Recontextualizing the social norms construct as applied to health promotion

Highlights

• •An increasing number of health promotion/behavior change programs focus on changing social norms.
• •However, in current usage, norms are typically not linked to the underlying social and cultural context (decontextualized).
• •The use of social norms to change health behavior could be improved if norms were understood as culturally embedded.
• •Social norms may represent underlying cultural meanings and values, power configurations, or shared cultural models.
• •Social norms may serve as cultural tools for the exercise of individual agency.
• •There may be multiple and competing normative options in a given situation.
• •Social norms can be public or private, and norms change over time for many reasons.
• •There are ways to use social norms in health promotion programs that can help reconnect norms to cultural context.
• •These are described in the article along with a revised definition of social norms.

Introduction

The construct of social norms has been widely used in health promotion efforts focusing on social/behavior change. Programs addressing family planning; reproductive health; water, sanitation and hygiene (WASH); and substance abuse and tobacco use among adolescents, among other issues, often focus on changing norms related to risk behaviors as a mediator of individual behavior change. Because this kind of formulation is increasingly common, there are important questions to consider regarding how social norms as a change construct has been defined and applied, the underlying assumptions, and factors that may be obscured or unaddressed in current usage. Addressing these issues may help inform health promotion initiatives that employ social norms so that they are more grounded in the multivalent social realities that shape behavior.

In this paper, we consider the issue from an anthropological perspective, focusing our attention on the way social norms are often decontextualized in social/behavior change applications with respect to the cultural and social worlds in which they are embedded and help shape. We have four goals, to: 1) broadly summarize the ways in which the social norms construct is defined and used across multiple disciplines as relevant to and applied in health behavior change applications; 2) offer a constructive critique of those definitions and applications; 3) propose a set of theoretical and applied considerations concerning social norms that could be integrated in the planning and application of norms-related interventions in order to increase the validity of social norms as a change construct in such interventions, decrease the possibility of unintended consequences, and increase intervention sustainability; and 4) offer a working definition that reconnects norms to their social-cultural contexts, better reflecting the multiplicity of ways that norms can influence behavior.

Current definitions and usage of the social norms construct: a brief overview

First, to provide a common grounding, it is useful to review and summarize the ways in which social norms are defined across social science disciplines as well as in social psychology, and in a range of health promotion interventions. This is a brief review; there are more extensive reviews elsewhere (including Chung & Rimal, 2016; Mackie, Moneti, Shakya, & Denny, 2015; Bicchieri & Muldoon, 2011). Our comments follow the review.

Common definitions and usage in the social sciences

Norms as facilitating individual self-interest

Several individual-centered definitions arose in the context of game theory and its focus on maximizing individual outcomes within systems. Drawing from classical Western liberalism's primacy of the individual, the assumption of rational individuals from traditional economics, and utilitarian philosophy, game theorists have described norms as practices built from the rational choices of self-interested individuals who want to maximize their desired outcomes in an interdependent social world, where the interests of many individuals may conflict (Binmore, 1994; Schelling, 1960; Schotter, 1981; Sugden, 1986; Ullman-Margalit, 1977). Norms thus “solve” a collective coordination problem. Moreover, norms are said to evolve when “societies shift from one coordination equilibrium to another” (Boyd & Richerson, 1994, p. 73). In a similar vein, norms have been explained as behavioral choices that maximize efficient attainment of social goals (Akerlof, 1976; Arrow, 1971).

Norms as properties of social networks

Moving “outward” from a focus on the rational individual, social network theorists have conceptualized social norms as pertaining to reference groups and networks (e.g., Storey, 2016). Normative change is explained via network dynamics, including the actions of influencers and communication through networks of individuals, as a specific type of innovation diffusion (Rogers, 2003) where the innovation is a behavior pattern. In U.S.-based social network studies, norms have been defined as beliefs about what is acceptable within a social network (Leahey, Doyle, Xu, Bihuniak, & Wing, 2015, citing; Rosenquist, Murabito, Fowler, & Christakis, 2010; Christakis & Fowler, 2008; Christakis & Fowler, 2007).

Norms as linked to socialization and social cohesion

With respect to sociological definitions, social norms have been construed as the outcome of socialization patterns within a group or society (as in Parsons, 1951; Bicchieri & Muldoon, 2011), linking them to issues of social identity and conformity. According to Parsons’ formulation (1951, 1937), norms are essentially a set of behavioral alternatives that are internalized and considered desirable by individuals as socialized actors. The particulars of this behavioral set are shared by other members of a given society, forming its value system, a sine qua non of social solidarity. Social norms are further said to be markers of group identity and membership, where individual members of a group are likely to adopt group norms as part of their self-categorization (Cancian, 1975; Tajfel & Turner, 2004; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987); thus they are a means to provide individuals a coherent framework for behavioral choices. Norms also appear in the sociological literature as facilitators of social control (Cialdini, 2007; Innes, 2003; Janowitz, 1975), and gender norms are often linked to criteria for the legitimization of social authority (e.g., Nanda & Warms, 1980/2013; Sanday, 1974; Krieger, 1984, drawing from Max Weber).

Recent conceptual trends

Norms as descriptive or injunctive

Several recent explications of the social norms construct are widely cited and represent parallel thinking around key conceptual dimensions. Cialdini and colleagues (Cialdini et al., 2006; Cialdini, Kallgren, & Reno, 1991) proposed the bifurcation of social norms into descriptive and injunctive norms, where the former refers to what is commonly done (the “is”), while the latter refers to behaviors that are believed to be approved or disapproved (the “ought”) by others within a reference group. The locus is still the individual and his/her reference group.

