Written by Mel Reichler Copyright 2002
Mel Reichler Associate Professor Emeritus,
Department of Sociology, Queens College Flushing 11367, NY
mwl@nyc.rr.com
A Critique of the Concept of norms
The concept of norm, unlike some others in sociology, appears to be unusually free of complexity. It seems to identify unambiguously a clear-cut and ubiquitous element of social life (Blake and Davis, 1964). The idea that every group has rules (or norms) has become so natural that it is hardly identified anymore as an idea, as an abstraction on the world. To speak of norms is no longer recognized as a way of conceptualizing the world; it appears to be a mere acknowledgement and reflection of social reality itself. Because the concept does not seem to be incumbered by a commitment to nay particular theory or to a particular point of view it is likely to be employed with considerable freedom. The concept is used with some feeling that it is relatively unencumbered, precise and simple. It is the argument of this paper that this is not so.
I
The Theoretical Context of the Concept Norm
The norm concept occupies a secure place in the conceptual scheme of contemporary sociology. It is generally understood to be one of a set of concepts that identify the ideational elements involved in social action (Homans, 1950:121-127). There is generally agreement that more than ideational components are involved in action. In the perspective of contemporary sociology, social behavior is not to be understood as a simple emanation or expression of norms; hence norms, of themselves, are not sufficient to explain behavior. The pluralism of contemporary sociology affirms, in fact, that social action is not ever determined by one type or kind of thing.
Although norms(and ideational elements in general) are only a part
of the mechanism determining behavior, much remains to be accomplished by way
of theory about the ways the different kinds of elements interact to determine
action (Blalock, 1969). In the absence of such theory it is difficult to
clarify whatever ambiguities remain associated with key concepts. Because the
norm idea has not been well integrated with elements of a more adequate and
comprehensive theory, there is a drift to expressionistic and emanationist treatments even where these have long been
recognized and condemned as adequate (Parsons, 1937). Without a solid mooring
in a grounded theory, the term cannot be held to a single fixed role, and it is
used in a variety of (at times conflicting) senses [1] (Rommetveit, 1954). It may be that, paradoxically, at least
some of the usefulness of the term in contemporary
sociology comes about because only its general import as an affirmation
of the importance of symbolic cultural and ideational elements is closed. It is
to the concept in this general context that this paper is oriented.
Problematics Associated with the Concepts Norm
Although its place in sociology is secure (no one has suggested abandoning the concept), a number of difficulties and problems have been noted in connection with the use of the concept of norm. Three are of general enough significance to be dealt with here.
1. Generally speaking, there appears to be a discrepancy
between the theoretical weight of the concept and its value in guiding vigorous
empirical work. The stress here is on rigorous. The concept is
paradigmatically used in two different ways in empirical research. In one kind
of study a norm is used to account for variations in the rate or level of some
other variable which is of primary interest. These variations are seen and
explained as an indirect consequence of the existence of the norm. In theses
studies some attempt will usually be make to document the assertion that the
norm involved really is a rule held by a group (Phillips, 1969). In another
kind of study, where the operations of a group is of interest, one or a few
norms are isolated and described in detail. Generally, in empirical work, only
a few norms are ever described or analyzed formally and completely, i.e.,
rigorously. Very few analyses of groups have appeared in which the concept of
norm is used systematically and and exhaustively as
the crucial structural element it is suggested to be in theory. Almost no
studies treat a group in terms of the entire set of norms which characterize
it. Gibbs (1965) makes this argument noting the "... paucity of systematic
studies of norm in general...." (cf. Gibbs, 1968).
If the rigorous empirical use of the term is limited, its theoretical weight is much less restrained. It is often credited as a core theoretical element (Williams, 1960; Sorokin, 1947; Parsons, 1951) Moreover, we often deal theoretically with systems of norms. In empirical work such systems of norms have almost never been fully described, although reference to such a system and some sampling of it are customary. Theorists threat the matter of the limited rigorous empirical use of the concept as relatively inconsequential matter whose explanation lies in entirely practical problems. In theory, a comprehensive set of activities of a group is dealt with using the concept; the few examples which are then illustratively offered are taken as establishing the general applicability of the term. There is a theoretical inflation of the value of the term over its established empirical utility.
2. The definition of the concept is not stable, and the meaning of the term in the vocabulary of sociology is not fixed. Even though the concept is of considerable vintage [although its general use is more recent (Williams, 1968)] it has proved remarkably difficult to fix its meaning except in a quite general sense. Everyone "knows" what the term means, but no one has been satisfied with any particular precise construction of that meaning. Gibbs presents seven definitions of the term(each from a reputable sociologist) which he notes agree on only two points, and these are on what the term does not refer to (Gibbs, 1965). Sometimes a norm is seen as a rule, sometimes as a pre- or proscription, sometimes as a standard, or sometimes it is taken to mean all three. (There are not insignificant differences; see von Wright, 1963: Miller and McNeill, 1969:667-668.) Definitions also vary in their identification of the locus of the concept (Williams, 1960:chapter 3), is established differently by different definitions. Finally conceptualizations vary in the way they set up the term: some position it for insertion into theory, others shape it so that it is directly useful in empirical work.
