American origins: Political and religious divides in US collective memory

Abstract

Origin stories are particularly influential collective memories, establishing a society in the minds of its members. National collective memories are frequently conceived as being shared by all members of the country, but subnational groups may differ in their images of the group’s past. Surveying 2000 Americans, we examined political and religious differences in foundational events selected for America’s origins. While there was considerable agreement across religious and political affiliations for the most important events, there were also critical differences. While all participants demonstrated a marked positivity bias in their origin stories, conservative participants more frequently omitted foundational atrocities from America’s origin story, and thus had the most positive stories. Secular participants were most likely to begin America with the independent state, whereas religious participants frequently began America with earlier colonization events. Origin stories accord with varying images of America and American identity.

Keywords

American origins charters collective memory mnemonic community social representations

Members of a social group will draw on shared memories and historical narratives to support their group identity and common purposes (Halbwachs, 1992). These collective memories differ from professional history in that group solidarity is the goal toward which remembering occurs. Collective remembering is thus always selective, and it may downplay inconvenient or unflattering historical facts (Renan, 1990; Wertsch and Roediger, 2008). One particularly influential class of collective memory is the origin story—how the group or nation came into existence (Liu and Hilton, 2005; Zerubavel, 2003). Origin stories ground collective identity by setting moral imperatives, shared missions, and a society’s raison d’être—in a word, they frequently serve as charters (Hilton and Liu, 2017; Malinowski, 1948). In this study, we examine how Americans represent their nation’s origin story, and whether these social representations vary by religious and political affiliation. We argue that such subnational identities pose different frames through which people construct collective memory. Origin stories themselves usually reflect political and religious values. Becoming enculturated as an American is to some extent to espouse a certain story about what “America” means, and how it originated (Van Engen, 2018, 2020). As we will show, there are multiple ways to frame that story. Although we focus on collective memory in the United States, the relation between memories at a variety of levels of social identity—national, subnational, supranational—is undoubtedly of more general significance.

The “sharedness” of collective memory within a group might take at least three forms (Moscovici, 1988). Collective memory could be hegemonic, that is, largely shared and consensual across a society; it could be emancipated, with different but compatible representations among different subgroups; or it could be polemical, in which members of different groups maintain conflicting narratives of the past. Some earlier research, discussed below, demonstrated remarkably stable and hegemonic representations of key American historical figures across several decades (Frisch, 1989). Other more recent work, however, suggests that contemporary Americans by and large do not agree about “when America was great” (Taylor et al., 2017), which could indicate a failure to converge on a hegemonic charter narrative. We present data drawn from a large national survey in which Americans suggested foundational events and a start year for their nation.

Collective memory

Collective memory may be distinguished from history in that it regards the past through the perspective of contemporary concerns; it represents a people’s living connection to their shared past (Halbwachs, 1992). This use of the past in service of group solidarity in the present differs from the professional historian’s goal to understand the past objectively, on its own terms. Historians recognize that in many ways the past is a “foreign country” alien to contemporary eyes (Lowenthal, 1985), and must rely on rigorous, professional standards of evidence for their claims. Halbwachs may have overdrawn the distinction between memory and history (Schwartz, 1997), but nevertheless, in collective memory, complexity and ambiguity tend to be pruned away in the service of a clear, committed narrative (Novick, 1999).

Rather than seeking an “accurate” depiction of the past per se, collective memory selectively reconstructs the past in a way that supports shared identities (Brooks, 1918). In this spirit, the philosopher and historian Ernst Renan defined a nation as a group of people who share a “rich legacy of memories … having common glories in the past and a will to continue them in the present” (Renan, 1990). Renan suggested that as most nations are founded on tremendous brutality, inconvenient memories must face oblivion or be made otherwise acceptable if people are to converge on a shared memory for “common glories.” That is, substantial collective forgetting is necessary to produce a useful collective memory (Hirst and Yamashiro, 2017). A primary task in nation building has thus been to manufacture selective convergence onto officially sanctioned renderings of the past (Hobsbawm and Ranger, 2013; Smith, 1998; Zerubavel, 1994). Some aspects of this task in the US context seem to have been quite successful. For instance, over 1000 undergraduates, from 1975 to 1988, produced nearly identical lists of important historical figures when asked who “popped into their heads” when they thought of American history between the founding of the nation and the Civil War (Frisch, 1989). Such a stable pantheon suggests that the industries for propagating a particular version of American collective memory did indeed produce a shared core of cultural referents and their associated values.

However, despite these efforts, national memory frequently does fragment along the lines of subnational communities (e.g. Schuman et al., 1994; Wertsch, 2008) or across generations (Schuman and Scott, 1989; Zaromb et al., 2014). The ongoing controversy over Confederate monuments in the American South poses a vivid recent example of such fragmentation, and the social disruption that may result from it (Fausset, 2017; Landrieu, 2018). Although Renan may have attempted to envision a nation as a mnemonic community (Zerubavel, 1997), and elites and memory activists put forth considerable effort to create such communities (Harris, 2006), a single collective memory shared by the entire nation may be more of an ideal than an empirical reality (Moscovici, 1988).

As evidence for this fragmentation, recent work on information sharing in social media has shown that ideological group boundaries represent powerful structural divides within social networks. Sharing and communication of moralized political information are intensive within clusters of like-minded people and infrequent between clusters (Brady et al., 2017). Such clustering in social networks impacts not only relatively transient phenomena like the sharing of political Tweets, but also convergence onto collective memories, which become shared within but not between network clusters (Coman et al., 2016). An interesting question in an age of intense ideological fragmentation, then, is how successful American “civic religion” is in promoting a hegemonic collective memory for national charter events, given that charter events represent not just history, as well as the values upon which a society is based. To what extent do ideological (political and religious) divides shape how Americans construct their nation’s charter?

