Cultural Models of Contention: How Do the Public Interpret the Repertoire of Contention?

Abstract

The literature on contention tends to conflate contentious actions and audience’s interpretation of those actions. This is problematic because interpretation is central to how contention unfolds and brings about social change. We theorise that interpretation is patterned by one or more cultural models of contention. These provide background assumptions about what actions count as political and what actions are legitimate. We show the fruitfulness of our approach in two survey studies of 1429 US citizens. It allows us to explore patterns in how the US public interpret contention. Furthermore, we investigate how interpretation varies across political and apolitical contexts, finding little variation between these. Finally, we study heterogeneity in how the public interpret contention, finding variation between individuals but also shared patterns. The article contributes to the literature on contention by providing a theoretical framework to study the public’s interpretation of contention and a fine-grained empirical analysis of this interpretation.

Keywords

contentious politics culture protests public repertoire of contention

Contentious politics is a central driver of change in liberal democracies. Actors use tactics such as public marching, pamphleteering and property destruction to put forward claims that challenge others or authorities (Tarrow, 2011). These tactics are often described through the metaphor of a ‘repertoire of contention’, a stock of tactics used by contentious actors and recognised by the audience to whom they are directed (Tilly, 1995, 2008). While the literature has mapped the use of these tactics, it has paid little attention to how they are interpreted by the audience (Wouters, 2019). The literature remains ‘movement-centric’ (McAdam and Boudet, 2012: 3). This is problematic because interpretation is central to how contention unfolds and brings about change.

The ‘Black Lives Matter’ contention sparked by the police murder of George Floyd exemplified this problem. People marched in the streets and held public meetings but also looted, destroyed property and attacked the police (Francis and Wright-Rigueur, 2021). The interpretation of the latter actions was deeply contested in the public (Drakulich and Denver, 2022; Shuman et al., 2022). They could be seen as political tactics that challenged the racialised exclusion of the protesters. They could also be seen as apolitical outbursts of social disorder (Boonen, 2020). A gap opened up between the actions of the rioters and their interpretations by the public. This put the rioters’ ability to promote social change into question (Gøtzsche-Astrup, 2022).

The distinction between the actions and their interpretation holds for contention in general because meaning does not inhere in actions but is ascribed to these through interpretation (Taylor, 1971). We must theoretically and analytically distinguish contentious actions from their interpretation. Several traditions have touched on the interpretation of contentious actions without tackling it directly. The literatures on framing and discursive opportunity structures analyse the interpretation of contention but focus on demands and not actions (e.g. Benford and Snow, 2000; Koopmans and Olzak, 2004; Snow et al., 2019). The literature on the media coverage of protest (e.g. Andrews and Caren, 2010; Koopmans, 2004; Martin et al., 2017; Seguin, 2015) has tackled how the traditional media frame protests and the challenges and opportunities posed by new media forms (Cottle, 2008). However, although the media influence the public’s interpretation of contention, we cannot infer interpretation from media coverage (Wouters, 2019). There has recently been a revival of interest in public perceptions of protest. This has focused on the perceived legitimacy and favourability of protest forms and movements (e.g. Andrews et al., 2016; Park and Einwohner, 2019; Wouters, 2019), missing other dimensions of interpretation. Furthermore, it has not clearly theorised how the interpretation of contentious actions is patterned in the public.

We theorise and analyse the public’s interpretation of contention. Drawing on the sociology of culture and cognition (Lizardo, 2017; Rotolo, 2021) and studies of the public’s perception of contention, we theorise that the interpretation of contention is patterned by tacit assumptions about contention, which congeal into one or more cultural models. These can be studied through the dimensions of legitimacy and legibility. Legitimacy concerns whether a tactic is seen as acceptable whereas legibility describes whether a tactic is recognisable as a political protest.

We use our approach to study how the US public, conceptualised as the general audience of citizens (Wouters, 2019), interpret contention. In two survey studies, one of which is population representative for age, gender, race and education, we explore the interpretation of contention in the dimensions of legitimacy and legibility. We ask respondents to imagine that a person has been shot and have them consider whether a series of actions reacting to this shooting are legitimate and legible. We find a meaningful pattern in these dimensions. Additionally, we investigate whether the legitimacy and legibility of the actions depend on their context, which we do not find strong support for. Furthermore, we use our dimensions to explore heterogeneity in the public’s interpretation. We find substantial variation in the interpretation of some tactics and widely shared interpretations of others.

The article makes two major contributions to the literature on contention. It provides a theoretical underpinning to the literature on the public perception of protest. It also presents a fine-grained empirical analysis of this interpretation in the United States. We offer one perspective on how social change is enabled and constrained by the public’s interpretation of a central democratic practice.

Patterns of Action and Interpretation in the Extant Literature

We are interested in how the public’s interpretation of contentious actions is patterned. While the literature on the public perception of protest has not engaged directly with this question, it has partly been dealt with under the heading of repertoires of contention. The repertoire describes what actions can be used and recognised as contentious in a given time and place. We first review the work on repertoires briefly, focusing on the meaning of actions. We argue that we should distinguish more sharply between the meaning that the actions have for the contentious actors themselves and how the public interpret these actions.

Analysing the repertoire means tracing the evolution of individual forms of contentious actions (e.g. Biggs, 2013; McCurdy et al., 2016; Traugott, 1993) and of repertoires as such (e.g. Tilly, 1995; Wada, 2012). These analyses assume that the repertoire of contention is meaningful for the actors that draw on it (Larson and Lizardo, 2019; Ring-Ramirez et al., 2014; Tarrow, 2011). Cultural scholars have added to this approach by analysing how actors invest tactics with different meanings (Doherty and Hayes, 2019; Gillan, 2020; Goodwin and Jasper, 1999). While these analyses are clearly interested in the actions’ meaning, they have focused on the contentious actors themselves and not the public. They remain ‘movement-centric’ (McAdam and Boudet, 2012: 3).

