Affect as Evidence of Care

Affect in students’ tweets was certainly a result of students’ values as well as the framework of the course, but their affective responses can also be viewed as an expression of care, a key aspect of historical empathy. Students may have been more willing to alter their affective response to sources about the Mongols because the lecture content addressed students’ concerns regarding the treatment of women and conquered peoples in the sources. Students cared about the impact of the Mongols’ violence on everyday people. Discovering that violence was often tempered by an egalitarian social structure and religious tolerance alleviated some of the students’ concerns.

Students’ care, and affect produced by that care, is a result of present-day values and concerns. Yet viewing students’ care as simply a matter of presentism dismisses the appropriateness of their care and students’ genuine concern for people in the past. Barton and Levstik unapologetically argue that students should care more for victims of the Trail of Tears than they do for the thought processes of Andrew Jackson. Kohlmeier too makes space for care by encouraging both “connection with” and “consideration of” the lives of past peoples. Endacott and Brooks see resonance between past and present as a practical tool for students to understand actions driven by strong emotions, not rational thought.

Students’ care — however present-minded — is a useful step towards historical empathy. Sentiment analysis and close reading of the Women Subset, a set of 985 tweets based on the common theme “women” (prominently visible in Figure 5), demonstrate that students’ affective responses are evidence of what they care about, who they care for, and what responsibilities they attached to caring for others. While care can hinder the growth of understanding (as I explore in the next section, Understanding & Care), it does not inevitably do so. Students’ expressions of care represent an essential piece of the affective models of historical empathy and as such carry the potential to motivate students to acquire more complex understandings of the lives of past peoples.

Figure 5. Word cloud demonstrating the most common words in the All Tweets Dataset.

Word cloud (decorative)

Students’ Care for Women in the Past

In the Women Subset, students expressed care about the treatment and roles of women in societies studied in the course. Their care exhibited as the celebration of situations and actions they associated with equality, respect, and complex portrayals of women. Care also was evident when students condemned actions they deemed inequitable, disrespectful, or involving stereotypes. They perceived equitable situations as normative and typically did not offer explanations for why fair treatment of women occurred. By contrast, students frequently sought to explain examples of injustice as a lack of care on the part of past peoples.

Word clouds containing terms with high frequencies in the Women Subset, but low frequencies in English provide a snapshot of what students cared about. To determine what topics and sources elicited care as well as the situations students celebrated or condemned, I divided the terms in the Women Subset based on tweets sentimentr scores. Filtering the tweets by positive or negative scores illuminated the language students used to express their perspectives on women in ancient and medieval societies. Terms that appeared in tweets with sentimentr scores greater than zero and tweets with scores less than zero indicated the topics and texts that evoked students’ care for women (Positive/Negative Women Subset, Figure 6). Terms appearing only in positive tweets served as evidence of what students’ praised (Positive Women Subset, Figure 7). Terms that only appeared in negative tweets suggested what students criticized (Negative Women Subset, Figure 8). Close reading adds nuance to conclusions from the word clouds and illuminates the reasons for students’ concerns.

Texts and Topics of Care

Words that appear in the Positive/Negative Women Subset (Figure 6) show topics that evoked students’ care for women. The terms “hatshepsut” (69 instances out of 688 instances), “lysistrata” (40), “gilgamesh” (37), “rome” (35), “quran” (17), and “mali” (10) refer to the primary sources discussed in the classes with the most tweets about women (Classes 03, 07, 04, 16, 21, and 23 respectively). The prominence of the terms “treated” (17), “portrayed” (13), “gender” (30) and “roles” (31) demonstrates students’ care about the treatment, portrayals, and expectations of women in ancient or medieval societies. The frequency of these terms also indicates that students’ perceptions of what should be celebrated or condemned depended on their definitions of fair and unfair treatment, portrayal, or expectations of women.