Norms as connecting individual perceptions and group behavior

Lapinski and Rimal (2005) argue that a key social dynamic involved in the behavioral power of norms is reliance on collective wisdom and experience as a decision-making heuristic or “shortcut.” The actual, prevailing norms of a group, community or culture (independent of individual perceptions) can be seen as collective norms, a whole-group phenomenon. However, this “master set” of norms, so to speak, may be interpreted differently by individuals, or not known to an equal degree across individuals within a group. The term perceived norms thus refers to a given individual's interaction with collective norms, as an agentic phenomenon. Individuals have perceptions regarding the prevalence of a norm (perceived descriptive norm), and whether or not the norm is associated with sanctions (perceived injunctive norm). At the collective level, descriptive norms can be gleaned (by outsiders) from data about the prevalence of some behavior pattern, while injunctive norms can be identified via the existence of actual sanctions (legal, social, policy) for violating a norm. Rimal and Real (2003) also position injunctive norms, outcome expectations and group identity as moderating the effect of descriptive norms on individual behavior, adding dimensionality and “thickness” to the construct.

Norms as behavioral codes

In subsequent work, Rimal and Lapinski (2015) situate norms in a continuum of behavioral codes that includes codified laws and traditions as well as what are more commonly described as norms. The defining feature is that norms are “socially negotiated and contextually dependent” (Rimal and Lapinski (2015), p. 394), where laws are not because they are “explicitly codified proscriptions that link violations with their corresponding punitive measures” (Rimal and Lapinski (2015), p. 394). Traditions, on the other hand, are more akin to norms except that they are stable over time. Chung and Rimal (2016) also cite the term “folkways” as an early definition of (descriptive) norms, and they refer to the term “mores” as a social version of injunctive norms because social sanctions are involved if there is a violation. Clarifying the differences between these terms is an issue addressed later in this article.

Identifying moderators between norms and behavior

Important in this recent work is the focus on moderators between norms and behavior, “delineating pathways of and conditions under which they operate” (Chung and Rimal (2016), p.395). Some of these moderators include social contextual factors such as salience – the degree to which a particular norm is activated or emphasized in a certain social context, the accessibility of particular attitudes about a norm within a specific social context, and social modeling of behavior where an individual engaging in a behavior is perceived and treated as favorable, as in Bandura's construct of a social model (Bandura, 1986; Chung & Rimal, 2016; Rimal & Lapinski, 2015) or a prototype in the prototype-willingness model (Gerrard, Gibbons, Houlihan, Stock, & Pomery, 2008).

Norms and expectations

Bicchieri's formulation (2006, p. 11) employs the descriptive/injunctive binary, with important modifications that echo some of the nuances identified by Rimal and colleagues, and thus go beyond just social expectations per se. Norms are behavioral rules that are “known to exist and apply to a class of situations,” that will be followed by individuals (conditional preference) if they believe others follow it (empirical expectations) or if they believe that enough others feel that the rule should be followed and are willing to apply sanctions if it is not (normative expectations). Mackie et al. (2015) note that empirical expectations may be based on what they call social proof, but also on social convention. Bicchieri likens norms to a “grammar” because norms “specify what is acceptable and what is not in a society or group” (Bicchieri, 2006; Bicchieri & Muldoon, 2011, p. 11).

Getting closer to context – norms as rooted in group frame of reference

Chung and Rimal (2016) include in their review a discussion of early work by social psychologist Muzafer Sherif (1936, 1937), who concluded that individual perceptions are “anchored around frames of reference provided by others” (Chung & Rimal, p. 4). This is an interesting formulation, bordering on what could be understood as culture, because individual judgments about and perceptions of a given situation are shaped and prioritized by the group reference frame, which, when internalized by individuals, endures beyond the actual presence of the group. This is very close to Goffman's theory of framing (1974), which in turn had a significant impact on anthropological descriptions of the connections between cognition and culture. As described below, it also touches on constructs of shared cognition and cultural models in psychological anthropology (see discussion below, and D'Andrade & Strauss, 1992; Holland, Lachicotte, Skinner, & Cain, 1998; Holland & Quinn, 1987; Shore, 1996).

Other types of norms

Recent and somewhat earlier literature also describe a range of other norm formulations, including: personal norms (Schwartz, 1977); moral norms as an individual construct separate from the social realm (Mackie et al. (2015); legal norms that are codified and have the force of the state/government behind them; and a range of other norm categories (Elster, 1989) such as consumption norms, norms against behavior “contrary to nature,” norms regulating the use of money, norms of reciprocity, norms of retribution, work norms, norms of cooperation, and norms of distribution.

Critique from within the social sciences

Bicchieri and Muldoon (2011) have criticized the treatment of social norms as akin to independent variables in social science, in which the focus is on sanctions, and where norms are viewed primarily through a functionalist lens. They also point out the difficulties in directly linking a norm to behavior because of the multiplicity of conditions which could govern the relationship. In claiming, however, that philosophy has taken a broader approach to norms than social science, their focus still remains on group-level interactions shaping, or shaped by, individual calculations and expectations, again hearkening back to classic liberal philosophy, in which individuals are the essential and formative root of society.