The Independent Role of Interpretation

Acting contentiously is a form of communicative action (Larson, 2013), often directed at an audience (Andrews et al., 2016; Wouters, 2019), which may ultimately be the public. To be understood, actors articulate their challenges through tactics that the audience can decode. The necessity of interpretation is implicitly acknowledged in the repertoire literature. Tarrow (2011: 39) notes that repertoires consist of what people do, know how to do and what others expect them to do. While the first two statements concern actions, the final one is about their interpretation. In effect, the literature assumes the existence of a pattern in the public’s interpretation of contention.

In making this assumption explicit, we argue that we need a new analytical perspective on the interpretation of contentious actions. Otherwise, we risk conflating the patterns of interpretation and contentious actions. In Tarrow’s terms, what others expect people to do is sometimes inferred from what people do and know how to do. This makes it difficult to deal with cases of misunderstanding and failed communication, as in the case of riots. Moreover, it obscures a fruitful line of research: how do the public distinguish contentious actions from each other? Why are certain actions seen as acceptable protests while others are not? In short, how do the public make sense of contention?

Cultural Models of Contention

We draw on recent developments in the sociology of culture and cognition to tackle these questions (e.g. Lizardo, 2017; Rotolo, 2021; Vaisey, 2009). We argue that the public’s interpretation of contention is structured by one or more cultural models of contention. This approach foregrounds the tacit and transposable assumptions about contentious actions that are widely shared in the public. We investigate these assumptions through the dimensions of legitimacy and legibility.

The Notion of a Cultural Model

Cultural models belong to the domain of personal culture, which is internalised in individuals and not externalised through public symbols, discourses or institutions. These models contain tacit assumptions about the world and specific domains in it, such as contention (Rotolo, 2021). The assumptions are transposable across contexts as they can be used to interpret a wide range of contentious actions, producing patterns in interpretation. We note that the concept of a cultural model incorporates both those assumptions that make up underlying schemas (non-declarative personal culture) and attitudes more directly accessible to the individual (declarative personal culture) (see Boutyline and Soter, 2021; Lizardo, 2017). We reflect on how to untangle these in the discussion.

Dimensions of Cultural Models of Contention

In speaking of cultural models of contention, we point to tacit assumptions about contentious actions that are widely shared, although not monolithic, in the US public. We follow Wang and Soule (2016) in seeing actions and tactics as synonymous and distinguish them from higher-level ‘forms’ such as rioting and demonstrating. While forms are important (Tilly, 2008), we focus on actions because they are the concrete manifestations of contention that the public interpret (see Hall et al., 1986).

We suggest studying the interpretation of contention in the two dimensions of legitimacy and legibility. The dimensions do not exhaust the complexity of interpretation. Rather, they capture aspects of cultural models whose practical logic is fuzzier than the dimensions suggest (Bourdieu, 1990; Lizardo, 2010). However, we expect the dimensions to capture some fundamental aspects of cultural models of contention. Faced with a contentious event, an observer makes two fundamental interpretations: whether the actions are acceptable and whether they are political. We construct these dimensions based on previous studies of public perceptions of contention. These have focused on legitimacy and hinted at the dimension of legibility. We also draw on sociological theories about communication, which distinguish between accepting and understanding communication (e.g. Luhmann, 1995: 147).

First, contentious actions can be seen as more or less legitimate, which is the focus of the literature on public perceptions of contention (e.g. Andrews et al., 2016; Hall et al., 1986; Olsen, 1968; Park and Einwohner, 2019). By legitimacy, we mean actions that are acceptable to carry out in society. When individuals orient themselves socially, they do so partly through moral boundaries between what is acceptable and what is not (Lamont and Molnár, 2002). This is also the case in the realm of politics (Tarrow, 2011). As Alexander (2006) has argued, the distinction is not merely a legal rule but a part of the public culture. We argue that it is also a part of the personal culture of individuals. This focus on acceptability foregrounds the normative aspects of legitimacy instead of legality (e.g. Palthe, 2014; Scott, 1995). Normative aspects are more likely to be internalised by members of the public (Scott, 1995: 74), driving their interpretation of contentious actions.

Second, an action can be more or less legible as political. This makes it possible to speak of a repertoire of contention, of a set of tactics used to put forward claims that challenge others. Turner (1969) made a similar point in arguing that actions must correspond with a ‘folk concept’ of protest to be classified as such. Although the literature has not neglected the dimension of legibility (Altheide and Gilmore, 1972; Jeffries et al., 1971; Tarrow, 1993: 71), it has rarely been analysed in its own right. Like Turner, we assume that individuals have a concept of what a political protest is, which informs how they interpret contentious actions. In other words, individuals can evaluate whether an action is political or not. We do not make any assumptions about what this concept consists of. This is ultimately an empirical question, which we raise in our discussion. As the debate on the political or apolitical nature of the recent Black Lives Matter riots suggests, this dimension is central to the public’s interpretation of contention. Looting and property destruction may not only be less legitimate but less legible compared with picketing or pamphleteering.

Our theoretical argument can be summarised in three points. First, extant approaches to the patterns of contention tend to marginalise interpretation. Second, the interpretation of contention is patterned by one or more cultural models of contention. Third, we can analyse the public’s interpretation of contention through the dimensions of legibility and legitimacy.

These points result in three empirical expectations, which would substantiate the fruitfulness of our model. First, we should observe meaningful and significant differences between the two dimensions. Results that would speak against our approach would be a lack of variation in how different tactics are evaluated, a lack of differences between dimensions or a meaningless pattern that did not resonate with the existing literature. Second, because the tacit assumptions about contention are transposable, we expect the interpretation of tactics to be relatively stable across contexts. Third, since there may be more than one cultural model in the public, we can investigate heterogeneity in interpretation on both observed (manifest) and unobserved (latent) categories.

