“I did not believe what I could not see.”
This chapter looks at the dynamics of people’s factual beliefs during the Syrian civil war. Specifically, it investigates people’s belief in things they heard during the war and their confidence in their ability to discern true and false information. To do so, it uses data from a significant sample of semi-structured interviews with Syrian refugees (N = 179) in Turkey that was collected by Schon (Reference Schon2020) and generously shared with me for this research. While these interviews focused chiefly on people’s survival strategies during the war, they also elicited rich data on their factual beliefs, truth discernment abilities, information diets, and wartime experiences while in Syria that enable them to shed light on the key ideas in this book. Additionally, since most interviewees gave detailed narrative answers, I was able to dig into their stories qualitatively and code parts of them up for quantitative analysis in new ways. Overall, the analysis suggests that it was hard for civilians to tell what was true and false in Syria, but that those with more exposure to the war and personal experience with the fighting were better positioned to do so. It also uncovers suggestive evidence that it was indirect exposure (“seeing”) and not direct exposure (“suffering”) to the fighting that is most responsible for these learning processes.
5.1 The Syrian Civil War and Its Informational Landscape
Syria was thrust into civil war in 2011 when the “Arab Spring” – the series of revolutionary uprisings that swept across the Middle East and North Africa region in 2010–11 – hit the country. After longstanding dictators in Tunisia and Egypt were ousted by protesters in early 2011, publics facing even some of the region’s most repressive regimes hoped for meaningful change. In Syria, a country that had been tightly controlled by the personalist dictatorships of Bashar al-Assad and his father since 1970 (Wedeen Reference Wedeen2015), protests finally broke out in mid-March 2011 in the southern city of Deraa and soon spread to other key cities in the country like Damascus, Aleppo, and Homs. The Assad regime responded to this mobilization ruthlessly, shooting into the crowds of protesters and arresting and torturing countless demonstrators in the early weeks and months of the uprising. As this vicious crackdown continued, the protests hardened into armed rebellion and began to seize territory from the regime. This only further escalated the government’s tactics, as it turned toward indiscriminate shelling and airstrikes as well as the unleashing of brutal militias from the dominant Alawite sect to “cleanse” rebel-held population centers.
The intensifying war drew in a variety of different actors. On the rebel side, the ostensibly more “moderate” rebel groups like the Free Syrian Army (FSA) that led the fight were soon joined by domestic and foreign jihadis, some released from prison by Assad specifically to discredit and undermine the rebels (Lister Reference Lister2016). Different rebel factions answered to different sponsors as well, with Turkey, some of the Gulf states, and the USA backing their own proxies. This fueled corruption and infighting in the rebel ranks. The emergence of ISIL also complicated the picture, as it seized large swaths of territory in both eastern Syria and northern and western Iraq in 2014 and fought against both the Syrian regime and other opposition factions. Kurdish armed groups such as the People’s Defense Units also emerged to defend Kurdish communities in northern and eastern Syria, though they faced some international criticism too as they displaced Arab residents living in mixed Arab-Kurdish areas.Footnote 1
Meanwhile, allied as it was with the region’s so-called “Shiʿa crescent,” the Syrian regime received considerable support from Iran as well as the powerful Lebanese Shiʿa militia Hezbollah as it attempted to crush the rebellion. The government was also aided by Russia, with which it had a longstanding relationship dating to the Cold War. Russia’s assistance to the regime was initially more limited in nature, but in 2015 – with various rebel factions surging and the government facing its most serious risk of collapse – it began a large-scale bombing campaign against the rebellion. This combination of robust Iranian and Russian support has proved decisive in propping up the government and allowing it to steadily reassert control over most of the country since 2015. As of this writing, Assad controls most (some 70 percent) of Syria’s territory,Footnote 2 with only a few relatively limited slices outside the regime’s grasp, including a Kurdish zone of control underwritten by US support in the country’s northeast, a Turkish “buffer zone” that is dominated by Islamist groups in the northwest, and the Israeli-ruled Golan Heights in the southeast. While the regime has been able to reestablish control over most of the territory, it has come at a terrible price: The Syrian civil war has been one of the deadliest conflicts of the post-Cold War era, with about 300,000 civilian deaths during the dispute and nearly twelve million people – over half the country’s prewar population – fleeing their homes since it began.Footnote 3