Figure 6. Words appearing in tweets with both positive and negative sentiment scores in the All Tweets Dataset, weighted by word count

Word cloud (decorative)

Positive Affect = Praiseworthy Values and Actions

Students identified the most positive treatment, portrayals, and roles for women (Figure 7) in their study of Egypt (15 instances out of 573 instances), the Germanic tribes (9 instances), and early Islam (8 instances). They used “equality” (20), “strength” (6), “wisdom” (6), and “badass” (5) as positive qualities in relation to these civilizations. Less frequent terms in the subset, such as “respected” (3) and “egalitarian” (3) reinforce the theme of “equality” in the Positive Women Subset. Students also admired men’s respectful behavior toward women, such as the “modest” (6 instances) attitude toward women promoted by Clement of Alexandria’s in “Letter to the Newly Baptized” (Class 17: Rome & Christianity).

Figure 7. Words appearing only in tweets with positive sentiment scores in the All Tweets Dataset, weighted by word count

Word cloud (decorative)

Complicated Intersections of Power and Equality

A closer reading of the 568 tweets with positive sentiment scores illustrates the nuances of students’ definitions of equality, respect, or positive characteristics of women. Power, treatment equal to that of men, and a woman’s impact on an event contributed to students’ perception of a society as fair. Power was largely a matter of access to political positions. For Ancient Egypt (Class 03), students focused on Hatshepsut’s status as king of Egypt as a marker of equality:

Power was not an aspect of equality, however, when it signaled limited roles for or perceptions of women. In Classes 04 and 07, where we discussed the Epic of Gilgamesh and Lysistrata, students objected that sex appeared to be women’s only form of power. While they acknowledged that sex was a way for women to wield influence over men, they did not see sexual power as equitable. Their language instead noted the hierarchies created by the intersection of sex and power:

Irony: Lysistrata hates playing to stereotypical, female sexual roles, yet chose to exploit such feminine stereotypes for power #hwc111 #c07 (Tweet from protected account; ID: 833597083156246528.)

I think its queer that the norm was for women to withhold sex from males to get what they want as though sex means power #hwc111 #c07 (Tweet from protected account; ID: 833680008002498560.)

Clear Definitions of Equal Treatment

Power was a complicated aspect of equality for students, but the definition of equal treatment had clearer boundaries. Students required evidence of precisely the same expectations or rights for men and women before they saw treatment as equitable. In their comments on Lysistrata, students’ picked up my suggestion that the author, Aristophanes, portrays both men and women as vulnerable to their sexual desires. They viewed the mutual offense given in the play as an indicator of equal treatment of men and women:

#hwc111 #c07 Primary source: men or woman, you WILL be laughed at in rhis play. EQUALITY IS REAL. (Tweet from protected account; ID: 834275659136458753.)

The different punishments prescribed for male and female adulterers in Tacitus’s Germania and Surah

Surah 4 of the Quran, though, drew confusion and criticism from students. The notion that a woman would be punished with shame or death, while a man could go free, struck students as deeply unfair:

@nugcn’s tweet from Class 18 displays the clear boundaries students established for equal treatment. If a society exhibited equality in one area (e.g., marriage rites) then it ought to practice equality in every aspect of society (“the punishment meted out…should also be the same”). Inconsistencies were irreconcilable with equal treatment.

Women’s Impact on Society

In addition to power and equal treatment, students identified women’s impact on their civilizations as markers of respect, equality, and fairness. Across the classes, they celebrated women’s ability to influence people or events through their actions:

Qualities like wisdom, strength, and support characterized women’s actions and influence in the students’ tweets. Students read these characteristics as components of women’s impact on society. Women’s advice was “vital” in the Epic of Gilgamesh. Tomyris’s badassery allowed her Scythian tribe to defeat the powerful Persian emperor Cyrus the great. Khadija changed her husband’s life through her love as well as social status and support (emotional and financial). The term “equality” does not appear in any of the tweets above, but students’ praise of women’s ability to effect change parallels their celebration of power and equal treatment.