Empirical Strategy

We investigate these expectations in two original survey studies of the US population. We use panel participants on the crowdsourcing marketplace MTurk and the participant-sourcing company Cloudresearch. This reaches our target population of ordinary individuals not particularly interested in politics or protest (Robinson et al., 2019).

Measuring Cultural Models of Contention

In studying cultural models, we need to decide what tactics to include and how to distinguish between the tactics’ legibility and legitimacy.

To build a repertoire of tactics, we started with the Dynamics of Collective Action (DOC) dataset, which contains about 23,000 New York Times articles covering protest events in the United States between 1960 and 1995. As the events are reported in the news media, it represents how most people would encounter a specific protest event (Koopmans, 2004). The events have been coded into the level of tactics, which is our focus (Wang and Soule, 2016).

In deciding what tactics to include, we adapted Wang and Soule’s (2016: 542) list of 57 tactics. We did not strive for a comprehensive list but a breadth in the included tactics. We combined some tactics (e.g. tactic 10 ‘Film showing’ and tactic 16 ‘Photo exhibiting’ into the tactic ‘Hosting an art exhibition’) and removed a few others that referred to specific historical issues (e.g. ‘Bed racing’). As the dataset only included tactics up until 1995, we included a few modern tactics (e.g. ‘Posting opinions on social media’). We ended up with the 41 tactics in Table 1. Further reduction in the number of tactics risked overinclusive categories and a loss of detail in our study of interpretation. Our list represents a broad slice of tactics although there are more possible tactics. As we rely on a 20th-century dataset for our tactics, our analysis is less fine-grained for online contention. To lessen the cognitive load on participants, we opted for a slider format rather than Likert-style response categories.

Table 1.

Tactics included in this study.

1. Ringing bells

2. Lighting candles

3. Putting up posters

4. Dancing

5. Having a public discussion

6. Going on a hunger strike

7. Setting off fireworks

8. Sending out leaflets

9. Sitting in meditation

10. Collecting signatures for a petition

11. Holding signs in public

12. Marching in the streets

13. Selling items marking the shooting

14. Holding a speech

15. Shouting slogans

16. Holding a silent vigil

17. Worshipping

18. Laying down wreaths

19. Holding a press conference

20. Hosting an art exhibition (e.g. showing films or exhibiting photos)

21. Recruiting members to organisations combating shootings

22. Lobbying politicians

23. Conducting an opinion poll

24. Passing torches

25. Meeting with politicians

26. Distributing goods (e.g. food, clothes)

27. Doing a sit-in

28. Physically attacking someone

29. Holding an art performance (e.g. a play, a musical, singing or drumming)

30. Posting your opinion on the shooting on social media

31. Verbally attacking someone

32. Blockading roads or buildings

33. Making loud noises in public

34. Taking over a building

35. Looting (e.g. stores, homes)

36. Damaging property

37. Kidnapping and holding people hostage

38. Disrupting meetings

39. Conducting a walkout

40. Filing a lawsuit

41. Sharing news coverage of the shooting on social media

We then needed to measure how each tactic scored on the dimensions of legibility and legitimacy. As tactics are never performed in a vacuum, we presented participants with an event that could result in the tactics. The shooting to death of an individual would be recognisable to the participants and could generate the breadth of tactics. We operationalised legibility by asking participants to indicate their opinion on whether each tactic, presented as a possible reaction to the shooting, was a political protest. They could grade the tactics on a scale from zero, indicating that it was surely not a political protest, to 100, indicating that it was surely a political protest. We operationalised legitimacy in a similar manner by asking people to indicate their opinion on whether each tactic was an acceptable reaction. The presentation of the tactics was randomised to avoid order effects. Each participant evaluated all 41 tactics on perceived legitimacy and legibility. We report instructions for participants and question wordings in online Appendix A.

Conducting the Studies and Demographics

Both studies received institutional review board approval, and all participants agreed to a consent form informing them of their rights prior to the survey.

We used the first study to investigate if the dimensions were fruitful in understanding the interpretation of contention. Between 15 October and 17 October 2019, 396 respondents completed the first survey. Furthermore, mapping the positions of the tactics in the two dimensions enabled a comparison with our second study.

In the second study, we attempted to replicate and extend the findings from the first study in a sample representative of the US population. We also investigated whether the assumptions about contention were transposable across contexts and potential heterogeneity among respondents. We varied the political nature of the context, reasoning that a stable pattern across contexts would be evidence of transposability, especially in the dimension of legibility. One-third of the respondents were asked to consider the fatal shooting of an individual in a private quarrel, while another third considered the shooting in a political quarrel. The last third of respondents were not given any specific context. In explicitly framing the shooting as political, we expected that reactions to this shooting would be easier to understand as political. At the end of the survey, we asked the participants about the context for the shooting. A total of 23.1% of respondents were unable to correctly remember the context. While this is high, we include the respondents in our analyses. Dropping respondents can create serious bias (Aronow et al., 2019). Furthermore, the question related to the context for the tactics rather than the dimensions. Respondents who failed this check could still evaluate the tactics in terms of legitimacy and legibility. For transparency, we report the tactics’ means for participants who passed this check in online Appendix B Table B1. We collected the data through the participant-sourcing platform Cloudresearch to be representative of the adult US population for age in decades, gender, race and education. Cloudresearch set quotas to match the 2010 US Census. Between 19 November and 22 November 2019, 1033 respondents completed the second questionnaire.

The distributions on key observable demographic outcomes are reported in Appendix B Table B2. Both samples show large variance on demographics (age, sex and race) and socio-economic variables (income, education). In total, 37.7% indicated they were Democrats, while 29.6% indicated they were Republicans and 32.7% indicated they were independent. About one-tenth of respondents indicated having protested during the last year, and between four-fifths (study 1) and two-thirds (study 2) reported having voted in the last election where they were eligible.