Bicchieri and Muldoon's review of the social norms construct (2011) concludes with an acknowledgment that many issues remain unresolved. All current models of social norms presuppose the existence of such norms and their related fabric of expectations in a population, without addressing their origins. There is also what they refer to as a “signaling” component of norms, where, for example, the norm common in the U.S. for brides to wear white “signals something,” and the extant literature on norms does not adequately address the signaling or its origins. Finally, there is not enough understanding on how to productively intervene with respect to socially harmful norms, such as norms justifying wife beating, a critique echoed in a very recent paper by Cislaghi and Heise (2018).

Social norms as used in health promotion programs

The use of social norms as a construct to promote health behavior change draws from the approaches reviewed above, though not yet incorporating some of the more recent theoretical work and – with some exceptions -- still largely focused on individual beliefs and expectations.

Norm correction and norm reinforcement approaches

Efforts in the U.S. to change social norms associated with health risks are often the focus of mass media and social marketing campaigns, for example with respect to smoking (Brownson et al., 1995; California Department of Health Services Tobacco Control Section, 1998), drinking and driving (Perkins, Linkenbach, Lewis, & Neighbors, 2010), and adolescent risk behavior (e.g., Haines, Perkins, Rice, & Barker, 2005), particularly binge drinking (Perkins, 2002; Perkins & Craig, 2003; Perkins, Haines, & Rice, 2005). In one such application, Yanovitsky and Stryker (2001) describe a norm-reinforcement approach for addressing adolescent binge drinking, defining norms as socially approved or disapproved behaviors. This intervention was based on a view that there is too much ambiguity surrounding drinking norms, often leaving individual adolescents without clear standards to which they can conform -- an assumption reflected in a paradigm underlying a significant amount of the adolescent risk prevention programming in the U.S. (e.g., Catalano et al., 2012; Hawkins, Catalano, & Miller, 1992). To address that kind of ambiguity, these types of programs take a “norms corrective” approach. This same approach has been used in small-group normative interventions as well (e.g., Reilly & Wood, 2008).

Within a social marketing framework, changing behavioral norms has also been described as a task of managing the environment surrounding a behavior or behaviors, in which multiple tools, including persuasion and sanctions (laws) are deployed (Rothschild, 1999).

Norms as expectations

In global health programs and campaigns, social norms have typically been equated with social expectations, (e.g., CARE, 2017), also common in the brief review of social science usage above. The term is seldom defined in practice settings, although that may be changing (e.g. CARE 2017, Institute for Reproductive Health, Georgetown U. and FHI 360, 2016). However, a recent USAID-supported review of gender norm changes deploys the norms construct throughout, including in such formulations as “traditional gender roles and norms” (Muralidharan et al., 2015) without any further explication. On the other hand, a very comprehensive UNICEF paper on social norms (Mackie et al., 2015) reviewed 173 publications and articles on social norms in global development, defining the construct as “what people in some group believe to be normal in the group, that is, believe to be a typical action, an appropriate action, or both.” In addition, the salience of expectations is again highlighted – “a social norm is held in place by the reciprocal expectations of the people within a reference group.” (Mackie et al., 2015, p. 7, citing; Paluck & Ball, 2010). As described in the UNICEF review, use of the construct in global health and development draws most prominently from the Reasoned Action Approach, in which individual behavioral intention is held to be an outcome of attitudes towards a behavior, perceived social norms (what an individual perceives are the behavioral expectations of her/his peer group), and perceived behavioral control (Fishbein & Ajzen, 2010).

Getting closer to context -- norms and the social environment

There are examples of social norms used in global health promotion interventions that are exceptions to, or at least modifications of, the pattern described thus far, and move closer to an anthropological view. California Department of Health Services Tobacco Control Section’s (1998) definition does refer to the broader social-structural context of norms. The USAID-funded Passages project (www.irh.org/projects/passages), addressing adolescent reproductive health, uses the term “normative environment” and specifically focuses on activities to change community-level norms, emphasizing the social aspect and distinguishing them from individual expectations (Passages Project, Feb./March 2016).The Grandmother Project/Change through Culture focuses on the cultural role of grandmothers in transmitting traditions and norms to girls as well as other family members (Newman, 2017), and links norms to important social processes and relationships that are beyond individual expectations and perceptions.

Table 1 below summarizes some of the most common definitions of social norms in the social science/social psychology literature and in program practice.

Table 1.

Definitions of social norms from social sciences, social psychology and social/behavior change program practice.

Discipline/Field of Use	How Defined	Selected References
Game theory	Practices derived from the rational choices of self-interested individuals within social systems; practices that maximize efficient attainment of social goals	Binmore, 1994; Schelling, 1960; Schotter, 1981; Sugden, 1986; Ullman-Margalit, 1977; Arrow, 1971; Akerlof, 1976
Sociology	Outcomes of socialization processes, links to group identity, conformity, and social control	Parsons, 1951, 1937; Bicchieri & Muldoon, 2011,Tajfel & Turner, 2004; Turner et al., 1987; Cialdini, 2007; Innes, 2003; Janowitz, 1975
Social network theory	Beliefs, shared expectations about what is acceptable within a social network	Leahey et al., 2015, citing Rosenquist et al., 2010; Christakis & Fowler, 2008; Christakis & Fowler, 2007
Recent formulations from social psychology	Bifurcation into descriptive and injunctive norms	Cialdini and colleagues (Cialdini et al., 1991; Cialdini et al., 2006
Recent formulations from communications for behavior change	Addition of collective norms to the descriptive/injunctive bifurcation; identification of norm moderators; norm salience; norms as social grammar; norms as linked to frame of reference	Chung & Rimal, 2016; Rimal & Lapinski, 2015; Bicchieri, 2006; Rimal & Real, 2003; Lapinski & Rimal, 2005.
Norms and social/behavior change - U.S.	Socially approved or disapproved behaviors; clear behavioral standards, voluntary behaviors shaped by environmental influences	Brownson et al., 1995; Perkins et al., 2010; Perkins et al., 2005; Perkins & Craig, 2003; Perkins, 2002; Yanovitsky & Stryker, 2001; Hawkins et al., 1992; Catalano et al., 2012, Rothschild, 1999
Norms and social/behavior change - global	Social expectations about others' behavior, individual beliefs about typical and/or appropriate behavior within a reference group	CARE, 2017; Mackie et al., 2015; Fishbein & Ajzen, 2010; Institute for Reproductive Health, Georgetown University, & FHI 360, 2016; Newman, 2017