Religiosity, politics, and national image

Prior work has suggested that politics and religion are frequently critical fault lines along which collective memories diverge, both in terms of what is included in the collective memory, and in how people evaluate events emotionally. Hanke et al. (2015) examined ratings of 40 historical figures in their 6902 participants, 37-country World History Survey. Participants rated these figures for perceived historical influence and emotional valence (i.e. whether their influence was positive or negative). In a latent profile analysis, participants clustered into four major domains of political culture. Latent profile analysis identifies discrete populations of participants, which are not defined by the researcher a priori, and which are based on characteristic configurations of values in the variables measured. In Hanke et al.’s (2015) case, these variables were the ratings of importance and emotional valence for each of the 40 historical figures. Latent profiles thus represented discrete populations defined by unique configurations of attitudes and beliefs concerning prominent figures of collective memory. Hanke et al. (2015) interpreted these four profiles as Secular or Religious Idealists, Political Realists, and Historical Indifferents. Idealists view politics primarily through the lens of values—human rights, pluralism, liberal democracy, and the culture of cooperation it requires, and so on while Realists view politics through the lens of a zero-sum competition between groups, power being of value in and of itself. The primary divide between Idealists and Realists had to do with their attitudes toward dictators, with Realists expressing less uniformly negative views of famous authoritarians like Mao and Stalin. The divide between Religious and Secular Idealists primarily concerned their attitudes toward prominent religious figures. Secular Idealists were less likely to assign historical influence or positive ratings to religious figures, such as the Buddha, Mohammed, and Pope John Paul II, than Religious Idealists. The two historical idealist profiles—religious and secular—were most prevalent in Western and historically Christian-majority countries, whereas Political Realists were more common in Muslim-majority and Asian countries, particularly China. Hanke et al.’s (2015) data suggest that a primary divide of interest within American collective memory might emerge between religious—in the US context, predominantly Christian—and secular Americans. Religiosity and political affiliation undoubtedly influence how Americans espouse and formulate exceptionalist beliefs, imagine America’s historical trajectory, and, in turn, remember America’s foundational events. What exactly they believe America to be about should be reflected in the origin stories they construct for it.

Collective and collected memory

Before continuing on to our survey, we must briefly address a broader methodological issue. There are scholars who would find our psychological approach to collective memory odd. After all, survey data are aggregated measurements drawn from individuals, rather than descriptions of social processes proper. Some have argued that the label “collective memory” has been inappropriately applied to two sorts of research cultures, only one of which, properly speaking, truly addresses “collective” memory (Olick, 1999). Such scholars distinguish traditionally sociological sorts of studies, such as the debates and controversies involved in constructing memorials or museums, as properly “collective,” from studies examining how lay people represent their groups’ pasts, which are characterized as “collected” memory studies.

We argue that psychological approaches examining how individuals remember their group’s past do contribute to broader collective memory scholarship. As Hirst and Manier (2008) argued, one powerful model for the study of collective memory is epidemiology—how do particular memories “infect” an entire population? Cognitive epidemiologists should study both individual level phenomena, such as what makes a particular representation more memorable and likely to be transmitted (e.g. Sperber and Hirschfeld, 2004), and community practices that impact the scale of exposure to particular renderings of the past, and how particular memories propagate. Insight at both scales is necessary to understand how a particular set of memories comes to be shared across an entire population (Vlasceanu et al., 2018). For example, Holocaust consciousness is maintained by an immense array of institutions, museums, media, and rituals (Novick, 1999). To be sure, what members of a given population remember about the Holocaust if asked is only a part of this complex activity of “remembering the Holocaust.” However, the case could be made that such social representations as held by individuals should be a critical component, a sine qua non, to a proper characterization of collective memory of the Holocaust. Of what interest would be museums and memorials if no one attended to or remembered them? Is not lay people’s representation of the Holocaust after going to a museum of just as much scholarly interest as the design choices that went into building the exhibits?

We now describe a survey in which we solicited events Americans believed to be foundational in their nation’s origin. We examined whether Americans converged onto a shared origin story, or whether they produced a diversity of events, and if so, whether there would be telling divides between Americans of different political and religious affiliations. Again, although we focus on America as a case study, we will argue that many of the phenomena we describe will be more generally applicable.

Methods

Participants

We surveyed 2000 anonymous participants on Amazon’s Mechanical Turk for a compensation of $3. Inclusion criteria included fluency in written and spoken English (English did not need to be the participants’ first language), habitation in the United States, and self-identification as an American. Ages ranged from 18 to 81, with a mean age of 36 years old. Forty-seven percent of participants identified as women and the remainder as men, except for five participants who reported “Other” for gender. Self-reported ethnicity indicated 77% White, 7.7% Black, 6.5% Asian, 5.2% Latino, 1.4% Mixed, and all other designators below 1%. Regarding religion, 50% of participants identified as Christian, with Christians groups as a proportion of the total equaling 21% Protestant, 17% Catholic, 4% Evangelical, 1% Orthodox, 1% Mormon, and 6% Christian-Other. Non-religious participants accounted for 44% of the sample (27% Atheist/Agnostic, 17% no affiliation). Other religions accounted for 4% of the sample (Jewish 2%, Buddhist 0.9%, Hindu 0.6%, Muslim 0.5%), and the remaining 2% of participants marked their religion as “Other.” Regarding political party affiliation, 41% of participants identified as Democrat, 21% as Republican, 28% as Independent, 8% as unaffiliated, and 1% as Other. No participants were categorically excluded. However, for specific analyses, if participants did not respond to a relevant item, or did not belong to the groups of interest, they were excluded. We will disclose these ad hoc exclusions where they are relevant. The anonymous survey was exempted from review by Washington University in St. Louis’ IRB. Subjects were informed about the study and their right to withdraw at the outset of the survey.

Survey tasks

The data reported here were drawn from a larger survey on American collective memory. Survey, data, and additional analyses are hosted on the Open Science Framework (OSF), as the “American Origins Collective Memory Project,” at https://osf.io/kr567/. We focus on two sets of questions for this inquiry.¹ In the first set, participants provided five historical events “important to the foundation of America.” We left the term “America” undefined and allowed participants to pick historical events corresponding to whichever concept of “America” they deemed relevant. They typed each event into one of five dialogue boxes. Because prior work on group differences in collective memory has shown that collective memories can vary both in terms of the events selected, and in attitudes toward the same events (e.g. Hanke et al., 2015; Zaromb et al., 2014), participants additionally rated events for emotional valence. Below each dialogue box participants could rate the event as “positive” or “negative.” We opted not to use a bipolar valence scale because such scales have an inherent ambiguity built into them; they conflate changes in valence with changes in arousal (Kensinger and Corkin, 2004). We opted for a binary valence rating in order to force a choice and simplify analyses. In the second question of interest, participants provided 1 year in response to the question, “When did America begin?”

Participants thus provided a set of origin events, some of which were positive and some negative, and a start year for America. These events in the aggregate will be referred to as participants’ American origin story. The two questions—five foundational events with ratings, and America’s start year—were the second and fourth items, respectively, on a 31-question survey. The first question asked for a one-sentence description of the origin of America, and the third question asked participants to provide 10 historical events that “all Americans should remember,” entered and rated in the same format as the origin events. All participants were exposed to these questions in the same order, thus systematic confounds in priming across participants are unlikely. Finally, from question number 14 onward, participants provided demographics. Data were collected over a 2-week period in November 2017.