Cultural Models in the US Public

Mapping the Legitimacy and Legibility of Contentious Tactics

We investigate our data by first mapping the tactics’ legitimacy and legibility. We then explore differences between legitimacy and legibility, transposability, and heterogeneity. Figure 1 plots the means for each tactic for the nationally representative sample, more or less replicating the figure from the first study, reported in online Appendix C Figure C1. It shows large variation in evaluations, with meaningful variation across the dimensions. We find it helpful to interpret the results in terms of the four quadrants.

Figure 1.

Means for legibility and legitimacy, Nat. Rep. study (N = 1033).

The quadrant defined by high legibility and high legitimacy contains conventional tactics such as petitioning, marching in the streets, distributing leaflets and recruiting individuals to join organisations. These tactics seem to be conventional protest tactics that are staples of US political society. In Tilly’s (2008: 145) terms, they belong to the ‘super-strong’ movement repertoire.

The quadrant defined by low legibility and low legitimacy contains two groups of tactics with similar mean scores. First, a small group generally seen as somewhat illegitimate and illegible (fireworks, selling items and making loud noises in the street). Second, a relatively large group of illegible and extremely illegitimate tactics, including looting, verbal attack, physical attack, hostage taking and damaging property. These tactics include actual or potential violence, usually associated with riots. Dancing sits as an isolated tactic with some legitimacy but the lowest legibility.

Below the midline for legibility and above the midline for legitimacy, we find tactics that are, on average, seen as legitimate but less legible as political protests. These tactics include highly legitimate tactics such as the silent vigil, meditation, worshipping and lighting candles, and a group of somewhat legitimate but less legible tactics such as holding an art performance or exhibition. We interpret these as associated with religious or artistic practices that are not clearly political.

The final quadrant contains five tactics generally perceived as illegitimate while still legible as political protest. These include the somewhat illegitimate actions of shouting slogans and going on hunger strike, and very illegitimate but legible actions of disrupting meetings, creating blockades and building takeovers.

The meaningful variation in the means indicates that this is a fruitful approach to studying the public’s interpretation of protest. We wanted to substantiate this by directly investigating our expectations about differences in dimensions, the transposability of interpretation and heterogeneity.

Differences between Dimensions

In the first study, the means for legibility and legitimacy for all tactics differed statistically significantly (p < .05) from each other. Using the means for each tactic on the two dimensions as observations in a regression, the summed means for each tactic correlated positively in the first study but not strongly (r = .346, p = .027). Each dimension explained only 9.68% of the variation in the other. These results indicate that respondents do on average distinguish between legitimacy and legibility in their interpretation of tactics. We plot the results and report the means and standard deviations, as well as the absolute differences and results of simple t-tests for differences in online Appendix C.

The nationally representative sample in study 2 strengthens the external validity of our claim. We used a two-sample t-test to explore if the means for each tactic differed in terms of their legitimacy and legibility. Table 2 reports the sample means and standard deviations for the tactics in each dimension as well as the absolute differences. There was a significant legitimacy-legibility difference for 38 of the 41 tactics. Thirty tactics had absolute differences of more than 10 points on the 0–100 scale, and eight tactics had differences of more than 25 points. There was a stronger legitimacy–legibility correlation than for the first study (r = .479, p = .002), although it does not reach the conventional level for a large correlation of r > .5. The variation in each dimension explained 21.0% of the variation of the other. In other words, knowing the exact variation in the tactics on one dimension, we can explain little more than one-fifth of the variation in the other dimension. This speaks against legitimacy and legibility being identical dimensions. In online Appendix D, we show that the tactics correlated positively and statistically significantly with each other for all pairwise comparisons between legitimacy and legibility ratings, indicating that legitimacy and legibility ratings are not entirely independent within respondents. However, the average size of the correlation is only moderate (r = .290).

Table 2.

Means and standard deviations for tactics for legitimacy and legibility, study 2 (N = 1033).

	Legitimacy	Legibility	Difference
Ringing bells	58.4 (34.0)	38.2 (33.8)	20.3***
Lighting candles	79.4 (27.5)	43.9 (36.7)	35.5***
Putting up posters	70.1 (30.0)	61.7 (30.9)	8.38***
Dancing	41.5 (37.1)	24.7 (29.3)	16.8***
Having a public discussion	78.3 (25.8)	62.6 (31.9)	15.7***
Going on a hunger strike	42.2 (34.4)	53.3 (35.8)	−11.0***
Setting off fireworks	32.4 (33.0)	31.1 (31.8)	1.26
Sending out leaflets	69.1 (29.9)	58.9 (31.8)	10.2***
Sitting in meditation	72.7 (30.2)	41.5 (35.5)	31.3***
Collecting signatures for a petition	76.2 (27.6)	74.0 (27.8)	2.24*
Holding signs in public	68.9 (29.8)	70.7 (28.0)	−1.85
Marching in the streets	59.9 (33.4)	72.2 (30.9)	−12.4***
Selling items marking the shooting	32.2 (33.3)	38.4 (34.8)	−6.21***
Holding a speech	74.3 (27.1)	66.5 (30.1)	7.78***
Shouting slogans	45.0 (34.4)	60.4 (34.3)	−15.4***
Holding a silent vigil	80.9 (26.8)	49.6 (36.8)	31.3***
Worshipping	72.1 (31.9)	36.1 (35.4)	36.0***
Laying down wreaths	77.2 (28.7)	41.6 (35.5)	35.6***
Holding a press conference	75.1 (27.0)	63.2 (31.6)	11.9***
Hosting an art exhibition (e.g. showing films or exhibiting photos)	56.3 (36.1)	37.6 (33.2)	18.7***
Recruiting members to organisations combating shootings	61.3 (35.2)	60.3 (34.1)	0.99
Lobbying politicians	65.7 (32.3)	70.8 (30.3)	−5.14***
Conducting an opinion poll	69.0 (31.2)	57.2 (32.8)	11.9***
Passing torches	52.1 (34.3)	44.5 (33.1)	7.57***
Meeting with politicians	75.2 (27.6)	68.2 (31.3)	7.06***
Distributing goods (e.g. food, clothes)	68.5 (33.3)	36.0 (34.4)	32.5***
Doing a sit-in	61.3 (32.6)	65.1 (32.3)	−3.88***
Physically attacking someone	15.9 (27.8)	32.8 (36.4)	−16.9***
Holding an art performance (e.g. a play, a musical, singing or drumming)	58.7 (35.6)	34.9 (32.7)	23.8***
Posting your opinion on the shooting on social media	67 (31.3.0)	52.5 (33.1)	14.5***
Verbally attacking someone	21.3 (28.6)	39.3 (35.8)	−17.9***
Blockading roads or buildings	25.6 (32.0)	53.2 (38.5)	−27.6***
Making loud noises in public	32.2 (32.4)	44.3 (34.7)	−12.0***
Taking over a building	20.2 (29.6)	50.1 (38.9)	−30.0***
Looting (e.g. stores, homes)	15.2 (27.9)	35.0 (38.1)	−19.8***
Damaging property	15.4 (27.3)	38.9 (38.0)	−23.5***
Kidnapping and holding people hostage	14.4 (28.0)	37.7 (39.5)	−23.3***
Disrupting meetings	28.0 (31.5)	52.7 (36.4)	−24.7***
Conducting a walkout	52.7 (34.8)	65.4 (33.0)	−12.6***
Filing a lawsuit	67.0 (30.8)	55.4 (33.9)	11.6***
Sharing news coverage of the shooting on social media	68.4 (29.9)	55.0 (33.4)	13.4***

