.3.1. Upgrading violations to remedied violations
The definitions of violation and remedied violation suggest three ways to upgrade violations to remedied violations. First, since a violation is remediless deviation, a way to control actions-violations is to prevent them, or at least lessen their probability. This can be accomplished by (1) making violations difficult or impossible, e.g., using "forcing functions" (Norman, 1990). Forcing function is an action constraint that prevents one from performing an operation unless another operation is performed before. Another approach is
warning of negative consequences, or asking a user to confirm a computer command with remediless consequences while setting defaults on the safe side.
Second, since violation is a deviation beyond remediable limits, widening these limits to embrace an action-violation is another approach. This can be achieved by providing for additional remedial acts. For example, a programmer may add additional reversals, such as multiple-step undoes. Another example is a "garbage can" which keeps deleted files for an indefinite time, thus giving a user an opportunity to undelete them if desired.
Third, violation may result from failure to perform a remedial act following remediable violation. To lessen the probability of such a failure, prompting a user or suggesting an appropriate remedial act can be used. A forcing function approach can be also used for this purpose. Sometimes remedies for certain remediable violations are available, but users are not aware of them. Also it is possible that a remedial act is a complex one, so a user needs to be guided during its performance. In these cases, on-line context-sensitive help, specifically oriented to the remedies, not only provides an ad hoc help, but also may help user learn the remedies and upgrade their performance.
4.3.2. Upgrading remedied violations to regular-norm acts
One of the main differences between regular norm and remedied violation is the agency that ought to perform a compensatory act making up for the deviation. While remedied violation obliges the actor to perform a remedial act, regular norm obliges the system or other actors to absorb/offset for tolerances. This consideration suggests that a remedied- violation act can be upgraded to a regular-norm one if the responsibility to perform a remedial act were shifted to the system or other actors. This shift means that the limits of tolerance are extended to include a deviation that previously was a remediable one. In technical systems, this can be achieved by automating the remedial act. The automation transforms a remedial act formerly to be performed by an actor into a system process offsetting the deviation. An example is a computer system automatically repairing and then performing a misspelled command, instead of just informing a user about the error and prompting to re-enter a command correctly. Another way of upgrading remedied violations can be improving user performance under the system’s influence. Constructive error messages, which explain what was done wrong and how to do it right, may promote user learning and avoid the errors in the future.
4.3.3. Upgrading regular-norm acts to ideal-norm acts
System designers should always look for any possibility of making specific tolerances unnecessary, thus upgrading a regular-norm act to the ideal-norm one. A classic example is using a blend of natural-language queries and menus (Tenant, Ross, and Thompson, 1983). The menus not only eliminated typing and confusion, assuring that every query act had an ideal-norm status but showed all range of possible queries. Another example is the natural- language query system INTELLECT (Shneiderman, 1992). Before performing the user’s natural-language queries, the system displayed them on the screen rephrased in SQL query language. With time, users switched from natural language to SQL, a more efficient query language, thus upgrading the status of their queries from acceptable regular-norm status to the ideal-norm status.
4.3.4. Pro-active versus Re-active User Performance
The above strategies deduced from the structural standard model are applicable to user pro- active performance in relation to the system. But users do not always act pro-actively. Almost as frequently, users re-act in response to the system’s "acts". The structural standard model is applicable to the responsive user acts as well. User responses to the system are guided by the response standards which are elements of the structural standard model. The same act, such as punching in the face, can be an appropriate response to a violent act and a violation in response to an acceptable act. In many cases, especially involving novice users, these standards are carried over from non-computer environments. From the activity theory perspective, "know your user" means, first of all, know the standards guiding user acts in both pro-active and re-active modes. Every user response to the system operation, prompt, message, or display is an act having a certain normative status, depending on the normative status of the system’s "act". For example, arrogant use of "hackers" terminology in prompts, error and help messages not only confuses and frustrates the user, but also may be perceived as offensive, and thus provoke an otherwise inappropriate user’s reaction. The structural standard model can be applied to user responsive acts in the same manner as user pro-active performance. The appropriate design strategies for upgrading normative statuses of user responses to the system can be deduced in the same way as for pro-active performance.
The above strategies of continuous performance upgrading are especially well suited for evolutionary iterative design or prototyping. Changes in design of a system may affect tolerance and remedy limits. Therefore, special care should be taken to prevent degradation of normative status of performance. The means are essentially the same. The strategy is to review every change of the design for possible negative side effects on normative status of the main operations and, if any, providing for upgrading of the latter to at least the previous level.
The concepts of tolerance and remedy limits, tolerance absorbing and offsetting, and remedial acts are especially relevant to the modern workplace with its "fuzzy" task definitions. Actors should be allowed to select performance strategies via trial-and-error exploration of the system’s functionality and limits (Rasmussen, 1990). Modern systems should be tolerant-supportive with extended tolerance limits and corresponding extended tolerance absorbing capacities and offsetting devices; with extended remedy limits and corresponding variety of possible remedial acts.
The main purpose of this paper was to illustrate applicability of the Aristotelian method of constructive attribution to contemporary science. This paper applied constructive attribution to the opposition of norm and deviation to unfold an abstract theoretical model, based on the opposition, into a concrete model of Structural Standard. The resulting model was concrete enough to support empirical investigation as well as practical applications. We demonstrated the effectiveness of this model with two applications: (1) theoretical
interpretation of experimental data on action strategies in conflicting norm situations and
(2) deduction of strategies for evolutionary design of “tolerant-supportive” systems.