The information environment in Syria reflects many of these dynamics. Before the war, the media channels in Syria were tightly controlled by the regime, with state-run television, radio, and newspapers – along with a handful of private outlets owned by regime-linked elites – dominating the landscape. The war has significantly changed this picture, bringing to prominence a wide range of other information sources. One trend has been the rise of scores of “citizen journalists” in Syria, who have used social media to document the war – particularly the regime’s brutal role in it – and broadcast it to the wider world. Major pro-opposition news websites, radio stations, and television channels have also emerged, often operating from nearby countries like Turkey due to government repression of their efforts in Syria. These new citizen journalists and opposition outlets have faced very difficult working conditions, with repression not only from the regime but from other parties in the conflict as well. Indeed, Islamist factions like Jaysh al-Islam and ISIL have spread their own propaganda and targeted local journalists who have been critical of their repressive behavior. And, while it has been less brutal, the Kurdish autonomous administration in the country’s northeast has pressured and censored critical media as well. Understandably, in this context, many Syrians have turned to local Facebook pages affiliated with their communities for news about the war. Overall, the conflict has thus yielded a less state-dominated information landscape, but one in which various powerful combatants – the government, the Islamist rebels, and the Kurdish groups – aim to project favorable narratives and control information within their areas of control and influence.
Given these realities, it is not surprising that there has been ample misinformation pushed within Syria about the conflict. Perhaps the most internationally infamous case of misinformation in the conflict is the regime’s denial of its use of chemical weapons on civilians in rebel-held areas like Khan Shaykhun, Ghouta, and elsewhere. More broadly, the government’s narrative portrays the conflict as a battle against hard-core religious terrorists that are seeking to destroy the country and wreck the order and security it positions itself as bringing to ordinary Syrians (Wedeen Reference Wedeen2019). Within this context, it has denied the many harms inflicted on civilian populations in its “counterterrorism operations” and falsely attributed some of its worst violence to the opposition. It has also sought to scare the country’s various minority groups – including the Alawites, the government’s nominal core supporters, as well as the Christians and Druze – into remaining loyal to the regime by ginning up false claims about imminent attacks on their communities. For example, the war correspondent Janine di Giovanni recounted a conversation with an aspiring pro-regime politician, Maria Saadeh, who exhibited such thinking (2016: 58):
She also refused to believe that the government had tortured, maimed, and killed civilians. When I listed the atrocities one by one she stopped me, putting down her cup of tea. There was an angelic smile on her face. “Do you think our president could put down his own people?” she asked me incredulously. “Gas his own people? Kill his own people? This is the work of foreign fighters. They want to change our culture.”
Meanwhile, while the opposition and its biases are diverse, pro-opposition outlets have also pushed misinformation denying atrocities against government supporters, as well as overstating the rebels’ performance and territorial gains in the conflict. False pro-opposition rumors have also repeatedly swirled about the death of Bashar al-Assad and other top government officials in the war. In sum, a wealth of misleading information has been promoted by different combatants and their supporters in Syria. As one analysis of the informational dimension of the war stated, “the procurement and publication of information is inseparable from the strategies and tactics of those pursuing interests within the conflict. It is a matter of managing perceptions of hope and glory, injustice and pain” (Powers and O’Laughlin Reference Powers and O’Loughlin2015: 175).
5.2 The Interview Data and Questions
To explore people’s factual beliefs in Syria, this chapter examines a set of semi-structured interviews (N = 179) with Syrian refugees in Turkey. The data are from Schon’s (Reference Schon2020) compelling book Surviving the War in Syria: Survival Strategies in a Time of Conflict and were generously shared with me for the purposes of this research. While much of the substance of these interviews focuses on civilians’ migration experiences and other types of survival strategies during the war, they also include rich data on how people navigated new information that they encountered during the conflict – as well as their wartime experiences and attitudes – that serve as excellent fodder for examining some of the main ideas in this book.