Note: results from two-sample t-tests. Standard deviations reported in parentheses. Significance levels: * < .05, ** < .01, *** < .001.

The dimensions of legitimacy of legibility nuance how we understand the public’s interpretation of tactics. For example, whereas blockading roads is generally seen to be moderately political, with a mean score of 53.2, it is generally seen as strongly illegitimate, with a mean legitimacy score of 25.6. Its pairwise correlation coefficient is .257. Contrast this with the act of lighting candles, which is somewhat politically illegible (mean = 43.9) but much more legitimate (mean = 79.4) than blockading roads. It also has a small pairwise correlation of .096. The two dimensions capture an important distinction in how the public interpret contention.

The Transposability of Interpretation

Our second expectation was that the tacit assumptions about contention are transposable. The pattern observed above ought to be relatively stable across political and private contexts in our nationally representative study.

The results support this expectation. Online Appendix Tables C1 and C2 and Figure C1 report the results of tests comparing the means in the private and political contexts. Overall, context did little to change most evaluations of legitimacy and legibility. The difference in legitimacy across contexts was statistically significant for seven tactics: collecting signatures for a petition, marching, selling items, lobbying politicians, conducting an opinion poll, meeting politicians and conducting a sit-in. The largest difference was for meeting with politicians at 7.6 points. For legibility, seven tactics differed across contexts. These were lighting candles, holding a silent vigil, worshipping, laying down wreaths, holding an art performance, posting on social media and filing a lawsuit. The largest difference was for the silent vigil at 7.57 points. In all cases, tactics were more legitimate and more legible in the political context than in the private context.

If we plot the means for the 14 tactics that differed on at least one dimension between contexts (Figure 2), it seems that each tactic is anchored around a position in the dimensions. While context does affect the interpretation of some tactics, it seems that interpretation remains patterned by a series of tacit assumptions. Having a public discussion is political and legitimate. This is the case when it is in a private, political or non-specific context. Likewise, looting or property destruction are neither legitimate nor clearly political even when responding to a politically motivated killing. In interpreting contentious actions, the public seem to draw on both contextual clues and their tacit assumptions about legitimacy and legibility.

Figure 2.

Graph of the 14 tactics differing significantly in at least one dimension between the two contexts.

Observed Differences in the Interpretation of Contentious Tactics

In this article, we focus on the public as a whole. However, we also see indications of heterogeneity among members of the public, indicated by the large standard deviations in Table 2. This may be an artefact of using sliders as a response format with respondents selecting the extremes of 0 and 100. However, the large variances could also reflect observed and unobserved differences across members of the public.

We approach this question of heterogeneity in two ways. First, we carry out a set of analyses, shown in online Appendix D, of the differences in legitimacy and legibility ratings across observed political, socio-economic and demographic characteristics. Second, we investigate possible heterogeneity stemming from unobserved characteristics through a latent profile analysis of the two dimensions, which we report in online Appendix E.

The pattern for observed differences tallies with previous research on group differences in legitimacy ascriptions (e.g. Hall et al., 1986; Park and Einwohner, 2019; Rodeghier et al., 1991). All else equal, belonging to a group with a higher education, having previous protest experience, having a democratic partisan identity and identifying as a male are all related to generally higher legitimacy ratings. Self-identified race also matters. Those who describe themselves as black indicated higher legitimacy ratings than those who described themselves as white. Our fine-grained image of protest tactics nuances this general picture. For example, while previous protest experience relates to higher legitimacy ratings for almost all tactics (38 out of 41), higher education was only related to higher ratings for non-violent tactics (see Hall et al., 1986; Park and Einwohner, 2019). On the contrary, identifying as black was related to higher acceptance of classical protest tactics and disruptive and violent tactics such as blockading or looting. These differences suggest that previous protest experience, education and race shape the interpretation of protest in substantially different ways.