The anthropological perspective: social norms as linked to culture, shared meaning, and shared structure

At least some of the conceptual gaps addressed by both Bicchieri and Muldoon, Cislaghi and Heise as well as in our own assessment exist because the social norms construct has not been adequately considered as a culturally embedded phenomenon – that is, a phenomenon that cannot be fully understood apart from the systems of interpretation, meaning, and acquisition through socialization that infuse social actions with motivational and affective force as well as representational content (as in representing values or incorporating symbols), together with the structures of power that are associated with those systems. A social norm that is merely a utilitarian calculation (see critique in Bourdieu, 1977 and in Ensminger & Knight, 1997), or held in place by the force of sanction or shared expectation, is a thin construct indeed, devoid of social depth or the “thickness” that animates the behaviors of human beings in society. From our perspective, people do not just behave based on calculations of group dynamics, or of expected responses. The mechanistic character of such formulations elides the multivalent dimensions of actual behavior and does not adequately facilitate an understanding of the processes of social/cultural change -- and may, in fact, impede an understanding of such change, including an understanding of normative change.

Moreover, with respect to norms as a group phenomenon, the notion of “reference group” appearing in much of this literature is problematic. Certainly most individuals participate in more than one reference group or network – whether a work group, religious group, activity group, club, age cohort, professional society, secret society or any other. Each of these groups will serve, to some extent, as a reference group with social norms unique to the group. Overlapping settings raise the question of which reference group is salient, and what norms (if any) pertain across groups as broadly cultural or societal, a situation that may be obscured by a focus on reference groups. Furthermore, norms may conflict with each other in specific settings. This may be crucial for the conscious manipulation of norms described later in this paper, allowing individuals to work with and around and not just under normative structures. Some of these issues have in fact been raised in the sociological and social-psychological literature on norms (e.g., Hechter & Opp, 2001; Mollborn, 2017), and by Chung and Rimal (2016) and Rimal and Lapinski (2015).

Norms, we argue, are also not analogous to a grammar (as argued by Bicchieri), which – in the tradition of anthropological linguistics – is more than a set of rules and expectations. Grammars are also a means of encoding a multiplicity of interpretive guidelines about the world, including, for example, hierarchy (via forms of address, formal vs. familiar grammatical constructions), agency (subject vs. object rules), time (rules for conjugation of verbs), and so on (Bauman & Sherzer, 1989; Hymes, 1972, 1974). Moreover, wearing a white wedding dress, as in Bicchieri's example, may be a signifier in some cultures because it is a complex social metaphor that encompasses and represents cultural models that could include deeply held beliefs about the course of life and the chronology and meaning of life stages, gender roles (e.g., purity for females before marriage), and a mélange of narratives and scripts for how the movement between life stages and roles should be marked in a social ritual (Turner, 1974; Van Gennep, 1909/1960) that can be understood or “read” by participants in the ritual. The so-called signaling function in this case is equivalent to a cultural, symbolic function.

Moreover, while Rimal and Lapinski (2015) and Chung and Rimal (2016) laudably explore definitional differences between norms and laws, traditions, folkways, and mores, these are difficult distinctions to make. Laws are usually norms that are or were so important at the time that lawmakers were considering the issue, that they became institutionalized and formalized as legal code. Furthermore, laws are usually socially and politically negotiated, albeit through formal processes, and the temporal difference between a tradition and a norm is hard to pin down – indeed, norms must also have temporal continuity in order that people recognize them and hold them as expectations. Framing the term “folkways” as an early definition of (descriptive) norms may also be confusing, as that term can also be seen as synonymous with traditions. Lastly, if “mores” are seen as a broader social version of injunctive norms because social sanctions result from a violation, it is not clear how the construct of mores in this sense differs from values.

Nine tenets of an anthropological approach to norms

Treating social norms as culturally embedded and contextualized follows from a general understanding that no construct of social behavior can be understood as a phenomenon-in-itself. Because some form of holism and integration is intrinsic to notions of culture and its ideational as well as practical manifestations, everything is embedded, although to different degrees. Moreover, because the notion of culture itself is complex and includes multiple foci, a culturally embedded perspective on norms generally includes the potential dimensions outlined below. The use of “potential” here is deliberate. We do not argue that all social norms incorporate all of these dimensions, just that they may, and therefore a valid, sustainable use of the construct for behavior change interventions should at least consider the following nine possibilities.