3. No typology which is both general and fruitful has been
widely accepted and used. The lack of a stable and complete typology of
norms is rooted, in some of the problems connected with its definition, and is
part of the general absence of models for the analysis of norms (
The Argument of the Paper
The problems which have been noted -- the question of the ideational-behavior locus, the issues of "rule" and "oughtness," the difficulties of fixing meaning and deriving a typology -- direct attention to some of the vagueness and uncertainty connected with the concept. It is not contradictory to argue then, while the concept has a secure position in the vocabulary of sociology, it is also problematic in a number of particulars. The situation in which an idea so seemingly fundamental, necessary, and elemental on the one hand is so unsettled and opaque on the other, is the core of the issue to which this paper is directed. Again, the proposition that every group has rules seems so elementary as to be incontrovertible. Yet this proposition represents a way of conceptualizing the social world. The consequences that flow from conceptualizing the world in this way are unclear.
The paper starts from the fact that the concept at various points is problematic. It assumes as a very tentative orienting hypotheses that the difficulties are rooted in the partition of the world that the concept introduces and the fundamental uses made of this partition, i.e., it assumes the concept is faulty. Under this hypothesis no simple modification of the concept or its uses, no reorganization of the attributes space associated with it (Barton, 1955; Gibbs, 1965) no change which does not drastically alter its role in sociology, will be successful in eliminating the problems that characterize it in use. It is assumed here that this hypothesis is correct. On the basis of this initial assumption, the body of this paper subjects the concept to critical analysis. The paper is of course, in no way a test of this basically untestable hypothesis; rather the hypothesis presents an opportunity to seriously examine the concept without a pre-commitment to it.
In the following section an attempt has been made to fix the
meaning of the term norm for the purposes of discussion here. In section III a
number of criticisms of the concept are presented. Every criticism of a concept
is, of course, an oblique criticism of the manner in which the concept is
employed in research and theory, that is, it is a criticism of the users of the
concept. But one may hold the primary fault to lie with the users because they
have been negligent or careless, or one may assign the blame to the concept
itself. To hold the concept at "fault" is
to question the usefulness of the way in which it partitions the world, the way
in which it suggests and commits one to fundamental uses. Concepts ripe for
such criticism saddle their users to a conception of reality which generates
paradoxical problems and theoretical dead ends.
None of the criticisms presented in the third section proves more than that greater care and clarity in the use of the term is demanded. They do not demonstrate that finer distinctions will not or cannot eliminate possible confusions. In Section IV the argument will be briefly presented that the problematics generated by the concept go deeper and are "intrinsic" to it and its uses.
The following qualification and clarification is in order. The criticism of the term should not be taken as a denial of its usefulness but rather a way of raising the question which construction of the concept is useful in which way. Given a definition which identifies norm as an observable it is undoubtedly true that exemplification can be found in the real world. NOrms in this sense exist. But this does not settle the real issue which is, does the concept catch a critical aspect of the world and does it construe this critical aspect in a constructive way. Essentially a critique should expose the limits of the usefulness of the concept.
Section II
A Working Definition of the Term Norm
The concept of norm is variously represented by a whole set of
definitions (for a representative sample see, (Gibbs, 1965; Gould, 1967). It is
to the set of reasonable definitions of the term, that is, to the manner
in which the term is used in sociology) that the criticism of this paper is
directed. Nevertheless, before criticisms can be advanced it is necessary to
make clear exactly what is being criticized; some particular conceptualization
must be designated as an object of analysis. The difficulties
created by the lack of consensus as to the meaning and use of the term
nor, complicate the fact that any selection will have self-fulfilling
consequences. The selection of a working definition might be facilitated,
however, if the differences between various definitions could be shown to be
primarily a result of varying weights assigned to the same or similar basic set
of elements.
The chart below summarizes the main elements indicated in common definitions of the term norm. These elements -- the form of the symbolic element, awareness of the element, quantitative specification, designation of the unit, specification of action and of actor, designation of the compulsory element, and the behavioral component -- constitute the categories from which the pieces of available definitions are drawn. The chart permits the development of a variety of definitions and families of definitions of the term norm. No significance should be given to the names of categories or to the designations of the entries; they serve here merely to identify components, not to characterize them analytically. The first row of entries constitutes the most restrictive, the third the most general, characterizations of the elements of each column.
Chart 1 Components of the Definition of Norm
┌─────────────┬────────────┬─────────┬─────────┬─────────────┬─────────────┬─────────────┬───────────────┐
│Form of sym- │Awareness of│ │ │ Action │ Actor │Compulsory │ Behavioral │
│bolic element│the element │ Quantity│ Unit │Specification│Specification│ element │ component │
├─────────────┼────────────┼─────────┼─────────┼─────────────┼─────────────┼─────────────┼───────────────│
│Prescription │ Stated of │ All or │ │ Specific │ Specific │ │Guides and is │
│ or │ statable │ Most │ Group │ pattern of │ actor in │ Must │Sanctioned │
│prescription │ │ ││ behavior │ situation │ │ │
├─────────────┼────────────┼─────────┼─────────┼─────────────┼─────────────┼─────────────┼───────────────│
│ Rule │ Held by │ Some or │Group or │ Acting or │ Person in │ Ought │Guides or is│
│ │ │ most │social │ Behaving │ status │ │Sanctioned │
│ │ │ │unit │ │ │ │ │
├─────────────┼────────────┼─────────┼─────────┼─────────────┼─────────────┼─────────────┼───────────────│
│ Idea │Charac- │ Some │Popula- │ Thinking │ People │Is expected │Is manifested │
│ │terizes │ │tion │ Feeling │ │ or│ │
│ │ │ ││ Behaving │ │supposed to │ │
└─────────────┴────────────┴─────────┴─────────┴─────────────┴─────────────┴─────────────┴───────────────┘
Consider two definitions, rather mechanically generated, using this chart. The first consists of all row 1 entries, the second of a selection from those in row 3.