The group differences in the dimension of legitimacy did not necessarily replicate themselves in the dimension of legibility, illustrated by partisan identity. Republicans indicated significantly lower legitimacy ratings than Democrats on 27 tactics. However, this was attenuated for legibility. Here, only 12 tactics were significantly less legible for Republicans and the correlation coefficients were smaller than for legitimacy. The differences for education, gender and age were similarly attenuated. Moreover, group legibility differences for race were not the same as in the dimension of legitimacy. Most notably, self-identified black people were not more likely to see violent tactics as protests. It is only by distinguishing between legitimacy and legibility that we can study these different patterns.

Our approach shines a new light on the heterogeneity of interpretation in the public by nuancing group differences in legitimacy and distinguishing these from differences in legibility.

Unobserved Differences: Latent Profile Analysis

We also investigate whether our approach is fruitful in analysing heterogeneity across unobserved differences in the population. We do so through a latent profile analysis. This is a person-centred approach that identifies latent subpopulations given a set of indicator variables (Spurk et al., 2020; Woo et al., 2018; for approaches that focus on cultural schemas see Boutyline, 2017; Boutyline and Soter, 2021). It is ‘person-centred’ because it clusters persons and not variables. We conducted two analyses for legitimacy and legibility respectively, using individual responses to the 41 tactics as indicators. We compared several models for each dimension to find the optimal model. Here, we focused on the relative fit criteria given in the Bayesian Information Criterion (BIC), bootstrapped likelihood tests, as well as theoretical plausibility (see Oberski, 2016; Spurk et al., 2020). We discuss the results here and report them in online Appendix E, which also shows the distributions of observed characteristics across profiles.

In the analysis of legitimacy, the best solution was a model with eight profiles shown in online Appendix E, Tables E1 and E2 (BIC = 394161.8). The most numerous profile 5 (N = 399) is in many ways an extreme version of the population means. Here, having a public discussion is very legitimate, with a score of 87.6, and looting very illegitimate, scoring 3.0. Many of the other profiles resemble this overall pattern, differing in certain tactics such as disruptive but non-violent ones. In profile 4 (N = 92) making loud noises in public is very illegitimate, with a score of 5.8, whereas it is somewhat legitimate in profile 1 (N = 126) with a score of 54.8. However, some profiles differ markedly from this general pattern, having legitimacy scores that are either very high (profile 7), very low (profile 8) or at the midpoint of the scale (profile 6). These profiles may be an artefact from the sliding response scale or reflect small subpopulations in the public. The latter is suggested in online Figure E1 by the high proportion of people with protest experience in profile 7.

Turning to legibility, the best solution was a model with seven profiles shown in online Appendix E, Tables E3 and E4 (BIC = 386585.3). The overall impression is again one of substantial overlap with some variation in individual tactics. Most profiles are characterised by high legibility scores on the conventional protest tactics belonging to the ‘super-strong’ repertoire such as marching in the streets (Tilly, 2008: 145). At the same time, there are differences in the legibility of disruptive tactics such as blockading roads or buildings. In profile 1 (N = 346), it is largely legible with a score of 76.2 whereas it is largely illegible in profile 3 with a score of 11.9 (N = 221). There is also variation in the degree to which artistic or religious tactics are legible. Laying down wreaths scores 15.7 in profile 1 but 76.3 in profile 4 (N = 116). Finally, as with legitimacy, we see small profiles characterised by legibility scores that are either very low, very high or at the midpoint of the scale. Online Figure E2 shows that individuals in these profiles differ across observable characteristics, suggesting that the profiles reflect actual subpopulations. For instance, highly educated individuals with previous protest experience are overrepresented in the high legibility profile 7.

The latent profile analyses help us further understand the shared patterns in how the public interpret contention. They suggest that variation in the interpretation of some tactics exists alongside widely shared assumptions about the legitimacy and legibility of other tactics, such as those found in the super-strong protest repertoire. It is only by distinguishing legitimacy from legibility that we can study these shifting patterns.

Developing the Cultural Models Approach

The cultural models approach unpacks the public’s interpretation of contention theoretically and empirically. While the results speak to the fruitfulness of this approach, they also raise questions about the transposability of interpretation, the dimensions of legibility and legitimacy, and the relations between tactics.

The Transposability of Cultural Models of Contention

The results suggest that there are a series of background assumptions in the US public about which tactics count as political and are acceptable. One could ask whether these assumptions would continue to pattern interpretation if the context were fleshed out even further.

In the dimension of legitimacy, we can supplement our results with the literature on the public perception of protest. Wouters (2019) found that tactics displaying worthiness heightened the audience’s support for protests. He replicated this pattern in the different contexts of a Black Lives Matter protest in the United States and an asylum seeker protest in Belgium. Simpson et al.’s (2018) study of the negative effects of violence on protest support also suggest a degree of stability, if not complete fixity across contexts (see Hsiao and Radnitz, 2021). Although support is not legitimacy, the results add further plausibility to the transposability of assumptions.

Turning to the dimension of legibility, our results indicate that legibility is as transposable as legitimacy. However, this dimension has rarely been systematically studied (for classical studies, see Altheide and Gilmore, 1972; Jeffries et al., 1971; Turner, 1969). In unpacking the transposability of legibility assumptions, future studies could engage more directly with how tactics are understood as political protests or not. Some tactics may be illegible because they are unfamiliar to audiences, needing more contextualisation to be interpreted as protests. Art exhibitions might fall in this category. Other tactics are illegible because they are defined in opposition to legible tactics and not because they are unfamiliar. For example, looting and property destruction are often contrasted to protest tactics such as marching in the streets and pamphleteering. If these tactics are to be interpreted as protests, it may require a reconfiguration of what counts as protests.

Dimensions

The dimensions of legitimacy and legibility are useful when studying the public’s interpretation of contention. However, there are likely nuances to the two dimensions and alternative dimensions. Furthermore, we do not claim to cover all modern protest tactics. For example, we do not focus much on digital and online tactics. A complementary approach could bring Hunzaker and Valentino’s (2019) concept-association approach to bear on the question of what constitutes a protest. Focusing less on the interpretation of individual actions and more on the concepts that constitute the notion of protest would unpack the dimension of legibility. Moreover, it would untangle the role that underlying schemas play in the cultural models (Lizardo, 2017).