Negative Affect = Critical Care

Students viewed some aspects of the primary sources, such as intersections of sex and power or imbalanced punishments, as unfair because ran contrary to their association of equality with (political) power, equal treatment, and influence on society. Yet terms unique to the Negative Women Subset (Figure 8) demonstrate that unfairness was not simply the opposite of fairness for students. Students instead identified distinct actions and qualities as inequitable or unjust. In addition, students frequently attempted to explain why something unjust occurred in a text or civilization. Their tendency to provide a reason for unfair situations, but not for fair ones, suggests they viewed unfairness as abnormal. Evidence of equality made sense, but inequality required deciphering.

Figure 8. Words appearing only in tweets with negative sentiment scores in the All Tweets Dataset, weighted by word count.

Word cloud (decorative).

Unlike the Positive Women Subset where Classes 03, 18, and 21 stood out as centerpieces of students’ positive reactions, no specific classes consistently drew students’ condemnation. Instead, students attached negative sentiment to particular episodes and themes in the primary sources. “Prostitute” (6 instances of 206) “harlot” (6), “primitive” (5), “spread” (4), and “robe” (3) are all tied to students’ discussion of the Epic of Gilgamesh. In the text, the character Shamhat is described as a “harlot” and commanded to “spread out [her] robe and perform the primitive task of womankind” in order to civilize the wild man Enkidu.

Likewise, the prevalence of “Ibn” (4) and “Surah” (4) signal negative responses to portions of Ibn Battuta’s Travels in Asia and Africa and Surah 4 from the Quran. Twenty-one of the 45 tweets about women from Class 23 discuss Battuta’s observation of women “going about naked” in the court of the Malian emperor. In Surah 4, students’ negative reactions focused on the punishment for adultery (discussed in the Positive Affect section above) and verse 4.34, which contains the words “beat [or discipline] them (women) lightly.” Contemporary interpretations of the verse typically do not condone domestic violence, but many students read the verse as permitting husbands to abuse their wives.

Along with specific episodes in the primary sources, students used “restricted” (5), “confined” (3), and “degrade” (or “degraded”/”degrading”) (12) to build their definitions of unfair treatment. “Mocking” (4) and “stereotypical” (3) signify students’ opposition to portrayals of women that ridiculed women or portrayed them using negative tropes. Physical violence drew their ire as well; words like “abducted” (3), “beat” (3), “strike” (3), and “strangle” (3) repeatedly appear in tweets with negative sentiment scores. Finally, students reacted negatively to situations they found “confusing” (4 times; “confused” appears 4 times), “ridiculous” (4), or “weird” (3).

Frustration with Limited Portrayals, Violence, and Systemic Inequalities

Examining the full text of the 312 tweets in the Negative Women Subset reveals frustrations hinted at by the prominent terms in the word cloud. Students expressed sadness, anger, and disgust at limited portrayals of women, systemic inequalities, and physical coercion. They saw narrow portrayals of women as especially potent forces for constricting women’s roles and rights:

Still slightly disturbed but I guess I have to get used to seeing women being depicted as having only 1 role: to have sex 🙁 (Tweet from protected account; ID: 834011038320242688.)

Students use of “only” expressed their perception that reducing women to a single role, purpose, or name signaled misogyny, objectification, and devaluation. Students also saw mockery and stereotyping as ways to reinforce limited views of women:

Female stereotype present; Francis’ mother as the more compassionate one and even staying at home. #hwc111 #c19 (Tweet from protected account; ID: 849492323658498049.)

In @Vid_hya’s tweet especially, mockery and stereotypes are tools to diminish women’s potential power in society. Students saw limiting women to a single role (wise, irrational, ridiculous, compassionate mother) as a way for societies to perpetuate inequality.