Coding

We calculated the proportion of origin events that were rated positively. This measure could fall along a scale from 0 (all negative events) to 1 (all positive events). A ratio of 0.5 indicated an equal number of positive and negative events, when aggregated across members of a group (because we solicited an odd number of events, no individual participant could have a ratio of 0.5, although a group of individuals could). We will refer to this valence score as the Foundations Rating. Foundations Ratings indicate the general emotional tenor of America’s origin story as participants constituted it in their retelling. Renan’s formulation of a nation as those sharing memories for “common glories” would lead us to predict positivity biases in Foundations Ratings (i.e. Foundations Ratings > 0.5).

In the content analysis, we coded participants’ free responses to the five origin event probes into event categories using the open source application OpenRefine that offers several tools for coding textual data (Verborgh and De Wilde, 2013). A simple example would be a clustering algorithm that suggests merging several variants of a term, for example, “Declaration of Independence,” which differ in spelling or capitalization. The user can examine the suggested members of a category cluster, and collectively recode them to represent the same category if desired. Three human coders were involved in reducing the event data, and ambiguous clustering choices were discussed and resolved by agreement of two of the three coders. Coders were blind to any information concerning which responses came from which participants. The primary question concerning clustering choices arose around the appropriate level of abstraction at which a category should be defined; participants frequently provided both relatively general events (e.g. “British Colonization”) and specific events (e.g. “Landing at Plymouth Rock”). Because the specificity of categories in collective memory is potentially interesting data, we opted not to cluster responses according to a pre-specified level of generality. Instead, we developed a heuristic: if more than 10 participants mentioned an event (e.g. “Boston Tea Party”), it got its own category. If fewer than 10 participants mentioned an event (e.g. “Cornwallis’ surrender at Yorktown”), it was clustered into the superordinate generic category (i.e. “Revolutionary War”). We judged this rule to be a reasonable compromise between coding strategies that multiplied event categories beyond reasonable limits, and those that led to the loss of interesting data by collapsing all events into their most generic representation. Algorithmic clustering in OpenRefine allowed the coders to reduce unique event categories from nearly 10,000 to approximately 500 events. Further reduction was accomplished by the coders’ human judgment, with controversial decisions resolved in discussion; for example, the simple terms “Indians,” “Native Americans,” and “First Nations people” were all collapsed into the “Native Americans” category, but the frequently mentioned but more specific “Native American Genocide,” “Native American Displacement,” and “Native American Migration into North America” were kept as separate categories.

Results

Emotional valence of American origin story

All means and mean differences are presented with 95% confidence intervals in brackets. Overall, participants provided mostly, but not unanimously, positive items for America’s foundational events, M = 0.68 [0.67, 0.69]. Our two social markers of primary interest were religion and political affiliation. We focused on the four largest religiosity groups (Catholic, Protestant, Atheist/Agnostic, and no religious affiliation) and the four most common political affiliations (Democrat, Republican, Independent, and Non-Affiliated). Unfortunately, our samples of other religious and political minorities were not of sufficient size to allow proper statistical comparisons, and the following analyses will not include them.

Religion and political affiliation tend to be correlated, and our sample bears out that tendency. Table 1 shows the number of people in our sample broken down by religiosity (rows) and political affiliation (columns). Christians are overrepresented in the Republican Party, and Atheists/Agnostics and the religiously unaffiliated dramatically underrepresented. Catholics and non-religious people were more or less proportionally represented in the Democratic Party in our sample, with Protestants slightly underrepresented, and Atheist/Agnostics slightly overrepresented. However, although political affiliation and religion are somewhat correlated, they are not reducible to one another in any straightforward way.

Table 1.

Crosstabs for political affiliation by religiosity.

Religiosity	Political affiliation				Total
Religiosity	Democrat	Republican	Independent	No affiliation	Total
Protestant	153 (220)	211 (116)	129 (147)	29 (40)	522
Catholic	138 (137)	104 (72)	73 (92)	11 (25)	326
Atheist/Agnostic	292 (222)	34 (116)	163 (148)	37 (40)	526
No religious ID	135 (139)	28 (73)	115 (93)	52 (25)	330
Total	718	377	480	129

The numbers in parentheses indicate the count expected if the cells were proportionally representative, and numbers outside parentheses indicate the actual count.

X²(9) = 274.56, p < 0.001.

We first examined whether there were group differences in overall Foundations Ratings. Table 2 provides mean Foundations Ratings for each religiosity × political affiliation cell. An overall pattern emerged in which Christian participants, both Protestant and Catholic, provided more positive origin stories than non-religious participants, and Republicans provided more positive origin stories than people from the other three political affiliations. In order to examine whether one of these two variables, religion or political affiliation, was the primary driver of these differences, we entered both political orientation and religion into a univariate analysis of variance (ANOVA) with Foundations Ratings as the predicted variable. With both social groupings entered into the model, only political party remained a significant predictor, F(3,1688) = 6.66, p < 0.001, ηp^2 =0.01. Neither religion, F(3,1688) = 1.83, p = 0.14, ηp^2 =0.003, nor an interaction between religion and political party, F(9,1688) = .55, p = 0.84, ηp^2 =0.003, reached statistical significance. Republicans provided more positive origin stories than Democrats, M_diff = 0.09 [0.05, 0.13], p < 0.001, than Independents, M_diff = 0.06 [0.02, 0.11], p = 0.004, and than those with no political affiliation, M_diff = 0.09 [0.02, 0.15], p = 0.007. These latter three groups did not differ statistically from one another, all p > 0.05.

Table 2.

Proportion of origin events that were positive, for political affiliation by religiosity.

Religiosity	Political affiliation				Mean
Religiosity	Democrat	Republican	Independent	No affiliation	Mean
Protestant	0.67 [0.63, 0.71]	0.76 [0.73, 0.80]	0.70 [0.65, 0.74]	0.72 [0.62, 0.81]	0.72 [0.70, 0.74]
Catholic	0.67 [0.62, 0.71]	0.75 [0.70, 0.80]	0.70 [0.65, 0.76]	0.65 [0.50, 0.81]	0.70 [0.67, 0.73]
Atheist/Agnostic	0.66 [0.63, 0.69]	0.73 [0.64, 0.82]	0.65 [0.62, 0.69]	0.60 [0.52, 0.68]	0.66 [0.64, 0.68]
No religious ID	0.63 [0.59, 0.67]	0.74 [0.65, 0.84]	0.68 [0.63, 0.73]	0.67 [0.60, 0.74]	0.66 [0.64, 0.69]
Mean	0.66 [0.64, 0.68]	0.76 [0.73, 0.78]	0.68 [0.66, 0.70]	0.66 [0.61, 0.70]

95% confidence intervals in brackets.