1. A norm is a prescription or proscription stated or statable by all or most of the members of a group which designates the specific pattern of behavior which an actor in a situation must perform, actually guiding the actor's relevant behavior, and sanctioned by members of a group.
2. A norm is an idea about the way members of some population are expected to act or feel or think in (some) situation.
Although some published definitions come close to the extremes
represented by these two definitions neither would be acceptable to most sociologists;
each would be rejected however for different reasons. The first, besides being
unwieldy is unduly restrictive. The term is limited to prescriptions or
proscriptions for specific patterns of behavior. Its application is restricted
to cases where the pre- and prescriptions are statable
by members of a group. It also limits the term to pre- and proscriptions which
are actually followed and are also sanctioned. The second definition is too
general. The form of the symbolic element is not limited and the locus of
expectations is not fixed in any way. Its openness would permit too many
different kinds of elements to be called norms.
These definitions represent what might be called maximal and minimal definitions of the term norm. In selecting some conceptualization where there is no consensus on such a definition, it would seem reasonable to choose some middle ground. In the analysis below the term norm shall designate an element which can be defined as follows:
A norm is a sanctioned, or conformed to, rule held by some members of a group or social unit specifying the way that an occupant of a status ought to behave in (some) situation.
Under this definition, a norm (1) is required to have more form than simply an idea but may be more fluid than a specific behavioral prescription or proscription; (2) must be held by some members of a group or unit (although they are not required to be able to state it); 3) specifies a way of behaving (but this does not have to be a concrete and detailed pattern of behavior); and (4) requires some behavioral manifestation (but either conforming behavior or sanctions for non-conformity will satisfy this requirement).[2]
The sections which follow are oriented to this conceptualization. In places where the analysis would have to be modified to take account of the differences between this and other definitions, the fact is noted.
Section III
Criticisms of the Concept of Norm
The concept of norm is open to a number of criticisms. Perhaps the most clearly recognized of these is that in use, the term is ambiguous. "Norm" often appears simply as a notational term representing the whole set of propositions which constitute the presuppositions for the valid study of social life e.g., that order in some quantity exists, that social life is symbolically grounded, that patterns of behavior are not instinctive, etc. (Brown, 1963). This use creates no special problems.
The ambiguity that arises when "norm" is used in other ways takes at least three forms.
1. The concept is ambiguous because it is used to refer to both a rule for behavior and to an actual pattern of behavior.[3]In use, the term is extended to cover both the ideational and the behavioral realm. This extension occurs easily because although norm is defined as an ideational element it is an idea of behavior and "points" or refers to behavior. Moreover a rule for behavior is a norm only to the extent that real behavior can be mapped onto it, only insofar as it can be associated with some actual manifested behavior, i.e., is sanctioned and/or conformed to. The extension appears unproblematic because regularities of behavior are "perceived" as manifesting or embodying a rule (whether or not the rule is visible).
But problems are generated when the term is used simultaneously to refer to a rule for behavior and behavior itself. Every statement put forth as a descriptive proposition becomes simultaneously (but covertly) a proposition affirming an explanation of that behavior. Where these uses are not distinguished the act of describing becomes simultaneously, even if not fully, an explanatory act. Explanation is absorbed and incorporated into description. The issue is not, of course, the acceptability of this kind of explanation but the fact that, whether it is acceptable or not, it is accomplished de facto and covertly and, given the ambiguities of the term, very difficult to avoid.
2. Connected with this dual reference is a somewhat different
ambiguity. The behavior referred to by the term norm is in fact two quite
different sorts of behavior. There is a conceptual ambiguity to the term which
has not been closed by empirical work. On the one hand a pattern of behavior is
specified as appropriate to some actor in a situation. The norm specifies some
way of behaving as appropriate to or required of some actor. The appearance of
this behavior will justify a proposition that a norm is operative. On the other
hand the term also specifies a quite different kind of behavior, namely the
reaction of an audience contingent on the performance of actor; the term is used where, in fact, the behavior
required of the actor is not forthcoming but sanctions are applied. This means
that what is usually taken as a unitary element is always a family of elements,
a more complex unit formed of at least the following:
1. a rule specifying a pattern of behavior as appropriate to an actor in a situation;
2. a rule specifying another pattern of behavior for an audience to the behavior; and
3. an expectation of each actor that the other holds the rule to be appropriate.
A still further ambiguity is involved in the term norm, one which may have serious consequences. The term norm is used without discrimination in two major conceptual modes (Kaplan, 1964:12-17). The first of these is descriptive. Norms under this use have the status of observables or indirect observables. Basically, they are constituents of the world as perceived. In this use norm refers to a set of statements or set of behaviors of some members of a group or social unit, and under this usage it is possible with training to observe norms (Homans, 1950:121-127). They are units of uniformity in the real world.