In addition to the subtler power wielded by limited portrayals of women, students regarded overt forms of violence and coercion as clear evidence of inequality. This was especially true in Class 16, in which students read the Roman historian Livy’s account titled “The Rape of the Sabines.” Livy recounts the Romans’ abduction of the daughters of a neighboring tribe, the Sabines, but does not mention any act of sexual violence. Regardless, students’ noted evidence of coercion in the text. They condemned the abduction as barbaric and censured the Romans for eliminating women’s power of choice:

Students saw the abduction of the Sabine women as an obvious example of gender inequality but did not limit their criticism to singular acts of mistreatment. Students also diagnosed systemic inequalities as symptoms of unfairness. They criticized marriage practices, patriarchal family structures, economic rights, and religion as facets of unfairness. Polygamy in early Islam and Mongol society seemed inequitable to some students. Patriarchal family structures in which women were granted respect only if they obeyed their fathers or husbands likewise were out of keeping with students’ definitions of fairness:

Women r given fair treatment when they obey husbands &/or religious dictums. But the text advocates 4 patriarchal fam structure  #c21 #hwc111 (Tweet from deleted account; ID: 850243297243045889.)

This last tweet, in fact, ties together two systems of inequality: family structure and religion. Other students also objected to instances in which a religious text described a male deity who exercised free reign over a woman’s life or bodies:

Ancient Egyptians and Romans would not have viewed their deities as beholden to human laws, but the students did. Religious justifications for inequality were as worthy of condemnation as human actions.

Inequality as a Lack of Care

Students assumed that men and women in the past shared their care for women. When they presumed shared concerns, students also expected past peoples to “care to” change things, in Levstik and Barton’s terminology. In their tweets, students insisted that caring for someone required taking action. They praised efforts to effect change but criticized acquiescence to the status quo. When people in the past failed to take action, students’ identified their failure as a lack of care.

Praise for Resistance and Incremental Change

Students primarily identified efforts to create change when women and men resisted inequitable situations. Rebellion was a particularly prominent theme in students’ responses to Lysistrata. Students’ praised the eponymous heroine for standing up for women and for her perceived similarity to women today:

In one instance, a student also praised a man’s efforts to effect change. @jeremycherngg read Enkidu’s stand against Gilgamesh as evidence of resistance to inequality:

Although dramatic rebellions frequently caught students attention, they also lauded incremental steps toward equity. In their responses to Class 21, which focused on early Islam, students interpreted Surah 4 as a relatively equitable text. Using their personal knowledge of the religion and context information I provided, students acknowledged the disparities present in inheritance laws especially. Nonetheless, they recognized how exceptional it was to read a text enshrining women’s social and economic freedoms:

Laws on inheritance/marriage indicate that women and men were treated justly and equally.  #hwc111#c21 (Tweet from deleted account; ID: 851396850150232064.)

One student also set the record straight regarding the punishment for adultery. They noted that although punishment for adultery is applied unequally in Surah 4, other passages from the Quran promote equal responsibility:  

@RenetteFLee the punishment for illegal sexual activities for both women AND men is mentioned in Surah 24 (2-5th ve… (Tweet from protected account; ID: 851377970442772480.)

The student emphasized the Quran’s general message of care for women and encouraged their classmates to look more closely at the assigned passages.

Criticism for Failure to Take Action

Students praised women’s rebellion against expectations, men’s willingness to oppose inequality, and systemic changes in society because they cared about what happened to women in the past. Evidence that a person or society was willing to work for change alleviated students’ concerns and confirmed their belief that care required action. When women or men failed to take action, students accused historical figures of a lack of care.

Students associated absence of care with a lack of morals and acquiescence to the status quo. They interpreted Romulus’s plan to abduct the Sabine women as evidence that Romulus himself had no morals or morality simply did not exist at the time:

The actions of the Sabine women likewise appeared to exhibit a lack of care. The fact that Romulus persuaded them to forgive their Roman captors disgusted and confused students:

The last tweet from @effnahjeeching’s suggests the Sabine women’s psychological trauma excused their acceptance of the Romans’ offer of land, money, and children. Yet for the most part, students saw women as responsible for rebelling against inequalities imposed on them. Although they do not explicitly state the Sabine women’s response was a lack of care, students’ negative responses to their actions demonstrate that they viewed acquiescence as an unacceptable response to inequality. Caring about other people required taking action, regardless of the time or place in which an event occurred.