This section first offers more information on how the interviews were carried out and the attributes of the sample before turning to the major questions used in the analysis. Overall, the data are a large and diverse convenience sample of Syrian refugees in Turkey, with ample variation in key characteristics. To collect the data, Schon first conducted test interviews with Syrian refugees living outside of the region and in Jordan (total N = 39). This allowed him to substantially develop his questionnaire and methodology. He then completed 179 semi-structured interviews in Turkey, using an Arabic translator where necessary. Since his major fieldwork trip occurred amid the 2016 Turkish coup attempt and sharp tensions between the Turkish government and Kurdistan Workers Party, Schon decided to avoid visiting the Syrian border areas and focus on collecting data from selected major cities in the country. Specifically, he conducted his fieldwork in Istanbul and Izmir, which offered “large Syrian refugee populations and relative safety” at the time (48). Schon used snowball sampling with multiple points of insertion in the target community in order to reach a variety of individuals, forging many of his connections by visiting Syrian schools and language academies as well as major Turkish universities with large numbers of Syrian students.
Descriptive statistics from Schon show that he achieved substantial diversity in his sample, Geographically, Schon reached civilians from various parts of Syria, including Damascus, Homs, Hama, Latakia, Idlib, Aleppo, Hasakah, and Deir ez-Zor. His participants fled at a wide range of times during the war, ranging from 2011 to 2016 when he conducted fieldwork. He also obtained meaningful variation in income, education, and other demographics among those he interviewed. That said, Schon’s sample reflects the dynamics of the population of Syrian refugees that went to Turkey instead of other countries when they left the conflict. In particular, those sampled by Schon tended to be Sunni Arab, anti-government, and come from the northern half of Syria – all of which make sense given the types of Syrians who migrated to Turkey (Pearlman Reference Pearlman2016).Footnote 4 Additionally, his sample tended to skew toward younger, male, and more educated respondents, which may have been a function of his use of universities as a major conduit for sampling. In sum, these interviews are not a random sample of Syrian refugees but do offer functional diversity with which to explore relationships of interest in this chapter. Moreover, like other batches of semi-structured interviews, they also provide a rich stock of narrative material to mine for insights about people’s thinking in war in ways that complement the survey-based analyses used elsewhere in this book.
One final feature of the interviews that should be noted is that not all respondents provided narrative responses. In particular, 129 respondents completed full in-person interviews with either the researcher, his interpreter, or both asking questions. The narrative answers of these individuals were recorded in full. However, fifty respondents preferred to complete the questionnaire privately and return their responses by email. These individuals only answered the close-ended questions in the survey and did not provide narrative material. This means that the quantitative elements of this chapter can take advantage of the full sample of 179 respondents, whereas the qualitative elements have slightly fewer respondents available (N = 129).
There were several parts of these interviews that were particularly relevant to our analysis. First, people were asked a set of questions about their engagement with new information in Syria, including whether they had confidence in their ability to differentiate true from false information. The specific question wording was “Do you think that you were able to tell the difference between rumors, or false information, and accurate information while you were in Syria?” This measure of truth discernment is the primary dependent variable in our analysis. Indeed, this builds on a recent argument that research on misinformation should focus on people’s truth discernment ability rather than their belief in false information per se, as the latter could be conflated with general skepticism as opposed to resistance to misinformation specifically (e.g., Batailler et al. Reference Batailler, Brannon, Teas and Gawronski2022, Guay et al. Reference Guay, Berinsky, Pennycook and Rand2022). In the sample of interest, 73.7 percent of those interviewed were confident in their ability to distinguish between true and false information while they were in Syria. Of course, it should be noted that the measure used here captures people’s self-reported confidence in their truth discernment ability – they may be over or underconfident about their ability to discern true and false information. This limitation must be acknowledged. Still, the data offer rare insights into people’s engagement with false information in war on a large scale. And the semi-structured nature of the interviews allows us to delve into the narratives of those who express high (or low) confidence in their discernment ability, helping corroborate any observed patterns with ample qualitative material.