In the second mode the term is placed more fully in the conceptual
realm; "norm" has the status of a construct or theoretical term. It
is used to explain observable. In this mode we treat behavior "as if"
it could be explained by norms -- "as if" it followed a rule. The
observations of uniformity of behavior constitute the occasions for the use of
the term, but the meaning of "norm" is not reducible to these
observations. When it is used as theoretical term, the concept of norm is
located in the sociologist's theoretical vocabulary without requiring either
that members of a group whose behavior is subject to analysis acknowledge that
a rule is operation or that the set of observations which are descriptively
implied by the term be present.
It may be that the use of concept as a simple observable is invalid because it involved too radical a simplification. At its position closest to observables the concept requires considerable abstraction and organization for a group of some distribution of individual idealizations. It is not, however, this aspect of the concept which is relevant here. The immediate criticism is directed to the fact that the concept has two distinct uses and that these are not distinguished in practice. This ambiguity makes it likely that a body of conflicting statements and propositions about norms will arise since in fact two different kinds of things are being referred to b the same term. In Section IV below it will be argued that this confusion of uses will account for man of the problems generated by the term.
A second criticism of the norm concept raises the issue of a different kind of imprecision. Within both the theoretical and descriptive uses the concept is not discriminating enough. It obscures rather than clarifies important qualitative distinctions. The term norm refers in practice -- sometimes intentionally and overtly, sometimes covertly and inadvertently -- both to rules for behavior and to "rules" for selecting rules themselves, that is higher-order "rules" (for articles dealing with a similar issue see (Scheff, 1967; Garfinkel, 1964). This extension occurs not only where the definition of the term permits it but also in cases where the definition would seem to preclude it, i.e., where norm is specifically defined as a prescription for behavior. By using the same term for both these kinds of "rules" we obscure their distinctiveness, and make it easy to assume that they all have the same character, that they are the same class of operators. The answer to the question "How shall we characterize second- and higher-order ideational structures in which the elements or parts are rules or sets of rules?" is prejudged by calling these structures "norms".
Thus far the form of what is taken to be shared under the concept
of norm has not been examined. The criticisms so far have not dealt with rule
conceptualization of behavior implied by the term. The concept of norm
associates a particular symbolic element (a rule for behavior) with behavior.
This construction would seem to rest on the presupposition that social behavior
is well or usefully conceptualized as flowing from or following rule. The
symbolic element standing immediately behind behavior is a rule where rule is
taken as a descriptively accurate representation of the symbolic element and
not simply as a way of speaking about whatever is there. In its general uses
the concept orients us to examine the world in terms of such objects, that is,
to comprehend behaviors as associated with these particular symbolic elements
(rules). It is not so much a matter of looking for such objects in the world as
it is of looking at the world as if it were
populated with such objects, of seeing the world in terms of such objects, of
seeing behavior and a rule for behavior as separately incomplete, as pieces or
parts of a unit, formed by their combination.
The concept designates a rule as the immediate symbolic element relevant to behavioral regularities. It specifies that the most immediate relevant symbolic element to some behavioral regularity has the form of an imperative, no matter how complex in combination, relating three elements: (a) behavior, or more generally, a feeling, belief or attitude, appearing in definitely discriminable chunks, and (b) a person, actor, or type of actor in (c) a discriminable social location or well defined social situation. The rule conceptualization of the symbolic elements standing behind behavior when closely examined seem to rest upon the unstated assumption that the behavior involved exists as a presumed piece and is produced as a piece, i.e., as a pattern of behavior. The problem of the rule appropriate to this situation, that is, to the selection of the "chunk" of behavior which fits this situation.
By conceptualizing the symbolic element immediately behind behavior as a rule, the degrees of freedom left in conceptualizing the nature of the organization of the social field composed of more complex ideational elements are severely restricted. This can be seen by asking the question: "What would the organization of rules of varying degrees of generality. The alternatives are (1) some dense hierarchical system of elements having the form of rules organized into a logically deductive system or (2) some system in which lower-order rules are specifications of, and subsumed under, higher-order elements of the same kind. Once the concept of norm is accepted as a lower-order element central to the workings of a social system, there is a significant limitation introduced on possible forms of social organization. Given the difficulties of supporting a strictly logically integrated view, there is a built-in bias in the norm conception to value-integrated theory of social systems. (This bias may be more generally based; see (Lee, 1959)).
The criticisms so far have focused upon what may be called the
formal aspects of the concept norm. But besides involving a specification of
behavior, the notion of norm associates an oughtness
(shouldness, mustness)
(Williams,1964) with that pattern and specifies also
the nature of a response to performance or non-performance, i.e., sanctions.
The manner in which these components are organized by the concept needs to be
examined.
The existence of a norm concerning an actor-situation-action unit means that for the actor and the audience the action available to an actor is to be considered limited. Considered as a prescription or proscription the norm depicts what this specific behavior will be; construed more generally a norm specifies that an actor's behavior should be limited to some class of behaviors (Goffman, 1961:91-95). Norms can be understood as conceptualizing the compulsory character of some act in a situation; they are a wa of representing that compulsory character -- the character of "compellingness" -- in a sociologically useful way. The concept is open to criticism for the way it specifies or accounts for this quality of compellingness.