  1. Keith C. Barton and Linda S. Levstik, Teaching History for the Common Good, Kindle Edition (Mahwah, N.J: Routledge, 2004), 228.

  2. Jada Kohlmeier, “‘Couldn’t She Just Leave?’: The Relationship Between Consistently Using Class Discussions and the Development of Historical Empathy in a 9th Grade World History Course.,” Theory & Research in Social Education 34, no. 1 (2006): 37.

  3. Jason L. Endacott and Sarah Brooks, “An Updated Theoretical and Practical Model for Promoting Historical Empathy,” Social Studies Research and Practice 8, no. 1 (2013): 42-43.

  4. I created the Women Subset by matching terms in a “Women Terms” list with tweets in the All Tweets dataset. The matched terms were: “(?i)wom.”, “(?i)^she$”, “(?i)^her$”, “(?i)wif.”, “(?i)wiv.”, “(?i)mothe.”, “(?i)daughte.”, “(?i)Hatshepsut”, “(?i)Shamhat”, “(?i)Ninsun”, “(?i)harlo.”, “(?i)prostitut.”, “(?i)Lysistrata”, “(?i)Cleonice”, “(?i)Lampito”, “(?i)Myrrhine”, “(?i)Tomyris”, “(?i)Sabin.”, “(?i)Khadij.”, “(?i)female.” The “(?i)” is the regex to ignore case, allowing me to search for terms that were capitalized or lowercase. The “.” at the end of each term allowed a search for variations on the pattern (e.g., “female.” also matched both “female” and “females” in the All Tweets dataset. The final version of the Woman Subset is comprised of all tweets containing any variation of the words on the Women Terms list.

  5. See Understanding & Care: Limitations & Possibilities in Affect & Historical Empathy.

  6. Source: Derived from All Tweets Dataset. See “All Tweets Word Cloud (Blue)” in R Script in Downloads & Datasets.

  7. I used sentimentr to determine the tweets in the Women Subset with scores greater than zero (positive) or less than zero (negative). I divided these tweets into a Positive Women Subset and Negative Women Subset and joined each subset with the word frequencies list of 150,000 English words shared by Taylor Arnold and Lauren Tilton, “Basic Text Processing in R,” Programming Historian, March 27, 2017, The join allowed me to filter for words with relative frequencies of less than 0.005 in the English language and raw frequencies greater than 3 in the Positive Women Subset and Negative Women Subset. The numbers were chosen after experimenting with lower relative frequencies; anything under 0.005 resulted in too few terms for analysis. In addition, I selected for words that only showed up in tweets with positive or negative scores rather than words that appeared in both the Positive Women Subset and Negative Women Subset.

  8. Source: Derived from All Tweets Dataset. See “Find Unique Terms in Positive, Negative, and Positive/Negative Subsets” in R Script in Downloads & Datasets.

  9. “Let your speech be gentle towards those you meet, and your greetings kind; be modest towards women, and let your glance be turned to the ground.” Clement of Alexandria, “Letter to the Newly Baptized,”, accessed Feb 27, 2019.

  10. Source: Derived from All Tweets Dataset. See “Find Unique Terms in Positive, Negative, and Positive/Negative Subsets” in R Script in Downloads & Datasets.

  11. Source: Derived from All Tweets Dataset. See “Find Unique Terms in Positive, Negative, and Positive/Negative Subsets” in R Script in Downloads & Datasets.

  12. “Among their bad qualities are the following. The women servants, slave-girls, and young girls go about in front of everyone naked, without a stitch of clothing on them. Women go into the sultan’s presence naked and without coverings, and his daughters also go about naked.” Ibn Battuta, “Travels in Asia and Africa,”, accessed Feb 27, 2019.

  13. Barton and Levstik, Teaching for the Common Good, 229.

  14. Kohlmeier’s anecdote of a student querying why Penelope couldn’t leave the home of Odysseus is a close parallel to my students’ responses to the Sabine women. Kohlmeier, “Why Couldn’t She Just Leave?,” 34.