Interviewees were also asked to generate examples of things they heard in Syria that they believed and that they did not believe. The question wording was: “Do you specifically remember any of these things that you heard? If yes, please share some examples of things that you heard that you did believe, as well as some examples of things that you did not believe.” This question was a useful way both of generating relevant data about people’s beliefs and of “priming the pump” and encouraging them to recall the major rumors and claims they encountered in the war. Overall, ninety-six people answered this question, providing a total of 182 claims they heard. I coded each of these claims for (1) their partisan thrust, which I categorized as broadly pro-regime, neutral, or pro-rebels in nature and (2) their primary topic, using an inductive approach to identify eight frequent topics based on a systematic and iterative effort to code the transcripts.
Figure 5.1 plots the distribution of things that respondents mentioned they heard during the war by their partisan thrust. There is ample variation on this score, with 24 percent of the claims mentioned classifiable as broadly pro-regime, 35 percent as broadly pro-rebels, and 42 percent as not clearly benefiting either side. This shows that respondents in the sample were exposed to new information that aligned with multiple different perspectives on the war. Meanwhile, roughly half the claims in the neutral category were believed, with lower shares of belief in the claims within the pro-rebels and especially the pro-regime categories. This suggests some disinclination to accept propaganda with a clear partisan thrust in the conflict, especially if it was pro-regime.
Figure 5.1 The distribution of claims heard in Syria, by partisan thrust
Figure 5.2 shows the distribution of claims by their main topic. The categories inductively generated by identifying the themes and topics that repeatedly emerged in the data were: territorial shifts, arrests, killings (individuals killed by a combatant or by parties unknown), fighting/battles (military engagements or activity between combatants), protests, intervention by external powers, the duration of the war, and information about leaders. As is clear, the claims that were mentioned focused on a wide variety of topics, with killings, fighting/battles, and territorial shifts being some of the most common themes among those that were identified. That said, roughly one-third of the claims did not focus on any of the identified topics, again speaking to their diversity – as it was difficult to place them into a relatively small number of buckets. There was variation in the extent to which claims within each of the different categories were believed as well. The main outlier here was the leadership category, which mainly consisted of rumors about Bashar al-Assad and other top regime officials being killed. Respondents said these rumors popped up persistently in Syria, but that they almost invariably knew they were false.
Figure 5.2 The distribution of claims heard in Syria, by primary topic
In order to explain variation in people’s perceived ability to tell true from false information in Syria, several explanatory variables are used. Recall that the central argument made in the book is that “seeing is disbelieving” – in other words, that civilians with greater exposure and proximity to the relevant events in war will tend to form more accurate perceptions of and be less vulnerable to false information about them. This analysis gets at the logic of the argument in multiple ways. First, the rich information on people’s conflict experiences collected in the interviews allows us to capture their degree of exposure to wartime violence in Syria. In particular, respondents were asked whether they had witnessed each of three different types of violent events while they were in Syria. These included witnessing a battle between armed groups, witnessing beatings or torture of other people, and witnessing a killing. These items were aggregated into an additive index of the number of different types of violent events that were witnessed by respondents in the sample. This measure reveals substantial variation, with roughly 24 percent of respondents having witnessed none of the events, 32 percent having witnessed one, 22 percent having witnessed two, and 22 percent having witnessed all three.