The concept of norm locates this compelling character of an action or class of acts in the joint action of two elements; it identifies two sources of the compelling character of an act or class of acts. The first of these is an oughtness sui generis associated with the act. The second is in the rewards or punishments (sanctions) for the performance of an act.
Under the first the compelling character of an act is assigned to the fact that the act is evaluated according to some value or other as good, proper, correct, or right-- that, in other words, it is morally correct. The force of compulsion has its source in the evaluation associated with the act. The core here is a moral element (Blake and Davis, 1964). It is not, of course, the sociologist who asserts the act is good. It is understood that this assignment is made by the population being studied. The people concerned evaluate the act, and the compulsion they feel in performing it has its source in that evaluation. The criticism that this conception invites lies in the fact that the sociologist accepts and utilizes the vocabulary of "morality" to conceptualize the compelling character of the act.
The concept of norm takes the moral (or evaluative element)as a given primitive notion, and in doing so it leaves what
is basically an everyday conception of the moral largely unexamined and
unmodified. The moral may be conceived in a variety of ways. It may be thought
of as a vocabulary for dealing with an interpersonal and intrapsychic
economy involving special kinds of valuables, e.g., certain self feelings. Or
morality may be conceptualized as a theoretical component in everyday life, and
the moral may be understood as a domain formed by special kinds of entities and
special kinds of operations. Moral conceptions permit a direct connection of
behavior to non-present entities and to periods of time, past and future, which
would be otherwise unpredictable and unmanageable (Bellah,1964). Morality also
establishes meanings and may be constructively dealt with as a way of making
certain kinds of emotional reactions meaningful. the
variety of possible alternative conceptualizations of the moral calls into
question the utilization of a simple conception of the moral as primitive
element. It is precisely such a simple representation of the moral that the
concept of norm introduces. The common-sense identification of "compulsion
with" is imported into sociology through this and related concept.
The second source of compulsion identified by the concept is sanction, the reward or punishment associated with performance or non-performance of the act. If the moral element is effective to the extent that man operates as a moral being, pursuing that which is extent that man operates as a moral being, pursuing that which is evaluated as good and eschewing the bad, The second component is operative as man behaves in a more animal fashion, avoiding that which is aversive and punishing (whether it be moral disapprobation or a slap on the face) and pursuing that which is rewarding (that doing good may be its own reward is, of course, not denied). The concept of norm assigns the compelling character of a pattern of behavior to the joint operation of these elements. The concept of norm establishes that an act has a compelling character because it is both evaluated and sanctioned.
This juxtaposition which would seem to weaken or make unnecessary
each of the individual elements, may be accounted for by the fact that
experience demonstrates that while either of these elements might be sufficient
to establish a compelling character for behavior with which they might be
associated, neither is necessary. Seen in This light the concept of norm
embodies and is based upon a theory of human nature and motivation. Although
much empirical and theoretical work has gone beyond such a bifurcated view of
man, such a view is still held firmly at the heart of sociological theory by
the concept of norm. The idea of sanction is the least developed part of the
whole norm concept. Every response to an act has a sanctioning aspect; positive
sanctions are often no more than "normal" response to an act --
continuation in a regular way of the interaction. No models for the measurement
of the sanction variable exist. Notwithstanding the work of Homans
(1961), the concept in all its senses leans clearly to the side of the moral
element: an act is effectively compelled because it is correct and right; it
ought to be performed in light of certain values. Sanctions flow because the
act has this character.
The criticism offered here is not directed primarily ar these beliefs per se but at the concept as "embodying" them so that they can neither be chosen nor rejected but only taken. One of the consequences of the use of the norm conceptualizations is that it binds the theory in which the term must function to a value-like treatment of order and integration. This is so since the basis of the "morality", the values which support norms, are the critical elements in the system. In this sense the notion of norm biases (but does not limit) its users to a value integrated picture of society. This commitment may unduly restrict the ability to interpret and analyze the compelling character which an act or action may have. The concept of norm is significantly committed to a particular theory of social organization.
The critique of norms which is being offered here does not depend upon does it require the denial that moral elements which are sanctioned exist. The criticism is that the concept of norm identifies and reduces compulsion to values and rewards and punishment, and that this operation unduly restricts the ability of sociologists to interpret and analyze the compelling character of an act or action. The fact that the category of rewards (or punishments) is entirely open does not in any way solve the problem. By absorbing compulsion into the language of evaluation on the one hand and an (at times Skinnerian) utilitarianism on the other, certain issues are difficult to utilize. (Background understandings as identified b ethnomethodologists are clearly masked by the concept of norms (Denezin, 1969).) To take only one example, recent work by Schelling (1964) make it imperative to find room in sociological theory for sets of expectations which operates as compelling elements in the absence of moral evaluation; these expectations function as norms do, to guide behavior without the latter's property of being rooted in evaluations. Moreover, the rewards and punishments involved are neither external in the sense that they are applied by others nor internal in the sense of feelings of guilt or shame. They are merely anticipations of failure, and their locus is in the structure of the environment itself.
It would be possible to just make room under the concept of norm
for what might be called non-moral or non-evaluative, expectations. Gibbs
(1965) has suggested precisely this. But besides inserting a quite distinct
element into what seems a relatively homogenous group and overburdening the
term with a quite foreign use, the expectations discussed by Schelling would seem to call into question the whole
structure upon which the concept is built. They would seem to call for an
interpretation of the language of evaluation and morality in terms of language
of compulsion. It would call for an examination, in other words, of
"value" and "reward" as a way of conceptualizing the forces
generating the forces generating limitations on acts and actions.