• 1.Social norms are instantiations of meanings and values within a broader cultural complex

Norms are often linked to cultural values. Monica Heintz (2009) frames norms as “implemented values.” In this view, it is inappropriate, and insufficient to assess norms just in terms of data on prevalence or adherence. “One does not steal candy from a child” as a norm is meaningful because of its implied values (e.g., children are vulnerable, it is bad to exercise power over the vulnerable). In addition, values and their normative practices are part of the cultural material that is transferred across generations through socialization processes. The link between norms and values, however, is neither fixed nor linear. Because norms represent social practice, they may persist even as values change – for example, in the U.S., values related to gender and household tasks have changed, but in practice women still perform most of the childcare and domestic labor. Or, a value may persist but the norm(s) for implementing it may change.

• 2.Norms can be viewed as social practice within a “web of meaning”

Culture construed as a “web of meaning” (Geertz, 1973) implies that social practice is symbolic, representing something. Norms in this sense are like scripted action in a dramatic production that expresses an underlying “plot.” Linguistic norms, for example, may thus represent complicity with social hierarchies or instrumental tools to manipulate hierarchies; as such they are an intentional performance. Importantly, symbolic representations are public actions; thus they also communicate, and reinforce acceptance of the norm and what it represents. We could say that norms are therefore akin to cultural scripts, which are linguistic and behavioral patterns of speech or behavior that are commonly recognized as representing broader semantic constructs – e.g., a script for how to be “polite” in a certain circumstance (Wierzbicka, 1994). Norms as meaningful representations may also play out as performance of specific social roles, as in the regularized behavior patterns associated with gender roles or in enacting a cultural persona, as described by Edberg with respect to narcotraffickers (Edberg, 2004; Edberg & Bourgois, 2013; Edberg, in press).

• 3Norms can be linked to power configurations

As behavioral rules or codes, norms can enact and maintain structures of power. For example, it was once normative to bow or prostrate oneself before a king or emperor. Clearly, this norm is an enactment of a power differential. It was and still is a linguistic norm in some cultures that female speech should be quieter, less direct, and less aggressive than male speech (e.g., Lakoff, 1975; Tannen, 1996 on these characteristics in some American cultures). This is a norm clearly linked to gender hierarchy. From this perspective, Heise (1998; 2011) has also linked gender hierarchy to norms facilitating violence victimization and HIV risk. Norms are, in a Foucaultian sense, a form of discourse, implicating power.

Within a given configuration of power, norms may be linked to specific strata within that configuration. Pierre Bourdieu's construct of the habitus (1977) is useful here -- referring to clusters of common behavior patterns, habits, expectations, and cultural capital (of a group or class) associated with position in a social hierarchy. The concept of intersectionality (Crenshaw, 1991) is also pertinent as it situates norms within multiple social identity constructs, each of which may be subject to practices of power. For example, in the United States, a person who is of African descent, a woman, and an immigrant has a complex identity that collides with multiple locations within a power hierarchy, involving norms that both express and reinforce that position. The connection between norms and power also draws from multiple sources in anthropology and social theory, including Weber (1968), Foucault (1980, 1979, 1976), and Bourdieu (1977). Along the same lines, norms can be viewed as ideological tools (a la Gramsci, 1971), where conformance to a norm, and beliefs associated with conformance, are social control mechanisms that serve to impose and maintain cultural hegemony. After all, it is most useful to the king if subjects bow to him (as a norm) and believe in his divine right to be king. At the same time, the beliefs associated with hegemonic norms may also be held by those at the top of the hierarchy – the king may also believe he is divinely ordained.

Finally, we must ask a broader question about power and the relationship of public health efforts to utilize social norms as a change construct in a global context. What structural hierarchy is promulgated when largely Western-funded institutions and NGOs are developing and implementing such efforts?

• 4.From a cognitive perspective, norms can be viewed as elements of cultural models

Cultural models are shared mental schemas or models about some social phenomenon (D'Andrade & Strauss, 1992; Holland et al., 1998; Holland & Quinn, 1987; Shore, 1996). As such they provide interpretational guidelines typically connected to behavioral scripts. As an example, we have a cultural model of a “classroom.” In order to understand something as a classroom, we employ expectations/norms related to appropriate behavior, dress, social relationships, etc. – as information processing tools. These tools guide interpretation (recognition of the situation as a classroom), motivation, and behavior. The meaning inherent in the model of a classroom (e.g., its place in life trajectories, its connection to social status) can in turn spur compliance with (or disruption of) classroom norms. Moreover, the meanings and cultural models are themselves shaped by social and structural context – “classroom” and what it means vis a vis a life trajectory will mean something very different to those who have limited access to classrooms and the life trajectories with which they are associated.

Cultural models (and their associated behavioral norms), however, may be known and shared to different degrees among a given group or culture. To address this variance, Kimball Romney, William Dressler, and other anthropologists introduced and have applied the constructs of cultural consensus, cultural competence and cultural consonance (Dressler, 2018; Romney, Weller, & Batchelder, 1986). These three constructs roughly translate into the degree to which a group shares a particular cultural model (consensus), the degree to which individuals within the group have thorough knowledge about the model (competence), and the degree to which individual behaviors or practices actually conform to the model, even if it is shared (consonance). The health implications of cultural models and their associated variance have been demonstrated in research by Dressler and colleagues (Dressler & Bindon, 2000; Dressler, Balieiro, & Dos Santos, 1999) showing correlations, for example, between high blood pressure and low cultural consonance – that is, when there is a significant variance between culturally shared models of a desirable lifestyle and a given individual's actual lifestyle There is some overlap here between these cultural models constructs and Rimal & Lapinski (2005) collective vs. perceived norm dichotomy.