We note that these political differences were not confounded by age differences or differences in educational attainment. For reports on analyses ruling out age and education as potential confounds, see supplementary online material on the OSF. Briefly, our sample showed no age difference between participants of differing political affiliations when accounting for religion, and age-matched subsamples examining only participants under 40 showed that young Republicans still produced the most positive origin stories. Further, although Democrats were overrepresented among holders of advanced degrees and Republicans and non-affiliated overrepresented among those with less than college degrees by approximately the same amount, the non-affiliated showed Foundations Ratings comparable to Democrats, despite having similar levels of educational attainment as Republicans. Neither age nor educational attainment independently predicted Foundations Ratings when accounting for political affiliation.

Despite nominally belonging to the same nation, people of differing political affiliation styled America’s origin story differently. Although all groups showed at least a moderate positivity bias (i.e. all groups had Foundations Ratings higher than 0.5), Republicans tended to produce the most positive origin stories. To explain this divergence, we next conducted a content analysis on origin events. The observed difference in Foundations Ratings could have emerged through at least two routes: different tendencies to exclude negative events from the origin story, and group differences in emotional evaluations of the same events.

Content analysis

Following the two waves of data reduction, the coders converged on 288 unique event categories. One hundred and seven of these event categories were mentioned by 10 or more people, and 22 of these event categories were mentioned by more than 5% of the total sample (i.e. 100 people; see Figure 1). In examining group differences, we focused on the 12 items mentioned by at least 10% of the full sample, as described below. For a full list of unique categories mentioned by at least 10 people, with mention frequency, see Supplemental Appendix 2 in the OSF repository.

Figure 1.

Frequency of mention for each of the 22 foundational event categories that were cited by at least 5% of the sample (i.e. 100 people).

Hegemony in collective memory

The 12 events mentioned by at least 10% of the group are listed in Table 3, with the proportion of participants from each political affiliation mentioning that event as well as the proportion of those mentions that were rated positively. All groups agreed on the top three events: Revolutionary War, Declaration of Independence, and Christopher Columbus’ discovery. As a measure of the extent to which groups prioritized the events similarly, as determined by frequency of mention, we conducted Kendall’s Tau rank correlations between every pair of groups (see Table 4). Overall, there was good agreement across political affiliations on the relative ranking of the most important events. This suggests a degree of hegemonic memory for American origins across political affiliations, but this agreement should not be overstated; as already indicated, even the most frequently mentioned event was produced by only slightly over half of participants, and the next most frequently mentioned items were mentioned by only about one-third of the sample.

Table 3.

Proportion of members of each political affiliation mentioning each of the top 12 most frequently mentioned events, with mean emotional valence rating as proportion of mentions rated as positive.

Event	Political affiliation
	RepublicanN = 425		DemocratN = 814		IndependentN = 569		No affiliationN = 159
	Mention	Valence	Mention	Valence	Mention	Valence	Mention	Valence
Revolutionary War	0.53 (0.02)	0.83 (0.03)	0.58 (0.02)	0.75 (0.02)	0.55 (0.02)	0.78 (0.02)	0.54 (0.04)	0.82 (0.05)
Declaration of Independence	0.37 (0.02)	0.99 (0.007)	0.33 (0.02)	0.99 (0.005)	0.35 (0.02)	0.99 (0.006)	0.4 (0.04)	0.98 (0.01)
Christopher Columbus	0.34 (0.02)	0.69 (0.04)	0.31 (0.02)	0.72 (0.03)	0.32 (0.02)	0.70 (0.03)	0.38 (0.04)	0.69 (0.07)
Civil War	0.29 (0.02)	0.39 (0.05)	0.29 (0.02)	0.41 (0.03)	0.27 (0.02)	0.43 (0.04)	0.21 (0.04)	0.44 (0.08)
Constitution	0.19 (0.02)	1.00 (0.009)	0.23 (0.01)	0.99 (0.006)	0.16 (0.02)	1.0 (0.007)	0.19 (0.03)	1.00 (0.02)
British Colonization	0.17 (0.02)	0.72 (0.05)	0.18 (0.01)	0.76 (0.03)	0.21 (0.02)	0.75 (0.04)	0.28 (0.03)	0.91 (0.07)
Boston Tea Party	0.18 (0.02)	0.79 (0.05)	0.16 (0.01)	0.74 (0.04)	0.15 (0.02)	0.81 (0.04)	0.17 (0.03)	0.78 (0.09)
Pilgrims	0.16 (0.02)	0.86 (0.04)	0.17 (0.01)	0.87 (0.03)	0.16 (0.02)	0.84 (0.04)	0.18 (0.03)	0.83 (0.08)
Slavery	0.06 (0.02)	0.02 (0.03)	0.13 (0.01)	0.03 (0.02)	0.12 (0.01)	0.02 (0.02)	0.16 (0.03)	0.12 (0.04)
Native American Genocide	0.05 (0.02)	0.00 (0.00)	0.10 (0.01)	0.00 (0.00)	0.14 (0.01)	0.00 (0.00)	0.15 (0.02)	0.00 (0.00)
British Taxes	0.09 (0.01)	0.35 (0.07)	0.09 (0.01)	0.19 (0.05)	0.10 (0.01)	0.13 (0.05)	0.10 (0.02)	0.38 (0.10)
World War II	0.08 (0.01)	0.34 (0.08)	0.08 (0.01)	0.28 (0.06)	0.10 (0.01)	0.33 (0.06)	0.09 (0.02)	0.29 (0.12)

SEM: standard error of mean.

SEM presented in parentheses.

Table 4.

Kendall’s Tau rank correlation coefficient for order of top 12 origin events (N = 12) between political affiliations.

	Political affiliation
	Democrat	Independent	No affiliation
Republican	0.87*	0.72*	0.77*
Democrat		0.74*	0.69*
Independent			0.69*

Indicates statistical significance at p < 0.001.

Political divides in collective memory

We next examined differences in events and ratings between political affiliations in an attempt to explain why Republicans’ origin stories were so much more “glorious” than those produced by Americans of other political affiliations. At least two explanations could exist for the difference: Republicans (1) could have omitted negative events from the American origin story more frequently and (2) if they provided the same events, they could have been more likely to evaluate them positively.