Section IV
Grounds for the Critique
It was asserted earlier that the problems and difficulties generated by the concept of norm could not be eliminated simply by more careful use of the concept or by a reorganization of the attribute space associated with it. This is equivalent to arguing that the weaknesses of the concept are "intrinsic
to it in some way, that the problematics connected with it are a function of its central meanings and the basic uses to which the concept is put, rather than the care with which these meanings are put to these uses or to particular details of meaning and use. If these propositions are correct, then if one wants to use the concept as it is used in sociology one must take a large quantity of uncertainty and ambiguity with it; the confusion cannot be eliminated -- only its source, location, and appearance shifted. Yet no single one of the criticisms presented in Section 3, nor the set as a whole, seems on the surface to show that these assertions are correct; no argument presented so far supports these assertions and makes them plausible. What follows is an attempt to complete this analysis.
Two sources of the confusions and ambiguities associated with the concept can be suggested. The first of these is the dual scientific status of the concept, its simultaneous use as an observable and a theoretical concept. The second source of problematics is the presuppositional structure upon which the concept rests and which it expresses.
One may begin looking at the first source of the problematics associated with the concept by asking the
question: "What is the general place and utility of the concept of norm in
sociology?" The sociologist is interested in the understanding of actual
patterns of behavior, social organization and social systems. How does this
concept facilitate that understanding? It helps, using the language of the
concept itself, by making available a way of dealing with behavior using
elements of the symbolic or ideational world. It refers behavior to ideas more
accurately; it refers behavior to a particular class of symbolic or ideational
world. It refers behavior to ideas more accurately; it refers behavior to a
particular class of symbolic or ideational elements, which it identifies as
rules for behavior. Its use value inheres in the (asymmetric) bridge it
establishes between the symbolic and behavioral worlds.
But if, in general, this is what the term does, it does it, as has been noted, in two distinct ways. The term has two distinct uses. It is used as both a theoretical and as an observable term. As a theoretical concept it is at the center of the contemporary theory of social organization. In this theoretical status it is "horizontally" connected to such other terms as culture, values, roles, and institutions. Its value as a theoretical concept lies precisely in the set of connections it has with these concepts.
But in its other use, its observational use, norms are the material which is ordered by and explained by theory, not part of the theory itself. In its observational uses its "vertical" connections (to observations) are essential. In this latter use the term refers not to some element with theoretical properties but to one which can be observed and measured, if not with the senses alone then with their extensions in the form of various instruments, questionnaires, interviews, etc. The concept of norm plays a double role in the discipline of sociology. If this simultaneous usage as both a theoretical and observational term is recognized, then one major source of the instability of definitions and typologies becomes clear. As a theoretical element its symbolic or ideational character is of critical importance. It is theoretically significant only when it can be generally and widely applied over a range of empirical instances. Its theoretical weight depends on its ability to be associated with other theoretical elements of a similar kind. As a theoretical term its maximum utility is likely to coincide with (1) the maximum degree of freedom from any particular behavioral specification or assignment and (2) the degree to which it make available the entire realm of symbolic structures which can be construed as having a "rule-like" character.
But the some term has an observational use. When used in this way,
the openness to, or freedom from, behavioral specification destroys its
utility. Precisely that freedom which gives it theoretical power minimizes its
empirical usefulness. In its observational uses it is precisely the unambiguity and clarity with which it is closed via the
behavioral would that makes it productive.
This situation in which the term has two distinct uses which make different and opposing demands on the construction of the concept accounts for the instability of definitions and for some of the difficulties in involved in the development of a standardized typology. The tension between norms as theoretical constructions and norms as observables is located in the fact the quite different concrete norms or rules may generate the same action. A similar problem is faced in mathematics with functions defined using the notion of rules.
The second source of confusion originates in the fact that the concept is base upon some common-sense understandings about the world which may be inadequate for sociological understanding. It is of course true that any concept begins from taken-for-granted aspects of the world. In the case of norm, it is, in terms of our sense of appropriate kinds of objects in the world, and limits on these objects, that bounds are placed on the conceptual entities identified by the term norm. But what is true about the concept of norm that is not true about some others is that concept identifies both its referent objects and their properties at the level at which they are identified in common-sense use. These common-sense understandings are precisely formulated and lent scientific status by the term, but they are not otherwise strengthened or transformed. These particular common-sense understandings may not be adequate to handle the actual complexity of sociological phenomena. The point may be clarified if the functions of the concept are examined very briefly again.
The concept of norm is a bridge concept. It is productive only to
the extent that there is a general functional link between ideational elements
(rules for behavior) and behavior. It is useful to a sociologist interested in
understanding behavior (not ideational elements for their own sake) only to the
extent that rule for behavior are effectively linked
to behavior, and to the extent also that this linkage between rules and
behavior is not unusual or extraordinary, but common and general. Other factors
are expected to intervene, and hence rules do not by themselves explain
behavior, but the effectiveness of the connection must be present and dominate.
We must believe that this link is effective in order for the concept to have
any appeal at all. But why do we believe this? What evidence is there to
support such a belief?