• 5Norms are cultural tools for the exercise of individual agency

There is an important aspect of individual agency implied in normative behavior that is not sufficiently addressed in the social science and public health literatures. Because norms can be considered as behavioral codes, rules, common patterns, ways of legitimating behavior, or valued patterns that are known to one degree or another among individuals in a cultural group, they can also be deployed by individuals to advance personal or social goals and to communicate messages. This is possible precisely because norms and their cultural meaning are commonly known – thus, intentional manipulation or violation of a norm is a communicative and social tool just as is normative conformance. This is also possible because there are a plethora of norms that may be applied to a given social situation, some of which will likely conflict with each other. Norms may be consciously manipulated by those in authority (e.g., political figures or those who hold power in a patriarchy) or those without authority but seeking power. Normative conflict can be used strategically to increase both the range of behavior that may be legitimated and therefore acceptable as well as to widen the social expectations for a given social situation that are produced by social norms. This view concords with that of Barth (1981), who framed social norms as agentic in the sense that individuals “engage in strategic decision making within the constraints presented by social norms and other institutions” (I98I: 30–31, quoted in Ensminger & Knight, 1997, p. 2). Krieger (1984) also described norms among Egyptian traditional urban women, not just as constraints, but as political assets that could be manipulated by taking advantage of normative conflict to gain power, to stretch what was legitimately seen as accepted behavior, and in so doing to craft self- or family member presentations that appeared to be within normatively acceptable boundaries, even if the behavior actually violated a norm. Moreover, people seeking to change culture may strategically and publicly seek to violate certain norms -- e.g., feminists in the 1960s–70s who rushed through doors so they could hold them open for men.

• 6Norms often refer to public vs. private behavior

Compliance with norms, perceived norms, and sanctions related to normative violation – key dimensions of the norms construct in the literature – typically refer to public behavior. Yet non-normative behavior may be quite common, though generally characterized by secrecy (Krieger, 1984; 2011) -- people try to ensure that their non-normative behavior does not become known to others, with perhaps the exception of very close family members or friends. Non-normative behavior often has public health implications, for example with respect to injection drug use or stigmatized sexual relationships. Even this distinction, however, must be qualified. Non-normative, private behavior may of course be normative within small subgroups, and for public health interventions that is important information.

• 7Norms may occur as reciprocal structures

Unidirectional assumptions that norms produce behavior ignore the many norms that are reciprocal, e.g., husbands are responsible for their wives’ behavior; wives should obey their husbands. As mentioned earlier, norms are often part of societal authority structures and have wide-ranging ramifications. For example, norms that justify and specify behavior of parents to children and children to parents are inherent in the jural authority of parents. Attempting to change just one of the norms may be either difficult or nonsensical if one is not aware of their connectivity, and of the potential cultural, social, or even legal consequences.

• 8Norms take multiple forms

The substance of what should be considered as a norm also varies. In addition to a dichotomy between ideal vs. behavioral norms, there is a distinction between formal vs. informal: Norms can refer to standards that are codified (e.g., “rules of conduct” for an organization, or at the most formal level as laws), or to standards that are widely understood or shared, as informal but “common knowledge” (an empirical question). Fried's summary here (Fried, 1953), though dated, is also useful as a reference: “Ideal norms are the moral evaluation of what is proper behavior,” what a group of people “may feel ought to be done [emphasis original],” not to be confused with observations of what most people seem to be doing. “Careful daily documentation of actual behavior will show that there are, beside the ideal norms mentioned by informants, usually alternative patterns or acceptable substitutes, though these may not be so clearly formulated by them as the ideal norms” (p. 286). Normative expectations, he argued, are in part based on ideal norms; however, individuals within a cultural group are aware that actual behavior does not always mirror ideal norms. Variation is in fact normative: “Part of the expectancies that people have of others involve an understanding, and tolerance of, some degree of variation” (Fried, 1953, p. 287).

Within these multiple forms, there is also a potential gender issue. Males and females may have different sets of ideal and behavioral norms – a situation that can produce discord within intimate relationships, among other consequences.

• 9There is both continuity and change in social norms

Finally, apart from any social/behavior change intervention, norms are not static and typically change over time, for multiple reasons. As Ensminger and Knight (1997) wrote, “The role of social norms is central to our understanding of how communities maintain and transmit dominant social practice” (p. 2, citing Ortner, 1984). The question of how social norms originate and either persist or change, then, is inseparable from how cultures develop and change – a much broader question. As part of that process, the relationship between any given norm and the deeper structure of power, meaning or social organization that it represents is fluid. That underlying structure may change, yet the norm remains because it has been ingrained as habit; or the norm changes, but the deeper structure does not. Moreover, if norms are equated to traditions, it must be understood that traditions inherently change, typically at any given point combining the old and the new (Adra, 2016). Tradition, as part of culture, is always emergent, and contingent, with potential consequences for social norms.

Moreover, understanding that norms may represent underlying values or power-structures, variance may exist between underlying values and current norms, such that programs to change norms may miss the mark entirely if those norms have shifted and no longer serve to instantiate a particular cluster of values or a structural node. Or, when behavior has shifted but norms and values have not yet caught up (cultural lag), which may be common in times of rapid social change, especially when spurred by technology -- as first proposed by Ogburn (1922) but still a salient concept (see Yoshida, 2010).

Table 2 below summarizes important ways, as described above, in which social norms have been characterized in the anthropological and related literature.

Table 2.

Dimensions and definitions of social norms as culturally embedded.