Proportional mention

Although all rank correlations for the top 12 events were strong and statistically significant, groups did prioritize—or de-prioritize—specific events differently. For example, slavery was the fifteenth most frequently mentioned event in the Republican sample (mentioned by only 8%), whereas, it was ninth in the Democrat (mentioned by 19%) and non-affiliated (mentioned by 27%) samples, and tenth in the Independent (mentioned by 20%) sample. Slavery and Native American Genocide were the two major origin events universally rated as negative. In order to determine whether differences in Foundations Ratings were due to the omission of these negative events, for each of the top 12 events, we conducted a univariate ANOVA with political affiliation as predictor and frequency of mention for that item as the predicted variable. Four of the top 12 events showed statistically significant group differences in frequency of mention—two of which were our negative events of interest: Slavery, F(3,1963) = 6.18, p < 0.001, η² = 0.09; Native American Genocide, F(3,1963) = 7.82, p < 0.001, η² = 0.01; British Colonization, F(3,1963) = 4, p = 0.007, η² = 0.006; and the Constitution, F(3,1963) = 4.21, p = 0.006, η² = 0.006. The remaining events were mentioned equally frequently by members of all political affiliations, all p > 0.05. We focus on between political group comparisons for the two foundational atrocities, as they relate most directly to our question concerning collective forgetting in the service of memory for “common glories.” Reports for British Colonization and the Constitution may be found on the OSF.

The two negative origin events, Slavery and Native American Genocide, were infrequently mentioned by members of all groups, with Slavery mentioned by 12% of participants on average, and Native American Genocide by 11%. However, these events were conspicuously absent from the Republican sample. Republicans cited slavery as a foundational event less frequently than Democrats, M_diff = 0.07 [0.03, 0.11], p < 0.001, than Independents, M_diff = 0.06 [0.02, 0.1], p = 0.004, and than the non-affiliated, M_diff = 0.11 [0.05, 0.16], p < 0.001. Republicans were also significantly less likely to mention Native American Genocide than Democrats, M_diff = 0.05 [0.01, 0.08], p = 0.01, Independents, M_diff = 0.09 [0.05, 0.12], p < 0.001, and the non-affiliated, M_diff = 0.10 [0.04, 0.15], p < 0.001. Democrats, Independents, and the non-affiliated did not differ from one another on either Slavery or Native American Genocide, all p > 0.05. The near absence of the two major atrocities from Republicans’ origin stories partially explains why their origin stories could be so much more positive overall.

Emotional evaluation

We next examined whether group differences in Foundations Ratings might additionally arise from differing emotional evaluations for particular events. Political affiliation groups did not show different emotional evaluations for any of the most frequently mentioned events, except for the Revolutionary War, F(3,674) = 2.77, p = 0.041, η² = 0.012. Republicans (M = 0.86 [0.79, 0.93]) rated this event more positively than Democrats (M = 0.74 [0.69, 0.79]), M_diff = 0.12 [0.04, 0.21], p = 0.004, and than Independents (M = 0.77 [0.71, 0.83]), M_diff = 0.09 [0.001, 0.18], p = 0.048, and did not differ from the unaffiliated (M = 0.77 [0.65, 0.89]), p = 0.2. Republicans were thus less likely to suggest atrocities as foundations of America, and were also more likely to rate the most frequently mentioned event for all groups, the Revolutionary War, as positive. These two factors together explain why Republicans’ origin stories showed stronger positivity biases than all other political groups.

Religious divides in collective memory

Although political affiliation was the primary driver of differences in Foundations Ratings, interesting differences in event prioritization and emotional evaluation emerged across religiosity groups as well. Whereas the primary difference between political groups concerned the extent to which major atrocities—slavery and Native American genocide—were included in or excluded from the origin story, the primary religious divide concerned attitudes toward European colonization, and a cluster of evidence suggesting that secular participants were more likely to begin America with the foundation of the independent state, whereas religious participants, particularly Catholics, were more likely to begin America with earlier phases of European colonization.

Secular participants cited the Revolutionary War, the Constitution, and British colonization more frequently than Christian participants. Atheist/Agnostics mentioned the Revolutionary War more frequently than Catholics, M_diff = 0.07 [0.006, 0.14], p = 0.03, than Protestants, M_diff = 0.07 [0.006, 0.13], p = 0.03, and than the religiously unaffiliated, M_diff = 0.08 [0.01, 0.15], p = 0.02. Catholics, Protestants, and the religiously unaffiliated did not differ from one another, all p > 0.05. Atheist/Agnostics were also the most likely to mention British colonization, citing it as an origin event more frequently than Catholics, M_diff = 0.09 [0.03, 0.14], p = 0.002, than Protestants, M_diff = 0.06 [0.01, 0.11], p = 0.01, and than the religiously unaffiliated, M_diff = 0.06 [0.01, 0.12], p = 0.02. Again, Catholics, Protestants, and the religiously unaffiliated did not differ from one another, all p > 0.05. Non-religious participants of both types were more likely to suggest the Constitution as an origin event. Those with no religious orientation mentioned it more frequently than both Catholics, M_diff = 0.06 [0.003, 0.13], p = 0.04, and Protestants, M_diff = 0.06 [0.003, 0.11], p = 0.04. Atheist/Agnostics likewise were marginally more likely to cite the Constitution as an origin event than Catholics, M_diff = 0.05 [–0.001, 0.11], p = 0.055, and than Protestants, M_diff = 0.05 [0.00, 0.1], p = 0.05.

In contrast, religious participants, particularly Catholics, were more likely to cite a much earlier event: Christopher Columbus’ “discovery.” Atheist/Agnostics were the least likely group to cite Christopher Columbus’ discovery as an origin of America, citing Christopher Columbus less frequently than Catholics, M_diff = 0.08 [0.01, 0.14], p = 0.02, than those with no religious affiliation, M_diff = 0.07 [0.009, 0.14], p = 0.03, and marginally less than Protestants, M_diff = 0.05 [–0.003, 0.11], p = 0.07. Secular participants were thus more likely to cite events related to the foundation of the independent state, particularly the American Revolution and the Constitution, as well as British colonization, which began relatively late in the European colonization of the Americas, but which is the wave of European immigration most directly linked with what would later become the independent nation. Religious participants were more likely to cite an event nearly 300 years prior to the American Revolution: Christopher Columbus’ discovery. This may, in part, reflect the arrival of Christians, and particularly Catholics, in America.

America’s start year

The tendency for Catholics to begin America with Columbus, rather than the much later American Revolution, is further corroborated by participants’ responses to the question, “When did America begin?” Participants had provided a year in response to this question. A range of dates from 0 (presumably referencing the birth of Christ) to 1977 emerged, but 1776 was strongly modal, with 76% of participants choosing that date. There was, however, a second, smaller bump at 1492, selected by 7% of participants. We binned participants into those who selected a start date for America prior to 1600 (i.e. starting with Columbus and other Catholic explorers), between 1600 and 1774 (English colonization), and after 1775 (the American Revolution), and examined whether religiosity groups differed in the frequency with which their members chose years within each bin. See Table 5 for frequency counts for each period bin, broken down by religiosity category. Catholics were three times more likely than Atheist/Agnostics to select a start date for America around 1492, and around 13% less likely to select 1776. Atheist/Agnostics, on the other hand, were much less likely to select the early date, associated with Christopher Columbus, with fewer than 6% choosing 1492. Atheist/Agnostics were the group that most frequently chose 1776. Protestants and those with no religious identity did not deviate substantially from expected distributions. For many Catholics, it seems, America begins with the arrival of Catholics in the New World, whereas for Atheist/Agnostics, America begins with the founding of the federal state.