If the application of the concept in its observational uses depended upon simple observations, then one might argue that elementary experience supported this belief. But this is not the case. A norm is a property of a group, not of an individual. Even for an individual an operative rule for behavior is likely to involve a considerable abstract summary of his, and others, anticipated behavior over a class of actors and a class of situation, a summary of a large number of particularized rules appropriate to sub-classes of actors and sub-classes of situations. In the form in which it is closest to an observational term it represents an idealization and abstraction -- a summary for a group as the unit of individual realizations and abstractions. Moreover, the belief that a rule attends behavior must stand in the face of considerable evidence that it is not so linked in many clear-cut and quite ordinary experiences. It is often the case that people are unaware of any given construction of such a rule only incoherently or superficially. if, finally, we were not convinced of the generality of the linkage, we would limit the term to instances where rules did in fact cover behavior, i.e., define a norm as a rule for behavior acknowledged by a population, and manifested in a corresponding sanctioned regularity in the relevant behaviors of the members of a group. But almost no definitions are so rigidly constructed (Gibbs, 1965).
If we believe, then, that rules are useful generally in handling human behavior, that it is profitable and sensible to conceptualize behavior as associated with a rule for behavior, it is not because our elementary experience makes such a proposition undeniable. If the basis for the belief in the link between behavior and rule for behavior is not to be found in experience, perhaps it can be located in other beliefs which make the link natural and unproblematic.
The concept of norm depends upon and incorporates into
sociological theory a set of common-sense beliefs. The first of these which has
already been noted is a theory that conceptualizes man as responding to
basically two kinds of motives. The second corresponds to a presupposition that
the key to understanding man's behavior scientifically (sociologically) corresponds
to the element by which his behavior is controlled civilly and religiously,
i.e., law. The concept of norm is related more strongly to "law" in a
legal sense than in the scientific sense (von Wright,1963).
The scientific conception built on this presupposition represents man
sociologically as an enlarged built on this presupposition represents man
sociologically as an enlarged civil or religious actor. Finally, and most
important here, the concept of norm rest upon a
set of assumptions about the relations between ideas and behavior.[4] These may be
summarized informally but not inaccurately by the propositions that: (1) the
ideational world and the behavioral world mirror one another or are mirror
images of one another; (2) behavior is reflected in or by ideas of behavior;
and (3) reflections when they are close must have the same "form" as
the behavior which they reflect.
Such a belief system supporting and expressed by the concept would
account precisely for those aspects the concept criticized in the third section
this paper. It would follow from this presuppositional
structure that the ideational reflection of behavior regularities are rules for
behavior. There is neither need nor basis for more complex symbolic theoretical
elements. Since most theoretical elements of sociology are currently drawn from
the pool of common sense ideational elements, there is neither need nor
justification for distinguishing rules for behavior for behavior from rules for
rules for behavior. The higher-order elements are elements of the same kind and
have the same form as the lower-order parts only they are correspondingly more
general, involving a larger class of actors and situations. There is no anxiety
abut using the same term to refer to behavior and ideas about behavior, or
rules for behavior, since they are "reflections" of one another. The
treatment of norm as if it were a simple observable makes sense within this
perspective. Since behavioral regularities are reflected accurately in rules
for behaving there is no problem if, in any particular instance of behavior we
construct the rule even if it is not recognized by the participants.
In this system of belief the problem of explaining behavior tends to be seen as
solved by finding the rule of which it is (an imperfect) reflection. Once the
rule is identified the only problem which remains is to account for its
selection and translation into behavior which is a matter of its valuational and utilitarian weights. Whatever variance
occurs from the rule pattern can be treated as a peculiarity of the particular
situation.
Every concept begins where a set of presuppositions leaves off. The presuppositions define the level at which the concept introduces a partition of the world, sets out objects and relations, etc. If this set of presuppositions is inappropriate and inadequate for dealing with the world, or if the presuppositional structure with which the concept articulates is contradicted by the other more general philosophical beliefs held by users, one may expect the problem to generate difficulties and problems. In this sense the concept can be said to be inherently or intrinsically problematic. It is this argument that is being advanced here about the concept norm.
Summary
Through its linkage of the behavioral and ideational realms the concept norm affirms a radical distinction between them. But the concept aligns these world; it does not integrate them. In both uses of the term the notion of norm is clearly associated primarily with the ideational realm. The only clear anchorage appears to be in the ideational or symbolic realm. It functions poorly as a bridge term by "pointing" to the behavioral realm. Its foundations in the behavioral field are muddied and obscure. The degree to which behavior must follow a prescription, the degree to which sanctions must be applied, what constitutes a clear and empirically discriminating conception of sanction, are empirically discriminating conception of sanction, are empirical questions for which no answers or generally accepted methods correspond.
The concept of norm is problematic in various degrees at various points. It has been subjected to criticism, and some attempt has been made to show that the weaknesses are rooted in the basic uses to which it is put and the presuppositional structure upon which it is based.