Culturally Embedded Dimensions of Norms – How Defined	Selected References
Instantiations of, or performance of cultural values	Monica Heintz (2009)
Behaviors linked to power configurations, as practices that represent and perpetuate power and its associated ideologies	Weber (1968), Foucault (1980, 1979, 1976), Bourdieu (1977),Lakoff, 1975; Tannen, 1996; Heise (1998; 2011); Goffman, 1974
Social practices within a “web of meaning,” as coded behaviors or “scripts” connected to broader context of meaning and performance of social roles/personas	Geertz, 1973; Wierzbicka, 1994; Edberg, 2004; Edberg, in press
Elements of (cognitive) cultural models, as practices flowing from/representing shared, cognitive cultural models about something	Shore, 1996; Holland et al., 1998; Holland & Quinn, 1987; D'Andrade & Strauss, 1992; Romney et al., 1986; Dressler, 2018
Cultural tools for the exercise of individual agency – where individuals can manipulate coexisting or competing rules and categories to advance personal goals	Barth, 1981; Ensminger & Knight, 1997; Krieger, 1984
Often public vs. private behavior – important are the settings in which specific norms are operative	This article.
Norms take multiple forms, as “sets” of potential behavioral rules that may apply (as alternatives). Ideal norms, as expressions of cultural ideology/values, may or may not result in behavior; behavioral norms express rules for what people usually do and guide behavior	Fried, 1953; Krieger, 1984; Nanda & Warms, 1980/2013

The idea of changing norms as a health promotion strategy

As can be seen, the meaning and practice of the norms construct is more complex than its typical usage for social/behavior change suggests. The anthropological perspectives presented above offer added dimensions to understanding norms and their relation to behavior, society, and culture, and in fact synthesize the way norms are treated within anthropological literature, even when a given work is not specifically about norms, per se. Consequently, the idea of “changing norms” as a go-to social/behavioral change strategy to improve health outcomes is much more problematic than it may seem. According to Mackie et al. (2015), to change the social norm of corporal punishment, for example, “one would seek to change the normative and empirical expectations among enough members of the community” (Mackie et al. (2015), P. 33). Given the issues we have raised in this paper, that approach may not be a straightforward task, for multiple reasons. A particular norm targeted for change is likely to be integrated with others as part of a behavioral/cognitive/emotional complex that is connected to cultural and structural nodes.

From the cultural embeddedness perspective, focusing on the social norm itself would not be enough. Corporal punishment as a behavior, for example, also represents something important to the population or society that engages in this practice. It is highly likely that corporal punishment is connected to patterns of authority, personhood, and the intergenerational continuity of those culturally significant patterns, and also potentially connected at a deeper level to notions about the function of the state, religious authority, and to the teleology of childrearing. What kind of child is a good parent (in society x) supposed to raise? What should be their qualities? Essentially, this is also a question of what kind of adult is the process of childrearing supposed to produce. Furthermore, this is tied to parents’ expectations for what their (changing) society will be like in the future, when their children reach adulthood. In that sense, corporal punishment is much more than a social pattern or expectation. It is a structural replicator and a meaningful representation, a signifier. It is connected to structural and cultural nodes and to an assessment of the future.

In a recent project for UNICEF assessing programs to prevent corporal punishment in Papua province, Indonesia (Edberg et al., 2016), data indicated that the programmatic attempts to reduce/prevent corporal punishment bumped up against just this sort of cultural underpinning. In fact, the most common aphorism heard was “there is gold at the end of a rattan stick,” where rattan sticks were commonly used to physically punish children. Resistance to changing these social norms was often based on what people saw as the ruination of children, as leading to badly-raised children with no discipline and no respect. In effect, the prevention efforts were perceived as asking parents to accept raising children who were not good, who did not conform to the cultural model or ideal. Even if a strategy was devised that affected individual expectations about public practice of corporal punishment, it is very possible, if not likely, that the deeper imperative to produce disciplined children would result in new norm variations that might have their own negative consequences or simply the continued practice of corporal punishment in private vs. public settings.

Similarly, research as part of a family planning project in Albania (undertaken by the second author) uncovered a stated norm that “real” men exercise self-control and are responsible. This was linked to adult male identity and the construct of manhood, and was expressed in the almost universal use of withdrawal as the family planning method of choice, with abortion as a backup method. Recognizing the difficulty in directly challenging a deeply-embedded norm, the project team chose not to make that challenge, but instead to reframe the issue so that an entirely different set of norms applied that could support healthier behavioral options. The unreliability of withdrawal was recast as an outcome of biological, scientific factors not within a man's control, thus taking withdrawal out of the purview of gender norms surrounding control and responsibility and linking it to norms regarding knowledge and education.