Another key proxy for exposure to the fighting is how long people actually spent in wartime Syria. Helpfully, the interviews contain extensive information on respondents’ migration decisions and trajectories out of Syria, including the timing of when they left their homes and left the country. This information is recorded at the daily level. Since the start of the conflict is often dated to March 15, 2011 (e.g., Schon Reference Schon2020), the number of days from this date until respondents exited Syria was used as the primary measure of temporal exposure.Footnote 5 There is considerable variation on this score, with respondents spending between 30 and 1,944 days in wartime Syria before leaving the country. Looking at this measure annually highlights the variation nicely as well: Approximately 4 percent of the people who were interviewed left Syria in 2011, 14 percent in 2012, 21 percent in 2013, 18 percent in 2014, 34 percent in 2015, and 8 percent in 2016.Footnote 6
As far as controlling for other factors, the relatively limited sample size from a quantitative perspective (N = 129, with fewer observations on some variables) argues against pushing the data too hard and including a large number of control variables. That said, the models do contain some key control variables. First, they include the binary variable wasta, which is an Arabic term for the degree of social connections or “juice” one has in society. This is one of the two primary pieces of the argument made by Schon (Reference Schon2020) to explain civilians’ decision-making and behavior in conflict. Here, the idea would be that those with more wasta will have greater confidence in their ability to tell true from false information in Syria. Second, they include an ordinal measure of respondents’ educational attainment, as this is a common predictor of belief in misinformation and conspiratorial thinking in the literature, and has in particular been linked to conspiracism in the Middle East and North Africa region (Gentzkow and Shapiro Reference Gentzkow and Shapiro2004). Third, the models contain a binary measure of whether respondents describe themselves as anti-Assad in their political orientation. Scholars have highlighted people’s political orientations and worldviews as an important predictor of their belief in misinformation in general (e.g., Miller, Saunders, and Farhart Reference Miller, Saunders and Farhart2016), as well as war and conflict in particular (e.g., Greenhill and Oppenheim Reference Greenhill and Oppenheim2017). Holding an anti-Assad ideology or worldview might make people more confident in their truth discernment ability if they firmly believed claims pushed by one side but not the other.
5.3 The Drivers of Discernment during the War
Table 5.1 displays the results of the analysis. Since the outcome – people’s confidence in their truth discernment ability – is binary, the models used are logistic regressions. As is apparent, both measures of individuals’ personal exposure to the fighting positively and significantly predict greater confidence in their truth discernment ability. Specifically, more days spent in wartime Syria and more types of violent events witnessed in Syria are both associated with a significantly greater probability of being confident in one’s truth discernment skills. Meanwhile, there is little evidence that people’s social connections (wasta), educational attainment, or political orientations explains variation in their discernment confidence.
Table 5.1 Predictors of Syrian refugees’ confidence in their ability to discern between true and false information during the war
| Discernment confidence | Discernment confidence | |
|---|---|---|
| War exposure | ||
| No. days in wartime Syria | 0.00Footnote *** | |
| (0.00) | ||
| No. event types witnessed | 0.48Footnote ** | |
| (0.22) | ||
| Other factors | ||
| Wasta | 0.04 | −0.51 |
| (0.52) | (0.47) | |
| Education | −0.04 | −0.09 |
| (0.19) | (0.18) | |
| Anti-Assad | −0.64 | −0.44 |
| (0.59) | (0.55) | |
| Constant | −0.29 | 1.12Footnote * |
| (0.90) | (0.67) | |
| Observations | 112 | 108 |
Note: Results from logistic regression models. Standard errors in parentheses.
*** p < 0.01, **p < 0.05, *p < 0.1.
To understand the substantive impacts of these two variables, the predicted probability of a respondent being confident in her truth discernment ability is plotted across each of their values. Figure 5.3 shows that their substantive influence on the outcome is quite large in real-world terms. In particular, the panel on the left shows that as the number of violent events witnessed by a civilian while they were in Syria grows from zero to three, the predicted probability that they are confident in their truth discernment goes up from about 60 percent to about 90 percent. Similarly, the panel on the right shows that as the number of days spent living in wartime Syria increases from about zero to about 2,000, the predicted probability of discernment confidence rises from about 30 percent up to about 90 percent. These are weighty shifts in people’s beliefs that they can discern truth and falsehood in the war, underscoring the power that exposure to the fighting can have.