Table 5.

America’s start date by religiosity category, count (expected count).

Religiosity	America’s start date
Religiosity	Pre-1600	1600–1774	Post-1775	Total
Atheist/Agnostic	27 (49)	49 (53)	447 (422)	523
Catholic	46 (28)	34 (30)	220 (242)	300
No religious ID	28 (30)	35 (32)	256 (257)	319
Protestant	52 (47)	48 (51)	406 (408)	506
Total	153	166	1329	1648

Numbers in parentheses represent expected count if group members were distributed proportionally, and numbers outside parentheses represent the actual counts.

X²(6) = 26.69, p < 0.001.

Emotional evaluations

Religious people were not only more likely to cite earlier colonization events; but also more likely to regard Christopher Columbus and British colonization positively than secular people, Christopher Columbus’ discovery, F(3,555) = 5.62, p = 0.001, η² = 0.03, and British colonization, F(3,283) = 4.99, p = 0.002, η² = 0.04.

Christopher Columbus’ discovery. Catholics and Protestants did not differ from one another in their valence ratings of Christopher Columbus, p = 0.8, and Atheist/Agnostics and participants with no religious affiliation did not differ from one another, p = 0.75. We, therefore, collapsed the groups, respectively, into Christian and non-religious categories. Christians (M = 0.79 [0.73, 0.85]) rated Christopher Columbus significantly more positively than did the secular groups (M = 0.63 [0.57, 0.69]) who were, on average, 15% less likely to rate Christopher Columbus’ discovery as a positive event, M_diff = 0.15 [0.08, 0.23], p < 0.001.

British colonization. A similar divergence between Christians and non-religious participants emerged regarding British colonization. Christians (M = 0.89 [0.83, 0.95]) were approximately 16% more likely to rate British colonization as positive, relative to secular participants (M = 0.73 [0.67, 0.79]), M_diff = 0.16 [0.07, 0.25], p < 0.001. For these two events, Christopher Columbus’ discovery and British colonization, the primary divide in collective memory seems to be between Christians, who regarded the European colonization of America in significantly more positive light, and secular participants, who were more likely to evaluate this aspect of America’s national origins negatively. We should mention that our religiosity categories did not interact with ethnicity, X²(24) = 19.36, p = 0.73; that is, the range of ethnicities in our sample was appropriately represented in each of the religiosity categories. This suggests that it was unlikely that the observed differences in attitudes toward European colonization were due to ethnic differences.

Discussion

Memory for shared glories

In line with Ernst Renan’s formulation of a nation as a people sharing memories for their “common glories,” we demonstrated that Americans tend to show a positivity bias when constructing their nation’s origin story. That is, Americans of different political and religious orientations all tended to provide more positive events than negative events at their nation’s foundation. As Renan pointed out in the nineteenth century, however, such a glorious, unifying national memory requires substantial forgetting for the atrocities upon which many nations are based. Such collective forgetting can occur through a variety of mechanisms—top down societal mechanisms, such as repressive erasure in which powerful state actors literally destroy historical evidence they wish the populace to forget (Connerton, 2008; Wertsch, 2008), or bottom up cognitive mechanisms that support the emergence of collective forgetting (Hirst et al., 2018). In the present instance of collective forgetting, one such cognitive mechanism may be particularly influential: narrative schematic templates (Hirst and Yamashiro, 2017).

National narratives are stereotyped scripts that a people may use while reconstructing memory for multiple specific historical episodes (Wertsch, 2002). These national narratives not only determine what aspects of the past people perceive as meaningful (i.e. which fit into the narrative), but will also determine what is “meaningless,” in that it cannot be made sense of using the narrative as a frame for understanding (Hirst and Yamashiro, 2017). The enslavement and genocide of millions simply cannot be fit meaningfully into an American narrative of a nation founded on life, liberty, and the pursuit of happiness, in which “all men are created equal,” and whose history has witnessed the spread of those ideals to a greater and greater proportion of humanity. Because foundational atrocities do not fit into the national narrative, they are omitted when telling the story of the nation’s origin. Further, while participants of all political affiliations mentioned America’s “original sins” relatively infrequently, these events were conspicuously absent from the Republicans’ origin stories. Republicans were, on average, between one-half and one-third as likely to include these events as Democrats, Independents, and the politically unaffiliated. Collective forgetting of America’s foundational atrocities characterized most Americans’ origin story, but Republican omissions were particularly thorough.

We should qualify carefully what we mean by “collective forgetting.” Total oblivion to posterity is one sort of forgetting (Connerton, 2008); for example, how many Americans know that in 1939 the German American Bund booked Madison Square Garden to host a rally for 20,000 American Nazis (Buder, 2017; Taylor, 2017)? However, we mean something closer to selective omissions made while fashioning a useful narrative. It is unlikely that any American is unaware of the historical facts of slavery and Native American genocide; most simply do not select them as events upon which America was built, although a strong case could be made for doing so. The presence or absence of the atrocities in a national origin story may reflect at least two different motives for recalling the past. A purified image of the nation’s origin allows the past to be used as a model for the present. Conservative political rhetoric is permeated by jeremiad narratives, in which a fallen and decadent present is contrasted unfavorably with mythologized origins, with accompanying exhortation to return to those morally righteous foundations (Bercovitch, 1978; Murphy, 2008). On the other hand, rehearsing past atrocities could provide motivation for a continual struggle toward an as-yet unattained social justice.

Concurrent with cultural tools such as jeremiad narratives, certain group-oriented moral motives may shape how people fashion their national narratives. For instance, there is some preliminary evidence that endorsement of the binding values of loyalty, authority, and purity, which tend to be more strongly endorsed by conservatives (Graham et al., 2009), may predict an inflated perception of the in-group’s historical influence (Churchill et al., 2019). If collective remembering were shaped by motives of loyalty or respect for authority rather than, say, desire to correct a past injustice, we would expect the omission of shameful events from narratives about the group’s past, and for such omission to be more common in more conservative people. Vice versa, if moral motives of justice motivate remembering, we would expect the inclusion of such events, to the extent that past injustices are perceived as being unresolved. This hypothesis about the relation between moral and political values and biases in collective remembering awaits more rigorous testing, but would be congruent with the data presented. If group-oriented moral motives and the use of national narratives do explain the patterns of selective forgetting we observed, we would expect selective pruning of atrocities from collective memory to be a more general phenomenon, beyond the specific case of American collective memory. When Ernst Renan made his point about forgetting in the service of national solidarity, after all, he was speaking not of America, but of France. The link between political conservatism and positivity biases in collective remembering should be examined in other national contexts.