The problem of changing norms that are seen as harmful

Social and behavior change programs typically aim to change norms that, from a public health or even broader values perspective, are seen as harmful. This too is complex. One must first identify whom the norms purportedly harm, what those norms represent, whether the underlying value or structure can be changed, or whether (as we advocate) there are internal or at least collaborative processes available to facilitate change. What is the ongoing relationship between norms and values? A static link? Does norm x always represent value x? What competing norms exist? And, how does an external party best address the normative representation of ideologies (such as gender dominance) that are themselves perceived as harmful by broader global standards but are core elements of a particular culture? Furthermore, there may be multiple practices related to the same norms, as representations of particular cultural nodes. Is it possible or even wise to try to change one of the behaviors? For example, a key aspect of social and symbolic organization and authority patterns in traditional urban Egyptian culture is the imperative to maintain the difference between male and female gender constructs. Female genital cutting, locally conceptualized as circumcision, exacerbates existing biological differences between the two gender roles by removing the clitoris, viewed as analogous to a male penis (Krieger, 1984). Many other practices (e.g., childrearing, gendered clothing and hair styles) also serve the same purpose of enhancing a prescribed sex and gender dimorphism – suggesting that a given practice may be amenable to change, since the same underlying structures and ideology are expressed in other ways (i.e., there is a sufficiency of representational alternatives). On the other hand, the ideology may be so powerfully enmeshed in practice that no one desires a change. Even female relatives have been shown to play an important role in forming and sustaining the masculinities of male relatives via constant socialization to masculinity norms (Ghannam, 2013). All of these practices, which could be seen as normative, support the binary gender model that is a key value underlying Egyptian cultural ideology.

How can the construct of social norms be used with increased validity for health promotion interventions and research?

There are varying degrees to which it may be necessary to address the cultural embeddedness of norms, depending upon project aims and resources. A central goal of this paper, however, is to elucidate the relationship between norms and culture in order to emphasize that this should be considered before applying the construct for health promotion. The literature and program materials we have reviewed, as well as our own experience, do not indicate that it has.

The following are some thoughts and guidelines for consideration when using norms as a mechanism for changing behavior. It is our recommendation that interventions intending to link norm changes to behavior change and improved health outcomes need to assess, through basic or formative research, the embedded nature of the social norms in question and the degree to which the health outcome is a direct consequence of just one or more selected norms, as opposed to an outcome of multiple, connected norms that share cultural and/or structural content. Understanding cultural embeddedness calls for answers to a number of questions derived from the nine possibilities (or “tenets”) outlined above, including the following:

• •What deeper values, ideologies or social roles might the norm or norms represent? What is said/understood about people who conform or don't conform to the norm? What sort of people do “norm-conformers” and “norm non-conformers” represent as characters in a scripted social play? How are those values and associated norms transmitted across generations? [Tenets #1 and #2]
• •Is the norm an expression of a hierarchy that would persist even if the norm itself was changed and that might be expressed in equally harmful behavior? [Tenet #3]
• •Are there sanctions for norm violation and if so, in which contexts? Who imposes the sanctions for norm violation and how consistent and unpleasant are they? Are there ways to get around the sanctions? [Tenet #3]
• •What is the phenomenon to which a particular norm is attached (e.g., norms related to a wedding, norms related to the conduct of business)? What is the prevalent cultural model (within a defined group) about that phenomenon as a whole and how do specific norms connect to the model? [Tenet #4]
• •Is conformance or non-conformance to a norm useful to particular individuals or subgroups as a means of advancing some specific aim(s)? [Tenet #5]
• •Are there practices that people do in private that differ from what they do in public, and how does this articulate with publicly expressed norms? [Tenet #6]
• •Is a particular norm entangled in reciprocal structures, such as husband-wife mutual obligations? If so, can that norm even be untangled and addressed separately? [Tenet #7]
• •Are there multiple, conflicting or competing norms that relate to the same social situation and represent the same or different value(s)? How and for whom do these come into play? [Tenets #5 and 8]
• •For whom is a particular norm most salient and valued? In contrast, are there individuals or categories of people for whom the norm is not valued? [Tenets #5 and 8]
• •How does a given health-related practice connect to one or more norms, and what normative form do these take – informal, formal, ideal, situational, public, private? [Tenet #8]
• •Instead of attempting to change norms, could the practice be reframed in a different cultural or ideological context so that an entirely different set of norms apply? [Tenets #1 and 8]
• •If a particular, common (normative) practice was eliminated, for example, by a law, would there be other ways in which people would seek to pursue underlying cultural goals facilitated by the prohibited practice? What health consequences would those alternate practices have? Are there already other ways in which people are pursuing the underlying goals? [Tenet #9]
• •How much change has already occurred in a norm/norms? Over what period of time? Have these changes occurred differentially among groups? How does the norm change link to broader cultural change processes? [Tenet #9]

It is our suggestion that practitioners consider these questions in the program/campaign development or formative research phase, if the focus is on norms change. In addition, these and related questions could serve as useful research questions in themselves pertinent to a given health behavior issue as a means of expanding the information available to program developers. The goal, as noted earlier, is to improve the practical effectiveness of social/behavior change interventions and to reduce the likelihood that norms change efforts will miss important determinants of behavior.

A revised definition

Finally, based on the review of norms in this paper from an anthropological perspective, we offer the following working definition: Social norms are formed from interactions between underlying structures and rationales expressed though regularized social behaviors known to most members of a social or cultural group at a particular point in time. Norms can: 1) generate expectations in order to guide behavior and its interpretation in a range of circumstances, and among a range of groups; 2) serve as social scripts for carrying out a way of life that is characteristic of a group or culture, and are therefore linked by most members of that group/culture to that way of life (as a marker, referential to identity and meaning); 3) act as “surface” manifestations of underlying values, ideologies and social structures; and 4) operate as social tools that individuals can deploy to sustain, manipulate or contest underlying values, hierarchies and social structures. Not every social norm has all of these characteristics, but the presence of any of these aspects of embeddedness may confound social norms interventions that do not consider them.

Author statement

Both authors for this publication have no competing interests with respect to the subject or material presented in this article.

Ethical statement

This paper is a review and commentary. Any research results or conclusions from research conducted by the authors are derived from research approved by the ethical review processes of the funding institutions.