Thick Description

Finally, we focus on an overall pattern in interpretation. We do not provide a thick description of the meaning of these tactics or the relations between them in the cultural models. In-depth interview studies could therefore complement our approach. These studies could also probe whether individuals have more than one cultural model, something that the quantitative approaches used here do not allow for (see Rotolo, 2021).

Conclusion

We theorise a cultural models approach that allows us to analyse the patterns in how the public interpret contention. We show the fruitfulness of this approach and its distinction between the dimensions of legitimacy and legibility in two studies of the US public. In following this line of research, we gain a new understanding of how the interpretation of contention is patterned. In liberal democracies in particular, contention works through the public (Alexander, 2006; Koopmans, 2004). This means that cultural models of contention in the public both enable and constrain attempts to promote social change. This is not to say that only legitimate and legible tactics can bring about change. Sometimes calls for radical change are best expressed through tactics that challenge the tacit assumptions among the public (Gillan, 2020). Rather, these assumptions are central to the potential for social change even when they are challenged.

Supplemental Material

sj-docx-1-soc-10.1177_00380385221109698 – Supplemental material for Cultural Models of Contention: How Do the Public Interpret the Repertoire of Contention?

Supplemental material, sj-docx-1-soc-10.1177_00380385221109698 for Cultural Models of Contention: How Do the Public Interpret the Repertoire of Contention? by Johan Gøtzsche-Astrup and Oluf Gøtzsche-Astrup in Sociology

Footnotes

Funding

The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: the project was funded through an internal research grant from Aarhus University.

ORCID iDs

Johan Gøtzsche-Astrup

Oluf Gøtzsche-Astrup

Supplemental material

Supplemental material for this article is available online.

Johan Gøtzsche-Astrup is a visiting research associate at the Department of Political Science, Aarhus University and an incoming Postdoctoral fellow at Yale’s Center for Cultural Sociology. His research centres on the constitution of contention from both a historical and contemporary perspective. His current work focuses on public constructions of activists and protests.

Oluf Gøtzsche-Astrup is Assistant Professor at the Department of Political Science, Aarhus University. His research focuses on political activism and radicalism, professional collaboration and public leadership.

References

Alexander

(2006) The Civil Sphere. New York, NY: Oxford University Press.

Altheide

Gilmore

(1972) The credibility of protest. American Sociological Review 37(1): 99–108.

Andrews

Caren

(2010) Making the news: Movement organizations, media attention, and the public agenda. American Sociological Review 75(6): 841–866.

Andrews

Beyerlein

Farnum

(2016) The legitimacy of protest: Explaining white southerners’ attitudes toward the civil rights movement. Social Forces 94(3): 1021–1044.

Aronow

Baron

Pinson

(2019) A note on dropping experimental subjects who fail a manipulation check. Political Analysis 27(4): 572–589.

Benford

Snow

(2000) Framing processes and social movements: An overview and assessment. Annual Review of Sociology 26: 611–639.

Biggs

(2013) How repertoires evolve: The diffusion of suicide protest in the twentieth century. Mobilization: An International Quarterly 18(4): 407–428.

Boonen

(2020) At the outer limits of democratic division: On citizenship, conflict and violence in the work of Chantal Mouffe and Étienne Balibar. International Journal of Politics, Culture, and Society 33(2): 529–544.

Bourdieu

(1990) The Logic of Practice. Stanford, CA: Stanford University Press.

10.

Boutyline

(2017) Improving the measurement of shared cultural schemas with correlational class analysis: Theory and method. Sociological Science 4: 353–393.

11.

Boutyline

Soter

(2021) Cultural schemas: What they are, how to find them, and what to do once you’ve caught one. American Sociological Review 86(4): 728–758.

12.

Cottle

(2008) Reporting demonstrations: The changing media politics of dissent. Media, Culture & Society 30(6): 853–872.

13.

Doherty

Hayes

(2019) Tactics and strategic action. In: Snow

Soule

Kriesi

, et al (eds) The Wiley Blackwell Companion to Social Movements. Hoboken, NJ: Wiley Blackwell, 271–288.

14.

Drakulich

Denver

(2022) The partisans and the persuadables: Public vies of Black Lives Matter and the 2020 protests. Perspectives on Politics. Epub ahead of print 1 March 2022. DOI: https://doi.org/10.1017/S1537592721004114.

15.

Francis

Wright-Rigueur

(2021) Black lives matter in historical perspective. Annual Review of Law and Social Science 17: 441–458.

16.

Gillan

(2020) Social Movements, Protest, and Practices of Social Change. In: Ritzer

Murphy

(eds) The Wiley Blackwell Companion to Sociology, 2nd edn. Hoboken, NJ: Wiley Blackwell, 301–318.

17.

Goodwin

Jasper

(1999) Caught in a winding, snarling vine: The structural bias of political process theory. Sociological Forum 14(1): 27–54.

18.

Gøtzsche-Astrup

(2022) A sociological perspective on the experience of contention. Sociological Theory. Epub ahead of print 2 May 2022. DOI: https://doi.org/10.1177/07352751221097115.

19.

Hall

Rodeghier

Useem

(1986) Effects of education on attitude to protest. American Sociological Review 51(4): 564–573.

20.

Hsiao

Radnitz

(2021) Allies or agitators? How partisan identity shapes public opinion about violent or nonviolent protests. Political Communication 38(4): 479–497.

21.

Hunzaker

MBF

Valentino

(2019) Mapping cultural schemas: From theory to method. American Sociological Review 84(5): 950–981.

22.