Figure 5.3 Predicted probability of confidence in discernment ability by war exposure
In addition, further analysis was done to boost confidence in the mechanisms at work here. Indeed, while the agreement and strength of the results around time in Syria and events witnessed are encouraging, these variables do not directly test the causal mechanisms in the book’s argument. To probe the underlying process further, I coded the narratives in all of the interviews for whether there was evidence that people actually relied on personal experience to vet information while they were in Syria or not.Footnote 7 Such evidence appeared in two forms. First, some refugees explicitly stated that this strategy was used to verify things in Syria. For example, one respondent from Damascus (T020) said about how he discerned truth from lies during the war, “you can base this on your own observations and experience.” Second, other interviewees implicitly demonstrated that they relied on their personal experience to figure out what was going on in Syria. For instance, a civilian from Aleppo (T075) related that “the Syrian regime forces used to say that the demonstrations were full of weapons. I didn’t believe that because I witnessed many of these demonstrations in my area and there were not weapons.” If either explicit or implicit evidence of personal experience reliance was found, a one was coded for this measure – if not, it was coded as a zero.Footnote 8 Ultimately, this coding revealed that 44 percent of the respondents relied on personal experience to navigate information while they were in Syria.
With this as the main independent variable of interest, the model above was rerun. Figure 5.4 shows the resulting predicted probability plot, which reveals that there is a significantly greater chance that an individual will be confident in their truth discernment ability if they rely on personal experience to vet information. Specifically, the probability goes up from around 71 percent that they will have confidence in their truth discernment ability if they do not use personal experience to 92 percent if they do. This suggests that reliance on personal experience to judge the veracity of new information – which gets at the heart of the book’s argument about what buffets people against misinformation in violent conflicts – is indeed promoting people’s (perceived) ability to tell apart truth and lies in situations like the war in Syria.
Figure 5.4 Predicted probability of confidence in discernment ability by evidence of reliance on personal experience
5.3.1 Qualitative Evidence from the Interviews
Of course, the interviews also contain a rich store of qualitative material, and a careful look at this material helps complement the quantitative findings. To begin with, there were a number of interviewees who explicitly emphasized the amount of time spent in Syria as essential for people’s ability to adjudicate information during the war. As stated by one civilian who hailed from Homs (T012), “I didn’t know how to check the information at first, but I learned as I had experience from the conflict.” Similarly, another individual from Damascus (T019) said “after five years of conflict, anyone has to have the experience to know. At the beginning of the conflict, you cannot distinguish fact from fiction.” Meanwhile, a third individual who came from Homs (T033) echoed these ideas, declaring that “after a duration, things get clearer. Even the children discover that the media often says false things.” Another from Damascus (T049) related his observation that people’s gullibility was highest at the beginning of the dispute, saying “at the beginning of the conflict, people believed everything they heard. Later, I went back and found that the people had changed.” In sum, temporal experience was often explicitly discussed by the respondents in the sample as a key ingredient in people’s ability to know what was happening in the fighting.
Moreover, a number of individuals underscored the significance of personally witnessing different kinds of events in the conflict for developing a clear understanding of what was going on. The Aleppo refugee mentioned above (T075) who did not believe the regime’s propaganda about the protests after seeing them was one such example. Likewise, another individual who left Aleppo (T100) evinced this type of thinking when discussing his factual beliefs clearly. When asked about examples of things he believed while in Syria, he recounted “When the FSA said that we will enter Aleppo to free it I believed that because I witnessed them.” In contrast, when asked about examples of things that he did not believe while in Syria, he answered “When I heard that the Syrian regime will close the zone which was the only exit between western Aleppo and eastern Aleppo. I didn’t believe that, but I discovered [it was] true because they closed it.” In other words, this individual relied on personal observation to check things that he heard in the war, either confirming his preexisting beliefs or updating them when necessary. In sum, these quotations demonstrate how civilians were able to evaluate new information about things like shifts in territorial control and the dynamics of protests based on their personal observations and experiences.
The interviews indicate that personal experience was similarly crucial in learning the truth about wartime atrocities and massacres. For instance, one Aleppo refugee (T028) related learning about a pair of harrowing atrocities that were perpetrated by the regime after personally witnessing their aftermath. He explained that:
I heard that in the park close to where I lived, people were saying that there were a bunch of dead bodies. I went there to check because I didn’t believe it. I witnessed the bodies. They had to burn the bodies in the park … I heard that a school full of children was targeted by an air strike. I went there and saw the dead bodies.