A re-examination of the concept of norm and the sociological
theory in which it functions is called for. The remarks of Noam
Chomsky in Language and Mind (19:222) might serve as guiding principles
of such an investigation. "We intend," he writes,
" too easily to assume that explanations must be transparent and
close to the surface." And he adds in another place in the same
essay,"...It seems to me that the essential weakness in the structuralist and behavioralist
approaches ... is the faith in the shallowness
of explanations, the belief that ... the most primitive of assumptions must be
adequate to explain whatever phenomena can be observed." The continued use
of the concept of norm requires a sustained disregard of problems, problems
which are generally acknowledged but attention to which is usually sacrificed
to the necessity which the term seems to meet. Perhaps an investment in an
examination of these problems will return
dividends to the whole of sociology.
/////////////////////////////////////////////////////////
norms.add
/////////////////////////////////////////////////
The argument of the paper;
We develop the argument in 3 sections;
1. In the first we fix the meaning of the concept of norm for the purposes of discussion.
2) In the second section we subject the concept of norm to a critical examination.
We try to show where it is that the concept is defective and faulty.
3. In the final section we try to show that the criticism of the concept is fundamental that no modification of the concept of its uses, no reorganization of the attribute space associated with it ( barton ....) will be successful...) We try to show that fine distinctions will not and cannot eliminate possible confusions; that the problematics generated by the concpet go deepder and are intrincic to it.
4. In the last section we relate the problem of norms in sociolgy to a set of more general issue that have appeared recently in the areas of cognitve science and artificial intelligence.
We show that the problem that socilogists encounter with the conept of norm represent particularizations in sociology of issues faced in cognitive science and computer science in the A.I field; by pointing out at least one alternative model for represneting and accoutngtin for order in social life we howpe to provide an rational and incentive for reexisinaing the fundamental character of social order which which the concept of norm plays a antered and perhaps reduced role.
At the end we will try to connect the difficulties with the concept of norm with other rule connected issues in contemporary AI and Cognitive science. We will show that the issue of norm is a particularization in sociology of a more general issue that has arisen recently in Ai and cognitive scinece. We will argue that the concept of norm in sociology represpresents a particularization in sociology of a much larger issue being faced by a number of diciplines simultaneiously.
The issues that arise in connection with the concept of norm are in fact much more genral issues that have surfaced in Artificial intelligence and cognitive science. New computational paradigms provicde sociology a way out of the dillemma of not being ablt to think about the world other than in terms of norms. Current work in AI and Cognitive science provide another model for conceptualizing regular behavior other than the norm or rule bound model.. They provide a basis for much richer and more complex models of behavior than the norm model permiteed.
The issues raised in this paper in connection with the concept of norms have appeared iin connection with arguements abourt the representaion of knowledge and learning iin artificial intelligence and cognitive scinece.
Rumelhart, McClelland et all argue "Computation apprahces to learning fall prdominantly into what mucht be called the "explict rule formulation" tradition... All of this owrk shares the assumption that hte goal of learning is to frormulate explity rules (propositions, production etc.) whichcapture poswerful generalizations in a succinct way." They offhandly summarize the approach of these approaches to learning--which aspply equally well to the concept of norms as they have beeen developed in sociology. " Fairly powerful mechanisms, usually with cond\siderable innate knowledge about a domain, and / or some starting set of primitive propositional representations, then formulate hypothetical general rules. e.g., by comparing particular cases and forumulating explcity geneeralizations." 32
Rumelhart and McClellland et al. take a quite differnt approach; the "do not assume that the goal of leanring is the formulation of explcit rules, " Rather we assume it is the acquisition of connection strn\engths which allow a network of simple units to act as though it knew the rules..." 32
The approach R and M and other have taken opens up a new way of looking at the regularlity of bheavior which which rules seemed the only possible concpetulziation. Rathher than norms producing the regularity we look for the regularity which neither takes the form of rules for behavior nor exists as such rules; the regularity exists ina more complext set of processing interconnections beh\trow indiviuduals which may be rperesented as a rule for one purpose or another w\but which does not exist as a rule.
Although R and M preference is for explanation ofpreviously rule existants by dissolving them into a set of simple connections through a long series of simple learning experiences. we do not have to accept that represenation conceptulziation as the appropriate one for sociolgoy.
On the other hand the possiblility of anothe r way of thinking about regularities in
interpersonal behavior is important. It opens the possibility of thinking about
regular patterns of behavior exhibited among humans and in groups in particular
without assuming without just assuming that
norms ( rules) are involved. Other notions elements of the framwork
of the connectionists may be reelvant to soociolgooy( I am thinkging of
the idea of multiple simultaneous constrinats) -- see
reicherl, a theory of friendship unpublished 1988);
The point is that there a\has appeared another gross way of seeing regularity
that escapes the simolicity of reductionishm
into individual imiitation and stresses pthe idea of patterning but apprahces
the matter without a committment to the notion of
rules as central and fundamental.
[1]To avoid repetition, and where
there will be no confusion introduced, I have used the works
"concept" and "term" interchangeably.
[2]The sanction for conforming behavior is most often merely the continuance in an expected manner of interaction itself. The term sanction has an empirical content different from interaction itself only where the response is discontinuous in some manner as is the case with negative sanctions (Goffman, 1961).
[4]The point is that the concept of norm analytically is supported by a certain world view. Needless to say, not every user would agree with this set of beliefs. In fact man would disagree. What is argued here is that these presuppositions define the limits and level of application of the concept. While a user need not agree with this view, he nevertheless by using the term "agrees" in a sense to treat the world as if it were the kind of world implied by these assumptions. Where in fact the user cleaves