A second major finding was that secular participants were more likely to choose events associated with the foundation of the independent state, and rated events associated with European colonialism negatively more frequently. Religious participants, on the other hand, tended toward unanimously positive evaluations of such colonization events. Religious participants, particularly Catholics, were relatively more likely to begin America with Columbus’ “discovery” in 1492. Such differences in priorities seem to index different images of the nation, although, to be sure, they suggest differences of emphasis rather than qualitatively discrete collective memories. In one telling of the American origin story, America begins with the arrival of the first Europeans, the first colonists, or, perhaps, the first Christians, in the Western hemisphere. In another way of telling the story, America begins with the founding of a constitutional government of an independent state. Separated by nearly 300 years, the set of events emphasized in these two versions of the origin story suggest two different images of what America is; a culture continuous with the Christian, European heritage, or a federal state with associations of democratic representation, the rule of law, individual rights, and an explicit break from European influence. In both cases, religious and secular Americans tend to create the nation’s origin in their own image.

Sharedness of collective memories

We step back from our discussion of American origins to comment on the empirical study of collective memory more generally. The distribution shown in Figure 1 makes clear an important point for researchers trying to characterize collective memories, in that it qualifies the extent to which we might say that a memory is hegemonic. One definition of collective memory is that collective memories are memories or historical narratives shared by members of a group, which bear on group identity (Hirst and Manier, 2008; Wertsch and Roediger, 2008). This definition would suggest that collective memories are shared more or less evenly across members of a group; Wertsch and Roediger (2008) imply this in saying that in a sense, having a particular collective memory defines a person as a member of a group. However, in our data, even the most frequently mentioned event, the Revolutionary War, was only chosen by 62% of participants; all other events were agreed upon as foundational much less frequently, with the next most frequently mentioned event, Declaration of Independence, mentioned by only 36% of the sample. Undoubtedly, most Americans would recognize all or most of the events listed if probed, but there is limited agreement about which events constitute America’s foundations. Relatedly, there does not seem to be a qualitative cutoff, at least as characterized by frequency of mention, between items that belong to the American origin story and those that do not; there is a smooth transition from the most frequently mentioned items to the least frequently mentioned items, out to a long tail (not presented in Figure 1) of events mentioned by fewer than 100 people. Further, the events mentioned seem to span over a period of 400 years. “Hegemony” within collective memory is a complicated construct that may be difficult to characterize empirically.

This lack of agreement on which events are the critical ones in the nation’s past is compatible with other recent empirical work on American collective memory. Taylor, et al. (2017) found that only two events were nominated by more than 10% of participants when asked to provide an event corresponding to America’s greatest year (Declaration of Independence: 20.8%, and World War II: 12.9%). They found more agreement when participants were probed for the 10 most important events shaping American identity, and the first five of their ten most frequently mentioned events were also present in our list of origin events mentioned by at least 5% of the sample (Civil War, 9/11, World War II, Declaration of Independence, and the Revolutionary War). That origin events would also figure prominently in identity defining memory is unsurprising given the role origin events play in national charters, as discussed in the introduction.

Interestingly, although Taylor et al.’s (2017) list of identity defining collective memories overlapped with our origin events considerably, nearly none of their national identity defining memories were represented among the events on which we found the most subnational divergence (slavery, Native American genocide, Christopher Columbus, British colonization, or the Constitution, with the exception of the Revolutionary War). Although our framing of origin stories as charters for a nation would seem to imply a critical identity-defining role for them, national identity may draw from a broader range of collective memory. How the goals and frames for collective remembering alter which collective memories people draw on remains an open question for future research.

Although Americans of different political and religious orientations nominally belong to the same nation, subnational identities can shape how people conceive of their nation’s origins. Differences in ideological motivation may motivate forgetting for foundational atrocities to different extents, and may lead people to center events relevant to their own political or religious image of the nation, as for American Catholics and Columbus or secular Americans and the federal government. America seems to have many origins.

Supplemental Material

Yamashiro_supplementary_file – Supplemental material for American origins: Political and religious divides in US collective memory

Supplemental material, Yamashiro_supplementary_file for American origins: Political and religious divides in US collective memory by Jeremy K Yamashiro, Abram Van Engen and Henry L Roediger in Memory Studies

Footnotes

Acknowledgements

We would like to thank David Tyus for his help in coding and preparing the data and Douglas Knox and Stephen Pentecost at Washington University in St. Louis’ Digital Humanities Workshop for assistance in OpenRefine. Early versions of these data were presented at the Primary Source: An Interdisciplinary Conference on Memory and Identity, at Fontbonne University, in May 2018.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from the James S. McDonnell Foundation to the third author.

Notes

Supplemental material

Supplemental material for this article is available online.

ORCID iD

Jeremy K Yamashiro

Author biographies

Jeremy K Yamashiro is a postdoctoral research associate in the Department of Psychological & Brain Sciences at Washington University in St. Louis. His research concerns social remembering, and takes two approaches. The first examines how humans in communication come to remember in the same way—that is, the socio-cognitive mechanisms underlying mnemonic convergence. The other line examines social representations of history, and explores how various group affiliations and moral profiles may impact how such memories vary within different subcommunities.

Abram Van Engen is an associate professor in the Department of English at Washington University in St. Louis. His first book, Sympathetic Puritans: Calvinist Fellow Feeling in Early New England (Oxford University Press, 2015), argues that a Calvinist theology of sympathy helped shape the culture of early New England, affecting definitions of English identity, conceptions of community, relations with Native Americans, and the development of American literature. His second book, City on a Hill: A History of American Exceptionalism (Yale University Press, 2020), traces how American culture has remembered and remade the Pilgrims and Puritans by offering a history of John Winthrop’s “city on a hill” sermon, A Model of Christian Charity, from 1630 to the present day.

Henry L Roediger III is the James S. McDonnell Distinguished University Professor in the Department of Psychological & Brain Sciences at Washington University in St. Louis. He studies human learning and memory using both laboratory techniques and more naturalistic approaches. His interests include applying cognitive psychology to enhance education; the genesis of false memories; the study of people with superior memories; the relation between confidence and accuracy of memories; and collective memory. Roediger is a fellow of the American Academy of Arts and Sciences and a fellow of several organizations of psychologists. He was also elected as a member of the Society of Experimental Psychologists in 1994, and he received the Howard Crosby Warren Medal from SEP for his work on illusory memories. He is a member of the US National Academy of Sciences.

References

Bercovitch

(1978) The American Jeremiad. Madison, WI: University of Wisconsin Press.