Similarly, a civilian from Homs (T033) recalled learning about a bloody massacre that occurred in the city’s central square and that the regime tried to conceal after he personally saw the carnage:
There is a big square in Homs called The Watch. There were many civilians holding a demonstration there. They were only sitting. The Syrian regime brought weapons and killed many people. At first, I didn’t believe this. The next morning, I learned that it was true. The regime sent trucks for the bodies … On the next day, the people burned the bodies. The regime killed them with weapons. Later, we discovered that many dead people were in the trees.
Terrible atrocities and cover-up efforts by their perpetrators in the war were thus often discoverable by local civilians who could ascertain what had happened either by seeing them or observing some signs that they occurred in the immediate area.
Meanwhile, still other interviewees simply conveyed the crucial role of personal witnessing in how they navigated new information in the conflict more broadly. For instance, another civilian who was displaced from Aleppo (T171) stated about his information habits: “I only depended on what I witnessed. I was not interested in this other information. I witnessed many events. At the beginning of the demonstrations, I [watched] many TV channels. This news was reported before I witnessed the event. They are all propaganda … for any event, I am curious. I follow it to the end.”
This quotation speaks to the civilian’s ability – as well as their motivation – to verify events locally in the war by relying on personal observation and investigation. Or as one refugee from Damascus (T007) put it – in perhaps the simplest and most pointed articulation of the dynamics described in this section – “I did not believe what I could not see.” In sum, there is an abundance of qualitative evidence in the interviews – across a variety of different issues, from protest dynamics to territorial shifts to atrocities and more – of civilians either implicitly using their personal experience to check things they heard in the war or explicitly recognizing its value for figuring out what was going on. This detailed qualitative material helps flesh out and substantially buttress the quantitative findings highlighted in this chapter.
5.4 Dissecting Experience: Observation or Victimization?
One secondary question these data can help speak to is what kinds of personal experiences are most beneficial for effectively navigating wartime information. In particular, the detailed data on people’s conflict exposure allows us to compare the impacts of personal observation vs. personal victimization on discernment. The quantitative analyses above used a scale that contained several items about the extent to which people personally observed or witnessed different types of violence during the war. However, in that section of the interviews there were also a number of items about whether people were victimized – that is, whether they themselves or their loved ones were harmed by combatants – in a variety of different ways in the war. Does resistance to misinformation in war come more from witnessing things, being harmed by them, or both equally?
To examine this question, one can compare the model with the scale of personal witnessing to an additive scale made from different items about people’s personal victimization in the dispute. In particular, there are eight different items that asked about personal victimization. These capture whether an individual: (1) had their personal property taken or destroyed, (2) had someone shoot at them or their home, (3) received a severe beating to the body by someone, (4) was attacked with a knife or blunt object, (5) was kidnapped or detained as a prisoner, (6) received a serious physical injury during a battle, (7) received a serious physical injury during a bombing, and (8) was betrayed and put at risk of death or injury. On average, the individuals in the sample had experienced 2.5 of these eight victimization events, with a significant amount of variation among them.Footnote 9
The results of the analysis are shown in Table 5.2. As is clear, personally witnessing more types of violent events has a positive and significant association with people’s confidence in their discernment ability during the conflict (p = 0.03). In contrast, being personally victimized in more ways during the conflict has a positive association with people’s confidence in their discernment ability, but the effect is not close to statistical significance (p = 0.2). In other words, while both of these measures push in the “right” direction, we can only confidently say that personal observation helps buttress individuals’ (perceived) discernment ability. Definitively answering this question is difficult with these data, but they point toward the idea that it is seeing rather than suffering from wartime violence that is the surest path to resilience toward misinformation in conflict.
Table 5.2 The influence of observation vs. victimization on Syrian refugees’ confidence in their discernment ability
| Discernment confidence | Discernment confidence | |
|---|---|---|
| War exposure | ||
| Observation scale | 0.48Footnote ** | |
| (0.22) | ||
| Victimization scale | 0.25 | |
| (0.19) | ||
| Other factors | ||
| Wasta | −0.51 | −0.49 |
| (0.47) | (0.46) | |
| Education | −0.09 | −0.02 |
| (0.18) | (0.18) | |
| Anti-Assad | −0.44 | −0.35 |
| (0.55) | (0.56) | |
| Constant | 1.12Footnote * | 0.94 |
| (0.67) | (0.75) | |
| Observations | 108 | 108 |
Note: Results from logistic regression models. Standard errors in parentheses.