The Aristotelian method of ontological construction prescribes the building of a four-layer structural representation for each ontological unit (Dubrovsky, 1999). Constrained by the length of this paper, we discussed the "norm-deviance-standard" unit for the layer of contraries only. With all four layers in place the power of the model more than quadruples. It gains additional power due to the synergy emerging from the congruence among the layers. While in the mode of ontological construction, one has to assure that the resulting four-layer system is congruent, in the empirical world of real systems incongruence among their structural layers can be the source of system development and evolution or degradation and destruction.
This paper discussed a layer of ontological unit of Social Standard originated by opposition of norm and deviance as contraries. Another layer can be originated by the type of opposition which Aristotle called "possession-privation, or opposition of appropriate acts versus inappropriate acts. Appropriate and inappropriate acts are defined in terms of social values, roles, and situations. Mutual correspondence, or congruence of all four layers, is necessary for effectiveness of the norms. In case of incongruence, society, organization, or group is compelled to change the norms in order to re-establish the congruence. Suppose an organization has developed polices and procedures perfectly appropriate for a certain situation. The workers conforming to the policies are rewarded while violators are punished. Also suppose that the situation has changed, e.g., the cold war is over or the world market has changed, and the old procedures are not appropriate any more. In other words, the two normative layers become incongruent. In order to survive, the organization must replace the old policies and procedures by the appropriate new ones. In other words, structural norms are themselves a source of social development and progress, and multiple-layer structural representation of norms and standards is of principal importance for understanding societal and social development and evolution.
To summarize, we suggest that the full power of the Aristotelian method of ontological construction as a whole and construction attribution as its stage could not be demonstrated by only one of four levels of a structural ontological representation. The method is much more powerful than this paper could demonstrate.
Becker, H.S. 1963. Outsiders. New York: Free Press.
Blake, J. and Davis, K. 1964. “Norms Values and Sanctions,". in Handbook of Modern
Sociology (R.E. Faris ed.) Chicago: Rand McNally, pp. 456-484.
Broom, L. and Selznick, P. 1963. Sociology, 3rd Edition. New York: Harper & Row. Brumbauch, P.S. 1982. "Quality control", in Handbook of Industrial Engineering (G.
Salvendy, ed.) New York: Wiley & Sons, pp. 8.3.1. - 8.3.23.
Dubrovsky, V. 1987. "Social exchange in group consensus development: Face-to-face versus electronic mail." in Proceedings of 31st Human Factors Society Annual Meeting. New York, NY: Human Factors Society, pp. 701-705.
Dubrovsky, V. 1996. "Duality: The third approach." in Proceeding of The Fortieth Annual Meeting of The International Society for The Systems Sciences (ISSS), (M.L.W. Hall, ed) Louisville, July 14-19, pp. 235-246.
Dubrovsky, V. 1999. "Beyond duality: From opposition to constructive attribution." (This volume).
Garfinkel, H. 1967. Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Gibbs, J.P. 1990. “The sociology of deviance and social control,” in Social Psychology: Sociological Perspectives. (M. Rosenberg and R. Turner, eds.). New Brunswick: Transaction Publishers, pp. 481-522.
Goffman, E. 1971. Relations in Public: Microstudies of the Public Order. New York: Basic
Jackson, J.M. 1966. “A conceptual and measurement model for norms and roles”. Pacific
Sociological Review Vol. 9: 35-47.
Jackson, J.M. 1988. Social Psychology, Past and Present: An Integrative Orientation.
Hillsdale, NJ: Lawrence Erlbaum.
Johnson, A.G. 1995. The Blackwell Dictionary of Sociology: A Users Guide to Sociological
Language. Malden, MA: Blackwell.
Konz, S.A. 1983. Work Design: Industrial Ergonomics. Columbus, OH: Grid Publishing. Labovitz, S. 1977. An Introduction to Sociological Concepts. New York: Wiley.
Norman, D.A. 1990. Design of Everyday Things. New York: Doubleday/Currency.
Polner, M. 1974. “Sociological and common-sense models of the labeling process,” in
Ethnomethodology. (R. Turner, ed.) Baltimore, MD: Penguin.
Rasmussen, J. 1990. "The role of error in organizing behavior," Ergonomics, Vol. 33, No.
Shchedrovitsky, G.P. 1975. “Fundamental concepts and categories of activity theory.” In Razrabotka ee Vnedreniye Avtomatizirovaneeh System v Proyektirovaniye: Teoriya ee Metodologiya, (B.V. Sazonov, ed.) Moscow: Stroyizdat, pp. 72-161.
Sherif, M. 1936. The Psychology of Social Norms. New York: Harper Brothers. Shneiderman, B. 1992. Designing the User Interface: Strategies for Effective
Human-Computer Interaction, 2nd ed. Reading, MA: Addison-Wesley.
Stouffer, S.A. 1949. "An analysis of conflicting social norms," American Sociological
Review, Vol 14, No 6: 707-717.
Tenant, H.R., Ross, K.M., and Thompson, C.W. 1983 "Usable natural language interfaces through menu-based natural language understanding," in CHI’83 Conference on Human Factors in Computing Systems. New York: ACM press, pp. 381-386.