Brady

Wills

Jost

, et al. (2017) Emotion shapes the diffusion of moralized content in social networks. Proceedings of the National Academy of Sciences of the United States of America 114(28): 7313–7318.

Brooks

(1918) On creating a usable past. The Dial 64: 337–341.

Buder

(2017) When 20,000 American Nazis descended upon New York city. The Atlantic, 10 October. Available at: https://www.theatlantic.com/video/index/542499/marshall-curry-nazi-rally-madison-square-garden-1939/

Churchill

Yamashiro

Roediger

III (2019) Binding values predict state narcissism. Memory, DOI: 10.1080/09658211.2019.1623261 .

Coman

Momennejad

Drach

, et al. (2016) Mnemonic convergence in social networks: the emergent properties of cognition at a collective level. Proceedings of the National Academy of Sciences of the United States of America 113(29): 8171–8176.

Connerton

(2008) Seven types of forgetting. Memory Studies 1(1): 59–71.

Fausset

(2017) Tempers flare over removal of Confederate statues in New Orleans. The New York Times, 7 May. Available at: https://www.nytimes.com/2017/05/07/us/new-orleans-monuments.html

Frisch

(1989) American history and the structures of collective memory: a modest exercise in empirical iconography. Journal of American History 75(4): 1130–1155.

10.

Graham

Haidt

Nosek

(2009) Liberals and conservatives rely on different sets of moral foundations. Journal of Personality and Social Psychology 96(5): 1029–1046.

11.

Halbwachs

(1992) On Collective Memory (trans. Coser

). Chicago, IL: University of Chicago Press.

12.

Hanke

Liu

Sibley

, et al. (2015) “Heroes” and “villains” of world history across cultures. PLoS ONE 10(2): 1–21.

13.

Harris

(2006) It takes a tragedy to arouse them: collective memory and collective action during the Civil Rights Movement. Social Movement Studies 5(1): 19–43.

14.

Hilton

Liu

(2017) History as the narrative of a people: from function to structure and content. Memory Studies 10(3): 297–309.

15.

Hirst

Manier

(2008) Towards a psychology of collective memory. Memory 16(3): 183–200.

16.

Hirst

Yamashiro

(2017) Social aspects of forgetting. In: Meade

Harris

Van Bergen

, et al. (eds) Collaborative Remembering: Theories, Research, and Applications. 1st ed. Oxford: Oxford University Press, pp. 76–99.

17.

Hirst

Yamashiro

Coman

(2018) Collective memory from a psychological perspective. Trends in Cognitive Science 22(5): 438–451.

18.

Hobsbawm

Ranger

(2013) The Invention of Tradition. New York: Cambridge University Press.

19.

Kensinger

Corkin

(2004) Two routes to emotional memory: distinct neural processes for valence and arousal. Proceedings of the National Academy of Sciences of the United States of America 101(9): 3310–3315.

20.

Landrieu

(2018) In the Shadow of Statues: A White Southerner Confronts History. New York: Viking.

21.

Liu

Hilton

(2005) How the past weighs on the present: social representations of history and their role in identity politics. British Journal of Social Psychology 44: 537–556.

22.

Lowenthal

(1985) The past Is a Foreign Country. New York: Cambridge University Press.

23.

Malinowski

(1948) Magic, Science, and Religion. Boston, MA: Beacon Press.

24.

Moscovici

(1988) Notes towards a description of social representations. European Journal of Social Psychology 18: 211–250.

25.

Murphy

(2008) Prodigal Nation: Moral Decline and Divine Punishment from New England to 9/11. New York: Oxford University Press.

26.

Novick

(1999) The Holocaust in American Life. New York: Houghton Mifflin.

27.

Olick

(1999) Collective memory: the two cultures. Sociological Theory 17(3): 333–348.

28.

Renan

(1990) What is a nation? In: Bhabha

(ed.) Nation and Narration (trans. Thom

). New York: Routledge, pp. 8–22.

29.

Schuman

Scott

(1989) Generations and collective memory. American Sociological Review 54: 359–381.

30.

Schuman

Rieger

Gaidys

(1994) Collective memories in the United States and Lithuania. In: Schwarz

Sudman

(eds) Autobiographical Memory and the Validity of Retrospective Reports. New York: Springer-Verlag, pp. 313–333.

31.

Schwartz

(1997) Collective memory and history: how Abraham Lincoln became a symbol of racial equality. Sociological Quarterly 38(3): 469–496.

32.

Smith

(1998) Nationalism and Modernism: A Critical Survey of Recent Theories of Nations and Nationalism. New York: Routledge.

33.

Sperber

Hirschfeld

(2004) The cognitive foundations of cultural stability and diversity. TRENDS in Cognitive Sciences 8(1): 40–46.

34.

Taylor

(2017) American Nazis in the 1930s—the German American Bund. The Atlantic, 5 June. Available at: https://www.theatlantic.com/photo/2017/06/american-nazis-in-the-1930sthe-german-american-bund/529185/

35.

Taylor

Burton-Wood

Garry

(2017) America was great when nationally relevant events occurred and when Americans were young. Journal of Applied Research in Memory and Cognition 6: 425–433.

36.

Van Engen

(2018) American exceptionalism and America first. Religion and Politics, 9 January. Available at: https://source.wustl.edu/2018/01/american-exceptionalism-america-first/

37.

Van Engen

(2020) City on a Hill: A History of American Exceptionalism. New Haven, CT: Yale University Press.

38.

Verborgh

De Wilde

(2013) Using OpenRefine. PACKT: Open Source. Available at: https://www.packtpub.com/big-data-and-business-intelligence/using-openrefine

39.

Vlasceanu

Enz

Coman

(2018) Cognition in a social context: a social-interactionist approach to emergent phenomena. Current Directions in Psychological Science 27(5): 369–377.

40.

Wertsch

(2002) Voices of Collective Remembering. New York: Cambridge University Press.

41.

Wertsch

(2008) Collective memory and narrative templates. Social Research 75(1): 133–156.

42.

Wertsch

Roediger

III (2008) Collective memory: conceptual foundations and theoretical approaches. Memory 16(3): 318–326.

43.

Zaromb

Butler

Agarwal

, et al. (2014) Collective memories of three wars in United States history in younger and older adults. Memory and Cognition 42: 383–399.

44.

Zerubavel

(1997) Social Mindscapes: An Invitation to Cognitive Sociology. Cambridge, MA: Harvard University Press.

45.

Zerubavel

(2003) Time Maps: Collective Memory and the Social Shape of the past. Chicago, IL: University of Chicago Press.

46.

Zerubavel

(1994) The death of memory and the memory of death: Masada and the holocaust as historical metaphor. Representations 45: 72–100.

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

0.24 MB