*** p < 0.01, **p < 0.05, *p < 0.1.
5.5 Discussion and Conclusion
This chapter investigated people’s susceptibility to misinformation in the Syrian civil war. In order to do so, it examined a sizable sample of semi-structured interviews with Syrian refugees in Turkey. These interviews were originally conducted by Schon (Reference Schon2020), but they were mined for new insights quantitatively and qualitatively in this chapter. After first providing a brief overview of the conflict and the various informational biases that characterize it, I described these interviews and the items used from them in detail. I then ran a series of models predicting people’s confidence in their ability to tell true from false information during the war, while also digging for qualitative insights about how they judged new information in Syria. The results of this analysis showed that individuals’ actual exposure to the conflict – conceptualized both in terms of how much time they spent in wartime Syria, and the range of violent events they witnessed there – is strongly associated with having higher confidence in their truth discernment ability. Moreover, a careful coding of the narratives in the interviews showed that actively using personal experience to vet new information is an important part of this story, speaking to some of the key mechanisms in the book’s argument. Finally, the quantitative findings were supplemented by ample qualitative material which showed that people indeed highlighted things like the amount they spent in the country, the witnessing of different types of events, and the use of direct and indirect experiential knowledge as important safeguards against the allure of misinformation.
These results build on and complement those from Chapter 3 (on the US drone campaign in Pakistan) and Chapter 4 (on the anti-ISIL air campaign in Iraq), driving toward the same overall point but from different angles. Several differences between the analyses are worth appreciating. First, the Pakistan and Iraq chapters (Chapters 3 and 4) focus on people’s belief in specific prominent misinformation claims or narratives in conflict – such as the idea that the US drone campaign in Pakistan is killing mostly civilians, or that the anti-ISIL airstrikes from 2014 to 2016 targeted allied militias – while this chapter analyzes people’s general ability to distinguish between truthful and fictitious information. Second, the type of conflict and violence in the cases is quite different: Examining people’s belief in misinformation about both the US drone campaign in Pakistan and the anti-ISIL Coalition air campaign in Iraq entails a focus on external interventions and aerial bombardment, while focusing on truth discernment in the Syrian civil war means claims about a wider range of combatants and types of violence. Third, different methods were employed in these cases, with a quantitative large-n analysis of survey and violent event data in Iraq, a case study approach with quantitative elements in Pakistan, and an analysis of detailed semi-structured interview data in Syria. Given these myriad differences, the fact that the findings of Chapters 3–5 converge and agree so strongly should seriously bolster our confidence in the central argument that exposure and proximity to the fighting are powerful antidotes to misinformation in war.
Finally, this chapter offers a prime opportunity to touch on the issue of wartime migration and displacement. Indeed, the book recognizes that people often flee from the front lines of war or from war-affected countries entirely, and leaves ample room for this process to occur. Of course, this could pose a problem if people with better information habits were less likely to do the fleeing, leaving those closer to the “action” with more accurate beliefs – but via a process of selection and not learning. While this would indeed be troubling, two key points militate against this concern. First, the bias it implies should run against the results in this book. Indeed, people with better prior information habits should be able to more accurately identify risks and thus be less exposed to violence in war. This would mean that the relationship between exposure and accuracy found in this book would be underestimated. Second, this book’s findings are borne out regardless of displacement. Indeed, this chapter demonstrated evidence of experiential learning among Syrian refugees, while Chapter 4 showed it among Iraqis who remained in Iraq during the ISIL conflict. And Chapter 3 showcased the accurate conflict perceptions of people from the tribal areas of Pakistan whether they remained in or fled from their communities. All of this suggests that the central story here is about the extent to which people are personally exposed or proximate to the fighting rather than whether they are displaced or not in a conflict per se.