Jeffries

Turner

Morris

(1971) The public perception of the Watts riot as social protest. American Sociological Review 36(3): 443–451.

23.

Koopmans

(2004) Movements and media: Selection processes and evolutionary dynamics in the public sphere. Theory and Society 33(3): 367–391.

24.

Koopmans

Olzak

(2004) Discursive opportunities and the evolution of right-wing violence in Germany. American Journal of Sociology 110(1): 198–230.

25.

Lamont

Molnár

(2002) The study of boundaries in the social sciences. Annual Review of Sociology 22: 167–195.

26.

Larson

(2013) Social movements and tactical choice. Sociology Compass 7(10): 866–879.

27.

Larson

Lizardo

(2019) Institutional movement logics and the changing shape of the US social movement field, 1960-1995. Social Forces 97(4): 1389–1422.

28.

Lizardo

(2010) Beyond the antinomies of structure: Levi-Strauss, Giddens, Bourdieu, and Sewell. Theory and Society 39(6): 651–688.

29.

Lizardo

(2017) Improving cultural analysis: Considering personal culture in its declarative and nondeclarative modes. American Sociological Review 82(1): 88–115.

30.

Luhmann

(1995) Social Systems. Stanford, CA: Stanford University Press.

31.

McAdam

Boudet

(2012) Putting Social Movements in Their Place: Explaining Opposition to Energy Projects in the United States, 2000–2005. New York, NY: Cambridge University Press.

32.

McCurdy

Feigenbaum

Frenzel

(2016) Protest camps and repertoires of contention. Social Movement Studies 15(1): 97–104.

33.

Martin

Rafail

McCarthy

(2017) What a story? Social Forces 96(2): 779–702.

34.

Oberski

(2016) Mixture models: Latent profile and latent class analysis. In: Robertson

Kaptein

(eds) Modern Statistical Methods for HCI: A Modern Look at Data Analysis for HCI Research. Cham: Springer, 278–287.

35.

Olsen

(1968) Perceived legitimacy of social protest actions. Social Problems 15(3): 297–310.

36.

Palthe

(2014) Regulative, normative, and cognitive elements of organizations: Implications for managing change. Management and Organizational Studies 1(2): 59–66.

37.

Park

Einwohner

(2019) Becoming a movement society? Patterns in the public acceptance of protest, 1985-2006. Sociological Focus 52(3): 186–200.

38.

Ring-Ramirez

Reynolds-Stenson

Earl

(2014) Culturally constrained contention: Mapping the meaning structure of the repertoire of contention. Mobilization 19(4): 405–419.

39.

Robinson

Rosenzweig

Moss

, et al. (2019) Tapped out or barely tapped? Recommendations for how to harness the vast and largely unused potential of the mechanical turk participant pool. PloS One 14(12): e0226394.

40.

Rodeghier

Hall

Useem

(1991) How education affects attitude to protest: A further test. The Sociological Quarterly 32(2): 277–288.

41.

Rotolo

(2021) Culture beneath discourse: A conceptual model for analyzing nondeclarative cultural knowledge. American Journal of Cultural Sociology. Epub ahead of print 11 June 2021. DOI: 10.1057/s41290-021-00135-0.

42.

Scott

(1995) Institutions and Organizations. Los Angeles, CA: SAGE.

43.

Seguin

(2015) Cascades of coverage: Dynamics of media attention to social movement organizations. Social Forces 94(3): 997–1020.

44.

Shuman

Hasan-Aslih

van Zomeren

, et al. (2022) Protest movements involving limited violence can sometimes be effective: Evidence from the 2020 BlackLivesMatter protests. Proceedings of the National Academy of Sciences 119(14): e2118990119.

45.

Simpson

Willer

Feinberg

(2018) Does violent protest backfire? Testing a theory of public reactions to activist violence. Socious: Sociological Research for a Dynamic World 4: 1–14.

46.

Snow

Vliegenthart

Ketelaars

(2019) The framing perspective on social movements: Its conceptual roots and architecture. In: Snow

Soule

Kriesi

, et al (eds) The Wiley Blackwell Companion to Social Movements. Hoboken, NJ: Wiley Blackwell, 392–410.

47.

Spurk

Hirschi

Wang

, et al. (2020) Latent profile analysis: A review and ‘how to’ guide of its application within vocational behavior research. Journal of Vocational Behavior 120: 103445.

48.

Tarrow

(1993) Modular collective action and the rise of the social movement: Why the French Revolution was not enough. Politics & Society 21(1): 69–90.

49.

Tarrow

(2011) Power in Movement: Social Movements and Contentious Politics. Cambridge: Cambridge University Press.

50.

Taylor

(1971) Interpretation and the sciences of man. The Review of Metaphysics 25(1): 3–51.

51.

Tilly

(1995) Popular Contention in Great Britain 1750–1834. Boulder: Paradigm Publishers.

52.

Tilly

(2008) Contentious Performances. Cambridge: Cambridge University Press.

53.

Traugott

(1993) Barricades as repertoire: Continuities and discontinuities in the history of French contention. Social Science History 17(2): 309–323.

54.

Turner

(1969) The public perception of protest. American Sociological Review 34(6): 815–831.

55.

Vaisey

(2009) Motivation and justification: A dual-process model of culture in action. American Journal of Sociology 114(6): 1675–1715.

56.

Wada

(2012) Modularity and transferability of repertoires of contention. Social Problems 59(4): 544–571.

57.

Wang

Soule

(2016) Tactical innovation in social movements: The effects of peripheral and multi-issue protest. American Sociological Review 81(3): 517–548.

58.

Woo

Jebb

Tay

, et al. (2018) Putting the ‘person’ in the center: Review and synthesis of person-centered approaches and methods in organization science. Organizational Research Methods 21(4): 814–845.

59.

Wouters

(2019) The persuasive power of protest: How protest wins public support. Social Forces 98(1): 403–426.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.58 MB